P1-73 Conventional Methodologies vs. Metabolomics for the Quantification of Spoilage of Beef Filets and Minced Beef

Monday, July 23, 2012
Exhibit Hall (Rhode Island Convention Center)
Anthoula Argyri, National Agricultural Research Foundation, Lycovrissi, Attikis, Greece
Efstathios Panagou, Agricultural University of Athens, Athens, Greece
Fady Mohareb, Cranfield University, Cranfield, United Kingdom
Conrad Bessant, Cranfield University, Cranfield, United Kingdom
George-John Nychas, Agricultural University of Athens, Athens, Greece
Introduction: Meat industry needs rapid analytical methods or tools for the quantification of meat spoilage and for the estimation of the product’s shelf life.

Purpose: To evaluate metabolomics as a potential tool to quantify meat spoilage.

Methods: In the context of SYMBIOSIS - EU project, ground meat and beef filets were stored aerobically, under modified atmosphere packaging (MAP) and under MAP with the presence of the volatile compounds of oregano essential oil (MAP/OEO) at 0, 5, 10 and 15°C. The microbial association, i.e., total viable counts, pseudomonads, Brochothrix thermosphacta, lactic acid bacteria, Enterobacteriaceae, yeasts/molds, was assessed in parallel with sensory analysis, pH measurements and the evolution of the metabolic products, as well as the compounds occurring in the meat substrate using HS/SPME-GC/MS and HPLC.

Results: The data collected by HPLC and GC/MS were correlated with microbial counts and sensory scores to estimate the shelf life of the beef, aiming mainly at the early detection of spoilage. Both parameters, i.e., temperature and packaging, were found to have a great impact on the evolution of end-products during storage that resulted in distinct profiles in both minced and beef filets. Correlation of the metabolic profile, monitored with these instruments and developed during beef storage, with the sensory discrimination of the samples, was performed with principal components analysis (PCA) and factorial discriminant analysis (FDA), whilst quantitative predictions of the different microbial groups were performed with partial least squares-regression (PLS-R). The performance of these models for the different microbial groups was 75.34 to 91.78% within the ± 20 % relative error.

Significance: Overall, it was shown that metabolic profiling derived from HS/SPME-GC/MS and HPLC analysis of meat combined with advanced computational analysis may be considered as a potential method to predict the spoilage of a meat sample regardless of packaging type and storage temperature.