Regulation by partition predicts enhancer target selection in mammalian genomes

Enhancers play a crucial role in regulating specific promoters across long genomic ranges in mammalian genomes. However, the process of selecting among the numerous promoters within their reach has yet to be fully understood. Researchers have now developed a method called SPIE-seq (Synthetic Perturbation by Inserted Enhancer) to investigate this phenomenon.

SPIE-seq involves piggyBac-mediated enhancer insertion and an allele-specific readout of cis-regulatory effects on expression. By analyzing data from 724 clonal lines with 29,954 mapped insertions, the team identified enhancer-dependent expression changes at 18,440 enhancer-gene pairs.

The study reveals that enhancer influence diminishes with increasing genomic distance, following a consistent pattern of average distance-decay scaling across all tested enhancers. Nevertheless, the shape of response curves varies between loci. The genomic context surrounding an enhancer, rather than its identity, plays a significant role in determining which target genes are activated.

When the same enhancer is inserted at different positions within the same locus, it can activate distinct promoters. Promoters that respond to enhancers are often associated with developmental features, yet target selection remains competitive. Even at loci with a high density of promoters, an enhancer predominantly activates only one or two of them.

The emergence of promoter skipping, where promoters are not activated due to competition from other nearby promoters, is a common occurrence that escalates as more competitors accumulate. This pattern suggests that promoters do not respond independently based solely on their intrinsic features. To explain this complex process, researchers have proposed the Regulation-by-Partition (RbP) model.

According to RbP, an enhancer allocates its output among the promoters based on factors such as genomic distance, contact with the enhancer, and intrinsic features. The RbP model accurately predicts target choice and shows remarkable transferability across species, cell types, and perturbation directions, outperforming existing enhancer-gene models.

The model's superiority is particularly evident at long genomic ranges. Ultimately, the study concludes that enhancer target choice is not determined by enhancer or promoter attributes alone but emerges from a distance-weighted competition among promoters within the regulatory landscape.