Livestock gene editing and precision breeding

Gene editing in farm animals is often presented as breeding made faster. The more accurate description is that it addresses a different class of trait. Conventional and genomic selection are extremely good at quantitative traits — milk yield, growth rate, feed efficiency — because those are built from many loci of small effect and respond well to selection on estimated breeding values. Editing is good at the opposite: single loci of large effect, and in practice at destroying them.

Why the successes are loss-of-function

The pattern in the traits that have actually worked is consistent. Resistance to porcine reproductive and respiratory syndrome virus comes from altering CD163, the macrophage receptor the virus requires for entry; removing the relevant domain leaves the animal without the door the virus uses. The SLICK coat phenotype, associated with a variant of the prolactin receptor gene PRLR, gives cattle a short sleek coat and better heat tolerance, and the causal variants are truncating. Polled cattle — hornless by genetics rather than by dehorning — are produced by introducing an allele that already exists in beef breeds.

The reason is mechanistic. Breaking a gene is a single, well-defined molecular operation with a predictable phenotype when the gene’s role is known. Improving a quantitative trait requires knowing which of hundreds of loci to change, in which direction, and by how much, in a genetic background where effects are not additive. Nothing about the editing tool supplies that knowledge, so editing does not compete with genomic selection there; it complements it.

Two routes, two problems

Editing reagents delivered into a zygote act while the embryo is already dividing, so different cells can receive different outcomes and the founder animal is frequently mosaic — carrying a mixture of edited, unedited and differently edited cells, sometimes including the germline, sometimes not. Mosaic founders have to be bred out, which costs a generation, and a generation in cattle is years. The alternative is to edit a somatic cell line in culture, screen and sequence individual clones until a clean genotype is confirmed, and then use nuclear transfer to produce an animal from that verified cell. This gives genetic certainty at the cost of cloning’s own low efficiency and its documented developmental and perinatal losses. Which route dominates depends on the species’ reproductive biology as much as on the edit.

Off-target editing and unintended insertions at the target site are real and are addressed by whole-genome sequencing of founders — a check that is now routine and is what the regulatory dossier is largely made of.

The binding constraint is legal

Molecular feasibility is not what determines whether these animals reach a farm. Jurisdictions diverge on the central question of whether an edited animal is regulated by the process used or by the characteristics of the product. England’s Genetic Technology (Precision Breeding) Act 2023 creates a distinct route for organisms whose changes could have arisen through traditional breeding, with animal provisions phased separately from plants. In the United States intentional genomic alterations in animals fall under FDA oversight as animal drugs, with a risk-based framework that has produced individual approvals rather than a category exemption. Because these regimes classify the same animal differently, market access — not the edit — sets the timeline.