Satb1and Related MicroRNA Expression Deregulation In Human Breast Cancer

Part I The expression pattern and prognostic significanceof SATB1in breast cancerPurpose: The aim of this study was to investigate the expression of SATB1in breastcancer, evaluate its role in predicting prognosis and compare with routinely usedbiomarkers: Ki67, c-erbB-2, ER and PR for patients with breast cancer.Methods: The SATB1expression and its clinical significance were examined in93breast invasive ductal carcinoma patients using immunohistochemistry. Association ofSATB1expression with other commonly used biomarkers and clinicopathologicparameters were assessed. Kaplan-Meier survival analysis was used to estimate theprognostic relevance of SATB1and the survival difference between groups was assessedby the log-rank test. Univariate and multivariate Cox regression analysis was performed toevaluate differences of all possible factors in the risk of death. Moreover, the sensitivityand specificity of the aforementioned biomarkers were identified by using receiveroperating characteristic (ROC) curve.Results:Immunohistochemistry results showed that subcellular staining of SATB1was heterogeneous, ranging from cytoplasmic to nuclear. SATB1staining was stronglyrelated to clinical stage, Ki67, ER and PR (all p <0.05). Otherwise, no association wasfound between SATB1expression and expression of c-erbB-2.The median follow-up timeof the patients enrolled in this study was72months(range11～79months). When survivalwas analyzed according to SATB1localization, we observed that patients with negativenuclear SATB1staining and positive cytoplasmic staining had a similar better outcomecompared with positive nuclear SATB1staining samples, suggesting that cytoplasmicpositive and nuclear negative SATB1expression have a similar prognostic significance.Kaplan-Meier survival analysis of93cases revealed a correlation between higher SATB1 nuclear expression levels and shorter overall survival times (p <0.001). Univariate andmultivariate analyses revealed that SATB1was an independent prognostic factor.Interestingly, analysis of survival in these patients classified according to NottinghamPrognostic Index (NPI) score as “good”(score＜3.4) or “poor”(score≥3.4) risk showed asignificant difference in survival time in the NPI either good or poor score subgroupsaccording to SATB1expression. Similarly, we tested the prognostic value of SATB1inlymph node-positive and node-negative subgroups and a significant association of SATB1with survival time in both node-positive and node-negative subgroups was found.Furthermore, ROC curves showed that the AUC(Area Under Curve)for SATB1, Ki67,c-erbB-2, ER and PR were0.859,0.662,0.637,0.586and0.498, respectively, and thesuperiority of SATB1in sensitivity and specificity as biomarker was demonstrated.Conclusion: Nuclear localization of SATB1is a robust independent biomarker forprognosis comparing with currently widely used diagnostic index in breast carcinoma.Part II Distinct expressions of microRNAs predicted to targetSATB1in human breast cancerPurpose: The aim of this study was to test whether microRNAs (miRNAs) play a rolein modulating SATB1expression and to clarify the role of these miRNAs in actual breastcancer by using clinically obtained breast cancer tissues.Methods: Firstly, four canonical target prediction databases, including TargetScan,PicTar, miRanda and miRDB, were utilized to predict the interaction between miRNA andthe3′-UTR of SATB1mRNA. Then we integrated the results of the prediction softwareprograms by displaying Venn diagram. Secondly, total RNA was extracted with miRNeasyFFPE Kit, then the TaqMan MicroRNA Assays and TaqMan MicroRNA ReverseTranscription Kit were used to detect candidate miRNA transcriptional levels in breastcancer tissues by relative real-time quantitative RT-PCR. And SATB1protein expressionlevels were assessed by immunohistochemistry(IHC). Finally, correlations between theexpression levels of these miRNAs and SATB1were analyzed.Results: Overall22miRNAs,24miRNAs,76miRNAs and75miRNAs were foundby TargetScan, PicTar, miRanda and miRDB, to target SATB1respectively, whereas only miR-7、miR-21、miR-23a、miR-23b、miR-34a and miR-155were found by all fourprograms to be predicted to regulate SATB1. Quantitative RT-PCR detection analysisshowed that expression levels of miR-34a were much higher in SATB1-negative than inSATB1-positive tumors (p<0.001). whereas lower miR-7expression showed a strong trendtoward an association with elevated SATB1expression (P=0.056). We then extended theanalysis of miR-34a expression in breast cancer tissues to include all scores for SATB1protein expression, assessed semiquantitatively by IHC. miR-34a expression levels weregradually decreased as SATB1protein scores increased. Moreove,our quantitative RT-PCRresults showed no differences in miR-21,miR-23a,miR-23b,and miR-155expressionbetween SATB1-negative and SATB1-positive tumors.Conclusion: SATB1maybe a novel target of miR-34a in human breast cancer.Down-regulation of miR-34a promotes human breast cancer to progress through reducedrepression of SATB1. Restoration of miR-34a may prove therapeutic in breast cancerpatients in which aberrant SATB1expression plays a role.Part III Inactivation of miR-34a by aberrant MIR34A CpGmethylation in human breast cancerPurpose: The aim of this study was to determine whether CpG methylation of theMIR34A promoter contributes to downregulation of mature miR-34a in patients withbreast cancer.Methods: Bisulfite conversion and purification of genomic DNA from eachformalin-fixed and paraffin-embedded tumor sample was performed by EpiTect Fast FFPEBisulfite Kit. Then, CpG methylation of the MIR34A promoter in the breast tumor DNAsamples was assessed by Methylation-specific PCR (MSP). The expression of maturemiR-34a was described previously.Results: Methylation status in the promoter region of the MIRN34A gene wasassessed in55patients aviable. Sixteen(29%) were unmethylated,9(16%) weremethylated and30(55%) were methylated/unmethylated. A significant inverse correlationbetween miR-34a methylation and expression of mature miR-34a (p <0.001) was detected.Conclusion: MIR34A represents a tumor suppressor gene which is inactivated by CpG methylation and subsequent transcriptional silencing in human breast cancer.