Work

Portable mid-infrared spectroscopy combined with chemometrics to detect toxic metabolites, aflatoxins in Aspergillus-infected peanuts

Year: 2024

Type: article

Abstract: A portable Mid-Infrared spectroscopy-based approach was developed for the non-destructive and rapid (∼1 min) detection of aflatoxins in Aspergillus -infected peanuts, effectively addressing the challenges in screening biological samples. In this study, a total of 274 kernels were inoculated with Aspergillus flavus ( A. flavus , NRRL 3357 and AF13) strains of fungus. Spectral data of inoculated peanut kernels were collected by a field-deployable FT-IR system, and further analyzed by chemometrics. Reference results of aflatoxins levels in inoculated peanuts were determined by uHPLC-MS/MS analysis. Interestingly, aflatoxin M 1 was detected in A. flavus -infected peanut kernels, which was confirmed by its fragmentation pattern. Supervised classification algorithms, developed by OPLS-DA (orthogonal partial least squares discriminant analysis) and SIMCA (Soft independent modeling of class analogies) both achieved high sensitivity (94.7%) in identifying fungal contamination with aflatoxin levels exceeding 3 ppb. In addition, the regression algorithm, developed by PLSR (partial least squares regression), showed a strong regression correlation (Rpre = 0.85) and excellent residual predictive deviation (RPD = 6.2) for quantifying total aflatoxins level in individual kernels. The integration of this FT-IR system with advanced predictive algorithms provided a rapid, user-friendly, and non-invasive solution for detecting A. flavus -infected peanuts with aflatoxins and quantifying aflatoxin levels. This proposed method would offer the feasibility of doing routine, in-plant aflatoxin screening "on the go," offering a cost-effective and scalable alternative to labor-intensive wet chemistry methods (e.g., uHPLC-MS/MS). This study would also inspire further innovation in non-destructive and high-throughput screening technologies for food safety applications.

Source: LWT

DOI: 10.1016/j.lwt.2024.117186