Towards detection of enhancer and promoter activities in single cells

Towards detection of enhancer and promoter activities in single cells

The majority of disaese-associated genetic variants fall into intergenic, regulatory regions rather than the genes they control. Hence, the path towards identifying the cell types and developmental stages that are altered by genetic variants inevitably leads through locating and quantitatively assessing these regulatory regions and the enhancers and promoters they harbor.

            However, studying the activity of regulatory regions genome-wide in specific cells has been challenging. State-of-the-art approaches have low sensitivity, require extensive input material, or are limited to cell lines rather than disease-relevant tissue samples. Also, the lack of benchmarks has caused inconsistent accuracy estimates and thus an inability to compare assays and computational methods.

            To address these issues, we devised a new, combined approach based on the enrichment of nuclear, capped RNAs (nucCAGE) and a light-gradient boosting machine algorithm (PRIMEloci). This allowed to detect transcriptionally active regulatory regions with unmatched sensitivity and to map enhancers and promoters enriched in fine-mapped eQTLs, pathogenic GWAS variants, and CRISPRi loci. Based on the FANTOM5 enhancer atlas, we drastically expanded the landscape of known enhancers, mapped them in a cell type-specific manner and outlined inter-organ mechanisms for genetic variants influencing physiological and disease traits.

            In parallel, we develop experimental and computational methods to record active regulatory elements over time and in vivo on the level of genomic DNA rather than RNA via DNA adenine methylation identification (DamID). As a first step, we benchmarked all existing software dealing with DamID data. This framework allowed developing the most sensitive model for DamID data analysis and aids ongoing efforts to record enhancer and promoter activities in single cells.

Together, our methods aim to link regulatory elements to their cell-type specific biological functions, facilitate variant interpretation and advance insights into gene regulation during development and disease progression.