Investigation of the beef supply-chain microbiome and pathogen controls

Abstract

Foodborne illness associated with pathogenic bacteria is a global public health and economic challenge. Understanding the ecology of foodborne pathogens within the meat industry is critical to mitigating this challenge. The diversity of microorganisms (pathogenic and non-pathogenic) that exists within the food and meat industries complicates efforts to understand pathogen ecology. Further, little is known about the interaction of pathogens within the microbiome throughout the whole meat production chain. Here, the combined use of a metagenomics approach and shotgun sequencing technology was evaluated as a tool to detect pathogenic bacteria in different sectors of the beef production chain. Environmental samples were obtained at different longitudinal processing steps of the beef production chain: cattle entry to feedlot (Arrival), exit from feedlot, cattle transport trucks, abattoir holding pens, and the end of fabrication system (Market-Ready). The log counts population per million reads for all investigated pathogens (Salmonella enterica, Listeria monocytogenes, generic Escherichia coli, Staphylococcus aureus, Clostridium (C. botulinum, C. perfringens), and Campylobacter (C.jejuni, C.coli, C.fetus)) were reduced from Arrival to Market-Ready samples mainly due to reduced diversity within the microbiome. Further, normalized counts for Salmonella enterica, E. coli, and C. botulinum were greater in Market-Ready samples. This indicates that the proportion of these bacteria increases within the remaining bacterial community, which is likely a result of a reduction or elimination of other bacteria via antimicrobial interventions applied during meat processing. Further characterization of the microbiome allowed for the identification of 63 virulence factors within 27 samples (31% of samples). From an ecological perspective, data indicated that shotgun metagenomics can be used to evaluate not only the microbiome of samples collected from the beef production system, but also observe shifts in pathogen populations during the beef production chain over time. However, our utilization of this approach presented challenges and highlighted a need for further refinement of this methodology. Specifically, identifying the origin of reads assigned to specific pathogen from a diverse environmental sample containing thousands other bacterial species can be difficult. Additionally, low coverage on pathogen whole genome is another limitation of current next generation sequencing technology for shotgun metagenomic data. Moreover, the identification of bacteria from metagenomic data relies heavily on the quality of public genome database, which still need to be improved. Our investigation demonstrates that although the metagenomic approach has promise, further refinement is needed before it can be used to confirm the presence of pathogens in environmental samples. A study was conducted to compare decontamination efficacy of a blend of sulfuric acid and sodium sulfate (SSS) or lactic acid (LA) against Salmonella on the surface of hot beef carcasses. A total of 60 pieces of beef briskets, obtained directly from unchilled beef carcasses, were cut into two sections (10 x 10 x 1 cm) and spot-inoculated with 200µl of inoculum, comprised of six-strain mixtures of Salmonella, and allowed 15 minutes for pathogenic attachment to reach a target level of approximately 5 to 6 log CFU/cm2. One brisket section (of the pair) remained untreated while the other section was treated with the compounds using a custom-built spray cabinet that sprays either SSS (21°C and 52°C) or LA (21°C and 52°C) at pressure of 15 psi for 5 seconds. Treated samples were transferred into Whirl-Pak filter bags and were held for 10 minutes, allowing pathogen bacterialcidal activity before sampling, plating, and counting. Unheated and heated SSS lowered (P < 0.05) means of the total bacterial counts on Tryptic Soy Agar (TSA) from 6.3 log CFU/cm2 to 4.6 and 4.3 log CFU/cm2, respectively. Likewise, unheated and heated LA reduced (P < 0.05) means of the total bacterial counts on TSA from 6.3 log CFU/cm2 to 4.7 and 4.4 log CFU/cm2, respectively. On Xylose lysine deoxycholate agar (XLD), initial counts of inoculated Salmonella (6.1 to 6.2 log CFU/cm2) were reduced (P < 0.05) by 2.0 to 4.2 log CFU/cm2 due to treatment with unheated SSS, by 2.3 to 3.9 log CFU/cm2 due to treatment with heated SSS, by (P < 0.05) 2.4 to 3.7 log CFU/cm2 and 3.8 log CFU/cm2 after treatment with unheated and heated LA, respectively. Overall, no (P > 0.05) chemical by temperature interaction effects on microbial reductions was detected when plated on either TSA or XLD agars. Heating chemical solutions lead to an additional 0.3 log CFU/cm2 reduction in total aerobic bacteria compared to unheated solutions. Less (0.3 log CFU/cm2) inoculated Salmonella were recovered on XLD agar from samples treated with LA compared to samples treated with SSS. However, such a small numeric unit change was likely not biologically important. These results indicated that both unheated and heated SSS and LA are effective interventions to reduce Salmonella inoculated onto hot beef carcass surface tissue.