Protein haze in white wine can be a serious quality defect because consumers perceive hazy wines as "spoiled". Unfortunately, a specific method for the detection, or selective treatment, of such proteins in affected wines does not exist. Herein we investigate on the development of an easy-to-use sensor device that allows detection of haze-forming proteins (HFPs). Such a device is expected to overcome the limitations of the "heat test" currently used to assess the protein content in wine and the amount of bentonite needed to remove such proteins. To this aim, three different approaches were explored. Firstly, an impedimetric immunosensor against chitinases was developed and its performance assessed. Secondly, the exploitation of the dual role of HFPs as biorecognition element and analyte to develop an impedimetric biosensor was evaluated, in what can be considered a very unique strategy, representing a new paradigm in biosensing. Lastly, Fourier transform infrared (FT-IR) spectra were collected for various wine samples and chemometric tools such as discrete wavelet transform (DWT) and artificial neural networks (ANNs) were used to achieve the quantification of HFPs. Detection of HFPs at the μg/L level was achieved with both impedimetric biosensors, whereas the FT-IR-based approach allowed their quantification at the mg/L level in wine samples directly. The sensitivity of the developed methods may enable the rapid assessment of wine protein content.
Keywords: Artificial neural networks; Biosensor; Chitinases; Haze-forming proteins; Thaumatin-like proteins; Wine.
Copyright © 2023 Elsevier B.V. All rights reserved.