Purpose: The aldo-keto reductase (AKR) superfamily members have been proposed with multiple roles in various tumors. Here, a comprehensive analysis on the integral role of AKR genes was conducted to evaluate the expression profile, regulation network, and prognostic significance in hepatocellular carcinoma (HCC).
Materials and Methods: Transcriptome datasets of HCC were obtained from the Cancer Genome Atlas (TCGA) and Gene Expression Omnibus. Univariate and multivariate Cox regression analyses were used to build a novel risk score model, and then were further used to identify independent prognostic factors for overall survival (OS) of HCC. A prognostic nomogram was developed and validated. The expression of these critical AKR members was also evaluated by quantitative real-time polymerase chain reaction and immunohistochemistry in HCC specimens.
Results: Eight differentially expressed AKR genes were identified in HCC. The dysregulation of most AKR genes was negatively correlated with DNA methylation, and a regulation network with transcription factors (TFs) was also established. Then, three critical AKR genes (AKR1B10, AKR1D1, and AKR7A3) were screened out to build a novel risk score model. Worse OS was observed in high-risk patients. Besides, a prognostic nomogram based on the model was further established and validated in both the TCGA and GSE14520 cohorts, which showed superior performance in predicting the OS of HCC patients. Notably, close correlations were identified between the risk score and tumor immune microenvironment, somatic mutation profiles, and drug susceptibilities of HCC. Finally, the upregulated AKR1B10 and downregulated AKR1D1 and AKR7A3 were further verified in HCC tumor and adjacent tissues from our institution.
Conclusion: The dysregulated AKR genes could be mediated by DNA methylation and TFs in HCC. The risk model established with superior prognostic performance further suggested the significant role of AKR genes involved in the progression of HCC.

Keywords: hepatocellular carcinoma, aldo-keto reductase, AKR, risk score model, nomogram, prognosis


中文翻译：

---

肝细胞癌中醛酮还原酶（AKR）超家族基因的调控网络及预后意义

目的：已提出醛酮还原酶 (AKR) 超家族成员在各种肿瘤中具有多种作用。在这里，对 AKR 基因的整体作用进行了综合分析，以评估肝细胞癌 (HCC) 中的表达谱、调控网络和预后意义。
材料和方法：HCC 的转录组数据集来自癌症基因组图谱 (TCGA) 和 Gene Expression Omnibus。单变量和多变量 Cox 回归分析用于构建新的风险评分模型，然后进一步用于确定 HCC 总生存期 (OS) 的独立预后因素。开发并验证了预后列线图。这些关键 AKR 成员的表达也通过定量实时聚合酶链反应和免疫组织化学在 HCC 标本中进行了评估。
结果：在 HCC 中鉴定出 8 个差异表达的 AKR 基因。大多数AKR基因的失调与DNA甲基化呈负相关，并且还建立了具有转录因子（TFs）的调节网络。然后，三个关键的 AKR 基因（AKR1B10、AKR1D1, 和AKR7A3 ) 被筛选出来建立一个新的风险评分模型。在高危患者中观察到更差的 OS。此外，基于该模型的预后列线图在 TCGA 和 GSE14520 队列中进一步建立和验证，在预测 HCC 患者的 OS 方面表现出优异的性能。值得注意的是，风险评分与肿瘤免疫微环境、体细胞突变谱和 HCC 药物敏感性之间存在密切相关性。最后，在我们机构的 HCC 肿瘤和邻近组织中进一步验证了上调的 AKR1B10 和下调的 AKR1D1 和 AKR7A3。
结论：失调的 AKR 基因可能由 HCC 中的 DNA 甲基化和 TF 介导。建立的具有优异预后性能的风险模型进一步表明 AKR 基因在 HCC 进展中的重要作用。

关键词：肝细胞癌，醛酮还原酶，AKR，风险评分模型，列线图，预后