Location: Meat Safety and Quality
2018 Annual Report
Objectives
Objective 1. Molecular characterization including whole genome sequencing and transcriptomic characterization of foodborne bacteria, including pathogens and commensals, exposed to various physiologically relevant conditions reflective of the production continuum.
Sub-objective 1.A: De novo, whole genome sequencing and metagenomic profiling of the microbial community present in bovine rectoanal mucosa (RAM) swab samples.
Sub-objective 1.B: Characterize the genomic, phenotypic and transcriptional differences present in clinically important STEC and Salmonella serotypes exposed to different physiological relevant conditions in order to identify virulence and regulatory control mechanisms.
Objective 2. Characterize the ecological niches and reservoirs to identify mechanisms of foodborne pathogen, especially biofilms, for their ability to colonize and persist leading to the development of intervention strategies.
Sub-objective 2.A: Molecular mechanisms of biofilm formation.
Sub-objective 2.B: Association between biofilm formation, antibiotic resistance, and sanitizer tolerance.
Objective 3. Development and validation of various antimicrobial resistance detection methodologies including culture and genomic techniques, such as whole genome sequencing.
Sub-objective 3.A: Evaluation of culture based methods for the detection of bacteria resistant to antimicrobials important to human medicine.
Sub-objective 3.B: Development of genomic methods for the detection of antimicrobial resistance elements.
Approach
The cost of food borne illness and the loss of productivity in the United States is reported to be greater than $14 billion a year. While research efforts have resulted in great strides in tracking contamination entry points and identifying mitigation strategies, outlier events continue to occur and complete prevention of foodborne pathogens entering the food chain remains an elusive goal. Attaining this goal is challenging in part because many of the target pathogens live in dynamic and complicated communities, likely not even causing disease in their host reservoir. In addition, a better understanding of the use of antimicrobial agents in animal production and the possible impact on foodborne pathogens acquiring resistance has become a top priority for many government agencies and health care advocates. The project described here will provide new information about these issues by helping to better understand the different colonization sites and how various pathogens survive and interact with their respective bacterial communities. Further, we will characterize population differences within these foodborne bacteria, focusing on those that enhance an organism’s ability to cause human illness. Ultimately, the overall aim of this project is to provide new information about pathogen (predominantly Shiga toxin-containing Escherichia coli (STEC) and Salmonella enterica) persistence and survival in a variety of environments that position them for entry into the food supply.
Progress Report
Under Objective 1, we continue to make progress in generating complete closed reference genome sequences of multidrug resistant bacteria isolated from cattle fecal swab samples. An additional 20 genome sequences were generated in FY2018 for multidrug resistant (MDR) Enterobacteriaceae isolated from cattle fecal swabs. These sequences will be uploaded to the National Center for Biotechnology Information (NCBI) open access DNA sequence database, GenBank, and will ultimately facilitate data analysis of metagenomic sequencing projects aimed at understanding the dynamics of cattle fecal shedding of antimicrobial resistant bacteria, or super-shedders of Shiga-toxigenic Escherichia coli (STEC) or Salmonella. Mock recto-anal mucosa communities will be constructed from a subset of the 80 isolates sequenced to date, and total genomic DNA will be isolated and sequenced from these communities using the MinION sequencing platform and data analysis pipeline.
Also, we sequenced the complete genome from 40 STEC O26:H11 carrying the Shiga toxin 2 (stx2) gene and 22 STEC O111:H8 strains. STEC O26 strains containing stx2 were initially identified in Europe in the 1990s, but have become increasingly reported in the United States and other countries. O26 strains carrying stx2 cause more severe disease in humans and are associated with an increased risk of hemolytic uremic syndrome, in comparison to strains that only contain stx1. While there are STEC O26:H11 stx2 containing strains that are found in different countries, our sequencing identified a group of strains that appear to have arisen in the United States and are not found elsewhere. These U.S. strains differ from the other STEC O26:H11 stx2 strain in several ways. One difference is that these strains have a different bacteriophage (an important mobile element that can integrate into a bacteria’s chromosomes and often carries virulence factors like Shiga toxins) that carries the stx2 gene and is inserted into a previously undefined location on the chromosome. Also, the size and gene content of a mobile virulence plasmid is different than other STEC O26:H11 stx2 strains. Further characterization of the U.S. strain to determine the significance of these genomic changes and their ability to cause disease still needs to be explored.
Genome comparison of the STEC O111:H8 strains showed that the stx2 gene was located in 30% of the strains and was evenly distributed among the different phylogenic groups except for three groups, where the strains from two groups only had the stx1 gene while the strains from the third group only had the stx2 gene. Interestingly, the third group with only the stx2 gene were all human clinical strains. Another strain contained 3 stx genes, two for stx1 and one for stx2, which is very rare as strains usually have either no, one or two stx genes. Very few STEC O111 strains had chromosomal rearrangements unlike STEC O157:H7 and O26:H11 and almost always the rearrangement was at the terminus of replication. Bacteriophage are important mobile elements that can integrate into a bacteria’s chromosomes and often carry virulence factors. The phages in STEC O111:H8 typically integrate at the same location on the chromosome, but there were also phage that integrated in previously undefined locations. Nineteen of the twenty-two strains have a plasmid that contained the virulence factors, enterohaemolysin and the serine protease espP. A version of this virulence factor containing plasmid is also found in other STEC including STEC O157:H7 and STEC O26:H11. However, in STEC O111 this plasmid is smaller in size partially due to a deletion of part of the virulence factor catalase/peroxidase and the absence of another virulence factor, toxin B.
To better understand what effect genome variability has on Salmonella enterica serotypes Montevideo, Anatum and Typhimurium, Biolog phenotyping microarrays were used to determine growth characteristic with different carbon, sulfur, phosphorous and nitrogen sources along with different pH and osmolality. There was a strong correlation between metabolic rates in these three serotypes that associated with a strain being from an environmental or clinical source.
Genome sequencing and comparative genomics of 13 Salmonella enterica serovar Montevideo (one of the CDC’s top 20 serovars attributed to human illness and also frequently isolated from healthy cattle and beef) along with 72 publicly available Montevideo sequences, showed strains of this serotype fall into four distinct clades. While strains from all four clades have been isolated from humans, cattle isolates were all found to group in clade I, while the majority of strains associated with human illness were members of clades II, III and IV, with the majority residing in clade IV. Results further showed distinct differences in gene content among members of the four clades, especially with regard to prophage distribution, secreted effector and virulence factors, metabolic island and fimbrial operon content. Taken together, these differences suggest that the success of clade IV strains as human pathogens may be attributed to a combination of fitness advantages for colonizing and persisting within food types that are not normally cooked, and virulence traits that aid in modulating the host immune system and avoiding clearance. Identification of these traits may facilitate the development of rapid screening tools specific for members of this clade, which could have important implications for improving public health.
Under Objective 2, we continue to make progress towards understanding the molecular mechanisms of biofilm formation and sanitizer resistance by common foodborne pathogens. We have serotyped and tested a wide collection of Salmonella enterica strains isolated from beef trim contamination for their biofilm forming ability on food contact surfaces of materials commonly used in the meat industry, post-sanitization survival and recovery growth ability, as well as their antibiotic resistance profiles. Our results showed that the majority of these Salmonella strains had strong biofilm forming ability, which was positively correlated to their high tolerance against sanitization, suggesting biofilm formation might contribute to meat product contamination by Salmonella enterica at processing plants. In addition, certain Salmonella serotypes were more commonly isolated from contaminated trim than others, and antibiotic resistance phenotype was not necessary associated with biofilm forming ability and sanitizer tolerance. More importantly, the effect of residual amount of common sanitizers on bacterial biofilm formation and sanitizer susceptibility was evaluated. Our results showed that even though enhanced biofilm formation by certain Salmonella strains could be observed after multiple passages of bacterial growth in medium without sanitizers, this developed phenotype was not stable. Conversely, strains of certain Salmonella serotypes with stable phenotype of enhanced biofilm forming ability were obtained after repeated exposure to low concentrations of common sanitizers, likely due to genomic alternation after such exposure. However, such developed higher biofilm forming ability did not affect strains’ antibiotic resistant phenotype. Currently, genomic analysis and comparison between the parental and daughter strains is ongoing to identify any genomic changes that could be responsible for the altered/stable phenotype of stronger biofilm formation. Furthermore, we have been continuously working with the meat industry to test and compare sanitizer effectiveness, when applied with different methods (foaming, fogging, and immersion, etc.), to remove biofilms from contact surfaces for ease of implementation by the commercial plants.
Under Objective 3, to address concerns regarding impact of antimicrobial use on the occurrence of antimicrobial resistant bacteria and antimicrobial resistance genes, we sampled cattle manure applied soils, soils with commercial fertilizer applied, and a no application control to compare the levels of antimicrobial resistant E. coli, Enterococcus, Salmonella. Also, we screened for 10 antimicrobial resistance genes important in human medicine. A total of 240 samples were collected over 10-months starting in early spring and finishing in late fall. Antimicrobial resistant E. coli, Enterococcus, and Salmonella levels were not impacted by manure or commercial fertilizer application. Levels of 2 antimicrobial resistance genes were slightly higher immediately following land application of cattle manure in fall, but returned to normal levels by the next sampling date, prior to planting in the spring. Progress has also been made in generating complete closed reference genome sequences of multidrug resistant bacteria isolated from cattle fecal swab samples. The genomic elements contributing to antimicrobial resistance in these strains have been identified using CARD (the Comprehensive Antibiotic Resistance Database) and the resulting genotypes and phenotypes will be uploaded to NCBI’s Pathogen Detection Isolate Browser, in order to enhance this database and increase our understanding of bacterial resistance genotypes and corresponding phenotypes.
Accomplishments
1. Salmonella Montevideo markers for identifying strains frequently involved in outbreaks. Salmonella Montevideo is on the CDC’s list of top 10 Salmonella causing foodborne illness in the U.S. It is also a type of Salmonella that is frequently associated with cattle and can be found in beef products. However, S. Montevideo outbreaks are more often attributed to contaminated nuts, seeds or spices. To determine if there was a genetic basis for this distinction, ARS researchers in Clay Center, Nebraska, sequenced 13 strains of S. Montevideo from live cattle, fresh beef and humans, and then performed a comparative analysis of these complete sequences, along with 72 partial Montevideo sequences. The results showed that S. Montevideo strains fall into four distinct groups, and that while human isolates could be found in all four groups, cattle strains were restricted to only one group. Additionally, the strains associated with cattle were found to lack a number of genes needed for creating human illness. Overall, these results suggest that S. Montevideo strains associated with foodborne illness outbreaks are primarily attributed to food types that are not normally cooked, and specific traits that make them better able to cause disease in these food types. Consequently, the markers now identified may be useful to screen for strains of S. Montevideo that are more commonly associated with foodborne illness outbreaks, resulting in a safer food supply for consumers.
Review Publications
Webb, H.E., Harhay, D.M., Brashers, M.M., Nightengale, K.K., Arthur, T.M., Bosilevac, J.M., Kalchayanand, N., Schmidt, J.W., Wang, R., Granier, S.A., Brown, T.R., Edrington, T.S., Shackelford, S.D., Wheeler, T.L., Loneragan, G.H. 2017. Salmonella in peripheral lymph nodes of healthy cattle at slaughter. Frontiers in Microbiology. 8:2214. https://doi.org/10.3389/fmicb.2017.02214.
Nguyen, S.V., Harhay, D.M., Bono, J.L., Smith, T.P.L., Fields, P.I., Dinsmore, B.A., Santovina, M., Wang, R., Bosilevac, J.M., Harhay, G.P. 2018. Comparative genomics of Salmonella enterica serovar Montevideo reveals lineage-specific gene differences that may influence ecological niche association. Microbial Genomics. 4:1-17. https://doi.org/10.1099/mgen.0.000202.
