Deregulation of miR‐1245b‐5p and miR‐92a‐3p and their potential target gene, GATA3, in epithelial–mesenchymal transition pathway in breast cancer

Abstract Background MicroRNAs (miRNAs) are small molecules that have prominent roles in tumor development and metastasis and can be used for diagnostic and therapeutic purposes. This study evaluated the expression of miR‐92a‐3p and miR‐1245b‐5p and their potential target gene, GATA3 in patients with breast cancer (BC). Materials and Methods In the search for BC‐related microRNAs, miR‐124b‐5p and miR‐92a‐3p were selected using Medline through PubMed, miR2disease, miRcancer and miRTarBase. Moreover, target gene GATA3 and their possible interaction in the regulating epithelial‐mesenchymal transition (EMT) and invasion was evaluated using in silico tools including miRTarBase, TargetScan, STRING‐db, and Cytoscape. The expression level of miR‐92a‐3p, miR1245b‐5p, and GATA3 were assessed on extracted RNAs of tumor and nontumor tissues from 36 patients with BC using qPCR. Additionally, clinical‐pathologic characteristics, such as tumor grade, tumor stage, lymph node were taken into consideration and the diagnostic power of these miRNAs and GATA3 was evaluated using the ROC curve analysis. Results In silico evaluation of miR‐92a‐3p and miR‐1245b‐5p supports their potential association with EMT and invasion signaling pathways in BC pathogenesis. Comparing tumor tissues to nontumor tissues, we found a significant downregulation of miR‐1245b‐5p and miR‐92a‐3p and upregulation of GATA3. Patients with BC who had decreased miR‐92a‐3p expression also had higher rates of advanced stage/grade and ER expression, whereas decreased miR‐1245b‐5p expression was only linked to ER expression and was not associated with lymph node metastasis. The AUC of miR‐1245b‐5p, miR‐92a‐3p, and GATA3 using ROC curve was determined 0.6449 (p = .0239), 0.5980 (p = .1526), and 0.7415 (p < .0001), respectively, which showed a significant diagnostic accuracy of miR‐1245b‐5p and GATA3 between the BC patients and healthy individuals. Conclusion MiR‐1245b‐5p, miR‐92a‐3p, and GATA3 gene contribute to BC pathogenesis and they may be having potential regulatory roles in signaling pathways involved in invasion and EMT pathways in BC pathogenesis, as a result of these findings. More research is needed to determine the regulatory mechanisms that they control.

findings.More research is needed to determine the regulatory mechanisms that they control.

K E Y W O R D S
breast cancer, epithelial-mesenchymal transition, GATA3, MicroRNAs

| INTRODUCTION
Breast cancer (BC), as a multifactorial disease, is the leading cause of death in women, globally. 1Despite the numerous studies on BC, the molecular mechanisms of this disease have not been fully known. 2,3e survival rate of the patient in the early stages of this cancer is high, therefore early diagnosis of BC is very crucial. 4,5Moreover, suggesting the potential biomarkers for early diagnosis is a promising strategy. 5,6croRNAs (miRNAs), as highly conserved single-strand short RNAs, link with 3´UTR regions in their target genes to regulate gene expression in the transcription level. 7Recent studies showed that the specific miRNAs act as potential oncogenes or tumor suppressors in human breast tumors.][10][11] The cancer invasion and metastasis program aim to investigate the underlying processes that are responsible for the majority of cancer deaths. 12,13ithelial-mesenchymal transition (EMT) is a leading cause of distant metastases in the epithelial cancers, such as BC.EMT able to disrupt cell-to-cell junctions, thereby leading to the gain of mesenchymal shape and loss of epithelial properties.As a result, cells are released from the parental epithelial tissues and regenerate new metastatic tissues.miRNAs can target EMT-inducing transcription factors such as basic helix loop helix (bHLH), SNAIL, TWIST, and ZEB and inhibit the EMT process. 14For example, miR-200 targets ZEB1 and ZEB2 mRNAs in a direct manner via the upregulation of E-cadherin in cancer cell lines and thus suppresses cell motility and EMT pathway. 15erefore, the prediction and identification of miRNAs contributing to EMT, can be introduced as diagnostic and therapeutic tools in the treatment of various cancers.
Recently, it has been reported that the weak prognosis in cancer patients is associated with the upregulation of miR-92a in blood or tumor tissues miR-92a.As, several studies reported that patients with high expression levels of miR-92a experience lower survival rates or faster tumor progression and metastasis in colorectal cancer [16][17][18] and nonsmall cell lung cancer. 19,20Versus, some new studies showed that the increased expression of miR-92a is closely related to the patient survival in these cancers.For example, in chronic leukemia, the overexpression of this miRNA was reported to cause higher survival rates in these patients. 21The role of this miRNA in BC is also unclear.Some results showed that an increased expression of this miRNA is associated with decreased tumor macrophage infiltration and better outcomes in BC. 22 On the contrary, another study indicated that the expression of miR-92a is directly related to the tumor size and increased TNM stage of BC. 23 Moreover, there are limited studies on the in-silico prediction of the role of miR-92a-3p in the regulating EMT and invasion in the BC.
The miR-1245b family consists of the miR-1245a and miR-1245b, both of which are expressed in breast tissue.Limited research has been conducted on their roles in various cancers.Recently, Yan et al showed that the inhibition of miR1245b-5p in osteosarcoma cancer significantly increased the invasiveness in this cancer. 24Moreover, MiR-1245 upregulation in BC represses homologous recombination (HR)-mediated repair and increases the cancer cell sensitivity to γirradiation by the targeting the tumor suppressor gene, BRCA2. 25ere is no information on miR-1245b and its potential role in the regulating the EMT pathway and BC invasion.This study evaluated the expression levels of miR-29a-3p, miR-1245b-5p, and GATA3 in the BC and normal tissues and investigate their association with clinicopathological characteristics of the tumors.We also evaluated the diagnostic power of GATA3 with miR-92a-3p and miR-1245b-5p through the ROC curve.In addition, using in silico tools, we addressed the possible role of these miRNAs in regulating other genes implicated in EMT and invasion in BC pathogenesis.

| In silico analysis
The protein-protein interactions of the potential target genes with selected miRNAs were investigated by STRING-db (https:// string-db.org),and results were visualized in Cytoscape V.1.7.3 software.The number of direct edges was used in Cytoscape to visualize and analyze the network for constructing the network graph.Furthermore, a bioinformatics analysis was performed to determine the possible function of the desired miRNAs in the regulating other genes involved in general signaling pathways related to invasion and EMT.
An enrichment assessment was performed for the target genes of desired miRNAs by DIANA miRPath v.3 tools. 32

| Preparation of normal and cancerous breast tissues
Both normal and cancerous breast tissue samples were collected from 36 women aged 27-85 undergoing surgery between October 2018 and April 2019 at Seyed-O-Shohada Hospital in Isfahan and Ayatollah Kashani hospital in Shahrekord, both in Iran.All patients had not taken any adjuvant therapy before surgery.The subjects under treatment with chemotherapy and radiotherapy were excluded of study.The written informed consent was signed by volunteers and frozen tissue specimens were stored at À80 C for further research.A professional pathologist characterized all the patients.The clinicopathological characteristics of these patients are detailed in Table 1.This study was approved by the Institutional Review Board (IRB) of the Shahrekord Universities of Medical Sciences.

| Total RNA isolation
The breast tissue samples were subjected to RNX-Plus solution (1 mg/50 mg breast tissue) based on the manufacturer's protocol (SinaClon BioSience, Iran).Tissue lysis was gently carried out with a homogenizer for 30 min at 37 C. Subsequently, RNA quality and quantity were evaluated using Nano drop and agarose gel.Only those samples with two distinct rRNA bands (28s, 18s) and 260/280 ratio of 1.8-2 were considered for further analysis.The specimens were kept at À80 C until further investigation.

| cDNA synthesis
Treated RNA with DNase I, was used for cDNA synthesis using a cDNA synthesis kit (Yektatajhis, Tehran, Iran cat No. Y. TA500).The reaction in a final volume of 10 μL containing 500 ng RNA, 0.5 μL RT enzyme, 2 μL 5 Â prime script buffer, 0.5 μL random 6mer, and 0.5 μL oligo dT primer was incubated for 60 min at 42 C, 60 min at 37 C, and 5 min at 70 C.Moreover, cDNA synthesis for miRNA was performed using a cDNA synthesis kit (Bonyakhteh, Tehran, Iran, cat.no: Bon 209001) through polyadenylation of the 3 0 end of all the RNAs.Briefly, elongation was performed in a polyadenylation reaction, with a final volume of 20 μL at 37 C for 30 and 20 min at 65 C.
Then the reaction to make cDNA was performed immediately after polyadenylation reaction and using the existing compounds in cDNA Synthesis Kit, for each sample of polyadenylated RNA.cDNA specimens were kept at À80 C.   method, followed by normalization for β2M and C/D box snoRNAs (SNORD) as an internal control for genes and miRNAs, respectively.

| Statistical analyses
Data analysis was performed with GraphPad Prism V.6 software (GraphPad; USA).Data were reported as mean ± standard deviation.
One-way analysis of variance (ANOVA) followed by Dunnett's multiple comparison test was carried out by analyzing several groups, and an independent Student's t-test was used to compare cancerous vs. control tissues.Pearson's correlation analysis was also performed for the miRNA-mRNA co-expression.Moreover, receiver operating characteristic (ROC) curve analysis was conducted to investigate the diagnostic power of GATA3 and miRNAs.p-values <.05 were considered statistically significant.

| In silico analysis
The KEGG and WikiPathways databases were searched for finding genes contributing to the EMT and invasion signaling pathways in BC cells.Moreover, the expression of searched genes was evaluated in the UniGene and the human protein atlas and the GATA3 gene was chosen for further in vitro investigation.GATA3 gene with ability to regulate cancer-related signaling pathways out of all the existing genes, is the sixth most important mutated gene in BC in the TCGA (The Cancer Genome Atlas Program) database (https://portal.gdc. cancer.gov) which has a high specific expression in breast tissue and limited studies have been done on it.Furthermore, using PicTar, miR-TarBase, miRWalk, and TargetScan Human Version 7.2 to find the probable target miRNAs for the GATA3 gene, miR-92a-3p and miR-1245b-5p were selected considering high binding score to GATA3 gene, their predicted significant relationship with the breast-related diseases and limited study in the EMT phenomenon in BC.Other important genes in the BC invasion which are potentially regulated by desired miRNAs are detailed in Table 2.Moreover, expressed genes in the breast tissue that contribute to EMT and invasion and are potential targets of the desired miRNAs were summarized in Table 3, and were subjected to STRING-db v11 and visualized by Cytoscape (Figures 1-5).The pathways associated with desired miRNAs were determined by miRPath v3.0 and the output was depicted in the heatmap (Figure 6).The expression levels of miR-1245-5p were also significantly decreased in the cancer tissue compared to the surrounding normal tissues (À4.35-fold, p < .01)(Figure 7B).Also, miR-1245b-5p showed a significant correlation with lymph node and ER status ( p < .05)and not with stage/grade of patients with BC ( p > .05).In addition, GATA3 mRNA was upregulated 5.46 folds in BC tissues ( p < .01)compared to healthy tissues (Figure 7C).

| miRNAs and GATA3 correlation with the clinical profile of BC
Analytical evaluation of the correlation of the GATA3 expression with clinicopathological features revealed that the GATA3 expression decreased ( p < .001,p < .0001)as the grade and stage increase (namely tumor progression).Moreover, the expression level of GATA3 was higher in ER + compared to ER À patients ( p < .001),and in the Lamin A subgroup compared to the other subgroups ( p < .0001).In regard to miR-92-3p, the expression of this miRNA further decreased in ER À patients ( p < .001).Furthermore, there was an inverse correlation between the grade, stage, and tumor size and the expression of this miRNA, suggesting that as the grade and stage increase, the miR-92-3p expression further decreased ( p < .0001).Moreover, the expression of this miRNA in patients whose lymph nodes were involved was significantly reduced compared to patients whose lymph was not affected ( p < .001).Among all the clinicopathological features, miR-1245b-5p was correlated with lymph and ER status (p < .05),and the expression of this miRNA decreased in ER negative patients and patients whose lymph was unaffected.

| Correlation among miRNAs and GATA3 expression
The relative expression levels of miR-1245b-5p and miR-92a-3p were compared to the GATA3 expression in all samples using Pearson's correlation coefficient analysis.The results showed a slight correlation and statistical insignificance between the expression levels of miR-92a-3p and GATA3 (Pearson's correlation = 0.3, p > .05).No correlation between miR-1245b-5p and GATA3 was observed.

| The ROC curve
The ROC curve was conducted to determine the specificity and sensitivity of GATA3 and the studied miRNAs during the differentiation of BC and control tissues.As shown in Figure 8, the area under the ROC curve (AUC) for GATA3 corresponded to 0.7415 (p < .0001),with the F I G U R E 1 Interactions of genes implicated in BC pathogenesis and predicted targets of miR-92a-3p and miR-1245b-5p using STRING db.As the line thickness increases, the interaction of proteins with each other also increases.It is predicted that GATA3 interacts with many proteins participating in the EMT and invasion pathways involved in BC pathogenesis.
F I G U R E 2 BC-related genes that may interact with miR-1245b-5p and miR-92a-3p were visualized in Cytoscape after inputting to STRINGdb.Genes in green, blue, or red circles display the targeted genes of miR-92a-3p, miR-1245b-5p, or both of them, respectively.
invasion and thus induces senescence.It sensitizes mesenchymal-TNBC cells to Dasatinib, thus inhibits tumor growth and suppresses cell migration. 39Guo et al. showed that the miR-539 overexpression suppressed tumor growth in BC by targeting the EGF receptor. 40Liu et al. concluded that miR-30e decreased metastasis, invasion, and chemotherapy resistance by suppressing IRSI. 8Because of the importance of miRNAs in the prediction, diagnosis, and treatment of various cancers, 9,10,41,42 we studied miR-1245b-5p and miR-92a-3p miRNAs in BC.Our bioinformatics results predicted that these two miRNAs could potentially bind to the 3'UTR region of GATA3.
Furthermore, the expression profile of these miRNAs and their potential target gene (GATA3) was assessed using qRT-PCR in the BC tissues.Moreover, the ROC curve analysis showed a significant diagnostic accuracy of miR-1245b and GATA3 between BC-patients and healthy individuals.
So far, the role of miR-92a-3p in BC has not been completely determined.The expression levels of this miRNA are even different in various breast tumor samples.Several studies have reported an increase or decrease in the expression of this miRNA.For example, a study showed that miR-92a-3p-3p expression was higher in the tumor tissue than in the normal tissue and suggested that the increased expression of this miRNA was associated with poor prognosis of BC. 12 In contrast, in our study, a decrease in the expression of this miRNA was observed in cancer tissues.Moreover, Moi et al. observed that the expression of miR-92a was significantly associated with several clinicopathological features, such as the grade of BC specimens. 43 well as, using the in-situ hybridization method, it was found that reducing the miR-92a-3p expression is inversely related to tumor grade and recurrence-free survival (RFC). 44Another study showed that tumors with a low miR-92a-3p expression are associated with The potential interactions of genes implicated in the EMT process using the STRING db.The thickness of the lines indicates the interaction level of genes with each other.
advanced tumor stages and poor patient survival. 23In agreement, our findings showed a significant invert correlation between the miR-92a-3p F I G U R E 6 The functions of miR-92a-3p and miR-1245b-5p in cancer progression were elucidated by gene ontology enrichments of their target genes via miRPath v3.0 based on TargetScane and microT-CDS.The color intensity (red to yellow) indicates the extent to which the genes of this pathway are regulated by the mentioned miRNAs.As the color intensity increases, the effect of that miRNA on the signaling pathway increases.
and also ER expression with miR-92a-3p expression in BC cells. 12In agreement, our results showed a decreased expression of miR-92a-3p in ER À BC patients.However, we not found a significant correlation between the miR-92a level and tumor size.
The activity and contribution of miR-1245b-5p in BC pathogenesis, as well as EMT and invasion, has yet to be fully characterized.Yang et al. found that the upregulation of miR-1245 was associated with breast and lung cancer progression. 45Another study found that upregulation of the c-Myc induced the miR-1245 expression, leading to BRACA2 suppression in BC. 25 Furthermore, the upregulation of miR-1245 has been shown to reduce NKG2D receptor expression in natural killer cells. 46Also, miR-1245 has been shown to accelerate colon cancer cell invasion and proliferation by targeting BRACA2. 47Weiyan Lou et al. demonstrated that miR-1245b-5p was downregulated in drug resistance BC patients. 48In agreement, our data showed the decreased expression of miR-1245b-5p in BC tissues compared to surrounding health tissues.Moreover, the assessment of clinical factors showed that the miR-1245b-5p expression was reversely associated with lymph vascular invasion and ER expression in BC patients.
However, no significant correlation was found between other clinical factors and the miR-1245b-5p expression.
7][58][59] A high level of GATA3 expression has been reported in luminal types of BC. 60  BC samples. 54,61,62In agreement, our results showed that the expression of GATA3 in the luminal subtype of BC was higher than that in other subtypes.However, the GATA expression was higher in the BC-affected patients with lymph node involvement than in BC-affected patients without lymph node involvement.
In addition, using in silico tools, several target genes of miR-92a-3p and miR-1245b involved in EMT and invasion pathways were predicted.Our results confirmed the potential contribution of studied miRNAs in regulating these signaling pathways.Moreover, the gene ontology analysis by DIANA miRPath v. 3 tools showed several pathways related to metabolism, biosynthesis, and transcription that potentially targeted by these miRNAs.These in silico results support the possible involvement of these miRNAs in BC pathogenesis.
Despite the limitations of this study including the low fund to evaluate a greater number of miRNAs and target genes related to BC, it is hoped that this study can be effective to provide novel information in future on the predictive diagnostic and therapeutic efficacy of these biomarkers and target genes in various types of cancer.

| CONCLUSION
Our findings showed an increased level of GATA3 expression, as well as, a decreased level of miR-1245b-5p and miR-92a-3p expression in BC tissues.Moreover, bioinformatics analysis confirmed miR-1245b-5p and miR-92a-3p as potential regulators of signaling pathways involved in BC pathogenesis especially invasion and EMT pathways.So, it seems that these genetic markers can be effective in the diagnosis and treatment of BC and other types of cancer.

A
set of primers were designed by Primer 3 software (F: 5'AT GCCTGCCGTGTGAAC3' and R: 5'ATCTTCAAACCTCCATGATG3' for β2M) and (F: 5'GAGACAGAGCGAGCAACG3' and R: 5'CTCGGGTCA CCTGGGTAG3' for GATA3) to amplify the β2M and GATA3 gene.The quantitative real-time RT-PCR was carried out using YTA SYBER Green qPCR Master Mix 2 Â (Cat# YT2551, Yekta Tajhiz Azma, Tehran, Iran) and BONmiR High-Specificity miRNA qPCR Core Reagent Kit (Cat# 209002, Bonyakhteh, Tehran, Iran) to detect the mRNA and miRNA levels.The thermal program for carrying out GATA3 gene amplification was set as 95 C for 3 min, and then 40 cycles for 15 s at 95 C, for 30s at 60 C, and for 20s at 72 C. Amplification of miR-92a was carried out for 2 min at 95 C, 40 cycles for 15 s at 95 C, and for 30s at 60 C. Results were analyzed according to the Livak or 2 ÀΔΔCt

T A B L E 1
The clinicopatological characteristics of BC patients and tissues.

3. 2 |
Expression analysis of miR-1245b-5p, miR-92a-3p, and GATA3 in the breast samples Expression analysis of miR-92-3p and miR-1245-5p in the cancerous and surrounding normal breast tissue samples was performed by qRT-PCR.A significant downregulation of miR-92a-3p in tumor tissues, especially in the primary grade compared to normal tissues was obtained (aÀ 3.22 fold reduction and p < .01)(Figure7A).Moreover, the miR-92a-3p expression showed a significant correlation with differentiation grade, stage, and expression level of estrogen receptor (ER) in tumor tissue (p < .0001,p < .0001,and p < .001,respectively).T A B L E 2 (Continued) level and the grade, stage, and lymph node status.As, the decreased level of miR-92a-3p accompany with the increase of grade and stage of tumor and also involving lymph node in patients with BC.Furthermore, Cun et al. reported a significant association between tamoxifen resistance, F I G U R E 5 The potential interactions of genes implicated in the EMT process with miR-92a-3p and miR-1245b-5p.(in databases including WikiPathways, MiRWalk, MirBase, Target scan, and MiRTarBase).The potential target genes of miR-92a-3p (green), miR-1245b-5p (purple), and common targets (pink) are shown in the picture.
Mehra et al. showed the expression of GATA3 was significantly related to the stage, grade, lymph node, tumor size, and the ER expression.Moreover, the highest level of GATA3 expression was reported in the luminal A subtype ofF I G U R E 7The expression levels of miR-92a-3p (A), miR-1245b-5p (B), and GATA3 (C) are measured by the formula 2 (ÀΔΔ CT) , and fold changes are presented as mean ± standard deviation.The asterisk shows the statistical significance between patients and healthy controls (** p < 0.01).