Combining genetic and single-cell expression data to expand the understanding of orofacial clefting

Abstract

Molecular malfunctions during craniofacial development can lead to non-syndromic cleft lip with or without cleft palate (nsCL/P), a congenital malformation that affects the upper lip and palate. Treatment involves multidisciplinary approaches, including surgery and speech therapy. In addition to an increased risk of morbidities such as cancer, neurological and cardiovascular diseases, these interventions can represent a burden for those affected. NsCL/P occurs between the fourth and tenth week of embryonic development and is one of the most common birth defects with a prevalence of about 1 in 1,000 live births. The etiology is described as multifactorial and polygenic, including environmental and numerous genetic factors. To date, various genetic studies have identified over 45 genomic loci that are associated with a risk for nsCL/P. However, most of the genetic risk variants map to non-coding regions of the genome, and the target genes and affected cell types are mostly unknown. The aim of the present thesis was to examine gene expression patterns of nsCL/P candidate genes to identify cell types that are potentially involved in the development of nsCL/P. For this purpose, published single-cell RNA sequencing (scRNA-seq) data from embryonic mice were re-analyzed. These data were then used to study the expression patterns of candidate genes that were identified by genome-wide association studies (GWAS) and whole-genome sequencing data. In addition to confirming gene expression patterns that were described in previous functional studies, this revealed that most candidate genes are specifically expressed in either one of two groups of cell types, epithelial or mesenchymal cells. After scRNA-seq data from the heads of human embryos became first available, a systematic analysis of the joint gene expression of nsCL/P GWAS candidate genes in these data showed that epithelial cells and HAND2⁺ pharyngeal arches are associated with nsCL/P candidate genes, complementing the previous research in murine scRNA-seq data. Co-expression network analysis in these cell types was then used to identify potential interactions between candidate genes and to prioritize candidate genes by combining the results with the initial GWAS data. The results were consistent with previously described gene-gene interactions and revealed potential new interactions and candidate genes. Together, these analyses determined nsCL/P-associated cell types and demonstrated a novel strategy for the prioritization of candidate genes based on a combination of GWAS and scRNA-seq data.