鉴定 RNA 修饰相关的可变剪接特征作为头颈部鳞状细胞癌的一个独立因素。
Identification of RNA Modification-Associated Alternative Splicing Signature as an Independent Factor in Head and Neck Squamous Cell Carcinoma.
机构信息
Department of Otorhinolaryngology Head and Neck Surgery, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, China.
出版信息
J Immunol Res. 2022 Sep 13;2022:8976179. doi: 10.1155/2022/8976179. eCollection 2022.
OBJECTIVE
Head and neck squamous cell carcinoma (HNSCC) is a highly heterotopic malignant tumor. Alternative splicing (AS) and RNA modification have been reported to be involved in tumorigenesis. Therefore, we constructed RNA modification-associated AS (RMA-AS) signature model to predict the prognosis of HNSCC.
METHODS
AS events and RNA-modified gene expression information were downloaded from TCGA-HNSCC database. Univariate Cox regression analysis was employed for analyzing prognosis-related AS events. RMA-AS events were obtained by constructing a coexpression network between RNA modification-associated genes and AS events using WGCNA package. The prognostic signatures were analyzed by LASSO, univariate Cox, and multivariate Cox regression. Kaplan-Meier survival analysis, proportional hazard model, and ROC curve were performed to verify the prognostic value. "ESTIMATE" R package, ssGSEA algorithm, and CIBERSORT method were used to detect immune microenvironment in HNSCC. Cytoscape was utilized to build a regulatory network of splicing factor-regulated RMA-AS.
RESULTS
There were 16,574 prognostic AS events and 4 differentially expressed RNA modification-associated genes in HNSCC. Based on RMA-AS events, we obtained a risk model consisting of 14 AS events, named RMA-AS_Score. The samples were divided into RMA-AS_Score high- and RMA-AS_Score low-risk groups, according to the risk score. The RMA-AS_Score high group was related to poor prognosis. Moreover, the RMA-AS_Score signature was an independent prognostic predictor and was related to tumor grade. Meanwhile, the AUC value of RMA-AS_Score was 0.652, which is better than other clinical characteristics. Besides, a nomogram prediction model of quantitative prognosis has also been developed, which has robust effectiveness in predicting prognosis. In addition, the prognostic signature was observably related to immune microenvironment and immune checkpoint. Finally, 14 splicing factors were identified and constructed into a network of splicing factor-regulated RMA-AS.
CONCLUSION
We identified the RMA-AS signature of HNSCC. This signature could be treated as an independent prognostic predictor.
目的
头颈部鳞状细胞癌（HNSCC）是一种高度异质性的恶性肿瘤。已有报道称，选择性剪接（AS）和 RNA 修饰参与了肿瘤的发生。因此，我们构建了 RNA 修饰相关 AS（RMA-AS）特征模型，以预测 HNSCC 的预后。
方法
从 TCGA-HNSCC 数据库中下载 AS 事件和 RNA 修饰基因表达信息。采用单因素 Cox 回归分析筛选与预后相关的 AS 事件。使用 WGCNA 包构建 RNA 修饰相关基因与 AS 事件之间的共表达网络，获得 RMA-AS 事件。通过 LASSO、单因素 Cox 和多因素 Cox 回归分析构建预后特征模型。采用 Kaplan-Meier 生存分析、比例风险模型和 ROC 曲线验证预后价值。使用“ESTIMATE”R 包、ssGSEA 算法和 CIBERSORT 方法检测 HNSCC 中的免疫微环境。使用 Cytoscape 构建剪接因子调控的 RMA-AS 调控网络。
结果
在 HNSCC 中，有 16574 个预后 AS 事件和 4 个差异表达的 RNA 修饰相关基因。基于 RMA-AS 事件，我们获得了一个由 14 个 AS 事件组成的风险模型，命名为 RMA-AS_Score。根据风险评分将样本分为 RMA-AS_Score 高风险组和 RMA-AS_Score 低风险组。RMA-AS_Score 高风险组与不良预后相关。此外，RMA-AS_Score 标志物是独立的预后预测因子，与肿瘤分级相关。同时，RMA-AS_Score 的 AUC 值为 0.652，优于其他临床特征。此外，还建立了一个用于定量预后预测的列线图预测模型，该模型在预测预后方面具有稳健的效果。此外，该预后标志物与免疫微环境和免疫检查点显著相关。最后，鉴定出 14 个剪接因子，并构建了一个剪接因子调控的 RMA-AS 网络。
结论
我们鉴定了 HNSCC 的 RMA-AS 特征。该标志物可作为独立的预后预测因子。
Zotero 插件