Volume 61, Number 7
OBSTETRICAL AND GYNECOLOGICAL SURVEY
Copyright © 2006
by Lippincott Williams & Wilkins
CME REVIEWARTICLE
21
CHIEF EDITOR’S NOTE: This article is part of a series of continuing education activities in this Journal through which a total
of 36 AMA/PRA category 1 creditsTM can be earned in 2006. Instructions for how CME credits can be earned appear on the
last page of the Table of Contents.
Clinical Proteomics: A Novel Diagnostic
Tool for the New Biology of Preterm
Labor, Part I: Proteomics Tools
Catalin S. Buhimschi, MD,* Carl P. Weiner, MD,†
and Irina A. Buhimschi, MD*
*Assistant Professor, Department of Obstetrics, Gynecology and Reproductive Science, Yale University School
of Medicine, New Haven, Connecticut; and †Professor and Chair, Department of Obstetrics and Gynecology,
University of Kansas School of Medicine, Kansas City, Kansas
The molecular mechanisms regulating myometrial contractility and preterm premature rupture of
the membranes leading to preterm birth are poorly understood. The completion of the human
genome sequence led to the development of functional genomics and gene array technology to
simultaneously identify candidate genes potentially involved in regulation of human parturition.
However, the study of living systems can now be expanded past genomics based on the rationale
that it is the protein products of the genes, not simply gene expression, that have effects and cause
disturbances at the cellular level. Therefore, identification of disease biomarkers, followed by a
description of their functional networks, has the potential to significantly aid the development of
new strategies for the prediction, diagnosis, and prevention of preterm birth. Interest in mass
spectrometry and its use as a new clinical diagnostic tool has grown rapidly and is poised to
become an important medical field for the next century.
Target Audience: Obstetricians & Gynecologists, Family Physicians
Learning Objectives: After completion of this article, the reader should be able to state the general
concept of proteomics, summarize the use of proteomics as a potential clinical tool as a biomarker of
disease, and recall that proteomics can be a means for understanding mechanisms of disease states.
Pregnancy is a special time during a woman’s
reproductive life as a result of the unique physiology
The authors have disclosed that they have no financial relationships with or interests in any commercial companies pertaining to
this educational activity.
The authors have disclosed that proteomic tools have not been
approved by the U.S. Food and Drug Administration for diagnosis
of human diseases and their application at this time remains for
research purposes only.
Lippincott Continuing Medical Education Institute, Inc. has
identified and resolved all faculty conflicts of interest regarding
this educational activity.
Reprint requests to: Catalin S. Buhimschi, MD, Department of
Obstetrics, Gynecology & Reproductive Science, Yale University
School of Medicine, 333 Cedar Street, LCI 804, P.O. Box 208063,
New Haven, CT 06520-8063. E-mail: catalin.buhimschi@yale.edu.
and the presence of a developing fetus. Despite an
impressive amount of effort and extensive research,
our knowledge of parturition and fetal physiology
remains limited. Scientists have exhaustively investigated over the past century “the timing of birth,” the
development, physiology, and pathophysiology of
the fetus, and its environment. Yet, our understanding of the biologic mechanisms that control the
events initiating term or preterm delivery remains
limited. As a direct consequence, we lack the therapeutic tools to block or circumvent the maladaptive
process.
Preterm delivery remains a major public health problem with lasting familial and societal repercussions (1).
481
482
Obstetrical and Gynecological Survey
Prevention strategies have failed and the prevalence of
preterm birth in the United States rose to an unprecedented 12.3% in 2003 (2). Preterm birth is associated
with almost 70% of neonatal deaths and up to 75% of
neonatal morbidity (1). It is critical that we develop a
working understanding of the highly controlled and
synchronized biochemical mechanisms that occur in the
myometrium as it interacts with the fetal–placental
compartments. For many years, research in preterm
labor has concentrated on identifying and localizing
individual factors when, similar to other pregnancyrelated disorders, preterm parturition involves complex
molecular protein to protein interactions functioning in
an interconnected cellular network regulated by receptors (3). Without a clear understanding of the intricate
pregnancy microenvironment, the complexity of clinical management of preterm labor increases exponentially. In the absence of robust diagnostic tools, therapy
cannot be provided or is delayed.
This article has 2 goals: first, to familiarize the
general obstetrician–gynecologist with the concept of
proteomics and second, to provide a general overview of the role of clinical proteomics in the identification of disease biomarkers and the generation of
protein profiles that make early diagnosis possible
and reveal fundamental mechanisms that should lead
to the first targeted treatments of preterm labor.
INTRODUCTION TO GENERAL
PROTEOMICS
Proteomics
Proteomics is the field of study that encompasses
knowledge of the structure, function, and expression
of all proteins in the biochemical or biologic context
of an organism at a given moment (4). Developed in
the postgenome era, the science of proteomics complements the genome initiative, which progressed
during the 20th century from the original description
of DNA to knowledge of the genes responsible for
specific human diseases. The ultimate goal of identifying and sequencing of the human genome has
become a reality (5). Tremendous advances in the
field of genomics exploded during the sequencing of
some 40,000 genes, paving the way for a new medical field, gene therapy (6,7). However, although
genomics represents a significant advancement, the
human genome fails to reflect the enormity and complexity of the human proteome. The concept of one
gene:one protein has fallen by the wayside. The
complex issue of posttranslational protein modification and variation in the sequence of amino acids can
be addressed only by full knowledge of the proteome, not the genome.
Unfortunately, mapping of the human proteome is
an almost impossible goal to achieve, at least in the
near future. Thus, it is appropriate to set more realistic, achievable goals. Until now, scientists have
tended to concentrate on accumulating information
about the nature of proteins and their absolute and
relative levels in cells or different biologic compartments (8–10). Such data can be useful, but information inherent to the broader definition of proteomics
must also be obtained if the true promise of preventing prematurity and its major complications is to be
realized (6).
Advances in protein analysis have provided valuable
perspectives on the posttranscriptional regulation of
gene expression and subtle protein–protein interactions
(Fig. 1) (11,12). For example, there is a poor correlation
between the abundance of mRNA and the amount of
protein in human tissues, suggesting that posttranscriptional regulation of gene expression is common (11).
This adds significantly to the difficulty of understanding a complex dynamic proteome.
Acquiring knowledge of the function of thousands of proteins is a challenge and the means to
support such endeavors must be provided. One
response to new ideas or approaches is to claim
they are not really new at all. In this instance, the
distinction between protein chemistry and proteomics may be difficult for the unfamiliar to
comprehend. Proteomics is not the science of protein biochemistry. Protein chemistry emphasizes
the importance of understanding protein structure
and function, and involves work toward complete
sequence analysis and a mathematical representation of how the structure enables function (4). In
contrast, proteomics is the study of multiprotein
systems and focuses on the interplay of multiple,
distinct proteins and their roles as part of a larger
network. The analysis is directed toward complex
mixtures and identification is based not necessarily
on complete sequencing, but rather on partial sequence analysis aided by a large database and matching tools. In this respect, proteomics is functional
biology, whereas protein chemistry is structural biology. There are 2 major issues confronting the field of
proteomics: first is its breadth, because the network of
proteins is often far larger than anticipated; and second,
it is more complicated than sequencing genomes. To
overcome these obstacles, new technologies are being
constantly developed.
Clinical Proteomics Y CME Review Article
483
Fig. 1. The cellular proteome at a relay from the genetic information to cellular function.
Proteomic Tools
Proteomics tools include protein separation and/or
identification of proteins in biologic samples coupled to
computational algorithms that allow the extraction of
the relevant information from the totality of data. In its
initial format, proteomics relied on high-resolution,
2-dimensional gel electrophoresis with isoelectric focusing in SDS-PAGE gels (2D-PAGE) This technique
was used to separate, identify, quantitate, and catalog a
large number of individual proteins present in complex
samples such as cerebrospinal fluid, plasma, seminal
fluid, and amniotic fluid (13–16). The first gel dimension allows separation according to protein charge by
isoelectric focusing, whereas the second dimension allows separation by protein size. After separation, the
proteins are visualized using gel staining procedures
such as Coomassie blue, silver staining, or fluorescent
tracers. Proteomics analysis using 2D-PAGE protein
separation is often criticized because the process of
image analysis necessary, to determine differential protein expression, can be laborious as a result of gel-to-gel
variations that confound the analytic process. To eliminate this weakness, fluorescence 2-dimensional difference gel electrophoresis (DIGE) was developed (17).
This technique allows multiple samples to be coseparated and visualized on one 2D-PAGE gel by labeling
with different fluorescent dyes (Cy2, Cy3, and/or Cy5).
Up to 3 images are captured on the gel using the Cy2,
Cy3, and/or Cy5 excitation wavelengths. The images
are then merged and differences between them deter-
mined using image analysis software. To overcome the
high cost of equipment as well as expendable supplies
such as the fluorescent dyes, many academic institutions have implemented this technique in their protein
core facility.
Recent advances in mass spectrometry (MS) have
allowed further refinements of the 2D-PAGE technique. Mass spectrometry now has the ability to identify
and characterize picomole quantities of gel-separated
proteins, making partial sequence analysis possible
(18). The available instrumentation is highly sensitive, robust, and reliable for the analysis of peptides
and integral proteins (19). Essentially, mass spectrometers consist of 3 parts: an ionization source, a
mass analyzer, and an ion detector (Fig. 2) (20). The
ionization source converts molecules into gas-phase
ions, which are then separated by the mass analyzer
and transferred to the ion detector. The mass spectrometer does not actually measure the molecular
mass of the sought proteins directly, but rather the
mass-to-charge ratio (m/z value) of the resulting
ions. In many cases, the ions encountered in mass
spectrometry have just one charge (z ⫽ 1), so the m/z
value is numerically equal to the molecular (ionic)
mass in Daltons.
The mass analyzer uses physical properties such as
their electric or magnetic field to separate ions by
their m/z ratios. They can also be separated by their
time of flight (TOF), the time it takes to reach the
detector. The rule of thumb is that the larger the m/z
484
Obstetrical and Gynecological Survey
Fig. 2. Components of a MALDI–TOF mass spectrometer with the data output (mass spectrum) shown below.
ratio of the ion, the longer it will take to reach the
detector. Two ionization techniques are commonly
used: electrospray ionization and matrix-assisted
laser desorption/ionization (MALDI). Electrospray
ionization creates ions by the application of a electric
potential to a flowing liquid such as a solvent, causing the fluid to charge and subsequently spray very
small droplets containing the analyte (20). The solvent is then removed, most frequently by heat, and
multiple charge ions are formed before their recognition by the ion detector.
In matrix-assisted laser desorption/ionization, sample molecules are bombarded with a laser beam to
induce ionization. The sample is premixed with a
highly energy-absorbing molecule (a matrix compound), which transforms the laser energy into the
excitation energy of the sample. This process leads to
subsequent ejection of the matrix compound and ions
into the gas phase of the mass analyzer so that they
can be detected by the ion detector.
The most commonly used instruments can be grouped
into either single-stage mass spectrometers or tandem
MS systems. Single-stage mass spectrometers, notably MALDI–TOF (matrix assisted laser desorption/
ionization time of flight), were used most frequently
for large-scale protein identification from species
with small or known genomes (18). Tandem MS
instruments such as triple quadruple, ion trap, and the
most recent advanced quadruple TOF (Q-TOF) allow
protein identification by sequence database searching
(19). The high accuracy of the Q-TOF technology
makes the combination of MALDI–Q-TOF configurations the best for de novo protein sequencing.
However, although accurate technology is essential
for novel protein discovery, greater automation, increased comprehensiveness, and friendlier technology
are equally important for the rapid and accurate diagnose of human diseases using proteomics.
The goal of making a rapid diagnosis with the least
amount of sample manipulation led to the development or surface-enhanced laser desorption/ionization
(SELDI). When used in conjunction with protein
chip arrays (21), SELDI allows for the isolation and
identification in complex biologic samples of peptides and proteins with specific properties. Protein
chip array assays using SELDI–TOF–MS technology
provide a valuable research tool as a result of the
multidimensional nature of protein separation, which
can be optimized for complex mixtures of proteins.
Moreover, SELDI can detect and quantitate both
proteins and posttranslationally modified forms of
these proteins in a single assay (22). The enhanced
separation is made possible by using a variety of active
surfaces on the protein chip arrays. The various array
Clinical Proteomics Y CME Review Article
surfaces interact differentially with constituents in the
biologic sample based on their hydrophobicity (H4,
H50 arrays), isoelectric point (SAX, WCX, CM10,
Q10 arrays), metal affinity (IMAC arrays), or ability to
bind to a specific antibody (PS10, PS20 arrays).
Briefly, a crude biologic sample is placed on a
small area of the protein chip array (spot). Specific
chemical interactions occur between the chip surface
and the biomolecules in the sample, depending on the
binding conditions and chip surface. The protein chip
surface preferentially binds some specific protein
structures while repelling others, thus allowing selection of a specific set of proteins that can be subsequently subjected to mass spectrometry for further
identification. By varying the chip surfaces, washing
conditions, incubation times, laser intensities, and
energy-absorbing molecules, an almost infinite number of experimental conditions can be designed for
the optimal separation of one protein or a group of
proteins from all the others.
Such advances in proteomics would not be possible
without the existence of large protein databases derived from organisms with known genomes. (23,24)
The ExPASy server (ExPASy Proteomics Server,
www.expasy.ch) and SwissProt database are perhaps
the best examples, providing links to other proteomics servers around the world and tools for further
sequence analysis. The continuous updating and validation of these databases are critical, because newly
discovered posttranslational modifications and variants
are published almost daily. Previously sequenced proteins and 2D-PAGE catalogs also stay at the core of
proteomic discovery. Several databases include mapping of biologic fluids such as plasma, urine, cerebrospinal fluid, or tissues such as heart, kidney, and
breast to allow experiments completed in the laboratory to be complemented by virtual experiments performed using bioinformatics tools (25). For example,
one way to identify proteins that have been separated
on gels (one-dimensional or 2-dimensional) is to
subject them to trypsin digestion. Each protein generates a unique combination of fragments (tryptic
digest fragments) whose masses allow insight into
the identity of protein (3). This strategy for protein
identification is known as peptide mass fingerprinting. Once there is a degree of confidence on the
sought protein’s identity, more definite identification
and confirmation can be achieved by de novo sequencing of tryptic fragments either by mass spectroscopy or using more traditional techniques such as
specific antiprotein antibodies.
The third essential tool of proteomics consists of
computer software that can first, differentiate among
485
thousands of proteins characteristic for a living system and second, accurately identify the sequence of
amino acids corresponding to a protein in the database. With the aid of specialized algorithms, the
software compares the data with the information in
the database. The investigator can analyze the results
and evaluate the quality of the data much faster than
with manual or visual discrimination. Given the wealth
of heterogeneous proteomic data and the numerous
bioinformatics tools available, scientists frequently are
faced with the dilemma of which computer tools to use.
Proteomics 2D-PAGE software such as Melanie offer
sophisticated state-of-the-art analysis for the identification, quantification, and matching of the gels. Most
combine comprehensive, advanced statistical and classification capabilities as well as versatile search engines
and reporting functions. Melanie, PD Quest, Phoretics
(for 2-dimensional gel analysis), and Ciphergen Biomarker Patterns software (for SELDI data) are just
few of the software programs available.
APPLICATIONS OF PROTEOMICS
Proteomics Has 4 Principal Parts
Data Mining
Identification of proteins in a biologic sample. The
goal is to catalog the proteome rather than to infer its
composition based on gene expression.
Differential Protein Expression Profiling
The identification of multiple proteins in a biologic
sample as a reflection of a particular state of the
organism or cell (disease state). Expression profiling
is essentially a more specialized form of mining and
involves a differential analysis in which the proteomes of 2 conditions of the biologic system are
compared (eg, disease–nondisease).
Protein–Network Mapping
Seeks to identify how proteins interact with one
another in living systems. In reality, it is this interaction that determines the function of a biologic
system. Proteomics offers the unique opportunity to
characterize complex networks using affinity-capture
techniques coupled with analytic proteomics methods. Protein mapping provides the opportunity to
assess the status of all participants in the pathway
simultaneously.
486
Obstetrical and Gynecological Survey
Protein Modification Mapping
Identifying how and where proteins are modified.
Characterization of posttranscriptional protein modification is one of the most ambitious goals of proteomics. Although several techniques are used to
detect modified proteins (eg, antibodies for specific
phosphorylated amino acid residues or fragments of
proteins), the precise sites of modification remain
largely unknown.
Over the last decade, proteome scientists have
largely focused their attention on 2 major areas:
expression proteomics, which seeks to quantify the
up- or downregulation of proteins, and functional
proteomics, which seeks to characterize protein activities, complex protein interactions, and signaling
pathways (26). In the past, studies designed to investigate the expression of different proteins sought to
first quantify and then compare the expression of
proteins in abnormal versus normal clinical conditions. The ultimate goal was the recognition of a
biomedical application; by this comparative approach, the newly identified proteins that differentiate the 2 conditions could be used as diagnostic
biomarkers. Today, an emerging field, clinical proteomics, seeks to apply the science of proteomics in
the search for biomarkers and the generation of protein profiles that can rapidly predict, diagnose, and
monitor treatment of human diseases, including preterm birth (14,27–30).
In part II, we will detail the emerging role of
proteomics in identifying the causes of spontaneous
preterm labor and birth.
REFERENCES
1. Stoll BJ, Hansen NI, Adams-Chapman I, et al. National Institute of Child Health and Human Development Neonatal Research Network. Neurodevelopmental and growth impairment
among extremely low-birth-weight infants with neonatal infection. JAMA 2004;292:2357–2365.
2. Hamilton BE, Martin JA, Sutton PD. Births: preliminary data for
2003. Natl Vital Stat Rep November 23, 2004;53(9).
3. Shankar R, Gude N, Cullinane F, et al. An emerging role for
comprehensive proteome analysis in human pregnancy research. Reproduction 2005;129:685–696.
4. Lieber CD. Introduction to Proteomics. Tools for the New
Biology. Totowa, NJ: Humana Press Inc, 2002.
5. Venter JC, Adams MD, Myers EW, et al. The sequence of the
human genome. Science 2001;291:1304–1351.
6. Banks RE, Dunn MJ, Hochstrasser DF, et al. Proteomics: new
perspectives, new biomedical opportunities. Lancet 2000;
356:1749–1756.
7. Zenclussen AC, Zenclussen ML, Ritter T, et al. The use of
gene therapy tools in reproductive immunology research. Curr
Gene Ther 2005;5:459–466.
8. Farag K, Hassan I, Ledger WL. Prediction of preeclampsia:
can it be achieved? Obstet Gynecol Surv 2004;59:464–482.
9. Park KH, Yoon BH, Shim SS, et al. Amniotic fluid tumor
necrosis factor-alpha is a marker for the prediction of earlyonset neonatal sepsis in preterm labor. Gynecol Obstet Invest
2004;58:84–90.
10. Lockwood CJ, Krikun G, Schatz F. Decidual cell-expressed
tissue factor maintains hemostasis in human endometrium.
Ann N Y Acad Sci 2001;943:77–88.
11. Anderson L, Seilhamer J. A comparison of selected mRNA
and protein abundances in human liver. Electrophoresis 1997;
18:533–537.
12. Monti M, Orru S, Pagnozzi D, et al. Interaction proteomics.
Biosci Rep 2005;25:45–56.
13. Wiederkehr F, Vonderschmitt DJ. [2-dimensional electrophoresis of cerebrospinal fluid in various neurological patients.] Schweiz Med Wochenschr 1985;115:368–373.
14. Chen JH, Chang YW, Yao CW, et al. Plasma proteome of
severe acute respiratory syndrome analyzed by twodimensional gel electrophoresis and mass spectrometry. Proc
Natl Acad Sci U S A 2004;101:17039–17044.
15. Fung KY, Glode LM, Green S, et al. A comprehensive characterization of the peptide and protein constituents of human
seminal fluid. Prostate 2004;61:171–181.
16. Vuadens F, Benay C, Crettaz D, et al. Identification of biologic
markers of the premature rupture of fetal membranes: proteomic approach. Proteomics 2003;3:1521–1525.
17. Unlu M, Morgan ME, Minden JS. Difference gel electrophoresis: a single gel method for detecting changes in protein
extracts. Electrophoresis 1997;18:2071–2077.
18. Patterson SD, Aebersold R. Mass spectrometric approaches
for the identification of gel-separated proteins. Electrophoresis 1995;16:1791–1814.
19. Gygi SP, Aebersold R. Mass spectrometry and proteomics.
Curr Opin Chem Biol 2000;4:489–494.
20. Yates JR 3rd. Mass spectrometry. From genomics to proteomics. Trends Genet 2000;16:5–8.
21. Kuwata H, Yip TT, Yip CL, et al. Bactericidal domain of lactoferrin: detection, quantitation, and characterization of lactoferrin in serum by SELDI affinity mass spectrometry. Biochem
Biophys Res Commun 1998;245:764–773.
22. Fung ET, Yip TT, Lomas L, et al. Classification of cancer types
by measuring variants of host response proteins using SELDI
serum assays. Int J Cancer 2005;115:783–789.
23. Hermjakob H, Apweiler R. The Proteomics Identifications Database (PRIDE) and the ProteomExchange Consortium: making proteomics data accessible. Expert Rev Proteomics 2006;
3:1–3.
24. Lo SL, You T, Lin Q, et al. SPLASH: systematic proteomics
laboratory analysis and storage hub. Proteomics 2006;6:
1758–1769.
25. Hochstrasser DF. Proteome in perspective. Clin Chem Lab
Med 1998;36:825–836.
26. Neubauer G, Gottschalk A, Fabrizio P, et al. Identification of
the proteins of the yeast U1 small nuclear ribonucleoprotein
complex by mass spectrometry. Proc Natl Acad Sci U S A
1997;94:385–390.
27. Petricoin EF, Ardekani AM, Hitt BA. Use of proteomic pattern
in serum to identify ovarian cancer. Lancet 2002;359:572–
527.
28. Papadopoulos MC, Abel PM, Agranoff D, et al. A novel and
accurate diagnostic test for human African trypanosomiasis.
Lancet 2004;363:1358–1363.
29. Pucci-Minafra I, Fontana S, Cancemi P, et al. Proteomic patterns of cultured breast cancer cells and epithelial mammary
cells. Ann N Y Acad Sci 2002;963:122–139.
30. Meehan KL, Holland JW, Dawkins HJ. Proteomic analysis of
normal and malignant prostate tissue to identify novel proteins
lost in cancer. Prostate 2002;50:54–63.