The motivation is stark. Epidemiological projections cited in the study estimate that roughly 589 million adults were living with diabetes in 2024, a figure expected to climb to about 852.5 million by 2050. Gout cases, meanwhile, are projected to reach 95.8 million globally by mid-century. Current drugs such as acarbose, sitagliptin and allopurinol work well but carry drawbacks, from gastrointestinal discomfort to rare but serious hypersensitivity reactions. That has pushed researchers toward food-derived bioactive peptides, which offer high bioavailability and low toxicity as candidates for nutritional intervention.

Goat milk proteins, dominated by caseins and whey proteins, are attractive raw material because goat milk is more digestible and less allergenic than cow milk. Yet most goat milk products remain traditional powders with low added value. The team, led by Xinru Yu and Guowei Shu, set out to change that by asking a question few had tackled: could a single goat milk hydrolysate inhibit enzymes relevant to both glucose regulation and uric acid production at the same time?

The answer required finding the right enzymes for the job. The researchers screened nine food-grade proteases, hydrolyzing reconstituted goat milk for five hours and measuring the degree of peptide bond cleavage along with inhibitory activity against alpha-glucosidase, which breaks down dietary carbohydrates; dipeptidyl peptidase-IV, or DPP-IV, the incretin-degrading enzyme targeted by sitagliptin; and xanthine oxidase, the purine-metabolism enzyme blocked by allopurinol. Alkaline protease produced the deepest hydrolysis, but papain stood out for inhibitory power, achieving up to 53.37 percent alpha-glucosidase inhibition and 97.08 percent DPP-IV inhibition, while a neutral protease delivered the strongest xanthine oxidase blockade.

Because single enzymes rarely produce a peptide mix that hits all three targets, the team combined the four best performers. After testing stability under simulated gastrointestinal digestion and heat treatments mimicking pasteurization, the papain plus neutral protease pairing emerged as the winner. The optimal ratio, determined on an activity-unit basis, was one part papain to two parts neutral protease. Response surface methodology with a Box-Behnken design then fine-tuned the process, revealing that lower enzyme dosage, moderate temperature and extended time favored activity. The final recipe: 3000 units of enzyme per gram of protein at 50 degrees Celsius for roughly 4.1 hours.

Validation experiments confirmed the model’s predictions. Under optimized conditions, the hydrolysate inhibited alpha-glucosidase by 73.05 percent, DPP-IV by 88.63 percent and xanthine oxidase by 80.69 percent, matching the predicted values within statistical error. Benchmarking against standard curves showed the hydrolysate performed comparably to 68 micrograms per milliliter of acarbose, 2.87 micrograms per milliliter of sitagliptin and 64.73 micrograms per milliliter of allopurinol in the respective in vitro assays, though the authors caution that such comparisons serve only as internal references for complex peptide mixtures.

To find the molecules behind the activity, the team filtered the hydrolysate below 10 kilodaltons, separated it by preparative chromatography and analyzed the most potent fraction by liquid chromatography-tandem mass spectrometry. Twenty-four peptides emerged, mostly from beta-, alphaS1- and alphaS2-casein, with molecular weights between 400 and 1400 daltons. Bioinformatic screening with PeptideRanker and the BIOPEP-UWM database narrowed the field to candidates with predicted activity against multiple targets. Two sequences rose to the top: the tripeptide LRF, predicted to inhibit both alpha-glucosidase and xanthine oxidase, and the seven-residue FLPYPYY, predicted to hit both alpha-glucosidase and DPP-IV.

Molecular docking revealed why these peptides work. FLPYPYY bound alpha-glucosidase with a predicted energy of minus 9.7 kilocalories per mole, anchoring itself through hydrogen bonds to Tyr733, Arg647, Glu767 and Gly766, while LRF formed hydrogen bonds with Asp649, Arg653, Glu788 and Gly766. FLPYPYY also docked into DPP-IV with minus 8.8 kilocalories per mole, engaging Trp629, Trp627 and Tyr547 through pi-pi interactions, and LRF bound xanthine oxidase at minus 7.2 kilocalories per mole. Redocking of co-crystallized ligands produced root-mean-square deviations below 2 angstroms, validating the protocol.

The docking snapshots were then stress-tested in 100-nanosecond molecular dynamics simulations using GROMACS with the Amber99SB-ILDN force field. All four peptide-enzyme complexes stabilized, with backbone RMSD values settling between 0.10 and 0.18 nanometers, radius of gyration and solvent-accessible surface area remaining steady, and free-energy landscapes showing well-defined low-energy basins. MM/PBSA calculations over the final 10 nanoseconds yielded binding free energies of minus 4.16 to minus 23.79 kilocalories per mole, with van der Waals and electrostatic terms favoring association and per-residue decomposition pinpointing energetic hotspots such as Phe735 and Tyr636 in alpha-glucosidase and Arg560 in DPP-IV.

The authors are careful to note the limits of the work. Docking and simulation cannot by themselves prove enzyme inhibition, and the identified peptides still need synthesis, concentration-response testing, IC50 determination and kinetic or structural validation before any functional-food claim can be made. Bioavailability, safety and physiological effects in humans remain open questions. Still, the study demonstrates a compelling pipeline: from a humble carton of goat milk to computationally vetted, dual-target peptide candidates, offering dairy producers a route to higher-value products and offering metabolic-disease researchers a fresh trove of molecules to investigate.

Subject of Research: Multifunctional enzyme-inhibitory peptides generated from goat milk proteins by optimized enzymatic hydrolysis

Article Title: Optimization of enzymatic hydrolysis for generating potential multifunctional peptides from goat milk: Bioactivity evaluation, structural characterization and molecular interaction analysis

Article References: Yu, X., Shu, G., Hu, W., Zhang, G., Xie, P., Song, K., & Li, G. (2026). Optimization of enzymatic hydrolysis for generating potential multifunctional peptides from goat milk: Bioactivity evaluation, structural characterization and molecular interaction analysis. Food Chemistry: X, 39, Article 104517. https://doi.org/10.1016/j.fochx.2026.104517

Image Credits: AI Generated

DOI: 10.1016/j.fochx.2026.104517

Keywords: goat milk, bioactive peptides, enzymatic hydrolysis, alpha-glucosidase, DPP-IV, xanthine oxidase, molecular docking, molecular dynamics, diabetes, hyperuricemia, functional food, response surface methodology

Collagen peptides are short chains of amino acids produced by enzymatically breaking down collagen from skin, bone, and cartilage. They have attracted attention for antioxidant, antihypertensive, and anti-aging properties, and previous work has shown that proline- and arginine-rich sequences at their C-terminal ends can inhibit dipeptidyl peptidase-IV (DPP-IV), an enzyme whose suppression prolongs the action of incretin hormones that stimulate insulin release. But the researchers argue that hitting DPP-IV alone is often not enough. Persistent hyperglycemia fuels overproduction of reactive oxygen species, and when antioxidant defenses falter, the resulting oxidative stress damages pancreatic beta cells, muddles insulin signaling, and derails glucose handling in the liver, accelerating the progression of type 2 diabetes.

To find a suitable partner for the peptides, the team screened four plant extracts, from blue plum, tremella, mulberry leaf, and bitter melon, alongside the cowhide collagen peptides themselves. In radical-scavenging assays, bitter melon extract stood out: its DPPH radical scavenging activity was nearly five times that of the collagen peptides, and its hydroxyl radical scavenging capacity exceeded the peptides by more than 85 percent. It also showed a stronger, though statistically not significant, tendency to inhibit alpha-amylase, the digestive enzyme that cleaves dietary starch into sugars. That dual profile, antioxidant power plus suppression of carbohydrate breakdown, made bitter melon the natural candidate for a combination formulation.

The interaction between the two components turned out to be a genuine molecular partnership rather than a simple mixture. Fluorescence spectroscopy showed that as bitter melon extract concentration rose, the peptides’ fluorescence was progressively quenched, and Stern-Volmer analysis indicated static quenching, meaning the two had formed non-fluorescent complexes. Thermodynamic calculations revealed positive enthalpy and entropy changes and negative free energy, signatures of a spontaneous, endothermic binding process driven primarily by hydrophobic interactions. Molecular docking of two representative collagen peptides, GPVGPPG and GPVG, against momordicine I, a key bioactive constituent of the extract, confirmed that hydrogen bonds and hydrophobic pi-alkane and pi-sigma contacts stabilize the assembly. Infrared spectroscopy and calorimetry backed this up: the amide I and amide II bands of the peptides shifted after binding, and the complex melted at higher temperatures, indicating enhanced thermal stability. Microscopy even showed the fragments reassembling into larger, flake-like structures.

What matters most for an oral supplement, however, is survival in the gut. The researchers subjected both the plain peptides and the complex to simulated gastric and intestinal digestion. Both withstood pepsin admirably, retaining over 80 percent of their activities. But trypsin in the intestinal phase proved far more destructive, significantly degrading the bioactivity of the plain peptides. The bitter melon complex, by contrast, retained significantly higher alpha-amylase inhibition and antioxidant capacity after intestinal digestion, suggesting the molecular partnership shields the peptides from enzymatic attack. Notably, DPP-IV inhibition was no different between the two after digestion, likely because continued hydrolysis of the proline-rich peptides keeps generating new DPP-IV-inhibiting fragments regardless.

Then came the animal trial. The team induced type 2 diabetes in mice with a high-sugar, high-fat diet followed by streptozocin injections, then treated the diabetic animals for four weeks with collagen peptides, the peptide-bitter melon combination, or metformin as a positive control. After four weeks, fasting blood glucose fell by nearly 38 percent in the peptide group and about 39 percent in the combination group, compared with roughly 52 percent for metformin. Glucose tolerance improved in both treatment groups, again with a slight edge for the combination, as measured by a standard oral glucose tolerance test.

The combination also tackled oxidative stress more effectively. Diabetic mice showed depressed total antioxidant capacity and reduced activities of glutathione peroxidase and superoxide dismutase in the liver, along with elevated malondialdehyde, a marker of lipid damage. Peptide supplementation partially reversed these changes, and the combination pushed them further toward normal, consistent with the in vitro radical-scavenging results. The liver told a similar story structurally: diabetic mice had depleted hepatic glycogen stores, swollen and vacuolated hepatocytes, and disrupted kidney tubules, all of which improved after treatment, with the combination group showing the clearest restoration of glycogen deposition and tissue architecture.

Mechanistically, the benefits appear to route through a well-known metabolic control axis. AMP-activated protein kinase, or AMPK, is a master sensor of cellular energy status whose activation suppresses hepatic gluconeogenesis, the liver’s internal production of new glucose. In the diabetic mice, the phosphorylation ratio of AMPK was significantly depressed. Both treatments restored it, with the combination showing the strongest effect, while simultaneously reducing the expression of FoxO1, a transcription factor that drives gluconeogenic genes, and PEPCK, the rate-limiting enzyme of that process. In other words, the supplement combination seems to flip the liver from a glucose-producing state toward a glucose-storing one.

The study also ventured into the gut microbiome, sequencing bacterial DNA from cecal contents and profiling cecal metabolites by untargeted metabolomics. Diabetic mice exhibited an elevated ratio of Firmicutes to Bacteroidetes, reduced microbial diversity, and shifts in genera such as Candidatus Arthromitus and Anaerostipes. The combination treatment partially reversed these dysbiotic patterns and shifted the overall community composition toward that of healthy controls. Metabolomics revealed 89 significantly altered metabolites, including stress-related steroid hormones like cortisol and aldosterone, which rose with diabetes and fell back toward normal with treatment, along with antioxidant compounds like ascorbate that were depleted in disease and partially restored. Association analysis linked these metabolic shifts to specific microbial phyla, hinting that the gut ecosystem participates in the formulation’s effects.

The authors are careful to frame the microbiota findings as correlational, offering mechanistic clues rather than proof of a causal host-microbe pathway, and the work remains at the mouse-model stage, with human efficacy, dosing, and safety still untested. Yet the study makes a compelling case that food-derived ingredients can be engineered into something greater than the sum of their parts. By pairing waste-stream collagen with a traditional medicinal plant, and letting noncovalent chemistry, digestive stability, AMPK signaling, and gut ecology all pull in the same direction, the researchers suggest that the future of glycemic management may partly lie not in a single magic molecule, but in thoughtfully assembled natural complexes. If subsequent studies confirm these effects in humans, cowhide, one of the meat industry’s most abundant leftovers, could find itself at the center of a new generation of functional foods for blood sugar control.

Subject of Research: Antihyperglycemic effects of cowhide collagen peptides combined with Momordica charantia extract in type 2 diabetes

Article Title: Enhanced antihyperglycemic effects of cowhide collagen peptides by the combination with Momordica charantia extract: AMPK–FoxO1 pathway engagement and gut microbiota remodeling

Article References: He, L., Wang, X., He, J., Gao, Y., Yu, Q., & Zhang, L. (2026). Enhanced antihyperglycemic effects of cowhide collagen peptides by the combination with Momordica charantia extract: AMPK–FoxO1 pathway engagement and gut microbiota remodeling. Journal of Agriculture and Food Research, 31, Article 103298. https://doi.org/10.1016/j.jafr.2026.103298

Image Credits: AI Generated

DOI: 10.1016/j.jafr.2026.103298

Keywords: collagen peptides, bitter melon, Momordica charantia, type 2 diabetes, AMPK, FoxO1, gut microbiota, antioxidant, DPP-IV, alpha-amylase, hepatic gluconeogenesis, functional foods