Non Technical Summary
Antimicrobial resistance (AMR) is a threat to human and animal health, and the food supply is a major route of AMR transmission from animals to humans. Despite its importance, it is impossible to effectively manage AMR genes (resistome) because appropriate tools to measure resistomes at broad scale do not exist. To this date methods available for tracking AMR genes across the food supply chain are labor-intensive or prohibitively expensive. Our team at Penn State has developed a cost-effective resistome profiling system based on a dual enzyme technology that had not been applied to investigate AMR. In essence, the method enriches AMR genes for sequencing in a targeted manner, which has the potential to revolutionize the surveillance and fight against AMR.In this integrated project we will first refine our method and validate it in collaboration with an external laboratory. Next, we will apply the method to answer an industry relevant question: How does feeding natural antimicrobials impacts the resistome in poultry? Since 2017, the use of medically important antibiotics for growth promotion has been prohibited, giving rise to alternative feed additives such as phytotherapeutics, probiotics and prebiotics for pathogen suppression and to maintain feed conversion and gut health. However, very little research has been done to understand the specific changes in the gut microbes caused by these products. In this project we will apply our refined method to answer this question and develop curricula and extension training on monitoring and mitigation of AMR in the food supply chain and offer technology transfer activities to facilitate the broad adoption of this novel method.?

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
711	7299	1040	60%
711	3220	1060	10%
712	3220	1060	10%
307	3220	1010	20%

Knowledge Area
711 - Ensure Food Products Free of Harmful Chemicals, Including Residues from Agricultural and Other Sources; 712 - Protect Food from Contamination by Pathogenic Microorganisms, Parasites, and Naturally Occurring Toxins; 307 - Animal Management Systems;

Subject Of Investigation
3220 - Meat-type chicken, live animal; 7299 - Research equipment and methods, general/other;

Field Of Science
1040 - Molecular biology; 1060 - Biology (whole systems); 1010 - Nutrition and metabolism;

Goals / Objectives
We propose to develop, validate, apply, and disseminate a comprehensive and cost-effective amplicon-based NGS AMRg surveillance method that our team recently created and successfully completed preliminary proof-of-concept experiments. The method, rhAMR (pronounced "rhammer"), is estimated to cost <$100/sample and will provide information on several 1,000s of genes of interest. rhAMR is scalable because it leverages a novel amplification technology that allows for massively parallel gene amplification. Once further developed and validated in this proposed project, the rhAMR method will be made publicly available to advance AMR surveillance research.We are proposing to achieve this goal through the following objectives:Obj. 1: [REFINE] Refine a recently developed cost-effective AMR profiling method leveraging rhAMPseq technology (rhAMR)Obj. 2: [VALIDATE] Validate rhAMR in collaboration with an external laboratoryObj. 3: [APPLY] Apply rhAMR to compare the effect of different feeding strategies on antimicrobial resistanceObj. 4: [DISSEMINATE] Develop curricula and extension training on monitoring and mitigation of AMR in the food supply chain and offertechnology transfer activities to facilitate the adoption of rhAMR by others

Project Methods
Objective 1: Develop a cost-effective AMR profiling method leveraging rhAMPseq technology1.a. In silico design and evaluation of a comprehensive AMR gene panel.Once a final comprehensive target set including genes from several databases has been curated, a FASTA file and BED files will be generated with specific gene locations of interest for amplificationIn silico validation will include primer sensitivity and specificity calculation based on primer binding affinity determined with BLAST against simulated metagenomes. NetPrimer, oli2go and other available primer analysis tools will be used to detect primer-dimers and secondary structures. Off-target binding will be checked against the RefSeq database. Results will be used to calculate precision and recall of the developed method1b. In vitro validation of rhAMR panel by using a mock microbiome.The developed panel will be validated in vitro using a mock microbiome comprised of bacteria that have been whole genome sequenced and are known to carry AMRg of public health relevance. Following in silico simulations and testing, bacterial strains will be grown each in their specific recommended media and conditions, and the mock microbiome will be assembled to test the designed panel in vitro. Bacteria enumeration will be performed in duplicate by serially diluting cultures and plating into agar plates followed by incubation in optimal conditions for each of the 30 microbial cultures and will inform our expected results. We will then extract DNA from the spike-in cultures and measure their concentration to determine genome-equivalents that were included in the community.The 35 mock microbiome versions (with 30 strains represented in different ratios) will undergo DNA extraction, rhAMR library preparation, sequencing, and data analysis will be performed.Analysis and interpretation: Data obtained from simulated metagenomes (as described in section 1.a.) will be used to calculate precision and recall of the developed method. Sensitivity, specificity, accuracy, agreement, PPV, and NPV of rhAMR-based AMR detection and how it compares to mock community ground truth will be calculated. The primer efficiency will be measured by calculating the regression coefficient for a linear regression model describing the relationship between the target gene read count and the dilution of the strain carrying the target gene.Expected results and outcomes: This objective will provide a novel, comprehensive and cost-effective AMRg profiling method for use in tracking and understanding AMRg dynamics. The experiments proposed here will provide data on the sensitivity, specificity, and LOD of the method for resistome characterization. The development of rhAMR will greatly improve our ability to perform comprehensive AMRg surveillance, detecting thousands of AMR genes in a cost-effective way.Objective 2: Validate rhAMR in collaboration with external laboratoryPanels developed and tested in Obj. 1 will be shipped to an external laboratory for further validation with biological samples and mock microbiomes (e.g., animal feces or bodily fluids).Method validation by external laboratory based on FDA Bioanalytical Method Validation GuidelinesA subset of 20-50 samples from which a diverse set of AMRg has been detected based on shotgun sequencing (considered "ground truth" and a reference for comparison with rhAMR). This subset of samples, and a subset of mock microbiomes produced in Obj. 1 will then be shipped to the Goodman lab, which will be blinded to the characteristics of each sample. The Goodman lab will then apply rhAMR to validate comparing against shotgun metagenomic data. Additionally, these samples will also be profiled with a targeted method - Illumina Ampliseq - which is currently the only other amplicon-based NGS AMRg profiling method available. Data will be analyzed by research personnel at Cornell following an established pipeline protocol developed at Penn State.Analysis and interpretation: Results obtained at the Goodman lab will be used to calculate sensitivity, specificity, accuracy, agreement, PPV, and NPV of rhAMR-based AMR detection and how it compares to target-based AMR detection (Ampliseq), and untargeted shotgun-based AMRg detection. Agreement between the three methods will be calculated based on Cohen's kappa coefficients , which quantify agreement beyond chance based on observed versus expected agreements and the McNemar's test for paired data. Limit of agreement plots (Bland-Altman plots) will be calculated to determine if there are systematic differences between two sets of observations.Expected results and outcomes: This objective will validate the method developed in Obj. 1 by experts in the field researching AMR from the food safety and animal health perspectives.Objective 3: Apply rhAMR to determine the effect of feed additives on the resistome.Characterize the resistome, microbiome, and performance in broilers fed various feed supplements.Sixteen replicates of each of 6 treatments randomly allocated at the pen level with a complete block design. A standard diet will be formulated according to published nutrient recommendations and will be referred to as the control. Commercially available feed supplements will be added to the control diet at their respective recommended levels. Feed and refuse weights will be recorded at the pen level to calculate total feed intake. Animals will be weighted at the end of the starter period and at the end of the grower period to calculate feed conversion ratio and coefficient of variation. Baseline diet will be fed and referred to as a negative control; a positive control diet will be formulated with bacitracin methylene disalicylate. Four other groups will assess the impact of a probiotic, an oregano-based, a saponin-based, and a capsaicin-based supplement.Samples will be collected throughout the production cycle from the environment and from a subset of slaughtered birds for assessment of microbiome and resistome profiles.Analysis and interpretation: The hypotheses that resistomes and microbiomes will differ among treatment groups will be tested by principal coordinate analysis, ANOSIM, PERMANOVA, adonis, and reference frames. Multiple hypothesis testing corrections will be performed using Bonferroni or Benjamini-Hochberg false discovery rate.Expected results and outcomes: Research activities performed in Obj. 3 will provide data quantifying the effects of different feeding strategies on broiler resistomes with the goal of providing critical data to aid in decreasing the load of AMRg in the poultry food value chain.Objective 4: Develop curricula and extension training focused on monitoring and mitigation of AMR in the food supply chain, and offer technology transfer activities to facilitate adoption of rhAMR methodology by other researchers4a. Develop an undergraduate-level course focused on AMR issues in the food supply chain4b. Develop videos on the topic of AMR geared toward the public to be offered through extension4c. Develop AMR detection methods data analysis workshop4d. Offer a poultry-focused AMR mitigation workshopExpected results and outcomes: By exposing undergraduate students to AMR related issues in the food supply we will provide the workforce with recent graduates aware of AMR and potential mitigation practices that can be applied in food production systems. Bioinformatics workshop participants will be familiar with the rhAMR-based and metagenomic-based data analyses and will be able to conduct their own data analyses using the provided workshop materials and program manuals. After completion of the antimicrobial resistance mitigation workshop, attendees will be more aware of the antimicrobial resistance problems and ways of mitigating AMRg transmission in the food value chain.