Researchers used the technique with chemometrics to detect foodborne pathogens after isolation from artificially spiked meat and poultry products.
Results are promising enough to test the identification of real-world samples, they said.
Identification process requires a timeframe of two hours including sample preparation and Raman spectroscopic analyses.
Chemometrics means performing calculations on measurements of chemical data, according to Camo, who analyse data and design experiments.
To identify meat-borne pathogens, 12 independent identification samples were prepared. Six originated from Columbia blood agar and three each from spiked minced beef and spiked chicken breast.
Each Raman spectrum of the test samples was taken as a basis to build up a three level classification model. First, the spectrum was analysed concerning a categorization into the Gram-positive or Gram-negative group.
Afterwards it was assigned to one of the genera and then one of the species was performed.
A database of 19 species (24 strains) of the most important and non-pathogenic bacteria associated with meat and poultry was created (E.coli, L. monocytogenes, P. aeruginosa, Salmonella spp., S. aureus or Y. enterocolitica).
Common methods based on pre-enrichment by bacterial cultivation with specific media can be laborious and time-consuming and other techniques require pre-cultivation, complicated separation techniques or are expensive.
Raman microspectroscopy with Raman excitation wavelengths in the visible wavelength region is a promising method to detect microorganisms on a single-cell level with minimal sample preparation, said the researchers.
Vibrational spectroscopic approaches (IR absorption and Raman spectroscopy) have shown their great potential to rapidly identify microorganisms on a single cell level with minimal sample preparation, said the researchers.
Pieces of minced beef and chicken breast were spiked with pathogenic microorganisms.
To separate the bacteria from structured meat surface for Raman spectroscopic measurements an additional isolation and concentration procedure was applied.
Microorganisms were prepared on three different meat-like media to account for the natural environment of meat.
The media were Columbia blood agar, which is a standard agar for microbial clinical purposes, brain heart infusion agar (main ingredients: beef heart and calf brain infusions) and Müller–Hinton agar (main ingredient: beef infusion).
Validation of the database spectra showed that 99.5% of the data were correctly assigned to the Gram-positive (‘Gram+’) or Gram-negative (‘Gram−‘) group.
The second level of the classification tree subdivides the Raman spectra into the respective genera before a separation on a species-level was carried out in a third step.
Accuracies on a species-level were determined for Listeria spp. (94.7%), Staphylococcus spp. (100%), E. coli type strains (95.8%), Pseudomonas spp. (99.4%), Salmonella spp. (98.9%) and Yersinia spp. (90.6%).
Source: Food Microbiology, Volume 38, April 2014, Pages 36–43
Online ahead of print, DOI: 10.1016/j.fm.2013.08.007
“Identification of meat-associated pathogens via Raman microspectroscopy”
Authors: Susann Meisel, Stephan Stöckel, Petra Rösch, Jürgen Popp