Vibrational Spectroscopy for Bacterial Identification
Vibrational spectroscopy has been extensively applied to the study of bacteria in food stuffs (Lu et al., 2011; Pahlow et al., 2015). The characteristic fingerprints generated contain information about the biochemical constitution of bacterial cells and enable the differentiation of the organisms at the species and even strain level.
Many studies have analyzed bacterial foodborne pathogens and were aimed at either the identification of bacterial species present in food products or the detection and in some cases quantification of a concrete species with pathogenic character. In further studies, various foodborne pathogenic bacterial species have been studied together and could be successfully differentiated and identified by SERS approaches (Fan et al., 2011; Sundaram et al., 2013; Xie et al., 2013). A spectral library of 19 bacterial species of the most important harmful and nonpathogenic bacteria associated with meat and poultry has been created and tested for the identification of spiked meat samples. The test samples were correctly assigned to their genus and in most cases down to the species level by Raman fingerprinting and a three-level classification model by means of support vector machines (Meisel et al., 2014).
Bacterial identification by FT-IR achieved 100% of correct species identification for Gram-positive bacteria and 80% for Gram-negative bacteria (Janbu et al., 2008; Sandt et al., 2006). IR and Raman spectroscopy have also been applied to differentiate strains at the subspecies level with the aim of typing different biotypes and serotypes. In this sense, an FT-Raman procedure was successful in discriminating different E. coli strains on whole apples and accurately differentiated the nonpathogens from pathogens, including E. coli O157:H7 (Yang and Irudayaraj, 2003). Likewise, strains of Y. enterocolitica have been distinguished into the main biotypes and serotypes (Kuhm et al., 2009). In this study, species of the genus Yersinia that cannot be differentiated by conventional biochemical methods exhibit specific IR fingerprints, allowing the clear discrimination at the species level. In addition, the presence of the ail gene, one of the main pathogenicity markers, was demonstrated using FT-IR and correct identification of isolates concerning the ail gene was achieved in 98.5%. Fan et al. (2011) successfully applied SERS to identify E. coli O157:H7 and Staphylococcus epidermidis in a mixed bacterial sample. Nevertheless, analysis of bacterial mixtures of different species remains challenging. Furthermore, bacterial identification generally includes cultivation procedures previous to analysis. This implies that only cultivable bacterial cells can be detected and the risk of the presence of injured but viable cells, spores, and already produced toxins remains. Al-Qadiri et al. (2008) applied FT-IR spectroscopy to detect sublethally heat-injured S. enterica var. typhimurium and L. monocytogenes cells. The studies of injured bacterial cells by vibrational spectroscopy have been reviewed and discussed by Lu et al. (2011). The detection of thermo-resistant spores of Bacillus spp. has also been successfully carried out by SERS (Alexander and Le, 2007; He et al., 2013). Further works have reported the application of Raman and FT-IR fingerprinting to detect the contamination of food products with mycotoxins, such as aflatoxins in maize kernels (Lee et al., 2014) and deoxynivalenol in ground wheat and barley (Liu et al., 2009).
To avoid time-consuming culturing steps, different approaches have been described to isolate bacterial strains from food matrices previous to IR or Raman analysis. The most promising practice is the implementation of nanoparticles that are targeted against the bacterial species of interest, concentrated, and then submitted to spectral analysis. In this sense, magnetic nanoparticles have been applied, functionalized with anti-E. coli O157:H7 or anti-Salmonella typhimurium antibodies (Ravindranath et al., 2009). The pathogens could be detected using a portable mid-IR spectrometer in complex food matrixes with a detection limit of 10(4)—10(5) CFU/mL. In further studies, bacterial cells have been captured by using antibody—antigen interaction with nanoparticles immobilized on SERS active surfaces and detected by direct analysis with spectral fingerprinting (Chae et al., 2013).
Besides bacterial species identification and detection of single pathogens, the monitoring of microbial spoilage of food products by FT-IR spectroscopy has been described for fish (Tito et al., 2012), meat (Ellis et al., 2002), and milk (Nicolaou and Goodacre, 2008), allowing the quantification of the microbial load by direct analysis on the food surface without any culturing or isolation processes.
