![3
Angelo D’Alessandro (ed.), High-Throughput Metabolomics: Methods and Protocols, Methods in Molecular Biology, vol. 1978,
https://doi.org/10.1007/978-1-4939-9236-2_1, © Springer Science+Business Media, LLC, part of Springer Nature 2019
Chapter 1
Sample Preparation and Reporting Standards
for Metabolomics of Adherent Mammalian Cells
Sarah Hayton, Robert D. Trengove, and Garth L. Maker
Abstract
Metabolomics is an analytical technique that investigates the small molecules present within a biological
system. Metabolomics of cultured cells allows profiling of the metabolic chemicals involved in a cell type-
specific system and the response of that metabolome to external challenges, such as change in environment
or exposure to drugs or toxins. The numerous benefits of in vitro metabolomics include a much greater
control of external variables and reduced ethical concerns. There is potential for metabolomics of mam-
malian cells to uncover new information on mechanisms of action for drugs or toxins or to provide a more
sensitive, human-specific early risk assessment in drug development or toxicology investigations. One way
to achieve stronger biological outcomes from metabolomic data is via the use of these mammalian cultured
cell models, particularly in a high-throughput context. With the sensitivity and quantity of data that
metabolomics is able to provide, it is important to ensure that the sampling techniques have minimal inter-
ference when it comes to interpretation of any observed shifts in the metabolite profile. Here we describe
a sampling procedure designed to ensure that the effects seen in metabolomic analyses are explained fully
by the experimental factor and not other routine culture-specific activities.
Key words Cell culture, Mammalian cells, Adherent cells, Experimental design, Sample preparation,
Quenching, Metabolite extraction, Extracellular, Intracellular, Metabolomics
1 Introduction
Metabolomics studies frequently state that the metabolome is a
closer reflection of the phenotype of an organism, tissue, or cell
than other “omics analyses,” such as proteomics, transcriptomics,
and genomics [1–3]. The small molecules that are shuffled around
the vast network of metabolic pathways in a biological system are
referred to as metabolites, with the “metabolome” made up of the
many thousands of different metabolites and their relative levels of
abundance. The goal of metabolomic profiling is to measure
changes in the metabolome of a given system in response to a chal-
lenge to normal cellular homeostasis, whether physical, chemical,
environmental, or other external stressor.](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-20-320.jpg)
![4
Metabolomics has recently become an attractive application
for untargeted, high-throughput screening analyses [4]. However,
due to the data-rich, hypothesis-generating outcomes of untar-
geted metabolomics, it is understandable that such an approach
might not be attractive to some investigators, particularly if a lim-
ited number of clinical or animal samples are available. In particu-
lar, handling the vast amounts of data can complicate meaningful
biological interpretation of the data. This can be partly addressed
by the use of cultured mammalian cells, as opposed to animal-
based samples, to more easily accommodate re-visiting of the sam-
ple set if novel or previously unknown metabolites are highlighted
by an untargeted study. The number of controls and replicates can
be easily manipulated in design of cell culture-based experiments,
benefiting the development, validation, and standardization of
untargeted, high-throughput metabolomic studies. The impor-
tance of experimental design and standardized reporting require-
ments for these studies of cultured mammalian cells has been
previously reviewed across a large number of published protocols
[5, 6].
The focus of this protocol is to give the investigator a sound,
high-throughput procedure to follow for handling and preparing
cultured cell samples preceding instrumental analysis for metabo-
lomics, which will have minimal possible interference on the
detectable metabolome. It is designed so that any of the multiple
platforms currently used for metabolomics can be applied to the
prepared samples thereafter (e.g., gas chromatography-mass spec-
trometry, liquid chromatography-mass spectrometry). Here we
provide a standardized protocol for untargeted metabolomic anal-
ysis applied to any cultured mammalian cell line, highlighting the
importance of adequate reporting of culture conditions to allow
for meaningful biological interpretation of metabolomic data and
comparison of results across multiple studies.
2 Materials
Cell lines should be sourced from reputable sources such as the
European Collection of Cell Cultures (ECACC) or the American
Type Culture Collection (ATCC). Examples of cell lines which
have been studied using this protocol:
●
● B50 rat neuroblastoma (ECACC 85042302).
●
● HepG2 human hepatocarcinoma (ECACC 85011430; ATCC
HB-8065).
●
● SH-SY5Y human neuroblastoma (ECACC 94030304; ATCC
CRL-2266).
2.1 Cell Lines
Sarah Hayton et al.](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-21-320.jpg)
![9
the samples can be reconstituted in the required mobile phase
solvent(s).
Reporting requirements:
Composition of extraction solution.
Description of extraction procedure (mechanical disruption, col-
lection details, etc.)
If known, expected recovery rate and stability of extracted metabo-
lites (see Note 9).
4 Notes
1. The quantity of cells that needs to be grown per sample extracted
for metabolomic studies should be optimized specifically for the
intended analytical instrument’s detection capabilities and will
vary depending on the specific cell being cultured. The exact
quantity may depend on whether the study is an untargeted
screen for metabolites or if a certain class or selection of metab-
olites is being targeted. For untargeted analysis, this is generally
assessed by observing the total ion chromatogram (TIC) of a
single sample, which is the combined signal of everything mea-
sured in that sample. An optimal number of cells per sample are
chosen based on the maximum number of detectable metabo-
lites that are confidently resolved from the background signal
while also minimizing the occurrence of any high-abundance
peaks that may cause overloading of the instrument detector,
potentially obscuring any less abundant metabolites with similar
chromatographic retention times. Once an optimal cell number
for seeding density is obtained, this should remain consistent
for the duration of the experiment, to minimize data manipula-
tion in later interpretation stages.
2. If a change of medium is required during the incubation period
after cell seeding, care should be taken not to disrupt the course
of cell growth, as much as is practicable. All control samples
should also have medium changed.
3. If multiple passages are required for repeated samples of the
same experimental setup, then the range of passage generations
used should be stated. Depending on the design for the specific
metabolomic experiment, as well as the characteristics of the
cells used, it may not be necessary to conduct experiments
across multiple passages, especially for high-throughput cell
metabolomics applications. This should be a part of the consid-
eration during the researcher’s experimental design [5].
Sample Preparation and Reporting Standards for Metabolomics of Adherent Mammalian…](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-26-320.jpg)
![10
4. The use of quality control samples within a metabolomic study,
particularly for large sample sets, is crucially important for accu-
rate correction and statistical interpretation of the data post-
analysis [7, 8]. For an untargeted study, the QC samples are
typically obtained from a pooled sample of all treatment groups,
set up in parallel to the experimental samples, and collected
simultaneously. If sample volume is sufficient, QC samples can
be obtained by pooling an aliquot of the experimental samples.
QC samples should be placed throughout the sequence during
instrumental analysis, meaning that any drift in metabolite
detection levels can be detected and modelled. This allows any
variation observed from drift phenomena to be removed from
statistical analysis of the data. The appropriate number of QC
samples should be determined in the experimental design
stages, to ensure that adequate sample material is prepared.
5. In cell culture experiments, especially when testing the response
of a system to a physical challenge, the amount of material avail-
able for metabolomic analysis will likely differ between sample
groups. It is therefore important to normalize metabolomic
data post-analysis to remove this sample variation and allow for
meaningful biological interpretation. A number of differing
techniques have been used, including protein or DNA concen-
tration, tissue weight, or cell number [9–12]. The most com-
mon method of normalization of sample variation is to the cell
number upon sample harvesting, which should be considered as
the standard approach when possible. A separate, parallel sam-
ple should be set up along with the sample for metabolite har-
vest and extraction, specifically for cell counting. This is
necessary due to the destructive nature of harvesting for metab-
olomic analysis, as well as the leakage of metabolites encoun-
tered during trypsinization of cells for counting [13, 14].
6. An ideal extraction procedure for metabolomics of cultured
cells should immediately quench metabolism in order to extract
and collect all metabolites present at the end point of the exper-
iment [15, 16]. PBS is a buffered, isotonic solution, so it causes
less membrane leakage, thereby reducing loss of intracellular
metabolites compared to other solvents that have been used in
quenching, such as methanol.
7. The composition of extraction solution can be altered depend-
ing on the chemical properties of the specific metabolites tar-
geted for analysis, to maximize recovery. The suggested
extraction solution, methanol and water in an 80:20 ratio, gives
a single-phase extraction supernatant and allows for the whole
extract to be collected. Also common is the use of a dual-phase
(polar and nonpolar) extraction using methanol, chloroform,
and water (or acetonitrile). The ratio of these solvents can vary
but is most commonly used at 1:1 methanol and chloroform,
Sarah Hayton et al.](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-27-320.jpg)
![13
Angelo D’Alessandro (ed.), High-Throughput Metabolomics: Methods and Protocols, Methods in Molecular Biology, vol. 1978,
https://doi.org/10.1007/978-1-4939-9236-2_2, © Springer Science+Business Media, LLC, part of Springer Nature 2019
Chapter 2
High-Throughput Metabolomics: Isocratic and Gradient
Mass Spectrometry-Based Methods
Travis Nemkov, Julie A. Reisz, Sarah Gehrke, Kirk C. Hansen,
and Angelo D’Alessandro
Abstract
Metabolomics has emerged in the past decade as a highly attractive and impactful technique for phenotype-
level profiling in diverse biological applications. Most recently, the dual developments of high-throughput
analytical techniques along with dramatically increased sensitivity of high-resolution mass spectrometers
have enabled the routine analysis of hundreds of unique samples per day. We have previously reported a
robust 3 min isocratic metabolomics platform for the quantification of amino acids and the key pathways
of central carbon and nitrogen metabolism. Building on this work, we describe here a 5 min reverse phase
gradient followed by global, untargeted profiling of the hydrophilic metabolome. In addition to observing
those metabolites measured in the 3 min run, the use of the longer gradient run here also allows for cover-
age of less polar compounds such as fatty acids and acylcarnitines, both key players in mitochondrial and
lipid metabolism, without a significant sacrifice in throughput.
Key words Untargeted metabolomics, Mass spectrometry, Isocratic, Gradient, High-throughput
1 Introduction
In the last two decades, the cost of sequencing the human genome
[1] has dropped from ~$300 million US dollars to ~$1000–1500
by use of whole-exome sequencing (www.genome.gov/sequenc-
ingcosts/). Technological advances have made the genome ame-
nable to large-scale studies which foster personalized medicine
initiatives around the globe. The introduction of high-throughput
metabolomics technologies promises to achieve a similar goal.
Similarly to high-throughput genome sequencing tools, it is
easy to anticipate how high-throughput metabolomics applications
will soon become instrumental for the characterization of clinically
relevant metabolic biomarkers requiring a final validation phase in
large prospective cohorts, such as in the case of large-scale person-
alized medicine initiative studies. For example, high-throughput
metabolomics approaches can be theoretically used to quantify](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-30-320.jpg)
![14
metabolic markers of disease such as homocysteine levels in homo-
cystinuria, an inborn error of metabolism [2], glucose levels in dia-
betes [3], and lactate [4] and succinate [5] levels as readouts for
base deficit and tissue hypoxia in trauma. Historically, targeted
metabolomics approaches that have been translated into clinical
practice are based on multiple reaction monitoring-mass spectrom-
etry (MRM-MS) [6], a highly sensitive method that requires pre-
selection of molecular targets of interest, therefore affording
detection and quantitation of only a handful of small molecules at
a time. The introduction of metabolomics strategies based on
high-resolution quadrupole-time of flight (QTOF) or quadrupole-
Orbitrap-
based instruments has enabled investigators to perform
untargeted “discovery mode” metabolomics analyses while simul-
taneously quantifying metabolites of interest against external cali-
bration curves or internal stable isotope-labeled standards.
Over the past 5 years, several approaches have been proposed
to achieve high-throughput metabolomics workflows, either based
on flow-injection mass spectrometry [7], ultra-high-pressure liq-
uid chromatography (UHPLC) interfaced to MS [8, 9], or matrix-
assisted laser desorption/ionization imaging mass spectrometry
(IMS) [10]. These approaches have made the detection and quan-
tification of over 100,000 features and thousands of named metab-
olites possible in less than 5 min through direct injection or
isocratic/gradient-based chromatography or using in situ imaging
approaches. Further advancements in terms of throughput and
sensitivity are anticipated upon the recent introduction of micro-
fluidic capillary electrophoresis coupled to MS for metabolomics
applications [11, 12]. Optimization of rapid extraction protocols
[13], automation [14], and, predictably, low-cost robotization of
sample extraction protocols [15] would enable affordable process-
ing of hundreds to a thousand samples per day in a cost-effective
manner.
Recently, we described a high-throughput isocratic
chromatography-
based method for rapid quantitation of hydro-
philic compounds from the central carbon and nitrogen pathways
[8], including amino acids [16]. Below we describe an evolution of
this method, where mobile phases are coupled with agents that
improve chromatographic separation and ease anionic ionization
of high-energy phosphate compounds (e.g., nucleoside triphos-
phates) to facilitate detection in negative ion mode. In addition,
we describe a high-throughput gradient-based method that in part
addresses the main limitation of the isocratic 3 min method, by
making some classes of hydrophobic compounds (including acyl-
conjugated carnitines, free fatty acids, and bile acids) amenable to
detection and quantification. This method affords the rapid quan-
titation of up to 600,000 features from a clinically relevant sample
(e.g., blood extracts), resulting in hundreds to thousands of named
compounds (level of confidence based on high-resolution intact
Travis Nemkov et al.](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-31-320.jpg)
![15
mass, transition fingerprints, and comparison of retention times
against our in-house standard library). Given the rapidity of these
approaches, 450 samples can be processed per day, making it real-
istic to anticipate the potential translation of such technology into
the field of clinical biochemistry. These approaches are also com-
patible with the use of stable isotope-labeled internal standards,
which allow for the correction of matrix-dependent ion suppres-
sion effects, as well as to monitor the stability of chromatographic
peak shape and retention times and the quality of signal intensity at
the MS and MS/MS level [8]. The use of stable isotope-labeled
standards also streamlines quantitation of compounds of interest,
thus normalizing matrix or batch effects as well as instrument-
dependent variability across platforms, provided the concentra-
tions of the labeled standards and the endogenous metabolite from
the tested matrix are within the linearity range of the MS instru-
ment. Stable isotope-labeled standards can be leveraged to quan-
tifymetabolicfluxesbydeterminingratiosofspiked-inisotopologues
against those deriving from alternative metabolism of stable iso-
tope tracers in fluxomics experiments, in like fashion to what we
described recently for lactate [17] and alanine [18] synthesis
downstream to [1,2,3-13
C3]glucose catabolism via glycolysis or the
pentose phosphate pathway in red blood cells.
2 Materials
Prepare all solutions using highest-quality solvents and reagents.
We recommend LC-MS grade solvents, such as Optima™ (Fisher).
Solutions can be stored at room temperature unless otherwise
stated.
1. Benchtop centrifuge capable of 18,213 × g with refrigeration
to 4 °C.
2. Refrigeration space (4 °C) with power outlet(s) to accommo-
date vortex(es), often a cold room or chromatography
refrigerator.
3. Vortex with foam microcentrifuge tube insert.
1. Water.
2. Water containing 0.1% v/v formic acid.
3. Acetonitrile.
4. Acetonitrile containing 0.1% v/v formic acid.
5. Ammonium acetate.
6. Sterile disposable vacuum filtration system (0.22 μm
membrane).
2.1 General
2.2 Mobile Phases
for Ultra-High-
Pressure Liquid
Chromatography
High-Throughput Metabolomics](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-32-320.jpg)
![23
4 Data Analysis
1. Quantify metabolite peak areas using software of choice. We
utilize Maven, which first requires file conversion from .raw to
.mzXML (accomplished using RawConverter), and assign
metabolite names using the KEGG database. Table 1 contains
a list of observed retention times for acylcarnitines, fatty acids,
and oxylipins observed on the 5 min gradient method.
Reported times are an average of 3–10 measurements in vari-
ous matrices (plasma, red blood cells, saliva, pancreatic tissue)
and commercial standards prepared in the background of
plasma or RBCs (see Note 7).
2. Utilize QC sample data to assess instrument performance and
stability throughout analytical runs. Coefficients of variation
(CV) for named metabolites should be 20%. If CV values are
above this threshold while other readouts (pump pressure,
total signal level, etc.) do not suggest technical issues with
instrument performance, consider the use of a batch- or drift-
correction software. When needed, we utilize the MetaboDrift
normalization algorithms separately on positive and negative
ion mode raw peak area data. Once the peak areas are appro-
priately normalized as needed for each polarity, the data can be
merged for subsequent calculations and visualizations.
3. Use the ratio of peak areas for stable isotope-labeled standards
and their unlabeled (light) counterparts to determine the abso-
lute values of endogenous metabolites in the extract. Back-
calculate the concentrations in the biological samples using the
following equations:
For biofluids:
Light Area Area Heavy DF
light heavy
[ ] = ( )[ ]×
/
where DF is dilution factor—10 for RBCs and blood and 25 for
serum and plasma as described above (see Note 8).
For tissues:
Light Area Area Heavy EC
light heavy
[ ] = ( )[ ]×
/ /
1
where EC is extraction concentration, described here as i15 mg/
mL (see Note 9).
For cells:
Light Area Area Heavy EC
light heavy
[ ] = ( )[ ]×
/ /
1
where EC is extraction concentration, described here as 2 × 106
cells/mL (see Note 10).
High-Throughput Metabolomics](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-40-320.jpg)
![26
in different biological models. Mol BioSyst
7(4):1024–1032
7. Zampieri M, Sekar K, Zamboni N, Sauer U
(2017) Frontiers of high-throughput metabo-
lomics. Curr Opin Chem Biol 36:15–23
8. Nemkov T, Hansen KC, D’Alessandro A
(2017) A three-minute method for high-
throughput quantitative metabolomics and
quantitative tracing experiments of central car-
bon and nitrogen pathways. Rapid Commun.
Mass Spectrom 31(8):663–673
9. Lewis MR, Pearce JTM, Spagou K, Green M,
Dona AC, Yuen AHY, David M, Berry DJ,
Chappell K, Horneffer-van der Sluis V, Shaw R,
Lovestone S, Elliott P, Shockcor J, Lindon JC,
Cloarec O, Takats Z, Holmes E, Development
NJK (2016) Application of ultra-performance
liquid chromatography-TOF MS for precision
large scale urinary metabolic phenotyping.
Anal Chem 88(18):9004–9013
10. Miura D, Fujimura Y, Yamato M, Hyodo F,
Utsumi H, Tachibana H, Wariishi H (2010)
Ultrahighly sensitive in situ metabolomic imag-
ing for visualizing spatiotemporal metabolic
behaviors. Anal Chem 82(23):9789–9796
11. Gasilova N, Yu Q, Qiao L, Girault HH (2014)
On-Chip spyhole mass spectrometry for
droplet-
based microfluidics. Angew Chem Int
Ed 53(17):4408–4412
12. Wouters S, De Vos J, Dores-Sousa JL, Wouters
B, Desmet G, Eeltink S (2017) Prototyping of
thermoplastic microfluidic chips and their
application in high-performance liquid chro-
matography separations of small molecules.
J Chromatogr A 1523:224–233
13. Gehrke S, Reisz JA, Nemkov T, Hansen KC,
D’Alessandro A (2017) Characterization of
rapid extraction protocols for high-throughput
metabolomics. Rapid Commun Mass Spectrom
31(17):1445–1452
14. Dias DA, Koal T (2016) Progress in metabolo-
mics standardisation and its significance in
future clinical laboratory medicine. EJIFCC
27(4):331–343
15. Liquid-handling Lego robots and experiments
for STEM education and research. [date
unknown]. http://journals.plos.org/plosbiol-
o g y / a r t i c l e ? i d = 1 0 . 1 3 7 1 / j o u r n a l .
pbio.2001413. Accessed 19 July 2017
16. Nemkov T, D’Alessandro A, Hansen KC
(2015) Three-minute method for amino acid
analysis by UHPLC and high-resolution quad-
rupole orbitrap mass spectrometry. Amino
Acids 47(11):2345–2357
17. Reisz JA, Wither MJ, Dzieciatkowska M,
Nemkov T, Issaian A, Yoshida T, Dunham AJ,
Hill RC, Hansen KC, D’Alessandro A (2016)
Oxidative modifications of glyceraldehyde
3-phosphate dehydrogenase regulate metabolic
reprogramming of stored red blood cells.
Blood 128(12):e32–e42
18. D’alessandro A, Dzieciatkowska M, Nemkov
T, Hansen KC (2017) Red blood cell pro-
teomics update: is there more to discover?
Blood Transfus 15(2):182–187
Travis Nemkov et al.](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-43-320.jpg)
![27
Angelo D’Alessandro (ed.), High-Throughput Metabolomics: Methods and Protocols, Methods in Molecular Biology, vol. 1978,
https://doi.org/10.1007/978-1-4939-9236-2_3, © Springer Science+Business Media, LLC, part of Springer Nature 2019
Chapter 3
High-Throughput Metabolomics Based on Direct Mass
Spectrometry Analysis in Biomedical Research
Raúl González-Domínguez, Álvaro González-Domínguez,
Carmen Segundo, Mónica Schwarz, Ana Sayago, Rosa María Mateos,
Enrique Durán-Guerrero, Alfonso María Lechuga-Sancho,
and Ángeles Fernández-Recamales
Abstract
Metabolomics based on direct mass spectrometry analysis shows a great potential in biomedical research
because of its high-throughput screening capability and wide metabolome coverage. This chapter contains
detailed protocols to perform comprehensive metabolomic fingerprinting of multiple biological samples
(serum, plasma, urine, brain, liver, spleen, thymus) by using complementary analytical platforms. The most
important issues to be considered are discussed, including sample treatment, metabolomic analysis, raw
data preprocessing, and data analysis.
Key words Metabolomics, Direct mass spectrometry analysis, Direct infusion, Flow injection,
High-throughput
1 Introduction
Nontargeted metabolomic analysis is very challenging because of
the difficulty of simultaneously characterizing the entire set of
metabolites present in biological systems in a comprehensive man-
ner. To this end, nuclear magnetic resonance (NMR) and mass
spectrometry (MS) are nowadays the most commonly employed
metabolomic platforms, with complementary analytical perfor-
mance and applicability [1, 2]. NMR is a rapid, nondestructive,
and very reproducible technique, which requires a relatively simple
sample pre-treatment step and provides important structural infor-
mation. However, its low sensitivity and spectral resolution seri-
ously compromise the detection of individual metabolites in
complex samples. On the other hand, the combination of mass
spectrometry with separation techniques, such as liquid chroma-
tography (LC), gas chromatography (GC), or capillary
1.1 The Potential
of Direct Mass
Spectrometry Analysis
for High-Throughput
Metabolomics](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-44-320.jpg)
![28
electrophoresis (CE), has demonstrated a great utility to perform
both qualitative and quantitative metabolomic analyses. As a con-
sequence of the inherent analytical bias introduced by these separa-
tion methods, the combination of complementary approaches has
been postulated as the most suitable strategy to maximize metabo-
lomic coverage [3, 4]. Nevertheless, the application of these
hyphenated approaches significantly increases overall analysis times,
which in turn can negatively affect technical stability (e.g., analyti-
cal drifts in mass accuracy and/or sensitivity).
To solve these limitations, the use of direct mass spectrometry
analysis has emerged in recent years for high-throughput metabo-
lomics [5, 6]. In this approach, sample extracts are directly intro-
ducedintothemassspectrometerwithoutpreviouschromatographic
or electrophoretic separation. For this purpose, the simplest instru-
mental configuration is direct infusion mass spectrometry (DIMS),
which employs a syringe pump to constantly deliver samples into
the MS system. A second alternative is flow injection mass spec-
trometry (FI-MS), based on the infusion of samples as a plug into
a stream of solvent delivered by a LC pump. In general, DIMS-
based metabolomics provides higher sensitivity and reproducibil-
ity, but as counterpart, FI-MS is a less sample volume consuming
technique and enables automation. Direct mass spectrometry anal-
ysis exhibits multiple advantages over conventional MS hyphen-
ated platforms thanks to the lack of a preceding separation step.
First, it is noteworthy the higher technical simplicity of DI/FI-MS
approaches, which are unaffected by common troubles usually
found in chromatography such as column clogging and overpres-
sure, changes in retention times and peak broadening due to sta-
tionary phase deterioration, or irregular baseline and signal drifts
due to mobile phase contamination or leaks in the LC system,
among other issues. Furthermore, direct introduction of samples
without applying a previous chromatographic/electrophoretic
step, which is inherently biased toward specific metabolite classes
depending on the separation mode employed, usually provides
more comprehensive metabolome coverage. Therefore, direct MS
analysis has demonstrated huge potential for high-throughput
metabolomic screening, of particular interest when dealing with
large sample populations [6].
Metabolomics based on direct MS analysis has gained great popu-
larity for mapping metabolic alterations associated with disease
pathogenesis and progression in a holistic manner. For instance,
various authors have previously described the application of MS
fingerprinting to investigate the underlying pathology of different
oncological disorders, including lung cancer [7–9], prostate cancer
[10], kidney cancer [11], pancreatic cancer [12], and colorectal
cancer [13–15]. Thereby, it has been demonstrated the occurrence
1.2 Application
of Direct MS-Based
Metabolomics
in Biomedical
Research
Raúl González-Domínguez et al.](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-45-320.jpg)
![29
of multiple metabolic dysregulations affecting to a wide range of
pathways depending on the cancer type and stage, such as impaired
energy metabolism, altered homeostasis of lipids, oxidative stress,
and many others. DI/FI-MS-based metabolomics has also been
employed to identify the characteristic metabotype associated with
abnormal glucose metabolism in patients with impaired glucose
tolerance [16] and insulin resistance [17] and in volunteers sub-
jected to an oral sugar challenge [18]. Other studies performed on
the last years dealt with the discovery of potential diagnostic bio-
markers for Chagas disease [19], schistosomiasis [20], aspergillosis
[21], Crohn’s disease [22], induced stress [23], and preeclampsia
[24]. However, it should be noted that the application of direct
MS has mainly focused on the metabolomic study of Alzheimer’s
disease (AD). González-Domínguez et al. optimized a metabolo-
mic approach based on a two-step extraction procedure and subse-
quent DIMS analysis of serum samples to elucidate pathological
hallmarks associated with AD development and progression, in
both human cohorts [25, 26] and transgenic APP×PS1 mouse
models [27, 28]. Later, this methodology was adapted to get
deeper insights into lipidomic [29] and phospholipidomic [30]
alterations occurring in this neurodegenerative disorder.
Furthermore, authors also described the implementation of this
DIMS-based metabolomic platform for the analysis of other bio-
logical matrices, such as urine [31], brain [32], and other periph-
eral organs (i.e., liver, kidney, thymus, spleen) [33]. Thereby, a
comprehensive characterization of AD-related metabolic distur-
bances was accomplished, comprising multiple impairments in the
homeostasis of neurotransmitters, various lipid classes (e.g., phos-
pholipids, glycerolipids, cholesterol, fatty acids), and antioxidants,
among others. Interestingly, these findings were then validated by
applying orthogonal hyphenated approaches, such as UHPLC-MS
[34–38], GC-MS [35–39], and CE-MS [40], thus demonstrating
the reliability of direct MS for metabolomic fingerprinting. In this
line, Lin et al. reported the use of DIMS techniques to address
metabolic alterations in the hippocampus [41] and cerebellum
[42] of CRND8 transgenic mice, revealing significant perturba-
tions in the regulation of neuroinflammatory processes. Finally, it
is also noteworthy the recent development of a novel metabolomic
approach based on flow injection analysis using an atmospheric
pressure photoionization mass spectrometer (FI-APPI-MS) with
the aim to complement the ionization capabilities of the electro-
spray source (ESI), usually employed in DIMS-based metabolo-
mics [43]. This methodology was successfully applied to serum
samples from AD patients [43] and APP×PS1 transgenic mice
[27], which provided complementary findings to those obtained
with DI-ESI-MS platforms.
Metabolomics Based on Direct MS Analysis](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-46-320.jpg)
![31
QSTAR XL Hybrid system (Applied Biosystems), equipped with
integrated syringe pump, Accela LC system (Thermo Fisher
Scientific), and syringe pump model KDS 100 (KD Scientific); (2)
Q-TOF mass spectrometer, model Xevo G2-S, coupled to an
ACQUITY UPLC™ system (Waters). For accurate mass measure-
ment, the QSTAR XL system is daily calibrated using renin and
taurocholic acid in positive and negative ion modes, respectively.
In the Xevo G2-S mass spectrometer, all spectra are acquired using
a reference lock mass (leucine enkephalin) to ensure accuracy and
reproducibility.
A cryogenic homogenizer model Freezer/Mill 6770 (SPEX
SamplePrep) and a pellet mixer (VWR International) are used to
perform metabolomic extraction of tissues. All samples are centri-
fuged in a centrifuge model Eppendorf 5804R. A vacuum mani-
fold (Visiprep, Supelco) is employed to perform solid phase
extraction (SPE) of urine samples.
3 Methods
A two-step extraction protocol is applied to serum and plasma
samples with the aim to fractionate the blood metabolome in two
complementary extracts: (1) aqueous extract containing more
polar compounds (i.e., low molecular weight metabolites and
phospholipids) and (2) lipophilic extract, mainly composed of neu-
tral lipids [25].
1. Thaw serum/plasma samples on an ice bath.
2. Add 400 μL of methanol/ethanol (1:1, v/v) to 100 μL of
serum/plasma into an Eppendorf tube placed on an ice bath.
3. Stir the mixture during 5 min, using a vortex mixed or an
orbital shaker (at 4 °C if possible), in order to precipitate
proteins.
4. Centrifuge samples at 4000 × g for 10 min at 4 °C.
5. Transfer the supernatant to a new tube, take it to dryness under
nitrogen stream (see Note 3), and reconstitute the resulting
residue with 100 μL of methanol/water (80:20, v/v) contain-
ing 0.1% (v/v) formic acid (polar extract) (see Note 4).
6. Add 400 μL of chloroform/methanol (1:1, v/v) to the protein
precipitate obtained in step 3 placed on an ice bath.
7. Shake vigorously during 5 min using a vortex mixer (at 4 °C if
possible).
8. Centrifuge samples at 10,000 × g for 10 min at 4 °C.
3.1 Sample
Treatment
3.1.1 Metabolomic
Extraction of Serum/
Plasma Samples
Metabolomics Based on Direct MS Analysis](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-48-320.jpg)
![32
9. Take to dryness the supernatant under nitrogen stream (see
Note 3) and reconstitute with 100 μL of
dichloromethane/
methanol (60:40, v/v) containing 0.1% (v/v) formic acid and
10 mM ammonium formate (lipophilic extract) (see Note 4).
To get a deeper insight into the blood circulating lipidome, serum/
plasma samples are treated following a modification of the Bligh-
Dyer extraction protocol, which allows analyzing neutral lipids
(e.g., diglycerides, triglycerides, cholesterol derivatives), usually
not detected by means of conventional metabolomic approaches
[29].
1. Thaw serum/plasma samples on an ice bath.
2. Mix 50 μL of serum/plasma with 150 μL of methanol contain-
ing 30 mM ammonium acetate in an Eppendorf tube placed
on an ice bath.
3. Shake vigorously during 1 min using a vortex mixer (at 4 °C if
possible).
4. Add 200 μL of pure chloroform.
5. Shake vigorously during 1 min using a vortex mixer (at 4 °C if
possible).
6. Centrifuge samples at 10,000 × g for 10 min at 4 °C.
7. Transfer the organic layer to a new tube and keep for further
analysis.
Direct MS fingerprinting of urine samples requires the application
of a pre-treatment step with the aim to reduce the high salinity of
this biological fluid, which may compromise further analysis due to
matrix effects (e.g., ion suppression). For this purpose, two differ-
ent approaches can be applied after thawing urine samples on an
ice bath, as describe elsewhere [31].
1. Sample dilution.
(a)
Centrifuge urine samples at 4000 × g for 10 min at 4 °C.
(b)
Dilute tenfold with methanol/water (1:1, v/v).
2. Mixed mode solid phase extraction (MM-SPE).
(a)
Condition SPE cartridges (Isolute Multimode, 500 mg of
sorbent) with 2 mL of pure methanol.
(b)
Load 1.5 mL of urine, previously centrifuged at 4000 × g
for 10 min at 4 °C.
(c)
Clean the SPE cartridge with 2 mL of deionized water.
(d)
Carry out the elution with: (1) 0.5 mL of methanol, (2)
0.5 mL of 10 mM ammonium acetate (pH = 3) in metha-
nol, and (3) 0.5 mL of 5% (v/v) ammonia in methanol.
3.1.2 Lipidomic
Extraction of Serum/
Plasma Samples
3.1.3 Metabolomic
Extraction of Urine
Samples
Raúl González-Domínguez et al.](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-49-320.jpg)
![33
All eluents are pumped through the SPE cartridges by using a vacuum
manifold (Visiprep, Supelco).
Extraction of tissue samples, including the hippocampus, brain
cortex, cerebellum, striatum, olfactory bulbs, liver, kidney, spleen,
and thymus, is carried out in accordance with the methodology
previously optimized by González-Domínguez et al. [33, 36].
1. Cryo-homogenize tissues to obtain a fine powder using a
Freezer/Mills 6770 homogenizer during 30 s at rate of 10
strokes per second (see Note 5).
2. Weigh 30 mg of homogenized tissue (or the entire sample for
smaller organs) in 1.5 mL Eppendorf tubes. Samples are kept
frozen until the addition of the extraction solvent.
3. Add 10 μL mg−1
of pre-cooled 0.1% (v/v) formic acid in meth-
anol (−20 °C).
4. Use a pellet mixer (VWR international) to disrupt cells during
2 min. Perform this extraction step inside an ice bath to avoid
sample heating by friction.
5. Centrifuge at 10,000 × g for 10 min at 4 °C.
6. Transfer the supernatant to a new tube and keep for further
analysis (polar extract).
7. Add 10 μL mg−1
of pre-cooled chloroform/methanol (2:1,
v/v), containing 0.1% (v/v) formic acid and 10 mM ammo-
nium formate (−20 °C), to the pellet obtained in step 5 (see
Note 6).
8. Repeat steps 4–6 to obtain the corresponding lipophilic
extracts.
For each biological matrix, prepare quality control (QC) samples
by pooling equal volumes of each individual sample. The analysis
of these QC samples allows monitoring the stability and perfor-
mance of the system along the analysis period [44]. These samples
must be analyzed at the start of the run in order to equilibrate the
analytical system as well as at intermittent points throughout the
sequence to monitor system stability.
Sample extracts are directly infused into the QTOF-MS system
(QSTAR XL, Applied Biosystems) by using an integrated syringe
pump operating at 5 μL min−1
flow rate. Mass spectra are obtained
by electrospray ionization (ESI) in positive and negative ionization
modes in separate runs, as described elsewhere [25, 27, 32, 33].
Briefly, full-scan spectra are acquired for 0.2 min in the m/z range
50–1100 Da, with 1.005 s scan time. The source temperature is
maintained at 60 °C, and high-purity nitrogen is used as curtain
and nebulizer gas at flow rates of 1.13 L min−1
and 1.56 L min−1
,
3.1.4 Metabolomic
Extraction of Tissue
Samples
3.1.5 Preparation
of Quality Control Samples
3.2 Direct MS
Metabolomic
Fingerprinting
3.2.1 Direct Infusion
Electrospray Ionization
Mass Spectrometry
(DI-ESI-MS)
Metabolomics Based on Direct MS Analysis](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-50-320.jpg)
![34
respectively. The ion spray voltage (IS), declustering potential
(DP), and focusing potential are fixed at 3300/−4000 V,
60/−100 V, and 250/−250 V in positive and negative ion modes,
respectively.
For FI-ESI-MS analysis, sample extracts are introduced into the
mass spectrometer (Xevo G2-S) by flow injection using an
ACQUITY UPLC™
system [18]. Methanol containing 0.1% (v/v)
formic acid and 10 mM ammonium formate is delivered at
200 μL min−1
as flow injection solvent. Mass spectra are obtained
in positive and negative ion modes, by injecting 5 μL of sample.
Full-scan spectra are acquired in the m/z range 50–1100 Da dur-
ing 1.5 min (scan time 0.5 s). Capillary and cone voltages are set at
3.0 kV and 30 V, respectively, and source temperature is main-
tained at 120 °C. Desolvation gas flow (high-purity nitrogen) is
fixed at 250 L h−1
, at 150 °C of temperature.
The application of FI-APPI-MS fingerprinting requires the cou-
pling of a high-resolution mass spectrometer equipped with an
atmospheric pressure photoionization source (QSTAR XL), a LC
system for flow injection of samples (Accela), and a syringe pump
to deliver the dopant reagent for photospray ionization. Following
the previously optimized protocol, methanol is used as flow injec-
tion solvent at 50/100 μL min−1
in positive and negative ion
modes, respectively [43]. On the other hand, toluene is delivered
at 20/40 μL min−1
as photoionization dopant, in both ionization
modes. Mass spectra are obtained in positive and negative ion
modes by injecting 5 μL of sample. Full-scan spectra are acquired
in the m/z range 50–1100 Da, with 1.005 s of scan time. The ion
spray voltage (IS), declustering potential (DP), and focusing
potential are fixed at 1500/−2300 V, 50/−50 V, and 250/−250 V
in positive and negative ion modes, respectively. The source tem-
perature is maintained at 400 °C, and the gas flows (high-purity
nitrogen) are fixed at 1.13 L min−1
for curtain gas, 1.50 L min−1
for nebulizer gas, 3.0 L min−1
for heater gas, and 1 L min−1
for
lamp gas.
In direct MS-based metabolomics, data preprocessing only implies
the application of a single peak detection step in order to convert
original files into a two-dimensional data matrix of spectral peaks
and their intensities. For this purpose, different software must be
employed depending on the MS system employed to acquire
metabolomic fingerprints.
1. Metabolomic data obtained by DI-ESI-MS and FI-APPI-MS
analysis are submitted to peak detection by using the
MarkerView™ software (Applied Biosystems). For this, all
peaks above the noise level (10 counts, determined empirically
3.2.2 Flow Injection
Electrospray Ionization
Mass Spectrometry
(FI-ESI-MS)
3.2.3 Flow Injection
Atmospheric Pressure
Photoionization Mass
Spectrometry (FI-APPI-MS)
3.3 Raw Data
Processing
Raúl González-Domínguez et al.](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-51-320.jpg)
![35
from experimental spectra) are selected and binned in intervals
of 0.1 Da. For FI-MS fingerprints, this processing step is lim-
ited to scans within the apex of infusion profiles [43].
2. Metabolomic data obtained by FI-ESI-MS analysis are sub-
mitted to peak detection by using the MarkerLynx™ soft-
ware (Waters). To this end, all peaks above the noise level
(200 counts, determined empirically from experimental
spectra) are selected and binned in intervals of 0.02 Da. This
processing step is limited to scans within the apex of infusion
profiles [18].
Multiple statistical packages can be used to analyze metabolomic
data. In this section, we focus on the MetaboAnalyst 3.0 web tool
(http://www.metaboanalyst.ca/), the SIMCA-P™ software (ver-
sion 11.5, UMetrics AB, Umeå, Sweden), and the STATISTICA
8.0 software (StatSoft, Tulsa, USA). Thereby, a conventional pipe-
line for metabolomic data analysis involves the following steps:
1. Data filtering by choosing masses present in at least 50% of
samples.
2. Estimation of missing values by means of the k-nearest neigh-
bor (KNN) algorithm.
3. Data filtering based on the interquartile range (IQR) in order
to remove variables showing little variance within the study
population.
4. Data normalization according to the total area sum.
5. Pareto scaling to reduce the relative importance of larger val-
ues, and logarithmic transformation in order to approximate a
normal distribution [45].
6. Application of multivariate statistical techniques with the aim
to visualize general trends and detect discriminant patterns
among the study groups: principal component analysis (PCA),
partial least squares discriminant analysis (PLS-DA).
7. Application of univariate statistical techniques to identify dis-
criminant metabolites among the study groups: t-test or one-
way analysis of variance (ANOVA) (see Note 7), with Bonferroni
or false discovery rate (FDR) correction for multiple testing.
4 Notes
1. Various anticoagulants can be used to obtain plasma samples,
including citrate, ethylenediaminetetraacetic acid (EDTA), and
heparin. Advantages and drawbacks of using each anticoagulant
have recently been reviewed by Hernandes et al. [46].
3.4 Data Analysis
Metabolomics Based on Direct MS Analysis](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-52-320.jpg)
![36
2. Preservatives can be added to enhance the stability of urine
samples (e.g., borate, sodium azide), which are prone to bacte-
rial contamination during collection and storage. However,
the general recommendation of the European Consensus
Expert Group is to avoid the use of additives [47].
3. The evaporation of sample extracts can also be carried out by
using vacuum concentrators (e.g., SpeedVac), if available.
4. Depending on the sensitivity of the MS system employed, sam-
ple extracts can be directly analyzed without performing a pre-
concentration step.
5. Smaller organs (e.g., hippocampus, striatum, olfactory bulbs)
can be directly extracted without prior cryo-homogenization.
6. This second extraction step is of great interest for studying
peripheral organs (i.e., liver, kidney, spleen, thymus), due to
the high content of neutral lipids in these tissues. However,
steps 7 and 8 can be omitted when brain tissue is analyzed, as
previously reported [32, 36].
7. Nonparametric methods must be used when variables do not
show normal distribution (checked by normal probability plots)
and variances are not homogeneous (checked by Levene’s test).
References
1. Emwas AHM, Salek RM, Griffin JL, Merzaban
J (2013) NMR based metabolomics in human
disease diagnosis: applications, limitations, and
recommendations. Metabolomics
9:1048–1072
2. Theodoridis G, Gika HG, Wilson ID (2011)
Mass spectrometry-based holistic analytical
approaches for metabolite profiling in systems
biology studies. Mass Spectrom Rev
30:884–906
3. González-Domínguez R, Sayago A, Fernández-
Recamales A (2017) Metabolomics in
Alzheimer’s disease: the need of complemen-
tary analytical platforms for the identification
of biomarkers to unravel the underlying pathol-
ogy. J Chromatogr B Analyt Technol Biomed
Life Sci 1071:75–92
4. Gonzalez-Dominguez A, Duran-Guerrero E,
Fernandez-Recamales A, Lechuga-Sancho AM,
Sayago A, Schwarz M, Segundo C, Gonzalez-
Dominguez R (2017) An overview on the
importance of combining complementary ana-
lytical platforms in metabolomic research. Curr
Top Med Chem 17:3289–3295
5. Draper J, Lloyd AJ, Goodacre R, Beckmann M
(2013) Flow infusion electrospray ionisation
mass spectrometry for high throughput, non-
targeted metabolite fingerprinting: a review.
Metabolomics 9:S4–S29
6. González-Domínguez R, Sayago A, Fernández-
Recamales A (2017) Direct infusion mass spec-
trometry for metabolomic phenotyping of
diseases. Bioanalysis 9:131–148
7. Guo Y, Wang X, Qiu L, Qin X, Liu H, Wang Y,
Li F, Wang X, Chen G, Song G, Li F, Guo S, Li
Z (2012) Probing gender-specific lipid metab-
olites and diagnostic biomarkers for lung can-
cer using Fourier transform ion cyclotron
resonance mass spectrometry. Clin Chim Acta
414:135–141
8. Cameron SJ, Lewis KE, Beckmann M, Allison
GG, Ghosal R, Lewis PD, Mur LA (2016) The
metabolomic detection of lung cancer bio-
markers in sputum. Lung Cancer 94:88–95
9. Lokhov PG, Kharybin ON, Archakov AI
(2012) Diagnosis of lung cancer based on
direct-infusion electrospray mass spectrometry
of blood plasma metabolites. Int J Mass
Spectrom 309:200–205
10. Lokhov PG, Dashtiev MI, Moshkovskii SA,
Archakov AI (2010) Metabolite profiling of
blood plasma of patients with prostate cancer.
Metabolomics 6:156–163
Raúl González-Domínguez et al.](https://image.slidesharecdn.com/5242326-250607152052-46177014/85/Highthroughput-Metabolomics-Methods-And-Protocols-1st-Ed-Angelo-Dalessandro-53-320.jpg)
Highthroughput Metabolomics Methods And Protocols 1st Ed Angelo Dalessandro
- 1. Highthroughput Metabolomics Methods And Protocols 1st Ed Angelo Dalessandro download https://ebookbell.com/product/highthroughput-metabolomics- methods-and-protocols-1st-ed-angelo-dalessandro-10484652 Explore and download more ebooks at ebookbell.com
- 2. Here are some recommended products that we believe you will be interested in. You can click the link to download. Highthroughput Mass Spectrometry In Drug Discovery 1 Chang Liu https://ebookbell.com/product/highthroughput-mass-spectrometry-in- drug-discovery-1-chang-liu-50753706 High Throughput Mass Spectrometry In Drug Discovery Chang Liu https://ebookbell.com/product/high-throughput-mass-spectrometry-in- drug-discovery-chang-liu-51310884 Highthroughput Image Reconstruction And Analysis Rao Ar Cecchi Ga Eds https://ebookbell.com/product/highthroughput-image-reconstruction-and- analysis-rao-ar-cecchi-ga-eds-2046986 Highthroughput Next Generation Sequencing Methods And Applications 1st Edition Kathleen E Steinmann https://ebookbell.com/product/highthroughput-next-generation- sequencing-methods-and-applications-1st-edition-kathleen-e- steinmann-2120052
- 3. Highthroughput Screening In Drug Discovery Methods And Principles In Medicinal Chemistry 1st Edition Jrg Hser Editor https://ebookbell.com/product/highthroughput-screening-in-drug- discovery-methods-and-principles-in-medicinal-chemistry-1st-edition- jrg-hser-editor-2164066 High Throughput Analysis For Early Drug Discovery 1st Edition James Kyranos Editor https://ebookbell.com/product/high-throughput-analysis-for-early-drug- discovery-1st-edition-james-kyranos-editor-2164632 Highthroughput Screening Methods In Toxicity Testing Steinberg https://ebookbell.com/product/highthroughput-screening-methods-in- toxicity-testing-steinberg-4171180 Highthroughput Phenotyping In Plants Methods And Protocols 1st Edition Benjamin J Cole https://ebookbell.com/product/highthroughput-phenotyping-in-plants- methods-and-protocols-1st-edition-benjamin-j-cole-4177116 Highthroughput Analysis A Tool For Combinatorial Materials Science 1st Edition Radislav A Potyrailo https://ebookbell.com/product/highthroughput-analysis-a-tool-for- combinatorial-materials-science-1st-edition-radislav-a- potyrailo-4204378
- 5. High-Throughput Metabolomics Angelo D’Alessandro Editor Methods and Protocols Methods in Molecular Biology 1978
- 6. For further volumes: http://www.springer.com/series/7651 Me t h o d s i n Mo l e c u l a r Bi o lo g y Series Editor John M. Walker School of Life and Medical Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK
- 7. High-Throughput Metabolomics Methods and Protocols Edited by Angelo D’Alessandro DepartmentofBiochemistryandMolecularGenetics,UniversityofColoradoDenver, Aurora,CO,USA
- 8. ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-4939-9235-5 ISBN 978-1-4939-9236-2 (eBook) https://doi.org/10.1007/978-1-4939-9236-2 © Springer Science+Business Media, LLC, part of Springer Nature 2019 This work is subject to copyright. All rights are reserved 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, express 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 Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature. The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A. Editor Angelo D’Alessandro Department of Biochemistry and Molecular Genetics University of Colorado Denver Aurora, CO, USA
- 9. v High-Throughput Metabolomics: So Much to Discover, So Little Time… Dum loquimur fugerit invida aetas… (Horace, Odes 1, 11, 8) The post-genomic era and the bioinformatic revolution that accompanied it fostered new strides in the fields of metabolomics and lipidomics. These “omics” approaches are often referred to—rightfully so—as the “closest to the phenotype” and perceived by the scientific community as novel, especially in comparison to genomics, transcriptomics, and pro- teomics. Despite the aggressive and largely successful efforts to rebrand this discipline, metabolomics—defined as the comprehensive analysis of small molecule metabolites—is perhaps the oldest analytical tool mankind managed to harness. History is full of records describing symptoms and metabolic characteristics of metabolic diseases such as (ante lit- teram) diabetes: “the sweet taste” and “capacity to attract ants” of urine have been docu- mented since the fifth century BCE in India and Greece, second century BCE in China. Centuries of advancements in the fields of chemistry and (clinical) biochemistry, recently accompanied by the introduction of tools like NMR and mass spectrometers, have simply provided a novel “magnifying lens” to expand our understanding of the small molecule world as a function of our attempts to “poke nature.” From this perspective, metabolomics is nothing but the next iteration of a discipline that scientists have been investigating for decades with much less sophisticated tools, often compensating for the technological gap with incredible rigor and acumen. Building on decades of advancements and empowered by novel analytical and bioinformatics tools, scientists have embraced the “new” field of metabolomics to generate a wealth of data from laboratory studies, some of which are slowly transitioning into the clinics. This transition can be significantly sped up owing to the opportunity to perform large-scale studies in a high- throughput fashion both at the discovery phase (e.g., high-throughput screening of novel drugs) and clinical testing (e.g., in large-scale prospective studies). In this view, this entry of the Methods in Molecular Biology series focuses on recent technological, computational, and biostatistical advances in the field of high-throughput metabolomics. Chapters encompass methods, platforms, and analytical strategies for steady-state measurements and metabolic flux analysis with stable isotope-labeled tracers, in biological matrices of clinical relevance and model organisms. Mass spectrometry-based or orthogonal methods are discussed, along with computational and statistical methods to address data sparsity in high- throughput metabolomics approaches. Finally, a few representative applications are discussed, including biodosimetry, sports and wellness, and personalized metabolomics. The main take-home message we wish to share with the interested reader is that high- throughput metabolomics tools can bring about the next generation of clinical biochemistry in a cost-effective, but not necessarily less rigorous fashion than current analytical approaches, exponentially advancing our capacity to investigate nature while easing the advent of personalized medicine. Preface
- 10. vi Prior to concluding this quick introduction to the contents of the book, I will take the chance to thank all the contributing authors for their support to this successful initiative and Dr. John Walker and David C. Casey (Springer Nature) and Julie Reisz Haines (University of Colorado Denver, Anschutz Medical Campus) for their invaluable editorial assistance. Conflict of Interest A.D. is founder and CSO of Omix Technologies, Inc. Aurora, CO, USA Angelo D’Alessandro Preface
- 11. vii Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Part I Methods 1 Sample Preparation and Reporting Standards for Metabolomics of Adherent Mammalian Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Sarah Hayton, Robert D. Trengove, and Garth L. Maker 2 High-Throughput Metabolomics: Isocratic and Gradient Mass Spectrometry-Based Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Travis Nemkov, Julie A. Reisz, Sarah Gehrke, Kirk C. Hansen, and Angelo D’Alessandro 3 High-Throughput Metabolomics Based on Direct Mass Spectrometry Analysis in Biomedical Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Raúl González-Domínguez, Álvaro González-Domínguez, Carmen Segundo, Mónica Schwarz, Ana Sayago, Rosa María Mateos, Enrique Durán-Guerrero, Alfonso María Lechuga-Sancho, and Ángeles Fernández-Recamales 4 Traveling Wave Ion Mobility Mass Spectrometry: Metabolomics Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Giuseppe Paglia and Giuseppe Astarita 5 Capillary Electrophoresis Mass Spectrometry as a Tool for Untargeted Metabolomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Ángeles López-Gonzálvez, Joanna Godzien, Antonia García, and Coral Barbas Part II Lipidomics 6 Overview of Lipid Mass Spectrometry and Lipidomics . . . . . . . . . . . . . . . . . . . 81 Simona Zarini, Robert M. Barkley, Miguel A. Gijón, and Robert C. Murphy 7 LC-MS/MS-MRM-Based Targeted Metabolomics for Quantitative Analysis of Polyunsaturated Fatty Acids and Oxylipins . . . . . . . . . . . . . . . . . . . 107 Xiaoyun Fu, Mikayla Anderson, Yi Wang, and James C. Zimring 8 Untargeted and Semi-targeted Lipid Analysis of Biological Samples Using Mass Spectrometry-Based Metabolomics . . . . . . . . . . . . . . . . . . . . . . . . 121 Julie A. Reisz, Connie Zheng, Angelo D’Alessandro, and Travis Nemkov 9 HPLC-MS/MS Methods for Diacylglycerol and Sphingolipid Molecular Species in Skeletal Muscle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 Kathleen A. Harrison and Bryan C. Bergman Contents
- 12. viii Part III Metabolomics of Animal and Plant Models 10 Quantification of d- and l-2-Hydroxyglutarate in Drosophila melanogaster Tissue Samples Using Gas Chromatography- Mass Spectrometry . . . . . . . . . . . . 155 Hongde Li and Jason M. Tennessen 11 Comprehensive LC-MS-Based Metabolite Fingerprinting Approach for Plant and Fungal-Derived Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Kirstin Feussner and Ivo Feussner 12 Untargeted Metabolomics of Plant Leaf Tissues . . . . . . . . . . . . . . . . . . . . . . . . 187 Federica Gevi, Giuseppina Fanelli, Lello Zolla, and Sara Rinalducci Part IV Tracing Experiments and Metabolic Flux Analysis 13 Analysis of Arginine Metabolism Using LC-MS and Isotopic Labeling . . . . . . . 199 Gretchen L. Seim, Emily C. Britt, and Jing Fan 14 Quantifying Intermediary Metabolism and Lipogenesis in Cultured Mammalian Cells Using Stable Isotope Tracing and Mass Spectrometry . . . . . . 219 Thekla Cordes and Christian M. Metallo 15 Insights into Dynamic Network States Using Metabolomic Data . . . . . . . . . . . 243 Reihaneh Mostolizadeh, Andreas Dräger, and Neema Jamshidi 16 Analysis of Endothelial Fatty Acid Metabolism Using Tracer Metabolomics . . . 259 Joanna Kalucka, Bart Ghesquière, Sarah-Maria Fendt, and Peter Carmeliet 17 Stable Isotope Tracers for Metabolic Pathway Analysis . . . . . . . . . . . . . . . . . . . 269 Sara Violante, Mirela Berisa, Tiffany H. Thomas, and Justin R. Cross Part V Data Processing in Metabolomics 18 Data Processing for GC-MS- and LC-MS-Based Untargeted Metabolomics . . . . . 287 Linxing Yao, Amy M. Sheflin, Corey D. Broeckling, and Jessica E. Prenni 19 El-MAVEN: A Fast, Robust, and User-Friendly Mass Spectrometry Data Processing Engine for Metabolomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301 Shubhra Agrawal, Sahil Kumar, Raghav Sehgal, Sabu George, Rishabh Gupta, Surbhi Poddar, Abhishek Jha, and Swetabh Pathak 20 Pre-analytic Considerations for Mass Spectrometry-Based Untargeted Metabolomics Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323 Dominik Reinhold, Harrison Pielke-Lombardo, Sean Jacobson, Debashis Ghosh, and Katerina Kechris Part VI Metabolic Measurements with Techniques Orthogonal to Mass Spectrometry 21 Temporal Metabolite, Ion, and Enzyme Activity Profiling Using Fluorescence Microscopy and Genetically Encoded Biosensors . . . . . . . . . . . . . 343 Douglas A. Chapnick, Eric Bunker, Xuedong Liu, and William M. Old 22 Microplate Assays for Spectrophotometric Measurement of Mitochondrial Enzyme Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355 Rachel C. Janssen and Kristen E. Boyle Contents
- 13. ix 23 Quantitative NMR-Based Metabolomics on Tissue Biomarkers and Its Translation into In Vivo Magnetic Resonance Spectroscopy . . . . . . . . . 369 Natalie J. Serkova, Denise M. Davis, Jenna Steiner, and Rajesh Agarwal Part VII Towards Personalized Metabolomics 24 Metabolomic Applications in Radiation Biodosimetry . . . . . . . . . . . . . . . . . . . . 391 Evagelia C. Laiakis 25 Metabolomics Analyses to Investigate the Role of Diet and Physical Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403 Pol Herrero, Miguel Ángel Rodríguez, Maria Rosa Ras, Antoni del Pino, Lluís Arola, and Núria Canela 26 Blood Biomarkers in Sports Medicine and Performance and the Future of Metabolomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431 Iñigo San-Millán 27 Personalized Metabolomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447 David P. Marciano and Michael P. Snyder Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457 Contents
- 14. xi Rajesh Agarwal • Department of Pharmaceutical Sciences, Skaggs School of Pharmacy and Pharmaceutical Sciences, University of Colorado Denver, Aurora, CO, USA Shubhra Agrawal • Elucidata, Inc., Cambridge, MA, USA Mikayla Anderson • Bloodworks Northwest Research Institute, Seattle, WA, USA Lluís Arola • Biochemistry and Biotechnological Department, Nutrigenomics Research Group, Universitat Rovira i Virgili, Tarragona, Spain; Biotechnological Area, EURECAT-Technological Center of Catalonia, Reus, Spain Giuseppe Astarita • Department of Biochemistry and Molecular and Cellular Biology, Georgetown University, Washington, DC, USA Coral Barbas • Facultad de Farmacia, Centro de Metabolómica y Bioanálisis (CEMBIO), Universidad CEU San Pablo, Madrid, Spain Robert M. Barkley • Department of Pharmacology, University of Colorado Denver, Aurora, CO, USA Bryan C. Bergman • Division of Endocrinology, Diabetes, and Metabolism, School of Medicine, University of Colorado Anschutz Medical Campus, Denver, CO, USA Mirela Berisa • Donald B. and Catherine C. Cancer Metabolism Center, Memorial Sloan Kettering Cancer Center, New York, NY, USA Kristen E. Boyle • Department of Pediatrics, Section of Nutrition, University of Colorado Anschutz Medical Campus, Aurora, CO, USA Emily C. Britt • Morgridge Institute for Research, Madison, WI, USA; Department of Nutritional Sciences, University of Wisconsin-Madison, Madison, WI, USA Corey D. Broeckling • Proteomics and Metabolomics Facility, Colorado State University, Fort Collins, CO, USA Eric Bunker • Department of Chemistry and Biochemistry, University of Colorado, Boulder, CO, USA Núria Canela • Technological Joint Unit of Omic Sciences, EURECAT-Technological Center of Catalonia, Universitat Rovira i Virgili, Reus, Spain Peter Carmeliet • Laboratory of Angiogenesis and Vascular Metabolism, VIB Center for Cancer Biology (CCB), VIB, Leuven, Belgium; Laboratory of Angiogenesis and Vascular Metabolism, Department of Oncology and Leuven Cancer Institute (LKI), KU Leuven, Leuven, Belgium Douglas A. Chapnick • Department of Chemistry and Biochemistry, University of Colorado, Boulder, CO, USA Thekla Cordes • Department of Bioengineering, University of California San Diego, La Jolla, CA, USA Justin R. Cross • Donald B. and Catherine C. Cancer Metabolism Center, Memorial Sloan Kettering Cancer Center, New York, NY, USA Angelo D’Alessandro • Department of Biochemistry and Molecular Genetics, University of Colorado Denver, Aurora, CO, USA Denise M. Davis • Department of Radiology, School of Medicine, University of Colorado Denver, Aurora, CO, USA Contributors
- 15. xii Antoni del Pino • Technological Joint Unit of Omic Sciences, EURECAT-Technological Center of Catalonia, Universitat Rovira i Virgili, Reus, Spain Andreas Dräger • Center for Bioinformatics Tübingen (ZBIT), University of Tübingen, Tübingen, Germany; Department for Computer Science, University of Tübingen, Tübingen, Germany; German Center for Infection Research (DZIF), Tübingen, Germany Enrique Durán-Guerrero • Instituto de Investigación Vitivinícola y Agroalimentario (IVAGRO), University of Cádiz, Puerto Real, Spain; Department of Analytical Chemistry, University of Cádiz, Puerto Real, Spain Jing Fan • Morgridge Institute for Research, Madison, WI, USA; Department of Nutritional Sciences, University of Wisconsin-Madison, Madison, WI, USA Giuseppina Fanelli • Department of Ecological and Biological Sciences (DEB), University of Tuscia, Viterbo, Italy Sarah-Maria Fendt • Laboratory of Cellular Metabolism and Metabolic Regulation, VIB Center for Cancer Biology (CCB), VIB, Leuven, Belgium; Laboratory of Cellular Metabolism and Metabolic Regulation, Department of Oncology and Leuven Cancer Institute (LKI), KU Leuven, Leuven, Belgium Ángeles Fernández-Recamales • Department of Chemistry, Faculty of Experimental Sciences, University of Huelva, Huelva, Spain; International Campus of Excellence CeiA3, University of Huelva, Huelva, Spain Ivo Feussner • Department of Plant Biochemistry, Albrecht-von-Haller-Institute for Plant Sciences, University of Goettingen, Goettingen, Germany; Service Unit for Metabolomics and Lipidomics, Goettingen Center for Molecular Biosciences (GZMB), University of Goettingen, Goettingen, Germany; Department of Plant Biochemistry, Goettingen Center for Molecular Biosciences (GZMB), University of Goettingen, Goettingen, Germany Kirstin Feussner • Department of Plant Biochemistry, Albrecht-von-Haller-Institute for Plant Sciences, University of Goettingen, Goettingen, Germany; Service Unit for Metabolomics and Lipidomics, Goettingen Center for Molecular Biosciences (GZMB), University of Goettingen, Goettingen, Germany Xiaoyun Fu • Bloodworks Northwest Research Institute, Seattle, WA, USA; Division of Hematology, Department of Internal Medicine, University of Washington School of Medicine, Seattle, WA, USA Antonia García • Facultad de Farmacia, Centro de Metabolómica y Bioanálisis (CEMBIO), Universidad CEU San Pablo, Madrid, Spain Sarah Gehrke • Department of Biochemistry and Molecular Genetics, University of Colorado Denver, Aurora, CO, USA Sabu George • Elucidata, Inc., Cambridge, MA, USA Federica Gevi • Department of Science and Technology for Agriculture, Forestry, Nature and Energy (DAFNE), University of Tuscia, Viterbo, Italy Bart Ghesquière • Metabolomics Expertise Center, VIB Center for Cancer Biology (CCB), VIB, Leuven, Belgium; Department of Oncology, Metabolomics Expertise Center, KU Leuven, Leuven, Belgium Debashis Ghosh • Department of Biostatistics and Informatics, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, CO, USA Miguel A. Gijón • Department of Pharmacology, University of Colorado Denver, Aurora, CO, USA Contributors
- 16. xiii Joanna Godzien • Facultad de Farmacia, Centro de Metabolómica y Bioanálisis (CEMBIO), Universidad CEU San Pablo, Madrid, Spain Álvaro González-Domínguez • Department of Pediatrics, Hospital Universitario Puerta del Mar, Cádiz, Spain; Institute of Research and Innovation in Biomedical Sciences of the Province of Cádiz (INiBICA), Cádiz, Spain Raúl González-Domínguez • Department of Chemistry, Faculty of Experimental Sciences, University of Huelva, Huelva, Spain; International Campus of Excellence CeiA3, University of Huelva, Huelva, Spain Rishabh Gupta • Elucidata, Inc., Cambridge, MA, USA Kirk C. Hansen • Department of Biochemistry and Molecular Genetics, University of Colorado Denver, Aurora, CO, USA Kathleen A. Harrison • Division of Endocrinology, Diabetes, and Metabolism, School of Medicine, University of Colorado Anschutz Medical Campus, Denver, CO, USA Sarah Hayton • Medical, Molecular and Forensic Sciences, Murdoch University, Murdoch, WA, Australia; Separation Science and Metabolomics Laboratory, Murdoch University, Murdoch, WA, Australia Pol Herrero • Technological Joint Unit of Omic Sciences, EURECAT-Technological Center of Catalonia, Universitat Rovira i Virgili, Reus, Spain Sean Jacobson • Center for Genes, Environment, and Health, National Jewish Health, Denver, CO, USA Neema Jamshidi • University of California, Los Angeles, Los Angeles, CA, USA Rachel C. Janssen • Department of Pediatrics, Section of Neonatology, University of Colorado Anschutz Medical Campus, Aurora, CO, USA Abhishek Jha • Elucidata, Inc., Cambridge, MA, USA Joanna Kalucka • Laboratory of Angiogenesis and Vascular Metabolism, VIB Center for Cancer Biology (CCB), VIB, Leuven, Belgium; Laboratory of Angiogenesis and Vascular Metabolism, Department of Oncology and Leuven Cancer Institute (LKI), KU Leuven, Leuven, Belgium Katerina Kechris • Department of Biostatistics and Informatics, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, CO, USA Sahil Kumar • Elucidata, Inc., Cambridge, MA, USA Evagelia C. Laiakis • Department of Oncology, Lombardi Comprehensive Cancer Center, Georgetown University, Washington, DC, USA; Department of Biochemistry and Molecular and Cellular Biology, Georgetown University, Washington, DC, USA Alfonso María Lechuga-Sancho • Department of Pediatrics, Hospital Universitario Puerta del Mar, Cádiz, Spain; Institute of Research and Innovation in Biomedical Sciences of the Province of Cádiz (INiBICA), Cádiz, Spain; Department of Mother and Child Health and Radiology, Faculty of Medicine, University of Cádiz, Cádiz, Spain Hongde Li • Department of Biology, Indiana University, Bloomington, IN, USA Xuedong Liu • Department of Chemistry and Biochemistry, University of Colorado, Boulder, CO, USA Ángeles López-Gonzálvez • Facultad de Farmacia, Centro de Metabolómica y Bioanálisis (CEMBIO), Universidad CEU San Pablo, Madrid, Spain Garth L. Maker • Medical, Molecular and Forensic Sciences, Murdoch University, Murdoch, WA, Australia; Separation Science and Metabolomics Laboratory, Murdoch University, Murdoch, WA, Australia David P. Marciano • Department of Genetics, Stanford University School of Medicine, Stanford, CA, USA Contributors
- 17. xiv Rosa María Mateos • Department of Pediatrics, Hospital Universitario Puerta del Mar, Cádiz, Spain; Institute of Research and Innovation in Biomedical Sciences of the Province of Cádiz (INiBICA), Cádiz, Spain Christian M. Metallo • Department of Bioengineering, University of California San Diego, La Jolla, CA, USA; Moores Cancer Center, University of California San Diego, La Jolla, CA, USA; Diabetes and Endocrinology Research Center, University of California San Diego, La Jolla, CA, USA; Institute of Engineering in Medicine, University of California San Diego, La Jolla, CA, USA Iñigo San-Millán • Division of Sports Medicine, University of Colorado School of Medicine, Aurora, CO, USA Reihaneh Mostolizadeh • Center for Bioinformatics Tübingen (ZBIT), University of Tübingen, Tübingen, Germany; Department for Computer Science, University of Tübingen, Tübingen, Germany; German Center for Infection Research (DZIF), Tübingen, Germany Robert C. Murphy • Department of Pharmacology, University of Colorado Denver, Aurora, CO, USA Travis Nemkov • Department of Biochemistry and Molecular Genetics, University of Colorado Denver, Aurora, CO, USA William M. Old • Department of Molecular, Cellular and Developmental Biology, University of Colorado, Boulder, CO, USA; Linda Crnic Institute for Down Syndrome, University of Colorado School of Medicine, Aurora, CO, USA Giuseppe Paglia • Institute for Biomedicine, EURAC Research, Bolzano, Italy Swetabh Pathak • Elucidata, Inc., Cambridge, MA, USA Harrison Pielke-Lombardo • Computational Bioscience Program, School of Medicine, University of Colorado Anschutz Medical Campus, Aurora, CO, USA Surbhi Poddar • Elucidata, Inc., Cambridge, MA, USA Jessica E. Prenni • Proteomics and Metabolomics Facility, Colorado State University, Fort Collins, CO, USA Maria Rosa Ras • Technological Joint Unit of Omic Sciences, EURECAT-Technological Center of Catalonia, Universitat Rovira i Virgili, Reus, Spain Dominik Reinhold • PPD, Wilmington, NC, USA Julie A. Reisz • Department of Biochemistry and Molecular Genetics, University of Colorado Denver, Aurora, CO, USA Sara Rinalducci • Department of Ecological and Biological Sciences (DEB), University of Tuscia, Viterbo, Italy Miguel Ángel Rodriguez • Technological Joint Unit of Omic Sciences, EURECAT- Technological Center of Catalonia, Universitat Rovira i Virgili, Reus, Spain Ana Sayago • Department of Chemistry, Faculty of Experimental Sciences, University of Huelva, Huelva, Spain; International Campus of Excellence CeiA3, University of Huelva, Huelva, Spain Mónica Schwarz • Institute of Research and Innovation in Biomedical Sciences of the Province of Cádiz (INiBICA), Cádiz, Spain; “Salus Infirmorum” Faculty of Nursing, University of Cádiz, Cádiz, Spain; Instituto de Investigación Vitivinícola y Agroalimentario (IVAGRO), University of Cádiz, Puerto Real, Spain Carmen Segundo • Institute of Research and Innovation in Biomedical Sciences of the Province of Cádiz (INiBICA), Cádiz, Spain; “Salus Infirmorum” Faculty of Nursing, University of Cádiz, Cádiz, Spain Contributors
- 18. xv Raghav Sehgal • Elucidata, Inc., Cambridge, MA, USA Gretchen L. Seim • Morgridge Institute for Research, Madison, WI, USA; Department of Nutritional Sciences, University of Wisconsin-Madison, Madison, WI, USA Natalie J. Serkova • Department of Radiology, School of Medicine, University of Colorado Denver, Aurora, CO, USA Amy M. Sheflin • Proteomics and Metabolomics Facility, Colorado State University, Fort Collins, CO, USA Michael P. Snyder • Department of Genetics, Stanford University School of Medicine, Stanford, CA, USA Jenna Steiner • Department of Radiology, School of Medicine, University of Colorado Denver, Aurora, CO, USA Jason M. Tennessen • Department of Biology, Indiana University, Bloomington, IN, USA Tiffany H. Thomas • Department of Pathology and Cell Biology, Columbia University Medical Center, New York, NY, USA Robert D. Trengove • Separation Science and Metabolomics Laboratory, Murdoch University, Murdoch, WA, Australia Sara Violante • Donald B. and Catherine C. Cancer Metabolism Center, Memorial Sloan Kettering Cancer Center, New York, NY, USA Yi Wang • Bloodworks Northwest Research Institute, Seattle, WA, USA Linxing Yao • Proteomics and Metabolomics Facility, Colorado State University, Fort Collins, CO, USA Simona Zarini • Department of Pharmacology, University of Colorado Denver, Aurora, CO, USA Connie Zheng • Department of Biochemistry and Molecular Genetics, University of Colorado Denver, Aurora, CO, USA James C. Zimring • Bloodworks Northwest Research Institute, Seattle, WA, USA; Division of Hematology, Department of Internal Medicine, University of Washington School of Medicine, Seattle, WA, USA; Department of Laboratory Medicine, University of Washington School of Medicine, Seattle, WA, USA Lello Zolla • Department of Science and Technology for Agriculture, Forestry, Nature and Energy (DAFNE), University of Tuscia, Viterbo, Italy Contributors
- 19. Part I Methods
- 20. 3 Angelo D’Alessandro (ed.), High-Throughput Metabolomics: Methods and Protocols, Methods in Molecular Biology, vol. 1978, https://doi.org/10.1007/978-1-4939-9236-2_1, © Springer Science+Business Media, LLC, part of Springer Nature 2019 Chapter 1 Sample Preparation and Reporting Standards for Metabolomics of Adherent Mammalian Cells Sarah Hayton, Robert D. Trengove, and Garth L. Maker Abstract Metabolomics is an analytical technique that investigates the small molecules present within a biological system. Metabolomics of cultured cells allows profiling of the metabolic chemicals involved in a cell type- specific system and the response of that metabolome to external challenges, such as change in environment or exposure to drugs or toxins. The numerous benefits of in vitro metabolomics include a much greater control of external variables and reduced ethical concerns. There is potential for metabolomics of mam- malian cells to uncover new information on mechanisms of action for drugs or toxins or to provide a more sensitive, human-specific early risk assessment in drug development or toxicology investigations. One way to achieve stronger biological outcomes from metabolomic data is via the use of these mammalian cultured cell models, particularly in a high-throughput context. With the sensitivity and quantity of data that metabolomics is able to provide, it is important to ensure that the sampling techniques have minimal inter- ference when it comes to interpretation of any observed shifts in the metabolite profile. Here we describe a sampling procedure designed to ensure that the effects seen in metabolomic analyses are explained fully by the experimental factor and not other routine culture-specific activities. Key words Cell culture, Mammalian cells, Adherent cells, Experimental design, Sample preparation, Quenching, Metabolite extraction, Extracellular, Intracellular, Metabolomics 1 Introduction Metabolomics studies frequently state that the metabolome is a closer reflection of the phenotype of an organism, tissue, or cell than other “omics analyses,” such as proteomics, transcriptomics, and genomics [1–3]. The small molecules that are shuffled around the vast network of metabolic pathways in a biological system are referred to as metabolites, with the “metabolome” made up of the many thousands of different metabolites and their relative levels of abundance. The goal of metabolomic profiling is to measure changes in the metabolome of a given system in response to a chal- lenge to normal cellular homeostasis, whether physical, chemical, environmental, or other external stressor.
- 21. 4 Metabolomics has recently become an attractive application for untargeted, high-throughput screening analyses [4]. However, due to the data-rich, hypothesis-generating outcomes of untar- geted metabolomics, it is understandable that such an approach might not be attractive to some investigators, particularly if a lim- ited number of clinical or animal samples are available. In particu- lar, handling the vast amounts of data can complicate meaningful biological interpretation of the data. This can be partly addressed by the use of cultured mammalian cells, as opposed to animal- based samples, to more easily accommodate re-visiting of the sam- ple set if novel or previously unknown metabolites are highlighted by an untargeted study. The number of controls and replicates can be easily manipulated in design of cell culture-based experiments, benefiting the development, validation, and standardization of untargeted, high-throughput metabolomic studies. The impor- tance of experimental design and standardized reporting require- ments for these studies of cultured mammalian cells has been previously reviewed across a large number of published protocols [5, 6]. The focus of this protocol is to give the investigator a sound, high-throughput procedure to follow for handling and preparing cultured cell samples preceding instrumental analysis for metabo- lomics, which will have minimal possible interference on the detectable metabolome. It is designed so that any of the multiple platforms currently used for metabolomics can be applied to the prepared samples thereafter (e.g., gas chromatography-mass spec- trometry, liquid chromatography-mass spectrometry). Here we provide a standardized protocol for untargeted metabolomic anal- ysis applied to any cultured mammalian cell line, highlighting the importance of adequate reporting of culture conditions to allow for meaningful biological interpretation of metabolomic data and comparison of results across multiple studies. 2 Materials Cell lines should be sourced from reputable sources such as the European Collection of Cell Cultures (ECACC) or the American Type Culture Collection (ATCC). Examples of cell lines which have been studied using this protocol: ● ● B50 rat neuroblastoma (ECACC 85042302). ● ● HepG2 human hepatocarcinoma (ECACC 85011430; ATCC HB-8065). ● ● SH-SY5Y human neuroblastoma (ECACC 94030304; ATCC CRL-2266). 2.1 Cell Lines Sarah Hayton et al.
- 22. 5 All cell culture media should be sterile-filtered (whether pre- or post-purchase) and medium and supplements within expiry peri- ods. It is highly recommended to follow ECACC or ATCC recom- mendations of growth conditions for the specific cell line. The above cell lines were cultured at 37 °C and 5% CO2 with the fol- lowing media and supplements: ● ● Dulbecco’s Modified Eagle’s Medium (DMEM; 4.5 g/L glucose). ● ● Ham’s F12 nutrient mixture. ● ● l-glutamine (2 mM, 1% v/v). ● ● Penicillin and streptomycin (combined 104 U/mL, 1% v/v). ● ● Fetal calf serum (heat-treated at 56 °C for 60 min; 5 or 10% v/v). All solvents used for cell culture treatments, PBS washes, and sam- ple preparation (e.g., water, methanol) should be of the highest purity possible, preferably LC-MS grade. The use of lower-quality solvents may introduce impurities that can interfere with mass spectrometry data. Suggested composition of solutions for cell handling: ● ● 0.25% trypsin-EDTA solution (2.5 g/L porcine trypsin and 0.2 g/L EDTA). ● ● Phosphate-buffered saline (PBS; 0.01 M phosphate buffer, 0.0027 M potassium chloride, and 0.137 M sodium chloride, pH 7.4). ● ● 0.4% w/v trypan blue dye in phosphate-buffered saline solu- tion. A hemocytometer and microscope or an automated cell counter can be used for cell counting. 3 Methods The entire workflow is summarized in Fig. 1. 1. Seed cells at the required density (number of cells) in the required number of culture plates (see Note 1). The volume of medium should be determined based on the size of the culture plate chosen. 2. Cells should be left to grow and adhere for 24–48 h before treatment, depending on the specific cell type. Cells should not be allowed to become over-confluent, as the metabolome will be affected by the depletion of supplements in the medium or potentially by differentiation of the cells (see Note 2). Reporting requirements: 2.2 Growth Conditions 2.3 Cell Handling 3.1 Cell Culture Sample Setup Sample Preparation and Reporting Standards for Metabolomics of Adherent Mammalian…
- 23. 6 Fig. 1 Workflow schematic of sample preparation procedure for metabolomic analyses of adherent mammalian cells Sarah Hayton et al.
- 24. 7 Name and source of cell line, including results from genetic valida- tion and mycoplasma testing. Generation (passage) number of cells used to set up experimental samples (see Note 3). Size and type of culture plates used and the seeding density as cell number. Specifics of medium and supplements used, including type, sup- plier, concentration, and serum percentage. Passaging procedures used during general cell culture. Inoculation procedures for any treatments/exposures (nature of exposure, exposure time, doses). Environmental conditions of cell incubation. Setup of parallel samples for cell counting and quality control sam- ples (see Note 4). 1. Immediately before harvesting for metabolomics, trypsinize and count one representative (duplicate) sample per replicate for the number of viable cells, via addition of 0.4% w/v trypan blue dye (see Note 5). Cells can be counted manually or using an automated cell counter, which should be validated before use. 2. Immediately following cell counting, remove all culture plates from incubation, and place on ice. Care should be taken not to dislodge cells from the plate surface while transporting. 1. Remove medium from all culture plates into centrifuge tubes, and centrifuge at low rcf (300 × g) for 10 min to pellet any cells that may be present. 2. Transfer medium to fresh tubes, and store on ice before snap- freezing, lyophilization, and storage at −80 °C, prior to extrac- tion and analysis of the extracellular metabolome. Freeze-drying unit used should be capable of reaching −80 °C temperature and 0.002 mBar vacuum. 1. Wash remaining adhered cells with cold phosphate-buffered saline (PBS; 4 °C) to remove any trace medium present. The volume used should be adequate to cover the cells in the cul- ture plate on a level surface. The cold PBS also acts to quench the metabolism of the cells (see Note 6) and so should be added to all culture plates at the same time. 2. Remove and discard this wash PBS. 3.2 Counting of Cells in Duplicate Samples 3.3 Sample Collection (Medium/ Extracellular) 3.4 Quenching Sample Preparation and Reporting Standards for Metabolomics of Adherent Mammalian…
- 25. 8 1. Add an equal volume of cold PBS to all plates for the collection of cells. Remove cells from the plate surface by scraping into the PBS, and collect into tubes appropriate for tissue lysis (nor- mally screw cap, o-ring sealed). Store tubes on ice until snap- freezing and lyophilization. 2. Snap-freeze both cell and medium samples in liquid nitrogen, and lyophilize (freeze-dry) to minimize degradation of metab- olites before extraction and analysis can take place. 1. Prepare metabolite extraction solution by the addition of required concentration of internal standard compounds (e.g., stable isotope-labeled chemicals) in an 80:20 methanol/water mixture (see Note 7). 2. Remove freeze-dried samples from −80 °C storage, and add the same volume of extraction solution to all samples. This required volume is determined by the amount of material col- lected and thus the amount of dried tissue present and the size of the storage tube. To prevent loss of sample during the con- centration process (step 6 below), no more than two-thirds of the maximum volume capacity of the tube should be used, e.g., a maximum of 1 mL total volume of extracted sample in a 1.5 mL microcentrifuge tube. 3. Extract metabolites from cell samples by vigorous sonication or tissue disruption to lyse any remaining intact cell mem- branes and release intracellular metabolites. Extracellular medium can be extracted by simple vortex mixing. 4. Centrifuge extracts at high rcf (e.g., 1.5 × 104 × g) for 10 min to pellet the sample debris. Collect the same volume of super- natant into a fresh tube for every sample. 5. The remaining pellet can undergo a second extraction step (repeat of steps 2–4) to maximize metabolite recovery from the biological material. Following centrifugation of the repeat extraction, the supernatant can be added to the first collected extract. 6. Evaporate methanol in the extraction solution via vacuum con- centration, and reconstitute the sample in the same volume of water (see Note 8). 7. Snap-freeze samples once again, and lyophilize by freeze- drying. Once completely dry, the samples can be stored at −80 °C until metabolomic analysis. 8. As water has been removed from the extracts, the samples can be reconstituted as desired, according to the chosen analytical platform. For example, for GC-based analysis the samples would be chemically derivatized, while for LC-based platforms, 3.5 Sample Collection (Cells/ Intracellular) 3.6 Metabolite Extraction Sarah Hayton et al.
- 26. 9 the samples can be reconstituted in the required mobile phase solvent(s). Reporting requirements: Composition of extraction solution. Description of extraction procedure (mechanical disruption, col- lection details, etc.) If known, expected recovery rate and stability of extracted metabo- lites (see Note 9). 4 Notes 1. The quantity of cells that needs to be grown per sample extracted for metabolomic studies should be optimized specifically for the intended analytical instrument’s detection capabilities and will vary depending on the specific cell being cultured. The exact quantity may depend on whether the study is an untargeted screen for metabolites or if a certain class or selection of metab- olites is being targeted. For untargeted analysis, this is generally assessed by observing the total ion chromatogram (TIC) of a single sample, which is the combined signal of everything mea- sured in that sample. An optimal number of cells per sample are chosen based on the maximum number of detectable metabo- lites that are confidently resolved from the background signal while also minimizing the occurrence of any high-abundance peaks that may cause overloading of the instrument detector, potentially obscuring any less abundant metabolites with similar chromatographic retention times. Once an optimal cell number for seeding density is obtained, this should remain consistent for the duration of the experiment, to minimize data manipula- tion in later interpretation stages. 2. If a change of medium is required during the incubation period after cell seeding, care should be taken not to disrupt the course of cell growth, as much as is practicable. All control samples should also have medium changed. 3. If multiple passages are required for repeated samples of the same experimental setup, then the range of passage generations used should be stated. Depending on the design for the specific metabolomic experiment, as well as the characteristics of the cells used, it may not be necessary to conduct experiments across multiple passages, especially for high-throughput cell metabolomics applications. This should be a part of the consid- eration during the researcher’s experimental design [5]. Sample Preparation and Reporting Standards for Metabolomics of Adherent Mammalian…
- 27. 10 4. The use of quality control samples within a metabolomic study, particularly for large sample sets, is crucially important for accu- rate correction and statistical interpretation of the data post- analysis [7, 8]. For an untargeted study, the QC samples are typically obtained from a pooled sample of all treatment groups, set up in parallel to the experimental samples, and collected simultaneously. If sample volume is sufficient, QC samples can be obtained by pooling an aliquot of the experimental samples. QC samples should be placed throughout the sequence during instrumental analysis, meaning that any drift in metabolite detection levels can be detected and modelled. This allows any variation observed from drift phenomena to be removed from statistical analysis of the data. The appropriate number of QC samples should be determined in the experimental design stages, to ensure that adequate sample material is prepared. 5. In cell culture experiments, especially when testing the response of a system to a physical challenge, the amount of material avail- able for metabolomic analysis will likely differ between sample groups. It is therefore important to normalize metabolomic data post-analysis to remove this sample variation and allow for meaningful biological interpretation. A number of differing techniques have been used, including protein or DNA concen- tration, tissue weight, or cell number [9–12]. The most com- mon method of normalization of sample variation is to the cell number upon sample harvesting, which should be considered as the standard approach when possible. A separate, parallel sam- ple should be set up along with the sample for metabolite har- vest and extraction, specifically for cell counting. This is necessary due to the destructive nature of harvesting for metab- olomic analysis, as well as the leakage of metabolites encoun- tered during trypsinization of cells for counting [13, 14]. 6. An ideal extraction procedure for metabolomics of cultured cells should immediately quench metabolism in order to extract and collect all metabolites present at the end point of the exper- iment [15, 16]. PBS is a buffered, isotonic solution, so it causes less membrane leakage, thereby reducing loss of intracellular metabolites compared to other solvents that have been used in quenching, such as methanol. 7. The composition of extraction solution can be altered depend- ing on the chemical properties of the specific metabolites tar- geted for analysis, to maximize recovery. The suggested extraction solution, methanol and water in an 80:20 ratio, gives a single-phase extraction supernatant and allows for the whole extract to be collected. Also common is the use of a dual-phase (polar and nonpolar) extraction using methanol, chloroform, and water (or acetonitrile). The ratio of these solvents can vary but is most commonly used at 1:1 methanol and chloroform, Sarah Hayton et al.
- 28. 11 with a smaller proportion of water or acetonitrile. This results in a separation of polar and nonpolar extraction phases, which can be collected separately. If multiple phase extractions and collec- tions are carried out, this should be included in the reporting. 8. Extracts need to be reconstituted in a majority aqueous solution to allow for successful lyophilization, which itself enhances the conservation of metabolites for any extended period of deep- freeze storage. 9. If known, the expected recovery rate and stability of the extracted metabolites, dependent on the class of compound, should be included in basic reported information. Any such information that is known from previous metabolomic analyses using a particular extraction or instrumental protocol will ben- efit the researcher during the post-instrumental stages and ensure that accurate biological interpretation of the data is rep- resented. It may benefit other researchers by allowing the com- parison of multiple metabolomic studies. References 1. Idle JR, Gonzalez FJ (2007) Metabolomics. Cell Metab 6:348–351 2. Blow N (2008) Biochemistry's new look. Nature 455:697–700 3. Fiehn O (2002) Metabolomics - the link between genotypes and phenotypes. Plant Mol Biol 48(1–2):155–171 4. Hayton S, Maker GL, Mullaney I, Trengove RD (2017) Untargeted metabolomics of neu- ronal cell culture: a model system for the toxic- ity testing of insecticide chemical exposure. J Appl Toxicol 37:1481–1492. https://doi. org/10.1002/jat.3498 5. Hayton S, Maker GL, Mullaney I, Trengove RD (2017) Experimental design and reporting standards for metabolomics of mammalian cell lines. Cell Mol Life Sci 74(24):4421–4441. h t t p s : / / d o i . o r g / 1 0 . 1 0 0 7 / s00018-017-2582-1 6. León Z, García-Cañaveras JC, Donato MT, Lahoz A (2013) Mammalian cell metabolo- mics: experimental design and sample prepara- tion. Electrophoresis 34(19):2762–2775 7. Dunn WB, Broadhurst D, Begley P, Zelena E, Francis-McIntyre S, Anderson N, Brown M, Knowles JD, Halsall A, Haselden JN, Nicholls AW, Wilson ID, Kell DB, Goodacre R, Consortium THSMH (2011) Procedures for large-scale metabolic profiling of serum and plasma using gas chromatography and liquid chromatography coupled to mass spectrome- try. Nat Protoc 6(7):1060–1083 8. Goodacre R, Broadhurst D, Smilde AK, Kristal BS, Baker JD, Beger R, Bessant C, Connor S, Capuani G, Craig A, Ebbels T, Kell DB, Manetti C, Newton J, Paternostro G, Somorjai R, Sjöström M, Trygg J, Wulfert F (2007) Proposed minimum reporting standards for data analysis in metabolomics. Metabolomics 3(3):231–241 9. Muschet C, Möller G, Prehn C, Hrabē de Angelis M, Adamski J, Tokarz J (2016) Removing the bottlenecks of cell culture metabolomics: fast normalization procedure, correlation of metabolites to cell number, and impact of the cell harvesting method. Metabolomics 12(10):151. https://doi. org/10.1007/s11306-016-1104-8 10. Hutschenreuther A, Kiontke A, Birkenmeier G, Birkemeyer C (2012) Comparison of extrac- tion conditions and normalization approaches for cellular metabolomics of adherent growing cells with GC-MS. Anal Methods 4:1953–1963 11. Cao B, Aa J, Wang G, Wu X, Liu L, Li M, Shi J, Wang X, Zhao C, Zheng T, Guo S, Duan J (2011) GC-TOFMS analysis of metabolites in adherent MDCK cells and a novel strategy for identifying intracellular metabolic markers for use as cell amount indicators in data normaliza- tion. Anal Bioanal Chem 400(9):2983–2993 12. Silva LP, Lorenzi PL, Purwaha P, Yong V, Hawke DH, Weinstein JN (2013) Measurement of DNA concentration as a normalization strat- Sample Preparation and Reporting Standards for Metabolomics of Adherent Mammalian…
- 29. 12 egy for metabolomic data from adherent cell lines. Anal Chem 85(20):9536–9542 13. Dettmer K, Nürnberger N, Kaspar H, Gruber MA, Almstetter MF, Oefner PJ (2011) Metabolite extraction from adherently growing mammalian cells for metabolomics studies: optimization of harvesting and extraction pro- tocols. Anal Bioanal Chem 399(3):1127–1139 14. Garcia-Canaveras JC, Lopez S, Castell JV, Donato MT, Laboz A (2016) Extending metabolome coverage for untargeted metabo- lite profiling of adherent cultured hepatic cells. Anal Bioanal Chem 408:1217–1230 15. Čuperlović-Culf M, Barnett DA, Culf AS, Chute I (2010) Cell culture metabolomics: applications and future directions. Drug Discov Today 15(15/16):610–621 16. Álvarez-Sánchez B, Priego-Capote F, Luque de Castro MD (2010) Metabolomics analysis II: preparation of biological samples prior to detection. Trends Analyt Chem 29(2):120–127 Sarah Hayton et al.
- 30. 13 Angelo D’Alessandro (ed.), High-Throughput Metabolomics: Methods and Protocols, Methods in Molecular Biology, vol. 1978, https://doi.org/10.1007/978-1-4939-9236-2_2, © Springer Science+Business Media, LLC, part of Springer Nature 2019 Chapter 2 High-Throughput Metabolomics: Isocratic and Gradient Mass Spectrometry-Based Methods Travis Nemkov, Julie A. Reisz, Sarah Gehrke, Kirk C. Hansen, and Angelo D’Alessandro Abstract Metabolomics has emerged in the past decade as a highly attractive and impactful technique for phenotype- level profiling in diverse biological applications. Most recently, the dual developments of high-throughput analytical techniques along with dramatically increased sensitivity of high-resolution mass spectrometers have enabled the routine analysis of hundreds of unique samples per day. We have previously reported a robust 3 min isocratic metabolomics platform for the quantification of amino acids and the key pathways of central carbon and nitrogen metabolism. Building on this work, we describe here a 5 min reverse phase gradient followed by global, untargeted profiling of the hydrophilic metabolome. In addition to observing those metabolites measured in the 3 min run, the use of the longer gradient run here also allows for cover- age of less polar compounds such as fatty acids and acylcarnitines, both key players in mitochondrial and lipid metabolism, without a significant sacrifice in throughput. Key words Untargeted metabolomics, Mass spectrometry, Isocratic, Gradient, High-throughput 1 Introduction In the last two decades, the cost of sequencing the human genome [1] has dropped from ~$300 million US dollars to ~$1000–1500 by use of whole-exome sequencing (www.genome.gov/sequenc- ingcosts/). Technological advances have made the genome ame- nable to large-scale studies which foster personalized medicine initiatives around the globe. The introduction of high-throughput metabolomics technologies promises to achieve a similar goal. Similarly to high-throughput genome sequencing tools, it is easy to anticipate how high-throughput metabolomics applications will soon become instrumental for the characterization of clinically relevant metabolic biomarkers requiring a final validation phase in large prospective cohorts, such as in the case of large-scale person- alized medicine initiative studies. For example, high-throughput metabolomics approaches can be theoretically used to quantify
- 31. 14 metabolic markers of disease such as homocysteine levels in homo- cystinuria, an inborn error of metabolism [2], glucose levels in dia- betes [3], and lactate [4] and succinate [5] levels as readouts for base deficit and tissue hypoxia in trauma. Historically, targeted metabolomics approaches that have been translated into clinical practice are based on multiple reaction monitoring-mass spectrom- etry (MRM-MS) [6], a highly sensitive method that requires pre- selection of molecular targets of interest, therefore affording detection and quantitation of only a handful of small molecules at a time. The introduction of metabolomics strategies based on high-resolution quadrupole-time of flight (QTOF) or quadrupole- Orbitrap- based instruments has enabled investigators to perform untargeted “discovery mode” metabolomics analyses while simul- taneously quantifying metabolites of interest against external cali- bration curves or internal stable isotope-labeled standards. Over the past 5 years, several approaches have been proposed to achieve high-throughput metabolomics workflows, either based on flow-injection mass spectrometry [7], ultra-high-pressure liq- uid chromatography (UHPLC) interfaced to MS [8, 9], or matrix- assisted laser desorption/ionization imaging mass spectrometry (IMS) [10]. These approaches have made the detection and quan- tification of over 100,000 features and thousands of named metab- olites possible in less than 5 min through direct injection or isocratic/gradient-based chromatography or using in situ imaging approaches. Further advancements in terms of throughput and sensitivity are anticipated upon the recent introduction of micro- fluidic capillary electrophoresis coupled to MS for metabolomics applications [11, 12]. Optimization of rapid extraction protocols [13], automation [14], and, predictably, low-cost robotization of sample extraction protocols [15] would enable affordable process- ing of hundreds to a thousand samples per day in a cost-effective manner. Recently, we described a high-throughput isocratic chromatography- based method for rapid quantitation of hydro- philic compounds from the central carbon and nitrogen pathways [8], including amino acids [16]. Below we describe an evolution of this method, where mobile phases are coupled with agents that improve chromatographic separation and ease anionic ionization of high-energy phosphate compounds (e.g., nucleoside triphos- phates) to facilitate detection in negative ion mode. In addition, we describe a high-throughput gradient-based method that in part addresses the main limitation of the isocratic 3 min method, by making some classes of hydrophobic compounds (including acyl- conjugated carnitines, free fatty acids, and bile acids) amenable to detection and quantification. This method affords the rapid quan- titation of up to 600,000 features from a clinically relevant sample (e.g., blood extracts), resulting in hundreds to thousands of named compounds (level of confidence based on high-resolution intact Travis Nemkov et al.
- 32. 15 mass, transition fingerprints, and comparison of retention times against our in-house standard library). Given the rapidity of these approaches, 450 samples can be processed per day, making it real- istic to anticipate the potential translation of such technology into the field of clinical biochemistry. These approaches are also com- patible with the use of stable isotope-labeled internal standards, which allow for the correction of matrix-dependent ion suppres- sion effects, as well as to monitor the stability of chromatographic peak shape and retention times and the quality of signal intensity at the MS and MS/MS level [8]. The use of stable isotope-labeled standards also streamlines quantitation of compounds of interest, thus normalizing matrix or batch effects as well as instrument- dependent variability across platforms, provided the concentra- tions of the labeled standards and the endogenous metabolite from the tested matrix are within the linearity range of the MS instru- ment. Stable isotope-labeled standards can be leveraged to quan- tifymetabolicfluxesbydeterminingratiosofspiked-inisotopologues against those deriving from alternative metabolism of stable iso- tope tracers in fluxomics experiments, in like fashion to what we described recently for lactate [17] and alanine [18] synthesis downstream to [1,2,3-13 C3]glucose catabolism via glycolysis or the pentose phosphate pathway in red blood cells. 2 Materials Prepare all solutions using highest-quality solvents and reagents. We recommend LC-MS grade solvents, such as Optima™ (Fisher). Solutions can be stored at room temperature unless otherwise stated. 1. Benchtop centrifuge capable of 18,213 × g with refrigeration to 4 °C. 2. Refrigeration space (4 °C) with power outlet(s) to accommo- date vortex(es), often a cold room or chromatography refrigerator. 3. Vortex with foam microcentrifuge tube insert. 1. Water. 2. Water containing 0.1% v/v formic acid. 3. Acetonitrile. 4. Acetonitrile containing 0.1% v/v formic acid. 5. Ammonium acetate. 6. Sterile disposable vacuum filtration system (0.22 μm membrane). 2.1 General 2.2 Mobile Phases for Ultra-High- Pressure Liquid Chromatography High-Throughput Metabolomics
- 33. 16 1. Water. 2. Acetonitrile. 3. Methanol. 4. 1 L glass bottle with cap, preferably new. 5. 1 L glass graduated cylinder, preferably new. 1. Bead beater (Next Advance, Storm 24). 2. Eppendorf™ Safe-lock tubes, 1.5 mL (catalog no. 022363204). 3. Glass beads, 1.0 mm diameter (NextAdvance SKU GB10). 1. Ultra-high-pressure liquid chromatography (UHPLC) system. 2. Analytical column: Kinetex® 1.7 μm C18 100 Å, UHPLC col- umn 150 × 2.1 mm (Phenomenex catalog no. 00F-4475-AN). 3. Guard column: SecurityGuard™ ULTRA cartridge-UHPLC C18 for 2.1 mm ID columns (Phenomenex catalog no. AJ08782) with SecurityGuard™ ULTRA holder (catalog no. AJ09000). 4. Electrospray ionization (ESI) or heated ESI (HESI) source. 5. Mass spectrometer capable of high-resolution untargeted anal- ysis with MS1 acquisition. Instrument should be interfaced with a UHPLC system. 6. Data visualization software. We recommend Maven (freely available at http://genomics-pubs.princeton.edu/mzroll/ index.php) preceded by use of RawConverter (http://fields. scripps.edu/rawconv) to transform Thermo. RAW files into the .mzXML format needed for Maven. 7. Driftand/orbatchcorrectionsoftware.WeutilizeMetaboDrift, an Excel macro add-in. 3 Methods 1. Mobile phases for positive ion mode may be used as supplied. Phase A is water with 0.1% formic acid, and phase B is acetoni- trile with 0.1% formic acid. 2. Prepare 25 mL of a 2 M stock of ammonium acetate in water. Vacuum filter resulting solution to remove any insoluble material. 3. Prepare 1 L of mobile phase A for negative mode (5% acetoni- trile, 95% water, 1 mM ammonium acetate) by first diluting 2 M ammonium acetate (0.5 mL) with water (950 mL) and then adding acetonitrile (50 mL). Ensure thorough mixing and complete solubility. 2.3 Extraction Solution 2.4 Tissue Sample Preparation 2.5 Mass Spectrometry Data Acquisition and Analysis 3.1 Preparation of Chromatography Mobile Phases Travis Nemkov et al.
- 34. 17 4. Prepare 1 L of mobile phase B for negative mode (95% aceto- nitrile, 5% water, 0.5 mM ammonium acetate) by first diluting 2 M ammonium acetate (2.5 mL) with water (to the 50 mL mark) and then adding acetonitrile to the 1 L mark. Ensure thorough mixing and complete solubility. 1. Prepare stock solutions to 1000× the desired final concentra- tion in extraction solution, which should be within 1–2 orders of magnitude of the typical biological range in the analyzed sample matrix. Experimentally determined reference ranges can be found by searching for the compound of interest in Human Metabolome Database (www.hmdb.ca) and clicking on the “Concentrations” tab. Stock solutions of stable isotope- labeled standards desired for analysis are weighed and dissolved in a solvent according to the chemical nature of the compound and manufacturer instructions. Small volumes (10 μL) of concentrated formic acid or concentrated aqueous ammonium formate may be added as needed to neutralize pH. Typical stock concentrations (in mM) are given in Table 1. 2. Store stock solutions at −20 °C until use. 1. Prepare 1 L of extraction solution by adding methanol (0.5 L), acetonitrile (0.3 L), and water (0.2 L) to a clean glass gradu- ated cylinder. 2. Pour into a clean glass bottle, cap, swirl to mix, and cool to subzero temperatures. We recommend overnight storage at −20 °C or, when time-restricted, at least 30 min at −80 °C. 3. Unused extraction solution may be stored capped at −20 °C. 4. Remove stable isotope-labeled standard stock solutions from −20 °C, and thaw on ice. Ensure that the solution is com- pletely thawed and that no precipitation has occurred due to freeze- thaw. If precipitation has occurred, vortex the solution until compound has redissolved. 5. Pipette stable isotope-labeled standard into pre-cooled extrac- tion solution at desired concentration. We recommend main- taining the total volume of internal standards at or less than 5% of the total extraction solution volume to avoid diluting the organic components and thereby potentially altering extrac- tion efficiency. 1. Thaw samples on ice or in refrigerator. Pre-cool labeled micro- centrifuge tubes by immersing in ice. We utilize 1.7 mL tubes; however tube size can vary to accommodate volume of extrac- tion and compatibility with rotor. 2. Briefly vortex thawed samples to homogenize; then transfer a 20 μL aliquot by micropipette to labeled and cooled 3.2 Preparation of Stable Isotope- Labeled Standard Stock Solutions 3.3 Preparation of Metabolite Extraction Solution Containing Stable Isotope-Labeled Standards 3.4 Metabolite Extraction from Biofluid Samples High-Throughput Metabolomics
- 36. 19 microcentrifuge tube. Remaining volumes may be refrozen and stored at −80 °C. 3. Add chilled extraction solution containing applicable standard(s) to each tube. For whole blood and red blood cell samples, add 180 μL of extraction solution so that the extrac- tion ratio is 1:10. For plasma, serum, cell media, and other fluids, add 480 μL of extraction solution so that the extraction ratio is 1:25. 4. Vortex extraction samples vigorously for 30 min at 4 °C. 5. Centrifuge for 10 min at 18,213 × g (or maximum speed) at 4 °C to pellet insoluble materials such as proteins, nucleic acids, and less polar lipids, and ensure that the resulting super- natant is completely clear. If there are still particulates or cloudiness, transfer supernatants to a new tube, and repeat centrifugation. 6. Transfer ~100 μL of extract supernatant into labeled and cooled autosampler vials. Please note that, at this step, 100 μL of the extract could be dried down (e.g., under nitrogen flow or SpeedVac) prior to resuspension in equal volume of pure water (with or without 0.1% formic acid or 10 mM ammonium acetate) or methanol. Though time-consuming, this process can be automated in 96 well-plate format and ensures signifi- cantly improved chromatographic retention of hydrophilic compounds in both the isocratic and gradient-based separations. 7. For use as a quality control, prepare a pooled sample of each biological matrix being analyzed by combining 10 μL of each extract supernatant into an autosampler vial. Aliquot size may be adjusted so that the final QC sample volume is sufficient for QC runs at the beginning, middle, and end of the sequence for both positive and negative ion polarity modes or every 10–20 sample runs depending on the size of the sample set. Remaining supernatants may be stored at −80 °C (see Note 1). 8. Prepare a blank sample by adding 100 μL of extraction solu- tion from step 3 into a new, labeled autosampler vial. 9. Ensure that the sample vials do not contain air bubbles and are kept cold (in range of −20 °C in freezer to LC autosampler at 7 °C) prior to analysis. Perform MS analysis on samples. 1. Label new 1.5 mL bead-beater-safe tubes (such as Eppendorf™ Safe-lock tubes), and cool on tube rack immersed in dry ice. 2. Quickly transfer tissue samples from −80 °C freezer to rack on dry ice. Prevent tissue samples from thawing. 3.5 Metabolite Extraction from Tissue Samples High-Throughput Metabolomics
- 37. 20 3. Transfer tissue to pre-chilled, tared tube. Quickly weigh and record the mass of each tissue to the nearest 0.1 mg; then immediately return tube to dry ice (see Note 2). 4. Add chilled extraction solution with applicable standard(s) to each tube to obtain a concentration of 15 mg tissue per mL of extraction solution (see Note 3). 5. For powderized tissue, proceed to step 6. For whole pieces of tissue, add one scoop of glass beads (~50–75 μL of volume) to each sample, and bead beat for 5 min at level 4 at 4 °C. 6. Vortex samples vigorously for 30 min at 4 °C. 7. Centrifuge for 10 min at 4 °C at 18,213 × g (or maximum speed) to pellet insoluble material (see Note 4). 8. If extraction was performed at a concentration greater than 15 mg/mL (in step 4), dilute an aliquot of supernatant with the same chilled extraction solution used for initial extraction (containing standards if applicable) to obtain a concentration of 15 mg/mL. Then repeat step 7. 9. Transfer ~100 μL extract supernatant into labeled and cooled autosampler vials. 10. For use as a quality control, prepare a pooled sample of each biological matrix being analyzed by combining 10 μL of each extract supernatant into an autosampler vial. Aliquot size may be adjusted so that the final QC sample volume is sufficient for QC runs at the beginning, middle, and end of the sequence for both positive and negative ion polarity modes or every 10–20 sample runs depending on the size of the sample set. Remaining supernatants may be stored at −80 °C (see Note 1). 11. Prepare a blank sample by adding 100 μL of extraction solu- tion from step 4 into a new, labeled autosampler vial. 12. Ensure that the sample vials do not contain air bubbles and are kept cold (in range of −20 °C in freezer to LC autosampler at 7 °C) prior to analysis. 13. Perform MS analysis on samples. 1. Place sample tubes with frozen cells on dry ice. 2. Add ice cold extraction solution with applicable standard(s) to each tube such that the final extraction concentration is 2 × 106 cells per mL of solution (see Notes 5 and 6). 3. Ensure cell pellet is released from the tube into suspension, and vortex samples vigorously for 30 min at 4 °C. 4. Centrifuge for 10 min at 18,213 × g (or maximum speed) to pellet insoluble material. 5. Transfer ~100 μL extract supernatant into labeled and cooled autosampler vials. 3.6 Metabolite Extraction from Cell Samples Travis Nemkov et al.
- 38. 21 6. For use as a quality control, prepare a pooled sample of each biological matrix being analyzed by combining 10 μL of each extract supernatant into an autosampler vial. Aliquot size may be adjusted so that the final QC sample volume is sufficient for QC runs at the beginning, middle, and end of the sequence for both positive and negative ion polarity modes or every 10–20 sample runs depending on the size of the sample set. Remaining supernatants may be stored at −80 °C (see Note 1). 7. Prepare a blank sample by adding 100 μL of extraction solu- tion from step 2 into a new, labeled autosampler vial. 8. Ensure that the sample vials do not contain air bubbles and are kept cold (in range of −20 °C in freezer to LC autosampler at 7 °C) prior to analysis. 9. Perform MS analysis on samples. 1. Prepare an isocratic UHPLC-MS positive ion method with the following chromatography conditions: Flow rate 0.25 mL/ min; solvent composition 95% A, 5% B from time zero until 3 min; column temperature 25 °C; and sample compartment temperature 7 °C. Utilize a C18 column (method has been optimized for C18 column in Subheading 2.5) with a C18 guard column. Phases for positive mode should be supple- mented with 0.1% (v/v) formic acid. 2. The mass spectrometry settings for this method are resolution 70,000 (Q Exactive) or 60,000 (Q Exactive HF), scan range 65–900 m/z, maximum injection time 200 ms, 2 microscans, automatic gain control (AGC) 3 × 106 ions, ESI source voltage 4.0 kV, capillary temperature 320 °C, and sheath gas 15, aux- iliary gas 5, and sweep gas 0 (all nitrogen, measured in arbi- trary units). Optimal ESI or HESI gas and voltage parameters may be determined by observing signal intensity as settings are changed. 3. If applicable, incorporate optimized gas and voltage settings into method. 4. Ensure polarity is set to positive, and save method. 5. Prepare a method for negative mode isocratic runs by repeat- ing the above steps with the following changes: Utilize nega- tive ionization solvents supplemented with 1 mM NH4OAc instead of formic acid, and set solvent composition to 100%. A. Save negative mode method. 6. Prepare run sequence starting with several blank runs, 1–2 injections of QC sample, and then randomized analytical sam- ples with QC samples inserted every 10–20 runs. End with several blank runs before changing polarities. We recommend collecting all data in one polarity mode and then switching to the other to minimize issues with column equilibration. 3.7 UHPLC-MS Data Acquisition: Isocratic Method High-Throughput Metabolomics
- 39. 22 7. Before beginning, ensure stable column back pressure, stable MS background signal, and the absence of air bubbles or leaks in the LC system. 1. Prepare a 5 min gradient UHPLC-MS positive ion method with the following chromatography conditions: Flow rate 0.45 mL/min, column temperature 45 °C, and sample com- partment temperature 7 °C. Solvent gradient is as follows: 0–0.5 min 5% B, 0.5–1.1 min 5–95% B, 1.1–2.75 min hold at 95% B, 2.75–3 min 95–5% B, and 3–5 min hold at 5% B. Utilize a C18 column (method has been optimized for Kinetex C18 column) and a C18 guard column. Phases for positive mode should be supplemented with 0.1% (v/v) formic acid. 2. Add mass spectrometry settings to the gradient UHPLC-MS method: resolution 70,000 (Q Exactive) or 60,000 (Q Exactive HF), scan range 65–900 m/z, maximum injection time 200 ms, microscans 2, automatic gain control (AGC) 3 × 106 ions, source voltage 4.0 kV, capillary temperature 320 °C, and sheath gas 45, auxiliary gas 15, and sweep gas 0 (all nitrogen). Optimal ESI or HESI gas and voltage parameters may be determined by observing signal intensity as settings are changed. 3. If applicable, incorporate optimized gas and voltage settings into method. 4. Ensure polarity is set to positive, and save method. 5. Prepare a method for negative mode 5 min gradient runs by repeating the above steps with the following changes: Utilize solvents supplemented with 1 mM NH4OAc (instead of formic acid) to favor negative ionization. Solvent gradient is as fol- lows: 0–0.5 min 0% B, 0.5–1.1 min 0–100% B, 1.1–2.75 min hold at 100% B, 2.75–3 min 100–0% B, and 3–5 min hold at 0% B. Save negative mode method. 6. Prepare run sequence starting with several blank runs, 1–2 injections of the QC sample, and then randomized analytical samples with QC samples inserted every 10–20 runs. End with several blank runs before changing polarities. We recommend collecting all data in one polarity mode and then switching to the other to minimize issues with column equilibration. With a sufficient number of blanks to ensure column equilibration when switching modes and solvents, the instrumentation can run continuously under automation. 7. Before beginning, ensure stable column back pressure and MS background signal after the column temperature set point has been reached. Ensure also the absence of air bubbles or leaks in the LC system, then run sequence. 3.8 UHPLC-MS Data Acquisition: Gradient Method Travis Nemkov et al.
- 40. 23 4 Data Analysis 1. Quantify metabolite peak areas using software of choice. We utilize Maven, which first requires file conversion from .raw to .mzXML (accomplished using RawConverter), and assign metabolite names using the KEGG database. Table 1 contains a list of observed retention times for acylcarnitines, fatty acids, and oxylipins observed on the 5 min gradient method. Reported times are an average of 3–10 measurements in vari- ous matrices (plasma, red blood cells, saliva, pancreatic tissue) and commercial standards prepared in the background of plasma or RBCs (see Note 7). 2. Utilize QC sample data to assess instrument performance and stability throughout analytical runs. Coefficients of variation (CV) for named metabolites should be 20%. If CV values are above this threshold while other readouts (pump pressure, total signal level, etc.) do not suggest technical issues with instrument performance, consider the use of a batch- or drift- correction software. When needed, we utilize the MetaboDrift normalization algorithms separately on positive and negative ion mode raw peak area data. Once the peak areas are appro- priately normalized as needed for each polarity, the data can be merged for subsequent calculations and visualizations. 3. Use the ratio of peak areas for stable isotope-labeled standards and their unlabeled (light) counterparts to determine the abso- lute values of endogenous metabolites in the extract. Back- calculate the concentrations in the biological samples using the following equations: For biofluids: Light Area Area Heavy DF light heavy [ ] = ( )[ ]× / where DF is dilution factor—10 for RBCs and blood and 25 for serum and plasma as described above (see Note 8). For tissues: Light Area Area Heavy EC light heavy [ ] = ( )[ ]× / / 1 where EC is extraction concentration, described here as i15 mg/ mL (see Note 9). For cells: Light Area Area Heavy EC light heavy [ ] = ( )[ ]× / / 1 where EC is extraction concentration, described here as 2 × 106 cells/mL (see Note 10). High-Throughput Metabolomics
- 41. 24 5 Notes 1. If/when saved extracts are retrieved for use, spin at max speed for 10 min at 4 °C before aliquotting to AS vial. 2. Tissue extraction method is optimized for 2–20 mg pieces of tissue (wet weight). For powdered tissues, weigh and extract in similar manner, but skip the bead-beating step. 3. If the sample mass is too large to accommodate the volume for 15 mg/mL (e.g., 20 mg), all samples may be extracted at an identical higher concentration and extracts diluted in step 8 of this section. 4. When centrifuging samples in the presence of glass beads, supernatant is not often fully clarified. A supernatant aliquot may be transferred to a new tube and respun to ensure com- plete isolation of insoluble material. 5. The cell extraction concentration of 2 × 106 cells/mL is opti- mized for average-sized eukaryotic cells. Extraction of metab- olites from cells with significantly smaller diameter (e.g., T cells, neurons, bacteria) is best performed at higher concentrations. 6. If cell counts are approximated and/or imprecise, the insolu- ble protein pellet remaining after metabolite extraction may be dissolved in an appropriate chaotrope (e.g., 8 M urea) and quantified by the Bradford or BCA assay for post hoc data normalization. 7. Retention time drift and degradation of peak shape of the acyl- carnitines sometimes occur during column age perhaps due to the presence of residual ammonium acetate. For robust reten- tion time matching, use of a column naïve to negative mode additives, such as ammonium salts, is recommended. 8. The concentration of heavy standard utilized here is the final concentration in the extraction. For example, if the extraction solution contains 1 μM of standard and 20 μL of sample are mixed with 180 μL of extraction solution, the resulting con- centration of heavy standard would be 0.9 μM. 9. If concentration of heavy standard is entered in μM units and the concentration of tissue extraction is measured in mg/mL, then the resulting concentration of light (endogenous metab- olite) will be in units of μmol of metabolite per mg of tissue. 10. If concentration of heavy standard is entered in μM units and the concentration of cells in extraction is measured in cells/ mL, then the resulting concentration of light (endogenous metabolite) will be in units of μmol of metabolite per cell. Travis Nemkov et al.
- 42. 25 Conflicts of Interest A.D. and T.N. are part of Omix Technologies, Inc., and A.D. is a consultant for Hemanext, Inc. References 1. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG, Smith HO, Yandell M, Evans CA, Holt RA, Gocayne JD, Amanatides P, Ballew RM, Huson DH, Wortman JR, Zhang Q, Kodira CD, Zheng XH, Chen L, Skupski M, Subramanian G, Thomas PD, Zhang J, Gabor Miklos GL, Nelson C, Broder S, Clark AG, Nadeau J, McKusick VA, Zinder N, Levine AJ, Roberts RJ, Simon M, Slayman C, Hunkapiller M, Bolanos R, Delcher A, Dew I, Fasulo D, Flanigan M, Florea L, Halpern A, Hannenhalli S, Kravitz S, Levy S, Mobarry C, Reinert K, Remington K, Abu-Threideh J, Beasley E, Biddick K, Bonazzi V, Brandon R, Cargill M, Chandramouliswaran I, Charlab R, Chaturvedi K, Deng Z, Di Francesco V, Dunn P, Eilbeck K, Evangelista C, Gabrielian AE, Gan W, Ge W, Gong F, Gu Z, Guan P, Heiman TJ, Higgins ME, Ji RR, Ke Z, Ketchum KA, Lai Z, Lei Y, Li Z, Li J, Liang Y, Lin X, Lu F, Merkulov GV, Milshina N, Moore HM, Naik AK, Narayan VA, Neelam B, Nusskern D, Rusch DB, Salzberg S, Shao W, Shue B, Sun J, Wang Z, Wang A, Wang X, Wang J, Wei M, Wides R, Xiao C, Yan C, Yao A, Ye J, Zhan M, Zhang W, Zhang H, Zhao Q, Zheng L, Zhong F, Zhong W, Zhu S, Zhao S, Gilbert D, Baumhueter S, Spier G, Carter C, Cravchik A, Woodage T, Ali F, An H, Awe A, Baldwin D, Baden H, Barnstead M, Barrow I, Beeson K, Busam D, Carver A, Center A, Cheng ML, Curry L, Danaher S, Davenport L, Desilets R, Dietz S, Dodson K, Doup L, Ferriera S, Garg N, Gluecksmann A, Hart B, Haynes J, Haynes C, Heiner C, Hladun S, Hostin D, Houck J, Howland T, Ibegwam C, Johnson J, Kalush F, Kline L, Koduru S, Love A, Mann F, May D, McCawley S, McIntosh T, McMullen I, Moy M, Moy L, Murphy B, Nelson K, Pfannkoch C, Pratts E, Puri V, Qureshi H, Reardon M, Rodriguez R, Rogers YH, Romblad D, Ruhfel B, Scott R, Sitter C, Smallwood M, Stewart E, Strong R, Suh E, Thomas R, Tint NN, Tse S, Vech C, Wang G, Wetter J, Williams S, Williams M, Windsor S, Winn-Deen E, Wolfe K, Zaveri J, Zaveri K, Abril JF, Guigó R, Campbell MJ, Sjolander KV, Karlak B, Kejariwal A, Mi H, Lazareva B, Hatton T, Narechania A, Diemer K, Muruganujan A, Guo N, Sato S, Bafna V, Istrail S, Lippert R, Schwartz R, Walenz B, Yooseph S, Allen D, Basu A, Baxendale J, Blick L, Caminha M, Carnes-Stine J, Caulk P, Chiang YH, Coyne M, Dahlke C, Mays A, Dombroski M, Donnelly M, Ely D, Esparham S, Fosler C, Gire H, Glanowski S, Glasser K, Glodek A, Gorokhov M, Graham K, Gropman B, Harris M, Heil J, Henderson S, Hoover J, Jennings D, Jordan C, Jordan J, Kasha J, Kagan L, Kraft C, Levitsky A, Lewis M, Liu X, Lopez J, Ma D, Majoros W, McDaniel J, Murphy S, Newman M, Nguyen T, Nguyen N, Nodell M, Pan S, Peck J, Peterson M, Rowe W, Sanders R, Scott J, Simpson M, Smith T, Sprague A, Stockwell T, Turner R, Venter E, Wang M, Wen M, Wu D, Wu M, Xia A, Zandieh A, Zhu X (2001) The sequence of the human genome. Science 291(5507):1304–1351 2. Scolamiero E, Cozzolino C, Albano L, Ansalone A, Caterino M, Corbo G, di Girolamo MG, Di Stefano C, Durante A, Franzese G, Franzese I, Gallo G, Giliberti P, Ingenito L, Ippolito G, Malamisura B, Mazzeo P, Norma A, Ombrone D, Parenti G, Pellecchia S, Pecce R, Pierucci I, Romanelli R, Rossi A, Siano M, Stoduto T, Villani GRD, Andria G, Salvatore F, Frisso G, Ruoppolo M (2015) Targeted metab- olomics in the expanded newborn screening for inborn errors of metabolism. Mol BioSyst 11(6):1525–1535 3. Danaei G, Finucane MM, Lu Y, Singh GM, Cowan MJ, Paciorek CJ, Lin JK, Farzadfar F, Khang Y-H, Stevens GA, Rao M, Ali MK, Riley LM, Robinson CA, Ezzati M (2011) National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980: systematic analysis of health examination sur- veys and epidemiological studies with 370 country-years and 2·7 million participants. Lancet 378(9785):31–40 4. Odom SR, Howell MD, Silva GS, Nielsen VM, Gupta A, Shapiro NI, Talmor D (2013) Lactate clearance as a predictor of mortality in trauma patients. J Trauma Acute Care Surg 74(4):999–1004 5. Lusczek ER, Muratore SL, Dubick MA, Beilman GJ (2017) Assessment of key plasma metabolites in combat casualties. J Trauma Acute Care Surg 82(2):309–316 6. D’Alessandro A, Gevi F, Zolla L (2011) A robust high resolution reversed-phase HPLC strategy to investigate various metabolic species High-Throughput Metabolomics
- 43. 26 in different biological models. Mol BioSyst 7(4):1024–1032 7. Zampieri M, Sekar K, Zamboni N, Sauer U (2017) Frontiers of high-throughput metabo- lomics. Curr Opin Chem Biol 36:15–23 8. Nemkov T, Hansen KC, D’Alessandro A (2017) A three-minute method for high- throughput quantitative metabolomics and quantitative tracing experiments of central car- bon and nitrogen pathways. Rapid Commun. Mass Spectrom 31(8):663–673 9. Lewis MR, Pearce JTM, Spagou K, Green M, Dona AC, Yuen AHY, David M, Berry DJ, Chappell K, Horneffer-van der Sluis V, Shaw R, Lovestone S, Elliott P, Shockcor J, Lindon JC, Cloarec O, Takats Z, Holmes E, Development NJK (2016) Application of ultra-performance liquid chromatography-TOF MS for precision large scale urinary metabolic phenotyping. Anal Chem 88(18):9004–9013 10. Miura D, Fujimura Y, Yamato M, Hyodo F, Utsumi H, Tachibana H, Wariishi H (2010) Ultrahighly sensitive in situ metabolomic imag- ing for visualizing spatiotemporal metabolic behaviors. Anal Chem 82(23):9789–9796 11. Gasilova N, Yu Q, Qiao L, Girault HH (2014) On-Chip spyhole mass spectrometry for droplet- based microfluidics. Angew Chem Int Ed 53(17):4408–4412 12. Wouters S, De Vos J, Dores-Sousa JL, Wouters B, Desmet G, Eeltink S (2017) Prototyping of thermoplastic microfluidic chips and their application in high-performance liquid chro- matography separations of small molecules. J Chromatogr A 1523:224–233 13. Gehrke S, Reisz JA, Nemkov T, Hansen KC, D’Alessandro A (2017) Characterization of rapid extraction protocols for high-throughput metabolomics. Rapid Commun Mass Spectrom 31(17):1445–1452 14. Dias DA, Koal T (2016) Progress in metabolo- mics standardisation and its significance in future clinical laboratory medicine. EJIFCC 27(4):331–343 15. Liquid-handling Lego robots and experiments for STEM education and research. [date unknown]. http://journals.plos.org/plosbiol- o g y / a r t i c l e ? i d = 1 0 . 1 3 7 1 / j o u r n a l . pbio.2001413. Accessed 19 July 2017 16. Nemkov T, D’Alessandro A, Hansen KC (2015) Three-minute method for amino acid analysis by UHPLC and high-resolution quad- rupole orbitrap mass spectrometry. Amino Acids 47(11):2345–2357 17. Reisz JA, Wither MJ, Dzieciatkowska M, Nemkov T, Issaian A, Yoshida T, Dunham AJ, Hill RC, Hansen KC, D’Alessandro A (2016) Oxidative modifications of glyceraldehyde 3-phosphate dehydrogenase regulate metabolic reprogramming of stored red blood cells. Blood 128(12):e32–e42 18. D’alessandro A, Dzieciatkowska M, Nemkov T, Hansen KC (2017) Red blood cell pro- teomics update: is there more to discover? Blood Transfus 15(2):182–187 Travis Nemkov et al.
- 44. 27 Angelo D’Alessandro (ed.), High-Throughput Metabolomics: Methods and Protocols, Methods in Molecular Biology, vol. 1978, https://doi.org/10.1007/978-1-4939-9236-2_3, © Springer Science+Business Media, LLC, part of Springer Nature 2019 Chapter 3 High-Throughput Metabolomics Based on Direct Mass Spectrometry Analysis in Biomedical Research Raúl González-Domínguez, Álvaro González-Domínguez, Carmen Segundo, Mónica Schwarz, Ana Sayago, Rosa María Mateos, Enrique Durán-Guerrero, Alfonso María Lechuga-Sancho, and Ángeles Fernández-Recamales Abstract Metabolomics based on direct mass spectrometry analysis shows a great potential in biomedical research because of its high-throughput screening capability and wide metabolome coverage. This chapter contains detailed protocols to perform comprehensive metabolomic fingerprinting of multiple biological samples (serum, plasma, urine, brain, liver, spleen, thymus) by using complementary analytical platforms. The most important issues to be considered are discussed, including sample treatment, metabolomic analysis, raw data preprocessing, and data analysis. Key words Metabolomics, Direct mass spectrometry analysis, Direct infusion, Flow injection, High-throughput 1 Introduction Nontargeted metabolomic analysis is very challenging because of the difficulty of simultaneously characterizing the entire set of metabolites present in biological systems in a comprehensive man- ner. To this end, nuclear magnetic resonance (NMR) and mass spectrometry (MS) are nowadays the most commonly employed metabolomic platforms, with complementary analytical perfor- mance and applicability [1, 2]. NMR is a rapid, nondestructive, and very reproducible technique, which requires a relatively simple sample pre-treatment step and provides important structural infor- mation. However, its low sensitivity and spectral resolution seri- ously compromise the detection of individual metabolites in complex samples. On the other hand, the combination of mass spectrometry with separation techniques, such as liquid chroma- tography (LC), gas chromatography (GC), or capillary 1.1 The Potential of Direct Mass Spectrometry Analysis for High-Throughput Metabolomics
- 45. 28 electrophoresis (CE), has demonstrated a great utility to perform both qualitative and quantitative metabolomic analyses. As a con- sequence of the inherent analytical bias introduced by these separa- tion methods, the combination of complementary approaches has been postulated as the most suitable strategy to maximize metabo- lomic coverage [3, 4]. Nevertheless, the application of these hyphenated approaches significantly increases overall analysis times, which in turn can negatively affect technical stability (e.g., analyti- cal drifts in mass accuracy and/or sensitivity). To solve these limitations, the use of direct mass spectrometry analysis has emerged in recent years for high-throughput metabo- lomics [5, 6]. In this approach, sample extracts are directly intro- ducedintothemassspectrometerwithoutpreviouschromatographic or electrophoretic separation. For this purpose, the simplest instru- mental configuration is direct infusion mass spectrometry (DIMS), which employs a syringe pump to constantly deliver samples into the MS system. A second alternative is flow injection mass spec- trometry (FI-MS), based on the infusion of samples as a plug into a stream of solvent delivered by a LC pump. In general, DIMS- based metabolomics provides higher sensitivity and reproducibil- ity, but as counterpart, FI-MS is a less sample volume consuming technique and enables automation. Direct mass spectrometry anal- ysis exhibits multiple advantages over conventional MS hyphen- ated platforms thanks to the lack of a preceding separation step. First, it is noteworthy the higher technical simplicity of DI/FI-MS approaches, which are unaffected by common troubles usually found in chromatography such as column clogging and overpres- sure, changes in retention times and peak broadening due to sta- tionary phase deterioration, or irregular baseline and signal drifts due to mobile phase contamination or leaks in the LC system, among other issues. Furthermore, direct introduction of samples without applying a previous chromatographic/electrophoretic step, which is inherently biased toward specific metabolite classes depending on the separation mode employed, usually provides more comprehensive metabolome coverage. Therefore, direct MS analysis has demonstrated huge potential for high-throughput metabolomic screening, of particular interest when dealing with large sample populations [6]. Metabolomics based on direct MS analysis has gained great popu- larity for mapping metabolic alterations associated with disease pathogenesis and progression in a holistic manner. For instance, various authors have previously described the application of MS fingerprinting to investigate the underlying pathology of different oncological disorders, including lung cancer [7–9], prostate cancer [10], kidney cancer [11], pancreatic cancer [12], and colorectal cancer [13–15]. Thereby, it has been demonstrated the occurrence 1.2 Application of Direct MS-Based Metabolomics in Biomedical Research Raúl González-Domínguez et al.
- 46. 29 of multiple metabolic dysregulations affecting to a wide range of pathways depending on the cancer type and stage, such as impaired energy metabolism, altered homeostasis of lipids, oxidative stress, and many others. DI/FI-MS-based metabolomics has also been employed to identify the characteristic metabotype associated with abnormal glucose metabolism in patients with impaired glucose tolerance [16] and insulin resistance [17] and in volunteers sub- jected to an oral sugar challenge [18]. Other studies performed on the last years dealt with the discovery of potential diagnostic bio- markers for Chagas disease [19], schistosomiasis [20], aspergillosis [21], Crohn’s disease [22], induced stress [23], and preeclampsia [24]. However, it should be noted that the application of direct MS has mainly focused on the metabolomic study of Alzheimer’s disease (AD). González-Domínguez et al. optimized a metabolo- mic approach based on a two-step extraction procedure and subse- quent DIMS analysis of serum samples to elucidate pathological hallmarks associated with AD development and progression, in both human cohorts [25, 26] and transgenic APP×PS1 mouse models [27, 28]. Later, this methodology was adapted to get deeper insights into lipidomic [29] and phospholipidomic [30] alterations occurring in this neurodegenerative disorder. Furthermore, authors also described the implementation of this DIMS-based metabolomic platform for the analysis of other bio- logical matrices, such as urine [31], brain [32], and other periph- eral organs (i.e., liver, kidney, thymus, spleen) [33]. Thereby, a comprehensive characterization of AD-related metabolic distur- bances was accomplished, comprising multiple impairments in the homeostasis of neurotransmitters, various lipid classes (e.g., phos- pholipids, glycerolipids, cholesterol, fatty acids), and antioxidants, among others. Interestingly, these findings were then validated by applying orthogonal hyphenated approaches, such as UHPLC-MS [34–38], GC-MS [35–39], and CE-MS [40], thus demonstrating the reliability of direct MS for metabolomic fingerprinting. In this line, Lin et al. reported the use of DIMS techniques to address metabolic alterations in the hippocampus [41] and cerebellum [42] of CRND8 transgenic mice, revealing significant perturba- tions in the regulation of neuroinflammatory processes. Finally, it is also noteworthy the recent development of a novel metabolomic approach based on flow injection analysis using an atmospheric pressure photoionization mass spectrometer (FI-APPI-MS) with the aim to complement the ionization capabilities of the electro- spray source (ESI), usually employed in DIMS-based metabolo- mics [43]. This methodology was successfully applied to serum samples from AD patients [43] and APP×PS1 transgenic mice [27], which provided complementary findings to those obtained with DI-ESI-MS platforms. Metabolomics Based on Direct MS Analysis
- 47. 30 2 Materials Prepare all samples and solutions by using high-purity solvents and reagents, including methanol, ethanol, chloroform, dichlorometh- ane, formic acid, ammonium formate, ammonium acetate, and toluene. Deionized water can be obtained from a Milli-Q Gradient system (Millipore, Watford, UK). For identification purposes, authentic metabolite standards must be purchased if available. Blood samples are extracted by venipuncture of the antecubital region after 8 h of fasting. All samples must be collected at the same time of the day in order to avoid the influence of the circa- dian rhythm. To obtain serum samples, blood is immediately cooled and protected from light for 30 min to allow clot retrac- tion. On the contrary, plasma collection requires the addition of anticoagulants to the corresponding blood collection tube (see Note 1). Then, blood samples are centrifuged at 4000 × g for 10 min at 4 °C, and the resulting serum/plasma is divided into aliquots and frozen at −80 °C until analysis. On the other hand, urine samples are directly collected in sterile plastic containers known not to release plasticizers or other compounds into the sample (see Note 2). Before storing at −80 °C, it is recommended to perform a mild centrifugation step at 2000 × g for 10 min at 4 °C in order to remove human cells and bacteria, which may break upon sample freezing. The study population recruitment must be performed in accordance with the principles contained in the Declaration of Helsinki. After reception, mice must be acclimated for 3 days in rooms with a 12-h light/dark cycle at 20–25 °C, with water and food available ad libitum. Then, mice are individually anesthetized by isoflurane inhalation and exsanguinated by cardiac puncture. These blood samples are processed as previously described (Subheading 2.2.1) to obtain serum and plasma. Subsequently, brain and other periph- eral organs (liver, kidney, spleen, and thymus) are rapidly removed and rinsed with saline solution (0.9% NaCl, w/v). Furthermore, brains are dissected into hippocampus, cortex, cerebellum, stria- tum, and olfactory bulbs. Urines are directly collected in sterile plastic containers and centrifuged at 2000 × g for 10 min at 4 °C (see Note 2). Finally, all these samples are snap-frozen in liquid nitrogen and stored at −80 °C until analysis. Handling of animals must be performed according to the directive 2010/63/EU stipu- lated by the European Community. Metabolomic methods described in this chapter have previously been validated in two high-resolution mass spectrometry systems: (1) quadrupole-time-of-flight (Q-TOF) mass spectrometer, model 2.1 Chemicals and Standards 2.2 Sample Collection 2.2.1 Human Samples 2.2.2 Animal Model Samples 2.3 Instrumentation Raúl González-Domínguez et al.
- 48. 31 QSTAR XL Hybrid system (Applied Biosystems), equipped with integrated syringe pump, Accela LC system (Thermo Fisher Scientific), and syringe pump model KDS 100 (KD Scientific); (2) Q-TOF mass spectrometer, model Xevo G2-S, coupled to an ACQUITY UPLC™ system (Waters). For accurate mass measure- ment, the QSTAR XL system is daily calibrated using renin and taurocholic acid in positive and negative ion modes, respectively. In the Xevo G2-S mass spectrometer, all spectra are acquired using a reference lock mass (leucine enkephalin) to ensure accuracy and reproducibility. A cryogenic homogenizer model Freezer/Mill 6770 (SPEX SamplePrep) and a pellet mixer (VWR International) are used to perform metabolomic extraction of tissues. All samples are centri- fuged in a centrifuge model Eppendorf 5804R. A vacuum mani- fold (Visiprep, Supelco) is employed to perform solid phase extraction (SPE) of urine samples. 3 Methods A two-step extraction protocol is applied to serum and plasma samples with the aim to fractionate the blood metabolome in two complementary extracts: (1) aqueous extract containing more polar compounds (i.e., low molecular weight metabolites and phospholipids) and (2) lipophilic extract, mainly composed of neu- tral lipids [25]. 1. Thaw serum/plasma samples on an ice bath. 2. Add 400 μL of methanol/ethanol (1:1, v/v) to 100 μL of serum/plasma into an Eppendorf tube placed on an ice bath. 3. Stir the mixture during 5 min, using a vortex mixed or an orbital shaker (at 4 °C if possible), in order to precipitate proteins. 4. Centrifuge samples at 4000 × g for 10 min at 4 °C. 5. Transfer the supernatant to a new tube, take it to dryness under nitrogen stream (see Note 3), and reconstitute the resulting residue with 100 μL of methanol/water (80:20, v/v) contain- ing 0.1% (v/v) formic acid (polar extract) (see Note 4). 6. Add 400 μL of chloroform/methanol (1:1, v/v) to the protein precipitate obtained in step 3 placed on an ice bath. 7. Shake vigorously during 5 min using a vortex mixer (at 4 °C if possible). 8. Centrifuge samples at 10,000 × g for 10 min at 4 °C. 3.1 Sample Treatment 3.1.1 Metabolomic Extraction of Serum/ Plasma Samples Metabolomics Based on Direct MS Analysis
- 49. 32 9. Take to dryness the supernatant under nitrogen stream (see Note 3) and reconstitute with 100 μL of dichloromethane/ methanol (60:40, v/v) containing 0.1% (v/v) formic acid and 10 mM ammonium formate (lipophilic extract) (see Note 4). To get a deeper insight into the blood circulating lipidome, serum/ plasma samples are treated following a modification of the Bligh- Dyer extraction protocol, which allows analyzing neutral lipids (e.g., diglycerides, triglycerides, cholesterol derivatives), usually not detected by means of conventional metabolomic approaches [29]. 1. Thaw serum/plasma samples on an ice bath. 2. Mix 50 μL of serum/plasma with 150 μL of methanol contain- ing 30 mM ammonium acetate in an Eppendorf tube placed on an ice bath. 3. Shake vigorously during 1 min using a vortex mixer (at 4 °C if possible). 4. Add 200 μL of pure chloroform. 5. Shake vigorously during 1 min using a vortex mixer (at 4 °C if possible). 6. Centrifuge samples at 10,000 × g for 10 min at 4 °C. 7. Transfer the organic layer to a new tube and keep for further analysis. Direct MS fingerprinting of urine samples requires the application of a pre-treatment step with the aim to reduce the high salinity of this biological fluid, which may compromise further analysis due to matrix effects (e.g., ion suppression). For this purpose, two differ- ent approaches can be applied after thawing urine samples on an ice bath, as describe elsewhere [31]. 1. Sample dilution. (a) Centrifuge urine samples at 4000 × g for 10 min at 4 °C. (b) Dilute tenfold with methanol/water (1:1, v/v). 2. Mixed mode solid phase extraction (MM-SPE). (a) Condition SPE cartridges (Isolute Multimode, 500 mg of sorbent) with 2 mL of pure methanol. (b) Load 1.5 mL of urine, previously centrifuged at 4000 × g for 10 min at 4 °C. (c) Clean the SPE cartridge with 2 mL of deionized water. (d) Carry out the elution with: (1) 0.5 mL of methanol, (2) 0.5 mL of 10 mM ammonium acetate (pH = 3) in metha- nol, and (3) 0.5 mL of 5% (v/v) ammonia in methanol. 3.1.2 Lipidomic Extraction of Serum/ Plasma Samples 3.1.3 Metabolomic Extraction of Urine Samples Raúl González-Domínguez et al.
- 50. 33 All eluents are pumped through the SPE cartridges by using a vacuum manifold (Visiprep, Supelco). Extraction of tissue samples, including the hippocampus, brain cortex, cerebellum, striatum, olfactory bulbs, liver, kidney, spleen, and thymus, is carried out in accordance with the methodology previously optimized by González-Domínguez et al. [33, 36]. 1. Cryo-homogenize tissues to obtain a fine powder using a Freezer/Mills 6770 homogenizer during 30 s at rate of 10 strokes per second (see Note 5). 2. Weigh 30 mg of homogenized tissue (or the entire sample for smaller organs) in 1.5 mL Eppendorf tubes. Samples are kept frozen until the addition of the extraction solvent. 3. Add 10 μL mg−1 of pre-cooled 0.1% (v/v) formic acid in meth- anol (−20 °C). 4. Use a pellet mixer (VWR international) to disrupt cells during 2 min. Perform this extraction step inside an ice bath to avoid sample heating by friction. 5. Centrifuge at 10,000 × g for 10 min at 4 °C. 6. Transfer the supernatant to a new tube and keep for further analysis (polar extract). 7. Add 10 μL mg−1 of pre-cooled chloroform/methanol (2:1, v/v), containing 0.1% (v/v) formic acid and 10 mM ammo- nium formate (−20 °C), to the pellet obtained in step 5 (see Note 6). 8. Repeat steps 4–6 to obtain the corresponding lipophilic extracts. For each biological matrix, prepare quality control (QC) samples by pooling equal volumes of each individual sample. The analysis of these QC samples allows monitoring the stability and perfor- mance of the system along the analysis period [44]. These samples must be analyzed at the start of the run in order to equilibrate the analytical system as well as at intermittent points throughout the sequence to monitor system stability. Sample extracts are directly infused into the QTOF-MS system (QSTAR XL, Applied Biosystems) by using an integrated syringe pump operating at 5 μL min−1 flow rate. Mass spectra are obtained by electrospray ionization (ESI) in positive and negative ionization modes in separate runs, as described elsewhere [25, 27, 32, 33]. Briefly, full-scan spectra are acquired for 0.2 min in the m/z range 50–1100 Da, with 1.005 s scan time. The source temperature is maintained at 60 °C, and high-purity nitrogen is used as curtain and nebulizer gas at flow rates of 1.13 L min−1 and 1.56 L min−1 , 3.1.4 Metabolomic Extraction of Tissue Samples 3.1.5 Preparation of Quality Control Samples 3.2 Direct MS Metabolomic Fingerprinting 3.2.1 Direct Infusion Electrospray Ionization Mass Spectrometry (DI-ESI-MS) Metabolomics Based on Direct MS Analysis
- 51. 34 respectively. The ion spray voltage (IS), declustering potential (DP), and focusing potential are fixed at 3300/−4000 V, 60/−100 V, and 250/−250 V in positive and negative ion modes, respectively. For FI-ESI-MS analysis, sample extracts are introduced into the mass spectrometer (Xevo G2-S) by flow injection using an ACQUITY UPLC™ system [18]. Methanol containing 0.1% (v/v) formic acid and 10 mM ammonium formate is delivered at 200 μL min−1 as flow injection solvent. Mass spectra are obtained in positive and negative ion modes, by injecting 5 μL of sample. Full-scan spectra are acquired in the m/z range 50–1100 Da dur- ing 1.5 min (scan time 0.5 s). Capillary and cone voltages are set at 3.0 kV and 30 V, respectively, and source temperature is main- tained at 120 °C. Desolvation gas flow (high-purity nitrogen) is fixed at 250 L h−1 , at 150 °C of temperature. The application of FI-APPI-MS fingerprinting requires the cou- pling of a high-resolution mass spectrometer equipped with an atmospheric pressure photoionization source (QSTAR XL), a LC system for flow injection of samples (Accela), and a syringe pump to deliver the dopant reagent for photospray ionization. Following the previously optimized protocol, methanol is used as flow injec- tion solvent at 50/100 μL min−1 in positive and negative ion modes, respectively [43]. On the other hand, toluene is delivered at 20/40 μL min−1 as photoionization dopant, in both ionization modes. Mass spectra are obtained in positive and negative ion modes by injecting 5 μL of sample. Full-scan spectra are acquired in the m/z range 50–1100 Da, with 1.005 s of scan time. The ion spray voltage (IS), declustering potential (DP), and focusing potential are fixed at 1500/−2300 V, 50/−50 V, and 250/−250 V in positive and negative ion modes, respectively. The source tem- perature is maintained at 400 °C, and the gas flows (high-purity nitrogen) are fixed at 1.13 L min−1 for curtain gas, 1.50 L min−1 for nebulizer gas, 3.0 L min−1 for heater gas, and 1 L min−1 for lamp gas. In direct MS-based metabolomics, data preprocessing only implies the application of a single peak detection step in order to convert original files into a two-dimensional data matrix of spectral peaks and their intensities. For this purpose, different software must be employed depending on the MS system employed to acquire metabolomic fingerprints. 1. Metabolomic data obtained by DI-ESI-MS and FI-APPI-MS analysis are submitted to peak detection by using the MarkerView™ software (Applied Biosystems). For this, all peaks above the noise level (10 counts, determined empirically 3.2.2 Flow Injection Electrospray Ionization Mass Spectrometry (FI-ESI-MS) 3.2.3 Flow Injection Atmospheric Pressure Photoionization Mass Spectrometry (FI-APPI-MS) 3.3 Raw Data Processing Raúl González-Domínguez et al.
- 52. 35 from experimental spectra) are selected and binned in intervals of 0.1 Da. For FI-MS fingerprints, this processing step is lim- ited to scans within the apex of infusion profiles [43]. 2. Metabolomic data obtained by FI-ESI-MS analysis are sub- mitted to peak detection by using the MarkerLynx™ soft- ware (Waters). To this end, all peaks above the noise level (200 counts, determined empirically from experimental spectra) are selected and binned in intervals of 0.02 Da. This processing step is limited to scans within the apex of infusion profiles [18]. Multiple statistical packages can be used to analyze metabolomic data. In this section, we focus on the MetaboAnalyst 3.0 web tool (http://www.metaboanalyst.ca/), the SIMCA-P™ software (ver- sion 11.5, UMetrics AB, Umeå, Sweden), and the STATISTICA 8.0 software (StatSoft, Tulsa, USA). Thereby, a conventional pipe- line for metabolomic data analysis involves the following steps: 1. Data filtering by choosing masses present in at least 50% of samples. 2. Estimation of missing values by means of the k-nearest neigh- bor (KNN) algorithm. 3. Data filtering based on the interquartile range (IQR) in order to remove variables showing little variance within the study population. 4. Data normalization according to the total area sum. 5. Pareto scaling to reduce the relative importance of larger val- ues, and logarithmic transformation in order to approximate a normal distribution [45]. 6. Application of multivariate statistical techniques with the aim to visualize general trends and detect discriminant patterns among the study groups: principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA). 7. Application of univariate statistical techniques to identify dis- criminant metabolites among the study groups: t-test or one- way analysis of variance (ANOVA) (see Note 7), with Bonferroni or false discovery rate (FDR) correction for multiple testing. 4 Notes 1. Various anticoagulants can be used to obtain plasma samples, including citrate, ethylenediaminetetraacetic acid (EDTA), and heparin. Advantages and drawbacks of using each anticoagulant have recently been reviewed by Hernandes et al. [46]. 3.4 Data Analysis Metabolomics Based on Direct MS Analysis
- 53. 36 2. Preservatives can be added to enhance the stability of urine samples (e.g., borate, sodium azide), which are prone to bacte- rial contamination during collection and storage. However, the general recommendation of the European Consensus Expert Group is to avoid the use of additives [47]. 3. The evaporation of sample extracts can also be carried out by using vacuum concentrators (e.g., SpeedVac), if available. 4. Depending on the sensitivity of the MS system employed, sam- ple extracts can be directly analyzed without performing a pre- concentration step. 5. Smaller organs (e.g., hippocampus, striatum, olfactory bulbs) can be directly extracted without prior cryo-homogenization. 6. This second extraction step is of great interest for studying peripheral organs (i.e., liver, kidney, spleen, thymus), due to the high content of neutral lipids in these tissues. However, steps 7 and 8 can be omitted when brain tissue is analyzed, as previously reported [32, 36]. 7. Nonparametric methods must be used when variables do not show normal distribution (checked by normal probability plots) and variances are not homogeneous (checked by Levene’s test). References 1. Emwas AHM, Salek RM, Griffin JL, Merzaban J (2013) NMR based metabolomics in human disease diagnosis: applications, limitations, and recommendations. Metabolomics 9:1048–1072 2. Theodoridis G, Gika HG, Wilson ID (2011) Mass spectrometry-based holistic analytical approaches for metabolite profiling in systems biology studies. Mass Spectrom Rev 30:884–906 3. González-Domínguez R, Sayago A, Fernández- Recamales A (2017) Metabolomics in Alzheimer’s disease: the need of complemen- tary analytical platforms for the identification of biomarkers to unravel the underlying pathol- ogy. J Chromatogr B Analyt Technol Biomed Life Sci 1071:75–92 4. Gonzalez-Dominguez A, Duran-Guerrero E, Fernandez-Recamales A, Lechuga-Sancho AM, Sayago A, Schwarz M, Segundo C, Gonzalez- Dominguez R (2017) An overview on the importance of combining complementary ana- lytical platforms in metabolomic research. Curr Top Med Chem 17:3289–3295 5. Draper J, Lloyd AJ, Goodacre R, Beckmann M (2013) Flow infusion electrospray ionisation mass spectrometry for high throughput, non- targeted metabolite fingerprinting: a review. Metabolomics 9:S4–S29 6. González-Domínguez R, Sayago A, Fernández- Recamales A (2017) Direct infusion mass spec- trometry for metabolomic phenotyping of diseases. Bioanalysis 9:131–148 7. Guo Y, Wang X, Qiu L, Qin X, Liu H, Wang Y, Li F, Wang X, Chen G, Song G, Li F, Guo S, Li Z (2012) Probing gender-specific lipid metab- olites and diagnostic biomarkers for lung can- cer using Fourier transform ion cyclotron resonance mass spectrometry. Clin Chim Acta 414:135–141 8. Cameron SJ, Lewis KE, Beckmann M, Allison GG, Ghosal R, Lewis PD, Mur LA (2016) The metabolomic detection of lung cancer bio- markers in sputum. Lung Cancer 94:88–95 9. Lokhov PG, Kharybin ON, Archakov AI (2012) Diagnosis of lung cancer based on direct-infusion electrospray mass spectrometry of blood plasma metabolites. Int J Mass Spectrom 309:200–205 10. Lokhov PG, Dashtiev MI, Moshkovskii SA, Archakov AI (2010) Metabolite profiling of blood plasma of patients with prostate cancer. Metabolomics 6:156–163 Raúl González-Domínguez et al.
- 54. 37 11. Lin L, Yu Q, Yan X, Hang W, Zheng J, Xing J, Huang B (2010) Direct infusion mass spec- trometry or liquid chromatography mass spec- trometry for human metabonomics? A serum metabonomic study of kidney cancer. Analyst 135:2970–2978 12. Ritchie SA, Akita H, Takemasa I, Eguchi H, Pastural E, Nagano H, Monden M, Doki Y, Mori M, Jin W, Sajobi TT, Jayasinghe D, Chitou B, Yamazaki Y, White T, Goodenowe DB (2013) Metabolic system alterations in pancreatic cancer patient serum: potential for early detection. BMC Cancer 13:416 13. Ritchie SA, Ahiahonu PW, Jayasinghe D, Heath D, Liu J, Lu Y, Jin W, Kavianpour A, Yamazaki Y, Khan AM, Hossain M, Su-Myat KK, Wood PL, Krenitsky K, Takemasa I, Miyake M, Sekimoto M, Monden M, Matsubara H, Nomura F, Goodenowe DB (2010) Reduced levels of hydroxylated, poly- unsaturated ultra long-chain fatty acids in the serum of colorectal cancer patients: implica- tions for early screening and detection. BMC Med 8:13 14. Li F, Qin X, Chen H, Qiu L, Guo Y, Liu H, Chen G, Song G, Wang X, Li F, Guo S, Wang B, Li Z (2013) Lipid profiling for early diagno- sis and progression of colorectal cancer using direct-infusion electrospray ionization Fourier transform ion cyclotron resonance mass spec- trometry. Rapid Commun Mass Spectrom 27:24–34 15. Williams MD, Zhang X, Park JJ, Siems WF, Gang DR, Resar LM, Reeves R, Hill HH Jr (2015) Characterizing metabolic changes in human colorectal cancer. Anal Bioanal Chem 407:4581–4595 16. Lokhov PG, Trifonova OP, Maslov DL, Balashova EE, Archakov AI, Shestakova EA, Shestakova MV, Dedov II (2014) Diagnosing impaired glucose tolerance using direct infu- sion mass spectrometry of blood plasma. PLoS One 9:e105343 17. Renner S, Römisch-Margl W, Prehn C, Krebs S, Adamski J, Göke B, Blum H, Suhre K, Roscher AA, Wolf E (2012) Changing meta- bolic signatures of amino acids and lipids dur- ing the prediabetic period in a pig model with impaired incretin function and reduced β-cell mass. Diabetes 61:2166–2175 18. González-Domínguez R, Mateos RM, Lechuga-Sancho AM, González-Cortés JJ, Corrales-Cuevas M, Rojas-Cots JA, Segundo C, Schwarz M (2017) Synergic effects of sugar and caffeine on insulin-mediated metabolomic alterations after an acute consumption of soft drinks. Electrophoresis 38:2313–2322 19. Antunes LC, Han J, Pan J, Moreira CJ, Azambuja P, Borchers CH, Carels N (2013) Metabolic signatures of triatomine vectors of Trypanosoma cruzi unveiled by metabolomics. PLoS One 8:e77283 20. Ferreira MS, de Oliveira DN, de Oliveira RN, Allegretti SM, Catharino RR (2014) Screening the life cycle of Schistosoma mansoni using high-resolution mass spectrometry. Anal Chim Acta 845:62–69 21. de Francisco TMG, Zaramella IF, Gasparetto JC, Cerqueira LB, Piantavini MS, Pontarolo R, Campos FR (2009) Rapid detection of asper- gillosis in immunocompromised patients using DIMS and chemometric analysis. Anal Methods 7:6346–6351 22. Jansson J, Willing B, Lucio M, Fekete A, Dicksved J, Halfvarson J, Tysk C, Schmitt- Kopplin P (2009) Metabolomics reveals meta- bolic biomarkers of Crohn's disease. PLoS One 4:e6386 23. Lorenzo-Tejedor M, de la Cámara C, Lopez- Anton R, Bailon R, Aguiló J, Bernal-Ruiz ML (2015) Direct infusion electrospray mass spec- trometry as a new non-invasive tool for serum metabolomics in induced-stress subjects. Eur J Psychiat 29:259–275 24. Anand S, Young S, Esplin MS, Peaden B, Tolley HD, Porter TF, Varner MW, D'Alton ME, Jackson BJ, Graves SW (2016) Detection and confirmation of serum lipid biomarkers for pre- eclampsia using direct infusion mass spectrom- etry. J Lipid Res 57:687–696 25. González-Domínguez R, García-Barrera T, Gómez-Ariza JL (2014) Using direct infusion mass spectrometry for serum metabolomics in Alzheimer’s disease. Anal Bioanal Chem 406:7137–7148 26. González-Domínguez R, García-Barrera T, Gómez-Ariza JL (2012) Metabolomic approach to Alzheimer’s disease diagnosis based on mass spectrometry. Chem Papers 66:829–835 27. González-Domínguez R, García-Barrera T, Vitorica J, Gómez-Ariza JL (2015) Application of metabolomics based on direct mass spec- trometry analysis for the elucidation of altered metabolic pathways in serum from the APP/ PS1 transgenic model of Alzheimer’s disease. J Pharm Biomed Anal 107:378–385 28. González-Domínguez R, García-Barrera T, Vitorica J, Gómez-Ariza JL (2015) Metabolomic research on the role of interleu- kin- 4 in Alzheimer’s disease. Metabolomics 11:1175–1183 29. González-Domínguez R, García-Barrera T, Gómez-Ariza JL (2014) Metabolomic study of Metabolomics Based on Direct MS Analysis
- 55. Exploring the Variety of Random Documents with Different Content
- 56. We had not journeyed far beyond Lincoln Park before we approached the State Asylum for the Acute Insane. From the beginning of my pilgrimage, I had kept a sharp lookout for Insane Asylums, always passing them after dark, but Mac argued that the public had by this time found me harmless, and advised me to call. So I did. A patient has arrived, some one called to an attendant. I was startled, but soon recovered my equilibrium, when I observed several doctors and nurses rush out of doors to a carriage at the porch. The lunatic having been safely deposited in one of the wards, the Superintendent then welcomed me, and persuaded me to accept his invitation to visit and inspect the institution. There was only one department that interested me. I had no sooner entered the kitchen than my omnivorous eye caught the pie- ocine stratum of a well-developed pie, and my curiosity led me to inquire if it were made by a lunatic. Why, most certainly, Professor! exclaimed the Superintendent. What's the matter with it? As far as appearances go, I think it's all right—doesn't look different from any other pie I've seen and eaten. Shouldn't think a crazy man could make a decent pie, though; did he do it all alone, without anybody watching him? Oh no, we employ a sane cook to supervise the cooking, explained the officer, much to my satisfaction. Will you have a piece? he asked. Y-y-y-y-yes, I said incredulously, if you are sure there is no danger of insanity being transferred to me by such a delectable agency. The head cook then butchered the great pie into quarters, and the Superintendent said, Help yourself, boys. I gathered up the juicy quarter, and saying, My good sir, you have heard of dog eat dog, you shall now witness Pye eat pie. I proceeded to devour it. I couldn't recollect ever having eaten better pie; I was almost prompted to ask the cook to slaughter another, but, instead, carried the remaining quarter out to Mac A'Rony.
- 57. When we had left the asylum, I could not help but remark the scrutiny with which each man regarded the other. At length we went into camp near a farm house, where we certainly acquitted ourselves in a manner to arouse the suspicions of any sane observer. We put our sleeping-bag on the ground outside of the tent, built a fire close to the tent on the windward side while a strong breeze was blowing, cooked creamed potatoes in the coffee pot, and steeped tea in the frying pan; and Coonskin tied all three donkeys and the dog to a small sapling by their tails. I felt sure that insanity was breaking out in our party in an aggravated form, and congratulated Cheese, Damfino and Don for not having eaten infected pie. Camp Lunatic, as we called it was visited by the owner of the farm, a hospitable German, who had a large family. He gave us a generous donation of corn-cobs for fuel, milk, butter, fresh eggs, and water, then introduced his wife and children. I asked him how he came to have such a large family. He explained that he had a large farm and couldn't afford hired help, and he thought the best way to remedy the difficulty was to rear boys to help him. He looked hopeful, although he had eight girls, no boys. Supper over, the farmer conferred on me every possible honor, even letting me hold his youngest girl, a child of ten months. He said, enthusiastically, he was going to name his boy after me; the wife smiled heroically. To cap the climax, I was asked to write my name in the big family Bible. The book was in German. My host opened it to a blank page, and, without comment, I inscribed my name underneath the strangely printed heading—Gestorben, thus pleasing the whole family. When we reached our tent, Barley began to find fault with me. What for did yuse want to write your name on de Gestorben page? he asked seriously. Dat means bad luck, dat does. And why? I inquired, puzzled. Gestorben is German and means death, yuse crazy loon! he returned. It's de lunatic pie dat's workin' already; wese all goin' crazy.
- 58. Next day was hot. In the afternoon my party rested three hours in the shade of a peach orchard, where we were treated to ice cream by the kind lady of the house close by. It was about 105 miles from Lincoln to Hastings, and we covered it in five days. Threading the villages of Exeter, Crete, Friend, and Dorchester, we arrived in Grafton, where I caught my courier in a dishonest trick, and discharged him. The party reached Hastings Thursday, June 17, where I purchased a saddle for Coonskin. Detained by a thunderstorm, we passed a miserable night in close quarters. Next morning, Mac pranced about like a circus donkey, and trailed to Kearney in a manner almost to wind his fellows. Before leaving Hastings, the Superintendent of the Asylum for the Chronic Insane, three miles out of town, telephoned me to stop and dine with him. On this occasion I rode into the asylum grounds without hesitation or nervousness. You must earn your grub, according to contract, Professor, said the Superintendent, when the greetings were over, pointing to a wood-pile in the rear of the building. As soon as I fairly began to comply with the suggestion his young lady secretary, the daughter of a deceased and much esteemed congressman, trained a camera on me and the axe and secured a picture. I was then notified I had more than earned my dinner, and was escorted into the family dining-room, where an enjoyable repast was accorded me, after which, some twenty wardens and matrons purchased photos at double price. Then I resumed the journey with more heartfelt blessings than had been expressed to me on similar occasions. The trail was superb. But an intensely hot spell followed, and made all of us perspire. Two days of hard travel brought us to the old Government Reservation of Ft. Kearney, established by Gen. Fremont on his historic overland trip to California in pioneer days. The fort has long since been abandoned. There the Mormons camped for a short period after leaving Council Bluffs. Next evening, I made my camp on the site of the notorious Dirty Woman's Ranch of early days, and spent a Sunday in delightful rest
- 59. and recreation in the shade of the grove of wide-spreading elms and cotton-woods that sighed mournfully over the deserted scene. larger A. Trail through the timber. B. He had caught a nice mess. C. Climbing Pike's Peak. We crossed the long, low bridge over the Platte, early in the morning. It required nearly an hour and all our wits and energies to get the donkeys across, even after blindfolding them. And when my
- 60. party ambled into Kearney, that sultry, dusty June day, grimy with dirt and perspiring, we all were in ripe condition for a swim. The little city looked to be about the size of Hastings, but did not show the same enterprise and thrift. In fact, the inhabitants ventured out in the broiling sun with an excusable lack of animation, and seemer to show no more interest in their local affairs than they did in Pye Pod's pilgrimage. It was here I first saw worn the Japanese straw helmet. It served as a most comfortable and effective sun-shade, and purchasing a couple, we donned them at once. Kearney is said to be the half-way point, by rail, between New York and San Francisco. My diary, however, showed I had covered fully two thousand miles of my overland journey; I had consumed 227 days, with only one hundred and thirty-four days left me, the prospects of accomplishing the feat in schedule time looked dubious enough. The great Watson Ranch, when my donkey party arrived, was experiencing its busiest season. But, while the male representatives were in the fields, the good matron in charge of the house made us welcome and treated us to cheering bowls of bread and milk. When Mr. Watson, Jr., arrived, he showed us about the place and enlightened me about alfalfa, of which he had over a thousand acres sown; fifty hired hands were busy harvesting it. For a week or two we had, for the most part, been trailing through the perfumed prairies at an invigorating altitude ranging from two thousand to nearly three thousand feet, inhaling the fresh, pure air, gazing on the flower-carpeted earth, and enjoying a constant shifting of panoramic scenes of browsing herds, and bevies of birds, and occasional glimpses of the winding Platte and the sand dunes beyond. The cities and villages, that formed knots in the thread of our travels on the plains, came into view like the incoming ships from the sea. At first one spied a white church-steeple in the distance like a pointed stake in the earth only a mile away, but soon the chimneys and roofs and finally door-yard fences would come into view, then what we thought a village, nearby, proved to be, as we journeyed onward, a town of much greater size seven or eight miles beyond the
- 61. point of calculation. The crossbars on the telegraph poles, along the straight and level tracks of the Union Pacific, formed in the eye's dim perspective a needle, as they seemed to meet with the rails on the horizon. Little bunches of trees, scattered miles apart and then overtopped by the spinning wheel of an air motor, indicated the site of a ranch-house where we might procure water. The trail ahead became lost in a sea of flowers and grasses. From time to time, as I dismounted to ease myself and little steed I picked from the stirrups a half dozen kinds of flowers, ensnared as my feet brushed through the grasses. Great beds of blood-red marshmallows; natural parterres of the wax-like blooms of the prickly pear; scattering stems of the flowery thistle with white corollas as large as tulips; and wild roses and daisies of all shades and colors—the white and pink, and the white wild roses being the first I ever saw; these with varicolored flowers of all descriptions were woven into the prairie grasses and likened the far-reaching plain to a great Wilton carpet enrolled from the mesa to the river. Some of the sunsets were gorgeous. At times, the western sky glowed like a prairie fire; and the sunrises were not less magnificent. Sometimes, we were overtaken by severe electric storms, and obliged to pitch the tent in a hurry. When the lightning illuminates the plains at night, the trees and the distant towns are brought into fantastic relief against the darkness, like the shifting pictures of a stereopticon. A flash of lightning to the right reveals a church or school-house, to the left, a bunch of cattle chewing the cud or grazing, ahead of us, a ranch house, and, sometimes, to the rear, a pack of cowardly coyotes, at a safe distance, either following my caravan, or out on a forage hunt. Often, as the trains swept by, the engineers would salute with a deafening blast of whistles, frightening the donkeys and entertaining the passengers. Some of the prairie towns which look large on the map have entirely disappeared. In one case, I found more dead citizens in the cemetery than live ones in the village. Frequently, as a means of diversion, I left the saddle to visit these white-chimney villages of the dead. Such might be considered a grave sort of
- 62. amusement, but really some of the gravestones contained interesting epitaphs. In one instance the following caught my eye: God saw best from us to sever Darling Michael, whom we love; He has gone from us forever, To the happy realms above. Imagine the shock to my sobered senses on reading these lines cut on a white-washed wooden slab, close by: Here lays Ezekiel Dolder, Who died from a jolt in the shoulder; He tried to shoot snipe While lighting his pipe, And now underneath his bones moulder. Just below the heartrending epitaph appeared in bold letters the satisfactory statement—This monument is pade fer. On the lonely plains, miles from habitation, a single grave fenced in with barbed wire in a circular corral, I discovered a mate to the preceding epitaph, which illustrates the utter abandon with which the rugged, dashing bronco buster regards the perils of riding a bucking wild horse. Here is buried my bronco, Ah Sam, Beside me—I don't give a damn! While bucking he killed me; On this spot he spilled me, And now the devil's I am. Sometime before parting with my courier, unknown to him we pitched camp one dark night in a graveyard. Barley was an early riser, and, as we know, as superstitious as he was gullible. He was the first out of the tent at dawn. Suddenly he rushed back, exclaiming: De Resurrection has came, fellows, an' wese de first livin' on earth agin. And with terror in his eyes and voice, dragged Coonskin and me to see a strange sight indeed. There, some forty feet from the tent, stood a towering crucifix with a figure of the Saviour, life size, looking down upon us, while about us were tablets and mounds: the
- 63. scene was so still and solemn no wonder that my awestricken courier thought the world had come to an end. On the 24th of June, after a hot and dusty trail across an arid waste, where only occasional patches of buffalo grass and cacti matted the earth in the place of the long prairie grass and flowers we were tramping in a few days before, my weary troop, jaded and hungry entered the little village of Overton.
- 64. CHAPTER XXXI. Narrow escape in quicksand TOC BY MAC A'RONY. And the ass turned out of the way, and went into the field; and Balaam smote the ass, to turn her into the way.—Book of Numbers. Shortly after reaching Overton, I took Pod with Coonskin and Don to pay our respects to Towserville, a large dog town so closely situated to Overton as to inspire a rivalry far more serious than that existing between Minneapolis and St. Paul. Overtonians complained of repeated raids made by prairie dogs of Towserville on their chickens and gardens. On the other hand, the Towser villians repudiated the calumny, then fled in confusion from the charge of shotguns and rifles. As our party approached with guns trained for a complimentary salute, I saw his honor, the Mayor, seated in his hallway. The roof of his mound towered above the other habitations, and was undoubtedly the City Hall. Copying after New York, each burrow in Towserville had a representative in the City Council. I'm sure we would have been welcomed cordially, had not Don wanted to be first to shake the Mayor's paw; his honor abruptly excused himself to avoid a scene, and his fellow townsdogs likewise, with the result that the above dogtown population rushed in and slammed the doors in our faces. The Professor was embarrassed. He had no visiting cards, so decided to leave at each door a sample box of cathartic pills; and a careful distribution was made. Next morning as we passed Towserville, his dogcellency, the Mayor, his alderdogs and towndogs looked regretful of their slight to us, as each stood at his door or sat with his housekeeper, the owl, on the roof of his dwelling, nodding and waving at us. Others, however, were prostrate, either from remorse or Pod's magnanimity. Sometime about noon, we approached the shallow current of the Platte, where we were unpacked and fed. We donks were almost
- 65. roasted from the sun's scorching rays. Close by was a deep well, but no bucket in which to draw water. So Coonskin hitched a syrup can to the rope and drew water for Pod and himself. Soon a drove of cattle, accompanied by two ranchmen and a boy, came down to the river to drink with us donks, just to show there was no hard feeling. The lad laid down to drink from the stream. Here, boy, come and have a drink of cold water! Pod called. That ain't fit to drink. Fitter'n that well water, answered the lad. Said Pod: I'd like to know the reason. Well, replied the lad, approaching, I dropped a dead jackrabbit in the well a week ago. Somehow the men had drunk so much of that cool well-water they hadn't room for dinner; too cool water I guess aint' good for one when heated. After the dishes were washed, Pod took off everything but his socks and collar-button, and wrote his newspaper letter, while Coonskin went prospecting. Pretty soon the latter returned with a sand turtle and, hitching it up in a rope harness, said he was going to keep it for a pet. He named it Bill. He said it would make a fine center-piece for the table; it would keep the Buffalo gnats and mosquitos and flies off the victuals, and if tied at the tent door no centipede or tarantula would dare enter. Pod thought it a good scheme. So, when we packed up, Bill was put in one of my saddle bags, without my knowing it. All new luggage was generally tied on to Damfino; I supposed the turtle was. After going a couple miles, I felt something mysterious crawling on my back. I looked around, but my master was in the way; so I up and kicked with all my might, determined to scatter that crawling thing to the four winds, but, instead, threw Pod completely over my head. Then I ran pell-mell down the desert trail, kicking and braying, with that terrible something gnawing my hair and bouncing and flopping with every jump I made. I ran fast and thought fast, and that thing stuck fast. Suddenly, I stopped, laid down, and tried to roll on it. This I couldn't do, on account of the saddle horn. But while I was still trying, the rest of the party came up, and solved the mystery by capturing the turtle, Bill; then they chained him on Damfino, and our
- 66. outfit moved on peacefully for several miles, the men talking merrily. Said Pod, Hitting the trail on the plains in summer isn't as comfortable as driving a city ice-wagon. Not much, Coonskin returned; but the donkeys and dog have their woes, too. Verily so, confirmed the Professor. For instance, there's Damfino; she thinks she's awfully persecuted. Being a female, she doesn't have much to say. But how about Mac? Doesn't he do more kicking than all the rest put together? Oh, well, Coonskin answered, you see Mac regards himself a pioneer and all the others mere tenderfeet. I couldn't help grinning at the simple debate. The fact of the case was, our caravan had been growing larger with every day's travel. New articles were continually added. Cheese and I generally carried the men; but to our saddles were hung guns, revolvers, cameras, and the lantern, not to mention a bundle of blankets; all of which, added to the burden of our thoughts, a nagging whip and a pair of spurs, and a million and one buffalo gnats, mastodon mosquitos, and other kindergarten birds of prey, tended to make us lose our mental equilibrium a dozen times a day. In my case, there was a lump of avoirdupois in the saddle ranging between 150 and 160 pounds. Sometimes Pod would get out of his seat and walk a mile or two, to relieve me. With Cheese it was much the same. But that old spinster, Damfino, bore a burden, increasing daily. She was large and strong, and couldn't appreciate fine sentiments, or fine stuffs either, even complaining of sand in the wind, and coughed and snorted continually. Her sawbuck saddle corset was laced tightly around her robust bust, and to this unhealthsome vesture were hung on both sides large canvas panniers, packed with canned goods, medicines, salves, ink, cow-bells, vegetables, ham and bacon, vinegar, old shoes, toilet articles, including currycomb, clothes, soap, flour, salt, baking- powder, cheese, coffee, tea, kerosine oil, matches, cooking tools, ammunition, folding kitchen range, and two dozen et ceteras. On top and lopping over the panniers were roped the tent and tent-poles, folding beds, canteens, musical instruments, axe, and axle-grease, five iron picket-pins, packages of photos (for sale), a tin wash basin,
- 67. two tin pails, extra ropes, a half dozen paper pads, and a dozen more et ceteras. Beneath all that burden, she ambled along without a murmur, swinging her ears to help her outwalk the rest, except Don, who kept up a dog-trot. A ranchman gave Pod some new potatoes one day (half of which I yanked out of the tent door at night and devoured), and in reply to his habitual inquiry, Where'll we stow 'em? Coonskin said, On Damfino, of course. When some canned goods were added to the list of poisons, my master was puzzled. Strap 'em on Damfino, advised Coonskin. Pod bought some canteens. Where'll we put these? he asked. Oh, hang 'em on Damfino somewhere, said the wise Sancho. One day a large package of chromos came, and the Professor was discouraged. How the d—l can we carry these? he asked with bewilderment. Why, ejaculated the valet chuckling, right on Damfino. Just then that silent old maid looked at the men; and I saw blood in her eye. Picture if you can our party trailing along the banks of the Platte that bright June afternoon. A few miles away loomed the cacti- covered sand-dunes, and between them and the river stared the desert of glistening alkali, sprinkled with cacti and sage, where an occasional steer was scratching an existence—and mosquito bites. We came to a muddy irrigation ditch, where the water had leaked out. Across it was an alfalfa field, and beyond that an adobe ranch house. We donks thought the mud in the ditch was stiff; the green field looked tempting. Damfino whispered that she would make a bolt for the field, if we would follow; and we said we would. At once she shied into the ditch, and the next minute was knee-deep in quicksand, and still sinking. Cheese and I stood riveted to the trail, while the men just gaped at Damfino with open mouths. Damfino, thinking she would soon be out of sight, brayed as she never brayed before. When Pod got his senses he yelled, Let's pull her out! What with? Every rope and strap's on Damfino, said the truthful valet, running around like a head with the chicken cut off.
- 68. Coonskin tried to reach a rope and, losing his balance, put a foot in the quicksand. Then, all excited, he attempted to pull his foot out, and got them both in. The Professor tried to reach a bridle-rein to his comrade, and went sprawling across the ditch on his corduroys and whiskers, his arms elbow-deep in the mire. This put Don in a panic. Seeing his master sinking, he grabbed his boots and pulled them off. Then he fastened his teeth in Pod's trousers, and I expected to see them come off too, but s' help me Balaam! the dog only pulled off one trouser leg, when Coonskin managed to free himself by crawling over Pod's corduroy road to dry land, and saved the day! At once, with a bridle-rein, the valet roped the Professor's feet and pulled him out, after which both men fastened the reins to Damfino's pack and tied the other ends to the saddles of Cheese and myself. Then that she- ass, wet and gray as a rat, with her burden, was dragged out of the ditch into the trail. Well, that quicksand pulled all the bad nature out of her, and she went a long time before she was tempted to leave the trail again. The men looked grateful as they wiped the brine from their faces, and Pod remarked, That was a narrow escape for all of us. Our donkey party came within two of going ass-under, sure.
- 69. CHAPTER XXXII. At Buffalo Bill's ranch TOC BY PYE POD.
- 70. It has come about that now, to many a Royal Society, the Creation of a World is little more mysterious than the cooking of a dumpling; concerning which last, indeed, there have been minds to whom the question, How the apples were got in, presented difficulties.—Sartor Resartus. It was noon at Big Springs, the last village on the Union Pacific Railroad in Nebraska, when I sat down to write in my dairy. I had just finished a combination breakfast and dinner, warranted to kill any appetite and keep it dead for twelve hours. Consequently I wrote under great pressure. Since striking Camp Coyote, I had shot prairie dogs, owls, jack- rabbits, and gophers innumerable, but on Wednesday, June 30, I killed my first rattlesnake. It was not the first we had seen, but the first to lie in our path. I wanted to shoot it's head off, but instead of it losing its head, I lost mine, and severed its vertebræ. The snake was three feet five, and possessed eight rattles and a button. Cookskin suggested that the button might come in handy in many ways. You know, Pod, you are always losing buttons. These dreaded reptiles abound on the plains, particularly in dogtowns, where they can dine on superfluous baby-dogs when families become too large. Three sorts of creatures, including the owl —animal, bird, and reptile—bunk together companionably, but have quarrels of their own, doubtless, like mankind in domestic affairs. At that season the South Platte was drained for irrigation in Colorado. I was riding peaceably along, watching its morbid current and the gray hills beyond, when suddenly my valet yelled to me, Look out, Pod, a rattler ahead! Coonskin was riding Cheese, who leaped to one side, but my own steed, blinded by his spectacle-frames, walked on and stepped over the coiled snake, which struck at my leg. Fortunately my canvas legging protected me from the reptile's fangs, which glanced off, letting him fall in the trail. Instantly I turned in my saddle and ended its miserable existence. The report of my revolver attracted some cowboys, who galloped up on their rope horses and accompanied us to their adobe house a few miles beyond. It was five in the afternoon, the day was hot, and
- 71. our journey long and dusty. They were a jolly lot. Thir ranch was a square sod structure, without a floor, and sparingly furnished, but cool and comfortable. We'll have hot biscuit for supper, said one of the cowboys. So you like cooking, I remarked; I pride myself on the dumplings I make, and my flapjacks are marvels of construction. Hang together well, I suppose, observed the cook, smiling and piling buffalo chips in the stove. I haven't tasted dumplings since I visited the World's Fair, said another. Well, declared the first speaker, my tenderfoot friend, your oven will soon be hot, and the flour, soda, shortening, and apples are on the shelf. Anything else you need, ask for it. I was in a bad fix; I remembered the parrot that got into trouble with the bull-terrier by talking too much. It requires a long time to steam dumplings; it will delay supper, I protested. We shan't turn you out, if it takes you all night, but we'll shoot the enamel off your front teeth if you don't make them apple dumplings, and do your best, said a cowboy. All right, boys, I'll try my luck, and you can save time by helping. Sure, all replied. Fetch me the shortening, I called. Right before your eyes, said one. Blamed if I can see it, I explained. The fellow put his hands on a cake of greasy-looking substance. That's soap, I said, remonstrating, with a chuckle. All we use for shortening, apologized the cook; don't see much butter or lard out on this here desert. I fell to with a will. Before long my dough was mixed. As I rolled it out with a tin can, I directed a cowboy to put in the apples and roll up the dough. Soon the dumplings were in the steamer, and the cook began to prepare other eatables for the meal. Then, my duty done, I watched two fellows throw the lariat, and shoot the fly specks off Coonskin's hat in midair.
- 72. At last, five hearty eaters sat down to dinner. The cook's hot biscuits, potatoes, bacon, eggs and coffee were delicious, and I devoured them greedily. But in the middle of our repast I turned my head in time to detect the cook meddling with the dumplings. Shouldn't take off the cover till they're done, I shouted; makes 'em heavy. Didn't take it off—lifted itself off, explained the man, regarding me first, then the steamer. Man alive, the dumplings are as big as cabbages. And 'tain't more'n likely they've got their growth yet, said Coonskin, who examined the wonders. Gracious! I exclaimed. How many apples did you cram into each dumpling? Only fifteen or twenty, the cook returned; awfully small, you know. That explains the size of them, said I. You've got a half dozen whole apples in each dumpling, and a peck or more in the steamer. Don't you know dried fruit swells? But how am I to keep the lid on the steamer, asked the hungry cook, wistfully eying the disappearing meal. Sit on it, you crazy loon, suggested a companion. And the fellow did. Presently there was a deafening report, and the cook was lifted off the steamer, while dumplings flew in every direction, striking the ceiling, and then, from heaviness, dropping on the floor. One broke my plate into a dozen pieces. Another hot and saucy dumpling shot through the bursted side of the steamer, hitting one of the cowboys in the eye. Just my luck, I said; they would have been as light as a feather. Light! exclaimed the injured fellow with a handkerchief against his scalded optic. It was the heaviest thing that ever hit me, let me tell you, and I've been punching cattle seven years. When the excitement was over, and we had found sufficient grub to complete our meal, all assembled in the cool outer air, where Coonskin and I entertained with our musical instruments until bedtime.
- 73. Next morning, on my suggestion, a cowboy threw his lariat round my body good-naturedly and pulled me over, but before I could right myself Don took three bounds and pulled the fellow down by the shoulder, frightening one and all. I shouted so loudly to the dog that I was hoarse for a week. That demonstration of Don's loyalty was a revelation to me. The man was not injured, although his coat was torn. The lack of energy and enterprise of the town of the western plains was both surprising and amusing. I expected a package of photos at Willow Island. When I called for it I was informed that the railroad station had burned a few months before, and that their express stopped at Cozad, which I had passed through. So I wrote to have the package forwarded to a station farther west. Gothenburg, the next town, was in a decline, the reaction of a boom. A traveler approaching it expects to find a business center. Many stores and dwellings were of brick, but whole rows were vacant at the time. The soothing melody of the squalling infant was only a memory to the village druggist; the itinerant butcher and milkman had ceased their daily rounds; and all that was left to distinguish the half-deserted village from the desert was an occasional swallow that went down the parched mouth of a chimney. There is another town characteristic of the plains. I had a letter to post at Paxton, but forgot it; some miles beyond, a ranchman whom we met said I would find a post-office at Korty, five miles further on. After traveling two hours, we could see no vestige of a village anywhere. Don ran ahead to the top of every sand hill and stood on his hind feet to have the first peep at the mysterious town. I came to the conclusion the ranchman had said twenty-five miles instead of five. Finally the trail approached the railroad. I see the town of Korty! my valet exclaimed. Where? I asked. There. Plain as day. Can't you see it? he asked, pointing straight ahead. I must confess I can't, I replied. Let me look over your finger. Then I saw it. It wasn't one hundred feet away. A single white- painted post stood beside the track, and on it was nailed a cross-bar,
- 74. lettered in bold type, Korty; underneath was a letter-box. That was the town. There was no section house, no water tank, no break in the wire fence, and there being, of course, no general delivery window in the post-office, I did not ask for my mail. On the way to North Platte, we passed the site of old Ft. McPherson, where Buffalo Bill, the celebrated scout, once lived and won his fame and title by providing buffalo meat for the Government, and also the site of a notorious Pawnee village, now called Pawnee Springs. We reached North Platte, situated at the confluence of the North and South Platte rivers, which form the great River Platte, Saturday afternoon, and spent Sunday in a manner to meet the approval of the most pious. That first evening I lectured from a large dry-goods box on a prominent corner. Sunday afternoon an old friend and classmate drove me into the country to the famous Scout's Rest Ranch, the estate of Mr. Cody (Buffalo Bill), where I saw a herd of buffalo and a cornfield of 500 acres. There is quite a contrast between your cornfield and mine, I said to the manager. How big a cornfield have you? Just a small one, I replied. One acher on each big toe. I see, only sufficient for your own use, came the response; your 'stock in' trade, as it were. Then the ranchman purchased a photo, and we two grown-up school boys drove back to town, in time to escape a thunder shower. The country between North Platte and Julesburg is a desolate and barren region. Occasionally we could see a ranch house, sometimes cattle grazing on I knew not what. There was plenty of alkali grass in the bottom lands of the Platte, and further back on the mesa, patches of the short and nutritious buffalo grass, half seared by the scorching sun. The railway stations, with one or two exceptions, consisted of water tanks and section houses, where water could be procured. At Ogalala we met a train-load of Christian Endeavorers, and had a chance to quench our thirst.
- 75. CHAPTER XXXIII. Fourth of July in the desert TOC BY MAC A'RONY. What a thrice double ass Was I, to take this drunkard for a god, And worship this dull fool! —Tempest. Where and how to celebrate the Fourth of July greatly concerned Pye Pod. The third was spent in Julesburg, a town in Colorado, two miles west of the boundary line; as Sunday was the Fourth, we naturally expected a lively programme for Saturday. We were disappointed. Everybody had gone off on an excursion, and Julesburg was dead. So my master, realizing the long journey before us, inquired as to the possibility of obtaining an extra donkey, and was told of one, some six miles from town. He rode in a buggy to a ranch right after lunch and brought back the prettiest damsel I ever saw. Her name was Skates; Pod said he so named her because she ran all the way and beat his pride-broken, wind-broken horse into town. I gave Skates a loving smile, but she gave me a look, which said, Keep your distance, young feller. So I did. But I lost my heart to that girl then and there. Pod noticed my leaning toward Skates, and asked me my intentions. I frankly told him. But what nonsense for a youth of four years, he remarked. Mac, be patient; wait until you are of age, at least. Time was precious, and we could not tarry. That afternoon we set out for Sterling, sixty miles into the desert, where, it was said, there would be a big time on the fifth. Monday dawned cloudy and threatening, as is usual with celebration days. The tent door was open, and Skates and I were
- 76. looking in, I waiting for a chance to pull a bag of eatables out of the tent for her. What is your programme for to-day? Pod asked his valet. No answer. The question was repeated; still no response. Then my master turned drowsily on his pillow, and beheld Coonskin with bloodshot eyes and the only whiskey bottle clasped lovingly to his breast. The valet wanted to say something, but his lips refused to speak. It was evident that his celebration had begun the night before. Pod sat up and rubbed his eyes to make sure he was not dreaming, and then asked the fellow why he drank all the emergency whiskey. R-r-r-r-r-r-r-rat-schnake bite-bite-bited me—d—drank whisky t'shave life, stammered the youth. H-h-h-hic-have shome, Prof. Pod looked mad. He up and dressed, and mixed soda and water and lemon juice, and made Coonskin drink it. Soon the tipsy fellow tried to dress, but finally gave it up and went to sleep. Two hours later he awoke quite sober, and came out to where Pod was currying me for the celebration, and showed him his programme. I haven't space to give it in full. One feature was an obstacle race, the prize for the winner being a quart bottle of snake-bit (whiskey). Coonskin said, as his excuse for drinking the whiskey, that he was certain of winning the race, but afraid the bottle might be broken before the event. Pod thought that reasonable enough, and forgave him; but he told me confidentially that he didn't know what he should do if he were bitten by a rattlesnake without whiskey at hand. I suggested, in such event, he should point a revolver at Coonskin's garret, where his brains ought to have been, and make him suck out the poison. The obstacle race began at eleven in the morning. The start was made from the tent door; the course and conditions were as follows: Run to the fifth fence-post down the trail, alongside the railroad track; crawl through the barbed-wire fence four times between different posts on the way back to the tent, without tearing clothes; creep through the legs of the little portable table (purchased in Julesburg) without rolling off an egg resting on it; run a hundred yards and unpicket one of the donkeys and ride it round the tent three times with a spoon in hand, holding an egg; ride the donk back
- 77. to his picket-pin and crawl between its hind legs without disturbing the animal's equilibrium; stand in the tent door and shoot some hair off one of the donkey's tails without touching the tail proper; then lead that donkey to the tent and hitch him to the turtle, Bill. Cheating, if detected, forfeited the prize. Well, while there were two starters, there was only one finisher. It seems that Coonskin shot a piece off Cheese's tail (improper, the donk said), and, in consequence, man and donk disappeared over the horizon, without leaving their future address or the date for their return. Coonskin rode Cheese into camp after dark. Then he rubbed axle- grease on Cheese's sensitive part, and prepared the delayed dinner. Next came fire-works—Roman candles, firecrackers, and pin-wheels —after which both men retired, fancying they had the jolliest Fourth ever witnessed by man or donkey in the history of the Colorado desert.
- 78. CHAPTER XXXIV. Bitten by a rattler TOC BY PYE POD. Sancho Panza hastened to his master's help as fast as his ass could go, and when he came up he found the knight unable to stir, such a shock had Rosinante given him in the fall.—Don Quixote. The casualty, which terminated our celebration on the fifth, seemed to portend bad luck. The metaphorical lightning first struck me. We struck camp, that hot July day, before the sun was an hour high, and a mile beyond trailed through a dog-town reservation. I had long been desirous of securing a prairie dog to have mounted; as a rule one can pick off these shy creatures only at long rifle range. This morning, stealing up behind a cornfield, I wounded a dog, then dropping my gun, ran to catch him before he could escape into his hole. Crawling through a barbed-wire fence without afterward appearing in dishabille is considered by a tenderfoot the feat of feats. Before I reached the hole half undressed the dog had tumbled into it. He must have made a mistake, however, for out the fellow came, and made for another hole. I grabbed him, but instantly dropped him, for he tried to bite me. Then, like a shot, he dived into the second hole, and I thrust my arm in to pull him out. But my hand came out quite as fast as it went in. It was bitten; and at the mouth of the hole I now detected for the first time the tail of a rattlesnake. That was an awful moment, What should I do? My whiskey was gone; I had no antidote for the poison. I rushed to where Coonskin was waiting with my outfit. Make for the house! he exclaimed. A ranch house stood some two miles away, but not a soul was in sight. Still, that seemed to be my only salvation; I realized a painful death was the only alternative. With a hundred other thoughts rushing into my head, I ran toward the distant house. Coonskin began picketing the donkeys, and promised to follow.
- 79. While racing madly through the cacti and sage, I thought of my past, from three months upward. Just when I had reached an episode, which almost ended my reckless career at the age of ten, I heard the sound of galloping hoofs, and, a moment later, a young woman reined her steed at my side, dismounted and gave me her horse. Into the saddle, quick, man! she cried. Mother has turpentine and whiskey. The horse will take the fence and ditch. Pull leather, stick to the saddle, never mind the stirrups! and to the horse—Git home, Topsy!—Run for your life, old girl! Like a flash, the big mare sped forward with the velocity of the wind. To pull leather, in the parlance of the cowboy, means to grip the saddle with the hands. For a cow-puncher to pull leather is deemed disgraceful; for Pod, it was excusable. Although the mare fairly flew, she did not travel half fast enough to suit me. With reins round the saddle-horn, I gripped the saddle with my left hand and sucked the bite on my right, but suddenly the mare took a hop-skip-and-jump over the fence and ditch; fell to her knees, and threw me over her head. When I sat up, I saw a woman in the door of the house, yet a half mile away, no doubt, wondering how a maniac happened to be on her daughter's steed. The next moment, Coonskin arived all out of breath, and assisted me to the house. Before we could fully explain the situation, the good woman disappeared, soon to return with a bottle of turpentine, which she turned nozzle down over the snake bite, while my valet poured whiskey down my throat. They say it takes a long time and much whiskey to affect one bitten by a rattler, but this case seemed to be an exception; in a few moments, my head was going round, and I prostrate on a couch. My kind nurse looked curiously at the turpentine, and finally said it was queer it didn't turn green, as it should in the case of a rattle-snake bite. A half hour passed and still there was no change. Then when I repeated my story of how the thing happened, she grinned, and said she guessed it was the prairie dog and not the snake that bit me, after
- 80. Welcome to our website – the perfect destination for book lovers and knowledge seekers. We believe that every book holds a new world, offering opportunities for learning, discovery, and personal growth. That’s why we are dedicated to bringing you a diverse collection of books, ranging from classic literature and specialized publications to self-development guides and children's books. More than just a book-buying platform, we strive to be a bridge connecting you with timeless cultural and intellectual values. With an elegant, user-friendly interface and a smart search system, you can quickly find the books that best suit your interests. Additionally, our special promotions and home delivery services help you save time and fully enjoy the joy of reading. Join us on a journey of knowledge exploration, passion nurturing, and personal growth every day! ebookbell.com