OL Oncology Letters 1792-1074 1792-1082 D.A. Spandidos 10.3892/ol.2016.4160 OL-0-0-4160 Articles AGR2 is associated with gastric cancer progression and poor survival ZHANGJUN 1 * JINYONGMING 1 * XUSHAONAN 1 ZHENGJIAYIN 2 ZHANGQI 3 WANGYUANYU 3 CHENJINPING 1 HUANGYAZENG 1 HEXUJUN 4 ZHAOZHONGSHENG 5 Department of Orthopaedics, Zhejiang Provincial People's Hospital, Hangzhou, Zhejiang 310014, P.R. China Department of Probability and Statistics, School of Mathematical Sciences, Peking University, Beijing 100871, P.R. China Department of Surgery, Zhejiang Provincial People's Hospital, Hangzhou, Zhejiang 310014, P.R. China Key Laboratory of Gastroenterology of Zhejiang, Hangzhou, Zhejiang 310014, P.R. China Department of Pathology, Zhejiang Provincial People's Hospital, Hangzhou, Zhejiang 310014, P.R. China Correspondence to: Dr Zhongsheng Zhao, Department of Pathology, Zhejiang Provincial People's Hospital, 158 Shangtang Road, Hangzhou, Zhejiang 310014, P.R. China, E-mail: zhaozhongsheng1950@163.com

Contributed equally

 03 2016 29 01 2016 11 3 2075 2083 03082014 05112015 Copyright: © Zhang et al. 2016 This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

Anterior gradient protein 2 (AGR2) has been reported as a novel biomarker with a potential oncogenic role. However, its association with the prognosis and survival rate of gastric cancer (GC) has not yet been determined. Therefore, the present study aimed to examine the expression and prognostic significance of AGR2 in patients with GC. Immunohistochemistry was used to analyze AGR2 and cathepsin D (CTSD) protein expression in 436 clinicopathologically characterized GC cases and 92 noncancerous tissue samples. AGR2 and CTSD expression were both elevated in GC lesions compared with noncancerous tissues. In 204/436 (46.8%) GC patients, high expression of AGR2 was positively correlated with the expression of CTSD (r=0.577, P<0.01). Furthermore, several clinicopathological parameters were significantly associated with AGR2 expression level, including tumor size, depth of invasion and TNM stage (P<0.05). Using Kaplan-Meier survival analysis, it was determined that the mean survival time of patients with low levels of AGR2 expression was significantly longer than those with high ARG2 expression (in stages I, II and III; P<0.05). For stage IV disease, no significant difference in survival time was identified. Multivariate survival analysis demonstrated that AGR2 was an independent prognostic factor and was associated in the progression of GC. The findings of the present study indicate that AGR2 expression is significantly associated with location and size of GC, depth of invasion, TNM stage, lymphatic metastasis, vessel invasion, distant metastasis, Lauren's classification, high CTSD expression and poor prognosis. Thus, AGR2 may be a novel GC marker and may present a potential therapeutic target for GC.

 anterior gradient protein 2 cathepsin D gastric cancer immunohistochemistry progression poor survival Introduction

Despite continuously improving therapies, gastric cancer (GC) still has the second highest mortality rate of all tumor cases in China, with a 5-year survival rate of ~20% (1). The incidence of GC is ~934,000, with 41% of new cases diagnosed in China (2). In 2008, GC was ranked second in men and fourth in women in terms of incidence, and the second in men and third in women for mortality in China (3). Due to its asymptomatic character and lack of specific symptoms in the early stages, GC is often diagnosed in the later stage of the disease (III or IV) with high rates of lymph node metastasis (4). This leads to its poor prognosis. Invasion and metastasis are regulated at multiple molecular levels and by angiogenesis factors (5). Considering that GC is, at present, a largely incurable malignant disease, improved understanding of cancer cells is essential for the development of novel detection and therapeutic strategies.

Anterior gradient 2 (AGR2) protein is elevated in numerous types of cancer, including breast (6,7), lung (8), ovarian (9), esophageal (10), prostate (11) and pancreatic cancer (12), and was reported to be associated the metastatic phenotype and poor prognosis of breast cancer (6). However, the association between AGR2 expression, and prognosis and survival in GC remains largely unknown.

Cathepsin D (CTSD) is a common aspartic lysosomal endopeptidase. Its overexpression is positively associated with gastric carcinoma (13–15), melanoma (16), ovarian cancer (17) and colorectal cancer (18). CTSD levels were reported to be higher in tumors compared with in adjacent noncancerous tissue in colorectal cancer (19,20). Furthermore, CTSD was reported to degrade and remodel the basement membrane and interstitial stroma surrounding primary breast cancer tumors (21), and stimulate apoptotic caspases or cooperate with tumor associated pathogenic lysosomal cysteine cathepsins (22). In addition, AGR2 was reported to promote in vitro and in vivo dissemination of cancer cells through posttranscriptional induction of two proteases, cathepsin B and CTSD (12).

The present study was conducted to investigate the expression of AGR2, and its association with the progression and prognosis of patients with GC. The potential association between AGR2 and CTSD expression in GC cancer progression was also analyzed. AGR2 expression was detected and positively associated with the CTSD expression level and specific clinicopathological parameters of patients with GC.

 Patients and methods <sec> <title>Patient collection and sample preparation

The procedures of patient collection and sample preparation were described previously (23,24). There were a total of 528 samples (436 cancer samples and 92 adjacent noncancerous tissue samples) collected during gastrectomies performed at the Zhejiang Provincial People's Hospital (Hangzhou, China).

 Construction of tissue microarray

Sample preparation was performed as previously described (25,26). Briefly, core tissue biopsies (2 mm in diameter) were obtained/sampled from each individual paraffin-embedded gastric tumor sample (donor blocks) and arranged in recipient paraffin blocks (tissue array blocks) using a trephine, as a previous study indicated that staining results obtained from different intratumoral areas in various tumors correlate well (27). Cases in which the tumor occupied >10% of the core area were selected for further investigation (28). Each block contained >3 internal controls consisting of nonneoplastic gastric mucosa. Sections (4-µm thick) were cut from each tissue array block, deparaffinized and dehydrated.

The project was approved by the Ethics Committee of Zhejiang Provincial People's Hospital and written consent was obtained from all participants.

 Immunohistochemistry (IHC)

IHC was performed for detecting AGR2 and CTSD in 528 tissues (with 92 control samples and 436 GC samples) (29,30). The procedures were performed using an EnVision kit (K4011 HRP, Rabbit (DAB+); Dako, Glostrup, Denmark) in accordance with previous studies (23,24). In brief, the slides were baked overnight at 60°C, followed by deparaffinization with xylene and rehydrated in graded alcohol. The sections were submerged into EDTA (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) and microwaved for 10 min for antigenic retrieval. Subsequently, 3% hydrogen peroxide (Invitrogen; Thermo Fisher Scientific, Inc.) in methanol was used to quench endogenous peroxidase activity, followed by incubation with 1% bovine serum albumin (Pierce Biotechnology, Inc., Rockford, IL, USA) to block nonspecific binding. Sections were incubated with anti-human rabbit monoclonal anti-AGR2 (cat. no. 2533-1; 1:500) and anti-CTSD antibodies (cat. no. 2487-1; 1:750) (Epitomics, Inc., Burlingame, CA, USA) overnight at 4°C. Normal goat serum (10000C; Invitrogen; Thermo Fisher Scientific, Inc.) was used as a negative control. Subsequent to rinsing with phosphate buffer (pH=7.2) three times, the slides were incubated with the secondary antibody (EnVision kit; Dako) for 20 min at room temperature and stained with diaminobenzidine (Vector Laboratories, Inc., Burlingame, CA, USA). All the slides were counterstained with hematoxylin (Vector Laboratories, Inc.), dehydrated and mounted with a coverslip using a standard medium. The slides were visualized using the Axioskop 40 microscope (Carl Zeiss AG, Oberkochen, Germany Two independent observers, who were blinded to the study design, were employed to review the results and score all of the samples, as previously described (23). The staining intensity was scored as 0 (no staining), 1 (weak staining, light yellow), 2 (moderate staining, yellow brown) and 3 (strong staining, brown), and the proportion of stained tumor cells was classified as 0 (≤5% positive cells), 1 (6–25% positive cells), 2 (26–50% positive cells) and 3 (≥51% positive cells). The expression of both proteins was considered low if the product of the staining intensity and proportion of stained tumor cells scores was ≤3 and high if the product was ≥4.

 Statistical analysis

Data analysis was conducted with SPSS software (version 16.0; SPSS, Inc., Chicago, IL, USA). Measurement data were analyzed using the Student's t-test, whereas χ2 or Fisher's exact tests were used to examine the correlation between AGR2 and CTSD expression, and their clinicopathological parameters. Furthermore, Spearman's rank correlation tests were used to analyze the association between AGR2 and CTSD expression. Comparisons between survival curves, which were calculated using the Kaplan-Meier method, were performed by univariate survival analysis using the log-rank test. The prognostic value of AGR2 and CTSD expression were assessed by stepwise multivariate analysis using the Cox proportional hazards regression model. In addition, correlation coefficients between protein expression levels and clinicopathological findings were estimated using the Pearson correlation method. Variables that were significant in the univariate analysis were included in the model with backward Cox regression (the Wald method). P<0.05 was considered to indicate a statistically significant difference.

 Results <sec> <title>AGR2 and CTSD expression are significantly higher in GC compared with adjacent noncancerous tissue

AGR2 was detected by IHC in 228/436 (52.3%) cases of GC, with 204 (46.8%) classified as high expression. By contrast, only 29 (31.5%) cases exhibited AGR2 expression in the noncancerous control group, all of which were classified as low expression. The IHC results indicated that AGR2 was predominantly located in the cytoplasm of GC cells (Fig. 1). Its expression was significantly higher in GC samples compared with control samples (P<0.05; data not shown). As expected, expression of CTSD was detected in 215/436 (49.3%) GC samples, with 138 (31.7%) exhibiting high expression. Only 27 (29.3%) cases of AGR2 expression were detected in the control group, all of which were low expression. CTSD was predominantly distributed in the cytoplasm in a similar manner to AGR2 (Fig. 2). Its expression was significantly higher in GC samples compared with the noncancerous control samples (P<0.05; data not shown).

 High AGR2 expression and CTSD expression are correlated in gastric cancer

High coincidental expression of AGR2 and CTSD in GC samples. Of the 204 patients with high expression of AGR2, 152 (74.5%) exhibited high expression of CTSD. The correlation was statistically significant (r=0.577, P<0.01).

 AGR2 and CTSD are associated with clinicopathological parameters

The Pearson correlation method was used to determine the association between AGR2 and CTSD expression with clinicopathological parameters of patients with GC. The results indicated that AGR2 was significantly associated with location of the tumor, tumor size, depth of invasion, TNM stage, Lauren's classification, vessel invasion, lymphatic metastasis, regional lymph nodes and distant metastasis of tumor (P<0.05; Table I); however, no significant correlation was identified with gender, age, grade of differentiation and histological type of the tumor (P>0.05; Table I). Furthermore, GC patients with deep tumor invasion (T3 and T4), high TNM stage (stages III and IV), vessel invasion, lymph node metastasis and distant metastasis exhibited significantly higher expression of AGR2 compared to those with superficial tumor invasion (T1 and T2), low TNM stage (stages I and II), and no vessel invasion, lymph node metastasis or distant metastasis (Table I). The associations between the clinicopathological parameters and CTSD expression were consistent with those of AGR2.

 AGR2 and CTSD expression are associated with prognosis

GC patients with low AGR2 expression had a significantly longer mean survival time (52.9 months) compared with those patients exhibiting high AGR2 expression (32.4 months). In agreement with this result, the 3- and 5-year cumulative survival rates were 90.9 and 57.3% versus 36.9 and 5.7%, respectively. These results indicated that high expression of AGR2 was correlated with significantly poorer prognosis compared with those exhibiting low AGR2 expression (P<0.05). Furthermore, the results of CTSD were in line with AGR2 regarding to mean survival time and 3- and 5-year cumulative survival rate. In addition, factors significantly associated with survival were assessed using univariate analysis, and it was found that age, tumor size, location, depth of invasion, TNM stage, Lauren's classification, vessel invasion, and lymph node and distant metastasis were significantly related to the prognosis while histological type and grade of differentiation were not (Table II). After stratifying by TNM stage, we found that low expression of AGR2 was significantly related longer mean survival time only in stage I, II and III. In particular, there was no significant difference in survival times between low and high expression AGR2 in stage IV (Fig. 3). Multivariate analysis was employed to further determine the correlation of the clinicopathological parameters identified by univariate analysis with the survival of GC patients. The results of Cox regression model indicated that depth of invasion, vessel invasion, lymph node metastasis, distant metastasis, Lauren's classification, and AGR2 expression were independent prognostic factors, whereas age, location and size of tumor, TNM stage and CTSD expression were not (Table III).

 Discussion

As a p52 suppressor inhibitor, AGR2 has been widely investigated in several types of human carcinogenesis (6–12); however, its exact functions and regulation have been largely unclear. In the present study, microarray tissue samples were initially used to evaluate the protein expression of AGR2 and CTSD in GC patients and its prognostic implications. Increased AGR2 expression in GC tissues was detected compared with adjacent noncancerous tissue. Significant associations were identified between AGR2 and location and size of tumor, TNM stage, depth of invasion, vessel invasion, lymph node and distant metastasis and Lauren's classification. Patients in late TNM stages (III and IV), with deep invasion (T3 and T4), presence of vessel invasion, and lymph node and distant metastasis exhibited the highest level of AGR2. These findings indicate that upregulation of AGR2 was involved in the progression of GC.

AGR2 is known as a stimulator of cancer cell proliferation, invasion and survival, chemotherapy resistance, metastasis and tumor growth (6,7,9,11,12). Secretion of AGR2 was been reported to correlate with metastasis and poor prognosis in breast cancer, and considered as a biomarker in prostate cancer (31–33). AGR2 upregulation was also detected in pancreatic carcinoma tissues (34,35). In GC tissues, higher expression in GC cells has previously been reported to be evident in the cytoplasm compared with non-tumor cells (36). Notably, AGR2 can be used as a suitable candidate gene for the detection of circulating tumor cells, a novel resource to identification of molecular markers, in patients with gastrointestinal cancer (37,38).

To the best of our knowledge, prognostic factors of GC includes invasion depth, TNM stage, and lymph node and distant metastasis (39). In the present study, AGR2 was identified as a novel independent prognostic factor that was significantly associated with poor prognosis in patients with GC. The current study also revealed that low expression of AGR2 was significantly associated with longer mean survival time in TNM stages I, II and III.

AGR2 shares structural characteristics with the protein disulfide isomerase (PDI) family. The PDI family has important roles on the cell surface, as the majority of surface proteins contain disulfide bonds in which they can modulate the activity of membrane receptors (and thus activate and regulate signaling pathways), adhesion molecules integrins, or even proteases, such as ADAM metallopeptidase domain 17 (40–45).

The current findings also indicated significant positive correlation between the expression of AGR2 and CTSD in GC tissues. In agreement with AGR2 expression, CTSD, which is known as an aspartic lysosomal endopeptidase, was significantly correlated with location and size of tumor, depth of invasion, vessel invasion, TNM stage, distant metastasis, lymph node metastasis, regional lymph node and Lauren's classification in the present study. Overexpression of CTSD has previously been reported in several types of human cancer, including GC (13–15), melanoma (16) and ovarian cancer (17). It may have a direct role in promoting tumor growth by basement membrane and interstitial stroma degradation and remodeling (21). Furthermore, CTSD is able to stimulate other enzymes and cooperate with certain cathepsins in the proteolysis process (22). CTSD has been reported to be upregulated by AGR2 in pancreatic cancer (12). The underlined mechanism may be the direct effect of AGR2 PDI activity in the ER during the processing of pro-cathepsins, as previously reported for the production of MUC2 in enterocytes (46).

Taken together, the present study indicated that upregulation of AGR2 may contribute to the expression of CTSD as the direct result of AGR2 PDI activity in the ER. The cross-talk of AGR2 and CTSD is possibly involved in the carcinogenesis, development and progression of GC.

The present study still has a number of limitations. First, the subjective nature of the scoring method could not be avoided and further investigation is recommended to evaluate the reproducibility of IHC scores system. Furthermore, standardization and quality control for IHC procedures is required prior to its clinical application (47).

 Acknowledgements

The authors thank Dr Qinchuan Wang, Dr Jie Ma, Dr Yingyu Ma and Dr Yingjie Xia for their help and discussion with immunohistochemistry staining and statistical analysis. This study was supported by grants from the Zhejiang Provincial Department of Science and Technology Research Foundation (grant nos. 2008C33040 and 2012C33069), the Natural Scientific Research Foundation of Zhejiang Province (grant no. LY14H060006) and the Natural Scientific Research Foundation of Zhejiang Medical College (grant no. 2011XZB01).

 References MurrayCJLopezADAlternative projections of mortality and disability by cause 1990–2020: Global Burden of Disease StudyLancet34914981504199710.1016/S0140-6736(96)07492-29167458 InoueMTsuganeSEpidemiology of gastric cancer in JapanPostgrad Med J81419424200510.1136/pgmj.2004.02933015998815 Shanghai Municipal Center for Disease Control and PreventionShanghai Cancer Report 2010ShanghaiSCDC2010 WHZFHHSurgical therapy of gastric cancer in chinaJ Pract Oncol2391932008 BogenriederTHerlynMAxis of evil: Molecular mechanisms of cancer metastasisOncogene2265246536200310.1038/sj.onc.120675714528277 BarracloughDLPlatt-HigginsAde Silva RudlandSBarracloughRWinstanleyJWestCRRudlandPSThe metastasis-associated anterior gradient 2 protein is correlated with poor survival of breast cancer patientsAm J Pathol17518481857200910.2353/ajpath.2009.09024619834055 LiuDRudlandPSSibsonDRPlatt-HigginsABarracloughRHuman homologue of cement gland protein, a novel metastasis inducer associated with breast carcinomasCancer Res6537963805200510.1158/0008-5472.CAN-04-382315867376 FritzscheFRDahlEDankofABurkhardtMPahlSPetersenIDietelMKristiansenGExpression of AGR2 in non small cell lung cancerHistol Histopathol22703708200717455144 ParkKChungYJSoHKimKParkJOhMJoMChoiKLeeEJChoiYLAGR2, a mucinous ovarian cancer marker, promotes cell proliferation and migrationExp Mol Med4391100201110.3858/emm.2011.43.2.01121200134 PohlerECraigALCottonJLawrieLDillonJFRossPKernohanNHuppTRThe Barrett's antigen anterior gradient-2 silences the p53 transcriptional response to DNA damageMol Cell Proteomics3534547200410.1074/mcp.M300089-MCP20014967811 ZhangJSGongAChevilleJCSmithDIYoungCYAGR2, an androgen-inducible secretory protein overexpressed in prostate cancerGenes Chromosomes Cancer43249259200510.1002/gcc.2018815834940 DumartinLWhitemanHJWeeksMEHariharanDDmitrovicBIacobuzio-DonahueCABrentnallTABronnerMPFeakinsRMTimmsJFAGR2 is a novel surface antigen that promotes the dissemination of pancreatic cancer cells through regulation of cathepsins B and DCancer Res7170917102201110.1158/0008-5472.CAN-11-136721948970 AllgayerHBabicRGrütznerKUBeyerBCTarabichiASchildbergWilhelm FHeissMMAn immunohistochemical assessment of cathepsin D in gastric carcinoma: its impact on clinical prognosisCancer80179187199710.1002/(SICI)1097-0142(19970715)80:2<179::AID-CNCR2>3.0.CO;2-P9217027 Del Manuel CasarJVizosoFJAbdel-LaaOSanzLMartínACorteDaniela MBongeraMMuñizGarcía JLFueyoAPrognostic value of cytosolyc cathepsin D content in resectable gastric cancerJ Surg Oncol861621200410.1002/jso.2004115048675 SakuTSakaiHTsudaNOkabeHKatoYYamamotoKCathepsins D and E in normal, metaplastic, dysplastic and carcinomatous gastric tissue: An immunohistochemical studyGut3112501255199010.1136/gut.31.11.12502253908 BartenjevIRudolfZStabucBVrhovecIPerkovicTKanskyACathepsin D expression in early cutaneous malignant melanomaInt J Dermatol39599602200010.1046/j.1365-4362.2000.00025.x10971728 LöschASchindlMKohlbergerPCathepsin D in ovarian cancer: Prognostic value and correlation with p53 expression and microvessel densityGynecol Oncol92545552200410.1016/j.ygyno.2003.11.01614766246 KiranaCShiHLaingECathepsin D expression in colorectal cancer: From proteomic discovery through validation using western blotting, immunohistochemistry and tissue microarraysInt J Proteomics2012245819201210.1155/2012/24581922919486 AdenisAHuetGZerimechFHecquetBBalduyckMPeyratJPCathepsin B, L and D activities in colorectal carcinomas: Relationship with clinico-pathological parametersCancer Lett96267275199510.1016/0304-3835(95)03930-U7585467 SzajdaSDSnarskaJJankowskaARoszkowska-JakimiecWPuchalskiZZwierzKCathepsin D and carcino-embryonic antigen in serum, urine and tissues of colon adenocarcinoma patientsHepatogastroenterology55388393200818613372 BerchemGGlonduMGleizesMBrouilletJPVignonFGarciaMLiaudet-CoopmanECathepsin-D affects multiple tumor progression steps in vivo: Proliferation, angiogenesis and apoptosisOncogene2159515955200210.1038/sj.onc.120574512185597 KrepelaECysteine proteinases in tumor cell growth and apoptosisNeoplasma48332349200111845977 ZhaoZSWangYYChuYQYeZYTaoHQSPARC is associated with gastric cancer progression and poor survival of patientsClin Cancer Res16260268201010.1158/1078-0432.CCR-09-124720028745 ShouZXJinXZhaoZSUpregulated expression of ADAM17 is a prognostic marker for patients with gastric cancerAnn Surg25610141022201210.1097/SLA.0b013e3182592f5622668812 LeeHSLeeHKKimHSYangHKKimWHTumour suppressor gene expression correlates with gastric cancer prognosisJ Pathol2003946200310.1002/path.128812692839 LeeHSLeeHKKimHSYangHKKimYIKimWHMUC1, MUC2, MUC5AC and MUC6 expressions in gastric carcinomas: Their roles as prognostic indicatorsCancer9214271434200110.1002/1097-0142(20010915)92:6<1427::AID-CNCR1466>3.0.CO;2-L11745219 ZhangDSalto-TellezMPuttiTCDoEKoayESReliability of tissue microarrays in detecting protein expression and gene amplification in breast cancerMod Pathol167984200310.1097/01.MP.0000047307.96344.9312527717 LeeHSChoSBLeeHEKimMAKimJHParkJKimJHYangHKLeeBLKimWHProtein expression profiling and molecular classification of gastric cancer by the tissue array methodClin Cancer Res1341544163200710.1158/1078-0432.CCR-07-017317634543 KolevYUetakeHIidaSIshikawaTKawanoTSugiharaKPrognostic significance of VEGF expression in correlation with COX-2, microvessel density and clinicopathological characteristics in human gastric carcinomaAnn Surg Oncol1427382747200710.1245/s10434-007-9484-717687613 MizokamiKKakejiYOdaSIrieKYonemuraTKonishiFMaeharaYClinicopathologic significance of hypoxia-inducible factor 1alpha overexpression in gastric carcinomasJ Surg Oncol94149154200610.1002/jso.2056816847924 FritzscheFRDahlEPahlSBurkhardtMLuoJMayordomoEGansukhTDankofAKnuechelRDenkertCPrognostic relevance of AGR2 expression in breast cancerClin Cancer Res1217281734200610.1158/1078-0432.CCR-05-205716551856 ZweitzigDRSmirnovDAConnellyMCTerstappenLWO'HaraSMMoranEPhysiological stress induces the metastasis marker AGR2 in breast cancer cellsMol Cell Biochem306255260200710.1007/s11010-007-9562-y17694278 ZhangYForootanSSLiuDBarracloughRFosterCSRudlandPSKeYIncreased expression of anterior gradient-2 is significantly associated with poor survival of prostate cancer patientsProstate Cancer Prostatic Dis10293300200710.1038/sj.pcan.450096017457305 Iacobuzio-DonahueCAAshfaqRMaitraAAdsayNVShen-OngGLBergKHollingsworthMACameronJLYeoCJKernSEHighly expressed genes in pancreatic ductal adenocarcinomas: A comprehensive characterization and comparison of the transcription profiles obtained from three major technologiesCancer Res6386148622200314695172 MissiagliaEBlaveriETerrisBWangYHCostelloENeoptolemosJPCrnogorac-JurcevicTLemoineNRAnalysis of gene expression in cancer cell lines identifies candidate markers for pancreatic tumorigenesis and metastasisInt J Cancer112100112200410.1002/ijc.2037615305381 BaiZYeYLiangBXuFZhangHZhangYPengJShenDCuiZZhangZProteomics-based identification of a group of apoptosis-related proteins and biomarkers in gastric cancerInt J Oncol38375383201121165559 Valladares-AyerbesMDíaz-PradoSReboredoMMedinaVIglesias-DíazPLorenzo-PatiñoMJCampeloRGHazMSantamarinaIAntón-AparicioLMBioinformatics approach to mRNA markers discovery for detection of circulating tumor cells in patients with gastrointestinal cancerCancer Detect Prev32236250200810.1016/j.cdp.2008.08.00218801625 SmirnovDAZweitzigDRFoulkBWMillerMCDoyleGVPientaKJMeropolNJWeinerLMCohenSJMorenoJGGlobal gene expression profiling of circulating tumor cellsCancer Res6549934997200510.1158/0008-5472.CAN-04-433015958538 ShiraishiNSatoKYasudaKInomataMKitanoSMultivariate prognostic study on large gastric cancerJ Surg Oncol961418200710.1002/jso.2063117582596 PerssonSRosenquistMKnoblachBKhosravi-FarRSommarinMMichalakMDiversity of the protein disulfide isomerase family: Identification of breast tumor induced Hag2 and Hag3 as novel members of the protein familyMol Phylogenet Evol36734740200510.1016/j.ympev.2005.04.00215935701 JordanPAGibbinsJMExtracellular disulfide exchange and the regulation of cellular functionAntioxid Redox Signal8312324200610.1089/ars.2006.8.31216677077 TuranoCCoppariSAltieriFFerraroAProteins of the PDI family: Unpredicted non-ER locations and functionsJ Cell Physiol193154163200210.1002/jcp.1017212384992 CouëtJde BernardSLoosfeltHSaunierBMilgromEMisrahiMCell surface protein disulfide-isomerase is involved in the shedding of human thyrotropin receptor ectodomainBiochemistry351480014805199610.1021/bi961359w8942642 LahavJGofer-DadoshNLuboshitzJHessOShaklaiMProtein disulfide isomerase mediates integrin-dependent adhesionFEBS Lett4758992200010.1016/S0014-5793(00)01630-610858494 WillemsSHTapeCJStanleyPLThiol isomerases negatively regulate the cellular shedding activity of ADAM17Biochem J428439450201010.1042/BJ2010017920345372 ParkSWZhenGVerhaegheCNakagamiYNguyenvuLTBarczakAJKilleenNErleDJThe protein disulfide isomerase AGR2 is essential for production of intestinal mucusProc Natl Acad Sci USA10669506955200910.1073/pnas.080872210619359471 von WasielewskiRMengelMWieseBRüdigerTMüller-HermelinkHKKreipeHTissue array technology for testing interlaboratory and interobserver reproducibility of immunohistochemical estrogen receptor analysis in a large multicenter trialAm J Clin Pathol118675682200210.1309/URLK-6AVK-331U-0V5P12428786

Immunohistochemical staining for anterior gradient protein 2 in gastric cancer and non-cancerous adjacent (EnVision method). (Aa-Ac) Strong staining in moderately differentiated papillary adenocarcinoma; (Ba-Bc) strong staining in poorly differentiated adenocarcinoma; (Ca-Cc) strong staining in moderately differentiated tubular adenocarcinoma; (Da-Dc) strong staining in mucinous adenocarcinoma; and (Ea-Ec) no staining in non-cancerous gastric mucosa. Original magnification: ×40 (Aa-Ea), ×100 (Ab-Eb), and ×400 (Ac-Ec).

Immunohistochemical staining for cathepsin D in gastric cancer and non-cancerous adjacent tissue (EnVision method). (Aa-Ac) Strong staining in poorly differentiated adenocarcinoma; (Ba-Bc) moderate staining in papillary adenocarcinoma; (Ca-Cc) strong staining in moderately differentiated tubular adenocarcinoma; (Da-Dc) strong staining in poorly differentiated tubular adenocarcinoma; (Ea-Ec) and non-cancerous gastric mucosa, poor staining in stromal cells, and no staining in columnar epithelium. Original magnifications: ×40 (Aa-Ea), ×100 (Ab-Eb) and ×400 (Ac-Ec).

Kaplan-Meier survival curves of gastric cancer patients with high and low AGR2 expression, stratified by TNM stage of the tumor (log-rank test). (A) Survival in stage I (z=40.266, P=0.000); (B) stage II (z=10.108, P=0.001); (C) stage III (z=10.396, P=0.001); and (D) stage IV (z=1.774, P=0.183) gastric cancer with low versus high AGR2 expression. Cum, cumulative; AGR2, anterior gradient protein 2.

Association of AGR2 and CTSD expression with clinicopathological parameters of patients with gastric cancer.

High AGR2 expression High CTSD expression
Clinicopathological parameter Total patients, n n (%) t/χ2/r-valuea P-value n (%) t/χ2/r-valuea P-value
Gender 0.556   0.456 0.148   0.700
  Male 311 142 (45.7)   138 (44.4)
  Female 125   62 (49.6)     58 (46.4)
Age range, years 3.632   0.057 3.401   0.065
  ≤60 237 101 (42.6)   97 (40.9)
  >60 199 103 (51.8)   99 (49.7)
Location of tumor 6.906   0.032 9.798   0.007
  Cardia   55   34 (61.8)   34 (61.8)
  Body 163   78 (47.9)   77 (47.2)
  Antrum 218   92 (42.2)   85 (39.0)
Tumor size, cm 23.336 <0.001 22.176 <0.001
  <5 256   95 (37.1)   91 (35.5)
  ≥5 180 109 (60.0) 105 (58.3)
Depth of invasion 70.250 <0.001 71.524 <0.001
  T1   57   4 (7.0)   4 (7.0)
  T2 109   36 (33.0)   33 (30.3)
  T3 244 143 (58.6) 137 (56.1)
  T4   26   21 (80.8)   22 (84.6)
TNM stage 168.125 <0.001 132.672 <0.001
  I   90   5 (5.6)   8 (8.9)
  II 104   21 (20.2)   22 (21.2)
  III 173 118 (68.2) 110 (63.6)
  IV   69   60 (87.0)   56 (81.2)
Vessel invasion 93.143 <0.001 61.702 <0.001
  Negative 183   36 (19.7)   42 (23.0)
  Positive 253 168 (66.4) 154 (60.9)
Lymphatic metastasis 108.752 <0.001 89.160 <0.001
  Negative 166   25 (15.1)   27 (16.3)
  Positive 270 179 (66.3) 169 (62.6)
Regional lymph nodes 126.361 <0.001 100.981 <0.001
  PN0 166   25 (15.1)   27 (16.3)
  PN1 136   74 (54.4)   73 (53.7)
  PN2   99   74 (74.7)   67 (67.7)
  PN3   35   31 (88.6)   29 (82.9)
Distant metastasis 38.614 <0.001 39.265 <0.001
  Negative 375 153 (40.8) 146 (38.9)
  Positive   61   51 (83.6)   50 (82.0)
Lauren's classification 153.612 <0.001 113.254 <0.001
  Intestinal 166   40 (17.9)   45 (27.1)
  Diffuse 270 164 (77.0) 151 (55.9)
Grade of differentiation   5.285   0.152   4.782   0.188
  Well   13     3 (23.1)     3 (23.1)
  Moderately 128   58 (45.3)   62 (48.4)
  Poorly 293 143 (48.8) 131 (44.7)
  Not     2   0 (0.0)   0 (0.0)
Histological type   3.671   0.299 5.059   0.168
  Papillary   16     9 (56.2)     9 (56.2)
  Tubular 326 148 (45.4) 143 (43.9)
  Mucinous   29   18 (62.1)   18 (62.1)
  Signet-ring cell   65   29 (44.6)   26 (40.0)

Calculated using Student's t-test, χ2 test or Fisher's exact test. AGR2, anterior gradient protein 2; CTSD, cathepsin D.

Univariate analysis of the correlation between clinicopathological parameters and survival of patients with gastric cancer.

Cumulative survival rate, %
Clinicopathological parameter 3-year 5-year Mean survival time, months Z-valuea P-value
Age range, years 14.745 <0.001
  ≤60 74 44 45.85
  >60 59 29 39.63
Location of tumor 7.849   0.020
  Cardia 55 24 37.76
  Body 67 39 43.22
  Antrum 71 39 44.13
Tumor size, cm 49.579 <0.001
  <5 78 49 47.50
  ≥5 52 21 36.63
Histological type 0.934   0.817
  Papillary 69 24 41.92
  Tubular 67 39 43.26
  Mucinous 79 29 44.35
  Gignet-ring cell 63 38 41.54
Grade of differentiation 0.617   0.432
  Well and moderately 73 36 44.12
  Poorly and not 64 38 42.45
TNM stage 370.398 <0.001
  I 96 94 58.09
  II 87 76 52.97
  III 61   7 37.70
  IV 16   1 23.26
Depth of invasion 135.118 <0.001
  T1 93 91 57.18
  T2 82 62 50.01
  T3 58 18 38.38
  T4 35   8 26.85
Lymph node metastasis 176.051 <0.001
  Negative 88 82 54.23
  Positive 54 12 36.30
Distant metastasis 141.372 <0.001
  Negative 75 43 46.23
  Positive 29   3 23.18
Vessel invasion 127.41 <0.001
  Negative 90 70 52.56
  Positive 51 16 36.26
Lauren's classification 239.586 <0.001
  Intestinal 93 66 54.12
  Diffuse 40   9 31.56
AGR2 expression 179.188 <0.001
  Low 91 57 52.85
  High 37   6 32.40
CTSD expression 113.445 <0.001
  Low 87 51 51.12
  High 40 12 33.70

Calculated using the log-rank test. AGR2, anterior gradient protein 2; CTSD, cathepsin D.

Multivariate analysis of the correlation between clinicopathological parameters and survival time of patients with gastric cancer.

Covariate Coefficient Standard error HR 95% CI P-value
Age range (>60 vs. ≤60) −0.277 0.148 0.758 0.567–1.014 0.062
Tumor location (cardia vs. others)   0.046 0.204 1.047 0.702–1.563 0.821
Tumor size (≥5 vs. <5 cm) −0.265 0.151 0.768 0.571–1.032 0.08
Lauren's classification (diffuse vs. intestinal) −0.649 0.194 0.523 0.357–0.765 0.001
Lymph node metastasis (positive vs. negative) −0.669 0.328 0.512 0.269–0.975 0.042
Vessel invasion (positive vs. negative) −0.614 0.205 0.541 0.362–0.809 0.003
Distant metastasis (positive vs. negative) −0.503 0.248 0.605 0.372–0.983 0.042
TNM stage (stages III and IV vs. I and II) 0.263 0.375 1.300 0.624–2.711 0.483
Depth of invasion (T3, T4 vs. T1, T2) −0.724 0.268 0.485 0.287–0.819 0.007
AGR2 expression (high vs. low) −0.805 0.189 0.447 0.309–0.647 <0.001
CTSD expression (high vs. low) −0.215 0.168 0.806 0.580–1.121 0.201

HR, hazard ratio; CI, confidence interval; AGR2, anterior gradient protein 2; CTSD, cathepsin D.