Abstract

In the 2007 report of World Health Organization (WHO) it was reported that in 2005 a great proportion of 1.8 million people died because of food and drinking water contamination (Velusamy et al., 2010). Fresh food product such as, fruits and vegetables carry a natural non-pathogenic epiphytic micro-flora, but during the food chain: harvest, transportation and further processing and handling the produce can be contaminated with pathogens from human or animal sources (Anon, 2002). While conventional methods used to evaluate pasteurization efficiency are based on cultivation in vitro, it has been ascertained that, under environmental stress conditions (e.g. nutrient limitation, pressure, temperature), a number of pathogens enter in a so-called Viable But Not Cultivable (VBNC) state, becoming eventually more resistant to stress and thus escaping to detection by cultivation methods. Improving health risk assessment associated with the increasing consumption of minimally processed fresh food products is a crucial need. To reach this objective, in the first part of my PhD project I set up and validated cultivation-independent bacterial viability assays, propidium monoazide quantitative PCR (PMA-qPCR) and flow cytometry (FCM), to monitor bacterial populations in food after Supercritical Carbon Dioxide (SC-CO2) treatment, that is one of the most promising non-thermal pasteurization technology in the age of the increasing demand for “ready-to-eat” and minimally-processed food products. The efficiency of SC-CO2 treatment was evaluated on bacterial liquid cultures, on bacteria spiked both on a synthetic solid substrate (LB agar) and on some fresh food products, including carrots, coconut and dry cured ham. The results indicated that the treatment is more efficient on bacteria spiked on LB agar, and that bacterial inactivation is accompanied by a reduction of their biovolume. Total bacterial inactivation on food products was reached for both Escherichia coli and Listeria monocytogenes, satisfying both the US and European requirements (CFSAN/FSIS, 2003; European Commission, 2005). Salmonella enterica was instead more resistant to treatment, suggesting future experiments consisting in the application of a combination between SC-CO2 and other techniques alternative to heat pasteurization, such as ultrasounds or Pulsed Electrical Field. FCM and PMA-qPCR data showed that a fraction of bacterial cells not detectable by plate counts maintained the integrity of their membrane (at least 102 cells/g for each bacterial species) suggested that the cells entered in a VBNC state. Comprehensively, the FCM assay showed the best performance as a bacterial viability test method, permitting to evaluate with high sensitivity the efficiency of treatment, to discriminate subpopulations of cells with different level of membrane permeabilization, and to identify variations in biovolume and alterations of the cellular surface. The method could be applied, with some adjustments, to any field where determining microbial viability status is of importance, including food, environment or in the clinic. Permeabilization of the cell membrane has been proposed to be the first event leading to cell inactivation or death after SC-CO2 treatment (Garcia-Gonzalez et al., 2007; Spilimbergo et al., 2009).The Permeabilization of membrane induced by SC-CO2 was also observed in Salmonella enterica (Kim et al., 2009a; Tamburini et al., 2013) and in Saccharomyces cerevisiae (Spilimbergo et al., 2010). Whether SC-CO2 has a direct effect on the bacterial membrane or permeabilization is a consequence of cell death remains an open question. In the second part of the Thesis to increase knowledge on the mechanism of bacterial inactivation mediated by SC-CO2 lipidomic profiles (HPLC-IT-ESI-MS), bacterial depolarization/permeabilization analysis (FCM) and gene expression studies of enzymes involved in phospholipids biosynthesis were performed on E. coli K12 MG1665. The data indicated that after 15 min of SC-CO2 treatment most of bacterial cells lost their membrane potential (95%) and membrane integrity (81% of permeabilized and 18% of partially-permeabilized cells). Bacterial permeabilization was associated to a 20% decrease of cellular biovolume and to a strong reduction (more than 50%) of all Phosphatidylglycerol (PG) membrane species, but without altering their average unsaturation index (1.30 ±0.02) and the average acyl chain on the glycerol backbone (33.30 ±0.03). The process acts more efficiently on PG than on PE (Phosphatidylethanolamine) head group phospholipids. Bacteria responded to treatment up-regulating the expression level of PssA gene, involved in PEs synthesis, since PssA activity is regulated by mole fraction of PGs and Cardiolin in the membrane. However still remains to understand why only PG species have been found to strongly decrease during the treatments. Further studies would be necessary, including phospholipid biosynthesis mutant analysis.