Expression Of MAGE-A1and MAGE-A3in Glioma And The Expressed Mechanism

Glioma is the most common primary neuroepithelial tumors of centralnervous system in human, which accounts for about45-50%primaryintracranial tumor. To date, the therapeutic effect of glioma is still not idealand the prognosis is poor. Immunotherapy of tumour is a therapeutic methodwith more specificity and less side effects, which has been the fourththerapeutic method after surgery, radiotherapy and chemotherapy. With thetumor immunotherapy rising, identification of ideal targets became critical.The ideal targets should have characteristics as follow: firstly, they expresshigyly in tumours, secondly, they do not express in normal tissue, or theirexpressions are strictly limited.Cancer/testis antigens (CTAs) express in testicular germ cell of testis andpartial placenta and testis is considered as immune privilege organ, thereforeCTAs are considered ideal targets for specific tumor immunotherapy.Melanoma-associated antigen gene-A (MAGE-A), as a subgroup of CTA,became research hotspot for its high tumour specificity.In recent years, our professor’s research group has been committed toMAGE-A gene. In this study, we detected the expression of MAGE-A1、MAGE-A3and MAGE-A11proteins in glioma tissues and normal braintissues by immunohistochemistry and analyzed the relationship between theexpressions of the three proteins and clinicopathologic characteristics.Expression of MAGE-A1and MAGE-A3in mRNA level was detected withRT-PCR method. Promoter methylation status of MAGE-A1and MAGE-A3was researched by MSP. The association of promoter methylation status andMAGE-A1and MAGE-A3gene expression levels was analyzed. AddingDNA methyltransferase inhibitor and/or histone deacetylase inhibitor to twoglioma cell lines, we analyzed the expression of MAGE-A1and MAGE-A3 genes before and after treating with two inhibitors in order to elucidateepigenetic mechanism of the two genes expression.Part Ⅰ. Expression of MAGE-A1，MAGE-A3and MAGE-A11proteinsand their correlation with the clinicopathological parametersObjective: To investigate the expression of MAGE-A1，MAGE-A3andMAGE-A11proteins in glioma tissues and the normal brain tissues, andexplore their correlation with the clinicopathological parameters and theprognosis.Methods: Expression of MAGE-A1， MAGE-A3and MAGE-A11proteins in78formalin-fixed and paraffin-embedded glioma and normal braintissues were detected with immunohistochemical staining. Then thecorrelation between MAGE-A1，MAGE-A3，MAGE-A11protein expressionand the clinicopathological parameters of glioma patients, including age,gender, pathological grades, KPS score and ki-67index was analyzed. Theassociation between MAGE-A1，MAGE-A3，MAGE-A11protein expressionand the prognosis of glioma was also researched.Results:1Expression of MAGE-A1,-A3, and-A11proteins in normal testis tissuesIn normal human testis tissue, MAGE-A1mainly located in thecytoplasm and membrane, however, MAGE-A3and-A11was observed inboth cytoplasm and nucleus of the primary spermatogonia and spermatocytes.2Expression of MAGE-A1,-A3, and-A11proteins in glioma and normalbrain tissuesIn normal brain tissues, the expressions of MAGE-A1,-A3, and-A11proteins were not detected. Fifty out of78(64.1%),40out of78(51.2%), and45out of78(57.6%) glioma specimens were detected high expression levelswith MAGE-A1,-A3, and-A11antibodies, respectively. In addition, theexpression levels of MAGE-A1(P=0.000) and-A11(P=0.000) proteins indifferent pathological grade were significantly different, which was increasedwith pathological grade ascending, however, there was no significantdifference between MAGE-A3（P=0.958）protein expression level and pathological grade. In grade Ⅰ-Ⅱ gliomas, MAGE-A1expressed mainlyin cytoplasm and membrane and MAGE-A3,-A11in nucleus of tumor cells.In grade Ⅲ-Ⅳ gliomas, MAGE-A1was mainly detected in cytoplasm andMAGE-A3,-A11mainly in both cytoplasm and nucleus of tumor cells.3Relationship between the expressions of MAGE-A1,-A3,-A11proteins andclinicopathologic characteristics of glioma patientsNo association was observed among age, gender, KPS score and theexpression levels of MAGE-A1,-A3, and-A11proteins. However, significantcorrelation were detected between expression of MAGE-A1(P=0.000) orMAGE-A11(P=0.030) protein and glioma pathological grade. There were nosignificant relations between MAGE-A3(P=0.069) protein expression andtumor grade.4Correlation between expression of MAGE-A1,-A3, and-A11proteins andki-67indexMAGE-A1（χ~2=4.965，P=0.026）and MAGE-A11（χ~2=7.123，P=0.008）proteins significantly correlated with ki-67index, however, MAGE-A3（χ~2=2.224，P=0.136）protein has no significant correlation with ki-67index.5Relationship between MAGE-A1,-A3, and-A11protein expressions andprognosis of glioma patientsThe survival of patients with high expression level of MAGE-A1（P=0.005）or-A11（P=0.019） protein was lower than that of patients withlow expression level, however, no significant difference was detected in thetwo group of MAGE-A3（P=0.304）.In univariate analysis, high pathologicalgrade (P=0.000), low KPS score (P=0.000), decreased age (P=0.014), highki-67labeling index (P=0.050), high MAGE-A1(P=0.005) and-A11(P=0.019) proteins expression levels correlated with poor outcome of patients.Further assessment with Cox’s multivariable analysis showed that highpathological grade (P=0.000), low KPS score (P=0.000), high MAGE-A1(P=0.007) and-A11(P=0.010) expression levels were statistical predictors forpoor overall survival. Conclusions：1MAGE-A1,-A3and-A11proteins may be ideal targets for tumorimmunotherapy.2MAGE-A1and MAGE-A11proteins may correlate with the proliferation ofglioma cells.3MAGE-A1，MAGE-A11protein，pathological grade and KPS score may beprognostic markers of glioma.Part Ⅱ Expression of MAGE-A1and MAGE-A3gene and analysis ofDNA methylation modeObjective: Expression of MAGE-A1and MAGE-A3genes andpromoter methylation status of the two genes were detected and theirrelationship was analyzed. Then the association of promoter methylationstatus or gene expression level and clinicopathological parameters was studed.Methods: Expressions of MAGE-A1and MAGE-A3genes in gliomawere detected with RT-PCR method, and the relationship between the genesexpression levels and clinicopathological parameters was analyzed.Methylation specific PCR was used to survey promoter methylation status ofMAGE-A1and MAGE-A3, then， the relationship of promoter methylationstatus of the two genes and their expression level of gene and protein wasstudied. Finally, we discussed the association between promoter methylationstatus of MAGE-A1, MAGE-A3and clinicopathological characteristics.Results:1Expression of MAGE-A1and MAGE-A3gene in normal brain tissues andglioma tissuesIn normal brain tissues, no MAGE-A1and MAGE-A3gene was detected.In glioma tissues,65.4%and38.5%specimens expressed MAGE-A1andMAGE-A3gene, respectively, and26.9%gliomas simultaneously expressedtwo genes. Only one of MAGE-A1or MAGE-A3gene was detected in76.9%glioma specimens and no one in23.1%glioma tissues.2Relationship of MAGE-A1or MAGE-A3gene and overall survival ofglioma patientsThe overall survival of MAGE-A1gene positive group was significantly lower than that of MAGE-A1gene negtive group（P=0.0463）, however, nosignificant difference was observed between the two group of MAGE-A3gene（P=0.797）.3The relationship between protein and gene of MAGE-A1and MAGE-A3MAGE-A1（r=0.402，P=0.000）and MAGE-A3（r=0.242，P=0.033）genes significantly correlated with MAGE-A1and MAGE-A3protein,respectively. However, the expressions of gene and protein were notone-to-one correspondence.4Promoter methylation status of MAGE-A1and MAGE-A3Demethylation rates of promoter were0%and64.1%of MAGE-A1genein normal brain and glioma tissues, and there was significant differencebetween them（χ~2=20.796，P=0.000）. Promoter demethylation status ofMAGE-A3were detected in none normal brain tissues and41%gliomaspecimens, and there was significant difference between them（χ~2=9.382，P=0.002）.5Association between promoter methylation status of MAGE-A1,MAGE-A3and their gene or protein expression levelMAGE-A1（χ~2=37.289，P=0.000）gene and protein （χ~2=14.736，P=0.000）was statistically associated with its promoter demethylation status,and MAGE-A1gene expression level in promoter demethylation group wassignificantly higher than that in promoter methylation group(0.1414±0.0690vs.0.0317±0.0066，P=0.000). MAGE-A3gene expression （χ~2=36.066，P=0.000）was significantly associated with its promoter demethylation status,however, no statistic correlation was found between MAGE-A3proteinexpression（χ~2=1.697，P=0.193） and its promoter demethylation status.MAGE-A3gene expression level in promoter demethylation group wassignificantly higher than that in promoter methylation (0.0666±0.0065vs.0.0071±0.0026，P=0.000).6Relationship between promoter methylaiton status of MAGE-A1andMAGE-A3genes and clinicopathological parametersNo association was found between gender（χ~2=0.778，P=0.378）, KPS score（χ~2=0.060，P=0.806）and promoter demethylaiton status of MAGE-A1.There was significant association between promoter demethylaiton status ofMAGE-A1and ag（eχ~2=7.306，P=0.007）and pathological grad（eχ~2=14.286，P=0.003）. Demethylation frequency of MAGE-A1gene in low grade gliomaswas significantly lower than that in high grade gliomas. The overall survivalof patients whose MAGE-A1promoter demethylated was lower than that ofpatients whose MAGE-A1promoter methylated, but the difference was notsignificant（P=0.179）. No statistic association was detected between gender（χ~2=0.464，P=0.496）, age（χ~2=0.1.009，P=0.315）, KPS scor（eχ~2=0.102，P=0.749）, pathological grade （χ~2=3.367， P=0.338） and promoterdemethylaiton status of MAGE-A3. The overall survival of patients whoseMAGE-A3promoter demethylated was lower than that of patients whoseMAGE-A3promoter methylated, but the difference was not statistic（P=0.201）.Conclusions:1No MAGE-A1and MAGE-A3gene expression was detected in normalbrain tissues, and majority of glioma tissues expressed MAGE-A1andMAGE-A3, therefore they may be ideal targets of tumor immunotherapy.2MAGE-A1gene may be considered as a potential indicator of poorprognosis.3MAGE-A1and MAGE-A3gene expressions were positively correlated withtheir protein expression, but they were not one-to-one correspondence and thefrequencies of their protein expression are higher than that of their geneexpression, which may attribute to the cross reaction of antibodies.4There was no promoter demethylation of MAGE-A1and MAGE-A3innormal brain tissues, however, majority of glioma specimens demethylated inpromoter of MAGE-A1and part in MAGE-A3.5DNA demethylation may be one of MAGE-A1and MAGE-A3geneexpression mechanism and there may be other mechanisms for their geneexpression.6Promoter methylation status of MAGE-A1or MAGE-A3may not be prognostic indicator.Part Ⅲ Expression of MAGE-A1and MAGE-A3gene in glioma celllines and the detection of DNA methylation and histone acetylation modesObjective: After stimulating U87and U251cell lines with DNAmethyltransferase inhibitor (5-aza-CdR) and/or Histone deacetylase inhibitor(TSA), we observed the MAGE-A1and MAGE-A3gene expressions in orderto illuminate mechanism of the two gene expressions.Methods: The two cell lines were treated with5-aza-CdR (0、2.5、5μmol/l) and/or TSA (0、0.5、1μmol/l) for72hours and then were cultured infresh medium without5-aza-CdR and TSA for48hours. The cells werecollected to extract total RNA for evaluate MAGE-A1and MAGE-A3geneexpressions. In group1to3cells, the final concentrations of TSA in mediumwere0、0.5、1μmol/l, respectively and no5-aza-CdR was contained in medium.In group4to6cells, the final concentrations of TSA in medium were0、0.5、1μmol/l, respectively and2.5μmol/l5-aza-CdR was contained in medium. Ingroup7to9cells, the final concentrations of TSA in medium were0、0.5、1μmol/l, respectively and5μmol/l5-aza-CdR was contained in medium.Results:1Expression of MAGE-A1and MAGE-A3genes in U87and U251Before treated with experimental drugs, there were no gene expressionsof MAGE-A1and MAGE-A3in U87and a few expressions in U251.2Expression of MAGE-A1and MAGE-A3genes in U87after treated with5-aza-CdR and/or TSAIn group1to3cells, no MAGE-A1and MAGE-A3gene expression wasfound. In group4to6cells, the two gene expression levels were higher thanthose in group1to3cells and increased with TSA concentration evaluated. Ingroup7to9cells, MAGE-A1and MAGE-A3gene expression levels werehigher than those in group4to6cells and increased with TSA concentrationevaluated. We analyzed the role of two experimental drugs with factoranalysis and found that there were significant differences in MAGE-A1andMAGE-A3gene expression levels when cells were treated with5-aza-CdR alone or TSA alone or the two drugs united. When treated with5μmol/l5-aza-CdR and1μmol/l TSA, the two gene expression levels were highest.Theresults indicated that TSA alone could not induce MAGE-A1and MAGE-A3to express,5-aza-CdR alone or two drugs united may cause MAGE-A1andMAGE-A3to express, further, the role of combined application of5-aza-CdRand TSA was stronger than anyone alone.3Expression of MAGE-A1and MAGE-A3genes in U251after treated with5-aza-CdR and/or TSAIn group1to3cells, a little MAGE-A1and MAGE-A3gene expressionswere found. In group4to6cells, the two gene expression levels were higherthan those in group1to3cells and increased with TSA concentrationevaluated. In group7to9cells, MAGE-A1and MAGE-A3gene expressionlevels were higher than those in group4to6cells and increased with TSAconcentration evaluated. We analyzed the role of two experimental drugs withfactor analysis and found that there were significant differences in MAGE-A1and MAGE-A3gene expression levels when cells were treated with5-aza-CdR alone or TSA alone or the two drugs united. When treated with5μmol/l5-aza-CdR and1μmol/l TSA, the two gene expression levels werehighest. The results indicated that TSA alone could not induce MAGE-A1andMAGE-A3to increase,5-aza-CdR alone or two drugs united may causeMAGE-A1and MAGE-A3to increase, further, the role of combinedapplication of5-aza-CdR and TSA was stronger than anyone alone.Conclusion: The mechanism of MAGE-A1and MAGE-A3geneexpressions included DNA methylation and histone acetylation, and DNAmethylation was more important.