Calibração pseudo-univariada utilizando MCR-ALS para quantificação de tirosina em albumina por espectroscopia NIR

Abstract

Albumin, a protein present in eggs, stands out for its high biological value and its profile of essential amino acids. Tyrosine, a non-essential aromatic amino acid, acts as a precursor of important neurotransmitters and hormones and is widely used in nutritional formulations. Tyrosine is crucial because it is a precursor of neurotransmitters such as dopamine, adrenaline, and noradrenaline, which are essential for mood, focus, memory, and stress response, in addition to participating in the production of thyroid hormones and melanin (skin and hair pigment). Considering the importance of this amino acid, this study aimed to quantify tyrosine in albumin by means of near-infrared spectroscopy (NIR), using pseudo-univariate calibration, and to evaluate the detection capability of the method through the analysis of figures of merit. Different tyrosine contents (m/m) were added to albumin samples, with percentage concentrations ranging from 0% to 100% tyrosine, and all measurements were performed in triplicate. The NIR spectra were pre-processed to minimize radiation scattering effects using standard normal variate (SNV), followed by curve smoothing using the Savitzky–Golay algorithm, and subsequently processed by Multivariate Curve Resolution–Alternating Least Squares (MCR-ALS). The relative concentrations obtained by MCR-ALS were correlated with the actual concentrations. The calibration curves obtained were evaluated based on figures of merit, including sensitivity, analytical sensitivity, coefficient of determination (𝑅2), correlation coefficient (R), linearity, limits of detection and quantification, and root mean square error. Curves considering 100% and 50% tyrosine were compared, with the restricted curve (0–50%) showing better linearity (𝑅2 = 0.9741) and higher sensitivity (S = 0.092). It was concluded that, under the evaluated conditions, the method is suitable for the quantification of tyrosine in albumin, with relatively low limits of detection (LD = 1.75%) and quantification (LQ = 5.31%), suggesting that the methodology can be applied to several purposes related to the quantification of this amino acid.