Content uploaded by Michiaki Unno
Author content
All content in this area was uploaded by Michiaki Unno on Apr 30, 2019
Content may be subject to copyright.
ORIGINAL ARTICLE
Mass spectrometry-based proteomic analysis of formalin-fixed paraffin-
embedded extrahepatic cholangiocarcinoma
Shimpei Maeda · Takanori Morikawa ·
Tatsuyuki Takadate · Takashi Suzuki · Takashi Minowa ·
Nobutaka Hanagata · Tohru Onogawa · Fuyuhiko Motoi ·
Toshihide Nishimura · Michiaki Unno
© 2015 Japanese Society of Hepato-Biliary-Pancreatic Surgery
Abstract
Background Extrahepatic cholangiocarcinoma is very diffi-
cult to diagnose at an early stage, and has a poor prognosis.
Novel markers for diagnosis and optimal treatment selection
are needed. However, there has been very limited data on
the proteome profile of extrahepatic cholangiocarcinoma.
This study was designed to unravel the proteome profile of
this disease and to identify overexpressed proteins using mass
spectrometry-based proteomic approaches.
Methods We analyzed a discovery set of formalin-fixed
paraffin-embedded tissues of 14 extrahepatic cholangio-
carcinomas using shotgun mass spectrometry, and compared
proteome profiles with those of seven controls. Then, selected
candidates were verified by quantitative analysis using sched-
uled selected reaction monitoring-based mass spectrometry.
Furthermore, immunohistochemical staining used a validation
set of 165 cases.
Results In total, 1,992 proteins were identified and 136 pro-
teins were overexpressed. Verification of 58 selected proteins
by quantitative analysis revealed 11 overexpressed proteins.
Immunohistochemical validation for 10 proteins showed posi-
tive rates of S100P (84%), CEAM5 (75%), MUC5A (62%),
OLFM4 (60%), OAT (42%), CAD17 (41%), FABPL (38%),
AOFA (30%), K1C20 (25%) and CPSM (22%) in extrahepatic
cholangiocarcinomas, which were rarely positive in controls.
Conclusions We identified 10 proteins associated with ex-
trahepatic cholangiocarcinoma using proteomic approaches.
These proteins are potential targets for future diagnostic bio-
markers and therapy.
Keywords Biomarker · Extrahepatic cholangiocarcinoma ·
Mass spectrometry · Proteomics · Scheduled selected reaction
monitoring
Introduction
Cholangiocarcinomas account for 3% of all gastrointestinal
cancers [1] and are classified according to their anatomic loca-
tion as intrahepatic and extrahepatic cholangiocarcinomas
(EHCC). EHCC is very difficult to diagnose at an early stage
and has a poor prognosis, which has improved only margin-
ally over the past 30 years [2]. Although complete surgical re-
section is the only opportunity for cure, the 5-year survival
rate after complete resection is 39.1% for perihilar EHCC,
and 44.0% for distal EHCC [3]. One of the factors responsible
for these poor outcomes of EHCC is limitations of diagnostic
modalities. Novel biomarkers for early diagnosis and optimal
treatment selection are needed; however, there have been very
limited data on the proteome profile of EHCC. Therefore, we
made an attempt to unravel the proteome profile of EHCC to
identify proteins overexpressed in EHCC compared with non-
cancerous bile duct tissues.
Mass spectrometry (MS) is reportedly valuable in both pre-
clinical and clinical research [4] as well as for biomarker
S. Maeda · T. Morikawa · T. Takadate · T. Onogawa · F. Motoi · M. Unno (✉)
Department of Surgery, Tohoku University Graduate School of Medicine, 1-1
Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-8574, Japan
e-mail: m_unno@surg1.med.tohoku.ac.jp
S. Maeda
Department of Surgery, South Miyagi Medical Center, Miyagi, Japan
T. Suzuki
Department of Pathology and Histotechnology, Tohoku University
Graduate School of Medicine, Sendai, Japan
T. Minowa · N. Hanagata
Nanotechnology Innovation Station, National Institute for Materials
Science, Tsukuba, Japan
T. Nishimura
Department of Surgery I, Tokyo Medical University, Tokyo, Japan
J Hepatobiliary Pancreat Sci (2015) ••:••–••
DOI: 10.1002/jhbp.262
discovery [5]. Shotgun proteomics is a method of identifying
proteins in complex mixtures using liquid chromatography
(LC) and MS to provide global proteome profiles [6]. Mean-
while, targeted proteomics based on selected reaction moni-
toring (SRM) is an appropriate method for accurate
identification and quantitation of proteins of interest [7].
Identification of scarce biomarkers in serum remains chal-
lenging because of the complexity and wide dynamic range
characterizing such samples [8]. Cancer-specific proteins exist
at high concentrations in tumor tissues compared with other
samples such as blood and bile juice. Thus, cancerous tissue
itself is an important source for biomarker discovery.
Formalin-fixed paraffin-embedded (FFPE) tissues have been
extensively collected and stored in hospitals for various pe-
riods of time. Those are readily available, pathologically
well-defined, and include all stages of cancer, even for rare
diseases. Newly developed technology has made it possible
to efficiently extract proteins from FFPE tissues, thereby
allowing proteomic analysis [9]. Although successful MS
analysis of FFPE tissue has recently been reported [10, 11],
there are no reports of proteomic analysis fromarchived FFPE
samples of EHCC.
We conducted a large-scale proteomic study to identify
novel proteins overexpressed in EHCC using both shotgun
and targeted proteomics. We further validated the candidate
proteins by immunohistochemical analysis.
Materials and methods
Tissues
We retrospectively retrieved EHCC samples from patients un-
dergoing resection between 1998 and2008 at Tohoku Univer-
sity Hospital. Those given neoadjuvant therapy were
excluded. Intrahepatic cholangiocarcinoma, carcinoma of
the gallbladder and carcinoma of the papilla of Vater cases
were also excluded. Non-cancerous bile duct tissues were ob-
tained from pancreatic cancer patients with pancreaticoduo-
denectomy. In total, 186 FFPE tissues, 165 EHCCs and 21
non-cancerous bile ducts, were examined. Before analysis,
hematoxylin and eosin stained sections from each sample
were evaluated by a pathologist. Clinicopathologic features
are shown in Table 1. According to the Union for Interna-
tional Cancer Control (UICC) 7th edition, the numbers of
stage I, II, III and IV EHCC tissues were 33, 64, 24 and 44,
respectively. For MS analyses, a discovery set of 21 samples
from early EHCC (stage I, n=7), advanced EHCC (stage II,
III and IV, n=7), and non-cancerous bile duct tissues
(n=7) were used. The remaining 165 samples including 151
EHCCs and 14 non-cancerous bile ducts served as a validation
set. Representative slide showing the largest diameter of each
carcinoma was used for both MS and immunohistochemical
analyses. Cells of invasive area were dissected and
immunohistochemically evaluated in the case of invasive carci-
noma. The study design and composition of the discovery and
validation sets are shown in Figure S1.
Ethics statement
Informed consent was obtained from individual patients. This
study was approved by the Tohoku University Ethics Com-
mittee, and conducted according to the principles expressed
in the Declaration of Helsinki.
Laser Micro Dissection and protein extraction
Cancerous lesions and non-cancerous bile duct epithelium
were identified on serial, hematoxylin and eosin-stained
sections. For MS analysis, 10-μm sections were attached to
DIRECTOR slides (Expression Pathology, Gaithersburg, MD,
USA), de-paraffinized three times with xylene for 5 min,
rehydrated with graded ethanol solutions and distilled water,
and then stained with hematoxylin [12]. Stained, uncovered
slides were air dried and about 30,000 cells (8 mm
2
)were
collected into the cap of a 0.2-mL polymerase chain reaction
(PCR) tube using Leica LMD6000 (Leica Microsystems
GmbH, Wetzler, Germany). 1-4 sections of carcinoma and
8-20 sections of non-cancerous bile duct were needed to obtain
30,000 cells. Peptides were extracted using a Liquid Tissue MS
Protein Prep kit (Expression Pathology) [9] according to the
manufacturer’s instructions.
Table 1 Clinicopathologic features
Factor No. patients
EHCC (n=165)
Median age (range) 67 (15–83) years
Sex Male: female 112: 53
Location Perihilar: distal 91: 74
T 1: 2: 3: 4 13: 68: 51: 33
N 0: 1 98: 67
M 0: 1 150: 15
Stage I: II: III: IV 33: 64: 24: 44
Histological type Papillary 10
Well differentiated 28
Moderately differentiated 108
Poorly differentiated 18
Adenosquamous 1
T2 includes T2a and T2b perihilar EHCC and T2 distal EHCC. Stage I
includes stage I perihilar EHCC and stage IA and IB distal EHCC, the
same applies to stage II, III andIV. The median ageof 14 patients (4 males
and 10 females) from whom non-cancerous bile ducts were obtained was
65 years (range 34–83). EHCC, extrahepatic cholangiocarcinoma.
J Hepatobiliary Pancreat Sci (2015) ••:••–••2
Exploratory shotgun analysis
LC-tandem mass spectrometry (MS/MS)
Peptide-mixture samples processed from FFPE tissues were
used for LC-MS/MS using a Finnigan LXQ linear ion-trap
mass spectrometer (Thermo Fisher, San Jose, CA, USA) [12].
A capillary reverse phase LC-MS/MS system (ZAPLOUS Sys-
tem; AMR, Tokyo, Japan), comprising a Paradigm MS4
(Michrom BioResources, Auburn, CA, USA), an HTC PAL
autosampler (CTC Analytics, Zwingen, Switzerland) and
Finnigan LXQ linear ion-trap mass spectrometer, was equipped
with an ADVANCE nanospray ionization source (Michrom
BioResources).
Data analysis and protein identification
Mascot software (version 2.2.03, Matrix Science, London,
UK) was used for a database search against Homo sapiens en-
tries in the Swiss-Prot 55.6 database (20,009 entries). Peptide
and fragment mass tolerances were 2.0 Da and 1.0 Da,
respectively, and trypsin specificity was applied with a maxi-
mum of two missed cleavages. Methionine oxidation and N-
formylation including formyl (K), formyl (R) and formyl
(N-terminus) were allowed as variable modifications. A P-
value less than 0.05 was considered to be significant in protein
identification. Reported results were obtained from triplicate
LC-MS runs for each sample.
Semi-quantitative comparison using spectral counting
To compare protein expressions across all tissue samples, we
used the spectral counting method. The number of peptide
spectra with high confidence (Mascot ion score, P<0.05)
served as the spectral count value. Fold changes in the
expressed proteins on a base-2 logarithmic scale were calcu-
lated using the protein ratio from spectral counting (Rsc)
[13]. Relative abundances of identified proteins were also ob-
tained by applying the normalized spectral abundance factor
(NSAF) [14]. Candidate proteins differing between groups
were chosen so that their Rsc would satisfy the ≥1 criterion,
corresponding to a fold change ≥2, and with statistical signif-
icance at P<0.05 by the G-test [15].
Targeted verification analysis by SRM mass spectrometry
Selected reaction monitoring-based MS analysis was con-
ducted using a discovery set. Proteins for SRM verification
were mainly selected from among candidates identified by
shotgun analysis and spectral counting. Furthermore, several
proteins that were not overexpressed in EHCC in our shotgun
analysis but were previously reported to be potential bio-
markers were added to the SRM list for verification [16–19].
All sequences were confirmed by the BLAST searches (Na-
tional Center for Biotechnology Information) and compared
with the Swiss-Prot human database.
An LC-MS/MS system was composed of a Paradigm MS4
(Michrom BioResources) connected to a 4000 QTRAPhybrid
system (AB Sciex, Foster City, CA, USA) operating in posi-
tive ion mode [7]. For all SRM studies, quadruples were oper-
ated under conditions of unit/unit resolution, and the collision
energy (CE) was determined using the equation: CE = 0.044 ×
m/z + 6 for doubly-charged precursor ions. The scheduled
SRM (sSRM) mode was used in this study, with the sSRM de-
tection window set at 180 sec. The peptide AGFAGDDAPR
(m/z 488.7) is a doubly-charged actin, beta (ACTB) peptide
and its specific SRM transition to the singly charged fragment
(m/z 630.3) served as the internal standard [20]. This internal
standard is referred to as the in-sample internal standard (ISIS)
since ACTB is a housekeeping protein [21]. Peak areas of
each transition were normalized using the equation: Normal-
ized peak area = peak area × (500,000/peak area of 488.7/
630.3). The averaged values of early EHCC and advanced
EHCC based on triplicate runs were each compared to those
of non-cancerous bile duct tissues, and an expression differ-
ence of at least two-fold was defined as overexpression.
Immunohistochemistry
A validation set of 165 samples was used. Sections (4-μm
thick) after de-paraffinization with xylene were rehydrated with
a graded ethanol series and distilled water. Protein S100-P
(S100P) (HPA019502, Sigma, St. Louis, MO, USA),
carcinoembryonic antigen-related cell adhesion molecule 5
(CEAM5) (HPA019758, Sigma), mucin-5AC (MUC5A)
(OBT1746, AbD Serotec, Oxford, UK), olfactomedin-4
(OLFM4) (ab96280, Abcam, Cambridge, MA, USA),
cadherin-17 (CAD17) (HPA023616, Sigma), keratin, type I cy-
toskeletal 20 (K1C20) (HPA027236, Sigma), and carbamoyl-
phosphate synthase (ammonia) (CPSM) (ab54586, Abcam)
immunostaining was achieved by heating slides in an autoclave
at 120°C for 5 min in citrate acid buffer (10 mM citric acid, pH
6.0). Similarly, antigen retrieval for fatty acid-binding protein,
liver (FABPL) (ab82157, Abcam) was performed in a micro-
wave oven for 15 min in a citric acid buffer. No antigen re-
trieval was carried out for ornithine aminotransferase (OAT)
(HPA040098, Sigma) or amine oxidase [flavin-containing] A
(AOFA) (NBP1-19796, Novus Biologicals, Littleton, CO,
USA). The dilutions of primary antibodies were as follows:
S100P, 1: 3000; CEAM5, 1: 40; MUC5A, 1: 50; OLFM4, 1:
100; CAD17, 1: 1500; K1C20, 1: 100; CPSM, 1: 100; FABPL,
J Hepatobiliary Pancreat Sci (2015) ••:••–•• 3
1: 300; OAT, 1: 100; AOFA, 1: 200. The sections were incu-
bated overnight at 4°C with one of the primary antibodies. Af-
ter blocking of endogenous peroxidase by methanol containing
0.3% hydrogen peroxidase, labeled antigens were detected
with an EnVision
+
kit (Dako, Glostrup, Denmark) and visual-
ized using 3,3’-diaminobenzidine tetrahydrochloride as a chro-
mogen. Sections were counterstained with hematoxylin.
Appropriate positive and negative tissue controls were used
throughout, in part with reference to the Human Protein Atlas
(URL: http://www.proteinatlas.org/).
After completely reviewing all slides of immunostained
sections for each sample, three of the authors (S. M., T.
Morikawa and T. S.) classified cases into two groups as de-
scribed previously [11, 22–27]: those in which ≥10% of cells
were positive for S100P, CEAM5, OLFM4, OAT, CAD17,
AOFA, CPSM and K1C20 constituted the positive group,
while the negative group was comprised of those in which
<10% of cells were positive. For MUC5A and FABPL, those
with strong and moderate immunoreactivity were categorized
as the positive group, while those with absent orweak staining
constituted the negative group.
Fisher’s exact test was used to assess the significance of dif-
ferences among staining patterns. Differences with P<0.05
were considered significant. Analyses were performed with
JMP software version 9.0 (SAS Institute, Cary, NC, USA).
Results
Proteome profiles identified by shotgun proteomics and
semi-quantitative comparison
A discovery set of 21 samples was used to identify proteins
showing different expressions in cancerous and non-cancerous
tissues. We identified 1,266 proteins in early EHCC, 1,143 in
advanced EHCC, and 1,095 in non-cancerous bile ducts. In to-
tal, 1,992 proteins were identified. The identified proteins were
compared semi-quantitatively using spectral counting. For pro-
teins identified in non-cancerous bile ducts and EHCC, Figure 1
shows a plot of each Rsc value against the corresponding pro-
tein (X-axis) in increasing order from left to right. A positive
value indicates greater expression in EHCC than non-
cancerous bile ducts. Proteins with Rsc ≥1 and a significant
difference by G-test were regarded as candidates for character-
ization of EHCC. A total of 136 of the 1,992 proteins identified
had P<0.05, indicating statistically significant overexpres-
sion. Meanwhile, ACTB, used as housekeeping protein, was
commonly expressed in cancerous and non-cancerous cells
with minimum variation.
Quantitative verification by SRM-based targeted proteomics
Selected reaction monitoring measurements were carried out
for the discovery set to verify the spectral counting results.
Preliminary analysis of the project control (mixtures of equal
aliquots of all patient samples) was conducted to select well
detectable SRM transitions and to confirm the retention time
of each peptide. Finally, an SRM assay, comprising 56 pro-
teins (102 peptides, 400 transitions), with sufficient sensitiv-
ity was developed. The 56 proteins included 48 proteins
found to be overexpressed by our shotgun analysis, seven
previously reported as potential biomarkers, and ACTB
(ISIS).
Selected reaction monitoring quantitative analysis revealed
that 11 proteins, S100P, CEAM5, MUC5A, OLFM4, OAT,
CAD17, FABPL, AOFA, K1C20, CPSM and HMCS2, were
Fig. 1 Protein ratio from spectral counting (Rsc) and normalized spectral abundancefactor (NSAF) values calculated for the proteins identified. Protein
expressions are compared between extrahepatic cholangiocarcinoma (EHCC) and non-cancerous bile ducts. Proteins significantly overexpressed in
EHCC are near the right side of the X-axis. ACTB is located near the center of the X-axis
J Hepatobiliary Pancreat Sci (2015) ••:••–••4
overexpressed by at least two-fold and seemed to be poten-
tially useful for detection of EHCC (Table S1). Figure 2 is a
scatter plot for the normalized peak area. The seven potential
biomarkers previously reported were not overexpressed in our
SRM analysis. These results are consistent with those of spec-
tral counting.
Validation by immunohistochemical analysis
Ten overexpressed proteins were further validated by im-
munohistochemical analysis using a validation set of 165
samples. HMCS2 was excluded because no appropriate an-
tibody was commercially available. We confirmed the spe-
cific expressions of 10 proteins by immunohistochemistry
(Fig. 3). Table 2 shows the positive rate for each protein
in the non-cancerous bile ducts and EHCC. All non-
cancerous bile duct samples were classified into the negative
group in evaluation of each protein except for one sample
with positive immunoreactivity for OAT. However, the
positive rates of nine proteins, including S100P (84%),
CEAM5 (75%), MUC5A (62%), OLFM4 (60%), OAT
(42%), CAD17 (41%), FABPL (38%), AOFA (30%)
and K1C20 (25%), differed significantly between non-
cancerous bile ducts and EHCC (P<0.05). Only in
CPSM, the difference did not reach statistical signifi-
cance (P= 0.07), although there was a tendency to have
a higher positive rate (22%).
Discussion
We identified overexpressed proteins in EHCC using shotgun
proteomics with spectral counting and targeted proteomics,
and then validated our findings immunohistochemically.
Shotgun and targeted proteomics revealed 11 proteins to be
overexpressed in EHCC as compared with non-cancerous bile
ducts. Ten of these 11 proteins were validated by immunohis-
tochemistry. The significant overexpressions of nine proteins
were also confirmed using a validation set, although the differ-
ence in CPSM did not reach statistical significance (P= 0.07).
The newly-identified markers of EHCC in our study are
OLFM4, OAT, CAD17, FABPL, AOFA and CPSM.
OLFM4, a member of the olfactomedin domain-containing
protein family, is an anti-apoptotic factor promoting tumor
growth. OLFM4 promotes pancreatic cancer cell proliferation
by favoring transition from S to G2/M phase [28], and facili-
tates cell adhesion [29]. OLFM4 expression is related to
differentiation and progression of gastric cancer [30].
OLFM4 was expressed in 60% of EHCC samples, while
Fig. 2 Scatter plot of the normalized peak area by selected reaction monitoring (SRM)-based quantitative analysis. Proteins overexpressed by at least
two-fold in early and/or advanced extrahepatic cholangiocarcinoma (EHCC) are shown
J Hepatobiliary Pancreat Sci (2015) ••:••–•• 5
Fig. 3 Representative immunohistochemical staining results for the indicated proteins in extrahepatic cholangiocarcinoma (EHCC) and non-cancerous
bile ducts. Positive staining is shown in brown. Scale bars represent 100 μm
Table 2 Immunohistochemical findings of the validation set
Non-cancerous
bile duct (n=14)
Stage I
(n=26)
Stage II
(n=62)
Stage III
(n=21)
Stage IV
(n=42)
Antibody EHCC (n=151) P-value
S100P 0 (0%) 127 (84%) <0.01 21 (81%) 52 (84%) 16 (76%) 38 (90%)
CEAM5 0 (0%) 113 (75%) <0.01 19 (73%) 47 (76%) 15 (71%) 32 (76%)
MUC5A 0 (0%) 94 (62%) <0.01 10 (38%) 40 (65%) 13 (62%) 31 (74%)
OLFM4 0 (0%) 91 (60%) <0.01 14 (54%) 34 (55%) 10 (48%) 33 (79%)
OAT 1 (7%) 63 (42%) <0.01 11 (42%) 18 (29%) 11 (52%) 23 (55%)
CAD17 0 (0%) 62 (41%) <0.01 15 (58%) 22 (35%) 7 (33%) 18 (43%)
FABPL 0 (0%) 57 (38%) <0.01 9 (35%) 15 (24%) 10 (48%) 23 (55%)
AOFA 0 (0%) 45 (30%) 0.01 10 (38%) 15 (24%) 8 (38%) 12 (29%)
K1C20 0 (0%) 38 (25%) 0.04 6 (23%) 11 (18%) 7 (33%) 14 (33%)
CPSM 0 (0%) 33 (22%) 0.07 6 (23%) 13 (21%) 1 (5%) 13 (31%)
EHCC, extrahepatic cholangiocarcinoma.
J Hepatobiliary Pancreat Sci (2015) ••:••–••6
non-cancerous bile ducts showed no immunoreactivity for
this protein.
OAT is a key enzyme in the pathway converting arginine
and ornithine into the major excitatory and inhibitory neuro-
transmitters glutamate and gamma-aminobutyric acid. Al-
though Miyasaka et al. demonstrated OAT overexpression
in hepatocellular carcinoma using the suppression subtractive
hybridization technique, little is known about the role of OAT
in carcinogenesis [31].
CAD17, a cadherin superfamily member, is an important
cell adhesion molecule and plays major roles in organ devel-
opment and maintenance of tissue integrity. A low CAD17
level is associated with a poor prognosis in gastric cancer
and intrahepatic cholangiocarcinoma, although the relation-
ship betweenCAD17 expression and prognosis in gastric can-
cer is controversial [24, 32].
FABPL is a family of small, highly conserved, cytoplas-
mic proteins that bind free fatty acids, and is considered to
play roles in fatty acid uptake, transport and metabolism.
However, precise roles of FABPL have yet to be elucidated.
Several reports suggest associations of FABPL with colon
and pancreatic cancers [33, 34].
AOFA is a mitochondrial enzyme that degrades amine
neurotransmitters, including dopamine, norepinephrine
and serotonin. Little is known about AOFA functions
in cancer. AOFA is reportedly overexpressed in poorly
as compared to well differentiated prostate cancer,
suggesting roles in progression and aggressiveness of
tumors [35].
CPSM, expressed mainly in intestinal epithelial and liver
cells, is a mitochondrial enzyme catalyzing the synthesis of
carbamoyl phosphate from ammonia and bicarbonate, and is
important for excess urea removal from cells. Previous reports
have shown CPSM overexpression in gastric cancer [36], and
under-expression in human hepatocellular carcinoma [37].
S100P, CEAM5, MUC5A and K1C20 are reportedly asso-
ciated with EHCC [38–41]. The fact that we identified previ-
ously reported biomarkers in an unbiased fashion suggests our
workflow to be extremely useful for biomarker research.
CEAM5, which is identical to clinically-used CEA, was
reported to be expressed in 79% of cholangiocarcinomas by
immunohistochemistry, similar to our findings (75%) [39].
Recently, Hamada et al. demonstrated S100P to be a poten-
tially novel biomarker of EHCC, and noted that detecting
S100P expression levels in brushing cytology had diagnostic
value [38].
Few reports have described MS-based proteomics analysis
on EHCC using tissue [18], bile [42] and serum samples [19].
Compared with those studies, the outstanding feature of this
investigation is that we analyzed EHCC tissue samples only,
because the etiopathogenesis of EHCC and intrahepatic chol-
angiocarcinoma may differ [43]. Additionally, we obtained
samples using Laser Micro Dissection (LMD) technology to
isolate cancerous cells of interest from abundant stroma con-
taining inflammatory cells and fibroblasts. Kawase et al. con-
ducted MS-based proteomic analysis using six paired
cancerous, including intrahepatic cholangiocarcinoma and
EHCC, and non-cancerous bile duct cases [18]. The
overexpressed proteins in their study, e.g. actinin-1 and
actinin-4, were not confirmed by SRM analysis in our sam-
ples. Factors for this different result may include homogeneity
of samples and LMD use, which presumably contribute to
more precise analysis of protein expression.
Discoveries of novel biomarkers are based on identifica-
tion of proteins with expressions that differ between disease
and control samples. Global shotgun proteomics has advan-
tages in terms of the number of proteins identified. With this
approach, the proteomes of complex mixtures can be analyzed
in a completely unbiased fashion with broad proteome cove-
rage, thereby increasing chances to discover novel biomarkers.
Then a rational approach to their prioritization before large
scale validation is needed, because comparisons of global
proteome profiles yield hundreds of candidate biomarkers.
Verifications of hundreds of candidates by affinity-based
methods, the broadly used Western blotor ELISA approaches,
are impractical, because development of reagents of suitable
specificity and affinity to support accurate detection and
quantitation of target proteins remains expensive and time
consuming. Furthermore, such methods are also hindered
by marked limitations in abilities to detect multiple proteins
in the same sample. Recently, quantification assays based
on SRM MS have been extensively investigated for protein
verification purposes. This technique provides two signifi-
cant advantages, that is, biomarker candidates can be
assessed simultaneously at high speed with good quantita-
tive accuracy for verification. We have conducted
MS-based verification using the leading-edge sSRM
method [44]. Based on retention times, the sSRM method
decreases the number of concurrent SRM transitions moni-
tored at any one time-point, offering improved reproducibility
and signal-to-noise ratio. Herein, we detected and quantified
56 proteins and 400 SRM transitions simultaneously in com-
plex mixtures. The impact of MS technologies on biomarker
discovery and clinical practice will be more important in near
future.
These 10 proteins may improve diagnostic capability and
have clinical utility. Because we conducted this proteomics
study using tumor tissues in consideration of the wide dy-
namic range, this study lacks the validation of these proteins
using serum and bile juice. However, detections of OLFM4,
FABPL, S100P, CEAM5 and MUC5A in serum have been
reported [40, 45–47]. In the near future, however, it might
be possible to diagnose EHCC by measuring these proteins
simultaneously using SRM-based MS.
J Hepatobiliary Pancreat Sci (2015) ••:••–•• 7
In conclusion, we identified several unique proteins, newly
discovered to be associated with EHCC, using MS-based pro-
teomics approaches with archived FFPE tissues. These pro-
teins are potential targets for future diagnostic biomarkers
and anticipated to facilitate unraveling the molecular events
that underlie this lethal disease. Many of these proteins are
poorly-documented and further investigation about their roles
in EHCC is needed.
Acknowledgments We thank Emiko Shibuya and Keiko Inabe (Tohoku
University) for technical assistance. We thank Brent Bell (Tohoku
University) for English editing.
Conflict of interest None declared.
Author contributions Study design: S.M., T. Morikawa, T.O. and M.U.
Acquisition of data: S.M., T.T., T.S., T. Minowa, N.H. and F.M. Analysis
and interpretation: S.M., T.O. and T.N. Manuscript drafted by: S.M. and
T. Morikawa. Revision: S.M., T.T., T.S., T. Minowa, N.H., T.O., F.M.,
T.N. and M.U. Statistical advice: T. Minowa, N.H. and T.N.
References
1. Khan SA, Thomas HC, Davidson BR, Taylor-Robinson SD. Chol-
angiocarcinoma. Lancet. 2005;366:1303–14.
2. Shaib Y, El-Serag HB. The epidemiology of cholangiocarcinoma.
Semin Liver Dis. 2004;24:115–25.
3. Ishihara S, Miyakawa S, Takada T, Takasaki K, Nimura Y, Tanaka M,
et al. Status of surgical treatment of biliary tract cancer. Dig Surg.
2007;24:131–6.
4. Pandey A, Mann M. Proteomics to study genes and genomes.
Nature. 2000;405:837–46.
5. Petricoin EF, Ardekani AM, Hitt BA, Levine PJ, Fusaro VA,
Steinberg SM, et al. Use of proteomic patterns in serum to identify
ovarian cancer. Lancet. 2002;359:572–7.
6. Washburn MP, Wolters D, Yates JR III. Large-scale analysis of the
yeast proteome by multidimensional protein identification technol-
ogy. Nat Biotechnol. 2001;19:242–7.
7. Anderson L, Hunter CL. Quantitative mass spectrometric multiple
reaction monitoring assays for major plasma proteins. Mol Cell
Proteomics. 2006;5:573–88.
8. Anderson NL, Anderson NG. The human plasma proteome: history,
character, and diagnostic prospects. Mol Cell Proteomics.
2002;1:845–67.
9. Prieto DA, Hood BL, Darfler MM, Guiel TG, Lucas DA, Conrads
TP, et al. Liquid Tissue: proteomic profiling of formalin-fixed
tissues. Biotechniques. 2005;38:32–5.
10. Hood BL, Darfler MM, Guiel TG, Furusato B, Lucas DA, Ringeisen
BR, et al. Proteomic analysis of formalin-fixed prostate cancer
tissue. Mol Cell Proteomics. 2005;4:1741–53.
11. Takadate T, Onogawa T, Fukuda T, Motoi F, Suzuki T, Fujii K, et al.
Novel prognostic protein markers of resectable pancreatic cancer
identified by coupled shotgun and targeted proteomics using
formalin-fixed paraffin-embedded tissues. Int J Cancer.
2013;132:1368–82.
12. Kawamura T, Nomura M, Tojo H, Fujii K, Hamasaki H, Mikami S,
et al. Proteomic analysis of laser-microdissected paraffin-embedded
tissues: (1) Stage-related protein candidates upon non-metastatic
lung adenocarcinoma. J Proteomics. 2010;73:1089–99.
13. Old WM, Meyer-Arendt K, Aveline-Wolf L, Pierce KG, Mendoza A,
Sevinsky JR, et al. Comparison of label-free methods for quantifying
human proteins by shotgun proteomics. Mol Cell Proteomics.
2005;4:1487–502.
14. Zybailov B, Coleman MK, Florens L, Washburn MP. Correlation of
relative abundance ratios derived from peptide ion chromatograms
and spectrum counting for quantitative proteomic analysis using
stable isotope labeling. Anal Chem. 2005;77:6218–24.
15. Sokal RR, Rohlf FJ. Biometry: the principles and practice of statis-
tics in biological research, 3rd edn. New York: W. H. Freeman
and Company; 1995.
16. Alvaro D. Serum and bile biomarkers for cholangiocarcinoma. Curr
Opin Gastroenterol. 2009;25:279–84.
17. Bonney GK, Craven RA, Prasad R, Melcher AF, Selby PJ, Banks RE.
Circulating markers of biliary malignancy: opportunities in
proteomics? Lancet Oncol. 2008;9:149–58.
18. Kawase H, Fujii K, Miyamoto M, Kubota KC, Hirano S, Kondo S,
et al. Differential LC-MS-based proteomics of surgical human
cholangiocarcinoma tissues. J Proteome Res. 2009;8:4092–103.
19. Wang X, Dai S, Zhang Z, Liu L, Wang J, Xiao X, et al. Characteri-
zation of apolipoprotein A-I as a potential biomarker for cholangio-
carcinoma. Eur J Cancer Care. 2009;18:625–35.
20. Nishimura T, Nomura M, Tojo H, Hamasaki H, Fukuda T, Fujii K,
et al. Proteomic analysis of laser-microdissected paraffin-embedded
tissues: (2) MRM assay for stage-related proteins upon non-
metastatic lung adenocarcinoma. J Proteomics. 2010;73:1100–10.
21. Ferguson RE, Carroll HP, Harris A, Maher ER, Selby PJ, Banks RE.
Housekeeping proteins: a preliminary study illustrating some limita-
tions as useful references in protein expression studies. Proteomics.
2005;5:566–71.
22. Aishima S, Fujita N, Mano Y, Kubo Y, Tanaka Y, Taketomi A, et al.
Different roles of S100P overexpression in intrahepatic cholangio-
carcinoma: carcinogenesis of perihilar type and aggressive behavior
of peripheral type. Am J Surg Pathol. 2011;35:590–8.
23. Nakanishi Y, Zen Y, Kawakami H, Kubota K, Itoh T, Hirano S, et al.
Extrahepatic bile duct carcinoma with extensive intraepithelial
spread: a clinicopathological study of 21 cases. Mod Pathol.
2008;21:807–16.
24. Takamura M, Yamagiwa S, Wakai T, Tamura Y, Kamimura H, Kato
T, et al. Loss of liver-intestine cadherin in human intrahepatic chol-
angiocarcinoma promotes angiogenesis by up-regulating metal-
responsive transcription factor-1 and placental growth factor. Int J
Oncol. 2010;36:245–54.
25. Hodorova I, Rybarova S, Vecanova J, Solar P, Domorakova I,
Adamkov M, et al. Comparison of expression pattern of monoamine
oxidase A with histopathologic subtypes and tumour grade of renal
cell carcinoma. Med Sci Monit. 2012;18:BR482–6.
26. Shimonishi T, Miyazaki K, Nakanuma Y. Cytokeratin profile relates
to histological subtypes and intrahepatic location of intrahepatic
cholangiocarcinoma and primary sites of metastatic adenocarcinoma
of liver. Histopathology. 2000;37:55–63.
27. Hashimoto T, Kusakabe T, Watanabe K, Sugino T, Fukuda T,
Nashimoto A, et al. Liver-type fatty acid-binding protein is highly
expressed in intestinal metaplasia and in a subset of carcinomas of
the stomach without association with the fatty acid synthase status
in the carcinoma. Pathobiology. 2004;71:115–22.
28. Kobayashi D, Koshida S, Moriai R, Tsuji N, Watanabe N.
Olfactomedin 4 promotes S-phase transition in proliferation of pan-
creatic cancer cells. Cancer Sci. 2007;98:334–40.
29. Liu W, Chen L, Zhu J, Rodgers GP. The glycoprotein hGC-1 binds
to cadherin and lectins. Exp Cell Res. 2006;312:1785–97.
30. Liu W, Zhu J, Cao L, Rodgers GP. Expression of hGC-1 is corre-
lated with differentiation of gastric carcinoma. Histopathology.
2007;51:157–65.
J Hepatobiliary Pancreat Sci (2015) ••:••–••8
31. Miyasaka Y, Enomoto N, Nagayama K, Izumi N, Marumo F,
Watanabe M, et al. Analysis of differentially expressed genes in hu-
man hepatocellular carcinoma using suppression subtractive hybrid-
ization. Br J Cancer. 2001;85:228–34.
32. Lee HJ, Nam KT, Park HS, Kim MA, Lafleur BJ, Aburatani H, et al.
Gene expression profiling of metaplastic lineages identifies CDH17
as a prognostic marker in early stage gastric cancer. Gastroenterol-
ogy. 2010;139:213–25. e3.
33. Lawrie LC, Dundas SR, Curran S, Murray GI. Liver fatty acid
binding protein expression in colorectal neoplasia. Br J Cancer.
2004;90:1955–60.
34. Sharaf RN, Butte AJ, Montgomery KD, Pai R, Dudley JT, Pasricha
PJ. Computational prediction and experimental validation associat-
ing FABP-1 and pancreatic adenocarcinoma with diabetes. BMC
Gastroenterol. 2011;11:5.
35. True L, Coleman I, Hawley S, Huang CY, Gifford D, Coleman R,
et al. A molecular correlate to the Gleason grading system for pros-
tate adenocarcinoma. Proc Natl Acad Sci U S A. 2006;103:10991–6.
36. Liu TH, Li DC, Gu CF, Ye SF. Carbamyl phosphate synthetase I. A
novel marker for gastric carcinoma. Chin Med J (Engl).
1989;102:630–8.
37. Kinoshita M, Miyata M. Underexpression of mRNA in human hepato-
cellular carcinoma focusing on eight loci. Hepatology. 2002;36:433–8.
38. Hamada S, Satoh K, Hirota M, Kanno A, Ishida K, Umino J, et al.
Calcium-binding protein S100P is a novel diagnostic marker of
cholangiocarcinoma. Cancer Sci. 2011;102:150–6.
39. Minato H, Nakanuma Y, Terada T. Expression of blood group-related
antigens in cholangiocarcinoma in relation to non-neoplastic bile ducts.
Histopathology. 1996;28:411–9.
40. Boonla C, Wongkham S, Sheehan JK, Wongkham C, Bhudhisawasdi
V, Tepsiri N, et al. Prognostic value of serum MUC5AC mucin in
patients with cholangiocarcinoma. Cancer. 2003;98:1438–43.
41. Rullier A, Le Bail B, Fawaz R, Blanc JF, Saric J, Bioulac-Sage P.
Cytokeratin 7 and 20 expression in cholangiocarcinomas varies
along the biliary tract but still differs from that in colorectal
carcinoma metastasis. Am J Surg Pathol. 2000;24:870–6.
42. Lankisch TO, Metzger J, Negm AA, Vobetakuhl K, Schiffer E,
Siwy J, et al. Bile proteomic profiles differentiate cholangiocarci-
noma from primary sclerosing cholangitis and choledocholithiasis.
Hepatology. 2011;53:875–84.
43. Welzel TM, Graubard BI, El-Serag HB, Shaib YH, Hsing AW,
Davila JA, et al. Risk factors for intrahepatic and extrahepatic
cholangiocarcinoma in the United States: a population-based case-
control study. Clin Gastroenterol Hepatol. 2007;5:1221–8.
44. Fillatre Y, Rondeau D, Jadas-Hecart A, Communal PY. Advantages
of the scheduled selected reaction monitoring algorithm in liquid
chromatography/electrospray ionization tandem mass spectrometry
multi-residue analysis of 242 pesticides: a comparative approach
with classical selected reaction monitoring mode. Rapid Commun
Mass Spectrom. 2010;24:2453–61.
45. Oue N, Sentani K, Noguchi T, Ohara S, Sakamoto N, Hayashi T,
et al. Serum olfactomedin 4 (GW112, hGC-1) in combination with
Reg IV is a highly sensitive biomarker for gastric cancer patients.
Int J Cancer. 2009;125:2383–92.
46. Lachmann RA, Werchan S, Schachtrup C, Haitsma JJ, Spener F,
Lachmann B. Liver-type fatty acid binding protein in serum and
broncho-alveolar lavage in a model of acute respiratory failure be-
cause of surfactant depletion--a possible marker for lung damage?
Clin Physiol Funct Imaging. 2006;26:371–5.
47. Wang Q, Zhang YN, Lin GL, Qiu HZ, Wu B, Wu HY, et al. S100P,
a potential novel prognostic marker in colorectal cancer. Oncol Rep.
2012;28:303–10.
Supporting information
Additional Supporting Information may be found in the online ver-
sion of this article at the publisher’s web-site:
Table S1 Proteins to which the scheduled SRM MS analysis was ap-
plied.
Fig. S1 Study design for identification and validation. Shotgun
proteomics with semi-quantitative spectral counting was conducted
to identify proteins with expression profiles differing between EHCC
and non-cancerous bile ducts. Selected candidates were verified by
quantitative analysis using scheduled SRM-based targeted proteomics.
The resulting proteins were then validated by immunohistochemical
analysis. EHCC, extrahepatic cholangiocarcinoma; LC-MS/MS,
liquid chromatography-tandem mass spectrometry; SRM, selected
reaction monitoring.
J Hepatobiliary Pancreat Sci (2015) ••:••–•• 9
