Collagen, long regarded as the inert structural scaffold of the extracellular matrix, is now recognised as a source of bioactive signaling molecules: enzymatic hydrolysis releases low-molecular-weight bioactive collagen peptides that act systemically. This narrative review synthesises in vitro, animal-model, and veterinary clinical evidence with explicit differentiation of evidence levels; literature was identified through structured searches of PubMed, Web of Science, and Google Scholar, covering peer-reviewed, English-language reports published between 2000 and January 2026, without a formal systematic protocol. Across rodent, porcine, and human pharmacokinetic studies, orally administered collagen hydrolysate is efficiently absorbed, with a fraction reaching the circulation as intact prolyl-hydroxyproline and hydroxyprolyl-glycine through the conserved PEPT1/PEPT2 transporters; reported bioavailability varies with source, dose, and method, so no single value generalises across species. Native collagen and larger collagen fragments engage structure-dependent receptors in vitro (α2β1 integrins, DDR1/DDR2, GPVI, LAIR-1/2) that require triple-helical or Gly-Pro-Hyp presentation, whereas it is not established that the di- and tripeptides that reach the circulation after oral dosing engage them; their systemic actions are partly attributable to intracellular routes, including the Keap1–Nrf2 axis, HDAC/HAT modulation, and glycine-dependent glutathione synthesis. The gut–collagen peptide axis, a model derived from rodent and cell-culture data, links microbial bile acid remodeling and FXR/TGR5 signaling to systemic effects. Current evidence supports validated use in canine and equine osteoarthritis; species-specific bioavailability studies in dogs and cats remain the priority.

Bioactive collagen peptides and collagen hydrolysates are dietary proteins whose degradation products also carry signaling activity. This second part of a two-part narrative review addresses the translational dimensions of their bioactivity, integrating veterinary clinical trials, controlled animal-model studies, and human biomedical data, with evidence strictly stratified by type. The gastrointestinal tract acts not only as the absorption site but as a target organ. In cell culture and rodent models, luminal collagen fragments restore tight-junction integrity, alter the microbiome, and shift the enterohepatic bile acid pool; increased secondary bile acid synthesis is proposed to engage the farnesoid X receptor (FXR) and Takeda G-protein-coupled receptor 5 (TGR5). None of these steps have been demonstrated in veterinary clinical patients. The accompanying GLP-1 and PYY response may reflect direct amino acid stimulation of enteroendocrine cells as much as a bile acid-dependent route, and current data do not separate the two. Veterinary trials show objective kinetic improvement in osteoarthritic dogs and horses. Taurine-responsive dilated cardiomyopathy is a reversible, diet-amenable condition; as the principal bile acid conjugation substrate in carnivores, taurine acts on the same bile acid pool as a mechanistically separate input. Recent data implicate prolyl-hydroxyproline in brown adipogenesis, collagen peptides in hippocampal neurogenesis, and taurine in platelet normalisation. The gut–collagen peptide axis is an evolving model linking dietary collagen to systemic metabolic, endocrine, and immune signaling, with implications for companion-animal, equine, and livestock practice.

Bile acid (BA) dysmetabolism is implicated in canine chronic inflammatory enteropathy (CIE), but fasting serum profiles across treatment-response groups remain poorly characterized. This exploratory observational study analyzed pretreatment serum from dogs subsequently classified as food-responsive enteropathy (FRE; n = 8) or immunosuppressant-responsive enteropathy (IRE; n = 5), with four unmatched purpose-bred Beagle controls. Targeted LC–MS/MS, performed using a published canine serum method, provided results for 15 BA species. All samples were analyzed in one batch by an analyst blinded to subsequent group assignment. Eight species quantifiable in every dog and ten derived variables were compared using Kruskal–Wallis tests with Benjamini–Hochberg correction across 18 omnibus tests. Cholic acid and unconjugated primary BA concentrations were higher in IRE than in controls and FRE. The total primary-to-secondary BA ratio showed a monotonic group ordering (control < FRE < IRE; H = 13.76, p = 0.0010, q = 0.0092; epsilon-squared = 0.840), with all Bonferroni-corrected pairwise comparisons significant. A post hoc four-scenario sensitivity analysis comprising the submitted zero-substitution analysis and three LOD/LOQ-based alternatives reproduced the same 18-test FDR-significant set and the same ratio inference. This analysis tests robustness to censored-value handling but is not independent validation. The findings suggest group-associated serum BA patterns, but the small sample, clinical heterogeneity, and unmatched controls preclude diagnostic or treatment-predictive interpretation.