Early in life, gut, immunity, growth, and brain development occur rapidly and vary considerably. Growing evidence suggests that probiotics, prebiotics, and synbiotics effectively support host health during early life. Bifidobacterium is the dominant microbiota in the infant gut, while human milk oligosaccharides (HMOs) function as natural prebiotics in breast milk. Synbiotics combining HMOs and Bifidobacterium have been proposed. This review systematically elucidates the interaction between HMOs and Bifidobacteria, providing theoretical support for synbiotics based on both. The mechanisms by which Bifidobacteria utilize HMOs are classified as “exogenous degradation” and “endogenous degradation”. Furthermore, this review summarizes the preclinical and clinical evidence regarding HMOs and Bifidobacterium-based synbiotics in promoting host health and preventing early-life diseases. Mechanistically, synbiotics primarily regulate host gut health and immune function through direct effects. Additionally, synbiotics exert their effects on non-gut health mainly through indirect mechanisms, manifested in the production of beneficial metabolites (e.g., short-chain fatty acids (SCFAs), indole derivatives, gamma-aminobutyric acid (GABA), sialic acid, and 5-hydroxytryptamine (5-HT)).

2-acetylfuran is a product of the Maillard reaction and is widely found, especially in heat-processed foods such as grain products, baked goods, and dairy products. Although 2-acetylfuran contributes to flavor, high concentrations may be toxic. Its target organs and dose–response relationships remain poorly characterized. In this study, transgenic zebrafish with fluorescently labeled immune and neural systems were used to assess the effects of 2-acetylfuran on immune and neural development. Wild-type zebrafish were employed to assess the toxicity of 2-acetylfuran on locomotor ability, gastrointestinal development, and liver function. The maximum non-lethal concentration (MNLC) and the 10% lethal concentration (LC₁₀) for zebrafish embryos were 0.844 and 0.889 μL/mL, respectively. Regarding immunotoxicity, at concentrations of 0.281, 0.844, and 0.889 μL/mL, 2-acetylfuran significantly reduced the numbers of neutrophils, T cells, and macrophages. Regarding locomotor and neurotoxicity, motor speed and total locomotor distance were significantly reduced at 0.844 and 0.889 μL/mL. These findings were consistent with neurodevelopmental assessments, in which 0.844 μL/mL 2-acetylfuran resulted in a significant increase in apoptotic cells in the central nervous system and markedly shortened peripheral motor nerve lengths. Regarding gastrointestinal toxicity, 0.844 and 0.889 μL/mL 2-acetylfuran significantly reduced the gastrointestinal area, while neutrophil counts showed no significant changes, suggesting a relatively mild effect on the gastrointestinal tract. Regarding hepatic toxicity, all tested concentrations of 2-acetylfuran primarily increased the delayed yolk sac absorption area. Furthermore, at 0.844 μL/mL, histological examination revealed hepatic pathological changes characterized by hepatocyte nuclear swelling, vacuolar degeneration, and hepatocyte necrosis. In summary, this study reveals the multi-organ toxicity profile of 2-acetylfuran in the zebrafish model, with particularly high sensitivity in the immune system and liver. This research provides theoretical support for risk assessment and process control of 2-acetylfuran in foods.

Human milk oligosaccharides (HMOs) have high structural diversity and extensive isomerism, posing major analytical challenges. Capillary electrophoresis–laser-induced fluorescence detection (CE-LIF) separates structurally similar isomers by charge-to-hydrodynamic radius ratio. Here, an APTS-derivatized CE-LIF method was developed and optimized for HMO separation and quantification. Using nine representative standards, key parameters (buffer, injection mode, capillary, voltage, and temperature) were systematically evaluated. Optimal performance was obtained with a TBE solution supplemented with 3% polyvinylpyrrolidone, 4 M urea, and 10% boric acid, electrokinetic injection, and separation at 30 kV and 30°C on a fluorocarbon-coated capillary. The method showed good linearity (R2 = 0.992–0.998), high sensitivity (LOD: 0.02–0.18 mg/100 g; LOQ: 0.06–0.60 mg/100 g), and satisfactory recoveries (90.4–105.5%). The validated CE-LIF method was successfully applied to human milk and infant formula, enabling simultaneous analysis of neutral and acidic HMO isomers, and providing a robust, high-resolution platform for routine HMO research and quality control.

Bovine colostrum (BC) is essential for neonatal survival and human health, owing to its rich bioactive content. However, highly abundant proteins constrain the proteome depth of BC. By integrating magnetic nanoparticle enrichment of low-abundance proteins with a DIA-MS workflow, we achieved an unprecedented depth in BC proteome profiling. Across 463 samples from 10 Chinese provinces, our rapid DIA-MS platform identified 4,558 proteins. Importantly, the identified data set represents an MMN-enriched and DIA-quantifiable proteome space rather than an absolute reconstruction of the entire colostrum proteome. A conserved core of 409 proteins (≥90% of samples) was enriched for immune-, defense-, and transport-related functions. Multivariate analysis revealed a distinctive Heilongjiang (HLJ) profile with reduced abundance of proteins annotated to acute-phase or inflammatory processes and increased abundance of proteins associated with cellular maintenance pathways. Machine learning identified a 9-protein HLJ marker panel that distinguished HLJ from other provinces with near-perfect accuracy (Area Under the Curve, AUC ≈0.99). This study provides the first comprehensive characterization of the BC proteome across 10 Chinese provinces and defines an HLJ-specific biomarker signature, offering tools for traceability and quality control of BC products.

Various Maillard reaction products (MRPs) are formed in infant formula during thermal processing and storage. However, safety assessments are often based on individual compounds, which may not adequately reflect the potential risks associated with exposure to multiple coexisting components in real food products. Therefore, this study used GC–MS to screen for typical MRPs in infant formula. Based on the detection results, glyoxal, 2-acetylfuran, 2-furfural, and 5-hydroxymethylfurfural (5-HMF) were selected to establish a mixed-exposure system. Subsequently, using zebrafish as a model, we systematically evaluated the toxic effects of two mixture concentrations, 1/9 maximum non-lethal concentration (MNLC) and 1/3 MNLC, on the immune system, nervous system, gastrointestinal tract, and liver. The results showed that these four MRPs coexisted in infant formula and that combined exposure induced significant biological damage even at low doses. Both mixture concentrations significantly reduced neutrophil and macrophage levels and promoted apoptosis in central nervous system cells. The higher-concentration mixture further reduced T-cell counts, suppressed motor neuron development, and decreased locomotor activity during the light phase. In addition, the higher-concentration mixture decreased the gastrointestinal area, increased the liver area, delayed yolk sac absorption, and caused marked histopathological damage. Compared with single-compound exposure, combined exposure exerted more pronounced effects on immune- and neuro-related endpoints, suggesting synergistic interactions among different MRPs and identifying the immune and nervous systems as the more sensitive targets of toxicity. In summary, mixed exposure to typical MRPs in infant formula may pose a greater biological risk than that predicted by evaluations based on individual components alone. This study provides experimental evidence for identifying combined toxicity and improving the safety risk assessment of MRPs in thermally processed infant foods.