Quantitative detection of Salmonella enterica and the specific interaction with Lactuca sativa

Salmonella is among the most commonly known bacterial pathogens to cause human illness. Often Salmonellosis is associated with the consumption of contaminated foods like meat, eggs or egg products. However, during the last decades an increase of outbreaks is recognized to be caused by human pathogenic bacteria in association with fresh produce. The use of manure for production of vegetables, e.g. lettuce, contributes significantly to the risk of contamination of fresh produce. Enteric pathogens like Salmonella associated with manure can come in close contact with plants like lettuce, and a better understanding of the interaction between lettuce and Salmonella serovars during cultivation is necessary to be able to take preventive actions to reduce the risk for human health.This thesis describes the development of detection methods of Salmonellaenterica and Escherichia coli O157:H7 for routine diagnostic screening in the food production chain. Next to that, it describes the physiological and molecular interaction between Salmonella serovars and lettuce. Background information concerning Salmonella serovars in association with lettuce, like history of produce-associated outbreaks, approaches to detect the pathogen in food samples, pathogenesis, plant responses and the molecular interaction between plants and human pathogens, is discussed in Chapter one.Chapter two concerns the comparison of different molecular methods to detect S. enterica ( invA -gene) or E.coli O157:H7 ( stx -1, stx- 2 and eae -gene) with respect to sensitivity, precision and accuracy. Two basic methods were selected, both based on real-time Taqman PCR, a method that generates fluorescence upon specific DNA amplification. The increase in fluorescence during PCR is directly correlated to the amount of target DNA present after each amplification cycle. The detection and quantification methods were improved by the addition of a general internal amplification control (IAC), viz. comprised of DNA coding forgreenfluorescent protein ( gfp ), that allowed the identification of false negative results. The IAC provided insight in amplification efficiency and enabled a more accurate quantification. Implementation of the IAC did not affect the precision of the methods, although the sensitivity was reduced 10-fold. At least 1 pg of target DNA (equal to 200 CFU) was detected and quantified with high precision and accuracy. Qualitative detection was feasible even down to 10fgof target DNA (equal to 2 CFU) per reaction using both methods in which the IAC was incorporated. The methods enable a reduction in assay time to two days to test food samples, compared to five days required for the standardized procedures.To improve molecular methods to detect the pathogen in environmental substrates, five commercially available DNA extraction methods were evaluated in Chapter three with respect to DNA extraction efficiency of S. Enteriditis from soil, manure and compost. An internal procedural control (GIPC) for DNA extraction and amplification was developed. The GIPC was based on the same control DNA as used for the IAC in Chapter 1, incorporating gfp containing bacterial cells ( E. coli harboring a gfp-plasmid) in the sample prior to DNA extraction. Inclusion of the GIPC permitted a more accurate quantification of S. Enteriditis after DNA extraction and amplification and reduced the possibility of false-negatives. Using this protocol, the optimal extraction method differed forsoil (Mobio soil DNA extraction kit), manure (Bio101 soil DNA extraction kit) and compost (Mobio fecal DNA extraction kit).With each method, at least 2000 CFU of added S. Enteriditis/100 mg substrate could be detected bydirect DNA extraction and subsequent S. enterica specific Taqman PCR. After bacterial enrichment, as little as 1 CFU/100 mg of original substrate was detected. Using this approach a more reliable quantification was obtained for S. enterica initially present in environmental substrates.In Chapter four the physiological and molecular interactions between the human pathogenicS. Dublinand the commercially available mini Roman lettucecvTamburo are described. Investigation of the localization of S.Dublinon/in lettuce plants revealed the presence of significant populations on the surface and inside the plants. The latter was evidenced from significant residual concentrations after highly efficient surface disinfection (99.81%) and fluorescence microscopy ofS.Dublinin cross-sections of lettuce at the root-shoot transition region. A reduction in biomass was observed upon colonization of lettuce plants withS.Dublincompared to water-inoculated plants. Next to this physiological response, there were clear differential gene expression profiles between non-colonized and colonized lettuce plants based on transcriptome analysis by cDNA-AFLP. To confirm the results, generally and differentially expressed genes were selected, identified by sequence analysis and analyzed by RT-PCR to present the specific gene expression profiles in time. Functional grouping of the expressed genes indicated a correlation between colonization of the plants and an increase in expressed pathogenicity-related genes. From these results it was evident that lettuce plants respond to the presence ofS.Dublinat a physiological and molecular level. In addition, it was confirmed thatSalmonella serovars can colonize the interior of lettuce plants, thus potentially imposing a human health risk when contaminated lettuce is processed and consumed.The fact that the lettuce plants responded to the colonization by Salmonella serovars suggested that differences in susceptibility between cultivars or differences in colonization efficiency between Salmonella serovars might be present. In Chapter five, the differential interaction of S. Typhimurium, S. Enteritidis, S . Dublin, S. Newport and S. Montevideo with lettuce cultivars Cancan, Nelly and Tamburo is presented, in terms of prevalence and degree of endophytic colonization of lettuce by the Salmonella serovars. Besides a significant interaction, significant differences among serovars, but not among lettuce cultivars, were obtained when lettuce was grown under axenic conditions. When grown on soil, all three evaluated serovars S. Typhimurium, S. Enteritidis andS . Dublinwere able to colonize lettuce epiphytically, but to a lower extent than on axenically grown plants. OnlyS. Dublinwas able to colonize the plants endophytically when these were grown on contaminated soil. Species richness and diversity of the endophytic microbial community, determined from DGGE gels with DNA from Salmonella -colonized lettuce Cancan and Nelly, were negatively correlated with the number of Salmonella CFU / gram of lettuce. No correlation was observed for cultivar Tamburo. Thus, the microflora of lettuce cultivars Cancan and Nelly appeared more antagonistic to Salmonella serovars than that of cultivar Tamburo.Besides plant-associated colonization, also the active movement of Salmonella serovars towards lettuce roots was assessed. Movement was visualized using a metabolism marker (tetrazolium) for chemotaxis. Reduction of this marker suggested the presence of an organic compound in the lettuce root exudates that was used as carbon source by the Salmonella serovars. Subsequent micro-array analyses with DNA extracted from a broth culture of Salmonella with or without exudates identified genes of S. Typhimurium that were induced by root exudates. These genes, trehalose-6-phosphate synthase ( Ots A; utilizes glucose-6-phosphate as substrate), hexose phosphate utilization protein ( Uhp C; sensor for external glucose-6-phosphate), putative effector protein ( Ssa H; regulator of secretion of the type III secretion system),and putative anti-silencer RNA( Drs A; regulator of transcription to express rcs A promoter, responsible for capsular polysaccharide synthesis), imply a relation with a sugar-like carbon source and thus suggest an association with chemotaxis. The results described in Chapter 5 reveal different plant and microbial factors that influence the colonization efficiency of Salmonella serovars. The serovar and cultivar, but indirectly also the rhizosphere and the endophytic microflora of lettuce were most influential with respect to the risk of colonization and thus the risk for human health.Finally, an extensive discussion concerning the research of Chapters two to five is described in Chapter six, including future perspectives of risk for human health, route of infection and risk reduction in the production chain of Salmonella -associated lettuce.