Expressions Of MAGE-A9, MAGE-A11, MAGE-C1and MAGE-C2in Breast Cancers And Their Expression Mechanism

Breast cancer is the most common malignancy in women. However,therapeutic options for the treatment of patients with this tumor are limited to3fundamental modalities: surgical resection, radiation therapy andchemotherapy. Against advanced carcinomas, therapeutic options are limitedto radiation therapy and chemotherapy; however, these modalities do not yieldresults. Therefore, cancer-specific immunotherapy may be expected to becomea novel treatment modality for breast carcinomas.Melanoma-associated antigens (MAGE) are a group of well-characterizedmembers of the cancer-testis antigen (CTA) family that are expressed invarious tumor cells, but not in healthy tissues except for the testis and placenta.The MAGE gene family encodes tumor-associated antigens, which arerecognized by CTLs in conjunction with MHC class I molecules of varioushaplotypes on the tumor cell surface. Thus, MAGE are appealing targets forcancer immunotherapy.In recent years, our professor’s research group has been committed to theresearch of MAGE family. In the present study, we investigated the expressionstatus of MAGE-A9, MAGE-A11, MAGE-C1and MAGE-C2by reversetranscriptional-polymerase chain reaction (RT-PCR) andimmunohistochemistry in60breast benign diseases specimens (includingfibroadenoma and adenosis),60primary breast cancer specimens and60tumor free breast specimens, analyzed their correlation with theclinicopathological parameters and the overall survival of breast cancerpatients. Furthermore, we added DNA methyltransferase inhibitor5-aza-CdRand/or histone deacetylase inhibitor TSA to two breast cancer cell lines andanalyzed the expression of MAGE-A9, MAGE-A11, MAGE-C1andMAGE-C2genes before and after treating with the two inhibitors by RT-PCR in order to elucidate epigenetic mechanism of the four genes expression.The main research contents and results were shown as follows:Part ⅠThe expression of MAGE-A9and MAGE-A11in breast cancertissues and their correlation with the clinicopathological parameters andthe prognosisObjective: To investigate the expression of MAGE-A9and MAGE-A11inbreast benign diseases tissues, tumor-free breast tissues and breast cancertissues, explore their correlation with the clinicopathological parameters andthe prognosis of the breast cancer patients.Methods: The expression of MAGE-A9and MAGE-A11was investigated in60breast benign diseases specimens,60tumor-free breast specimens and60breast cancer specimens by RT-PCR and immunohistochemistry, then thecorrelation between MAGE-A9, MAGE-A11expression and theclinicopathological parameters of breast cancer patients, including age of thepatients, tumor size, pathological types, histology grades, clinical stages,metastasis of axillary lymph nodes, tumor embolus, estrogen receptor (ER),progestrogen receptor (PR) and HER-2status was analyzed. The associationbetween MAGE-A9and MAGE-A11protein expression and the prognosis ofbreast cancer patients was also analyzed.Results:1Expression of MAGE-A9and MAGE-A11mRNA in breast benign diseasesspecimens, tumor-free breast specimens, breast cancer specimens and thecorrelation between MAGE-A9, MAGE-A11expressions andclinicopathological parameters of breast cancer patientsOverall, no MAGE-A9and MAGE-A11expression was found in60tumor-free breast specimens and60breast benign diseases specimens.Expression of MAGE-A9and MAGE-A11mRNA was detected in45%and66.7%of the breast cancer specimens, respectively.MAGE-A9and MAGE-A11mRNA expression was positively associatedwith ER and HER-2expression. MAGE-A9and MAGE-A11mRNAexpression was more frequent in ER-positive breast carcinomas (A9:25/45, 55.6%; A11:34/45,75.6%) compared with ER-negative breast carcinomas(A9:2/15,13.3%; A11:6/15,40%)(A9: χ2=4.865，P=0.027; A11: χ2=5.742，P=0.017). MAGE-A9and MAGE-A11mRNA expression was also morefrequent in HER-2-positive breast carcinomas (A9:25/48,52.1%; A11:36/48,75%) compared with HER-2-negative breast carcinomas (A9:2/12,16.7%;A11:4/12,33.3%)(A9: χ2=6.16，P=0.013; χ2=4.266，P=0.039).No correlation was found between MAGE-A9mRNA expression and theage of the patients (χ2=2.372，P=0.366), tumor size (χ2=3.125，P=0.246),pathological type (χ2=5.104，P=0.08), clinical stage (χ2=3.326，P=0.190)，histological grade (χ2=1.304，P=0.562), lymphatic metastasis (χ2=0.180，P=0.672), tumor embolus (χ2=2.095，P=0.148)and PR (χ2=1.358，P=0.244)status.No correlation was found between MAGE-A11mRNA expression and theage of the patients (χ2=0.399，P=0.895), tumor size (χ2=2.687，P=0.263),pathological type (χ2=5.312，P=0.064), clinical stage (χ2=2.461，P=0.305)，histological grade (χ2=0.525，P=0.893), lymphatic metastasis (χ2=1.250，P=0.264), tumor embolus (χ2=0.549，P=0.459)and PR (χ2=0.601，P=0.438)status.2Expression of MAGE-A9and MAGE-A11proteins in normal human testistissuesAs the positive control, MAGE-A9and MAGE-A11antibodies were firstlyused to stain human normal testicular tissue sections. Both MAGE-A9andMAGE-A11expressions were mainly observed on primary spermatocytes andspermatogonia, and both the nucleus and the cytoplasm were stained.3Expression of MAGE-A9and MAGE-A11protein in breast benign diseasesspecimens, tumor-free breast specimens, breast cancer specimens and thecorrelation between MAGE-A9, MAGE-A11expression andclinicopathological parameters of breast cancer patientsIn most breast cancer cells, MAGE-A9and MAGE-A11staining werefound in cytoplasm but occasionally also observed in nucleus. Overall, noMAGE-A9and MAGE-A11immunoreactivity was observed in60tumor-free breast specimens and60breast benign diseases specimens, while43.3%(26out of60) and63.3%(38out of60) breast cancer specimens were foundpositive with MAGE-A9and MAGE-A11antibodies, respectively.There was generally concordance between MAGE-A9(r=0.967，κ=0.966，P=0.000) and MAGE-A11(r=0.929，κ=0.927，P=0.000) mRNA and proteinexpression.MAGE-A9and MAGE-A11protein expression was positively associatedwith ER and HER-2expression. MAGE-A9and MAGE-A11proteinexpression was more frequent in ER-positive breast carcinomas comparedwith ER-negative breast carcinomas (A9: χ2=4.344，P=0.037; A11: χ2=4.311,P=0.038). MAGE-A9and MAGE-A11protein expression was also morefrequent in HER-2-positive breast carcinomas compared with HER-2-negativebreast carcinomas (A9: χ2=4.671, P=0.031; A11: χ2=5.967, P=0.015).No correlation was found between MAGE-A9protein expression and theage of the patients (χ2=3.621，P=0.209), tumor size (χ2=4.188，P=0.116),pathological type (χ2=4.186，P=0.122), clinical stage (χ2=2.658，P=0.273)，histological grade (χ2=1.006，P=0.618), lymphatic metastasis (χ2=0.045，P=0.832), tumor embolus (χ2=2.807，P=0.094)and PR (χ2=1.067，P=0.302)status.No correlation was found between MAGE-A11protein expression and theage of the patients (χ2=0.548，P=0.808), tumor size (χ2=2.943，P=0.231),pathological type (χ2=3.257，P=0.203), clinical stage (χ2=4.232，P=0.117)，histological grade (χ2=0.964，P=0.732), lymphatic metastasis (χ2=3.062，P=0.08), tumor embolus (χ2=1.386，P=0.239)and PR (χ2=1.270，P=0.260)status.4Correlation between MAGE-A9and MAGE-A11expression and theprognosis of breast cancer patientsPatients with MAGE-A9or MAGE-A11expression had a worse prognosisthan the patients without MAGE-A9(χ2=5.348, P=0.021) or MAGE-A11(χ2=4.192, P=0.041) expression.Conclusions: 1Expression of MAGE-A9and MAGE-A11mRNA was detected in45%and66.7%of the breast cancer specimens, respectively. The expression rate ofMAGE-A9and MAGE-A11protein in breast cancer was43.3%and63.3%,respectively, while no MAGE-A9and MAGE-A11was observed intumor-free breast specimens and breast benign diseases specimens, suggestingthat MAGE-A9and MAGE-A11were tumor specific antigens.2MAGE-A9, MAGE-A11mRNA and protein expression were more frequentin ER or HER-2positive breast carcinomas compared with ER or HER-2negative breast carcinomas.3Both MAGE-A9and MAGE-A11expression were significantly associatedwith reduced overall survival, suggesting that MAGE-A9and MAGE-A11may be potential markers of a poor prognosis for breast cancer patients.Part ⅡThe expression of MAGE-C1and MAGE-C2in breast cancertissues and their correlation with the clinicopathological parameters andthe prognosisObjective: To investigate the expression of MAGE-C1and MAGE-C2inbreast benign diseases tissues, tumor-free breast tissues and breast cancertissues, explore their correlation with the clinicopathological parameters andthe prognosis of the breast cancer patients.Methods: The expression of MAGE-C1and MAGE-C2was investigated in60breast benign diseases specimens,60tumor-free breast specimens and60breast cancer specimens by RT-PCR and immunohistochemistry, then thecorrelation between MAGE-C1and MAGE-C2expression and theclinicopathological parameters of breast cancer patients was analyzed. Theassociation between MAGE-C1and MAGE-C2protein expression and theprognosis of breast cancer patients was also analyzed.Results:1Expression of MAGE-C1and MAGE-C2mRNA in breast benign diseasesspecimens, tumor-free breast specimens, breast cancer specimens and thecorrelation between MAGE-C1and MAGE-C2expression andclinicopathological parameters of breast cancer patients Overall, no MAGE-C1and MAGE-C2expression was found in60tumor-free breast specimens and60breast benign diseases specimens.Expression of MAGE-C1and MAGE-C2mRNA was detected in43.3%and61.7%of the breast cancer specimens, respectively.MAGE-C1and MAGE-C2mRNA expression was positively correlatedwith tumor grade. MAGE-C1and MAGE-C2mRNA expression was morefrequent in high-grade tumors (grade III) compared with low-grade tumors(grade I)(C1: χ2=8.832, P=0.01; C2: χ2=8.575, P=0.01).Furthermore, MAGE-C2mRNA expression was also significantlyassociated with tumor embolus and histological type. Higher frequencies ofMAGE-C2mRNA expression was seen in invasive ductal breast cancercompared to invasive lobular breast cancer and medullary breast cancer(χ2=11.707，P=0.002). MAGE-C2expression was also more frequent in breastcarcinomas with positive tumor embolus compared to breast carcinomas withnegative tumor embolus (χ2=6.094，P=0.014).No correlation was found between MAGE-C1mRNA expression and theage of the patients (χ2=0.403，P=0.907), tumor size (χ2=0.976，P=0.671),pathological type (χ2=2.256，P=0.352), clinical stage (χ2=0.684，P=0.765)，lymphatic metastasis (χ2=0.102，P=0.75), tumor embolus (χ2=0.008，P=0.93),ER (χ2=0.611，P=0.434), PR (χ2=0.16，P=0.689) and HER-2(χ2=0.307，P=0.58) status.No correlation was found between MAGE-C2mRNA expression and theage of the patients (χ2=0.228，P=1.0), tumor size (χ2=0.579，P=0.815), clinicalstage (χ2=4.059，P=0.115)，lymphatic metastasis (χ2=2.303，P=0.129), ER(χ2=0.004，P=0.947), PR (χ2=0.000，P=1.0) and HER-2(χ2=2.366，P=0.124)status.2Expression of MAGE-C1and MAGE-C2proteins in normal human testistissuesAs the positive control, MAGE-C1and MAGE-C2antibodies were firstlyused to stain human normal testicular tissue sections. Both MAGE-C1andMAGE-C2expressions were mainly observed on primary spermatocytes and spermatogonia, and both the nucleus and the cytoplasm were stained.3Expression of MAGE-C1and MAGE-C2proteins in breast benign diseasesspecimens, tumor-free breast specimens, breast cancer specimens and thecorrelation between MAGE-C1and MAGE-C2expression andclinicopathological parameters of breast cancer patientsIn most breast cancer cells, MAGE-C1and MAGE-C2staining were foundin cytoplasm but occasionally also observed in nucleus. Overall, noMAGE-C1and MAGE-C2immunoreactivity was observed in60tumor-freebreast specimens and60breast benign diseases specimens, while38.3%and58.3%breast cancer specimens were found positive with MAGE-C1andMAGE-C2antibodies, respectively.There was generally concordance between MAGE-C1(r=0.902，κ=0.897，P=0.000) and MAGE-C2(r=0.933，κ=0.931，P=0.000) mRNA and proteinexpression.MAGE-C1and MAGE-C2protein expressions were positively correlatedwith tumor grade. MAGE-C1and MAGE-C2protein expressions were morefrequent in high-grade tumors (grade III) compared with low-grade tumors(grade I)(C1: χ2=6.233, P=0.038; C2: χ2=7.471, P=0.017).Furthermore, MAGE-C2mRNA protein was also significantly associatedwith tumor embolus and histological type. Higher frequencies of MAGE-C2protein expression was seen in invasive ductal breast cancer compared toinvasive lobular breast cancer and medullary breast cancer (χ2=10.889，P=0.004). MAGE-C2protein expression was also more frequent in breastcarcinomas with positive tumor embolus compared to breast carcinomas withnegative tumor embolus (χ2=5.503，P=0.019).No correlation was found between MAGE-C1protein expression and theage of the patients (χ2=1.115，P=0.729), tumor size (χ2=1.659，P=0.435),pathological type (χ2=3.173，P=0.206), clinical stage (χ2=0.383，P=0.87)，lymphatic metastasis (χ2=0.012， P=0.914), tumor embolus (χ2=0.099，P=0.753), ER (χ2=0.533，P=0.465), PR (χ2=0.097，P=0.755) and HER-2(χ2=1.422，P=0.233) status. No correlation was found between MAGE-C2protein expression and theage of the patients (χ2=0.525，P=0.903), tumor size (χ2=0.359，P=0.936),clinical stage (χ2=4.022，P=0.15)，lymphatic metastasis (χ2=1.143，P=0.285),ER (χ2=0.429，P=0.513), PR (χ2=0.07，P=0.791) and HER-2(χ2=1.936，P=0.164) status.4Correlation between MAGE-C1and MAGE-C2expressions and theprognosis of breast cancer patientsPatients with MAGE-C1or MAGE-C2expression had a worse prognosisthan the patients without MAGE-C1(χ2=4.213, P=0.04) or MAGE-C2(χ2=4.467, P=0.035) expression.Conclusions:1Expression of MAGE-C1and MAGE-C2mRNA was detected in43.3%and61.7%of the breast cancer specimens, respectively. The expression rate ofMAGE-C1and MAGE-C2protein in breast cancer was38.3%and58.3%,respectively, while no MAGE-C1and MAGE-C2was observed in tumor-freebreast specimens and breast benign diseases specimens, suggesting thatMAGE-C1and MAGE-C2can be used not only as appropriate candidates fortargets of cancer-specific immunotherapy but also as tumor biomarkers forbreast cancer.2MAGE-C1and MAGE-C2expression correlated significantly with tumorgrade, MAGE-C2-positive expression was also significantly associated withpresence of tumor embolus and histological type, suggesting that there was anassociation between MAGE-C1, MAGE-C2expression and features of a moreaggressive clinical behavior in breast cancer, and some carcinoma might bemore suitable for CTA based immunotherapy than others, depending on cellmorphology.3Both MAGE-C1and MAGE-C2expression were significantly associatedwith reduced overall survival, suggesting that MAGE-C1and MAGE-C2maybe potential markers of a poor prognosis for breast cancer patients. Part Ⅲ The expression mechanism of MAGE-A9, MAGE-A11,MAGE-C1and MAGE-C2in breast cancersObjective: We examined the influence of the DNA methylase inhibitor5-aza-2’-deoxycytidine (5-aza-CdR) together with the histone deacetylaseinhibitor trichostatin A (TSA) on the expression of MAGE-A9, MAGE-A11,MAGE-C1and MAGE-C2genes in two breast cancer cell lines.Methods: Cell lines were divided into four groups:(1)5-aza-CdR group:Cells were stimulated by2.5μmol/L and5μmol/L5-aza-CdR for72hours,respectively;(2) TSA group: Cells were stimulated by0.3μmol/L and0.5μmol/L TSAfor24hours, respectively;(3)5-aza-CdR+TSA group: Cells werestimulated by0.5μmol/L TSA for24hours after the48-hour5μmol/L5-aza-CdR incubation;(4) Control group: Cells were mock-treated with thesame volume of culture medium. Then the cells were collected to extract totalRNA for evaluate MAGE-A9, MAGE-A11, MAGE-C1and MAGE-C2geneexpressions.Results:1Expression patterns of MAGE-A9and MAGE-A11genes in MCF-7andMDA-MB-231cells before and after treatment with5-aza-CdR and/or TSATreatment of the two breast cancer cells with2.5μmol/L and5μmol/L5-aza-CdR alone could induce the expression of MAGE-A9and MAGE-A11,and the expression was enhanced with5-aza-CdR concentration increased. Weobserved robust expression of both genes after addition of0.5μmol/L TSA for24hours following48hours of5μmol/L5-aza-CdR treatment. However,0.3μmol/L and0.5μmol/L TSA treatment alone had no influence on MAGE-Agene expression in the two cell lines.2Expression patterns of MAGE-C1and MAGE-C2genes in MCF-7andMDA-MB-231cells before and after treatment with5-aza-CdR and/or TSATreatment of the two breast cancer cells with2.5μmol/L and5μmol/L5-aza-CdR alone could only induce the expression of MAGE-C2, and theexpression was increased with5-aza-CdR concentration evaluated. Weobserved robust expression of MAGE-C2genes after addition of0.5μmol/LTSA for24hours following48hours of5μmol/L5-aza-CdR treatment.However,0.3μmol/L and0.5μmol/L TSA treatment alone had no influence on MAGE-C2gene expression in the two cell lines.However, addition of2.5μmol/L and5μmol/L5-aza-CdR alone or0.3μmol/L and0.5μmol/L TSA alone could not activate MAGE-C1geneexpression in the two breast cancer cells.5μmol/L5-aza-CdR in combinationwith0.5μmol/L TSA also failed to activate MAGE-C1gene expression in thetwo breast cancer cells.Conclusions:1Treatment of MCF-7, MDA-MB-231cells with5-aza-CdR could induce theexpression of MAGE-A9, MAGE-A11and MAGE-C2, suggesting that DNAmethylation is an important mechanism in the regulation of MAGE-A9,MAGE-A11and MAGE-C2.2DNA methylation and histone deacetylation appear to act as synergisticlayers for the silencing of MAGE-A9, MAGE-A11and MAGE-C2genes, andDNA methylation is dominant for the stable maintenance of a silent state.3MAGE-C1expressions could not be reactivating in the two breast cancercells despite treatment with5-aza-CdR and/or TSA, suggesting that DNAmethylation and histone acetylation are not the only mechanism in theregulation of MAGE gene expression.