To elucidate the pathway and mechanism of Maillard reaction (MR) in forming meaty flavor in 'cysteine-xylose-glutamic acid 'system, Cys-Amadori and Glu-Amadori intermediates were prepared using a two-stage variable temperature. The effects of pH (5-9), reaction temperature (85-125 °C), and reaction time (60-140 min) on the content of Amadori rearrangement products (ARPs) were studied. Quantitative characterization of Cys-Amadori and Glu-Amadori intermediates was performed using LC-MS/MS. It was found that ARPs readily degrade to form heterocyclic compounds, such as nitrogen-containing compounds, under conditions of reaction temperature 100-125 °C, pH > 7, and reaction times between 100 and 140 min, resulting in the formation of roasted flavor. Conversely, when the reaction temperature was 85-115 °C, pH < 6, and reaction time was 80-120 min, ARPs underwent 1, 2-enolization to produce sulfur-containing compounds, which facilitated the formation of broth flavor. Therefore, the flavor profile of roasted/broth in the 'cysteine-xylose-glutamic acid' MR system can be controlled by changing the reaction conditions.

Furaneol and sotolone are caramel aroma isomers with distinct sensory qualities, yet how their differential perception emerges from molecular to neural levels remains unclear. This study integrates sensory analysis, molecular dynamics, and EEG to address this gap. Sensory evaluation showed furaneol was more pleasant and had a tenfold lower detection threshold than sotolone. Simulations revealed furaneol binds stably to OR5M3 via TYR-257, while sotolone interacts selectively with OR8D1 via HIS-159/ASN-206. EEG identified that furaneol uniquely enhanced frontal θ power (4-8 Hz), associated with cognitive engagement, whereas both odorants increased α/β power. These results demonstrate that perceptual differences originate from receptor-specific binding and are cortically encoded in distinct oscillatory patterns. This multi-level approach provides a mechanistic framework linking molecular interactions to perception, supporting the rational design of flavors.

The perception of taste is shaped by stimulus intensity and individual sensitivity, but the neural mechanisms governing this interaction remain poorly understood. Combining psychophysics and multi-level EEG to investigate umami (MSG) processing, we identified a universal neural signature of intensity: increasing MSG concentration systematically suppressed the P200 event-related potential and weakened beta-band functional connectivity between the orbitofrontal (OFC) and dorsolateral prefrontal (DLPFC) cortices. This universal response is sculpted by innate sensitivity, dictating distinct neural strategies. High-sensitivity individuals showed neural efficiency with earlier processing and proactive gain control; low-sensitivity individuals showed effortful compensation Multivariate pattern analysis classified sensitivity groups from brain activity in a 119-320 ms window at 59.9% accuracy (chance 33.3%). These findings reveal that sensory experience reflects two dissociable processes (universal encoding and personalized strategy) and support EEG as an objective tool for food sensory evaluation.

The brewing conditions are critical factors influencing the nutrient, flavor characteristic and bioactive components of tea. In this work, the effects of different brewing conditions on the nutritional composition, flavor profile, sensory characteristics, and antioxidant capacity of Xanthoceras sorbifolium bud black tea (XBT) were investigated. Results showed that brewing at 90 ℃ for 30 min (XBTI-1) was identified as the optimal condition to contributes high quality of XBT due to higher content of flavonoids, soluble sugars, and total polyphenols, accompanied by the well-balanced nutrients, intense umami, and outstanding richness in XBT infusion. Moreover, 1-octen-3-one, 2-methylpropanal, and 2-methylbutanal were found as flavor markers in 16 key odorants with relative odor activity value (ROAV > 1.0) from 75 volatile organic compounds in XBT infusion. Furthermore, XBTI-1 displayed the strongest antioxidant activity with Antioxidant Potency Composite (APC) index of 0.984. Collectively, 90 ℃/30 min is recommended as the ideal brewing condition for XBT.

Hybrid dry-fermented sausage analogues with texturized pea proteins (TPPs) are emerging, yet flavor formation mechanisms remain unclear. We combined quantitative descriptive analysis with complementary HS-SPME-GC-MS/HS-GC-IMS volatilomics, UHPLC-MS/MS untargeted metabolomics, and marker-gene microbiome sequencing across sausages with different fermentation and ripening stages to map key aroma and their potential microbial and metabolic drivers. Sensory data showed rising fruity, cocoa-chocolate and nutty notes. In total, 47 volatiles were identified by GC-MS and 40 by GC-IMS. Screening of odorants based on relative odor activity value (rOAV) consistently highlighted seven odorants, with a shift from hexanal-dominated raw profiles to linalool-dominated processed profiles, indicating suppression of aldehyde-derived off-notes and enrichment of terpene/ester notes. Metabolomics detected 2467 metabolites, dominated by lipids and organic acids, and short-peptide enrichment suggested intensified proteolysis supplying aroma precursors. Bacterial succession exceeded fungal variation, with Latilactobacillus and Staphylococcus as core taxa. The integrated dataset provides practical markers and microbial/process cues to enhance flavor quality of sustainable hybrid fermented meats.

Background: The rapid growth in consumers' demand for food sensory quality has driven the rapid development of flavor analysis technology. There has been significant progress in the development of bionic sensors based on mammalian sensory systems, but challenges remain in their stability, accuracy, and adaptability in complex food matrices. Hydrogels combined with bionic sensors provide a powerful platform to overcome challenges. However, there is currently a lack of systematic reviews on hydrogel-based bionic sensors for food flavor detection. Scope and approach: This review describes the types and functional characteristics of hydrogels and explains their key roles in the development of bionic sensors and the application progress in flavor detection. It examines the current situation of the integration of hydrogel bionic sensors with emerging technologies such as machine learning, microfluidics, 3D printing, and organoid bionic technology. The challenges and potential future directions of hydrogel bionic sensing technology in food flavor field are also discussed. Key findings and conclusions: Hydrogels, as fixed carriers for the sensor's sensitive elements, are crucial for maintaining receptor functional activity, achieving signal conversion and amplification, and preventing nonspecific adsorption. This makes bionic sensors highly stable, accurate, and adaptable for flavor detection. The integration of emerging technologies with sensors can further enhance their portability, bionic properties and intelligence level. Even so, there are still challenges in building a truly intelligent micro food detection system. Ongoing research helps to promote the development and application of next-generation intelligent flavor analysis systems in future food and intelligent food manufacturing.

Reconstructing the precise biosynthesis of structurally complex natural esters, such as monoterpene esters, in engineered microbes remains a major challenge, owing to the limited repertoire of highly selective alcohol acyltransferases and the lack of compatible pathway modularity. Here, we establish a dual-substrate microbial platform to profile the activities of alcohol acyltransferase (AAT) to synthesize three distinct classes of monoterpene esters: monoterpenyl esters, monoterpenoate esters, and monoterpenyl monoterpenoate esters, enabling access to both natural and non-natural monoterpene ester biosynthetic pathways. Through structure-guided critical residue engineering and dual-substrate molar ratio tuning, we achieve selective biosynthesis of >C2 acyl-CoA-derived monoterpene esters, despite competing intracellular acetyl-CoA. Coculture engineering further redistributed metabolic fluxes between acyl-CoA and alcohol precursors, yielding 11.50 g/L linalyl acetate and 3.16 g/L geranyl butyrate in 1-L bioreactor. This study expands the biosynthetic space of monoterpene esters and provides a versatile strategy to control AAT selectivity, offering a plug-and-play, scalable framework for ester biomanufacturing.

In the modern goji berry industry, accurate and rapid discrimination of geographical origins was crucial for traceability and quality assessment. This study employs multiple chemometric approaches to establish a region-tracing discrimination system based on flavor profile variations across different geographical origins. HS-GC-IMS analysis identified 62 volatile compounds, with ketones and esters as the primary contributors. Eighteen compounds were identified as characteristic markers for geographical origin discrimination of goji berries. Among these, methyl hexanoate ROAV (QH=8.54, GS=9.26, NX-XR =16.5, NX-ZN=8.6), ethyl pentanoate ROAV (QH=8.09, GS=6.17, NX-XR=10.4, NX-ZN=3.7) hexanal ROAV (QH=2.65, GS =11.3, NX-XR =19.9, NX-ZN=5.4) and butyraldehyde ROAV (QH=4.45, GS =7.88, NX-XR =9.52, NX-ZN=5.99) were identified as key components determining the characteristic aromas of different region groups. Electronic tongue analysis indicated that umami and sweetness were the main taste attributes distinguishing samples from the NX region from those of other origins, whereas bitterness was the key attribute differentiating QH from the other origins. The feedforward neural network (FNN) model based on electronic tongue data outperformed the support vector machine (SVM) model, achieving complete discrimination among goji berries from four geographical origins and 100% classification accuracy for blind samples. The region tracing model constructed in this study provides a theoretical basis and technical support for verifying the geographical origin of goji berries.

As natural flavor enhancers, umami peptides are hindered by inefficient traditional identification. In this study, an interpretable multimodal model named Umami-Multi was constructed for umami peptide prediction. The model achieved excellent performance (ACC = 0.9517) by adaptive interaction and fusion of sequence, physicochemical and structural features with a cross-modal gated attention fusion network, feature selection with SENet-based attention, and hyperparameter optimization with multi-fidelity Bayesian optimization. The umami peptide EDALNVNR (threshold: 0.23 mg/mL) with umami-enhancing and saltiness-enhancing effects was screened and identified from Pixian Douban by integrating peptidomics, virtual screening and sensory validation. Model visualization and molecular simulation revealed that the umami characteristics of umami peptides were largely dependent on key amino acids (glutamic acid and aspartic acid), core physicochemical properties (hydrophilicity), critical functional groups (carboxyl groups), and major interaction forces (hydrogen bonds). This study provided the basis for umami formation clarifications and the technical support for novel umami peptides screening.

Drying affects goji berry (Lycium barbarum) quality, yet the involvement of aquaporins (LbAQPs) remains largely unexplored. Four accessions were evaluated to investigate drying impacts on fruit quality and transcriptomic profiles. Drying significantly increased flavonoid content and induced 509 common differentially expressed genes (DEGs) enriched in heat and water deprivation pathways. Comparative analysis between fast- and slow-drying accessions highlighted genes linked to secondary metabolism and stress response. Fruit traits, such as peel thickness and cell size, alongside PIP1.3Lb07G01036 and TIP3.1Lb0600692, significantly influenced water loss rates, while heat-induced TIP3.2Lb03G02112 and PIP2.4Lb08G01903 were associated with drying duration. Furthermore, dual-luciferase assays confirmed that transcription factors like ERF_Lb10G01263 and bZIP_Lb03G00095 activate, while MYB_Lb07G00723 represses, the expression of LbTIP3;2 and LbNIP6;1. Collectively, these LbAQPs and their upstream regulators provide vital genetic resources for developing goji berry cultivars optimized for the drying process.

BACKGROUND:Pumpkin is an important food source. Its flavor significantly impacts consumer acceptance and market competitiveness. Machine learning (ML) can capture non-linear patterns and multivariate relationships in volatile organic compound (VOC) datasets, linking VOC profiles to sensory attributes. In this study, sensory evaluation, gas chromatography-time-of-flight mass spectrometry (GC-ToF/MS) and ML were integrated to identify the flavor profiles of raw and steamed pumpkins and the VOC markers associated with their typical sensory attributes. RESULTS:Raw pumpkin's sensory profile was dominated by green and cucumber aromas, whereas that of steamed pumpkin was characterized by chestnut, creamy, sweetness, and potato aromas. Based on five classification ML algorithms, using a consensus voting strategy, 25 important VOCs were identified across raw and steamed pumpkins. Alcohols were the primary VOCs in raw pumpkin, whereas 2,3-butanedione and methional were the main ML-highlighted VOCs associated with steamed pumpkin. ExtraTrees achieved the highest performance in linking VOCs to sensory attributes. Shapley Additive Explanations (SHAP) summary plots showed that alcohol compounds contributed substantially to the cucumber and green aromas. 2,3-Butanedione and methional emerged as major drivers of creamy, sweet, and potato aromas. CONCLUSION:The study contributes to the understanding of the sensory profile and VOC composition of pumpkins. The identified key VOCs provided reference compounds for sensomics research, establishing a basis for aroma breeding and product processing. © 2026 Society of Chemical Industry.

To mitigate microbial contamination and oxidative browning of Agaricus bisporus, a novel oregano essential oil (OEO)-loaded Pickering emulsion (DCNO-PE) with long-acting antibacterial and antioxidant activity was fabricated using amphiphilic nanocellulose (DCN) from Xanthoceras sorbifolium shells as the stabilizer. DCN was extracted via deep eutectic solvent treatment combined with high-pressure homogenization, therefore exhibiting amphiphilicity and excellent aspect ratio. These distinct characteristics facilitated the formation of a stable oil-in-water interface and stable intermolecular interactions in DCNO-PE. Preservation assays revealed that DCNO-PE inhibited microbial proliferation, reduced total plate count by 1.31 log CFU/g, and suppressed the dominant spoilage bacterium Serratia marcescens, while alleviating oxidative stress and decreasing browning index by 21.25%. By integrating antibacterial and antioxidant properties through sustained OEO release, DCNO-PE extended the shelf-life of Agaricus bisporus to 12 days, demonstrating great potential for edible fungus preservation and promoting the development of natural food preservatives and functional packaging materials.

This study systematically explores the spatiotemporal Electroencephalography (EEG) characteristics of astringency induced by different concentrations of tannic acid (TA). Utilizing spectral decomposition and time-frequency domain analytical techniques, it revealed that the neural responses triggered by astringency stimuli exhibit distinct temporal dynamics and spatial specificity. The results indicated that under astringency stimulation, the power spectral density in the delta band undergoes a dynamically enhanced change. As TA concentration increased, EEG signal intensity exhibited a progressively enhanced positive correlation trend, reflecting the brain's dynamic discrimination of astringency intensity. The neural response to astringency stimulation demonstrated distinct spatial directionality, with pronounced neural activation patterns observed in frontal lobe regions. This study underscores the specificity and complexity of the brain's response to varying intensities of astringency stimulation, while providing a novel approach for evaluating astringency intensity through physiological signals.

Table grapes are produced for direct consumption compared to wine grapes for winemaking. Aroma significantly influences table grape quality, and harvest timing is crucial. This study investigated the aroma profiles of three table grapes (Kyoho (JF), Shine Muscat (YGMG), and Muscat of Alexandria (YLSD)) at six ripening stages using sensory evaluation and instrumental analysis. As ripeness increased, JF and YLSD grapes shifted from green and floral to fruity and sweet, while YGMG grapes intensified in sweet, fruity, green, and floral. Biomarkers for distinguishing ripening stages were identified, such as furfural, neral, and (Z)-2-hexen-1-ol for JF, (Z)-3-hexen-1-ol, (E)-2-hexenal, and linalool for YGMG, and furfural, 3-methylbutanal, and nerolidol for YLSD. Machine learning based on E-nose enabled rapid classification and prediction of flavor, with the SVC model achieving the highest accuracy (0.9394), which may facilitate quality grading. This study provides a theoretical basis and technical support for monitoring grape aroma quality.

Food-derived extracellular vesicles (FDEVs) are nanoscale membrane-bound structures naturally present in plant- and animal-derived foods, carrying diverse bioactive components including lipids, proteins, small RNAs, and metabolites. Increasing evidence suggests that FDEVs may contribute to dietary bioactivity; however, their physiological relevance remains difficult to establish due to the limited understanding of how food processing and gastrointestinal conditions influence their biological functions. Recent studies have highlighted that food processing can reshape multiple aspects of FDEV properties, including membrane integrity, particle characteristics, aggregation behavior, and cargos accessibility. These processing-induced alterations may determine the stability, transformation, and biological availability of FDEVs during gastrointestinal digestion and subsequent interactions with host tissues and microbiota. Nevertheless, current investigations often examine processing effects, digestive fate, and biological outcomes in separate experimental systems, resulting in a lack of integrated understanding of structure-function relationships under physiologically relevant conditions. This review critically evaluates the current knowledge regarding the impact of food processing on FDEV structure, gastrointestinal fate, and potential bioactivities. Particular attention is given to the challenges associated with dose relevance, vesicle characterization, biodistribution assessment, and the interpretation of cargos-mediated effects. By integrating evidence across processing, digestion, and biological response studies, this review proposes a processing-digestion-bioactivity framework as a conceptual model for organizing current evidence, identifying knowledge gaps, and guiding future studies toward physiologically meaningful mechanisms.

In this thesis, the high-yielding ethyl acetate yeast and lactic acid degrading bacteria in the brewing process of Light-Flavored Baijiu were taken as the research objects, and the interaction and metabolism differences of microorganisms in the co-cultivation process were compared through the establishment of mixed-bacteria fermentation systems with different ratios, and provide a theoretical basis for regulating flavor formation during Light-Flavored Baijiu fermentation. The main research content and results were as follows: During fermentation, high-yielding ethyl acetate yeast (J2) inhibited the growth of lactic acid-degrading bacteria (R12), but its development was not affected by lactic acid-degrading bacteria; untargeted metabolomics analysis indicates that the inhibitory effect of high-yield ethyl acetate yeast on lactic acid-degrading bacteria was achieved through many different metabolic pathways. Among them, the cysteine and methionine metabolism pathway has the strongest metabolic effect. High-yield ethyl acetate yeast secretes substances such as cystathionine, n-butylamine, tryptophan, alanine, and triethyl orthoacetate into the extracellular space through these metabolic pathways, inhibiting the growth of lactic acid-degrading bacteria. Lactic acid-degrading bacteria were able to influence the metabolism of ethanol, organic acids, and flavor substances in high-yield ethyl acetate yeast. Lactic acid-degrading bacteria could promote the metabolism of ethanol, citric acid, succinic acid, and tartaric acid in high-yield ethyl acetate yeasts, effectively lowering the content of ethyl lactate with an appropriate proportion of inoculation; it also stimulates the metabolism of some new esters, alcohols, aldehydes, and change the content of some esters and phenols in the co-culture system due to the metabolism of lactic acid degrading bacteria themselves.

Kokumi taste, a sensory attribute that enhances flavor complexity, continuity, and mouthfulness without being a basic taste, has gained significant scientific attention. This review synthesizes advances in kokumi peptide research, focusing on their flavor-enhancing mechanisms, receptor interactions (notably calcium-sensing receptors), natural sources, and food applications. We critically evaluate global research progress and key debates, such as receptor synergy mechanisms and limitations of current flavoring mechanism. Future directions emphasize establishing a global kokumi peptide database, developing artificial intelligence-driven prediction platforms, improving in vivo validation, and optimizing industrial-scale production. This review aims to provide a theoretical foundation for food flavor science and to facilitate the integration of kokumi peptides into the development of functional foods and flavor enhancers.

Background Traditional meat products generally contain high levels of salt, and long-term consumption increases the risk of chronic diseases. Most existing studies achieve salt reduction in meat products through salt substitutes, while research on personalized, intelligent, precise saltiness regulation mechanisms involving oral processing, trigeminal stimulation, and multisensory integration remains limited and requires further optimization. Scope and Methods This structured narrative review summarizes recent advances in salt reduction in meat products, covering the spatiotemporal distribution of salt, oral processing dynamics, Na+ release kinetics, salt taste receptor signaling, multisensory integration, low-salt protein network re-engineering, and machine learning and data-driven decision-support for customized formulation. Key Findings and Conclusions To ensure the quality of salt-reduced, meat products and improve acceptance among general consumers and special populations, this review proposes a salt perception programming framework that shifts salt reduction from empirical formulation adjustment toward coordinated optimization of structure, release, and perception. Current evidence supports salt spatial distribution, multisensory enhancement, physical processing, enzymatic cross-linking, and hydrocolloid incorporation as practical strategies for improving saltiness perception and maintaining product quality in salt-reduced meat products. By contrast, machine learning, data-driven decision-support frameworks, and virtual saltiness customization represent promising long-term approaches, but remain largely conceptual without validation in actual low-sodium meat products. Moreover, sensory responses should not be regarded as direct surrogates for long-term Na+ intake or health outcomes. Future research should prioritize quantitative investigation of the coupling between Na+ release kinetics and multisensory perception and establish robust experimental evidence for data-driven decision-support systems for personalized salt-reduction design.

Meat analogs represent a rapidly expanding sector of the food industry; however, it remains unclear whether increased volatile diversity in plant-based products translates into beef-like sensory authenticity, particularly due to the lack of integrated sensory-instrumental evaluation in commercially available systems. Here, five commercial plant-based patties were compared with conventional beef (TM) under standardized cooking conditions using combined sensory mapping and instrumental analyses. Sensory resemblance to beef was not determined by volatile abundance, but by specific sensory drivers and reaction-derived aroma markers. Napping-UFP showed that S1 and S2 were positioned closer to TM in perceptual space, with S1 additionally exhibiting meat-like textural attributes (tight, springy, chewy), whereas other samples were associated with softer structures and oxidation-related off-notes. Instrumental analyses revealed method-dependent similarity patterns: E-nose PCA showed partial odor overlap between S2 and TM, while E-tongue PCA indicated distinct taste-response distributions rather than convergence toward beef. SPME-GC-TOF-MS identified 58 volatile compounds in TM and 87-108 in plant-based samples; despite this higher diversity, beef was characterized by lipid-oxidation-related aldehydes (e.g., hexanal, nonanal) and reaction-derived compounds associated with cooked meat aroma, whereas plant-based patties were dominated by furans, esters, and plant-derived oxidation products. These results demonstrate that improving meat-like authenticity requires targeting reaction-specific aroma pathways and texture architecture rather than increasing overall volatile complexity, providing a mechanistic basis for next-generation formulation strategies.

Smoked Lapsang Souchong tea (SLST) is renowned for its unique smoky aroma and relatively low bitterness, attributed to the pinewood smoke withering (SW) process. This study investigated the role of SW in reducing SLST bitterness through an integrated approach. Electronic tongue analysis, sensory evaluation, and cell-based calcium imaging assay confirmed that SW significantly reduced tea bitterness. Untargeted metabolomics revealed SW substantially decreased key bitter phenolics, including 10 monomeric catechins, 14 glycosylated flavonols, and 6 galloylated phenolic acids. Concurrently, key pinewood smoke odorants, guaiacol and α-terpineol, directly inhibited TAS2R14 activation by pre-smoke withered tea extract, reducing intracellular calcium signaling by 37% and 21%, respectively. Quantum chemistry calculations further demonstrated that these odorants readily interact with bitter compounds primarily through intermolecular hydrogen bonding and van der Waals forces, suggesting a potential mechanism for inhibiting bitter taste signaling. This study provides novel insights into how smoke withering refines tea flavor by attenuating bitterness.