Overexpression of circulating MiR-30b-5p identifies advanced breast cancer

Breast cancer (BrC) is the second most frequent cancer worldwide and the most commonly diagnosed cancer among women [1]. Despite improvements in early detection and treatment, BrC remains the leading cause of cancer-related death in women, mainly due to the development of recurrent and/or metastatic disease [1, 2]. In fact, at diagnosis, approximately 5% of patients present distant metastases and up to 10–15% of patients develop distant metastases within the first 3 years [3].

Metastatic BrC arises following the proliferation and dissemination of malignant cells from its primary location to distant organs [4]. Metastases often display an organ-specific pattern of spread, with bone constituting the most common site for BrC, followed by lungs, liver, and brain [3]. The development of metastases is the major prognostic factor for BrC patients, as metastatic BrC is incurable in most cases [5]. Thus, adequate patient monitoring, using minimally invasive biomarkers, that might anticipate the diagnosis of progressing disease, might allow for earlier implementation of adequate therapy, improving BrC patients’ survival and quality of life.

MicroRNAs (miRNAs) are endogenous, highly conserved small non-coding RNAs of approximately 22 nucleotides in length involved in genes’ posttranscriptional regulation [6]. Recently, the value of miRNAs as cancer biomarkers has been emphasized, emerging as promising diagnostic, prognostic and predictive biomarkers for BrC [7], including advanced BrC [8, 9]. Moreover, several lines of evidence implicate miRNAs either as promotors or suppressors of BrC metastization, by targeting multiple signaling pathways and important proteins that are major players at different steps of the metastatic process [8, 9].

Specifically, the role of miR-200 family (miR-200f) members in the initial steps of the metastatic cascade, by regulating ZEB1 and ZEB2 expression and consequently the E-cadherin expression has been described by several groups [10, 11]. Likewise, the miR-30 family (miR-30f) members have been previously reported to promote tumor invasion and metastases formation in melanomas [12], whereas in BrC downregulation was implicated in the metastatic process [13, 14]. Conversely, miR-182-5p upregulation was associated with epithelial-like state in BrC cell lines and macrometastases formation in vivo [15].

Nonetheless, we have previously shown that miR-30b-5p, miR-30c-5p, miR-182-5p and miR-200b-3p expression levels associated with patients prognosis in luminal BrC [16].

Thus, we set out a study aiming to assess the biomarker potential of the previously identified miRNAs expression for predicting BrC progression and dissemination in liquid biopsies.

BrC remains the most common malignancy in women and a major cause of morbidity and mortality [1]. Although biological features are routinely used for BrC diagnosis and prognosis assessment, patients with similar clinicopathological features often show different clinical outcomes [22]. Therefore, identification of biomarkers providing more accurate prognostic information for BrC patients, complementing currently used parameters, will have a major impact. Consequently, assessment of specific miRNAs expression deregulation, which has been associated with several mechanisms underlying BrC aggressiveness, might be a potential source for biomarkers [7, 23]. Most studies addressing miRNAs expression and miRNA-target validation have been performed in cancer cell lines lacking tumor-host interactions [7, 24]. Thus, tissue analysis might allow for a broader insight into biologically and clinically relevant miRNAs which may serve as prognostic biomarkers.

In previous work, we reported a panel of miRNAs related to luminal BrC patients’ prognosis, suggesting these miRNAs as potential progression biomarkers [16]. Hence, we aimed to evaluate the biomarker potential of miR-30b-5p, miR-30c-5p, miR-182-5p and miR-200b-3p for progression/metastization in BrC.

In the testing cohort, only miR-30b-5p and miR-200b-3p expression were significantly higher in metastatic lesions compared to matched primary BrC tissues. Although increased miR-182-5p expression levels have previously been detected in paired metastatic lymph nodes when compared with primary tumors [15], we were not able to confirm the same in our testing set of BrC tissues. This might be due to the small number of samples, the restricted molecular subtype of the primary tumor and the variability of metastases localization.

MiR-30b-5p and miR-200b-3p were further evaluated in a larger set of tumor tissue samples from BrC patients (validation cohort #1). It should be recalled that the stability of miRNAs in FFPE tissues holds an enormous potential [25], especially in BrC patients in which late relapses frequently occur, as demonstrated in the testing cohort and validation cohort #1. Herein, miR-30b-5p expression levels were found to be significantly higher in metastatic lesions compared to matched primary BrC tissues in the validation cohort #1, despite the patients studied are mostly luminal, and the rather limited number of non-bone metastases. Although downregulation of miR-30f members and its role as tumor suppressor during BrC invasion and metastization have been previously reported [13, 14], miR-30b-5p upregulation in BrC metastatic tissue samples has not been described thus far. The role of miR-30b-5p remains controversial. On one hand, expression of miR-30b-5p and miR-30c-5p has been associated with increased cell viability and resistance to apoptosis [26], and miR-30b-5p was found to be upregulated in several cancers [27‐29] and associated with metastases in melanoma [12]. On the other hand, miR-30b-5p was also associated with decreased migration and invasiveness in colorectal cancer [30, 31], and miR-30a was reported to be downregulated in primary nasopharyngeal carcinoma tissues but overexpressed in the corresponding metastases [32], although miR-30f members were shown to inhibit early steps of the metastatic process [33]. Our data clearly support an oncogenic role for miR-30b-5p in BrC.

Progression of solid malignancies is the result of a multi-step cascade in which tumor cells undergo widespread modifications to successfully migrate and colonize other organs. In this process, the dynamic ability to first undergo epithelial-to-mesenchymal transition (EMT) and subsequently mesenchymal-to-epithelial transition (MET) is an important feature of metastatic cells [34]. In support of the MET hypothesis, several studies have shown that metastatic lesions and the corresponding primary breast tumor have a similar epithelial nature [35, 36]. MiR-30b-5p modulation might be important in this plastic process. Downregulation of miR-30f members might lead to EMT initiation enabling cells to metastasize, while subsequent upregulation might be associated with MET, facilitating re-adaptation of the epithelial phenotype and colonization, crucial to develop macroscopic metastases. In fact, a recent study showed that downregulation of miR-30f members in primary BrC of patients without evidence of distant metastases was associated with poor relapse-free survival, which might be associated with the ability of decreased miR-30f levels to prompt EMT initiation. However, miR-30f expression levels were only evaluated in primary BrC tissue and were not assessed in metastatic lesions. Nevertheless, Kenny et al. [37] classified BrC cell lines into four morphological categories. BT-474 cell line, derived from a primary BrC, was included in Mass category characterized by tightly cohesive colonies with strong cell–cell adhesion, whereas MDA-MB-231, a metastatic cell line derived from a pleural effusion, was allocated to Stellate category characterized by limited cell–cell interactions and lacked E-cadherin expression, which are characteristics of EMT. Thus, these different phenotypes are in agreement with the previous hypothesis that decreased miR-30b-5p expression levels might lead to EMT initiation enabling cells with motility and invasive features. This might explain lower miR-30b-5p expression in MDA-MB-231 cells, and subsequent upregulation associated with MET, enabling re-adaptation of the epithelial phenotype, which was observed in BT-474 cells. Therefore, additional studies are needed to ascertain the miR-30b-5p functional role in BrC and the downstream transcriptional gene targets that might mediate distant metastases.

Knowledge of determining patterns of metastatic organ tropism might provide useful information for clinical evaluation of disease stage and to monitor progression. Hence, comparative analyses of miR-30b-5p expression according to the metastatic site were performed. Interestingly, bone metastases disclosed significantly higher miR-30b-5p expression levels compared to other metastases and, remarkably, primary BrC cases that metastasized to bone also displayed increased levels compared to those that did not. These results strongly suggest that not only miR-30b-5p plays a role in metastization, but it also predisposes tumor cells to home at specific organ sites, especially promoting bone colonization by tumor cells. Nevertheless, it should be recalled that miRNAs expression is highly context- and tissue-dependent, and thus, ideally, miRNA expression in normal tissues more prone to receive metastatic cells should also be assessed. We could then ascertain whether differential expression of miRNAs in the primary tumors versus metastatic tissues might be a consequence of their modulation in the metastatic microenvironment. Moreover, the mechanisms underlying tumor cell tropism to the bone and the extent to which metastatic cells miRNA’s profile differ according to their location may add valuable insights into disease development and clinical management. In fact, miR-30f members were reported to inhibit BrC bone metastases in an experimental model [38]. Nevertheless, these results were derived from triple negative BrC cell lines, which represent a (very) limited subset of BrC patients who do not commonly develop bone metastases. Nonetheless, it is well-recognized that BrC patients prone to develop bone metastases are mainly those which display luminal features [39, 40], and, thus this experimental model can hardly be considered representative of the clinically apparent heterogeneity.

Circulating miRNAs are stable in body fluids and their assessment might provide valuable diagnostic, prognostic and therapeutic prediction information, allowing for noninvasive testing and potential individual treatment optimization [7]. Recently, miR-30b-5p expression levels were shown to distinguish BrC patients from healthy controls in liquid biopsies [41], although miR-30b-5p expression levels have also been associated with aged [42]. Remarkably, we found that miR-30b-5p could discriminate patients with advanced BrC from those with localized BrC with high sensitivity, but modest specificity and overall accuracy, and no association with age was disclosed. Although the limited size of cohort #2 should be acknowledged, our results suggest that miR-30b-5p might identify, at diagnosis, patients who are more likely to endure disease progression. Particularly, circulating miR-30b-5p levels might provide a useful tool for early detection of BrC metastases, if these findings are further proven in a larger set of patients.

Several other miRNAs have been implicated in BrC invasion and metastasis [7]. MiR-10b was found highly expressed in tissue samples from patients with metastatic BrC [43] and, more recently, found to be significantly more expressed in tissues from patients with stage III and IV BrC compared to early-stage disease [44]. Moreover, circulating miR-10b combined with miR-373 might identify BrC lymph node metastasis with 72% sensitivity and 94.3% specificity [45], and miR-21 overexpression was significantly correlated with lymph node metastases, advanced clinical stage and poor prognosis [46]. Notwithstanding the tissue series size (n = 113), no stage IV patients were included in this study [46]. Similarly, circulating miR-21 discriminated stage IV BrC patients with visceral metastasis from those with stage I, II and III disease, with 86% specificity and 70% sensitivity [47]. Nonetheless, as far as we know, these metastasis-related miRNAs were only evaluated in primary BrC tissue and were not assessed in a larger tissue series of primary tumors and the corresponding metastatic lesions. Of note, in our study, circulating miR-30b-5p expression levels identified advanced disease with higher sensitivity, although with limited specificity.

Our results indicate that miR-30b-5p is overexpressed in metastatic BrC, suggesting an important role in tumor dissemination. Interestingly, bone metastases and their correspondent primary tumors displayed higher miR-30b-5p expression levels, suggesting a role in modulation of metastatic organ tropism. Importantly, advanced BrC patients disclosed significantly higher circulating miR-30b-5p expression levels compared to patients with localized BrC, suggesting that circulating miR-30b-5p levels might identify BrC patients at higher risk of disease progression, providing a useful clinical tool for patient monitoring, entailing earlier and more effective treatment. Nonetheless, validation in larger multicentric cohorts is needed to further sustain our findings.