Glycoprotein screening in colorectal cancer based on differentially expressed Tn antigen

  • Authors:
    • Hongyun Wei
    • Zongyong Cheng
    • Chunhui Ouyang
    • Yu Zhang
    • Yanyan Hu
    • Shuijiao Chen
    • Chunlian Wang
    • Fanggen Lu
    • Jie Zhang
    • Yongjun Wang
    • Xiaowei Liu
  • View Affiliations

  • Published online on: July 14, 2016     https://doi.org/10.3892/or.2016.4937
  • Pages: 1313-1324
Metrics: Total Views: 0 (Spandidos Publications: | PMC Statistics: )
Total PDF Downloads: 0 (Spandidos Publications: | PMC Statistics: )


Abstract

Colorectal cancer (CRC) is one of the most common cancers worldwide, and the identification of new biomarkers for CRC is valuable for its diagnosis and treatment. We aimed to screen differentially expressed glycoproteins (especially O-glycoproteins) and to identify diagnostic or therapeutic candidates for colorectal cancer (CRC) based on different Tn antigen expression levels. Fresh cancer tissues and adjacent healthy tissues were obtained from CRC patients and classified into three groups based on their Tn antigen expression: CRC with negative Tn expression (CRC Tn‑), CRC with positive Tn expression (CRC Tn+) and normal control without Tn expression (NC). Protein extractions were separated and identified by iTRAQ technology. Glycoproteins and O-glycoproteins were selected using UniProt and DAVID. Deep bioinformatic analysis, including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KO), was used to annotate this O-glycoprotein interaction network. Subsequently, two O‑glycoproteins were verified by western blotting and immunohistochemistry in either LS174T cells or CRC tissues. We found that 330 differentially expressed proteins were identified by iTRAQ between CRC Tn‑ and NC tissues, 317 between CRC Tn+ and NC tissues, and 316 between CRC Tn‑ and Tn+ tissues. Of the 316 proteins, 55 glycoproteins and 19 O‑glycoproteins were identified and analyzed via deep informatics. Namely, different Tn antigen expression levels in CRC led to differential protein expression patterns, especially for glycoproteins and O‑glycoproteins. Decorin and SORBS1, two representative functional O-glycoproteins, were significantly downregulated in the CRC Tn+ tissues compared with the level in the CRC Tn‑ or NC tissues. Based on this deep bioinformatic analysis, Decorin and SORBS1 are hypothesized to be involved in the TGF‑β and PPAR‑γ signaling pathways, respectively.

Introduction

Colorectal carcinoma (CRC) is one of the most common cancers worldwide. However, the early diagnosis rate of CRC is low, and surgery is less effective in patients with more advanced CRC (stage III), for whom the mortality rate is high and the prognosis is poor. Early diagnosis depends on a combination of early screening of non-specific biomarkers such as CEA and CA199 and pathological endoscopic biopsies (1,2). Consequently, most patients are diagnosed at the intermediate or late stages and deprived of the opportunity for surgery. Therefore, the lack of specific biomarkers for earlier diagnosis and therapeutic targets leads to high CRC mortality worldwide (24), and CRC is one of the primary causes of cancer-related death in the world. Intervention in the early stage could cure colorectal cancer, and the identification of CRC biomarkers may be a convenient way to achieve this objective.

Recent studies show that the pathogenesis of cancer is highly related to glycosylation (5,6). The increased expression of N-acetylgalactosamine (GalNAc) on the surface of cancer cells is one of the characteristics of the altered glycosylation pathway (7). Mucins are highly O-glycosylated proteins that play pivotal roles in cell adhesion, inflammation, cell proliferation, apoptosis and tumorigenesis (811). The mucin-associated carbohydrate Tn antigen is composed of a GalNAc residue covalently O-linked to a serine/threonine (12,13), which is shielded in healthy tissues and benign tumors but occurs in ~90% of all human primary carcinomas (14). The synthesis of sialyl-Tn antigen (sTn) is regulated by the sialyltransferase ST6GalNAc, which competes against O-glycan elongating glycosyltransferases and prevents tumors from exhibiting longer O-glycans (12,15). Abnormal Tn antigen expression has been observed in lung, breast, ovarian and CRC (1620). Tn antigen is also associated with malignancy and represents a potential marker of malignant transformation (17,21,22). Recent studies have shown that higher Tn antigen expression is observed in CRC patients (14,23,24). Tn antigen may promote cancer metastasis by secretion of transforming growth factor-β (TGF-β) in the tumor microenvironment (25). In addition, Tn was found to be expressed in moderately differentiated tumors but not in poorly differentiated carcinomas (26), and a deficiency of core-1-derived O-glycans could delay the onset and progression of breast cancer (18). Furthermore, Tn antigen is related to poor prognosis in patients with breast cancer (27). Above all, we hypothesized that the Tn antigen plays a central role in the early diagnosis and the development of CRC. However, the Tn-related proteins participating in its pathogenic progression remain unclear. It is therefore worthwhile to analyze the expression of Tn-related proteins. In the present study, we used iTRAQ to screen differentially expressed proteins between Tn-positive and Tn-negative CRC tissues and to identify accessible proteins as useful diagnostic biomarkers or new therapeutic candidates for CRC.

Materials and methods

Patient information and sample preparation

A series of primary colon resections were prospectively collected between November 2011 and March 2012 from the Department of Minimally Invasive Surgery of the Second Xiangya Hospital in Changsha. Each resection contained both colon tumor samples and adjacent healthy tissues, and lymph nodes were also collected. Some of the samples were kept at −80°C until analysis, and the remaining samples were fixed in 4% paraformaldehyde and sent to the Pathology Department for further analysis. CRC samples were screened for Tn staining via immunochemical examination by Dr Lijun Xia from the Oklahoma Medical Research Foundation (OMRF; Oklahoma City, OK, USA). Based on the Tn antigen expression, the samples were divided into three groups as follows: CRC with negative Tn expression (CRC Tn), CRC with positive Tn expression (CRC Tn+) and normal control without Tn expression (NC). No patient was treated with any kind of chemotherapy. The patient information is shown in Table I.

Table I

Patient characteristics.

Table I

Patient characteristics.

Pt. no.Mean age (years)GenderHistological assessmentTNMTnCategory
164MaleRectal tubulovillous adenocarcinomaTis0Tn
250FemaleModerately-poorly differentiated rectal adenocarcinomaT3N1bM0 (IIIB) Tn/normal
381MaleModerately differentiated colon adenocarcinomaT3N0M0 (IIA)Tn
456FemaleModerately differentiated rectal adenocarcinomaT4aN1aM0 (IIIB)+ Tn+/normal
570MaleModerately and poorly differentiated rectal adenocarcinomaT3N0M0 (IIA)+ Tn+/normal
652MaleModerately differentiated colon adenocarcinomaT3N0M0 (IIA)+Tn+

This study was approved by the Institutional Ethics Committee of the Second Xiangya Hospital of Central South University. All participants provided written informed consent.

Protein extraction and iTRAQ labeling

One gram of fresh tissues was ground in liquid nitrogen and suspended in 5 ml of lysis buffer with supplements (7 mM urea, 2 mM sulfourea, 0.1% PMSF, 65 mM DTT) on ice for 30 min, followed by centrifugation for 15 min at 24,1488 × g. The supernatant was collected, and the protein concentration was determined using the Bradford Protein Assay kit (Bio-Rad, USA).

For peptide labeling, a peptide mixture from each group was labeled with iTRAQ tags according to the kit's protocol (Applied Biosystems, Foster City, CA, USA). Protein samples from adjacent healthy tissues were labeled with reagent 114. Tn CRC and Tn+ CRC samples were labeled with reagent 117 and 118, respectively.

2D LC-MS/MS

The peptide mixture was reconstituted and acidified with buffer A (10 mM KH2PO4, pH 2.6, and 25% acetonitrile) (ACN; Thermo Fisher Scientific, Inc., Fair Lawn, NJ, USA) and loaded onto a strong cation exchange (SCX) chromatography column (The Nest Group, Inc., Southborough, MA, USA) on a 20AD high-performance liquid chromatography (HPLC) system (Shimadzu, Kyoto, Japan). The peptides were eluted at a flow rate of 200 µl/min with a gradient of 0–80% buffer B for 60 min. The elution was monitored by absorbance at 214/280 nm, and fractions were collected every minute. The collected fractions were combined and desalted on a C18 cartridge (Waters Corp., Milford, MA, USA).

All samples were resuspended in 50 µl HPLC buffer A (5% ACN and 0.1% formic acid), loaded in the ZORBAX 300SB-C18 reverse-phase column (5 µm, 300 Å, 0.1×150 mm; Michrom BioResources, Auburn, CA, USA). The peptides were separated for >90 min using a linear gradient of 5–35% HPLC buffer B (95% ACN and 0.1% formic acid) at a flow rate of 300 nl/min. The MS analysis was performed using a QSTAR XL analyzer (Applied Biosystems) coupled with a 20AD HPLC system (Shimadzu). Precursor ions were selected across the mass range of 400–1,800 m/z. Four precursors were selected for MS/MS analyses across the mass range of 100–2,000 m/z.

Data analysis and bioinformatics

Protein identification and relative iTRAQ quantification were performed with the Paragon algorithm in the ProteinPilot Software 3.0 (revision 114732; Applied Biosystems). The results were further processed using the Pro Group algorithm (Applied Biosystems), where isoform-specific quantification was adopted to trace the differences between the expression levels of various isoforms. Quantitative ratios of reporter ions, calculated by comparing the peak areas of each of these reporter ions in the mass spectrum, were used to evaluate the expression change of the protein in different samples. Normalization was performed during the quantification to correct the experimental bias. To accept proteins as showing differential expression between different groups, we strictly followed the following criteria. It was mandatory for proteins to be identified with >95% confidence, and the protein confidence threshold cutoff was 1.3 (unused ProtScore). Proteins with significant P-values (P<0.05) were considered differentially expressed. The false discovery rate (FDR) for protein detection was calculated as FDR = (2 × reverse)/(forward + reverse). UniProt was used to annotate the proteins in the biological process and molecular function categories in the human UniProtKB/Swiss-prot database (version 3.52, November 2008).

Gene Ontology and pathway analysis

Gene Ontology (GO) was used to annotate the proteins by biological process and molecular function. We translated the genes into putative amino acid sequences and aligned these genes against a set of protein sequences from the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Each protein was assigned to KEGG orthology (KO), and differentially enriched KO pathways were identified.

Cell culture and immunofluorescence staining of the Tn antigen

LS174T cells (a gift from Dr Lijun Xia, OMRF), were cultured in Dulbecco's modified Eagle's medium (DMEM; Hyclone, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS; Zhejiang Tianhang Biological Technology Stock Co., Ltd., Zhejiang, China) and antibiotics (100 U/ml penicillin and 100 mg/ml streptomycin; Gibco, USA).

For immunofluorescence staining of the Tn antigen, LS174T cells on cover slips were fixed in 4% paraformaldehyde (Sigma), permeabilized with 0.1% Triton X-100 (Promega) at room temperature for 15 min, and blocked with 1% BSA (Gibco) for 30 min at room temperature. The slides were incubated with a mouse monoclonal antibody against Tn (made by OMRF) at 1:50 overnight at 4°C, followed by a 1-h incubation at room temperature with the secondary antibody (anti-mouse antibody, 1:2,000; Jackson ImmunoResearch, West Grove, PA, USA). The sections were counterstained with DAPI for 2 min and then visualized and photographed under a confocal laser scanning microscope (LaVision BioTec GmbH, Bielefeld, Germany). To confirm the expression of Decorin or SORBS1, 3×106 LS174T cells were harvested, and the total proteins were prepared for subsequent western blotting.

Western blotting

The extracted proteins from tissues or cells were resolved by SDS-PAGE gradient gel (5% spacer gel and 8% separation gel). After electrophoresis, samples were transferred to a polyvinylidene fluoride membrane, blocked at room temperature for 1 h with 5% milk, and incubated with a mouse antibody against Decorin (1:500; Abcam, USA), a rabbit antibody against SORBS1 (1:1,000; Abgent, USA) or a mouse antibody against β-actin (1:1,000; Abcam) at 4°C overnight. The membranes were washed with buffer three times, incubated with an HRP-conjugated anti-rabbit secondary antibody (1:5,000) or anti-mouse secondary antibody (1:20,000) (both from Abcam) at room temperature for 1 h, and visualized with an ECL kit (GE Healthcare). Specific signals were quantified from exposed X-ray film using a scanner with BandScan 4.30 densitometry software and are expressed as integrated intensity units relative to the β-actin signals.

Immunohistochemistry assay in tissues

To detect the immunohistochemical expression of Decorin or SORBS1, the embedded tissues were cut into 5-µm thin sections. These sections were incubated with an antibody against Decorin (1:50) or SORBS1 at 1:100 overnight at 4°C after blocking with 2% BSA. Subsequently, the sections were treated with an HRP-conjugated anti-rabbit (1:5,000) or anti-mouse secondary antibody (1:20,000) at room temperature for 1 h. The DAB Detection kit (EliVision Super DAB; Maixin-Bio, Fuzhou China) for blocking non-specific binding and antibody detection was used. The sections were then counterstained with hematoxylin. The immunostaining results were detected under a light microscope.

Statistics

All observations were confirmed by at least three independent experiments. Image-Pro Plus 6.0 was used to analyze the results from IHC, and Quantity One v4.6.2 was used for western blotting. For quantitative data, t-test and ANOVA were used for statistical analysis. The analyses were performed using SPSS Statistics 20 (SPSS Inc. Chicago, IL, USA). P<0.05 was considered statistically significant.

Results

Tn antigen expression and clinicopathological factors in CRC patients

Sixteen patients with a confirmed diagnosis of CRC were included. The enrolled patients included three males and three females, with ages ranging from 50 to 81 years (mean age, 62). Most were diagnosed with moderately differentiated cancer. The Tn antigen expression was detected in 3 patients, as shown in Table I.

Identification of Decroin and SORBS1 as differentially expressed glycoproteins in CRC by iTRAQ

A total of 1,051 non-redundant proteins in the genome (UniProt) were significantly identified from the number of MS/MS spectra and number of peptides using a 2% false discovery rate (FDR) as the cutoff in the triplicate independent experiments. Of these, 750 had quantitative information, and more than two peptides were matched to these proteins. There were 330 differentially expressed proteins between CRC Tn and NC tissues, 317 between CRC Tn+ and NC tissues, and 316 between CRC Tn and CRC Tn+ tissues. Among the 316 proteins, there were 55 glycoproteins and 19 O-glycoproteins. Twelve of the 19 O-glycoproteins were also differentially expressed between the CRC Tn/CRC Tn+ tissues and normal mucosal tissues (Tables II and III).

Table II

Differentially expressed proteins in Tn+ compared with Tn− tissues.

Table II

Differentially expressed proteins in Tn+ compared with Tn− tissues.

Pt. no.AccessionName Tn+:Tn PValTn+:Tn EFTn+:Tn
1P24821Tenascin0.619441092010.0417615696791.1803209782
2P27797Calreticulin1.29419600960.101053297521.1912419796
3O43852Calumenin1.52756595610.139537304641.2022639513
4P02768Serum albumin2.01372408870.130205497151.213389039
5P05787Keratin, type II cytoskeletal 80.816582381730.384063005451.1168630123
6P05783Keratin, type I cytoskeletal 180.47424200177 0.000499245012181.213389039
7P11047Laminin subunit γ-10.654636085030.184094905851.2473829985
8P12110Collagen α-2(VI) chain1.25892496110.0072906902061.2705739737
9Q01082Spectrin β chain, brain 11.36772894860.184956699611.2941960096
10P01023 α-2-macroglobulin0.666806817050.342034399511.3061709404
11P02766Transthyretin0.698232412340.210423797371.3182569742
12P02763α-1-acid glycoprotein 11.67494297030.495774805551.3304539919
13Q9Y4L1Hypoxia upregulated protein 11.51356101040.0651060268281.3304539919
14P55268Laminin subunit β-20.711213529110.171349093321.3304539919
15P02452Collagen α-1(I) chain0.554625689980.0194047093391.3427649736
16Q12864Cadherin-170.744732022290.212115794421.3551889658
17P05164 Myeloperoxidase0.648634374140.120273999871.3677289486
18P02787 Serotransferrin1.83653795720.195306792861.4190570116
19Q9Y6R7IgG Fc-binding protein1.30617094040.299269795421.4190570116
20P08294Extracellular superoxide dismutase [Cu-Zn]0.642687678340.303986102341.4321880341
21P10153Non-secretory ribonuclease1.35518896580.134217604991.4454400539
22P69905Hemoglobin subunit α0.42461958528 0.00315799703821.4454400539
23P01833Polymeric immunoglobulin receptor3.34194993972.79E-061.4859360456
24P02675Fibrinogen β chain0.510505020620.463126987221.4859360456
25P08123Collagen α-2(I) chain0.461317598820.209063097831.4859360456
26P98160Basement membrane-specific1.38038396840.354511290791.4996850491
27P01857Ig γ-1 chain C region1.35518896580.506120383741.4996850491
28P00738Haptoglobin0.452897608280.105233103041.4996850491
29P08238Heat shock protein HSP90-β1.73780095580.325008213521.5135610104
30P01876Ig α-1 chain C region1.64437198640.167511001231.5135610104
31P05155Plasma protease C1 inhibitor0.636795520780.381509512661.5135610104
32P01009 α-1-antitrypsin0.575439870360.518361985681.5135610104
33P02788 Lactotransferrin1.38038396840.407394200561.5275659561
34P10645Chromogranin A2.67916798590.0201653502881.5703630447
35P10909Clusterin0.679203629490.502867579461.5995579958
36P01871IgM chain C region0.660693526270.679463028911.5995579958
37Q9BX66Sorbin and SH3 domain-containing protein 10.474242001770.0988838672641.6292959452
38P01877Ig α-2 chain C region0.666806817050.442894190551.6443719864
39P22105Tenascin-X0.794328212740.498078793291.6595870256
40P02750Leucine-rich α-2-glycoprotein0.501187205310.268613100051.6904410124
41P51888Prolargin1.31825697420.956294119361.7060819864
42P62807Histone H2B type 1-C/E/F/G/I0.487528502940.0987804904581.7060819864
43P02671Fibrinogen α chain0.0824138075111.61E-061.7060819864
44P04217 α-1B-glycoprotein0.405508488420.160428106781.7060819864
45P16070CD44 antigen0.679203629490.490876197811.7218689919
46P02790Hemopexin1.73780095580.307444691661.7378009558
47Q02817Mucin-25.3951058388 0.00256116990931.9054609537
48O00748Cocainee sterase1.92309200760.16036160291.7538809776
49P04196Histidine-rich glycoprotein0.465586096050.197445899251.8197009563
50P62937Peptidyl-prolyl cis-trans isomerase A3.0478949547 0.000119985597851.9230920076
51P07585Decorin0.7311391230.6357188821.99843504
52Q16363Laminin subunit α-40.711213529110.705770373341.9230920076
53P02751Fibronectin0.816582381730.244894698261.213389039
54P12109Collagen α-1(VI) chain0.801678121090.506826519971.2359470129
55P23229Integrin α-60.809095919130.581975102421.4190570116

[i] iTRAQ ratio indicates the relative quantification of the differentially expressed proteins. Statistical analysis of iTRAQ ratio was performed.

Table III

Differentially expressed glycosylated proteins in Tn+ compared with Tn− tissues

Table III

Differentially expressed glycosylated proteins in Tn+ compared with Tn− tissues

Pt. no.UnusedAccessionNamePeptides (95%)Differential expression
Tn+:Tn Tn:normal Tn+:normal
189.82P02768Seruma lbumin842.0137240890.1940885930.440554887
2101.57P05787Keratin, type II cytoskeletal 81080.8165823821.7218689921.445440054
345.92P05783Keratin, type I cytoskeletal 18420.4742420024.0550851821.958845019
437.56Q01082Spectrin β chain, brain1261.3677289490.6918309930.963828981
540.35P02452Collage α-1(I) chain340.554625691.0665960310.586138189
623.09P02787 Serotransferrin141.8365379570.4875285030.862978518
78.5P98160Basement membrane-specific heparan sulfate proteoglycan core protein41.3803839680.5058246850.704693079
810.12P08238Heat shock protein HSP90-β221.7378009561.0964779851.923092008
924.54P01876Ig α-1chain C region221.6443719860.44874540.724435985
106P05155Plasma protease C1 inhibitor30.6367955211.8535319571.224616051
1122.08P10645Chromogranin-A142.6791679860.15995580.420726597
1210.04Q9BX66Sorbin and SH3 domain-containing protein 170.4742420020.5058246850.248885706
136P02750Leucine-rich α-2-glycoprotein30.5011872050.7726805810.390840888
1421.35P62807Histone H2B type 1-C/E/F/G/I320.4875285030.903649390.432513803
156P16070CD44 antigen30.6792036291.9230920081.342764974
165.43P02790Hemopexin31.7378009560.5248075130.990831971
1714.3Q02817Mucin-285.3951058390.3133285941.614359021
1837.01P02751Fibronectin210.8165823820.9638289810.787045777
192.99P07585Decorin20.7311391230.3250873090.260615289

Sixty-five proteins were differentially expressed in CRC Tn tissue compared with NC, while 74 proteins were significantly altered in CRC Tn+ compared with NC. Forty proteins displayed differential expression between CRC Tn and CRC Tn+, and 25 of these proteins were simultaneously significantly differentially expressed in CRC Tn and Tn+ tissues compared with NC (Fig. 1).

Among the significantly differentially expressed proteins, there were a few glycosylated proteins, especially O-glycosylated proteins. Fifty-five glycosylated proteins exhibited differential expression between CRC Tn+ and CRC Tn, including 19 O-glycosylated proteins (Tables II and III). Among the differentially expressed O-glycosylated proteins, 12 were significantly differentially regulated in groups, including Keratin 8 (gene: KRT8), Keratin 18 (gene: KRT18), Decorin (gene: DCN), Sorbin and SH3 domain-containing protein 1 (SORBS1), and CD44 antigen (gene: CD44) (Tables II and III).

Differential expression of Decorin and SORBS1 in CRC tissues or cell lines with different expression levels of the Tn antigen

We investigated whether the expression of Decorin and SORBS1 is related to the Tn antigen expression level in CRC tissues. The relative quantification of Decorin using iTRAQ showed that Decorin was more highly expressed in normal tissues than in CRC tissues, especially Tn+ tissues (Fig. 2A), as in the results for SORBS1 using iTRAQ. To verify the function of Decorin and SORBS1 in carcinogenesis, we used western blotting and IHC to quantify their expression in NC, CRC Tn and CRC Tn+ tissues. As shown in Fig. 2B–E, Decorin expression was significantly decreased in the CRC Tn+ tissues compared with the CRC Tn and NC tissues (P<0.05). A significant difference was also found in the expression of SORBS1 among Tn+, Tn and NC tissues, as shown in Fig. 3.

To verify the Tn antigen expression levels in Tn+ cells, we demonstrated that Tn antigen was significantly increased in Tn+ cells compared with Tn cells via immunofluorescence assay (Fig. 4A). We also detected the expression levels of Decorin and SORBS1 in the LS174T cells with different Tn expression levels and found that the expression of Decorin was significantly decreased in the Tn+ LS174T cells compared with that noted in the Tn cells (P<0.05) (Fig. 4B and C). Surprisingly, SORBS1 was not detected in the LS174T cells.

Functional analysis of 12 differentially expressed O-glycosylated proteins related to CRC

Functional analysis of the O-glycosylated proteins showed that the proteins participated in metabolic processes, in the regulation of biological processes, in the formation of membrane proteins of cells and organelles, and in the binding to other proteins (Table IV). In the present study, we paid particular attention to the analysis of metabolic processes, especially for SORBS1 and Decorin. SORBS1 was correlated with cell-cell adhesion via paxillin (PXN), the PPAR signaling pathway and the insulin signaling pathway via INS/IRS1. It could also interact with the Cbl proto-oncogene (CBL). Decorin participated in the TGF-β pathway via TGFB1/3 and connected with JUN, neurocan (NCAN), syndecan (SDC) and FOS-like protein-1 (FOSL1), as shown in the network in Fig. 5 and Table V.

Table IV

Functional analysis of differentially expressed O-glycosylated proteins.

Table IV

Functional analysis of differentially expressed O-glycosylated proteins.

FunctionsNo. of genesNo. of proteinsProteins (%)
Metabolic processes271263.16
Regulation of biological processes411052.63
Cell parts631157.89
Organelles181052.63
Organelle parts261473.68
Binding341368.42

Table V

KEGG pathway analysis of O-glycosylated proteins.

Table V

KEGG pathway analysis of O-glycosylated proteins.

KEGG-controlled vocabulariesKEGG no.GeneProtein
Focal adhesionHsa04510FN1Fibronectin
Regulation of actin cytoskeletonHsa04810FN1Fibronectin
ECG-receptor interactionHsa04512FN1Fibronectin
Small cell lung cancerHsa05222FN1Fibronectin
Bacteria invasion of epithelial cellsHsa05100FN1Fibronectin
AmoebiasisHsa05146FN1Fibronectin
Pathways in cancerHsa05200FN1Fibronectin
HSP90AB1Heat shock protein HSP90-β
Prostate cancerHsa05215HSP90AB1Heat shock protein HSP90-β
Antigen processing and presentationHsa04612HSP90AB1Heat shock protein HSP90-β
NOD-like receptor signaling pathwayHsa04621HSP90AB1Heat shock protein HSP90-β
Protein processing in endoplasmic reticulumHsa04141HSP90AB1Heat shock protein HSP90-β
Progesterone-mediated oocyte maturationHsa04914HSP90AB1Heat shock protein HSP90-β
Adherens junctionHsa04520SORBS1SORBS1
Insulin signaling pathwayHsa04910SORBS1SORBS1
PPAR signaling pathwayHsa03320SORBS1SORBS1
TGF-β signaling pathwayHsa04350DCNDecorin
Correlation of Decorin and SORBS1 expression in CRC with clinicopathological factors

To ascertain whether the expression of Decorin and SORBS1 was related to neoplasm staging, we compared the expression of Decorin and SORBS1 with TNM stages. Decorin and SORBS1 were not significantly increased in TNM stage 0-II compared with stage III (Table VI, P>0.05). However, a significant correlation was found between the expression of the two proteins and the tumor differentiation stages. Decorin and SORBS1 were significantly increased in moderately differentiated tissues compared with poorly differentiated tissues (Table VI, P<0.05).

Table VI

Expression of Decorin/SORBS1 in different TNM stages and differentiation statuses (mean ± SD).

Table VI

Expression of Decorin/SORBS1 in different TNM stages and differentiation statuses (mean ± SD).

Tumor featuresSamplesDecorinSORBS1
TNM 0-IIa51.58±0.052.02±0.77
TNM III41.05±0.451.38±0.53
Moderately differentiatedb61.69±0.292.23±0.41
Poorly differentiated30.87±0.201.07±0.09

a Compared with TNM III, the expression of the two proteins was not significantly increased (P>0.05);

b expression of the two proteins was significantly increased in the moderately differentiated group (P<0.05).

Discussion

Since the Tn antigen participates in the development of CRC (24,28,29), we investigated specific Tn antigen-related biomarkers as novel indicators for diagnostic or therapeutic targets. We analyzed the differentially expressed proteins between CRC Tn+ and CRC Tn tissues with the iTRAQ-based proteome technique. In the present study, 25 proteins were significantly expressed only in the CRC groups compared with the NC tissues, indicating that these proteins may participate in the development of CRC. The expression of 10 proteins was differently regulated in CRC Tn tissues compared with CRC Tn+ tissues but not in NC tissues. Considering that the Tn antigen was expressed mainly in moderately differentiated and not poorly differentiated tumors, we investigated the potential effects of these 10 proteins on the evaluation of the prognosis of carcinoma. Thirteen proteins were significantly differentially regulated among three groups, demonstrating the potential role of these 13 proteins in the initiation and progression of CRC.

Given that the Tn antigen is related to the disarrangement of O-glycosylation, we chose 12 differentially expressed O-glycosylated proteins for further study. Among these O-glycoproteins, Keratin 8, Keratin 18, Decorin, Sorbin, SORBS1 and CD44 antigen were reported to correlate with the initiation and progression of CRC. As yet, few studies have investigated the relationship between SORBS1, Decorin, the Tn antigen and the development of CRC (30,31).

SORBS1 belongs to a growing family of proteins containing a Sorbin homology (SoHo) domain and three SH3 domains in the C-terminal region. It plays an important role in adhesion between cells and the matrix via the SH3 domain (32,33). A study showed that c-Abl kinase, the product of the c-abl proto-oncogene, mediated the effects of cell adhesion on cell cycle progression or gene expression (34). Another study found that c-Abl could bind the SH3 domain of SORBS1 (30). Considering the abnormal adhesion between cancer cells, we hypothesized that SORBS1 may play a role in preventing CRC metastasis. Furthermore, SORBS1 together with vinculin was found to inhibit cell migration (35). p53 promoted the expression of SORBS1 in EB-1 cells (36,37). As SORBS1 is dependent on p53, mutations which correlate with cancer metastasis via the downregulation of adhesion proteins, we deduced that SORBS1 reduction may participate in the development and progression of carcinoma. In the present study, we found that SORBS1 expression was significantly downregulated in the CRC Tn+ tissues compared with levels in the CRC Tn and the healthy tissues, especially in poorly differentiated tissues. From the GO network, it was observed that SORBS1 also is connected with the proto-oncogene Cbl (38). According to the functional analysis, SORBS1 participates in the PPAR-γ pathway, which has been identified as a 'driver gene' in CRC initiation and progression (39). Thus, we demonstrated that downregulation of SORBS1 plays a potential role in the development of CRC via the PPAR-γ pathway (Fig. 6).

Moreover, Decorin controls the expression of p21, affects the activation of EGFR, promotes the apoptosis of cancer cells, and plays a key role in the immune response (40). Downregulation of Decorin is related to poor prognosis in breast cancer, and a similar role was found in lung cancer and lymphoma (27,4143). However, highly-expressed Decorin suppressed the development and metastasis of breast cancer (43). Few studies on the relationship between Decorin and CRC have been established. Bi et al found that the Decorin expression level was significantly reduced which may lead to tumorigenesis in mice (40). Our results showed that Decorin was decreased in the CRC Tn+ tissue compared with this level in the CRC Tn and the control tissues, supporting the aforementioned conclusion. Furthermore, functional analysis revealed that Decorin is involved in the TGF-β pathway, which has been found to be associated with the onset and progression of intestinal cancer (4446). In addition, Decorin is linked with JUN, NCAN, SDC1 and FOSL1. NCAN is involved in the modulation of cell adhesion and migration (47), and SDC1 functions as an integral membrane protein and participates in cell proliferation, cell migration and cell-matrix interactions via its receptor for extracellular matrix proteins. Altered SDC1 expression has been detected in several different tumor types (4850). The FOS gene family consists of four members: FOS, FOSB, FOSL1, and FOSL2. These genes encode leucine zipper proteins that can dimerize with proteins of the JUN family, thereby forming the transcription factor complex AP-1 (51). As such, the FOS proteins have been implicated as regulators of cell proliferation, differentiation, and transformation (52,53). Therefore, we conclude that downregulated Decorin could lead to the development of cancer via the TGF-β pathway as our results confirmed (Fig. 6).

To verify our hypothesis that abnormal O-glycosylation may control the development of carcinoma via influencing the expression of SORBS1 and Decorin, we also tested their expression in vitro. The expression of Decorin was in accordance with the results obtained from western blotting. Notably, SORBS1 was not detected in the Tn+ colorectal cells. As it is not a secretory protein, it is not possible that it was excreted outside the cell. We speculated that the SORBS1 protein level might be too low to be detected in vitro. Further studies with a larger number of samples are needed to clarify the relationship between Decorin, SORBS1 and neoplasm staging.

Abbreviations:

CRC

colorectal cancer

SORBS1

SH3 domain protein 1

iTRAQ

isobaric tags for relative and absolute quantitation

Acknowledgments

We would like to thank Elsevier for their language editing service. This study was supported by the National Natural Science Foundation of China (nos. 81172299 and 81570504).

References

1 

Kim EH and Misek DE: Glycoproteomics-based identification of cancer biomarkers. Int J Proteomics. 2011:6019372011. View Article : Google Scholar : PubMed/NCBI

2 

Chen W, Liu Q, Tan SY and Jiang YH: Association between carcinoembryonic antigen, carbohydrate antigen 19–9 and body mass index in colorectal cancer patients. Mol Clin Oncol. 1:879–886. 2013.

3 

Yoon YS, Keum N, Zhang X, Cho E and Giovannucci EL: Circulating levels of IGF-1, IGFBP-3, and IGF-1/IGFBP-3 molar ratio and colorectal adenomas: A meta-analysis. Cancer Epidemiol. 39:1026–1035. 2015. View Article : Google Scholar : PubMed/NCBI

4 

Woo HD, Kim K and Kim J: Association between preoperative C-reactive protein level and colorectal cancer survival: A meta-analysis. Cancer Causes Control. 26:1661–1670. 2015. View Article : Google Scholar : PubMed/NCBI

5 

Taniguchi N and Kizuka Y: Glycans and cancer: Role of N-glycans in cancer biomarker, progression and metastasis, and therapeutics. Adv Cancer Res. 126:11–51. 2015. View Article : Google Scholar : PubMed/NCBI

6 

Meany DL and Chan DW: Aberrant glycosylation associated with enzymes as cancer biomarkers. Clin Proteomics. 8:72011. View Article : Google Scholar : PubMed/NCBI

7 

Brooks SA: The involvement of Helix pomatia lectin (HPA) binding N-acetylgalactosamine glycans in cancer progression. Histol Histopathol. 15:143–158. 2000.PubMed/NCBI

8 

Hollingsworth MA and Swanson BJ: Mucins in cancer: Protection and control of the cell surface. Nat Rev Cancer. 4:45–60. 2004. View Article : Google Scholar

9 

Haugstad KE, Gerken TA, Stokke BT, Dam TK, Brewer CF and Sletmoen M: Enhanced self-association of mucins possessing the T and Tn carbohydrate cancer antigens at the single-molecule level. Biomacromolecules. 13:1400–1409. 2012. View Article : Google Scholar : PubMed/NCBI

10 

Conze T, Carvalho AS, Landegren U, Almeida R, Reis CA, David L and Söderberg O: MUC2 mucin is a major carrier of the cancer-associated sialyl-Tn antigen in intestinal metaplasia and gastric carcinomas. Glycobiology. 20:199–206. 2010. View Article : Google Scholar

11 

Haugstad KE, Stokke BT, Brewer CF, Gerken TA and Sletmoen M: Single molecule study of heterotypic interactions between mucins possessing the Tn cancer antigen. Glycobiology. 25:524–534. 2015. View Article : Google Scholar :

12 

Julien S, Videira PA and Delannoy P: Sialyl-tn in cancer: (how) did we miss the target? Biomolecules. 2:435–466. 2012. View Article : Google Scholar : PubMed/NCBI

13 

Sousa BL, Silva Filho JC, Kumar P, Pereira RI, Łyskowski A, Rocha BA, Delatorre P, Bezerra GA, Nagano CS, Gruber K, et al: High-resolution structure of a new Tn antigen-binding lectin from Vatairea macrocarpa and a comparative analysis of Tn-binding legume lectins. Int J Biochem Cell Biol. 59:103–110. 2015. View Article : Google Scholar

14 

Mazal D, Lo-Man R, Bay S, Pritsch O, Dériaud E, Ganneau C, Medeiros A, Ubillos L, Obal G, Berois N, et al: Monoclonal antibodies toward different Tn-amino acid backbones display distinct recognition patterns on human cancer cells. Implications for effective immuno-targeting of cancer. Cancer Immunol Immunother. 62:1107–1122. 2013. View Article : Google Scholar : PubMed/NCBI

15 

Barrow H, Tam B, Duckworth CA, Rhodes JM and Yu LG: Suppression of core 1 Gal-transferase is associated with reduction of TF and reciprocal increase of Tn, sialyl-Tn and Core 3 glycans in human colon cancer cells. PLoS One. 8:e597922013. View Article : Google Scholar : PubMed/NCBI

16 

Holmberg LA, Oparin DV, Gooley T, Lilleby K, Bensinger W, Reddish MA, MacLean GD, Longenecker BM and Sandmaier BM: Clinical outcome of breast and ovarian cancer patients treated with high-dose chemotherapy, autologous stem cell rescue and THERATOPE STn-KLH cancer vaccine. Bone Marrow Transplant. 25:1233–1241. 2000. View Article : Google Scholar : PubMed/NCBI

17 

Matsumoto Y, Zhang Q, Akita K, Nakada H, Hamamura K, Tsuchida A, Okajima T and Furukawa K, Urano T and Furukawa K: Trimeric Tn antigen on syndecan 1 produced by ppGalNAc-T13 enhances cancer metastasis via a complex formation with integrin α5β1 and matrix metalloproteinase 9. J Biol Chem. 288:24264–24276. 2013. View Article : Google Scholar : PubMed/NCBI

18 

Song K, Herzog BH, Fu J, Sheng M, Bergstrom K, McDaniel JM, Kondo Y, McGee S, Cai X, Li P, et al: Loss of core 1-derived O-glycans decreases breast cancer development in mice. J Biol Chem. 290:20159–20166. 2015. View Article : Google Scholar : PubMed/NCBI

19 

Hofmann BT, Schlüter L, Lange P, Mercanoglu B, Ewald F, Fölster A, Picksak AS, Harder S, El Gammal AT, Grupp K, et al: COSMC knockdown mediated aberrant O-glycosylation promotes oncogenic properties in pancreatic cancer. Mol Cancer. 14:1092015. View Article : Google Scholar : PubMed/NCBI

20 

Loureiro LR, Carrascal MA, Barbas A, Ramalho JS, Novo C, Delannoy P and Videira PA: Challenges in antibody development against Tn and sialyl-Tn antigens. Biomolecules. 5:1783–1809. 2015. View Article : Google Scholar : PubMed/NCBI

21 

Tavares C, de Oliveira JT, Lopes C, Carvalheira J, Matos AJ, Rutteman G, Reis CA and Gärtner F: Mucin 6 and Tn antigen expression in canine mammary tumours: Correlation with pathological features. J Comp Pathol. 147:410–418. 2012. View Article : Google Scholar : PubMed/NCBI

22 

Lescar J, Sanchez JF, Audfray A, Coll JL, Breton C, Mitchell EP and Imberty A: Structural basis for recognition of breast and colon cancer epitopes Tn antigen and Forssman disaccharide by Helix pomatia lectin. Glycobiology. 17:1077–1083. 2007. View Article : Google Scholar : PubMed/NCBI

23 

Zamri N, Masuda N, Oura F, Yajima Y, Nakada H and Fujita-Yamaguchi Y: Effects of two monoclonal antibodies, MLS128 against Tn-antigen and 1H7 against insulin-like growth factor-I receptor, on the growth of colon cancer cells. Biosci Trends. 6:303–312. 2012.

24 

Ju T, Aryal RP, Kudelka MR, Wang Y and Cummings RD: The Cosmc connection to the Tn antigen in cancer. Cancer biomarkers: Section. Dis Markers. 14:63–81. 2014.

25 

Takamiya R, Ohtsubo K, Takamatsu S, Taniguchi N and Angata T: The interaction between Siglec-15 and tumor-associated sialyl-Tn antigen enhances TGF-β secretion from monocytes/macrophages through the DAP12-Syk pathway. Glycobiology. 23:178–187. 2013. View Article : Google Scholar

26 

Berriel E, Hill M, Barcia JJ, Ubillos L, Gonzalez M, Detjen G, Rondan M, Navarrete H and Osinaga E: Simple mucin-type cancer associated antigens are pre-cancerous biomarkers during 1,2-dimethylhydrazine-induced rat colon carcinogenesis. Oncol Rep. 14:219–227. 2005.PubMed/NCBI

27 

Troup S, Njue C, Kliewer EV, Parisien M, Roskelley C, Chakravarti S, Roughley PJ, Murphy LC and Watson PH: Reduced expression of the small leucine-rich proteoglycans, lumican, and decorin is associated with poor outcome in node-negative invasive breast cancer. Clin Cancer Res. 9:207–214. 2003.PubMed/NCBI

28 

Oura F, Yajima Y, Nakata M, Taniue K, Akiyama T, Nakada H, Yamamoto K and Fujita-Yamaguchi Y: Susceptibility to proteases of anti-Tn-antigen MLS128 binding glycoproteins expressed in human colon cancer cells. Biosci Trends. 9:49–55. 2015. View Article : Google Scholar : PubMed/NCBI

29 

Yu X, Du Z, Sun X, Shi C, Zhang H and Hu T: Aberrant Cosmc genes result in Tn antigen expression in human colorectal carcinoma cell line HT-29. Int J Clin Exp Pathol. 8:2590–2602. 2015.PubMed/NCBI

30 

Lin WH, Huang CJ, Liu MW, Chang HM, Chen YJ, Tai TY and Chuang LM: Cloning, mapping, and characterization of the human sorbin and SH3 domain containing 1 (SORBS1) gene: A protein associated with c-Abl during insulin signaling in the hepatoma cell line Hep3B. Genomics. 74:12–20. 2001. View Article : Google Scholar : PubMed/NCBI

31 

Farace C, Oliver JA, Melguizo C, Alvarez P, Bandiera P, Rama AR, Malaguarnera G, Ortiz R, Madeddu R and Prados J: Microenvironmental modulation of Decorin and Lumican in temozolomide-resistant glioblastoma and neuroblastoma cancer stem-like cells. PLoS One. 10:e01341112015. View Article : Google Scholar : PubMed/NCBI

32 

Kioka N, Ueda K and Amachi T: Vinexin, CAP/ponsin, ArgBP2: A novel adaptor protein family regulating cytoskeletal organization and signal transduction. Cell Struct Funct. 27:1–7. 2002. View Article : Google Scholar : PubMed/NCBI

33 

Zhang M, Kimura A and Saltiel AR: Cloning and characterization of Cbl-associated protein splicing isoforms. Mol Med. 9:18–25. 2003.PubMed/NCBI

34 

Lewis JM, Baskaran R, Taagepera S, Schwartz MA and Wang JY: Integrin regulation of c-Abl tyrosine kinase activity and cytoplasmic-nuclear transport. Proc Natl Acad Sci USA. 93:15174–15179. 1996. View Article : Google Scholar : PubMed/NCBI

35 

Zhang M, Liu J, Cheng A, Deyoung SM, Chen X, Dold LH and Saltiel AR: CAP interacts with cytoskeletal proteins and regulates adhesion-mediated ERK activation and motility. EMBO J. 25:5284–5293. 2006. View Article : Google Scholar : PubMed/NCBI

36 

Kostic C and Shaw PH: Isolation and characterization of sixteen novel p53 response genes. Oncogene. 19:3978–3987. 2000. View Article : Google Scholar : PubMed/NCBI

37 

Mandai K, Nakanishi H, Satoh A, Takahashi K, Satoh K, Nishioka H, Mizoguchi A and Takai Y: Ponsin/SH3P12: An l-afadin- and vinculin-binding protein localized at cell-cell and cell-matrix adherens junctions. J Cell Biol. 144:1001–1017. 1999. View Article : Google Scholar : PubMed/NCBI

38 

Hyakuna N, Muramatsu H, Higa T, Chinen Y, Wang X and Kojima S: Germline mutation of CBL is associated with moyamoya disease in a child with juvenile myelomonocytic leukemia and Noonan syndrome-like disorder. Pediatr Blood Cancer. 62:542–544. 2015. View Article : Google Scholar

39 

Michalik L, Desvergne B and Wahli W: Peroxisome-proliferator-activated receptors and cancers: Complex stories. Nat Rev Cancer. 4:61–70. 2004. View Article : Google Scholar : PubMed/NCBI

40 

Bi X, Pohl NM, Qian Z, Yang GR, Gou Y, Guzman G, Kajdacsy-Balla A, Iozzo RV and Yang W: Decorin-mediated inhibition of colorectal cancer growth and migration is associated with E-cadherin in vitro and in mice. Carcinogenesis. 33:326–330. 2012. View Article : Google Scholar :

41 

Campioni M, Ambrogi V, Pompeo E, Citro G, Castelli M, Spugnini EP, Gatti A, Cardelli P, Lorenzon L, Baldi A, et al: Identification of genes down-regulated during lung cancer progression: A cDNA array study. J Exp Clin Cancer Res. 27:382008. View Article : Google Scholar : PubMed/NCBI

42 

Goldoni S, Seidler DG, Heath J, Fassan M, Baffa R, Thakur ML, Owens RT, McQuillan DJ and Iozzo RV: An antimetastatic role for decorin in breast cancer. Am J Pathol. 173:844–855. 2008. View Article : Google Scholar : PubMed/NCBI

43 

Reed CC, Waterhouse A, Kirby S, Kay P, Owens RT, McQuillan DJ and Iozzo RV: Decorin prevents metastatic spreading of breast cancer. Oncogene. 24:1104–1110. 2005. View Article : Google Scholar : PubMed/NCBI

44 

Li J, Liang H, Bai M, Ning T, Wang C, Fan Q, Wang Y, Fu Z, Wang N, Liu R, et al: miR-135b promotes cancer progression by targeting transforming growth factor beta receptor II (TGFBR2) in colorectal cancer. PLoS One. 10:e01301942015. View Article : Google Scholar : PubMed/NCBI

45 

Biswas S, Guix M, Rinehart C, Dugger TC, Chytil A, Moses HL, Freeman ML and Arteaga CL: Inhibition of TGF-beta with neutralizing antibodies prevents radiation-induced acceleration of metastatic cancer progression. J Clin Invest. 117:1305–1313. 2007. View Article : Google Scholar : PubMed/NCBI

46 

de Miranda NF, van Dinther M, van den Akker BE, van Wezel T, ten Dijke P and Morreau H: Transforming growth factor β signaling in colorectal cancer cells with microsatellite instability despite biallelic mutations in TGFBR2. Gastroenterology. 148:1427–1437. 2015. View Article : Google Scholar

47 

Nischalke HD, Lutz P, Krämer B, Söhne J, Müller T, Rosendahl J, Fischer J, Berg T, Hittatiya K, Fischer HP, et al: A common polymorphism in the NCAN gene is associated with hepatocellular carcinoma in alcoholic liver disease. J Hepatol. 61:1073–1079. 2014. View Article : Google Scholar : PubMed/NCBI

48 

Kim SY, Choi EJ, Yun JA, Jung ES, Oh ST, Kim JG, Kang WK and Lee SH: Syndecan-1 expression is associated with tumor size and EGFR expression in colorectal carcinoma: A clinicopathological study of 230 cases. Int J Med Sci. 12:92–99. 2015. View Article : Google Scholar : PubMed/NCBI

49 

Zaragosi LE, Dadone B, Michiels JF, Marty M, Pedeutour F, Dani C and Bianchini L: Syndecan-1 regulates adipogenesis: New insights in dedifferentiated liposarcoma tumorigenesis. Carcinogenesis. 36:32–40. 2015. View Article : Google Scholar

50 

Mundt F, Heidari-Hamedani G, Nilsonne G, Metintas M, Hjerpe A and Dobra K: Diagnostic and prognostic value of soluble syndecan-1 in pleural malignancies. BioMed Res Int. 2014:4198532014. View Article : Google Scholar : PubMed/NCBI

51 

Galvagni F, Orlandini M and Oliviero S: Role of the AP-1 transcription factor FOSL1 in endothelial cells adhesion and migration. Cell Adhes Migr. 7:408–411. 2013. View Article : Google Scholar

52 

Zhu W, Li J, Su J, Li J, Li J, Deng B, Shi Q, Zhou Y and Chen X: FOS-like antigen 1 is highly expressed in human psoriasis tissues and promotes the growth of HaCaT cells in vitro. Mol Med Rep. 10:2489–2494. 2014.PubMed/NCBI

53 

Liang X, Liu Y, Zeng L, Yu C, Hu Z, Zhou Q and Yang Z: miR-101 inhibits the G1-to-S phase transition of cervical cancer cells by targeting Fos. Int J Gynecol Cancer. 24:1165–1172. 2014. View Article : Google Scholar : PubMed/NCBI

Related Articles

Journal Cover

September 2016
Volume 36 Issue 3

Print ISSN: 1021-335X
Online ISSN:1791-2431

Sign up for eToc alerts

Recommend to Library

Copy and paste a formatted citation
APA
Wei, H., Cheng, Z., Ouyang, C., Zhang, Y., Hu, Y., Chen, S. ... Liu, X. (2016). Glycoprotein screening in colorectal cancer based on differentially expressed Tn antigen. Oncology Reports, 36, 1313-1324. https://doi.org/10.3892/or.2016.4937
MLA
Wei, H., Cheng, Z., Ouyang, C., Zhang, Y., Hu, Y., Chen, S., Wang, C., Lu, F., Zhang, J., Wang, Y., Liu, X."Glycoprotein screening in colorectal cancer based on differentially expressed Tn antigen". Oncology Reports 36.3 (2016): 1313-1324.
Chicago
Wei, H., Cheng, Z., Ouyang, C., Zhang, Y., Hu, Y., Chen, S., Wang, C., Lu, F., Zhang, J., Wang, Y., Liu, X."Glycoprotein screening in colorectal cancer based on differentially expressed Tn antigen". Oncology Reports 36, no. 3 (2016): 1313-1324. https://doi.org/10.3892/or.2016.4937