THE LANDSCAPE OF REGULATORY GENETIC VARIATION IN EARLY NEURAL DIFFERENTIATION

PINKASZ M; SCHOR IE

Congreso; SAIB-SAMIGE Joint Meeting 2021; 2021

Sociedad Argentina de Investigación en Bioquímica y Biología Molecular

The genomes of all living organisms are affected by genetic variability that in many cases can have functional effects and even phenotypic impact, including genetic susceptibility to different pathologies. Non-coding variants, for example those affecting transcriptional regulatory elements (such as promoters or enhancers), make an important contribution to phenotypic variability by modulating the expression levels of a number of genes. On the other hand, although biological systems are subject to different types of perturbations, their functions may not be altered since there are molecular mechanisms that give them robustness. These control mechanisms are particularly important during embryonic development, which typically proceeds in a stereotypic way. To study the impact that regulatory genetic variants have on key processes of animal development, we used single-cell RNA-seq data from cell differentiation of human embryonic stem cells (hESCs) (Yao et at. Cell stem cell,2017) to identify key regulators driving this process, and analyzed the distribution of single nucleotide variations (SNVs) in their promoters. We used the R package CORTO to infer a gene regulatory network based on gene co-expression, and found 484 regulators and 3.457 targets involved in this process. The identity of these two groups was supported by an over representation analysis of gene sets against Gene Ontology and Reactome databases. We selected promoter regions for these genes using the UCSC.hg38 genome annotation and searched for SNVs in these regions with information taken from NCBI dbSNP (common_all, build ID=151, genome assembly GRCh38), finding a total of 8.391 SNVs for regulators and 54.860 SNVs for targets. To identify the subset of potentially functional variants, we used the Combined Annotation Dependent Depletion tool (CADD) with a score cutoff of 15, resulting in 1.025 potential deleterious SNVs for regulators and 3.705 for targets. Strikingly, while regulators have a significant higher percentage of deleterious variants within their promoters than targets, the minor allele frequency (MAF) of these variants where significantly smaller, suggesting different selective forces acting in the promoter regions of these two groups. In order to further analyze the SNV distributions in terms of regulatory hierarchy, we performed differential expression analysis between differentiation stages in progenitors and neural cells, resulting in 64 and 98 differentially expressed regulators, that are candidate for driving cell state transitions. This first analysis identified promising variants in regulator genes driving neural differentiation that could have an impact in neural differentiation and function in humans. We have selected a set of regulators in progenitors and neural cells with multiple potentially deleterious variants in their promoters to experimentally test the effects of the different variants and combinations in transcriptional activity.