AS was a universal regulatory mechanism for gene expression that allowed a single gene to produce multiple unique mRNAs (Baralle & Giudice, 2017). A genome-wide study estimates that 90–95% of genes have experienced AS (Pan, Shai, Lee, Frey, & Blencowe, 2008). One of the molecular signs of cancer was AS abnormality (Ladomery, 2013). Previously, some studies exploring AS signatures have shown that aberrant AS guides many genes involved in cancer occurrence, transformation, and metastasis, and can be used as biomarkers and therapeutic targets for cancer diagnosis, prediction and prognosis (Oltean & Bates, 2014). In recent years, the development of bioinformatics methods and high-throughput sequencing technologies has helped to develop more reliable biomarker diagnostic methods (Gao, Zhong, Patel, Alur, & Vyas, 2017). These have provided effective help for fully understanding AS events in TNBC. Here we integrate the splicing map of TNBC patients with the clinical factors of TCGA to fully investigate the prognostic value of AS.
The evidence presented thus far supports that specific abnormal AS plays a significant part in the initiation, progression and metastasis of breast cancer. For example, ESRP1 and hnRNPM can alter the splicing pattern of exons in CD44 to produce different specific subtypes and correlate with the interstitial state of breast tumors. Among them, CD44v subtype is associated with TNBC carcinogenesis. It promotes the dryness of cancer cells by activating PDGFRβ/Stat3 cascade and PFKFB4-mediated glucose metabolism, which is an important factor in tumor metastasis (Xu et al., 2014; H. Zhang et al., 2019). Similarly, the splicing factor SRSF1, which is up-regulated in human breast tumor cells, promotes AS of the tumor gene MDM2 to produce a MDM2-ALT1 subtype with tumorigenic properties (Comiskey, Jacob, Singh, Tapia-Santos, & Chandler, 2015). Due to the importance of AS in cancer biology, more and more studies have focused on the clinical relevance of AS in malignant tumors.
In this study, we used several biomedical analysis methods to integrate AS event profiles and clinical information from TNBC patients into prognostic-associated AS. A splicing prognostic marker that can divide TNBC patients into subgroups with different survival outcomes was constructed. First of all, the AS signatures of 115 patients with TNBC were analyzed, followed by a comprehensive survival analysis and a powerful prognostic predictor. Nearly half of the AS events (1,428 AS events from 1,383 genes) were ES, indicating that AS is common phenomenon in TNBC, and ES is the most common type of splicing. This shows that AS has great potential in applications. ANAPC7 (APC7) and Nedd4L are central genes in PPI network analysis. It is worth noting that APC has been identified as essential for the pathogenesis of breast cancer (Khan, Arafah, Shaik, Mahale, & Alanazi, 2018). Some reports show APC plays a central part in inhibiting Wnt signaling pathways that control TNBC cell proliferation and differentiation (De et al., 2016; Lang et al., 2017; Lv et al., 2019). Guarnieri AL presented novel findings indicating that Nedd4L had a very important role in suppressing breast cancer (Guarnieri et al., 2018). In addition, we sought to study the underlying mechanisms of prognostic AS events in TNBC. It is worth noting that the transcription of these genes can be modified through exosomes-related pathways to modify protein markers, thereby participating in processes such as cell proliferation, migration and apoptosis. This was discovered from the CC aspect of the GO analysis in our current work. Functional enrichment analysis indicated that the ubiquitin-mediated proteolytic pathway is an important pathway of interference, consistent with studies of AS in breast and colorectal cancer (Xiong et al., 2018; Zhang, Duan, Cun, & Yang, 2019).
Based on the overall survival of the AS events, the prognostic features are the focus of our current study to facilitate monitoring the prognosis of patients with TNBC. Studies have shown that a variety of molecules can be used as a special diagnosis and independent prognostic marker for tumors. Recently, many biomarkers for breast cancer have been developed. A study developed a prognostic signature for the 19 gene associated with clinical prognosis in patients with BRCA (Su, Miao, Ye, Cui, & He, 2019). In addition, BCL2 and CD82 are both considered as potential and reliable biomarkers for breast cancer diagnosis (Wang et al., 2019). For TNBC, which is more difficult to cure, studies have also shown that there are related prognostic markers (Cai et al., 2019).
Here, we used a multivariate Cox regression model to screen out a series of AS events to promote clinical metastasis. The prognostic markers we present show an ideal performance in predicting the patient's clinical outcome. Ultimately, we achieved a combination of all available types of AS. The AUC of ROC for final prognostic predictor was 0.8, which was much higher than all prognostic indicators established using only one type of AS, indicating that the application of predictive power enhances the prognostic potential of TNBC patients. SF is an important regulator of AS events. Changes in AS events occur in a variety of tumors, suggesting that SFs may be important molecules in splicing disorders in cancer (Climente-Gonzalez, Porta-Pardo, Godzik, & Eyras, 2017). More and more people believe that many SFs have changed and participated in the development of BRCA cancer cells through various mechanisms (Anczukow et al., 2012; Dolfini, Andrioletti, & Mantovani, 2019; Gokmen-Polar et al., 2019). In our study, we not only explored the relationship between AS events and tumorigenesis, but also emphasized the potential role of SF in TNBC. A potential SF-AS related network was constructed between prognostic SF and the most significant survival-related AS events. AS events with good prognosis were positively correlated with splicing factor expression, while AS events with poor prognosis were negatively correlated with splicing factor expression. For example, ALYREF is an RNA binding protein that ligates to transcription. It has been reported that ALYREF may not only be a molecular marker for early detection of lymph node metastasis, but may also be effective in preventing oral squamous cell carcinoma metastasis (Saito et al., 2013). However, the role of ALYREF in TNBC has not been explored. Therefore, whether the down-regulation of some specific SFs leads to a reduction in favorable prognostic AS events and an increase in poor prognosis AS events requires further verification by functional experiments. But we have raised important questions about the potential key SFs in TNBC. Up-regulation or down-regulation of SFs expression may lead to abnormal splicing and differential expression of splice variants. Upregulation of certain oncogenic SFs could promote TNBC progression. Because of the exploration of prognostic prediction models, more personalized methods for different patients could precisely target and regulate AS in TNBC patients, providing a wealth of biomarker candidates and potential targets for TNBC treatment.