Editorial: Phage-based interventions in livestock: from genomics to translational applications

Le résumé fourni par la source

Animal production systems are simultaneously essential to global food security; however, the extensive use of antimicrobials in these systems has significantly contributed to the emergence and spread of antimicrobial resistance (AMR) (Clifford Astbury et al., 2025;Stehling et al., 2023). Food-producing animals account for most of the global antimicrobial consumption, sustaining selective pressure across food-producing animal sectors (Schar et al., 2020;Van Boeckel et al., 2015). This pressure has selected multidrugresistant Escherichia coli, Salmonella enterica, Campylobacter spp., Streptococcus spp., and other pathogens in animal, human, and environmental reservoirs (Izydorcyzk & Church, 2026;Tiseo et al., 2020). As the antibiotic pipeline fails to keep pace, bacteriophages have re-emerged as one of the most credible alternatives for veterinary use (Calero-Cáceres & Balcázar, 2025;Topa-Pila et al., 2026).Yet the path from an isolated phage to a validated veterinary product is long. It requires well-characterized phages and lysins, rational cocktail design, an understanding of prophage ecology and horizontal gene transfer, evolution-informed strategies to manage resistance, curated phage biobanks that enable precision selection, and regulatory frameworks (Bhale et al., 2026;Calero-Cáceres et al., 2026;Kaneko et al., 2026;Melaku et al., 2025;Mora-Domínguez & Calero-Cáceres, 2025). This Research Topic assembles twelve contributions that span this full arc: novel phage and endolysin discovery for animal-associated pathogens, prophage genomic surveillance, evolution-informed phage engineering, alternative in vivo models, speciesfocused translational syntheses, and integrative One Health roadmaps.Four original studies expand the phage and lysin repertoire against economically and zoonotically important pathogens of animal production. Wen et al. isolated phage WEN7 from hospital wastewater using a multidrug-resistant Pseudomonas aeruginosa PAO1-SZ1 host; the phage lysed 19 of 29 clinical P. aeruginosa isolates (65.5%), showed a burst size of ~165 PFU/cell and a 66,379 bp genome free of resistance and virulence determinants, remained stable across temperature, pH, UV, and ethanol stress, and displayed bactericidal activity in both milk and hospital wastewater -matrices that bridge clinical, food-safety, and environmental deployment. Zhao et al. characterized phage vB_RanS_GDF21 targeting Riemerella anatipestifer, the etiological agent of duck infectious serositis; the phage carries a 46,925 bp genome free of resistance and virulence genes, and its recombinant endolysin LysGDF21 both inhibited biofilm formation and disrupted mature biofilms, showing broad antibacterial activity across the R. anatipestifer panel. Shang et al. identified Ply691, a prophage-encoded lysin from Streptococcus suis SC267 with an unusual four-domain architecture (N-terminal Amidase-5, two central CW-7 binding domains, C-terminal Glucosaminidase), temperature adaptability from 4 to 37 °C and alkaline tolerance across pH 7-10, lytic activity against 11 S. suis serotypes, and, in a murine bacteremia model, 100% survival at 2 mg per mouse administered intraperitoneally one hour post-infection, with substantial reduction of bacterial burden across major organs and no detectable resistance emergence. Lu et al. isolated and genomically characterized Myoviridae phage LQ5 (171,908 bp, carrying endolysin and holin lysis-mediator genes) against avian pathogenic E. coli (APEC) serotype O₇₈; in a chick challenge model, phage administration reduced hepatic and splenic bacterial loads more effectively than amoxicillin.Two contributions turn to the bacterial genome itself as both a source of information and a substrate for engineering. Quelal-Madrid et al. performed the first systematic characterization of prophage diversity in 142 S. enterica genomes from poultry and clinical sources, using Phigaro, PHASTEST, and a terminase-based custom BLAST database. They found Peduovirus pro483 prevalent in S. Infantis, distinct cytosine-specific methyltransferase-bearing prophages in related S. Enteritidis, and Enterobacteria phage ST104 in S. Typhimurium encoding SieA and SieB superinfection-exclusion proteins that may confer competitive advantage within microbial communities. The study positions prophages as active drivers of serovar adaptation, virulence potential, and phage-resistance in a Latin American poultry system, and highlights that any deployed lytic-phage cocktail must be designed with awareness of the resident prophage landscape.Luo et al. take an explicitly evolution-informed approach by developing a sequential positive-negative selection strategy -omitting conventional liquid enrichment to preserve viral diversity -to isolate two TolC-dependent phages, PTolC-28 and PTolC-69, directly from poultry-farm environments. Both exploit the outer-membrane exit duct of the AcrAB-TolC multidrug efflux pump as their receptor, enabling two distinct therapeutic modes. PTolC-28 drives "phage steering" in a strain-specific manner: in isolate GDW21C03, phage escape was mediated by a >60% reduction in tolC mRNA (rather than coding-sequence mutation), producing collateral re-sensitization to fluoroquinolones (~5.3fold MIC drop), tetracyclines, and aminoglycosides. PTolC-69, in contrast, exhibited robust phage-antibiotic synergy with doxycycline and florfenicol, reducing effective antibiotic dose 8-fold in vitro and, in a chick infection model, reducing lung and spleen bacterial loads by nearly two orders of magnitude relative to either agent alone.Translating phage efficacy from bench to production animal is limited by ethical, economic, and methodological constraints of vertebrate studies. Ruiz-Santamaría et al. address this bottleneck for Campylobacter, the leading cause of foodborne zoonosis in the EU, by validating Galleria mellonella as a bridging host for phage evaluation. Their fourphage cocktail lysed all 13 tested Campylobacter strains in vitro. In vivo virulence screening revealed marked strain dependence -only 5 of 13 strains reduced larval survival below 50% -and efficacy was optimized on the most virulent isolate, C. jejuni CJE065. At a Multiplicity of Infection (MOI) of 10, the cocktail raised larval survival from 25.5% to 57.5% (p < 0.001), and combination with erythromycin or ciprofloxacin further increased survival to 88.8% and 83.8%, respectively (p < 0.001). Beyond specific efficacy data, the study establishes an invertebrate screening platform that reduces reliance on vertebrate trials while enabling systematic optimization of dosing and phage-antibiotic combinations before farm-scale validation.Three reviews synthesize the state of phage-based intervention across biologically and epidemiologically distinct animal systems. Xu et al. comprehensively review phagebased interventions in livestock, poultry, and aquaculture, framing modern phage strategies along three mechanistic axes -direct bacterial killing, immune modulation, and antigen delivery -and distinguishing lytic phage therapy from phage-display-derived interventions and CRISPR-Cas-enabled theranostic platforms. Beyond cataloguing applications, they critically examine translational bottlenecks (host-range constraints, pharmacokinetics, microbiota effects, and regulatory fragmentation) and openly note that most in vivo evidence remains preclinical.Jespersen et al., in a mini-review, focus on Salmonella Dublin -a host-adapted, zoonotic, and characteristically intracellular bovine pathogen with strong regional clonality. They argue that this clonality is itself a design advantage for geographically tailored cocktails, while cataloguing the persistent gaps in receptor knowledge, phage co-evolution data, and rumen-compatible administration routes. They also address a distinctive constraint for this serotype: because S. Dublin survives and replicates inside macrophages, effective phage therapy will depend on phages that internalize efficie

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.