Oncogenic MicroRNAs: miR-155, miR-19a, miR-181b, and miR-24 enable monitoring of early breast cancer in serum

Left untreated, early-stage breast cancer (EBC) can progress into a deadly disseminated cancer and thus early detection is of utmost importance. In addition to BRCA mutations that occur within a relatively small patient cohort there exists other pathology risk factors such as hormonal receptors, HER2, Ki-67, and grade III that aide in determining the potential risk of disease progression or relapse. MicroRNAs (miRs) are additional factors that are related to tumor growth and are detectable in a patient specimen [1]. MicroRNAs are 18–24 nucleotide-long non-coding RNAs each of them able to inhibit hundreds of mRNAs. MicroRNAs are produced as genes and function to inhibit their target mRNAs based on complementarity, thus take part in normal cellular processes as well as BC oncogenesis [2]. MicroRNAs are exported outside the cells and due to their stability are readily detectable in peripheral blood samples. This feature can be very important for monitoring of miRs secreted from the expanding EBC tissue.

Work by others involving 76 BC samples and 10 control tissues and utilizing microarray technique followed by Northern blot validations demonstrated that miRs are deregulated in BC with a defined ‘signature’ often observed; decreased: miR-10b, miR-125b, miR-145, increased: miR-21, and miR-155 [3]. For example, expression of miR-155 in BC (normalized median = 1.75, range 0.95-11.45) and in control tissue (1.37, 1.11-1.88) was found significantly different according to ANOVA [3]. Moreover, additional oncomiRs such as miR-17-92 cluster is also deregulated in BC and contributes to transformed BC characteristics [4] possibly through significant changes in target gene expression [5]. Several oncomiRs are potentially involved in aggressiveness including let-7b, miR-30c, miR-148a, miR-181a, miR-181a*, and miR-181b based on next generation sequencing study of several cases of highly invasive micropapillary carcinoma [6].

Notably, several miRs including miR-24 and miR-181b displayed co-upregulation in sera and tumor tissue of BC patients (N = 13) compared to controls (N = 10) using sequencing approach [7]. While several miRs are repeatedly deregulated others are stable and show very little variability in sera such as: let-7a, miR-16, miR-93, miR-103, miR-192, and miR-451 in a study of gastric cancer patients [8] or: let-7a and miR-195 in BC patient plasma as compared to normal controls [9]. Let-7a levels are stable likely because it is widely expressed in somatic cells, while less expressed in rare stem cells and undifferentiated precursors [10]. In agreement, let-7a was found slightly downregulated in several malignancies including breast cancer samples [11] and upon BC tumorigenesis [12, 13].

Based on above-mentioned data the miR-155, miR-19a, miR-181b, and miR-24 represent very good candidates to monitor tumor growth in BC patients as serum molecular biomarkers (albeit there are possibly other oncomiRs involved in BC pathogenesis, aggressiveness, or tumor growth). A direct role of these oncomiRs in BC pathogenesis is supported by studies on their target genes respectively. For example, the oncomiR miR-155 is overexpressed in BC and functions to downregulate Suppressor of cytokine signaling-1 (SOCS1) leading to persistent activation of the Signal Transducer and Activator of Transcription-3 (STAT3) [14]. Indeed, upregulation of miR-155 upon diagnosis of BC and a decrease of its levels upon therapy was previously observed based on data using the spike-in control [15]. Interestingly, the decline of serum miR-155 levels following therapy was observed much earlier than that of other tumor-biomarkers [15]. Next, miR-19a, a member of the miR-17-92 cluster, affects BC pathogenesis by multiple mechanisms including inhibition of tumor suppressor PTEN [16]. Involvement of miR-181b is suggested in aggressive BC due to its role in DNA damage response by downregulating ATM and signal efficiency of PARP1 inhibitors in the treatment of triple-negative BC [17]. Recently, expression of miR-19a and miR-181b has been interlinked by a study showing that miR-19a-mediated inhibition of SOCS1 activates miR-181b expression (through STAT3) [18]. In BC pathogenesis, role of TGFβ-mediated epithelial-to-mesenchymal transition has been repeatedly suggested with a focus on miR-24 role in this process [19]. Taken together, microRNAs selected for this study are key molecules involved in tumor growth and aggressiveness of BC.

We herein studied expression of miR-155, miR-19a, miR-181b, and miR-24 in sera of 63 EBC patients and evaluated statistically significance of their expression throughout the therapy and in relation to clinical risk stratification. Our work suggests that the serum-oncomiR expression associates with EBC at diagnosis especially in high-risk patients and can be potentially useful for disease monitoring.

We herein present evidence that oncomiR levels are significantly different at initial diagnosis of EBC compared to time points following specific treatments (after surgery and after adjuvant therapy). Our results suggest that surgery had tremendous impact on three serum oncomiR levels while adjuvant therapy significantly decreased all of the four oncomiRs studied. This suggests that there exist relationship between EBC elimination throughout the therapy and serum oncomiR levels.

OncomiRs are often deregulated in solid cancers and leukemias and function in the gene-reprogramming of the cell to establish the phenotypic outcome. As such, their levels are accordingly detectable in patient sera at diagnosis and are thus strongly considered for diagnostic screening in cancer patients and their relatives [2]. Our other data confirm that BC tumor tissues indeed produced the studied oncomiRs and that their elevated levels in the sera reflect this (Additional file 2: Results 2). In addition, detection of oncomiRs can be useful for relapse detection however much longer follow up is needed. In our patient cohort, we observed BC relapse in three patients and this resulted in changes in oncomiR levels (Additional file 2: Results 3). However, as our cohort consisted of a low number of relapsed patients it did not allow more thorough analysis of oncomiRs. MiR-155 is one of the most studied miRs in BC patients and its levels were found consistently increased by others in the serum samples [15]. However, miR-155 production by a tumor is highly responsive to surgical therapy and therefore its levels decline very rapidly (similarly to our data). Contrary to this, miR-19a appears to be more stable and detectable following the surgical removal (as compared to miR-155, miR-181b, or miR-24). Finally, the levels of all studied oncomiRs decreased following the adjuvant therapy (time point III).

Our study demonstrated that the levels of oncomiRs for the high-risk patients’ drop less rapidly than in case of the low-risk patients. In addition, high-risk patients, unlike low-risk, do not achieve normal oncomiR levels following the adjuvant therapy (time point III). For example, miR-17-92 cluster appears to be very important factor in aggressive triple-negative BC [16]. Similarly, additional microRNA miR-181b was associated with more aggressive BC characterized by grading, Ki-67, and triple negativity (its pathogenic effect was linked to downregulation of ATM and influence on DNA damage response [17]). miR-24 has been associated with BC pathogenesis [19], but has not been so far suggested to be involved in BC aggressiveness. We also consider and cannot fully exclude the possibility that oncomiR levels drop less rapidly in high-risk patients due to post-surgical presence of tumor cells that are not however detectable by clinically-utilized techniques. However, both careful surgical removal of surrounding non-tumorous tissue as well as very low number of relapsed EBC patients suggests that the elevation reflects rather increased stability and possibly also slower turnover of the oncomiRs in the patient sera.

Longitudinal multivariate analysis of oncomiR expression in the EBC patient cohort strongly supports the clinical relevance of the four oncogenic microRNAs. In part it helps to associate observed deregulations in the serum samples with aggressive clinical parameters such as HER2, Ki-67, and grade III. This is consistent with previous analyses that were based on expression-data within tumor tissue [17] and some of the data with regard to miR-155 in sera were also consistent with others [20]. Our work thus provides important link in connecting biological function of miRs within the tumor with their levels in the sera.

Our data provide evidence that the expression of four BC pathogenesis-related oncomiRs (miR-155, miR-19a, miR-181b, and miR-24) in sera are increased at diagnosis of EBC patients. These oncomiR levels decrease following combined therapy. Importantly, high-risk EBC patients show notably delayed and less-pronounced decrease of oncomiR expression following surgical tumor removal. Elevated expression of the oncomiRs in sera was also observed in primary tumor tissues. Upon relapse, serum levels of some oncomiRs are increased indicating their potential for the EBC patient monitoring. Multivariate analysis of the oncomiR expression in the EBC patient cohort suggested that adverse clinical characteristics associate with elevated oncomiRs in the EBC sera. OncomiR-profiling in sera is very promising tool to be incorporated in further studies and clinical practice as well.