Controls Required for a Regulatory Element Knockout

A defensible enhancer or promoter knockout needs, at minimum: a non-targeting control processed through the same pipeline, two independent perturbations of the element, a positive control element known to matter in the same cell type, and matched clonal controls. Add a distal control element when specificity is claimed, and a rescue when the claim is mechanistic.

Everything below explains why each one earns its place, and what fails in its absence.

The row most often missing from submitted manuscripts is the positive control element — and it is the one that converts "we saw nothing" into "we can detect something".

A single guide or a single breakpoint produces a single data point with its own idiosyncrasies: the guide may cut inefficiently, the breakpoint may sit on a motif, the clone may carry an off-target edit. Two perturbations of the same element with different molecular positions control for all three at once.

If the two agree, the effect is attributable to the element. If they disagree, the discordance itself is informative — usually pointing to a second functional element near one breakpoint, which is exactly what you want to know before writing the paper.

The positive control must be an element with a known, large, cell-type-appropriate effect. Common choices:

• • A promoter or enhancer whose deletion is known to reduce a highly expressed gene in that lineage.
• • A core promoter motif disruption in a gene with robust expression.
• • A validated essential enhancer from the literature for that cell type.

Two requirements are frequently violated: the control must be measured in the same experiment, and it must produce an effect above the effect size you are trying to detect. A positive control that changes expression by 25% cannot validate an assay being used to claim a 30% effect is absent.

When you claim that the effect is specific to your element, a second deletion somewhere else in the genome is not sufficient — it controls for "deleting DNA", not for "deleting regulatory DNA". A stronger design deletes a different, active element in the same locus or a nearby element that is active but not expected to regulate the target gene.

This is the control that distinguishes "this element regulates this gene" from "this region contains regulatory activity".

Editing campaigns produce their own baseline shifts: transfection, single-cell cloning, and expansion under selection can each alter expression. Clones of the same genotype from the same campaign share that history; parental cells do not.

Practical approach: carry at least two non-targeting clones through the identical pipeline and use them as the comparison group, rather than comparing edited clones to a parental line that never saw a guide.

Re-inserting the element — wild-type, and where relevant a variant version — at the same locus tests whether the phenotype tracks the sequence rather than the editing process. It is technically demanding and frequently worth it: rescue converts a correlation between an edit and a phenotype into a causal statement, and reviewers treat it accordingly.

Where re-insertion is impractical, reverting the edit (for variant work) or complementing with an exogenous construct achieves a similar purpose at lower fidelity.

EDITGENE delivers element knockout clones with matched non-targeting controls, two independent perturbations per element, validated positive control elements, and sequenced junctions with confirmed zygosity — so the control structure is in place before the first experiment.

Send us the elements and the cell background. Contact EDITGENE to request a project consultation or quote.

Control requirements scale with the claim: a screening hit needs fewer controls than a mechanistic publication claim. The hierarchy above describes a publication-grade minimum.

Even a complete control set does not exclude redundancy with untested elements; see our discussion of enhancer redundancy and shadow enhancer controls.

• • Fulco CP, et al. Systematic mapping of functional enhancer–promoter connections with CRISPR interference. Science. 2016;354:769–773.
• • Canver MC, et al. BCL11A enhancer dissection by Cas9-mediated in situ saturating mutagenesis. Nature. 2015;527:192–197.
• • Tycko J, et al. Mitigation of off-target toxicity in CRISPR-Cas9 screens for essential non-coding elements. Nature Communications. 2019;10:4063.