Distribution, Diversity and Fate of Salmonella in Natural Biofilms
MetadataShow full metadata
Salmonella enterica strains represent important enteric pathogens that are typically transmitted to humans via food and drinking water contaminated with feces of vertebrate animals. The intestinal tract of vertebrates is typically presumed to be the native habitat of salmonellae, however, recent studies frequently detected Salmonella strains in water, sediments, animals (i.e., fish, turtles) and biofilms even in supposedly clean habitats such as Spring Lake, the spring-fed headwaters of the San Marcos River, Texas. We therefore proposed to monitor these potential human pathogens as they persist in or move through such ecosystems using a combination of traditional enrichment culture techniques in combination with molecular detection and identification tools that allow for highly sensitive, high-resolution analyses of salmonellae. An initial study [Chapter 2, published in Systematic and Applied Microbiology 34, 353-359 (2011)] assessed the diversity and distribution of salmonellae in freshwater biofilms at a fine scale (i.e. in 20 locations from a 324 cm2 area) for two sites in San Marcos, TX, a concrete storm water overflow channel (City Park) and a concrete surface in the spring-fed headwaters of the San Marcos River (Spring Lake) between April and September 2009. The study demonstrated the presence of salmonellae in natural biofilms and a significant micro-heterogeneity with differences in diversity and persistence of salmonellae during the season. The composition of Salmonella strains in the area analyzed changed in time with large differences between early (April, June) and late sampling times (September) within and among sites, except for one strain (S12) that was present at almost all sampling times at both sites, though often at different locations within the area analyzed. Follow-up studies [Chapter 3, published in Microbial Ecology DOI: 10.1007/s00248-012-0106-y (2013)] identified 4 selected strains as serovars Give, Thompson, Newport and -:z10:z39, and confirmed their pathogenicity in feeding studies with the nematode Caenorhabditis elegans demonstrating that pathogenic salmonellae were isolated from heterogeneous aquatic biofilms. Cells of these isolates inoculated into water or biofilms declined numerically within 2 days, reaching the detection limit of our qPCR-based quantification technique (i.e. 103 cells ml-1); however, cells persisted and stayed viable in biofilms in high numbers for some time. The fourth chapter [ accepted by FEMS Microbiology Ecology] focused on the analyses of the diversity of Salmonella in biofilm and water samples from the spring and slough arms of Spring Lake during the drought of 2011, with only one potential run-off event at the beginning of the study. Salmonellae were detected in semi-selective enrichment cultures by end-point PCR during the entire sampling period (11 sampling events during 2 months). From the spring arm site, 73% of the biofilms and 41% of the water samples were positive for salmonellae, while only 9% of the biofilms and 23% of the water samples were positive from the slough arm site. Salmonellae could be isolated from all positive samples, with higher diversity in biofilms compared to water samples, and more strains obtained from the spring arm than from the slough arm. Differences between sites were generally caused by less frequently detected isolates, while the majority of isolates that were present in both biofilms and water from both sites was represented by three strains only. Quantification attempts by qPCR directly in samples without prior enrichment did not result in a reliable detection of salmonellae, suggesting that numbers in all samples were below the detection limit. One of the strains isolated from biofilms was used to assess the potential of fish to transfer salmonellae from heterogeneous aquatic biofilms into feces using controlled aquarium studies with suckermouth catfish (Hypostomus plecostomus) and biofilms on tiles inoculated with salmonellae [Chapter 5]. Neither the presence of fish nor inoculation with salmonellae had detectable effects on the abundance of the microbial community, i.e. all DAPI-stained cells. Numbers of salmonellae quantified by qPCR and by in situ hybridization in water and biofilms, however, decreased fast from an initial value representing about 20% of the DAPI-stained cells to less than 0.01% within 3 days indicating that salmonellae are not persisting in high numbers in these environments, but probably present in low numbers. The results presented in this thesis indicate long-term persistence of Salmonella at considerable diversity, albeit in low numbers, in both water and heterogeneous aquatic biofilms, even in the absence of concurrent runoff that could be expected to contribute to contamination.