Journal of Proteomics 151 (2017) 2–11

Contents lists available at ScienceDirect

Journal of Proteomics

journal homepage: www.elsevier.com/locate/jprot

Proteomic analysis of ovarian cancer cells during

epithelial-mesenchymal transition (EMT) induced by epidermal growth
factor (EGF) reveals mechanisms of cell cycle control
Mariana Lopes Grassi a,b, Camila de Souza Palma a,b, Carolina Hassibe Thomé a,b, Guilherme Pauperio Lanfredi a,b,
Aline Poersch a, Vitor Marcel Faça a,b,⁎
a
Dept. Biochemistry and Immunology, Ribeirão Preto Medical School, University of São Paulo, Ribeirão Preto, SP, Brazil
b
Cell-Based Therapy Center, Ribeirão Preto Blood Center, Ribeirão Preto Medical School, University of São Paulo, Ribeirão Preto, SP, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: Epithelial to mesenchymal transition (EMT) is a well-orchestrated process that culminates with loss of epithelial
Received 29 January 2016 phenotype and gain of a mesenchymal and migratory phenotype. EMT enhances cancer cell invasiveness and
Received in revised form 20 May 2016 drug resistance, favoring metastasis. Dysregulation of transcription factors, signaling pathways, miRNAs and
Accepted 7 June 2016
growth factors including EGF, TGF-beta and HGF can trigger EMT. In ovarian cancer, overexpression of the
Available online 6 July 2016
EGFR family is associated with more aggressive clinical behavior. Here, the ovarian adenocarcinoma cell line
Keywords:
Caov-3 was induced to EMT with EGF in order to identify specific mechanisms controlled by this process.
Epithelial to mesenchymal transition Caov-3 cells induced to EMT were thoroughly validated and a combination of subcellular proteome enrichment,
Epidermal growth factor GEL-LC-MS/MS and SILAC strategy allowed consistent proteome identification and quantitation. Protein network
Cell cycle analysis of differentially expressed proteins highlighted regulation of metabolism and cell cycle. Activation of rel-
Ovarian cancer evant signaling pathways, such as PI3K/Akt/mTOR and Ras/Erk MAPK, in response to EGF-induced EMT was val-
idated. Also, EMT did not affected the proliferation rate of Caov-3 cells, but led to cell cycle arrest in G1 phase
regulated by increased levels of p21Waf1/Cip1, independently of p53. Furthermore, a decrease in G1 and G2
checkpoint proteins was observed, supporting the involvement of EGF-induced EMT in cell cycle control.
Biological significance: Cancer is a complex multistep process characterized by accumulation of several hallmarks
including epithelial to mesenchymal transition (EMT), which promotes cellular and microenvironmental chang-
es resulting in invasion and migration to distant sites, favoring metastasis. EMT can be triggered by different ex-
tracellular stimuli, including growth factors such as EGF. In ovarian cancer, the most lethal gynecological cancer,
overexpression of the EGFR family is associated with more aggressive clinical behavior, increasing mortality rate
caused by metastasis. Our proteomic data, together with specific validation of specific cellular mechanisms dem-
onstrated that EGF-induced EMT in Caov-3 cells leads to important alterations in metabolic process (protein syn-
thesis) and cell cycle control, supporting the implication of EGF/EMT in cancer metastasis, cancer stem cell
generation and, therefore, poor prognosis for the disease.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction inducing the acquisition of a more motile mesenchymal phenotype

Cancer is a complex multistep process characterized by accumula-
tion of several genetic and epigenetic alterations that stimulate in ma-
lignant cells a number of hallmarks that promote uncontrolled
proliferation, migration and invasion [1]. During malignant tumor pro-
gression, epithelial to mesenchymal transition (EMT) arises as a crucial
mechanism to enhance cancer cell invasiveness and drug resistance,
⁎ Corresponding author at: Department of Biochemistry and Immunology, Ribeirão
Preto Medical School, University of São Paulo, Av. Bandeirantes, 3900 - Monte Alegre,
CEP 14049-900 Ribeirão Preto, SP, Brazil.
E-mail address: vitor.faca@fmrp.usp.br (V.M. Faça).
that progresses to detachment from the main epithelial tumor, invasion
of the surrounding stroma and culminating with cells entering in the
circulatory system to finally metastasize to distant sites [2].
Initially described as a distinct cell differentiation process in embryo-
genesis, EMT involves the loss of epithelial markers - such as the adher-
ents junction proteins E-cadherin and β-catenin, and the tight junction
proteins Claudin and Occludin - cytoskeleton reorganization and gain of
mesenchymal markers, including fibroblast-like cell morphology and
other proteins such as N-cadherin, vimentin, fibronectin and matrix me-
talloproteinase [3,4]. These morphogenetic changes are well-orches-
trated by a number of master transcription regulators that involve
three families of transcription factors: the zinc finger proteins Snail

http://dx.doi.org/10.1016/j.jprot.2016.06.009
1874-3919/© 2016 Elsevier B.V. All rights reserved.
 M.L. Grassi et al. / Journal of Proteomics 151 (2017) 2–11
and Slug, the zinc finger and homeodomain proteins Zeb1 and Zeb2 and
the helix-loop-helix proteins Twist (TWIST1 and TWIST2). All together,
these transcription factors repress epithelial markers, specially the ex-
pression of E-cadherin, as well as enhance the expression of mesenchy-
mal genes [5]
Several studies have shown that dysregulation of transcription fac-
tors, signaling pathways, miRNAs and growth factors can trigger the ini-
tiation of EMT. Growth factors such as transforming growth factor
(TGF), hepatocyte growth factor (HGF) and epidermal growth factor
(EGF) have been widely associated with EMT induction, since these
stimuli can be directly produced by tumor-associated stroma cells. In
addition to stimulation by growth factors, the effective implementation
of EMT is also dependent on a variety of signal transduction pathways,
including PI3K, MAPK, Wnt/β-catenin and Notch signaling [6,7].
EGF has contributed effectively as a potent EMT inducer in different
types of solid tumors, including gynecological cancer such as endome-
trial, ovarian and cervical cancer, significantly favoring the poor progno-
sis and increasing mortality rate caused by metastasis. In ovarian cancer,
the most lethal gynecological cancer, overexpression of the EGFR family
has been associated with more aggressive clinical behavior [8,9].
In addition, it has been shown that EGF stimulation can induce EMT
in ovarian surface epithelium (OSE) through inhibition of keratin ex-
pression and enhancing of pro-MMP 2/9 expression, with activation of
ERK and integrin-linked kinase (ILK) pathways. Moreover, EGF binds
to the EGF receptor (EGFR), inducing EMT through phosphorylation of
the Janus kinase/signal transducer and activator of transcription (JAK/
STAT3) pathway with involvement of ERK, increasing the
chemoresistance in ovarian carcinoma cell lines [10,11].
In order to obtain a proteomic perspective about the complex molec-
ular mechanisms of EMT, we performed a multiple-compartment anal-
ysis of the proteomic alterations during EMT induced by EGF in the
ovarian adenocarcinoma cell line Caov-3. Here, we report that EGF-in-
duced EMT activated signaling pathways relevant for growth and cell
proliferation, represented by increased phosphorylation of Akt, ERK1/2
and S6 ribosomal protein, which leads to alterations in the cell prolifer-
ation pattern. We also show that the underlying mechanisms of EMT in-
volve cell cycle arrest through activation of p21Waf1/Cip1 independent of
p53 with decreased phosphorylation status of checkpoint proteins such
as cyclin D1, Rb and CDK1/cdc2. Altogether, these results not only con-
tribute for broadening the comprehension of the molecular basis of
EMT, but also highlight important pathways that can serve as targets
to control cancer progression and metastasis.
2. Material and methods
2.1. Cell culture
Human ovarian adenocarcinoma cell line Caov-3 was acquired from
the ATCC and cultured in Dulbecco's Modified Eagle Medium (DMEM,
Life Technologies, Carlsbad, CA, USA) supplemented with 10% fetal bo-
vine serum (FBS, Thermo Scientific, Marietta, OH, USA), 100 U/ml peni-
cillin and 100 μg/ml streptomycin (Life Technologies). Cells were
maintained at 37 °C in a humidified incubator in an atmosphere of 5%
CO2.
2.2. EMT induction
Caov-3 cells were seeded at 1.5 × 106 cells on a 100 mm plate in
DMEM supplemented with 10% FBS. Cells were washed in pre-warmed
PBS and the media was replaced by DMEM without FBS. After 24 h,
DMEM without FBS and containing 10 ng/ml EGF (R&D Systems) was
added and cells were maintained in culture for 96 h to induce EMT.
Each 24 hour media was replaced by a new DMEM solution containing
10 ng/ml EGF was replaced.
3
2.3. Western Blot
Caov-3 cells were washed in cold PBS and lysed with CelLytic
(Sigma-Aldrich) containing the 5% (v/v) protease inhibitor cocktail
(Sigma-Aldrich). Three sonication cycles were carried out for 5 min
each in an ultrasonic bath (Unique, São Paulo, SP, Brazil) with cooled
water. Lysates were centrifuged at 20,000 ×g for 30 min at 4 °C, to ob-
tain the total cell lysates. The protein concentration was determined ac-
cording to the Bradford method (Bio-Rad, Hercules, CA, USA). Proteins
were submitted to SDS-PAGE and transferred to polyvinylidene fluoride
membranes (GE Healthcare, Life Sciences, Pittsburgh, PA, USA). Mem-
branes were blocked with 3% bovine serum albumin (BSA) in 0.05%
Tween-TBS and incubated with the specific antibodies. Rabbit anti-
Vimentin (#5741), rabbit anti-E-cadherin (#3195), rabbit anti-N-
cadherin (#4061), mouse anti-Snail (#3895), rabbit anti-EGFR
(#2646), rabbit anti-phospho-EGFR (Tyr1068) (#3777), rabbit anti-
p21Waf1/Cip1 (#2947), rabbit anti-p53 (#2527), rabbit anti-phospho-Rb
(Ser807/811) (#9308), mouse anti-cdc2 (#9116), rabbit anti-cyclin
D1 (#2978), rabbit anti-phospho-cyclin D1 (#3300), rabbit anti-mTOR
(#2983), rabbit anti-phospho-Erk 1/2 (Thr202/Tyr204) (#4370), rabbit
anti-Akt (#9272), rabbit anti-phospho-Akt (Ser 473) (#4060), rabbit
anti-GAPDH (#2118), and horseradish peroxidase-conjugated second-
ary antibody goat anti-mouse IgG were purchased from Cell Signaling
(Beverly, MA, USA). The protein-antibody complex was detected using
the ECL™ Prime Western Blotting Detection Reagents (GE Healthcare,
Life Sciences) and chemiluminescent was detected using a CCD-camera
based system (ImageQuant 350, GE Healthcare, Uppsala, Sweden).
2.4. Multiple reaction monitoring (MRM) analysis
Aliquots of 100 μg of proteins from total cell lysates, quantitated by
Bradford assay, were diluted in 50 μl Urea 8 M, Tris-HCl 0.1 M, pH 8.5.
Protein cysteine residues were reduced with 50 μg dithiothreitol at
37 °C for 30 min and then alkylated with 250 μg iodoacetamide for
30 min, protected from light. Protein solution was then diluted in
740 μl Tris-HCl 0.2 M, pH 8.0 to reduce urea concentration (b1.5 M)
and in-solution trypsin digestion was performed by adding 4 μg trypsin
(Promega), incubated overnight at 37 °C. Samples were then desalted
using solid-phase extraction columns Oasis HLB (Waters) according to
the manufacturer's instructions. Peptides were eluted in 52.5% ACN/
0.1% formic acid and dried in SpeedVac (Thermo Scientific, Marietta,
OH, USA). Before MRM analysis, samples were reconstituted in 50 μl of
5% acetonitrile/0.1% formic acid solution, agitated thoroughly, centri-
fuged for 20 min at 20,000 ×g and then transferred to injection vials.
Each sample was injected in triplicate through the LC-MS/MS Xevo
TQs system (Waters). Chromatographic separation was performed in a
UPLC (I-class, Waters) using a C18 column (1.8 μm particle size, 100 Å
pore size, 1 mm × 150 mm, Waters) in a linear gradient of 5 to 30% ace-
tonitrile over 15 min at 100 μl/min in a formic acid:acetonitrile:water
solvent system. Detection of proteotypic peptides was programmed in
a scheduled method using 3–5 transitions per peptide and monitoring
windows of 2 min. Detailed information about proteotypic peptides,
transitions, collision energies and dwell times are presented in Supple-
mentary Table 1. Both method development and data analysis were
conducted using Skyline software [12].
2.5. Immunofluorescence microscopy
Caov-3 cells grown on glass coverslips were rinsed briefly in PBS,
fixed in PBS containing 3% formaldehyde, and incubated in a solution
containing 0.3% triton x-100, 22.52 mg/ml glycine and 1% BSA in PBS
for 1 h at room temperature. Cells were incubated overnight with spe-
cific antibodies. Rabbit anti-E-cadherin (#3195) and rabbit anti-
Vimentin (#9854) were purchased from Cell Signaling (Beverly, MA,
USA). Mouse anti-N-cadherin (#33-3900) was purchased from Thermo
Fisher Scientific (Waltham, MA, USA). Secondary antibody anti-rabbit
4 M.L. Grassi et al. / Journal of Proteomics 151 (2017) 2–11
IgG (Alexa Fluor® 488-conjugated) (#4412) was purchased from Cell
Signaling (Beverly, MA, USA). Images were captured on Leica CTR
6000 Fluorescence, with sequential fluorophore excitation.
2.6. RTK activity array
Caov-3 cells were washed in cold PBS and lysed with buffer from
PathScan® RTK Signaling Antibody Array kit (Cell Signaling) containing
5% (v/v) protease inhibitor cocktail (Sigma-Aldrich). To extract proteins,
three sonication cycles were carried out for 5 min each in an ultrasonic
bath (Unique, São Paulo, SP, Brazil) with cooled water. Lysates were
centrifuged at 20,000 ×g for 30 min at 4 °C, and protein concentration
was determined according to the Bradford method (Bio-Rad, Hercules,
CA, USA). 50 μg aliquots of total cell extracts were used, according to
the manufacturer's instructions. The protein-antibody complex was de-
tected using the ECL Western Blotting Detection Reagents (GE
Healthcare, Life Sciences) and signals were detected using a CCD-Cam-
era (Image Quant LAS 4000 mini, Uppsala, Sweden).
2.7. SILAC labeling
Caov-3 cells were cultured in SILAC DMEM (SILAC™ Advanced
DMEM, Thermo Fisher Scientific) containing heavy lysine (13C6-L-ly-
sine) (0.1 mg/ml) and supplemented with 10% dialyzed FBS, 2 mM L-
glutamine, 4.5 mg/ml D-glucose and 0.01% (w/v) penicillin/streptomy-
cin. To incorporate heavy lysine, cells were incubated for more than
five passages and the incorporation rate was measured by a MRM meth-
od that quantifies 13C6-L-lysine peptides of actin-beta (ACTB) protein.
Incorporation was superior to 96%.
2.8. Subcellular fractionation
Nuclear, cytoplasmic and membrane fractions were obtained by
using a successive detergent extraction method as described by Adam
et al. [13] with modifications. Caov-3 (1 × 107) cells were washed
twice with PBS and lysed with 100 μl hypotonic buffer M (50 mM
HEPES pH 7.4, 10 mM NaCl, 5 mM MgCl2, 0.1 mM EDTA, 1 mM
Na3VO4, 1 mM NaF and 1 mM Na4P2O7 ∙ 10H2O) containing 5% (v/v)
protease inhibitor cocktail (product number P8340; Sigma-Aldrich).
Samples were vortexed for 20 s, passed through a 25-gauge needle
(20 times) and homogenized (D-130 tissue homogenizer, Biosystems,
São José dos Pinhais, PR, Brazil) for 1 min. At this point, equal amounts
of protein from Caov-3 cells induced to EMT/non-induced and Caov-3
cells cultured in SILAC containing heavy lysine (13C6-L-lysine) were
mixed. The nuclear pellet was obtained by centrifugation at 500 × g
for 20 min at 4 °C. Supernatant was centrifuged at 16,000 × g for
15 min at 4 °C, collected, and designated as the cytoplasmic fraction.
The pellet was treated with buffer A (25 mM 2-(N-morpholino)-
ethanesulfonic acid, 150 mM NaCl, pH 6.5) and combined with equal
volume of buffer A containing 2% Triton X-100 (v/v) with protease
and phosphatase inhibitors. After 60 min of incubation, samples were
centrifuged at 16,000 ×g for 20 min at 4 °C. The supernatant was collect-
ed and designated as the membrane fraction. Protein extraction of the
nuclear pellet was performed with lysis solution containing 8 M urea,
2% CHAPS and protease inhibitor cocktail. Three sonication cycles
were carried out for 5 min each in an ultrasonic bath (Unique, São
Paulo, SP, Brazil) with cooled water. Samples were centrifuged at
20,000 ×g for 30 min at 4 °C and supernatant was collected and stored
at −80 °C. Western Blot analysis was performed to determine the en-
richment efficiency.
2.9. In-gel digestion and LC-MS/MS analysis
50 μg of the cytoplasmic and nuclear fractions and 30 μg of the mem-
brane fraction were dissolved in 30 μl Laemmli sample buffer [14] con-
taining dithiothreitol (1 mg/mg of total protein), boiled for 5 min to
reduce disulfide bonds, and alkylated with iodoacetamide (5 mg/mg
of total protein). Samples were then separated using a 10% SDS-PAGE
(Bio-Rad, Hercules, CA, USA). Gel lanes corresponding to each fraction
were cut into 5 pieces of approximately 1.2 cm each, washed, and
digested with trypsin as described elsewhere [15]. Tryptic peptides
were successively extracted with 0.1% formic acid and 50% acetonitrile,
and then 70% acetonitrile, and dried by SpeedVac. Peptide mixtures
were dissolved in 10 μl 5% acetonitrile/0.1% formic and desalted using
ZipTip columns (Supelco, Sigma-Aldrich) according to the
manufacturer's instructions. Samples were eluted in 52.5% acetoni-
trile/water in 0.1% formic acid, dried again, and re-suspended in 15 μl
5% acetonitrile/0.1% formic acid for LC-MS analysis. Samples were ana-
lyzed by LC-MS/MS using an LTQ-Orbitrap Velos mass spectrometer
coupled with a nano-LC (EASY-nLC II, ThermoScientific). Chromatogra-
phy was performed using an in-house packed 75-μm inner diameter
(New Objective) × 25-cm long C18 column packed with Magic C18
resin, at 600 nl/min with 90 min linear gradients from 5 to 40% acetoni-
trile in 0.1% formic acid. MS/MS scans of the ten most abundant doubly
or triply charged ions in the FT-MS scan were selected for data-depen-
dent analysis in the linear ion-trap. Peptides and proteins were identi-
fied with a local installation of the Labkey Server (2013.3), using the
X!Tandem search engine (2013.2.01 release) [16], and PeptideProphet
[17] and ProteinProphet [18] algorithms for statistical validation of
data and protein grouping. MS data were searched against a human pro-
teome database (Uniprot April 2014 - 68,561 entries). Search parame-
ters for tryptic peptides included up to two miss-cleavages, mass
tolerance of 0.5 Da for fragment ions, fixed cysteine modification with
carbamidomethylation (+ 57.02146), variable methionine oxidation
(+15.99491), and variable lysine modification (+6.020129) to account
for both heavy and light SILAC labels. Only peptides with a
PeptideProphet score above 0.9 were considered for protein quantifica-
tion. The list of proteins was generated with a ProteinProphet cut-off
value of 0.9, representing an overall protein false discovery rate of ap-
proximately 1% based on the ProteinProphet estimate. Proteins were
quantitated as previously described, using the Q3 algorithm (1.22a re-
lease) to measure SILAC peak intensities [19,20].
2.10. Proteomic data analysis
The proteomic data was evaluated and correlated with several dif-
ferent bioinformatics tools. Protein-protein interactions and networks
were evaluated using STRING-DB (string-db.org) [21]. Gene ontology
annotation was obtained using PANTHER tool (phanterdb.org) [22].
Protein expression information based on immunohistochemistry (IHC)
for normal and cancer tissues was obtained from ProteinAtlas
(proteinatlas.org) [23].
2.11. Cell cycle and proliferation rate
Cell cycle analysis was assessed using the DNA-binding dye
propidium iodide (PI) (Sigma-Aldrich) and cell proliferation rate was
measured by monitoring the incorporation of 5-bromo-2′-deoxyuridine
(BrdU) (Invitrogen, Thermo Fisher Scientific), according to the
manufacturer's instructions. Both assays were analyzed by flow cytom-
etry (BD FACSCanto™, BD Biosciences).
3. Results
3.1. EGF receptor (EGFR) activation in Caov-3 cells
Since studies have reported a significant association between in-
creased EGFR expression/phosphorylation and advanced stages of
OvCa in patients, we aimed to evaluate the activation of EGFR specifical-
ly in Caov-3 ovarian cancer cells using initially a brief stimulation with
EGF. Although Caov-3 cells exhibited a basal phosphorylation of EGFR,
the stimulation with EGF was able to increase the EGFR phosphorylation
 M.L. Grassi et al. / Journal of Proteomics 151 (2017) 2–11 5

after 15 min even at the lowest concentration of 1 ng/ml, while the total
amount of EGFR remained unchanged (Fig. 1A). On the other hand,
MRM analysis demonstrated that only after 96 h of stimulation with
10 ng/ml EGF, Caov-3 cells showed a significant increase of EGFR abun-
dance (Fig. 1B). Concomitantly, the proteins Ephrins A2, A3 and B2,
which compose the largest subfamily of receptor protein-tyrosine ki-
nases (RTK), together with the proto-oncogene tyrosine-protein kinase
Src, also showed high levels of phosphorylation in response to EGF ex-
posure in Caov-3 cells, based on our RTK activity array analysis (Fig.
1C). Overall, the phosphorylation profiles of these proteins indicate
that the Caov-3 cell line is functional and responsive to the growth fac-
tor EGF.
3.2. Morphological and functional evaluation of EMT induced by EGF in
Caov-3 cells
In order to ensure EMT induction by EGF in Caov-3 ovarian cancer
cells, specific tools were applied to characterize the changes observed
during this process. Contrast-phase microscopy showed clearly mor-
phological alterations indicated by the acquisition of a spindle-shaped
and loss of cell-cell contact after EGF stimulation in Caov-3 cells (Fig.
2A). MRM analysis demonstrated that the newly acquired mesenchy-
mal morphology was supported by alterations in several proteins in-
volved in EMT. Increasing levels of Vimentin (VIM), N-cadherin
(CDH2) and Annexin A2 (ANXA2) proteins followed by the EMT-related
transcription factor SNAIL (SNAI1) were observed after 96 h of EMT in-
duction by EGF. As expected, significant reduction of E-cadherin (CDH1)
epithelial marker abundance was observed (Fig. 2B). Changes were not
detected for the SLUG (SNAI2), another EMT-related transcription
factor.
In addition, changes in the levels of Vimentin and E-cadherin pro-
teins were also evaluated by immunofluorescence microscopy to con-
firm EMT induction. In agreement with MRM analysis, fluorescence
images showed the dynamic change of Vimentin filaments extended
throughout the entire spindle-shaped cell after EGF-induced EMT,
with higher fluorescent intensity than the control cells, as well as a
weaker and discontinuous distribution of E-cadherin near the cell mem-
brane with a lesser intensity than the control cells (Fig. 2C).
3.3. Proteomic profile of EGF-induced EMT in Caov-3 cells
After proper confirmation that EGF induces molecular and cellular
changes associated with EMT in Caov-3 cells, an in-depth quantitative
proteomic analysis was performed to elucidate in more detail the coor-
dinated protein changes as well as cellular process altered during the
EMT. Initially, conventional SILAC strategy, containing dialyzed fetal bo-
vine serum, was adopted for quantitative analysis. However Caov-3 cells
grown in SILAC media had a significantly different pattern EMT when
compared to those induced to EMT in regular DMEM media. Therefore,
we designed a strategy based on mixing Caov-3 cells induced to EMT (or
the corresponding control not induced to EMT) grown in regular DMEM
media containing natural (“light”) lysine, with reference Caov-3 cells
grown in heavy SILAC media (13C6-lysine).
After mixing equal amounts (300 μg) of control or EMT-induced
Caov-3 cells with Caov-3 heavy labeled cells, samples were fractionated
by differential centrifugation to obtain subcellular extracts enriched in
cytoplasmic, nuclear or membrane proteins. Initially, the quality of sub-
cellular fractionation was confirmed by Western Blotting, using anti-
bodies to detect proteins exclusive to each subcellular compartment
(data not shown). Each subcellular fraction was further fractionated
by SDS-PAGE and gel lanes were cut into five bands, generating a total
of 30 fractions. A schematic description of our experimental workflow
is presented in Supplemental Fig. 1. Each fraction was subjected to in-
gel trypsin digestion and individually analyzed by LC-MS/MS. Around
450,000 MS/MS scans were collected, providing confident identification
for 3018 proteins represented by unique gene names with b1% FDR

Fig. 1. Profile of EGF receptor (EGFR) activation in ovarian adenocarcinoma cell line Caov-3. A) Western blotting shows the activation of EGFR (Tyr1068) in Caov-3 cells after 15 min of
stimulation with two different doses of EGF: 1 and 10 ng/ml. GAPDH protein was used as endogenous control. B) MRM analysis shows that Caov-3 cells stimulated with EGF maintain
high levels of EGFR when compared to untreated cells, after 96 h of stimulation. Bars indicate the sum of the means of each transition of two independent replicates. Statistical
significance was assessed by Student's t-test. *p b 0.05. Detailed method parameters are shown in Supplementary Table 1. C) RTK activity array analysis using the PathScan® RTK
Signaling Antibody Array kit (cell signaling), showed differential phosphorylation in important pathways in response to EGF treatment.
6 M.L. Grassi et al. / Journal of Proteomics 151 (2017) 2–11

Fig. 2. Characterization of EMT induction in Caov-3 cells by EGF. A) Morphological changes observed by contrast phase microscopy in Caov-3 cells induced to EMT by EGF 10 ng/ml after
96 h. B) EGF-induced EMT increased levels of known mesenchymal related proteins (VIM, CDH2, ANX2, SNAI1), as well as decreased CDH1, but not changed levels of the transcription
factor Slug (SNAI2), base on MRM analysis after 96 h of induction. Bars indicate the sum of the means of each transition of two independent experiments. Statistical significance was
assessed by Student's t-test. *p b 0.05. Detailed method parameters are shown in Supplementary Table 1. C) Immunofluorescence microscopy of Caov-3 cells induced to EMT shows a
higher fluorescent intensity of the dynamic change of Vimentin filaments extended throughout the cells, while E-cadherin fluorescence intensity appears weaker than the control cells.
Caov-3 CTRL: non-induced; Caov-3 EGF: induced with EGF 10 ng/ml during 96 h.

(False Discovery Rate). The complete list of protein identification is pre-
sented in Supplementary Table 2. The subsets of proteins identified in
each subcellular compartment were complementary, since proteins de-
tected in a single subcellular compartment contributed to 13% to 19%
additional identifications and only 28% of all proteins were identified
in all three subcellular compartments.
Our quantitative analysis elected 206 differentially expressed pro-
teins, based on a stringent criteria where only proteins that were ob-
served with at least two peptides with associated quantitative
information and changes N 1.5-fold in at least one subproteome were
considered for further evaluation. The full list of regulated proteins is
presented in Supplementary Table 3.
Of note, AnxA6 (Annexin A6) was downregulated in nuclear and
membrane fractions and upregulated in the cytoplasm, PSMA5 (protea-
some subunit alpha type 5) was downregulated in membrane fraction
and upregulated in cytoplasm and ITGA2 (integrin alpha 2) was upreg-
ulated only in membrane and nuclear fractions. These differences show
the important feature of analyzing multiple subcellular compartments
in parallel, potentially highlighting protein translocation processes
such those mentioned above.
3.4. Protein-protein interaction (PPI) network analysis and validation of
proteomic findings using ProteinAtlas
Initially, using STRING, we performed a PPI network analysis with
the 206 differentially expressed proteins during EMT induced by EGF.
From that set of regulated protein, we selected a subset composed of
72 proteins (Fig. 3A) for which STRING established a PPI network.
After, using PANTHER, we identified major cellular processes arising
from EMT induced by EGF in Caov-3 cells. Two distinct cellular
processes were significantly changed: 1) metabolism (63.3%), specifi-
cally protein synthesis, and 2) cellular processes (38.9%) including cell
cycle and cell-cell communication (Fig. 3B). In fact, metabolic
reprogramming is one of the most important hallmarks of cancer and
very favorable not only for tumor survival and growth but also for inva-
sion and metastasis. In addition, cell cycle regulation has been closely
linked to EMT since Snail transcription factor family members can in-
duce cell cycle arrest in G1 phase and the main cell cycle regulation
pathways seem to be dependent on cell-cell adhesion [24–29].
In order to validate if the 72 regulated proteins from our PPI network
are relevant for ovarian cancer, we relied on ProteinAtlas IHC data to
verify expression in normal ovarian and/or ovarian cancer tissue. As ob-
served in Fig. 3C, most proteins were upregulated by EMT in at least one
subcellular compartment and several of these proteins appeared highly
stained in most of ovarian cancer tumor tissues analyzed by
ProteinAtlas, such as P4HA1 (Prolyl 4-hydroxylase subunit alpha-1),
HNRNPA1 (heterogeneous nuclear ribonucleoprotein A1), RAB7A
(Ras-related protein Rab-7a), RPL24 (60S ribosomal protein L24) and
EPS15 (epidermal growth factor receptor substrate 15). Moreover, sever-
al proteins downregulated in our proteomic data are highly expressed in
normal ovarian tissues, suggesting that these proteins could potentially
represent metastasis suppressors XRCC6 (X-ray repair cross-
complementing protein 6), EWSR1 (RNA-binding protein EWS), PSMC2
(26S protease regulatory subunit 7), RAD23B (UV excision repair protein
RAD23 homolog B) and EFTUD2 (116 kDa U5 small nuclear ribonucleo-
protein component). In this context, information available in the Protein
Atlas database represented an external validation for our data and was
highly relevant for future investigation of proteins potentially related to
EMT and metastasis, since both up- and downregulation of proteins
might represent important processes for ovarian cancer progression.
 M.L. Grassi et al. / Journal of Proteomics 151 (2017) 2–11 7

Fig. 3. Proteins regulated by EGF-induced EMT in Caov-3 cells. A) From a total of 206 differentially expressed proteins, 72 were able to establish a PPI network based on STRING analysis. The
network indicates a major hub composed specially by ribosomal proteins (RPs). B) PANTHER analysis indicated that these 72 proteins present in the PPI network are involved in metabolic
(protein synthesis) and cellular processes (cell cycle and cell-cell adhesion). Dotted red circle indicates EGF growth factor added together with the 206 regulated proteins data to connect
these proteins with EMT process induction. C) The heat map shows the distribution of the 72 differentially expressed proteins among the three different subcellular compartments (left
panel). Color scale represents each protein log2 ratio from EMT-induced/control experiments. Immunohistochemical data from ovarian cancer and normal tissues obtained from The
Human Protein Atlas database (right panel) was used as an external indicator of relevance of regulated proteins identified in our study. The scale (0–100%) refers to the percentage of
immunohistochemically stained breast cancer samples and the color gradient represents the staining intensity: high (dark blue), medium (blue), low (light blue), not detected (white)
or not available.

Fig. 4. Analysis of Ras/Erk MAPK and PI3K/Akt/mTOR signaling pathways and cell proliferation during EGF-induced EMT in Caov-3 cells. A) EMT induced by EGF in Caov-3 cells activates the
Ras/Erk MAPK pathway, with high levels of Erk 1/2 (Thr202/Tyr204), and the PI3K/Akt pathway indicated by high levels of Akt (Ser473) favored by deactivation of PTEN (Ser308), the main
PI3K/Akt phosphatase inhibitor, while mTOR level did not change. GAPDH protein was used as endogenous control. B) Analysis by flow cytometry of BrdU incorporation indicates no
significant change of proliferation rate for Caov-3 cells induced to EMT by EGF. Bars indicate the mean ± standard deviation of two independent experiments. Statistical significance
was assessed by Student's t-test. ns = no significant.
8 M.L. Grassi et al. / Journal of Proteomics 151 (2017) 2–11

3.5. Induction of EMT by EGF in Caov-3 cells implicates in Akt and ERK 1/2 our results from PPI-network analysis, indicating that EGF-induced
activation without changing proliferation pattern EMT indeed affects cell cycle control.

Based on the PPI network obtained from proteomic data, the induc-
tion of EMT in Caov-3 cells by EGF promotes regulation of several ribo-
somal proteins (RPs) and proteins related to cell cycle control. To test
whether these molecular changes promoted in Caov-3 cells by EMT
would result in modulation of the cell growth and proliferation patterns,
we evaluated the activation of key signaling pathways, such as PI3K/
Akt/mTOR and ERK 1/2 MAP kinase. We also measured the cellular pro-
liferation rate by BrdU incorporation. Fig. 4A indicates that AKT activa-
tion was stronger in EMT-induced Caov-3 cells compared to control
cells, with increased levels of total and phosphorylated Akt (Ser473)
and reduced levels of PTEN, the major negative regulator of this kinase.
Independent of Akt activation, a decreasing level of mTOR was ob-
served, followed by high levels in ERK 1/2 (Thr202/Tyr204) phosphory-
lation (Fig. 4A). Although EMT induction resulted in evident
phosphorylation of specific kinases involved in cell proliferation, BrdU
incorporation unexpectedly demonstrated that Caov-3 cells prolifera-
tion rate did not change after EGF-induced EMT (Fig. 4B).
3.6. EGF-induced EMT promotes cell cycle arrest in Caov-3 cells
Since Caov-3 ovarian cancer cells showed no significant change in
proliferation rate after EGF-induced EMT, we performed Western Blot-
ting analysis for proteins relevant to cell cycle regulation. We also per-
formed cell cycle analysis using flow cytometry. As demonstrated in
Fig. 5, EGF-induced EMT had a measurable impact over Caov-3 cell
cycle. The remarkable increase in p21 protein levels indicates accumula-
tion of cells in G1 phase and consequent blockage of S phase, while p53
protein was not detected (Fig. 5A and B). Concomitantly, we verified a
reduction in the phosphorylation levels of cyclin D1, retinoblastoma
protein (Rb) and cyclin-dependent kinase 1 (CDK1)/cdc2 protein in re-
sponse to EMT induced by EGF (Fig. 5C). Altogether this data supports
4. Discussion
The role of EMT in cancer invasion and metastasis is strongly sup-
ported by several cellular models. However, monitoring this process
during cancer progression is still difficult since cells undergoing EMT
share many molecular and morphological characteristics with the sur-
rounding stromal cells. Here we show an effective in vitro induction of
EMT by EGF treatment in ovarian cancer cells. The concentration of
EGF used in this study (10 ng/ml) is considerably below what is found
in general literature for EMT induction, usually N 50 ng/ml, but greater
than the know regular concentration of EGF in circulation (40pM–
240 pg/ml) [30]. In fact, EGF stimulation in ovarian surface epithelial
cells has demonstrated to be involved in matrix remodeling mediated
by ERK and ILK/GSK-3β pathways [11] and combined targeting of
endothelin and EGF receptors has shown to enhance antitumor activity
in ovarian cancers [31].
In our model, EGF-stimulated Caov-3 cells showed high EGFR activ-
ity followed by high phosphorylation levels of Ephrins A2, A3 and B2
(Fig. 1C). These RTKs have played crucial roles in cell morphology, adhe-
sion, migration and invasion by modifying the actin cytoskeleton reor-
ganization and regulating the activities of integrins and cell-cell
adhesion molecules. Several studies correlating the role of EPH recep-
tors in EMT process have emerged. In hepatocellular carcinoma, for ex-
ample, the EPHA receptor favored EMT through the activation of HGF
signaling pathway, since HGF induced EMT-like morphological changes
as well as upregulation of SNAIL and N-cadherin [32]. Notably, EPHA2
overexpression was able to upregulate EMT molecular markers N-
cadherin and Snail, and downregulate E-cadherin in gastric cancer
cells [33]. Hence, these evidences suggest that elevated expression of
EPH/Ephrin may be correlated with increased invasiveness in different
types of aggressive tumors, including ovarian cancer.

Fig. 5. EGF-induced EMT in Caov-3 cells promote cell cycle arrest through regulation of checkpoint proteins. A) Cell cycle analysis of Caov-3 cells by PI incorporation shows that EMT
induced by EGF blocks G1/S transition phase increasing the G1 cell population. Bars indicate the mean ± standard deviation of three independent experiments. Statistical significance
was assessed by two way ANOVA. ***p b 0.0001. B) Regulation of G1 cell-cycle arrest in Caov-3 cells was mediated by high levels of the cyclin-dependent kinase inhibitor p21Cip1/Waf1,
an inhibitor of the G1-to-S phase, independent of p53 expression. C) Decreasing or non-alteration in phosphorylation levels of G1/S checkpoint proteins cyclin D1 (Thr286) and Rb
(Ser807/811) followed by a decrease in G2/M checkpoint protein CDK1/cdc2 supports the effect of EGF-induced EMT in Caov-3 cells. GAPDH protein was used as endogenous control.
 M.L. Grassi et al. / Journal of Proteomics 151 (2017) 2–11
Morphological and molecular changes, such as the acquisition of a
spindle shaped morphology and increased levels of EMT-related pro-
teins (Vimentin, N-cadherin, Annexin A2, Snail), support the efficient
EMT induction with EGF (Fig. 2). These molecular changes observed in
our EMT model are relevant for a metastatic context, since in high-
grade serous ovarian carcinomas, the low N-cadherin mRNA expression
has been associated with worse survival and the transcription factor
SNAIL is commonly overexpressed in later stage ovarian tumors (III
and IV) compared [34–36].
Based on this valid model for EMT, proteomic and bioinformatics
analysis of the three subcellular fractions from EGF-induced Caov-3
cells revealed several differentially expressed proteins. The organization
of such data into functional classes highlighted groups of proteins previ-
ously linked to cellular metabolism, tumor progression and EMT (Fig. 3;
Supplementary Table 3). In fact, metabolic reprogramming is one of the
most important hallmarks of cancer and very beneficial not only for
tumor survival and growth but also invasion and metastasis.
Several proteins differentially expressed in our Caov-3 model have
been already associated with EMT and metastasis. AnxA6 depletion in
the invasive BT-549 breast cancer cells results in strong inhibition of
cell motility and invasiveness, contributing to cancer progression [37,
38]. A recent study showed that ITGA2, when modulated by the miR-
128, suppressed cell migration, invasion and EMT in MG-63 osteosarco-
ma cells [39]. In addition, NT5E (5′-nucleotidase or CD73), PRMT1 (pro-
tein arginine methyl transferase 1) and Ostf1b (osmotic stress
transcription factor 1b) proteins have been considered as new regula-
tors of EMT by favoring the tumor progression in patients with gallblad-
der cancer, enhancing TWIST1 methylation in non-small cell lung
cancer and promoting cell migration and cytoskeleton rearrangement
in HEK293 cells, respectively [40–42].
It is also important to highlight that most of the proteins observed as
regulated in our study are detected as moderate or intense in tumor
samples, according ProteinAtlas external data validation. More interest-
ing, some of the proteins that are downregulated were strongly stained
in normal tissue. Based on these observations, one could speculate that
proteins such as XRCC6, EWSRS1, PSMC2, RAD23B and EFTUD2 can act
as tumor suppressors. Indeed, loss of the canonical non-homologous
end-joining (cNHEJ) composed by XRCC6 and RB1 has been recently
implicated in genomic instability and cancer progression [43]. EWS-fu-
sion-proteins (EFPS) play a role in different tumorigenic processes and
is directly associated with mesenchymal stem cell phenotype [44].
Thus, this set of downregulated proteins seems extremely interesting
in terms of targets for tumor control.
Recently, alterations in translational control and cancer etiology
have outlined a new perspective in cancer research since it has been
known that genetic alterations in several components of the transla-
tional apparatus (translation factors, ribosomes, RPs) and cancer are in-
terconnected, favoring increased cancer susceptibility [45]. As we show
in Fig. 3B, the great number of regulated RPs creates a dense network in
our data set, suggesting a link between the EMT process induced by EGF
and ribosome biogenesis/protein translational regulation in ovarian
cancer. Following a traditional view, most of the studies have reported
that dysregulation of protein synthesis during cancer development
merely results in cell growth and proliferation. However, protein dys-
regulation can be one of the mechanisms that favor cancer develop-
ment, since important oncogenic signaling pathways are intrinsically
related to the translational machinery at almost every stage of cancer
initiation [46].
Akt and ERK1/2 phosphorylation status indicates the activation of
PI3K/Akt/ERK1/2 pathway after EGF-induced EMT in Caov-3 cells.
PI3K/Akt pathway controls ribosome biogenesis by modulating the
transcription of Pol I and III [47], and the loss of PTEN causes upregula-
tion of this pathway (Fig. 4). Moreover, the Ras/Erk MAPK pathway, as
well PI3K/Akt, regulates several downstream effector pathways, includ-
ing mTORC1, by Erk- and p90RSK-dependent inhibitory phosphoryla-
tion of tuberous sclerosis 2 (TSC2) [48]. Ras pathway also stimulates
9
translation through eIF4E phosphorylation by Erk-activated MNK1 and
MNK2, favoring translation and tumor development [49]. In this con-
text, although these signaling pathways appear activated by EGF during
the induction of EMT in Caov-3 cells, it seems that decreasing mTOR
levels either is sufficient to maintain the cell proliferation pattern un-
changed or is controlling autophagy, once mTOR inhibition promotes
autophagy, which is required for the management of metabolic alter-
ations and can limit tumor progression [50]. Together these alterations
suggest a remarkable connection between EMT process and translation-
al control in ovarian cancer since these events share the same classical
signaling pathways.
In addition to the role of ribosomal stabilization and maintenance of
the efficiency and accuracy of translation, the RPs have also demonstrat-
ed secondary functions, or extra-ribosomal functions, related to other
cellular processes including proliferation, cell cycle and gene expression
[51–53]. Overexpression of RP L41 can arrest the cell cycle at the G1
phase in human lung carcinoma H1299 cells through an increase in
p21Cip1/Waf1 levels, while constitutive overexpression of RP L7 in Jurkat
T-lymphoma cells leads to G1 arrest via the modulation of cell cycle pro-
gression related proteins [54,55]. In our data, EGF-induced EMT was
able to impair the transition from G1 to S phase by maintaining high
levels of p21Cip1/Waf1, which block the G1/S transition, independent of
p53 modulation (Fig. 5). Clearly, the loss or inhibition of p53 protein
prevents cellular senescence, by controlling cell cycle checkpoints, and
apoptosis. In fact, relevant cell cycle proteins such as cyclin D1, Rb and
CDK1/cdc2 had a significant decrease after EGF-induced EMT, and its ex-
pression has also been predictive of poor clinical outcome in patients
with ovarian cancer [56]. About half of all human cancers, the p53
tumor suppressor is either lost or mutated, resulting in the expres-
sion of a transcriptionally inactive mutant protein. Curiously, studies
showed that a point mutation in TP53 gene (resulting in a chain ter-
mination signal likely to truncate the p53 peptide at amino acid 135)
and substantial copy-number changes in Caov-3 cell line are key
characteristics related to high-grade serous ovarian cancer, although
it also can reflect the difficulty for detecting the p53 protein using
antibodies [57,58].
EGF has the ability to stimulate differentiation and proliferation of
several cell types by binding to its transmembrane receptor EGFR [59].
Surprisingly, the presence of EGF as an EMT inducer did not affect the
proliferation rate of Caov-3 cells during the acquisition of the mesen-
chymal phenotype and suggests a new response profile very similar to
TGF-β anti-proliferative effect discussed in different studies [60–62].
Furthermore, the conversion to a mesenchymal phenotype, which
denotes a dramatic actin cytoskeleton reorganization, may be in-
compatible with a highly proliferative state, as we previously report-
ed in a recent study where MCF7 cells exhibited reduced capacity of
cell proliferation after EMT induction by SNAIL-overexpressing,
supporting the idea that cell proliferation is not essential for tumor
malignancy [26,29].
5. Conclusions
Overall, our study demonstrated that EGF, a potent growth factor
that is tightly connected to cancer development and progression, in-
duced EMT in Caov-3 cells as well as important alterations in metabolic
process (protein synthesis) and cell cycle control. Although these
changes result from complex reprograming of protein expression, our
data contribute to support the implication of the EMT highlighted path-
ways that are likely linked to in vivo cancer metastasis and cancer stem
cell generation. Further clinical studies that correlate RP expression,
translational control and tumor progression are therefore worthwhile
to uncover new therapeutic targets for cancer treatment and improved
patient outcome.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.jprot.2016.06.009.
10 M.L. Grassi et al. / Journal of Proteomics 151 (2017) 2–11
Acknowledgments
We thank São Paulo Research Foundation (FAPESP, grant EMU-
2004/09448-5), the Centre of Flow Cytometry of Faculty of Pharmaceu-
tical Sciences of Ribeirão Preto – University of Sao Paulo, for providing
access to the equipment flow cytometer BD FACSCanto I and the Central
de Análises Químicas Instrumentais –Instituto de Química de São Carlos
– University of São Paulo, for providing access to the LTQ-Orbitrap Velos
Instrument. This research was supported by FAPESP (Young Scientist
Grant – Proc. No. 2011/0947-1), CNPq (Proc. No. 308561/2014-7),
Center for Cell Based Therapy - CTC-CEPID (Proc. FAPESP 2013/08135-
2) and CISBi-NAP. M.L.G, C.S.P. and C.H.T. received fellowships from
FAPESP Proc. Nos. 2013/08755-0, 2012/09682-4 and 2013/07675-3 re-
spectively. A.P. received fellowship from CAPES. C.S.P., G.P.P. and
V.M.F. received fellowships from CNPq, Proc. Nos. 154138/2014-2,
130142/2015-8 and 308561/2014-7 respectively.
References
[1] D. Hanahan, R.A. Weinberg, Hallmarks of cancer: the next generation, Cell 144
(2011) 646–674.
[2] R. Kalluri, R.A. Weinberg, The basics of epithelial-mesenchymal transition, J. Clin. In-
vest. 119 (2009) 1420–1428.
[3] E.D. Hay, An overview of epithelio-mesenchymal transformation, Acta Anat. (Basel)
154 (1995) 8–20.
[4] J.M. Lee, S. Dedhar, R. Kalluri, E.W. Thompson, The epithelial-mesenchymal transi-
tion: new insights in signaling, development, and disease, J. Cell Biol. 172 (2006)
973–981.
[5] A. Puisieux, T. Brabletz, J. Caramel, Oncogenic roles of EMT-inducing transcription
factors, Nat. Cell Biol. 16 (2014) 488–494.
[6] M.A. Huber, N. Kraut, H. Beug, Molecular requirements for epithelial-mesenchymal
transition during tumor progression, Curr. Opin. Cell Biol. 17 (2005) 548–558.
[7] L. Larue, A. Bellacosa, Epithelial-mesenchymal transition in development and can-
cer: role of phosphatidylinositol 3′ kinase/AKT pathways, Oncogene 24 (2005)
7443–7454.
[8] H. Niikura, H. Sasano, S. Sato, A. Yajima, Expression of epidermal growth factor-re-
lated proteins and epidermal growth factor receptor in common epithelial ovarian
tumors, Int. J. Gynecol. Pathol. 16 (1997) 60–68.
[9] J.M. Bartlett, S.P. Langdon, B.J. Simpson, M. Stewart, D. Katsaros, P. Sismondi, et al.,
The prognostic value of epidermal growth factor receptor mRNA expression in pri-
mary ovarian cancer, Br. J. Cancer 73 (1996) 301–306.
[10] P. Yue, X. Zhang, D. Paladino, B. Sengupta, S. Ahmad, R.W. Holloway, et al., Hyperac-
tive EGF receptor, Jaks and Stat3 signaling promote enhanced colony-forming abil-
ity, motility and migration of cisplatin-resistant ovarian cancer cells, Oncogene 31
(2012) 2309–2322.
[11] N. Ahmed, S. Maines-Bandiera, M.A. Quinn, W.G. Unger, S. Dedhar, N. Auersperg, Molec-
ular pathways regulating EGF-induced epithelio-mesenchymal transition in human
ovarian surface epithelium, Am. J. Physiol. Cell Physiol. 290 (2006) C1532–C1542.
[12] B. MacLean, D.M. Tomazela, N. Shulman, M. Chambers, G.L. Finney, B. Frewen, et al.,
Skyline: an open source document editor for creating and analyzing targeted prote-
omics experiments, Bioinformatics 26 (2010) 966–968.
[13] R.M. Adam, W. Yang, D. Di Vizio, N.K. Mukhopadhyay, H. Steen, Rapid preparation of
nuclei-depleted detergent-resistant membrane fractions suitable for proteomics
analysis, BMC Cell Biol. 9 (2008) 30.
[14] U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of
bacteriophage T4, Nature 227 (1970) 680–685.
[15] C.H. Thomé, G.A. dos Santos, G.A. Ferreira, P.S. Scheucher, C. Izumi, A.M. Leopoldino,
et al., Linker for activation of T-cell family member2 (LAT2) a lipid raft adaptor pro-
tein for AKT signaling, is an early mediator of alkylphospholipid anti-leukemic activ-
ity, Mol. Cell. Proteomics 11 (2012) 1898–1912.
[16] B. MacLean, J.K. Eng, R.C. Beavis, M. McIntosh, General framework for developing
and evaluating database scoring algorithms using the TANDEM search engine, Bio-
informatics 22 (2006) 2830–2832.
[17] A.I. Nesvizhskii, A. Keller, E. Kolker, R. Aebersold, A statistical model for identifying
proteins by tandem mass spectrometry, Anal. Chem. 75 (2003) 4646–4658.
[18] A. Keller, A.I. Nesvizhskii, E. Kolker, R. Aebersold, Empirical statistical model to esti-
mate the accuracy of peptide identifications made by MS/MS and database search,
Anal. Chem. 74 (2002) 5383–5392.
[19] V. Faca, M. Coram, D. Phanstiel, V. Glukhova, Q. Zhang, M. Fitzgibbon, et al., Quanti-
tative analysis of acrylamide labeled serum proteins by LC-MS/MS, J. Proteome Res.
5 (2006) 2009–2018.
[20] V.M. Faca, K.S. Song, H. Wang, Q. Zhang, A.L. Krasnoselsky, L.F. Newcomb, et al., A
mouse to human search for plasma proteome changes associated with pancreatic
tumor development, PLoS Med. 5 (2008) e123.
[21] D. Szklarczyk, A. Franceschini, S. Wyder, K. Forslund, D. Heller, J. Huerta-Cepas, et al.,
STRING v10: protein-protein interaction networks, integrated over the tree of life,
Nucleic Acids Res. 43 (2015) D447–D452.
[22] H. Mi, A. Muruganujan, P.D. Thomas, PANTHER in 2013: modeling the evolution of
gene function, and other gene attributes, in the context of phylogenetic trees,
Nucleic Acids Res. 41 (2013) D377–D386.
[23] M. Uhlén, E. Björling, C. Agaton, C.A. Szigyarto, B. Amini, E. Andersen, et al., A human
protein atlas for normal and cancer tissues based on antibody proteomics, Mol. Cell.
Proteomics 4 (2005) 1920–1932.
[24] R.A. Cairns, I.S. Harris, T.W. Mak, Regulation of cancer cell metabolism, Nat. Rev. Can-
cer 11 (2011) 85–95.
[25] T. Han, D. Kang, D. Ji, X. Wang, W. Zhan, M. Fu, et al., How does cancer cell
metabolism affect tumor migration and invasion? Cell Adhes. Migr. 7 (2013)
395–403.
[26] S. Vega, A.V. Morales, O.H. Ocaña, F. Valdés, I. Fabregat, M.A. Nieto, Snail blocks the
cell cycle and confers resistance to cell death, Genes Dev. 18 (2004) 1131–1143.
[27] F.E. Turner, S. Broad, F.L. Khanim, A. Jeanes, S. Talma, S. Hughes, et al., Slug regulates
integrin expression and cell proliferation in human epidermal keratinocytes, J. Biol.
Chem. 281 (2006) 21321–21331.
[28] J.L. Walker, A.K. Fournier, R.K. Assoian, Regulation of growth factor signaling and cell
cycle progression by cell adhesion and adhesion-dependent changes in cellular ten-
sion, Cytokine Growth Factor Rev. 16 (2005) 395–405.
[29] C.S. Palma, M.L. Grassi, C.H. Thomé, G.A. Ferreira, D. Albuquerque, M.T. Pinto, et al.,
Proteomic analysis of epithelial to mesenchymal transition reveals crosstalk be-
tween SNAIL and HDAC1 in breast cancer cells, Mol. Cell. Proteomics (2016).
[30] B. Marquèze-Pouey, S. Mailfert, V. Rouger, J.M. Goaillard, D. Marguet, Physiological
epidermal growth factor concentrations activate high affinity receptors to elicit cal-
cium oscillations, PLoS One 9 (2014) e106803.
[31] L. Rosanò, V. Di Castro, F. Spinella, G. Tortora, M.R. Nicotra, P.G. Natali, et al., Com-
bined targeting of endothelin A receptor and epidermal growth factor receptor in
ovarian cancer shows enhanced antitumor activity, Cancer Res. 67 (2007)
6351–6359.
[32] T. Nagai, T. Arao, K. Furuta, K. Sakai, K. Kudo, H. Kaneda, et al., Sorafenib inhibits the
hepatocyte growth factor-mediated epithelial mesenchymal transition in hepatocel-
lular carcinoma, Mol. Cancer Ther. 10 (2011) 169–177.
[33] J. Huang, D. Xiao, G. Li, J. Ma, P. Chen, W. Yuan, et al., EphA2 promotes epithelial-
mesenchymal transition through the Wnt/β-catenin pathway in gastric cancer
cells, Oncogene 33 (2014) 2737–2747.
[34] B. Davidson, C.G. Tropé, R. Reich, Epithelial-mesenchymal transition in ovarian car-
cinoma, Front. Oncol. 2 (2012) 33.
[35] L. Quattrocchi, A.R. Green, S. Martin, L. Durrant, S. Deen, The cadherin switch in ovar-
ian high-grade serous carcinoma is associated with disease progression, Virchows
Arch. 459 (2011) 21–29.
[36] S. Elloul, M.B. Elstrand, J.M. Nesland, C.G. Tropé, G. Kvalheim, I. Goldberg, et al., Snail,
Slug, and Smad-interacting protein 1 as novel parameters of disease aggressiveness
in metastatic ovarian and breast carcinoma, Cancer 103 (2005) 1631–1643.
[37] C. Enrich, C. Rentero, S.V. de Muga, M. Reverter, V. Mulay, P. Wood, et al., Annexin
A6-linking Ca(2+) signaling with cholesterol transport, Biochim. Biophys. Acta
2011 (1813) 935–947.
[38] A.M. Sakwe, R. Koumangoye, B. Guillory, J. Ochieng, Annexin A6 contributes to the
invasiveness of breast carcinoma cells by influencing the organization and localiza-
tion of functional focal adhesions, Exp. Cell Res. 317 (2011) 823–837.
[39] X. Liu, Z. Liang, K. Gao, H. Li, G. Zhao, S. Wang, et al., MicroRNA-128 inhibits EMT of
human osteosarcoma cells by directly targeting integrin α2, Tumour Biol. (2015).
[40] L. Xiong, Y. Wen, X. Miao, Z. Yang, NT5E and FcGBP as key regulators of TGF-1-in-
duced epithelial-mesenchymal transition (EMT) are associated with tumor progres-
sion and survival of patients with gallbladder cancer, Cell Tissue Res. 355 (2014)
365–374.
[41] S. Avasarala, M. Van Scoyk, M.K. Karuppusamy Rathinam, S. Zerayesus, X. Zhao, W.
Zhang, et al., PRMT1 is a novel regulator of epithelial-mesenchymal-transition in
non-small cell lung cancer, J. Biol. Chem. 290 (2015) 13479–13489.
[42] K.P. Lai, A.Y. Law, M.C. Lau, Y. Takei, W.K. Tse, C.K. Wong, Osmotic stress transcrip-
tion factor 1b (Ostf1b) promotes migration properties with the modulation of epi-
thelial mesenchymal transition (EMT) phenotype in human embryonic kidney
cell, Int. J. Biochem. Cell Biol. 45 (2013) 1921–1926.
[43] R. Cook, G. Zoumpoulidou, M.T. Luczynski, S. Rieger, J. Moquet, V.J. Spanswick, et al.,
Direct involvement of retinoblastoma family proteins in DNA repair by non-homol-
ogous end-joining, Cell Rep. 10 (2015) 2006–2018.
[44] F. Tirode, K. Laud-Duval, A. Prieur, B. Delorme, P. Charbord, O. Delattre, Mesenchy-
mal stem cell features of Ewing tumors, Cancer Cell 11 (2007) 421–429.
[45] D. Silvera, S.C. Formenti, R.J. Schneider, Translational control in cancer, Nat. Rev.
Cancer 10 (2010) 254–266.
[46] J.P. Le Quesne, K.A. Spriggs, M. Bushell, A.E. Willis, Dysregulation of protein synthesis
and disease, J. Pathol. 220 (2010) 140–151.
[47] M.J. James, J.C. Zomerdijk, Phosphatidylinositol 3-kinase and mTOR signaling path-
ways regulate RNA polymerase I transcription in response to IGF-1 and nutrients,
J. Biol. Chem. 279 (2004) 8911–8918.
[48] X.M. Ma, J. Blenis, Molecular mechanisms of mTOR-mediated translational control,
Nat. Rev. Mol. Cell Biol. 10 (2009) 307–318.
[49] S. Pyronnet, H. Imataka, A.C. Gingras, R. Fukunaga, T. Hunter, N. Sonenberg, Human
eukaryotic translation initiation factor 4G (eIF4G) recruits mnk1 to phosphorylate
eIF4E, EMBO J. 18 (1999) 270–279.
[50] R. Mathew, V. Karantza-Wadsworth, E. White, Role of autophagy in cancer, Nat. Rev.
Cancer 7 (2007) 961–967.
[51] H. Naora, Involvement of ribosomal proteins in regulating cell growth and apopto-
sis: translational modulation or recruitment for extraribosomal activity? Immunol.
Cell Biol. 77 (1999) 197–205.
[52] M. Thapa, A. Bommakanti, M. Shamsuzzaman, B. Gregory, L. Samsel, J.M. Zengel,
et al., Repressed synthesis of ribosomal proteins generates protein-specific cell
cycle and morphological phenotypes, Mol. Biol. Cell 24 (2013) 3620–3633.
[53] M.S. Lindström, Emerging functions of ribosomal proteins in gene-specific transcrip-
tion and translation, Biochem. Biophys. Res. Commun. 379 (2009) 167–170.
 M.L. Grassi et al. / Journal of Proteomics 151 (2017) 2–11
[54] M.J. Kim, Y.A. Yoo, H.J. Kim, S. Kang, Y.G. Kim, J.S. Kim, et al., Mitochondrial ribosomal
protein L41 mediates serum starvation-induced cell-cycle arrest through an in-
crease of p21(WAF1/CIP1), Biochem. Biophys. Res. Commun. 338 (2005)
1179–1184.
[55] F. Neumann, U. Krawinkel, Constitutive expression of human ribosomal protein L7
arrests the cell cycle in G1 and induces apoptosis in Jurkat T-lymphoma cells, Exp.
Cell Res. 230 (1997) 252–261.
[56] A. Bali, P.M. O'Brien, L.S. Edwards, R.L. Sutherland, N.F. Hacker, S.M. Henshall, Cyclin
D1, p53, and p21Waf1/Cip1 expression is predictive of poor clinical outcome in se-
rous epithelial ovarian cancer, Clin. Cancer Res. 10 (2004) 5168–5177.
[57] S. Domcke, R. Sinha, D.A. Levine, C. Sander, N. Schultz, Evaluating cell lines as tumour
models by comparison of genomic profiles, Nat. Commun. 4 (2013) 2126.
[58] Y. Yaginuma, H. Westphal, Abnormal structure and expression of the p53 gene in
human ovarian carcinoma cell lines, Cancer Res. 52 (1992) 4196–4199.
11
[59] Y. Yarden, M.X. Sliwkowski, Untangling the ErbB signalling network, Nat. Rev. Mol.
Cell Biol. 2 (2001) 127–137.
[60] L. Levy, C.S. Hill, Smad4 dependency defines two classes of transforming growth fac-
tor {beta} (TGF-{beta}) target genes and distinguishes TGF-{beta}-induced epitheli-
al-mesenchymal transition from its antiproliferative and migratory responses, Mol.
Cell. Biol. 25 (2005) 8108–8125.
[61] J. Seoane, Escaping from the TGFbeta anti-proliferative control, Carcinogenesis 27
(2006) 2148–2156.
[62] A. Klotzsche-von Ameln, A. Muschter, S. Mamlouk, J. Kalucka, I. Prade, K. Franke,
et al., Inhibition of HIF prolyl hydroxylase-2 blocks tumor growth in mice through
the antiproliferative activity of TGFβ, Cancer Res. 71 (2011) 3306–3316.