Background

During embryonic development, cells with the same genome must selectively activate specific genes at different times and locations. A key element responsible for this precise regulation is the enhancer. Enhancers are cis-regulatory sequences where transcription factors and co-factors bind, and they can function thousands of base pairs to over one megabase away from the promoter of their target gene. The concept of enhancers as 'regulators relatively independent of distance and direction' was established in 1981 when a viral-derived sequence was shown to enhance the expression of the β-globin gene from 1,400 base pairs away.

However, enhancer function is not entirely independent of distance. As linear genomic distance increases, the probability of enhancer-promoter interaction decreases, and the average expression level and transcriptional precision of the target gene tend to decline. The mechanism by which enhancers select precise targets across intervening genes and insulator boundaries remains incompletely understood. This Nature Genetics Perspective does not report a new single experiment but synthesizes the molecular mechanisms and developmental and evolutionary advantages of long-range enhancer regulation.

Key Findings

The authors argue that long-range regulation cannot be fully explained by simple DNA loop formation. The first axis involves cohesin and CCCTC-binding factor (CTCF)-mediated loop extrusion. The ring-shaped cohesin complex pulls chromatin together, reducing the search space for distant enhancers and promoters. In the Sonic hedgehog (SHH) locus, cohesin is required for long-range enhancer function, and topologically associating domains (TADs) can mitigate the loss of regulatory strength with increased distance. However, acute removal of CTCF or cohesin still allows many enhancer-promoter contacts and transcription to persist, suggesting that loop extrusion is not a universal requirement.

The second axis is the cooperative action of multiple enhancers. A 2025 synthetic genome study showed that multiple enhancers working together can compensate for the reduced individual regulatory activity at long distances. A 2026 study reported that when the synergistic effect between regulatory elements is sufficient, cohesin dependency can be reduced. Enhancer clusters, super-enhancers, linker proteins such as LDB1, and 'facilitator' sequences are proposed to increase the frequency and duration of productive contacts.

Multivalent interactions between transcription factors and mediator complexes, as well as local transcriptional condensates, are also considered potential mechanisms. In this model, two DNA sequences do not need to physically touch but can share a concentrated transcriptional machinery to promote expression. The TATA·DPE motifs, CpG composition, and adjacent regulatory sequences in the promoter act as selective filters determining which enhancer signals are accepted. Ultimately, long-range regulation is interpreted as the combined result of 3D contact, sequence compatibility, cooperativity, and transcriptional burst dynamics.

Implications and Outlook

The structure of placing enhancers at a distance may appear inefficient, but it may confer advantages for developmental genes. If multiple tissue-specific enhancers are dispersed across a broad regulatory region, they can integrate distinct signals independently while controlling the same promoter. Redundant regulatory elements serve as a buffer to maintain expression if some enhancers are damaged, and they also allow for the addition of new enhancers or modification of existing sequences. The authors suggest that such a structure may have enhanced the diversity and evolvability of gene expression programs.

There are also clear medical implications. Variants in SHH long-range enhancers are linked to polydactyly, and the loss of a long-range enhancer in human neural progenitor cells can cause craniofacial developmental disorders. Structural variants or non-coding mutations that disrupt TAD boundaries or create incorrect enhancer-promoter connections can lead to disease, even if the coding region is intact.

However, some of the functional advantages proposed in the paper are closer to hypotheses than validated conclusions. Cell population-based contact maps average transient interactions, and physical proximity does not necessarily mean transcriptional activation. Combining single-cell and real-time imaging, precise genome rearrangement, and long-range high-throughput reporter assays will be necessary to clarify how the frequency and duration of contacts actually alter transcriptional bursts.