Methods
in
Molecular Biology™
Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfi...
143 downloads
724 Views
12MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Methods
in
Molecular Biology™
Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK
For other titles published in this series, go to www.springer.com/series/7651
1
2 3 4 5
Lipidomics Volume 1: Methods and Protocols
Edited by
Donald Armstrong University at Buffalo, Buffalo, NY, USA University of Florida, Gainesville, FL, USA
Editor Donald Armstrong University of Buffalo Buffalo, NY, USA and University of Florida Gainesville, FL, USA
The cover image are overlapping snapshots of a single peroxidized lipid, taken from computer stimulations. Peroxidation mofidies the local conformational preferences of acyl chains and increases their mobility, with implications for structural and dynamic properties of the membrane. ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-60761-321-3 e-ISBN 978-1-60761-322-0 DOI 10.1007/978-1-60761-322-0 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2009927725 © Humana Press, a part of Springer Science+Business Media, LLC 2009 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Cover illustration: Snapshots of a single PLPC and a peroxidized analogue (13-tc), taken at 5 ns intervals. Molecules are oriented along the membrane and superimposition was done on the phosphorus and oxygen atoms. Images were obtained from Chapter 18, Vol. 1. Cover design: Karen Schulz Printed on acid-free paper Humana Press is part of Springer Science+Business Media (www.springer.com)
sdfsdf
Preface Lipidomics is a sub-discipline of metabolomics and is defined as the large-scale study of non-water-soluble metabolites (lipids and lipidome) that utilize system-level analysis to characterize lipids and their interacting moieties (1). A literature search at the end of 2008 showed there were 200 articles published on lipidomics encompassing glycerophospholipids, sphingolipids, polyunsaturated fatty acids, glycolipids, sterol lipids, and proteolipids. It has been predicted that the combination of these lipid classes totals between 1,000 and 2,000 molecules. Lipids can also act as second messengers, or mitogens, and participate in profiling and signaling via specialized microdomains that have large amounts of lipids. When lipids are disturbed, their metabolites probably contribute to disease. In prostate cancer, for example, cyclooxygenase and lipooxygenase are upregulated reducing angiogenesis and tumor growth. Early separation and identification of lipids started with TLC, and as technology advanced, it progressed to the use of GC and HPLC. Technical improvements to HPLC include reversed-phase methods, ESI, evaporative light scattering, electrochemical detection, APCI, suppressed conductivity and multi-dimensional electrophoresis. Other technologies coupled to chromatographic methods, such as MS/MS-MSn/MALDI/ TOF, NMR, and MRM, provide a powerful approach to the global analysis of complex lipid mixtures, understanding structural changes through biophysical approaches and the effects of lipids on physiology, i.e. atherosclerosis. This has given us a clearer understanding of human and animal pathology, i.e. diabetes, cancer, neurodegeneration, and infectious disease. A new approach to measure oxidized lipids, referred to as “oxidative lipidomics,” has recently been described which provides methodology for separation and identification of these highly reactive lipids, especially in mitochondria. Many novel techniques are described in these volumes, including an imaging lipidomics approach. For another lipidomic approach of a lipid-derived radical technique, the reader is referred to Iwabashi, H., 2008. Advanced Protocols in Oxidative Stress I, volume 477, Chapter 6, Humana Press. In that same volume, a lipidomics technique for sphingolipids is also described, i.e. Wilder, AJ and Cowart, LA, Chapter 28. The present volumes have taken a “shotgun” approach and are divided into seven parts in order to include as many different varieties of technology as possible. Chapters by international experts present a wide variety of reviewed as well as unpublished data on isolation techniques, structural analysis, lipid rafts, lipid trafficking and profiling, biomarkers, lipid peroxidation, biostatistics applied to lipids, software tools, and bioinformatics. These studies range from simple systems, such as in yeast, to complex biological models. The ever increasing utilization of lipidomics will lead to more powerful technology, improved diagnostic–prognostic capabilities for medical disorders, and for the identification of new classes of lipids.
v
vi
Preface
I thank my son, Dennis Armstrong, and my grandson, David Armstrong of On-Staff Technology, Inc., for assistance with technical support, information technology, and multi-media services. Buffalo, NY Gainesville, FL
Donald Armstrong
Reference 1. Wenk, M.R. 2005. The emerging field of lipidomics. Nature Reviews 4: 595–610.
vii
Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
v ix
Part I Shotgun and Global Lipidomics 1 Examination of the Brain Mitochondrial Lipidome Using Shotgun Lipidomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Michael A. Kiebish, Xianlin Han, and Thomas N. Seyfried 2 Tracking the Glycerophospholipid Distribution of Docosahexaenoic Acid by Shotgun Lipidomics . . . . . . . . . . . . . . . . . . . . . . . . Todd W. Mitchell 3 Global Analysis of Retina Lipids by Complementary Precursor Ion and Neutral Loss Mode Tandem Mass Spectrometry . . . . . . . . . . . Julia V. Busik, Gavin E. Reid, and Todd A. Lydic 4 Combining Lipidomics and Proteomics of Human Cerebrospinal Fluids . . . . . . . Alfred N. Fonteh and Rachel D. Fisher
3
19
33 71
Part II Analytical Approaches 5 Lipid Profiling Using Two-Dimensional Heteronuclear Single Quantum Coherence NMR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Engy A. Mahrous, Robin B. Lee, and Richard E. Lee 6 Capabilities and Drawbacks of Phospholipid Analysis by MALDI-TOF Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Beate Fuchs, Ariane Nimptsch, Rosmarie Süß, and Jürgen Schiller 7 Lipidomics of the Red Cell in Diagnosis of Human Disorders . . . . . . . . . . . . . . . Peter J. Quinn , Dominique Rainteau , and Claude Wolf 8 HPLC/MS/MS-Based Approaches for Detection and Quantification of Eicosanoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Susanna L. Lundström, Fabio L. D´Alexandri, Kasem Nithipatikom, Jesper Z. Haeggström Åsa M. Wheelock, and Craig E. Wheelock 9 Brain Phosphoinositide Extraction, Fractionation, and Analysis by MALDI-TOF MS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Roy A. Johanson and Gerard T. Berry 10 Lipidomic Analysis of Biological Samples by Liquid Chromatography Coupled to Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . . . Giuseppe Astarita, Faizy Ahmed, and Daniele Piomelli 11 Lipidomic Analysis of Human Meibum Using HPLC–MS . . . . . . . . . . . . . . . . . . Igor A. Butovich
vii
89
103 127
161
189
201 221
viii
Contents
12 Lipid Geographical Analysis of the Primate Macula by Imaging Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Timothy J. Garrett and William W. Dawson 13 A Novel Role for Nutrition in the Alteration of Functional Microdomains on the Cell Surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wooki Kim , Robert S. Chapkin, Rola Barhoumi, and David W.L. Ma 14 Lipidomic Analysis of Prostanoids by Liquid Chromatography–Electrospray Tandem Mass Spectrometry . . . . . . . . . . . . . . . . . Anna Nicolaou, Mojgan Masoodi, and Adnan Mir 15 Qualitative and Quantitative Analyses of Phospholipids by LC–MS for Lipidomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hiroki Nakanishi, Hideo Ogiso, and Ryo Taguchi 16 Determination of Fatty Acid Profiles and TAGs in Vegetable Oils by MALDI-TOF/MS Fingerprinting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zeev Wiesman and Bishnu P. Chapagain 17 Dynamics of Adipose Tissue Development by 2 H 2O Labeling . . . . . . . . . . . . . . . Etienne Pouteau, Carine Beysen, Nabil Saad, and Scott Turner 18 Analysis of Lipid Particles from Yeast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Melanie Connerth, Karlheinz Grillitsch, Harald Köfeler, and Günther Daum 19 Mammalian Fatty Acid Elongases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Donald B. Jump 20 Membrane Lipidomics and the Geometry of Unsaturated Fatty Acids : From Biomimetic Models to Biological Consequences . . . . . . . . . . . . . . . . . . . . . Carla Ferreri and Chryssostomos Chatgilialoglu 21 OnLine Ozonolysis Methods for the Determination of Double Bond Position in Unsaturated Lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Michael C. Thomas, Todd W. Mitchell, and Stephen J. Blanksby 22 Comprehensive Quantitative Analysis of Bioactive Sphingolipids High-Performance Liquid Chromatography-Tandem Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jacek Bielawski, Jason S. Pierce, Justin Snider, Barbara Rembiesa, Zdzislaw M. Szulc, and Alicja Bielawska 23 The Use of Charged Aerosol Detection with HPLC for the Measurement of Lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marc Plante, Bruce Bailey, and Ian Acworth
247
261
271
287
315 337 359
375
391
413
443
469
Part III Lipid Maps 24 Non-invasive Mapping of Lipids in Plant Tissue Using Magnetic Resonance Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485 Thomas Neuberger, Hardy Rolletschek, Andrew Webb, and Ljudmilla Borisjuk
Contents
ix
25 Mapping the Lipolytic Proteome of Adipose Tissue Using Fluorescent Suicide Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497 Maximilian Schicher, Manfred Kollroser, and Albin Hermetter 26 Identifying the Spatial Distribution of Vitamin E, Pulmonary Surfactant and Membrane Lipids in Cells and Tissue by Confocal Raman Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513 J. Renwick Beattie and Bettina C. Schock Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
sdfsdf
Contributors Ian Acworth • ESA Biosciences, Inc., Chelmsford, MA, USA Faizy Ahmed • Agilent Technologies, Irvine/Agilent Analytical Discovery Facility, University of California – Irvine, Irvine, CA, USA Giuseppi Astarita • Department of Pharmacology, University of California – Irvine, Irvine, CA, USA Bruce Bailey • Applications Department, ESA Biosciences, Inc., Chelmsford, MA, USA Rola Barhoumi • Department of Nutritional Sciences, University of Toronto Faculty of Medicine, Toronto, ON, Canada; Center for Environmental and Rural Health, Texas A&M University, College Station, TX, USA J. Renwick Beattie • School of Medicine, Dentistry and Biomedical Sciences, Queen’s University, Belfast, UK Gerard T. Berry • Division of Genetics, Children’s Hospital, Boston, MA, USA Carine Beysen • KineMed, Inc., Emeryville, CA, USA Alicja Bielawska • Department of Biochemistry and Molecular Biology, Lipidomic Core Mass Spectrometry Lab, Medical University of South Carolina, Charleston, SC, USA Jacek Bielawski • Department of Biochemistry and Molecular Biology, Lipidomic Core Mass Spectrometry Lab, Medical University of South Carolina, Charleston, SC, USA Stephen J. Blanksby • School of Chemistry, University of Wollongong, Wollongong, NSW, Australia Ljudmilla Borisjuk • Leibniz-Institut fur Pflanzengenetik und Kulturpflanzenforschung (IPK), Gaterslaben, Germany Julia V. Busik • Department of Physiology, Michigan State University, East Lansing, MI, USA Igor A. Butovich • Department of Ophthalmology, University of Texas Southwestern Medical Center, Dallas, TX, USA Bishnu P. Chapagain • The Phyto-Lipid Biotechnology Laboratory, Department of Biotechnology Engineering, The Institutes for Applied Research, Ben-Gurion University of the Negev, Beer-Sheva, Israel Robert S. Chapkin • Program in Integrative Nutrition & Complex Diseases, Genomics & Bioinformatics Facility Core Center for Environmental and Rural Health, Kieberg Biotechnology Center, Texas A&M University, College Station, TX, USA Chryssostomos Chatgilialoglu • ISOF, Consiglio Nazionale delle Ricerche, Bologna, Italy Melanie Connerth • Institute of Biochemistry, Graz University of Technology, Austria
xi
xii
Contributors
Fabio Luiz D’Alexandri • Department of Medical Biochemistry and Biophysics, Division of Physiological Chemistry II, Karolinska Institutet, Stockholm, Sweden; Department of Parasitology, Department of Biochemical Sciences, University of San Paulo, San Paulo, Brazil Günther Daum • Institute of Biochemistry, Graz University of Technology, Graz, Austria William W. Dawson • Department of Ophthalmolgy, College of Medicine, University of Florida Health Sciences Center, Gainesville, FL, USA Carla Ferreri • ISOF-BioFree Radicals, Consiglio Nazionale delle Riceriche, Bologna, Italy Rachel D. Fisher • Scripps College, Claremont, CA, USA Alfred N. Fonteh • Molecular Neurology Program, Huntington Medical Research Institutes, Pasadena, CA, USA Beate Fuchs • Medical Department, Institute of Medical Physics and Biophysics, University of Leipzig, Leipzig, Germany Timothy J. Garrett • Department of Medicine, College of Medicine, University of Florida Health Science Center, Gainesville, FL, USA Karlheinz Grillitsch • Institute of Biochemistry, Graz University of Technology, Graz, Austria Jesper Z. Haeggström • Department of Medical Biochemistry and Biophysics, Division of Physiological Chemistry II, Karolinska Institutet, Stockholm, Sweden Xianlin Han • Department of Internal Medicine, Washington University School of Medicine, St. Louis, MO, USA Albin Hermetter • Institute of Biochemistry, Graz University of Technology, Graz, Austria Roy A. Johanson • Department of Neurology, Thomas Jefferson University Medical Center, Philadelphia, PA, USA Donald B. Jump • Department of Nutrition and Exercise Sciences, The Linus Pauling Institute, Oregon State University, Corvallis, OR, USA Michael A. Kiebish • Biology Department, Boston College, MA, USA Wooki Kim • Department of Nutritional Sciences, University of Toronto Faculty of Medicine, Toronto, ON, Canada; Center for Environmental and Rural Health, Texas A&M University, College Station, TX, USA Harald Köfeler • Institute of Biochemistry, Graz University of Technology, Graz, Austria Manfred Kollroser • Institute of Forensic Medicine, Medical University of Graz, Graz, Austria Richard E. Lee • Department of Pharmaceutical Sciences, College of Pharmacy, University of Tennessee Health Science Center, Memphis, TN, USA Robin B. Lee • Department of Pharmaceutical Sciences, College of Pharmacy, University of Tennessee Health Sciences Center, Memphis, TN, USA Susanna L. Lundström • Department of Medical Biochemistry and Biophysics, Division of Physiological Chemistry II, Karolinska Institute, Stockholm, Sweden Todd A. Lydic • Department of Physiology, Michigan State University, MI, USA David W.L. Ma • Department of Nutritional Sciences, University of Toronto Faculty of Medicine, Toronto, ON, Canada; Center for Environmental and Rural Health, Texas A&M University, College Station, TX, USA
Contributors
xiii
Engy A. Mahrous • Department of Pharmaceutical Sciences, College of Pharmacy, University of Tennessee Health Sciences Center, Memphis, TN, USA Mojgan Masoodi • Division Pharmaceutics and Pharmaceutical Chemistry, School of Pharmacy, Life Sciences, University of Bradford, West Yorkshire, UK Adnan Mir • Division Pharmaceutics and Pharmaceutical Chemistry, School of Pharmacy , Life Sciences, University of Bradford, West Yorkshire, UK Todd W. Mitchell • School of Health Sciences, University of Wollongong, Wollongong, NSW, Australia Hiroki Nakanishi • Department of Metabolome, Graduate School of Medicine, The University of Tokyo and Core Research for Evolutional Science and Technology, Saitama, Japan Thomas Neuberger • Department of Bioengineering, Pennsylvania State University, University Park, PA, USA Anna Nicolaou • Division of Pharmaceutics and Pharmacological Chemistry, School of Pharmacy, Life Sciences, University of Bradford, West Yorkshire, UK Ariane Nimptsch • Medical Department, Institute of Medical Physics and Biophysics, University of Leipzig, Leipzig, Germany Kasem Nithipatikom • Department of Pharmacology and Toxicology, Medical College of Wisconsin, WI, USA Hideo Ogiso • Department of Metabolome, Graduate School of Medicine, The University of Tokyo and Core Research for Evolutional Science and Technology, Saitama, Japan Jason S. Pierce • Department of Biochemistry and Molecular Biology, Lipidomic Core Mass Spectrometry Lab, Medical University of south Carolina, Charleston, SC, USA Daniele Piomelli • Department of Pharmacology, University of California – Irvine, Irvine, CA, USA Marc Plante • Applications Department , ESA Biosciences, Inc., Chelmsford, MA, USA Etienne Pouteau • Nutrition and Health Department, Nestle Research Center, Lausanne, Switzerland Paul J. Quinn • Life Sciences, King’s College, London, UK Dominique Rainteau • UMRS 893-Faculte de Medicine, Pierre et Marie Curie, CHU Saint Antoine, Paris, France Gavin E. Reid • Departments of Chemistry, Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI, USA Barbara Rembiesa • Department of Biochemistry and Molecular Biology, Lipidomics Core Mass Spectrometry Lab, Medical University of South Carolina, Charleston, SC, USA Hardy Rolletschek • Leibniz-Institut fur Pflanzengenetik und Kulturpflanzenforschung ( IPK ), Gaterslaben, Germany Nabil Saad • KineMed, Inc., Emeryville, CA, USA Maximilian Schicher • Institute of Biochemistry, University of Graz, Graz, Austria Jürgen Schiller • University of Leipzig Medical Faculty, Leipzig, Germany Bettina C. Schock • School of Medicine, Dentistry and Biomedical Sciences, Queen’s University of Belfast, Belfast, UK
xiv
Contributors
Thomas N. Seyfried • Department of Biology, Boston College, Chestnut Hill, MA, USA Justin Snider • Department of Biochemistry and Molecular Biology, Lipidomics Core Mass Spectrometry Lab, Medical University of South Carolina, Charleston, SC, USA Rosmarie Süß • Medical Department, Institute of Medical Physics and Biophysics, University of Leipzig, Leipzig, Germany Zdzislaw M. Szulc • Department of Biochemistry and Molecular Biology, Lipidomics Core Mass Spectrometry Lab, Medical University of South Carolina, Charleston, SC, USA Ryo Taguchi • Department of Metabolome, Graduate School of Medicine, The University of Tokyo and Core Research for Evolutional Science and Technology, Saitama, Japan Michael C. Thomas • School of Chemistry and Health Sciences, University of Wollongong, Wollongong, NSW, Australia Scott Turner • KineMed, Inc., Emeryville, CA, USA Andrew Webb • Department of Bioengineering, Pennsylvania State University, PA, USA Åsa M. Wheelock • Lung Research Lab L4:01, Respiratory Medicine Unit, Department of Medicine, Karolinska Institutet, Stockholm, Sweden Craig E. Wheelock • Department of Medical Biochemistry and Biophysics, Division of Physiological Chemistry II, Karolinska Institutet, Stockholm, Sweden Zeev Wiesman • Phyto-Lipid Biotechnology Lab, Department of Biotechnology Engineering, The Institutes for Applied Research, Ben Gurion University of the Negrev, Beer-Sheva, Israel Claude Wolf • HDR, Laboratoire de Spectrometrie de Masse, Faculte de Medicine, Pierre et Marie Curie, Universite Paris, Paris, France
Part I Shotgun and Global Lipidomics
sdfsdf
Chapter 1 Examination of the Brain Mitochondrial Lipidome Using Shotgun Lipidomics Michael A. Kiebish, Xianlin Han, and Thomas N. Seyfried Summary Contamination from subcellular organelles and myelin has hindered attempts to characterize the lipidome of brain mitochondria. A high degree of mitochondrial purity is required for accurate measurements of the content and molecular species composition of mitochondrial lipids. We devised a discontinuous Ficoll and sucrose gradient procedure for the isolation and purification of brain mitochondria free from any detectable contamination. Shotgun lipidomics was used to analyze the lipid composition of the brain mitochondria. These procedures can be used to determine whether intrinsic lipid abnormalities underlie mitochondrial dysfunction associated with neurological and neurodegenerative diseases. Key words: Nonsynaptic, Synaptic, Mitochondria, Brain, Lipidome
1. Introduction The brain contains two major populations of mitochondria that can be isolated by discontinuous gradients. These populations include the nonsynaptic (NS) mitochondria, which are largely derived from neuronal and glial cell bodies, and the synaptic (Syn) mitochondria, which originate from the synaptic nerve terminal of neurons (1, 2). These mitochondria populations differ in calcium homeostasis, enzymes involved in the TCA cycle, electron transport chain activities, and glutamate metabolism (2–5). Differences between NS and Syn mitochondria may underlie the complexity of neural metabolism (1). A better understanding of brain bioenergetics can be obtained with accurate information on the mitochondrial lipidome. In contrast to extensive research on mitochondria DNA and proteins in disease pathology, little attention has been focused on Donald Armstrong (ed.), Lipidomics, Methods in Molecular Biology, vol. 579, DOI 10.1007/978-1-60761-322-0_1, © Humana Press, a part of Springer Science + Business Media, LLC 2009
3
4
Kiebish, Han, and Seyfried
the role of mitochondrial membrane lipids (6). Mitochondrial lipids can influence electron transport chain activities and create a membrane environment conducive for an efficient proton gradient (7). In addition, mitochondrial lipids regulate membrane fluidity and can affect the morphological structure of mitochondria (8, 9). Also, lipid composition can influence numerous mitochondrial enzyme activities to include creatine kinase, adenine nucleotide transporter, voltage dependent ion channel, and carnitine acyltransferase (7, 10, 11). Thus, the maintenance of the mitochondrial lipidome is critical for mitochondrial functionality. Shotgun lipidomics uses a high-throughput platform which allows for the rapid quantitation of lipid molecular species using multidimensional mass spectrometry (12). This lipidomic approach has the capabilities of detecting glycerophospholipid subclasses which contain alkyl ether, vinyl ether, or ester linkages (13–15). In addition, shotgun lipidomics has demonstrated increased detection sensitivity to quantitating cardiolipin, a mitochondrial specific phospholipid (16, 17). Cardiolipin is a complex phospholipid, containing four acyl chains, three glycerol moieties, and two phosphate groups (8). We also used shotgun lipidomics and tandem MS to measure all detectable molecular species of cardiolipin in the C57BL/6J mouse brain (1). Difficulty in removing subcellular contamination from brain mitochondrial populations has hindered the accurate analysis of mitochondrial lipid composition. This is especially the case for lipid rich myelin membranes, which can remain bound to brain mitochondria thus distorting analysis of lipid composition and content. Mitochondria can be isolated by Percoll, Nycodenz, metrizamide, Ficoll, and sucrose discontinuous gradients, however, the purity of these mitochondrial preparations differ depending on the intended use of these mitochondrial fractions (18–24). We used discontinuous Ficoll and sucrose gradients to isolate NS and Syn brain mitochondria free from detectable contamination. This procedure will be useful for analyzing differences in brain mitochondria in a variety of neurodegenerative and neurological disorders that involve alterations in energy metabolism.
2. Materials 2.1.Equipment (See Also Note 1)
1. Ultracentrifuge.
2.2. Reagents
1. Mitochondrial isolation buffer. 0.32 M sucrose, 10 mM Tris–HCl, and 1 mM EDTA–K (pH 7.4).
2.2.1. Mitochondrial Isolation
2. Mass Spectrometer.
2. Nanopure deionized water.
Examination of the Brain Mitochondrial Lipidome Using Shotgun Lipidomics
5
3. TE buffer (1 mM EDTA–K and 10 mM Tris–HCl, pH 7.4). 4. 20% Ficoll stock solution made with MIB, 12% and 7.5% Ficoll working solutions. 5. 1.6 M sucrose stock containing 1 mM EDTA–K and 10 mM Tris–HCl (pH 7.4); 0.8 M, 1.0 M, 1.3 M, and 1.6 M sucrose gradient working solutions. 6. Bovine serum albumin (BSA). 7. Mitochondrial isolation buffer containing 0.5 mg/mL BSA. 8. 6 mM Tris–HCl (pH 8.1). 2.2.2. Mass Spectrometry
1. 1,2-dimyristoleoyl-sn-glycero-3-phosphocholine (14:1–14:1 PtdCho). 2. 1,2-dipalmitoleoyl-sn-glycero-3phosphoethanolamine (16:1–16:1 PtdEtn). 3. 1,2-dipentadecanoyl-sn-glycero-3-phosphoglycerol (15:0–15:0 PtdGro). 4. 1,2-dimyristoyl-sn-glycero-3-phosphoserine (14:0–14:0 PtdSer). 5. N-lauroryl sphingomyelin (N12:0 CerPCho). 6. 1,1¢,2,2¢-tetramyristoyl cardiolipin (T14:0 Ptd2Gro). 7. Heptadecanoyl ceramide (N17:0 Cer). 8. 1-heptadecanoyl-2-hydroxy-sn-glycero-3-phosphocholine (17:0 LysoPtdCho). 9. Deuterated cholesterol (d6-chol). 10. Chloroform. 11. Methanol. 12. Lithium chloride.
3. Methods 3.1. Brain Mitochondrial Isolation
Mice of the C57BL/6 strain (4 months of age) were sacrificed by cervical dislocation and the cerebral cortex was dissected. Mitochondria were isolated in a cold room (4°C) and all reagents were kept on ice. The isolation procedure employed a combination of gradients and strategies as previously described (2–4, 25–28) (Fig. 1). The cerebral cortexes (a pool of 6 per sample) were diced on an ice-cold metal plate and then placed in 12 mL of mitochondria isolation buffer (MIB; 0.32 M sucrose, 10 mM Tris–HCl, and 1 mM EDTA–K (pH 7.4)). The pooled cerebral cortexes were homogenized using a Potter Elvehjem homogenizer with a Teflon coated pestle attached to a hand-held drill.
6
Kiebish, Han, and Seyfried
Fig. 1. Procedure used for the isolation and purification of NS and Syn mitochondria from mouse cerebral cortex.
Examination of the Brain Mitochondrial Lipidome Using Shotgun Lipidomics
7
Samples were homogenized using 15 up and down strokes at 500 rpm. The homogenate was centrifuged at 1,000 × g for 5 min. The supernatant was collected and the pellet was washed twice by centrifugation, collecting the supernatants each time. The supernatants were pooled and centrifuged at 1,000 × g for 5 min. The collected supernatant was then spun at 14,000 × g for 15 min. The supernatant was discarded and the pellet, which contained primarily NS mitochondria, synaptosomes, and myelin, was resuspended in 12 mL MIB and was layered on a 7.5%/12% discontinuous Ficoll gradient. Each Ficoll gradient layer contained 12 mL for a total volume of 36 mL. The Ficoll gradients were made from a 20% Ficoll stock with MIB. The gradient was centrifuged at 73,000 × g for 36 min (4°C) in a Sorvall SW 28 rotor with slow acceleration and deceleration (Optima L-90K Ultracentrifuge). The centrifugation time used permitted sufficient acceleration and deceleration to achieve maximum g force (28), and to prevent synaptosomal contamination of the mitochondrial fraction below the 12% Ficoll layer. Crude myelin collected at the MIB/7.5% Ficoll interface was discarded. Synaptosomes were collected at the 7.5%/12% interface and were resuspended in MIB and centrifuged at 16,000 × g for 15 min. The Ficoll gradient-purified NS mitochondria (FM) were collected as a pellet below 12% Ficoll. 3.2. Mitochondria Purification 3.2.1. Purification of NS Mitochondria
3.2.2. Purification of Syn Mitochondria
The FM pellet, containing NS mitochondria, was resuspended in MIB containing 0.5 mg/mL bovine serum albumin (BSA) and was centrifuged at 12,000 × g for 15 min. The resulting pellet was collected and resuspended in 6 mL of MIB. The resuspended FM pellet was layered on a discontinuous sucrose gradient containing 0.8 M/ 1.0 M/ 1.3 M/ 1.6 M sucrose. The volumes for the sucrose gradient were 6 mL/ 6 mL/ 10 mL/ 8 mL, respectively. The gradients were made from a 1.6 M sucrose stock containing 1 mM EDTA–K and 10 mM Tris–HCl (pH 7.4). The discontinuous sucrose gradient was centrifuged at 50,000 × g for 2 h (4°C) in a Sorvall SW 28 rotor using slow acceleration and deceleration to prevent disruption of the gradient. Purified NS mitochondria were collected at the interface of 1.3 M and 1.6 M sucrose. NS mitochondria were collected and resuspended in (1:3, v/v) TE buffer (1 mM EDTA–K and 10 mM Tris–HCl, pH 7.4) containing 0.5 mg/mL BSA and centrifuged at 18,000 × g for 15 min. The pellet was then resuspended in MIB and centrifuged at 12,000 × g for 10 min. The pellet was again resuspended in MIB and centrifuged at 8,200 × g for 10 min. Synaptosomes were burst by homogenization in 6 mM Tris–HCl (pH 8.1) using five up and down strokes. The homogenized synaptosomes were transferred to a 15 mL conical tube and then placed on a rocker for 1 h (4°C). The burst synaptosomes were
8
Kiebish, Han, and Seyfried
centrifuged at 10,000 × g for 10 min. The pellet was resuspended in 6 mL of MIB. The resuspended pellet was layered on a discontinuous sucrose gradient and centrifuged following the same procedure as described above for NS mitochondria. 3.3. Mass Spectrometry 3.3.1. Sample Preparation for Mass Spectrometric Analysis
3.3.2. Instrumentation and Mass Spectrometry
An aliquot of the mitochondrial preparation was transferred to a disposable culture borosilicate glass tube (16 × 100 mm). Internal standards were added based on protein concentration and included 16:1–16:1 PtdEtn (100 nmol/mg protein), 14:1–14:1 PtdCho (45 nmol/mg protein), T14:0 Ptd2Gro (3 nmol/mg protein), 15:0–15:0 PtdGro (7.5 nmol/mg protein), 14:0–14:0 PtdSer (20 nmol/mg protein), 17:0 LysoPtdCho (1.5 nmol/mg protein), N12:0 CerPCho (20 nmol/mg protein), N17:0 Cer (5 nmol/mg protein). This allowed the final quantified lipid content to be normalized to the protein content and eliminated potential loss from the incomplete recovery. The molecular species of internal standards were selected because they represent < 0.1% of the endogenous cellular lipid mass as demonstrated by ESI/MS lipid analysis. A modified Bligh and Dyer procedure was used to extract lipids from each mitochondrial preparation as previously described (29). Each lipid extract was reconstituted with a volume of 500 mL/mg protein (which was based on the original protein content of the samples as determined from protein measurement) in CHCl3/MeOH (1:1, v/v). The lipid extracts were flushed with nitrogen, capped, and stored at –20°C for ESI/MS analysis. Each lipid solution was diluted approximately 50-fold immediately prior to infusion and lipid analysis. A triple-quadrupole mass spectrometer (Thermo Scientific TSQ Quantum Ultra, Plus, San Jose, CA, USA), equipped with an electrospray ion source and Xcalibur system software, was utilized as previously described (30). The first and third quadrupoles serve as independent mass analyzers using a mass resolution setting of peak width 0.7 Thomson while the second quadrupole serves as a collision cell for tandem MS. The diluted lipid extract was directly infused into the ESI source at a flow rate of 4 mL/min with a syringe pump. Lipid classes were analyzed in three different modes: negative-ion ESI, negative-ion ESI plus lithium hydroxide, and positive-ion ESI plus lithium hydroxide (Fig. 2). Typically, a 2-min period of signal averaging in the profile mode was employed for each mass spectrum. For tandem MS, a collision gas pressure was set at 1.0 mTorr, but the collision energy varied with the classes of lipids as described previously (14, 30). Typically, a 2- to 5-min period of signal averaging in the profile mode was employed for each tandem MS spectrum. All the mass spectra and tandem MS spectra were automatically acquired by a customized sequence subroutine operated under
Examination of the Brain Mitochondrial Lipidome Using Shotgun Lipidomics
9
Fig. 2. Schematic of the shotgun lipidomics procedure used for the analysis of the mitochondrial lipidome (modified from (12)).
Xcalibur software. Data processing of 2D MS analyses including ion peak selection, data transferring, peak intensity comparison, and quantitation was conducted using self-programmed MicroSoft Excel macros (30). 3.4. Results 3.4.1. Mitochondrial Purity
Ficoll as well as sucrose discontinuous gradients were used to purify NS and Syn mitochondria (Fig. 1). As Ficoll gradientpurified NS mitochondria (FM) contained markers for cytoskeletal (b-actin) and membrane (SNAP25, PCNA, tuberin, PLP, and calnexin) contamination, we further purified mitochondria using a discontinuous sucrose gradient (see also Note 2). None of these markers were present in the Ficoll and sucrose discontinuous gradient-purified mitochondria, which contained only mitochondrial-enriched markers representing the inner mitochondrial membrane (Complex IV, subunit IV) and the outer mitochondrial membrane (MAO-A) (Fig. 3a). Cholera toxin b immunostaining is a sensitive procedure for detecting gangliosides with the GM1a structure in cells and tissues (31). The toxin can have slight cross reactivity with GD1a. GM1a and a low amount of GD1a was found in the TH, My, and FM fractions, indicating the presence of myelin and microsomal membranes in these subcellular fractions (Fig. 3b). Only a trace amount of GM1a was detected in the NS mitochondria and no GM1a was detected in Syn mitochondria. These findings attest to the high degree of mitochondrial purity achieved with the isolation procedure.
10
Kiebish, Han, and Seyfried
Fig. 3. Distribution of protein markers on Western blots (a) and of gangliosides on thin-layer chromatography (b) in subcellular fractions from mouse cerebral cortex. Subcellular fractions included total homogenate (TH), crude myelin (My), Ficoll gradient-purified NS mitochondria (FM), Ficoll and sucrose gradient purified nonsynaptic mitochondria (NS), and Ficoll and sucrose gradient-purified synaptic mitochondria (Syn). Western blots were performed to determine the distribution of specific protein markers for the inner mitochondria membrane (complex IV, subunit IV), outer mitochondrial membrane (monoamine oxidase A), myelin (proteolipid protein), synaptosomal membrane (SNAP25), cytoskeleton (β-actin), nuclear membrane (proliferating cell nuclear antigen), Golgi membrane (Tuberin), and microsomal membrane (calnexin). GM1a was visualized on TLC plates with cholera toxin b immunostaining as described in Methods. Std. is GM1a.
3.4.2. Mitochondrial Lipid Composition
We used shotgun lipidomics to evaluate lipid content and distribution of fatty acid molecular species in the NS and Syn mitochondria. The lipid classes were listed according to their relative abundance (Table 1) (see also Note 3). Although the content of most lipids was similar in the NS and Syn mitochondria, the content of Ptd2Gro was lower whereas the content of PtdSer and Cer were higher in the Syn mitochondria than in the NS mitochondria. The myelin-enriched lipids, sulfatides, and cerebrosides, were not detected in either NS or Syn mitochondria. No major changes in lipid molecular species were found between NS and Syn mitochondria (see also Note 4). The major molecular species (>2%) of anionic, weak anionic, and weak polar lipids are found in Tables 2–4.
Examination of the Brain Mitochondrial Lipidome Using Shotgun Lipidomics
11
Table 1 Lipid composition of C57BL/6J mouse brain mitochondria Lipid
Non-synaptic
Synaptic
Ethanolamine glycerophospholipids Phosphatidylethanolamine Plasmenylethanolamine
187.4 ± 12.1 164.9 ± 10.0 22.5 ± 2.2
211.7 ± 21.3 184.6 ± 20.3 27.0 ± 1.0
Choline glycerophospholipids Phosphatidylcholine Plasmenylcholine Plasmanylcholine
129.9 ± 7.7 119.6 ± 5.3 1.2 ± 0.1 9.1 ± 3.2
156.3 ± 26.1 137.4 ± 17 2.4 ± 1.1 16.5 ± 8.5
Cholesterol
139.0 ± 46.7
126.7 ± 31.2
Cardiolipin
52.7 ± 4.5
39.9 ± 3.4
Phosphatidylinositol
9.4 ± 0.8
10.2 ± 0.9
Phosphatidylglycerol
7.1 ± 0.5
6.4 ± 0.7
Sphingomyelin
5.3 ± 1.2
6.5 ± 0.6
Phosphatidylserine
4.6 ± 1.5
14.1 ± 3.0
Lysophosphatidylcholine
2.7 ± 0.6
3.3 ± 0.4
Ceramide
0.7 ± 0.2
1.6 ± 0.2
Values are expressed as mean nmol/mg protein ±S.D (N = 3) Significantly different values from NS mitochondria at *: P < 0.02; **: P < 0.005 as determined from the two-tailed t-test
Table 2 Mass content of major molecular species of anionic lipids as detemined by shotgun lipidomics [M − 2H]−
Major species
Non-synaptic
Synaptic
713.0
18:2–18:1–18:1–16:1 18:1–18:1–18:1-16:2 18:2–18:1–18:0–16:2 18:2–18:2–18:0–16:1
0.71 ± 0.02
0.53 ± 0.04
714.0
18:1–18:1–18:1–16:1 18:2–18:1–18:1–16:0
1.47 ± 0.23
1.03 ± 0.12
715.0
18:1–18:1–18:1–16:0 8:0–18:1–18:1–16:1
0.68 ± 0.04
0.61 ± 0.08
724.0
20:4–18:2–18:1–16:1
0.95 ± 0.08
0.66 ± 0.08
Cardiolipin
(continued)
12
Kiebish, Han, and Seyfried
Table 2 (continued) [M − 2H]−
Major species
Non-synaptic
Synaptic
725.0
20:4–18:2–18:1–16:0 20:4–18:1–18:1–16:1
2.22 ± 0.17
1.56 ± 0.22
726.0
20:4–18:1–18:1–16:0 20:3–18:1–18:1–16:1 20:3–18:1–18:1–16:1
1.74 ± 0.13
1.31 ± 0.09
727.0
18:2–18:1–18:1–18:1
2.48 ± 0.33
1.59 ± 0.18
728.0
18:1–18:1–18:1–18:1
3.72 ± 0.72
2.21 ± 0.15
735.0
20:4–20:4–18:2–16:1
0.56 ± 0.06
0.39 ± 0.02
736.0
20:4–20:4–18:1–16:1
1.41 ± 0.16
1.10 ± 0.12
737.0
20:4–20:4–18:1–16:0 22:6–18:1–18:1–16:1 22:6–18:2–18:1–16:0
2.11 ± 0.07
1.54 ± 0.17
738.0
20:4–18:2–18:1–18:1 22:6–18:1–18:1–16:0
2.68 ± 0.37
2.02 ± 0.12
739.0
20:4–18:1–18:1–18:1
3.30 ± 0.45
2.48 ± 0.23
740.0
20:4–18:1–18:1–18:0 20:3–18:1–18:1–18:1
1.07 ± 0.14
0.75 ± 0.03
748.0
20:4–20:4–18:2–18:2 20:4–20:4–20:4–16:0 22:6–20:4–18:1–16:1 22:6–22:6–16:0–16:0 22:6–18:2–18:2–18:2
1.38 ± 0.09
1.11 ± 0.17
749.0
20:4–20:4–18:2–18:1
1.99 ± 0.19
1.46 ± 0.20
750.0
20:4–20:4–18:1–18:1
3.24 ± 0.24
2.31 ± 0.28
751.0
22:6–18:1–18:1–18:1 20:4–20:3–18:1–18:1
2.74 ± 0.10
2.36 ± 0.26
752.0
22:6–18:1–18:1–18:0
0.54 ± 0.17
0.46 ± 0.07
760.0
22:6–20:4–18:2–18:2 22:6–20:4–20:4–16:0 22:6–22:6–18:1–16:1
0.76 ± 0.05
0.58 ± 0.09
761.0
22:6–20:4–18:2–18:1
1.72 ± 0.08
1.27 ± 0.05
762.0
22:6–20:4–18:1–18:1
2.94 ± 0.23
2.36 ± 0.13
763.0
22:6–20:4–18:1–18:0 22:6–20:3–18:1–18:1
1.03 ± 0.05
0.78 ± 0.07
773.0
22:6–20:4–20:4–18:1
1.42 ± 0.12
1.10 ± 0.10
Cardiolipin
(continued)
Examination of the Brain Mitochondrial Lipidome Using Shotgun Lipidomics
13
Table 2 (continued) [M − 2H]−
Major species
Non-synaptic
Synaptic
774.0
22:6–22:6–18:1–18:1 22:6–20:4–20:3–18:1
1.81 ± 0.09
1.56 ± 0.17
786.0
22:6–22:6–20:4–18:1 22:6–20:4–20:4–20:3
1.33 ± 0.08
1.11 ± 0.16
797.0
22:6–22:6–22:6–18:1 22:6–22:6–20:4–20:3
0.50 ± 0.04
0.50 ± 0.05
Cardiolipin
Phosphatidylinositol 857.5
16:0–20:4
1.12 ± 0.09
1.17 ± 0.11
881.5
18:2–20:4 16:0–22:6
0.23 ± 0.01
0.26 ± 0.06
885.5
18:0–20:4
6.68 ± 0.41
7.04 ± 0.69
909.5
18:0–22:6
0.37 ± 0.02
0.48 ± 0.03
Phosphatidylglycerol 719.5
16:0–16:1
0.29 ± 0.02
0.31 ± 0.02
721.5
16:0–16:0
0.49 ± 0.04
0.49 ± 0.03
743.5
16:1–18:2
0.22 ± 0.04
0.19 ± 0.04
745.5
16:0–18:2
0.54 ± 0.10
0.48 ± 0.05
747.5
16:0–18:1
3.41 ± 0.13
3.07 ± 0.43
769.5
18:2–18:2
0.82 ± 0.04
0.64 ± 0.10
771.5
18:1–18:2
0.29 ± 0.03
0.26 ± 0.02
773.5
18:1–18:1
0.47 ± 0.12
0.38 ± 0.02
775.5
18:0–18:1
0.55 ± 0.10
0.60 ± 0.14
Phosphatidylserine 786.5
18:0–18:2
0.29 ± 0.07
0.47 ± 0.11
788.5
18:0–18:1
0.24 ± 0.21
1.61 ± 0.77
810.5
18:0–20:4
0.10 ± 0.08
0.35 ± 0.15
834.5
18:0–22:6
3.13 ± 1.22
9.59 ± 1.60
836.5
18:0–22:5
0.22 ± 0.27
0.31 ± 0.20
838.6
20:0–20:4
0.28 ± 0.07
0.79 ± 0.04
18:0–22:4
14
Kiebish, Han, and Seyfried
Table 3 Mass content of major molecular species of weak anionic lipids as detemined by shotgun lipidomics [M − H]−
Major species
Non-synaptic
Synaptic
Ethanolamine Glycerophospholipids 716.5
D16:0–18:1
5.11 ± 1.14
4.82 ± 0.68
742.5
D18:0–18:2 D18:1–18:1 D16:0–20:2
5.07 ± 0.62
5.09 ± 0.88
746.5
P16:0–22:6 D18:0–18:0 P18:2–20:4
9.03 ± 1.39
9.89 ± 1.72
762.5
D16:0–22:6
9.39 ± 0.81
11.08 ± 1.00
764.5
D16:0–22:5 D18:1–20:4
8.78 ± 0.52
8.35 ± 0.86
766.5
D18:0–20:4 D16:0–22:4
37.49 ± 2.42
36.39 ± 4.86
768.6
D18:1–20:2 D16:0–22:3 D18:0–20:3
11.49 ± 2.19
16.42 ± 1.11
774.5
P18:0–22:6 P18:1–22:5 D18:0–20:0
10.37 ± 1.44
13.25 ± 1.63
776.6
P18:0–22:5 P18:1–22:4
7.29 ± 2.19
8.24 ± 2.24
788.5
D18:1–22:6
5.95 ± 0.44
6.47 ± 0.55
790.5
D18:0–22:6 D18:1–22:5
58.18 ± 1.87
58.66 ± 8.62
792.6
D18:1–22:4 D18:0–22:5
3.62 ± 0.29
4.90 ± 0.38
794.6
D20:0–20:4 D18:1–22:3 D18:0–22:4
6.53 ± 0.46
15.63 ± 2.58
796.6
D20:0–20:3 D18:0–22:3
6.85 ± 1.34
9.14 ± 0.48
564.5
N18:0
0.66 ± 0.15
1.54 ± 0.19
592.6
N20:0
0.04 ± 0.02
0.03 ± 0.00
646.6
N24:1
0.02 ± 0.02
0.04 ± 0.02
648.6
N24:0
0.02 ± 0.00
0.01 ± 0.01
Ceramide
Examination of the Brain Mitochondrial Lipidome Using Shotgun Lipidomics
15
Table 4 Mass content of major molecular species of weak polar lipids as detemined by shotgun lipidomics [M + Li]+
Major species
Non-synaptic
Synaptic
Choline glycerophospholipids 740.6
D16:0–16:0
10.00 ± 1.27
13.58 ± 1.32
766.6
D16:0–18:1
33.31 ± 1.31
37.34 ± 3.07
768.6
D16:0–18:0
2.47 ± 0.26
3.36 ± 0.93
782.7
A16:0–20:0
7.42 ± 2.81
13.18 ± 7.40
788.6
D18:2–18:2 D16:0–20:4
23.37 ± 1.28
23.13 ± 2.55
792.6
D18:0–18:2 D18:1–18:1
4.15 ± 0.42
3.89 ± 0.28
794.6
D18:0–18:1
4.39 ± 0.46
6.43 ± 0.99
812.6
D16:0–22:6 D18:2–20:4
13.94 ± 0.79
14.08 ± 1.45
814.6
D18:1–20:4 D16:0–22:5
5.66 ± 0.72
5.51 ± 1.07
816.6
D18:2–20:2 D18:0–20:4
11.36 ± 0.90
12.21 ± 2.03
840.6
D18:0–22:6 D20:2–20:4
2.07 ± 1.96
4.14 ± 1.02
709.6
N16:0
0.27 ± 0.32
0.27 ± 0.13
735.6
N18:1
0.26 ± 0.02
0.29 ± 0.08
737.6
N18:0
1.99 ± 0.38
2.76 ± 0.17
765.6
N20:0
2.33 ± 0.41
2.61 ± 0.32
793.7
N22:0
0.16 ± 0.27
0.32 ± 0.13
821.7
N24:0
0.20 ± 0.14
0.08 ± 0.02
Sphingomyelin
Lysophosphatidylcholine 490.3
14:0
0.06 ± 0.01
0.04 ± 0.02
504.3
A16:0
0.06 ± 0.03
0.04 ± 0.01
516.3
16:1
0.09 ± 0.04
0.13 ± 0.07
518.3
16:0
0.89 ± 0.21
1.22 ± 0.14
542.3
18:2
0.06 ± 0.02
0.03 ± 0.02
544.3
18:1
0.41 ± 0.13
0.54 ± 0.12 (continued)
16
Kiebish, Han, and Seyfried
Table 4 (continued) [M + Li]+
Major species
Non-synaptic
Synaptic
Lysophosphatidylcholine 546.4
18:0
0.32 ± 0.08
0.44 ± 0.08
566.3
20:4
0.16 ± 0.09
0.16 ± 0.03
574.4
20:0
0.04 ± 0.02
0.07 ± 0.09
590.3
22:6
0.24 ± 0.12
0.27 ± 0.03
592.3
22:5
0.08 ± 0.03
0.09 ± 0.03
4. Notes 1. Additional equipments and supplies needed include a 4°C temperature-controlled room, Potter Elvehjem homogenizer with a Teflon coated pestle (10 mL and 25 mL), handheld drill (max speed 500 rpm), table top centrifuge, 1.5 mL Eppendorf tubes, 15 mL and 50 mL conical tubes, Sorvall SW 28 rotor, Ultra-Clear™ ultracentrifuge tubes (Part 344058 Beckman Coulter), and culture borosilicate glass tube (16 × 100 mm). The mass spectrometer and ultracentrifuge used were a triple-quadrupole mass spectrometer (Thermo Scientific TSQ Quantum Ultra, Plus, San Jose, CA, USA), equipped with an electrospray ion source and Xcalibur system software and an Optima L-90K ultracentrifuge, respectively. 2. The analysis of highly purified brain mitochondria by shotgun lipidomics presents an innovated high-throughput approach to analyzing the mitochondrial lipidome. The major obstacle that arises from analyzing brain mitochondria is obtaining a highly purified fraction free from contamination. Although numerous types of discontinuous gradients can be utilized, we found that a Ficoll gradient in addition to a sucrose discontinuous gradient could obtain a highly purified brain mitochondria preparation suitable for lipidomic analysis (1). 3. Differences can exist in content, molecular species, glycerophospholipid subclass, or total fatty acid distribution in mitochondrial lipids. These differences can be readily detected using a shotgun lipidomics platform with minimal sample processing (14). Changes in any or all aspects of the mitochondrial
Examination of the Brain Mitochondrial Lipidome Using Shotgun Lipidomics
17
lipidome will likely change membrane fluidity, efficiency of the proton gradient, regulation of specific enzyme activities, as well as the overall bioenergetic efficiency of mitochondria (7). Alterations in brain energy metabolism in neurological and neurodegenerative diseases is well established (32–34). Analysis of the brain mitochondrial lipidome in diseased tissues or mouse models can provide new insight into the role of mitochondria lipid alterations during disease pathogenesis. 4. The development of a high-throughput approach to analyze the brain mitochondrial lipidome is a new tool to study the cause(s) for altered energy metabolism in diseased brain (35). We have also used this approach to provide lipidomic evidence supporting the Warburg theory of cancer in a series of mouse brain tumors (36). It is our opinion that new insight on brain function and pathogenesis will be realized with further investigations of the brain mitochondrial lipidome. References 1. Kiebish M.A., Han X., Cheng H., Lunceford A., Clarke C.F., Moon H., Chuang J.H. and Seyfried T.N. (2008) Lipidomic analysis and electron transport chain activities in C57BL/6J mouse brain mitochondria. J Neurochem 106, 299–312. 2. Lai J.C., Walsh J.M., Dennis S.C. and Clark J.B. (1977) Synaptic and non-synaptic mitochondria from rat brain: isolation and characterization. J Neurochem 28, 625–631. 3. Brown M.R., Sullivan P.G. and Geddes J.W. (2006) Synaptic mitochondria are more susceptible to Ca2+overload than nonsynaptic mitochondria. J Biol Chem 281, 11658– 11668. 4. Dagani F., Gorini A., Polgatti M., Villa R.F. and Benzi G. (1983) Synaptic and non-synaptic mitochondria from rat cerebral cortex. Characterization and effect of pharmacological treatment on some enzyme activities related to energy transduction. Farmaco [Sci] 38, 584–594. 5. Villa R.F., Gorini A., Geroldi D., Lo Faro A. and Dell’Orbo C. (1989) Enzyme activities in perikaryal and synaptic mitochondrial fractions from rat hippocampus during development. Mech Ageing Dev 49, 211– 225. 6. Wallace D.C. (2001) A mitochondrial paradigm for degenerative diseases and ageing. Novartis Found Symp 235, 247–263; discussion 263–266. 7. Daum G. (1985) Lipids of mitochondria. Biochim Biophys Acta 822, 1–42.
8. Hoch F.L. (1992) Cardiolipins and biomembrane function. Biochim Biophys Acta 1113, 71–133. 9. Stuart J.A., Gillis T.E. and Ballantyne J.S. (1998) Remodeling of phospholipid fatty acids in mitochondrial membranes of estivating snails. Lipids 33, 787–793. 10. Rostovtseva T.K. and Bezrukov S.M. (2008) VDAC regulation: role of cytosolic proteins and mitochondrial lipids. J Bioenerg Biomembr 40, 163–170. 11. Campbell A.M. and Chan S.H. (2008) Mitochondrial membrane cholesterol, the voltage dependent anion channel (VDAC), and the Warburg effect. J Bioenerg Biomembr 40, 193–197. 12. Han X. and Gross R.W. (2005) Shotgun lipidomics: electrospray ionization mass spectrometric analysis and quantitation of cellular lipidomes directly from crude extracts of biological samples. Mass spectrometry reviews 24, 367–412. 13. Yang K., Zhao Z., Gross R.W. and Han X. (2007) Shotgun lipidomics identifies a paired rule for the presence of isomeric ether phospholipid molecular species. PLoS ONE 2, e1368. 14. Han X. and Gross R.W. (2005) Shotgun lipidomics: multidimensional MS analysis of cellular lipidomes. Expert review of proteomics 2, 253–264. 15. Han X., Holtzman D.M. and McKeel D.W., Jr. (2001) Plasmalogen deficiency in early Alzheimer’s disease subjects and in animal models: molecular characterization using electrospray
18
Kiebish, Han, and Seyfried
ionization mass spectrometry. J Neurochem 77, 1168–1180. 16. Cheng H., Mancuso D.J., Jiang X., Guan S., Yang J., Yang K., Sun G., Gross R.W. and Han X. (2008) Shotgun lipidomics reveals the temporally dependent, highly diversified cardiolipin profile in the mammalian brain: temporally coordinated postnatal diversification of cardiolipin molecular species with neuronal remodeling. Biochemistry 47, 5869–5880. 17. Han X., Yang K., Yang J., Cheng H. and Gross R.W. (2006) Shotgun lipidomics of cardiolipin molecular species in lipid extracts of biological samples. J Lipid Res 47, 864–879. 18. Zischka H., Lichtmannegger J., Jagemann N., Jennen L., Hamoller D., Huber E., Walch A., Summer K.H. and Gottlicher M. (2008) Isolation of highly pure rat liver mitochondria with the aid of zone-electrophoresis in a free flow device (ZE-FFE). Methods Mol Biol 424, 333–348. 19. Sims N.R. (1990) Rapid isolation of metabolically active mitochondria from rat brain and subregions using Percoll density gradient centrifugation. J Neurochem 55, 698–707. 20. Sims N.R. and Anderson M.F. (2008) Isolation of mitochondria from rat brain using Percoll density gradient centrifugation. Nat Protoc 3, 1228–1239. 21. Dagani F., Zanada F., Marzatico F. and Benzi G. (1985) Free mitochondria and synaptosomes from single rat forebrain. A comparison between two known subfractionation techniques. J Neurochem 45, 653–656. 22. Taylor S.W., Warnock D.E., Glenn G.M., Zhang B., Fahy E., Gaucher S.P., Capaldi R.A., Gibson B.W. and Ghosh S.S. (2002) An alternative strategy to determine the mitochondrial proteome using sucrose gradient fractionation and 1D PAGE on highly purified human heart mitochondria. J Proteome Res 1, 451–458. 23. Stocco D.M. and Hutson J.C. (1980) Characteristics of mitochondria isolated by rate zonal centrifugation from normal liver and Novikoff hepatomas. Cancer Res 40, 1486–1492. 24. Graham J.M. (2001) Purification of a crude mitochondrial fraction by density-gradient centrifugation. Curr Protoc Cell Biol Chapter 3, Unit 3 4. 25. Lai J.C. and Clark J.B. (1976) Preparation and properties of mitochondria derived from synaptosomes. Biochem J 154, 423–432.
26. Mena E.E., Hoeser C.A. and Moore B.W. (1980) An improved method of preparing rat brain synaptic membranes. Elimination of a contaminating membrane containing 2´,3´-cyclic nucleotide 3´-phosphohydrolase activity. Brain Res 188, 207–s31. 27. Rendon A. and Masmoudi A. (1985) Purification of non-synaptic and synaptic mitochondria and plasma membranes from rat brain by a rapid Percoll gradient procedure. J Neurosci Methods 14, 41–51. 28. Battino M., Bertoli E., Formiggini G., Sassi S., Gorini A., Villa R.F. and Lenaz G. (1991) Structural and functional aspects of the respiratory chain of synaptic and nonsynaptic mitochondria derived from selected brain regions. J Bioenerg Biomembr 23, 345–363. 29. Cheng H., Guan S. and Han X. (2006) Abundance of triacylglycerols in ganglia and their depletion in diabetic mice: implications for the role of altered triacylglycerols in diabetic neuropathy. J Neurochem 97, 1288–300. 30. Han X., Yang J., Cheng H., Ye H. and Gross R.W. (2004) Toward fingerprinting cellular lipidomes directly from biological samples by two-dimensional electrospray ionization mass spectrometry. Anal Biochem 330, 317–331. 31. Brigande J.V., Platt F.M. and Seyfried T.N. (1998) Inhibition of glycosphingolipid biosynthesis does not impair growth or morphogenesis of the postimplantation mouse embryo. J Neurochem 70, 871–882. 32. Petrozzi L., Ricci G., Giglioli N.J., Siciliano G. and Mancuso M. (2007) Mitochondria and neurodegeneration. Biosci Rep 27, 87–104. 33. Beal M.F. (2005) Mitochondria take center stage in aging and neurodegeneration. Ann Neurol 58, 495–505. 34. Calabrese V., Scapagnini G., Giuffrida Stella A.M., Bates T.E. and Clark J.B. (2001) Mitochondrial involvement in brain function and dysfunction: relevance to aging, neurodegenerative disorders and longevity. Neurochem Res 26, 739–64. 35. Bowling A.C. and Beal M.F. (1995) Bioenergetic and oxidative stress in neurodegenerative diseases. Life Sciences 56, 1151–1171. 36. Kiebish M.A., Han X., Cheng H., Chuang J.H. and Seyfried T.N. (2008) Cardiolipin and electron transport chain abnormalities in mouse brain tumor mitochondria: Lipidomic evidence supporting the Warburg theory of cancer. J Lipid Res 49, 2545–2556
Chapter 2 Tracking the Glycerophospholipid Distribution of Docosahexaenoic Acid by Shotgun Lipidomics Todd W. Mitchell Summary Docosahexaenoic acid (DHA, 22:6 n-3) is an omega-3 fatty acid with a 22 carbon acyl chain containing six cis double bonds and is predominantly found in membrane glycerophospholipids. Dietary consumption of DHA has been positively linked with the prevention of numerous pathologies and consequently, it has been the focus of extensive research over the last four decades. Nevertheless, our understanding of its molecular mode of action is not well understood. One likely mechanism is through DHA’s influence on cell membranes and the proteins embedded within them. This influence may be altered depending on the glycerophospholipid head group DHA is esterified to and its fatty acid partner, i.e., the specific glycerophospholipid molecule. Accordingly, an understanding of the exact glycerophospholipid distribution of DHA within a tissue is important if we wish to gain further insight into its role in the prevention of disease. In this chapter a rapid, shotgun lipidomic approach for identifying the molecular glycerophospholipid distribution of DHA is described. Key words: Docosahexaenoic acid, ESI–MS, Phospholipid, Shotgun lipidomics, Lipid
1. Introduction Docosahexaenoic acid (DHA, 22:6 n-3) is a long-chain polyunsaturated omega-3 fatty acid with a 22 carbon acyl chain that contains six cis double bonds (Fig. 1). It is found in most animals, particularly in glycerophospholipids (GPLs) and is abundant in fish. DHA has been the focus of a large amount of research over the last few decades with interest in this essential fatty acid initiated by the famous work of Bang and Dyerberg in the 1970s (1). From this pioneering work, a link between the consumption of omega-3 fatty acids, in particular DHA and eicospentaenoic acid Donald Armstrong (ed.), Lipidomics, Methods in Molecular Biology, vol. 579, DOI 10.1007/978-1-60761-322-0_2, © Humana Press, a part of Springer Science + Business Media, LLC 2009
19
20
Mitchell
Fig. 1. Structure of dochosahexaenoic acid.
(EPA, 20:5 n-3) and a reduced risk of cardiovascular disease was identified. Since that time research has implicated DHA in the prevention of several prevalent chronic diseases affecting modern society, e.g., cardiovascular disease (2), depression (3), type 2 diabetes (4, 5), obesity (6), and cancer (7). In fact, dietary consumption of this simple lipid has been associated with the prevention of a myriad of pathologies (8). Although DHA’s mechanism/s of action are not well known, its high preference for phospholipids has ensured that its effect on membrane physics is of particular interest. The high number of possible structural conformations of DHA (9, 10) leads to a reduction in membrane packing and stability (11). DHA is also known to increase membrane permeability (12) and propensity for fusion (13). With such an extensive influence on the physical properties of membranes it is not surprising that phospholipid DHA content has been linked to the activity of numerous membrane proteins, e.g., Na + K + ATPase (14). Nevertheless, studies on protein kinase C suggest that, at least in some cases, DHA’s precise molecular glycerophospholipid distribution may be of more importance than its influence on bulk membrane properties (15, 16). In 2000, Farkas and co workers were able to identify several phosphatidylcholine (GPCho) and phosphatidylethanolamine (GPEtn) molecules containing DHA in vertebrate brains (17). Although successful, the method employed for tracking the molecular GPL distribution of DHA was extremely laborious, requiring numerous chromatographic and derivatization steps (18). More recently, the relatively simple approach of shotgun lipidomics has been applied to the identification of DHA-containing glycerophospholipids in various tissues from mice and naked mole rats (19). Shotgun lipidomics describes the ability to determine the lipid content, from classes to subclasses and even individual molecules directly from a crude lipid extract by electrospray-ionization mass spectrometry (ESI–MS) without need for prior separation or derivatization (20–22). This technique is highly sensitive, specific, and can produce quantitative data with the addition of appropriate internal standards to the lipid extract (23). In this technique, GPLs are initially separated by their charge polarity, i.e., by selecting either the positive or negative ion mode. These molecules are easily viewed using a simple MS scan, in
Tracking the Glycerophospholipid Distribution of Docosahexaenoic Acid
21
which all GPLs carrying a common charge are selected from a lipid extract. The power in ESI–MS is unveiled in its ability to select one molecular ion, and through collision-induced dissociation (CID), observe the compositional fragments of that ion. From these fragments the phospholipid head group as well as the carbon chain length and unsaturation of each fatty acid (FA) can be determined (24). The production of these characteristic fragments forms the basis of the separation methods exploited by the shotgun lipidomics approach. A triple quadrupole or a quadrupole time of flight mass spectrometer (with enhanced duty cycle) permits precursor ion scanning allowing the analysis of specific phospholipids based on a characteristic fragment ion. An example of such is the glycerol backbone of GPLs at m/z 153.0 that identifies phosphtidylserine (GPSer), Phosphatidic acid (GPA), phosphatidylglycerol (GPGro), phosphatidylinositol (GPIns), and cardiolipin in negative ions, or a FA scan to separate out GPLs based on their FA moieties, e.g., DHA (20, 25). Alternatively, triple quadrupole mass spectrometers also provide the ability to perform neutral loss scans to identify a common neutral fragment, such as the loss of didehydroalanine (87 Da) from GPSer in negative ions or phosphoethanolamine (141 Da) from GPEtn in positive ions. Such scans focusing on the head group fragments reduce the incidence of isobaric interference from other GPLs and are used for the quantification of GPLs by comparing to an internal standard, and varying the collision energy depending on the headgroup (25). In this chapter, a step-by-step description of how to utilize these various mass spectrometric scan techniques to identify the molecular glycerophospholipid distribution of DHA within tissues will be presented.
2. Materials 2.1. Equipment
1. Glass–glass homogenizers. 2. Tube rotator. 3. Vortex mixer. 4. UV 1601 spectrophotometer (Shimadzu Scientific Instruments, Colombia, U.S.A.). 5. Waters QuattroMicro™ triple quadrupole mass spectrometer (Waters, Manchester, U.K.).
2.2. Reagents and Solvents
All solvents must be of minimum HPLC grade. 1. Ammonium acetate.
22
Mitchell
2. Ammonium molybdate. 3. Butylated hydroxyltoluene (BHT). 4. Chloroform. 5. Hydrochloric acid. 6. Methanol. 7. Phospholipid Standards (all purchased from Avanti Polar Lipids):
(a) Dinonadecanoyl phosphatidylcholine, GPCho (19:0/ 19:0).
(b) Diheptadecanoyl phosphatidylserine, GPSer (17:0/17:0).
(c) Diheptadecanoyl phosphatidylethanolamine, GPEtn (17:0/17:0).
(d) Diheptadecanoyl phosphatidylglycerol, GPGro (17:0/ 17:0).
(e) Diheptadecanoyl phosphatidic acid, GPA (17:0/17:0).
(f) Heptadecanoyl eicosatetraenoyl phosphatidylinositol, GPIns (17:0/5Z,8Z,11Z,14Z-20:4).
8. Potassium dihydrogen phosphate. 9. Perchloric acid. 10. Stannous chloride. 2.3. Supplies
1. Pasteur pipettes. 2. Pyrex screw cap 15 mL test tubes. 3. Plastic cuvettes.
3. Methods A workflow outlining the procedures required for the identification and relative quantification of glycerophospholipids containing DHA is shown in Fig. 2. 3.1. Total Lipid Extraction
Total lipids are extracted from tissues according to traditional methods (26) with slight modifications to enhance the compatibility of the extracts with mass spectrometric analysis as described previously (27). In detail: 1. Weigh tissue and homogenize in 2 mL methanol:chlorofrorm (1:2 v/v) containing 0.01% butylated hydroxytoluene (BHT) using a glass–glass homogenizer. 2. Add internal standard mixture at 4 mL /g tissue (see Note 1).
Tracking the Glycerophospholipid Distribution of Docosahexaenoic Acid
23
Fig. 2. Workflow for tracking the glycerophospholipid distribution of docosahexaenoic acid.
24
Mitchell
3. Add further methanol:chlorofrorm (1:2 v/v) containing 0.01% BHT to ensure that the total methanol:chlorofrorm volume is 20× the tissue weight. 4. Vortex and mix in a test tube rotator for a minimum of 4 h or preferably overnight at 4°C. 5. Add 500 mL 0.15 M ammonium acetate and vortex for at least 15 s. 6. Centrifuge for 5 min at 2,000 × g. There should be two phases with chloroform at the bottom containing the lipids. Proteins float in the water/methanol phase or interface. 7. Aspirate a glass pipette with a small amount of chloroform and expel. Insert glass pipette gently to the bottom of the tube and take out the chloroform without removing any of the water/methanol phase or proteins. Expel the chloroform/ lipid mixture into a new test tube. 8. Add 2 mL methanol:chloroform (1:2 v/v) to original homogenate tube and vortex for at least 15 s. 9. Centrifuge for 5 min at 2,000 × g. 10. Aspirate glass pipette with a small amount of chloroform and expel. Insert glass pipette gently to the bottom of the tube and take out the chloroform without removing any of the water/methanol phase or proteins, combine with first chloroform extract. 11. Add 500 mL of 0.15 M ammonium acetate to the combined chloroform extract and vortex for at least 15 s. 12. Centrifuge for 5 min at 2,000 × g. 13. Insert glass pipette gently into the aqueous phase without removing any of the organic phase, and discard this aqueous phase and protein layer as waste. 14. Dry down under nitrogen at 37°C. 15. Resuspend phospholipids in 2 mL methanol:chloroform (2:1 v/v) and vortex. 16. Store in glass vial at −80°C. 3.2. Phosphorus Assay
The total phospholipid concentration of lipid extracts is determined by phosphorous assay (28). The concentration of phosphorous in the lipid extract is quantified by comparison of absorbance values (680 nm) with a standard reference curve (see Note 2). 1. Take 100 mL aliquots of each lipid extract (in duplicate), dry under nitrogen and resuspend in 0.8 mL of 72% (w/v) perchloric acid. 2. Heat at 190°C for 45 min. 3. Place on ice and add 5 mL of water and 500 mL each of ammonium molybdate (8%, w/v) and stannous chloride (0.005% dilution of 40% (w/v) SnCl2 in HCl).
Tracking the Glycerophospholipid Distribution of Docosahexaenoic Acid
25
4. Make up to 10 mL by the addition of water and allow color to develop for 10 min. 5. Measure absorbance at 680 nm. 6. Calculate phospholipid content using the following equation:
Phospholipid content =
µg phosphorous × 780 , 30.97
where 780 is the assumed average mass of phospholipids in grams and 30.97 is the molecular weight of phosphorous in grams. 3.3. Mass Spectrometry 3.3.1. Instrumentation 3.3.2. Identification of Phospholipids Containing DHA
The following discussion is based on the use of a Waters QuattroMicro™ triple quadrupole mass spectrometer (Waters, Manchester, U.K.) equipped with a z-spray electrospray ion source and controlled by Micromass Masslynx version 4.0 software. Performing this step will identify all phospholipids containing a DHA moiety, providing a targeted approach for later quantitative analysis. To identify anionic glycerophospholipids (GPA, GPGro, GPSer, GPIns, and GPEtn; see Note 3) containing DHA by precursor ion scanning in negative ion mode: 1. Dilute lipid extracts to a final phospholipid concentration of 40 mM with the addition of methanol:chloroform (2:1 v/v). 2. Set capillary voltage to 3,000 V, source temperature to 80°C, desolvation temperature to 120°C and Cone voltage to 50 V. Nitrogen drying gas is used at a flow rate of 320 L/h. 3. Infuse samples into the electrospray ion source at a flow rate of 10 mL/min using the instrument’s on-board syringe pump. 4. Set the argon collision gas at a pressure of 3 mTorr and accelerate the ions at a collision energy offset of 35 eV 5. Set quadrupole 3 (Q3) to m/z 327.3 and scan quadrupole 1 (Q1) over a mass range of m/z 740–920 For the identification of GPCho containing DHA by neutral loss scanning in positive ion mode: 1. Dilute lipid extracts to a final phospholipid concentration of 40 mM with the addition of methanol:chloroform (2:1 v/v) and add aqueous lithium acetate (to a final concentration of 200 mM). 2. Set capillary voltage to 3,000 V, source temperature 80 °C, desolvation temperature 120°C and Cone voltage to 35 V. Nitrogen drying gas is used at a flow rate of 320 L/h. 3. Infuse samples into the electrospray ion source at a flow rate of 10 mL/min using the instrument’s on-board syringe pump. 4. Set the argon collision gas at a pressure of 3 mTorr and accelerate the ions at a collision energy offset of 35 eV
Fig. 3. (a) A precursor ion scan for anionic glycerophospholipids containing DHA (m/z 327.3) and (b) a neutral loss scan for GPCho containing DHA (387.3 Da; GPCho – trimethylamine – DHA) in total lipids extracts from mouse skeletal muscle. GPA phosphatidic acid, GPEtn phosphatidylglycerol, GPSer phosphatidylserine, GPCho phosphatidylcholine, DHA docosahexaenoic acid.
26 Mitchell
Tracking the Glycerophospholipid Distribution of Docosahexaenoic Acid
27
5. Scan Q1 over a mass range of m/z 750–850 and scan Q3 at an m/z offset of Q1–387.3. This offset corresponds to the loss of both trimethylamine and DHA (see Note 4). 3.4. Results 3.4.1. Structural Characterization of DHA Containing Glycerophospholipids
The above scans provide spectra of all anionic phospholipid and GPCho molecules containing DHA as shown in Fig. 3a, b respectively. Product ion spectra can now be obtained from each of the DHAcontaining phospholipid molecules to complete their structural characterization. These spectra are obtained using the same instrument settings and sample preparations as described for both positive and negative ion analysis above with the exception that: 1. Q1 is set to the m/z of the deprotonated or lithiated molecular ion, and 2. Q3 is scanned over an appropriate range to identify all fatty acid moieties present, i.e., m/z 200–350 in negative ion mode for deprotonated ions or an m/z range that is between 400 and 200 Da less than the lithiated molecular ion in positive ion mode. If GPEtn or GPCho ethers are identified as containing DHA, their identity can be confirmed by the presence of ions indicative of the loss of the ether-linked acyl chain as an alcohol from lithiated ions as described by Hsu and Turk (29, 30). Alternatively, the identity of ether lipids can be confirmed by ozone-induced dissociation (OzID) (31) that is described in detail in Chapter 21.
3.4.2. Quantification of DHA Containing Glycerophospholipids
In order to remove isobaric interference across glycerophospholipid classes, quantification of molecular phospholipids is performed using precursor ion and neutral loss scans of head group-specific fragments (25). It is unlikely that all GPL classes will contain DHA and therefore the specific head group scans required are determined by the previous identification of molecular GPLs containing DHA. GPA, GPGro, and GPIns specific scans are performed in negative ion mode: 1. Dilute lipid extracts to a final phospholipid concentration of 40 mM with the addition of methanol:chloroform (2:1 v/v). 2. Set capillary voltage to 3,000 V, source temperature 80°C, desolvation temperature 120°C and Cone voltage to 50 V. Nitrogen drying gas is used at a flow rate of 320 L/h. 3. Infuse Samples into the electrospray ion source at a flow rate of 10 mL/min using the instrument’s on-board syringe pump. 4. Set the argon collision gas at a pressure of 3 mTorr. 5. Q1 and Q3 settings and collision energies are set as listed in Table 1. GPCho, GPSer, and GPEtn specific scans are performed in positive ion mode:
28
Mitchell
Table 1 Mass spectrometer scan parameters used for the relative quantification of GPA, GPGro, and GPIns GPL class
Scan type
Q1 scan range (m/z) Q2
Collision energy offset (eV)
GPA/GPGro Precursor ion 650–830
m/z 153.0 50
GPIns
m/z 241.0 45
Precursor ion 860–920
GPL glycerophospholipid, GPA phosphatidic acid, GPGro phosphatidylglycerol, GPIns phosphatidylinositol, m/z mass-to-charge ratio
Table 2 Mass spectrometry scan parameters used for the relative quantification of GPCho, GPEtn, and GPSer GPL class Scan type
Q1 scan range (m/z) Q2
Collision energy offset (eV)
GPCho
Precursor ion 780–840
m/z 184.1
35
GPEtn
Neutral loss
700–800
Q1–141.5 Da 25
GPSer
Neutral loss
750–840
Q1–185.4 Da 22
GPL glycerophospholipid, GPCho phosphatidylcholine, GPEtn phosphatidylethanolamine, GPSer phosphatidylserine; m/z, mass-to-charge ratio
1. Dilute lipid extracts to a final phospholipid concentration of 40 mM with the addition of methanol:chloroform (2:1 v/v). The formation of protonated GPSer and GPEtn ions can be enhanced by the addition aqueous ammonium acetate (to a final concentration of approximately 50 mM) (32). 2. Set capillary voltage to 3,000 V, source temperature 80°C, desolvation temperature 120°C, and Cone voltage to 35 V. Nitrogen drying gas is used at a flow rate of 320 L/h. 3. Infuse samples into the electrospray ion source at a flow rate of 10 mL/min using the instrument’s on-board syringe pump. 4. Set the argon collision gas at a pressure of 3 mTorr. 5. Q1 and Q3 settings and collision energies are set as listed in Table 2. Glycerophospholipids are then quantified by comparing their peak areas, obtained from averaging a minimum of 100 scans with the appropriate internal standard for each class after correction
Tracking the Glycerophospholipid Distribution of Docosahexaenoic Acid
29
Table 3 Molecular glycerophospholipids containing DHA identified in mouse skeletal muscle using shotgun lipidomics GPL
Proportion of Total GPL (%)
GPCho (16:0/22:6)
16.5 ± 0.4
GPCho (16:1/22:6)
0.6 ± 0.1
GPCho (18:0/22:6)
2.2 ± 0.2
GPCho (18:1/22:6)
0.9 ± 0.1
GPCho (18:2/22:6)
0.6 ± 0.1
GPEtn (16:0/22:6)
4.6 ± 0.3
GPEtn (18:0/22:6)
8.3 ± 0.5
GPEtn (18:1/22:6)
2.1 ± 0.1
GPEtn (18:2/22:6)
0.8 ± 0.1
GPSer (18:0/22:6)
3.6 ± 1.1
GPA (18:0/22:6)
0.2 ± 0.0
Total DHA-containing GPL
40.5 ± 1.3
GPL glycerophospholipid, GPCho phosphatidylcholine, GPEtn phosphatidylethanolamine, GPSer phosphatidylserine, GPA phosphatidic acid. Data are presented as mean ± SE (n = 4)
for isotope contributions as described by Deeley et al. (27) (see Note 5). To achieve this, the isotopic ion distribution of each phospholipid can be calculated from isotope models and the area of the monoisotopic peak multiplied by the calculated correction factor. This calculation must start with the smallest observed phospholipid so that any contribution of its isotope peaks to the area of phospholipids of greater m/z can be subtracted before subsequent isotope calculations are performed. Where two or more isomeric phospholipids are identified the relative abundance of each isomer can be determined from comparison of the combined abundances of the two fatty acid carboxylate ions arising from each lipid with the combined peak area of all carboxylate anions present in the product ion spectrum. Neither the relative position of the acyl chains on the glycerol backbone (33) (often called the sn-position) nor the position of double bonds (31, 34, 35) can be rigorously assigned from these data. A list of DHA-containing glycerophospholipids detected in mouse skeletal muscle using this technique is shown in Table 3.
30
Mitchell
4. Notes 1. An internal standard mixture in methanol:chloroform (2:1, v/v) should be prepared with the concentration of each internal standard reflecting the concentration of each phospholipid class within the tissue, e.g., for skeletal muscle use GPCho (19:0/19:0), 250 mM; GPEtn (17:0/17:0), 188 mM; GPSer (17:0/17:0), 125 mM; GPA (17:0/17:0), 25 mM; GPGro (17:0/17:0), 25 mM, and GPIns (17:0/20:4), 25 mM. 2. The standard used for the phosphorous assay is potassium dihydrogen phosphate (KH2PO4) at 20 mg/mL and the curve constructed using 1, 2, 5, and 10 mg of phosphorous. 3. GPEtn is technically a zwitterionic phospholipid, however it can be easily deprotonated (particularly at elevated pH) to form an anion. 4. Other ions characteristic of the neutral loss of DHA are also produced by the collision-induced dissociation of lithated GPCho; however, the neutral loss described here produces the most abundant of these ions under the described conditions. 5. Isotope corrections are required to account for (a) The greater contribution of isotopic ions to the total abundance of larger phospholipids. (b) The contribution of isotope peaks of one phospholipid to the area of the monoisotopic peak of a larger one. References 1. Dyerberg J, Bang HO, Hjorne N. (1975) Fatty acid composition of the plasma lipids in Greenland Eskimos. Am J Clin Nutr. 28(9):958–66. 2. McLennan P, Abeywardena M. (2005) Membrane basis for fish oil effects on the heart: Linking natural hibernators to prevention of human sudden cardiac death. J Membr Biol. 206(2):102. 3. Hibbeln JR, Salem N, Jr. (1995) Dietary polyunsaturated fatty acids and depression: when cholesterol does not satisfy. Am J Clin Nutr. 62(1):1–9. 4. Andersen G, Harnack K, Erbersdobler HF, Somoza V. (2008) Dietary eicosapentaenoic acid and docosahexaenoic acid are more effective than alpha-linolenic acid in improving insulin sensitivity in rats. Ann Nutr Metab. 52(3):250–6. 5. Carpentier YA, Portois L, Malaisse WJ. (2006) n-3 Fatty acids and the metabolic syndrome. Am J Clin Nutr. 83(6):S1499–504. 6. Li J-J, Huang CJ, Xie D. (2008) Anti-obesity effects of conjugated linoleic acid, docosahex-
aenoic acid, and eicosapentaenoic acid. Mol Nutr Food Res. 52(6):631–45. 7. Chapkin RS, Seo J, McMurray DN, Lupton JR. (2008) Mechanisms by which docosahexaenoic acid and related fatty acids reduce colon cancer risk and inflammatory disorders of the intestine. Chem Phys Lipids. 153(1):14–23. 8. Stillwell W. (2008) Docosahexaenoic acid: a most unusual fatty acid. Chem Phys Lipids. 153(1):1–2. 9. Feller SE. (2008) Acyl chain conformations in phospholipid bilayers: a comparative study of docosahexaenoic acid and saturated fatty acids. Chem Phys Lipids. 153(1):76–80. 10. Feller SE, Gawrisch K, MacKerell AD, Jr. (2002) Polyunsaturated fatty acids in lipid bilayers: intrinsic and environmental contributions to their unique physical properties. J Am Chem Soc. 124(2):318–26. 11. Stillwell W, Wassall SR. (2003) Docosahexaenoic acid: membrane properties of a unique fatty acid. Chem Phys Lipids. 126(1):1–27.
Tracking the Glycerophospholipid Distribution of Docosahexaenoic Acid
12. Huster D, Arnold K, Gawrisch K. (1998) Influence of Docosahexaenoic Acid and Cholesterol on Lateral Lipid Organization in Phospholipid Mixtures. Biochemistry. 37(49):17299–308. 13. Ehringer W, Belcher D, Wassall SR, Stillwell W. (1990) A comparison of the effects of linolenic (18:3[Omega]3) and docosahexaenoic (22:6[Omega]3) acids on phospholipid bilayers. Chem Phys Lipids. 54(2):79–88. 14. Turner N, Else PL, Hulbert AJ. (2003) Docosahexaenoic acid (DHA) content of membranes determines molecular activity of the sodium pump: implications for disease states and metabolism. Naturwissenschaften. 90(11):521–3. 15. Giorgione JR, Kraayenhof R, Epand RM. (1998) Interfacial membrane properties modulate Protein Kinase C Activation: Role of the Position of Acyl Chain Unsaturation. Biochemistry. 37(31):10956–60. 16. Slater SJ, Kelly MB, Taddeo FJ, Ho C, Rubin E, Stubbs CD. (1994) The modulation of protein kinase C activity by membrane lipid bilayer structure. J Biol Chem. 269(7):4866–71. 17. Farkas T, Kitajka K, Fodor E, et al. (2000) Docosahexaenoic acid-containing phospholipid molecular species in brains of vertebrates. Proc Natl Acad Sci USA. 97(12):6362–6. 18. Takamura H, Kito M. (1991) A highly sensitive method for quantitative analysis of phospholipid molecular species by high-performance liquid chromatography. J Biochem. 109:436–9. 19. Mitchell TW, Buffenstein R, Hulbert AJ. (2007) Membrane phospholipid composition may contribute to exceptional longevity of the naked mole-rat (Heterocephalus glaber); a comparative study using shotgun lipidomics. Exp Gerontol. 42(11):1053–62. 20. Ekroos K, Chernushevich IV, Simons K, Shevchenko A. (2002) Quantitative profiling of phospholipids by multiple precursor ion scanning on a hybrid quadrupole time-of-flight mass spectrometer. Anal Chem. 74:941–9. 21. Han X, Gross RW. (2003) Global analyses of cellular lipidomes directly from crude extracts of biological samples by ESI mass spectrometry: a bridge to lipidomics. J Lipid Res. 44(6): 1071–9. 22. Han X, Gross RW. (2005) Shotgun lipidomics: electrospray ionization mass spectrometric analysis and quantitation of cellular lipidomes directly from crude extracts of biological samples. Mass Spectrom Rev. 24:367–412. 23. Han X, Yang K, Yang J, Fikes KN, Cheng H, Gross RW. (2006) Factors influencing the electrospray intrasource separation and selective ionization of glycerophospholipids. J Am Soc Mass Spectrom. 17(2):264–74.
31
24. Pulfer M, Murphy RC. (2003) Electrospray mass spectrometry of phospholipids. Mass Spectrom Rev. 22:332–64. 25. Brugger B, Erben G, Sandhoff R, Wieland FT, Lehmann WD. (1997) Quantitative analysis of biological membrane lipids at the low picomole level by nano-electrospray ionisation tandem mass spectrometry. Proc Natl Acad Sci U S A. 94:2339–44. 26. Folch J, Lees M, Sloane-Stanley GH. (1957) A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem. 226:497–509. 27. Deeley JM, Mitchell TW, Nealon JR, et al. (2008) Human lens lipids differ markedly from those of commonly used experimental animals. Biochim Biophys Acta. 1781(6–7):288–98. 28. Mills GL, Lane PA, Weech PK. (1984) A guide to lipoprotein technique. In: Burdon RH, Kippenberg PHv, eds. Laboratory Techniques in Biochemistry and Molecular Biology. Elsevier Science, New York, pp. 240–1. 29. Hsu FF, Turk J. (2000) Characterization of phosphatidylethanolamine as a lithiated adduct by triple quadrupole tandem mass spectrometry with electrospray ionization. J Mass Spectrom. 35(5):595–606. 30. Hsu FF, Turk J, Thukkani AK, Messner MC, Wildsmith KR, Ford DA. (2003) Characterization of alkylacyl, alk-1-enylacyl and lyso subclasses of glycerophosphocholine by tandem quadrupole mass spectrometry with electrospray ionization. J Mass Spectrom. 38(7): 752–63. 31. Thomas MC, Mitchell TW, Harman DG, Deeley JM, Nealon JR, Blanksby SJ. (2008) Ozone-induced dissociation: Elucidation of double bond position within mass-selected lipid ions. Anal Chem. 80(1):303–11. 32. Thai TP, Rodemer C, Worsch J, Hunziker A, Gorgas K, Just WW. (1999) Synthesis of plasmalogens in eye lens epithelial cells. FEBS Lett. 456(2):263–8. 33. Ekroos K, Ejsing Christer S, Bahr U, Karas M, Simons K, Shevchenko A. (2003) Charting molecular composition of phosphatidylcholines by fatty acid scanning and ion trap MS3 fragmentation. J Lipid Res. 44(11):2181–92. 34. Thomas MC, Mitchell TW, Blanksby SJ. (2006) Ozonolysis of phospholipid double bonds during electrospray ionization: A new tool for structure determination. J Am Chem Soc. 128(1):58–9. 35. Thomas MC, Mitchell TW, Harman DG, Deeley JM, Murphy RC, Blanksby SJ. (2007) Elucidation of double bond position in unsaturated lipids by ozone electrospray ionization mass spectrometry. Anal Chem. 79(13):5013–22.
sdfsdf
Chapter 3 Global Analysis of Retina Lipids by Complementary Precursor Ion and Neutral Loss Mode Tandem Mass Spectrometry Julia V. Busik, Gavin E. Reid, and Todd A. Lydic Summary Despite an increasing recognition of the causative and diagnostic role of lipids in the onset and progression of retinal disease, information on the global lipid profile of the normal retina is quite limited. Here, a “shotgun” tandem mass spectrometry approach involving the use of multiple lipid class-specific precursor ion and neutral loss scan mode experiments has been employed to analyze lipid extracts from normal rat retina, obtained with minimal sample handling prior to analysis. Redundant information for the identification and characterization of molecular species in each lipid class was obtained by complementary analysis of their protonated or deprotonated precursor ions, or by analysis of their various ionic adducts (e.g., Na+, NH4+, Cl–, CH3OCO2–). Notably, “alternative” precursor ion or neutral loss scan mode MS/ MS experiments are introduced that were used to identify rat retina lipid molecular species that were not detected using “conventional” scan types typically employed in large-scale lipid-profiling experiments. This chapter outlines the principles and advantages of utilizing complementary/redundant identification of lipid species as a strategy to overcome inherent challenges and limitations of shotgun lipid analysis, and provides examples of the application of this strategy in the analysis of the retina lipidome. Key words: Retina, Lipidomics, Lipid analysis, Electrospray ionization, Tandem mass spectrometry
1. Introduction Retina has a unique fatty acid profile with the highest levels of long chain polyunsaturated fatty acids (LCPUFA), including docosahexaenoic acid (DHA22:6n3), and arachidonic acid (ARA20:4n6), observed in the body (1–5). Of these, DHA22:6n3 is the most abundant fatty acid in both neural and vascular elements of the retina (2), retinal pigment epithelial cells (6, 7) and retinal
Donald Armstrong (ed.), Lipidomics, Methods in Molecular Biology, vol. 579, DOI 10.1007/978-1-60761-322-0_3, © Humana Press, a part of Springer Science + Business Media, LLC 2009
33
34
Busik, Reid, and Lydic
photoreceptor outer segment disc membranes (8, 9). Extensive studies clearly demonstrate the important role of lipids in retinal health and disease (5). However, most studies to date have focused on the role of LCPUFA’s, and DHA in particular, measuring total fatty acid levels without obtaining structural information. The reason for this is likely methodological – LCPUFA are very abundant in the retina and relatively easy to measure by well developed high-performance liquid chromatography (HPLC) or gas chromatography (GC) techniques. The limitations of the traditional techniques have precluded comprehensive complex lipid analysis from the limited amount of retinal material that can be obtained from animal models such as rats and mice. Thus, there is surprisingly little information available regarding the lipid composition of the normal retina, and only limited information describing changes in global lipid profiles between normal and diseased retinal tissue (See Note 1). Recent advances in the application of electrospray ionization (ESI) and matrix assisted laser desorption/ionization (MALDI) (10–20) techniques, coupled with the use of tandem mass spectrometry methods employing selective precursor ion and neutral loss scan mode analysis strategies, have enabled the development of “shotgun” lipidomics approaches for rapid and sensitive monitoring of the molecular compositions and abundances of individual lipid species in unfractionated lipid extracts (10–17). While shotgun approaches allow for high-throughput analysis of multiple lipid classes without prior chromatographic separation of lipid analytes, the large number of lipid molecular species present in crude extracts presents a significant challenge for the analyst. In addition to possible overlap of the molecular ions and the 13C isotope peaks of numerous lipid species, even greater extract complexity can arise due to the presence of individual lipid species in multiple ionic forms, via adduction with a variety of cationic (e.g., +H+, +Na+, +NH4+) or anionic (e.g., −H–, +Cl–) species in positive or negative ion modes, respectively, that may be present in small amounts following extraction of lipids from tissues or cells. Additional complexity may also be observed for certain lipids due to the presence of adducts formed by reaction of the solvents and buffer additives that are commonly employed for sample analysis. For example, phosphocholine-containing lipids are readily adducted with methylcarbonate (CH3OCO2–) anions (21), formed by the in vitro reaction of hydroxide or bicarbonate salts (22, 23) with methanol, effectively increasing the number of lipid species observed in negative ion mode analysis. In our studies, relatively limited tissue availability has prompted us to develop strategies for attaining a thorough accounting of the global lipid composition of retina without requirement for multiple sample fractionation or processing steps, such as chromatographic separation, or destruction of glycerophospholipids
Global Analysis of Retina Lipids
35
for enhanced sphingolipid analysis. Utilizing a triple quadrupole mass spectrometer to perform multiple precursor ion and neutral loss scan mode MS/MS experiments, we have found that complementary/redundant detection of a given lipid class based on the unique fragmentation behaviors of various ionic forms (e.g.,[M + H]+, [M − H]−, [M + Na]+, [M + NH4]+, [M + Cl]−, [M + CH3OCO2]−) of various lipid classes facilitates (a) the ability to identify molecular lipid species that may not be detected when only one scan mode is used, (b) a more thorough accounting of the various lipid species that may be present in multiple ionic forms when “absolute quantification” is desired, and (c) simplification of relative quantification by selection of an MS/MS scan mode in which lipid species are present in only one ionic form. Furthermore, the ability to identify lipids in more than one ionic form greatly increases the confidence for peak identification, even for ions observed at low (R(in Munster-3B). 34:
37–47
DLATVYVDVLK
1,235.6881
618.348
52–64
DYVSQFEGSALGK
1,400.6692
700.838
65–69
QLNLK
615.3824
308.195
70–83
LLDNWDSVTSTFSK
1,612.7853
806.896
86–101
EQLGPVTQEFWDNLEK
1,932.9337
966.971
102–107
ETEGLR
704.3573
352.682
108–112
QEMSK
622.2865
311.647
113–118
DLEEVK
732.3774
366.692
113:
121–130
VQPYLDDFQK
1,252.6208
626.814
126: 127:
132–140
WQEEMELYR
1,283.5725
642.29
132: 134:E->K(in Fukuoka).
143–147
VEPLR
613.3668
307.187
148–155
AELQEGAR
873.4424
437.225
158–164
LHELQEK
896.4836
448.745
160:E->K(in Norway). 163:
165–173
LSPLGEEMR
1,031.519
516.263
167: 168:L->R(in Zaragoza). 171:
178–184
AHVDALR
781.4315
391.219
180: 184:R->P(indbSNP
185–195
THLAPYSDELR
1,301.6484
651.328
189:P->R.
202–206
LEALK
573.3606
287.184
207–212
ENGGAR
603.2845
302.146
213–219
LAEYHAK
831.4359
416.222
220–230
ATEHLSTLSEK
1,215.6215
608.314
231–239
AKPALEDLR
1,012.5785
506.793
240–250
QGLLPVLESFK
1,230.7092
615.858
251–262
VSFLSALEEYTK
1,386.7151
693.861
61:A->T
92:
222:
APOA1_HUMAN was subjected to tryptic digestion using Peptidemass (http://www.expasy.org/ tools/peptide-mass.html). The number of missed cleavages was set at 0 and all cysteines were treated with iodoacetamide to form carbamidomethyl-cysteine. Methionines were oxidized to methionine sufoxide and only peptide masses >500 Da are displayed. Monoisotopic masses ([M + H+], m/z = 1) and the masses with a charge state (m/z) of 2 (default for the ion trap mass spectrometer) are shown. All known variants are also shown.
Combining Lipidomics and Proteomics of Human Cerebrospinal Fluids
83
This approach allows for the complete understanding of biological pathways and should be useful in detecting changes in protein expression for brain diseases. Any change in lipids in combination with proteins should better illuminate strategies for biomarker discovery or molecular targets for drugs.
4. Notes 1. Lipids with polyunsaturated fatty acids are susceptible to oxidation. Enzymes at room temperature may also rapidly degrade some lipid mediators and signaling lipids. It is important to protect lipids from oxidation by adding BHT and by storing samples under inert conditions. We flush air out of our sample tubes before storage at −80°C. Once lipids are extracted, they are similarly stored in organic solvents until required for LC/MS2. In cases where specific enzyme inhibitors are known, these can be incorporated into the samples to avoid degradation of lipids. 2. While ESI is normally used for lipids, others have reported efficient use of other ionization methods. These include chemical ionization or photo ionization under atmospheric conditions (58). We have used the triple quadrupole mass spectrometer for most of our studies. However, iontrap mass spectrometers and high resolution linear FT/MS have also been used for lipidomic studies (59). 3. Several databases can be used to search mass spectrometer data of lipids. So far, these are proprietary and not universally available. These may also be instrument- and/or method-specific especially when attempts are made to identify molecular species of lipids. With accurate mass detection and increase in sensitivity of instruments, it may be possible to link individual lipidomic studies with the lipid maps database (http://www. lipidmaps.org/) for identification of lipid molecular species. 4. High-abundant proteins may be removed from samples to increase the dynamic range. This allows detection of the less abundant proteins in CSF. 5. Sample preparation is a key step in shotgun sequencing of proteins. For denaturation, urea or guanidine HCl can be used in place of Rapigest™. However, these have to be removed before mass spectrometry since these compounds may reduce the ionization of peptides of interest. 6. Other proteolytic enzymes and other buffer conditions can also be used instead of, or in combination, with trypsin. Such combinations may increase sequence coverage and increase the number of proteins detected in a sequencing experiment (52).
84
Fonteh and Fisher
7. 2-Dimensional liquid chromatography can be used to improve the separation and detection of peptides. Normally, the first dimension is a cation exchange resin that is followed by a reverse phase column. The Pepfinder kit (ThermoFisher, San Jose, CA) has been effectively used to improve extraction of peptides, removal of salts, and thus enables detection of thousands of proteins in human CSF without prior removal of abundant proteins (52). 8. The default charge state of the ion trap mass spectrometer is 2. Ions of charge state 1 or 2 may appear when a full scan is performed on the triple quadrupole mass spectrometer. For some peptides, the major charge state may be 1 or it may be 2 for others. We use the more intense ion for MS/MS and MRM studies.
Acknowlegments We thank Susan Onami and Elena Oborina for technical help. We appreciate the support and critical discussion with Drs. Michael Harrington, Andreas Hulmer, and Roger Biringer. This work was supported in part by NIH grants RO1# NS43295; institutional support (HMRI), the Norris, Jamison and Glide Foundations and donations from the Dunlevey, Hezlep and Posthuma families. References 1. Fonteh, A. N., Harrington0, R. J., Huhmer, A. F., Biringer, R. G., Riggins, J. N., and Harrington, M. G. (2006) Identification of disease markers in human cerebrospinal fluid using lipidomic and proteomic methods. Dis Markers 22(1–2), 247–272. 2. Albers, J. J. (1978) Effect of human plasma apolipoproteins on the activity of purified lecithin:cholesterol acyltransferase. Scand J Clin Lab Invest Suppl 150, 48–52. 3. Beffert, U., Danik, M., Krzywkowski, P., Ramassamy, C., Berrada, F., and Poirier, J. (1998) The neurobiology of apolipoproteins and their receptors in the CNS and Alzheimer’s disease. Brain Res Brain Res Rev 27(2), 119–142. 4. Haffner, S., Applebaum-Bowden, D., Wahl, P. W.et al. (1985) Epidemiological correlates of high density lipoprotein subfractions, apolipoproteins A-I, A-II, and D, and lecithin cholesterol acyltransferase. Effects of smoking, alcohol, and adiposity. Arteriosclerosis 5(2), 169–177.
5. Bazan, N. G., and Allan, G. (1996) Plateletactivating factor is both a modulator of synaptic function and a mediator of cerebral injury and inflammation. Adv Neurol 71, 475–482. 6. Brady, H. R., and Serhan, C. N. (1996) Lipoxins: putative braking signals in host defense, inflammation and hypersensitivity. Curr Opin Nephrol Hypertens 5(1), 20–27. 7. Fonteh,vv A. N., and Harrington, M. G. (2004) Remodeling of arachidonate and other polyunsaturated fatty acids in Alzheimer’s disease. In: Fonteh AN, Wykle RL, editors. Arachidonate Remodeling and Inflammation. Birkhauser Verlag 145–168. 8. Ford-Hutchinson, A. W. (1990) Leukotriene B4 in inflammation. Crit Rev Immunol 10(1), 1–12. 9. Lefkowith, J. B. (1988) Essential fatty acid deficiency inhibits the in vivo generation of leukotriene B4 and suppresses levels of resident and elicited leukocytes in acute inflammation. J Immunol 140(1), 228–233.
Combining Lipidomics and Proteomics of Human Cerebrospinal Fluids
10. O’Banion, M. K. (1999) COX-2 and Alzheimer’s disease: potential roles in inflammation and neurodegeneration. Expert Opin Investig Drugs 8(10), 1521–1536. 11. Peroutka, S. J. (2005) Neurogenic inflammation and migraine: implications for the therapeutics. Mol Interv 5(5), 304–311. 12. Pruzanski, W., and Vadas, P. (1989) Phospholipase A2 and inflammation. Ann Rheum Dis 48(11), 962–963. 13. Serhan, C. N. (1996) Inflammation. Signalling the fat controller. Nature 384(6604), 23–24. 14. Zurier, R. B. (1993) Fatty acids, inflammation and immune responses. Prostaglandins Leukot Essent Fatty Acids 48(1), 57–62. 15. Axelrod, J. (1995) Phospholipase A2 and G proteins. Trends Neurosci 18(2), 64–65. 16. Bazan, N. G., Allan, G., and Rodriguez de Turco, E. B. (1993) Role of phospholipase A2 and membrane-derived lipid second messengers in membrane function and transcriptional activation of genes: implications in cerebral ischemia and neuronal excitability. Prog Brain Res 96, 247–257. 17. Bonventre, J. V. (1992) Phospholipase A2 and signal transduction. J Am Soc Nephrol 3(2), 128–150. 18. Clark, J. D., Schievella, A. R., Nalefski, E. A., and Lin, L. L. (1995) Cytosolic phospholipase A2. J Lipid Mediat Cell Signal 12(2–3), 83–117. 19. Dennis, E. A. (2000) Phospholipase A2 in eicosanoid generation. Am J Respir Crit Care Med 161(2 Pt 2), S32–S35. 20. Diez, E., Chilton, F. H., Stroup, G., Mayer, R. J., Winkler, J. D., and Fonteh, A. N. (1994) Fatty acid and phospholipid selectivity of different phospholipase A2 enzymes studied by using a mammalian membrane as substrate. Biochem J 301(Pt 3), 721–726. 21. Farooqui, A. A., Antony, P., Ong, W. Y., Horrocks, L. A., and Freysz, L. (2004) Retinoic acid-mediated phospholipase A2 signaling in the nucleus. Brain Res Brain Res Rev 45(3), 179–195. 22. Fonteh, A. N., Bass, D. A., Marshall, L. A., Seeds, M., Samet, J. M., and Chilton, F. H. (1994) Evidence that secretory phospholipase A2 plays a role in arachidonic acid release and eicosanoid biosynthesis by mast cells. J Immunol 152(11), 5438–5446. 23 Kramer, R. M., Stephenson, D. T., Roberts, E.F., and Clemens, J. A. (1996) Cytosolic phospholipase A2 (cPLA2) and lipid mediator release in the brain. J Lipid Mediat Cell Signal 14(1–3), 3–7.
85
24. Kudo, I., and Murakami, M. (2002) Phospholipase A2 enzymes. Prostaglandins Other Lipid Mediat 68–69, 3–58. 25. Cocco, L., Martelli, A. M., Gilmour, R. S., Rhee, S. G., and Manzoli, F. A. (2001) Nuclear phospholipase C and signaling. Biochim Biophys Acta 1530(1), 1–14. 26. Cockcroft, S. (2006) The latest phospholipase C, PLCeta, is implicated in neuronal function. Trends Biochem Sci 31(1), 4–7. 27. Manzoli, L., Billi, A. M., Martelli, A. M., and Cocco, L. (2004) Regulation of nuclear phospholipase C activity. Acta Biochim Pol 51(2), 391–395. 28. Matsushima, H., Shimohama, S., Kawamata, J., Fujimoto, S., Takenawa, T., and Kimura, J. (1998) Reduction of platelet phospholipase C-delta1 activity in Alzheimer’s disease associated with a specific apolipoprotein E genotype (epsilon3/epsilon3). Int J Mol Med 1(1), 91–93. 29. Minke, B. (2001) The TRP channel and phospholipase C-mediated signaling. Cell Mol Neurobiol 21(6), 629–643. 30. Nakahara, M., Shimozawa, M., Nakamura, Y., et al. (2005) A novel phospholipase C, PLC(eta)2, is a neuron-specific isozyme. J Biol Chem 280(32), 29128–29134. 31. Rhee, S. G. (2001) Regulation of phosphoinositide-specific phospholipase C. Annu Rev Biochem 70, 281–312. 32. Shimohama, S., Sasaki, Y., Fujimoto, S.et al. (1998) Phospholipase C isozymes in the human brain and their changes in Alzheimer’s disease. Neuroscience 82(4), 999–1007. 33. Stewart, A. J., Mukherjee, J., Roberts, S. J., Lester, D., and Farquharson, C. (2005) Identification of a novel class of mammalian phosphoinositol-specific phospholipase C enzymes. Int J Mol Med 15(1), 117–121. 34. Cockcroft, S. (2001) Signalling roles of mammalian phospholipase D1 and D2. Cell Mol Life Sci 58(11), 1674–1687. 35. Jenkins, G. M., and Frohman, M. A. (2005) Phospholipase D: a lipid centric review. Cell Mol Life Sci 62(19–20), 2305–2316. 36. Klein, J. (2005) Functions and pathophysiological roles of phospholipase D in the brain. J Neurochem 94(6), 1473–1487. 37. Xie, Z., Ho, W. T., Spellman, R., Cai, S., and Exton, J. H. (2002) Mechanisms of regulation of phospholipase D1 and D2 by the heterotrimeric G proteins G13 and Gq. J Biol Chem 277(14), 11979–11986. 38. Balazy, M. (2004) Eicosanomics: targeted lipidomics of eicosanoids in biological systems. Prostaglandins Other Lipid Mediat 73(3–4), 173–180.
86
Fonteh and Fisher
39. Fitzpatrick, F. A., and Murphy, R. C. (1988) Cytochrome P-450 metabolism of arachidonic acid: formation and biological actions of “epoxygenase”-derived eicosanoids. Pharmacol Rev 40(4), 229–241. 40. Lands, W. E. (1993) Eicosanoids and health. Ann N Y Acad Sci 676, 46–59. 41. Murphy, R. C. (2001) Free-radical-induced oxidation of arachidonoyl plasmalogen phospholipids: antioxidant mechanism and precursor pathway for bioactive eicosanoids. Chem Res Toxicol 14(5), 463–472. 42. Serhan, C. N., Lu, Y., Hong, S., and Yang, R. (2007) Mediator lipidomics: search algorithms for eicosanoids, resolvins, and protectins. Methods Enzymol 432, 275–317. 43. Smith, W. L. (1989) The eicosanoids and their biochemical mechanisms of action. Biochem J 259(2), 315–324. 44. Bazan, N. G., Squinto, S. P., Braquet, P., Panetta, T., and Marcheselli, V. L. (1991) Platelet-activating factor and polyunsaturated fatty acids in cerebral ischemia or convulsions: intracellular PAF-binding sites and activation of a fos/jun/AP-1 transcriptional signaling system. Lipids 26(12), 1236–1242. 45. Benveniste, J., Chignard, M., Le Couedic, J. P., and Vargaftig, B. B. (1982) Biosynthesis of platelet-activating factor (PAF-ACETHER). II. Involvement of phospholipase A2 in the formation of PAF-ACETHER and lyso-PAFACETHER from rabbit platelets. Thromb Res 25(5), 375–385. 46. Snyder, F. (1994) Metabolic processing of PAF. Clin Rev Allergy 12(4), 309–327. 47. Farooqui, A. A., Litsky, M. L., Farooqui, T., and Horrocks, L. A. (1999) Inhibitors of intracellular phospholipase A2 activity: their neurochemical effects and therapeutical importance for neurological disorders. Brain Res Bull 49(3), 139–153. 48. Berenbaum, F. (1995) Phospholipase A2 inhibitors: a challenge for the future. Rev Rhum Engl Ed 62(6), 409–414. 49. Glaser, K. B., Lock, Y. W., and Chang, J. Y. (1991) PAF and LTB4 biosynthesis in the human neutrophil: effects of putative inhibitors of phospholipase A2 and specific inhibitors of 5- lipoxygenase. Agents Actions 34(1–2), 89–92. 50. Bligh, E. A., and Dyer, W. T. (1959) A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37, 911–917.
51. Fonteh, A. N. (1999) Assessment of arachidonic acid distribution into phospholipids of inflammatory cells. Methods Mol Biol 120, 77–89. 52. Biringer, R. G., Amato, H., Harrington, M. G., Fonteh, A. N., Riggins, J. N., and Huhmer, A. F. (2006) Enhanced sequence coverage of proteins in human cerebrospinal fluid using multiple enzymatic digestion and linear ion trap LC-MS/MS. Brief Funct Genomic Proteomic 5(2), 144–153. 53. Wolters, D. A., Washburn, M. P., and Yates, III J.R. (2001) An automated multidimensional protein identification technology for shotgun proteomics. Anal Chem 73(23), 5683–5690. 54. Amanchy, R., Kalume, D. E., and Pandey, A. (2005) Stable isotope labeling with amino acids in cell culture (SILAC) for studying dynamics of protein abundance and posttranslational modifications. Sci STKE 267, l2. 55. Romijn, E. P., Christis, C., Wieffer, M. et al. (2005) Expression clustering reveals detailed co-expression patterns of functionally related proteins during B cell differentiation: a proteomic study using a combination of onedimensional gel electrophoresis, LC-MS/MS, and stable isotope labeling by amino acids in cell culture (SILAC). Mol Cell Proteomics 4(9), 1297–1310. 56. Fonteh, A. N., Biringer, R. G., Huhmer, A. F., Rush, J. D., and Harrington, M. G. (2005) Use of [13C],[15N]-peptide standards to quantify enzymes of the Cyclooxygenase and Lipoxygenase pathways in human cerebrospinal fluids. American Society for Mass Spectrometry: Conference Proceeding 5–6(http://www. asms.org/asms05pdf/A052054.pdf). 57. Huhmer, A. F., Biringer, R. G., Amato, H., Fonteh, A. N., and Harrington, M. G. (2006) Protein analysis in human cerebrospinal fluid: Physiological aspects, current progress and future challenges. Dis Markers 22(1–2), 211–234. 58. Lee, S. H., and Blair, I. A. (2007) Targeted chiral lipidomics analysis by liquid chromatography electron capture atmospheric pressure chemical ionization mass spectrometry (LC-ECAPCI/ MS). Methods Enzymol 433, 159–174. 59. Schwudke, D., Liebisch, G., Herzog, R., Schmitz, G., and Shevchenko, A. (2007) Shotgun lipidomics by tandem mass spectrometry under data-dependent acquisition control. Methods Enzymol 433, 175–191.
Part II Analytical Approaches
sdfsdf
Chapter 5 Lipid Profiling Using Two-Dimensional Heteronuclear Single Quantum Coherence NMR Engy A. Mahrous, Robin B. Lee, and Richard E. Lee Summary The use of NMR spectroscopy in lipid research has been traditionally reserved for the analysis and structural elucidation of discrete lipid molecules. Although NMR analysis of organic molecules provides a plethora of structural information that is normally unattainable by most other techniques, its use for global analysis of mixed lipid pools has been hampered by its relatively low sensitivity and overlapping of signals in the spectrum. However, the last few decades have witnessed great advancements in NMR spectroscopy that generally resulted in greater sensitivity and offered more flexibility in sampling techniques. The method discussed in this chapter describes the use of NMR for global lipidome analysis. This methodology benefits from the quantitative nature of this technique together with the abundance of the structural information it can offer, while partially overcoming the problems of low sensitivity and overlapping signals through isotope-enrichment and the use of multidimensional NMR, respectively. We have applied this method successfully to the mycobacterial lipidome as an example of an organism with a very complex and chemically diverse lipid pool. The same concept is applicable to a wide range of prokaryotes that can grow in the laboratory in well-defined growth media. Key words: NMR spectroscopy, Mycobacteria, 2D-HSQC, Cell wall lipids, Polyketides, Glycolipids
1. Introduction The genus Mycobacteria is characterized by a unique cell wall rich in complex lipids, polyketides, and polysaccharides. Historically, numerous studies of individual mycobacterial lipids and the way they participate in pathogenesis and virulence have been conducted (1–3). Early attempts for global analysis of the mycobacterial lipidome began in the 1960s as a search for a taxonomical tool to assign newly discovered species in their right taxonomical order and to help identify clinical isolates at Donald Armstrong (ed.), Lipidomics, Methods in Molecular Biology, vol. 579, DOI 10.1007/978-1-60761-322-0_5, © Humana Press, a part of Springer Science + Business Media, LLC 2009
89
90
Mahrous, Lee, and Lee
the species level (4–7). A major breakthrough in this field was achieved by Dobson and Minnikin who reported five 2D-TLC systems in the 1980s to analyze the mycobacterial lipidome (8). Despite its wide application the methodology remained less than satisfactory. Problems associated with poor staining, low sensitivity, and limited resolution on one hand and the recent advancement of bioanalytical tools on the other, have fueled a research effort for new methodologies to study the mycobacterial lipidome. Recently, two new methodologies based on mass spectroscopy have been described. The first uses Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) (9). The second uses prior separation with HPLC followed by analysis with electrospray ionization HPLC/ ESI-MS (10). These analytical techniques are highly sensitive, however, they require significant sample manipulation to overcome the phenomena of ion suppression within lipid mixtures thus are best used for the analysis of readily ionizable molecules (9, 10). The method discussed in this chapter focuses on the application of NMR spectroscopy as a complementary procedure that can be utilized to rapidly analyze the composition of the mycobacterial lipid pool directly from crude lipid extracts (11). It offers a great tool for investigating the mode of action of antituberculosis drugs, determining gene function, understanding the stringent response of Mycobacteria under stress conditions that resemble the in vivo environment, and investigating differences between species or strains of Mycobacteria. To maximize its benefit, this lipid-profiling technique can be used in a complementary fashion with other genomic and proteomic methods. As it is the case with other “omics” techniques, the great utility of lipidomics methods is that they provide a wealth of information regarding the total composition of the lipid pools, from which a small number of candidate molecules can then be defined for more comprehensive follow-up studies through purification and isolation of such molecules of interest. For introduction of this method, we applied the technique to the lipid pool of Mycobacterium tuberculosis, the etiological agent of tuberculosis, which currently infects one-third of the world population and claimed 1.7 million lives worldwide in 2006 (12). The cell wall of M. tuberculosis, with its associated lipids, is believed to play an important role in its pathogenesis and virulence. The currently accepted structural and functional model of the mycobacterial cell wall describes it as a three-layer structure (13). The innermost layer consists of the plasma membrane, followed to the outside by the peptidoglycan layer, which is then covalently attached to the outer mycolyl arabinogalactan complex. The unusual long chain mycolic acids form a permeability barrier where a mixture of lipids, proteins, and glycolipids are loosely bound to the outside of the cell wall. Some of these molecules are surface
Lipid Profiling Using Two-Dimensional Heteronuclear Single Quantum
91
exposed and are free to interact with the host immune system (14–16), which may contribute to virulence and pathogenicity. This population of lipids can be easily extracted and analyzed by NMR spectroscopy. As NMR measures properties of nuclei rather than the physical property of the entire molecule, NMR signals are less likely to be affected by the different physicochemical properties exhibited by different classes of lipids (molecular weight, polarity, ionization potential, etc.). This uniformity of response is advantageous because it limits artifacts and confers quantitative representation of all lipids regardless of their structure. However, several challenges exist for using NMR to study complex pools of metabolites. First is the low sensitivity associated with NMR analysis. This can be partially overcome by enriching the cell lipidome with 13 C isotope through the replacement of the carbon source in the growth media with 13C-labeled precursors. Second, is the complexity and frequent overlapping of the NMR signals in the onedimensional spectrum. This problem can be addressed through the use of two- and three-dimensional NMR. Two-dimensional 1 H-13C heteronuclear single quantum coherence (HSQC) NMR can be used to rapidly generate lipid profiles, while more sophisticated three-dimensional experiments can be used to resolve the lipid profile and help assign signals when necessary. These experiments require no specialized equipment and can be performed using a standard high-field NMR spectrometer. Third, data analysis of such complex data requires extra consideration and efforts. To ensure data quality, signals in the NMR need to be referenced in both their chemical shift and relative intensity to an internal reference that produces a discrete signal in the NMR spectrum. To enhance reproducibility, external references are required to calibrate the NMR magnet and account for any observed variation due to instability of the magnet overtime. The data must then be normalized based on these internal and external standards using NMR software that confers robust analysis of large data sets. Specific NMR acquisition and data processing methods will not be discussed here as they are highly dependent on the instrumentation and software available. The following sections describe methods for sample preparation and analysis of mycobacterial lipid profiles using 2D NMR spectroscopy.
2. Materials 2.1. Equipment
1. High-field nuclear magnetic resonance spectrometer. 2. NMR processing software.
92
Mahrous, Lee, and Lee
2.2. Reagents and Supplies 2.2.1. Cell Culture
1. Sterile solution of Difco Middlebrook 7H9 Becton Dickinson and company, Sparks, MD. 2. Sterile 10× ADG supplement (2% glucose, 2% glycerol, 0.8% sodium chloride, 5% Bovine serum albumin (BSA) fraction V, Calbiochem, La Jolla, CA). 3. Sterile 10× solution of 5% Bovine serum albumin/0.8% sodium chloride. 4. Sterile solution of 10% 13C6 glucose Cambridge Isotope Laboratories, Andover, MA. 5. Sterile solution of 10% 13C3 glycerol Cambridge Isotope Laboratories, Andover, MA. 6. Sterile solution of 10% Tween-80, Calbiochem, La Jolla, CA.
2.2.2. Lipid Extraction
1. Deuterium oxide D2O 99% Cambridge Isotope Laboratories, Andover, MA. 2. Deuterated chloroform CDCl3 99% Cambridge Isotope Laboratories, Andover, MA. 3. Deuterated methanol CD3OD Cambridge Isotope Laboratories, Andover, MA.
2.2.3. Mycolic Acid Extraction
1. 1 M methanolic solution of KOH. 2. Diethyl ether. 3. Concentrated HCl.
3. Methods 3.1. Cell Culture and Preparation of Cell Pellet
To generate a good 2D HSQC spectrum in a short acquisition time, it is necessary to enrich the cell lipidome with 13C isotope. This can be achieved through the replacement of the carbon source in the growth media with 13C-labeled lipid precursors such as U-13C2 acetate, U-13C6 glucose, or U-13C3 glycerol. To attain a well-enriched sample, it is best to use a defined medium and limit any carbon sources to 13C as much as possible. Culture can also be passed in 13 C medium to further increase enrichment. Although the procedure discussed here focuses mainly on profiling the overall free cell wall lipid populations, the method can be easily modified to target a specific lipid population. The global 13C-enrichement strategy using 13C glucose and glycerol can be replaced by a biosynthetically guided approach that uses specific 13C-labeled precursors in order to enrich the signals for certain lipid targets. Cells can be grown or treated under various conditions to probe the response of the bacteria to different environmental conditions such as
Lipid Profiling Using Two-Dimensional Heteronuclear Single Quantum
93
nutrient starvation, drug treatment, pH, oxidative stress, or media composition. The time 13C label is incorporated can be varied based on the conditions tested. For example, studies involving the response of bacteria to drug treatment may use either cells fully labeled prior to drug treatment to observe changes in the complete lipidome or the 13C precursors may be added at the time of drug treatment to observe incorporation of label, and hence determine which lipids are synthesized in the presence of the drug. After cells have been grown under the desired conditions, the cell pellet must be washed with D2O, prior to extraction of the lipids, to ensure the removal of water and media components that may interfere with signals in the NMR spectrum. The methods below describe the standard protocol for the preparation of a mycobacterial sample (see Note 1). Generally, a 20–50 ml culture grown to an OD600 of 0.4–0.6 is sufficient to generate a good lipid profile. 1. 50 ml of 13C-labeled Middlebrook 7H9 media is inoculated from a midlog culture to an OD600 = 0.05 for slow growing species or OD600 = 0.005 for fast growing species (see Notes 1–3). 2. The cells are incubated at 37°C with continuous shaking at 250 rpm until an OD600 of 0.4–0.6 is reached. The cells are harvested by centrifugation at 3,700 × g for 10 min. 3. The supernatant is carefully discarded and the cells are washed twice with a volume D2O equal to 20% of the original culture volume (10 ml for a 50 ml culture). Centrifuge at 3,700 × g for 10 min and discard the supernatant. 4. Use 1 ml of D2O to resuspend the pellet and transfer the sample to a preweighed microcentrifuge tube. Centrifuge the cells at 3,700 × g for 10 min. 5. Discard the supernatant and carefully remove all remaining D2O from the cell pellet. Determine the weight of the wet pellet. 3.2. Extraction and Analysis of the Total Free Lipid Pool
In these experiments, the use of deuterated solvents is necessary to suppress solvent peaks and optimize the signal to noise ratio. The lipid pool of mycobacterial cells are extracted directly from the washed cell pellet by a 2:1 mixture of CDCl3:CD3OD, a solvent system commonly used to extract surface lipids of mycobacteria. Internal standards are incorporated into the extraction mixture to facilitate NMR data processing and allow normalization of data collected at different times. Ideally, compounds are selected that provide two sets of discrete internal reference peaks located in an uncongested area of the spectrum. For mycobacterial lipid profiling, dimethyl sulfoxide (DMSO) and/or 1,3,5 trimethyl benzene can be added to the CDCl3:CD3OD mixture prior to extraction. A lipid profile is generated directly from the
94
Mahrous, Lee, and Lee
crude extract using a 2D-HSQC pulse sequence to resolve the signals into two dimensions (the F1 proton dimension and the F2 carbon dimension). This inverse detection sequence helps maximize sensitivity to heteronuclei and increase signal to noise ratio. The isotope enrichment strategy greatly enhances the signal intensity and confers a good spectrum in a relatively short acquisition time. A good quality 2D spectrum is attainable in 20 min. However, observation of some less abundant lipid species benefits from a longer acquisition time of 90 min. Importantly, such maps are not attainable from normally grown (unlabeled) cells using larger quantities of crude lipid in the same volume as these lipids precipitate at high concentrations in the NMR solvents. After generation of a baseline lipid profile, to facilitate interpretation of the data, it is necessary to define biomarker signals of known lipid species using purified standards. In the lipid profile, each molecule is represented by multiple signals. Although many signals in the 2D spectrum still overlap, each lipid molecule generally has at least one discrete signal present. These discrete signals are designated as biomarker peaks. Therefore, the presence or absence and the quantity of a certain molecule can be readily determined by the identification of these discrete and diagnostic biomarker peaks within the 2D spectrum. When possible it is useful to identify and utilize multiple biomarker signals for molecules of interest to confer a greater degree of confidence in the data interpretation and hence limit the possibility of artifacts from unidentified overlapping species. Due to the complexity of the spectra and incomplete spectral assignments in the literature, it is important to purify a set of standards and analyze them under the same conditions as the lipid profile, in order to assign biomarkers effectively. For some molecules, their abundance in M. tuberculosis is not sufficient to purify enough sample. This can often be resolved by utilizing the increased expression of the targeted lipids in other mycobacterial species, from which they can then be isolated in suitable amounts. If necessary 3D HCCH TOCSY and 3D HCCH COSY can then be used to confirm assignment of biomarker signals in the total lipid profile. This is especially useful for confirming the shift of any signals that may vary slightly from standards that were purified from other mycobacterial species, which occurs due to slight variations in structural composition or substitution. After biomarkers have been assigned in the total lipid profile, spectral sets can be compared and lipids quantified by processing the data using software such as Felix or NMR Pipe. 1. Add 2 ml of CDCl3 and 1 ml of CD3OD per 1 mg of wet pellet weight. Mix the sample to homogeneity and extract at 37°C with continuous shaking at 250 rpm for 90 min.
Lipid Profiling Using Two-Dimensional Heteronuclear Single Quantum
95
2. After extraction, add an extra 1 ml of CD3OD for every milligram of original pellet weight and centrifuge at 3,700 × g for 10 min. Due to the water released from the cells during lysis, the additional CD3OD is necessary to allow the cell debris to pellet and to form a single phase solution. If the solution is still biphasic after centrifugation, CD3OD can be added dropwise with mixing until a single phase is formed after centrifugation. 3. Remove the supernatant to a clean microcentrifuge tube, being careful not to disrupt the pellet. Recentrifuge at 3,700 × g for 10 min and carefully transfer the supernatant to a fresh tube. 4. Transfer an appropriate volume of the supernatant to a clean dry NMR tube and acquire a 2D HSQC spectrum according to the NMR spectrometer used (Notes 4 and 5). 3.3. Preparation of the Mycolic Acid Pool
One important population of lipids in Mycobacteria is the mycolic acids. These unusual long chain fatty acids form a pseudomembrane layer where a mixture of free lipids, proteins, and glycolipids are embedded in a parallel inclusion. They are tethered to the cell wall by the arabinogalactan and therefore additional steps are required to isolate them from the cell pellet to complete the lipid profile in global lipid analysis studies. 13C enrichment significantly reduces the amount of sample required and they can easily be isolated from the left over cell debris after extraction of the free lipids. 1. The remaining pellet left after free lipid extraction is washed with 1 ml of D2O, which is subsequently removed by centrifugation and decantation. 2. A solution of 1 M methanolic KOH is added to the washed pellet (2 ml per 100 mg of original weight of the wet cell pellet). The methanolic solution is then transferred to a glass screw capped tubes and stirred for 16 h at 80°C. 3. The methanolysate is then cooled down and acidified by adding concentrated HCl dropwise to reach pH 4, to precipitate the free mycolic acids. The acidified methanolic solution is then diluted 3× with deionized water and extracted with one volume of diethyl ether. The top ether layer is carefully removed and transferred to another tube. 4. The extraction step is repeated and the ether extract is pooled and washed collectively with one volume of deionized water. The water is finally removed and the ether layer is dried under an atmosphere of inert gas. 5. The lipid film produced upon dryness is then dissolved in 1:1 solution of CD3OD: CDCl3 and transferred to a standard NMR tube.
96
Mahrous, Lee, and Lee
3.4. Alternative Extraction Methods
Although it is advantageous to monitor the whole cell lipidome at once, in some cases lipids of low abundance are better observed in a more limited subpopulation. The general extraction method discussed previously in Subheading 3.2 can be easily modified, through the incorporation of simple extraction or fractionation processes, to focus on specific or less abundant subpopulations. For example, methods include successive extraction of cells with deuterated solvents of increasing polarity, use of scavenger resins to pull down and fractionate anionic lipids and cationic lipids (17, 18), and the use of solvent partitions to remove very abundant signals or to separate the polar and apolar lipid pools. Separation of the polar and apolar lipids can be easily achieved through a simple modification to this method. As water is released from cells during extraction with 2:1 CDCl3:CD3OD, the relative ratio of chloroform to methanol to water in the solvent mixture results in partitioning of the sample into two phases as described in Subheading 3.2. The two phases can simply be collected and analyzed separately rather than adding more CD3OD to achieve a monophasic solution. 1. Add 2 ml of CDCl3 and 1 ml of CD3OD per 1 mg of wet pellet weight. Mix the sample to homogeneity and extract at 37°C with continuous shaking at 250 rpm for 1 h, 30 min. After extraction, centrifuge the sample at 3,700 × g for 10 min. A 2-layer suspension will be formed with the cellular debris at the interface. 2. Remove the aqueous upper layer carefully and transfer to a clean tube. This constitutes the bulk of the polar cell wall lipids. 3. Transfer the lower organic to a clean microcentrifuge tube. Add 2 ml of CDCl3 per original cell pellet weight to the extraction mixture. Vortex the tube for 30 s, then centrifuge the mixture at 3,700 × g for 10 min. 4. Transfer the chloroformic extract (lower layer) to a clean NMR tube. This constitutes the apolar lipid pool.
3.5. Results
1. Figure 1 shows the 2D HSQC lipid profile of M. bovis BCG. Signals in the HSQC spectrum can be categorized into three main regions; the most up-field region (d1H,–0.5–3.0 ppm) representing mainly the aliphatic chain of the lipid molecules, the far down-field region (d1H, 5.2–8.5 ppm) representing unsaturated and aromatic substructures and the middle region (d1H, 3.2–5.4 ppm), representing signals from sugars attached to glycolipids and phospholipids. Two sets of biomarker signals for menaquinone and trehalose dimycolates are labeled on the HSQC spectrum for illustration.
Lipid Profiling Using Two-Dimensional Heteronuclear Single Quantum
97
Fig. 1. 2D HSQC lipid profile of M. bovis BCG: 1H-13C HSQC spectrum of the 2:1 v/v CDCl3:CD3OD extract of M. bovis BCG. The extract was obtained as described in the text Subheading 3.2. For illustration, three biomarker signals for menaquinones are shown (signals Ia–c), and two biomarker signals are shown for trehalose dimycolate (signals IIa,b) including the diastereotopic CH2. Signal III indicates the signal for DMSO, which was included as an internal standard during the extraction. The total sample preparation time including extraction was 2 h. Total NMR acquisition time was 1 h and 40 min (ni = 256, nt = 10).
2. Figure 2 demonstrates the utility of this technique to probe gene function by detecting changes in the lipid pool as a result of gene knockout. In this experiment the lipid profiles of the hypervirulent M. tuberculosis HN-878 was compared to that of HN878-▵Dpks, a mutant strain where the pks 1–15 gene responsible for the production of the virulence factor phenolicglycolipid (PGL) was effectively disrupted. The three biomarker signals corresponding to M. tuberculosis PGL can be observed in the lipid profile of the clinical isolate HN-878, but are missing in the lipid profile of the mutant strain as a result of pks1–15 gene knockout. PGL is generally a difficult molecule to isolate from M. tuberculosis due to its low abundance, which makes application of this technique particularly useful for PGL studies. Indeed, this method was found to be applicable for rapidly screening clinical isolates for the presence or absence of PGL expression. 3. Figure 3 shows the 2D 1H-13C HSQC spectrum of the mycolic acid pool extracted from M. bovis BCG grown in Middlebrook 7H9 supplemented with 13C glucose and 13C glycerol under normal growth conditions. Each class of
98
Mahrous, Lee, and Lee
a
c
e
b
d
f
Fig. 2. Demonstration of lipid profiling to probe gene function. Biomarker signals for phenolicglycolipid (PGL) were defined using the 1H-13C HSQC spectrum acquired from a sample of PGL purified from M. liflandii [shown as circled cross peaks in spectra (a) and (b)]. These signals were readily detected in extracts from wild type M tuberculosis: HN878 (c) and (d) but not in the knockout strain of HN878 ▵∆ pks (e) and (f). For simplicity, relevant smaller regions are shown from the whole HSQC spectra: two biomarker aromatic signals are shown in solid circles in panels (a), (c), (e) (dH 6.7–7.4, dC 115–133) and the benzylic CH2 signal is shown in dashed circles in panels (b), (d), (f) (dH 2.3–3.4, dC 33.8–49.4). Due to differences in the glycosidation pattern between PGL molecules produced by M. tuberculosis and M. liflandii, a small difference in the chemical shift of the aromatic protons at the ortho-position to the glycosidic linkage was observed in their respective 2D-HSQC spectrum. Adapted from (11) with permission.
mycolic acids exhibit unique biomarker signals, which can be used to detect the relative abundance of a certain class in the mycolic acid pool. 4. Figure 4 demonstrates the differences in the polar and apolar lipid profiles of M. bovis BCG compared to the total lipid profile: total lipids A, apolar lipids B, polar lipids C. By acquiring the 2D HSQC spectra of the polar and apolar lipid fractions separately, some overlapping signals were effectively resolved. Noticeably, the abundant triglycerides and phosphatidyl inositol mannosides signals were eliminated from the polar lipid pool allowing the detection of many glycolipids of relatively low abundance.
Lipid Profiling Using Two-Dimensional Heteronuclear Single Quantum
99
Fig. 3. 2D-HSQC mycolic acid profiles for M. bovis BCG. 1H-13C HSQC spectrum of mycolic acids, extracted from M. bovis BCG as described in text Subheading 3.3. Examples for signals common to all mycolic acid are shown by arrow [(a), bCH–OH (b), aCH–COOH)]. Signals indicative to specific classes of mycolic acids are circled in dashed lines.
4. Notes 1. The handling of mycobacterial species should follow the biosafety recommendation for each specific species. Mycobacterium tuberculosis must be handled at Biosafety Level 3. Handling of extracts should be performed in accordance with institutional guidelines. The sterilization ability of the extraction methods should be verified before samples are removed from the BSL3 laboratory. In this case, it was found that extraction of 200 mg of cells with 600 ml of a 2:1 mixture of CDCl3:CD3OD for 90 min followed by two centrifugation steps at 3,700 × g for 10 min resulted in a sterile sample. Samples were prepared in triplicate, dried under sterile conditions, resuspended in 7H9 and the entire contents of a single preparation were plated on 7H11 media or inoculated into 7H9 broth. After 6 weeks, no growth on plates or in broth was observed. 2. Composition of 13C-7H9 Media (a) 45 ml Sterile Middlebrook 7H9 media (7H9, 0.05% Tween 80) (b) 5 ml Filter sterilized solution of BSA (5%) and NaCl (0.8%) (c) 1 ml Filter sterilized solution of 13C-glucose 0.1 gm/ml (d) 1 ml Filter sterilized solution of 13C-glycerol 0.1 gm/ml (e) the growth
100
a
Mahrous, Lee, and Lee
b
c
Fig. 4. Separation of the polar and apolar lipid profiles of M. bovis BCG. 1H-13C HSQC spectra generated from M. bovis BCG grown in 13C-labeled 7H9 (a) total lipid, (b) apolar lipids, and (c) polar lipids. Expanded views showing detail of the biomarkers for mycobactins, mycoside B, (dashed box, dH 6.68–7.28, dC 113–132.3) and glycerol-containing phospholipids (solid box, dH 5.0–5.44, dC 67–78.8) have been shown in these spectra as examples of how lipids can be separated through simple fractionation. Only mycoside B is present in the apolar lipid spectrum while mycobactin signals are observed in the polar spectrum only. The glycerol-containing phospholipids are overlapping in the total lipid pool, however, when the abundant triglyceride and PIMs are removed into the apolar solvent, less-abundant species were better presented in the polar lipid spectrum.
rate of bacteria is not affected by the use of a labeled carbon source. 3. Examples of fast growing species include the saprophytic mycobacterial species M. smegmatis and M. phlei. Examples of slow growing species include M. tuberculosis, M. bovis, and M avium.
Lipid Profiling Using Two-Dimensional Heteronuclear Single Quantum
101
4. For the NMR data presented in this chapter, a Varian Inova 500 MHz NMR spectrometer was used. The spectrometer was equipped with a triple resonance trpfg 5 mm probe, which was tuned, matched, and shimmed for each individual sample. The 1H-13C HSQC spectra were acquired with a standard pulse sequence with 10 or 12 transients and 256 increments. 5. Extraction of lipids from a 20–50 ml culture usually generates approximately 0.5–0.7 ml of extract, which is appropriate for use with 5 mm sample probe and standard 5 mm NMR tubes. However, microsampling inserts and other low volume NMR probes require far less sample volume and may be used if analysis of smaller samples is required.
Acknowledgments We thank Dr Clifton Barry III, (National Institutes of Health) for providing M. tuberculosis HN-878 and HN878-▵D pks strains, Dr. Pamela L.C. Small, (University of Tennessee, Knoxville) for providing both M. marinum and M. liflindii strains, and Dr. Wei Li (University of Tennessee, Memphis) for technical assistance. We acknowledge financial support for this work from National Institutes of Health grant AI076938. References 1. Goren, M. B., O. Brokl, and W. B. Schaefer. 1974. Lipids of putative relevance to virulence in Mycobacterium tuberculosis: phthiocerol dimycocerosate and the attenuation indicator lipid. Infection and immunity 9:150–158. 2. Goren, M. B., O. Brokl, and W. B. Schaefer. 1974. Lipids of putative relevance to virulence in Mycobacterium tuberculosis: correlation of virulence with elaboration of sulfatides and strongly acidic lipids. Infection and immunity 9:142–149. 3. Kato, M., M. Kusunose, K. Miki, K. Matsunaga, and Y. Yamamura. 1959. The mechanism of the toxicity of cord factor. The American review of respiratory disease 80:240–248. 4. Etemadi, A. H. 1967. [Mycolic acids. Structure, biogenesis and phylogenetic value]. Exposes annuels de biochimie medicale 28:77–109. 5. Etemadi, A. H. 1967. [Structural and biogenetic correlations of mycolic acids in relation to the phylogenesis of various genera
of Actinomycetales]. Bulletin de la Societe de chimie biologique 49:695–706. 6. Lechevalier, M. P., A. C. Horan, and H. Lechevalier. 1971. Lipid composition in the classification of nocardiae and mycobacteria. Journal of bacteriology 105:313–318. 7. Minnikin, D. E., L. Alshamaony, and M. Goodfellow. 1975. Differentiation of Mycobacterium, Nocardia, and related taxa by thin-layer chromatographic analysis of wholeorganism methanolysates. Journal of general microbiology 88:200–204. 8. Dobson, G., D. E. Minnikin, S. M. Minnikin, M. Parlett, M. Goodfellow, M. Ridell, and M. Magnusson. 1985. Systematic analysis of complex mycobacterial lipids. In Chemical methods in bacterial systematics. M. Goodfellow, D. E. Minnikin(eds). Academic, London. 237–2654. 9. Jain, M., C. J. Petzold, M. W. Schelle, M. D. Leavell, J. D. Mougous, C. R. Bertozzi, J. A. Leary, and J. S. Cox. 2007. Lipidomics reveals control of Mycobacterium tuberculosis
102
Mahrous, Lee, and Lee
virulence lipids via metabolic coupling. Proceedings of the National Academy of Sciences of the United States of America 104:5133–5138. 10. Shui, G., A. K. Bendt, K. Pethe, T. Dick, and M. R. Wenk. 2007. Sensitive profiling of chemically diverse bioactive lipids. Journal of lipid research 48:1976–1984. 11. Mahrous, E. A., R. B. Lee, and R. E. Lee. 2008. A rapid approach to lipid profiling of mycobacteria using 2D HSQC NMR maps. Journal of lipid research 49:455–463. 12. 2008. WHO report 2008 Global tuberculosis control surveillance, planning, financing. WHO. 17–20. 13. Brennan, P. J. and H. Nikaido. 1995. The envelope of mycobacteria. Annual review of biochemistry 64:29–63. 14. Puzo, G. 1990. The carbohydrate- and lipidcontaining cell wall of mycobacteria, phenolic glycolipids: structure and immunological properties. Critical reviews in microbiology 17:305–327. 15. Karakousis, P. C., W. R. Bishai, and S. E. Dorman. 2004. Mycobacterium tuberculosis cell envelope
lipids and the host immune response. Cellular microbiology 6:105–116. 16. Ortalo-Magne, A., A. Lemassu, M. A. Laneelle, F. Bardou, G. Silve, P. Gounon, G. Marchal, and M. Daffe. 1996. Identification of the surface-exposed lipids on the cell envelopes of Mycobacterium tuberculosis and other mycobacterial species. Journal of bacteriology 178:456–461. 17. Villeneuve, C., G. Etienne, V. Abadie, H. Montrozier, C. Bordier, F. Laval, M. Daffe, I. Maridonneau-Parini, and C. Astarie-Dequeker. 2003. Surface-exposed glycopeptidolipids of Mycobacterium smegmatis specifically inhibit the phagocytosis of mycobacteria by human macrophages. Identification of a novel family of glycopeptidolipids. The Journal of biological chemistry 278:51291–51300. 18. Villeneuve, C., M. Gilleron, I. MaridonneauParini, M. Daffe, C. Astarie-Dequeker, and G. Etienne. 2005. Mycobacteria use their surface-exposed glycolipids to infect human macrophages through a receptor-dependent process. Journal of lipid research 46:475–483.
Chapter 6 Capabilities and Drawbacks of Phospholipid Analysis by MALDI-TOF Mass Spectrometry
1
2 3
Beate Fuchs, Ariane Nimptsch, Rosmarie Süß, and Jürgen Schiller
4
Summary
5
The important roles of lipids particularly certain phospholipids in signal transduction processes and as important disease markers are becoming increasingly evident. Unfortunately, however, sensitive methods of lipid analysis are established to a much lesser extent than, e.g., methods of protein analysis. Mass spectrometry (MS) is an increasingly used technique of lipid analysis and electrospray ionization (ESI) MS is the so far most established ionization method. Although matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF MS) was so far primarily used for protein analysis, however, this method has itself proven to be very useful in the field of lipid analysis, too. This chapter gives an overview of methodological aspects of MALDI-TOF MS in lipid research and summarizes the specific advantages and drawbacks of this soft-ionization method. In particular, suppression effects of some lipid classes, especially those with quaternary ammonia groups such as phosphatidylcholine, will be highlighted and possible ways to overcome this problem (use of different matrices, separation of the relevant lipid mixture prior to analysis) will be discussed on the example of an organic liver extract.
6 7 8 9 10 11 12 13 14 15 16 17 18
Key words: Glycerophospholipids, Lipid analysis, MALDI-TOF MS, Lipid extracts, Liver, Matrix, Thin-layer chromatography, Mass spectrometry
19 20
1. Introduction
21
Over decades lipids were primarily considered as the cellular “packing material” of more important contents, for instance, enzymes, nucleic acids, etc., and as energy-rich “fuel” in nutrition (1). Nowadays, however, lipids such as diacylglycerols and particularly phospholipids (PLs) such as phosphatidic acids or phosphoinositides are known to represent important secondmessenger molecules involved in cellular communication (2). Donald Armstrong (ed.), Lipidomics, Methods in Molecular Biology, vol. 579, doi 10.1007/978-1-60761-322-0_6, © Humana Press, a part of Springer Science + Business Media, LLC 2009
103
22 23 24 25 26 27 28
104 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 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76
Fuchs et al.
Additionally, lipids such as lysophosphatidylcholines (LPCs) were also recognized as important disease markers, for instance, in atherosclerosis or rheumatoid arthritis (3, 4). Accordingly, in addition to terms as “proteomics” or “genomics”, the term “lipidomics” (5, 6) was also recently introduced. In order to make a long story short: The interest in lipids and their analysis has significantly increased during the last decade. Surprisingly, however, in comparison to the analysis of proteins, there are currently only much less developed protocols of lipid analysis. One potential reason is the considerable diversity of lipids as “lipid” – in the most general definition – relates to all compounds that may be isolated from body fluids or biological tissues by extraction with organic solvents due to their apolar character (1). This diversity of phospholipids is not only coming from differences in the headgroup (e.g., choline or ethanolamine) but also from the linkage type between the apolar alkyl chains and the glycerol (diacyl, alkyl-acyl-, or alkenyl-acyl) and, finally, the large variability of potential fatty acyl residues (7). Therefore, complex lipid patterns with thousands of different lipid species can be expected if crude organic extracts from biological samples are analyzed. The best analytical approach to assess this diversity is still an open issue (8): Although chromatographic techniques are highly established, they have the disadvantage – equally if liquid chromatography, in particular, high-performance liquid chromatography (HPLC) (9) or thin-layer chromatography (TLC) (10) is used – that several runs are normally required in order to obtain complete compositional information: Normal phase chromatography is needed for the separation of lipids according to their headgroups, whereas reversed phase chromatography is usually used for the differentiation of lipids according to their fatty acyl compositions (8). Another problem is the detection of the lipids within the obtained fractions. If, for instance, a UV detector is used, lipids with unsaturated fatty acyl residues are primarily detected, whereas this kind of detector is not suitable for the detection of lipids with saturated fatty acyl residues as they lack the UV absorption of olefinic residues (1). Mass spectrometry (MS) is increasingly used for the analysis of complex lipid mixtures (11), in particular, for lipidomic studies (5). Although methods such as atmospheric pressure chemical ionization (APCI) (12) or electrospray ionization (ESI) (13) are nowadays considered to be the methods of choice for lipids, there is growing evidence that matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF MS) is also capable of providing important information about the composition of an unknown mixture. The particular advantages of MALDI MS are the simple performance and the high sensitivity (14) down to a few attomoles of the analyte. The use
1.1. Advantages and Disadvantages of MALDI-TOF MS in the Lipid Field
Capabilities and Drawbacks of Phospholipid Analysis
105
of MALDI-TOF MS for lipid and phospholipid analysis has been recently reviewed (15, 16) and compared with other methods of lipid analysis (17).
77
Although a comprehensive treatise of the fundamentals of MALDI-TOF MS is clearly outside the scope of this chapter (for a detailed description see (18)), some important aspects representing the most important differences between MALDI MS and other MS methods have to be shortly discussed: 1. A solid sample is used in MALDI-TOF MS. The lipid sample is mixed directly on the sample plate (the “target”) or prior to its deposition with the matrix solution and is subsequently allowed to crystallize (15). The matrix is required to absorb the energy emitted from the laser and enables the ionization of the analyte, even if this analyte does not exhibit a major absorption at the laser wavelength (often 337 nm). Although there are UV- and IR lasers in use (18), the majority of commercially available MALDI devices are equipped with UV lasers and, therefore, IR lasers will not be considered here to a larger extent. However, completely different matrix compounds in comparison to UV lasers are required if IR lasers are used. One common IR MALDI matrix is, for instance, glycerol that provides intense IR absorption (19). Equally if IR or UV MALDI is used: The actual analyte is not the sample as such but cocrystals between the matrix and the sample of interest. As these cocrystals are never completely homogenous, the reproducibility of MALDI measurements is often doubted (20). However, as many laser shots, each leading to an individual mass spectrum, are averaged for the final mass spectrum, these inhomogeneities may be significantly minimized. Additionally, the matrix, that normally comprises an aromatic, rather hydrophobic residue, as well as the lipid are both readily soluble in organic solvents and there is no need to use water as additional solvent (as, for instance, in the case of water-soluble analytes). Therefore, the homogeneity of matrix/lipid crystals is higher than that of matrix/protein crystals (21). Finally, the use of solid samples enables to record spatially resolved mass spectra. This “mass spectrometric imaging” is an important topic of modern research and attracts currently significant interest in medical diagnosis, for instance, in order to differentiate different types of cancer (22).
80
2. Soft ionization MS techniques as MALDI provide – in contrast to traditional electron impact (EI) MS – no radical cations by the abstraction of an electron from the analyte of interest (cf. Fig. 1), but “quasimolecular” ions. Thereby the analyte is cationized by the addition of a proton or an alkali metal ion (positive ions) or the loss of a proton (negative ions) (23). Protons are normally provided by the matrix that is often an
117
78 79
81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 118 119 120 121 122 123
106
Fuchs et al.
Fig. 1. Schema showing the events of ion formation upon electron impact ionization and quasimolecular ion generation that is most relevant to the MALDI ionization process. Please note that radical cations are generated upon EI ionization. 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148
organic acid, whereas alkali metal ions are present as such in virtually all biological samples. The tendency of the analyte to add or loose a proton depends, however, on the basicity of the analyte in comparison to the acidity of the used matrix or vice versa. Although 2,5-dihydroxybenzoic acid (DHB) is the most common matrix for the analysis of lipids (24), it should be noted that for special problems, other matrices as paranitroaniline (PNA) can be advantageously used (25). This will be illustrated below in more detail. A survey of the characteristics of electron impact ionization and positive as well as negative quasimolecular ion formation is given in Fig. 1. In the light of these aspects it is evident that providing information on the detection limits of MALDI-TOF MS makes sense only if the applied matrix is simultaneously mentioned (26). 3. Although MALDI is relatively robust against impurities of the sample as salts or buffer components (23), great care is needed if detergents such as Triton or Brij, for instance, are used for the extraction of lipids from the biological material as these molecules are also rather apolar and cannot be easily separated from the lipids of interest. The use of commercially available “detergent removal” kits normally fails in the case of lipids because these kits were primarily developed for the purification of proteins that are normally more polar in comparison to the applied detergent. Due to that problem, the use of organic solvents or mixtures is recommended for lipid
Capabilities and Drawbacks of Phospholipid Analysis
107
extraction. Additionally, the extraction with organic solvents removes simultaneously salts and other low molecular compounds of high polarity that might affect the ion yield.
149
4. MALDI-TOF MS offers an extremely high sensitivity and detection limits in the atto- (10−18) and even low zeptomole (10−21) range were already reported in the case of lipids – in particular for lyso-phosphatidylcholine (14). Although these detection limits seem only realistic for special samples in combination with highly sophisticated matrices, sensitivities in the femtomole (10−15) range can be easily achieved by standard MALDI-TOF MS. This means that 1 ml of a 1 mg/ml lipid sample is absolutely sufficient to detect the lipid of interest (26). This high sensitivity clearly requires the use of extremely pure chemicals: It is much simpler to detect very small amounts of the lipid of interest than – vice versa – NOT to detect impurities in the solvents, the matrix compounds, etc.
152 153
2. Materials 2.1. Equipment and Supplies (See Also Note 1)
2.2. Reagents
150 151
154 155 156 157 158 159 160 161 162 163 164 165
166
1. Bruker “Autoflex” MALDI-TOF mass spectrometer (Bremen, Germany) equipped with reflectron, “delayed extraction facility” and N2 laser emitting at 337 nm. The capability to record positive and negative ion spectra should be available.
167 168
2. MALDI targets made from stainless steel or from aluminum with gold-coated surface (see Note 2).
171
3. Micropipetts (see Note 3).
173
4. Glass (Hamilton) syringes of different sizes.
174
5. Small glass vessels for single use for mixing matrix and sample or for diluting stock solutions of lipids (available, for instance, from Knauer, Berlin, Germany).
175 176
6. TLC silica gel 60 plates (Merck, Darmstadt, Germany) and TLC trough (CAMAG, Muttenz, Switzerland).
178 179
7. Heat gun or common hair dryer (see Note 4).
180
8. Standard laboratory centrifuge.
181
1. Some selected phospholipid samples of known composition and with known concentration should be used as the first samples. As palmitoyl and oleoyl residues are highly abundant in samples of biological origin (16), it is recommended to use stock solutions of 1-palmitoyl-2-oleoyl-sn-phosphatidylcholine, -ethanolamine, and -glycerol. These are available from Avanti Polar Lipids, Alabaster, AL, USA. These compounds
182
169 170 172
177
183 184 185 186 187 188
108
Fuchs et al.
are normally abbreviated by POPC, POPE, and POPG. A stock concentration of 1 mg/ml lipid in chloroform is recommended for best performance and to optimize instrumental parameters.
189 190 191 192 194
2. 2,5-Dihydroxybenzoic acid (DHB) and para-nitroaniline (PNA) of highest available purity (see Note 5).
195 196
3. Chloroform, methanol, ethanol, triethylamine, and distilled water of highest commercially available quality (see Note 6).
197
4. A small piece of fresh bovine liver that is best obtained from a local butcher or slaughterhouse.
193
198
199
200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230
3. Methods 3.1. Sample Processing 3.1.1. Artificial Lipid Samples of Known Composition
1. If sufficient amounts of samples are available, the use of stock solutions in the mg/ml concentration range is recommended. Therefore, dilute the available lipid samples to about 1 mg/ml with CHCl3. Mix one equivalent of the lipid standard solution with one equivalent of DHB matrix solution (0.5 mol/l corresponding to about 77 mg/ml in CH3OH). Trifluoroacetic acid (TFA) is often suggested as an additive because it enhances the yield of H+ adducts due to its considerable acidity (23). However, TFA may also have deleterious effects because lipids containing alkenyl ether groups (“plasmalogens”) are readily hydrolyzed in the presence of TFA (27, 28). Therefore, the use of TFA is discouraged – particularly if unknown lipid mixtures are to be analyzed. Prepare the same samples in the presence of PNA (0.17 mol/l in chloroform/methanol (2:1, v/v)). 2. Apply the prepared samples to the MALDI target. Do not be surprised if a very large spot is formed and the sample “spreads” out over the sample plate: In contrast to water, organic solvents possess only a rather small surface tension (H2O: ~73 × 10−3 N/m; CH3OH ~23 × 10−3 N/m). This prevents the formation of a small spot on the metal surface of the sample plate. Do not be worried! Simply leave some space between the individual samples in order to avoid mixing of the different samples during deposition onto the MALDI target. Avoid touching the MALDI target with the pipette tip or the needle of the Hamilton syringe in order not to affect homogeneous crystallization. The MALDI target should also not be touched with the fingers. 3. Evaporate the solvent quickly by drying the sample plate with a hair dryer and load the prepared sample plate directly into the mass spectrometer. Avoid long-term exposition of the
3.1.2. Preparation of the Liver Extract
3.1.3. Separation of the Crude Lipid Extract by Thin-Layer Chromatography
Capabilities and Drawbacks of Phospholipid Analysis
109
TLC plate to air: Due to the large surface of the lipid film on the MALDI target, lipids may be easily oxidized by air. This is particularly a problem if diluted samples are analyzed. It has also been shown that the content of residual water in the organic solvents influences the quality of the mass spectra (29). This may be surprising as completely “dry” samples are investigated by MALDI-TOF MS under high vacuum conditions. However, there is increasing evidence that the solvent or solvent mixture determines the homogeneity of the crystallization between the matrix and the analyte. Therefore, the quality of mass spectra depends dramatically on the used solvent system.
231
1. Cut the liver tissue into small pieces with a scalpel. Add the 20-fold amount (by weight) of Bligh & Dyer solvent (30) mixture (CHCl3:CH3OH:H2O = 1:1:0.9) and stir or vortex the resulting mix vigorously for a few minutes. Subsequently centrifuge the sample (10 min, ca. 1000 g, 20°C) in order to improve the separation of the organic (bottom) and the H2O/CH3OH phase (see also Note 7).
243
2. Remove the organic (lower) phase by a Hamilton syringe and transfer it to another unused vial. Do not try to get the complete organic phase in order to avoid the introduction of impurities from the aqueous layer. This organic phase is used for all further experiments and the lipid concentration is of the order of 6–8 mg/ml. The lipid yield can be easily determined by evaporation of the solvent under reduced pressure and subsequent weighing of the residual material. If needed, a bovine liver extract is also commercially available from AVANTI Polar Lipids, Alabaster, AL, USA. Another established method to determine the PL concentration is the classical phosphate determination according to Bartlett (31).
250
3. The obtained lipid extract may be directly used for MALDITOF MS by simply diluting the organic solution with the prepared matrix solution (1:10, v/v).
262 263
1. For the separation of phospholipids according to differences of the headgroup normal phase chromatography is used, whereas reversed phase chromatography is used for the separation according to differences in the fatty acyl compositions (8). Use HPTLC silica gel 60 plates (10 cm × 10 cm in size with aluminum or glass backs) (Merck, Darmstadt, Germany). Prerun the plates with the needed solvent mixture in order to remove impurities. Dry the plate carefully.
265 266
2. Apply 0.5–1 ml (corresponding to a total lipid amount of 4–8 mg) of the liver extract by using a standard Hamilton syringe as small spots with 1 cm space between the individual spots and
273
232 233 234 235 236 237 238 239 240 241 242
244 245 246 247 248 249 251 252 253 254 255 256 257 258 259 260 261
264
267 268 269 270 271 272 274 275
110
Fuchs et al.
at a distance of 1.5 cm from the bottom edge of the TLC plate. Also apply a lipid mix of known composition (about 1–2 mg per PL). Dry the plate carefully prior to development.
276 277 278
3. Develop the plates in a commercially available vertical ascending TLC chamber (CAMAG, Switzerland) with chamber saturation using chloroform, ethanol, water, triethylamine (30:35:7:35, v/v/v/v) as solvent mixture (32). This should result in good separation quality of the individual PL classes. The required time of development is about 35 min: The total length of the run is about 6 cm under these conditions and is a good compromise between resolution and the required time. The separation is performed at room temperature (22 ± 2°C) and 50 ± 5% relative humidity.
279 280 281 282 283 284 285 286 287 288
4. Subsequent to development, TLC plates are dried under a stream of warm air and the lipids are visualized by spraying with a solution of primuline (Direct Yellow 59) in acetone according to (33). Upon irradiation by UV light individual PLs become detectable as violet spots and can be easily identified by comparison with the reference mixture. The binding between the dye and the lipid is noncovalent and does not affect the quality of the mass spectra. The monoisotopic mass of primuline is 475.0 and, therefore, a very small signal is sometimes detected at m/z = 498 (Na+ adduct) in the positive ion mass spectra.
289 290 291 292 293 294 295 296 297 298 299
5. Mark the spots of interest (under UV light) by circling with a soft pencil.
300 301
6. The spots of interest are carefully scratched off from the TLC plate into different small glass vessels and the PLs are eluted from the silica gel by the addition of a mixture of 75 ml CHC13, 75 ml methanol, and 75 ml 0.9% NaCl in water and intense vortexing (see Note 8). Afterward, samples are centrifuged (ca. 1000 g) as described above to enhance phase separation. The organic layers are evaporated to dryness and the residual material redissolved in 20 ml matrix solution (DHB or PNA) and directly used for MALDI-TOF MS.
302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320
3.2. Recording MALDI-TOF Mass Spectra
1. Due to the large number of MALDI devices that are available from different suppliers, it is impossible to describe the necessary experimental parameters in detail. Therefore, please consult the manual of your device for a suitable data file to start with. You should start with “delayed extraction conditions” (DE) and use the reflectron of the device. This results in higher resolution and mass accuracy than the common linear mode that provides, however, higher sensitivity. 2. It is recommended to start in all cases with the analysis of a known sample of known concentration in order to check if the
3.3. Quantitative Data Analysis
Capabilities and Drawbacks of Phospholipid Analysis
111
device is properly working and all parameters are adequately set. This known sample may also be used to check the mass accuracy, i.e., the quality of the applied – often default – calibration, as well as the resolution achievable by the used instrumental settings. Please note, however, that the applied laser intensity has the most pronounced effect on spectral quality.
321
3. Always take care that the MALDI target is carefully dried before it is inserted into the mass spectrometer in order to avoid a significant decrease of the vacuum – and the delay for waiting until high vacuum is re-established.
327
4. It is an incorrect assumption that spectral quality, in particular the signal-to-noise ratio, may be enhanced by increasing the laser intensity. Although absolute signal intensities may be enhanced at elevated laser intensities, the quality of the baseline simultaneously gets poor and the achievable resolution is simultaneously diminished. Therefore, always set the laser intensity as high as needed but as low as possible.
331 332
5. Try to move the laser randomly over the sample plate in order to average nonhomogeneous spots resulting from the sample preparation. It is, however, not possible to improve the signal-to-noise ratio by averaging a larger number of individual laser shots (19) because the level of unspecific chemical background noise forms the limiting criterion.
338 339
6. Try to optimize the required parameters always with a wellknown sample and use these parameters afterward for the unknown samples, i.e., the organic liver extract in our case.
344
7. DHB is completely stable under high vacuum conditions, while PNA tends to sublime at longer times in vacuum (25). Therefore, it is advisable to perform measurements with the PNA matrix relatively quickly. If no signal is detectable, try to add somewhat additional PNA matrix.
347
1. Quantitative evaluation of MALDI mass spectra is still a challenging task. However this does not only hold for MALDI MS but also for other soft-ionization techniques: The ion yield in EI mass spectra primarily depends on the ionization potential of the functional groups present in the analyte of interest. The ionization energy is rather constant for e.g., carbonyl groups – independent of their chemical environment. Therefore, the concentration of different compounds containing carbonyl groups can be easily determined by using the signal intensities. The comparison of different lipid classes with different acidities is particularly tricky if lipid mixtures are analyzed as the ion yield may be very different for the individual lipid classes (26). The very best way to obtain quantitative compositional information is normally the separation of the individual lipid
352
322 323 324 325 326 328 329 330
333 334 335 336 337
340 341 342 343 345 346 348 349 350 351
353 354 355 356 357 358 359 360 361 362 363 364 365
112
Fuchs et al.
classes and to add an internal standard to each fraction. As the structure of this internal standard should be as similar as possible to the analyte of interest, deuterated lipids are the reference compounds of choice (34). Unfortunately, however, this approach requires some prior knowledge about the composition of the sample to warrant that the standard is added in a suitable amount and particularly to avoid the addition of an excess of the standard that might result in severe signal suppression. Due to these problems, other simpler methods of quantitative analysis were also suggested.
366 367 368 369 370 371 372 373 374 375
2. Absolute peak intensities can hardly be used as quantitative concentration measures in MALDI-TOF mass spectra because they are affected by many different parameters, for instance, the applied laser intensity, the matrix to analyte ratio, the presence of impurities, etc.
376 377 378 379 380
3. Comparison of the intensity of the peak of interest to a defined matrix peak. Although common matrices have normally much lower masses than lipids, nearly all matrices, at least the ionic ones, tend to undergo photochemical reactions upon laser irradiation in the gas phase leading to peaks at higher m/z ratios (24). Although some applications of this method to apolar lipids as triacylglycerols (16) were described, the applicability of this method seems quite limited because the matrix peak pattern is influenced by many parameters.
381 382 383 384 385 386 387 388 389
4. Using the signal-to-noise (S/N) ratio. The achievable S/N ratio increases with the lipid concentration over a large concentration range. Only by using very large amounts of lipids, i.e., an inappropriate ratio between lipid and matrix, the S/N ratio decreases. Although this method seems to be applicable for all substances, it was so far only used for rather polar species, in particular lysophospholipids (35) and phosphoinositides (36). The majority of MALDI devices calculates the S/N ratio directly from the obtained spectra. Therefore, this is an easily accessible parameter.
390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411
3.4. Results
1. A coarse overview about the shape of the positive (left) and negative (right) ion MALDI-TOF mass spectra of different phospholipids (as well as their characteristic headgroups) in the presence of different matrix compounds is shown in Fig. 2. 1-Palmitoyl-2-oleoyl-sn-phosphatidylcholine (POPC, (2a,2e)), -phosphatidylethanolamine (POPE, (2b, 2f)), and -phosphatidylglycerol (POPG, (2c, 2g)) were chosen as defined PLs because they are rather abundant in biological materials but differ in their charge states (1): POPC and POPE are zwitterionic PLs, whereas POPG is a negatively charged compound at physiological pH. In (2d, 2h) the spectra of a 1:1:1 mixture of these three PLs are shown in order to illustrate the problems of mixture analysis by
Capabilities and Drawbacks of Phospholipid Analysis
113
Fig. 2. Positive (left) and negative ion (right) MALDI-TOF mass spectra of 1-palmitoyl-2-oleoyl-sn-phosphatidylcholine (a), (e), 1-palmitoyl-2-oleoyl-sn-phosphatidylethanolamine (b), (f), 1-palmitoyl-2-oleoyl-sn-phosphatidylglycerol (c), (g) and a 1:1:1 mixture of these three compounds (d), (h). Positive ion spectra were recorded with a 0.5 M solution of DHB in methanol, whereas a 0.17 M solution of PNA (in CHCl3 and CH3OH) was used as matrix for the negative ion spectra. Lipid sample solutions (1 mg/ml) were diluted 1:1 (v/v) with the corresponding matrix and afterwards spotted onto the MALDI target. All peaks are marked according to their m/z ratios and the structures of both matrices are shown in the figure. Please note that the PC can not be detected as negative ion and the presence of a characteristic fragment ion in the POPE and POPG spectra (b, c).
MALDI-TOF MS. It is evident that the spectra differ significantly: POPC gives – in the presence of DHB – two signals at m/z = 760.6 and 782.6 in the positive ion mode according to the generation of the H+ and the Na+ adduct (1a) (21). Due to the permanent positive charge of the quaternary ammonia group (37), PC cannot be detected as negative ion under the applied conditions (1e).
Please note that the PC does not give any fragmentation products under these conditions – even the loss of the choline moiety may be prevented by careful control of the laser intensity (38). In contrast, the PE (1b) as well as the PG (1c) gives significant yields of a fragment ion that corresponds to the loss of the polar headgroup. As the fatty acyl composition is the same, both compounds give the same fragment at m/z = 577.5 (38). The structure of this fragment is also shown in (2c). Additionally, it is evident that the POPE with the monoisotopic mass of 717.5 g/mol gives only rather small yields of the H+ adduct (m/z = 718.5) but significant amounts of the Na+ adduct (m/z = 740.5) as well
412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429
114 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477
Fuchs et al.
as the Na+ adduct subsequent to the exchange of one H+ by one Na+ (m/z = 762.5). This is caused by the exchangeable protons of the –NH3+ group. The reason why POPE and POPG give more pronounced fragmentation than POPC is not yet known. However, it might be possible that the stability of the H+ adduct of POPE is not high enough to allow these ions to reach the detector. A similar observation was already made in the case of triacylgylcerols (39), where no H+ adducts were detectable at all. Using PNA as an alkaline matrix, POPE is also detectable as negative ion at m/z = 716.5 (2f). Please note that the detectability of POPE as negative ion (data not shown) in the presence of DHB would be extremely weak due to the acidic properties of this matrix. POPG can be detected as negative ion in the presence of DHB (data not shown) as well as PNA (2g, m/z = 747.5) as negative ion because of its enhanced acidity in comparison to POPE. Regarding mixture analysis significant differences are obtained between the positive and the negative ion spectra: In the positive ion mode (2d), the spectrum of the mixture is clearly dominated by the POPC. The POPE is only detectable with low intensity although it is present in the same amount as the POPC. POPG is not detectable at all under these conditions. Therefore, the presence of PCs prevents the detection of other PL species in mixtures (37). Due to this problem, a simple method to remove PC from lipid mixtures has been recently suggested (40). In contrast, the POPC is completely absent in the negative ion spectrum (2h) and only POPE and POPG are detectable in the mixture. Although this is indeed a very simple example, it is evident that mixture analysis by MALDI-TOF MS must be regarded with great caution. 2. The problem of signal suppression is still more evident if the organic liver extract is investigated (Fig. 3). The liver extract was chosen as an educational example because it can be easily prepared with good reproducibility and is even commercially available. The left hand spectra represent the positive ion spectra, while the spectra at the right hand represent the negative ion spectra. Spectra (3a) and (3c) were recorded with DHB as matrix, while PNA was used for the spectra shown in (3b) and (3d). It is obvious that the positive ion spectra are, beside minor intensity differences of the individual adducts, virtually identical equally if DHB or PNA is used as matrix (3a, 3b). Both spectra are dominated by different PC species and a more detailed assignment of the individual peaks is given in Table 1. The negative spectra, however, differ significantly and it is evident at the first glance that the spectrum recorded with the DHB matrix (3c) exhibits rather poor quality. Although peaks of a few PI species (the signal at m/z = 885.5, for instance, corresponds to PI 18:0/20:4) can be easily identified, there are no further lipids detectable.
Fig. 3. Positive (left) and negative ion (right) MALDI-TOF mass spectra of an organic bovine liver extract. Traces (a) and (c) were recorded with DHB, whereas (b) and (d) were recorded in the presence of PNA. Due to the large variety of individual lipid classes in the extract, a relatively concentrated lipid solution was used (about 6 mg/ml) and diluted 1:10 with the matrix. All peaks are marked according to their m/z ratios. Note the poor quality of the negative ion spectrum recorded with DHB as matrix.
Capabilities and Drawbacks of Phospholipid Analysis 115
116
Fuchs et al.
Table 1 Overview of the positive ions detected in the MALDI-TOF mass spectra of the liver extract and the corresponding assignments Peak position (m/z)
Assignment of molecular mass
577.5
Fragment of PE 16:0/18:1/PG 16:0/18:1
603.5
Fragment of PE 16:0/18:1
702.5
Oxidation product of PS 18:0/18:1 (+H+)
703.6
SM 16:0 (+H+)
715.0
4 DHB - 2 H+ + Na+ + 2 K+
721.0
4 DHB - 3 H+ + 3 Na+ + K+
721.5
Oxidation product of PI 18:0/20:4 (+H+)
725.6
SM 16:0 (+Na+)
727.0
4 DHB - 4 H+ + 5 Na+
743.5
Oxidation product of PI 18:0/20:4 (+Na+)
753.5
Oxidation product of PI 18:0/18:1
760.6
PC 16:0/18:1 (+H+)
766.5
PE 18:0/18:2 (+Na+)
768.5
PE 18:0/18:1 (+Na+)
771.5
PG 16:0/18:1 (+H+)
777.5
Oxidation product of PI 18:0/18:1 + H+
782.6
PC 16:0/18:1 (+Na+)
786.6
PC 18:0/18:2 (+H+)
788.5
PE 18:0/18:2 (−H+ + 2 Na+)
788.6
PC 18:0/18:1 (+H+) or PE 18:1/20:4 (+Na+)
790.5
PE 18:0/20:4 (+Na+)
793.5
PG 16:0/18:1 (+Na+)
808.6
PC 18:0/18:2 (+Na+)
810.6
PC 18:0/18:1 (+Na+) and PC 18:0/20:4 (+H+) or PS 18:0/18:2 (+H+)
812.5
PE 18:0/20:4 (−H+ + 2Na+)
812.6
PS 18:0/18:1 (+H+) and PC 18:0/20:3 (+H+)
832.6
PC 18:0/20:4 (+Na+)
834.6
PC 16:0/22:6 (+H+) or PS 18:0/18:1 (+ Na+)
(continued)
Capabilities and Drawbacks of Phospholipid Analysis
117
Table 1 (continued) Peak position (m/z)
Assignment of molecular mass
856.6
PC 16:0/22:6 (+Na+) or PS 18:0/18:1 (−H+ + 2 Na+)
885.5
PI 18:0/18:2 (+ H+)
887.5
PI 18:0/18:1 (+ H+)
909.5
PI 18:0/20:4 (+ H+) and PI 18:0/18:1 (+ Na+)
911.5
PI 18:0/20:3 (+ H+)
931.5
PI 18:0/20:4 (+ Na+)
933.5
PI 18:0/20:3 (+ Na+)
PC phosphatidylcholine, PI phosphatidylinositol, PS phosphatidylserine, PE phosphatidylethanolamine Please note that all discussed PL– besides PC – contain functional groups showing exchange with the solvents and/or ions of the matrix solution leading to complex peak patterns. Please note that in the case of acidic PL (PS and PI) the corresponding sodium salt is considered as the neutral molecule.
The peaks between about m/z = 720 and 780 are stemming from the applied DHB matrix (24). The spectrum recorded in the presence of PNA (3d) is characterized by a much higher quality and provides the additional advantage that PE species are detectable simultaneously with the PI species. The additional signals at m/z = 742.5 and 766.5 correspond to PE 18:0/18:2 and PE 18:0/20:4. Please note that in the case of mixtures, negative ions have the considerable advantage that their assignment is less problematic as there is no superposition between the different proton and sodium adducts and differences of the fatty acyl composition. Each PE species results in a single peak. Of course, the differentiation of isobaric ions cannot be made by the provided simple mass spectra (MS1). For instance, the signal at m/z = 742.5 might not only represent PE 18:0/18:2 but also PE 18:1/18:1 that have the same masses. However, this differentiation can be easily made by recording the corresponding “postsource decay” (PSD) mass spectra, i.e., by observing the characteristic fragments generated during the travel of the ions from the source to the detector. Under these conditions the released fatty acyl residues can be observed as negative ions allowing the unequivocal differentiation of the above-mentioned compounds: The anion of stearic acid would be detected at m/z = 283, while oleic acid and linoleic acid are detected at m/z = 281 and 279, respectively (41). Three different PL species (PC, PE, and PI) may, therefore, already be identified from the analysis of the total extract without separation into the individual fractions. Nevertheless, for the
478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504
118 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524
Fuchs et al.
identification of minor components a previous separation is indispensable and can be performed by TLC or HPLC. We have chosen TLC as it is a quite simple and often used method of PL separation. 3. In Fig. 4 a typical TLC plate of the separated liver PLs subsequent to primuline staining is shown. The left lane (4a) corresponds to a standard mixture of known composition and the right lane (4b) corresponds to the liver extract. From the TLC it is evident that the liver extract consists of five lipid classes, namely sphingomyelin (SM), phosphatidylcholine (PC), phosphatidylethanolamine (PE), quite small amounts of phosphatidylinositol (PI), and very small amounts of phosphatidylserine (PS). Please note that the observed slight differences in the migration properties between the standard and the liver extract are caused by differences in the fatty acyl compositions of the individual lipid classes. After visual inspection of the TLC plate, the identified lipid classes were scraped off from the TLC plate, the lipids eluted from the silica gel, and the obtained fractions individually characterized by MALDITOF MS.
Fig. 4. Videoimage of a typical HPTLC plate of a reference PL mixture (a) and the liver extract (b) subsequent to primuline staining (33). 1.69 mg of each PL were applied onto the TLC plate as standard (a) and the total amount of liver lipids was 6 mg (b). The detailed reasons why the PS results in a relatively diffuse spot are yet unknown. LPC lyso-Phosphatidylcholine, PC phosphatidylcholine, PI phosphatidylinositol, PE phosphatidylethanolamine, PS phosphatidylserine, SM sphingomyelin.
Capabilities and Drawbacks of Phospholipid Analysis
119
4. The obtained spectra are shown in Fig. 5. It is evident that the quality of these mass spectra differs significantly. The best spectra regarding the signal-to-noise ratio are stemming from the SM (5a), PC (5b), and PE (5f). The qualities of the PI (5d,5e; negative and positive ion mode, respectively) and particularly the PS fraction (5c) are much worse. This is not only caused by their rather small contribution to the total amounts of lipids, but also by their high polarity and negative charge. Nevertheless it should be noted that the PS could not be identified neither in the positive nor the negative ion spectra of the total liver extract. This is a clear indication that separation is necessary if minor species of a mixture are of interest. All assignments are additionally provided in Table 1. Although the identification of an unknown compound is normally simpler as negative ion because the related adduct pattern is less complicated, the contribution of DHB matrix signals is much more pronounced in the negative ion spectrum (5d) and all the peaks at m/z = 845, 851, and 857 are stemming from the applied DHB matrix (24). This may lead to the suppression of minor species and is also the reason why no convincing negative ion spectrum of PS could be obtained. Please note that in some fractions oxidation of lipids is obvious. For instance the peak at m/z = 863.5 (5d) is caused by PI 18:0/18:1 and the intense peak at m/z = 753.5 corresponds to an oxidation product of that PI at the double bond of the oleoyl residue under generation of the corresponding aldehyde that is schematically shown in Fig. 5. The same mechanism is also valid in the case of the PS, where the peak at m/z = 702.5 is derived from the oxidation of PS 18:0/18:1 at m/z = 812.6. It is also not surprising that the minor fractions are most sensitive to oxidation. Details of this oxidation process are available in (16). Although oxidation is normally an unwanted process, it should be noted that these fragmentation products are sometimes also helpful because they allow the assignments of the positions of the double bonds. Of course, there are still open issues: First, the question of the most suitable matrix for lipid analysis by MALDI-TOF MS is not yet sufficiently answered. We have shown here that the matrix has a tremendous influence on spectral quality as well as the detectabilities of the individual lipid classes. Therefore, it is a strong hope that other still more suitable matrix compounds for the analysis of lipids will be discovered – maybe compounds that allow the detection of all lipid classes even in crude mixtures with higher sensitivity (14), lower matrix background (42), and improved reproducibility (43). Second, some caution is necessary if spectra are recorded from lipid fractions that were purified by means of chromatography (44). There is evidence that lipids differing in their fatty acyl compositions are not released from the TLC plate to the same extent and
525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572
Fig. 5. Typical MALDI-TOF mass spectra of the lipid extract shown in Fig. 3 subsequent to TLC separation into the individual lipid classes. Subsequent to the TLC run, the individual fractions were scraped off and re-eluted from the silica gel of the TLC plate for MALDI-TOF MS. Traces correspond to the following lipid classes: SM (a), PC (b), PS (c), PI (d), (e) and PE (f). All spectra are positive ion spectra, the only negative ion spectrum was obtained from the PI fraction (trace (5d)). Please note that the contribution of matrix peaks (marked by asterisks) is much higher in the negative than the positive ion mode. The most common oxidation pathway of lipids leading to characteristic fragmentation at the position of the double of an oleoyl residue is also shown.
Capabilities and Drawbacks of Phospholipid Analysis
121
saturated or moderately unsaturated lipid “stick” more tightly to the TLC plate. This is confirmed by the slight intensity differences of the PC species in trace (3a) and (5b). Therefore, some lipids may be lost during the extraction process. However, this problem may be overcome by the direct “scanning” of the TLC plate by MALDI-TOF MS without previous extraction of the separated fractions. This method has been recently reported by two different approaches (19, 32) and it is expected that it will develop significantly in the future. There are strong indications that this technique will also be applicable to complex lipids from, e.g., stem cells and other important cell lines (45).
4. Notes
573 574 575 576 577 578 579 580 581 582 583
584
1. Of course all commercially available MALDI-TOF devices can be used independent of the manufacturer. If product or company names are given here, this only means that the indicated products were used for performing the presented experiments. This is sometimes important as different nomenclatures may be used. For instance, “delayed extraction” and “energy lag focusing” are synonyms and mean the same. However, MALDI devices that are exclusively capable of recording linear mode spectra (23), are less suitable for lipid analysis because resolution as well as mass accuracy is reduced in comparison to reflector mode spectra. Isotopic resolution is, however, often important because this spectral feature enables the estimation of the molecular formula – at least in the case of smaller molecules.
585
2. Under the conditions described here, the matrix and the analyte give a relatively “thick” layer of hundreds of mm on the target. Only the upper layers are actually relevant for laser ablation. Therefore, the MALDI target material is less important than by using very thin layers. We are using by default gold-coated targets as they are expected to have a lower content of catalytically active transition metals than stainless steel (16).
598
3. Normally, the use of plastic pipettes (as well as other plastic material) is strongly discouraged because CHCl3 is a rather aggressive solvent that releases impurities as plasticers from the plastic material that may interfere with the signals of the analyte of interest (23). However according to our experience “grey” original Eppendorf pipette tips (up to a volume of 20 ml) may be used without problems. It is, however, recommended to check the potential contribution of impurities by using a sample of known composition and known concentration. Unfortunately, although CHCl 3 is
605
586 587 588 589 590 591 592 593 594 595 596 597 599 600 601 602 603 604 606 607 608 609 610 611 612 613 614 615
122 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661
Fuchs et al.
a hazardous solvent, all attempts to replace it, for instance, by hexane and isopropanol, were not very successful so far in lipid research. 4. Different solvents are normally required to solubilize matrix and lipid. In the majority of cases, the lipid stock solutions will be prepared in CHCl3 or CHCl3/CH3OH mixtures. Unfortunately, the most common MALDI matrix – DHB – is nearly insoluble in CHCl3, but highly soluble in CH3OH. Due to the different volatilities of both solvents (the boiling point of CHCl3 is 61.2°C and that of CH3OH 64.5°C), the lipid will crystallize prior to the DHB resulting in rather inhomogeneous cocrystals. This problem can be minimized by drying the native matrix/sample mixtures rapidly under a warm stream of air. The extent of lipid oxidation induced by air-drying of the sample is practically negligible. 5. Special attention should be paid to the salt content of the matrix as well as the solvents. Changes of the salt content may lead to changes of the peak patterns and affect the ratio between H+ and Na+ adducts. Additionally, the peaks stemming from the DHB matrix are also influenced by changes of the salt content (37). Please note that a spectrum of DHB crystallized from CH3OH differs from that in the presence of salts. 6. All used solvents should be of highest quality! Due to the high sensitivity of MS even very minor impurities are detected: The detection of analytes in small amounts is often a minor problem than NOT detecting impurities stemming from the solvents or the used reagents. 7. It is not implied that under the used experimental conditions all lipids are completely extracted: Some lipids may also stick to the proteins that precipitate at the interphase between the aqueous and the organic layer. If protein-rich samples are investigated, higher ionic strength, i.e., a high salt concentration is recommended in order to reduce the loss of lipids due to the binding to the protein. Complete extraction of lipids from biological tissues is a science of its own and the extraction must be optimized in dependence on the tissue or body fluid of interest. Under the recommended conditions of centrifugation using moderate g-values, special centrifuge glasses are not absolutely required, but common test tubes may also be used. 8. The addition of salt (NaCl) is recommended in order to enhance the dissociation of the lipids from the silica gel. The use of distilled water instead of NaCl solution may lead to the loss of lipids. This particularly holds for negatively charged lipids that are otherwise lost.
Capabilities and Drawbacks of Phospholipid Analysis
123
Abbreviations and Acronyms APCI Atmospheric Pressure Chemical Ionization DE Delayed Extraction DHB 2,5-Dihydroxybenzoic Acid EI Electron Impact ESI Electrospray Ionisation HPLC High-Performance Liquid Chromatography IR Infrared LPC Lyso-Phosphatidylcholine MALDI Matrix-Assisted Laser Desorption and Ionization MS Mass Spectrometry m/z mass over charge PC Phosphatidylcholine PE Phosphatidylethanolamine PG Phosphatidylglycerol PI Phosphatidylinositol PL Phospholipid PNA Para-Nitroaniline PO Palmitoyl-oleoylPS Phosphatidylserine PSD Post Source Decay rpm rotations per minute S/N Signal to Noise SM Sphingomyelin sn Stereospecific Numbering TFA Trifluoroacetic Acid TLC Thin-Layer Chromatography TOF Time-of-Flight UV Ultraviolet
Acknowledgments
662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691
692
This work was supported by the German Research Council (DFG Schi 476/5–1 and FU 771/1-1) and the Federal Ministry of Education and Research (Grant BMBF 0313836). The kind and helpful advice of Dr. Suckau and Dr. Schürenberg (Bruker Daltonics, Bremen) is particularly gratefully acknowledged.
699
References 1. Christie, W. W. (2003) Lipid Analysis, Oily Press, Bridgwater. 2. Wymann, M. P. and Schneiter, R. (2008) Lipid signalling in disease. Nat. Rev. Mol. Cell. Biol. 9, 162–176. 3. Matsumoto, T., Kobayashi, T., and Kamata, K. (2007) Role of lysophosphatidylcholine (LPC)
693 694 695 696 697 698
in atherosclerosis. Curr. Med. Chem. 14, 3209– 3220. 4. Fuchs, B., Schiller, J., Wagner, U., Häntzschel, H. and Arnold, K. (2005) The phosphatidylcholine/lysophosphatidylcholine ratio in human plasma is an indicator of the severity of rheumatoid arthritis: investigations by 31P
700 701 702 703 704 705 706 707
124 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766
Fuchs et al.
NMR and MALDI-TOF MS. Clin. Biochem. 38, 925–933. 5. Wenk, M.R. (2005) The emerging field of lipidomics. Nat. Rev. Drug Discov. 4, 594–610. 6. Han, X. (2007) An update on lipidomics: progress and application in biomarker and drug development. Curr. Opin. Mol. Ther. 9, 586–591. 7. Fahy, E., Subramaniam, S., Brown, H. A., Glass, C. K., Merrill, A. H. Jr, Murphy, R. C., Raetz, C. R., Russell, D. W., Seyama, Y., Shaw, W., Shimizu, T., Spener, F., van Meer, G., VanNieuwenhze, M. S., White, S. H., Witztum, J. L. and Dennis, E. A. (2005) A comprehensive classification system for lipids. J. Lipid Res. 46, 839–861. 8. Peterson, B. L. and Cummings, B. S. (2006) A review of chromatographic methods for the assessment of phospholipids in biological samples. Biomed. Chromatogr. 20, 227–243. 9. Patton, G. M. and Robins, S. J. (1998) Separation and quantitation of phospholipid classes by HPLC. Methods Mol. Biol. 110, 193–215. 10. Touchstone, J. C. (1995) Thin-layer chromatographic procedures for lipid separation. J. Chromatogr. B. 671, 169–195. 11. Wolf, C. and Quinn, P. J. (2008) Lipidomics: practical aspects and applications. Prog. Lipid Res. 47, 15–36. 12. Byrdwell, W. C. (2001) Atmospheric pressure chemical ionization mass spectrometry for analysis of lipids. Lipids 36, 327–346. 13. Pulfer, M. and Murphy, R. C. (2003) Electrospray mass spectrometry of phospholipids. Mass Spectrom. Rev. 22, 332–364. 14. Northen, T. R., Yanes, O., Northen, M. T., Marrinucci, D., Uritboonthai, W., Apon, J., Golledge, S. L., Nordström, A., and Siuzdak, G. (2007) Clathrate nanostructures for mass spectrometry. Nature 449, 1033–1036. 15. Schiller, J., Süß, R., Fuchs, B., Müller, M., Zschörnig, O. and Arnold, K. (2007) MALDI-TOF MS in lipidomics. Front. Biosci. 12, 2568–2579. 16. Schiller, J., Süß, R., Arnhold, J., Fuchs, B., Leßig, J., Müller, M., Petkovic, M., Spalteholz, H., Zschörnig, O. and Arnold, K. (2004) Matrix-assisted laser desorption and ionization time-of-flight (MALDI-TOF) mass spectrometry in lipid and phospholipid research. Prog. Lipid Res. 43, 449–488. 17. Roberts, L. D., McCombie, G., Titmana, C. M. and Griffin J. L. (2008) A matter of fat: An introduction to lipidomic profiling methods. J. Chromatogr. B. 871, 174–181.
18. Hillenkamp, F. and Peter-Katalinic, J. (2007) MALDI MS – A Practical Guide to Instrumentation. Methods and Application. Wiley, Weinheim. 19. Rohlfing, A., Müthing, J., Pohlentz, G., Distler, U., Peter-Katalinić, J., Berkenkamp, S. and Dreisewerd, K. (2007) IR-MALDI-MS analysis of HPTLC-separated phospholipid mixtures directly from the TLC plate. Anal. Chem. 79, 5793–5808. 20. Cohen, L. H. and Gusev, A. I. (2002) Small molecule analysis by MALDI mass spectrometry. Anal. Bioanal. Chem. 373, 571–586. 21. Schiller, J., Arnhold, J., Benard, S., Müller, M., Reichl, S. and Arnold, K. (1999) Lipid analysis by matrix-assisted laser desorption and ionization mass spectrometry: A methodological approach. Anal.Biochem. 267, 46–56. 22. McDonnell, L. A. and Heeren, R. M. (2007) Imaging mass spectrometry. Mass Spectrom. Rev. 26, 606–643. 23. Fuchs, B., Arnold, K. and Schiller, J. (2008) Mass spectrometry of biological molecules. In: Encyclopedia of Analytical Chemistry (Meyers, R. A., ed.), Wiley, Chichester, pp. 1–39. 24. Schiller, J., Süß, R., Fuchs, B., Müller, M., Petkovic, M., Zschörnig, O. and Waschipky, H. (2007) The suitability of different DHB isomers as matrices for the MALDI-TOF MS analysis of phospholipids: which isomer for what purpose? Eur. Biophys. J. 36, 517–527. 25. Estrada, R. and Yappert, M. C. (2004) Alternative approaches for the detection of various phospholipid classes by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. J. Mass Spectrom. 39, 412–422. 26. Gellermann, G. P., Appel, T. R., Davies, P. and Diekmann, S. (2006) Paired helical filaments contain small amounts of cholesterol, phosphatidylcholine and sphingolipids. Biol. Chem. 387, 1267–1274. 27. Murphy, R.C. (2002) Mass spectrometry of phospholipids: Tables of molecular and product ions. Illuminati Press, Denver. 28. Fuchs, B., Müller, K., Göritz, F., Blottner, S. and Schiller, J. (2007) Characteristic oxidation products of choline plasmalogens are detectable in cattle and roe deer spermatozoa by MALDI-TOF mass spectrometry. Lipids 42, 991–998. 29. Harvey, D. J. (1995) Matrix-assisted laser desorption/ionization mass spectrometry of phospholipids. J. Mass Spectrom. 30, 1333–1346. 30. Bligh, E. G. and Dyer, W. J. (1959) A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 3, 911–917.
Capabilities and Drawbacks of Phospholipid Analysis
31. Bartlett, G. R. (1959) Phosphorus assay in column chromatography. J. Biol. Chem. 234, 466–468. 32. Fuchs, B., Schiller, J., Süß, R., Schürenberg, M. and Suckau, D. (2007) A direct and simple method of coupling matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF MS) to thin-layer chromatography (TLC) for the analysis of phospholipids from egg yolk. Anal. Bioanal. Chem. 389, 827–834. 33. White, T., Bursten, S., Frederighi, D., Lewis, R. A. and Nudelman, E. (1998) High-resolution separation and quantification of neutral lipid and phospholipid species in mammalian cells and sera by multi-one-dimensional thin-layer chromatography. Anal. Biochem. 10, 109–117. 34. Zschörnig, O., Richter, V., Rassoul, F., Süß, R., Arnold, K. and Schiller, J. (2006) Analysis of human blood plasma by MALDI-TOF MS – Evaluation of critical parameters. Anal. Lett. 39, 1101–1113. 35. Petkovic, M., Schiller, J., Müller, J., Müller, M., Arnold, K. and Arnhold, J. (2001) The signalto-noise ratio as the measure for the quantification of lysophospholipids by matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry. Analyst 126, 1042–1050. 36. Müller, M., Schiller, J., Petkovic, M., Oehrl, W., Heinze, R., Wetzker, R., Arnold, K. and Arnhold, J. (2001) Limits for the detection of (poly-)phosphoinositides by matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF MS). Chem. Phys. Lipids 110, 151–164. 37. Petkovic, M., Schiller, J., Müller, M., Benard, S., Reichl, S., Arnold, K. and Arnhold, J. (2001) Detection of individual phospholipids in lipid mixtures by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry: phosphatidylcholine prevents the detection of further species. Anal. Biochem. 289, 202–216. 38. Al-Saad, K. A., Zabrouskov, V., Siems, W. F., Knowles, N. R., Hannan, R. M. and Hill, H.
125
H. Jr. (2003) Matrix-assisted laser desorption/ ionization time-of-flight mass spectrometry of lipids: ionization and prompt fragmentation patterns. Rapid Commun. Mass Spectrom. 17, 87–96. 39. Gidden, J., Liyanage, R., Durham, B. and Lay, J. O. Jr. (2007) Reducing fragmentation observed in the matrix-assisted laser desorption/ionization time-of-flight mass spectrometric analysis of triacylglycerols in vegetable oils. Rapid Commun. Mass Spectrom. 21, 1951–1957. 40. Johanson, R. A., Buccafusca, R., Quong, J. N., Shaw, M. A. and Berry, G. T. (2007) Phosphatidylcholine removal from brain lipid extracts expands lipid detection and enhances phosphoinositide quantification by matrixassisted laser desorption/ionization timeof-flight (MALDI-TOF) mass spectrometry. Anal. Biochem. 362, 155–167. 41. Fuchs, B., Schober, C., Richter, G., Süß, R. and Schiller, J. (2007) MALDI-TOF MS of phosphatidylethanolamines: different adducts cause different post source decay (PSD) fragment ion spectra. J. Biochem. Biophys. Methods 70, 689–692. 42. Guo, Z. and He, L. (2007) A binary matrix for background suppression in MALDI-MS of small molecules. Anal. Bioanal. Chem. 387, 1939–1944. 43. Li, Y.L., Gross, M.L. and Hsu, F.F. (2005) Ionic-liquid matrices for improved analysis of phospholipids by MALDI-TOF mass spectrometry. J. Am. Soc. Mass Spectrom. 16, 679–682. 44. DeLong, C. J., Baker, P. R., Samuel, M., Cui, Z. and Thomas, M. J. (2001) Molecular species composition of rat liver phospholipids by ESI-MS/MS: the effect of chromatography. J. Lipid Res. 42, 1959–1968. 45. Fuchs, B., Schiller, J., Süß, R., Zscharnack, M., Bader, A., Müller, P., Schürenberg, M., Becker, M., Suckau, D. (2008) Analysis of stem cell lipids by offline HPTLC-MALDI-TOF MS. Anal. Bioanal. Chem. 392, 849–860.
767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813
sdfsdf
Chapter 7 Lipidomics of the Red Cell in Diagnosis of Human Disorders
1
2 3
Peter J. Quinn, D. Rainteau, and C. Wolf
4
Summary
5
Applications of tandem mass spectrometry in the field of lipid clinical chemistry are considered. Haemato logical and biochemical advantages are presented favoring the choice of red blood cell membranes as a starting material in a wide variety of biomedical fields. Practical considerations are discussed with respect to methods of sampling, storage, and lipid extraction of red blood cells. The chapter describes the capabilities of a direct infusion of raw lipid extracts in the electro-spray ionization source compared with the more sophisticated method of high-performance liquid chromatography coupled with hybrid tandem mass spectrometry. Both methods have been evaluated and have been shown to be suitable for diagnosis and/or monitoring for a variety of human disorders.
6 7 8 9 10 11 12 13
Key words: Lipidomics, Diagnosis, Erythrocyte, Membrane, Essential fatty acid deficit, Peroxisomal defects, Plasmalogen, Oxidized phosphatidylethanolamines
14 15
1. Introduction
16
The human red blood cell (RBC) membrane is an object pathological investigation because it can provide a convenient and reliable indica tion of a range of pathological conditions. This work is devoted to the description of the development of routine methodologies for lipid profiling of the human red cell membrane. One develop ment is the routine profiling of the fatty acid substituents of the complex lipids of the erythrocyte membrane. Due to the rigorous and precise homeostatic regulatory mechanisms that operate in the membrane it is possible, for example, to detect essential fatty acid (EFA) deficit in children (i.e., exocrine pancreatic insufficiency of mucoviscidosis) as well as in adults (i.e. extensive enterectomy in advanced inflammatory bowel diseases). Other more subtle Donald Armstrong (ed.), Lipidomics, Methods in Molecular Biology, vol. 579, doi 10.1007/978-1-60761-322-0_7, © Humana Press, a part of Springer Science + Business Media, LLC 2009
127
17 18 19 20 21 22 23 24 25 26 27 28
128
Quinn, Rainteau, and Wolf
alterations in the molecular species of RBC phospholipids are also investigated for long-term nutritional consequences of sup plemented diets or biosynthetic defects of ether–lipids associated with peroxisomal enzyme blockade. The biosynthetic defects of ether–lipids together with very long chain fatty acid accumula tion are needed for the diagnosis of 12 out of 17 neurological conditions associated with one of the multiple forms of peroxisomal deficit (1). The complexity of the clinical diagnosis of peroxi somal disorders in children is an incentive for more widespread applications of lipidomic analysis. The analysis is also recommended to monitor the outcomes of diet and supplementation treatments because EFA profiling of the RBC membrane is a much more reliable index of deficit than profiling fatty acids of serum lipoproteins. This is because of a relatively slow turnover of fatty acyl components of complex membrane lipids as compared to renewal of neutral lipids and phos pholipids circulating in serum. Serum lipolytic enzymes (secretory PLA2, Lecithin-Cholesterol Acyl Transferase, Lipoprotein Lipase) are relatively active compared with low activities of phospholipase or acyltransferase in human erythrocyte membranes and the cells remain in circulation with a normal lifetime of 120 days. As a result of the relatively stable RBC fatty acid composition variations of a few percent of fatty acid molar ratios can be a significant indica tor of EFA deficit. For example, normal proportions of 20:4n-6 is 11–14% but values less than 10% may be considered as a deficit. The original analytic protocols of fatty acid profiling were based on analysis of hydrolyzed lipids and production of volatile methyl ester derivatives (FAME) using gas–liquid chromatography. The introduction of liquid chromatography coupled with mass spec trometry allows the identification of molecular species of mem brane phospholipids with relatively slow turnover rates thereby improving precision in indexing the EFA profile. Phosphatidylcho lines, for example, are subject to a different turnover rate compared with phosphatidylethanolamines populating the inner leaflet of the RBC membranes which normally are not accessible to circulating enzymes. However, phosphatidylethanolamines are subjected to enzymes from the inner side of the membrane. In addition phos phatidylcholine species are comprised of lower proportions of EFA than phosphatidylethanolamines which represent a reservoir enriched in EFA which expands under conditions of nutritional supplementation. Phosphatidyserines and phosphatidylinositols are enriched in the, so called, “distal” EFA such as 20:4n-6 and 22:6n3. These particular EFA are generated as the end products of bio synthetic pathways in the liver. The pathways comprised of acyl chain elongation and unsaturation lead to the interconversion of shorter chain and less unsaturated precursor EFA, such as 18:2n-6 and 18:3n-3, respectively. It is anticipated that future develop ments will enable distinction between nutritional deficits of EFAs from those derived by hepatic interconversion. Besides global liver
Lipidomics of the Red Cell in Diognosis of Human Disorders
129
failure it is expected that specific elongase/desaturase steps are crit ically impaired in pathology such as insulin resistance and diabetes, obesity, and hepatosteatosis (2). New therapeutic developments are aimed to remedy these defects. The measurement of kinetics of refilling with EFA into particular molecular species after enriched diets may also provide useful clinical information. Other applica tions include immunomodulation correlated with a shift in the bal ance of n-3/n-6 EFA. The assumption that inflammation may be prompted by an inappropriate eicosanoid synthesis in which there is obvious medical interest has not been supported yet by total FA methyl ester profiling studies. However, it is anticipated that with application of recent lipidomics methods described in this Chap ter evidence of a significant linkage will be established between abnormalities in the FA composition of a few molecular species of precursor phospholipids and the inflammatory condition. Current lipidomic methods using mass spectrometry are able to resolve the molecular species composition of the five major phosphol ipid classes present in human erythrocyte membranes: phosphatidyl cholines (PC), sphingomyelins (SM), phosphatidylethanolamines (PE), phosphatidylserines (PS), and phosphatidylinositols (PI). Cho lesterol and detailed glycolipid composition can also be determined but discussion of this is outside the scope of this presentation.
2. Materials Lipidomics studies of patient erythrocyte membranes were conducted at CHU Saint Antoine. The MS facility at the hospital center includes two complementary tandem MS2 equipments (a) a hybrid tandem QTrap 2000 (Applied Biosystems/MDS SCIEX) is coupled with high-pressure liquid chromatography (HPLC Agilent series 1100). (b) The triple quadrupole (TQ) API3000 (Applied Biosystems/MDS SCIEX) system is presently used for direct infusion (Harvard Apparatus syringe pump 11 Plus). Mass spectra are quantified after translation of the propri etary file format.wiff (Analyst, Applied Biosystems) to .cdf format using the translation module included in the software. Then mass spectra are computed for peak identification and deisotopisation using the software LIMSA (3).
3. Methods 3.1. The Use of RBC Membranes for Medical Investigations
As a tissue indicator of metabolic or nutritional disturbance or disease states, the human erythrocyte offers a number of advan tages. The cells are easily obtained without inordinate invasion
130
Quinn, Rainteau, and Wolf
and can be drawn at suitable intervals for monitoring purposes. They are nondividing cells with a single, relatively stable cell membrane. This section details the processes of haematopoiesis and the origin and turnover of the constituent lipids of the mem brane. This is followed by a consideration of sampling, storage, and processing of lipids for lipidomic analysis. 3.1.1. Haematopoiesis and RBC Lifetime
Haematopoietic stem cells of the bone marrow give rise to all the cellular components of blood. The pool of these multipotent cells is maintained by retention of multipotency by a propor tion of the daughter cells. Differentiation of the stem cells takes place in a stepwise manner in which changes in gene expression that place constraints on cell destiny move the cells closer to toward a mature erythrocyte. These changes are reflected in the proteins detected on the cell surface. The differentiation of erythrocytes from myeloid progenitor cells is augmented by hematopoietic growth factors and principally, erythropoi etin, produced in the kidneys. Prior to and immediately after leaving the bone marrow, the differentiating erythrocytes pass through a reticulocyte stage. Reticulocytes represent about 1% of circulating red blood cells. This proportion of reticulocytes is commonly determined to be used as a measurement of RBC turnover rate. In turn, the percentage is an indication of RBC lifetime with consequences on membranogenesis and the lipid incorporation rate at the time of measurement. Erythrocytes develop from committed stem cells through reticulocytes to mature erythrocytes in about 7 days and survive in circulation for a total of about 120 days. Ageing erythrocytes become damaged due to irreversible changes in the cell membrane, some of them related to mem brane lipids such as the outer exposure of phosphatidylserines and phosphatidylethanolamines, which are recognized by macrophages in the spleen, bone marrow, and liver and are removed from the circulation by phagocytosis. Much of the important breakdown products (iron) are then recirculated in the body but little is known of recirculation of lipids and EFA. Some physiological conditions (e.g., gender and altitude acclimation) influence turnover of RBCs and this, in turn may interfere with the lipid composition of the membranes. These changes have not been yet detailed by the recent lipidomic methods described below. Adult humans have approximately 2–3 × 1013 red blood cells in circulation at any given time (women have about 4–5 × 106 erythrocytes/mL of blood and men about 5–6 × 106). Eryth rocytes from adults and children with reduced vitamin E intake and with low vitamin E serum levels that are relatively sensitive to oxidants in vitro have a reduced lifetime. The lifetime of eryth rocytes in premature infants is also shorter than those in normal infants (4).
Lipidomics of the Red Cell in Diognosis of Human Disorders
131
Polycythemias are characterized by an increased concentra tion of circulating erythrocytes, some such as polycythemia vera is an acquired stem cell disease and others are congenital and are caused by mutations of the Erythropoitin-receptor gene, haemo globin variants, 2,3-bisphosphoglycerate mutase deficiency, or by disturbances in renal oxygen sensing. Acquired polycythemias can occur secondary to hypoxia at high altitudes, or primarily through acquired mutations in the EPO-receptor signaling system (JAK2 mutations). Alternatively they may be caused by pulmonary or renal diseases which perturb fatty acid composition of RBC with numerous intricate causes. Because they are relatively long lived in the circulation, their membrane lipid composition is more representative of the longer term lipid homeostatic situation than is the lipids of the plasma. The latter varies in lipid composition according to short-term dietary and physiological circumstances. 3.1.2. Regulation of Phospholipid Synthesis During Membrane Biogenesis
The cell membrane of the mature erythrocyte is not synthesized de novo but it is transformed from its progenitor cells by a proc ess of membrane differentiation. The primary site of synthesis of membrane lipids and proteins is the endoplasmic reticulum where most of the enzymes required for these tasks are located. The biosynthesis of membrane phospholipids and their incorpo ration into the membrane takes place on the cytosolic monolayer of the progenitor endoplasmic reticulum. Phospholipid flippases are needed to translocate newly synthesized phospholipids to the opposite leaflet of the membrane (luminal in the progenitor to become extracellular in the mature RBC). Phosphatidylcho line transferase operates to concentrate molecular species of this phospholipid class to the protoplasmic leaflet of the membrane but there are no transferase specific for PE, PS, and PI, so these phospholipid classes are predominant in the cytoplasmic leaflet (to become intracellular in the mature RBC). This asymmetry is preserved when transition vesicles derived from the endoplas mic reticulum fuse with membranes of other organelles of the endomembrane system including, ultimately, the plasma membrane. In addition, soluble lipid exchange proteins which are specific for particular membrane lipids, such as glucosylceramides, are present which translocate these lipids from the site of synthesis to the cytoplasmic surface of the plasma membrane. During passage of membrane through the endomembrane system of precursor RBC cell, the lipid composition changes as it is transformed into different morphologically distinct membrane. This process involves retailoring the molecular species of lipids that comprises the bilayer matrix. This process can involve both complete turnover of the lipid where the lipid is removed from the membrane and replaced by another molecular species of the same lipid class and partial turnover, when only a component of
132
Quinn, Rainteau, and Wolf
the lipid is exchanged for another. Complete turnover of lipids is mediated by phospholipase C-type enzymes in which the products of hydrolysis no longer possess an amphipathic character (diacylg lycerols and water-soluble phosphate esters). Partial turnover is of two types. The first is where the fatty acyl chains are exchanged one for another to remodel the molecular species of phospholipid. These types of reaction involve phospholipase A-type enzymes with an intermediate lysophospholipid product that is strongly amphipathic in character. The hydrolysis of fatty acids and reacyla tion of the lysophospholipid is known as the Lands cycle. The sec ond type of partial turnover is catalyzed by phospholipase D-type enzymes in which the intermediate product is phosphatidic acid. The reaction involves an exchange of bases on phosphatidic acid and its significance is concerned mainly with the biosynthesis of phosphatidylserine from phosphatidylethanolamine. The partial turnover of molecular species of phosphatidylinositols may also be achieved by this mechanism. During differentiation of reticulocytes into mature eryth rocytes fatty acids provided from diet and presented from the digestive system and/or after interconversion in the liver are incorporated into membrane lipids. This process takes place during the many steps of cellular differentiation from the early erythroid progenitors (burst-forming-units-erythroid) to late erythroid progenitors (colony-forming units-erythroid) and to morphologically recognizable erythroid precursors. Ultimately the endomembrane system including the nucleus is lost and the mature erythrocyte is no longer in possession of the complete machinery for producing membrane lipids. The lipid composition of the erythrocyte membrane is thereon dependent on removal of lipids by metabolic turnover or the limited remodeling of the existing membrane lipids using substrates and circulating enzymes that are provided from the plasma (secretory PLA2, LCAT, etc.). 3.1.3. Incorporation of Fatty Acids into Membrane Phospholipids
The key building block of membrane phospholipids is diacylg lycerol. The fatty acids of the glycerides reflect, to a large extent, the nutritional and dietary circumstances of the patient and are modified by pathological conditions. There are two principal pathways for de novo phospholipid biosynthesis. The fatty acid of triacylglycerol at the C3 position is hydrolyzed to form 1, 2-diacylglycerol. This may react with CDP derivatives of choline or ethanolamine to form PC and PE, respectively. Alternatively, 1, 2-diacylglycerol may be phosphorylated at the sn-3 position by the action of diacylglycerol kinase to form phosphatidic acid. Phosphatidic acid, in turn, reacts with CTP to form CDP-dig lyceride. The product can react directly with serine or inositol for PS and PI, respectively. Thus the origin of the fatty acids origi nates from the metabolic stores of triacylglycerol. In the case of sphingolipid biosynthesis from sphingosine bases, the fatty acids
Lipidomics of the Red Cell in Diognosis of Human Disorders
133
amide bonded to the precursor to form ceramide are distinctive from those acylated to the phospholipid molecular species. These particularly long-chain, more saturated fatty acids are preferentially utilized for biosynthesis of the sphingolipids. 3.1.4. Remodeling of Phospholipids by a Deacylation/ Reacylation Cycle
As indicated above, remodeling of membrane phospholipids via acyl exchange reactions in the Lands cycle is an active pathway in most tissues. The molecular species of membrane phospholipids is a reflection of the nature of the fatty acid substrates and the physical signals generated by the bilayer structure which govern the type of fatty acids that can be incorporated. The Lands cycle is highly active and plays an important role in membrane differ entiation. Operation of the cycle requires phospholipase A2/CoA ligase/acyltransferase enzyme components. The activity of these enzymes in mature erythrocytes, however, is very low compared with those in other tissues such as the liver. Moreover, the rela tively high concentrations of serum albumin (>30 g/L) strongly bind lysoderivatives formed in the RBC membranes. This means that in erythrocytes the turnover of molecular species of phos pholipids is relatively slow so that reliable long-term homeostasis can be characterized from a lipidomic analysis of the erythrocyte membrane. Recently the apoptosis-associated changes in the glyc erophospholipid composition of hematopoietic progenitor cells have been monitored by lipidomic methods (MALDI-TOF and 31 P-NMR) (5). The study has shown that withdrawal of growth factor induces an increase in the ratio of ether-linked glycero phospholipids to diacyl-glycerophospholipids during apoptosis as well as the relative decrease in the membrane of diacyl-phosphati dylcholine (PC) levels.
3.1.5. Membrane Lipids in Human Hemoglobinopathies
The lipid molecular species composition of the red blood cell is preserved within relatively narrow limits throughout the lifetime of the cell. Retailoring of the constituent lipids is accomplished by uptake of fatty acids from the serum which, in turn, reflects the intake of lipids in the diet. This intake can be highly variable so that the process of lipid turnover must be selective in that the fatty acids incorporated into complex membrane lipids are tightly regulated. The biophysical sensors of membrane lipid composi tion are, as yet, unknown but modulation of the processes of acyl chain turnover must take place. Only a few members of the enzyme pathways involved have so far been identified, and little is known about how regulation of catalysis is achieved through interaction of the enzymes with their lipid environment that would clarify the mechanisms that govern this selectivity. Never theless, changes in membrane lipids provide useful indicators of a range of hemoglobinopathies (6). It may be expected that phospholipid turnover and repair is higher in hemoglobinopathies because damage to the mem
134
Quinn, Rainteau, and Wolf
brane resulting from free radical reactions is often a feature of these conditions. A number of reports indicate evidence of lipid oxidation in RBCs from sickle cell or thalassemia patients, ((7, 8)) suggesting that phospholipid repair is not efficient enough to maintain the proper lipid molecular species composition in these cells. Analysis of the lipid molecular species of density fraction ated sickle cells reveals a complex picture. Fractions of greater density which are comprised of more saturated molecular species of phospholipids have more variable proportions of unsaturated phospholipids comparing one patient with another. This obser vation infers that the processes responsible for turnover and repair of membrane phospholipids differ from one cell to another because all cells share the same serum substrate pool. It is note worthy that when the exchange of phospholipids between RBCs is facilitated by lipid exchange, the densest fractions are those that undergo the greatest change of density (9). Erythrocytes are unable to replace proteins that are irrevers ibly damaged by reactive oxygen species. Thus, incorporation of fatty acids in erythrocyte membrane phospholipids is affected when cell suspensions are oxidatively stressed in vitro. Since complex pathways are employed in lipid repair, turnover dam age to any one member of these pathways has repercussions in membrane lipid composition. The relative sensitivity of the indi vidual proteins toward oxidant stress must vary but not much is known of where the vulnerabilities occur. Other than direct oxidative damage of these proteins, alterations in the ionic balance of the cytosol may also play a role. The recently identified acylCoA:Lysophosphatidylcholine acyltransferase (LPCAT) family, for example, which exhibits an EF-hand motif at the C-terminus, was shown to be modulated by calcium and magnesium (10). Since both cytosolic calcium and oxidant stress are altered in sub populations of sickle RBCs, this suggests that LPCAT may act differently in subpopulations of sickle RBCs, leading to insuf ficient or miss repair of membrane lipids. Overall, oxidant stress and alterations in the cytosol of erythrocytes from patients with hemoglobinopathies will lead to the inability to maintain a proper lipid composition, which in turn leads to alterations in membrane function, including ion and water transport across the bilayer and altered red cell density distributions. The alteration in the sickle cell membrane lipids may also be understood as a consequence of the shorter erythrocyte lifespan and accelerated membranogenesis in the progenitor cells. 3.2. Practical Considerations for Lipidomic Analysis of Human Erythrocytes
Blood are the most convenient liquid “tissue” for analysis because obtaining samples is less invasive and samples can be taken at frequent intervals for monitoring purposes. The physiological state of the subject must be standardized because blood lipids vary accordingly between subpopulations. Erythrocytes must be
Lipidomics of the Red Cell in Diognosis of Human Disorders
135
isolated from other blood cells and serum lipoproteins by differen tial centrifugation methods under conditions where lipid changes to the erythrocyte membrane are minimized. Once isolated, the erythrocytes can be stored before lipid extraction and analysis or alternatively the extraction of lipids may proceed immediately on isolation of the cells and the lipid extract stored until examined. This section deals with the practical aspects of sample handling prior to lipidomic analysis. 3.2.1. Blood Sample Collection and Erythrocyte Storage
The first question to be decided is what volume of blood needs to be collected and under what conditions to preserve the integrity of blood cells? Drawing of a 1 ml sample of blood is likely to be sufficient for lipidomic analysis. This volume can be reduced to 0.2 ml in new born babies without compromising the analysis. It is advisable to adapt the blood volume and the tube capacity to minimize the air in the container in contact with the sample. Collection of the sample into a vacutainer on a chelator like EDTA is recommended to reduce the availability of metal cations (Fe2+, Cu2+, etc.) activating lipid peroxidation. This prevents the chance of reactive oxygen species oxidizing unsaturated molecular species of lipid. The blood sample should not be shaken vigorously to minimize the damage to the erythrocytes; vesicles are known to be released from stored erythrocyte membranes after prolonged shaking and can be recovered in the plasma after centrifugation. Because the lipid composition of the released membranes is very different from the plasma the plasma composition is corrupted by minor contamination due to the presence of fragmented membrane vesicles. Likewise, the composition of membrane lipids can be compromised by contamination by lipoproteins adsorbed from plasma. For this reason it is good practice to “wash” the pelleted RBC with five volumes of isotonic sodium chloride, especially if the sample is of a small volume (under 1 ml). The storage conditions before analysis are important. Freezing whole blood causes extensive membrane fragmentation and the resulting plasma/RBC contamination after hemolysis. Freezing RBC only after separation of plasma from the blood cells is rec ommended in order to avoid exchange protein and enzyme activ ities which alter RBC membrane lipids. The activity of the serum enzymes is slow in EDTA-containing samples and a time of 2–4 h for transportation after drawing the sample is reasonable. If samples are to be stored for longer periods of time this should be at −20°C rather than −70°/−80°C. Storage at 4°C is not suffi cient to inhibit all interfering enzyme activities: the sharp decrease in the unesterified membrane cholesterol content under the influ ence of the enzyme L.C.A.T. upon storage for 21 days at −4°C under the standard blood bank conditions is an example. Unlike storage of samples for proteomic studies there is no advantage in storage at the lower temperatures in maintaining lipid integrity.
136
Quinn, Rainteau, and Wolf
The major cause of lipid deterioration is peroxidation and oxy gen gas is more soluble at low temperatures. Also oxygen-derived radicals are more stable at low temperatures, so very cold tempera tures are of no advantage. The oxygen should be chased from the sample by gassing with a dense inert gas like argon; nitrogen is a less satisfactory substitute because, being less dense, it leaks out of the tube unless the container has been flame-sealed. Nitrogen is also less effective in displacing oxygen than argon. An example of LCMS2 method to trace the oxidized phosphatidylethanolamines is given in the last paragraph of this chapter. 3.2.2. Lipid Extraction
Lipidomics studies often refer to studies of crude lipid extracts inferring that there is no requirement for the complete separation of protein, sugar, and lipid classes in the sample subject to analysis. Tandem mass spectrometry has considerably simplified the preanalytical steps as compared to relatively “low tech” procedures such as thin-layer chromatography which are more demanding in terms of purity. This is explained by the high specificity of MS and relatively weak influence of the matrix for monitoring the characteristic transition of a parent molecular ion to a fragment prod uct (tandem MS). Electrospray (ESI) with intrasource separation has led to strategies using 2D mass spectrometry in, so called, shotgun lipidomics and quantitation of cellular lipidomes directly from “crude” extracts of biological samples (11). The intrasource separation procedure is based on the activity of an external electric field to induce separation of cations from negative ions in the infusate while different ionization of molecular species that possess differential electrical propensities can be induced in either the positive- or negative-ion mode during the electrospray ioniza tion process. Intrasource separation and selective ionization are expected to simplify lipid purification prior to MS and result in greater accuracy of the analysis (12). Preanalytical separation steps and enrichment of lipid sample, however, are still required in many cases prior lipidomic final analyses. Ion suppression is a major complication with phospholipids. It can be viewed as the competition for ionization and transfer of molecules from the solvent droplets forming the spray to the gas phase. Because many different molecules are introduced simulta neously in the ESI source and because the volume of the drop lets is extremely small the positive or negative electrical charge in excess are scarce and will be captured by more affine molecular species. Similarly the surface of the droplet is extremely limited and surfactant molecules such as phospholipids are competing to be transferred to the gas phase and enter the mass spectrometer interface orifice. Phospholipids have a high propensity to suppress ionization of coeluting molecules such as drugs (13) or ganglio sides (14). In biological samples, abundant phospholipid molecu lar species compete with less abundant species and examination
Lipidomics of the Red Cell in Diognosis of Human Disorders
137
of diluted samples is recommended for characterization of minor species of phospholipids in the presence of abundant species. Moreover the ionization and transfer to the gas phase is dependent on the saturation and acyl chain length of phospholipid molecular species. This dependence and other ion suppression phenomenon are all increased as a function of the total lipid concentration (15). Inclusion in the infusate of ammonia or solvent-soluble trimethylamine (TEA), diisopropyl-amine (DIPA), or piperidine enhances phospholipid ionization in the negative mode and displaces Na+ and K+ counterions in the positive mode. However ion pairing of PC and SM with amines TEA or DIPA (+101) should be taken into account in the quantitation of the molecular species. To maintain equipment performance during long sampling lists, it is recommended that extracts are clarified. Indeed, insoluble lipid precipitates are frequently observed in solvent mixtures used for chromatography or infusion where phospholipid high concentra tion and limited solubility occurs in the solvent system required for HPLC separation. Precipitation of nonsoluble phospholipids can be reduced by heating the sample vial, connection tubes, and column with heating tape and in a thermostated sampler main tained at around 30–45°C. It should be noted that temperature may profoundly alter the performances of chromatography sys tems and modifies the retention times of components eluted from columns. On a reverse stationary phase grafted with octadecyl (C18) acyl chains the temperature reduces the troublesome ionic interactions and helps to maintain peak symmetry, so reducing the need for ionic pairing (16). On the other hand, retention is delayed on polar stationary phases (e.g., diol-silica) and useful degrees of resolution of phospholipid classes can be achieved at the expense of long analysis time. In erythrocyte membranes the proteins are associated with the lipids by both hydrophobic interactions and polar bonds. At the lipid/water interface hydrogen- and ionic-binding between headgroups are prevalent. Inside the bilayer core hydropho bic binding is responsible for the compactness and ordering of acyl chains and sterol rings. Therefore both polar and nonpolar solvents should be used for extraction of lipids bound to pro teins. Quantitative extraction requires a gradual procedure to separate the lipids from proteins by a method that avoids the rapid denaturation of the proteins and the trapping of lipids sur rounded by aggregated proteins. The clumping into aggregates of denaturated protein reduces lipid extractability into nonpolar nonwater-soluble solvents. A comparison of solvent activities and a description of the most common lipid extraction procedures can be found in the Cyberlipid website: (http://www.cyberlipid. org). Briefly, an aqueous suspension of lipid-containing biologi cal material is first mixed with a polar water-soluble solvent such as methanol (or isopropanol or ether). In a second step the less-
138
Quinn, Rainteau, and Wolf
polar nonwater-soluble solvent (chloroform, hexane) is added in a proportion which forms a single phase for the ternary system (water/water-soluble solvent/nonwater-soluble solvent). In this environment the proteins form a fine precipitate which does not tend to clump and does hinder lipid extraction from RBCs. Lipid omics studies by MS may be conducted directly from this ternary solvent mixture. However, a subsequent step is usually the parti tioning of the filtered previous extract mixture into two immis cible layers (17, 18) after the addition of a large extra volume of aqueous buffer. This partition helps to separate water-soluble constituents acting as an ESI ion-suppressor (proteins, sugar, salts) into an upper layer phase separated from the lower chlo roform layer containing the water-insoluble lipids. Because ionic strength and acidic buffer pH favors the desorption of anionic lipids (e.g., PIPx, gangliosides) bound to basic protein residues it is recommended for extraction protocols with partitioning of the polar lipids into the chloroform layer. Anionic lipids, however, tend to partition into the aqueous-alcoholic layer at pH greater than 4. Excessive acidification results in the cleavage of vinyl ether bond of plasmalogens. In tissues such as the red blood cell mem branes where both acid-resistant and plasmenyl lipids are to be examined, two separate extracts (one extraction protocol with acid and the other without acid addition) can be prepared. In solvent extraction mixtures with chloroform most of the lipids are finally extracted into the lower dense layer which separates from the aqueous-alcoholic upper layer. Proteins accumulate at the aqueous/chloroform layer interface with unfolded sequences oriented upward or downward according to their hydrophilic/ hydrophobic affinities. Indeed a turbid suspension is observed if abundant proteins or low ionic strength or other amphiphiles of the biological preparation stabilize the interface of chloroform droplets. When emulsions are formed (e.g., in some lipid extracts prepared in the presence of high bile acid concentrations), the separation of the upper/lower phases is long lasting and diffi cult. Protocols derived from the Folch procedure (17) have been devised to speed up the separation of layers such as centrifugation, cooling, alteration of chloroform/methanol ratio, or addition of saline buffer. When the lower chloroform layer is turbid it is rec ommended that the extract is percolated through a dehydrating agent to remove water. Barium oxide or anhydrous sodium sul fate is often employed for this purpose. Contamination by water jeopardizes the integrity of lipid extracts after solvent evaporation because a residue of acidic water remains which decomposes vinyl ether bonds. Lipid oxidation is also enhanced at the air interface because evaporation of water takes a long time. To minimize degradation and set up automatic preanalytical facilities for a simultaneous treatment of numerous samples, methods have been described using an automat robot with Solid
Lipidomics of the Red Cell in Diognosis of Human Disorders
139
Phase Extraction (SPE) protocols in which there is no need for partition the solvent/water mixture into two superposed layers. Extraction by SPE reduces the volume of solvent required. The principles of lipid SPE have been reviewed elsewhere (19). Station ary phases are designed for quantitative extraction of phospholip ids on reversed phase (20), or separation of neutral/phospholipid on straight phase (21). The recovery of anionic lipids has to be assessed against standards for polar stationary phases. A recent innovation for lipid extraction suited to highthroughput lipidomics using methyl tert-butyl ether has been described by Matyash et al. (22). Rigorous testing demonstrated that the protocol delivers similar or improved recoveries of spe cies of most of all major lipid classes compared with the ”goldstandard” Folch or Bligh and Dyer recipes. Moreover, because methyl tert-butyl ether forms the upper, less dense layer where lipids partition after separation, it allows faster and cleaner lipid recovery and is well suited for automated shotgun profiling using pipettage robots. 3.2.3. Storage of Lipid Extracts
Liquid chromatography–mass spectrometry (LC–MS) analysis has a high sample throughput relative to the methods described previously for the lipid extraction. Samples are pooled after extraction and stored before analysis as a series. The storage of lipids in organic solvent is critical if not protected against the chemical alteration by water and traces of oxygen. Intrinsic pro tective mechanisms and association with natural compounds act in living organism to defend the vulnerable double bonds of vinyl ether and polyunsaturated fatty acids against attack by reactive oxygen species. The removal of these lines of defense by isola tion and extraction of lipids requires addition of antioxidants. It is therefore usual to add a radical scavenger such as butylated hydroxy-toluene (BHT, 2,6-di-tert-butyl-p-cresol) and/or toco pherol to the extract. Being soluble in solvents of low polarity they can be easily added to the chloroform extract and their oxi dation can act as a potential index of damage to the biological lipids. Nevertheless, it is necessary to ensure that the concen tration of such agents does not exceed a threshold of 0.0001 (w/w) when antioxidants like BHT have been shown to act as prooxidants at higher concentrations. Oxidation can also be exac erbated by contamination with metal cations such as copper or/ and iron which catalyze production of free radicals in Fentontype reactions. Contamination can be avoided by using single-use glass vessels with PTFE fittings and it is recommended for all preanalytical steps in Lipidomics. Brown glass is recommended to prevent exposure of lipid extracts to UV light, for example, when performing operations on a UV-illuminated sterile bench, which can be particularly deleterious. Elevated temperatures also favor double-bond migration and diene conjugation. The isomeriza
140
Quinn, Rainteau, and Wolf
tion proceeds after the initial abstraction of an electron from an intermediary − CH2 methylene group intercalated between two double-bonds. Odd electron radicals derived of lipids induce an extensive peroxidation to other lipids especially in concentrated solutions. It is ideally recommended that lipid extracts are stored at low temperatures (−20°C) diluted in a solvent solution contained in an opaque (brown glass) vessel purged with argon gas and sealed by PTFE tape. The anionic phospholipids such as PI and PS present in RBC membrane extracts are protected in the long term from the head-group and fatty acid cleavage in the form of ammonium salts. Evaporation of the solvent prior to analysis is performed under a stream of oxygen-free dry nitrogen at a temperature not exceeding 40°C. SpeedVac is an effective alternative to this method since numerous samples can be evapo rated simultaneously in vials which can be subsequently trans ferred to an automatic sampler module of an HPLC system. Up to 40 samples can be evaporated simultaneously before they are loaded into the cooled sampler module. Cooling of the injec tion module prevents evaporation of solvent after the septum is perforated by the sampling needle which allows the reinjection of sample for further analysis using different data acquisition modes. Caution should be exercised, however, as temperatures lower than about 10°C may result in precipitation of the fraction of the most insoluble lipids concentrated in the type of solvent system used for HPLC (heating instead of cooling the sampler may be an alternative if sphingomyelins of RBC membranes are studied). Insoluble lipid aggregation results in the splitting of the eluted peaks into micellar and monomeric compounds. More efficient solvents for lipid solubilization like chloroform/methanol binary mixtures (1/1 or 2/1) can interfere with chromatography selec tive retention, the metal tubing of the equipment may deteriorate due to exposure to HCl formed from chloroform and finally the transfer of molecules from the ESI spray to the gas phase can be decreased. Using double layered PTFE-sealed vials and extensive purg ing with argon extracts containing highly polyunsaturated glyc erophospholipids such as human RBC membranes were found to remain unchanged for up to 6 months storage at subzero tem peratures. Sphingolipids are expected to be stable for much longer periods of time since they are generally more saturated. Sterols, steroids, and bile acids are also innately resistant to chemical altera tion as indicated by analyses of paleofecal specimens (23). By con trast, compounds such as 7-dehydrocholesterol or ergosterol with conjugated double bonds in the B-ring require more cautious handling among which is recommended low illumination, inert atmosphere, and analysis as soon as possible after isolation. The last paragraph in the chapter shows the measurement of oxidized phosphatidylethanolamines by LCMS2 after long-term storage.
3.3. Results 3.3.1. Analysis of RBC Phospholipids by Tandem MS2
Lipidomics of the Red Cell in Diognosis of Human Disorders
141
In this section the use of simple tandem MS2 to analyze most of the phospholipids extracted from human RBC membrane following a straightforward protocol will be described. Examination of the raw lipid extracts obtained by the method of Rose-Oklander (24) and methyl tert butyl ether (22) offers many advantages in terms of time and manpower per sample. The method developed for basic research can be now automated and fits appropriately within the context of biomedical applications. The utility of the method can also be compared to methods requiring more sophisticated equipment such as hybrid tandem MS2, Q-Trap, and Q-TOF, coupled with HPLC which are described below. To illustrate the method, results obtained after the direct infusion of the total extract of RBC membrane phospholipids into the ElectroSpray Ionisation (ESI) source of a triple quad rupole (TQ) spectrometer are presented. More comprehensive analyses are obtained if an HPLC separation of phospholipid class and subclass is conducted prior to MS2. However, incorporation of a chromatography step extends the time required for analysis by the length of the chromatographic run. Furthermore, comple mentary HPLC equipment requires pumps, filters, and seals that must be carefully maintained for running long series of samples in clinical applications. Examining RBC extracts for some particular medical applications may require a resolution not achieved by a direct infusion setup. An example of such demanding applica tion is the separation of diacyl/ether phospholipids described in below. The direct infusion of raw phospholipid extracts into ESI–MS has replaced the multiple steps previously used for the separa tion of phospholipid classes and fatty acid composition: recovery of individual phospholipids from TLC plates; phosphorus assay, and fatty acid analysis by gas chromatography. The MS-based method allows detailed analysis of lipid species in the field of biomedicine. The principle applied in tandem MS2 is the cleavage of a spe cific product ion from the phospholipid parent ion, e.g., serine from phosphatidylserine (PS) or phosphorylcholine from phos phatidylcholine and sphingomyelin (SM). Figure 1 shows the specific product ions resulting of collision-induced dissociation (CID) in the tandem MS2. As compared to electron impact fragmentography (energy = 70 eV) the collision of accelerated parent ions through a quadru pole filled with argon or nitrogen has a relatively low energy ( 315
40
0.5
20-hydroxy-LTB4
7.4
351 > 195
40
0.5
20-carboxy-LTB4
7.4
365 > 347
40
2.5
TBX2
8.1
369 > 195
40
0.5
PGF2a
8.6
353 > 309
40
0.5
11-dehydro-TBX2
9.3
367 > 305
40
0.5
LXA4
10.4
351 > 115
40
5.0
LTB5
11.5
333 > 195
40
0.5
LTD4
11.7
495 > 177
40
0.5
LTE4
12.0
438 > 333
50
250
15-dPGJ2
16.9
315 > 271
40
25
20-HETE
17.0
319 > 275
40
5.0
13-HODE
18.6
295 > 195
40
0.5
9-HODE
18.9
295 > 171
40
0.5
15-HETE
19.4
319 > 219
40
2.5
15-KETE
20.6
317 > 113
40
0.5
12-HETE
21.0
319 > 179
40
0.5
5-HETE
22.6
319 > 115
40
0.5
14,15-EET
23.3
319 > 219
40
2.5
11,12-EET
25.0
319 > 167
40
25
5-KETE
25.1
317 > 273
40
25
8,9-EET
25.1
319 > 155
40
2.5
DHA
30.9
327 > 283
50
5.0
AA
31.3
303 > 259
40
5.0
Segment
Analyte
1
6-keto-PGF1a
2
3
4
3.2. Results
Rt (min)
1. We compared differences in separation and detection of eicosanoids under acidic (pH 4), neutral, and basic (pH 8) conditions (Fig. 1a–c, Table 3). Results clearly showed that analytes demonstrated improved performance (i.e., narrower peak shapes, reduced tailing, and improved baseline resolution) under weak acidic conditions. However, a basic mobile phase can potentially be useful for certain eicosanoid targets.
HPLC/MS/MS-Based Approaches for Detection and Quantification
179
Fig. 1. Chromatographic separation of eight eicosanoids ((1) 6-keto PFG1a, (2) LTB4, (3) 15-HETE, (4) 5-KETE, (5) 12-HETE, (6) 5-HETE, (7) 14,15-EET, and (8) 11,12-EET). Gradient parameters are described in Table 3. (a) Gradient run with an additive of 0.1% Acetic acid. (b) Gradient run with an additive of 0.01% NH4+solution (25%). *This peak contains all eight eicosanoids. (c) Gradient run without an additive (neutral conditions). (d) Gradient run without additive, but preconditioned under the same acidic conditions as in a.
Optimal resolution was achieved after preconditioning of the column at acidic pH followed by a neutral gradient (Fig. 1d). These conditions were therefore used for further development of the method. 2. Twenty-four arachidonic and linoleic acid metabolites were analyzed by LC/MS/MS using a 45 min pH gradient (Table 1) and scanning in SRM mode (Table 2). LOD values from 0.5 to 250 pg/mL were observed (Table 2). Figure 2 shows a total ion chromatogram of the analyzed compounds and ion chromatograms obtained for eicosanoids detected 6–10 min into the run. Compounds with similar properties (i.e., PGs, LTs, HETEs, HODEs, KETEs, and EETs) have close retention times but can readily be distinguished by differences in fragmentation pattern. An example of this is shown in Fig. 3
180
Lundström et al.
Table 3 Gradient parameters for Fig. 1a Time (min)
1 (%)b
2 (%)c
3 (%)d
0.0
85
15
0
0.5
85
15
0
1.0
70
30
0
5.0
45
55
0
10.0
25
75
0
10.5
0
100
0
15.0
0
100
0
16.0
85
e
15
100f
21.0
85e
15e
100f
e
a Solvent (1) and (2) were run under neutral conditions or with an additive of either 0.1% Acetic acid or 0.01% NH4+(25% aqueous solution) depending on desired pH with a flow rate of 350 mL/min. b 1: Water. c 2: Acetonitrile:Methanol (88:12). d 3: Water:Acetonitrile:Methanol (85:13:2), 0.1% Acetic acid. e Used parameters for chromatograms a–c in Fig. 1. f Used parameters for chromatogram d in Fig. 1.
Fig. 2. ESI/MS/MS chromatogram obtained for the eicosanoids in Table 2. Details for all SRM parameters are provided in Table 2. (a) Total ion chromatogram. (b–g) Compounds detected in Segment 1: (b) 6-keto PGF1a, (c) 20-hydroxy-LTB4, (d) 20-carboxy-LTB4, (e) TXB2, (f) PGF2a, and (g) 11-dehydro-TXB2.
HPLC/MS/MS-Based Approaches for Detection and Quantification
181
Fig. 3. MS2 analysis of 5-HETE (m/z 319) (a) and of 12-HETE (m/z 319) (b). The fragment ions at m/z 115 and 179 are specific for 5-HETE and 12-HETE, respectively, and can be used in SRM analysis for compound identification and subsequent quantification.
in which the MS2 spectra of 5-HETE (a) and 12-HETE (b) are shown. The dehydrated ion at m/z 301 is present in both spectra, as well as the ion at m/z 257. However, scanning in SRM mode for the diagnostic fragment peaks at m/z 115 for 5-HETE and at m/z 179 for 12-HETE enabled the routine detection and quantification of these compounds.
4. Notes 1. The use of new instrumentation such as a quadrupole ion trap time of flight (QITTOF) or FTICR/MS will result in lower LODs in the range of attomol or even zeptomol
182
Lundström et al.
levels of constituent quantities. The use of these techniques for eicosanoid quantification can greatly increase our ability to detect and quantify a variety of eicosanoids in biological samples that have not been detected previously. This ability is important for clinical applications where the amount of sample available is often a limiting factor. Furthermore, the development of multi-isotope imaging mass spectrometry (MIMS) opens a new field of opportunities, enabling studies of the interaction, detection, and quantification of eicosanoids within a single cell. 2. There are specific challenges with the quantification of CysLTs, which are best analyzed directly by online SPE coupled to the HPLC/MS/MS run in positive mode (7, 95). This procedure is recommended because CysLTs are sensitive to oxidation and offline SPE extraction can result in decreased recoveries. 3. Eicosanoids are sensitive to oxidation and consequently standards should always be stored at low temperatures (e.g., £ −80°C) under an inert atmosphere (e.g., argon) in the absence of light, moisture, and active surfaces. Samples should be kept in a temperature-controlled environment both before as well as during the run (a common feature of many autosamplers). 4. It is important to use HPLC and MS grade solvents for the preparation of chromatographic mobile phases to reduce background noise and keep the instrument in a good working condition. The aqueous phase should be frequently replaced (once per week) in order to reduce the possibility of microorganism growth. In addition, the system should be regularly flushed with organic solvent (e.g., methanol). 5. The limit for the number of compounds that can be identified in a single run depends upon the HPLC separation as well as the ion dwell times (based on the number of analytes in each SRM, the interscan delays, and the peak width of the analytes). Accordingly, this number will vary greatly with different instruments. 6. System stability is particularly important when operated with large sample sets where the MS conducts multiple scans within a single run.
Acknowledgments We gratefully acknowledge the assistance of Malin Nording, Katrin Georgi, Jun Yang, and Pavel Aronov for helpful discussions and critical reading of the manuscript. This research was
HPLC/MS/MS-Based Approaches for Detection and Quantification
183
supported by The Åke Wibergs Stiftelse, the Fredrik and Ingrid Thurings Stiftelse, The Royal Swedish Academy of Sciences, The Swedish Council for Strategic Research, and The Swedish Research Council and the European Commission. C.E.W. was supported by a fellowship from the Centre for Allergy Research (Cfa). S. L. L. was supported by the Osher Initiative for severe Asthma. References 1. Funk CD. (2001) Prostaglandins and leukotrienes: advances in eicosanoid biology. Science 294, 1871–1875. 2. Conrad DJ. (1999) The arachidonate 12/15 lipoxygenases. A review of tissue expression and biologic function. Clin Rev Allergy Immunol 17, 71–89. 3. Newman JW, Watanabe T, Hammock BD. (2002) The simultaneous quantification of cytochrome P450 dependent linoleate and arachidonate metabolites in urine by HPLCMS/MS. J Lipid Res 43, 1563–1578. 4. Natarajan R, Reddy MA. (2003) HETEs/ EETs in renal glomerular and epithelial cell functions. Curr Opin Pharmacol 3, 198–203. 5. Schmelzer KR, Inceoglu B, Kubala L, et al. (2006) Enhancement of antinociception by coadministration of nonsteroidal anti-inflammatory drugs and soluble epoxide hydrolase inhibitors. Proc Natl Acad Sci U S A 103, 13646–13651. 6. Samuelsson B, Dahlen SE, Lindgren JA, Rouzer CA, Serhan CN. (1987) Leukotrienes and lipoxins: structures, biosynthesis, and biological effects. Science 237, 1171–1176. 7. Schmelzer KR, Wheelock AM, Dettmer K, Morin D, Hammock BD. (2006) The role of inflammatory mediators in the synergistic toxicity of ozone and 1-nitronaphthalene in rat airways. Environ Health Perspect 114, 1354–1360. 8. Hoagland KM, Maier KG, Moreno C, Yu M, Roman RJ. (2001) Cytochrome P450 metabolites of arachidonic acid: novel regulators of renal function. Nephrol Dial Transplant 16, 2283–2285. 9. Roman RJ. (2002) P-450 metabolites of arachidonic acid in the control of cardiovascular function. Physiol Rev 82, 131–185. 10. Natarajan R, Nadler JL. (2004) Lipid inflammatory mediators in diabetic vascular disease. Arterioscler Thromb Vasc Biol 24, 1542–1548. 11. Hussey HJ, Tisdale MJ. (1996) Inhibition of tumour growth by lipoxygenase inhibitors. Br J Cancer 74, 683–687.
12. Nie D, Tang K, Szekeres K, Trikha M, Honn KV. (2000) The role of eicosanoids in tumor growth and metastasis. Ernst Schering Res Found Workshop 31, 201–217. 13. Kumlin M. (1996) Analytical methods for the measurement of leukotrienes and other eicosanoids in biological samples from asthmatic subjects. J Chromatogr A 725, 29–40. 14. Powell WS, Funk CD. (1987) Metabolism of arachidonic acid and other polyunsaturated fatty acids by blood vessels. Prog Lipid Res 26, 183–210. 15. Tsukamoto H, Hishinuma T, Mikkaichi T, et al. (2002) Simultaneous quantification of prostaglandins, isoprostane and thromboxane in cell-cultured medium using gas chromatography-mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 774, 205–214. 16. Tsikas D. (1998) Application of gas chromatography-mass spectrometry and gas chromatography-tandem mass spectrometry to assess in vivo synthesis of prostaglandins, thromboxane, leukotrienes, isoprostanes and related compounds in humans. J Chromatogr B Biomed Sci Appl 717, 201–245. 17. Baranowski R, Pacha K. (2002) Gas chromatographic determination of prostaglandins. Mini Rev Med Chem 2, 135–144. 18. Murphy RC, Barkley RM, Zemski Berry K, et al. (2005) Electrospray ionization and tandem mass spectrometry of eicosanoids. Anal Biochem 346, 1–42. 19. Deems R, Buczynski MW, Bowers-Gentry R, Harkewicz R, Dennis EA. (2007) Detection and quantitation of eicosanoids via high performance liquid chromatography-electrospray ionization-mass spectrometry. Methods Enzymol 432, 59–82. 20. Blewett AJ, Varma D, Gilles T, Libonati JR, Jansen SA. (2008) Development and validation of a high-performance liquid chromatography-electrospray mass spectrometry method for the simultaneous determination of 23 eicosanoids. J Pharm Biomed Anal 46, 653–662.
184
Lundström et al.
21. Masoodi M, Nicolaou A. (2006) Lipidomic analysis of twenty-seven prostanoids and isoprostanes by liquid chromatography/electrospray tandem mass spectrometry. Rapid Commun Mass Spectrom 20, 3023–3029. 22. Wilson ID, Plumb R, Granger J, Major H, Williams R, Lenz EM. (2005) HPLC-MSbased methods for the study of metabonomics. J Chromatogr B Analyt Technol Biomed Life Sci 817, 67–76. 23. Zhang JH, Pearson T, Matharoo-Ball B, et al. (2007) Quantitative profiling of epoxyeicosatrienoic, hydroxyeicosatetraenoic, and dihydroxyeicosatetraenoic acids in human intrauterine tissues using liquid chromatography/electrospray ionization tandem mass spectrometry. Anal Biochem 365, 40–51. 24. Han X, Gross RW. (2005) Shotgun lipidomics: electrospray ionization mass spectrometric analysis and quantitation of cellular lipidomes directly from crude extracts of biological samples. Mass Spectrom Rev 24, 367–412. 25. Wheelan P, Murphy RC. (1997) Quantitation of 5-lipoxygenase products by electrospray mass spectrometry: effect of ethanol on zymosan-stimulated production of 5-lipoxygenase products by human neutrophils. Anal Biochem 244, 110–115. 26. Yue H, Jansen SA, Strauss KI, et al. (2007) A liquid chromatography/mass spectrometric method for simultaneous analysis of arachidonic acid and its endogenous eicosanoid metabolites prostaglandins, dihydroxyeicosatrienoic acids, hydroxyeicosatetraenoic acids, and epoxyeicosatrienoic acids in rat brain tissue. J Pharm Biomed Anal 43, 1122–1134. 27. Luria A, Weldon SM, Kabcenell AK, et al. (2007) Compensatory mechanism for homeostatic blood pressure regulation in Ephx2 genedisrupted mice. J Biol Chem 282, 2891–2898. 28. Bylund J, Ericsson J, Oliw EH. (1998) Analysis of cytochrome P450 metabolites of arachidonic and linoleic acids by liquid chromatography-mass spectrometry with ion trap MS. Anal Biochem 265, 55–68. 29. Masoodi M, Mir AA, Petasis NA, Serhan CN, Nicolaou A. (2008) Simultaneous lipidomic analysis of three families of bioactive lipid mediators leukotrienes, resolvins, protectins and related hydroxy-fatty acids by liquid chromatography/electrospray ionisation tandem mass spectrometry. Rapid Commun Mass Spectrom 22, 75–83. 30. Hishinuma T, Suzuki K, Saito M, et al. (2007) Simultaneous quantification of seven prostanoids using liquid chromatography/tandem mass spectrometry: the effects of arachidonic acid on prostanoid production in mouse bone
marrow-derived mast cells. Prostaglandins Leukot Essent Fatty Acids 76, 321–329. 31. Oliw E, Wollheim FA. (2000) [New treatment of pain and fever in rheumatoid arthritis and arthrosis. The first cyclooxygenase-2 inhibitors show promising results]. Lakartidningen 97, 2942–2944. 32. Kempen EC, Yang P, Felix E, Madden T, Newman RA. (2001) Simultaneous quantification of arachidonic acid metabolites in cultured tumor cells using high-performance liquid chromatography/electrospray ionization tandem mass spectrometry. Anal Biochem 297, 183–190. 33. Yang P, Chan D, Felix E, et al. (2006) Determination of endogenous tissue inflammation profiles by LC/MS/MS: COX- and LOXderived bioactive lipids. Prostaglandins Leukot Essent Fatty Acids 75, 385–395. 34. Porter NA, Caldwell SE, Mills KA. (1995) Mechanisms of free radical oxidation of unsaturated lipids. Lipids 30, 277–290. 35. Yin H, Porter NA. (2005) New insights regarding the autoxidation of polyunsaturated fatty acids. Antioxid Redox Signal 7, 170–184. 36. Williams MV, Lee SH, Blair IA. (2005) Liquid chromatography/mass spectrometry analysis of bifunctional electrophiles and DNA adducts from vitamin C mediated decomposition of 15-hydroperoxyeicosatetraenoic acid. Rapid Commun Mass Spectrom 19, 849–858. 37. Hammonds TD, Blair IA, Falck JR, Capdevila JH. (1989) Resolution of epoxyeicosatrienoate enantiomers by chiral phase chromatography. Anal Biochem 182, 300–303. 38. Lee SH, Williams MV, DuBois RN, Blair IA. (2003) Targeted lipidomics using electron capture atmospheric pressure chemical ionization mass spectrometry. Rapid Commun Mass Spectrom 17, 2168–2176. 39. Lee SH, Blair IA. (2007) Targeted chiral lipidomics analysis by liquid chromatography electron capture atmospheric pressure chemical ionization mass spectrometry (LC-ECAPCI/ MS). Methods Enzymol 433, 159–174. 40. Lee SH, Williams MV, Blair IA. (2005) Targeted chiral lipidomics analysis. Prostaglandins Other Lipid Mediat 77, 141–157. 41. Suzuki N, Hishinuma T, Saga T, et al. (2003) Determination of urinary 12(S)hydroxyeicosatetraenoic acid by liquid chromatography-tandem mass spectrometry with column-switching technique: sex difference in healthy volunteers and patients with diabetes mellitus. J Chromatogr B Analyt Technol Biomed Life Sci 783, 383–389.
HPLC/MS/MS-Based Approaches for Detection and Quantification
42. Wei S, Brittin JJ, Falck JR, et al. (2006) Chiral resolution of the epoxyeicosatrienoic acids, arachidonic acid epoxygenase metabolites. Anal Biochem 352, 129–134. 43. Kiss L, Bier J, Roder Y, Weissmann N, Grimminger F, Seeger W. (2008) Direct and simultaneous profiling of epoxyeicosatrienoic acid enantiomers by capillary tandem column chiralphase liquid chromatography with dual online photodiode array and tandem mass spectrometric detection. Anal Bioanal Chem 392, 717–726. 44. Dole M, Mack, L.L., Hines, R.L., Mobley, R.C., Ferguson, L.D. and Alice, M.B.. (1968) Molecular beams of macroions. J Chem Phys 49, 2240–2249. 45. Fenn JB, Mann M, Meng CK, Wong SF, Whitehouse CM. (1989) Electrospray ionization for mass spectrometry of large biomolecules. Science 246, 64–71. 46. Taylor G. (1964) Disintegration of water drops in an electric field. Proc R Soc London, A 280, 383–397. 47. Cech NB, Enke CG. (2001) Practical implications of some recent studies in electrospray ionization fundamentals. Mass Spectrom Rev 20, 362–387. 48. Kingsley PJ, Rouzer CA, Saleh S, Marnett LJ. (2005) Simultaneous analysis of prostaglandin glyceryl esters and prostaglandins by electrospray tandem mass spectrometry. Anal Biochem 343, 203–211. 49. Cui L, Isbell MA, Chawengsub Y, Falck JR, Campbell WB, Nithipatikom K. (2008) Structural characterization of monohydroxyeicosatetraenoic acids and dihydroxy- and trihydroxyeicosatrienoic acids by ESI-FTICR. J Am Soc Mass Spectrom 19, 569–585. 50. Horning EC, Horning MG, Carroll DI, Dzidic I, Stillwell RN. (1973) New picogram detection system based on a mass spectrometer with an external ionization source at atmospheric pressure. Anal Chem 45, 936–943. 51. Caroll DI, Dzidic I, Stillwell RN, Haegele KD, Horning EC. (1975) Atmospheric pressure ionization mass spectrometry: Corona discharge ion source for use in liquid chromatograph-mass spectrometer-computer analytical system. Anal Chem 47, 2369–2373. 52. Montuschi P, Martello S, Felli M, Mondino C, Chiarotti M. (2004) Ion trap liquid chromatography/tandem mass spectrometry analysis of leukotriene B4 in exhaled breath condensate. Rapid Commun Mass Spectrom 18, 2723–2729. 53. Koga D, Santa T, Fukushima T, Homma H, Imai K. (1997) Liquid chromatographicatmospheric pressure chemical ionization
185
mass spectrometric determination of anandamide and its analogs in rat brain and peripheral tissues. J Chromatogr B Biomed Sci Appl 690, 7–13. 54. Rosenberg E. (2003) The potential of organic (electrospray- and atmospheric pressure chemical ionisation) mass spectrometric techniques coupled to liquid-phase separation for speciation analysis. J Chromatogr A 1000, 841–889. 55. Cai SS, Syage JA. (2006) Comparison of atmospheric pressure photoionization, atmospheric pressure chemical ionization, and electrospray ionization mass spectrometry for analysis of lipids. Anal Chem 78, 1191–1199. 56. Locke DC, Dhingra BS, Baker AD. (1982) Liquid-phase photoionization detector for liquid chromatography. Anal Chem 54, 447–450. 57. Robb DB, Covey TR, Bruins AP. (2000) Atmospheric pressure photoionization: an ionization method for liquid chromatography-mass spectrometry. Anal Chem 72, 3653–3659. 58. Cai SS, Syage JA. (2006) Atmospheric pressure photoionization mass spectrometry for analysis of fatty acid and acylglycerol lipids. J Chromatogr A 1110, 15–26. 59. Short LC, Cai SS, Syage JA. (2007) APPIMS: effects of mobile phases and VUV lamps on the detection of PAH compounds. J Am Soc Mass Spectrom 18, 589–599. 60. Ferguson RE, MCCulloh KE, Rosenstock HM. (1965) Observation of the products of ionic collision processes and ion decomposition in a linear, pulsed time-of-flight mass spectrometer. J Chem Phys 42, 100. 61. Paul W, Steinwedel HS. (1960) US Patent, 2939952. 62. Kita Y, Takahashi T, Uozumi N, Shimizu T. (2005) A multiplex quantitation method for eicosanoids and platelet-activating factor using column-switching reversed-phase liquid chromatography-tandem mass spectrometry. Anal Biochem 342, 134–143. 63. Paul W. (1990) Electromagnetic traps for charged and neutral particles (nobel lecture). Angew Chem 29, 739–748. 64. Stafford Jr GC, Kelley PE, Syka JEP, Reynolds WE, Todd JFJ. (1984) Recent Improvements in Analytical Application of Advanced Ion-Trap Technology. Int J Mass Spectrom Ion Processes 60, 85–98. 65. Louris J, Brodbelt-Lustig J, Cooks R, Glish G, Van Berkel G, McLuckey S. (1990) Ion isolation and sequential stages of mass spectrometry in
186
Lundström et al.
a quadrupole ion trap mass spectrometer. Int J Mass Spectrom Ion Processes 96, 117–137. 66. Liminga M, Oliw E. (2000) Qualitative and quantitative analysis of lipoxygenase products in bovine corneal epithelium by liquid chromatography-mass spectrometry with an ion trap. Lipids 35, 225–232. 67. Kiss L, Roder Y, Bier J, Weissmann N, Seeger W, Grimminger F. (2008) Direct eicosanoid profiling of the hypoxic lung by comprehensive analysis via capillary liquid chromatography with dual online photodiode-array and tandem mass-spectrometric detection. Anal Bioanal Chem 390, 697–714. 68. Hevko JM, Murphy RC. (2001) Electrospray ionization and tandem mass spectrometry of cysteinyl eicosanoids: leukotriene C4 and FOG7. J Am Soc Mass Spectrom 12, 763–771. 69. Hong S, Lu Y, Yang R, Gotlinger KH, Petasis NA, Serhan CN. (2007) Resolvin D1, protectin D1, and related docosahexaenoic acidderived products: Analysis via electrospray/ low energy tandem mass spectrometry based on spectra and fragmentation mechanisms. J Am Soc Mass Spectrom 18, 128–144. 70. Thomson JJ. (1897) Cathode rays. Philos Mag 44, 29. 71. Stephens WE. (1946) Pulsed mass spectrometer with time dispersion. Phys Rev 69, 691. 72. Shui G, Bendt AK, Pethe K, Dick T, Wenk MR. (2007) Sensitive profiling of chemically diverse bioactive lipids. J Lipid Res 48, 1976–1984. 73. Dickinson JS, Murphy RC. (2002) Mass spectrometric analysis of leukotriene A4 and other chemically reactive metabolites of arachidonic acid. J Am Soc Mass Spectrom 13, 1227–1234. 74. Harks EG, Peters PH, van Dongen JL, van Zoelen EJ, Theuvenet AP. (2005) Autocrine production of prostaglandin F2alpha enhances phenotypic transformation of normal rat kidney fibroblasts. Am J Physiol Cell Physiol 289, C130–137. 75. Laude DA. Electrospray Ionization/Fourier Transform Ion Cyclotron Resonance Mass Spectrometry. In: Cole RB, ed. Elctrospray Ionization Mass Spectrometry: Fundamentals, Instrumentation, and Applications. New York: Wiley, 1997:291–320. 76. Page JS, Masselon CD, Smith RD. (2004) FTICR mass spectrometry for qualitative and quantitative bioanalyses. Curr Opin Biotechnol 15, 3–11. 77. Wu Q, Van Orden S, Cheng X, Bakhtiar R, Smith RD. (1995) Characterization of cytochrome c variants with high-resolution FTICR mass spectrometry: correlation of
fragmentation and structure. Anal Chem 67, 2498–2509. 78. Flora JW, Hannis JC, Muddiman DC. (2001) High-mass accuracy of product ions produced by SORI-CID using a dual electrospray ionization source coupled with FTICR mass spectrometry. Anal Chem 73, 1247–1251. 79. Hofstadler SA, Sannes-Lowery KA, Griffey RH. (1999) Infrared multiphoton dissociation in an external ion reservoir. Anal Chem 71, 2067–2070. 80. Stensballe A, Jensen ON, Olsen JV, Haselmann KF, Zubarev RA. (2000) Electron capture dissociation of singly and multiply phosphorylated peptides. Rapid Commun Mass Spectrom 14, 1793–1800. 81. Flora JW, Muddiman DC. (2004) Determination of the relative energies of activation for the dissociation of aromatic versus aliphatic phosphopeptides by ESI-FTICR-MS and IRMPD. J Am Soc Mass Spectrom 15, 121–127. 82. Pfister SL, Spitzbarth N, Nithipatikom K, Edgemond WS, Falck JR, Campbell WB. (1998) Identification of the 11,14,15- and 11,12, 15-trihydroxyeicosatrienoic acids as endothelium-derived relaxing factors of rabbit aorta. J Biol Chem 273, 30879–30887. 83. Pfister SL, Spitzbarth N, Zeldin DC, Lafite P, Mansuy D, Campbell WB. (2003) Rabbit aorta converts 15-HPETE to trihydroxyeicosatrienoic acids: potential role of cytochrome P450. Arch Biochem Biophys 420, 142–152. 84. Falck JR, Barma DK, Mohapatra S, et al. (2004) Asymmetric synthesis of the stereoisomers of 11,12,15(S)-trihydroxyeicosa5(Z),8(Z),13(E)-trienoic acid, a potent endothelium-derived vasodilator. Bioorg Med Chem Lett 14, 4987–4990. 85. Kerwin JL, Torvik JJ. (1996) Identification of monohydroxy fatty acids by electrospray mass spectrometry and tandem mass spectrometry. Anal Biochem 237, 56–64. 86. Nakamura T, Bratton DL, Murphy RC. (1997) Analysis of epoxyeicosatrienoic and monohydroxyeicosatetraenoic acids esterified to phospholipids in human red blood cells by electrospray tandem mass spectrometry. J Mass Spectrom 32, 888–896. 87. Nakamura T, Henson PM, Murphy RC. (1998) Occurrence of oxidized metabolites of arachidonic acid esterified to phospholipids in murine lung tissue. Anal Biochem 262, 23–32. 88. Wheelan P, Zirrolli JA, Murphy RC. (1993) Low-energy fast atom bombardment tandem mass spectrometry of monohydroxy substituted unsaturated fatty acids. Biol Mass Spectrom 22, 465–473.
HPLC/MS/MS-Based Approaches for Detection and Quantification
89. Harkewicz R, Fahy E, Andreyev A, Dennis EA. (2007) Arachidonate-derived dihomoprostaglandin production observed in endotoxin-stimulated macrophage-like cells. J Biol Chem 282, 2899–2910. 90. Hall LM, Murphy RC. (1998) Electrospray mass spectrometric analysis of 5-hydroperoxy and 5-hydroxyeicosatetraenoic acids generated by lipid peroxidation of red blood cell ghost phospholipids. J Am Soc Mass Spectrom 9, 527–532. 91. Currie LA. (1995) Nomenclature in evaluation of analytical methods including detection and quantification capabilities. Pure and Appl Chem 67, 1699–1723. 92. Revised Assessment of Detection and Quantitation Approaches; 2004. Report No.: EPA821-B-04–005. 93. Assigning Values to Nondetected/Nonquantified Pesticide Residues in Human Health Dietary Exposure Assessments; 2000. Report No.: OPP-00570A 94. Song WL, Lawson JA, Wang M, Zou H, FitzGerald GA. (2007) Noninvasive assessment of the role of cyclooxygenases in cardiovascular health: a detailed HPLC/MS/MS method. Methods Enzymol 433, 51–72.
187
95. Kishi N, Mano N, Asakawa N. (2001) Direct injection method for quantitation of endogenous leukotriene E4 in human urine by liquid chromatography/electrospray ionization tandem mass spectrometry with a column-switching technique. Anal Sci 17, 709–713. 96. Saunders RD, Horrocks LA. (1984) Simultaneous extraction and preparation for highperformance liquid chromatography of prostaglandins and phospholipids. Anal Biochem 143, 71–75. 97. Folch J, Lees M, Sloane Stanley GH. (1957) A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem 226, 497–509. 98. Smith KR, Pinkerton KE, Watanabe T, Pedersen TL, Ma SJ, Hammock BD. (2005) Attenuation of tobacco smoke-induced lung inflammation by treatment with a soluble epoxide hydrolase inhibitor. Proc Natl Acad Sci U S A 102, 2186–2191. 99. Yang J, Schmelzer K, Georgi K, Hammock B. (2009) Quantitative profiling method for oxylipin metabolome by liquid chromatography electrospray ionization tandem mass spectrometry. Anal Chem submitted.
sdfsdf
Chapter 9 Brain Phosphoinositide Extraction, Fractionation, and Analysis by MALDI-TOF MS
1
2 3
Roy A. Johanson and Gerard T. Berry
4
Summary
5
Matrix-assisted laser desorption and ionization time-of-flight (MALDI-TOF) mass spectrometry (MS) can provide rapid, sensitive determinations of lipids from small tissue samples in both single determinations and automated high-throughput assays. MALDI-TOF MS is a sensitive, high-throughput technique for the determination of lipids such as the phosphoinositides, PtdIns (phosphatidylinositol), PIP (phosphatidylinositol-4-phosphate, and PIP2 (phosphatidylinositol-4,5-bisphosphate), but in crude extracts the signals are weak or not observed due in large part to ion suppression by phosphatidylcholine and other cationic lipids. A rapid separation step using a small column of a strong cation exchange (SCX) gel can be utilized easily and effectively to adsorb or capture cationic lipids from chloroform/methanol lipid extracts and provide substantially improved signals for the phosphoinositides. In this review, we describe the use of fractionation of a crude lipid extracts using cation exchange columns in conjunction with MALDI-TOF MS and appropriate internal standards to quantify the levels of phosphoinositides in small mammalian brain samples.
6 7 8 9 10 11 12 13 14 15 16 17
Key words: MALDI-TOF MS, Phosphoinositides
1. Introduction
18
19
The utilization of mass spectrometry (MS) in studies of lipids has been expanding in recent years as instrumentation has improved and new methods of lipid sample preparation and handling have been developed. MS has been used increasingly to characterize changes in lipid composition related to lipid metabolism and lipid-mediated cellular signaling and structural changes (1–3). MALDI-TOF MS can provide rapid, sensitive determinations of lipids from small tissue samples in both single determinations and
Donald Armstrong (ed.), Lipidomics, Methods in Molecular Biology, vol. 579, doi 10.1007/978-1-60761-322-0_9, © Humana Press, a part of Springer Science + Business Media, LLC 2009
189
20 21 22 23 24 25 26 27
190
Johanson and Berry
automated high-throughput assays, and is relatively insensitive to impurities such as buffer salts compared with other MS methods (3–7). Both qualitative and quantitative determinations of lipids can be performed using MALDI-TOF MS (3). When MALDI-TOF MS is used to study crude lipid extracts from biological samples, signals for many of the lipids are weak or not observed. This is due in large part to ion suppression by phosphatidylcholine and other cationic lipids such as lysophosphatidylcholine which has been shown to markedly decrease the sensitivity and reproducibility of signals obtained for the noncationic lipids in a lipid mixture (6, 8). Fractionation of the compounds in crude extracts prior to MALDI-TOF MS analysis (8), such as by thin-layer chromatography (9), is necessary in order to overcome the effect of ion suppression with detection of lipids such as the phosphoinositides in crude lipid extracts from biological samples. The phosphoinositides, phosphatidylinositol (PtdIns), and its phosphorylated derivatives such as phosphatidylinositol-4,5-bisphosphate (PIP2), are important biologically because the polyphosphoinositides play key roles in intracellular signaling and in the localization of proteins to cellular membranes. Advances in analytical methods for phosphoinositides will improve the ability to study the metabolic disruptions and test hypotheses in appropriate biological models. The lipid extraction procedure described here was developed and optimized with the primary goal of determining the levels of the phosphoinositides, PtdIns, PIP, and PIP2 in small samples (10–250 mg) of brain tissue. Consequently, strong acid conditions (1 M HCl) are used in the extraction in order to recover the PIP2 in the organic extract phase (10). We have reported that a rapid separation step using a small column of a strong cation exchange (SCX) gel (RediSep® SCX silica gel) can be utilized easily and effectively to adsorb or capture cationic lipids including phosphatidylcholine species from chloroform/ methanol lipid extracts from mammalian brain (4). In this review we describe the extraction of lipids from tissue followed by fractionation of the crude lipid extracts using cation exchange columns and then MALDI-TOF MS with internal standards for the determination of the levels of PIP2 and other phosphoinositides in brain samples.
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 61 62 63 64 65 66
67
2. Materials
68
2.1. Equipment
69
1. MALDI-TOF MS instrument. 2. Glass Teflon (Potter-Elvehjem) homogenizers.
2.2. Reagents
2.3. Supplies
Brain Phosphoinositided Extraction, Fractionation, and Analysis
191
1. MALDI–MS grade 2,5-dihydroxybenzoic acid (DHB) from Fluka (Switzerland).
70
2. Soybean PtdIns from Avanti Polar Lipids (Alabaster, AL).
72
3. Bovine liver PtdIns from Avanti Polar Lipids (Alabaster, AL).
73
4. Synthetic 1-stearoyl-2-arachidonoyl-phosphatidylinositol-4,5bisphosphate (PIP2) from Avanti Polar Lipids (Alabaster, AL).
74
5. Synthetic dipalmitoyl-PIP2 was obtained from both Sigma (St.Louis, MO) and Cayman Chemical (Ann Arbor, MI).
76 77
6. RediSep® SCX silica gel was from Teledyne Isco (Lincoln, NE).
78
C57BL/6 mice were obtained from Charles River (Wilmington, MA).
79
3. Methods 3.1. Tissue Harvest
3.2. Lipid Extraction 3.2.1. Solvent Preparation
71
75
80
81
The murine brains are harvested as quickly as possible, immediately frozen in liquid N2, and stored and maintained at −80°C until the homogenization is started in order to minimize the appearance of degradation products in the mass spectra (11–15). Adult murine brains were removed and clamp frozen in liquid N2 within 40–45 s after the mouse was sacrificed. The frozen brains were stored at −80°C. Embryonic day 18.5 (E18.5) brains are obtained from fetuses in pregnancies timed by the appearance of vaginal plugs following mating.
82
The extraction procedure is based on methods described by Bligh and Dyer (16) and Schacht (10) and is carried out using an extraction solvent containing a ratio of CHCl3:CH3OH:H2O:conc.HCl such that the final ratio of CHCl3:CH3OH:H2O:conc.HCl, including the H2O in the sample, was 10:20:7:1. Adult brains are estimated to be 80% H2O by weight and fetal brains are estimated to be 85% H2O in calculating how much solvent to use. 1. Extraction solvent CHCl3:CH3OH:HCl:H2O, 10:20:1:3
91
2. Wash solvent CH3OH:H2O:HCl, 20:19:1 3. SCX column solvent CHCl3:CH3OH:H2O, 20:9:1 4. SCX column salt elution CHCl3:CH3OH:500 mM NaCl in H2O, 20:9:1
83 84 85 86 87 88 89 90
92 93 94 95 96 97 98 99 100 101 102 103 104 105 106
192 107 108
Johanson and Berry
3.2.2. Preparation of SCX Column
109 110 111
1. A column containing Teledyne ISCO RediSep® SCX silica gel is packed in a glass–wool plugged 5¾ in. long Pasteur pipette. The column bed is packed to 5 cm total height, including taper (at full bore, 1 ml = approximately 3.7 cm). This column volume is satisfactory for extracts from up to 200 mg wet wt brain tissue.
114
2. Prior to use, the column is washed/regenerated with three column volumes freshly mixed acetonitrile/1 M HCl (100 ml HCl/ml acetonitrile) as recommended by the manufacturer.
115 116
3. The column is equilibrated using five column volumes of the column solvent, CHCl3:CH3OH:H2O, 20:9:1 (17–19).
117
4 After use, the column is washed with CH3OH prior to storage.
112 113
118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152
3.2.3. Extraction of Lipids
1. Aliquots of the internal standards, soybean PtdIns and/or dipalmitoyl PIP2, that are dissolved in CHCl3 and/or CH3OH, are first pipetted into glass–teflon homogenizers (see Note 1) and the solvent evaporated. To prevent degradation of phospholipids (11–15), the homogenizers are then chilled on dry ice. Frozen brain samples are weighed into the chilled homogenizers (see Note 2). Extraction solvent, 6.8 ml CHCl3:CH3OH:H2O:conc HCl (10:20:3:1)/mg wet wt adult brain (7.22 ml/mg wet wt for fetal brain), is added to the samples. The still-frozen murine brain samples are rapidly homogenized as the homogenizer and tissue are warmed by mechanical friction and hand warmth to above freezing temperature. Defrosting and homogenizing the brain tissue works best when the homogenizer capacity is 4–8 times the total volume of the sample plus extraction solvent. Care should be taken to minimize evaporative losses by minimizing exposure to air and transferring to glass centrifuge tubes (see Note 3) and capping for comparison, immediately after homogenization. All subsequent steps are carried out at room temperature. CHCl3 and H2O are added to obtain the separation of phases. First 2.0 ml CHCl3/mg wet wt (2.12 ml/mg wet wt for fetal brain) is added followed by moderate vortex mixing. Then 2.0 ml H2O/ mg wet wt (2.12 ml/mg wet wt for fetal brain) is added followed by mixing. The phases are separated by centrifugation at 100 × g and the lower CHCl3 phase recovered. If the volume of fluffy emulsion that fills more than ~10% of the upper portion of the lower phase, it can generally be reduced by rocking gently to break up the emulsion and recentrifuging to separate the phases (see Note 4). 2. The upper phase and precipitate are reextracted with 2.0 ml CHCl3/mg wet wt (2.12 ml/mg for fetal brain), and the lower phase recovered and combined with the first lower phase (see Note 5). The combined lower phases are then washed 2 times by shaking with 1.5 volumes CH3OH:H2O:HCl, 20:19:1, separating the phases by centrifugation, and recovering the lower phase. If unfractionated extract is to be saved, it should be stored at −80°C.
3.2.4. Lipid Fractionation
3.3. Mass Spectrometry
3.4. Data Analysis
Brain Phosphoinositided Extraction, Fractionation, and Analysis
193
1. Sufficient methanol and H2O is added to the washed chloroform phase lipid extract from the lipid extraction procedure to bring the CHCl3:CH3OH:H2O, ratio to 20:9:1. The sample (0.5–1.5 ml) is passed through a 1.3 ml column of RediSep ® SCX that is packed in a glass–wool plugged Pasteur pipette. Prior to use, the column should be washed/regenerated with freshly mixed acetonitrile/1 M HCl as recommended by the manufacturer, and equilibrated in CHCl3:CH3OH:H2O, 20:9:1 (17–19). The lipid containing flow through fraction is collected.
153
2. The H2O/CH3OH phase is separated from the CHCl3 phase by adding 26% volume H2O:HCl, 10:1.
163
3. The extracted sample should be dried thoroughly under a stream of N2 and stored under N2 at −80°C.
165 166
For an overview of mass spectrometry, see the review by Baldwin (20), and the description on the American Society for Mass Spectrometry website, http://www.asms.org/whatisms/. MALDI-TOF mass spectrometry was carried out using Ettan MALDI-ToF/Pro (GE Healthcare, Piscataway, NJ). The appropriate aliquots of dissolved samples were prepared, dried using N2, and then dissolved at 2.5–15 mg lipid/ml in 0.5 M DHB in CHCl3:CH3OH, 1:1. Aliquots of 0.1 ml were spotted on the target slides. The MALDI-TOF MS measurements were carried out at 20 kV or −20 kV in the reflectron mode using selective accumulation.
167
The S/N was calculated using the equation S/N = (I−(Nmax + Nmin)/2)/(Nmax−Nmin) where I = the maximum intensity of a peak, and Nmax and Nmin are the maximum and minimum of pure noise. Excel (Microsoft, Redmond, WA) and SigmaPlot (Systat Software, Inc, Point Richmond, CA) were used for graphing and linear regression analysis. Since the amounts of sample and internal standard mixed with DHB and loaded on the target spots are operationally maintained within a relatively narrow optimum range, the levels of 16:0–18:2 PtdIns and 18:0–20:4 PtdIns can be readily set up so that, in effect, the range of the samples being assayed will be within the range of the standards. A range of external standard levels was selected for calibration curves that spanned the range of 18:0–20:4 PtdIns levels found in the optimum range of sample levels with MALDITOF MS analysis. Since there is a peak in the mass spectra of the brain extract at m/z 834.5, it is best to set the 16:0–18:2 PtdIns in the soy PtdIns at the high end of the scale to minimize interference by the m/z 834.5 peak. When the ratio of I885/I833 is plotted vs. the increasing level of an external standard, it has been demonstrated that a linear plot is obtained (4).
178
154 155 156 157 158 159 160 161 162 164
168 169 170 171 172 173 174 175 176 177
179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197
194
Johanson and Berry
The first-order regression for the results can be expressed as I885/I833 = m(18:0−20:4 PtdIns)+b where I885 is the intensity at m/z 885 from variable amounts of 18:0–20:4 PtdIns in the external standard or sample, and I833,s is the intensity at m/z 833 for the amount of internal standard, 16:0–18:2 PtdIns, used in running the standard curve (s). For a different amount of internal standard (a) the relationship between the intensity at m/z 833 from a mg 16:0–18:2 PtdIns compared to the intensity at m/z 833 from s mg 16:0–18:2 PtdIns (I833,a) can be expressed as I833,a/ a = I833,s. By rearranging this equation to I833,s=(s/a)I833,a and then substituting into the regression equation and rearranging, Eq. 1 is obtained.
198 199 200 201 202 203 204 205 206 207 208 209
18 : 0 − 20 : 4 PtdIns = a (I 885 / I 883,a ) / ms − b / m.
210 211 212
3.5. Results
(1)
1. Mass spectra of the standards that were used are shown in Fig. 1. Peaks at several m/z were observed in the mass spectra
Fig. 1. Negative mode mass spectra of purified bovine liver PtdIns and soybean PtdIns. Separate samples (40 mg each) in CHCl3 were dried and dissolved in 10 ml of 0.5 M DHB in CHCl3:CH3OH, 1:1. Aliquots of 0.1 ml containing approximately 400 ng PtdIns (440 pmol bovine PtdIns or 470 pmol soybean PtdIns) were spotted on each spot of the target slides. The mass spectra for the bovine PtdIns and the soybean PtdIns are shown in Panel a and b, respectively.
Brain Phosphoinositided Extraction, Fractionation, and Analysis
195
of both the purified bovine PtdIns (Panel A) and purified soybean PtdIns (Panel B). The mass spectra of the bovine PtdIns showed peaks at m/z 836.5, 857.5, 859.5, 861.5, 863.5, 883.5, 885.5, and 887.5 and the averages of the results for the relative intensities of the peaks in five mass spectra were 3, 2, 3, 10, 9, 9, 37, and 26%, respectively. All of these peaks could be accounted for by the analysis of the acyl species provided by the vendor, Avanti Polar Lipids. A similar analysis of soybean PtdIns showed peaks at m/z 831.5, 833.5, 855.5, 857.5, 859.5, 861.5, and 883.5. The 16:0–18:2 PtdIns peak intensity at m/z 833.5 was 46% of the summed intensities of the peaks for soybean PtdIns and the S/N ratio of the peak was 72. For our calibration curve analyses, the external and internal standard masses for the bovine 18:0–20:4 PtdIns and the soybean 16:0–18:2 PtdIns were calculated using the fractions 0.37 and 0.46, respectively, and the manufacturer’s specified concentration, 10 mg/ml. The results obtained based on the specified concentration were consistent, lot to lot, and consistent with published values (4).
213
2. Mass spectra run in the positive ion mode showed that the cation exchange minicolumns effectively removed the cationic lipids including the phosphatidylcholine species from the lipid extracts. Mass spectra of brain extracts run in the negative ion mode are shown for extracts from adult murine brain (Fig. 2) and fetal brain (Fig. 3). Peaks for PIP and PIP2 were barely discernable in the extract from the adult brain prior to the removal of the cationic lipids, and after the cationic lipids were removed, clear peaks were obtained. A similar improvement in the mass spectra was obtained for the extracts from the fetal brain.
233
3. When mass spectra of a dilution series of PtdIns was run using a fixed amount of an added standard of closely similar chemistry and mass, the mass spectra peaks increased linearly with increasing PtdIns (see Fig. 4). When the samples were spiked with a fixed amount of murine brain extract, the curve obtained was displaced upward by the PtdIns in the extract. The curve in the presence of the added brain extract was parallel to that obtained without added extract indicating that the presence of nonphosphoinositide compounds in the extract did not alter the comparative responses observed for the internal standard and the analytes in MALDI-TOF MS.
244
214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 234 235 236 237 238 239 240 241 242 243 245 246 247 248 249 250 251 252 253 254 255
196
Johanson and Berry
Fig. 2. Mass spectra obtained in negative ion mode of lipid extract fractions from adult murine brains. The CHCl3:CH3OH:HCl lipid extracts were prepared and then fractionated using RediSep® SCX silica gel as described in Subheading 3. The internal standards (indicated by IS), 2.00 mg soybean PtdIns and 0.50 mg dipalmitoyl PIP2 per g wet wt brain, were added before the homogenization step for the extracts shown in panel a and b. Dried samples of the extracts were dissolved in 0.5 M DHB in CHCl3:CH3OH, 1:1 and 0.1 ml aliquots were spotted on the target slides. Panel a shows the mass spectra obtained in the negative ion mode of the unfractionated extract, and panel b shows the spectra obtained of the lipids remaining in the extract following passage through the SCX column.
Fig. 3. (continued) of unfractionated extract, and panel b shows the spectra obtained for an extract that had been treated with an SCX column. For the mass spectra shown in Panels c and d, soybean PtdIns containing 141 mg 16:0–18:4 PtdIns and 150 mg synthetic dipalmitoyl PIP2 were added as internal standards to a pool of 4 E18.5 brains with a total wet weight of 290 mg prior to the homogenization and extraction procedures. Panel c shows the mass spectra of a crude extract with internal standards and Panel d shows the SCX treated extract with internal standards. The inset in Panel d shows a portion of the spectra for an extract that was prepared using precautions to minimize contamination by sodium ion by using plastic ware and thoroughly rinsed glassware including all disposable glassware in the extraction procedure. For this extract, two internal standards, soybean PtdIns containing 38 mg 16:0–18:2 PtdIns and 40 mg synthetic dipalmitoyl PIP2, had been added to a 83 mg wet wt E18.5 brain before the homogenization and extraction procedures.
Brain Phosphoinositided Extraction, Fractionation, and Analysis
197
Fig. 3. Negative ion mode mass spectra of lipid extract fractions from E18.5 murine brains. The CHCl3:CH3OH:HCl lipid extracts were prepared and then fractionated using RediSep® SCX silica gel as described in Subheading 3. The dried samples of the extracts were dissolved in 0.5 M DHB in CHCl3:CH3OH, 1:1 and 0.1 ml aliquots were spotted on the target slides. Panel a shows the mass spectra obtained in the negative ion mode
198
Johanson and Berry
Fig. 4. PtdIns calibration curves. Samples containing a constant amount (9.2 mg) of the internal standard, soybean 16:0–18:2 PtdIns, and varying amounts of an external standard, bovine 18:0–20:4 PtdIns, were prepared without and with added SCX treated murine brain extract. The mixed samples were dried, dissolved in 20 ml 0.5 M DHB in CHCl3:CH3OH, 1:1 and 0.1 ml aliquots were spotted on the target slides. The maximum intensities for the mass spectral peaks at m/z 833.5 for 9.2 mg 16:0–18:2 PtdIns and m/z 885.5 for the 18:0–20:4 PtdIns were determined. The plots show the ratios of the signal intensities of the peaks at m/z 885.5 for variable amounts of 18:0–20:4 PtdIns (I885) to the intensities at m/z 833.5 from 9.2 mg 16:0–18:2 PtdIns (I833,9.2 mg) in the soybean PtdIns. The data points shown are the mean ± SD from five determinations. In Panel a, mixtures of the indicated amounts of bovine liver 18.0–20.4 PtdIns with 9.2 mg 16:0–18:2 PtdIns are shown by the filled circles and the results when an aliquot of SCX treated extract from one adult murine brain proportional to that obtained from 7 mg wet wt tissue was added to identical mixtures of bovine PtdIns and soybean PtdIns is shown by the open circles. In Panel b, the filled squares are the results for mixtures of
Brain Phosphoinositided Extraction, Fractionation, and Analysis
199
4. Notes
256
1. Glass Teflon homogenizers are recommended for homogenization because they work well in homogenizing the small, 50–500 mg, frozen tissue samples as they are thawing and also because they produce a minimal amount of aeration with concomitant changes in the solvent composition from the proportionally more rapid evaporation of CHCl3.
257 258
2. Since evidence of enzymatic degradation products of the lipids such as the appearance of lysophosphatidylcholine and increased diacylglycerols were evident in the mass spectra when the tissue was allowed to thaw even briefly before homogenization, the homogenizer should be maintained at dry ice temperatures until the homogenization is started in order to keep the time span to an absolute minimum between when the tissue becomes defrosted and when it is homogenized and solubilized in the acidic organic solvent.
263
3. When mass spectra are to be measured in the positive ion mode, glassware should be used throughout the preparation of the lipid extracts because of known problems with strong organic solvents dissolving or leaching various compounds such as plasticizers from plastic labware that appear as artifact peaks in mass spectra run in the positive ion mode (3). When only negative ion mode mass spectra are to be determined, plastic ware such as polypropylene tubes and “Eppendorf” tips are suitable, as well as other plastic ware that is compatible with CHCl3 keeping in mind that unusual peaks from compounds which come from the plastic may appear randomly.
272
4. When the phases are separated following the initial extraction, there is a tendency for a fluffy emulsion to form in the upper portion of the lower phase especially if the mixing is too vigorous. This fluffy emulsion can fill most of the lower phase. If the layers are gently rocked to disperse the emulsion and recentrifuged, the emulsion will break down and the fluffy layer will become smaller. This may take more than one repeat of the rocking and centrifuging to get this layer to a minimum width band in the top of the lower phase thus maximizing the recovery of the lower phase.
283
5. The upper phase can be recovered for determinations of metabolites such as inositol which partition to this phase.
292
Fig.4. (continued) bovine and soybean PtdIns identical to the samples shown by the filled circles in Panel a, and the open squares show the results when SCX treated extract from 12 mg wet wt E18.5 brain was added to the mixtures. The lines and equations are shown for the first order regressions where I885 is the signal intensity at m/z 885.5 from varying amounts of 18:0–20:4 PtdIns, I833,9.2 mg is the signal intensity at m/z 833.5 from 9.2 mg 16:0–18:2 PtdIns, and PtdIns is mg of 18:0–20:4 PtdIns.
259 260 261 262 264 265 266 267 268 269 270 271 273 274 275 276 277 278 279 280 281 282 284 285 286 287 288 289 290 291 293
200
Johanson and Berry
294
References
295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346
1. Han, X. and Gross, R. W. (2005) Shotgun lipidomics: electrospray ionization mass spectrometric analysis and quantitation of cellular lipidomes directly from crude extracts of biological samples. Mass Spectrom. Rev. 24, 367–412. 2. Han, X. and Cheng, H. (2005) Characterization and direct quantitation of cerebroside molecular species from lipid extracts by shotgun lipidomics. J. Lipid Res. 46, 163–175. 3. Schiller, J., Suss, R., Arnhold, J., Fuchs, B., Lessig, J., Muller, M., Petkovic, M., Spalteholz, H., Zschornig, O., and Arnold, K. (2004) Matrix-assisted laser desorption and ionization time-of-flight (MALDI-TOF) mass spectrometry in lipid and phospholipid research. Prog. Lipid Res. 43, 449–488. 4. Johanson, R. A., Buccafusca, R., Quong, J. N., Shaw, M. A., and Berry, G. T. (2007) Phosphatidylcholine removal from brain lipid extracts expands lipid detection and enhances phosphoinositide quantification by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry. Anal. Biochem. 362, 155–167. 5. Milne, S., Ivanova, P., Forrester, J., and Alex, Brown, H. (2006) Lipidomics: an analysis of cellular lipids by ESI-MS. Methods 39, 92–103. 6. Petkovic, M., Schiller, J., Muller, M., Benard, S., Reichl, S., Arnold, K., and Arnhold, J. (2001) Detection of individual phospholipids in lipid mixtures by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry: phosphatidylcholine prevents the detection of further species. Anal. Biochem. 289, 202–216. 7. Schiller, J., Arnhold, J., Benard, S., Muller, M., Reichl, S., and Arnold, K. (1999) Lipid analysis by matrix-assisted laser desorption and ionization mass spectrometry: A methodological approach. Anal. Biochem. 267, 46–56. 8. Muller, M., Schiller, J., Petkovic, M., Oehrl, W., Heinze, R., Wetzker, R., Arnold, K., and Arnhold, J. (2001) Limits for the detection of (poly-)phosphoinositides by matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF MS). Chem. Phys. Lipids 110, 151–164. 9. Schiller, J., Suss, R., Fuchs, B., Muller, M., Zschornig, O., and Arnold, K. (2003)
Combined application of TLC and matrixassisted laser desorption and ionization timeof-flight mass spectrometry (MALDI-TOF MS) to phospholipid analysis of brain. Chromatographia 57, S297–S302. 10. Schacht, J. (1981) Extraction and purification of polyphosphoinositides. Methods Enzymol. 72, 626–631. 11. Arthur, G. and Sheltawy, A. (1980) The presence of lysophosphatidylcholine in chromaffin granules. Biochem. J. 191, 523–532. 12. Christie, W. W. (1993) (Ed.), Advances in Lipid Methodology – Two, Oily Press, Dundee, 1993, pp. 195–213. (available online at http:// www.lipidlibrar y.co.uk/topics/extract2/ index.htm) 13. Hauser, G., Eichberg, J., and Gonzalez-Sastre, F. (1971) Regional distribution of polyphosphoinositides in rat brain. Biochim. Biophys. Acta 248, 87–95. 14. Hauser, G. and Eichberg, J. (1973) Improved conditions for the preservation and extraction of polyphosphoinositides. Biochim. Biophys. Acta 326, 201–209. 15. Nishihara, M. and Keenan, R. W. (1985) Inositol phospholipid levels of rat forebrain obtained by freeze-blowing method. Biochim. Biophys. Acta 835, 415–418. 16. Bligh, E. G. and Dyer, W. J. (1959) A rapid method for total lipid extraction and purification. Can. J. Biochem. Physiol. 37, 911–917. 17. Hendrickson, H. S. and Ballou, C. E. (1964) Ion exchange chromatography of intact brain phosphoinositides on diethylaminoethyl cellulose by gradient salt elution in a mixed solvent system. J. Biol. Chem. 239, 1369–1373. 18. Kiselev, G. V. (1982) Preparative isolation of polyphosphoinositide fractions from ox brain. Biochim. Biophys. Acta 712, 719–721. 19. Low, M. G. (1990) Purification of Phosphatidylinositol 4-Phosphate and Phosphatidylinositol 4,5-Bisphosphate by Column Chromatography. In: Methods in Inositide Research, (Irvine, R. F. ed.) Raven Press, New York, pp.145–151. 20. Baldwin, M. A. (2005) Mass spectrometers for the analysis of biomolecules. Methods Enzymol. 402, 3–48.
Chapter 10 Lipidomic Analysis of Biological Samples by Liquid Chromatography Coupled to Mass Spectrometry
1
2 3
Giuseppe Astarita, Faizy Ahmed, and Daniele Piomelli
4
Summary
5
Lipidomics studies the large-scale changes in nonwater-soluble metabolites (lipids) accompanying perturbations of biological systems. Because lipids are involved in crucial biological mechanisms, there is a growing scientific interest in using lipidomic approaches to understand the regulation of the lipid metabolism in all eukaryotic and prokaryotic organisms. Lipidomics is a powerful tool in system biology that can be used together with genomics, transcriptomics, and proteomics to answer biological questions arising from various scientific areas such as environmental sciences, pharmacology, nutrition, biophysics, cell biology, physiology, pathology, and disease diagnostics. One of the main challenges for lipidomic analysis is the range of concentrations and chemical complexity of different lipid species. In this chapter, we present a lipidomic approach that combines sample preparation, chromatographic, and intrasource ionization separation coupled to mass spectrometry for analyzing a broad-range of lipid molecules in biological samples.
6 7 8 9 10 11 12 13 14 15 16
Key words: Lipidomics, Lipids, Liquid chromatography mass spectrometry, Fatty acids, Phospholipids, Cholesterol, Lipid profile, Lipid biomarkers, Large-scale analysis
17 18
1. Introduction
19
Lipids are natural molecules that are insoluble or partially soluble in water. These hydrophobic or amphipathic molecules can be either biosynthesized or absorbed from the environment and are vital for the life of all eukaryotic and prokaryotic organisms. Lipids play crucial biological roles through three general mechanisms (1) they affect the cellular membrane structures and protein–membrane interactions, (2) they provide a source of energy through processes of oxidation, and (3) they serve as signaling molecules, binding to
Donald Armstrong (ed.), Lipidomics, Methods in Molecular Biology, vol. 579, doi 10.1007/978-1-60761-322-0_10, © Humana Press, a part of Springer Science + Business Media, LLC 2009
201
20 21 22 23 24 25 26 27
202
Astarita, Ahmed, and Piomelli
plasma membrane or nuclear receptors mediating transmembrane signaling and cell-to-cell communication (1). The development of mass spectro-metry (MS) techniques marked the beginning of a new era for the study of lipids, opening a series of unprecedented experimental opportunities. Indeed, the implementation of atmospheric-pressure ionization techniques such as electrospray ionization (ESI) and atmospheric pressure chemical ionization (APCI), capable of coupling liquid chromatography (LC) with MS, made it possible to separate and analyze even the most hydrophobic lipids with much greater accuracy than ever before possible. Such technological advances have contributed to the advancement of lipidomics, the discipline that studies the largescale changes in lipid composition accompanying perturbations of biological systems (see Note 1). The ultimate goal of lipidomics is to understand the role of lipids in the biology of living organisms. It represents a rapidly evolving tool in system biology, which integrates multidisciplinary sets of data derived from molecular-profiling techniques such as genomics, transcriptomics, and proteomics. Therefore, there is a growing scientific interest in using lipidomics to answer various biological questions, arising from living organisms with all degree of biological complexity, such as animals, plants, fungi, protists, bacteria, archaea, and viruses. Lipidomic approaches can be used to investigate the following main research areas:
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
1.1. Ecophysiology
Lipidomics can be applied to study the impact of both abiotic environmental factors (e.g., climate, radiation, toxins, gravity, CO2 and oxygen levels, insolation, light-dark cycle, habitat) and biotic environmental factors (e.g., plants, animals, pathogens, and micro-organisms) on lipid metabolism (2–5).
1.2. Nutrition
Lipidomics is applied to study the effect of nutrients (e.g., carbohydrates, fats, proteins, vitamins, minerals, water, and beverages), nutraceuticals (e.g., antioxidants, fibers, omega-3 fats), food additives and fertilizers on lipid metabolism (6, 7).
1.3. Pharmacology and Toxicology
Lipidomics can be utilized to study the effects of pharmacological treatments (e.g., medications, vaccinations) and other synthetic products (e.g., cosmetics, contaminants, drugs of abuse, chemicals) on lipid metabolism (9, 10).
1.4. Genetics, Transcriptomics, and Proteomics
Lipidomic approaches can be used to study the effects of genetic diversity (e.g., genotypes, epigenetic regulation, mutations, and polymorphism), messenger RNA expression profiles and protein diversity (e.g., isoforms, post-translational modifications, cofactors) on lipid metabolism. Furthermore, lipidomics can be applied to investigate the biological functions of genes and proteins by
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
Lipidomic Analysis of Biological Samples by Liquid Chromatography
1.5. Biophysics
1.6. Cell Biology
1.7. Biochemistry and Molecular Biology
1.8. Physiology and Psychology
1.9. Pathology and Disease Diagnostics
203
studying the lipid profiles associated with genetic manipulation (i.e., gene overexpression or knock down) in biological systems. Indeed, fluctuations in lipid composition can be used to uncover alterations in the transcriptome and proteome (11, 12).
72
Lipidomic approaches can be used to investigate the effects of lipid composition on biophysical parameters (e.g., fluidity, compressibility), the biological functions of membrane structures (e.g., lipid rafts), as well as lipid–protein and lipid–nucleic acid interactions (13).
76
Lipidomics can study the role of lipid metabolism in critical cellular processes (e.g., cell cycles, survival/death, morphology, organelles) together with the circadian regulation of biological processes (e.g., development, aging, hormone production, hunger, thirsty, sleep) (14–17).
81
Lipidomic approaches can be used to understand the biochemical mechanisms for the biosynthesis and the metabolism of lipids in living organisms. In this area of research, lipidomics may lead to the discovery of novel lipid molecular species and lipid-related biochemical pathways (e.g., enzymes, proteins, receptors, and genes) (18, 19).
86
Lipidomics can be applied to understand the physiological roles played by endogenous lipids in crucial biological processes (e.g., learning and memory, immune response, pain, and inflammation) (20, 21). Also, lipidomic approaches can unveil the role of lipids in the sensory perception (chemoreception, photoreception, mechanoreception, and thermoreception), in mental processes and behavior (e.g., cognition, emotion, personality, social, and sexual behaviors), and physical activity (e.g., exercise) (22).
92 93
Lipidomic strategies can be used to investigate the role of lipid metabolism in the pathology of plant and animal diseases. Indeed, epidemiological studies revealed that many human diseases are characterized by specific alterations in lipid metabolism (e.g., cancer, obesity, diabetes, insomnia, depression, stress, trauma, dementia, as well as infectious, cardiovascular, and neurodegenerative diseases) (23, 24). Therefore, lipidomics can be used to profile lipid composition of biological samples for disease diagnosis and drug discovery. In fact, lipid composition can provide a “snapshot” of the biological state of an organism and, consequently, be considerate as an index (biomarker) of healthy or diseased state (25, 26). Furthermore, such lipid biomarkers can serve also as indicators of pharmacologic responses to a therapeutic intervention (24).
100
73 74 75
77 78 79 80
82 83 84 85
87 88 89 90 91
94 95 96 97 98 99
101 102 103 104 105 106 107 108 109 110 111 112 113
204
Astarita, Ahmed, and Piomelli
One of the main challenges of lipidomic analysis is the range of concentrations and chemical complexity of lipid compounds in biological samples (27) (Scheme 1). In fact, a comprehensive lipidomic analysis is expected to take into consideration “structural lipids” (e.g., phospholipids, which serve both as building blocks of the cell membranes and as precursors for signaling lipids), “storage lipids” (e.g., triacylglycerols, which are hydrolyzed to produce either energy or signaling lipids) and the less abundant, but equally important, “signaling lipids” (e.g., fatty acids and their derivatives) (Scheme 1). Therefore, there is a need to develop analytical approaches that allow for the comprehensive analysis of structural, storage, and signaling lipids. In this chapter, we present current methodologies utilized in our laboratory for lipidomic analysis of biological samples. We describe in some detail an analytical approach that combines sample preparation, chromatographic, and intrasource ionization separation coupled to mass spectrometry for analyzing the lipid composition of cells, biological fluids and tissues (see Note 2).
114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132
2. Materials 133
2.1. Equipment
1. Analytical balance.
134
2. Chemical fume-hood.
135
3. Homogenizer.
136
4. Vortex.
137
5. Centrifuge.
138 139
6. Pierce Reacti-Therm III Heating/Stirring Module Thermo Fisher Scientific (Somerset, NJ, USA).
140
7. Spectrophotometer for protein measurement.
141
8. Agilent 1200-LC system (with autosampler) coupled to IonTrap XCT or single quadrupole 1946D MS detectors and interfaced with ESI or APCI (Agilent Technologies).
142 143 144
9. Gas: ultra-high purity compressed helium (for MS fragmentation) and high-purity N2 (for drying samples and for atmospheric pressure ionization functioning).
145 146 147 148 149 150 151 152 153
2.2. Reagents
A representative list of internal standards may include the following lipids. 1. Fatty acyls Fatty acid: heptadecanoic acid from Nu-Chek Prep (Elysian, MN, USA); d8-arachidonic acid from Cayman Chemicals (Ann Arbor, MI, USA);
Lipidomic Analysis of Biological Samples by Liquid Chromatography
Scheme 1. Lipid classes. Chemical classification of lipids. (a) Fatty acyls are fatty acids and their derivatives (oxygenated, amides, esters). (b) Glycerolipids are fatty acid esters of glycerol and comprise mono-, di-, and tri-acylglycerols. (c) Glycerophospholipids contain phosphoric acid in ester form with a glycerolipid. (d) Sphingolipids contain a common sphingoid base moiety. (e) Sterol lipids contain a fused four-ring core (27).
205
206
Astarita, Ahmed, and Piomelli
Prostaglandin: d4-prostaglandin E2 from Cayman Chemicals (Ann Arbor, MI, USA). Fatty-acid ethanolamide: heptadecenoylethanolamide (synthesized as previously reported, see (28)).
154 155 156 157
2. Glycerolipids Triacylglycerol: Trinonadecenoin from Nu-Chek Prep; Diacylglycerol: dinonadienoyl-sn-glycerol from Nu-Chek Prep; Monoacylglycerol: monoheptadecanoyl-sn-glycerol from Nu-Chek Prep; d8-2-arachidonoyl-sn-glycerol from Cayman Chemicals.
158 159 160 161 162 163
3. Glycerophospholipids Phosphatidylethanolamine: 1,2-diheptadecanoyl-sn-glycero3-phosphoethanolamine from Avanti polar Lipids; Phosphatidylglycerol: 1,2-diheptadecanoyl-sn-glycero-3phosphoglycerol from Avanti Polar Lipids; Phosphatidylcholine: 1,2-diheptadecanoyl-sn-glycero-3phosphocholine from Avanti Polar Lipids; Phosphatidylserine: 1,2-diheptadecanoyl-sn-glycero-3-phosphoserine from Avanti Polar Lipids; Phosphatidylinositol: 1,2-dipalmitoyl-sn-glycero-3-phosphoinositol from Avanti Polar Lipids.
164 165 166 167 168 169 170 171 172 173 174
177 178
4. Sphingolipids Ceramide: N-Lauroyl-ceramide from Avanti Polar Lipids; Sphyngomyelin: N-Lauroyl-sphingomyelin from Avanti Polar Lipids.
179
5. Sterol lipids
180
Cholesterol: d7-cholesterol from Avanti Polar Lipids.
181
6. Solvents and chemicals Water, methanol, chloroform (HPLC grade) are purchased from Thermo Fisher Scientific (Somerset, NJ, USA). Acetic acid and ammonium acetate are from Sigma (Saint Louis, Missouri, USA).
175 176
182 183 184 185 186
2.3. Supplies
1. LC columns.
188
2. Glass Vials (8 ml, 40 ml, 1.5 ml for autosampler and LC analysis).
189
3. Glass pipettes (5 ml, 10 ml).
190
4. Glass Pasteur pipettes.
191
5. Caps with Teflon-liner.
192
6. Conical insert for reducing the volume of the autosampler vials.
193
7. Vial racks.
194
8. Vial trays.
195
9. Dry ice.
187
Lipidomic Analysis of Biological Samples by Liquid Chromatography
207
3. Methods 3.1. Sample Preparation for Lipidomics Analysis
3.1.1. Cells
196
Sample preparation includes the extraction of lipids from the biological matrix and the removal of any nonlipid contaminants from the extract (see Note 3). We use a modified Folch procedure for lipid preparation of biological samples from cells, biological fluids, and tissues (see Notes 4 and 5) (Scheme 2). Lipid molecular species are quantified by normalizing the individual molecular ion peak intensity with an internal standard for each lipid class. Therefore, a mixture of nonendogenous lipid species used as internal standards for each lipid class before the extraction process (see Subheading 2.2). These internal standards allow the lipid levels to be normalized for both extraction efficiency and instrument response.
197
1. Label 8-ml glass vials based on the number of tissue samples to analyze.
209 210
2. Wash cells with phosphate-buffered saline (PBS, 1×), remove all PBS.
211
3. Add 1 ml of methanol containing the internal standards to each well keeping the plate on ice.
213 214
4. Scrape and collect the cells in 8-ml glass vials (see Note 6).
215
5. Sonicate in ice for 10 s (5 pulses (×2) at 200 V).
216
6. Save 20 ml aliquot for protein measurements. Protein concentration is measured using the Bradford protein concentration assay (Bio-Rad Laboratories Inc., Hercules, CA) or the BCA protein assay (Pierce, Rockford, IL).
217
7. Add 2-ml chloroform and vortex for 10 s.
221
8. Wash with 0.75 ml of water (or better 0.7% KCl solution) and vortex for 10 s.
222 223
9. Centrifuge 1,000 × g for 15 min at 4°C to separate the mixture into two phases with a protein disk at their interface. The lower phase is mainly chloroform and contains most of the lipids; the upper phase is methanol and water containing more polar metabolites.
224
10. Take the organic (bottom) layer using a glass Pasteur pipette and transfer into another 8-ml glass vial. Discard the protein disk and the upper (aqueous) phase.
229
11. Dry down using a gentle N2 steam.
232
12. Resuspend in 50–100 ml chloroform/methanol (1:3, vol:vol).
233
13. Transfer into 1.5-ml glass vials with the 0.2-ml conical inserts and proceed to LC/MS analysis.
234
14. Normalize lipids for mg protein (mol/mg protein).
236
198 199 200 201 202 203 204 205 206 207 208
212
218 219 220
225 226 227 228 230 231
235
208
Astarita, Ahmed, and Piomelli
Scheme 2. Lipidomic approach. Flow chart of the strategy used for a broad-range analysis of lipids from biological sample. MAG monoacylglycerol, FAE fatty acid ethanolamide, LPC lysophosphatidylcholine, FA fatty acid, oxFA oxygenated fatty acids, DAG diacylglycerol, TAG triacylglycerol, SP sphingolipid, PC phosphatidylcholine, PE phosphatidylethanolamine PS phosphatidylserine, PI phosphatidylinoitol; PG phosphatidylglycerol; PA phosphatidic acid.
239
Tissues are rapidly collected and snap-frozen in liquid N2. 1. Label 8-ml glass vials according to the number of tissue samples to analyze.
240 241
2. Add 1 ml of methanol containing the internal standards in each vial, while keeping the vials in ice.
242
3. Weigh the frozen tissues (10–100 mg) (see Note 6) and transfer them into the previously prepared vials containing methanol with internal standards.
237 238
243 244
3.1.2. Tissues
246
4. Homogenize the tissues keeping the vials in an ice bath and collect 20 ml aliquots for protein measurements (see Note 7).
247
5. Add 2 ml of chloroform and vortex for 10 s.
248 249
6. Wash with 0.75 ml of water (or better 0.7% KCl solution) and vortex for 10 s (see Note 8).
250
7. Centrifuge at 1,000 × g for 10 min at 4°C to the mixture into two phases with a protein disk at their interface. The lower phase is mainly chloroform and contains most of the lipids; the upper phase is methanol and water containing more polar metabolites (see Note 9).
245
251 252 253 254
Lipidomic Analysis of Biological Samples by Liquid Chromatography
3.1.3. Biological Fluids Plasma and Serum
Cerebrospinal Fluid (CSF)
209
8. Prepare another set of 8-ml glass vials using the same labeling system as described before.
255
9. After centrifugation, collect the lower (organic) phase using a glass Pasteur pipette.
257
10. Re-extract the protein disk and the upper (aqueous/methanol) phase with 2-ml of chloroform.
259 260
11. Centrifuge and add together the two organic phases.
261
12. Evaporate the solvent to dryness in the vials using a gentle N2 stream.
262
13. Resuspend in 50–100 ml chloroform/methanol (1:3, vol:vol).
264 265
14. Transfer the resuspended lipid solution to the 1.5-ml vials with the 0.2-ml conical inserts and proceed to LC/MS analysis.
266
15. Normalize lipid amount per grams of tissue (mol/g) or per mg protein (mol/mg protein).
268 269
For plasma preparation, blood is centrifuged in EDTA-containing tubes at 1,000 × g for 10 min at 4°C, and the top layer (plasma) is recovered using a glass Pasteur pipette. For serum preparation, blood is immediately centrifuged in glass tubes at 1,000 × g for 10 min at room temperature and the top layer (serum) is recovered using a glass Pasteur pipette.
270
CSF samples are checked for blood contamination by measuring the total cell count, total protein, CSF/serum albumin and IgG quotients, and determination of oligoclonal bands by isoelectric focusing and silver staining. Normal cell counts, normal CSF/ serum albumin ratios, and no oligoclonal bands indicate healthy blood–brain barrier function and lack of intrathecal immunoglobulin G synthesis. 1. Label 8-ml glass vials according to the number of tissue samples to analyze.
276
2. Transfer 0.2 ml of plasma/serum/CSF samples into the 8-ml vial in ice.
285
3. Add three volumes of ice-cold acetone containing internal standards.
287 288
4. Vortex for 10 s.
289
5. Shake and refrigerate sample for 30 min.
290
6. Centrifuge at 1,000 × g for 10 min at 4°C to pellet out the precipitated proteins.
291
7. Take the supernatant and evaporate the excess acetone under N2 stream.
293 294
8. Add 1-ml of methanol and vortex for 10 s.
295
9. Add 2-ml of chloroform and vortex for 10 s.
296
256 258
263
267
271 272 273 274 275 277 278 279 280 281 282 283 284 286
292
210
Astarita, Ahmed, and Piomelli
298
10. Wash with 0.8 ml of water (or better 0.7% KCl) and vortex for 10 s.
299
11. 11 Centrifuge at 1,000 × g for 10 min at 4°C.
300 301
12. Collect the lower phase with a glass Pasteur pipette and transfer to a clean 8-ml glass vial.
302
13. Evaporate the elnates to dryness under N2 stream.
303 304
14. Resuspend in 50–80 ml of a solution chloroform/methanol (1:3, vol:vol).
305
15. Prepare 1.5-ml vials with conical glass inserts.
306 307
16. Transfer the resuspended lipid solution to the 1.5-ml vials with 0.2-ml conical insert and proceed to LC/MS analysis.
308
17. Normalize lipid amount per ml of biological fluid (mol/ml).
297
318 319
Blood can be fractionated in plasma and blood cells, which is made of white blood cells (WBCs) and red blood cells (RBCs). Blood cells are normally discarded when collecting the plasma. However, the same procedure used for plasma collection, also allows the recovery of the buffy coat (mainly WBCs) and the RBCs, which can be used to measure biomarkers for dietary fat (29) and diseases (30). 1. Fractionate whole blood samples by centrifuging in EDTA at 1,000 × g for 10 min at 4°C. This will separate the blood into an upper plasma layer, a lower RBCs layer, and a thin interface (buffy coat) containing the WBCs.
320
2. Recover the plasma.
321
3. Recover the WBCs, wash with PBS (1×) three times and centrifuge discarding the supernatant; freeze in distilled water (1:1, vol:vol).
309
Blood
310 311 312 313 314 315 316 317
322 323
4. Recover the RBCs, wash with PBS (1×) three times and centrifuge discarding the supernatant; freeze in distilled water (1:1, vol:vol).
324 325 326 327 328 329
3.2. LC/MS Analysis of Lipids
Mobile phase A is methanol containing 0.25% acetic acid and 5 mM ammonium acetate; mobile phase B is water containing 0.25% acetic acid and 5 mM ammonium acetate. Lipids are identified based on their retention times and MSn properties.
3.2.1. Small Lipids Analysis
Small lipid molecules are separated using a reversed-phase Zorbax XDB Eclipse C-18 column (50 × 4.6 mm i.d., 1.8 mm particle size, 80 Å of porous diameter, Agilent Technologies). Detection and analysis is controlled by Agilent Chemstation and Bruker Daltonics software. 1. Fatty acyls Lipids are eluted using a linear gradient from 90% A to 100% B in 2.5 min at a flow rate of 1.5 ml/min with column
330 331 332 333 334 335 336 337 338
Lipidomic Analysis of Biological Samples by Liquid Chromatography
211
Fig. 1. Analysis of small lipids by LC/MS. Representative LC/MS chromatograms of fatty acids extracted from biological samples.
3.2.2. Large Lipids Analysis
t emperature at 40°C. ESI is in the negative mode, capillary voltage is set at −4 kV and fragmentor voltage is 100 V. N2 is used as drying gas at a flow rate of 13 l/min and a temperature of 350°C. Nebulizer pressure is set at 60 psi. We use commercially available fatty acyls as reference standards. They are analyzed monitoring the mass-to-charge ratio (m/z) of the deprotonated molecular ions [M − H]− in the selected-ion monitoring mode (Fig. 1).
339
Large lipid molecules are separated using a reversed-phase Poroshell 300SB C-18 column (2.1 × 75 mm i.d., coating layer of 0.25 mm on total particle diameter of 5 mm, 300 Å of porous diameter, Agilent Technologies). Lipids are identified based on their retention times and MSn properties. Detection and analysis is controlled by Agilent Chemstation and Bruker Daltonics software. 1. Glycerolipids, glycerophospholipids, sphingolipids A linear gradient is applied from 85% A to 100% B in 5 min at a flow rate of 1.0 ml/min with column temperature set at 50°C.
346 347
340 341 342 343 344 345
348 349 350 351 352 353 354 355
212
Astarita, Ahmed, and Piomelli
The capillary voltage is set at 4.0 kV and skimmer voltage at 40 V. N2 is used as drying gas at a flow rate of 10 l/min, temperature at 350°C and nebulizer pressure at 60 psi. Helium is used as collision gas, and fragmentation amplitude is set at 1.2 V. MS detection is both in the positive and in the negative ionization modes. Ion charge control is on, smart target set at 50,000 and max accumulation time at 50 ms, scan range of 100–1,500 amu, 26,000 m/z per second.
356 357 358 359 360 361 362 363 364 365 366
3.2.3. Sterol Lipids Analysis
Lipids are separated using a linear gradient from 75% A to 100% B in 4-min period at a flow rate of 1.0 ml/min with column temperature at 50°C. APCI is set in positive mode. Drying gas is set at 350°C and a flow of 8 l/min. Nebulizer gas pressure is set at 30 psi and vaporizer temperature at 475°C. Capillary voltage is set at 300 V with the corona current set at 5 mA.
3.3. Results: Chromatographic and Intrasource Ionization Separation of Lipid Molecules
Lipids exist in nature in a wide variety of chemical complexities and dynamic range of concentrations (Scheme 1). In order to simplify the analysis in biological tissues, lipids are schematically divided into three main classes (a) small lipids, defined here as molecules containing one aliphatic group such as fatty acids and their derivatives (amides, esters, oxygenated compounds); (b) large lipids, molecules containing two or more aliphatic groups, such as phospholipids, diacylglycerols, triacylglycerols, sphingolipids; and (c) sterol lipids, molecules containing a rigid four-ring backbone such as cholesterol and its derivatives. Therefore, in order to analyze the different classes of lipids by LC/ MS, two separate chromatographic approaches are applied, using different reversed-phase C-18 stationary phases (Scheme 2). Furthermore, because lipid classes with different functional groups have characteristic ionization efficiencies, a combination of ESI set in either positive or negative mode, and APCI set in positive mode is used (Scheme 2).
3.3.1. Small Lipids Analysis
To separate lipids containing one fatty acyl group, a reversed-phase C-18 column packed with conventional porous silica particles of small spherical diameter (sub-2 mm) is used. Fatty acyl species are separated both by chain length and by degree of unsaturation of their fatty acid chains. For example, fatty acids containing shorter or more unsaturated acyl chains elute earlier than those with longer and more saturated chains (Fig. 1). Generally, in positive ESI mode small lipids are detected as protonated molecular ions or sodium and ammonium adducts. In contrast, in negative mode small lipids are detected as deprotonated molecular ions (Fig. 1).
3.3.2. Large Lipids and Sterol Lipids Analysis
To separate large and sterol lipids, a reversed-phase C-18 column packed with superficially porous particles (Poroshell,
367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399
Lipidomic Analysis of Biological Samples by Liquid Chromatography
213
Agilent-Technologies, coating layer of 0.25 mm on total particle diameter of 5 mm) is used (18, 31). This allows for fast flow rates and good peak shapes (Fig. 2). Usually, because of diffusion limits in totally porous silica, large lipid molecules give tailing peaks at high flow rates. However, superficially coated columns allow for faster diffusion at the surface, allowing high flow rates and good peak shape. Indeed, the thin shell allows the slowly diffusing hydrophobic macromolecules and the rigid structures of sterol lipids to rapidly penetrate the superficial packing material (since the solid core prevents further diffusion). To decrease the retention times, a high flow velocity is applied. To decrease mobile phase viscosity and avoid exceeding the column back-pressure limits, a relatively high column temperature is used. A combination of high temperature and high flow velocities improves the separation speed, resulting in better peak shape of the lipid analytes. Notably, lipids are stable at high temperatures using high flow rates (32). Although lipids are separated when differing in a single fatty acyl chain, their combinatorial nature makes only a partial separation of the isomeric species possible (Fig. 2). Therefore, to obtain more information on the lipid structure, LC separation is coupled with MSn fragmentation data (Fig. 3). Generally, large lipids are detected in the positive ESI mode as sodium or ammonium adducts or as deprotonated molecular ions in the negative mode. For sterol lipids, which are highly hydrophobic and hard to ionize, APCI is used in the positive mode and the protonated molecular ions are detected after loss of water.
Fig. 2. Analysis of large lipids by LC/MS. Representative LC/MS chromatograms of phosphatidylethanolamines (PE) extracted from biological samples.
400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426
Fig. 3. Analysis of large lipids by MS/MS. Representative MS/MS spectrum of a selected phosphatidylethanolamine (1-stearoyl,2-docosahexaenoyl-sn-glycero-3-phosphoethanolamine) derived from biological samples.
214 Astarita, Ahmed, and Piomelli
Lipidomic Analysis of Biological Samples by Liquid Chromatography
215
3.4. Maintenance and System Suitability Test for LC/MS Analysis
In the following sections we report a general procedure for the maintenance and system suitability testing used to validate the LC/MS lipidomic analysis.
427
3.4.1. LC/MS Maintenance
To avoid contaminations, routinely preventive maintenance is performed. 1. Replacing the spray needle and electron multiplier; cleaning ionization spray chamber or other accessible MS components.
430 431
2. Replacing inline filters and frits, the injector needle and capillaries or other accessible LC components; flushing the system with a mixture of cyclohexane/acetonitrile/isopropanol (1/1/2, v/v/v).
434
1. To check for the LC column status, assure that the column has a constant backpressure, which usually is a guarantee of good column performance. Increased pressure indicates column contamination or fouling.
438 439
2. To check for contaminations, blank samples are run before and between biological samples.
442
3. To check for accuracy of quantification, quality control samples are run at the end of the run (three concentrations that are representative of the concentration range of the analyte of interest).
444 445
4. To check for linearity of the detection response, calibration curves are run before running the samples.
448
5. To avoid sample cross-contaminations, the injector needle is washed automatically between each sample injection.
450 451
Lipid extracts are generally stored in a freezer at –80°C. They are solubilized in chloroform–methanol solutions using glass vials closed with Teflon-lined caps and secured with Parafilm. To prevent oxidation, air is removed by flushing the vials or tubes with N2 before closing them. It was shown that after storage up to 4 years at –80°C, the blood lipid composition is practically unchanged (33). If storage is brief, lipids can be stored at –20°C.
452
Contaminants can be detected as extra-peaks or high background noise in LC/MS chromatograms. They strongly affect the specificity and sensitivity of our analysis. During sample preparation, common sources of contamination are mineral oils, grease, detergents, and plasticizers from plastics, including lipid molecules such as oleamide (34). Plastic pipettes, tips, beakers, and vials can leach contaminants into organic solutions. Therefore, all operations are generally carried out in glass and all vials or tubes are closed with screw caps including a Teflon-covered liner. Furthermore, all
460 461
3.4.2. Quality Assurance
3.4.3. Storage of Lipid Extracts
3.4.4. Contaminations
428 429
432 433 435 436 437
440 441 443
446 447 449
453 454 455 456 457 458 459
462 463 464 465 466 467 468
216
Astarita, Ahmed, and Piomelli
operators must wear gloves during the procedures to prevent any contaminations by skin surface lipids. Change gloves frequently and keep vials closed or covered with aluminum foil.
469 470 471
472
473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495
4. Notes 1. Because lipids are a set of small-molecule metabolites, lipidomics is considered to be part of metabolomics, which is the large-scale study of all metabolites (both water-soluble and water-unsoluble) in biological organisms. The distinction originated as consequence of the metabolome (complete set of small-molecule metabolite) complexity, which required the development of analytical approaches specific for nonwater soluble metabolites (lipids) (35). 2. The described fast lipidomic approach is suitable for the determination of a broad-range of lipid alterations occurring in biological samples. The combination of the chromatographic resolving power in conjunction with the ionization source selection and the mass detection can be used to analyze even the lipids present at very low concentration. In contrast, the direct infusion of the lipid extract into the MS detector is subject to ionization suppression effects and loss of sensitivity and accuracy. Furthermore, because lipids may differ in mass by only two units, a partial chromatographic separation helps avoid the isotopic effects, which affect the actual mass abundance (36). 3. Particular attention should be given to sample preparation: It is worth remembering that there is no good LC/MS analysis without a good sample preparation.
507
4. Alternative extraction procedures that use less toxic organic solvents such as methyl-tert-butyl ether (37), hexane–isopropanol, and ethyl acetate/ethanol mixtures have been proposed for a wide range of tissues (38, 39). Surprisingly, it is not always made clear in the laboratory environment that methanol and chloroform are toxic and potentially carcinogenic (38, 40, 41). Furthermore, the methanol/ chloroform mixture is extremely irritating to skin and eyes. Therefore, it is particularly important to train students and new laboratory personnel to handle organic solvents with gloves in a chemical fume-hood, avoiding health-hazard by accidental spills, skin contact and breathing of vapors.
508 509
5. For the recovery of acidic phospholipids such as gangliosides and phosphoinositides, alternative extraction methods have
496 497 498 499 500 501 502 503 504 505 506
Lipidomic Analysis of Biological Samples by Liquid Chromatography
217
been suggested which use strong HCl solutions instead of water during the washing step of the Folch procedure42.
510
6. Lipid composition is altered during thawing at room temperature. Therefore, to avoid tissue degeneration (1) cells are kept on ice or (2) tissue samples are cut and weighted while still frozen.
512
7. Sometimes it is useful to normalize the lipid levels in tissue samples by protein amount. Indeed, very small amount of tissue are often difficult to weigh without thawing them and, consequently, altering the lipid composition. Therefore, the samples are directly added to methanol (without the weighting step) and prior to extraction, 20 ml aliquots from the homogenate solutions are taken for protein measurements, which can be conducted using the Bradford protein concentration assay (Bio-Rad Laboratories Inc., Hercules, CA, USA) or the BCA protein assay (Pierce, Rockford, IL, USA).
516 517
8. The lipid extraction requires a ratio of chloroform, methanol, and water of 8:4:3. In these conditions, after centrifugation and phase separation, the approximate proportion of chloroform, methanol, and water in the upper phase is 3:48:47 by volume. In the lower phase, the respective proportion is 86:14:1.
526
9. To avoid contaminations from the upper aqueous phase into the pipette tip during the recovery of the bottom phase, insert the glass Pasteur pipette through the upper phase with gentle positive-pressure (i.e., gentle bubbling). Also, carefully withdraw the bottom phase through the pipette from the bottom of the vial. Furthermore, to avoid the interface or upper phase, it is better not to recover the entire bottom phase, but leaving the last drops (5–10% of the total organic phase) in the vials.
532 533
Acknowledgments
511 513 514 515
518 519 520 521 522 523 524 525 527 528 529 530 531
534 535 536 537 538 539 540
541
The contribution of the Agilent Technologies/University of California Irvine Analytical Discovery Facility, Center for Drug Discovery and the Agilent Technologies Foundation are gratefully acknowledged. This work was supported by grants from the National Institute of Health (R21DA-022702, R01DK-073955, R01 DA-012413, R01DA-012447, RR274–297/3504008, RR274–305/3505998, 1RL1AA017538 to D.P.).
542 543 544 545 546 547 548
218
Astarita, Ahmed, and Piomelli
549
References
550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606
1. Piomelli D, Astarita G, Rapaka R. (2007) A neuroscientist’s guide to lipidomics. Nat. Rev. Neurosci. 8, 743–754. 2. Welti R, Shah J, Li W, Li M, Chen J, Burke JJ, Fauconnier ML, Chapman K, Chye ML, Wang X. (2007) Plant lipidomics: discerning biological function by profiling plant complex lipids using mass spectrometry. Front. Biosci. 12, 2494–2506. 3. Nicholson JK, Holmes E, Wilson ID. (2005) Gut microorganisms, mammalian metabolism and personalized health care. Nat. Rev. Micro. 3, 431–438. 4. Kell DB, Brown M, Davey HM, Dunn WB, Spasic I, Oliver SG. (2005) Metabolic footprinting and systems biology: the medium is the message. Nat. Rev. Micro. 3, 557–565. 5. Lin CY, Viant MR, Tjeerdema RS. (2006) Metabolomics: Methodologies and applications in the environmental sciences. J. Pest. S. 31, 245–251. 6. Whitfield PD, German AJ, Noble PM. (2004) Metabolomics: an emerging post-genomic tool for nutrition. Br J Nutr. 92, 549–555. 7. German JB, Roberts M, Watkins SM. (2003) Personal metabolomics as a next generation nutritional assessment. J. Nutr. 133, 4260–4266. 8. Han X. (2007) An update on lipidomics: progress and application in biomarker and drug development. Curr. Opin. Mol. Ther. 9, 586–591. 9. Rapaka RS. (2005) Targeted lipidomics and drug abuse research. Prostaglandins Other Lipid Mediat. 77, 219–222. 10. Jung K, Astarita G, Zhu C, Wallace M, Mackie K, Piomelli D. (2007) A key role for diacylglycerol lipase-alpha in metabotropic glutamate receptor-dependent endocannabinoid mobilization. Mol. Pharmacol. 72, 612–621. 11. Chen C, Shah YM, Morimura K, Krausz KW, Miyazaki M, Richardson TA, Morgan ET, Ntambi JM, Idle JR, Gonzalez FJ. (2008) Metabolomics reveals that hepatic stearoylCoA desaturase 1 downregulation exacerbates inflammation and acute colitis. Cell Metab. 7,135–147. 12. Epand RM. (2008) Proteins and cholesterolrich domains. Biochim. Biophys. Acta. 1778, 1576–1582. 13. Gaspar ML, Aregullin MA, Jesch SA, Nunez LR, Villa-García M, Henry SA. (2007) The emergence of yeast lipidomics. Biochim. Biophys. Acta. 1771, 241–254.
14. Rabinowitz JD. Cellular metabolomics of Escherchia coli. (2007) Expert Rev. Proteomics. 4, 187–98. 15. Meer GV. (2005) Cellular lipidomics. EMBO J. 24, 3159–3165. 16. Hunt AN. (2006) Dynamic lipidomics of the nucleus. J. Cell. Biochem. 97, 244–251. 17. Astarita G, Ahmed F, Piomelli D. (2008) Identification of biosynthetic precursors for the endocannabinoid anandamide in the rat brain. J. Lipid Res. 49, 48–57. 18. Astarita G, Rourke BC, Andersen JB, Fu J, Kim JH, Bennett AF, Hicks JW, Piomelli D. (2006) Postprandial increase of oleoylethanolamide mobilization in small intestine of the Burmese python (Python molurus). Am. J. Physiol. Regul. Integr. Comp. Physiol. 290, R1407–1412. 19. Walker JM, Krey JF, Chen JS, Vefring E, Jahnsen JA, Bradshaw H, Huang SM. (2005) Targeted lipidomics: fatty acid amides and pain modulation. Prostaglandins Other Lipid Mediat. 77, 35–45. 20. Serhan CN. (2005) Mediator lipidomics. Prostaglandins Other Lipid Mediat. 77, 4–14. 21. Kaddurah-Daouk R, McEvoy J, Baillie RA, Lee D, Yao JK, Doraiswamy PM, Krishnan KR. (2007) Metabolomic mapping of atypical antipsychotic effects in schizophrenia. Mol. Psychiatry. 12, 934–945. 22. Adibhatla R, Hatcher J, Dempsey R. (2006) Lipids and lipidomics in brain injury and diseases. AAPS J. 8, E314-E321. 23. Wenk MR. (2005) The emerging field of lipidomics. Nat. Rev. Drug Discov. 4, 594–610. 24. Ackermann B, Hale J, Duffin K. (2006) The role of mass spectrometry in biomarker discovery and measurement. Curr. Drug Metab. 7, 525–539. 25. Fonteh A, Harrington R, Huhmer A, Biringer RG, Riggins JN, Harrington MG. (2006) Identification of disease markers in human cerebrospinal fluid using lipidomic and proteomic methods. Dis Markers. 22, 39–64. 26. Fahy E, Subramaniam S, Brown HA, Glass CK, Merrill AH Jr, Murphy RC, Raetz CR, Russell DW, Seyama Y, Shaw W, Shimizu T, Spener F, van Meer G, VanNieuwenhze MS, White SH, Witztum JL, Dennis EA. (2005) A comprehensive classification system for lipids. J Lipid Res. 46, 839–861. 27. Fu J, Astarita G, Gaetani S, Kim J, Cravatt BF, Mackie K, Piomelli D. (2007) Food intake regulates oleoylethanolamide formation and
Lipidomic Analysis of Biological Samples by Liquid Chromatography
degradation in the proximal small intestine. J. Biol. Chem. 282, 1518–1528. 28. Fuhrman BJ, Barba M, Krogh V, Micheli A, Pala V, Lauria R, Chajes V, Riboli E, Sieri S, Berrino F, Muti P. (2006) Erythrocyte Membrane Phospholipid Composition as a Biomarker of Dietary Fat. Ann. Nutr. Met. 50, 95–102. 29. Keshavan MS, Mallinger AG, Pettegrew JW, Dippold C. (1993) Erythrocyte membrane phospholipids in psychotic patients. Psychiatry Res. 49, 89–95. 30. Kirkland JJ, Truszkowski FA, Dilks CH, Engel GS. (2000) Superficially porous silica microspheres for fast high-performance liquid chromatography of macromolecules. J. Chromatogr. A. 890, 3–13. 31. Barroso B, Bischoff R. (2005) LC-MS analysis of phospholipids and lysophospholipids in human bronchoalveolar lavage fluid. J. Chromatogr. B. 814, 21–28. 32. Hodson L, Skeaff CM, Wallace AJ, Arribas GLB. (2002) Stability of plasma and erythrocyte fatty acid composition during cold storage. Clin. Chim. Acta. 321, 63–67. 33. Lau O, Wong S. (2000) Contamination in food from packaging material. J. Chromatogr. A. 882, 255–270. 34. German JB, Gillies LA, Smilowitz JT, Zivkovic AM, Watkins SM. (2007) Lipidomics and lipid profiling in metabolomics. Curr. Opin. Lipidol. 18, 66–71.
219
35. Han X, Gross RW. (2005) Shotgun lipidomics: multidimensional MS analysis of cellular lipidomes. Expert Rev. Proteomics. 2, 253–264. 36. Matyash V, Liebisch G, Kurzchalia TV, Shevchenko A, Schwudke D. (2008) Lipid extraction by methyl-tert-butyl ether for high-throughput lipidomics. J. Lipid Res. 49, 1137–1146. 37. Hara A, Radin NS. (1978) Lipid extraction of tissues with a low-toxicity solvent. Anal. Biochem. 90, 420–426. 38. Lin J, Liu L, Yang M, Lee M. (2004) Ethyl acetate/Ethyl alcohol mixtures as an alternative to Folch reagent for extracting animal lipids. J. Agric. Food Chem. 52, 4984–4986. 39. Boorman GA. (1999) Drinking water disinfection byproducts: review and approach to toxicity evaluation. Environ. Health Perspect. 107, 207–217. 40. Greim H, Reuter U. (2001) Classification of carcinogenic chemicals in the work area by the German MAK Commission: current examples for the new categories. Toxicology. 166, 11–23. 41. Wenk M, Lucast L, Di Paolo G, Romanelli AJ, Suchy SF, Nussbaum RL, Cline GW, Shulman GI, McMurray W, De Camilli P. (2003) Phosphoinositide profiling in complex lipid mixtures using electrospray ionization mass spectrometry. Nat. Biotechnol. 21, 813–817.
607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640
sdfsdf
Chapter 11
1
Lipidomic Analysis of Human Meibum Using HPLC–MSn
2
Igor A. Butovich
3
Summary
4
High-pressure liquid chromatography–mass spectrometry (HPLC–MS) has become a de facto standard analytical tool in lipidomic analyses of complex biological samples. This technique offers the best combination of selectivity and sensitivity among the currently available analytical methods, and provides not only the retention times of analytes, but also their m/z values, from which molecular masses of the compounds can be deduced. Further enhancement of the technique comes from the fact that some of the MS instruments (also known as ion traps, or MSn instruments) are capable of multistage fragmenting of the analytes, thus enabling the researcher to perform their structural elucidation. These capabilities make HPLC–MSn an ideal tool for analyzing small, complex lipid samples such as human meibum. Meibum is a very complex lipid mixture which is secreted onto the ocular surface by Meibomian glands. Meibum plays a critical role in the biochemistry and physiology of the human ocular surface. However, despite all efforts, its (bio)chemical composition remains elusive. In this chapter, several HPLC–MSn methods developed for lipidomic analysis of human meibum will be discussed. Considering the nature of analytes (all of which are hydrophobic compounds poorly soluble in water, and most of which are electroneutral), the only MS technique used in the study will be atmospheric pressure chemical ionization (APCI) MSn in both the positive and the negative ion modes. Electrospray ionization technique, though useful in phospholipid analyses, was found to be inadequate for analyzing less polar compounds, such as wax esters. As the data provided in this chapter will show that meibum is composed predominantly of nonpolar lipids of wax ester, cholesteryl ester, and triacylglycerol families with no appreciable amounts of more polar lipids present, APCI MSn seems to be a method of choice for lipidomic analysis of meibum and similar lipid mixtures.
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Key words: HPLC, Mass spectrometry, Atmospheric pressure chemical ionization, Ion trap, Polar lipids, Nonpolar lipids, Wax esters, Cholesteryl esters, Cholesterol, Phospholipids, Ceramides, Oleamide, Triacylglycerols
24 25 26
1. Introduction
27
Meibomian glands (MG) that are located at the perimeters of eyelids of humans and other mammals were originally described by H. Meibom in 1666 (1). MG are a major source of lipid Donald Armstrong (ed.), Lipidomics, Methods in Molecular Biology, vol. 579, doi 10.1007/978-1-60761-322-0_11, © Humana Press, a part of Springer Science + Business Media, LLC 2009
221
28 29 30
222 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 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77
Butovich
material (often referred to as meibum (2)) that is considered to be an important part of a so-called tear film lipid layer (TFLL) – an outermost part of tear film (TF) which covers the entire ocular surface, including cornea and conjunctiva (3, 4). TF and TFLL play a critical role in protecting the ocular surface from dehydration (5). Among other functions of TF and TFLL are antimicrobial, nutritional, and lubricating ones. The TF is also important in maintaining visual acuity as its thickness and uniformity affects refractive properties of cornea. Lipid composition of human TF and TFLL has been a subject of intensive research efforts for decades (see, for example, (2, 6–19)). However, considering relatively small amounts of samples that can be collected from human subjects without harming them (typically, less than 1 mg of dry meibum), and their extremely complex nature, a complete lipidomic analysis of meibum is yet to be completed. Earlier efforts in the area produced evidence that the vast majority of lipids found in meibum were of nonpolar nature (2, 3, 7–16). Among those reported were wax esters, cholesterol (Chl) and Chl esters, hydrocarbons, tri- and di-acylglycerols, di- and triesters of very complex nature (12), and many unidentified compounds. The fraction of nonpolar compounds typically was reported to comprise 60% or more of the entire lipid pool. The rest of the compounds were proposed to be of a more polar nature and included ceramides, sphingosines, various phospholipids, free fatty acids and fatty acid amides, monoacyl glycerols, etc. (16–22). It is noteworthy that all this body of data was generated over a period of 40 years and came from laboratories with extremely varying backgrounds (ophthalmic, biochemical, and chemical). The techniques utilized in those studies also varied with most common approaches being gas chromatography and/or gas–liquid chromatography with flame ionization detection and/or mass spectrometric detection. Also popular in the earlier studies was thin-layer chromatography, which later has been replaced by high performance liquid chromatography (HPLC). Later, attempts were made to utilize nuclear magnetic resonance spectrometry to study phospholipids (23) and infrared spectroscopy to study the composition and conformations of certain nonpolar lipids (24, 25). In recent years, mass spectrometry (MS) with or without HPLC has become a de facto standard for lipidomic analysis of complex lipid mixtures (16, 17, 26–31). Various MS techniques have been tried in meibum analyses with varying degree of success. Currently, its most advanced incarnation is HPLC–MS. This technique combines the sensitivity and selectivity of MS with the ability of HPLC to separate complex lipid mixtures thus improving the chances of correct identification of unknowns.
Lipidomic Analysis of Human Meibum Using HPLC–MSn
223
Several MS methods compatible with HPLC are available for a lipid chemist to choose from when designing a study. Those include, among others, electrospray ionization (ESI), atmospheric pressure chemical ionization (APCI), and atmospheric pressure photo ionization (APPI) techniques. However, a choice of an MS technique may be a determining factor in detecting a compound (or a set of compounds) or missing it altogether. ESI, for example, is very effective in detecting relatively polar lipids which include phospholipids, ceramides, and glycolipids (32). However, it is less than perfect in detecting nonpolar lipids such as wax esters and hydrocarbons. APCI, on the other hand, was successfully implemented in analyses of nonpolar lipids of meibum (29–31) and aqueous tears (31), and is generally considered a better technique than ESI for analyzing nonpolar compounds. APPI, being a newer technique than both ESI and APCI, is gaining ground as an effective tool in analyzing nonpolar lipids (33). In terms of its specificity, it is closer to APCI than to ESI, and, theoretically, could be used for analyzing even more hydrophobic (nonpolar) compounds than APCI could. Yet, it is less robust than the APCI technique because special care must be taken to avoid damaging the ultraviolet lamp of an APPI ion source. Among the three, APCI seems to be better suited for analyzing lipids of human meibum by HPLC–MS as it is very robust, sensitive, and compatible with a wide variety of HPLC solvents, and tolerates relatively high flow rates of a typical HPLC system. Another important consideration is a choice between an ion trap and a triple quadrupole mass detector. Without going into a lengthy discussion on the subject, it is sufficient to mention that triple quadrupole instruments are better suited for quantitation of the analytes, while ion traps offer unsurpassed capabilities for structural elucidation of unknown compounds. In particular, triple quadrupole mass spectrometers are capable of only MS and MS/MS (or MS2) experiments, while ion traps can easily perform multistage fragmentation up to MS5 and above. This comes at a price of a lower sensitivity of ion traps, and their narrower dynamic range. However, these disadvantages were found to be a nonissue in the lipidomic analysis of meibum, because collected samples were large enough to offset both the lower sensitivity and the narrower dynamic range. Thus, the capability of an ion trap to perform MSn analyses – an indispensable tool in studying of unknown compounds present in complex mixtures – made it a clear winner in this “competition.” In this manuscript the author will describe the methods of analyses of human meibum based on the ion trap APCI technique in conjunction with normal phase HPLC (NP HPLC).
78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123
224
Butovich
124
2. Materials
125
2.1. Equipment
126 127 128 129 130
1. An Alliance 2695 HPLC Separations Module (Waters Corp., Milford, MA) consisting of a quaternary gradient HPLC pump, thermostated column compartment, vacuum degasser, pulse dumper, and a built-in autoinjector. The HPLC system is controlled through an Empower (v.1.0) software installed on a PC-compatible computer operated under Windows XP.
143
2. A Lichrosphere Diol (3.2 × 150 mm, 5 mm) HPLC column (Phenomenex, Torrance, CA) for NP HPLC experiments (see Note 1, below). A Lichrosphere Si-60 silica gel (3.2 × 150 mm, 5 mm) HPLC column (from Sigma-Aldrich) for NP HPLC separation of phospholipids. 3. A Finnigan LCQ Deca XP Max ion trap mass spectrometer capable of performing MSn analyses from ThermoFisher Scientific (formerly, ThermoElectron) (Waltham, MA). The spectrometer must be equipped with an APCI ion source (see Note 2, below). An Xcalibur software (v.1.4 SR1) is used to operate the spectrometer and analyze the MS and HPLC data.
144
4. An AB135-S microbalance from Mettler (Toledo, OH).
131 132 133 134 135 136 137 138 139 140 141 142
147
1. HPLC- or spectroscopy grade n-hexane, propan-2-ol, acetonitrile, chloroform, and ethanol, mostly from Burdick&Jackson (Muskegon, MI).
148
2. Glacial acetic acid.
149 150
3. MilliQ-grade deionized water (18 MW) or bottled HPLCgrade water for preparing HPLC mobile phases.
151
4. 5 mM ammonium formate in HPLC-grade water.
152
5. Lipid standards used in the study were purchased from NuChek Prep, Inc. (Elysian, MN), Avanti Polar Lipids (Alabaster, AL), and Sigma-Aldrich (St. Louis, MO).
145 146
2.2. Reagents and Solvents
153 154
2.3. Supplies
156
1. An Eppendorf Easypet electric pipettor with glass pipettes ranging from 1 to 25 mL.
157 158
2. Positive displacement Digital Microdispensers (25 and 100 mL) from Drummond Sci. Co. (Broomall, PA).
159 160
3. HPLC-style microsyringes (10, 50, and 100 mL capacity) manufactured by SGE (Australia).
161
4. Block sample heater (Lab-Line Instruments, Melrose Park, IL).
162 163
5. One-milliliter flat-bottom clear glass shell vials with polyethylene snap caps (Waters Corp).
155
Lipidomic Analysis of Human Meibum Using HPLC–MSn
225
6. Total Recovery® glass HPLC vials with Teflon® liners (Waters Corp).
164
7. Borosilicate glass vials (19 × 65 mm, 11.3 mL) with caps equipped with Teflon® liners (VWR, Batavia, IL).
166
8. Glass solvent bottles with either glass stoppers or caps with Teflon® liners.
168 169
9. High-purity compressed nitrogen.
170
10. High-purity liquid nitrogen in a high-pressure tank.
171
3. Methods
3.1. Sample Collection and Handling
165 167
172
For general information on handling and storing lipids, the reader is advised to visit an Avanti Web page (http://www.avantilipids. com/storageandhandlingoflipids.html).
173 174
Disclosure: Human meibum was collected in a clinical setting from healthy volunteers using a procedure approved by the University of Texas Southwestern Medical Center Institutional Review Board and conducted in accordance with the Declaration of Helsinki. 1. Meibum is expressed from an eyelid of a volunteer using a plastic conformer and a Q-tip. The plastic conformer, placed between an eyelid and the eye protects the ocular surface, while the Q-tip is used to press the eyelid against the conformer to extrude meibum from an orifice of a MG. Immediately, the excreta are picked up with a polished platinum spatula. Care must be taken to avoid contamination of the sample with aqueous tears and cosmetic products. Meibum has a melting point of about 32–33°C (29) and, upon contact with the spatula, immediately solidifies at room temperature to assume a waxy structure. Its color is, typically, off-whitish to pale yellowish. Thus, it is easy to see if the sample has been successfully collected. Both upper and lower eyelids of both the eyes are used to collect the sample.
176 177
2. The solidified sample collected from one eyelid is immediately transferred into a preweighed flat-bottom HPLC glass vial filled with ~1.5 mL of chloroform: methanol solvent mixture (2:1, v/v; solvent CM). Dry weight of the vial is measured with precision of 0.01 mg or better before adding chloroform. An average meibum sample readily dissolves in 1.5 mL of the solvent mixture. The procedure is repeated for all four eyelids. All the samples are transferred into the same vial.
195 196
175
178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194
197 198 199 200 201 202
226
Butovich
3. Next, the sample solution is brought to dryness under a stream of nitrogen. The temperature of the sample is maintained by a thermostated sample heater at no more than 35°C. The vial with dry sample is weighed on the same microbalance and its weight is used to calculate the weight of the sample in the vial. On average, ~0.7 mg of meibum could be collected from four eyelids of a volunteer. The repeatability of the weight determination should be better than 0.01 mg. When the samples are to be stored for an extended period, the vials with dry material are flashed with dry nitrogen, sealed, and stored at −80°C. The dry samples were shown to be stable for at least 6 months to a year.
203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221
3.2. Sample Analyses 3.2.1. Nonpolar Lipid Analyses Using Isocratic NP HPLC–APCI MS 3.2.1.1. HPLC Procedures
2. A 100 mg/mL stock solution of meibum is prepared by dissolving the dry meibum material in an appropriate volume of a n-hexane:propan-2-ol solvent mixture (1:1, v/v, HP) and stored in a borosilicate 19 × 65 mm glass vial at −80°C under nitrogen. Just before the HPLC–MS analysis, the sample (50– 100 mL) is placed into a TotalRecovery® HPLC vial. Note that the cap must not have a silicon backing to a Teflon® liner as the tiny silicon particles picked up by the autoinjector’s needle will contaminate the sample and produce strong HPLC and MS signals. (Remember: if in doubt, leave the cap out, or remove the liner and its backing altogether. This will not affect the results of experiments if the sample is to be immediately injected).
222 223 224 225 226 227 228 229 230 231 232 233 234
3. Routinely, between 1 and 10 mL of the sample is injected. It was found that larger volumes of the stock solution would impact the shape and retention times of the HPLC peaks, while higher concentrations of the meibum solutions could overwhelm the detector and contaminate the ion source and/ or the HPLC column. Note that vacuum degassing of the solvent is advised to achieve the most stable baseline and avoid problems with air bubbles, but is not necessary if the sample concentration is sufficiently high.
235 236 237 238 239 240 241 242 243 244 245 246 247 248
1. A n-hexane:propan-2-ol:acetic acid solvent mixture (94:5:1, v/v/v, HPA) is used (see Note 3, below). The Diol column is pre-equilibrated in this solvent at 30°C for 15 min or until the back pressure and the MS signal stabilized. A 30-min analysis is performed isocratically at the flow rate of 0.3 mL/min and a column temperature of 30°C. The entire effluent is directed to the APCI ion source.
3.2.1.2. Mass Spectrometric Procedures
The mass spectrometer equipped with the APCI ion source is operated in the positive ion mode. Before the meibum analyses, the spectrometer is tuned using a 1 mg/mL behenyl oleate (BO) stock solution in the HPA solvent mixture. The sample is infused
Lipidomic Analysis of Human Meibum Using HPLC–MSn
227
at a flow rate of ~10 mL/min using the mass spectrometer’s builtin syringe pump. The Autotune routine of the Xcalibur software is engaged to tune to a signal with m/z value of 591.5 (a proton adduct of BO). The final MS parameters should be as follows: Source voltage
3 kV
Source current
5 mA
Vaporizer temperature
350°C
Sheath gas flow rate
20 arb. units
Capillary voltage
12 V
Capillary temperature
350°C
Tube lens voltage
−40 V
Multipole 1 offset
−2 V
Lens voltage
−41 V
Multipole 2 offset
−11 V
Entrance lens
−75 V
The parameters should be saved in an Autotune file and are used for meibum analyses and in experiments with standard nonpolar lipids. 3.2.1.3. Results HPLC–MS Analyses of Nonpolar Lipid Standards
Human meibum was reported to contain nonpolar lipids of various classes. Therefore, a range of standard lipids should be tested to determine their retention times under the conditions of NP HPLC analysis on a Diol column. A typical test mixture contains 50 mM each of behenyl oleate (a wax ester, MW 590, BO), cholesteryl oleate (a cholesteryl ester, MW 650, Chl-O) and free cholesterol (MW 386, Chl), triolein (a triacylglycerol, MW 885, TO), dipalmitin (a diacylglycerol, MW 569, DP), C18-ceramide (MW 566, C18-Cer), oleamide (a fatty acid amide, MW 281, OA), and 1-monomyristoyl glycerol (a monoacyl glycerol, MW 303, MG). Based on the published data, discussed in Subheading 1, compounds of these classes are expected to be present in human meibum. The lipids may be dissolved in either chloroform or the HP solvent mixture. No difference in the retention times of the analytes were observed provided that the sample injection volume did not exceed 10 mL. Note that to accommodate for the higher flow rates typical of HPLC-style experiments, the flow of sheath gas needs be increased to 40 arbitrary units. The rest of the MS parameters should remain as optimized in the Autotune routine. A sample NP chromatogram of the mixture is presented in Fig. 1. The retention times of the test lipids were found to be
249 250 251 252
253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277
228
Butovich
Fig. 1.Total ion chromatogram of a mixture of standard lipids (TIC) and reconstructed chromatograms of its individual components. Depicted are: BO (ion m/z 591; retention time 3.4 min); Chl-O (m/z 369, 3.5 min); Chl (m/z 369; 6.9 min); TO (m/z 885; 3.6 min); 1,2-DP (m/z 551; 3.9 min) and 1,3-DP (m/z 551; 6.4 min); C18-Cer (m/z 548; 13.7 min); OA (m/z 282; 19.2 min), and 1-MG (m/z 285; 22.8 min).
HPLC–MS Analyses of Nonpolar Components of Meibum
Lipidomic Analysis of Human Meibum Using HPLC–MSn
229
Chl-O » TO » BO