Table_6_Development and validation of an autophagy-related long non-coding RNA prognostic signature for cervical squamous cell carcinoma and endocervical adenocarcinoma.xlsx

BackgroundIn this study, we aimed to investigate the signature of the autophagy-related lncRNAs (ARLs) and perform integrated analysis with immune infiltration in cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC).
Methods and resultsThe UCSC Xena and HADb databases provided the corresponding data. The ARLs were selected via constructing a co-expression network of autophagy-related genes (ARGs) and lncRNAs. Univariate Cox regression analysis combined with LASSO regression and multivariate Cox regression analysis were utilized to screen lncRNAs. The ARL risk signature was established by Cox regression and tested if it was an independent element bound up with patient prognosis. We used the xCell algorithm and ssGSEA to clarify the pertinence between immune infiltration and the expression of ARLs. Finally, we predicted the sensitivity of drug treatment as well as the immune response. Results indicated that the three prognostic ARLs (SMURF2P1, MIR9-3HG, and AC005332.4) possessed significant diversity and constituted the ARL signature. Risk score was an individual element (HR = 2.82, 95% CI = 1.87–4.30; p < 0.001). Immune infiltration analysis revealed significant increases in central memory CD8+ T cells, endothelial cells, CD8+ naive T cells, and preadipocytes in the high-risk group (p < 0.05). There were 10 therapeutic agents that varied significantly in their estimated half-maximal inhibitory concentrations in the two groups. According to the experimental validation, we found that SMURF2P1 belongs to the co-stimulatory genes and might assume greater importance in the development of cervical adenocarcinoma. MIR9-3HG and AC005332.4 belonged to the tumor-suppressor genes and they may play a more positive role in cervical squamous cell carcinoma.
ConclusionsThis research explored and validated a novel signature of the ARLs, which can be applied to forecast the prognosis of patients with CESC and is closely associated with immune infiltration.

【背景】本研究旨在探究自噬相关长链非编码RNA（autophagy-related lncRNAs, ARLs）的特征，并针对宫颈鳞状细胞癌与宫颈内腺癌（cervical squamous cell carcinoma and endocervical adenocarcinoma, CESC）开展免疫浸润整合分析。
【方法与结果】本研究从UCSC Xena数据库与HADb数据库获取对应数据。通过构建自噬相关基因（autophagy-related genes, ARGs）与长链非编码RNA的共表达网络，筛选出自噬相关长链非编码RNA。采用单变量Cox回归分析结合最小绝对收缩和选择算子（Least Absolute Shrinkage and Selection Operator, LASSO）回归与多变量Cox回归分析完成长链非编码RNA的筛选。通过Cox回归构建自噬相关长链非编码RNA风险特征模型，并验证其是否为与患者预后密切相关的独立影响因素。本研究使用xCell算法与单样本基因集富集分析（single-sample gene set enrichment analysis, ssGSEA）明确免疫浸润与自噬相关长链非编码RNA表达之间的相关性，最终预测药物治疗敏感性与免疫应答情况。
结果显示，3种预后相关自噬相关长链非编码RNA（SMURF2P1、MIR9-3HG及AC005332.4）具有显著表达差异，共同构成了自噬相关长链非编码RNA风险特征模型。风险评分是独立预后影响因素（风险比HR=2.82，95%置信区间95%CI=1.87~4.30；p<0.001）。免疫浸润分析显示，高风险组的中枢记忆性CD8+T细胞、内皮细胞、初始CD8+T细胞及前脂肪细胞水平显著升高（p<0.05）。两组间共有10种治疗药物的估算半最大抑制浓度（half-maximal inhibitory concentration, IC50）存在显著差异。经实验验证发现，SMURF2P1属于共刺激基因，在宫颈腺癌发生发展中可能发挥更为关键的调控作用；MIR9-3HG与AC005332.4属于抑癌基因，在宫颈鳞状细胞癌中可能发挥更为积极的抗肿瘤作用。
【结论】本研究探究并验证了一种新型的自噬相关长链非编码RNA风险特征模型，该模型可用于预测宫颈鳞状细胞癌与宫颈内腺癌患者的预后，且与免疫浸润状态密切相关。