Division of Biostatistics and Bioinformatics, Department of Environmental and Public Health Sciences, University of Cincinnati College of Medicine, Cincinnati, OH 45267, USA.
AbbVie Inc., North Chicago, IL 60064, USA.
Int J Mol Sci. 2021 Apr 22;22(9):4371. doi: 10.3390/ijms22094371.
In the connectivity map (CMap) approach to drug repositioning and development, transcriptional signature of disease is constructed by differential gene expression analysis between the diseased tissue or cells and the control. The negative correlation between the transcriptional disease signature and the transcriptional signature of the drug, or a bioactive compound, is assumed to indicate its ability to "reverse" the disease process. A major limitation of traditional CMaP analysis is the use of signatures derived from bulk disease tissues. Since the key driver pathways are most likely dysregulated in only a subset of cells, the "averaged" transcriptional signatures resulting from bulk analysis lack the resolution to effectively identify effective therapeutic agents. The use of single-cell RNA-seq (scRNA-seq) transcriptomic assay facilitates construction of disease signatures that are specific to individual cell types, but methods for using scRNA-seq data in the context of CMaP analysis are lacking. Lymphangioleiomyomatosis (LAM) mutations in TSC1 or TSC2 genes result in the activation of the mTOR complex 1 (mTORC1). The mTORC1 inhibitor Sirolimus is the only FDA-approved drug to treat LAM. Novel therapies for LAM are urgently needed as the disease recurs with discontinuation of the treatment and some patients are insensitive to the drug. We developed methods for constructing disease transcriptional signatures and CMaP analysis using scRNA-seq profiling and applied them in the analysis of scRNA-seq data of lung tissue from naïve and sirolimus-treated LAM patients. New methods successfully implicated mTORC1 inhibitors, including Sirolimus, as capable of reverting the LAM transcriptional signatures. The CMaP analysis mimicking standard bulk-tissue approach failed to detect any connection between the LAM signature and mTORC1 signaling. This indicates that the precise signature derived from scRNA-seq data using our methods is the crucial difference between the success and the failure to identify effective therapeutic treatments in CMaP analysis.
在药物重定位和开发的连通图（CMap）方法中，通过患病组织或细胞与对照之间的差异基因表达分析来构建疾病的转录特征。假设疾病转录特征与药物或生物活性化合物的转录特征之间存在负相关，表明其具有“逆转”疾病过程的能力。传统 CMap 分析的一个主要限制是使用源自批量疾病组织的特征。由于关键驱动途径很可能在只有一部分细胞中失调，因此批量分析产生的“平均”转录特征缺乏有效识别有效治疗剂的分辨率。单细胞 RNA-seq（scRNA-seq）转录组分析的使用有助于构建针对单个细胞类型的特异性疾病特征，但在 CMap 分析背景下使用 scRNA-seq 数据的方法却缺乏。TSC1 或 TSC2 基因中的淋巴管平滑肌瘤病（LAM）突变导致 mTOR 复合物 1（mTORC1）的激活。mTORC1 抑制剂西罗莫司是唯一被 FDA 批准用于治疗 LAM 的药物。由于治疗中断后疾病复发，并且有些患者对药物不敏感，因此迫切需要治疗 LAM 的新疗法。我们开发了使用 scRNA-seq 分析构建疾病转录特征和 CMap 分析的方法，并将其应用于未经处理和西罗莫司治疗的 LAM 患者的肺组织 scRNA-seq 数据的分析中。新方法成功表明 mTORC1 抑制剂（包括西罗莫司）能够逆转 LAM 转录特征。模拟标准批量组织方法的 CMap 分析未能检测到 LAM 特征与 mTORC1 信号之间的任何联系。这表明，使用我们的方法从 scRNA-seq 数据中得出的精确特征是在 CMap 分析中成功识别有效治疗方法与失败之间的关键区别。
Zotero 插件
