Quantitative proteomics: challenges and opportunities in basic and
applied research
Schubert, O. T., Röst, H. L., Collins, B. C., Rosenberger, G., & Aebersold, R. (2017). Quantitative proteomics:
challenges and opportunities in basic and applied research.
Nature Protocols
,
12
(7), 1289-1294.
https://doi.org/10.1038/nprot.2017.040
Published in:
Nature Protocols
Document Version:
Peer reviewed version
Queen's University Belfast - Research Portal:
Link to publication record in Queen's University Belfast Research Portal
Publisher rights
Copyright 2017 Nature Research. This work is made available online in accordance with the publisher’s policies. Please refer to any
applicable terms of use of the publisher.
General rights
Copyright for the publications made accessible via the Queen's University Belfast Research Portal is retained by the author(s) and / or other
copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated
with these rights.
Take down policy
The Research Portal is Queen's institutional repository that provides access to Queen's research output. Every effort has been made to
ensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in the
Research Portal that you believe breaches copyright or violates any law, please contact openaccess@qub.ac.uk.
Download date:09. Aug. 2022
1
Quantitative proteomics: Challenges
and opportunities in basic and applied research
Olga T. Schubert
1
, Hannes L. Röst
2,5
, Ben C. Collins
3,5
, George Rosenberger
3,5,
Ruedi Aebersold
3,4,
*
1 Department of Human Genetics, University of California Los Angeles, Los Angeles, CA 90095, USA
2 Department of Genetics, Stanford University, Stanford, CA 94305, USA
3 Department of Biology, Institute of Molecular Systems Biology, ETH Zurich, 8093 Zurich, Switzerland
4 Faculty of Science, University of Zurich, 8057 Zurich, Switzerland
5 Authors contributed equally
*Corresponding author (aebersold@imsb.biol.ethz.ch)
Summary
In this perspective, we discuss developments in mass spectrometry-based proteomic
technology in the last decade from the viewpoint of our laboratory. We also reflect on existing
challenges and limitations, and explore the current and future role of quantitative proteomics in
molecular systems biology, clinical research and personalized medicine.
2
Introduction
Proteins constitute a large part of the molecular machinery of the cell and are the major class of
biomolecules targeted by drugs. Organized in functional modules and networks, they carry out
cellular functions and determine phenotypes by means of coordinated activities of a multitude of
molecular species
1
. Traditional biochemical methods for studying proteins have been highly
biased towards a relatively small subset of proteins for which high quality, mainly antibody-
based assays have been available
2
. Over the past two decades, mass spectrometry (MS)-
based methods have emerged as the method of choice for the confident and near exhaustive
identification and quantification of the proteins contained in a biological sample and have
significantly contributed to unraveling cellular signaling networks, to elucidating the dynamics of
protein-protein interactions in different cellular states, and to improved diagnosis and molecular
understanding of disease mechanisms. Overall, MS-based proteomics can reveal the
quantitative state of a proteome and thereby provides insights into the biochemical state of the
respective cell or tissue. In the following paragraphs, we will discuss important concepts and
developments in proteomic technology and explore the current and future role of quantitative
proteomics in molecular systems biology as well as clinical research and personalized medicine.
Evolution of MS-based quantitative proteomics
MS-based proteomics can be broadly grouped into top-down proteomics where intact proteins
are measured and bottom-up proteomics where peptides are measured as surrogates for the
respective protein; in this commentary, we will focus on bottom-up proteomics. The typical
bottom-up proteomics workflow starts with trypsin digestion of a protein sample into short
peptides which are then separated by liquid chromatography either directly or after biochemical
fractionation (Figure 1A)
3
. As peptides elute from the chromatography column, they are
subjected to electrospray ionization
4,5
and are directly sprayed into the mass spectrometer,
where two levels of MS measurement take place in tandem
3
. At the first level, a mass analyzer
measures the mass-to-charge ratio (m/z) of peptide molecular ions (MS1). At the second level,
m/z values of fragment ions resulting from the fragmentation of specific peptide ions are
detected (MS2). The specific fragment ion pattern of each peptide ion together with its m/z value
enable confident identification of peptides present in the sample. Identified peptide sequences
can then be mapped to proteins and the signal intensities of either peptides or fragment ions
can be used to estimate relative abundance changes across samples.
3
Figure 1. Standard MS-based proteomics workflow and acquisition schemes. (A) Proteins can be
extracted from various biological samples, such as bacterial or mammalian cell culture, tissues or bodily
fluids. They are then enzymatically digested into peptides, which are then subjected to reverse-phase
liquid chromatography, ionized with electrospray ionization and sprayed into the mass spectrometer. The
time dimension in B and C is this chromatographic time. (B) Different acquisition schemes for tandem MS
sample the proteome in distinct ways. While the most widely used untargeted (also referred to as shotgun
or data-dependent acquisition, DDA) is relatively simple and applicable to any sample without requiring
prior knowledge, resulting data can suffer from missing data points due to the stochastic sampling
process. In contrast, targeted acquisition acquires peptide and fragment ion data in a highly consistent
manner allowing accurate and sensitive quantification, but is limited to a relatively small, pre-defined set
of peptides. Data-independent acquisition (DIA) acquires data of all detectable fragment ions in a sample
in a systematic and consistent manner, but due to the relatively large peptide ion isolation windows (m/z
dimension) the resulting data is more complex than for the other two acquisition schemes. (C) DIA data
can be analyzed in different ways, either directly analyzing the multiplexed MS2 spectra or first extracting
a subset of informative fragment ion signals (requires prior knowledge) and using these to derive
quantitative data for specific peptides
36,40
.
Discovery-driven
Data-dependent
Hypothesis-driven
Targeted
Data-independent
Time
m/z
B
Time
m/z
C
Time
Intensity
MS2
MS2
MS2
MS2
MS2
Time
Quantitative
MS2 signal
A
m/z
Liquid
chromatography
Peptides Proteins
Tandem mass
spectrometry
Time
Eluent
+
Biological or
clinical samples
MS2
MS1
4
To account for technical variability at various stages of sample handling and during the actual
measurement, in the mid-90s, we and others started to develop strategies based on isotopic
labeling
6,7
, including chemical isotopic labeling
8
, metabolic isotopic labeling
9
, and isobaric
tagging
10,11
. Another important application of isotopic labeling in MS is the use of labeled spike-
in peptides or proteins of known concentration that enable the determination of absolute
concentrations of proteins in a sample, for example, in terms of number of molecules per cell or
nanograms per milliliter of blood
12
. While label-based approaches are still the gold-standard for
quantification by MS-based proteomic methods
13
, the past years have seen label-free
approaches becoming more popular thanks to simpler experimental design and sample
preparation
6,14
. Among the developments enabling this transition are the advance of
commercially available high resolution and fast scanning instruments, such as the introduction
of the Orbitrap (2005)
15
and continuous improvement of time-of-flight mass spectrometers
16
,
combined with improvements in software for aligning multiple MS runs
17,18
. Another more recent
trend, starting in 2006
19
, is label-free absolute quantification, where the absolute concentrations
of all proteins measured in a sample are estimated based on summarized ion counts, which can
then be converted into a meaningful unit by comparison to the total amount of protein that was
injected into the mass spectrometer or by correlation to a set of spiked-in reference peptides of
known concentration
20-23
.
Regardless of whether label-based or label-free strategies are used, bottom-up proteomic
methods have traditionally been divided into discovery proteomics and targeted proteomics
(Figure 1B). Discovery proteomics (also known as shotgun proteomics and exemplified by data-
dependent acquisition, DDA) has its strength in identifying thousands of proteins per run.
However, in complex samples, we have often been faced by limitations regarding repeatability
of peptide identification and consistency of quantification
24,25
. Recent developments in
chromatographic performance and MS hardware alleviate some of these concerns and allow
high-quality quantitative measurements of near-complete proteomes, even in highly complex
samples such as human cell lines and tissues
26-29
.
About a decade ago, in order to overcome the limited scalability and reproducibility of discovery
proteomics in studies aiming to quantify proteins in cohorts consisting of large numbers of
samples, we and others started exploring the capabilities of targeted proteomics (exemplified by
selected/multiple reaction monitoring, S/MRM
30,31
, and more recently parallel reaction
monitoring, PRM
32,33
). Targeting methods provide consistent and accurate quantification, even