Professional Documents
Culture Documents
Full Chapter A Practical Guide To Metabolomics Applications in Health and Disease From Samples To Insights Into Metabolism Learning Materials in Biosciences 1St Edition Julijana Ivanisevic PDF
Full Chapter A Practical Guide To Metabolomics Applications in Health and Disease From Samples To Insights Into Metabolism Learning Materials in Biosciences 1St Edition Julijana Ivanisevic PDF
https://textbookfull.com/product/hydrogen-peroxide-metabolism-in-
health-and-disease-1st-edition-hampton/
https://textbookfull.com/product/preventing-occupational-
exposures-to-infectious-disease-in-health-care-a-practical-guide-
amber-hogan-mitchell/
https://textbookfull.com/product/metabolomics-practical-guide-to-
design-and-analysis-1st-edition-ron-wehrens-editor/
https://textbookfull.com/product/superconductivity-from-
materials-science-to-practical-applications-paolo-mele/
Environmental Metabolomics: Applications in field and
laboratory studies to understand from exposome to
metabolome 1st Edition Diana Alvarez-Munoz (Editor)
https://textbookfull.com/product/environmental-metabolomics-
applications-in-field-and-laboratory-studies-to-understand-from-
exposome-to-metabolome-1st-edition-diana-alvarez-munoz-editor/
https://textbookfull.com/product/managing-health-in-the-genomic-
era-a-guide-to-family-health-history-and-disease-risk-1st-
edition-vincent-henrich/
https://textbookfull.com/product/a-practical-approach-to-using-
statistics-in-health-research-from-planning-to-reporting-1st-
edition-adam-mackridge/
https://textbookfull.com/product/neutrons-and-synchrotron-
radiation-in-engineering-materials-science-from-fundamentals-to-
applications-clemens/
https://textbookfull.com/product/frailty-and-kidney-disease-a-
practical-guide-to-clinical-management-carlos-guido-musso/
Learning Materials in Biosciences
Julijana Ivanisevic
Martin Giera Editors
A Practical Guide
to Metabolomics
Applications in
Health and Disease
From Samples to Insights into Metabolism
Learning Materials in Biosciences
Learning Materials in Biosciences textbooks compactly and concisely discuss a specific
biological, biomedical, biochemical, bioengineering or cell biologic topic. The textbooks in
this series are based on lectures for upper-level undergraduates, master’s and graduate
students, presented and written by authoritative figures in the field at leading universities
around the globe.
The titles are organized to guide the reader to a deeper understanding of the concepts
covered.
Each textbook provides readers with fundamental insights into the subject and prepares
them to independently pursue further thinking and research on the topic. Colored figures,
step-by-step protocols and take-home messages offer an accessible approach to learning
and understanding.
In addition to being designed to benefit students, Learning Materials textbooks represent
a valuable tool for lecturers and teachers, helping them to prepare their own respective
coursework.
Julijana Ivanisevic • Martin Giera
Editors
A Practical Guide
to Metabolomics
Applications in Health
and Disease
From Samples to Insights into Metabolism
Editors
Julijana Ivanisevic Martin Giera
Faculty of Biology & Medicine Center for Proteomics & Metabolomics
University of Lausanne Leiden University Medical Center
Lausanne, Switzerland Leiden, The Netherlands
# The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG
2023
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the
whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws
and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book are
believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a
warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that
may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
The coronavirus pandemic was a serious challenge for our society. Lockdowns and
restrictions affected all of us in different ways. Particularly, undergraduate students sud-
denly had to shift toward online lessons and courses, which severely impacted their
(practical) training. However, this shift has also accelerated and led to permanent changes
in our teaching systems with online content and learning becoming increasingly important.
This is when the idea of a metabolomics book accompanied with online content and open-
access datasets was born. Thereby, the main aim of this book is to facilitate online learning
for undergraduate students in the field of metabolomics and to provide guidelines and
materials for practical course or tutorial instructors.
As metabolomics has matured and become a routinely applied tool in (patho-) biochem-
istry, many excellent books cover the biological as well as chemical facets of this inter-
disciplinary analytical science. However, the content aimed at undergraduate students with
a clear teaching focus and practical examples is lacking. With the presented book A
Practical Guide to Metabolomics Applications in Health and Disease: From Samples to
Insights into Metabolism, we aim to bridge this gap by providing the datasets and step-by-
step protocols used for their treatment and embedded in the context of specific case studies
to students and course supervisors to facilitate self-study and training, respectively.
Following a general introduction, the book first covers the hottest topics in the field,
from the overlooked sample handling and preparation to highly debated data processing
and metabolite annotation. The following two chapters present and discuss innovative
strategies for data interpretation in the biochemically and physiologically relevant context.
The presented protocols oriented on specific steps in metabolomics workflow are followed
by the presentation of five stories told through metabolomics-derived data acquired in the
context of specific physiological conditions (e.g., exercise and healthy aging) or disease
(e.g., inflammation and cancer). Finally, the last three chapters present the case studies
accomplished by the emerging mass spectrometry-assisted imaging technology, which is
likely to pave the future of tissue and single-cell metabolomics. Practical examples from
v
vi Preface
our laboratories with easy-to-follow instructions enable course instructors and students
alike in designing educative training and working with “real-life” data.
Part I Introduction
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Martin Giera and Julijana Ivanisevic
Part III From Data Processing to Polar and Lipid Metabolite Identification
5 METLIN Tandem Mass Spectrometry and Neutral Loss Databases
for the Identification of Microbial Natural Products and Other
Chemical Entities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Wilasinee Heim, Aries Aisporna, Linh Hoang, H. Paul Benton, and
Gary Siuzdak
6 Untargeted GC-MS Data Processing and Metabolite Identification
Using eRah . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Sara M. de Cripan, Trisha Arora, Adrià Olomí, Jasen P. Finch, and
Xavier Domingo-Almenara
vii
viii Contents
The metabolome comprises the entity of all small molecules (metabolites) involved in
energy metabolism (as fuels), maintenance of cell structure (as building blocks) and
metabolic signaling, as messenger molecules. Metabolomics strives to identify, analyze,
and quantify all metabolites of a given biological system. Consequently, metabolomics is
defined as “the (quantitative) study of all small molecules (metabolites) of a biological
system such as cells, tissues, or organisms.” Metabolomics is significantly intertwined with
analytical chemistry, biochemistry, and lately systems biology. Historically, metabolomics
evolved from the centuries-old effort to describe and define the molecular fundaments of
life. Ultimately it is a biochemical tool set to shed light on the molecular fundaments of
metabolism. Importantly, metabolomics was only made possible by significant technologi-
cal developments, most importantly mass spectrometry (MS) and nuclear magnetic reso-
nance spectroscopy (NMR) [1]. Today, metabolomics can be divided into several subfields
and categories. On one hand, technological characteristics define targeted and untargeted
M. Giera (✉)
LUMC, Leiden, The Netherlands
Metabolomics Unit, University of Lausanne, Lausanne, Switzerland
e-mail: m.a.giera@lumc.nl
J. Ivanisevic (✉)
LUMC, Leiden, The Netherlands
Metabolomics Unit, University of Lausanne, Lausanne, Switzerland
Lausanne, Switzerland
e-mail: julijana.ivanisevic@unil.ch
metabolomics analysis. On the other hand, subfields, for example, lipidomics, computa-
tional metabolomics, plant metabolomics, and many more, have evolved in recent years.
Targeted metabolomics refers to the (quantitative) analysis of large panels of known
metabolites. In most cases, targeted methods make use of commercially available standard
materials which are used to establish the analytical assays. Most targeted metabolomics
panels are run by liquid chromatography tandem mass spectrometry (LC-MS/MS) or NMR
with some applications also being developed for gas chromatography mass spectrometry
(GC-MS) [2]. In turn, the observed chromatographic (or spectroscopic) signals are
identified based on specific retention times (or chemical shifts in NMR) and characteristic
mass spectrometric features (multiplicity) in comparison with genuine chemically
characterized standard materials. In the case of untargeted metabolomics, compound
identification takes place post-analysis. In other words, biological samples are most
frequently analyzed using either LC-MS/MS, GC-MS, or NMR, and subsequently metab-
olite identification is initiated [3]. Several chapters of this book will give detailed
instructions on how such data must be preprocessed and treated for successful metabolite
identification and biological interpretation. Nevertheless, it is important to point out that
untargeted metabolomics analysis usually results in hundreds to thousands of chro-
matographic signals for which mass spectrometric data has been obtained. While some
of these data might be matched against existing databases (e.g., METLIN) or genuine
synthetic standard materials, others are only putatively or ambiguously identified or remain
unknown. Consequently this has led to the introduction of a classification system for the
level of metabolite identification by the Metabolomics Standards Initiative (MSI) [4]. The
original classification was comprised of four levels ranging from level 1 (unambiguously
identified analyte) to level 4 (unidentified analyte). In this context, metabolic features
should be distinguished from metabolites. A metabolic feature is basically a signal (or an
ion) characterized by retention time and (tandem) mass spectrum (mass to charge or m/z
ratio) which are recorded in untargeted metabolomics analysis. Additionally, the precise
ion mobility (or collision cross section (CCS)) value of an ion reflecting its physicochemi-
cal properties can also be acquired nowadays (depending on the instrument setup). The
process of metabolite annotation is the assignment of metabolite candidates to metabolic
features, and a metabolite is an unambiguously identified chemical structure partaking in
metabolism. Several important definitions are summarized in Box 1.1.
(continued)
1 Introduction 5
The elucidation of the DNA structure in 1953 and the subsequent publication of molecular
biology’s central dogma in 1958 set the stage for our present understanding of the
information flow in biological systems [1]. It basically states that information flows from
DNA (genome) to RNA (transcriptome) to protein (proteome), and when the proteome
level has been reached, it cannot flow back, while this is possible between DNA and RNA.
Remarkably, at this time, metabolites (metabolome) had not been considered a central part
of information flow but pure biochemical bystanders. This view has dramatically changed
during the last decennia [5]. For example, several important classes of signaling molecules,
for example, fatty acyl esters of hydroxy fatty acids (FAHFA) [6], the eicosanoids or
oxysterols, and other oxidized lipids, have been identified [7, 8]. Consequently, metabolites
and the metabolome have become an integral part of our understanding of how information
in biological systems is transferred and its relation to phenotypic observations (Fig. 1.1).
Moreover, today the metabolome is considered as the phenotype at the molecular level, and
it has become accepted that information can flow back to the above-indicated layers of
biochemical organization. For example, posttranslational modifications represent a
proteome-metabolome interaction; the same is true for post-transcriptional modifications
(transcriptome—metabolome interaction) and epigenetics where the interaction between
DNA and metabolites plays a significant role [5]. Even more so, it is becoming increasingly
accepted that the metabolome (along with lipidome) is organelle- and cell type-specific. In
other words, a specific cell type, for example, a macrophage or neutrophil or even subsets
such as M1 or M2 macrophages, presents with a specific metabolome that is closely linked
to their function and phenotype.
6 M. Giera and J. Ivanisevic
Fig. 1.1 Biological information flow and the role of metabolites, taken with permission from [5]
As outlined above, metabolomics as a scientific field has largely been made possible by the
rise of advanced analytical technologies. While Williams and co-workers were among the
first to recognize the significance of individual biochemical signatures in the late 1940s, it
was Dalgliesh et al. in 1966 who for the first time used gas chromatography coupled to
mass spectrometry (GC/MS) to study urinary metabolites. Subsequently, Horning and
colleagues introduced the term metabolic profiles and, together with Linus Pauling and
Arthur Robinson, developed GC/MS methods to simultaneously monitor dozens of
metabolites present in biological samples during the 1970s. However, the most significant
cornerstone of metabolomics is electrospray ionization (ESI), which was mainly developed
by John Fenn in the late 1980s. Today, a plethora of different analytical technologies are
applied to deepen our understanding of the metabolome [1]. Below we will briefly describe
these technologies, and an overview is given in Fig. 1.2.
GC-MS was among the first technologies applied in metabolomics analysis. Moreover,
GC-MS is one of the first examples of online coupling a separation device (gas
1 Introduction 7
Fig. 1.2 Main analytical technologies used for metabolomics analysis (figure made with biorender.
com)
Fig. 1.3 The basic components of a GC-MS system (figure made with biorender.com)
are consecutively evaporated and transported through the column. In the case of GC-MS,
the column outlet is introduced into the MS via a transfer line. The transfer line is basically
a heated piece of capillary, ensuring that no analytes are trapped by a cold bridge between
GC and MS. After entering the MS, ionization in most cases takes place in an electron
ionization source (EI). Electrons expelled from a heated filament are accelerated into the
ionization chamber. The kinetic energy of the so produced electrons can be controlled and
is usually set to 70 eV. In the ionization chamber, neutral molecules, eluting from the GC
column, are interacting with the accelerated electrons. When the accelerated electrons
originating from the filament get into close proximity of a neutral molecule, an electron
is being expelled, resulting in the formation of a positively charged molecule and its
fragmentation. The so produced charged molecules and their fragments can now be filtered
by a mass analyzer (usually a quadrupole) and quantified as electric current through means
of an electron multiplier (Fig. 1.3).
The early success of GC-MS is related to the fact that coupling separation and detection
technology is rather straightforward as the carrier gas automatically transports the separated
molecules into the ionization chamber and can simply be pumped away. Moreover, for
many decades, significant mass spectrometric libraries for compound identification only
existed for EI-MS spectra. This is due to the fact that EI-MS spectra can be generated in a
very reproducible way on a single-stage MS instrument. The main variable influencing the
obtained EI-MS spectra is the applied ionization energy, and 70 eV has become the most
widely accepted standard. Today the NIST GC-MS library contains more than 350,000
searchable spectra. Additionally, EI-MS spectra follow distinct fragmentation rules (not to
1 Introduction 9
be discussed here) which to some extent allow for the identification of unknowns. How-
ever, besides these advantages, GC/MS does come with a couple of limitations. It is limited
to vaporizable components, which also restricts its application to low molecular weight
components (<750 Da). Moreover, EI is a very hard ionization technique that results in
hundreds of fragment ions and very complex spectra, and last but not least, thermal stability
is a prerequisite for GC/MS analysis. Throughout the years, researchers have developed
strategies to address these limitations. In order to increase thermal stability of functional
groups and to allow analysis of hard-to-vaporize molecules, numerous derivatization
strategies for hydrophilic functional groups have been developed; examples include
trimethylsilylation, methylation, acetylation, and many more (examples are shown in
Fig. 1.4) [9]. A typical protocol for GC/MS-based metabolomics would combine
methoximation and trimethylsilylation [10]. To allow for a milder ionization, so-called
chemical ionization (CI) has been developed. CI makes use of a reagent gas, typically
ammonia or methane. Within the ionization chamber, the number of reagent gas molecules
by far exceeds the number of analyte molecules. In turn, the reagent gas will be ionized by
the available free electrons whereafter a cascade of chemical steps leads to the ionization of
the analyte molecules. In other words, the reagent gas works as an intermediate buffering
the energies within the collision cell. Of note, unlike EI, CI can also produce negatively
charged ions, a fact that has led to the development of electron capture negative ionization
(ECNI). Particularly in combination with highly electrophilic derivatization, for example,
pentafluorobenzyl bromide, GC-ECNI-MS is a very selective and sensitive approach for
the analysis of carboxylic acids [11].
10 M. Giera and J. Ivanisevic
As outlined above GC-MS was the first hyphenated technique applied in metabolomics
research. However, its limitations hampered its application particularly for larger as well as
hydrophilic molecules. The solution to this issue was the development of electrospray
ionization (ESI) and liquid chromatography mass spectrometry (LC-MS). Liquid chroma-
tography (LC) is a partition-based separation technology. Partition of molecules takes place
between a stationary phase and a liquid eluent. For a detailed description of the fundamen-
tal principles of LC, please refer to dedicated literature. A schematic of a LC-MS system is
shown in Fig. 1.5. High-pressure LC pumps deliver a gradient of a weak and a strong eluent
to a stationary phase which is connected to the MS detector. Modern LC systems usually
operate at pressures between 200 and 800 bar. The most frequently applied eluent
combinations in LC-MS analysis are methanol or acetonitrile as organic modifiers and
water with 0.1% formic acid or MS-compatible buffer salts as aqueous phase. Given the
fact that metabolomics analysis of bodily fluids or cells deals with hundreds to thousands of
metabolites, it is understandable that superior chromatographic performance is desired.
Due to the fact that LC separation is influenced by specific factors, for example, extra
column volume, particle size, and flow rate (familiarize yourself with the van Deemter
curve), ultra-high-pressure LC (UPLC) in combination with either sub 2 μm particle
columns or core-shell materials have become standard. When it comes to the available
1 Introduction 11
Fig. 1.5 Schematic of an LC-MS system, LC stationary phases and typical settings for RP-based
metabolomics analysis (figure partially made with biorender.com)
stationary phases there are two main choices: on one hand, reversed-phase
(RP) chromatography mainly employing C18 materials and hydrophilic interaction chro-
matography (HILIC) using diverse HILIC materials. The difference between the two
stationary phases mainly lies in their respective mode of analyte interaction. RP materials
are ideal for the retention of lipophilic components, and the stationary phase/analyte
interaction is mainly based on nonpolar van der Waals interactions. Compound elution is
accomplished by increasing amounts of an organic modifier that competes with the
stationary phase and hence will move components along the separation column. HILIC
on the other hand is ideal for the retention of very polar, hydrophilic components. HILIC
stationary phases, for example, diol or amine-functionalized silica are polar phases that
strongly retain polar metabolites. For example, glucose, a very polar metabolite will elute
in the void volume of a RP column, while it will have strong retention on a HILIC
phase [12].
The hyphenation of LC and MS was mainly accomplished by ESI. Unlike for GC-MS
where a gas flow can directly be introduced into an ionization chamber does LC work with
large amounts of liquid eluents. Moreover, MS detection only works with ionized
components and takes place under high vacuum conditions while elution from the LC col-
umn takes place at atmospheric pressure. In turn, the main challenge was to ionize solved
12 M. Giera and J. Ivanisevic
Fig. 1.6 The basic parts of an ESI source and the CRM model for the formation of gas phase ions
(figure made with biorender.com)
analyte molecules and transfer them into the MS analyzer without breaking its high
vacuum. Notably, aqueous liquids result in very high vapor volumes when being
evaporated. Early devices basically build on technology already applied in GC-MS analy-
sis (not discussed here). However, the real breakthrough was the development of the ESI
source by Fenn and colleagues. Today, 90% of all LC-MS applications are run with an ESI
interface. In ESI, the continuous LC flow is pumped through a charged metal capillary
resulting in its dispersion. This process is referred to as electrospray. The resulting fine
aerosol subsequently undergoes extensive desolvation, and the charged liquid droplets start
to shrink. At the so-called Rayleigh limit, the electrostatic repulsion of like charges, in the
steadily decreasing droplets, overcomes the surface tension holding the droplet together,
and the droplets basically explode forming significantly smaller droplets. There are two
major theories how gas phase ions are ultimately formed after this process: the ion
evaporation model (IEM) and the charge residue model (CRM). The IEM argues that the
field strength of the continuously shrinking droplets assists field desorption at one point.
The CRM on the other hand argues that all solvent will eventually be evaporated in
shrinking fission cycles ultimately leaving the charged analytes that were contained in
the droplets. Figure 1.6 outlines the main parts of an ESI inlet. Other ionization
technologies are atmospheric pressure chemical ionization (APCI) and atmospheric pres-
sure photoionization (APPI).
Following the generation of gas phase molecules, these can be directed toward the mass
analyzer. Today many types of mass analyzers exist; besides others these include
1 Introduction 13
quadrupoles (Q), triple quadrupoles (QqQ), ion traps (IT), Orbitraps, time-of-flight (TOF),
and hybrid instruments such as quadrupole-time-of-flight (QTOF) instruments, or linear
ITs. Nowadays, almost all metabolomics research involves the generation of tandem mass
spectra (MS/MS). In order to generate MS/MS data, the ions generated during ESI are sent
to a collision cell for their fragmentation. Unlike EI, ESI is a soft ionization technique
resulting mainly in the formation of molecular ions. However, to obtain characteristic
molecular information, analyte fragmentation is of greatest importance not only for com-
pound identification but also for quantification. A collision cell is being introduced after the
first mass filter. For example, in a QqQ system, the second Q will be the collision cell. In a
QTOF instrument, a collision cell will be present between the Q mass filter and the TOF
mass analyzer. In all cases the purpose of the collision cell is analyte fragmentation. This is
typically achieved by the introduction of a collision gas, for example, argon. The charged
molecules passing the first mass filter will enter the collision cell and collide with collision
gas molecules; depending on the kinetic energy and the collision gas density, this will
cause the molecules to fragment, producing so-called fragment ions. The so produced
fragment ions can now be passed through another mass filter and made visible by an
electron multiplier. The entity of all produced fragment ions is called a tandem mass
spectrum or MS/MS spectrum. Importantly and just like in EI fragmentation, the produced
fragment ions are compound-specific and can be used for analyte identification and
quantification. METLIN is the biggest MS/MS database with almost one million physically
analyzed components (https://metlin.scripps.edu/). A dedicated chapter of this textbook
will explain in detail how to match and identify metabolites using XCMS and METLIN.
Within the field of metabolomics, QqQ instruments are used for wide targeted analysis,
whereas particularly QTOF instruments are being applied for untargeted metabolomics
assay. The reasons for this are as follows. QqQ instruments have become the gold standard
for targeted quantitative analysis, mainly due to the fact that operating QqQ instruments in
multiple reaction monitoring (MRM) mode significantly lowers background noise and
increases selectivity and thereby also sensitivity (think of the signal-to-noise ratio as the
main driver of sensitivity). The choice for QTOF instruments in untargeted workflows is
the fact that QTOF instruments typically have a high mass resolving power > 30,000 and
are capable of very fast acquisition of tandem mass spectra (up to 200 Hz). While high mass
resolution significantly improves analyte identification, as in limiting the relevant search
space in terms of analyte mass, the high scan speed allows for excellent coverage of the LC
dimension, generating sufficient data points to define a chromatographic peak and simulta-
neously allow to acquire a large number of MS/MS spectra. Figure 1.7 shows the basic
design of QqQ and QTOF mass analyzers and exemplifies the difference in mass
resolution.
Importantly, QqQ instruments can be operated in several scan modes; for simplicity we
will here only focus on the most widely used mode of multiple reaction monitoring (MRM)
analysis. MRM analysis is illustrated in Fig. 1.8. During MRM scanning, the first quadru-
pole selects an analyte-specific precursor ion, frequently the molecular ion of the com-
pound under investigation. Subsequently, this ion is being fragmented in Q2 (collision cell)
14 M. Giera and J. Ivanisevic
Fig. 1.7 Basic components of a QqQ (left) and QTOF (right) MS system. Below, comparison of
mass resolution for the negative molecular ion of glucose, left, QqQ, right TOF MS. QTOF picture
reused with permission from [13], QqQ picture courtesy of Sciex, reused with permission
Fig. 1.8 The operation principle of MRM analysis exemplified for the analysis of caffeine in a
human saliva sample. Note the substantial increase from a total ion current chromatogram on the left
to a single observed peak in MRM mode
1 Introduction 15
producing analyte-specific fragment ions. Of these, one or several analyte-specific ions are
selected in Q3; in other words they are allowed to pass Q3 (hence the name mass filter) to
reach the detector. During this process selectivity and background noise are constantly
improved, which in turn leads to increased signal to noise and hence sensitivity.
internal standards as well as producing highly reproducible data. Recently, NMR has also
become an attractive alternative for the analysis of lipoproteins. The disadvantages of NMR
analysis include its lower sensitivity when compared to MS-based approaches (factor
100–1000), the rather high amounts of biological samples needed (e.g., approximately
250 μL blood plasma), and its high investment and maintenance cost. Nevertheless, NMR
is an important cornerstone of modern metabolomics analysis and the most important
available technology for small metabolite structural elucidation. For further details on
NMR-based metabolomics, please refer to the dedicated chapters in this textbook.
Analytical
technology Advantage Disadvantage
GC-MS High separation efficiency Analyte breakdown
Informative MS1 spectra Derivatization necessary
Highly standardized Hard ionization
Large libraries
Relatively cheap
Matrix effects are limited
LC-MS(MS) Large molecular coverage Expensive
Large libraries Fragmentation can be instrument-
Highest sensitivity specific
Neutral molecules can be hard to ionize/
analyze
Matrix effects
NMR Very robust Very expensive
Highly standardized Limited sensitivity
Absolute quantitative Complex spectra
De novo substance
identification
Large coverage
During the last decade, several metabolomics subfields have arisen, focusing on either
specific organisms, e.g. plant metabolomics, or analyte classes, for example, lipidomics.
Other disciplines gaining significance are computational metabolomics, spatial
metabolomics (e.g., mass spectrometry imaging) or single-cell metabolomics. Moreover,
there is a general trend to integrate and match several omics layers in so-called multi-omics
1 Introduction 17
approaches. The most important field might be lipidomics that will be discussed in more
detail below.
1.4.1 Lipidomics
Lipidomics, the comprehensive study of lipids is one of the most recent and also significant
metabolomics subfields. The importance of lipid analysis can be exemplified by the fact
that more than 70% of human plasma metabolites are lipids. Moreover, lipids play very
important roles as signaling molecules. A prime example is the eicosanoids, metabolites of
the fatty acid arachidonic acid. The eicosanoids play very important roles in inflammation,
bronchoconstriction, as well as blood clotting and form the target of nonsteroidal anti-
inflammatory drugs (NSAID). Due to their outstanding biological importance, lipids have
taken center stage in many disease areas, for example, neurodegenerative diseases, inflam-
matory diseases, as well as cancer.
The term lipid is rather loosely defined mainly based on a substance solubility in organic
solvents. According to the IUPAC gold book, a lipid is defined as “. . .A loosely defined
term for substances of biological origin that are soluble in nonpolar solvents. . . .” Never-
theless, this rather practical definition is reflected in sample preparation techniques for
lipidomics analysis making use of organic solvents such as methyl-tert-butyl ether or
isopropanol. The main analytical technologies as described above for metabolomics are
also being used for lipidomics analysis. However, there are some peculiarities specific to
lipidomics. Unlike most metabolites, lipid classes are characterized by a building block
structure. In other words, a specific head group (e.g., glycerol or phosphatidylcholine) is
combined with fatty acid side chains to form a complex lipid species. This fact has
important implications for lipid analysis: (1) to a certain extend, de novo lipid identification
can to the most part principally be accomplished by MS/MS analysis without the need for
extensive libraries; (2) lipids present an extremely diverse class of molecules with numer-
ous possibilities in terms of head group and side chain combinations; (3) isomers pose a
significant challenge in lipid analysis—consider, for example, geometric double-bond
isomers, positional isomers, or enantiomers and diastereomers. This latter fact has led to
significant efforts in the separation of lipids based on LC but also ion mobility approaches.
As described above, lipids can be separated using either RP or HILIC-based LC
separations. Importantly, RP separates lipids mainly on their side chain composition,
whereas HILIC separates lipids based on their head groups. A very innovative and rapid
separation technology for lipids is ion mobility, allowing to separate lipid classes in
milliseconds within the gas phase, right before entering the MS. An overview of several
lipid classes and their preferred mode of analysis is given below (Fig. 1.9). For further
details on lipidomics sample preparation and analysis, please refer to the dedicated chapters
of this textbook.
18 M. Giera and J. Ivanisevic
Fig. 1.9 The lipid building block concept (examples of head groups and side chains), lipid analysis
technologies, and a phosphatidylcholine lipid with two oleic acid side chains; the circle highlights a
stereocenter in 5-hydroxy-arachidonic acid
1 Introduction 19
For the past decades, during the era of biochemical genetics, the focus of metabolism
research was on genes and their function. However, nowdays, in the post-genomics era of
biochemistry, we are focusing back on metabolites and their function in diverse metabolic
processes [14]. In other words, we are “rethinking” or revisiting metabolism using
metabolomics tools. This “comeback” is due to recent technological developments which
allow us to measure metabolites with more specificity and sensitivity than ever before. For
this reason, metabolomics is often qualified as a next-generation metabolic profiling.
Although we are still far from omics scale metabolite analysis, we have made an important
step-forward, toward the analysis of a broad range of metabolites, from highly abundant
nutrients involved in energy production and storage to low abundant signaling molecules.
Importantly, metabolomics approaches have allowed us to discover new biological and
potent signaling roles of well-known energy metabolites and compounds we previously
thought to be plain by-products of energy metabolism [15, 16]. As David Wishart nicely
resumed “While the metabolites may be small. . .their influence in human physiology and
disease is truly profound—they are perhaps the body’s most important signaling
molecules” [17]. These functionally diverse metabolites are also chemically highly diverse.
As Gary Siuzdak noted, “there are no limits on how metabolites can be assembled.” In
addition, metabolites span an extremely wide concentration range in biological matrices, at
least 12 orders of magnitude [18]. This is why we often claim that this high chemical
diversity and wide concentration ranges in which metabolites can be present in a biological
sample represent the challenge and the aim of metabolomics. This exceptional diversity
(compared to other levels of biochemical organization, e.g., genome and proteome) also
represents a technological challenge, as there is no technique or combination of techniques
which can be used to englobe metabolite diversity present in a cell, tissue, or biofluid. A
prominent example is lipid diversity in human blood. More than a thousand lipid species
have been reported in human blood plasma, spanning more than nine orders of magnitude,
from oxylipins (such as leukotrienes and prostaglandins) present in low picomolar
concentrations to highly abundant cholesterol measured at millimolar levels [19]. Beyond
the highly conserved primary metabolites required for survival, a variety of “exotic”
compounds, mainly products of specialized plant and/or microbial metabolism, can also
be found in human blood, depending on exposures due to our lifestyle, and the measure-
ment sensitivity, numerous food constituents and additives, phytochemicals, microbial
products, toxins, pollutants, cosmetic products, drugs, and drug metabolites [20].
As underlined in the first part of the introduction, during the past decade, mass
spectrometry-based approaches have made the most significant imprint in metabolomics,
allowing for the measurement of a wide panel of polar and lipid metabolites with high
specificity, sensitivity, accuracy, and precision [21, 22]. Mass spectrometers coupled to
different (and usually complementary) separation techniques (e.g., LC, GC, capillary
20 M. Giera and J. Ivanisevic
electrophoresis, ion mobility) serve today as working horses for the measurement of many
hundreds and up to thousand(s) of metabolites (including lipids!) from only minimal
amounts of biological samples (e.g., 5 μL of blood plasma).
— — — Oli jälleen toinen kerta, jolloin minä luulin että kaikki oli
mennyttä. Silloin hymyili äiti: »Ei mikään ole hukassa!»… Hänellä on
sydämessään ymmärtämisen lahja. Ei kukaan omista sellaista
ymmärtämyksen taitoa, kuin nainen, jolla on tämä taito. Ei kukaan
muu voi nähdä niin syvälle sieluun. Siksi minä jumaloin äitiäni. Hän
on ainoa nainen, jota jumaloin!… Kaikkia muita kohtaan olen oppinut
uskottomaksi hänestä, joka läksi luotani… Häntä minä katson
suurimmaksi kanssarikollisekseni kaikessa mitä olen rikkonut.
Thora Thammers oli hypähtänyt kiveltä, jolla hän oli istunut. Hänen
huulensa olivat valkoiset, kuin kuihtuneet. Hän oli huomannut että
jotain oli tulossa — oli huomannut sen äänestä, joka muuttui
muuntumistaan, kunnes hän ei sitä enää tuntenut.
Nyt vasta tunsi hän hänet samaksi, joka hän oli ollut muinoin. Oli
kuin hän olisi astunut hänen eteensä muuttumatonna pitkäin,
menneitten aikojen takaa — Thora katsoi häneen sanatonna
kauhusta.
— Minun täytyi sanoa se, sanoi hän vihdoin hiljaa, — se oli kerran
sanottava!
Mitä te ajattelette?
— Huomennako?
*****
Jos sitten sattui että sen, johon hänen sydämensä oli kiintynyt,
täytyi matkustaa liian pian, voi hän itkeä vuolaita kyyneliä — oli kuin
kuolema olisi ollut lähellä.
8.
Tuli talvi.
Kesä lupaili käydä oikein hauskaksi. Kaikki olivat hyvillä mielin. Oli
jo alettu maalata koskea.
Oli ehditty ensi päiviin, jolloin oli niin lämmin, että voi täyttä totta
istuskella ulkona.
Thamar rouva oli varsin viehättävä, kun hän nyt istui kysellen
mieheltänsä kaikenmoisia pikkuseikkoja. Hänellä oli erityinen kyky
keksiä tuontapaisia kysymyksiä — eikä kenenkään olisi tarvinnut
niitä arkailla, sillä hän ei koskaan tarkannut vastausta.
Mutta erästä seikkaa ei tirehtööri ollut koskaan oivaltanut. Hänelle
oli niin mieleen tuo, että hänen oli aina päätettävä kaikesta ja että
hänen vaimonsa oli niin tyytyväinen kaikkiin hänen järjestelyihinsä.
Nyt oli kaikki toisin. Nyt oli hänen huolehdittava kaikesta, vaikkei
hän vielä ollut täysin ehtinyt sitä huomata, ollen tuollaisen rakkauden
pauloissa, joka häikäisee ja orjuuttaa, niin kauan kuin se kestää.
Joskus hän sentään tuli ajatelleeksi, kuinka välttämättömiltä hänestä
nyt tuntuivat kaikki nuo vähäiset huolenpidon ilmaukset, joita hän ei
ennen tullut koskaan osoittaneeksi, ja hänen mielessänsä heräsi
kuin hämmästelevä tunne, että hän oli laiminlyönyt jotain, jota ei
voinut enää koskaan hyvittää.
Hän nauroi. Sehän kuului juuri samaan asiaan! Niin, hänen oli ollut
mahdoton olla niitä näkemättä — ja huomaamatta että nykyajan
naiset pyrkivät ratkaisemaan saman tehtävän kuin muinoin
kuningatar Kraaka, mutta vastakkaiseen suuntaan: olla puettu ja
kumminkin alaston. Oli ihmeellistä nähdä, kuinka kunnialliset naiset
kilvan jäljittelivät kuoseja, jotka olivat siveettömäin naisten keksintöjä
tai heitä varten keksittyjä. Erikoisen älykkäältä ei se hänestä
vaikuttanut, mutta kaiketi onkin älykkäisyys jo vanhentunut
ominaisuus… Ennenmuinoinhan poltettiin viisaita naisia!… Erääseen
toiseenkin seikkaan oli hänen huomionsa kiintynyt, nimittäin tuohon
erinomaisen kauhistuttavaan tapaan, millä he kantoivat helmojansa.
Minne olikaan joutunut sulottarien perintö?
Hän jätti lauseensa kesken. Hän tuli ajatelleeksi jotakin, jonka oli
kuullut kerran rouva Thamarilta: että jos ei kenenkään olisi vaikea
olla, niin eihän tietäisikään kuinka hyviä päiviä itse viettää. — Ja
mistä runoilijat sitten kirjoittaisivat? Kaikki ihmiset menehtyisivät
ikävään. — — — Tuota sieti todellakin miettiä!… Ja kun nyt kerran
maailman meno oli järjestetty sillä tavoin… Tietysti oli Luoja
asettanut kaikki sillä tavoin kuin tahtoi sen olevan. —
Mutta väliin voi käydä niin, että lehtori ei enää tiennyt puheensa
alkua eikä loppua, ja nytkään ei hän itse eivätkä naiset tienneet,
kuinka hän oli tullut puhuneeksi — analogia-todistuksista ja sitten
lopettaneeksi puheensa herttaisella myönnytyksellä, että henkisesti
voimakkaat naiset kohoavat ympäristönsä yläpuolelle.
— On kyllä.
Ensi kerran loi nyt Don Miguel rouva Lissiin katseen, joka ei ollut
ehdottomasti ihaileva.
9.