Liu Mian, Khushbu Rooh Afza, Chen Pei, Hu Hui-Yu, Tang Neng, Ou-Yang Deng-Jie, Wei Bo, Zhao Ya-Xin, Huang Peng, Chang Shi
Department of General Surgery, Xiangya Hospital Central South University, Changsha, China.
Clinical Research Center for Thyroid Disease in Hunan Province, Changsha, China.
Front Oncol. 2021 Oct 13;11:705929. doi: 10.3389/fonc.2021.705929. eCollection 2021.
Alternative splicing (AS) plays a key role in the diversity of proteins and is closely associated with tumorigenicity. The aim of this study was to systemically analyze RNA alternative splicing (AS) and identify its prognostic value for papillary thyroid cancer (PTC).
AS percent-splice-in (PSI) data of 430 patients with PTC were downloaded from the TCGA SpliceSeq database. We successfully identified recurrence-free survival (RFS)-associated AS events through univariate Cox regression, LASSO regression and multivariate regression and then constructed different types of prognostic prediction models. Gene function enrichment analysis revealed the relevant signaling pathways involved in RFS-related AS events. Simultaneously, a regulatory network diagram of AS and splicing factors (SFs) was established.
We identified 1397 RFS-related AS events which could be used as the potential prognostic biomarkers for PTC. Based on these RFS-related AS events, we constructed a ten-AS event prognostic prediction signature that could distinguish high-and low-risk patients and was highly capable of predicting PTC patient prognosis. ROC curve analysis revealed the excellent predictive ability of the ten-AS events model, with an area under the curve (AUC) value of 0.889; the highest prediction intensity for one-year RFS was 0.923, indicating that the model could be used as a prognostic biomarker for PTC. In addition, the nomogram constructed by the risk score of the ten-AS model also showed high predictive efficiency for the prognosis of PTC patients. Finally, the constructed SF-AS network diagram revealed the regulatory role of SFs in PTC.
Through the limited analysis, AS events could be regarded as reliable prognostic biomarkers for PTC. The splicing correlation network also provided new insight into the potential molecular mechanisms of PTC.
可变剪接（AS）在蛋白质多样性中起关键作用，且与肿瘤发生密切相关。本研究旨在系统分析RNA可变剪接（AS）并确定其对甲状腺乳头状癌（PTC）的预后价值。
从TCGA SpliceSeq数据库下载430例PTC患者的AS剪接百分率（PSI）数据。我们通过单变量Cox回归、LASSO回归和多变量回归成功鉴定出无复发生存期（RFS）相关的AS事件，然后构建了不同类型的预后预测模型。基因功能富集分析揭示了RFS相关AS事件涉及的相关信号通路。同时，建立了AS与剪接因子（SFs）的调控网络图。
我们鉴定出1397个RFS相关的AS事件，这些事件可作为PTC潜在的预后生物标志物。基于这些RFS相关的AS事件，我们构建了一个包含十个AS事件的预后预测特征，该特征可以区分高风险和低风险患者，并且具有很高的预测PTC患者预后的能力。ROC曲线分析显示十个AS事件模型具有出色的预测能力，曲线下面积（AUC）值为0.889；对一年RFS的最高预测强度为0.923，表明该模型可作为PTC的预后生物标志物。此外，由十个AS模型的风险评分构建的列线图对PTC患者的预后也显示出较高的预测效率。最后，构建的SF-AS网络图揭示了SFs在PTC中的调控作用。
通过有限的分析，AS事件可被视为PTC可靠的预后生物标志物。剪接相关网络也为PTC潜在的分子机制提供了新的见解。
