Content uploaded by Isabel Sá-Correia
Author content
All content in this area was uploaded by Isabel Sá-Correia on Aug 08, 2014
Content may be subject to copyright.
943
Review
www.expert-reviews.com ISSN 1478-9450
© 2010 Expert Reviews Ltd
10.1586/EPR.10.76
Proteomic analysis based on
2D electrophoresis
Genome-wide approaches have become obliga-
tory tools in biology research since the advent
of whole-genome sequencing. This is especially
true for microbiologists given that most available
genome sequences belong to microorganisms, as
is the case for the first bacterium, Haemophilus
influenzae, and the first eukaryote, Saccharomyces
cerevisiae, whose genomes were fully sequenced
in 1995 [1] and in 1996 [2], respectively. Having
much smaller genomes than higher eukaryotes,
research on microorganisms, particularly on pro-
karyotes, now relies on the availability of hun-
dreds of bacterial genome sequences. According
to the Genomes Online Database (GOLD) [101]
there were, in September 2009, more than 1000
fully sequenced genomes and over 5000 on going
genome projects, of which more than 4000
relate to bacterial and 450 to fungal genomes,
the remainder referring to parasites, plants and
other multicellular eukaryotes.
Soon after the first genome sequences were
made available, genome-wide expression screen-
ing tools were developed, both at the transcrip-
tome [3] and proteome [4] levels. Unavoidably,
the first organisms for which these genome-wide
screens were carried out were microorganisms.
In particular, the yeast S. cerevisiae is considered
a ‘work horse’ for the development of new tools
and technological advances in this field [5,6].
Unlike any other genome-wide approach, the
history of proteomics started 20 years before the
release of the first bacterial genome sequence,
Isabel Sá-Correia†1 and
Miguel C Teixeira1
1Institute for Biotechnolog y and
Bioengine ering, Biological Scie nces
Research Group, Centro de Engenharia
Biológica e Q uímica, Instituto Superior
Técnico, Technical Unive rsity of Lisbon,
Av. Rovisco Pais, 1049- 001 Lisboa,
Portugal
†Author for corresp ondence:
Tel.: +351 218 417 682
Fax: +351 218 419 199
isacorreia @is t.utl.pt
Quantitative proteomics based on 2D electrophoresis (2-DE) coupled with peptide mass
fingerprinting is still one of the most widely used quantitative proteomics approaches in microbiology
research. Our view on the exploitation of this global expression analysis technique and its
contribution and potential to push forward the field of molecular microbial physiology towards a
molecular systems microbiology perspective is discussed in this article. The advances registered in
2-DE-based quantitative proteomic analysis leading to increased protein resolution, sensitivity and
accuracy, and the promising use of 2-DE to gain insights into post-translational modifications at a
proteome-wide level (considering all the proteins/protein forms expressed by the genome) are
focused on. Given the progress made in this field, it is foreseen that the 2-DE-based approach to
quantitative proteomics will continue to be a fundamental tool for microbiologists working at a
genome-wide scale. Guidelines are also provided for the exploitation of expression proteomics
data, based on useful computational tools, and for the integration of these data with other
genome-wide expression information. The advantages and limitations of a complete 2-DE-based
expression proteomics analysis, envisaging the quantification of the global changes occurring in
the proteome of a given cell depending on environmental or genetic manipulations, are discussed
from the microbiologist’s perspective. Particular focus is given to the emerging field of
toxicoproteomics, a new systems toxicity approach that offers a powerful tool to directly monitor
the earliest stages of the toxicological response by identifying critical proteins and pathways that
are affected by, and respond to, a chemical stress. The experimental design and the bioinformatics
analysis of data used in our laboratory to gain mechanistic insights through expression proteomics
into the responses of the eukaryotic model Saccharomyces cerevisiae or of Pseudomonas strains
to environmental toxicants are presented as case studies.
KEYWOR DS:$ELEC TROPHORESISsEXPRESSIONPROTEOMICSsMICROBIALPROTEOMICSsPHOSPHOPROTEOMICSsQUANTITATIVE
PROTEOMICSsTOXICOPROTEOMIC S
2D electrophoresis-based
expression proteomics:
a microbiologist’s perspective
Expert Rev. Proteomics 7(6), 943–953 (2010)
944
Review
Expert Rev. Proteomics 7( 6), (2010)
Sá-Correia & Teixeira
with the development of a revolutionary technique, denomi-
nated ‘2D electrophoresis’ (2-DE), in 1975 [7]. At this time,
there were constraints limiting the applicability of this tool, and
the term ‘proteome’ was not created until 1995 [8,9]. Although
2-DE allowed large-scale proteome-wide protein separation and
quantification, it was not possible to carry out the subsequent
step of high-throughput protein identification before acquiring
knowledge of the genome sequence of a specific organism. Today,
such identification is fairly easy and relatively inexpensive, relying
on peptide mass fingerprinting [10] or, when necessary, tandem
mass spectrometry (MS/MS). This approach is commonly based
on the measurement of the trypsin-digested peptide mass:charge
ratio and subsequent comparison with the theoretical mass:charge
peptide ratios for all the translated open reading frames contained
in each sequenced genome. This, of course, is only feasible with
the progressive release of genome sequences. Therefore, the first
expression-proteomics experiments were really carried out in
1995 [8,11], using the information on the genome sequences that
were then almost completed. The field of proteomics, the large-
scale study of proteins, includes a set of techniques whose aim is
exclusive and independent of other genome-wide approaches, such
as genomics, transcriptomics or metabolomics.
Today, it is possible to use this approach even for species whose
genomes are not sequenced. For example, the proteomic analysis
of the response to phenol provided important clues regarding the
mechanisms behind the regulation of phenol catabolism in the
Pseudomonas spp. M1, which exhibits exceptional biodegradation
ability towards phenol and other toxic and recalcitrant compounds,
even though the genome of this strain was not available [12]. A total
of 56 of the 87 proteins assigned matched the P. aeruginosa pro-
teins present in databases. Remarkably, of the genome sequences
available, P. aeruginosa was found to be the closest species to the
strain M1 [12]. However, the remaining 31 proteins of interest had
no registered homologs and could not be identified. In such cases,
other MS approaches, such as peptide fragmentation fingerprinting
or de novo sequencing, may be required.
The design of a quantitative proteomics experiment to identify
the global response of a microbial strain to a particular chemi-
cal stress involves cell exposure to stress, proteome extraction
and eventual proteome fractionation, solubilization of proteins
in the isoelectric-focusing buffer, protein separation in the first
and second dimensions, staining of proteins, comparative gel
analysis, identification of proteins whose content varies, data
analysis and identification of mechanisms of toxicity and stress
response (a scheme of these various steps can be seen in [13]). 2-DE
can resolve complex protein mixtures by charge using isoelectric
focusing (IEF; using an immobilized pH gradient – commercial
immo bilized pH gradient gel strips that separate complex protein
samples more reproducibly and in higher amounts [from µg to
mg]), and then by mass using sodium docecyl sulfate–polyacryl-
amide gel electrophoresis (SDS–PAGE). Sample application for
IEF can be achieved by in-gel rehydration or by cup or paper
loading, cup/paper loading being very helpful for separating very
basic proteins [14]. After separation, resolved proteins from 2D
gels can be manually excised as plugs and digested overnight with
trypsin. The peptide fragments can be analyzed in an external
MS platform, releasing the microbiology laboratories from the
financial burden of sustaining such platforms and associated staff,
a set-up that is far from its main objectives. This option is pivotal
to maintain the feasibility of using expression proteomics as a
routine approach in many microbiology laboratories. Moreover,
once a proteome reference map is produced, it can be used to
readily identify protein spots in every gel produced using the same
protocol and organism, thus facilitating downstream analysis.
Alternatives to this classical method have been developed,
such as the multidimensional protein identification technology
(MudPIT) approach [15], which relies mostly on multidimensional
liquid chromatography (LC) for protein separation and MS for
protein identification. Especially since it became possible to use
MS or MS/MS to perform relative and absolute protein quanti-
fication in complex samples, LC-MS approaches have become
proteomics procedures [16,17]. However, LC-MS-based proteomics
relies on more-sophisticated equipment and requires qualified
users. Actually, only a few MS-based quantitative proteomics tech-
niques can accurately be considered established and require quite
expensive commercially available kits (e.g., isotope-coded affinity
tag [ICAT] and isobaric tag for relative and absolute quantifica-
tion, as reviewed elsewhere [16,17]). Although isotope-coded affinity
tag technology is present in most proteomics facilities today, in
many cases, the use of these novel approaches remains confined to
the laboratories of their inventors and collaborators [18]. LC separa-
tion approaches have gained increased relevance in the analysis of
samples that are difficult to study using 2-DE, for instance, using
very hydrophobic integral membrane proteins. However, LC-MS
approaches still lack the high resolving power of 2-DE, in par-
ticular, to resolve protein species [19]. Most proteins occur in 2-DE
gels as many spots, corresponding to different levels and natures
of post-translational modifications, and the relative intensity of
each of these protein species can be individually quantified. This
protein species level is ignored by the high-throughput bottom-up
LC-MS approaches. Theoretically, 2-DE should be able to resolve
up to 10,000 proteins in a single gel [20] , 2000 being routinely
detected and quantified using fluorescent staining, with each spot
representing amounts as low as 1 ng of protein [21]. In addition,
2-DE is not limited to simultaneous comparisons between two
or three samples. Indeed, new samples can always be compared
with those analyzed before using an identical protocol, and can
be stored as gel images on a computer, which is not possible using
LC-MS approaches [21]. The recent development of various fluores-
cent dyes for protein labeling, some of them exhibiting specificity
towards given protein classes, the development of the difference
gel electrophoresis (DIGE) technology, and the improvement of
software packages for 2-DE gel analysis have further increased
the sensitivity, reliability, reproducibility and range of applicabil-
ity of 2-DE-based proteomics (highlighted in FIGURE 1) [22]. This
experimental approach was recently explored by our group to
compare the expression profiles of clonal isolates of Burkholderia
cenocepacia retrieved from a patient with cystic fibrosis, who was
chronically infected during 4 years with the same strain. This
patient died with the cepacia syndrome and the clonal variants
www.expert-reviews.com 945
Review
Microbial 2-DE proteomics
exhibited differential antibiotic susceptibility. This comparative
analysis revealed mechanistic insights into the adaptive strategies
employed by these bacteria to deal with the stressing conditions
of the cystic fibrosis lung, including those related to aggressive
antibiotic therapy [Madeir a et al., Unpublish ed Data]. In the same con-
text of microbial clinical proteomics, the DIGE technology was
also used to study the response to the antibiotic ciprofloxacin in
Streptococcus uberis [23] .
In this review, current advances in sample preparation (pro-
teome extraction a nd proteome prefractionation or fraction
enrichment), proteome separation by 2-DE, data acquisition and
data analysis using specialized software will be addressed. The
synergy of our microbiologist’s perspective and the technological
advances in the field of quantitative proteomics is emphasized
to support an ambitious systems biolog y approach to micro-
bial physiology issues and to extend the knowledge gathered to
more-complex and less-accessible organisms. 2-DE-based expres-
sion proteomics, envisaging the quantitative analysis of the global
changes occurring in the proteome of a given cell depending on
environ mental and genetic manipulations, has contributed deeply
to elucidate scientific issues of molecular and cellular microbiol-
ogy and microbial physiology in different microbial systems, as
exemplified throughout this paper. The expertise of our labora-
tory in exploring 2-DE-based proteomics applied to bacteria of
the Pseudomonas genus and to S. cerevisiae is exploited herein as
case studies.
Advances in the sensitivity & accuracy of 2-DE-based
quantitative proteomics
One of the major drawbacks associated with 2-DE-based proteomics
is the reduced capacity to detect low-abundance proteins. However,
the development of fluorescence protein staining has contributed
Figure 1. Exploitation of 2D electrophoresis-based proteomics in the study of the microbial stress response. Quantitative
expression proteomics (e.g., using the DIGE technology) can be used to assess proteome-wide cellular responses but also, upon
prefractionation, organelle- or subcellular structure-specific stress responses. Selecting appropriate strains, for example, deletion mutants
devoid of a transcription factor or a kinase, and using specific fluorescent dyes, such as ProQ Diamond for the labeling of
phosphoproteins, it is possible to use 2-DE-based proteomics to identif y transcription factor regulons or protein kinase targets at a global
scale. When studying the effect of pro-oxidant chemicals, it is also possible to quantify changes at the level of proteome carbonylation
(using DNP labeling of proteins) or S-thiolation (using NEM labeling of proteins), through redox proteomics approaches.
DIGE: Difference gel electrophoresis; DNP: 2,4-dinitrophenylhydrazine; NEM: N-ethylmaleimide.
Global s
Respo
Control
Environmental Stress
Wild-type
∆transcription_factor
Wild-type
∆transcription_factor
∆protein_kinase Organe
structure-
stress re
Transcrip"on factor
kinase targets
Phosphoproteomics
Pre-frac"
Redox proteomics
"on DIGE
Cy5
Cy3
Cy2
Soluble proteome
oQ Diamond
Flamingo/
Sypro Ruby
DTN-labelled
abelled
Flamingo/
Sypro Ruby
∆protein_kinase
vs
Wild-type
∆transcription factor
∆protein kinase
Wild-type
∆transcription factor
∆protein kinase
Control vs environmental stress DIGE
Redox proteomics
Phosphoproteomics
Proteome-wide
kinase targets
Transcription factor
regulation Membrane proteome
Soluble proteome
Prefractionation
Organelle- or
structure-specific
stress response
Global stress
response
Cy2
Cy3
Cy5
Flamingo/Sypro Ruby
DNP labeled
NEM labeled
Proteome oxidation
profile
ProQ Diamond
Flamingo/Sypro
Ruby
Expert Rev. Proteomics 7( 6), (2010)
946
Review Sá-Correia & Teixeira
to overcoming this limitation. Fluorescent dyes used frequently
for in-gel protein staining include SYPRO® Ruby (Molecular
Probes, OR, USA) and Flamingo™ (BioRad, CA, USA), whose
sensitivity dynamic ranges embark 1–1000 ng of protein. This is
a remarkable achievement when compared with classical protein-
staining methods, based on Coomassie® Blue (Merck, Darmstadt,
Germany; dynamic range: 30–900 ng) or silver nitrate (dynamic
range: 1–10 ng), which are still important to produce preparative
gels for protein identification. Although quite sensitive, the very
narrow dynamic range of silver nitrate staining makes it completely
unsuitable for quantitative proteomics. Protein staining with DIGE
cyanine (Cy) dyes has even higher sensitivity, allowing the detec-
tion of proteins over a 20,000-fold concentration range [22]. DIGE
relies on prelabeling of different protein samples with different
fluorophores – for example, sample A (Cy3), sample B (Cy5) and
standard (Cy2) – followed by sample mixing, separation by 2-DE
and scanning of the gel in three wavelengths specific for each dye
(FIGURE 1). Since up to three samples can be separated in the same
gel, the DIGE approach also makes a perfect alignment of protein
spots possible. The use of a selected standard protein extract, to be
separated together with pairs of the samples under study in each
of the several gels required for a complete experiment, contributes
to the optimization of gel-to-gel spot matching.
The relative quantification of protein concentration in 2-DE
gels has also been significantly improved with the advances in the
computational tools used for image analysis. Specialized software
has been developed and continuously improved to carry out the
electronic alignment of stained proteins, comparisons of the intensi-
ties of the same protein spots and normalization against the total
protein concentration (usually assumed to be represented by the
total spot volume in a given gel), calculating a ratio (fold change) for
each protein. There are some inherent limitations to this approach,
including: the difficulty to properly identify spots when they are
partially superimposed; the possibility to obtain mismatched spots;
and the difficulty to quantify the intensity of very intense – satu-
rated – or very subtle spots. To try to overcome these limitations,
the whole analysis has to be verified spot by spot, making it a
very time-consuming process. Recent powerful software packages
(e.g., those made available by Progenesis, BioRad, GE Healthcare
[USA] and other companies) include refined statistical analyses of
the gels. Average spot intensities for each condition are compared by
their normalized volumes using one-way ANOVA between-group
tests. Only significant spots, both statistically (p-value < 0.05) and
biologically (those exhibiting intensity variation), are considered
for further analysis. The possibility to carry out automatic statisti-
cal analysis through the available software packages increases the
degree of confidence in the assessment of protein-content variation,
making it acceptable to consider statistically meaningful variations
in protein concentrations as low as 1.2-fold.
Increasing the spectrum of proteins that can be
analyzed through 2-DE-based proteomics
One of the most noticeable shortcomings of 2-DE is the lim-
ited range of proteins that can be analyzed. Extremely acidic or
basic proteins, very low- or very high-molecular-weight proteins
and, more importantly, low-abundance proteins are not covered.
Indeed, proteins can range from 10 to 1000 kDa, from 3 to 10 in
pI, and from 100 to 100,000,000 copies per cell, with the dynamic
range of protein concentrations in cell-extract samples reaching up
to six orders of magnitude [24]. Continuous efforts have been made
to improve the resolution power of 2-DE gels, as exemplified in a
recent analysis of the Helicobacter pylori proteome [25]. In this case,
gel porosity was manipulated to allow the separation of very low-
molecular-weight proteins [25]. It is important to recognize that
no experimental design is, so far, capable of entirely resolving all
proteins/protein forms present in a cell. However, this problem can
be partially overcome with prefractionation of the protein sample,
together with the development of novel detergents, surfactants and
chaotropes that allow the solubilization of very-hydrophobic pro-
teins, in particular, integral membrane proteins that are difficult
to separate by 2-DE [21,26 ,27]. Fractionation approaches to isolate
proteins specifically associated with organelles or other cell struc-
tures (FIG URE 1) have been used with success in the analysis of sub-
proteomes from the nucleus, ribosomes, membranes, chloroplasts,
mitochondria, lysosomes, phagosomes and the cell wall [28 –30],
further allowing the concentration of proteins present in lower
relative abundance in whole-cell extracts [26 –28].
A good example of the advantages of prefractionation in
Gram-negative bacteria is provided by the recent study of the
global response to phenol stress in the environmental model
strain Pseudomonas putida KT244, which was carried out in our
group [31]. A first study led essentially to the identification of
soluble proteins that are up- or down-regulated following cell
exposure to phenol [31], and, using a motif-finding algorithm
under development, motifs of biological relevance were identi-
fied in the co-regulated gene promoters [32]. The extension of this
analysis to the membrane proteome involved the fractionation of
the total proteome in the soluble proteome and the subproteomes
corresponding to inner-membrane and outer-membrane proteins
and their solubilization by the detergent dodecylmaltoside, which
is compatible with subsequent protein resolution by 2-DE (FIGUR E 2)
[33]. This study allowed the further identification of the coordinate
increase of the concentration of protein subunits of known or
putative solvent efflux pump systems (TtgA, TtgC, Ttg2A, Ttg2C
and PP_1516–7) and decrease in the content of porins OprB,
OprF, OprG and OprQ, consistent with an adaptive response to
reduce intracellular phenol concentrations [33]. Protein fraction-
ation was essential to the identification of these protein content
changes, as detailed in the article, and to the understanding of
the global response to phenol stress.
A possible drawback of prefractionation is the requirement for
very large amounts of cells per sample to reach, at the end of the
process, sufficient protein quantity to allow a 2-DE based pro-
teomics analysis. Although this is a serious problem in the case of
medical research when relying on the availability of small-scale
biopsies, this is not usually a critical issue for most microbiological
samples, since it is, in general, possible to obtain high amounts of
protein extracts in a relatively short period of time at a low cost.
Of course fractionation cannot fully solve the problem of protein
concentration range in a complex mixture, since all subfractions
www.expert-reviews.com 947
Review
Microbial 2-DE proteomics
exhibit a wide range of protein content as
well. For example, nuclei-enriched frac-
tions contain very high concentrations of
histones but a low abundance of transcrip-
tion factors. However, prefractionation
tremendously increases the total amount of
proteins that can be detected and quantified
in the various fractions [33]. Although 2-DE
has been used with some success to separate
membrane-enriched fractions, it has proven
unsuitable to detect integral membrane
proteins, in particular, multispanner solute
transport systems [27,33].
2-DE-based proteomics of
post-translational modifications
One of the most significant advantages
of 2-DE is the straightforward separation
of protein species that only differ in their
post-translational modifications. Over the
years, the assessment of a number of post-
translational modifications has become
possible at the proteome-wide level using
2-DE [28].
For instance, the development of ProQ
Emerald, a fluorescent dye specific for
glyco protein staining, made the study
of the glycoproteome possible [3 4]. ProQ
Emerald is conjugated to glycoproteins
prior to 2-DE and, upon separation, is
detected using a fluorescence scanner. The
same gel can be sequentially post-stained
with SYPRO Ruby to allow the comparison
between the glycoproteome and the ‘whole’
proteome [34] . Although glycoproteomics
is still a relatively unexplored field, this
‘omics’ approach may provide very signifi-
cant results given that glycosylation often
determines biological function, sub cellular
loca lization, immunogenicity, solubility
and the stability of proteins [34, 35].
Protein phosphorylation is one of the
most widespread regulatory mechanisms
in nat ure, and this post-tra nslational
mod ificat ion can be studied through
phosphoproteomics. The occurrence of
phosphoproteins is readily detectable in
2-DE gels, as multiple pI variants of a
single protein. Phosphoproteomics using 2-DE was first based
on the use of pre-electrophoretic 32P labeling of proteins or
western blotting using antiphosphoamino acid antibodies [36] .
Both approaches are highly laborious and, although radioac-
tivity detection is highly sensitive, it raises a number of issues
regarding safet y and waste disposal. 2-DE-based phosphopro-
teomics was later attempted by the comparison of duplicate
protein extracts, either treated or not with a protein phospha-
tase [37] . More recently, the use of the fluorescent dye ProQ
Diamond, specific for phosphoproteins, made possible the
easy and selective detection and quantification of the phos-
phoproteome, following separation by 2-DE [38] . A combina-
tion of ProQ Diamond phosphoprotein staining and SYPRO
Ruby staining of all proteins in the same 2-DE gel can be used
Figure 2. Comprehensive high-throughput proteome-wide data analysis, based
on expression profile and biological function. (A) Visualization of the results
obtained by protein clustering based on expression profiles (B) or shared ‘gene ontology’
terms, using different computational approaches, indicated in the figure, and applied to
the yeast proteome response to 2,4-dinitrophenylhydrazine-induced stress [4 4].
GO: Gene ontology; ROS: Reactive oxygen species.
FIGURE 2 B is modified from [105].
0 10 20 30 40
Oxygen and ROS metabolic process
Response to unfolded protein
Hydrogen transport
Regulation of pH
pH reduction
Carbohydrate metabolic process
Purine nucleotide metabolic process
Amino acid metabolic process
Protein frequency (%)
Frequency in
the genome
Frequency in
the dataset
GO-based clustering
Expression profile-based clustering
Normalized volume Distance
Protein dendrogram
2.5
2.0
1.5
1.0
0.5
0.0
3
2
1
0
-1
-2
-3
662
308
512
655
532
490
457
754
735
407
313
270
805
666
705
294
681
720
484
495
819
678
702
470
292
736
680
660
709
559
599
309
512
541
677
522
727
575
Expert Rev. Proteomics 7( 6), (2010)
948
Review Sá-Correia & Teixeira
(FIG URE 1) . This approach is called ‘multiplex proteomics’, since
it allows the quantitative dichromatic fluorescence detection of
different classes of proteins in 2-DE gels [38].
Proteins modified by oxidative stress, for example, by carbonyl-
ation [39,40] or S-thiolation [41], can also be quantified upon separa-
tion by 2-DE [42]. Current approaches in this field of 2-DE-based
redox proteomics are mostly suitable for evaluating irreversible pro-
tein modifications when they occur in a high frequency. However,
the use of affinity selection methods to trap subproteomes can
increase the possibility of evaluating reversible protein-oxidation
events and selectively increases the concentration of oxidized pro-
teins to detectable levels [43]. Changes occurring in the carbonylated
proteome of yeast cells exposed to hydrogen peroxide can be assessed
using 2-DE [39,40]. Upon prederivatization of protein extracts with
the carbonyl reactive compound 2,4-dinitrophenylhydrazine, pro-
teins can be separated by 2-DE, transferred into a polyvinylidene
fluoride or nitro cellulose membrane, and 2,4-dinitrophenylhydra-
zine (DNP)-labeled proteins can be detected and quantified using
specific anti-DNP antibodies [39,40] . Oxidized proteins can also
be purified with N-(6-(biotinamido)hexyl)-3´-(2´-pyridyldithio)-
propionamide and then separated and quantified through 2-DE,
or labeled with N-(14C)ethylmaleimide, followed by separation
through 2-DE and quantified by autoradiography, for the detec-
tion of oxidized protein thiols [41]. Other thiol-specific reagents
such as maleimides, iodoacetic acid, iodoacetamide, thiosulfates
and others can alternatively be used, and covalently attached to
fluorescent molecules such as fluorescein for detection [42]. These
experimental approaches may shed light on the mechanisms mediat-
ing thiol oxidation as a physiological event and as a result of oxida-
tive stress. The perception that oxidative stress has the power to
activate or denature proteins through S-thiolation or carbonylation
highlights the fact that the assessment of protein concentration per se
is not enough information, and that the study of post-translational
modifications at a proteome-wide scale is pivotal for understanding
biological phenomena.
Toxicoproteomics: a systems microbiology perspective
Microbiologists have an interest in exploring expression pro-
teomics as a genome-wide expression approach in order to gain
mechanistic insights into the global microbial responses to an
alteration of the environment and/or the genetic background.
Studies in the field of microbial toxicoproteomics are an example
of the potential of these research topics in the application of pro-
teomics in toxicology. This systems toxicity approach offers a pow-
erful tool to directly monitor the earliest toxicological response by
identifying critical proteins and pathways that are affected by, and
respond to, adverse chemical and environmental exposures using
quantitative proteomics. The data are analyzed and explored using
bioinformatics tools to extract significant hypotheses/conclusions,
and their integration with data coming from diverse genome-
wide approaches is instrumental. The assessment of proteomic
expression changes in response to toxicants has the potential to
provide both earlier and more sensitive biomarkers and signatures
of a toxic response than traditional toxicological methods. The
uncovering of new biomarkers is essential for mechanistic research
and risk assessment. It is expected that further coordinated stud-
ies on the mechanisms governing the toxicity of, and resistance
to, environmental toxicants may guide the development of rapid
inexpensive and reliable assays suitable for the large-scale toxicity
assessment of chemical compounds with an impact on environ-
mental or human health. Indeed, there is a pressing need for such
assays, as the majority of chemicals in commercial use have not
been comprehensively tested for human toxicity.
The budding yeast S. cerevisiae, used in the production of
bread, beer and wine and as a cell factory in biotechnology, is an
important and robust model species for the study of the genet-
ics and physiolog y of less-accessible eukaryotes. Indeed, yeast
shares basic molecular mechanisms and signaling pathways
with higher eukaryotes. Yeast is easy and inexpensive to grow,
its genetic manipulation is straightforward and the full genome
sequence has been available since 1996 [2]. This yeast species pio-
neered the development and testing of a number of postgenomic
experimental approaches and computational tools. Furthermore,
yeast-specific large-scale biological resources and genome-wide
strategies are available, together with huge amounts of informa-
tion based on functional genomics strategies, facilitating the bio-
logical interpretation of genome-wide data. Two case studies in
the field of yeast toxicoproteomics, currently under study in our
laboratory, focus on the agricultural fungicide mancozeb, linked
to the development of Parkinsonism symptoms and cancer, and
the widely used auxin-like herbicide 2,4-dichlorophenoxyacetic
acid (2,4-D). The widespread and intensive use and misuse of
pesticides has led to a growing number of resistant fungal and
weed species and may give rise to a number of toxicological and
environmental problems. The understanding of the underlying
mechanisms is crucial to dealing with their consequences.
To gain insights into the yeast global response to 2,4-D, the
variations occurring at the protein-expression levels in response to
this herbicide were examined [44], together with the comparative
analysis of the yeast transcriptome [45]. The responsive proteins were
first clustered based on their expression profile using the SameSpots
gel analysis software from NonLinear Dynamics (Newcastle-upon-
Tyne, UK) (FIGUR E 2A) , and were grouped using the FatiGO soft-
ware [102], which orders the groups of proteins associated with gene
ontology (GO) terms, based on the over-representation of each GO
term in the dataset in comparison with the genome (FIGURE 2B). Over-
represented GO terms included the carbohydrate catabolic process
and amino acid biosynthetic process, correlating with increased
energy requirement and with a dramatic reduction of the intracel-
lular amino acid concentration [44]. These adaptive mechanisms
presumably occur as a response to the deleterious effects exerted by
2,4-D on plasma membrane lipid organization and permeability,
leading to nutrient import inhibition. The current model suggests
that the presence of 2,4-D exerts a repressing effect over the target
of the rapamycin pathway [46], a central controller of cell growth
in all eukaryotes, which regulates the balance between protein
synthesis and degradation in response to nutrient quantity and
quality. Moreover, the observed increased expression of Vma1p and
Vma2p, which are two subunits of the peripheral catalytic sector of
the multimeric H+-ATPase, present in the membrane of the yeast
www.expert-reviews.com 949
Review
Microbial 2-DE proteomics
vacuole, guided the analysis of the role of the V-ATPase in 2,4-D
stress tolerance, particularly in the recovery of the intracellular
pH [47] and of the amino acid vacuolar and cytosolic pools [44] .
2D electrophoresis-based quantitative proteomics was a lso
explored in order to obtain insights into the mechanisms of man-
cozeb toxicity and tolerance in yeast [48]. The transcription factors
predicted to regulate the expression of the genes encoding the pro-
teins whose abundance varied in response to mancozeb-induced
stress were identified using the YEASTRACT database [49,103], a
publicly available curated repository of more than 48,000 regula-
tory associations between transcription factors and target genes
in S. cerevisiae. YEASTR ACT also includes computational tools
for the analysis of genome-wide expression results (microarray and
quantitative proteomics data). Remarkably, approximately 90% of
the genes encoding proteins whose content increased in response
to mancozeb were found to be targets of the transcription factor
Yap1, one of the main regulators of the oxidative stress response
in yeast, which is consistent with the hypothesized pro-oxidant
action of this fungicide (FIGURE 3B). It is expected that this proteomic
analysis may be extended to the phytopathogenic fungi that are the
natural targets of this agricultural fungicide and may also provide
clues to understanding mancozeb toxicity in humans. Indeed,
70% of the proteins identified as having an altered expression level
in response to mancozeb toxicity have human orthologs [48], while
the whole genome only includes approximately 30% proteins with
human homologs. Proteomics data were more-recently integrated
with results from a genome-wide screening of the EUROSCARF
yeast collection (comprising approximately 5000 single-deletion
mutants in which each yeast nonessential gene was individually
deleted). Results revealed that 286 genes are required to provide
protection against mancozeb toxicity [50]. Both expression pro-
teomics and chemogenomics results point to a possible role of
the agricultural fungicide as a pro-oxidant agent. Mancozeb was
found not to induce endogenous reactive oxygen species or free-
radical accumulation, but to lead to massive oxidation of pro-
tein cysteines, due to its action as a thiol-reactive compound [50].
Mancozeb-induced protein S-thiolation is also consistent with
the pivotal role of Yap1 in the yeast proteome-wide response to
mancozeb. Indeed, Yap1p has two molecular redox centers: one
triggered by thiol-reactive compounds and the other by reactive
oxygen species. In both cases, Yap1p is activated and localized
to the cell nucleus, thus mediating the upregulation of its target
genes that depend on the triggered redox center [51]. Physical and
genetic interactions of yeast proteins, whose abundance changed
in response to mancozeb, and of their human orthologs were pre-
dicted using the STRING information system (FIGURE 3A) [104] . The
generated networks of protein–protein associations of yeast and
human proteins share a high degree of similarity, suggesting that
the main conclusions of this study can be considered powerful
indications to expand the knowledge obtained in yeast to humans.
The data obtained in this study indicate that the expression level
of some target genes represents some interesting early biological
indicators of a toxic response and might constitute a useful tool
for toxicity assessment and the analysis of the presence of such
toxicants in the environment. The number of target genes whose
expression is significantly changed and the nature of the affected
biological processes emphasize the lack of risk assessment and the
relevance of research related to the biological effects of pesticides
prior to their commercial release and wide use in agriculture.
Expert commentary
Expression proteomics is instrumental to responding to the chal-
lenges of postgenomic microbiology research. 2-DE-based pro-
teomic analysis has been widely used over the last two decades,
particularly following the genome sequencing of a continuously
increasing number of microorganisms, and is still the most-
commonly used approach for expression proteomic studies [21,52],
despite its limitations [53,54] and the attempted development of
alternatives [16]. Although 2-DE has biophysical limitations that
constrain the resolution of proteins at extreme pH, hydropho-
bicity and mass, it exhibits a high resolving power and is an
extremely versatile separation technique that readily couples with
MS. The continued use of 2-DE-based expression proteomics,
especially in microbiology research, is apparently due to the fact
that this approach is compatible with the routine of modern
microbiolog y laboratories of international level. 2-DE-based
methodologies are well established, easy to implement and
continue to be a paradigm of proteomics research, especially in
research laboratories whose aim is not analytical proteomics and
the development of associated technology, but rather to use the
existing tools to address biological questions [12,13 ,18,31,33 ,44 ,48] .
Unlike LC-based proteomics, 2-DE only requires basic biochem-
istry knowledge and appropriate experimental skills. Moreover,
the necessary equipment and reagents for IEF and high-resolution
SDS–PAGE are relatively inexpensive. Furthermore, the advances
in the field of 2-DE-based proteomics reviewed in this paper
highlight the fact that part of the most critical constraints tradi-
tionally associated with this approach have been dealt with, and
it continues to be the methodology of choice to resolve and iden-
tify post-translational modifications. Altogether, these advances
have put 2-DE-based proteomics back on top as a great tool to
answer biological questions and to guide more profound studies
of molecular microbial physiology.
Five-year view
The important role of 2-DE-based proteomics, within the reach
of most microbiology laboratories, has pushed forward research
devoted to dealing with its limitations and has guaranteed the lon-
gevity of 2-DE-based proteomics coupled to peptide mass finger-
printing. 2-DE allies high-throughput proteome analysis with the
simplicity and versatility of its basic procedures. This versatility
makes it possible to extend the experimental set-up put together
for the quantitative proteomic analysis applied to the study of the
molecular physiology of our favorite microbe, and to the study of
various microbes, providing genome-wide responses to key issues
at the level of the effectors of a given response – the proteins.
The perception that post-translational modifications have the
potential to activate or reduce the biological activity of proteins
highlights the importance of studying these post-translational
modifications at a proteome-wide scale and not only the alteration
Expert Rev. Proteomics 7( 6), (2010)
950
Review Sá-Correia & Teixeira
of the protein content in response to a specific modification of
the microbial environment or genetic background. In this sense,
tools developed to quantitatively study subproteomes, phospho-
proteomes, glycoproteomes and redox-modified proteomes have
increased the possible applications of 2-DE to address key microbi-
ological issues. These approaches are expected to become standard
procedures in microbiology laboratories to help delineate the sig-
naling pathways and key players, particularly transcription factors,
protein kinases and phosphatases, underlying the proteome-wide
changes in both the protein content and protein modification
induced by environmental stress and/or alteration of the genetic
background. Further technological advances are still required
to lead to the final goal of reaching 100% protein coverage in
proteome-wide studies, allowing the complete separation and
chemical characterization of all protein species in a given organ-
ism. The continuous development of databases and computational
tools that allow the full exploitation of results coming from expres-
sion proteomics and from other omics approaches in a systems
biology perspective will also be instrumental in the coming years.
This will require the coordinate activity of multi disciplinary teams
with expertise in biological sciences and functional genomics,
bioinformatics and computational biology.
Acknowledgements
The authors would like to acknowledge all those who have, over the years,
contributed to implementing and exploiting their expression proteomics plat-
form, namely Pedro M Santos, Tânia Simões, Catarina Roma-Rodrigues,
Andreia Madeira, Nuno P Mira and Sandra C dos Santos.
Figure 3. Comprehensive high-throughput proteome-wide data analysis, based on protein–protein interaction and
transcriptional regulation data. Visualization of the results obtained by protein clustering based on protein–protein interactions (A) or
shared transcription factors (B), using different computational tools indicated in the figure, and applied to the yeast proteome response
to mancozeb-induced stress [49]. Emphasis is given to the important role of the transcription factor Yap1.
Modified from [106,107] .
Proteasome
General
stress
response
V-ATPase
Msn2
Skn7
Msn4
Yap1
Sod2 Gre2 Tsa1 Grx1 Ahp1 Tsa2 Trr1 Sod1
Transcription factor-based clustering
Protein/genetic interaction-based clustering
www.expert-reviews.com 951
Review
Microbial 2-DE proteomics
Financial & competing interests disclosure
Expression proteomics work in our laboratory has been financially sup-
ported by FEDER and Fundação para a Ciência e Tecnologia (contracts
ERA-PTG/SAU/ 0001/ 2008, in the conte xt of EraNet Pathogenomics
ADHR ES Signature Project, PTDC/AGR-A LI /102608/20 08, PTDC /
SAU-FC F / 7176 0 /20 06, PTDC /SAU-M II/69591/ 2006 , PTD C /
BIO/72063/2006 and PTDC/BIO / 66151/2006). The authors have no
other relevant affiliations or financial involvement with any organization
or entity with a financial interest in or financial conflict with the subject
ma tter or m ate r ial s dis cus sed in t he ma nus cript a par t from
those disclosed.
No writing assistance was utilized in the production of this manuscript.
References
Papers of special note have been highlighted as:
sOFINTEREST
1 Fleischmann RD, Adams MD, White O
et al. Whole-genome random sequencing
and assembly of Haemophilus influenzae
Rd. Science 269(5223), 496–512 (1995).
2 Goffeau A, Barrell BG, Bussey H et al. Life
with 6000 genes. Science 274(5287), 546,
563–547 (1996).
3 DeRisi JL, Iyer VR, Brown PO. Exploring
the metabolic and genetic control of gene
expression on a genomic scale. Science
278(5338), 680 –686 (1997).
4 Shevchenko A, Jensen ON,
Podtelejnikov AV et al. Linking genome
and proteome by mass spectrometry:
large-scale identification of yeast proteins
from two dimensional gels. Proc. Natl
Acad. Sci. USA 93(25), 14440 –14445
(1996).
5 Kolkman A, Slijper M, Heck AJ.
Development and application of proteomics
technologies in Saccharomyces cerevisiae.
Trends Biotechnol. 23(12), 598–604
(2005).
6 Suter B, Auerbach D, Stagljar I. Yeast-based
functional genomics and proteomics
technologies: the first 15 years and beyond.
Biotechniques 40(5), 625– 644 (2006).
7 O’Farrell PH. High resolution two-
dimensional electrophoresis of proteins.
J. Biol. Chem. 250(10), 4007–4021 (1975).
8 Wasinger VC, Cordwell SJ, Cerpa-Poljak A
et al. Progress with gene-product mapping
of the Mollicutes: Mycoplasma genitalium.
Electrophoresis 16(7), 1090–1094 (1995).
9 Wilkins MR , Sa nchez JC, Gooley A A
et al . Progress wit h proteome projects:
why all proteins expressed by a genome
should be identified and how to do it.
Biotechnol. Genet . Eng. Rev. 13, 19–50
(1996).
10 Gevaert K, Vandekerck hove J.
Protei n identificat ion methods in
proteomics. Elect rophoresis 21(6),
1145–1154 (2000 ).
11 Boucherie H, Dujardin G, Kermorgant M,
Monribot C, Slonimski P, Perrot M.
Two-dimensional protein map of
Saccharomyces cerevisiae: construction of a
gene-protein index. Yeast 11(7), 601–613
(1995).
12 Santos PM, Roma V, Benndorf D,
von Bergen M, Harms H, Sá-Correia I.
Mechanistic insights into the global
response to phenol in the phenol-
biodegrading strain Pseudomonas sp. M1
revealed by quantitative proteomics. OMICS
11(3), 233–251 (2007).
13 Teixeira MC, Santos PM, Rodrigues C,
Sá-Correia I. Teaching expression
proteomics : from the wet-lab to the laptop.
Biochem. Mol. Biol . Educ. 37(5), 279–286
(2009).
14 Gorg A, Obermaier C, Boguth G et al.
The current state of two-dimensional
electrophoresis with immobilized pH
gradients. Electrophoresis 21(6), 1037–1053
(2000 ).
15 Washburn MP, Wolters D, Yates JR 3rd.
Large-scale analysis of the yeast proteome
by multidimensional protein identification
technology. Nat. Biotechnol. 19(3),
242–247 (2001).
16 Sechi S. Quantitative Proteomics by Mass
Spectrometry. Walker JM (Ed.). Humana
Press, NJ, USA (2007).
17 Chen G, Prama nik BN. Application of
LC/MS to proteomics studies : current
stat us and f uture prospects. Drug
Discov. Today 14(9–10), 465 –471
(2009).
18 Nesaty y VJ, Suter MJ. Analysis of
environmental stress response on the
proteome level. Mass Spectrom. Rev. 27(6),
556–574 (2008).
19 Nagele E, Vollmer M, Horth P, Vad C.
2D-LC/MS techniques for the
identification of proteins in highly complex
mixtures. Expert Rev. Proteomics 1(1),
37–46 (2004).
20 Klose J. Protein mapping by combined
isoelectric focusing and electrophoresis
of mouse tissues. A novel approach to
test ing for i nduced point mutations in
mammals. Humangenetik 26(3), 231–243
(1975).
21 Lopez JL. Two-dimensional
electrophoresis in proteome expression
analysis. J. Chromatogr. B Analyt. Technol.
Biomed. Life Sci. 849(1–2), 190–202
(2007).
s 2EVIEWOFTHEUNIQUEFEATURESOF$
ELECTROPHORESIS$%ASABASIC
PROTEOMICSTOOLINCLUDINGADESCRIPTIONOF
THEPROTOCOLSTHECHALLENGESTHE
LIMITATIONSTHEPOTENTIALANDTHEFUTURE
OF$%INPROTEOMICS
22 Minden J. Comparative proteomics and
difference gel electrophoresis. Biotechniques
43(6), 739–743 (2007).
Key issues
s 2D electrophoresis still remains the main platform for quantitative proteomics, being the most amenable approach for most
microbiology- driven laboratories.
s The use of fluorescent dyes for protein detection and quantification, and especially coupled to the difference gel electrophoresis (DIGE)
technology, has increased the sensitivity and accuracy of 2-DE-based quantitative proteomics.
s Prefractionation of protein extracts is crucial to increase the spectrum of analyzable proteins, in particular, those occurring in low
abundance in the whole cell.
s 2-DE-based proteomics tools applied to the study of post-translational modifications (e.g., redox, phosphor- and glyco-proteomics)
have the power to tremendously increase our global understanding of signal transduction pathways affecting most
biological processes.
s The development and use of appropriate computational tools to integrate proteomics data with knowledge gathered from other
‘omics’ approaches is crucial to guide fur ther studies on integrative microbiology.
Expert Rev. Proteomics 7( 6), (2010)
952
Review Sá-Correia & Teixeira
s $ESCRIPTIONANDDISCUSSIONOFTHE
PRINCIPLESOFQUANTITATIVEPROTEOMICSAND
THEDEVELOPMENTOFDIFFERENCEGEL
ELECTROPHORESIS$)'%
23 Poutanen M, Varhi mo E , Kalkk inen N,
Sukura A, Varmanen P, Savijoki K.
Two-dimensional difference gel
electrophoresis analysis of
Streptococcus uberis in response to
mutagenesis-inducing ciprofloxacin
challenge. J. Proteome Res. 8(1), 246–255
(2009).
24 Corthals GL, Wasinger VC,
Hochstrasser DF, Sanchez JC. The
dynamic range of protein expression: a
challenge for proteomic research.
Electrophoresis 21(6), 1104–1115 (2000).
25 Jungblut PR, Schiele F, Zimny-Arndt U
et al. Helicobacter pylori proteomics by
2-DE/MS, 1-DE-LC/MS and functional
data mining. Proteomics 10(2 ), 182–193
(2010).
26 Roma-Rodrigues C, Santos PM,
Benndorf D, Rapp E, Sá-Correia I.
Response of Pseudomonas putida KT2440
to phenol at the level of membrane
proteome. J. Proteomics 73(8), 1461–1478
(2010).
27 Braun RJ, Kinkl N, Beer M,
Ueffing M. Two-d imensional
electrophoresis of membrane proteins.
Anal . Bioanal. Chem. 389(4), 1033–1045
(2007).
28 Cordwell SJ, Nouwens AS, Verri lls NM,
Bassea l DJ, Walsh BJ. Subproteomic s
based upon protein cellular locat ion a nd
relative solubi lities in conjunction with
composite two-di mensiona l
electrophoresis gels. Electrophoresis 21(6 ),
1094–1103 (2000).
29 Taylor SW, Fahy E, Ghosh SS. Global
organellar proteomics. Trends Biotechnol.
21(2), 82–88 (2003).
30 Pardo M, Ward M, Bains S et al.
A proteomic approach for the study of
Saccharomyces cerevisiae cell wall
biogenesis. Electrophoresis 21(16),
3396–3410 (2000 ).
31 Santos PM, Benndorf D, Sá-Correia I.
Insights into Pseudomonas putida KT2440
response to phenol-induced stress by
quantitative proteomics. Proteomics 4(9),
2640 –2652 (2004).
32 Mendes ND, Casimiro AC, Sa ntos PM,
Sa-Correia I, Oliveira AL, Freitas AT.
MUSA: a parameter free algorithm for the
identification of biologically significant
motifs. Bioinformatics 22(24), 2996 –3002
(2006).
33 Roma-Rodrigues C, Santos PM,
Benndorf D, Rapp E, Sa-Correia I.
Response of Pseudomonas putida KT2440 to
phenol at the level of membrane proteome.
J. Proteomics 73(8), 1461–1478 (2010).
s /PTIMIZATIONOFAPROTOCOLTOALLOWTHE
$%BASEDQUANTITATIVECOMPA RISONOF
THEPseudomonas putidaMEMBRANE
PROTEOMEFOLLOWINGFRACTIONATIONINTO
INNERANDOUTERMEMBRANEFRACTIONS
34 Steinberg TH, Pretty On Top K,
Berggren K N et al. Rapid and simple
single nanogra m detection of
glycoproteins in polyacrylamide gels and
on electroblot s. Proteomics 1(7), 841–855
(2001).
35 Imperiali B, O’Connor SE. Effect of
N-linked glycosylation on glycopeptide and
glycoprotein structure. Curr. Opin. Chem.
Biol. 3 (6), 643–649 (1999).
36 Thingholm TE, Jensen ON, Larsen MR.
Analytical strategies for
phosphoproteomics. Proteomics 9(6),
1451–1468 (2009).
37 Yamagata A, Kristensen DB, Takeda Y
et al. Mapping of phosphorylated proteins
on two-dimensional polyacr ylamide gels
using protein phosphatase. Proteomics 2(9),
1267–1276 (2002).
38 Steinberg TH, Agnew BJ, Gee KR et al .
Global quantitative phosphoprotein
analysis using multiplexed proteomics
technology. Proteomics 3(7), 1128–1144
(2003).
s /NEOFTHElRSTEXAMPLESOF
PHOSPHOPROTEOMICA NALYSISUSING$%
ANDTHEMULTIPLEXTECHNOLOGY
39 Costa VM, A morim MA, Quintanilha A,
Morada s-Ferreira P. Hydrogen peroxide-
induced carbonylation of key metabolic
enzymes in Saccharomyces cerevisiae: t he
involvement of the oxidative stress
response regulators Yap1 and Skn7. Free
Radic. Biol . Med . 33(11), 1507–1515
(2002).
40 Noda Y, Berlett BS, Stadtman ER,
Aponte A, Morga n M, Shen RF.
Identification of enzymes and regulatory
proteins in Escherichia coli that are oxidized
under nitrogen, carbon, or phosphate
starvation. Proc. Natl Acad. Sci. USA
104(47), 18456–18460 (2007).
41 Le Moan N, Clement G, Le Maout S,
Tacnet F, Toledano MB. The Saccharomyces
cerevisiae proteome of oxidized protein
thiols: contrasted functions for the
thioredoxin and glutathione pathways.
J. Biol. Chem. 281(15), 10420–10430
(2006).
42 Sheehan D. Detection of redox-based
modification in two-dimensional
electrophoresis proteomic separations.
Biochem. Biophys. Res. Commun. 349(2),
455–462 (2006).
s 2EVIEWONTHEMETHODOLOGIESFOR
REDOXPROTEOMICSBASEDON$%
43 Sheehan D, McDonagh B, Barcena JA.
Redox proteomics. Expert Rev. Proteomics
7(1), 1–4 (2010).
44 Teixeira MC, Santos PM, Fernandes A R,
Sá-Correia I. A proteome analysis of the
yeast response to the herbicide
2,4-dichlorophenoxyacetic acid. Proteomics
5(7), 1889–1901 (2005).
45 Teixeira MC, Fernandes AR, Mira NP,
Becker JD, Sá-Correia I. Early
transcriptional response of Saccharomyces
cerevisiae to stress imposed by the herbicide
2,4-dichlorophenoxyacetic acid. FEMS
Yeast Res. 6(2), 230–248 (2006).
46 Teixeira MC, Duque P, Sá-Correia I.
Environmental genomics: mechanistic
insights into toxicity of and resistance to
the herbicide 2,4-D. Trends Biotechnol .
25(8), 363–370 (2007).
47 Fernandes AR, Durão PJ, Santos PM,
Sá-Correia I. Activation and significa nce of
vacuolar H+-ATPase in Saccharomyces
cerevisiae adaptation and resistance to the
herbicide 2,4-dichlorophenox yacetic acid.
Biochem. Biophys. Res. Commun. 312(4),
1317–1324 (2003).
48 Santos PM, Simões T, Sá-Correia I.
Insights into yeast adaptive response to the
agricultural fungicide mancozeb:
a toxicoproteomics approach. Proteomics 3,
657–670 (2009).
s /NEOFTHElRSTEXA MPLESOFT HE
SUCCESSFULUSEOF$% BASED
PROTEOMICSINT HEEMERGINGlELD
OFTOX ICOPROTEOMICS
49 Teixeira MC, Monteiro P, Jain P et al.
The YE ASTRACT database: a tool for the
analysis of transcription regulatory
associations in Saccharomyces cerevisiae.
Nucleic Acids Res. 34(Database issue),
D446 –D451 (2006).
50 Dias PJ, Teixeira MC, Telo JP, Sá-
Correia I. Insights into the mechanisms of
toxicity and tolerance to the agricultural
fungicide mancozeb in yeast as suggested
by a chemogenomic approach. OMICS
14(2), 211–227 (2010).
51 Azevedo D, Tacnet F, Delaunay A,
Rodrigues-Pousada C, Toledano MB.
Two redox centers within Yap1 for H2O2
and thiol-reactive chemicals signaling. Free
Radic. Biol. Med. 35(8), 889–900 (2003).
www.expert-reviews.com 953
Review
Microbial 2-DE proteomics
52 Rabilloud T. Two-dimensional gel
electrophoresis in proteomic s: old, old
fashioned, but it still climbs up the
mountains. Proteomics 2(1), 3–10 (2002).
53 Gygi SP, Corthals GL, Zhang Y, Rochon Y,
Aebersold R. Evaluation of two-
dimensional gel electrophoresis-based
proteome ana lysis technology. Proc. Natl
Acad. Sci. USA 97(17), 9390–9395 (2000).
54 Righetti PG, Castagna A, Antonucci F
et al. Critical survey of quantitative
proteomics in two-dimensiona l
electrophoretic approaches. J. Chromatogr.
A 1051(1–2), 3–17 (2004).
s #RIT ICA LREVIEWOFTHEPROTOCOLS
APPLIEDTOQUANTITAT IVEPROTE OMICS
FOCU SEDONTHEADVANCESOF
$%BASEDAPPROACHES
Websites
101 GOLD, Genomes OnLine Databa se v 3.0.
GOLD indexing information
http://genomesonline.org/gold_statistics.
htm
102 Babelomics 4. Functiona l analysis
http://babelomics.bioinfo.cipf.es/
functional.html
103 YEASTR ACT
www.yeastract.com
104 STRING 8.3. Known and predicted
protein–protein interactions
http://string.embl.de/
105 FatiGO
http://fatigo.bioinfo.cipf.es
106 STRING
http://string.embl.de/
107 YEASTR ACT
www.yeastract.com
