Recently, Associate Professor Haibin Yuan and Professor Huaixiang Tian's team from the Faculty of Flavor, Flagrance and Cosmetics comprehensively summarized the latest research progress on the interactions between dairy proteins and aroma compounds. The related findings, titled "Interactions between dairy proteins and aroma compounds: A comprehensive review of mechanisms, processing effects, and analytical methods," were published in the international food field top journal Food Hydrocolloids (Chinese Academy of Sciences Zone 1, TOP journal, IF=12.4). Associate Professor Haibin Yuan is the first author, Professor Huaixiang Tian is the co-corresponding author, and Master's student Can Tang participated in the literature review and writing work.
The flavor perception of dairy products largely depends on the interactions between dairy proteins (DP) and dairy aroma compounds (DAC), which collectively determine the binding, retention, and release behavior of volatile molecules in the complex dairy matrix. Different from traditional reviews that merely focus on static equilibrium constants and molecular-level affinities, this paper innovatively proposes a "multi-scale dynamic coupling mechanism." It argues that the aroma behavior in real dairy products cannot be explained solely by molecular-level chemical affinity but is fundamentally regulated by the confinement effects of the processing-driven colloidal matrix.
The review comprehensively outlines the non-covalent (hydrophobic, hydrogen bonding, electrostatic, van der Waals) and covalent interaction mechanisms between dairy proteins (especially disordered casein and globular whey proteins) and aroma molecules. On this basis, it delves into how industrial processing variables (such as heat treatment, pH adjustment, ionic strength, fermentation, homogenization, and spray drying) dynamically reshape protein conformation and matrix microstructure, thereby transforming simple thermodynamic binding into complex, diffusion-controlled matrix entrapment mechanisms. Furthermore, the article systematically evaluates the advantages and limitations of analytical techniques such as spectroscopy, chromatography-mass spectrometry, thermodynamic and kinetic measurements in studying this field.
More prospectively, the article highlights the introduction of emerging computational tools like molecular dynamics simulations and machine learning. It explores how geometric deep learning frameworks can overcome the pain points of traditional experiments, enabling cross-scale flavor prediction from molecular-level parameters to macroscopic sensory perception. This review provides a highly valuable theoretical foundation and a multi-scale industrial guidance framework for achieving flavor stability, targeted aroma delivery, and off-flavor reduction in dairy products through rational protein design and optimization of processing parameters.
