Liu Guoqing, Liao Wang, Lv Xiangwen, Zhu Miaomiao, Long Xingqing, Xie Jian
Department of Cardiology, The First Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi, China.
Department of Cardiology, The First People's Hospital of Yulin, Yulin, Guangxi, China.
Front Mol Biosci. 2024 Aug 21;11:1448705. doi: 10.3389/fmolb.2024.1448705. eCollection 2024.
Hypoxia has been found to cause cellular dysfunction and cell death, which are essential mechanisms in the development of acute myocardial infarction (AMI). However, the impact of hypoxia-related genes (HRGs) on AMI remains uncertain.
The training dataset GSE66360, validation dataset GSE48060, and scRNA dataset GSE163956 were downloaded from the GEO database. We identified hub HRGs in AMI using machine learning methods. A prediction model for AMI occurrence was constructed and validated based on the identified hub HRGs. Correlations between hub HRGs and immune cells were explored using ssGSEA analysis. Unsupervised consensus clustering analysis was used to identify robust molecular clusters associated with hypoxia. Single-cell analysis was used to determine the distribution of hub HRGs in cell populations. RT-qPCR verified the expression levels of hub HRGs in the human cardiomyocyte model of AMI by oxygen-glucose deprivation (OGD) treatment in AC16 cells.
Fourteen candidate HRGs were identified by differential analysis, and the RF model and the nomogram based on 8 hub HRGs and were constructed, and the ROC curves verified its good prediction effect in training and validation datasets (AUC = 0.9339 and 0.8141, respectively). In addition, the interaction between hub HRGs and smooth muscle cells, immune cells was elucidated by scRNA analysis. Subsequently, the HRG pattern was constructed by consensus clustering, and the HRG gene pattern verified the accuracy of its grouping. Patients with AMI could be categorized into three HRG subclusters, and cluster A was significantly associated with immune infiltration. The RT-qPCR results showed that the hub HRGs in the OGD group were significantly overexpressed.
A predictive model of AMI based on HRGs was developed and strongly associated with immune cell infiltration. Characterizing patients for hypoxia could help identify populations with specific molecular profiles and provide precise treatment.
缺氧已被发现可导致细胞功能障碍和细胞死亡，这是急性心肌梗死（AMI）发生发展的重要机制。然而，缺氧相关基因（HRGs）对AMI的影响仍不确定。
从GEO数据库下载训练数据集GSE66360、验证数据集GSE48060和单细胞RNA数据集GSE163956。我们使用机器学习方法在AMI中识别枢纽HRGs。基于识别出的枢纽HRGs构建并验证了AMI发生的预测模型。使用单样本基因集富集分析（ssGSEA）探索枢纽HRGs与免疫细胞之间的相关性。采用无监督一致性聚类分析来识别与缺氧相关的稳健分子簇。单细胞分析用于确定枢纽HRGs在细胞群体中的分布。实时定量聚合酶链反应（RT-qPCR）通过对AC16细胞进行氧糖剥夺（OGD）处理，验证了枢纽HRGs在AMI人心肌细胞模型中的表达水平。
通过差异分析鉴定出14个候选HRGs，并构建了基于8个枢纽HRGs的随机森林（RF）模型和列线图，ROC曲线验证了其在训练和验证数据集中具有良好的预测效果（AUC分别为0.9339和0.8141）。此外，通过单细胞RNA分析阐明了枢纽HRGs与平滑肌细胞、免疫细胞之间的相互作用。随后，通过一致性聚类构建了HRG模式，HRG基因模式验证了其分组的准确性。AMI患者可分为三个HRG亚组，A组与免疫浸润显著相关。RT-qPCR结果显示，OGD组的枢纽HRGs显著上调。
建立了基于HRGs的AMI预测模型，且与免疫细胞浸润密切相关。对缺氧患者进行特征分析有助于识别具有特定分子特征的人群，并提供精准治疗。
