OPTIMIZATION OF AN ANALYTICAL METHOD FOR AMINO ACID PROFILING IN PROTEINS USING HPLC-PDA/FLD AND CHEMOMETRICS

The accurate quantification of amino acids is critical for evaluating the nutritional quality and functional properties of food proteins. Hydrolysis is a fundamental step in analytical approaches, and reliable results require optimized conditions. In this study, chemometric tools were applied to evaluate the effects of acid hydrolysis parameters and establish optimal conditions. A Box–Behnken design (BBD) was proposed to fit a model to optimize acid hydrolysis process parameters. The goal was to achieve the maximation of protein hydrolysis yield followed by OPA derivatization and to compare two HPLC detectors: fluorescence (FLD) and photodiode array (PDA).
The experimental design consisted of a three-factor, three-level BBD with 15 experiments, including three replicates at the central point. The variables studied were reaction temperature (100–115 °C; V₁), reaction time (18–24 h; V₂), and solid-to-liquid ratio (20:1–50:1, m:v; V₃). Modeling and coding were carried out in Excel and Octave software, while ANOVA and F-tests were applied to assess significance and predictive power. Samples of bovine serum albumin (BSA) were used for experiments and hydrolysis yield was expressed as the proportion of recovered amino acids relative to total protein content. Randomized triplicate experiments were analyzed by HPLC (Shimadzu) and a Kinetex EVO C18 column (Phenomenex).
The BBD produced a polynomial model without lack of fit (Fcalc < Ftab) for each detector: PDA, Y = 69.6 – 4.4V₁ – 6.3V₃ + 4.3V₁² (R2 = 0.9452); FLD, Y = 73.2 – 5.7V₁ – 8.0V₃ + 4.7V₁² (R2 = 0.8684). In both models, V₁ and V₃ were significant at low levels, while V₂ was not statistically significant. The optimal hydrolysis conditions defined by BBD were 100 °C, 18 h, and a solid-to-liquid ratio of 20:1 (m:v).
A key finding was the performance difference between the detection methods. While the PDA detector provided a model with a higher regression coefficient, suggesting greater precision for predicting yields, the FLD detector measured a significantly higher absolute yield of amino acids (91.7% vs. 84.6%) from the same hydrolyzed samples. This indicates that the FLD's superior sensitivity allows for a more complete quantification of the amino acids released by the hydrolysis process.
In summary, this optimization work provided a robust hydrolysis protocol for generating accurate amino acid profiles. The choice of detector should be purpose-driven: FLD is optimal for maximizing detection sensitivity, whereas PDA is preferred for analytical precision, both being good options for correct interpretation of food protein quality.