Bone Metabolism Laboratory, Rheumatology Division Faculdade de Medicina FMUSP, Universidade de Sao Paulo, Sao Paulo, SP, Brazil.
Instituto do Coracao (InCor), Hospital das Clinicas HCFMUSP, Faculdade de Medicina, Universidade de Sao Paulo, Sao Paulo, SP, Brazil.
Aging (Albany NY). 2021 Sep 7;13(17):20992-21008. doi: 10.18632/aging.203505.
Despite the well-established association of gene expression deregulation with low muscle mass (LMM), the associated biological mechanisms remain unclear. Transcriptomic studies are capable to identify key mediators in complex diseases. We aimed to identify relevant mediators and biological mechanisms associated with age-related LMM. LMM-associated genes were detected by logistic regression using microarray data of 20 elderly women with LMM and 20 age and race-matched controls extracted from our SPAH Study (GSE152073). We performed weighted gene co-expression analysis (WGCNA) that correlated the identified gene modules with laboratorial characteristics. Gene enrichment analysis was performed and an LMM predictive model was constructed using Support Vector Machine (SVM). Overall, 821 discriminating transcripts clusters were identified (|beta coefficient| >1; -value <0.01). From this list, 45 predictors of LMM were detected by SVM and validated with 0.7 of accuracy. Our results revealed that the well-described association of inflammation, immunity and metabolic alterations is also relevant at transcriptomic level. WGCNA highlighted a correlation of genes modules involved in immunity pathways with vitamin D level (R = 0.63, = 0.004) and the Agatston score (R = 0.51, = 0.02). Our study generated a predicted regulatory network and revealed significant metabolic pathways related to aging processes, showing key mediators that warrant further investigation.
尽管基因表达失调与肌肉减少症（LMM）之间存在明确的关联，但相关的生物学机制仍不清楚。转录组学研究能够识别复杂疾病中的关键介质。我们旨在确定与年龄相关的 LMM 相关的相关介质和生物学机制。通过使用从我们的 SPAH 研究（GSE152073）中提取的 20 名 LMM 老年女性和 20 名年龄和种族匹配的对照的微阵列数据，使用逻辑回归检测与 LMM 相关的基因。我们进行了加权基因共表达分析（WGCNA），将鉴定的基因模块与实验室特征相关联。进行了基因富集分析，并使用支持向量机（SVM）构建了 LMM 预测模型。总体而言，鉴定了 821 个有区别的转录物簇（|beta 系数|>1；-值<0.01）。在这个列表中，通过 SVM 检测到 45 个 LMM 的预测因子，并以 0.7 的准确率进行了验证。我们的结果表明，炎症、免疫和代谢改变的良好描述性关联在转录组水平上也是相关的。WGCNA 突出了参与免疫途径的基因模块与维生素 D 水平（R = 0.63， = 0.004）和 Agatston 评分（R = 0.51， = 0.02）之间的相关性。我们的研究生成了一个预测的调控网络，并揭示了与衰老过程相关的重要代谢途径，显示出需要进一步研究的关键介质。
