A biotech startup has used gene editing to create beagle puppies that lack the main protein responsible for dog allergies — a proof-of-concept for genetically hypoallergenic pets.
New York-based Kindred Companion Sciences targeted Can f 1, a protein secreted in dog saliva and dander that accounts for most human dog allergies. Using CRISPR, the team made a single-base insertion in the Can f 1 gene, creating a frameshift that disrupts the protein’s production. Published in The CRISPR Journal.
Why a frameshift and not a deletion
The editing choice is more deliberate than it appears.
DNA is read in three-letter codons. Inserting a single base shifts the reading frame for everything downstream, so every subsequent codon is misread and a stop signal typically appears quickly — producing a truncated, non-functional fragment that the cell degrades.
It is a reliable way to eliminate a protein with a minimal edit. Deleting the whole gene would require cutting at two points and removing the intervening sequence, which is more disruptive and more likely to affect neighbouring regulatory elements.
“We didn’t introduce any foreign DNA … all we did was direct it at this specific site,” said CEO Matt Walker — a distinction that matters for how such animals would be regulated, since introducing genetic material from another organism is treated differently from altering an existing sequence.
What Can f 1 is
The protein belongs to the lipocalin family, small proteins that bind and transport hydrophobic molecules such as lipids and pheromones. Its exact function in dogs is not fully established, which is part of why deleting it was considered reasonable.
It is produced mainly in the tongue and salivary glands and reaches skin and fur through grooming, then disperses on shed dander. That distribution explains why dog allergen is so pervasive indoors and why it is detectable in homes and public buildings where no dog has ever lived.
What they found
The edits were made in two genetically identical beagle twins, Alfie and Bailey, born in September 2024.
Western blot testing detected no Can f 1 protein in the pups’ saliva or dander, and skin-prick tests showed no allergic reaction to extracts from the edited dogs — while control poodles and goldendoodles carried high allergen levels.
The control choice makes a point of its own: poodles and goldendoodles are widely marketed as hypoallergenic, and the finding that they carry high allergen levels is consistent with prior research showing no reliable allergen difference between so-called hypoallergenic breeds and others.
Researchers reported no off-target mutations or chromosomal rearrangements, and at nearly two years old both dogs are described as developing normally.
What the skin-prick result does and does not show
Skin-prick testing measures whether a person’s IgE antibodies recognise an extract, which is a genuine test of allergenicity and a limited one.
Dogs produce several other allergens — Can f 2 through Can f 6 among them — and while Can f 1 accounts for most reactivity, sensitised individuals vary in which proteins they react to. Someone primarily sensitised to a different allergen would not benefit.
Real-world exposure is also different from a skin-prick test: living with an animal means sustained contact with a complex mixture, and whether that produces symptoms is the question a controlled challenge study would need to answer.
Why it matters, and the caveats
About 15% of people worldwide have dog allergies, and this approach tackles the allergen at its genetic source rather than treating a person’s immune response.
That inversion is the conceptually interesting part. Allergy management has always addressed the person — antihistamines, immunotherapy, avoidance. Removing the allergen from the animal changes what is being modified.
What would have to happen next
Two beagles demonstrate that the edit works. Turning that into available animals raises questions the paper does not resolve.
Breeding is the practical route. The edit is heritable, so an edited dog bred conventionally passes the altered gene to offspring — meaning the population could expand without editing each animal, though establishing lines across breeds would take generations.
Regulation is unsettled. Gene-edited animals for food have been assessed by agencies under frameworks built for veterinary drugs, a process that has proved slow and contested. Companion animals fit that framework awkwardly, since the intended benefit is to a human who does not consume the product, and no established pathway exists.
There is also a demand question worth stating plainly. Allergy is one reason people do not own dogs, and the market for a specific edited breed at whatever premium this commands is unproven — a proof-of-concept in beagles is some distance from a business.
Still, this is early work in just two animals, and gene-editing companion animals raises ethical and welfare questions the field will have to grapple with. A protein whose function is not fully understood has been deleted from an animal that cannot consent, for a benefit accruing entirely to humans — and long-term consequences in dogs that are not yet two years old remain unknown. Research news, not veterinary or medical advice.