The Analysis Of The Luminex Test Of Serum Protein Biomarkers And The Pathwayarray Analysis Of The Inflammation&Metabolism Tissue Proteins In Gastric Cancer With H.pylori Infection

Object:Gastric cancer is one of the primary malignant tumor of the global risks to human health,in2014, about1.1million new cases worldwide of people (not include not hospitalizedpatients), only in China,2014for490,000new cases, new gastric cancer cases worldwideaccount for44.6%；The occurrence of gastric cancer is closely related to helicobacter pylori(Hp, H.Pylori) infection[1,2], this paper applied Luminex[3]technology to detect theexpression of serum protein markers related to Hp infection in stomach cancers, to explorecorrelation analysis the relation among serological markers,early diagnosis of gastric cancerand tumor stage; Application of PPA (Protein PathwayArray)[4]technology studyinflammation metabolism related protein expression of helicobacter pylori infection in gastriccancer tissue to find the key proteins.Methods:In the first part,We used Luminex technology to construct9serum marker detectionkits,and test9serum proteins expression in285cases of gastric cancer patients and238healthy physical examination,(208/285Hp positive77/285Hp negative in gastric cancerpatients;92/238Hp positive146/238Hp negative in healthy check-ups).All gastric cancerpatients and healthy check-ups were harvested serum in static centrifugal tube in the morningbefore breakfeast.All of them were checked with gastroscope and chest X-ray film (femalepatients with breast check) to excluld other tumor and inflammation or infection state.Thegastric cancer patients and healthy check-upswere randomized to detect serum and performcorrelation analysis test results and diagnosis of gastric cancer and tumor stages, the diagnosismodel is established. We also applied a random forest (random forest,RF), SVM (supportvector machine,SVM), Logistic regression and COX regression analysis and diagnosis of riskmodels The second part, we applied Protein Pathway Array technologyanalyzed of Hpinfection in stomach tissue inflammation, metabolic pathways related proteins is, andapplication of Ingenuity Pathway Analysis (IPA)[5]Analysis software research PPA screening of key proteins in the key signaling pathways of Hp infection in gastric cancer.Findings:In the first part, based on the basis of previous studies[6,7]and literature review[8~12],we built9serum protein markers Luminex kits successfully (Pepsonigen–Ⅰ[8], Pepsonigen-II[7], ICAM-1[9], COX-2[6], ADAM8[6], VEGF[10], EGFR[11], beta-Catenin[7], Hemoglobin[12]), This platform is quick and joint determination of serum levels of these proteins, to domore than just operations simple and fast, high sensitivity and reproducibility, detection range,saving the time and amount of specimen.In this study Luminex method was applied to test thenine serological protein285cases of gastric cancer patients and238cases of healthycheck-ups, and ELISA method was used to test the IgG to H. pylori[7]in serum, totally10serum proteins were tested.From the statistical correlation analysis of the10proteins, beta-catenin, COX2and VEGF have significant correlation.All the gastric cancer patients and thehealthy physical examination were randomly divided into exercise group and the controlgroup, nested cross validation[13]of10proteins were performed, five proteins (IgG to H.pylori, ADAM8, Pepsonigen-Ⅰ, Pepsonigen-II, VEGF) score were higher, the importance ofconsists of them the best combination of proteins to be included in the statistical algorithms,Random Forests (Random Forests, RF)[14], Support Vector Machine (Support Vector Machine,SVM)[15], Logisitic regression algorithm were carried on the diagnosis accuracy, sensitivity,specific degree detection, after training we got the accuracy, sensitivity and specificity of RFwere82.5%,86.0%,86.0%, the accuracy, sensitivity and specificity of the SVM algorithmwere86.1%,88.6%,88.6%, the accuracy, sensitivity and specificity of logistic regressionwere78.7%,82.9%,73.7%in the validation group.The results of the three algorithms is betterthan conventional tumor markers[16]. Especially the SVM and the RF performance is superiorto the results of Logistic regression. R software survival kit for log-rank analysis of theprotein expression in serum of gastric cancer patients, concluded that the prognosis of patientswith obvious differences in protein cutoff value (P<0.05), through analysis of PGI and PGII,HB, VEGF, Beta-Catenin able to obtain cutoff cases data can be divided into high-risk andlow-risk groups, other proteins were not given satisfactory cutoff cut-off point, low risk groupis significantly better than the lifetime of the high-risk groups. raw a nomograph relating to that5proteins were used to assess prognosis.In the second part, we applied PPA to study inflammation, metabolic pathway proteinexpression in gastric cancers tissue with and without Hp infection.85cases of patients wereincluded and harvested gastric cancer tissue samples,52cases were identified as Hpserological positive (61.2%). We applied SAM tools after statistical analysis, between Hppositive gastric cancer with Hp negative there were11protein expression differently (P<0.05and False Discovery Rate <5%). In the11proteins, there are7proteins presented highexpression in Hp positive gastric cancer, including BAK1, IL6, HIF-1alpha, IL8, IL10,DACH1, HSP70, there are4proteins present lower expression in Hp negative gastric cancer,including CDC2, L-selectin, P27and DPYD. Through the IPA system analyses at the sametime the seven different expression of proteins in the upstream regulatory proteins,6proteinshad statistical significance in common upstream regulatory proteins, including BCL2(including downstream BAK1, CDK1, CDKN1B, CXCL8, HIF1A, IL10and IL6, p=1.22e-14), magnolol (including CDKN1B, CXCL8, downstream HIF1A, IL10and IL6, p=5.46e-14), epinephrine (including CDK1, CDKN1B, downstream CXCL8, IL10, IL6and SELL,e-13p=4.61), RET (including CDKN1B, CXCL8, downstream HSPA1A/HSPA1B, IL10, IL6and SELL, p=4.35e-12), PI3K (complex)(including BAK1, CDKN1B, downstream CXCL8,HIF1A, IL10, IL6and SELL, p=1.18e-11) and mutated BRAF (including CDK1, CDKN1B,downstream CXCL8, HIF1A and IL6, p=1.50e-11), and6upstream regulatory proteins anddownstream proteins between each other mutual inhibition or synergy.Conclusion:In research carried out independently in making Luminex assays to establish diagnosticanalysis of Luminex platform, detection of gastric cancer with this platform and Hp infectionin healthy people associated with serum protein markers for gastric cancer, got a serumprotein combination for the diagnosis of gastric cancer, is a continuation of previous studies.The combined detection of RF and SVM model for screening for gastric cancer with highaccuracy, sensitivity and specificity of was further applied to the diagnosis of gastric cancerscreening works laid the Foundation in the future; through analysis of PGI and PGII, HB,VEGF, Beta-Catenin able to obtain cutoff cases data can be divided into high-risk andlow-risk groups, other proteins were not given satisfactory cutoff cut-off point, Low dangerous group of survival period obviously good vs high dangerous Group obtained hasNomograph combines5species protein of dangerous degree on patients of prognosis fordiscriminant; on Hp infection and non-infection gastric cancer organization in theinflammation, and metabolism signal related protein of expression analysis found obviouslydifferences protein BAK1, and HIF-1Alpha, and IL6, and IL8, and IL10, and DACH1, andHSP70of expression, and found its upstream key signal conduction pathwayBCL2,magnolol,epinephrine,RET, and PI3K (complex) as well as BRAF, which is the startingpoint of future research and targeted therapy.