Each tumor tissue as same as normal tissues has specific expression signatures. Every of these amounts that linked a biological case with a potential hazard, is called biomarker. There are several main attributes for an ideal biomarker, such as 1) non-invasive survey (present in the body fluid or peripheral tissues), 2) easy detection assay, 3) affordable and rapid test, and 4) related to specific result. Accordingly, miRNAs have cancer specific expressions, small size (rapid and cost effective experiment), resistance potency to degradation, and can be measured in serum or plasma samples easily (circulating miRNAs) (38, 39). Furthermore, miRNAs dysregulation disrupts the expression of oncogene or tumor suppressor genes and microenvironment balance agents, hence they are considered metastasis factors and prognosis predictors. (40). On the other hand, nowadays BC has exceeded lung cancer as the most commonly diagnosed cancer, recently (41). Therefore, many of researches concentrate to detect DEMs as candidate biomarkers that can be utilized to BC predisposition or management and reduce the disease burden.
Evangelista et al. examined miRNA and mRNA microarray hybridization on different cell lines according to molecular subtypes and reported miR-193 and miR-210 as potential biomarkers in BC. Also, CCND1 and RUNX3 were exhibited as miRNAs targets. Despite they considered several cell lines for each molecular subtype, they didn’t utilize any serum or tissue samples to further confirm their predictions. Hence, sensitivity and specificity rate were not screened. At the next step, they validated their findings via 513 breast tissues (456 cancer tissue and 57 normal adjacent) from TCGA database. Their miRNAs selection protocol included ANOVA test (p.value <= 0.01), and significant differentiation between two separate cell lines, (p.value <= 0.05). But their TCGA report, was only regarded second step of the protocol. In addition, they skipped the other significant miRNAs that resulted from one individual cell line. Eventually, they didn’t examine the miR-193 and miR-210 survival rate (42).
Furthermore, another study conducted by Mahmoudian et al. reported miR-25-3p, miR-29a-5p, miR-105-3p, miR-181b1-5p, miR-335-5p, and miR-339-5p as candidate biomarkers in BC. They extracted those miRNAs from literature reviews and BC experimental studies, but they didn’t clarify their selection criteria. The qPCR finally results from 50 BC tumor tissues and 50 adjacent non-tumor controls, showed p.value <0.0001 for miRNAs. However, they had to consider bigger sample size and high-throughput data analysis to support their findings more reliable. Then the survival estimation was performed and Kaplan Meier analysis rejected prognostic values of two miRNAs (miR-25-3p with p.value = 0.058 and miR-339-5p with p.value= 0.36). At last, all miRNAs diagnostic values were significant (p.value < 0.0001, 0.77 <= AUC), but they were not assessed in serum samples (43).
Orlandella et al. showed miR-622 was downregulated in plasma and tissue samples of patients with BC, and NUAK1 was validated as a direct target of miR-622. They selected miR-622 from other studies as a tumor suppressor and utilized two GEO microarray datasets to determine the mir622-NUAK1 correlation and TCGA data was used to reveal survival rate between different molecular subtypes. But none of them were implemented to predict the role of miR-622 in BC. Moreover, the significance of miR-622 was only examined on 20 BC specimen and 39 BC serum samples through real-time PCR, diagnosis potency was not conveyed as well. Finally, they didn’t determine other mir622 targets, because their main goal was to investigate the mir622- NUAK1 controversary (44).
In further study was done in 2021, it was highlighted that miR-30b-5p plays a role as a significant diagnostic biomarker in BC. They utilized miR-30b-5p through other researches, because miR-30b-5p was non evaluated in tissue and plasma concurrently and its prognosis and functions are still controversial, particularly in BC investigations. Then they assisted their findings with 769 BC tissues and 74 healthy normal ones from TCGA database. However, they didn’t use microarray datasets to further support their results, especially for serum samples that TCGA inspection was not possible. Furthermore, miR-30b-5p expression was measured in three cohorts, first BC tissues, the second and third plasma samples, with qPCR. Also, ROC curve analysis confirmed sensitivity and specificity value in the tissue and plasma samples but the relapse-free survival analysis was not performed. Finally, the targets of miR-30 family or miR-30b were briefly discussed, however miR-30b-5p direct targets and interactions were not displayed clearly (45).
Li et al. emphasized the overexpression of miR-19b in BC tissues and introduced it as a prognostic candidate. miR-19b and its targets were selected from researches. However, the authors didn’t validate the miR-19b by any microarray and RNA sequencing datasets. They examined 124 cancer tissues and 124 adjacent normal tissues by qPCR, though serum samples were not imported. miR-19b Prognostic value was showed significant efficacy (log-rank P=0.002) on cancer tissues but diagnosis potency was not measured. At last, according to previous studies, PI3K/AKT members were chosen as targets of miR-19b and p-PI3K, p-AKT, and CCND1 showed upregulation along with miR-19b overexpression (46).
Finally, Zhu et al. reported miR-1908-3p as an oncogenic miRNA that can be used to diagnose and prognose of BC. They didn’t use high-throughput data analysis to find their specific miRNA, but they chose miR-1908-3p from other experiments. Then, they confirmed miR-1908-3p by TCGA database. However, they didn’t mention to the sample size. Also, any type of microarray and RNA-seq serum samples were not recruited to support serum assessments. In order to prognostic prediction, Caplan Meier survival plot just indicated miR-1908 and didn’t distinguish miR-1908-3p. Afterward, to validate the miR-1908-3p in-silico findings, 50 tumor tissues and 60 cancer serum samples were examined by qPCR and its sensitivity and specificity quantities were displayed through the serum sample appraisal (47).
In the present study, we analyzed three serum samples (GSE73002, GSE113740, and GSE113486), two tissue samples (GSE53179 and GSE57897) and one whole blood samples (GSE83270) datasets from GEO microarray series. Then WGCNA was used to filter hub miRNAs from DEMs and TCGA was applied to confirm significance of hub miRNAs. Also, we constructed miRNA-mRNA interactions and defined hsa-miR-34a-5p, hsa-miR-501-3p, hsa-miR-1307-3p, hsa-miR-450b-5p, hsa-miR-532-5p, and hsa-miR-151a-5p as core miRNAs. At the next step, qPCR supported our in-silico results and ROC curve and overall survival analysis validated hsa-miR-34a-5p, hsa-miR-501-3p, hsa-miR-1307-3p, hsa-miR-450b-5p, hsa-miR-532-5p, and hsa-miR-151a-5p as candidate diagnosis and prognosis biomarkers.
Numerous studies have reported aberrant expression of these miRNAs associated with BC. Analysis of the miRNA-mRNA network from multiple studies reveals that miR-34a-5P, with the cooperation of specific proteins, plays a role in breast cancer and other gynecological cancers as well (48-50). As an example, hsa-miR-34a-5p was discovered in two separate studies published in 2017 and 2020 as a potential marker of metastasis (51, 52) and chemo-resistance (53, 54) in patients with BC. In a study published in 2019, by Hadavi et al. In a study published in 2019, by Hadavi et al. the role of miR-501-3p in BC confirmed (55). This research reported a negative correlation between miR-501-3p and its target genes, suggesting that this miRNA could be a new candidate for TNBC targeted therapies. Additionally, the role of this miRNA in other cancers and disorders has also been identified by another researches (56-58). It was also confirmed that miR-1307-3p plays a role in tumorigenesis of multiple cancers (59-61). Han et al. identified miR-1307-3p as an oncogene and its role in the progression of BC (62). The role of miR-1307-3p in the early diagnosis of BC has also been affirmed (63). Another novel miRNA discovered as a result of our research is miR-450b-5p, whose role in BC has not been studied but has been extensively studied in other cancers such as cervical cancer (64), nasopharyngeal carcinoma (65), hepatocellular carcinoma (66), etc. Moreover, miR-502-5p was concerned in BC. MiR-532-5p promotes proliferation and migration in BC, according to a study by L et al. (67), and its role as an upregulated biomarker in early stages was revealed by Tsai et al (68). Finally, the role of deregulated miR-151a-5p in various malignancies has been demonstrated by the findings of many investigations. The data provided by Ragle Aure et al. showed that the overexpression of miR-151a-5p increases the proliferation of BC cell lines (69). However, miR-151a-5p was confirmed as a novel biomarker for BC diagnosis and prognosis (70). In addition to BC, the function of this miRNA in other disorders and cancers has been reviewed and presented (71-73).