FUTURE TRENDS
The spectral fingerprinting techniques described here all represent accurate and fast methods in comparison to other bacterial identification techniques. However, the challenge of food security and foodborne pathogen detection is that very low doses of a critical strain may cause serious infectious diseases after multiplying rapidly in a few hours or days. Unfortunately, recent identification methods also require multiplication of bacterial cells to reach a detectable concentration. In addition, one disadvantage of most fingerprinting techniques is that isolated bacterial strains are required. That means that culturing steps are still necessary to isolate and concentrate the bacterial strains from the food matrices. In this sense, approaches that apply spectral fingerprinting directly to a sample without or with minimal sample pretreatment are challenging. In clinical routine analysis, direct bacterial identification by MALDI-TOF MS has been achieved for urine and blood samples (Clark et al., 2013). Similar approaches could be applied to liquid food samples, as already demonstrated for contaminated water or the rinse water after washing lettuce and cotton cloth (Holland et al., 2000). Nevertheless, a critical point of food matrices is the presence of a mixture of different bacteria and, until now, the application of MALDI-TOF MS fingerprinting for microbial mixtures has not yet been demonstrated. For that to occur, specific biomarker proteins have to be determined that allow an unequivocal identification of the corresponding strain on the basis of a few peaks. The identification of biomarker proteins is also of interest for bacterial typing and the correlation to virulence factors, toxin production, and/or antibiotic resistance that could significantly improve the risk assessment in the food chain.
Shortening the culturing process has also been the focus of several studies based on vibrational spectroscopy. A real-time detection and identification approach to food pathogens has been described and was based on SERS measurement of labeled immunoassay reagents in cultural enrichment vessels while culturing is ongoing. The approach allowed sensitive detection of the pathogens E. coli, Salmonella, and Listeria in complex food matrices (Weidemaier et al., 2015).
Promising future prospects are those applications where the high specificity of molecular fingerprints is combined with the high selectivity of nanomaterials, such as aptamers, nanoparticles, or further recognition molecules. The use of magnetic nanoparticles allows the separation of target strains from the food matrix and nontarget strains and the posterior analysis by spectral fingerprinting.
Future trends will probably focus strongly on the analysis of a food product without any culturing step. An interesting study has been carried out aimed at the direct detection and quantification of the microbial load in meat and milk, using MALDI-TOF analysis (Nicolaou et al., 2012). Such rapid screening tests are essential for an effective risk assessment in the food control sector and are required for the fast and accurate detection of food contaminants, such as bacterial pathogens, as well as mycotoxins, drug residues, heavy metals, etc. To realize onsite analysis and field tests, portable and automated instruments are the objectives of ongoing research. A portable and automated optofluidic SERS system has been successfully applied to detect food and water contaminants (Yazdi and White, 2012).
Finally, the spectral fingerprinting techniques described can be combined with further analytical techniques, since this can give much more complete information as a single technique, especially in the field of bacterial strain characterization. As an example, vibrational spectroscopy is commonly implemented into microscope systems to recover cells after enrichment on agar and following spectral analysis. More interestingly, the combination of LIBS, vibrational spectroscopy, and mass spectrometry has great potential to unite all the information and create a “whole-organism spectral fingerprint.”
