Reprogramming endogenous regulatory DNA to fine-tune gene expression

Regulatory DNA sequences orchestrate cell-type-specific gene expression by facilitating transcription factor binding, yet their precise effects and reprogrammability remain challenging to delineate. Now, in Cell, Martyn et al. develop a method called variant effects from flow-sorting experiments with CRISPR targeting screens (Variant-EFFECTS), which measures the quantitative effects of changes to regulatory DNA on gene expression in endogenous contexts without the need for genetic engineering of reporters.

Variant-EFFECTS uses pooled prime editing to introduce hundreds of noncoding edits to regulatory sequences in cells. The cells are then labeled with RNA FlowFISH or a fluorescent antibody targeted to a gene of interest, and sorted on the basis of levels of fluorescence. To demonstrate the usefulness of the method, the authors performed tiling mutagenesis screens targeting the promotor and/or enhancer regions for two genes involved in immune responses (PPIF and IL2RA) in THP-1 cells and Jurkat T cells. This revealed several motif instances with strong effects on gene-expression levels, including some missed by massively parallel reporter assays. Variant-EFFECTS can also dissect the context-specific effects of transcription factor motifs by inserting them across cell types. Moreover, the authors used Variant-EFFECTS data to benchmark deep-learning models that predict gene regulatory signals directly from DNA sequence, which revealed limitations to these models for predicting effect sizes of variants.