| BackgroundThe morbidity and mortality rates of esophageal cancer (EC) ranks eighth and sixth worldwide,respectively. Newly diagnosed EC in China accounts for half of all EC cases in the world each year. InChina, esophageal squamous cell carcinoma (ESCC) accounts for nearly90%of EC. In recent years,despite significant advances in diagnostics and therapeutics, the5-year overall survival rate of ESCCranges from15%to25%. In sharp contrast, the5-year survival rate for ESCC patients at early stages ishigher than90%. Most ESCC patients were diagnosed when the disease has developed to late stage, whichwas one of the main causes of worse prognosis for ESCC. In addition, both lack of effective treatment andmetastatic cancer cells led to the less than one year survival of most ESCC patients after clinical diagnosis,of which60-70%of the ESCC patients died of systemic spread and metastasis. The development of ESCCimplicates multiple causes, multiple steps, and was a complex process involving multiple molecularaberration. Although previous studies have identified a number of molecular changes associated with EC,there is still a lack of specific molecular changes specifically associated with ESCC. Furthermore, there isurgent need for identification of novel molecular biomarkers for monitoring high-risk populations, earlydiagnosis and prognosis of ESCC.14-3-3σ, also called HME-1(Human mammary-epithelim-specific marker1or stratifin), is a newcandidate tumor suppressor gene. Among the14-3-3protein family members,14-3-3σ is directly associatedwith tumor development. Multiple literature has reported the low expression of14-3-3σ in breast cancer,lung cancer, colon cancer, liver cancer, stomach cancer, prostate cancer, ovarian cancer, nasopharyngeal,head and neck tumor. In contrast, incresed expression of14-3-3σ has also been reported in pancreaticcancer, colorectal cancer, head and neck cancer, lung cancer, and14-3-3σ has the potential for prognosticprediction. In addition, few investigations have studied the correlation of14-3-3σ with ESCC. Differentialexpression of14-3-3σ may be one of key elements in the multi-stage evolution of ESCC and has prognosticvalue for ESCC. AimsThis study characterized the expression of14-3-3σ in the multi-stage development of ESCC, i.e.normal esophageal epithelium (NEE), low grade intraepithelial neoplasia (LGIN), high grade intraepithelialneoplasia (HGIN) and ESCC, and determined the correlation of14-3-3σ and clinical factors and theirprognostic value in ESCC.1Materials and methods1.1Clinical tissue samplesThe first cohort of fresh ESCC samples, including52male and28female patients were collected fromLinzhou Cancer Hospital, Henan, China from2010to2011. The second cohort of biopsy samplescomprised50fresh biopsy samples collected from Huaihe Hospital, Henan University, China and110formalin-fixed biopsy samples collected from Linzhou Cancer Hospital, China from2010to2012. Forretrospective study of14-3-3σ expression in ESCC, formalin-fixed, paraffin embedded archival specimensof82primary ESCC patients obtained from Huaihe Hospital, Henan University, China between2005and2010, and an ESCC tissue microarray (TMA, Shanghai Outdo Biotech Co., Ltd.) comprising86patientswere used. The median follow-up period in this cohort was42.3months (range,1–60months). None ofESCC patients received radiotherapy or chemotherapy before surgery.An immortalised esophageal epithelial cell line (NEC), three ESCC cell lines (EC1, EC109, EC9706),a paclitaxel-resistant subline EC9706/PTX and a cisplatin-resistant subline EC9706/CDDP derived fromparental EC9706were maintained in RPMI1640supplemented with10%fetal bovine serum,100units/mlpenicillin G, and100ug/ml streptomycin at37℃in a5%CO2incubator.1.2Methods1.2.1Western blotProteins were extracted from tissue or cell samples in lysis buffer (8M urea,4%CHAPS,40mM DTT)and the concentration was measured by Bradford assay.40μg proteins were separated by12%SDS-PAGEand transferred onto a PVDF membrane. The blots were blocked with5%skim milk in TBST for1hourand then incubated with monoclonal mouse anti-14-3-3σ primary antibody (1:5000) overnight at4℃followed by incubation with HRP-conjugated goat anti-mouse antibody. The signal was visualized usingenhanced chemiluminescence reagent and quantified by densitometry using Quantity One software. 1.2.2.Immunohistochemistry (IHC)IHC was used to estimate the14-3-3σ protein expression in110biopsy samples and in168ESCCsamples from patients with long-term survival data. Avidin biotin-horseradish peroxidase complex (ABC)method was used for IHC analysis.1.2.3Semi-quantitative estimation IHC resultThe staining intensity was graded as0(absent),1(weak),2(moderate),3(strong) intensity and thepercentage of positively stained cells was graded as0(<10%positive cells),1(10–25%),2(26-50%),3(51-75%),4(76–100%). Semi-quantitative estimation for each slide or core was performed usingcomposite scores by multiplying the staining intensity and positivity scores (overall score range,0–12). Thecomposite score for each patient was further simplified by dichotomizing it to low expression (overall scoreof≤6) or high expression (score of＞6).1.3Development of SVM model for ESCC prognosisThe programs were coded using Matlab software. Eight ESCC patients were excluded due toincomplete information and the remaining168ESCC patients were divided into128patients used for SVMtraining and40patients for performance assessment of the SVM classifier.1.4Statistical analysesAll statistical analyses were performed with SPSS17.0software. Survival curves were estimated byKaplan-Meier method and differences between curves were tested by log-rank tests. The significance ofprognostic factors on survival was studied by Cox regression model. Receiver operating characteristic(ROC) curve analysis was used to estimate the predictive values of the clinicopathological parameters andSVM classifier. The chi-square test or Fisher’s exact test was used to evaluate the associations between14-3-3σ expression and clinicopathological parameters. P <0.05was considered statistically significant.2Results2.114-3-3σ protein levels were attenuated progressively from NEE to LGIN, HGIN, and ESCC.Gamma test analysis shows that the14-3-3σ expression was negatively correlated with ESCC precancerouslesions progress (Gamma value=-6.163, P=0.000） There was a significant difference in14-3-3σexpression between NEE vs HGIN, NEE vs ESCC, LGIN vs ESCC(P＜0.05). Compared with adjacentnon-cancerous tissue,14-3-3σ expression in ESCC of TNM Ⅰ-Ⅳ significantly decreased and was correlated with clinical progression.2.214-3-3σ expression was undectable in NEC cell line, and moderate in EC cell lines EC1, EC109and EC9706, and maximal in EC9706/PTX and EC9706/CDDP sublines among multiple cell lines withvarious biological behaviour.2.314-3-3σ immunostaining was located predominantly in the cytoplasm and plasma membrane.Chi-square analysis showed that14-3-3σ was negatively correlated with histological grade (P＜0.05),whereas no significant difference between14-3-3σ and gender, age, TNM stage, lymph node metastasis,primary tumor sites(P＞0.05).2.4Kaplan-Meier survival analysis showed that decreased expression of14-3-3σ was linked withESCC survival and the5-year overall survival rate in ESCC patients with low and high14-3-3σ proteinwere25.62%and43.47%, respectively,(log-rank test, χ2=8.448, P=0.004).2.5Multivariate Cox regression analysis indicated that T stage and14-3-3σ were independentprognostic factors of ESCC（P <0.05）；.2.6ESCC SVM classification model（ESCC-SVM）for ESCC prognosis based on SVM algorithmsincluded gender, sex, T stage, histological grade, lymph node metastasis, clinical stage and14-3-3σ. Thepredictive potency was significantly better than the other prognostic factors.3Conclusions3.1The aberrant expression of14-3-3σ occurrs in early stages of ESCC, especially in HGIN, indictingthat14-3-3σ could be a potential biomarker candidate for early detection.3.2The expression of14-3-3σ was negatively correlated with ESCC prognosis and was anindependent prognostic factor of ESCC, which indicates that14-3-3σ could become a potential prognosticbiomarker for ESCC.3.3The ESCC-SVM model for ESCC prognosis with potent predictive value developed in our studywould help inform decisions and tailor therapy to ESCC individuals. |
PDF Full Text Request