Derivation of marker gene signatures from human skin and their use in the interpretation of the transcriptional changes associated with dermatological disorders.
Derivation of marker gene signatures from human skin and their use in the interpretation of the transcriptional changes associated with dermatological disorders.
复制标题
DOI：
10.1002/path.4864
复制
发表时间：
2017-04
期刊：
影响因子：
--
通讯作者：
Freeman TC
中科院分区：
文献类型：
--
作者：
Shih BB;Nirmal AJ;Headon DJ;Akbar AN;Mabbott NA;Freeman TC
许多研究探索了与皮肤病相关的转录景观的改变，以了解这些疾病的本质。然而，由于构成该组织的许多特殊细胞类型缺乏良好的标记集，因此数据解释是一个重大挑战，这些细胞的组成可能在疾病期间发生根本性的改变。在这里，我们试图导出定义构成人类皮肤的各种细胞类型和结构的表达特征，并演示如何使用它们来帮助解释来自该器官的转录组数据。确定了两个大型正常皮肤转录组数据集，一个是 RNA-seq（n = 578），另一个是微阵列（n = 165），经过质量控制并分别进行基于网络的分析，以识别稳健共表达基因的簇。然后结合生物信息学分析、文献和专家评审来确定这些簇的生物学意义。经过分析之间的交叉比较后，定义了 20 个基因特征。这些包括毛囊、腺体（皮脂腺、汗液、大汗腺）、角质形成细胞、黑素细胞、内皮细胞、肌肉、脂肪细胞、免疫细胞和许多途径系统的表达特征。我们将此资源统称为 SkinSig。然后使用 SkinSig 分析 18 种皮肤状况的转录组数据集，提供这些数据的上下文解释。例如，传统分析表明，随着年龄的增长，角化和脂肪代谢会减少；我们更准确地将这些变化定义为由于毛囊和皮脂腺的丧失所致。 SkinSig 还强调了皮肤病中各种细胞类型的过度/不足，反映出炎症性疾病中免疫细胞的涌入以及其他细胞类型的相对减少。总体而言，我们的分析证明了这种新资源在定义皮肤细胞类型和附属器的功能概况以及改进疾病数据的解释方面的价值。 © 2016 作者。 《病理学杂志》由 John Wiley & Sons Ltd 代表大不列颠及爱尔兰病理学会出版。
Numerous studies have explored the altered transcriptional landscape associated with skin diseases to understand the nature of these disorders. However, data interpretation represents a significant challenge due to a lack of good maker sets for many of the specialized cell types that make up this tissue, whose composition may fundamentally alter during disease. Here we have sought to derive expression signatures that define the various cell types and structures that make up human skin, and demonstrate how they can be used to aid the interpretation of transcriptomic data derived from this organ. Two large normal skin transcriptomic datasets were identified, one RNA‐seq (n = 578), the other microarray (n = 165), quality controlled and subjected separately to network‐based analyses to identify clusters of robustly co‐expressed genes. The biological significance of these clusters was then assigned using a combination of bioinformatics analyses, literature, and expert review. After cross comparison between analyses, 20 gene signatures were defined. These included expression signatures for hair follicles, glands (sebaceous, sweat, apocrine), keratinocytes, melanocytes, endothelia, muscle, adipocytes, immune cells, and a number of pathway systems. Collectively, we have named this resource SkinSig. SkinSig was then used in the analysis of transcriptomic datasets for 18 skin conditions, providing in‐context interpretation of these data. For instance, conventional analysis has shown there to be a decrease in keratinization and fatty metabolism with age; we more accurately define these changes to be due to loss of hair follicles and sebaceous glands. SkinSig also highlighted the over‐/under‐representation of various cell types in skin diseases, reflecting an influx in immune cells in inflammatory disorders and a relative reduction in other cell types. Overall, our analyses demonstrate the value of this new resource in defining the functional profile of skin cell types and appendages, and in improving the interpretation of disease data. © 2016 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of Pathological Society of Great Britain and Ireland.