Microplastics (MPs) and nanoplastics (NPs) have become ubiquitous contaminants in food production systems, raising increasing concerns regarding their potential effects on gastrointestinal and systemic health. This review summarizes current evidence on the occurrence, sources, and biological effects of MPs/NPs within food production chains and their implications for the human gut. Experimental in vitro studies demonstrate that MPs/NPs can induce oxidative stress, inflammation, genotoxicity, metabolic dysregulation, and epithelial barrier dysfunction, while in vivo animal studies show disruption of gut homeostasis, microbiota alterations, systemic translocation, and hepatic, metabolic, and immune disturbances. Co-exposure to environmental pollutants and food-associated chemicals may further enhance toxicity through synergistic mechanisms. Although human evidence remains limited, MPs have been detected in human tissues, including the liver, indicating the possibility of systemic exposure. Emerging evidence also suggests complex interactions between MPs/NPs and the gut microbiome, although the clinical significance of these findings remains uncertain. Therefore, current knowledge is derived predominantly from experimental models, and substantial uncertainties remain regarding human exposure levels, dose-response relationships, and long-term health consequences. Further well-designed human studies and standardized exposure assessment methods are essential to clarify the health risks associated with dietary MPs/NPs.
Given the perishable nature of seafood, accurate and rapid freshness monitoring is essential for ensuring food safety and public health. In this study, a ratiometric fluorescent tag was proposed for detecting ammonia to assess seafood freshness, using D-penicillamine-capped silver/copper nanoclusters (DPA-AgCuNCs) and fluorescein isothiocyanate (FITC) as dual-emission probes. The synthesized DPA-AgCuNCs with aggregation-induced emission (AIE) properties exhibited high sensitivity toward ammonia, with a limit of detection of 2.95 ppm (3σ/s). By integrating orange-emitting DPA-AgCuNCs with green-emissive FITC, the fabricated ratiometric fluorescent tag exhibited a distinct orange-to-green fluorescence transition with increasing ammonia concentration, enabling direct visual assessment of freshness. A fluorescence colour card was further established for rapid preliminary differentiation among fresh, less fresh, and spoiled samples. Convolutional neural network (CNN)-assisted image analysis was introduced as a complementary approach to improve the objectivity and consistency of freshness classification, achieving precision values of 95.24%, 100%, and 100% for fresh, less fresh, and spoiled shrimp, respectively. For on-site monitoring, the CNN was integrated into a smartphone application (FreshSense), constructing a simple platform for rapid monitoring of seafood freshness. This proof-of-concept system is rapid, on-site, and non-destructive, demonstrating its potential for practical screening in food quality and safety.
Rice bran is an abundant agricultural by-product rich in proteins, polysaccharides, and polyphenols, exhibiting potential for Pickering emulsion stabilizers. However, these nutrients predominantly exist as insoluble rice bran natural complexes (IRBNC), which are underutilized due to their compact structure, limited wettability, and poor interfacial activity. To overcome these limitations and optimize the interfacial properties of IRBNC for enhanced emulsification, three catechin monomers with varying numbers of phenolic hydroxyl groups and galloyl moieties—epicatechin (EC), epigallocatechin (EGC), and epicatechin gallate (ECG) were selected as structural models to modify IRBNC. Appropriate concentrations of EC (25 mg/g), EGC (50 mg/g), and ECG (25 mg/g) decreased the ordered structure of proteins in IRBNC, shifted particle wettability from highly hydrophobic (127.015°) to amphiphilic (ranging from 86.976° to 95.695°), loosened the compact layered microstructure, thereby improving interfacial activity (equilibrium interfacial tension: 20.43 to 16.19 mN/m). Consequently, Pickering emulsion stabilized by IRBNC modified with appropriate concentrations of EC, EGC, and ECG exhibited smaller average droplet size (from 59.94 μm to 46.32 μm) and lower creaming index values (from 23.9% to 14.5%) compared with unmodified IRBNC. Moreover, appropriate catechin modification inhibited the formation of lipid oxidation products in IRBNC-stabilized Pickering emulsions during storage. Notably, modification efficiencies followed a structure-dependent trend (ECG > EGC > EC), with ECG exhibiting the strongest effect due to its galloyl substitution and higher phenolic hydroxyl density. Overall, appropriate catechin modification, particularly with ECG, represents an effective strategy to tailor the structural and interfacial properties of IRBNC for enhanced Pickering emulsion stabilization.
The increasing generation of fruit waste represents a major environmental and economic challenge for the agri-food sector, while also offering significant opportunities for sustainable valorisation within the framework of the circular economy. Fruits rejected for cosmetic reasons, together with processing by-products, are rich sources of dietary fibre, sugars and bioactive compounds that can be converted into value-added products. This review aims to provide a comprehensive overview of microbial fermentation as a sustainable strategy for fruit waste valorisation, with particular emphasis on the role of lactic acid bacteria, acetic acid bacteria, yeasts and microbial consortia in improving the nutritional, technological and functional properties of fruit-derived substrates. Current evidence demonstrates that microbial fermentation not only preserves fruit biomass but also enhances its value through the degradation of complex carbohydrates, the biotransformation of phenolic compounds, the production of bioactive metabolites and the generation of functional ingredients. This review further discusses the contribution of viable microorganisms and postbiotic components to product functionality, as well as the importance of microbial viability, processing conditions and storage stability for the development of safe and functional fermented foods. In addition, it highlights recent advances in the use of microbial consortia and identifies the mechanisms governing microorganism-plant matrix interactions as a key area requiring further investigation. Overall, the available literature supports fermentation as a versatile and effective strategy for transforming fruit waste into high-value food products. By integrating current knowledge on microbial metabolism, functionality and processing challenges, this review identifies the main scientific and technological gaps and outlines future research priorities, including microbial consortium design, process scale-up, regulatory considerations and life cycle assessment, to facilitate the industrial implementation of sustainable fruit waste biorefineries.
Previous and present studies confirm that glucose and glycine serve as key precursors for pyrazines contributing to the roasty aroma, but their formation pathways remain unclear. The formation pathways of pyrazines and their derivatives were elucidated by combining CAMOLA with carbonyls profiling. Continuous accumulation of early-stage Maillard products and melanoidins occurred during the glucose-glycine reaction at 160 °C, with 5-methyl-2-furanmethanol, hexanoic acid, and hexanoic acid pentyl ester identified as reliable reaction markers. Eight pyrazines were identified as temperature- and time-dependent products, which were dominated by unlabeled isotopologues from the recombination of glucose fragments. Products of aldol reaction and keto-enol tautomerism of 1-hydroxy-2-butanone and 1,3-dihydroxypropanone, together with glyoxal and methylglyoxal, reacted with glycine to produce aminocarbonyl intermediates. Differences in the structure and binding sites of aminocarbonyl intermediates were responsible for the formation of diverse pyrazines and their derivatives. This work provides a theoretical foundation for the targeted regulation of roasty aroma.
Despite considerable efforts to identify bitter agonists for the ∼25 putatively functional human bitter taste receptors, four receptors remained orphan until now. Whether the problem of identifying agonists for those receptors is due to technical issues related to functional heterologous expression or to the lack of suitable activators is unknown. A similar problem occurs with some variants of bitter taste receptors considered nonfunctional. While some researchers believe that these variants are nonfunctional, others speculate that agonist specificity may have shifted toward yet unknown bitter compounds. Efficient cell-surface localization is generally required for canonical TAS2R signaling in heterologous functional assays and is an important determinant of receptor responsiveness. To address the lack of responsiveness in heterologous expression assays, we quantitatively assessed cell-surface localization of these receptors in living mammalian cells using the HiBiT protein tagging system established in our laboratory. HiBiT-tagged putatively nonfunctional and orphan receptors were expressed in HEK 293T-Gα16gust44 cells and subjected to luminescence measurements using a membrane-impermeable detection reagent. It was observed that all four orphan receptors, TAS2R19, -R42, -R45, and -R60, as well as the putatively nonfunctional receptor variants of TAS2R9, and -R38, were present at the cell surface, albeit at grossly different levels. Immunofluorescence experiments confirmed these results. We conclude that some orphan bitter taste receptors do not show obvious signs of routing or misfolding problems in heterologous cells, suggesting the existence of yet-to-be-discovered bitter activators, particularly for the two receptors, TAS2R19 and TAS2R60, which exhibit pronounced cell-surface localization.
Flavor characteristics strongly influence consumer acceptance and preference for aquatic products, yet the molecular basis underlying flavor variation among shrimp strains remains unclear. This study combined electronic tongue analysis, untargeted metabolomics and transcriptomic analysis to characterize the flavor profile of a red variant of Litopenaeus vannamei. Results revealed enhanced saltiness and richness in the red variant compared to the wild-type strain. Metabolomic analysis identified ten potential flavor peptides, three of which (IKRDF, FAEKI and PLEGK) were confirmed to exhibit saltiness. Molecular docking further suggested potential interactions of the flavor peptides with the salty receptor TRPV1 and the umami receptor T1R1. Integrated omics analyses highlighted glycine, serine and threonine metabolism as a key contributor, with the accumulation of dimethylglycine (DMG) reflecting alterations in one-carbon metabolism and redox-related metabolic processes. Together, these findings provide novel insights into peptide-mediated taste formation and metabolic remodeling in the red variant of L. vannamei.
How the chemical structures of odorant molecules determine biologically oriented behaviors remains a central challenge at the intersection of food informatics and biology. To identify structural features of odorant molecules that are associated with mouse first-encounter approach-avoidance responses, this study used a published human flavor semantic descriptor database only as a stratification tool to select a semantically diverse set of odorants; the behavioral labels came from the mice's own first-encounter approach-avoidance responses. By integrating multidimensional molecular representations with machine learning algorithms, we established a predictive method for odor-driven behavioral preference. The results showed that, among the prespecified model-representation pipelines evaluated in this study, the Logistic Regression model based on structural keys and physicochemical descriptors achieved the highest mean F1-score (0.812 ± 0.016). Model interpretability analysis indicated that ester groups, aromatic rings, ethers, and branched carbon motifs were more likely to be associated with approach behavior, whereas thioethers, sulfur-containing heterocycles, disulfides, and certain carbonyl/heteroatom environments were more likely to be associated with avoidance behavior. Behavioral experiments using structural analogs provided supporting evidence for an association between local structural modification and shifts in first-encounter approach-avoidance behavior in mice. This study establishes an interpretable olfactory prediction methodology. This methodology provides an interpretable computational framework to support the screening and prioritization of candidate odorant molecules and may inform subsequent human sensory evaluation and product-level validation.
Understanding the long-term evolution of condensed tannins (CTs) in wine remains challenging due to the molecular complexity of aged wines. In this study, an advanced analytical workflow combining tannin-fraction isolation, thioglycolysis, UHPLC-UHRMS with AcquireX™ data-dependent acquisition, and Feature-Based Molecular Networking (FBMN) driven by the t-distributed stochastic neighbor embedding (t-SNE) algorithm was applied for the first time to wines. The studied wines are authentic red wines aged for 17 years. This exceptionally rare sample set, stored under four closures of differing oxygen permeability, provided a unique opportunity to assess the applicability of this workflow for the in-depth structural characterization of condensed tannin evolution markers in a complex wine matrix. Using this approach, 139 evolution markers were annotated. Among these, 122 markers were assigned to levels 2-3, while 17 markers were assigned to level 4. Overall, 72 of the 139 annotated markers were previously unreported, including 55 markers within the levels 2-3 group and 17 within the levels 3-4 group. These markers arose from interactions between tannins and anthocyanins, pyranoanthocyanins, and aromatic aldehydes with high oxidation levels of up to 5. Exploratory univariate (ANOVA) and multivariate statistical methods including Principal Component Analysis (PCA) and Hierarchical Cluster Analysis (HCA) revealed correlative trends between marker abundance and closure oxygen permeability within this specific sample set. Overall, this work demonstrates the power of this transferable analytical workflow for comprehensive structural characterization of condensed tannin evolution markers in complex matrix form long-aged red wines and provides new molecular insights into their chemical evolution.
Developing non-thermal technologies to enhance health-promoting compounds is important for producing high-value functional foods. This study compared the effects of electron beam irradiation (EBI) applied to dry broccoli seeds before germination (FZZZ) with those of direct irradiation of sprouts (FZYM) on glucosinolate (GSL) enhancement. Low-dose FZZZ (2 kGy) increased total GSL content to 6556.85 μg/g, compared with 3174.58 μg/g in the control, whereas low-dose FZYM decreased total GSL content. Sprouts grown from irradiated seeds also showed increased vitamin C, soluble sugars, total free amino acids, and volatile sulfur-containing compounds. Metabolomic and amino acid analyses indicated that FZZZ increased the availability of sulfur amino acid precursors and remodeled methionine- and phenylpropanoid-related metabolism, whereas FZYM mainly induced stress-associated changes in lipid and nucleotide metabolism. qRT-PCR revealed selective regulation of genes involved in downstream GSL modification and hydrolysis rather than a coordinated activation of the entire GSL biosynthetic pathway. Overall, seed-stage EBI at 2 kGy is a promising strategy for enhancing GSL content in broccoli sprouts while maintaining growth and nutritional quality.
Astringency is an important mouthfeel attribute in many beverages, yet its neurophysiological correlates remain poorly understood. It also remains unclear how beverage matrix shapes frontal hemodynamic responses to astringency over time. This study used functional near-infrared spectroscopy (fNIRS) to characterize prefrontal hemodynamic responses to proanthocyanidin (PAC)-induced astringency in three matrices: water, ethanol, and red wine. To capture temporal aspects of the perceptual experience, responses were assessed separately during the induction and aftertaste phases. Sensory evaluation showed that perceived astringency increased with PAC concentration, although both the magnitude and time course of this increase varied across matrices. During the induction phase, frontal HbO responses were generally limited and less consistent. By contrast, the aftertaste phase was associated with broader and stronger prefrontal HbO responses. Region-of-interest analysis further revealed distinct matrix-related response patterns in channels corresponding anatomically to the orbitofrontal cortex (OFC), frontopolar area (FPA), and dorsolateral prefrontal cortex (DLPFC), suggesting that lingering astringency may involve distributed evaluative processing rather than a simple representation of perceived intensity. Overall, prefrontal HbO responses to astringency differed between the induction and aftertaste phases and varied across beverage matrices. This work extends current understanding of astringency as a matrix-sensitive sensory process and provides useful neurophysiological evidence concerning the frontal hemodynamic correlates of persistent mouthfeel attributes in food systems.
Flavourzyme-assisted two-stage hydrolysis is a recognized debittering strategy for protein hydrolysates, yet peptidomic-level evidence for the underlying mechanism-particularly the role of pH control mode-remains limited. This study compared pH-stat and free-fall pH Flavourzyme treatments (3 and 4 h) of an Alcalase-derived rice protein hydrolysate (A6; degree of hydrolysis by the o-phthalaldehyde method [DH-OPA] 24.69%) using nanoLC-ESI-Q-TOF de novo sequencing to characterize peptide-profile changes associated with bitterness reduction. Formal sensory evaluation (15 trained panelists, paired comparison with Bonferroni correction for 10 comparisons) of the ultrafiltered (<2500 Da) fractions placed A6-s3 (pH-stat, 3 h) at the least-bitter end of the ranking, although it did not differ significantly from A6-s4 after multiplicity correction. The apparent ranking differed from that observed in the preliminary evaluation of unfiltered hydrolysates, which used a smaller panel and a different sensory design. De novo sequencing identified 963-1226 peptides per hydrolysate. A6-s3 exhibited the lowest mean Q value (1430 cal mol-1; Kruskal-Wallis, p < 10-22) and the lowest predicted bitter peptide proportion (49.6%). A6 and A6-s3 shared only one exact PEAKS-assigned sequence among >1900 combined detections, indicating extensive differences in the detectable peptide profiles. Database-assisted annotation against the Oryza sativa proteome showed a markedly lower assignment rate for A6-s3 (0.2%) than for A6 (52.8%, predominantly glutelins); this was treated as a descriptive analytical observation because I/L ambiguity, de novo sequencing uncertainty, stochastic data-dependent acquisition (DDA) sampling, and database coverage may contribute. A6-s3 also showed the lowest C-terminal strongly hydrophobic residue frequency (62.3%; I, L, F, W, Y, P) and the highest N-terminal amino acid diversity (Shannon entropy, 3.91 bits). These findings indicate that maintaining pH during Flavourzyme treatment was associated with a more favorable bitterness-related peptide profile. The results are compatible with established terminal processing accompanied by broader changes in the detectable peptide population, but they do not establish or quantify a distinct contribution from internal cleavage.
Understanding the causal associations among microorganisms, flavor metabolites, and sensory attributes is essential for the precise regulation of the flavor of fermented milk. In this study, 22 commercial fermented milks were characterized using sensory evaluation, microbial analysis, and volatile metabolite profiling. Five sensory phenotypes were developed, each exhibiting unique flavor profiles. However, microbial community structures differed only slightly among phenotypes, suggesting that the community structure itself could not fully explain sensory diversification. Therefore, generalized microbe-phenotype triangulation analysis was performed and identified multiple causal taxa associated with different sensory phenotypes, while the mmvec model further revealed strong associations between certain microorganisms and characteristic aroma compounds. Overall, this study established a novel data-driven framework for exploring potential causal associations among microorganisms, flavor metabolites, and sensory phenotypes in fermented milks, representing a preliminary effort to move beyond conventional correlation-based analyses.
Monooctanoylgalloylglycerols (MOGGs) are monoacylated galloyl structured lipids that combine the phenolic hydroxyl groups of gallic acid with a medium fatty-acyl chain. Their lipase-catalyzed preparation is limited by competing acyl transfer, hydrolysis, and secondary acylation reactions. This work compared several Candida antarctica lipase B-catalyzed routes using Lipozyme® 435, including transesterification of propyl gallate with mono/diacylglycerols, hydrolysis/alcoholysis of dioctanoylgalloylglycerols (DOGG), one-pot conversion of propyl gallate, glycerol, and octanoic acid (OA), and direct esterification of 1-o-galloylglycerol (GG) with OA. Only GG direct esterification gave substantial MOGG formation. Optimization using response surface methodology (RSM) identified 65 °C, an OA/GG molar ratio of 3:1, 72 h, and 30 g/L enzyme as suitable conditions, giving 62.7 ± 2.1% MOGG yield and 13.4 ± 0.3% DOGG formation. The purified MOGG fraction was characterized by FT-IR, Raman spectroscopy, NMR spectroscopy, and ESI-HRMS/MS. FT-IR and Raman spectra showed retention of galloyl features together with the introduction of octanoyl-chain signatures, while NMR and ESI-HRMS/MS confirmed the formation of two positional MOGG isomers. Docking was used to compare substrate accommodation within the lipase active site. These results identify GG direct esterification as the preferred route for controlled monoacylation of galloyl structured lipids and define the main reaction constraints affecting MOGG production.
Developing dry antimicrobial delivery systems that combine bioactive protection with biofilm-matrix disruption remains a major challenge. Building on a previously established protease-modulated pea protein isolate (PPI)-carvacrol nanoemulsion system, this study developed protease-assisted microcapsules designed to preserve carvacrol and promote the removal of preformed biofilms. Carvacrol-loaded nanoemulsions were prepared at pH 3.5, 7.0, and 10.0 and supplemented with pepsin or trypsin before spray-drying or freeze-drying with maltodextrin. Protease-specific interfacial modification markedly affected emulsion stability, powder structure, and encapsulation performance. At pH 7.0 and 10.0, trypsin reduced droplet size from 284.60 to 232.52 nm and from 149.13 to 140.38 nm, respectively. The resulting spray-dried microcapsules exhibited high encapsulation efficiency (>96%) and low surface carvacrol contents (0.95-1.24 mg/g). In contrast, pepsin caused pronounced destabilization under acidic conditions, particularly after freeze-drying, yielding porous powders with an encapsulation efficiency of 48.18% and a surface carvacrol content of 69.51 mg/g. Trypsin-assisted microcapsules achieved 90-99% removal of preformed Listeria innocua biofilm biomass within 1 h, whereas enzyme-free and pepsin-containing formulations generally remained below 40%. Microscopic observations confirmed extensive disruption and detachment of the biofilm structure. After one year at 4 °C, trypsin-loaded formulations retained high biofilm biomass removal activity (85-99% after 2 h). These findings demonstrate that protease-assisted microencapsulation can couple interfacial regulation during particle formation with enzyme-mediated biofilm-matrix disruption after rehydration, providing a sustainable carvacrol delivery platform with durable antibiofilm functionality.
Structural evolution of wheat endosperm particles during comminution can markedly influence flour processing and digestibility. While previous studies have mainly focused on differences among grinding endpoints, stage-dependent refinement under shear-dominant conditions remains unclear. This study characterized the structural and functional evolution of wheat endosperm particles during shear-dominant comminution. Coarse wheat endosperm particles were collected after predefined shear durations during shear-dominant comminution and examined using scanning electron microscopy and confocal laser scanning microscopy (CLSM), as well as based on particle-related physicochemical indices, protein structural parameters, rapid visco analyzer properties, and in vitro starch digestibility. CLSM image-based deep-learning analysis was performed to characterize microstructural features at different comminution stages. The refinement process could be operationally divided into a volume grind stage (RM-S40) and a surface grinding stage (S50-S70). Comminution increased the proportion of damaged starch (from 0.22% to 7.97%), decreased the particle size, and progressively enhanced the specific surface area. Additionally, comminution increased the peak viscosity (from 1517 cP to > 2300 cP) and rapidly digestible starch proportion (from 11.29% to 32.83%) and decreased the resistant starch proportion (from 74.32% to 50.99%). Analysis of protein-related indices further indicated the disruption and partial reorganization of gluten aggregation during shear comminution. The ResNet18 model achieved a test accuracy of 0.952. The resulting image-derived descriptor exhibited good consistency with key physicochemical and functional indicators. Thus, the refinement of wheat endosperm during shear-dominant comminution was not uniform and was accompanied by stage-dependent microstructural evolution and corresponding functional changes.
Gelatin is a natural biopolymer for biodegradable food packaging films, but its inherent hydrophilicity and unsatisfactory mechanical properties limit its application in packaging lipid-rich foods and edible oils. This study aimed to investigate the effects of cellulose I/Ⅱ nanocrystals (CNC-I, CNC-Ⅱ) and their lauric acid-modified derivatives (LCNC-I, LCNC-Ⅱ) on the physicochemical properties of gelatin-based films. The results showed that both CNCs and LCNCs were uniformly dispersed in the gelatin matrix. Compared with CNCs, the addition of LCNCs further increased the microstructural compactness, structural disorder, and disulfide bond content of gelatin-based films, and improved their hydrophobicity, mechanical properties, as well as UV, oxygen, and water vapor barrier properties. These improvements were attributed to the plasticizing effect of LCNCs, together with strengthened physical entanglement and interfacial compatibility between LCNCs and the gelatin matrix. Notably, 10% LCNC-I showed superior performance in reducing film solubility (25.51 ± 1.15%), prolonging dissolution time (1085.71 ± 31.16 s), and enhancing hydrophobicity (105.0 ± 1.3°). By contrast, 10% LCNC-Ⅱ exhibited greater potential in enhancing tensile strength (51.29 ± 0.79 MPa) and improving oxygen barrier properties. In camellia oil packaging and storage tests, compared with commercial films, gelatin-based films incorporated with 10% LCNCs effectively retarded oil oxidation and rancidity during storage; especially, films incorporated with LCNC-Ⅱ exhibited a slightly better protective effect. These findings demonstrate that LCNCs can serve as promising reinforcements to optimize the overall performance of gelatin films for food packaging.
Probiotic microorganisms are increasingly exploited for their capacity to generate bioactive metabolites with potential health-promoting properties. In this study, we investigated the in vitro antioxidant activity of sixteen probiotic bacteria and characterized the extracellular metabolomic profiles of the spent fermentation broths (cell-free supernatants) derived from two strains showing markedly different antioxidant behaviors. Antioxidant potential was quantified using the DPPH radical scavenging assay, and total phenolic content was assessed by the Folin-Ciocalteu method. Among the tested strains, Lactiplantibacillus plantarum 8VEG3C displayed the most favorable antioxidant profile within the tested panel, with the lowest IC50 value (2.47% v/v) and the highest phenolic content, whereas Lacticaseibacillus paracasei LC01 showed a substantially weaker antioxidant activity. To explore the molecular basis of these functional differences, untargeted extracellular metabolomic analysis was performed by liquid chromatography-tandem mass spectrometry (LC-MS/MS), coupled with multivariate statistical analysis and putative, database-supported annotation. The two strains generated clearly distinguishable extracellular signatures. The L. plantarum 8VEG3C supernatant was enriched in fatty acyl-related metabolites, including oleic acid, 7-oxooctadecanoic acid and γ-dodecalactone, together with aromatic amino acid-derived compounds such as indole lactate. In contrast, L. paracasei LC01 was enriched in amino acid- and small peptide-related metabolites including the putatively annotated 2-(6-hydroxyhexanoylamino)-3-phenylpropanoic acid, compatible with a more prominent involvement of the proteolytic system of lactic acid bacteria. These findings support the use of untargeted metabolomics as a functional screening platform for the rational selection of probiotic strains with postbiotic potential.
