Item No. 1 of 1

INVESTIGATOR: Greene, A.

PERFORMING INSTITUTION:
CLEMSON UNIVERSITY
CLEMSON, SOUTH CAROLINA 29634

UNINTENTIONAL CREATION OF A HEAT RESISTANT BACTERIAL BIOTA AND ITS IMPACT ON FOOD SAFETY

NON-TECHNICAL SUMMARY: Rendering processes waste animal tissues into animal feed ingredients that are recycled to poultry, swine, cattle and pets. With more than 8 billion chickens, 100 million hogs and 35 million cattle currently slaughtered annually, the rendering industry provides a significant benefit to society by processing billions of pounds of perishable waste tissues into stable value-added products while protecting the environment. The majority of rendered products are utilized as animal feed ingredients. In previous studies in this laboratory, a very low population of a heat resistant biota has been isolated from rendered animal products. The question is if recycling of materials has been a selective process creating a mobile biota of thermophilic microorganisms. It is important to learn more about this biota for cautionary monitoring and/or predictive actions of this biota to ensure human and animal health and safety. This proposed study will confirm the identify of the adapted thermophiles through genome sequencing, determine if these organisms could be pathogenic to animals and/or humans, and determine if horizontal gene transfer of virulence factors, such as antibiotic resistance, has occurred.

OBJECTIVES: Major Goals: Measure and assess the structure and complexity of the biodiversity present in rendered animal products that are used for animal feed and pet food ingredients.Objective 1: Sequence the genomes of resistant strains of Salmonella isolated from rendered animal products to define and interpret the level of dynamic gene-transfer between living heat resistant populations of bacteria after rendering and known reference pathogenic strains of bacteria.Objective 2: Identify and characterize the functional elements of post rendering biota, which may encode enzymes and other bacterial by-products with industrial uses.

APPROACH: Heat resistant bacterial isolates were previously collected from rendered animal products by this laboratory. These isolates, which tested as positive for Salmonella by FDA BAM procedures and 16S rRNA sequencing, will be further examined in this study.Nucleic acid extraction and sequencingEach of the samples collected post-rendering will be subject to a dual DNA extraction method that targets both intact genomic DNA and fragmented DNA (typically found in rendered products) using a silica-guanidine thiocyanate-based DNA extraction which has been validated for regulatory use by the U.S.Food and Drug Administration (Myers et al., 2001) and a kit based approach using the Rendered meat purification system from Invitrogen (Yancy, 2009) which has been demonstrated to be comprehensive in targeting a diverse population of organisms(Myers et al., 2010). Cultured samples will undergo the same procedures and purified DNA will be used as substrate for Illumina sequencing library preparations.Genomic DNA from each of the above extractions will be assessed for quality and integrity using the Qbit DNA fluorimenter (Life Technologies) and Bioanalyzer 2100 (Agilent). Dual samples will be normalized and combined for library preparation. Each site sample and time point will be individually indexed and sequencing adapters added with the Nextera DNA sample preparation kit (Illumina) following the manufacturer's recommended procedure. Indexed samples will be combined in equi-molar ratios as determined by quantitative PCR (qPCR) for sequencing. A total of 36 samples of post-rendered material and 36 samples of cultured material will be deeply sequenced across 4 lanes of HiSeq2500 (~5 million read pairs per samples) using the 2x101bp PE run type.Read preprocessing and analysisPost sequencing, raw trace data (.fq) files will subject to read preprocessing to remove adapter sequences and low quality bases with the Trimmomatic software (Lohse et al., 2012). The rendered metagenome will be input into a specified metagenome assembler tailored for discrimination of highly conserved sequences shared between different species such as MetaVelvet, Meta-IDBA, or Ray Meta (Boisvert et al, 2012; Namiki et al., 2012; Peng et al., 2011). Assembled sequence reads will be assessed for function and taxonomic placement with the MEGAN4 software (Huson et al., 2011). Homology with pathogenic bacteria will be performed through intersection with the National Microbial Pathogen Database Resource (NMPDR) (McNeil et al., 2007). Through these approaches we will enable the dissection of this biota to address any implications of potential selection in these animal feed products.The results of this study will be reported in a journal manuscript submitted for publication in a peer-reviewed journal often monitored by the target audience. The success of this exploratory project will measured depending on the results of the scientific study. If evidence that selection has created a resistant biota, this knowledge will allow future investigatations into means of destroying this biota. If evidence suggests no selection has occurred, this information will be a confirmation of the safety of the rendering industry.

PROGRESS: 2015/08 TO 2020/07
Target Audience:The target audience included personnel involved with animal feed production and safety. Changes/Problems:The thermally resistant isolates which had been collected in 2011 became unculturable after sequential passes in standard microbial media and conditions. All of the project objectives were completed by this time but additional desired information regarding which genes were turned could not be obtained since the isolates died. The final portion of the project to identify the bacteria which produced potentially beneficial lipase and protease enzymes could not be completed due to the COVID-19 pandemic shutting down our university laboratory. The isolates were frozen and will be examined upon return to the laboratory. What opportunities for training and professional development has the project provided?This project allowed research training for a post-doctoral student in microbiology techniques. This project also allowed professional development of the same post-doctoral student in project management. How have the results been disseminated to communities of interest?Results were presented to the communities of interest through in-person meetings on-campus. What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

IMPACT: 2015/08 TO 2020/07
What was accomplished under these goals? 1) Measure and assess the structure and complexity of the biodiversity present in rendered animal products that are used for animal feed and pet food ingredients. A total of 106 random bacterial isolates were collected directly from rendered animal feed ingredients. These isolates were collected by plating samples on Brain Heart Infusion agar followed by re-streaking randomly selected colonies on Tryptic Soy Agar. After Gram staining and examining morphology, obvious yeast and mold isolates were removed. A library of bacterial organisms was collected and stored at -80C until identification work could be conducted. Of these, 20% were identified as Staphylococcus epidermidis, 20% as Staphylococcus warneri, 16% as Bacillus lichenformis, 12% as Staphylococcus aureus, 6% as Escherichia coli O157:H7, 4% each of Bacillus subtilis, Staphylococcus xylosus and Escherichia coli, and 2% each of Klebsiella trevisanii, Staphylococcus simulans, Bacillus lentimorbus, Bacillus marinus, Bacillus spp., Bacillus pumilus and Staphylococcus cohnii. No Salmonella was identified in these samples. 2) Sequence the genomes of heat and radiation resistant strains of Salmonella isolated from rendered animal products to define and interpret the level of dynamic gene-transfer between living heat resistant populations of bacteria after rendering and known reference pathogenic strains of bacteria; In 2011, three bacterial isolates (A, B and C) were recovered from thermally processed feed ingredients. Because of preliminary laboratory thermal trials results indicating these isolates were heat-resistant, A, B and C were examined further through sequencing. PacBIO raw sequencing data for the isolates were downloaded and processed. The canu assembler (http://canu.readthedocs.org/en/latest/index.html) was used to produce three separate de novo assumblies for the three isolates. Polishing of each assembly was completed by aligning raw data for each sample to its generated assembly using pbalign (https://github.com/PacificBiosciences/pbalign) and then quiver (https://github.com/PacificBiosciences/GenomicConsensus). The statistics for the final 3 assemblies of the bacterial isolates indicated A had 4 scaffolds with the longest scaffold being 4,840,853 bp, B had 2 scaffolds with the longest scaffold being 4,965,959 pb and C had 2 scaffolds with the longest scaffold being 4,814,980 bp. Taxonomy identification for each of the three assemblies was conducted using four online methods: KmerFinder (https://cge.cbs.dtu.dk//services/KmerFinder/), Reads2Type (http://www.cbs.dtu.dk/~dhany/reads2type.html), SpeciesFinder (https://cge.cbs.dtu.dk//services/SpeciesFinder/), and ribosomal MultiLocus Sequence Typing (rMLST) (http://pubmlst.org/rmlst/). All results came back to the general result of Salmonella enterica. KmerFinder produced subspecies information for each sample as Salmonella enterica serovar Newport USMARC S3124 1 uid213895, Salmonella typhimurium DT104 uid223287 (enterica), and Salmonella enterica serovar Newport USMARC S3124 1 uid213895 for Isolates A, B and C, respectively. Total gene counts were 4,959,5073 and 4,730 for Isolates A, B and C, respectively, as compared to 4,553 in the reference S. typhimurium LT2. A, B and C isolates had 415, 157 and 418 genes missing as compared to the reference as shown by genes without a blastn hit to the respective genes. A, B and C isolates had 790, 603 and 579 genes inserted into the genome as compared to the S. typhimurium LT2 reference. These results indicated there were 31, 74 and 16 genes in A, B, and C, respectively, that were common to the reference Salmonella. Full gene profiling was conducted on each isolate and circos plots were created. SPI-3 Alignments were conducted to produce mummerplots. More than 300 genes were annotated as regulatory genes in Salmonella and of those, 14 regulators including SpvR, FruR, IHF, PhoP/PhoQ, SsrA/SsrB, SlyA, Hnr, RpoE, SmpB, CsrA, RpoS, CRP, OmpR/EnvZ, and Hfq which are required for virulence regulation. As is typical with Salmonella, virulence factors, such as adhesion, invasion, and toxin genes were shown to be clustered in certain areas of the chromosome known as "Salmonella pathogenicity islands." For plasmid virulence genes, SpvRABCD, Pef-BACDI and TraT were examined. B contained the spvB and spvD genes of the SpvRABCD series. Most virulence plasids are not self-transmissable but tra (transfer) genes allow plasmids to be transferred to additional strains by conjugation. A and B isolates contained multiple tra genes including TraT. These results indicated genes potential for thermal resistence and ability to transfer that resistance. Further work was to be conducted on the isolates to determine which genes are being expressed under thermal stress conditions. However, upon repeated sub-culture of the isolates in standard commercial media, the three isolates (A, B, and C) which had been collected in 2011 each eventually became non-viable. Returning to the samples of bacterial isolates stored in the freezer, again, the three isolates became non-viable upon three to six sub-cultures. This occurred on four repeated trials using all of the remaining frozen cultures we had of these isolates. All reference Salmonella isolates (also frozen under the same conditions) remained viable through multiple sub-cultures when grown in sub-samples of the same batches of prepared media and with incubation under the same conditions. This suggested a metabolic defect in the three isolates collected in 2011 or a failure to thrive in the media and conditions provided. To continue this study, additional thermally resistant isolates were sought from new samples of feed ingredients. However, no thermally resistant Salmonella were isolated from 42 independently collected thermally processed feed ingredient samples which were thermally treated in the laboratory under the same condistions as A, B and C isolates. Although the three isolates (A, B and C) isolated in 2011 had genetics indicating potential thermal resistance, these isolates apparently had a metabolic defect and/or inability to grow in standard Salmonella growth media. No similar isolates have been obtained during the past five years of this study. Therefore, with the inability to grow A, B and C isolates for further examination and no additional heat resistant Salmonella isolates found, this part of the study was ended. 3) Identify and characterize the functional elements of post rendering biota, which may encode enzymes and other bacterial by-products with industrial uses. The next phase of the study was to determine if there were any beneficial enzymes which could be isolated from organisms found in high lipid and protein content feed ingredients. Due to the recycling of feed ingredients through the intestinal tract of animals, it was hypothesized there could have been an unintentional yet beneficial creation of lipases and proteases in non-pathogenic bacterial strains. A total of 100 bacterial strains graded as having strong lipolytic and proteolytic activity as determined by visual analysis on tributyrin and casein plate agar were randomly collected. Of these, 74 were identified as Gram positive isolates (59 bacilli and 15 cocci). The remaining 26 Gram negative strains were tested using standard Salmonella growth and identification methods and determined to not be Salmonella. These 100 isolates were to be forwarded for 16S identification and further gene identification when the COVID-19 pandemic caused our laboratory to be closed. Therefore, this part of the study was ceased and will have to be completed after return to the laboratory is allowed.

PUBLICATIONS (not previously reported): 2015/08 TO 2020/07
No publications reported this period.