Proteomics Research in Cardiovascular Medicine and Biomarker Discovery

JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY

VOL. 68, NO. 25, 2016

PUBLISHED BY ELSEVIER ON BEHALF OF THE AMERICAN

ISSN 0735-1097/$36.00

COLLEGE OF CARDIOLOGY FOUNDATION

http://dx.doi.org/10.1016/j.jacc.2016.10.031

FOCUS SEMINAR: GENETICS STATE-OF-THE-ART REVIEW

Proteomics Research in Cardiovascular Medicine and Biomarker Discovery Maggie P.Y. Lam, PHD,a Peipei Ping, PHD,a Elizabeth Murphy, PHDb

ABSTRACT Proteomics is a systems physiology discipline to address the large-scale characterization of protein species within a biological system, be it a cell, a tissue, a body biofluid, an organism, or a cohort population. Building on advances from chemical analytical platforms (e.g., mass spectrometry and other technologies), proteomics approaches have contributed powerful applications in cardiovascular biomedicine, most notably in: 1) the discovery of circulating protein biomarkers of heart diseases from plasma samples; and 2) the identification of disease mechanisms and potential therapeutic targets in cardiovascular tissues, in both preclinical models and translational studies. Contemporary proteomics investigations offer powerful means to simultaneously examine tens of thousands of proteins in various samples, and understand their molecular phenotypes in health and disease. This concise review introduces study design considerations, example applications and use cases, as well as interpretation and analysis of proteomics data in cardiovascular biomedicine. (J Am Coll Cardiol 2016;68:2819–30) Published by Elsevier on behalf of the American College of Cardiology Foundation.

O

ver the past 20 years, there has been accel-

molecular changes at the protein layer, in particular,

erating

can provide independent insights into disease mech-

growth

in

the

application

of

proteomics in cardiovascular biomedicine,

anisms in the following 3 areas.

building on pioneering studies with translational

First, it has gradually emerged from large-scale

significance (1,2) and mechanistic insights (3–8).

studies that transcript and protein changes in a

The majority of human genes functions to create

system may, at times, correspond poorly—with

proteins, which are the molecular workhorses that

transcript levels explaining as little as 10% to 30% of

carry out virtually all metabolic, signaling, and phys-

variations in protein abundance. Although the exact

iological functions in life. Many human diseases

contribution of transcripts to protein-level abun-

may be characterized by the elicited changes in

dance is debated, it is evident that in the heart and

proteome configurations. In complex late onset dis-

other organs, a large number of post-transcriptional

eases

environmental

modulators can alter disease-driver protein expres-

components exist, such that individual genetic vari-

sion and function during pathogenesis, without

in

particular,

considerable

ants are often poorly predictive of disease states.

changes

Telltale molecular signatures of acquired cardiovas-

modulators include components of the nonsense-

cular diseases may instead manifest through inter-

mediated decay pathway, long noncoding RNAs,

mediate

and microRNAs (11–14).

phenotypes,

including

transcript

and

protein abundance. Beyond what may be learned from

gene

and

transcript

information

alone,

in

transcript

abundance

(9,10).

These

Second, plasma proteins provide an accessible readout of the status of potentially all tissues, and

Listen to this manuscript’s audio summary by JACC Editor-in-Chief Dr. Valentin Fuster. From the aNIH BD2K Center of Excellence and Department of Physiology, Medicine and Bioinformatics, University of California, Los Angeles, California; and the bSystems Biology Center, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland. This work was supported by NIH K99-HL127302 to Dr. Lam; NIH R37-HL063901 and NIH U54-GM114833 to Dr. Ping; and NIH ZIA-HL006059 and NIH ZIA-HL002066 to Dr. Murphy. Manuscript received August 3, 2016; revised manuscript received October 20, 2016, accepted October 21, 2016.

2820

Lam et al.

JACC VOL. 68, NO. 25, 2016 DECEMBER 27, 2016:2819–30

Proteomics in Cardiovascular Research

ABBREVIATIONS

are the sources of many current biomarkers

in plasma may be viewed as classical resident

AND ACRONYMS

in use (Table 1). Clinically useful biomarkers

plasma proteins, which tend to occupy the higher

can

end of the abundance spectrum. These resident

AP-MS = affinity purificationmass spectrometry

FDR = false discovery rate

include

circulating

proteins

that

have been secreted or leaked directly into

proteins

the plasma from resident cells (e.g., myo-

such as albumin, immunoglobulins, transferrins,

include

extremely

abundant

species,

cytes) following diseases or injuries (15), and

and fibrinogens, which are found in the milligram/

are thus spatially uncoupled from the tran-

milliliter concentration range and account for >90%

script change at the tissue of origin. The

of plasma protein content. Other plasma proteins

tissues of origin of circulating proteins may

come from secretion or leakage from various organs

modification

be unknown, inaccessible, or invasive to

before being diluted into the 5 liters of peripheral

RNA-seq = ribonucleic acid-

procure,

hunt

blood, and are found in the microgram/milliliter

sequencing

for these potential biomarkers via trans-

to nanogram/milliliter range. Very low-abundance

TMT = tandem mass tag

cript

proteins, including interleukin-6 and -12, tumor

MS = mass spectrometry PDH = pyruvate dehydrogenase

PTM = post-translational

making

it

impractical

measurements,

instead

to

of

protein

measurements.

necrosis factor-alpha, and other cytokines, circulate

Third, disease processes may be mediated not by

in the picogram/milliliter range, and are typically

the altered abundance of gene products, but rather

masked by higher-abundance proteins in untargeted

by other functional parameters of the proteome,

proteomics analyses (Figure 1). Adding to this

including protein post-translational modifications

complexity, plasma proteins may be proteolytically

(PTMs), protein-protein interactions, and protein

processed, creating numerous degradation product

degradation, which take place after the protein

species, and repetitive sequences in proteins can

molecules are synthesized and cannot be predicted

give a spurious appearance of higher abundance.

from genetic information a priori.

These challenges are compounded by the absence of

Proteomics

to

a protein equivalent of the polymerase chain reac-

measure protein function on a large scale, and hence

tion, such that amplification of minute samples is

interrogate

not possible.

abuts

technologies the

allow

molecular

physiological

layer

that

In the following 3 segments of this review, we will

the complexity of the human proteome also presents

discuss 3 aspects of proteomics applications in car-

a

analytical

challenge

(16).

closely However,

daunting

phenotypes

researchers

large-scale

diovascular biomedical research: 1) study design for

characterization. For example, circulating proteins

for

biomarker research and discovery, with discussions

in human plasma, found at a concentration of

on approaches to circumvent analytical challenges; 2)

w70 mg/ml, are estimated to comprise at least

characterization of multidimensional protein param-

10,000 distinct protein species (17) over a concen-

eters in disease mechanism research; and 3) inter-

tration range of >10 orders of magnitude (10-billion-

pretation and validation of proteomics data or

fold differences) (18). Approximately 3,000 proteins

methodologies.

T A B L E 1 Selected Protein Biomarkers for Cardiovascular Diseases

Protein Biomarker

Abbreviation

Disease Relevance

Required Assay Sensitivity (Estimate)

Discovery Period*

Ref. #

Apolipoprotein A-I

APOA

Cardiovascular event risk

w1 mg/ml

Mid-1980s

(96)

Apolipoprotein B

APOB

Cardiovascular event risk

w1 mg/ml

Mid-1980s

(97)

B-type natriuretic peptide

BNP

Heart failure, acute coronary syndrome

w100 pg/ml

Early 2000s

(98)

C-reactive protein

CRP

Cardiovascular event risk

w10 mg/ml

Late 1990s

(99) (100)

Creatine kinase-myocardial band

CK-MB

Acute myocardial infarct, myocardial necrosis

w1 ng/ml

1960s to 1970s

Cystatin-C

CST3

Cardiovascular event risk

w1 mg/ml

2000s

(101)

Fibrinogen

FBN

Cardiovascular event risk

w1 mg/ml

1980s

(99)

Lipoprotein-associated phospholipase A2

Lp-PLA2

Coronary heart disease risk

w100 ng/ml

2000s

(102) (103)

Myeloperoxidase

MPO

Ischemic heart disease; acute coronary syndrome

w10 ng/ml

1980s

Myoglobin

MYO

Myocardial infarction, necrosis

w10 ng/ml

Late 1970s

(104)

Serum amyloid A

SAA

Coronary artery disease

w10 mg/ml

1990s

(105)

Troponin I

cTnI

Myocardial injury, myocardial infarction

w10 pg/ml

1970s to 1990s

(106)

Troponin T

cTnT

Myocardial injury, myocardial infarction

w10 pg/ml

1970s to 1990s

(107)

*Periods of initial discovery or assay development (108,109).

Lam et al.

JACC VOL. 68, NO. 25, 2016 DECEMBER 27, 2016:2819–30

Proteomics in Cardiovascular Research

F I G U R E 1 Sensitivity and Coverage of Various Proteomics Study Designs

Proteomics Approach

Discovery (Shotgun)

Proteome coverage

High - up to 10,000 proteins may be analyzed

Number of samples

Typical sensitivity limit

Low

µg/ml

Medium

ng/ml

Reference concentration in plasma

10-3 (mg/ml)

g/ml Albumin

Example technology Platforms

Notes

High-resolution MS (time-of-flight or Orbitrap instruments)

+ Most unbiased view of proteins

Antibody-based protein arrays

+ Middle ground between scope and sensitivity

LDL 10-6 (µg/ml)

Targeted Discovery

Middle depends on size of target panels or array used

CRP

ng/ml

Aptamer-based protein arrays

Medium pg/ml

CKMB

Data-independent acquisition mass spectrometry

10-9 (ng/ml)

Targeted

Low - typically limited to 10s of proteins

ng/ml

TNF

Multiple-reaction monitoring mass spectrometry (triple-quadruple instruments)

IL-6

ELISA panels

NT-pro BNP

High pg/ml 10-12 (pg/ml)

– Quantification may not be as accurate as targeted approaches

– May require special reagents; target may not be available in commercial panels

+ Excellent specificity and quantification precision – May require method development in specialty laboratories

Properties of 3 common proteomics study designs are listed. Discovery (shotgun), targeted, and targeted discovery proteomics approaches take different strategies to address proteome coverage, number of samples, and sensitivity limits. Discovery proteomics experiments have high coverage (up to 10,000 proteins in a single sample), but have limitations in throughput; fewer samples or subjects can be analyzed. Targeted discovery approaches focus on analyzing a panel of high-potential targets in sufficient numbers of samples. Targeted proteomics is able to achieve the highest sensitivity, which allows the detection of low-abundance plasma markers, such as tumor necrosis factor (TNF) and interleukin (IL)-6, but at the expense of scope and throughput. Graph (center) shows the typical sensitivity range of discovery, targeted discovery, and targeted approaches, juxtaposed with the concentrations of selected plasma proteins and disease markers. Example technological platforms and additional notes are shown on the right. CKMB ¼ creatine kinase-myocardial band; CRP ¼ C-reactive protein; ELISA ¼ enzyme-linked immunosorbent assay; LDL ¼ low-density lipoprotein; MS ¼ mass spectrometry; NT-proBNP ¼ N-terminal pro–B-type natriuretic peptide.

STUDY DESIGNS FOR BIOMARKER

experimental designs. The advantages and disad-

RESEARCH AND DISCOVERY

vantages of technological platforms should therefore be evaluated comprehensively on the basis of the

With precision medicine initiatives demanding a

scope and depth of analysis required, and determined

broader pool of measurable proxies of individual

prior to the commencement of chemical analysis.

characteristics, the call for investigations to discover

Three common designs, all of which can provide both

new candidate biomarkers will likely continue to in-

qualitative and quantitative data on proteins, are: 1)

crease. Successful screens of protein biomarker can-

global discovery; 2) targeted; and 3) targeted discov-

didates require obtaining data: 1) with good coverage

ery. Each has distinct strengths and may be better

of all the proteins present in a sample; and 2) in a

suited to particular experimental goals. Each may be

sufficient number of subjects to ensure adequate

performed in suitable technological platforms that

power for discovery. Unlike microarrays or ribonu-

offer different throughputs, analytical depth, and

cleic acid-sequencing (RNA-seq), a single proteomics

sensitivity. We provide here a brief summary on the

experiment often cannot simultaneously obtain high

strengths

coverage and throughput, necessitating tradeoffs in

approach.

and The

considerations Central

for

Illustration

each

type

presents

of an

2821

2822

Lam et al.

JACC VOL. 68, NO. 25, 2016 DECEMBER 27, 2016:2819–30

Proteomics in Cardiovascular Research

C E NT R AL IL L U STR AT IO N Overview of Proteomics Analyses for Cardiovascular Diseases

The Role of Proteomics in the Search For Biomarkers and Mechanisms Of Acquired Cardiovascular Diseases

Sample collection

Data processing and bioinformatics

Data acquisition

Extract proteins from human cohorts or clinical models

Following the acquisitions of large-scale datasets, data are processed:

Three major study approaches:

Cardiac tissue

Data interpretation to extract insights from datasets:

Identify proteins from mass spectra using search engine

Global discovery: Analyze large number of proteins but with lower throughput and sensitivity

Blood/plasma samples

Data interpretation and analysis

Identify interaction networks and/or altered cellular pathways

Deduce relative and/or absolute quantities across samples

Targeted: Analyze fewer proteins but with higher throughput and sensitivity

Connect identified molecular features to phenotypes Recommend identified protein signatures for validation by complementary translational studies

Targeted discovery: Analyze a mid-sized panel of pre-selected proteins Lam, M.P.Y. et al. J Am Coll Cardiol. 2016;68(25):2819–30.

This figure lists major components in common proteomics workflows, as well as their associated experimental considerations in biomarker discovery and disease studies. From top to bottom, protein samples are collected from the plasma of human cohorts or cardiac tissues in animal models according to study goals (either biomarker discovery or mechanistic studies). Three major study approaches (discovery, targeted, and targeted discovery) take different strategies between proteome coverage and analytical throughput, and use different technological platforms, including mass spectrometry and protein arrays. Following the acquisition of largescale datasets, data are processed to identify the protein species present and to deduce their relative quantities across samples. Subsequently, a number of statistical and bioinformatics workflows are used in data interpretation to extract insights from datasets. Network analysis casts proteins in the context of interaction networks or altered cellular pathways. Statistical learning and modeling methods connect the identified molecular features to orthogonal phenotypes, identify signatures, and offer information on subject classification or predictive analysis. The identified protein signatures will require validation, which can be achieved by complementary translational studies, including in vitro biochemical analysis and large cohorts. PTM ¼ post-translational modifications.

overview of data acquisition and analysis; see Table 2

resulting mass spectra are then searched against a

for

protein sequence database (19) or spectral library

a

glossary

of

selected

technological

terms

commonly in use.

(20,21) for protein identification. Various downstream data analyses, such as spectral counting and network

GLOBAL DISCOVERY PROTEOMICS. Global discov-

analysis may be performed to extract other qualita-

ery approaches offer an unbiased view of the protein

tive and quantitative information from the data. A

species that may be found in a sample. Shotgun pro-

major strength of shotgun proteomics is that it casts

teomics is 1 of the most widely used and standardized

the widest net toward potential biomarkers (covering

global discovery strategies. In shotgun analysis, pro-

thousands of proteins) (22) and does not require the

teins are first enzymatically digested (e.g., using

development of individual assays for each target

trypsin) into peptides in vitro, and then subjected to

(23,24). Two weaknesses are its lower overall sensi-

tandem

tivity,

mass

spectrometry

(MS)

analysis.

The

which

may

bias

it

against

biologically

Lam et al.

JACC VOL. 68, NO. 25, 2016 DECEMBER 27, 2016:2819–30

2823

Proteomics in Cardiovascular Research

T A B L E 2 Glossary of Selected Technologies Used in Cardiovascular Proteomics

Terms

Description

Amino acid sequence

A segment of the primary structure of the protein where individual amino acids are linked by peptide bonds.

Aptamer array

A non-MS proteomics technology where single-stranded DNA oligonucleotides with affinity for specific protein targets are used as a means to detect and quantify up to w1,000 different protein targets.

Liquid chromatography A common front-end separation technology in proteomics experiments to fractionate complex sample mixture. It is used to resolve and enrich peptides by their chemical properties such as size, hydrophobicity, or charge, and the like. MS

An analytical technology to measure the mass of an analyte (e.g., peptide, protein, amino acid, and so on) in the format of mass-to-charge ratio.

Mass spectrum

The data output of MS. A mass spectrum is a plot displaying the relative intensity or abundance of the analyte ions (y-axis) and their mass-to-charge ratio (x-axis).

Mass/charge (m/z)

The principal measurement of a mass spectrometer: the mass of the ionized analytes divided by the charge they carry. Also commonly referred to as mass-to-charge ratio.

Multiple-reaction monitoring

A targeted proteomics approach in which a particular type of mass spectrometer is programmed to only measure a few peptides and their fragment ions of a target protein of interest, while filtering out all other peptides and proteins in the sample, thus enhancing specificity and sensitivity of the detection/quantification of the targets. Also known as SRM.

Search engine

A software tool used in sequence database search to identify proteins from MS data (e.g., SEQUEST) (110).

Sequence database search

A method to identify proteins from MS data by matching the collected experimental mass spectra to theoretical mass spectra generated from a protein sequence database, such as UniProt (19) or RefSeq (111).

Spectral library search

A method to identify peptides by matching MS data to previously known experimental spectra in a library, such as COPaKB (21) or PeptideAtlas (20).

Shotgun proteomics

A method to identify and quantify proteins on a large scale, in which proteins are first digested into short segments (peptides) that are amenable to sequence determination via tandem MS analysis. The identity of each present protein is then inferred from the sequenced peptides. Also known as bottom-up proteomics.

Tandem mass spectrometry

An analytical method involving the fragmentation of peptides (or other analytes) via multiple stages of MS analyses to obtain sequence/structural information of the analytes. Also known as MS/MS or MS2.

TMT/iTRAQ

Commercially available chemical labels containing stable isotopes, which can be used to label protein samples for multiplexed analysis and comparison.

iTRAQ ¼ isobaric tags for relative and absolute quantitation; MS ¼ mass spectrometry; SRM ¼ selected reaction monitoring; TMT ¼ tandem mass tags.

significant proteins at low abundance (see Improving

for each protein being targeted, which requires sub-

Detection

and

stantial technical expertise from specialized labora-

Throughput section), and less accurate quantification

tories. However, once developed, other investigators

compared with targeted proteomics.

and facilities in clinical and research settings may

TARGETED PROTEOMICS. Targeted proteomics of-

readily adapt the assays. Due to the labor and costs

fers the highest sensitivity and specificity among MS-

required to validate assays, the number of proteins

based proteomics approaches. MS-based targeted

that can be realistically analyzed in a targeted

of

Low-Abundance

Proteins

approaches require the programming of a specific

biomarker search may be limited (w100 or fewer).

instrument (e.g., a triple quadrupole mass spectrom-

Accordingly, targeted proteomics is most suitable for

eter), and development of methods (e.g., multiple

characterizing subsets of likely candidate biomarkers

reaction monitoring) to monitor only a limited num-

with prior disease implications. A limited number of

ber of analytes (via predetermined peptide mass or

non–MS-based antibody assays are also available for

charge and fragment information) to achieve its high

targeted quantification of low-abundance proteins in

sensitivity. It was demonstrated that targeted MS, in

the plasma (e.g., enzyme-linked immunosorbent

conjunction with stable isotope dilution, was able to

assays).

verify a panel of cardiovascular biomarker candidates

TARGETED

at a limit of quantification ranging from 2 to 15 ng/ml

ground between a targeted and an untargeted

(25).

Targeted

MS

is

also

applicable

to

DISCOVERY

APPROACHES. A

middle

post-

approach is sometimes sought, which may be called

translationally modified proteins with PTM site

targeted discovery. The goal is to attain the best of

specificity (26), with the absolute quantity of 14

both worlds: focusing on a finite panel of targets,

phosphorylation sites on cardiac troponin I as a

while maintaining sufficient throughput to analyze a

notable example (27). Ongoing effort is being devoted

large number of samples. Oftentimes, this comes at

to the development of targeted protein assays aimed

the expense of scope (number of proteins analyzed)

at cardiac biomarker analysis (28–30). A disadvantage

and flexibility, but, on the upside, panels of “likely

of targeted proteomics is that the practice of only

bets” can be preselected by manufacturers or by the

using 1 or a few peptides in the experiments may not

investigators to increase the probability of discovery

represent the behavior of the full-length protein (e.g.,

(31). A number of MS-based (e.g., sequential window

due to proteolysis). Specific assays must be developed

acquisition of all theoretical mass spectra) (32) and

2824

Lam et al.

JACC VOL. 68, NO. 25, 2016 DECEMBER 27, 2016:2819–30

Proteomics in Cardiovascular Research

non–MS-based affinity proteomics (e.g., antibody-

result quantified >3,000 human plasma proteins in 16

and aptamer-based arrays) (31,33,34), are being used

patients and discovered 333 regulated proteins (38),

to discover candidate biomarkers. Antibody-based

including known markers of spontaneous myocardial

arrays have been used to discover markers for

infarction, such as creatine kinase-myocardial band

ischemic stroke, using a proteomics chip of 92 high-

and troponin. Using 6-plex TMT labels, the proteomic

potential cardiovascular proteins (35), including

analysis of the plasma samples of w1,000 human

B-type natriuretic peptide, growth differentiation

subjects has been recently demonstrated (39). The

factor-15, matrix metalloproteinase-12, and others.

popularization of new 10-plex TMT labels will further

Aptamer arrays that use nucleotides in place of anti-

increase throughput.

bodies for target binding have been used to discover protein myocardial infarct markers and to evaluate a panel of blood proteins as risk factors for prediction of coronary heart diseases (33,34,36,37).

MULTIDIMENSIONAL PROTEIN PARAMETERS IN DISEASE MECHANISM RESEARCH Contemporary proteomics applications in basic dis-

LOW-ABUNDANCE

ease mechanism research often incorporate addi-

PROTEINS AND THROUGHPUT. To improve detec-

tional experimental techniques that target increasing

IMPROVING

DETECTION

OF

for

numbers of “dimensions” of protein properties.

MS-based analysis, commercial immunodepletion

Whereas previous proteomics studies focused largely

columns, including the Human IgY14 and SuperMix

on

columns (Sigma-Aldrich, St. Louis, Missouri), are

regulation in quantity as primary drivers of changes

used to remove top- to medium-abundance proteins.

in function), there has been a growing appreciation of

Simplifying the sample mixture helps unmask low-

other, equally significant proteome parameters that

abundance proteins for detection, but many pro-

modulate cardiovascular biology. These parameters

teins or fragments of interest may be inadvertently

include post-translational regulation, protein-protein

codepleted by the process. Typically, a sample con-

interactions, and protein turnover, each of which may

taining 100 ml of plasma may only yield w140 mg of

be altered in disease and present potential thera-

tion

of

low-abundance

proteins

in

plasma

gene

product

abundance

(up-

and

down-

proteins after 98% of high-abundance species are

peutic targets (9). These parameters are aiding in ef-

depleted, an amount sufficient for a limited number

forts to fully comprehend the disease proteome as a

of replicate analyses. To target challenging proteins

dynamic,

in plasma and other samples, additional preprocess-

velopments in these areas may expand the parameter

ing steps, including size-exclusion chromatography,

space from which future biomarkers may be found.

multidimensional

PROTEIN

liquid

chromatography,

and

multidimensional

entity.

POST-TRANSLATIONAL

Ongoing

de-

REGULATION.

immunocapture may be used, but necessitate greater

Post-translational regulation through covalent modi-

starting amounts of samples (e.g., 0.01 to 1.00 ml of

fication of proteins is a major mechanism of cardiac

plasma) and additional processing time.

signaling. Over 200 protein PTMs that can alter the

To achieve multiplexing, minimize experimental

chemistry and interaction surface of a protein, and

variations, and increase throughput, multiple protein

thereby its structure, activity, binding partners, or

samples can be barcoded with stable isotopes and

subcellular localization, are known (40). Although

analyzed in a single experiment at the expense of

over 500,000 PTM sites are known to exist (41),

costs. Commercially available isobaric tags for rela-

available site-specific antibodies target relatively few

tive and absolute quantitation and tandem mass tag

of them. MS-based proteomics techniques do not

(TMT) labels, small amine-reactive molecules cova-

require separate reagents (i.e., antibodies) for each

lently linked to differential mass reporters and mass

site, and hence can be effectively used to examine

balancers, can be used to tag peptides such that up to

new sites or sites for which antibodies are not

10 sets of barcoded samples may be coanalyzed in 1

available.

MS experiment. During tandem MS, the mass re-

For a number of reasons, PTM studies often require

porters separate from the balancers and act as surro-

additional sample preparation and enrichment con-

gates for comparison of protein quantity of each

siderations (e.g., with titanium dioxide) (42) before

sample. This multiplexing strategy has been profit-

data acquisition. First, modified proteins frequently

ably used to identify potential infarct markers in a

exist in lower abundance than the total protein pop-

study that analyzed the plasma of 16 patients at

ulation (although in some instances, such as in the

multiple time points following septal ablation treat-

case of secreted glycoproteins, almost every protein

ment. With the aid of 4-plex isobaric tags for relative

molecule is modified). For example, if only 10% of a

and absolute quantitation labels, the proteomics

protein species is modified, then the modified

Lam et al.

JACC VOL. 68, NO. 25, 2016 DECEMBER 27, 2016:2819–30

Proteomics in Cardiovascular Research

proteins are effectively 10 times more difficult to

histones. A family of enzymes known as histone

detect than the total (modified plus unmodified)

acetyltransferases catalyzes the addition of the acetyl

population. Second, whereas identifying a protein

group to proteins, whereas histone deacetylases

only requires detection of any of its multiple pro-

revert the modification. Although protein acetylation

teolytic peptides along its primary structure, PTM

was classically described on histones, other proteins

analysis requires the precise detection of the peptide

are also modified. Protein acetylation participates in

containing the PTM site, which sometimes can lie in a

nonchromatin processes, including metabolism and

region of the protein that is not amenable to MS

protein localization, as well as protein stability

detection (e.g., sequences without trypsin digestion

(45,46), and can be mapped with proteomics tech-

sites). Third, many PTMs are chemically labile and

niques following enrichment by anti–acetyl-lysine

biologically transient, making their capture and

antibodies (62,63). Over one-third of mitochondrial

enrichment before signal measurement essential.

proteins are subjected to acetylation, including a

PTMs under intense investigation for their roles in cardiovascular (43,44),

diseases

acetylation

include

(45,46),

phosphorylation

SUMOylation

large number of metabolic enzymes. Acetylation in mitochondria may function in cardioprotective con-

(47),

texts (46), whereas hyperacetylation of succinate

glycosylation (48,49), S-nitrosylation (50,51), ubiq-

dehydrogenase in mitochondria may drive energy

uitination (52), and others (53,54). The remainder of

imbalance in the development of heart failure (63).

this section highlights proteomics approaches to un-

Reactive

oxygen

and

nitrogen

species

can

derstand protein phosphorylation, acetylation, and

confer covalent modifications (e.g., S-nitrosylation,

oxidative modifications.

S-sulfhydration, and S-glutathionylation) on cys-

Multiple major kinase pathways operate in the

teine residues in redox-sensitive proteins. These

heart, including protein kinase A, protein kinase B,

modifications can change the activity and binding

mitogen-activated protein kinase or extracellular

surfaces of target proteins, and thus directly partic-

signal-regulated kinase, protein kinase C, glycogen

ipate in cardiovascular biology in a number of dis-

synthase kinase 3b, calcium/calmodulin-dependent

ease contexts. For example, it is known that nitric

protein kinase II, and others; hence, it is no sur-

oxide serves as a critical signaling messenger in

prise that phosphoproteomics is commonly applied

ischemic preconditioning. Nitric oxide signals partly

in the study of heart diseases. A well-known

by causatively increasing protein S-nitrosylation,

example of a phosphorylated protein essential for

which

cardiac energetics is pyruvate dehydrogenase (PDH)

dioprotection by reducing other, pernicious forms of

(55), which converts pyruvate to acetyl coenzyme A

cysteine oxidation on the target proteins during

for respiration, and whose activity is dictated by its

reperfusion

phosphorylation by PDH kinases. PDH phosphoryla-

modification en masse, biotin switch proteomics are

tion thus serves as a major regulatory checkpoint of

used that first chemically block free thiols, then uses

cardiac energy metabolism. Classically, phosphory-

a specific reducing agent to reduce different reduc-

lation sites were mapped by phosphorus 32 labeling

tive states of oxidative PTMs (e.g., ascorbic acids

and Edman degradation. Up to w50,000 phosphor-

may be used to reduce only S-nitrosylated thiols, but

has

been

injury

proposed

(64,65).

to

To

mediate

detect

car-

oxidative

ylation sites may now be identified in a single study

not other modifications), and finally labels the freed

using

thiols with biotin or isotope-labeled tags (cys-TMT or

shotgun

proteomics

(56)

combined

with

enrichment strategies. Mapping in the heart has

iodo-TMT) (51,66).

unearthed a large number of phosphorylation sites, including those common to multiple tissues and

PROTEIN TEMPORAL DYNAMICS. Protein homeosta-

many that are unique to the cardiovascular system

sis and proteolysis are broadly implicated in heart

(57). Targeted mapping efforts may also be directed

diseases and injuries. Classical studies of protein

toward selected cellular compartments (e.g., mito-

turnover required pulse-chase of radioisotopes and

chondria) (58), pathways (e.g., those targeted by

isolation of target proteins. With recent advances,

phosphodiesterase 9A/phosphodiesterase 5, calcium/

proteomics studies can measure turnover dynamics

calmodulin-dependent protein kinase II) (59,60),

of thousands of proteins in multiple tissues and or-

specific proteins (e.g., the myosin-binding protein C

ganisms (67,68). Combining stable isotope labeling,

phosphoproteome) (61), or multiple specific sites on

mathematical modeling, and shotgun proteomics,

the same proteins (26). Acetylation of histones plays a key role in gene

recent investigations into the turnover of w5,000 cardiac

proteins

in

mouse

cardiac

hypertrophy

expression by modulating the accessibility of tran-

(69,70) revealed novel correlations between prote-

scriptional elements to DNA wound around the

ome profiles and phenotypic functional limitations.

2825

2826

Lam et al.

JACC VOL. 68, NO. 25, 2016 DECEMBER 27, 2016:2819–30

Proteomics in Cardiovascular Research

For example, although the hypertrophic heart is

INTERPRETATION AND VALIDATION OF

known to switch in fuel preference from fatty acid to

PROTEOMICS DATA

glucose, the expression of glycolytic proteins does not change accordingly (71). Instead, the turnover of

Several considerations are provided here regarding

most glycolytic proteins was found to accelerate

how to interpret proteomic results, as well as how to

drastically

use new workflows and technologies.

in

hypertrophy

without

concomitant

expression changes. Thus static expression profiling

SCOPE AND SENSITIVITY. With incomplete coverage,

experiments alone would not have discovered this

the old adage that the absence of evidence is not

aspect of cardiac pathophysiology had there been no a

evidence of absence is especially true for proteomics

priori knowledge of increased glycolysis, whereas a

experiments. The number of proteins and peptides in

large-scale experiment of protein turnover might

a shotgun experiment is an important indicator of

have led to the generation of relevant hypotheses. A

the depth of coverage in data acquisition. Hence, if

possible connection between turnover and function is

a known biomarker is not reported or reproduced in

that old proteins may have reduced activity compared

a discovery-based experiment in the cohort, a perti-

with young, functional proteins. For example, the

nent question is whether the technological platform

glycolysis enzyme triose-phosphate isomerase can

used has sufficient sensitivity and precision to sup-

become

port reproducible measurements at endogenous

chemically

deamidated

upon

catalytic

cycling. In another example, the stem cell factor or KIT ligand has been found to lose biological potency

concentration. The predictive value of a putative biomarker de-

by 50-fold after chemical deamidation (72).

pends not only on the magnitude of its difference in

PROTEIN-PROTEIN INTERACTIONS. Physical liaison

normal versus diseased or at-risk individuals in the

between proteins is a primary means whereby infor-

population, but also on the effect size, defined as

mation is transduced along signaling pathways. The

the number of SDs between 2 measured means.

target specificity and activity of protein kinase A are

Hence, for the purpose of protein biomarker dis-

markedly regulated by which of the w40 A-kinase

covery, the coefficient of variation of measurement

anchoring proteins it primarily interacts with (73), to

is essential for determining whether there is suffi-

mediate kinase signaling, and which can change

cient power to distinguish significant differences

during heart disease (74). Protein-protein interaction

between healthy and diseased samples. Beyond the

analysis is pursued to characterize signaling net-

precision characteristics of the technological plat-

works, discover the components of protein com-

forms, sample processing, as well as biological var-

plexes, and identify the contexts in which a protein

iations in human cohorts (both intrasubject and

may function in health and disease (75). A powerful

intersubject), may also impart significant variance in

approach to analyze protein-protein interaction is

protein measurements. Thus, individual putative

affinity purification-MS (AP-MS) (76,77), which pre-

protein markers often require lengthy verification

cipitates bait proteins, along with their interacting

and validation pipelines that involve large cohorts

partners, using antibodies or another purification

(83). Proteomics assays have reported typical co-

method. Using unbiased AP-MS, a recent study

efficients of variation in the range of <5% for

discovered the association of the cardiac transcrip-

intralaboratory measurements (34,84) to <20% for

tion factor TBX5 with the nucleosome remodeling

interlaboratory measurements (85). Other measure-

deacetylase repressor complex, revealing a novel

ments of analytical performance that can have a

mechanism for TBX5 in cardiac development and

drastic impact on the likelihood of payoff in down-

congenital heart diseases (78). Although it is the

stream verification and validation processes include

current mainstay of interactome research, AP-MS

the limit of detection, limit of quantification, and

specificity can vary. Even with careful washing,

linearity of measurement, which are covered at

nonspecific binding occurs and contaminants can

length elsewhere (83).

remain. Thus, identified interactors should be veri-

FALSE DISCOVERY RATE. As in other large-scale ap-

fied by complementary methods. New methods are

proaches, both the number of identified proteins and

currently being developed to ameliorate this chal-

the significance of quantitative comparisons in a

lenge, including in silico filtering of interactome data

quantitative proteomics experiment may become in-

(79,80) and cross-linking MS (81). The latter uses

flated if controls for false discovery rates (FDRs) are

reactive chemicals to physically link together pro-

not adequate. In protein identification, the FDR of a

teins in apposition for confident MS identification

database search provides an estimate for the propor-

(reviewed in Leitner et al. [82]).

tion of incorrect protein identification in the result list.

Lam et al.

JACC VOL. 68, NO. 25, 2016 DECEMBER 27, 2016:2819–30

Proteomics in Cardiovascular Research

An FDR of 1% is commonly deemed acceptable,

procedures, or in peripheral blood versus proximal

meaning that in a profiling experiment claiming to

tissue fluids, such as coronary sinus effluents, to filter

identify 5,000 proteins, up to 50 may be mis-

out spurious associations. An in-depth discourse on

identifications. FDR is sometimes calculated with the

biomarker verification and validation can be found

aid of decoy databases, which contain scrambled or

elsewhere (83).

reversed protein sequences that are compared to experimental data, along with real protein sequences. The number of identified decoy sequences provides an estimate of the number of false discoveries among regular database sequence matches. For quantitative proteomics comparisons between samples, as many hypotheses as the number of proteins compared are being tested per sample pair. Multiple-testing corrections (e.g., Bonferroni corrections) are used to control for type I errors (incorrectly rejecting the null hypothesis) that arise from making multiple comparisons (1 per protein per sample pair). However, it must be noted that even infinitesimal p values are no indication that results are biomedically important (86), and do not obviate the need to validate with follow-up studies, including in vitro experimentation or in cohorts.

OUTLOOK With continued technological progress, new interactions of proteomics with genetics and genomics are now becoming possible. Proteomics studies applied to large cohorts can be integrated with genetic polymorphisms to identify protein quantitative trait loci (pQTL) that control protein abundance, and perhaps thereby control phenotypic traits. Proteogenomics studies, combining RNA-seq and proteomics data from common datasets, have been undertaken in cancer research to identify the protein coding consequences of cancer variants and mutations (91), whereas a recent multiomics study in 80 mouse genetic backgrounds demonstrated that multiple types of omics data can complement each other to characterize

COMPARISON

WITH

EXISTING

mitochondrial

functional

regulation

(92).

TECHNIQUES. In

Further integration of proteomics and genomics

addition to technical considerations, criteria to

methods will allow examination of nonstandard gene

nominate a differentially expressed protein as a

products, including short open reading frames, un-

candidate biomarker can vary. Currently, no gold

annotated transcripts, and alternative splicing iso-

standard exists for what constitutes a good candidate

form proteins. By greatly expanding the depth of the

(87). Decision justifications in published reports range

proteome that may be experimentally accessed, the

from formal statistical inferences of individual pro-

development of isoform proteomics, in particular,

teins (88), panels of proteins after feature selection

will have important implications for cardiovascular

(89), to additional criteria, such as cutoffs in the raw

research. These isoforms may present with differen-

magnitude of fold change (e.g., >5-fold changes) in at

tial expression levels or patterns, localizations, in-

least n number of patients (90), to more custom

teractions, and PTMs in different cell types and

modeling that adjust for covariates in risk assess-

during disease progression, including in ischemic

ments (31,35). In determining whether the presented

cardiomyopathy (93) and hypertrophy (94). Although

candidates are worthy of prioritized verification

RNA-seq has discovered many alternative isoforms at

studies, it stands to reason that the results of dis-

the transcript level, the majority of isoform tran-

covery experiments should be carefully scrutinized

scripts have not been validated at the protein level.

for both nominal significance (e.g., p values), and for

Many alternatively spliced transcripts may lead to

the magnitude and variance of biological changes

frame-shifts and be targeted by nonsense-mediated

(effect size and confidence interval).

decay. The functional significance of many tran-

Ultimately, a candidate biomarker or protein with

script isoforms may soon be addressed by proteomics

potential biomedical significance needs to be validated

using custom sequence databases translated from

by orthogonal approaches in independent cohorts or

RNA-seq transcripts.

models. In silico methods may be used to triage a list of

Broader applications will also be facilitated by

differentially expressed proteins and prioritize candi-

continued

dates for validation, for example, by determining their

coverage. For example, newly developed neutron

improvements

in

throughput

and

tissue expression specificity and half-lives. Promising

encoding (NeuCode) techniques will allow over 30

biomarkers may be further credentialed with targeted

samples to be combined for MS analysis (95), leading

discovery approaches that quantify the status of the

to a 5-fold increase in throughput for investigating

candidates in larger numbers of subjects and condi-

large cohorts. With new affinity proteomics plat-

tions. For instance, protein profiles may be compared

forms, such as protein and aptamer arrays, we have

in

already seen further development of preselected

paired

controls

before

and

after

unrelated

2827

2828

Lam et al.

JACC VOL. 68, NO. 25, 2016 DECEMBER 27, 2016:2819–30

Proteomics in Cardiovascular Research

cardiovascular-specific protein panels (33), which

innovative breakthroughs in cardiovascular medicine

could allow the disease status of the likeliest candi-

for years to come.

dates to be screened across large sample pools. In summary, with core proteomics technologies having

REPRINT REQUESTS AND CORRESPONDENCE: Dr.

now matured to provide high protein and sample

Maggie P.Y. Lam, David Geffen School of Medicine at

coverage, new applications are increasingly available

UCLA, Department of Physiology & Bioinformatics,

to interrogate the dynamic parameters that connect

675 Charles E. Young Drive, MRL Building, Suite 1-

gene expression to physiological functions. We envi-

625, Los Angeles, California 90095. E-mail: magelpy@

sion that proteomics-based discovery-driven plat-

ucla.edu. OR Dr. Elizabeth Murphy, National Heart,

forms

molecular

Lung, and Blood Institute, NIH, Building 10, Room

phenotyping of individuals in the era of precision

8N202, 10 Center Drive, Bethesda, Maryland 20892.

medicine, and will provide ample opportunities for

E-mail: [email protected].

will

continue

to

support

REFERENCES 1. McDonough JL, Labugger R, Pickett W, et al. Cardiac troponin I is modified in the myocardium of bypass patients. Circulation 2001;103:58–64. 2. Labugger R, Organ L, Collier C, et al. Extensive troponin I and T modification detected in serum from patients with acute myocardial infarction. Circulation 2000;102:1221–6. 3. Weekes J, Wheeler CH, Yan JX, et al. Bovine dilated cardiomyopathy: proteomic analysis of an animal model of human dilated cardiomyopathy. Electrophoresis 1999;20:898–906. 4. Ping P, Zhang J, Pierce WM, et al. Functional proteomic analysis of protein kinase C ε signaling complexes in the normal heart and during cardioprotection. Circ Res 2001;88:59–62. 5. Edmondson RD, Vondriska TM, Biederman KJ, et al. Protein kinase C ε signaling complexes include metabolismand transcription/ translation-related proteins: complimentary separation techniques with LC/MS/MS. Mol Cell Proteomics 2002;1:421–33. 6. Lindsey ML, Goshorn DK, Comte-Walters S, et al. A multidimensional proteomic approach to identify hypertrophy-associated proteins. Proteomics 2006;6:2225–35. 7. Sun J, Picht E, Ginsburg KS, et al. Hypercontractile female hearts exhibit increased S-nitrosylation of the L-type Ca2þ channel a1 subunit and reduced ischemia/reperfusion injury. Circ Res 2006;98:403–11. 8. Ge Y, Rybakova IN, Xu Q, et al. Top-down highresolution mass spectrometry of cardiac myosin binding protein C revealed that truncation alters protein phosphorylation state. Proc Natl Acad Sci U S A 2009;106:12658–63. 9. Larance M, Lamond AI. Multidimensional proteomics for cell biology. Nat Rev Mol Cell Biol 2015;16:269–80. 10. Liu Y, Beyer A, Aebersold R. On the dependency of cellular protein levels on mRNA abundance. Cell 2016;165:535–50. 11. Vogel C, Abreu Rde S, Ko D, et al. Sequence signatures and mRNA concentration can explain two-thirds of protein abundance variation in a human cell line. Mol Syst Biol 2010;6:400. 12. Tian Q, Stepaniants SB, Mao M, et al. Integrated genomic and proteomic analyses of gene

expression in mammalian cells. Mol Cell Proteomics 2004;3:960–9. 13. Schwanhäusser B, Busse D, Li N, et al. Global quantification of mammalian gene expression control. Nature 2011;473:337–42. 14. Lundberg E, Fagerberg L, Klevebring D, et al. Defining the transcriptome and proteome in three functionally different human cell lines. Mol Syst Biol 2010;6:450. 15. Marshall KD, Edwards MA, Krenz M, et al. Proteomic mapping of proteins released during necrosis and apoptosis from cultured neonatal cardiac myocytes. Am J Physiol Cell Physiol 2014; 306:C639–47. 16. Van Eyk JE. Proteomics: unraveling the complexity of heart disease and striving to change cardiology. Curr Opin Mol Ther 2001;3:546–53.

patient plasma by targeted mass spectrometry and stable isotope dilution. Mol Cell Proteomics 2009;8:2339–49. 26. Lam MP, Lau E, Scruggs SB, et al. Site-specific quantitative analysis of cardiac mitochondrial protein phosphorylation. J Proteomics 2013;81: 15–23. 27. Zhang P, Kirk JA, Ji W, et al. Multiple reaction monitoring to identify site-specific troponin I phosphorylated residues in the failing human heart. Circulation 2012;126:1828–37. 28. Percy AJ, Yang J, Hardie DB, et al. Precise quantitation of 136 urinary proteins by LC/MRMMS using stable isotope labeled peptides as internal standards for biomarker discovery and/or verification studies. Methods 2015;81:24–33.

Plasma Proteome Database as a resource for proteomics research: 2014 update. Nucleic Acids Res 2014;42:D959–65.

29. Fu Q, Chen Z, Zhang S, et al. Multiple and selective reaction monitoring using triple quadrupole mass spectrometer: preclinical large cohort analysis. Methods Mol Biol 2016;1410: 249–64.

18. Anderson NL, Anderson NG. The human plasma proteome: history, character, and diagnostic prospects. Mol Cell Proteomics 2002;1: 845–67.

30. Percy AJ, Byrns S, Chambers AG, Borchers CH. Targeted quantitation of CVD-linked plasma proteins for biomarker verification and validation. Expert Rev Proteomics 2013;10:567–78.

19. Pundir S, Magrane M, Martin MJ, et al. Searching and navigating UniProt databases. Curr Protoc Bioinformatics 2015;50:1.27.1–1.27.10.

31. Gerstein HC, Paré G, McQueen MJ, et al.,

17. Nanjappa V, Thomas JK, Marimuthu A, et al.

20. Kusebauch U, Deutsch EW, Campbell DS, et al. Using PeptideAtlas, SRMAtlas, and PASSEL: comprehensive resources for discovery and targeted proteomics. Curr Protoc Bioinformatics 2014;46:13.25.1–28. 21. Zong N, Li H, Li H, et al. Integration of cardiac proteome biology and medicine by a specialized knowledgebase. Circ Res 2013;113:1043–53. 22. Riley NM, Hebert AS, Coon JJ. Proteomics moves into the fast lane. Cell Syst 2016;2:142–3. 23. Bensimon A, Heck AJ, Aebersold R. Mass spectrometry-based proteomics and network biology. Annu Rev Biochem 2012;81:379–405. 24. Zhang Y, Fonslow BR, Shan B, et al. Protein analysis by shotgun/bottom-up proteomics. Chem Rev 2013;113:2343–94. 25. Keshishian H, Addona T, Burgess M, et al. Quantification of cardiovascular biomarkers in

Outcome Reduction With Initial Glargine Intervention Trial Investigators. Identifying novel biomarkers for cardiovascular events or death in people with dysglycemia. Circulation 2015;132: 2297–304. 32. Sajic T, Liu Y, Aebersold R. Using dataindependent, high-resolution mass spectrometry in protein biomarker research: perspectives and clinical applications. Proteomics Clin Appl 2015;9: 307–21. 33. Ngo D, Sinha S, Shen D, et al. Aptamer-based proteomic profiling reveals novel candidate biomarkers and pathways in cardiovascular disease. Circulation 2016;134:270–85. 34. Ganz P, Heidecker B, Hveem K, et al. Development and validation of a protein-based risk score for cardiovascular outcomes among patients with stable coronary heart disease. JAMA 2016; 315:2532–41. 35. Lind L, Siegbahn A, Lindahl B, et al. Discovery of new risk markers for ischemic stroke using a

Lam et al.

JACC VOL. 68, NO. 25, 2016 DECEMBER 27, 2016:2819–30

Proteomics in Cardiovascular Research

novel targeted proteomics chip. Stroke 2015;46: 3340–7.

with cysteine-reactive tandem mass tags: short communication. Circ Res 2012;111:1308–12.

CysTMT6 switch assay. Methods Mol Biol 2013; 1005:169–79.

36. Sabatine MS. Using aptamer-based technology to probe the plasma proteome for cardiovascular disease prediction. JAMA 2016;315:2525–6.

52. Zong N, Ping P, Lau E, et al. Lysine ubiquitination and acetylation of human cardiac 20S proteasomes. Proteomics Clin Appl 2014;8:590–4.

37. Gramolini A, Lau E, Liu PP. Identifying lowabundance biomarkers: aptamer-based prote-

53. Fert-Bober J, Giles JT, Holewinski RJ, et al. Citrullination of myofilament proteins in heart

67. Kim TY, Wang D, Kim AK, et al. Metabolic labeling reveals proteome dynamics of mouse mitochondria. Mol Cell Proteomics 2012;11: 1586–94.

omics potentially enables more sensitive detection in cardiovascular diseases. Circulation 2016;134: 286–9.

failure. Cardiovasc Res 2015;108:232–42.

38. Keshishian H, Burgess MW, Gillette MA, et al. Multiplexed, quantitative workflow for sensitive biomarker discovery in plasma yields novel candidates for early myocardial injury. Mol Cell Proteomics 2015;14:2375–93. 39. Cominetti O, Nuñez Galindo A, Corthésy J, et al. Proteomic biomarker discovery in 1000 human plasma samples with mass spectrometry. J Proteome Res 2016;15:389–99. 40. Olsen JV, Mann M. Status of large-scale analysis of post-translational modifications by mass spectrometry. Mol Cell Proteomics 2013;12: 3444–52. 41. Huang KY, Su MG, Kao HJ, et al. dbPTM 2016: 10-year anniversary of a resource for posttranslational modification of proteins. Nucleic Acids Res 2016;44:D435–46. 42. Thingholm TE, Jørgensen TJ, Jensen ON, et al. Highly selective enrichment of phosphorylated peptides using titanium dioxide. Nat Protoc 2006; 1:1929–35. 43. Lundby A, Andersen MN, Steffensen AB, et al. In vivo phosphoproteomics analysis reveals the cardiac targets of b-adrenergic receptor signaling. Sci Signal 2013;6:rs11. 44. Schechter MA, Hsieh MK, Njoroge LW, et al. Phosphoproteomic profiling of human myocardial tissues distinguishes ischemic from non-ischemic end stage heart failure. PLoS One 2014;9:e104157. 45. Karamanlidis G, Lee CF, Garcia-Menendez L, et al. Mitochondrial complex I deficiency increases protein acetylation and accelerates heart failure. Cell Metab 2013;18:239–50. 46. Nguyen TT, Wong R, Menazza S, et al. Cyclophilin D modulates mitochondrial acetylome. Circ Res 2013;113:1308–19. 47. Yang W, Paschen W. SUMO proteomics to decipher the SUMO-modified proteome regulated by various diseases. Proteomics 2015;15:1181–91. 48. Barallobre-Barreiro J, Gupta SK, Zoccarato A, et al. Glycoproteomics reveals decorin peptides with anti-myostatin activity in human atrial fibrillation. Circulation 2016;134:817–32. 49. Parker BL, Palmisano G, Edwards AV, et al. Quantitative N-linked glycoproteomics of myocardial ischemia and reperfusion injury reveals

54. Lindsey ML, Iyer RP, Zamilpa R, et al. A novel collagen matricryptin reduces left ventricular dilation post-myocardial infarction by promoting scar formation and angiogenesis. J Am Coll Cardiol 2015;66:1364–74. 55. Kerbey AL, Randle PJ, Cooper RH, et al. Regulation of pyruvate dehydrogenase in rat heart. Mechanism of regulation of proportions of dephosphorylated and phosphorylated enzyme by oxidation of fatty acids and ketone bodies and of effects of diabetes: role of coenzyme A, acetylcoenzyme A and reduced and oxidized nicotinamide-adenine dinucleotide. Biochem J 1976;154:327–48. 56. Sharma K, D’Souza RC, Tyanova S, et al. Ultradeep human phosphoproteome reveals a distinct regulatory nature of Tyr and Ser/Thrbased signaling. Cell Rep 2014;8:1583–94. 57. Lundby A, Secher A, Lage K, et al. Quantitative maps of protein phosphorylation sites across 14 different rat organs and tissues. Nat Commun 2012;3:876. 58. Deng N, Zhang J, Zong C, et al. Phosphoproteome analysis reveals regulatory sites in major pathways of cardiac mitochondria. Mol Cell Proteomics 2011;10. M110.000117. 59. Lee DI, Zhu G, Sasaki T, et al. Phosphodiesterase 9A controls nitric-oxide-independent cGMP and hypertrophic heart disease. Nature 2015;519: 472–6. 60. Scholten A, Preisinger C, Corradini E, et al. Phosphoproteomics study based on in vivo inhibition reveals sites of calmodulin-dependent protein kinase II regulation in the heart. J Am Heart Assoc 2013;2:e000318. 61. Kooij V, Holewinski RJ, Murphy AM, et al. Characterization of the cardiac myosin binding protein-C phosphoproteome in healthy and failing human hearts. J Mol Cell Cardiol 2013;60:116–20. 62. Svinkina T, Gu H, Silva JC, et al. Deep, quantitative coverage of the lysine acetylome using novel anti-acetyl-lysine antibodies and an optimized proteomic workflow. Mol Cell Proteomics 2015;14:2429–40. 63. Horton JL, Martin OJ, Lai L, et al. Mitochondrial protein hyperacetylation in the failing heart. JCI Insight 2016;2:e84897. 64. Sun J, Aponte AM, Menazza S, et al. Additive cardioprotection by pharmacological post-

early remodeling in the extracellular environment. Mol Cell Proteomics 2011;10. M110.006833.

conditioning with hydrogen sulfide and nitric oxide donors in mouse heart: S-sulfhydration vs. Snitrosylation. Cardiovasc Res 2016;110:96–106.

50. Wang SB, Foster DB, Rucker J, et al. Redox regulation of mitochondrial ATP synthase: implications for cardiac resynchronization therapy. Circ Res 2011;109:750–7.

65. Sun J, Nguyen T, Aponte AM, et al. Ischaemic preconditioning preferentially increases protein Snitrosylation in subsarcolemmal mitochondria. Cardiovasc Res 2015;106:227–36.

51. Kohr MJ, Aponte A, Sun J, et al. Measurement of S-nitrosylation occupancy in the myocardium

66. Murray CI, Chung HS, Uhrigshardt H, et al. Quantification of mitochondrial S-nitrosylation by

68. Price JC, Holmes WE, Li KW, et al. Measurement of human plasma proteome dynamics with 2H2O and liquid chromatography tandem mass spectrometry. Anal Biochem 2012;420:73–83. 69. Lam MP, Wang D, Lau E, et al. Protein kinetic signatures of the remodeling heart following isoproterenol stimulation. J Clin Invest 2014;124: 1734–44. 70. Lau E, Cao Q, Ng DC, et al. A large dataset of protein dynamics in the mammalian heart proteome. Sci Data 2016;3:160015. 71. Kolwicz SC Jr., Tian R. Glucose metabolism and cardiac hypertrophy. Cardiovasc Res 2011;90: 194–201. 72. Hsu YR, Chang WC, Mendiaz EA, et al. Selective deamidation of recombinant human stem cell factor during in vitro aging: isolation and characterization of the aspartyl and isoaspartyl homodimers and heterodimers. Biochemistry 1998;37: 2251–62. 73. Burgers PP, van der Heyden MA, Kok B, et al. A systematic evaluation of protein kinase A-A-kinase anchoring protein interaction motifs. Biochemistry 2015;54:11–21. 74. Aye TT, Soni S, van Veen TA, et al. Reorganized PKA-AKAP associations in the failing human heart. J Mol Cell Cardiol 2012;52:511–8. 75. Pankow S, Bamberger C, Calzolari D, et al.

DF508 CFTR interactome remodelling promotes rescue of cystic fibrosis. Nature 2015;528:510–6. 76. Morris JH, Knudsen GM, Verschueren E, et al. Affinity purification-mass spectrometry and network analysis to understand protein-protein interactions. Nat Protoc 2014;9:2539–54. 77. Huttlin EL, Ting L, Bruckner RJ, et al. The BioPlex network: a systematic exploration of the human interactome. Cell 2015;162:425–40. 78. Waldron L, Steimle JD, Greco TM, et al. The cardiac TBX5 interactome reveals a chromatin remodeling network essential for cardiac septation. Dev Cell 2016;36:262–75. 79. Rizzetto S, Priami C, Csikász-Nagy A. Qualitative and quantitative protein complex prediction through proteome-wide simulations. PLoS Comput Biol 2015;11:e1004424. 80. Goldfarb D, Hast BE, Wang W, et al. Spotlite: web application and augmented algorithms for predicting co-complexed proteins from affinity purification–mass spectrometry data. J Proteome Res 2014;13:5944–55. 81. Gingras AC, Gstaiger M, Raught B, et al. Analysis of protein complexes using mass spectrometry. Nat Rev Mol Cell Biol 2007;8:645–54. 82. Leitner A, Faini M, Stengel F, et al. Crosslinking and mass spectrometry: an integrated technology to understand the structure and function of molecular machines. Trends Biochem Sci 2016;41:20–32.

2829

2830

Lam et al.

JACC VOL. 68, NO. 25, 2016 DECEMBER 27, 2016:2819–30

Proteomics in Cardiovascular Research

83. Rifai N, Gillette MA, Carr SA. Protein biomarker discovery and validation: the long and uncertain path to clinical utility. Nat Biotechnol 2006;24: 971–83. 84. Percy AJ, Chambers AG, Yang J, et al. Multiplexed MRM-based quantitation of candidate cancer biomarker proteins in undepleted and nonenriched human plasma. Proteomics 2013;13: 2202–15. 85. Abbatiello SE, Schilling B, Mani DR, et al. Large-scale interlaboratory study to develop, analytically validate and apply highly multiplexed, quantitative peptide assays to measure cancerrelevant proteins in plasma. Mol Cell Proteomics 2015;14:2357–74. 86. Goodman SN. Aligning statistical and scientific reasoning. Science 2016;352:1180–1. 87. de Lemos JA, Rohatgi A, Ayers CR. Applying a big data approach to biomarker discovery: running before we walk? Circulation 2015;132:2289–92. 88. Keshishian H, Burgess MW, Gillette MA, et al. Multiplexed, quantitative workflow for sensitive biomarker discovery in plasma yields novel candidates for early myocardial injury. Mol Cell Proteomics 2015;14:1–45.

93. Kong SW, Hu YW, Ho JW, et al. Heart failureassociated changes in RNA splicing of sarcomere genes. Circ Cardiovasc Genet 2010;3:138–46. 94. Song HK, Hong SE, Kim T, et al. Deep RNA sequencing reveals novel cardiac transcriptomic signatures for physiological and pathological hypertrophy. PLoS One 2012;7:e35552. 95. Hebert AS, Merrill AE, Bailey DJ, et al. Neutron-encoded mass signatures for multiplexed proteome quantification. Nat Methods 2013;10: 332–4. 96. Stubbs P, Seed M, Lane D, et al. Lipoprotein(a) as a risk predictor for cardiac mortality in patients with acute coronary syndromes. Eur Heart J 1998; 19:1355–64. 97. McQueen MJ, Hawken S, Wang X, et al., INTERHEART study investigators. Lipids, lipoproteins, and apolipoproteins as risk markers of myocardial infarction in 52 countries (the INTERHEART study): a case-control study. Lancet 2008; 372:224–33. 98. de Lemos JA, McGuire DK, Drazner MH. Btype natriuretic peptide in cardiovascular disease. Lancet 2003;362:316–22.

89. Yin X, Subramanian S, Hwang SJ, et al. Protein biomarkers of new-onset cardiovascular disease:

99. Emerging Risk Factors Collaboration. C-reactive protein, fibrinogen, and cardiovascular disease prediction. N Engl J Med 2012;367:1310–20.

prospective study from the systems approach to biomarker research in cardiovascular disease initiative. Arterioscler Thromb Vasc Biol 2014;34: 939–45.

100. Puleo PR, Guadagno PA, Roberts R, et al. Early diagnosis of acute myocardial infarction based on assay for subforms of creatine kinaseMB. Circulation 1990;82:759–64.

90. Addona TA, Shi X, Keshishian H, et al. A pipeline that integrates the discovery and verification of plasma protein biomarkers reveals candidate markers for cardiovascular disease. Nat Biotechnol 2011;29:635–43.

101. Shlipak MG, Sarnak MJ, Katz R, et al. Cystatin C and the risk of death and cardiovascular events among elderly persons. N Engl J Med 2005;352: 2049–60.

91. Mertins P, Mani DR, Ruggles KV, et al. Proteogenomics connects somatic mutations to signalling in breast cancer. Nature 2016;534: 55–62. 92. Williams EG, Wu Y, Jha P, et al. Systems proteomics of liver mitochondria function. Science 2016;352:aad0189.

102. Daniels LB, Laughlin GA, Sarno MJ, et al. Lipoprotein-associated phospholipase A2 is an independent predictor of incident coronary heart disease in an apparently healthy older population: the Rancho Bernardo Study. J Am Coll Cardiol 2008;51:913–9. 103. Brennan ML, Penn MS, Van Lente F, et al. Prognostic value of myeloperoxidase in patients

with chest pain. N Engl J Med 2003;349: 1595–604. 104. Kavsak PA, MacRae AR, Newman AM, et al. Effects of contemporary troponin assay sensitivity on the utility of the early markers myoglobin and CKMB isoforms in evaluating patients with possible acute myocardial infarction. Clin Chim Acta 2007;380:213–6. 105. Johnson BD, Kip KE, Marroquin OC, et al. Serum amyloid A as a predictor of coronary artery disease and cardiovascular outcome in women: the National Heart, Lung, and Blood InstituteSponsored Women’s Ischemia Syndrome Evaluation (WISE). Circulation 2004;109:726–32. 106. Adams JE III, Bodor GS, Dávila-Román VG, et al. Cardiac troponin I. A marker with high specificity for cardiac injury. Circulation 1993;88: 101–6. 107. Reichlin T, Hochholzer W, Bassetti S, et al. Early diagnosis of myocardial infarction with sensitive cardiac troponin assays. N Engl J Med 2009; 361:858–67. 108. Danese E, Montagnana M. An historical approach to the diagnostic biomarkers of acute coronary syndrome. Ann Transl Med 2016;4:194. 109. Ladenson JH. A personal history of markers of myocyte injury [myocardial infarction]. Clin Chim Acta 2007;381:3–8. 110. Eng JK, McCormack AL, Yates JR. An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. J Am Soc Mass Spectrom 1994;5: 976–89. 111. O’Leary NA, Wright MW, Brister JR, et al. Reference sequence (RefSeq) database at NCBI: current status, taxonomic expansion, and functional annotation. Nucleic Acids Res 2016;44: D733–45.

KEY WORDS mass spectrometry, molecular phenotyping, post-translational modifications, protein arrays, protein dynamics, protein signatures