Background
Gallbladder cancer (GBC) is a prevalent malignancy of the biliary tract and is the fifth common cancer of the gastrointestinal tract worldwide [
1]. In majority of the cases, it manifests at an advanced and unresectable stage [
1,
2]. Early detection is incidental, with complete surgical resection of the gallbladder being the only available curative option. The prognosis is dismal with a five-year survival rate of 32 % for lesions confined to the gallbladder mucosa and a one year survival rate of 10 % for advanced stages [
2]. To date, various markers including carbohydrate antigen 19–9 (CA19-9) and carcinoembryonic antigen (CEA) have been explored in the diagnosis of GBC. However, these markers lack specificity and sensitivity [
3]. Targeted therapy for GBC is limited with bevacizumab which is a vascular endothelial growth factor (VEGF) inhibitor [
4]. Apart from bevacizumab, potential therapeutic targets such as estrogen receptor [
5], hedgehog signaling [
6] and mTOR inhibitors [
7] are pending clinical validation. This highlights an immediate need for identification of novel therapeutic targets to improve treatment options and disease outcome.
Mass spectrometry-based proteomic analysis in tandem with isobaric tags for relative and absolute quantitation (iTRAQ) labeling has been employed for the identification of potential biomarkers in several cancers. We have used similar approaches in the past to identify potential biomarkers in esophageal squamous cell carcinoma [
8], hepatocellular carcinoma [
9] and head and neck squamous cell carcinoma [
10]. Similar proteomic strategies have been employed by other groups to identify potential biomarkers in GBC using bile, serum and cell line-based models [
11‐
14]. However, limited effort has been made to identify potential therapeutic targets in GBC. In this study, we used high-resolution mass spectrometry coupled with iTRAQ-based labeling approach to identify proteins which can serve as potential diagnostic markers and/or therapeutic targets. Using a panel of GBC cell lines, we identified a total of 3,653 proteins of which 654 were found to be overexpressed and 387 were downregulated in invasive GBC cell lines as compared to the non-invasive GBC cell line. Amongst these, macrophage migration inhibitory factor (MIF) was found to be overexpressed in two of the invasive GBC cell lines.
MIF is a pro-inflammatory cytokine which plays a key role in innate and adaptive immunity and is associated with inflammatory conditions including cancer. It is secreted by a variety of cells including immune and epithelial cells [
15]. MIF has been reported to be overexpressed in multiple cancers, including gastric adenocarcinoma [
16], head and neck squamous cell carcinoma [
17], esophageal squamous cell carcinoma [
18], colorectal [
19], pancreatic [
20], ovarian [
21], and prostate [
22] cancers. Knockdown of
MIF in a murine ovarian cancer cell line, ID8 has been shown to decrease tumor growth and increase the survival in tumor transplanted mice [
21]. Similar results were demonstrated in mice grafted with colorectal carcinoma transplants, administered with anti-MIF therapeutics, using either MIF-antibodies or the MIF antagonist (S, R)-3-(4-hydroxyphenyl)-4,5-dihydro-5-isoxazole acetic acid methyl ester (ISO-1) [
19]. Pharmacological inhibition of MIF using the MIF irreversible inhibitor, 4-iodo-6-phenylpyrimidine (4-IPP) has shown a decrease in tumor aggressiveness in head and neck squamous cell carcinomas [
17] and lung adenocarcinomas [
23]. The role of MIF in tumorigenesis has been characterized in other cancers however its function in GBC is yet to be established. In this study, we have assessed the role of MIF as a potential therapeutic target in GBC.
Methods
Cell culture
The GBC cell lines, OCUG-1 and NOZ were obtained from Health Science Research Resources Bank, Osaka, Japan. TGBC2TKB, TGBC24TKB and G-415 were purchased from RIKEN Bio Resource Center, Ibaraki, Japan. SNU-308 was obtained from Korean Cell Line Bank, Seoul, Korea. GB-d1 was authenticated by short tandem repeat analysis. The properties and culture conditions of the GBC cell lines, TGBC2TKB, SNU-308, G-415, TGBC24TKB, NOZ, OCUG-1 and GB-d1 are provided in Additional file
1. All cell lines were maintained in humidified incubator with 5 % CO
2 at 37 °C.
Protein extraction and iTRAQ labeling
Each cell line was grown to ~80 % confluence, serum starved for 8 h and lysed in 0.5 % SDS-containing buffer. Protein concentration was measured using the BCA method [
24]. Equal amount of protein from each cell line was then split into two and treated as technical replicates. Peptides from each sample were differentially labeled using iTRAQ 8-plex reagent (iTRAQ Reagents Multiplex kit, Applied Biosystems/MDS Sciex, Foster City, CA) as described earlier [
25]. Briefly, 100 μg of proteins, in replicate, was treated with 2 μl of reducing agent (TCEP, tris (2-carboxyethyl) phosphine) at 60 °C for 1 h and alkylated with 1 μl of cysteine blocking reagent, MMTS (methyl methanethiosulfate) for 10 min at room temperature. Protein samples were digested using sequencing grade trypsin (Promega, San Luis Obispo, CA) at a 1:20 enzyme to protein ratio for 12 h at 37 °C. Peptides from each cell line were labeled with 8 iTRAQ reagents in 60 μl of isopropanol at room temperature as follows – TGBC24TKB (reporter ion m/z 113 and 114), OCUG-1 (reporter ion m/z 115 and 116), NOZ (reporter ion m/z 117 and 118) and GB-d1 (reporter ion m/z 119 and 121). After 2 h, the reaction was quenched by adding 100 μl of water to each sample. The samples were then pooled and vacuum dried.
Strong cation exchange chromatography
The iTRAQ labeled peptides were fractionated using strong cation exchange chromatography as previously described [
8]. Briefly, the pooled iTRAQ-labeled sample was reconstituted with solvent A (10 mM KH
2PO
4, 25 % acetonitrile, pH 2.8). The pH of the sample was adjusted to 2.8 using ortho-phosphoric acid. The peptides were loaded onto a PolySULFOETHYL A column (PolyLC, Columbia, MD) (5 μm, 200 Å, 200x 2.1 mm) using Agilent 1260 Infinity series binary HPLC system (Agilent Technologies, Santa Clara, CA). Peptides were loaded at a flow rate of 250 μl/min and washed for 8 min with solvent A. A 35 min gradient from 0 % to 60 % solvent B (350 mM KCl in solvent A, pH 2.8) was used for fractionation. The peptides were detected at a wavelength of 214 nm using a variant wavelength detector module of HPLC system. A total of 96 fractions were collected and further pooled into 24 fractions based on chromatographic peaks. The pooled fractions were vacuum dried and desalted using C
18 StageTips and stored at −20 °C till further analysis.
LC-MS/MS analysis
Peptide fractions were analyzed on an LTQ-Orbitrap Velos mass spectrometer (Thermo Scientific, Bremen, Germany) interfaced with Proxeon Easy nLC II system (Thermo Scientific, Bremen, Germany). Peptides were loaded onto trap column (75 μm x 2 cm, Magic C18AQ, 5 μm, 100 Å, Michrom Biosciences Inc., Auburn, CA) using solvent A (0.1 % formic acid) at a flow rate of 3 μl/min and resolved on an analytical column (75 μm x 10 cm, Magic C18AQ, 3 μm, 100 Å, Michrom Biosciences Inc, Auburn, CA) at a flow rate of 350 nl/min using a linear gradient of 7 – 30 % acetonitrile over 80 min. The MS and MS/MS scans were acquired at a mass resolution of 60,000 and 15,000 at 400 m/z, respectively. Full MS scans were acquired in m/z range of 350 – 1800. For each cycle, twenty most abundant precursor ions with charge state ≥2 were sequentially isolated. The fragmentation was carried out using higher energy collision dissociation as the activation method with 40 % normalized collision energy. Isolation width was set to 2 m/z. Singly charged precursor ions and precursors with unassigned charge states were rejected. The acquired ions were dynamically excluded for 45 s. The automatic gain control for full MS and MS/MS was set to 1x10
6 and 5x10
4 ions, respectively. The maximum ion accumulation time was set to 100 ms for MS and 300 ms for MS/MS scans. The lock mass option was enabled using polysiloxane ion (m/z, 445.120025) from ambient air for internal calibration as described [
26].
Data analysis
The raw data obtained was processed using Proteome Discoverer (version 1.4) software suite (Thermo Fisher Scientific, Bremen, Germany) and searched using Sequest and Mascot (version 2.2.0, Matrix Science, London, UK) search algorithms against human protein database NCBI RefSeq (Release 63 containing 71,434 protein sequences and known contaminants). The search parameters included: trypsin as the proteolytic enzyme with two missed cleavages allowed, oxidation at methionine as the dynamic modification, alkylation (methylthio) at cysteine and iTRAQ 8-plex modification at N-terminus of the peptide and lysine as static modifications. Precursor and fragment mass tolerance were set to 20 ppm and 0.05 Da, respectively. The peptide and protein data were extracted using high peptide confidence and top one peptide rank filters. The data were also searched against a decoy database to calculate the false discovery rate (FDR). Peptide spectrum matches (PSMs) at 1 % FDR were used for protein identifications. iTRAQ quantitation was done by taking the average of the reporter ion intensities from the technical replicates. The ratios, invasive neoplastic/non-invasive neoplastic, were obtained as follows – 115 + 116 (OCUG-1)/113 + 114 (TGBC24TKB), 117 + 118 (NOZ)/113 + 114 (TGBC24TKB) and 191 + 121 (GB-d1)/113 + 114 (TGBC24TKB).
Proteins identified in this study were classified based on their subcellular localization, molecular function and biological process using Human Protein Reference Database (HPRD;
http://www.hprd.org) which is a Gene Ontology (GO) compliant database [
27,
28]. The top canonical pathways associated with the differentially expressed proteins in this study were identified through the use of QIAGEN’s Ingenuity Pathway Analysis (IPA®, http://
www.qiagen.com/ingenuity).
Accessibility of proteomic data
Immunohistochemistry
Tissue microarrays (TMAs) were constructed at Lab Surgpath, Mumbai using the paraffin blocks of gallbladder adenocarcinoma and cholecystitis cases obtained from Cancer Hospital and Research Institute, Gwalior, India with the approval from Institutional Human Ethics Committee and informed consent of the patients. The tissue microarrays were constructed with 29 cases of gallbladder adenocarcinoma and 16 cholecystitis cases. For this, two cores of 2 mm size was taken from each paraffin block and embedded to a recipient paraffin block.
IHC was carried out on both cholecystitis and gallbladder adenocarcinoma cases. A semi-quantitative assessment was performed to evaluate the immunoreactivity as described previously [
31]. Briefly, the formalin fixed paraffin embedded tissue sections were deparaffinised and antigen retrieval was carried out using heat-induced epitope retrieval by incubating the slides for 20 minutes in antigen retrieval buffer (0.01 M Trisodium citrate buffer, pH 6). Endogenous peroxidases were quenched using a blocking solution followed by washes with wash buffer (PBS with 0.05 % Tween-20). The sections were incubated with anti-MIF antibody (sc-20121, Santa Cruz Biotechnology, Dallas, TX) at 1:50 dilution overnight at 4 °C in a humidified chamber. The slides were incubated with appropriate horseradish peroxidase conjugated rabbit secondary antibody for 30 minutes at room temperature. Excess secondary antibody was removed using wash buffer followed by addition of DAB substrate. The signal was developed using DAB chromogen (DAKO, Glostrup, Denmark). Tissue sections were then observed under the microscope. The immunohistochemical labeling was assessed by an experienced pathologist. The intensity of staining was scored on a grading scale ranging from 0 to 3+, where 0 represented negative staining, 1+ represented weak staining, 2+ represented moderate staining and 3+ represented strong staining. To determine the statistical significance of MIF expression in gallbladder adenocarcinoma and cholecystitis, Chi-square test was carried out using R version 3.1.0.
Western blotting
Whole cell extracts of GBC cells, were prepared using modified RIPA lysis Buffer (Merck Millipore, Billerica, MA) containing protease inhibitors (Roche, Indianapolis, IN) and phosphatase inhibitors (Thermo Scientific, Bremen, Germany). Rabbit polyclonal anti-MIF was obtained from Santa Cruz (sc-20121, Santa Cruz Biotechnology, Dallas, TX). β-Actin was used as a loading control. Western blot analysis was performed as previously described [
32] using 30 μg protein lysates.
Processing of conditioned media
Each cell line was grown to ~80 % confluence, washed multiple times with PBS to remove any adherent serum from the cells and then grown in serum-free medium for 8 h. Post-starvation, the conditioned media was collected for each cell line, centrifuged at 800 × g for 10 min to remove any cellular debris. The supernatant was filtered using a 0.22 μm filter (Merck Millipore, Billerica, MA). The filtered supernatant was subsequently concentrated using 3 kDa cut-off filters (Merck Millipore, Billerica, MA). Protein concentration was estimated by BCA assay [
24]. Western blot analysis was performed as previously described [
32] using 30 μg protein lysates.
Cell viability assays
The GBC cells were seeded in a 96-well plate at a density of 1x10
4 cells/well. The cells were vehicle - treated or treated with MIF-antagonist [(S,R)-3-(4-hydroxyphenyl)-4,5-dihydro-5-isoxazole acetic acid methyl ester (ISO-1) (EMD Millipore, Billerica, MA) (0 to 500 μM) or 4-iodo-6-phenylpyrimidine (4-IPP) (Tocris Bioscience, Bristol, UK) (0 to 500 μM) for 48 h in complete medium at 37 °C in 5 % CO
2 incubator. After 48 h, the medium was aspirated, the cells were rinsed and MTT assays were performed as previously described [
33]. All experiments were performed in triplicate.
siRNA transfection
ON-TARGETplus SMARTpool control siRNA and
MIF siRNA were purchased from Dharmacon (Lafayette, CO). The GBC cells were transfected with 10 nM of
MIF siRNA or control siRNA using RNAiMAX (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. Transfection was carried out as previously described [
32]. Cells were subjected to invasion assay and viability assay 48 h post-transfection, unless otherwise stated.
GBC cell lines were transfected with either MIF siRNA or control siRNA. 3x103 cells/well were seeded in 6-well plates. Cell colonies were allowed to grow for 14 days, before the colonies were fixed with methanol and stained with 4 % methylene blue (Sigma, St. Louis, MO). The number of colonies per dish was counted. Similarly, the colony forming ability of the GBC cells were monitored in the presence of MIF antagonists, ISO-1 and 4-iodo-6-phenylpyrimidine (4-IPP). All experiments were performed in triplicate.
Cell invasion assays
Cell invasion assays were performed in a transwell system using cell culture inserts for 24-well plates with translucent polyethylene terephthalate membrane containing 8 μm pores (BD Biosciences, NJ). The upper compartment of the culture insert was coated with Matrigel (BD Biosciences, San Jose, CA). GBC cells (2x104) were seeded into the transwell chambers in presence of serum-free medium. Complete media was added to the lower compartment and the cells were incubated at 37 °C in 5 % CO2 incubator for 48 h. Post-incubation, the upper surface of the membrane was wiped with a cotton-tip applicator to remove non-migratory cells. Cells that migrated to the lower side of membrane were fixed and stained using 4 % methylene blue (Sigma, St. Louis, MO). The number of invaded cells was counted using a light microscope. All experiments were done in duplicates and repeated thrice.
Statistical analysis
Paired t-test was carried out to evaluate the difference between control and treated groups. P ≤ 0.05 was considered to indicate statistical significance.
Discussion
Early diagnosis and treatment of GBC requires elucidation of molecular events associated with tumor progression and aggressiveness in GBC. Mass spectrometry has emerged as a reliable tool to identify differentially regulated proteins across different conditions enabling the discovery of potential biomarkers and therapeutic targets in cancer. In this study, quantitative proteomic analysis of a panel of GBC cell lines led to the identification of more than 1,000 differentially expressed proteins - 654 of which were found to be overexpressed and 387 were downregulated in invasive GBC cell lines. MIF, a pro-inflammatory cytokine, was found to be overexpressed >3-fold in two of the invasive GBC cell lines as compared to the non-invasive cell line, TGBC24TKB.
Tumor growth and metastasis is often accompanied by chronic inflammation, a condition commonly observed in cholecystitis and in the development of GBC. The ability of MIF to suppress anti-inflammatory pathways makes it a molecule of choice to be investigated for such conditions. MIF enhances its activity by inducing other inflammatory cytokines including TNF-alpha and IL-1 [
39]. MIF has been reported to act as an antagonist of glucocorticoids regulating its anti-inflammatory effects [
15]. MIF exerts its effects via the CD74/CD44 receptor complex. MIF has also been reported to activate the chemokine receptors CXCR2 and CXCR4 [
40] to exert its chemokine-like function. Overexpression of MIF has been reported in multiple human cancers [
17,
18,
20,
22]. MIF contributes to tumor development, progression and tumor cell survival through inhibition of p53-mediated apoptosis. This is achieved through the sustained activation of the ERK signaling pathway [
41]. MIF also exerts its pro-survival and anti-apoptotic effects through the activation of PI3K/Akt cascade [
42]. MIF causes an increased transcription of cyclin D leading to hyperphosphorylation of Rb and hence augmenting cellular proliferation [
43]. Recent studies indicate that MIF leads to HIF-1α activation under hypoxic conditions leading to enhancement of cancer growth and metastasis [
44]. In addition, MIF has been suggested as a potential biomarker for hepatocellular carcinoma, colorectal cancer, gastric cancer and non-melanoma skin cancer [
45]. CD74, which acts as a receptor to MIF, was also found to be overexpressed in two of the GBC cell lines used in the proteomic experiment in this study. The expression of CD74 has been linked to several cancers [
46]. The co-receptor of MIF, CD44 was also found to be overexpressed in all the GBC cell lines. Long-term activation of CD44 has been reported to play a key role in tumor progression [
34].
Studies have shown that vaccination of human subjects with autologous tumor cells modified to secrete granulocyte-macrophage colony stimulating factor (GM-CSF) and antibody-based blockade of cytotoxic T-lymphocyte-associated antigen-4 (CTLA4) results in a humoral response against multiple angiogenic cytokines, including MIF. This antibody-based inhibition of MIF attenuates macrophage Tie-2 (TEK) expression and matrix metalloproteinase-9 (MMP9) production. This and studies by others indicate that blockade of VEGF, angiopoietins, and MIF may be effective in tumor regression [
47,
48]. Taken together, these findings suggest that MIF can be explored as a therapeutic target in GBC.
For a molecule to act as a therapeutic target it is essential that it has to be expressed in the cancer tissue. In this study, tissue microarray-based immunohistochemical staining revealed overexpression of MIF in more than 72 % of the gallbladder adenocarcinoma cases. These findings suggest MIF as a potential therapeutic target in GBC.
ISO-1 is an antagonist of MIF which binds to the hydrophobic catalytic pocket of MIF and inhibits its tautomerase activity thereby counteracting glucocorticoid-inhibited TNF release as well as inhibiting the cytokine action of MIF on PLA2 activity. Inhibition of MIF using ISO-1 has been demonstrated to provide protection from septic shock induced by endotoxins [
38].
In vitro and
in vivo MIF inhibition using its specific inhibitor has been shown to be potentially effective in multiple cancers [
19,
23,
49‐
54]. In our study, knockdown of endogenous MIF expression using ISO-1 or its specific siRNA showed a significant decrease in cellular proliferation, invasion and colony forming ability of GBC cell lines. As evidenced from the current study and in agreement with literature, relatively high concentrations of ISO-1 are potentially effective in rendering cellular death. This property of ISO-1 has hindered the use of this antagonist in clinical settings. Meanwhile, the small molecule inhibitor 4-IPP has been reported to be ~5-10 times more potent than the MIF antagonist, ISO-1. The antagonist, 4-IPP acts as a suicide substrate to MIF through covalently modifying the catalytically active N-terminal proline [
23]. In this study, we demonstrate that the inhibition of MIF activity using 4-IPP decreased cellular proliferation, invasion and colony forming ability of GBC cell lines and was more potent than the prototypic MIF antagonist, ISO-1. Taken together, these studies provide experimental evidence of the potency of the MIF inhibitors in multiple cancers including GBC. We suggest that targeted MIF therapy might be effectively combined with antibody-based therapy to improve patient outcome in other cancers including GBC. Further clinical investigations of these inhibitors are needed to establish their role as a therapeutic target in cancer.
Acknowledgements
We thank the Department of Biotechnology (DBT), Government of India for research support to the Institute of Bioinformatics. IOB is supported by DBT Program Support on Neuroproteomics and infrastructure for proteomic data analysis (BT/01/COE/08/05). We thank the "Infosys Foundation" for the research support to the Institute of Bioinformatics. This work was supported by the Science and Engineering Research Board, Department of Science and Technology, Government of India grant “miRNAs in chronic tobacco-induced oral cancer (SR/S0/HS-02081/2012)”; NCI’s Clinical Proteomic Tumor Analysis Consortium initiative (U24CA160036) and FAMRI-funded 072017_YCSA. P.K. Tiwari acknowledges research support from the Indian Council of Medical Research (ICMR) and DBT, Government of India. Harsha Gowda is a Wellcome Trust/DBT India Alliance Early Career Fellow. Juan Carlos Roa acknowledges research support from the National Fund for Scientific and Technological Development (FONDECYT 1130204), and CONICYT- FONDAP 15130011, Government of Chile. Pamela Leal acknowledges research support from the National Fund for Scientific and Technological Development (FONDECYT 1151008) and Postdoc Research Fellowship (CONICYT-BECAS CHILE 74130044), Government of Chile. Mustafa A. Barbhuiya is a recipient of Senior Research Fellowship from ICMR, India. Gajanan Sathe is a recipient of Senior Research Fellowship from the Council for Scientific and Industrial Research (CSIR), India. Remya Raja is a recipient of Research Associateship from DBT, Government of India. Santosh Renuse and Nazia Syed are recipients of Senior Research Fellowship from University Grants Commission (UGC), Government of India. We thank Dr. S. K. Shankar of National Institute of Mental Health and Neuro Sciences for providing the use microscope facility.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
AP, HG and AC designed the study. RG, DS and JCR provided cell lines and scientific input. TS, PLR, MAB, RR, NS and VN performed the experiments and wrote the manuscript. SR, GS, SMP and NAS performed LC-MS/MS. AHP and PG carried out data analysis. SN constructed the tissue microarrays. MAB and PKT provided the samples for immunohistochemistry. VS scored the tissue microarrays. BP and TSKP edited the manuscript. All authors have given final approval of the version to be published. All authors have read and approved the final manuscript.