Identification of long non-coding RNA signatures in triple-negative breast cancer

Breast cancer (BC) is the most common type of cancer and the leading cause of cancer death among women all over the world. [1]. Triple-negative breast cancer (TNBC) is a particular subtype of breast cancer, characterized by poor prognosis because of its aggressive biological behavior and lack of molecular targets for therapy [2]. It is defined by the absence of estrogen receptor (ER) and progesterone receptor (PR) expression and without amplification of human epidermal growth factor receptor 2 (HER2) [3]. The treatment methods for TNBC are very limited owing to the lack of decisive therapeutic targets. Hence, it is necessary to explore new targeted approaches and make efforts to improve the outcomes of TNBC.

In recent years, long non-coding RNAs (lncRNAs) have drawn an increased attention because of their functions in the human diseases including cancers. LncRNAs are a class of RNA transcripts with a length of > 200 nucleotides that do not encode proteins. They are involved in diverse biological processes such as cell proliferation, differentiation, chromosome remodeling, epigenetic modulation, transcriptional and posttranscriptional modifications [4]. Studies have revealed that lncRNAs play an important role in cancer biology and the expression level or mutation of specific lncRNA genes are implicated in the development and progression of cancer. Moreover, a large number of lncRNAs are deregulated in multiple tumors including breast cancer, making them possibly as diagnostic and prognostic biomarkers or as potential therapeutic targets for cancer [5, 6]. Several lncRNAs have been reported to regulate TNBC progression. For instance, lncRNA LINP1 is overexpressed and enhances double-strand DNA break repair in TNBC. Blocking LINP1 increases sensitivity of BC cell response to radiotherapy [7]. LINK-A facilitates the activation of BRK kinase, thus activates normoxic HIF1α signaling in TNBC, promoting breast cancer glycolysis reprogramming and tumorigenesis [8]. With the development of RNA sequencing and genomic technologies as well as computational techniques, more and more lncRNAs have been discovered. However, the studies about lncRNAs and TNBC are very few by far, and the expression profile, functions and mechanisms of lncRNAs in TNBC remains to be extensively explored [9, 10]. Thus, we mined and analyzed data from several databases, hoping to highlight signatures of lncRNAs in TNBC and provide foundation for further studies.

Among all the breast cancer subtypes, TNBCs account for approximately 15% to 20% of all diagnosed breast cancer cases and are more prevalent in younger women (age < 40 years) [15]. TNBC is a complex and heterogeneous disease and the outcomes of patients are relatively worse than those of other subtypes. Only 30–45% of TNBC patients can achieve a pathological complete response (pCR) and survival rates similar to other types of breast cancer [16]. The poor prognosis of TNBC is mainly due to the lack of effective targets for treatment. Therefore, it is crucial to find new therapeutic targets for the improvement of TNBC prognosis.

LncRNAs play an important role in carcinogenesis. Many lncRNAs are dysregulated in tumors, and they are promising diagnostic biomarkers and potential therapeutic targets for cancers [17‐19]. In this study, we identified a number of TNBC-associated lncRNA genes through bioinformatic methods. Most of them are novel lncRNAs, many of which even do not have an official name. All of the overlapped lncRNAs in both microarrays have not been studied in BC except MEG3. So, they are good targets for future research. There are also some lncRNAs which have been extensively studied previously. For example, the most dysregulated lncRNA gene, BCAR4, has been found to be overexpressed in breast tumor tissue in previous studies and was associated with poor survival of breast cancer patients [20, 21]. Furthermore, it has been proved that BCAR4 can promote breast cancer cell migration and invasion through noncanonical hedgehog signaling pathway [21]. MEG3 is a tumor suppressor lncRNA gene, its expression is decreased in multiple tumors including lung cancer, gastric cancer, hepatocellular carcinoma, glioma etc. [22]. In breast cancer, it can inhibit cell proliferation, invasion and angiogenesis by sponging microRNAs and regulating signaling transduction such as AKT and TGF-β pathway [23, 24]. H19 is also one of the major lncRNA genes in cancer, but it has long been a controversy whether it is oncogenic or tumor-suppressive. H19 plays a role in tumor initiation and progression, the mechanisms, however, vary among cancer types [25, 26]. In breast cancer, H19 involves in tumor growth and metastasis through interaction with protein and microRNAs [27]. The mechanisms of lncRNA regulation in TNBC have not been clarified by far. Previous studies have shown that they can be regulated by some important signaling pathways. For example, LINP1 expression is activated by the EGF signaling and repressed by the p53 pathway in TNBC [7]. The expression level of lncRNAs can also be altered by epigenetic modification. For example, the promoter-associated CpG island of LOC554202 was hypermethylated, thus leading to the down-regulation of LOC554202 in TNBC cells [28]. In addition, lncRNA’s expression can be regulated by its biodegradation rate or transcription rate [29].

There were some similar studies published previously [30‐32]. But these studies only mined data from one microarray without making in-depth analysis. Our study comprehensively analyzed two datasets and made further analysis. We also identified numerous abnormally expressed PCGs in TNBC. And by establishing gene co-expression network, we found the PCGs whose expression profiles are correlated with that of lncRNA genes. Many of these PCGs were enriched in biological processes and pathways which are important for tumorigenesis and cancer progress. The 50 top genes ranked by degree in the network were selected as hub-genes, among which, low expression of three genes (NAP1L2, CRABP2 and SYNPO2) while high expression of two genes (MKI67 and COL4A6) was associated with poor RFS of TNBC patients. The products of hub PCGs mainly function as protein binding molecule and were involved in important biological processes and signaling pathways in cancer (CDK1, MKI67, CENPF, COL4A6, DACH1, etc.). As they are highly correlated with lncRNA genes, they may be the targets through which the TNBC-associated lncRNAs can influence the onset and progress of TNBC.

In summary, numerous lncRNA were dysregulated in TNBC and many of them are possibly involved in cancer development. The specific function of these lncRNAs needs further exploration. Nevertheless, our study illuminates a comprehensive understanding of lncRNA signatures in TNBC and suggests its important role. These dysregulated lncRNAs can be promising biomarkers for diagnosis or prognosis and may be potential targets for therapy. We hope these findings can draw more attention to lncRNAs in TNBC research and provide orientations for future studies.