FTIR for food bacteria diferentiation

The bacteria which contribute to food spoilage can be identified by a high-throughput FTIR spectroscopic procedure that is amenable to automation and will be useful for the food industry. The original technique was developed by scientists in Norway in 2010 and has now been applied to the characterisation of 59 fungal strains belonging to 19 species and 10 genera. Volha Shapaval from Nofima AS, Ås, and colleagues described the procedure in Journal of Applied Microbiology.

The spores were grown on malt extract agar for 5-6 days then collected on cotton tips for secondary growth in an automated micro-titre plate system with a capacity of 200 samples, using malt extract broth as growth medium. The second growth step took a further 5 days but the overall growth time of 10-11 days remains shorter than that required for traditional morphological methods, although it is longer than genetic methods.

The mycelia were washed with water, homogenised and dried before analysis by FTIR spectroscopy. The sample preparation does not introduce any chemicals that might affect the spectra.

A cluster analysis revealed that all of the fungal strains clustered in one of two groups, according to the phylum. This analysis was used to devise an artificial neural network model for classification which had a success rate of 94% at the genus level and 93.9% at the species level. All strains of Eurotium herbariorum and Mucor hiemalis, and the genera Aspergillus and Penicillium were correctly identified at the species level. The success rates for other species were lower, with an overall degree of identification varying from 75-100%.

The method will be useful in food industries for which moulds present problems, like bakeries, cheesemaking, and the juice and soft drinks industries. It could be refined by using different culture media to increase the success rate and could be transferred to a robotics platform to provide a fast automated procedure.