Computational Derivation Of A Molecular Framework For Hair Follicle Biology From Disease Genes

[InBeforeTheCure]

Computational derivation of a molecular framework for hair follicle biology from disease genes
Rachel K. Severin, Xinwei Li, Kun Qian, Andreas C. Mueller & Lynn Petukhova

Abstract:

Knowledge about genetic drivers of disease increases the efficiency of interpreting patient DNA sequence and helps to identify and prioritize biological points of intervention. Discoveries of genes with single mutations exerting substantial phenotypic impact reliably provide new biological insight, although such approaches tend to generate knowledge that is disjointed from the complexity of biological systems governed by elaborate networks. Here we sought to facilitate diagnostic sequencing for hair disorders and assess the underlying biology by compiling an archive of 684 genes discovered in studies of monogenic disorders and identifying molecular annotations enriched by them. To demonstrate utility for this dataset, we performed two data driven analyses. First, we extracted and analyzed data implicating enriched signaling pathways and identified previously unrecognized contributions from Hippo signaling. Second, we performed hierarchical clustering on the entire dataset to investigate the underlying causal structure of hair disorders. We identified 35 gene clusters representing genetically derived biological modules that provide a foundation for the development of a new disease taxonomy grounded in biology, rather than clinical presentations alone. This Resource will be useful for diagnostic sequencing in patients with diseases affecting the hair follicle, improved characterization of hair follicle biology, and methods development in precision medicine.

Excerpt:

Hippo signaling has been extensively studied in the contexts of cancer and development, and has been shown to influence tumor or organ size through the regulation of cell proliferation and apoptosis21,22. In fact, the pathway was originally given its name because genetic perturbations thereof generated “hippopotamus-sized” organs22. Interestingly, the most common form of hair loss, androgenetic alopecia (i.e. male pattern baldness; male pattern baldness), has long been characterized as a process of organ miniaturization, whereby hair follicles continue to cycle but undergo a reduction in size, resulting in a transition from thick terminal hair to fine vellus hair23. While Hippo signaling has yet to be specifically implicated in male pattern baldness, there is preliminary genetic evidence that is consistent with such a hypothesis. The largest male pattern baldness GWAS performed to date included a gene-based analysis that identified 112 autosomal genes with genome-wide significant association (Bonferroni correction of α < 2.769e-06)24, four of which are annotated to participate in Hippo signaling within the Kyoto Encyclopedia of Genes and Genomes (KEGG; pathway hsa04390), including WNT6, WNT10A, WNT3, and CTNNB1. We performed pathway enrichment analysis of these 112 genes and identified hsa04390:Hippo signaling pathway at a significance level of p = 0.049. Three of these genes reside at loci that were also associated with male pattern baldness in an independent GWAS (WNT6, WNT10A, WNT3)25. Our analysis of genes that establish a hair follicle cellular signaling network identified the Hippo pathway and a set of 59 genes that link this pathway to Wnt, Notch, Hedgehog and p53 signaling pathways (Fig. 1). A definitive role for Hippo signaling in the pathogenesis of male pattern baldness awaits further investigation.