Milena Portela dos Santos 1, Cícero Alves Lopes Júnior1
UV-Vis spectrophotometry is a consolidated technique used to identify and quantify proteins based on the attenuation of electromagnetic radiation measurement by an absorbing substance. Over the past few years, there has been a notable surge of spectrophotometer prototype development looking for a low cost, sensible, precisely and accurate analyses, mainly using alternative components such as LEDs (Light Emitting Diodes), light or color sensors associated with the 3D printing technology. In this context, the development and application of a platform for the quantification of total proteins in biofortified cowpea cultivars (Vigna unguiculata (L.) Walp) is a promising strategy for investigating the protein content of this most consumed legume in the world.
The spectrophotometer consists of an LED as the radiation source and an LDR (Light Dependent Resistor) as the light detector, which convert a light signal into an electrical signal. The system was controlled by an Arduino. For the Arduino application, a code was acquired for controlling light intensity and data input – which required conversion to absorbance values. The external part of the prototype was modeled using Autodesk Fusion 360 software and printed on a Creality K1 printer.
The analytical curve was made with six standard BSA solutions in the range of 20 to 95 using Tris-base buffer (pH = 7.2) as solvent. The protein quantification was performed using the Bradford method, where the Coomassie Brilliant Blue G-250 dye interacts with the positive functional groups in the hydrophobic pocket of the protein, resulting in a color change from reddish-brown to blue. In addition, its application as a Picture Box for protein quantification using Digital Image Colorimetry (DIC) was evaluated – which demands images captured by a smartphone. For statistical evaluation, the Partial Least Squares (PLS) model was used with evaluation of RMSEC (Root Mean Square Error of Calibration), RMSEP (Root Mean Square Error of Prediction), and RMSECV (Root Mean Square Error of Cross-Validation) parameters.
For digital image acquisition, a Samsung A15 smartphone with a 50-megapixel camera was used. Two photos were taken of each blank sample and two of each test sample, with analyses performed in duplicate, totaling 44 images. The processing of the images was carried out in ChemoStat software, in which the photos were cropped to dimensions of 30 × 30 pixels, and data from the red and blue channels were extracted for application of the PLS model.
For the calibration curve obtained using the spectrophotometer, the linear equation y = 0.0027x – 0.0364 and r = 0.9969 were obtained, in addition to the limit of quantification (LOQ) and the limit of detection (LOD) were = 0.0885 mg/L and 0.0265 mg/L, respectively. For the DIC method, the best model results were achieved with cross-validation using 4 latent variables, R2 = 0.998, RMSEC = 1.233 mg/L, RMSEP = 8.1395 mg/L, and RMSECV = 25.0808 mg/L.
Agradecimentos: Universidade Federal do Piauí and CNPq