Phage and microbiome preparations in animals

A bacteriophage is a virus that infects bacteria. It attaches to a specific surface structure on a specific host, injects its genome, and — if it is lytic — redirects the cell to build progeny and bursts it. That is the whole mechanism, and every property that makes phage attractive in animal production, and every property that makes it commercially awkward, follows from the specificity of the first step.

Specificity cuts both ways

Attachment requires a receptor: a lipopolysaccharide structure, an outer membrane protein, a pilus, a capsule component. These vary between species and between strains within a species, so a given phage typically kills some strains of one bacterial species and not others. Against a broad-spectrum antibiotic this is the advantage — a phage aimed at Salmonella leaves the surrounding gut community intact, which matters when the community is itself a colonisation barrier, and it applies no selective pressure on unrelated organisms.

It is also the difficulty. A useful product must be a cocktail covering the strains actually circulating in the target population, which means the coverage of a fixed formulation is a statement about the bacterial population at the time it was assembled. Circulating strains turn over; the cocktail ages. Manufacturers address this by broadening cocktails and by periodic reformulation, but the underlying issue does not go away, and coverage claims are only meaningful against a named strain collection.

Resistance without the antibiotic analogy

Bacteria resist phage readily, and mostly by the simplest route: losing or altering the receptor. A mutant that no longer displays the surface structure cannot be attached to. Restriction–modification systems and CRISPR-Cas adaptive immunity provide further layers.

The difference from antibiotic resistance is that receptor loss is rarely free. Lipopolysaccharide structures, outer membrane porins and capsules are there because they do something, so the escape mutant is often less fit, less able to colonise, or less virulent — and in several documented cases more sensitive to the host’s own defences. This is why the honest framing is not that resistance does not occur, but that its fitness cost differs, and that phage cocktails are usually built to require independent receptors so that a single mutation cannot escape the whole preparation.

Getting there alive

Delivery is the practical constraint. An orally administered phage passes the stomach or the proventriculus, where acid inactivates many phage types, and then must reach a target that may be in the gut lumen, in the mucus layer, or intracellular — the last of which phage generally cannot reach at all. Encapsulation and buffering address the acid step. Beyond that, titre in the gut falls with transit, and clearance means that effect ends when dosing does. Application to carcasses, hides, water lines and surfaces avoids all of this entirely, and is where phage has the strongest evidence in production.

Direct-fed microbials work on a different principle again — competitive exclusion, acid production, mucosal signalling — and their claims should be read the same way: a strain that colonises and a strain that transits are not the same product, and the trials that separate them are the ones worth reading.