What Fragment Size to Delete When Removing an Enhancer

For an isolated enhancer, delete the accessible footprint plus a buffer of roughly 50–150 bp on each side — typically a 300 bp to 1 kb deletion for a single element. Larger deletions (2–5 kb) are appropriate only when you are deliberately removing a cluster or a region containing multiple redundant elements, and they must then be interpreted as region-level perturbations rather than element-level ones.

The number that matters is not the size per se; it is whether the deleted interval contains anything other than the element you intend to study.

1. Define the element. Use the intersection of accessibility and active marks in your cell type, not a database annotation alone — annotations are cell-type-agnostic and often wider than the functional core.
2. Note the summit. ATAC and factor-occupancy summits mark the functionally critical bases; the deletion must contain the summit, not merely overlap the peak.
3. Add buffer. 50–150 bp per side accounts for nucleosome context and imperfect footprint boundaries. Buffer beyond that adds risk without adding coverage.

A 400-bp deletion containing the summit of a 600-bp accessible region is usually a better experiment than a 3-kb deletion containing the whole region and two neighbours.

The third row is the one that most often invalidates a clean result: deleting an interval that contains a bidirectional promoter produces a phenotype that has nothing to do with the enhancer.

• • Super-enhancer clusters. Removing the whole cluster answers "does the region matter"; the constituent elements should then be deleted individually in follow-up, as described in our piece on super-enhancer validation design.
• • Redundancy suspected. If shadow enhancers are present, single-element deletions may be silent; the informative design is a series (single, double, triple), not a larger single deletion. Our article on redundancy controls sets out that logic.
• • Regulatory domain hypotheses. Testing whether an entire domain is required is a legitimate question, provided the claim is stated at domain level.

In each case the interpretation changes: you are testing a region, and the manuscript should say so.

• • Place breakpoints in unique sequence. Junction PCR and Sanger confirmation fail in repeats; check uniqueness before ordering.
• • Keep at least one breakpoint away from the summit if you intend to re-insert the element later; otherwise the donor design has no unique landing sequence.
• • Check for SNPs in the deleted interval in your cell line, particularly if the locus is being studied for a variant; an unintended heterozygous deletion in a diploid line will show up as a dosing artefact.
• • Confirm zygosity by digital PCR or sequencing across the junction; a heterozygous deletion reported as a knockout understates effects by roughly half.
• • Design two independent deletions with different breakpoints spanning the same element. Concordance is what converts a deletion into evidence.

For a typical 500–800 bp accessible enhancer with a single summit, in a diploid cell line:

• • Deletion 1: −250 bp to +250 bp around the summit (≈500 bp).
• • Deletion 2: −350 bp to +150 bp, offset breakpoints (≈500 bp).
• • Controls: a non-targeting clone from the same campaign, and a deletion of an element known to be dispensable in the same cells.

Two independent deletions of the same core, plus a matched control, is a stronger package than one 3-kb deletion — and it costs less to interpret.

EDITGENE designs and builds enhancer deletion panels with mapped footprints, sequenced breakpoints, confirmed zygosity, and clone-matched controls — single elements, clusters, and combinatorial series for redundancy testing.

Send us the locus and the annotation you are starting from. Contact EDITGENE to request a project consultation or quote.

Recommended sizes are indicative; the appropriate interval depends on footprint width, local element density, and the claim being made. Deletions larger than the element test a region, not an element.

Zygosity, junction sequence, and copy number should be confirmed for every clone used in interpretation.

• • Canver MC, et al. BCL11A enhancer dissection by Cas9-mediated in situ saturating mutagenesis. Nature. 2015;527:192–197.
• • Fulco CP, et al. Systematic mapping of functional enhancer–promoter connections with CRISPR interference. Science. 2016;354:769–773.
• • Kvon EZ, et al. Progressive loss of function in a limb enhancer during snake evolution. Cell. 2016;167:633–642.
• • Osterwalder M, et al. Enhancer redundancy provides phenotypic robustness in mammalian development. Nature. 2018;554:239–243.