The Salmonella enterica serotype Hadar pangenome: Population structure and dynamics of a zoonotic pathogen

The Salmonella enterica serotype Hadar pangenome: Population structure and dynamics of a zoonotic pathogen | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article The Salmonella enterica serotype Hadar pangenome: Population structure and dynamics of a zoonotic pathogen Hattie Webb, Kaitlin Tagg, Arancha Peñil-Celis, G Stapleton, Zachary Ellison, and 23 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5896627/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 24 Jan, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract The bacterial accessory genome, comprised of plasmids, phages, and other mobile elements, underpins the adaptability of bacterial populations. Pangenome (core and accessory) analysis of pathogens can reveal epidemiological relatedness missed by using core-genome methods alone. Employing a k -mer-based Jaccard Index approach to compute pangenome relatedness, we explored the population structure and epidemiology of Salmonella enterica serotype Hadar (Hadar), an emerging zoonotic pathogen in the United States (U.S.) linked to both commercial and backyard poultry. Hadar populations underwent substantial shifts between 2019 and 2020 in the U.S., driven by the expansion of a lineage carrying a previously uncommon prophage-like element. Phylogenetic and pangenomic relatedness, coupled with epidemiological data, suggest this lineage emerged from extant populations circulating in commercial poultry, with subsequent dissemination into backyard poultry environments. We demonstrate the utility of pangenomic approaches for mapping vertical and horizontal diversity and informing complex dynamics of zoonotic bacterial pathogens. Biological sciences/Microbiology/Microbial genetics/Bacterial genetics Biological sciences/Computational biology and bioinformatics/Genome informatics Biological sciences/Genetics/Microbial genetics/Bacterial genetics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction The accessory genome, comprising plasmids, prophages, genomic islands, and other mobile genetic elements (MGE), is a key component of bacterial evolution [ 1 ]. While typically excluded from phylogenetic or source attribution analyses [ 2 , 3 ], there is growing interest in the discriminatory and predictive power of the accessory genome for epidemiological investigations [ 4 – 7 ]. For zoonotic pathogens like Salmonella enterica with numerous transmission routes [ 8 – 10 ], analysis of the pangenome (accessory and core genome) has proven useful for enhanced surveillance, outbreak investigation, and microevolutionary exploration [ 11 – 13 ]. The added public health value of pangenome data, however, depends on the unique genomic structure and microbial ecology of each Salmonella serotype and should be assessed within the context of serotype-specific population analyses. High-resolution pangenomic analyses, coupled with epidemiological and source information, are likely to be particularly informative for serotypes linked to multiple sources and transmission pathways or for clonal lineages that exhibit limited variability in their core genome [ 4 ], such as S. enterica serotype Hadar (herein referred to as Hadar). Hadar is transmitted to people via contaminated food and contact with animals and has caused several United States (U.S.) outbreaks in the last decade, linked to either ground turkey consumption or contact with backyard poultry (i.e., privately-owned, non-commercial poultry such as chickens, ducks, or turkeys) [ 14 , 15 ]. Although Hadar is considered a highly clonal serotype, exhibiting limited variability by core-genome multilocus sequence typing (cgMLST) [ 14 ], strains transmitted by these two different sources were historically differentiable (allele range 25–50). However, in 2020, despite decreased reporting of enteric illness during the early years of the coronavirus disease 2019 (COVID-19) pandemic, an emergent Hadar strain was linked with both ground turkey consumption and backyard poultry contact. These outbreaks resulted in > 900 human illnesses compared to < 500 total reported cases of Hadar in all years prior to 2020 [ 14 , 16 ]. Traceback investigations were not able to determine the epidemiological connection suggested by the detection of indistinguishable strains (determined by cgMLST) from two ostensibly distinct sources: commercial poultry and backyard poultry [ 14 , 15 ]. This emergent strain, now responsible for > 2000 human illnesses, continues to cause outbreaks into 2024; it has been designated by the U.S. Centers for Disease Control and Prevention (CDC) as a Reoccurring, Emerging, or Persisting (REP) strain REPTDK01, with a cgMLST range of 0–26 allele differences [ 17 ]. Given the limitations in discriminatory power of cgMLST for this strain, we employed k -mer-based Jaccard Index (JI) to compute pangenome relatedness [ 13 ] of Hadar along the U.S. farm-to-fork continuum (Farm-to-Fork Continuum). We explored and assessed the value of the pangenome for delineating strains, for attributing human cases to transmission vehicles, and for a general understanding of the epidemiological and evolutionary dynamics that underpin Hadar disease incidence and environmental persistence. In addition, we built a foundational landscape of the vertical and horizontal diversity and dynamics of this serotype and offer support for the incorporation of the accessory genome for differentiating strains transmitted via different pathways. Methods Data collection A total of 3384 U.S. Hadar genomes were included in this analysis (Supplementary Table S1 ), collected between 1990 and 2023 (August 30th ) from national surveillance systems and ad hoc sampling. Hadar genomes from ill humans with exposure information available were categorized as follows: “backyard poultry contact” – when contact was confirmed within seven days of illness onset (contact is defined as direct interaction with chickens, ducks, turkeys, geese, guinea fowl, or quail; direct contact with the environment where backyard poultry live and roam; consumption of eggs or meat obtained from backyard poultry; or residence with a household member who directly interacted with backyard poultry) [ 15 ], “turkey consumption” – where ground turkey was consumed within seven days of illness onset, and “unknown” – where exposure information was not available, or when neither backyard poultry contact nor turkey consumption was reported. Genomes from non-human sources were categorized according to the commodity from which they were sampled, for example, “commercial poultry” or “swine”. “Other” was used to categorize samples from unknown food, animal, or environmental source types. National surveillance systems Salmonellosis is a nationally notifiable disease in the United States, and isolates obtained from patients are routinely submitted to public health laboratories (PHLs) as part of the CDC’s national enteric disease surveillance network, PulseNet USA [ 18 ]. Since 2019, PHLs have performed whole genome sequencing (WGS) on all Salmonella isolates they receive and upload sequence data to a centralized national database for genetic analysis, including computed serotyping [ 18 , 19 ], and to the National Center for Biotechnology Information (NCBI) under the BioProject PRJNA230403. Additionally, public health departments routinely collect demographic information for all laboratory-confirmed cases of salmonellosis. For cases included in multistate outbreak investigations, public health officials conduct additional patient interviews, whenever possible, with Supplementary standardized questionnaires to obtain further details about foods eaten and animal contact before illness onset [ 14 ]. Approximately 5% of isolates detected by PHL also fall within the CDC arm of the National Antimicrobial Resistance Monitoring System (NARMS), a structured collection of enteric isolates from all 50 U.S. states used to monitor temporal trends in antimicrobial resistance (AR) ( https://www.cdc.gov/narms/index.html ). CDC NARMS has been routinely generating WGS data for this smaller subset of Salmonella isolates since 2016. WGS data for 2494 Hadar isolates collected between January 1st , 2016, and August 30th, 2023, were included in this analysis (Supplementary Table S1 ). For years prior to routine WGS (2005–2015), all Hadar isolates in PulseNet USA’s national database with WGS data available were included (n = 55); these represent a small proportion of total isolates collected from this time period that were sequenced for various special interest projects. The U.S. Food and Drug Administration (FDA) arm of NARMS routinely collects WGS data on Salmonella isolated from retail meats (chicken, ground turkey, ground beef, pork) purchased from U.S. grocery stores ( https://www.fda.gov/animal-veterinary/national-antimicrobial-resistance-monitoring-system/about-narms ). Sequencing data and source information are uploaded to the NCBI under the BioProject PRJNA292661. The following NCBI Pathogen Detection query (August 30th, 2023) identified 300 Hadar genomes (Supplementary Table S1 ) that were included in this analysis: https://www.ncbi.nlm.nih.gov/pathogens/isolates/#PRJNA292661%20AND%20Hadar . The U.S. Department of Agriculture’s Food Safety and Inspection Service (USDA-FSIS) routinely collects WGS data on Salmonella isolated from regulated food and animal products within U.S. food processing facilities ( https://www.fsis.usda.gov/science-data/sampling-program/sampling-results-fsis-regulated-products ). Sequencing data and source information are uploaded under NCBI BioProject PRJNA242847. Additionally, the USDA-FSIS arm of NARMS routinely collects WGS data from Salmonella isolated from the intestinal content of food animals at slaughter ( https://www.fsis.usda.gov/science-data/national-antimicrobial-resistance-monitoring-system-narms ) and data is uploaded under NCBI BioProject PRJNA292666. An August 30th, 2023 NCBI Pathogen Detection query identified 367 Hadar genomes from USDA-FSIS product sampling and 102 from NARMS sampling (Supplementary Table S1 ) for inclusion in this study: https://www.ncbi.nlm.nih.gov/pathogens/isolates/#Hadar%20AND%20collected_by:USDA-FSIS . Ad hoc sampling systems To expand source type representation along the farm-to-fork continuum, Hadar genomes isolated from North America were included from ad hoc sampling systems. The FDA’s Office of Regulatory Affairs (ORA), Center for Food Safety and Applied Nutrition (CFSAN), and Center for Veterinary Medicine (CVM) perform ad hoc WGS on human food and animal food (including imported) product sampling and upload sequencing data to the GenomeTrakr project at NCBI (BioProject PRJNA186035). Twenty genomes (Supplementary Table S1 ) collected between 2003 and 2022 were selected and included in this analysis. An additional nine isolates representing all sequenced Hadar collected from sick animals as part of FDA-CVM’s Veterinary Laboratory Investigation and Response Network (Vet-LIRN) AMR monitoring program were also included. USDA’s Animal and Plant Health Inspection Service (APHIS) provides ongoing animal disease surveillance and animal disease diagnostic services through the National Veterinary Services Laboratories (NVSL; https://www.aphis.usda.gov/labs/about-nvsl ) and the National Animal Health Laboratory Network (NAHLN; https://www.aphis.usda.gov/labs/nahln ). Thirty-two Hadar genomes (Supplementary Table S1 ) collected from chickens or turkeys from 2018 until 2023 as part of on-farm monitoring or for diagnostic purposes were included in this analysis. Three Hadar genomes previously sequenced and published by USDA’s Agricultural Research Service (ARS) [ 20 ], and two Hadar genomes collected from wild ducks by the National Wildlife Health Center were also included (Supplementary Table S1 ). Additional Hadar genomes were available on NCBI, but source information availability (through NCBI or personal communication) was a requirement for inclusion in this analysis. Non-U.S. genomes A dataset of global non-U.S. Hadar genomes was generated from EnteroBase [ 21 , 22 ] for comparative analysis against the pangenome of the U.S. collection. All genomes with predicted serotype “Hadar” (EnteroBase employs SISTR1 [ 23 ] and SeqSero2 [ 24 ]) isolated in any country other than the U.S. were downloaded (n = 1145) (accessed December 21st, 2023) (Supplementary Table S2 ). Genomic analysis Short reads with a base call quality score ≥ 28 and coverage ≥ 40x were assembled using shovill v.1.0.9 ( https://github.com/tseemann/shovill ) and resulting contigs with < 10% of the average genome coverage were excluded from the final assemblies. Serotype was confirmed using SeqSero 2.0 v1.2.1 [ 24 ], sequence type (ST) was determined using mlst ( https://github.com/tseemann/mlst ), core SNP (single nucleotide polymorphism) cluster was obtained from NCBI Pathogen Detection’s Isolate Browser ( https://www.ncbi.nlm.nih.gov/pathogens/ ), and allele code was calculated using cgMLST [ 18 ]. “Condensed allele code,” which collapses allele codes to the third digit (e.g., Salmonella spp. Allele codes SALM1.0–6771.1.1.30.1.21 and SALM1.0–6771.1.1.30.1.44 would be collapsed into SALM1.0–6771.1.1), was used to simplify representation of allele codes. Genomes of the same condensed allele code are expected to differ by less than ~ 15 allele loci (Williams in preparation ). Accessory (non-core) genome elements were detected using PanGraph (see Pangenome characterization ) [ 25 ] and characterized using PlasmidFinder [ 26 ] (90% identity, 60% gene coverage) for plasmid replicons, MOBscan [ 27 ] for conjugative relaxases, COPLA [ 28 ] for Plasmid Taxonomic Unit (PTU) designation [ 29 ], and Bakta [ 30 ] for gene annotation. AR determinants, including acquired genes and chromosomal mutations, were detected using staramr v.0.4.0 ( https://github.com/phac-nml/staramr?tab=readme-ov-file#mlsttsv ), which employs the ResFinder database (updated 30JUL2020; 90% identity, 50% gene coverage) and the Salmonella spp. PointFinder scheme [ 31 ]; predicted AR was determined by staramr according to ResFinder and PointFinder results. Assignment of draft Illumina contigs to plasmids or chromosomes was performed using MOB suite [ 32 ]. Long-read sequencing was performed on 35 selected isolates from each Jaccard Index (JI) group (see Jaccard Index calculation ), chosen strategically to maximize connectivity to other internal nodes and to best achieve JI-group representation. Eighteen Hadar isolates from people or food products were sequenced on the Oxford Nanopore GridION sequencing platform (Supplementary Table S3); reads were assembled using an in-house pipeline, as previously described [ 33 ]. Seventeen isolates collected from food or animal samples were sequenced using the 10-kb SMRTLink template preparation protocol (Pacific BioSciences, CA), as previously described [ 34 ]. Complete genomes were annotated by Bakta [ 30 ]. Long-read data are uploaded under BioSample numbers listed in Supplementary Table S3. An additional 18 previously published Hadar genomes [ 35 ] were also included in the analysis (Supplementary Table S3). Jaccard Index calculation The exact JI was used as a measure of similarity between all genome pairs as previously reported [ 13 ]. Briefly, each complete genome assembly was converted into a set of k -mers. JI was calculated as the ratio of shared k -mers over the total number of different k -mers between the two sets (including shared k -mers, SNP k -mers differing by a single base pair, and indel k -mers differing between the datasets and excluding duplicated k -mers). BinDash [ 34 ]⁠ was employed to calculate JI, using parameters minhashtype = -1 (to compute the exact JI between highly similar genomes using the complete set of k -mers, rather than an estimated JI based on a subset of k -mers) and k -mer length (k) = 21 (as previously defined as optimum in [ 13 ]). Network visualization and community detection The adjacency matrix of pairwise genome similarities generated by BinDash was used to construct an undirected network. Gephi v10 [ 36 ] was employed to visualize the network, using the ForceAtlas2 algorithm for the layout. To define the final components for study, referred to as JI-groups, a range of JI thresholds was assessed, and network sparsification was optimized according to transitivity and density, as previously described [ 13 ]. Transitivity plateaued between JI 0.986 and 0.990. The final JI threshold was set in the middle of this range, at 0.988, balancing the density of communities (defined by a minimum of five genomes) and the number of singletons. The Louvain method, implemented in Gephi, was used to define the JI-groups by using resolution 1.5. Once the main JI groups are defined (containing a minimum of five genomes), they can be further dissected into several subgroups within the network using a more stringent JI and the same community detection algorithm [ 13 ]. The nodes of the network, representing genomes, were colored according to metadata and genetic determinants of interest. Edges between nodes were included whenever the corresponding JI value met or exceeded the user-defined threshold. Network figures were generated using the igraph package in R ( https://r.igraph.org/articles/igraph.html ). Pangenome characterization PanGraph [ 25 ] identifies blocks of homologous sequence and was used to detect indels specific to each JI-group. PanGraph was run on all genomes using parameters `α = 20` and `β = 20`. The parameter α controls the cost of splitting a block into smaller units, where a value of 20 was chosen to minimize excessive fragmentation of the graph. The parameter β controls the diversity cost and was set to 20, establishing a sequence diversity threshold of 20%. Only homologous sequences (pancontigs), larger than 250 bp, present in ≥ 85% of the members of each JI-group and not present in all JI-groups, were retained as “core” pancontigs. Core pancontigs for each JI-group were mapped with BLASTn against a reference genome (Supplementary Table S1 ) from their respective JI-group (sequenced by long-read technology, when available) to order the pancontigs and detect the regions they form. For instance, a prophage might be composed of several pancontigs, and scaffolding those contigs against a reference genome helped reconstruct and identify that element as an indel. The term “prophage” was used to refer to chromosomally-integrated regions that contained at least five phage-related genes according to PhageScope [ 37 ] or PHASTEST [ 38 ]. For pangenome comparison between U.S. and non-U.S. datasets, gene prediction of the assembled genomes was performed with Prokka [ 39 ]. Annotated assemblies in GFF3 format were used as input for pangenome calculation using Roary v3.13 [ 40 ], with 80% minimum percent identity and coverage. Pangenome gene categories were defined as: core genes (shared by 80–100% of the genomes); shell genes (15–79%); and cloud genes (0–14%). Heaps’ law was used to evaluate pangenome openness and closeness, using the script available at https://github.com/SethCommichaux/Heap_Law_for_Roary . Phylogenetic analysis cgMLST-based phylogenetic trees were generated using BioNumerics v7.6.3 [ 3 ]. Snippy v4.4.5 ( https://github.com/tseemann/snippy ) was used to detect core genome single nucleotide polymorphisms (cg-SNPs) in three datasets: JI-C chromosomes, JI-C PTU-I1 plasmids, and PTU-I1 plasmids from JI-E, JI-G, and other enterobacteria (RefSeq200). In all cases, the PTU-I1-containing genome SAL-20-VL-OH-OSU-0008 was used as reference. Alignments generated with Snippy were used to construct maximum-likelihood (ML) phylogenetic trees based on cg-SNPs by using IQ-TREE [ 41 ]. All trees generated in this study were rooted at midpoint and visualized with iTol v6 [ 42 ]. To complement cg-SNP analysis of PTU-I1 plasmids, AcCNET (Accessory Genome Constellation Network [ 43 ]) was used to build proteome networks and assess relatedness of plasmids at the protein level; proteins were clustered if they shared greater than 80% identity and 80% coverage. Statistical analysis Statistical analyses were performed using genomes collected through NARMS (CDC, FDA, FSIS), PulseNet (CDC), and FSIS national surveillance systems from years 2016 through 2023, in line with the introduction of routine sequencing for NARMS, PulseNet, and FSIS surveillance isolates. Corrected Cramer’s V was used to measure strength of associations between all categorical variables [ 44 ]; chi-squared tests of independence were used to test associations between specific epidemiological and genomic variables (Bonferroni adjusted significance value: p < 0.005); odds ratios (OR) (95% confidence intervals (CI)) were used to quantify the strength and direction of significant associations. For statistical tests involving a specific JI-group, the comparison group was always “all other JI-groups”. All tests were calculated using the stats subpackage of SciPy v1.14.1 implemented in Python v3.11.7 ( https://docs.scipy.org/doc/scipy/reference/stats.html ). JI-groups with less than 20 genomes were not analyzed for statistical associations. Only NARMS surveillance data collected by CDC, FDA, and FSIS (cecal sampling) were used to assess shifts in pangenome group abundance over time, as the isolates in the NARMS dataset were systematically collected and were more robust against large outbreaks and changes to regulatory testing practices than were the surveillance isolates from the PulseNet and FSIS product sampling datasets. Results Pangenome structure of United States Hadar population Hadar genomes self-organized into 18 clusters by JI (JI threshold = 0.988), labeled JI-A through R (Fig. 1 ); less than 5% of genomes (n = 158/3387) did not cluster with a JI-group and were considered singletons (Fig. 1 ). The three largest groups JI-A, JI-B, and JI-C were further divided into subgroups using an increased JI threshold (Supplementary Fig. S1 ). JI-A subgroups A1-15 were defined at JI = 0.995; JI-B subgroups B1-6 and JI-C subgroups C1-9 were defined at JI = 0.992. The MGE that define each JI-group include large plasmids (> 30 kb), prophages, AR regions, or regions of unknown function (Fig. 2 ). In some cases, two JI-groups differed only by the presence of a large plasmid (e.g., JI-A and JI-C; JI-B and JI-G; JI-D and JI-E), while others displayed more differences in their pangenome content (e.g., JI-I) (Fig. 2 ). ST (sequence type, based on 7 core loci), NCBI SNP cluster [ 45 ], and cgMLST allele code (based on n = 3002 core loci) [ 3 ] were separately visualized on the network to contextualize the pangenome with core lineage information. Over 98% of Hadar genomes in this analysis are ST33 (n = 3326/3384); only JI-I (ST473), JI-L (ST5130 and ST9222), and JI-Q (ST473) contained genomes of a different ST (Supplementary Fig. S2 a). NCBI SNP cluster aligned well with JI-groups; PDS000158107 was the most common cluster, encompassing the largest groups JI-A, JI-B and JI-C (Fig. 3 a). cgMLST allele codes also aligned well with JI-groups, with the majority of groups (n = 12/18) containing a single condensed allele code (Fig. 3 b). Despite being in the same NCBI SNP cluster, JI-A and JI-C separate from JI-B by condensed allele code (Fig. 3 a and 3 b). Both NCBI SNP cluster and cgMLST suggest membership within certain JI-groups is due to convergence in pangenome content rather than core genome similarity. Plasmids were common in U.S. Hadar genomes, with 60% (n = 2047/3384) containing one or more Col-like plasmids and 22% (n = 740/3387) carrying at least one large (> 30 kb) conjugative plasmid (Fig. 3 c, Supplementary Fig. S2 b). IncI1 was the most common replicon, detected in three different PTUs: PTU-I1, present in JI-C and JI-E, newly identified PTU-NA (IncI1, MOB P ) present in JI-J and JI-N, and newly identified PTU-NA (IncI1, MOB P ), present in JI-I (Fig. 2 , Fig. 3 c). JI-I also contained PTU-E78, a recently identified non-mobilizable PTU (Fig. 2 , 3 c). However, nearly 30% of genomes (n = 1011/3384) contained neither plasmid replicons nor MOB relaxase genes (Supplementary Table S1 ); these genomes predominantly fell into JI-A (Supplementary Fig. S2 b). PanGraph analysis revealed that integrated MGEs were also common in several JI-groups, including prophages and integrated conjugative elements (ICEs) (Fig. 2 and Supplementary File 1). Over 90% (n = 3055/3384) of genomes contained at least one AR determinant; predicted resistance to aminoglycosides (specifically, streptomycin) and tetracyclines was the most common profile, mediated by aph(3'')-Ib , aph(6)-Id , and tet(A) ), all integrated in the chromosome (Fig. 3 d, Supplementary Table S1 ). Predicted resistance to penicillins was less common (4%, n = 127/3384) and was predominantly mediated by bla TEM−1 (Supplementary Fig. S2 c and S2d). While rare, cephalosporin resistance mediated by bla CMY−2 was detected in groups JI-C and JI-E (0.4%, n = 12/3384; Supplementary Fig. S2 c and S2d). Members of JI-D, JI-I and JI-Q were predicted to be pansusceptible, with no known AR determinants detected (Supplementary Table S1 ). Genetic and epidemiological differences between most abundant pangenome groups The dominant pangenome groups changed substantially between 2016 and 2023, most notably between 2019 and 2020 (Fig. 4 , Supplementary Fig. S2 e and S2f). This shift was particularly pronounced for human and retail meat samples, where JI-A and JI-C were rare prior to 2020 yet comprise between 56% and 100% of samples collected in years 2020–2023. JI-B was the most common group detected in retail meat and animal (cecal) sampling prior to 2020 but decreased in detection substantially in 2020–2023; JI-B was not detected at all in 2023 retail meat sampling. Groups JI-D and JI-E made up more than half of human Hadar samples in 2016 and 2017 but have not been detected since 2019; these groups were not found in retail meat or animal sampling throughout the study years (Fig. 4 ). JI-A and JI-C are the most common JI-groups in all three sampling systems from 2020–2023. JI-A and JI-C are indistinguishable by cgMLST-based phylogeny (Fig. 5 : Ring 1) but differ in their pangenome by carriage of a ~ 100 kb PTU-I1 (IncI1) plasmid, which underpins the separation of these two JI-groups (Figs. 2 and 3 c). Most JI-A and JI-C genomes fall within a comparatively tight “emergent” clade that forms the CDC-defined REPTDK01 strain (Fig. 3 f, Fig. 5 ), associated with ground turkey consumption and backyard poultry contact based on previous multistate outbreak investigations [ 17 ]. Of interest, two temporally “ancestral” JI-A genomes isolated from wild ducks in 1990 are positioned in a clade adjacent to the emergent genomes (Fig. 5 ). This emergent clade invariably contains an ~ 8 kb prophage, labeled here prophage 1 (Supplementary File 1), that forms part of the core pangenome of JI-A and JI-C (Fig. 2 ). Prophage 1 was detected as early as 2004 in singleton Hadar genomes (imported “sweet good without custard or cream filling” from Pakistan), was seen in genomes from swine and commercial poultry samples from 2015, yet remained uncommon until the 2020 emergence of REPTDK01 (Fig. 5 , Supplementary Fig. S3). According to PHASTEST, prophage 1 is related to filamentous phages I2-2 and Ike, and contains a protein with N-terminal homology to the zonular occludens toxin protein (Zot) (Supplementary Fig. S4). The phage-encoded Zot proteins in Vibrio cholerae [ 46 ] and Campylobacter spp. [ 47 , 48 ] have a demonstrated pathogenic role attributable to a C-terminal enterotoxic domain [ 49 ]. While homology with Zot proteins does not imply toxigenic function, the Hadar Zot-like protein identified here was bioinformatically predicted to contain toxigenic regions using ToxinPred3.0 [ 50 ], hinting at a putative role in pathogenesis. Thus, prophage 1 presence is notable both from an epidemiological and biological perspective, and its pathogenic and adaptive capacity is being assessed with functional analysis. JI-B is the second most abundant pangenome group, predominantly encompassing genomes from commercial poultry (Fig. 1 , Fig. 3 e). A smaller group, JI-G, is indistinguishable from JI-B phylogenetically (Fig. 5 : Ring 1) but can be differentiated by the presence of PTU-I1 plasmids (Fig. 2 ). JI-B (and JI-G) genomes appear more diverse in their core genome relative to those from other dominant pangenome groups (e.g., JI-A, JI-C, JI-D and JI-E) (Fig. 5 ), which may be a reflection of time and environmental factors—genomes in JI-B were isolated as early as 2011 from poultry sources across the country (Supplementary Table S1 ). Analysis of JI-B subgroups did not reveal any geographic association (Supplementary Fig. S1 b) or link to specific processing facilities. Of note, genomes from human samples that were part of a 2019 multistate Hadar outbreak linked to ground turkey consumption (internal CDC investigation) all fell into JI-B or JI-G, suggesting Hadar strains from these groups are transmitted via food. In contrast, groups JI-D and JI-E were almost always from ill humans (rather than animal or meat samples) (Fig. 3 e), often with reported contact with backyard poultry (Supplementary Table S1 ). JI-D and JI-E genomes display relatively little core diversity (Fig. 5 ) and differ from each other only by the carriage of PTU-I1 (IncI1) plasmids (Fig. 2 ). They differ from other JI-groups phylogenetically in that they are encompassed in a single clade by core SNP analysis (Fig. 5 ); and they differ pangenomically in that they lack a common AR region (“AMR-encoding Tn 1.1”, Fig. 2 ) and are the only groups to carry PTU-X1 (IncX1) plasmids (Fig. 2 ). Genomes in these groups were part of 2016 and 2017 multistate outbreaks linked to contact with backyard poultry [ 51 ]. Two small pangenome groups, JI-H and JI-K, are of interest because of their connectivity to JI-A in the network, indicating pangenomic relatedness (Fig. 1 ). JI-H genomes are all from commercial chicken sampling or from ill humans (no exposure information available), representing a statistically significant “chicken-source cluster” (Supplementary Table S1 ; p < 0.00001, chi-squared) that is unique among the more common commercial turkey source. JI-K genomes were all isolated throughout 2023, are almost exclusively from turkey product samples (n = 11/12) and are predominantly from a single state (n = 8/12 were isolated in CA) (Supplementary Table S1 ). JI-K genomes carry prophage 1, along with two other larger prophages unique to this group (prophage 6.2 and prophage 10; Fig. 2 ), potentially representing recent divergence from REPTDK01. Several pangenome groups harbor PTU-I1 (IncI1) plasmids, including JI-C, JI-E and JI-G (Fig. 2 ). PTU-I1 (IncI1) plasmids are common in avian environments, often carry AR genes, and may play a role in virulence and growth inhibition of competing bacteria [ 52 , 53 ]; thus, their presence and diversity in this dataset were of interest. Core plasmid SNP analysis coupled with AcCNET plasmid proteome analysis were used to assess the relatedness of PTU-I1 (IncI1) plasmids between and within JI-groups (Fig. 6 ). PTU-I1 plasmids from all three JI-groups were surprisingly diverse in their core and proteome and intermingled phylogenetically with PTU-I1 plasmids from other Enterobacteriaceae species (Fig. 6 a and 6 b, Supplementary Table S5). Plasmids from the same JI-C subgroups clustered together phylogenetically (Fig. 6 a and Supplementary Fig. S5a) and proteomically (Fig. 6 c), indicating that plasmid content is responsible for JI-C subgrouping. Of note, the largest JI-C subgroup, JI-C1, likely represents a multiyear clonal expansion event, given the tight relatedness of its plasmids and chromosomal genome (Fig. 6 a and Supplementary Fig. S5). Hadar pangenome offers increased discriminatory power for retrospective and prospective public health investigations REPTDK01 was clearly detectable in the pangenome network—98% (n = 2148/2194) of these genomes fell into JI-A, JI-C, JI-N and JI-R (Fig. 3 f, Supplementary Table S1 )—genetically corroborating and adding confidence to the REPTDK01 definition using pangenomic data. Additionally, REPTDK01 was further stratified by JI-grouping and JI-subgrouping, revealing clear epidemiological patterns. For example, while JI-A itself was not statistically associated with either commercial or backyard poultry (Supplementary Table S4), JI-A2 contained predominantly commercial poultry-related genomes from the U.S. and Canada (n = 42/68), and none of the human clinical cases in this group (n = 24/68) reported backyard poultry contact. In contrast, JI-A3 was almost exclusively genomes from human clinical cases (n = 27/28), a third of which reported backyard poultry contact, and zero commercial poultry-related genomes fell into this group (Supplementary Table S1 and Fig. S1 a). JI-N genomes were all human clinical—mostly isolated from the northeast (n = 4/6)—and may represent a closely related subcluster of illnesses that differ from JI-A REPTDK01 strains only by the carriage of a large plasmid (PTU-NA, IncI1) (Fig. 2 ). JI-C was significantly associated with backyard poultry (p < 0.00001, chi-squared; Supplementary Table S4), representing a subgroup of REPTDK01 (defined by the carriage of PTU-I1 plasmids) that was likely transmitted to humans via animal contact rather than food. More specifically, epidemiological traceback data available for clonal subgroup JI-C1 disproved the involvement of a single backyard poultry supply store chain or hatchery, instead suggesting a common reservoir of Hadar upstream of hatcheries. Coupling pangenome data and epidemiological data, REPTDK01 strains can be further differentiated for both retrospective and prospective investigations. Several other non-REPTDK01 pangenome groups are statistically associated with a specific source or exposure. JI-B and JI-G are each significantly associated with commercial turkey (p < 0.00001, chi-squared); JI-B genomes had 17.5 times (95% CI: 13.7–22.3), and JI-G genomes had 5.9 times (95% CI: 2.1–17.1) higher odds of being from commercial turkey compared with all other JI-groups (Supplementary Table S4). Coupled with the absence of human cases reporting backyard poultry contact in these groups, it is likely that Hadar strains from JI-B and JI-G are acquired through foodborne transmission. In contrast, JI-D and JI-E were each significantly associated with backyard poultry contact (p < 0.00001, chi-squared); JI-D genomes had 2.6 times (95% CI: 1.9–3.6) and JI-E genomes had 5.2 times (95% CI: 2.7–10.6) greater odds of backyard poultry contact, relative to all other JI-groups (Supplementary Table S4). The stark lack of genomes from commercial poultry sources (only JI-D had a single commercial chicken source genome), and the predominance of backyard poultry-associated outbreak genomes in these groups (n = 140/191 in JI-D, n = 35/40 in JI-E), strongly suggests JI-D and JI-E strains of Hadar are transmitted through animal contact. It is important to note that cgMLST differentiates JI-B and JI-G genomes from JI-D and JI-E (Fig. 3 a). Thus, the pangenome analysis performed here provides additional genomic confidence in these attributions. A handful of small JI-groups contained genomes from humans with limited epidemiological information, but with one or two genomes from a known source. Specifically, both JI-F and JI-J contain a genome from raw dog food (containing duck) obtained from ad hoc pet food sampling (see Methods) (Supplementary Table S1 ). JI-L contains two genomes from imported shrimp (Ecuador) isolated in 2022 (Supplementary Table S1 ). Given the close relatedness of genomes within JI-groups (median average nucleotide identity within JI-groups is ≥ 99.95, Supplementary Fig. S6), the presence of the pet food and imported food genomes alongside genomes from human samples is suggestive of an epidemiological connection, though without exposure information reported by these ill people this link cannot be confirmed. Prospectively, the relatedness of additional human cases found to be within the JI-F and JI-J groups could inform which food items to assess during supplementary interviews of ill people included in an outbreak investigation. As mentioned above, several pairs of JI-groups differ only by the presence of PTU-I1 (IncI1) plasmids: JI-A and JI-C (plasmid present), JI-D and JI-E (plasmid present), JI-B and JI-G (plasmid present). We further assessed these pairs for epidemiological patterns associated with plasmid presence, including source of isolation, geographic region, and patient demographics (age, sex, site of infection, hospitalization), but no variables were significantly different between paired groups ( V 0.005, chi-squared). However, PTU-I1 (IncI1) plasmids were independently associated with backyard poultry-related sources (PTU-I1 n = 208, no PTU-I1 n = 526) when compared with commercial poultry sources (PTU-I1 n = 32, no PTU-I1 n = 699), and when compared with all other sources (PTU-I1 n = 305, no PTU-I1 n = 2345) (Supplementary Table S1 ; p < 0.00001, chi-squared). Thus, PTU-I1 plasmids have statistical support to serve as a genetic marker to distinguish strains transmitted via backyard poultry contact versus those more likely attributed to another source, which is of particular value for differentiating REPTDK01 strains that can be transmitted via several pathways. U.S. Hadar pangenome structure reflects a subset of global diversity A dataset of Hadar genomes (n = 1145) from 33 countries other than the U.S., isolated from 1950 through 2023, was used to assess differences in pangenome structure between separate geographical locations (Supplementary Fig. S7, Supplementary Table S2 ). The non-U.S. dataset partially overlapped with U.S. genomes: 33% of non-U.S. genomes clustered within JI-groups identified in the U.S. pangenome data, while 47% formed distinct JI-groups not present in the U.S. dataset (Supplementary Fig. S8, Supplementary Table S6). The non-U.S. dataset contained 3095 genes absent from the U.S. pangenome, while the U.S. dataset contained 1628 genes absent from the non-U.S. dataset (Supplementary Fig. S9). Both datasets exhibited moderately open pangenomes (Heaps’ law γ value ~ 0.2) and shared a core of 4187 genes. Notably, separate analysis of each dataset revealed similar core gene counts, further highlighting the robustness of the core genome across different geographic populations (Supplementary Fig. S9). Furthermore, the non-US dataset displays a larger number of cloud genes, suggesting a higher diversity within its accessory genome. Separate analyses of genomes from the United Kingdom (U.K.) (n = 484) and France (n = 306) were performed since they represented more than half of the non-U.S. genomes. Of 18 JI-groups defined in the U.S. dataset, the U.K. and France datasets shared only seven (170 genomes, 35%) and six (74 genomes, 24%) JI-groups, respectively. Seventeen U.K. JI-groups (228 genomes, 47%) and nine France JI-groups (162 genomes, 53%) were distinct from those isolated in the U.S (Supplementary Fig. S10 and S11, Supplementary Table S2 ). While no temporal shift was observed for pangenome groups from U.K. data (Supplementary Fig. S10), a notable increase in genomes belonging to a novel group, JI-S, was observed in the France dataset, beginning in 2019 (Supplementary Fig. S11). JI-S genomes contain a prophage closely related to prophage 1, highlighting an intriguing parallel dynamic to the recent proliferation of prophage 1-containing groups JI-A and JI-C in the U.S. Thus, these analyses suggest Hadar pangenomic diversity is largely geographically defined, with potentially important genetic overlaps that will be further investigated. Discussion Identifying the molecular mechanisms underlying shifts in bacterial populations is key to understanding the adaptive forces that drive evolution of human bacterial pathogens. Analysis of the Hadar pangenome confirmed known, and revealed unknown, epidemiological and microevolutionary dynamics. Before 2020, two distinct lineages separately dominated in commercial poultry (JI-B and JI-G) and backyard poultry environments (JI-D and JI-E). In 2020, an emergent lineage closely related to previously circulating strains became dominant, displacing the historical commercial poultry lineage. Around the same time, coinciding with a surge in backyard poultry ownership during the COVID-19 pandemic [ 16 ], this same emergent lineage became dominant among backyard poultry-associated human cases – confirming through high-resolution pangenomic analysis a link between two presumably separate industries. Epidemiological and biological evidence suggest the presence of a novel phage in the emergent lineage may have contributed to its recent expansion. Interestingly, a similar genetic shift underpinned by an emergent phage-containing lineage was seen in the French genomes analyzed here, suggesting this phenomenon is not restricted to the U.S. The adaptive capacity of this prophage in Hadar, and specifically, the putative pathogenic role of the phage-encoded Zot-like protein, is still being evaluated in U.S. Hadar genomes. These new findings can be leveraged to mitigate further spread of this emergent strain in a number of ways. First, comparative plasmid analysis revealed a clonal subcluster of this lineage (JI-C1) that points to a reservoir upstream of backyard poultry suppliers and hatcheries, one that likely interfaces with commercial poultry. Backyard poultry hatchery practices, such as drop-shipping and outsourcing to larger commercial hatcheries to meet demand [ 54 , 55 ], could explain this connection. This data can inform conversations between industry and government stakeholders, as it promotes collective action with the goal of eliminating shared reservoirs affecting multiple industries. Second, functional analyses to determine the contribution of prophage 1 to avian gut colonization could inform intervention strategies in both commercial and backyard poultry settings; for example, by minimizing bacterial burden in birds, which is considered a control strategy to reduce risk of transmission to humans [ 56 ]. Third, this analysis highlighted the importance of known MGE (e.g., PTU-I1 plasmids) and identified previously uncharacterized MGE (e.g., prophage 1) that can potentially be incorporated into source attribution models and molecular case definitions. For example, PTU-I1 plasmids could serve as a genetic marker that distinguishes backyard poultry-related strains from those transmitted via other sources. More accurate prediction of foodborne versus animal contact transmission pathways and refinement of outbreak and REP strain case definitions both contribute to timelier epidemiological traceback, and ultimately, a reduction in human illness [ 14 ]. More generally, this analysis enabled high-resolution genomic linking of human cases with potential sources (e.g., pet food, imported shrimp), raising suspicion of specific vehicles to refine supplemental interviews or traceback efforts when exposure information is limited, and no transmission vehicles are otherwise suspected. Additionally, avenues were identified for investigation of ecological dynamics that underpin persistence of Hadar in different environments. For example, PTU-I1 and other large plasmids are associated with backyard poultry rather than commercial poultry environments; and some JI-groups (with unique MGE profiles) display a unique chicken association rather than the more common turkey signal. Along with highlighting the previously unreported role of prophages in Hadar diversification and microevolution, this broad description of MGE in the U.S. Hadar population is foundational information for pathogen risk modeling, especially as it pertains to carriage of AR. The presence of “risky” MGE related to AR, virulence, or colonization capacity, can be proactively monitored through existing surveillance programs, and any emergent threats addressed before they become systematically disseminated, as has previously occurred with Salmonella serotypes Infantis [ 57 ] and Reading [ 58 ]. By nature of the pangenomic approach employed here, exactly when and where this persisting REP strain arose was not determined; however, a molecular clock analysis is underway to explore the rapid rise and subsequent diversification of this lineage. Further, while source of human illnesses with unknown exposures, or those with multiple exposures (e.g., both commercial and backyard poultry), cannot be definitively determined using this approach, the findings from this study will be assessed within ongoing source attribution modeling to estimate the added value of inclusion of accessory genome content. Further, while efforts were made to obtain genomes representing diverse environments (wildlife, imported foods, commercial poultry production, backyard poultry environments, ill humans), several sources are underrepresented (e.g., live animals on farm) or absent (e.g., hatcheries), potentially missing pangenomic groups that are dominant in these spaces. Expanded analyses that include genomes from underrepresented sources, coupled with deeper investigation into the global diversity of Hadar, will fill important gaps in the pangenome landscape described here. Unraveling pathogen epidemiology and microevolutionary dynamics is highly complex, and the plethora of available data is both an opportunity and a challenge. Leveraging existing genomic data, we demonstrate the value of JI-based pangenomic analysis for delineating a highly clonal serotype and uncover actionable data to mitigate the spread of an emergent, and potentially more pathogenic, lineage of Hadar. We paint a pangenome landscape of this previously understudied serotype, highlighting the importance of known and unknown MGE, and revealing surprising geographic patterns and dynamics. These findings will inform future risk and source attribution modeling, reducing public health burdens and mitigating impacts on implicated food and animal industries. Declarations Acknowledgements The authors would like to acknowledge state and local public health departments and laboratories for isolation and sequencing of Hadar genomes included in this analysis. The authors thank FDA colleagues Olgica Ceric, Beilei Ge, Claudine Kabera, as well as University of Minnesota Professor Timothy Johnson, for their valuable expertise. This work was supported by the Centers for Disease Control and Prevention (Contract No. 75D30123P18303 to FdlC). This work was also supported by the Spanish Ministry of Science and Innovation MCIN/AEI/10.13039/501100011033 (PID2020-117923GB-I00 to FdlC and MPGB). Author Contributions KAT, APC, HEW, GSS, MKS, KB, MPGB and FdlC – Conceptualization. KAT, APC, HEW, GSS, ZE, ML, JYK, MS, CL, BH, BRMS, DM, SM, KM, JH, JMW, JMB and KB – Data collection. KAT, APC, HEW, GSS, MS, CL, BH, BRMS, KB and UD – Data curation. KAT, APC, HEW, MKS, SRS, MPGB and FdlC – Methodology. KAT, APC, HEW, MKS, SRS, and MPGB – Analysis. KAT, APC, HEW, MKS, SRS, MPGB and FdlC – Visualization. KAT, APC, HEW, GSS, KB, MPGB and FdlC – Writing - original draft. KAT, APC, HEW, GSS, ZE, ML, JYK, MS, GT, CL, BH, BRMS, MKS, DM, SM, KM, JH, JMW, CS, JMB, SS, KB, JPF, UD, SRS, MPGB and FdlC – Writing - review and editing. All authors read and approved the final manuscript. Competing Interests The authors declare that they have no competing interests. Disclaimer The views expressed in this article are those of the authors and do not necessarily reflect the official policy of the Agencies within the U.S. Department of Health and Human Services (CDC, FDA) and the U.S. Department of Agriculture (FSIS, APHIS), or the U.S. Government. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Health and Human Services or the U.S. Department of Agriculture. References Arnold BJ, Huang IT, Hanage WP (2021) Horizontal gene transfer and adaptive evolution in bacteria. Nat Rev Microbiol 20:206–218 Jagadeesan B et al (2019) The use of next generation sequencing for improving food safety: Translation into practice. Food Microbiol 79:96–115 Leeper MM et al (2023) Evaluation of whole and core genome multilocus sequence typing allele schemes for Salmonella enterica outbreak detection in a national surveillance network, PulseNet USA. Front Microbiol 14:1254777 Castillo-Ramírez S (2022) Beyond microbial core genomic epidemiology: towards pan genomic epidemiology. Lancet Microbe 3:e244–e245 Liu CM et al (2023) Using source-associated mobile genetic elements to identify zoonotic extraintestinal E. coli infections. One Health 16:100518 Mcnally A et al (2016) Combined analysis of variation in core, accessory and regulatory genome regions provides a super-resolution view into the evolution of bacterial populations. PLoS Genet 12:e1006280 Guillier L et al (2020) AB_SA: Accessory genes-Based Source Attribution – tracing the source of Salmonella enterica Typhimurium environmental strains. Microb Genom 6:mgen000366 Batz MB et al (2021) Recency-weighted statistical modeling approach to attribute illnesses caused by 4 pathogens to food sources using outbreak data, United States. Emerg Infect Dis 27:214–222 Beshearse E et al (2021) Attribution of illnesses transmitted by food and water to comprehensive transmission pathways using structured expert judgment, United States. Emerg Infect Dis 27:182–195 Interagency Food Safety Analytics Collaboration. Foodborne illness source attribution estimates for Salmonella, Escherichia coli O157, and Listeria monocytogenes – United States (2022) GA and D.C.: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Food and Drug Administration, U.S. Department of Agriculture's Food Safety and Inspection Service. December 13, 2024 Commichaux S et al (2023) Assessment of plasmids for relating the 2020 Salmonella enterica serovar Newport onion outbreak to farms implicated by the outbreak investigation. BMC Genomics 24:165 Pan H et al (2022) Comprehensive assessment of subtyping methods for improved surveillance of foodborne Salmonella . Microbiol Spectr 10:e02479–e02422 Peñil-Celis A et al (2024) Mobile genetic elements define the non-random structure of the Salmonella enterica serovar Typhi pangenome. mSystems 9:e0036524 Brandenburg JM et al (2024) Salmonella Hadar linked to two distinct transmission vehicles highlights challenges to enteric disease outbreak investigations. Epidemiol Infect 152:e86 Stapleton GS et al (2024) Multistate outbreaks of salmonellosis linked to contact with backyard poultry-United States, 2015–2022. Zoonoses Public Health 71:708–722 Nichols M et al (2021) Salmonella illness outbreaks linked to backyard poultry purchasing during the COVID-19 pandemic: United States, 2020. Epidemiol Infect 149:e234 Centers for Disease Control and Prevention (2024) Data Summary: Persistent Strain of Salmonella Hadar (REPTDK01) Linked to Backyard Poultry and Ground Turkey. Available from: https://www.cdc.gov/salmonella/php/data-research/reptdk01.html [Accessed December 2nd, 2024] Ribot EM, Freeman M, Hise KB, Gerner-Smidt P, PulseNet (2019) Entering the age of next-generation sequencing. Foodborne Pathog Dis 16:451–456 Tolar B et al (2019) An overview of PulseNet USA databases. Foodborne Pathog Dis 16:457–462 Mcmillan EA et al (2019) Antimicrobial resistance genes, cassettes, and plasmids present in Salmonella enterica associated with United States food animals. Front Microbiol 10:832 Zhou Z, Alikhan NF, Mohamed K, Fan Y, Achtman M (2020) The EnteroBase user's guide, with case studies on Salmonella transmissions, Yersinia pestis phylogeny, and Escherichia core genomic diversity. Genome Res 30:138–152 Achtman M et al (2021) Genomic diversity of Salmonella enterica -The UoWUCC 10K genomes project [version 2; peer review: 2 approved]. Wellcome Open Res 5 Yoshida CE et al (2016) The Salmonella In Silico Typing Resource (SISTR): An open web-accessible tool for rapidly typing and subtyping draft Salmonella genome assemblies. PLoS ONE 11:e0147101 Zhang S et al (2019) SeqSero2: Rapid and improved Salmonella serotype determination using whole-genome sequencing data. Appl Environ Microbiol 85:e01746–e01719 Noll N, Molari M, Shaw LP, Neher RA (2023) PanGraph: scalable bacterial pan-genome graph construction. Microb Genom 9:mgen001034 Carattoli A et al (2014) In silico detection and typing of plasmids using PlasmidFinder and plasmid multilocus sequence typing. Antimicrob Agents Chemother 58:3895–3903 Garcillán-Barcia MP, Redondo-Salvo S, Vielva L (2075) & De La Cruz, F. MOBscan: automated annotation of MOB relaxases. Methods Mol. Biol. (2020) Redondo-Salvo S et al (2021) COPLA, a taxonomic classifier of plasmids. BMC Bioinformatics 22:390 Redondo-Salvo S et al (2020) Pathways for horizontal gene transfer in bacteria revealed by a global map of their plasmids. Nat Commun 11:3602 Schwengers O et al (2021) Bakta: rapid and standardized annotation of bacterial genomes via alignment-free sequence identification. Microb Genom 7:000685 Bortolaia V et al (2020) ResFinder 4.0 for predictions of phenotypes from genotypes. J Antimicrob Chemother 75:3491–3500 Robertson J, Nash JHE (2018) MOB-suite: software tools for clustering, reconstruction and typing of plasmids from draft assemblies. Microb Genom 4:e000206 Tagg KA et al (2024) Azithromycin-resistant mph(A) -positive Salmonella enterica serovar Typhi in the United States. J Glob Antimicrob Resist 39:69–72 Li C et al (2024) The spread of pESI-mediated extended-spectrum cephalosporin resistance in Salmonella serovars—Infantis, Senftenberg, and Alachua isolated from food animal sources in the United States. PLoS ONE 19:e0299354 Webb HE et al (2022) Genome sequences of 18 Salmonella enterica serotype Hadar strains collected from patients in the United States. Microbiol Resour Announc 11:e0052222 Bastian M, Heymann S, Jacomy M, Gephi (2009) An open source software for exploring and manipulating networks. Proc. Internat. AAAI Conf. Web Social Media 3, 361–362 Wang RH et al (2024) PhageScope: a well-annotated bacteriophage database with automatic analyses and visualizations. Nucleic Acids Res 52:D756–D761 Wishart DS et al (2023) PHASTEST: faster than PHASTER, better than PHAST. Nucleic Acids Res 51:W443–W450 Seemann T (2014) Prokka: rapid prokaryotic genome annotation. Bioinformatics 30:2068–2069 Page AJ et al (2015) Roary: rapid large-scale prokaryote pan genome analysis. Bioinformatics 31:3691–3693 Minh BQ et al (2020) IQ-TREE 2: New models and efficient methods for phylogenetic inference in the genomic era. Mol Biol Evol 37:1530–1534 Letunic I, Bork P (2024) Interactive Tree of Life (iTOL) v6: recent updates to the phylogenetic tree display and annotation tool. Nucleic Acids Res 52:W78–W82 Lanza VF, Baquero F, De La Cruz F, Coque TM (2016) AcCNET (Accessory Genome Constellation Network): comparative genomics software for accessory genome analysis using bipartite networks. Bioinformatics 33:283–285 Bergsma W (2013) A bias-correction for Cramér’s V and Tschuprow’s T. J Korean Stat Soc 42:323–328 The NCBI Pathogen Detection Project [Internet]. Bethesda (MD): National Library of Medicine (US), National Center for Biotechnology Information (2016) May Available from: https://www.ncbi.nlm.nih.gov/pathogens/ Fasano A et al (1991), Vibrio cholerae produces a second enterotoxin, which affects intestinal tight junctions. Proc. Natl. Acad. Sci. U S A 88, 5242–5246 Liu F, Lee H, Lan R, Zhang L (2016) Zonula occludens toxins and their prophages in Campylobacter species. Gut Pathog 8:43 Mahendran V et al (2016) Examination of the effects of Campylobacter concisus zonula occludens toxin on intestinal epithelial cells and macrophages. Gut Pathog 8:18 Di Pierro M et al (2001) Zonula occludens toxin structure-function analysis. Identification of the fragment biologically active on tight junctions and of the zonulin receptor binding domain. J Biol Chem 276:19160–19165 Rathore AS, Choudhury S, Arora A, Tijare P, Raghava GPS (2024) ToxinPred 3.0: An improved method for predicting the toxicity of peptides. Comput Biol Med 179:108926 2016 Salmonella Outbreak Linked to Live Poultry in Backyard Flocks Available from: https://archive.cdc.gov/#/details?url=https://www.cdc.gov/salmonella/live-poultry-05-16/index.html [Accessed December 2nd, 2024] Kaldhone PR et al (2019) Evaluation of incompatibility group I1 (IncI1) plasmid-containing Salmonella enterica and assessment of the plasmids in bacteriocin production and biofilm development. Front Vet Sci 6:298 Foley SL, Kaldhone PR, Ricke SC, Han J (2021) Incompatibility group I1 (IncI1) plasmids: Their genetics, biology, and public health relevance. Microbiol Mol Biol Rev 85:e00031–e00020 Nichols M et al (2018) Preventing human Salmonella infections resulting from live poultry contact through interventions at retail stores. J Agric Saf Health 24:155–166 Behravesh CB, Brinson D, Hopkins BA, Gomez TM (2014) Backyard poultry flocks and salmonellosis: A recurring, yet preventable public health challenge. Clin Infect Dis 58:1432–1438 Galán-Relaño Á et al (2023) Salmonella and salmonellosis: An update on public health implications and control strategies. Animals 13:3666 Centers for Disease Control and Prevention (2024) Data Summary: Persistent Strain of Salmonella Infantis (REPJFX01) Linked to Chicken. Available from: https://www.cdc.gov/salmonella/php/data-research/repjfx01.html [Accessed December 2nd, 2024] Hassan R, Buuck S, Noveroske D (2019) Multistate outbreak of Salmonella infections linked to raw turkey products — United States, 2017–2019. MMWR Morb Mortal Wkly Rep 68:1045–1049 Additional Declarations There is NO Competing Interest. Supplementary Files TaggetalSalmonellaHadarSupplementaryInformation.docx TaggetalSalmonellaHadarSupplementaryTables.xlsx Cite Share Download PDF Status: Published Journal Publication published 24 Jan, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5896627","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":408350319,"identity":"15a24094-e140-42f7-be21-b003b826cd3f","order_by":0,"name":"Hattie Webb","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAo0lEQVRIiWNgGAWjYBACAwYGNmaGCgbGBhK1nCFZC2MbKVrMJZKPPS6cZyfb38B+8TEPMVosZ6SlG8/clmw84wBPsTFRWgxu55hJ8247kLiBgSdNcgZxWvK/SfPOIU1LDps0bwNIC/sxiQ9Eabn/zNx4xjGgXw7zMBsQp+XM4WePC2qAIdbe/vBBAjFaEICZx4A0DUDA/oBkLaNgFIyCUTAyAAC4azJjSthkvAAAAABJRU5ErkJggg==","orcid":"","institution":"Centers for Disease Control and Prevention","correspondingAuthor":true,"prefix":"","firstName":"Hattie","middleName":"","lastName":"Webb","suffix":""},{"id":408350320,"identity":"8a014188-7179-4992-8a8f-b880b4bbd006","order_by":1,"name":"Kaitlin Tagg","email":"","orcid":"","institution":"Centers for Disease Control and Prevention","correspondingAuthor":false,"prefix":"","firstName":"Kaitlin","middleName":"","lastName":"Tagg","suffix":""},{"id":408350321,"identity":"554b828d-65c8-40e7-acc3-737798f4c958","order_by":2,"name":"Arancha Peñil-Celis","email":"","orcid":"","institution":"Instituto de Biomedicina y Biotecnología de Cantabria","correspondingAuthor":false,"prefix":"","firstName":"Arancha","middleName":"","lastName":"Peñil-Celis","suffix":""},{"id":408350322,"identity":"4b68d85e-410c-4dea-8548-38d39d11903c","order_by":3,"name":"G Stapleton","email":"","orcid":"","institution":"Centers for Disease Control and Prevention","correspondingAuthor":false,"prefix":"","firstName":"G","middleName":"","lastName":"Stapleton","suffix":""},{"id":408350323,"identity":"a48a27b4-1d8c-4382-9a2d-b056b94930e2","order_by":4,"name":"Zachary Ellison","email":"","orcid":"","institution":"Centers for Disease Control and Prevention","correspondingAuthor":false,"prefix":"","firstName":"Zachary","middleName":"","lastName":"Ellison","suffix":""},{"id":408350324,"identity":"2087e3b7-3cac-4ac8-962d-7ca2eaa41a7b","order_by":5,"name":"Molly Leeper","email":"","orcid":"","institution":"Centers for Disease Control and Prevention","correspondingAuthor":false,"prefix":"","firstName":"Molly","middleName":"","lastName":"Leeper","suffix":""},{"id":408350325,"identity":"72a45b37-3604-4c07-b8d3-c68b634ec87f","order_by":6,"name":"Justin Kim","email":"","orcid":"","institution":"Centers for Disease Control and Prevention","correspondingAuthor":false,"prefix":"","firstName":"Justin","middleName":"","lastName":"Kim","suffix":""},{"id":408350326,"identity":"9935668c-ca88-4fc0-a040-48498cf1deab","order_by":7,"name":"Mustafa Simmons","email":"","orcid":"","institution":"United States Department of Agriculture","correspondingAuthor":false,"prefix":"","firstName":"Mustafa","middleName":"","lastName":"Simmons","suffix":""},{"id":408350327,"identity":"b4756813-0503-45c5-ad14-dc030c6ca5e6","order_by":8,"name":"Glenn Tillman","email":"","orcid":"","institution":"United States Department of Agriculture","correspondingAuthor":false,"prefix":"","firstName":"Glenn","middleName":"","lastName":"Tillman","suffix":""},{"id":408350328,"identity":"cb0dde53-9f89-4ed4-8a59-cf853183dd11","order_by":9,"name":"Cong Li","email":"","orcid":"","institution":"U.S. Food and Drug Administration, Center for Veterinary Medicine","correspondingAuthor":false,"prefix":"","firstName":"Cong","middleName":"","lastName":"Li","suffix":""},{"id":408350329,"identity":"a30cabf7-2780-4280-bd93-34ce255b3e87","order_by":10,"name":"Shaohua Zhao","email":"","orcid":"","institution":"FDA, Center for Veterinary Medicine","correspondingAuthor":false,"prefix":"","firstName":"Shaohua","middleName":"","lastName":"Zhao","suffix":""},{"id":408350330,"identity":"85c11596-1e4a-4682-aebc-0d2d33a55401","order_by":11,"name":"Beth Harris","email":"","orcid":"","institution":"United States Department of Agriculture","correspondingAuthor":false,"prefix":"","firstName":"Beth","middleName":"","lastName":"Harris","suffix":""},{"id":408350331,"identity":"32f84baa-4ddc-4450-a494-6c6b964302ea","order_by":12,"name":"Brenda Morningstar-Shaw","email":"","orcid":"","institution":"United States Department of Agriculture","correspondingAuthor":false,"prefix":"","firstName":"Brenda","middleName":"","lastName":"Morningstar-Shaw","suffix":""},{"id":408350332,"identity":"942ddca5-3000-4de6-9940-82b58e7e0f43","order_by":13,"name":"Molly Steele","email":"","orcid":"","institution":"Centers for Disease Control and Prevention","correspondingAuthor":false,"prefix":"","firstName":"Molly","middleName":"","lastName":"Steele","suffix":""},{"id":408350333,"identity":"12f83ece-598a-4b67-9f7c-f5b95012a7dd","order_by":14,"name":"Daniel Mallal","email":"","orcid":"","institution":"Colorado Department of Public Health and Environment","correspondingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"","lastName":"Mallal","suffix":""},{"id":408350334,"identity":"822909c1-25db-46e3-9fef-c3c392b576c9","order_by":15,"name":"Shannon Matzinger","email":"","orcid":"","institution":"Colorado Department of Public Health and Environment","correspondingAuthor":false,"prefix":"","firstName":"Shannon","middleName":"","lastName":"Matzinger","suffix":""},{"id":408350335,"identity":"83d497cc-7281-40d3-a6e3-6e7f7eb9c335","order_by":16,"name":"Kathy Manion","email":"","orcid":"","institution":"Montana Public Health Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Kathy","middleName":"","lastName":"Manion","suffix":""},{"id":408350336,"identity":"45bdfc6e-bbe7-472c-920c-3ee6255e3d3d","order_by":17,"name":"John Hergert","email":"","orcid":"","institution":"Utah Public Health Laboratory","correspondingAuthor":false,"prefix":"","firstName":"John","middleName":"","lastName":"Hergert","suffix":""},{"id":408350337,"identity":"1f6d8935-bf52-4517-8e47-7e33b4cd2ffa","order_by":18,"name":"Jennifer Wagner","email":"","orcid":"","institution":"Utah Public Health Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Jennifer","middleName":"","lastName":"Wagner","suffix":""},{"id":408350338,"identity":"ccdd56f0-9120-4438-8f9c-f5877520b832","order_by":19,"name":"Colin Schwensohn","email":"","orcid":"","institution":"Centers for Disease Control and Prevention","correspondingAuthor":false,"prefix":"","firstName":"Colin","middleName":"","lastName":"Schwensohn","suffix":""},{"id":408350340,"identity":"d88148e7-7964-4e63-8ac9-a72e0010c1d3","order_by":20,"name":"Joshua Brandenburg","email":"","orcid":"","institution":"Centers for Disease Control and Prevention","correspondingAuthor":false,"prefix":"","firstName":"Joshua","middleName":"","lastName":"Brandenburg","suffix":""},{"id":408350342,"identity":"fc3d5400-7a6f-4ce9-92b4-9a3e1d096e49","order_by":21,"name":"Sheryl Shaw","email":"","orcid":"","institution":"United States Department of Agriculture","correspondingAuthor":false,"prefix":"","firstName":"Sheryl","middleName":"","lastName":"Shaw","suffix":""},{"id":408350344,"identity":"02e42b8a-f6ac-49e6-8f4f-9c5a0697bfda","order_by":22,"name":"Katherine Benedict","email":"","orcid":"","institution":"Centers for Disease Control and Prevention","correspondingAuthor":false,"prefix":"","firstName":"Katherine","middleName":"","lastName":"Benedict","suffix":""},{"id":408350347,"identity":"edc5d25b-40e6-4251-9be6-86530908313c","order_by":23,"name":"Jason Folster","email":"","orcid":"","institution":"Centers for Disease Control and Prevention","correspondingAuthor":false,"prefix":"","firstName":"Jason","middleName":"","lastName":"Folster","suffix":""},{"id":408350348,"identity":"1787adc1-aa68-4f13-bbcc-70e580095c7d","order_by":24,"name":"Uday Dessai","email":"","orcid":"","institution":"United States Department of Agriculture","correspondingAuthor":false,"prefix":"","firstName":"Uday","middleName":"","lastName":"Dessai","suffix":""},{"id":408350351,"identity":"14eb5efc-d662-48b1-bf64-01098ed8c709","order_by":25,"name":"Santiago Redondo Salvo","email":"","orcid":"https://orcid.org/0000-0003-0089-6030","institution":"University of Cantabria","correspondingAuthor":false,"prefix":"","firstName":"Santiago","middleName":"Redondo","lastName":"Salvo","suffix":""},{"id":408350352,"identity":"7af1a491-b7c6-4d2b-b4ef-f90f74254acd","order_by":26,"name":"M. Pilar Garcillán-Barcia","email":"","orcid":"https://orcid.org/0000-0001-7058-5428","institution":"Spanish National Research Council","correspondingAuthor":false,"prefix":"","firstName":"M.","middleName":"Pilar","lastName":"Garcillán-Barcia","suffix":""},{"id":408350355,"identity":"a7f8a6af-d725-4b23-87d8-03f5a08dba7d","order_by":27,"name":"Fernando de la Cruz","email":"","orcid":"https://orcid.org/0000-0003-4758-6857","institution":"Universidad de Cantabria","correspondingAuthor":false,"prefix":"","firstName":"Fernando","middleName":"de la","lastName":"Cruz","suffix":""}],"badges":[],"createdAt":"2025-01-24 15:01:05","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5896627/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5896627/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-025-68026-3","type":"published","date":"2026-01-24T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":82261892,"identity":"9454f379-fcc3-4120-9a7d-bb3f7c1c6f7f","added_by":"auto","created_at":"2025-05-08 12:17:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":147081,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistribution of \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eSalmonella\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e Hadar genomes by JI.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe network contains 3384 nodes, connected when JI ≥ 0.988. Eighteen groups (JI-A to JI-R) are labeled, singleton genomes that do not associate with a JI-group are displayed around the outside of the network. Genomes are colored according to JI-group. Counts and percentages of genomes within each JI-group are listed.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-5896627/v1/002bb8ebcb5921fbb1516ca7.png"},{"id":82261666,"identity":"84c9e47f-f5e9-4e12-8c25-a298cf356ed4","added_by":"auto","created_at":"2025-05-08 12:09:39","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":147131,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDifferential distribution of accessory genome elements in JI-groups.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe heatmap displays the presence or absence of indels detected using PanGraph [25]. Indels larger than 5 kb (and their variants) were included in the analysis. Each column represents a JI-group (labeled along the bottom axis). Each row corresponds to an indel; presence in the corresponding JI-group is indicated in black, and absence indicated in white. The left bar categorizes the indels as “plasmid”, “prophage”, “ICE”, “other”, or “unknown”, as represented in the legend. Variants of named indels are indicated with a digit (e.g. Prophage 7.2).\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-5896627/v1/b7521cad99bf2f5d57ea0a19.png"},{"id":82261671,"identity":"66dd03f6-1bae-4e87-be0f-45ff341a1eb5","added_by":"auto","created_at":"2025-05-08 12:09:39","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":329919,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistribution of Hadar genomes by variables of interest.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe networks contain 3384 nodes, connected when JI ≥ 0.988. Eighteen groups (JI-A to JI-R) are labeled. Legends are left of each network. a) NCBI SNP cluster across JI-groups (obtained from Pathogen Detection Isolate Browser [45]. b) Condensed allele code across JI-groups. Grey nodes indicate genomes with no assigned allele code. c) Plasmid taxonomic units (PTUs) across JI-groups. Grey nodes indicate genomes with no known PTU. d) Most common antimicrobial resistance (AR) genes conferring predicted aminoglycoside and tetracycline resistance. Grey nodes indicate genomes without selected AR genes. e) Genomes isolated from different sources. f) Genomes determined to be REPTDK01 strains according to CDC-defined cgMLST allele range. Grey nodes indicate genomes not assigned to REPTDK01.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-5896627/v1/c6cfda3fd3d4e3e3a566fdf2.png"},{"id":82261893,"identity":"cdaaca8f-9590-47d9-96bb-db36345b0a04","added_by":"auto","created_at":"2025-05-08 12:17:39","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":62253,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAbundance of JI-groups over time.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNARMS data from CDC (humans), FDA (retail meats) and FSIS (animals) from 2016-2023 are included. Years are displayed on the x-axis and counts of isolates according to JI-group are displayed on the y-axis.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-5896627/v1/8266bd2e8c3801148d6f64cf.png"},{"id":82261896,"identity":"8988f18f-9c4f-4431-83fc-ffe2f7ada43e","added_by":"auto","created_at":"2025-05-08 12:17:39","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":346162,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCore genome MLST-based phylogenetic tree.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e3363 Hadar genomes are included in the tree (21/3384 could not be processed through BioNumerics (Supplementary Table S1)): Ring 1 displays JI-group, Ring 2 displays time range (1990-2019 versus 2020-2023), Ring 3 displays source, and Ring 4 displays presence of prophage 1 detected in this study. The large clade colored in green represents REPTDK01. Two “ancestral” genomes collected in 1990 are shown with orange branches and an asterisk at approximately 1 o’clock. A list of genomes included in this tree is available in Supplementary Table S1. Tree was generated using BioNumerics v7.6.3 and visualized in iTol v6 [42].\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-5896627/v1/5dbf157bec457337470258df.png"},{"id":82261894,"identity":"24de9c04-7951-4f5f-b86a-6bc2938ed978","added_by":"auto","created_at":"2025-05-08 12:17:39","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":488876,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCore and protein analysis of PTU-I1 plasmids.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ea) Maximum likelihood (ML) core genome phylogenetic tree of 512 PTU-I1 plasmids from the Hadar dataset (see PTU-I1 plasmids of JI-C, JI-E and JI-G in column “PTU” of Supplementary Table S1) and 259 PTU-I1 plasmids from RefSeq200 Enterobacterales hosts (Supplementary Table S5), generated using IQ-TREE v2 [41]. The tree was midpoint rooted and visualized in iTol v6 [41], UFBootstrap values \u0026gt; 80% are indicated by circles on the corresponding nodes, branch length scale represents substitutions per site. Ring 1 displays the plasmid host genus, Ring 2 displays the JI-group of plasmids found in Hadar, Ring 3 displays the JI-subgroup of the JI-C plasmids. b) Proteome network of PTU-I1 plasmids colored by JI-group. The proteins of the PTU-I1 plasmids were clustered at 80% identity and 80% coverage using AcCNET [43]. The larger nodes correspond to plasmids and are colored according to the JI-group of the Hadar plasmids (JI-C, JI-E, JI-G), or in grey if they belong to other Enterobacterales. The smaller nodes represent homologous protein clusters (HPCs) and are colored in grey. Both kinds of nodes are connected if a plasmid contains a member in the corresponding protein cluster. HPCs present in a single plasmid were removed. c) Proteome network of PTU-I1 plasmids colored by JI-C-subgroup. The network was constructed as indicated in Fig. 6b. The JI-C Hadar plasmids are colored based on their JI-subgroup.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-5896627/v1/4e0c0a1b42faef551619dcec.png"},{"id":101839210,"identity":"4e06f64c-96fc-4eef-8d1b-cf20f2ac744a","added_by":"auto","created_at":"2026-02-04 08:12:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2239776,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5896627/v1/547b08a6-bee0-4ef8-98e3-6fc51e85ec8a.pdf"},{"id":82261679,"identity":"07ffe686-752d-409d-b187-efe802283f9a","added_by":"auto","created_at":"2025-05-08 12:09:40","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2230637,"visible":true,"origin":"","legend":"","description":"","filename":"TaggetalSalmonellaHadarSupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-5896627/v1/ae140cc47666399009ad3218.docx"},{"id":82261672,"identity":"3aaf9c05-9892-4ac3-a3a7-6e8a51af8177","added_by":"auto","created_at":"2025-05-08 12:09:39","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":567358,"visible":true,"origin":"","legend":"","description":"","filename":"TaggetalSalmonellaHadarSupplementaryTables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5896627/v1/1d057f5c3edbaf55bb34b227.xlsx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"The Salmonella enterica serotype Hadar pangenome: Population structure and dynamics of a zoonotic pathogen","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe accessory genome, comprising plasmids, prophages, genomic islands, and other mobile genetic elements (MGE), is a key component of bacterial evolution [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. While typically excluded from phylogenetic or source attribution analyses [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], there is growing interest in the discriminatory and predictive power of the accessory genome for epidemiological investigations [\u003cspan additionalcitationids=\"CR5 CR6\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. For zoonotic pathogens like \u003cem\u003eSalmonella enterica\u003c/em\u003e with numerous transmission routes [\u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], analysis of the pangenome (accessory and core genome) has proven useful for enhanced surveillance, outbreak investigation, and microevolutionary exploration [\u003cspan additionalcitationids=\"CR12\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. The added public health value of pangenome data, however, depends on the unique genomic structure and microbial ecology of each \u003cem\u003eSalmonella\u003c/em\u003e serotype and should be assessed within the context of serotype-specific population analyses. High-resolution pangenomic analyses, coupled with epidemiological and source information, are likely to be particularly informative for serotypes linked to multiple sources and transmission pathways or for clonal lineages that exhibit limited variability in their core genome [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e], such as \u003cem\u003eS. enterica\u003c/em\u003e serotype Hadar (herein referred to as Hadar).\u003c/p\u003e \u003cp\u003eHadar is transmitted to people via contaminated food and contact with animals and has caused several United States (U.S.) outbreaks in the last decade, linked to either ground turkey consumption or contact with backyard poultry (i.e., privately-owned, non-commercial poultry such as chickens, ducks, or turkeys) [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Although Hadar is considered a highly clonal serotype, exhibiting limited variability by core-genome multilocus sequence typing (cgMLST) [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], strains transmitted by these two different sources were historically differentiable (allele range 25\u0026ndash;50). However, in 2020, despite decreased reporting of enteric illness during the early years of the coronavirus disease 2019 (COVID-19) pandemic, an emergent Hadar strain was linked with both ground turkey consumption and backyard poultry contact. These outbreaks resulted in \u0026gt;\u0026thinsp;900 human illnesses compared to \u0026lt;\u0026thinsp;500 total reported cases of Hadar in all years prior to 2020 [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Traceback investigations were not able to determine the epidemiological connection suggested by the detection of indistinguishable strains (determined by cgMLST) from two ostensibly distinct sources: commercial poultry and backyard poultry [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. This emergent strain, now responsible for \u0026gt;\u0026thinsp;2000 human illnesses, continues to cause outbreaks into 2024; it has been designated by the U.S. Centers for Disease Control and Prevention (CDC) as a Reoccurring, Emerging, or Persisting (REP) strain REPTDK01, with a cgMLST range of 0\u0026ndash;26 allele differences [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eGiven the limitations in discriminatory power of cgMLST for this strain, we employed \u003cem\u003ek\u003c/em\u003e-mer-based Jaccard Index (JI) to compute pangenome relatedness [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] of Hadar along the U.S. farm-to-fork continuum (Farm-to-Fork Continuum). We explored and assessed the value of the pangenome for delineating strains, for attributing human cases to transmission vehicles, and for a general understanding of the epidemiological and evolutionary dynamics that underpin Hadar disease incidence and environmental persistence. In addition, we built a foundational landscape of the vertical and horizontal diversity and dynamics of this serotype and offer support for the incorporation of the accessory genome for differentiating strains transmitted via different pathways.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eData collection\u003c/p\u003e \u003cp\u003eA total of 3384 U.S. Hadar genomes were included in this analysis (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e), collected between 1990 and 2023 (August 30th ) from national surveillance systems and ad hoc sampling. Hadar genomes from ill humans with exposure information available were categorized as follows: \u0026ldquo;backyard poultry contact\u0026rdquo; \u0026ndash; when contact was confirmed within seven days of illness onset (contact is defined as direct interaction with chickens, ducks, turkeys, geese, guinea fowl, or quail; direct contact with the environment where backyard poultry live and roam; consumption of eggs or meat obtained from backyard poultry; or residence with a household member who directly interacted with backyard poultry) [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e], \u0026ldquo;turkey consumption\u0026rdquo; \u0026ndash; where ground turkey was consumed within seven days of illness onset, and \u0026ldquo;unknown\u0026rdquo; \u0026ndash; where exposure information was not available, or when neither backyard poultry contact nor turkey consumption was reported. Genomes from non-human sources were categorized according to the commodity from which they were sampled, for example, \u0026ldquo;commercial poultry\u0026rdquo; or \u0026ldquo;swine\u0026rdquo;. \u0026ldquo;Other\u0026rdquo; was used to categorize samples from unknown food, animal, or environmental source types.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eNational surveillance systems\u003c/h2\u003e \u003cp\u003eSalmonellosis is a nationally notifiable disease in the United States, and isolates obtained from patients are routinely submitted to public health laboratories (PHLs) as part of the CDC\u0026rsquo;s national enteric disease surveillance network, PulseNet USA [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Since 2019, PHLs have performed whole genome sequencing (WGS) on all \u003cem\u003eSalmonella\u003c/em\u003e isolates they receive and upload sequence data to a centralized national database for genetic analysis, including computed serotyping [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e], and to the National Center for Biotechnology Information (NCBI) under the BioProject PRJNA230403. Additionally, public health departments routinely collect demographic information for all laboratory-confirmed cases of salmonellosis. For cases included in multistate outbreak investigations, public health officials conduct additional patient interviews, whenever possible, with Supplementary standardized questionnaires to obtain further details about foods eaten and animal contact before illness onset [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Approximately 5% of isolates detected by PHL also fall within the CDC arm of the National Antimicrobial Resistance Monitoring System (NARMS), a structured collection of enteric isolates from all 50 U.S. states used to monitor temporal trends in antimicrobial resistance (AR) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cdc.gov/narms/index.html\u003c/span\u003e\u003cspan address=\"https://www.cdc.gov/narms/index.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). CDC NARMS has been routinely generating WGS data for this smaller subset of \u003cem\u003eSalmonella\u003c/em\u003e isolates since 2016. WGS data for 2494 Hadar isolates collected between January 1st \u003csub\u003e,\u003c/sub\u003e 2016, and August 30th, 2023, were included in this analysis (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). For years prior to routine WGS (2005\u0026ndash;2015), all Hadar isolates in PulseNet USA\u0026rsquo;s national database with WGS data available were included (n\u0026thinsp;=\u0026thinsp;55); these represent a small proportion of total isolates collected from this time period that were sequenced for various special interest projects.\u003c/p\u003e \u003cp\u003eThe U.S. Food and Drug Administration (FDA) arm of NARMS routinely collects WGS data on \u003cem\u003eSalmonella\u003c/em\u003e isolated from retail meats (chicken, ground turkey, ground beef, pork) purchased from U.S. grocery stores (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.fda.gov/animal-veterinary/national-antimicrobial-resistance-monitoring-system/about-narms\u003c/span\u003e\u003cspan address=\"https://www.fda.gov/animal-veterinary/national-antimicrobial-resistance-monitoring-system/about-narms\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Sequencing data and source information are uploaded to the NCBI under the BioProject PRJNA292661. The following NCBI Pathogen Detection query (August 30th, 2023) identified 300 Hadar genomes (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) that were included in this analysis: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/pathogens/isolates/#PRJNA292661%20AND%20Hadar\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/pathogens/isolates/#PRJNA292661%20AND%20Hadar\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eThe U.S. Department of Agriculture\u0026rsquo;s Food Safety and Inspection Service (USDA-FSIS) routinely collects WGS data on \u003cem\u003eSalmonella\u003c/em\u003e isolated from regulated food and animal products within U.S. food processing facilities (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.fsis.usda.gov/science-data/sampling-program/sampling-results-fsis-regulated-products\u003c/span\u003e\u003cspan address=\"https://www.fsis.usda.gov/science-data/sampling-program/sampling-results-fsis-regulated-products\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Sequencing data and source information are uploaded under NCBI BioProject PRJNA242847. Additionally, the USDA-FSIS arm of NARMS routinely collects WGS data from \u003cem\u003eSalmonella\u003c/em\u003e isolated from the intestinal content of food animals at slaughter (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.fsis.usda.gov/science-data/national-antimicrobial-resistance-monitoring-system-narms\u003c/span\u003e\u003cspan address=\"https://www.fsis.usda.gov/science-data/national-antimicrobial-resistance-monitoring-system-narms\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and data is uploaded under NCBI BioProject PRJNA292666. An August 30th, 2023 NCBI Pathogen Detection query identified 367 Hadar genomes from USDA-FSIS product sampling and 102 from NARMS sampling (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) for inclusion in this study: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/pathogens/isolates/#Hadar%20AND%20collected_by:USDA-FSIS\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/pathogens/isolates/#Hadar%20AND%20collected_by:USDA-FSIS\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eAd hoc sampling systems\u003c/h3\u003e\n\u003cp\u003eTo expand source type representation along the farm-to-fork continuum, Hadar genomes isolated from North America were included from ad hoc sampling systems. The FDA\u0026rsquo;s Office of Regulatory Affairs (ORA), Center for Food Safety and Applied Nutrition (CFSAN), and Center for Veterinary Medicine (CVM) perform ad hoc WGS on human food and animal food (including imported) product sampling and upload sequencing data to the GenomeTrakr project at NCBI (BioProject PRJNA186035). Twenty genomes (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) collected between 2003 and 2022 were selected and included in this analysis. An additional nine isolates representing all sequenced Hadar collected from sick animals as part of FDA-CVM\u0026rsquo;s Veterinary Laboratory Investigation and Response Network (Vet-LIRN) AMR monitoring program were also included.\u003c/p\u003e \u003cp\u003eUSDA\u0026rsquo;s Animal and Plant Health Inspection Service (APHIS) provides ongoing animal disease surveillance and animal disease diagnostic services through the National Veterinary Services Laboratories (NVSL; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.aphis.usda.gov/labs/about-nvsl\u003c/span\u003e\u003cspan address=\"https://www.aphis.usda.gov/labs/about-nvsl\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and the National Animal Health Laboratory Network (NAHLN; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.aphis.usda.gov/labs/nahln\u003c/span\u003e\u003cspan address=\"https://www.aphis.usda.gov/labs/nahln\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Thirty-two Hadar genomes (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) collected from chickens or turkeys from 2018 until 2023 as part of on-farm monitoring or for diagnostic purposes were included in this analysis. Three Hadar genomes previously sequenced and published by USDA\u0026rsquo;s Agricultural Research Service (ARS) [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e], and two Hadar genomes collected from wild ducks by the National Wildlife Health Center were also included (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Additional Hadar genomes were available on NCBI, but source information availability (through NCBI or personal communication) was a requirement for inclusion in this analysis.\u003c/p\u003e\n\u003ch3\u003eNon-U.S. genomes\u003c/h3\u003e\n\u003cp\u003eA dataset of global non-U.S. Hadar genomes was generated from EnteroBase [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e] for comparative analysis against the pangenome of the U.S. collection. All genomes with predicted serotype \u0026ldquo;Hadar\u0026rdquo; (EnteroBase employs SISTR1 [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] and SeqSero2 [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]) isolated in any country other than the U.S. were downloaded (n\u0026thinsp;=\u0026thinsp;1145) (accessed December 21st, 2023) (Supplementary Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eGenomic analysis\u003c/p\u003e \u003cp\u003eShort reads with a base call quality score\u0026thinsp;\u0026ge;\u0026thinsp;28 and coverage\u0026thinsp;\u0026ge;\u0026thinsp;40x were assembled using shovill v.1.0.9 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/tseemann/shovill\u003c/span\u003e\u003cspan address=\"https://github.com/tseemann/shovill\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and resulting contigs with \u0026lt;\u0026thinsp;10% of the average genome coverage were excluded from the final assemblies. Serotype was confirmed using SeqSero 2.0 v1.2.1 [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e], sequence type (ST) was determined using mlst (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/tseemann/mlst\u003c/span\u003e\u003cspan address=\"https://github.com/tseemann/mlst\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), core SNP (single nucleotide polymorphism) cluster was obtained from NCBI Pathogen Detection\u0026rsquo;s Isolate Browser (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/pathogens/\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/pathogens/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), and allele code was calculated using cgMLST [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. \u0026ldquo;Condensed allele code,\u0026rdquo; which collapses allele codes to the third digit (e.g., \u003cem\u003eSalmonella\u003c/em\u003e spp. Allele codes SALM1.0\u0026ndash;6771.1.1.30.1.21 and SALM1.0\u0026ndash;6771.1.1.30.1.44 would be collapsed into SALM1.0\u0026ndash;6771.1.1), was used to simplify representation of allele codes. Genomes of the same condensed allele code are expected to differ by less than ~\u0026thinsp;15 allele loci (Williams \u003cem\u003ein preparation\u003c/em\u003e). Accessory (non-core) genome elements were detected using PanGraph (see \u003cem\u003ePangenome characterization\u003c/em\u003e) [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] and characterized using PlasmidFinder [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e] (90% identity, 60% gene coverage) for plasmid replicons, MOBscan [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] for conjugative relaxases, COPLA [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e] for Plasmid Taxonomic Unit (PTU) designation [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e], and Bakta [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e] for gene annotation. AR determinants, including acquired genes and chromosomal mutations, were detected using staramr v.0.4.0 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/phac-nml/staramr?tab=readme-ov-file#mlsttsv\u003c/span\u003e\u003cspan address=\"https://github.com/phac-nml/staramr?tab=readme-ov-file#mlsttsv\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), which employs the ResFinder database (updated 30JUL2020; 90% identity, 50% gene coverage) and the \u003cem\u003eSalmonella\u003c/em\u003e spp. PointFinder scheme [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]; predicted AR was determined by staramr according to ResFinder and PointFinder results. Assignment of draft Illumina contigs to plasmids or chromosomes was performed using MOB suite [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eLong-read sequencing was performed on 35 selected isolates from each Jaccard Index (JI) group (see \u003cem\u003eJaccard Index calculation\u003c/em\u003e), chosen strategically to maximize connectivity to other internal nodes and to best achieve JI-group representation. Eighteen Hadar isolates from people or food products were sequenced on the Oxford Nanopore GridION sequencing platform (Supplementary Table S3); reads were assembled using an in-house pipeline, as previously described [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Seventeen isolates collected from food or animal samples were sequenced using the 10-kb SMRTLink template preparation protocol (Pacific BioSciences, CA), as previously described [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Complete genomes were annotated by Bakta [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Long-read data are uploaded under BioSample numbers listed in Supplementary Table S3. An additional 18 previously published Hadar genomes [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e] were also included in the analysis (Supplementary Table S3).\u003c/p\u003e \u003cp\u003eJaccard Index calculation\u003c/p\u003e \u003cp\u003eThe exact JI was used as a measure of similarity between all genome pairs as previously reported [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Briefly, each complete genome assembly was converted into a set of \u003cem\u003ek\u003c/em\u003e-mers. JI was calculated as the ratio of shared \u003cem\u003ek\u003c/em\u003e-mers over the total number of different \u003cem\u003ek\u003c/em\u003e-mers between the two sets (including shared \u003cem\u003ek\u003c/em\u003e-mers, SNP \u003cem\u003ek\u003c/em\u003e-mers differing by a single base pair, and indel \u003cem\u003ek\u003c/em\u003e-mers differing between the datasets and excluding duplicated \u003cem\u003ek\u003c/em\u003e-mers). BinDash [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]⁠ was employed to calculate JI, using parameters minhashtype = -1 (to compute the exact JI between highly similar genomes using the complete set of \u003cem\u003ek\u003c/em\u003e-mers, rather than an estimated JI based on a subset of \u003cem\u003ek\u003c/em\u003e-mers) and \u003cem\u003ek\u003c/em\u003e-mer length (k)\u0026thinsp;=\u0026thinsp;21 (as previously defined as optimum in [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]).\u003c/p\u003e \u003cp\u003eNetwork visualization and community detection\u003c/p\u003e \u003cp\u003eThe adjacency matrix of pairwise genome similarities generated by BinDash was used to construct an undirected network. Gephi v10 [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e] was employed to visualize the network, using the ForceAtlas2 algorithm for the layout. To define the final components for study, referred to as JI-groups, a range of JI thresholds was assessed, and network sparsification was optimized according to transitivity and density, as previously described [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Transitivity plateaued between JI 0.986 and 0.990. The final JI threshold was set in the middle of this range, at 0.988, balancing the density of communities (defined by a minimum of five genomes) and the number of singletons. The Louvain method, implemented in Gephi, was used to define the JI-groups by using resolution 1.5. Once the main JI groups are defined (containing a minimum of five genomes), they can be further dissected into several subgroups within the network using a more stringent JI and the same community detection algorithm [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. The nodes of the network, representing genomes, were colored according to metadata and genetic determinants of interest. Edges between nodes were included whenever the corresponding JI value met or exceeded the user-defined threshold. Network figures were generated using the igraph package in R (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://r.igraph.org/articles/igraph.html\u003c/span\u003e\u003cspan address=\"https://r.igraph.org/articles/igraph.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003cp\u003ePangenome characterization\u003c/p\u003e \u003cp\u003ePanGraph [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] identifies blocks of homologous sequence and was used to detect indels specific to each JI-group. PanGraph was run on all genomes using parameters `α\u0026thinsp;=\u0026thinsp;20` and `β\u0026thinsp;=\u0026thinsp;20`. The parameter α controls the cost of splitting a block into smaller units, where a value of 20 was chosen to minimize excessive fragmentation of the graph. The parameter β controls the diversity cost and was set to 20, establishing a sequence diversity threshold of 20%. Only homologous sequences (pancontigs), larger than 250 bp, present in \u0026ge;\u0026thinsp;85% of the members of each JI-group and not present in all JI-groups, were retained as \u0026ldquo;core\u0026rdquo; pancontigs. Core pancontigs for each JI-group were mapped with BLASTn against a reference genome (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) from their respective JI-group (sequenced by long-read technology, when available) to order the pancontigs and detect the regions they form. For instance, a prophage might be composed of several pancontigs, and scaffolding those contigs against a reference genome helped reconstruct and identify that element as an indel. The term \u0026ldquo;prophage\u0026rdquo; was used to refer to chromosomally-integrated regions that contained at least five phage-related genes according to PhageScope [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e] or PHASTEST [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eFor pangenome comparison between U.S. and non-U.S. datasets, gene prediction of the assembled genomes was performed with Prokka [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. Annotated assemblies in GFF3 format were used as input for pangenome calculation using Roary v3.13 [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e], with 80% minimum percent identity and coverage. Pangenome gene categories were defined as: core genes (shared by 80\u0026ndash;100% of the genomes); shell genes (15\u0026ndash;79%); and cloud genes (0\u0026ndash;14%). Heaps\u0026rsquo; law was used to evaluate pangenome openness and closeness, using the script available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/SethCommichaux/Heap_Law_for_Roary\u003c/span\u003e\u003cspan address=\"https://github.com/SethCommichaux/Heap_Law_for_Roary\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003cp\u003ePhylogenetic analysis\u003c/p\u003e \u003cp\u003ecgMLST-based phylogenetic trees were generated using BioNumerics v7.6.3 [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Snippy v4.4.5 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/tseemann/snippy\u003c/span\u003e\u003cspan address=\"https://github.com/tseemann/snippy\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was used to detect core genome single nucleotide polymorphisms (cg-SNPs) in three datasets: JI-C chromosomes, JI-C PTU-I1 plasmids, and PTU-I1 plasmids from JI-E, JI-G, and other enterobacteria (RefSeq200). In all cases, the PTU-I1-containing genome SAL-20-VL-OH-OSU-0008 was used as reference. Alignments generated with Snippy were used to construct maximum-likelihood (ML) phylogenetic trees based on cg-SNPs by using IQ-TREE [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. All trees generated in this study were rooted at midpoint and visualized with iTol v6 [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. To complement cg-SNP analysis of PTU-I1 plasmids, AcCNET (Accessory Genome Constellation Network [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]) was used to build proteome networks and assess relatedness of plasmids at the protein level; proteins were clustered if they shared greater than 80% identity and 80% coverage.\u003c/p\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eStatistical analyses were performed using genomes collected through NARMS (CDC, FDA, FSIS), PulseNet (CDC), and FSIS national surveillance systems from years 2016 through 2023, in line with the introduction of routine sequencing for NARMS, PulseNet, and FSIS surveillance isolates. Corrected Cramer\u0026rsquo;s \u003cem\u003eV\u003c/em\u003e was used to measure strength of associations between all categorical variables [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]; chi-squared tests of independence were used to test associations between specific epidemiological and genomic variables (Bonferroni adjusted significance value: p\u0026thinsp;\u0026lt;\u0026thinsp;0.005); odds ratios (OR) (95% confidence intervals (CI)) were used to quantify the strength and direction of significant associations. For statistical tests involving a specific JI-group, the comparison group was always \u0026ldquo;all other JI-groups\u0026rdquo;. All tests were calculated using the stats subpackage of SciPy v1.14.1 implemented in Python v3.11.7 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://docs.scipy.org/doc/scipy/reference/stats.html\u003c/span\u003e\u003cspan address=\"https://docs.scipy.org/doc/scipy/reference/stats.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). JI-groups with less than 20 genomes were not analyzed for statistical associations. Only NARMS surveillance data collected by CDC, FDA, and FSIS (cecal sampling) were used to assess shifts in pangenome group abundance over time, as the isolates in the NARMS dataset were systematically collected and were more robust against large outbreaks and changes to regulatory testing practices than were the surveillance isolates from the PulseNet and FSIS product sampling datasets.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003ePangenome structure of United States Hadar population\u003c/p\u003e \u003cp\u003eHadar genomes self-organized into 18 clusters by JI (JI threshold\u0026thinsp;=\u0026thinsp;0.988), labeled JI-A through R (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e); less than 5% of genomes (n\u0026thinsp;=\u0026thinsp;158/3387) did not cluster with a JI-group and were considered singletons (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The three largest groups JI-A, JI-B, and JI-C were further divided into subgroups using an increased JI threshold (Supplementary Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). JI-A subgroups A1-15 were defined at JI\u0026thinsp;=\u0026thinsp;0.995; JI-B subgroups B1-6 and JI-C subgroups C1-9 were defined at JI\u0026thinsp;=\u0026thinsp;0.992. The MGE that define each JI-group include large plasmids (\u0026gt;\u0026thinsp;30 kb), prophages, AR regions, or regions of unknown function (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). In some cases, two JI-groups differed only by the presence of a large plasmid (e.g., JI-A and JI-C; JI-B and JI-G; JI-D and JI-E), while others displayed more differences in their pangenome content (e.g., JI-I) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eST (sequence type, based on 7 core loci), NCBI SNP cluster [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e], and cgMLST allele code (based on n\u0026thinsp;=\u0026thinsp;3002 core loci) [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e] were separately visualized on the network to contextualize the pangenome with core lineage information. Over 98% of Hadar genomes in this analysis are ST33 (n\u0026thinsp;=\u0026thinsp;3326/3384); only JI-I (ST473), JI-L (ST5130 and ST9222), and JI-Q (ST473) contained genomes of a different ST (Supplementary Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003ea). NCBI SNP cluster aligned well with JI-groups; PDS000158107 was the most common cluster, encompassing the largest groups JI-A, JI-B and JI-C (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). cgMLST allele codes also aligned well with JI-groups, with the majority of groups (n\u0026thinsp;=\u0026thinsp;12/18) containing a single condensed allele code (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). Despite being in the same NCBI SNP cluster, JI-A and JI-C separate from JI-B by condensed allele code (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). Both NCBI SNP cluster and cgMLST suggest membership within certain JI-groups is due to convergence in pangenome content rather than core genome similarity.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003ePlasmids were common in U.S. Hadar genomes, with 60% (n\u0026thinsp;=\u0026thinsp;2047/3384) containing one or more Col-like plasmids and 22% (n\u0026thinsp;=\u0026thinsp;740/3387) carrying at least one large (\u0026gt;\u0026thinsp;30 kb) conjugative plasmid (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec, Supplementary Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003eb). IncI1 was the most common replicon, detected in three different PTUs: PTU-I1, present in JI-C and JI-E, newly identified PTU-NA (IncI1, MOB\u003csub\u003eP\u003c/sub\u003e) present in JI-J and JI-N, and newly identified PTU-NA (IncI1, MOB\u003csub\u003eP\u003c/sub\u003e), present in JI-I (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). JI-I also contained PTU-E78, a recently identified non-mobilizable PTU (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). However, nearly 30% of genomes (n\u0026thinsp;=\u0026thinsp;1011/3384) contained neither plasmid replicons nor MOB relaxase genes (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e); these genomes predominantly fell into JI-A (Supplementary Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003eb). PanGraph analysis revealed that integrated MGEs were also common in several JI-groups, including prophages and integrated conjugative elements (ICEs) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and Supplementary File 1).\u003c/p\u003e \u003cp\u003eOver 90% (n\u0026thinsp;=\u0026thinsp;3055/3384) of genomes contained at least one AR determinant; predicted resistance to aminoglycosides (specifically, streptomycin) and tetracyclines was the most common profile, mediated by \u003cem\u003eaph(3'')-Ib\u003c/em\u003e, \u003cem\u003eaph(6)-Id\u003c/em\u003e, and \u003cem\u003etet(A)\u003c/em\u003e), all integrated in the chromosome (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed, Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Predicted resistance to penicillins was less common (4%, n\u0026thinsp;=\u0026thinsp;127/3384) and was predominantly mediated by \u003cem\u003ebla\u003c/em\u003e\u003csub\u003eTEM\u0026minus;1\u003c/sub\u003e (Supplementary Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003ec and S2d). While rare, cephalosporin resistance mediated by \u003cem\u003ebla\u003c/em\u003e\u003csub\u003eCMY\u0026minus;2\u003c/sub\u003e was detected in groups JI-C and JI-E (0.4%, n\u0026thinsp;=\u0026thinsp;12/3384; Supplementary Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003ec and S2d). Members of JI-D, JI-I and JI-Q were predicted to be pansusceptible, with no known AR determinants detected (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eGenetic and epidemiological differences between most abundant pangenome groups\u003c/p\u003e \u003cp\u003eThe dominant pangenome groups changed substantially between 2016 and 2023, most notably between 2019 and 2020 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, Supplementary Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003ee and S2f). This shift was particularly pronounced for human and retail meat samples, where JI-A and JI-C were rare prior to 2020 yet comprise between 56% and 100% of samples collected in years 2020\u0026ndash;2023. JI-B was the most common group detected in retail meat and animal (cecal) sampling prior to 2020 but decreased in detection substantially in 2020\u0026ndash;2023; JI-B was not detected at all in 2023 retail meat sampling. Groups JI-D and JI-E made up more than half of human Hadar samples in 2016 and 2017 but have not been detected since 2019; these groups were not found in retail meat or animal sampling throughout the study years (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). JI-A and JI-C are the most common JI-groups in all three sampling systems from 2020\u0026ndash;2023.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eJI-A and JI-C are indistinguishable by cgMLST-based phylogeny (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e: Ring 1) but differ in their pangenome by carriage of a\u0026thinsp;~\u0026thinsp;100 kb PTU-I1 (IncI1) plasmid, which underpins the separation of these two JI-groups (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). Most JI-A and JI-C genomes fall within a comparatively tight \u0026ldquo;emergent\u0026rdquo; clade that forms the CDC-defined REPTDK01 strain (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef, Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), associated with ground turkey consumption and backyard poultry contact based on previous multistate outbreak investigations [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Of interest, two temporally \u0026ldquo;ancestral\u0026rdquo; JI-A genomes isolated from wild ducks in 1990 are positioned in a clade adjacent to the emergent genomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This emergent clade invariably contains an ~\u0026thinsp;8 kb prophage, labeled here prophage 1 (Supplementary File 1), that forms part of the core pangenome of JI-A and JI-C (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Prophage 1 was detected as early as 2004 in singleton Hadar genomes (imported \u0026ldquo;sweet good without custard or cream filling\u0026rdquo; from Pakistan), was seen in genomes from swine and commercial poultry samples from 2015, yet remained uncommon until the 2020 emergence of REPTDK01 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, Supplementary Fig. S3). According to PHASTEST, prophage 1 is related to filamentous phages I2-2 and Ike, and contains a protein with N-terminal homology to the zonular occludens toxin protein (Zot) (Supplementary Fig. S4). The phage-encoded Zot proteins in \u003cem\u003eVibrio cholerae\u003c/em\u003e [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e] and \u003cem\u003eCampylobacter\u003c/em\u003e spp. [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e] have a demonstrated pathogenic role attributable to a C-terminal enterotoxic domain [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. While homology with Zot proteins does not imply toxigenic function, the Hadar Zot-like protein identified here was bioinformatically predicted to contain toxigenic regions using ToxinPred3.0 [\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e], hinting at a putative role in pathogenesis. Thus, prophage 1 presence is notable both from an epidemiological and biological perspective, and its pathogenic and adaptive capacity is being assessed with functional analysis.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eJI-B is the second most abundant pangenome group, predominantly encompassing genomes from commercial poultry (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee). A smaller group, JI-G, is indistinguishable from JI-B phylogenetically (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e: Ring 1) but can be differentiated by the presence of PTU-I1 plasmids (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). JI-B (and JI-G) genomes appear more diverse in their core genome relative to those from other dominant pangenome groups (e.g., JI-A, JI-C, JI-D and JI-E) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), which may be a reflection of time and environmental factors\u0026mdash;genomes in JI-B were isolated as early as 2011 from poultry sources across the country (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Analysis of JI-B subgroups did not reveal any geographic association (Supplementary Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eb) or link to specific processing facilities. Of note, genomes from human samples that were part of a 2019 multistate Hadar outbreak linked to ground turkey consumption (internal CDC investigation) all fell into JI-B or JI-G, suggesting Hadar strains from these groups are transmitted via food.\u003c/p\u003e \u003cp\u003eIn contrast, groups JI-D and JI-E were almost always from ill humans (rather than animal or meat samples) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee), often with reported contact with backyard poultry (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). JI-D and JI-E genomes display relatively little core diversity (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) and differ from each other only by the carriage of PTU-I1 (IncI1) plasmids (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). They differ from other JI-groups phylogenetically in that they are encompassed in a single clade by core SNP analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e); and they differ pangenomically in that they lack a common AR region (\u0026ldquo;AMR-encoding Tn 1.1\u0026rdquo;, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) and are the only groups to carry PTU-X1 (IncX1) plasmids (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Genomes in these groups were part of 2016 and 2017 multistate outbreaks linked to contact with backyard poultry [\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eTwo small pangenome groups, JI-H and JI-K, are of interest because of their connectivity to JI-A in the network, indicating pangenomic relatedness (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). JI-H genomes are all from commercial chicken sampling or from ill humans (no exposure information available), representing a statistically significant \u0026ldquo;chicken-source cluster\u0026rdquo; (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e; p\u0026thinsp;\u0026lt;\u0026thinsp;0.00001, chi-squared) that is unique among the more common commercial turkey source. JI-K genomes were all isolated throughout 2023, are almost exclusively from turkey product samples (n\u0026thinsp;=\u0026thinsp;11/12) and are predominantly from a single state (n\u0026thinsp;=\u0026thinsp;8/12 were isolated in CA) (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). JI-K genomes carry prophage 1, along with two other larger prophages unique to this group (prophage 6.2 and prophage 10; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), potentially representing recent divergence from REPTDK01.\u003c/p\u003e \u003cp\u003eSeveral pangenome groups harbor PTU-I1 (IncI1) plasmids, including JI-C, JI-E and JI-G (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). PTU-I1 (IncI1) plasmids are common in avian environments, often carry AR genes, and may play a role in virulence and growth inhibition of competing bacteria [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e]; thus, their presence and diversity in this dataset were of interest. Core plasmid SNP analysis coupled with AcCNET plasmid proteome analysis were used to assess the relatedness of PTU-I1 (IncI1) plasmids between and within JI-groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). PTU-I1 plasmids from all three JI-groups were surprisingly diverse in their core and proteome and intermingled phylogenetically with PTU-I1 plasmids from other Enterobacteriaceae species (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb, Supplementary Table S5). Plasmids from the same JI-C subgroups clustered together phylogenetically (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea and Supplementary Fig. S5a) and proteomically (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ec), indicating that plasmid content is responsible for JI-C subgrouping. Of note, the largest JI-C subgroup, JI-C1, likely represents a multiyear clonal expansion event, given the tight relatedness of its plasmids and chromosomal genome (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea and Supplementary Fig. S5).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eHadar pangenome offers increased discriminatory power for retrospective and prospective public health investigations\u003c/p\u003e \u003cp\u003eREPTDK01 was clearly detectable in the pangenome network\u0026mdash;98% (n\u0026thinsp;=\u0026thinsp;2148/2194) of these genomes fell into JI-A, JI-C, JI-N and JI-R (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef, Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e)\u0026mdash;genetically corroborating and adding confidence to the REPTDK01 definition using pangenomic data. Additionally, REPTDK01 was further stratified by JI-grouping and JI-subgrouping, revealing clear epidemiological patterns. For example, while JI-A itself was not statistically associated with either commercial or backyard poultry (Supplementary Table S4), JI-A2 contained predominantly commercial poultry-related genomes from the U.S. and Canada (n\u0026thinsp;=\u0026thinsp;42/68), and none of the human clinical cases in this group (n\u0026thinsp;=\u0026thinsp;24/68) reported backyard poultry contact. In contrast, JI-A3 was almost exclusively genomes from human clinical cases (n\u0026thinsp;=\u0026thinsp;27/28), a third of which reported backyard poultry contact, and zero commercial poultry-related genomes fell into this group (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e and Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ea). JI-N genomes were all human clinical\u0026mdash;mostly isolated from the northeast (n\u0026thinsp;=\u0026thinsp;4/6)\u0026mdash;and may represent a closely related subcluster of illnesses that differ from JI-A REPTDK01 strains only by the carriage of a large plasmid (PTU-NA, IncI1) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). JI-C was significantly associated with backyard poultry (p\u0026thinsp;\u0026lt;\u0026thinsp;0.00001, chi-squared; Supplementary Table S4), representing a subgroup of REPTDK01 (defined by the carriage of PTU-I1 plasmids) that was likely transmitted to humans via animal contact rather than food. More specifically, epidemiological traceback data available for clonal subgroup JI-C1 disproved the involvement of a single backyard poultry supply store chain or hatchery, instead suggesting a common reservoir of Hadar upstream of hatcheries. Coupling pangenome data and epidemiological data, REPTDK01 strains can be further differentiated for both retrospective and prospective investigations.\u003c/p\u003e \u003cp\u003eSeveral other non-REPTDK01 pangenome groups are statistically associated with a specific source or exposure. JI-B and JI-G are each significantly associated with commercial turkey (p\u0026thinsp;\u0026lt;\u0026thinsp;0.00001, chi-squared); JI-B genomes had 17.5 times (95% CI: 13.7\u0026ndash;22.3), and JI-G genomes had 5.9 times (95% CI: 2.1\u0026ndash;17.1) higher odds of being from commercial turkey compared with all other JI-groups (Supplementary Table S4). Coupled with the absence of human cases reporting backyard poultry contact in these groups, it is likely that Hadar strains from JI-B and JI-G are acquired through foodborne transmission. In contrast, JI-D and JI-E were each significantly associated with backyard poultry contact (p\u0026thinsp;\u0026lt;\u0026thinsp;0.00001, chi-squared); JI-D genomes had 2.6 times (95% CI: 1.9\u0026ndash;3.6) and JI-E genomes had 5.2 times (95% CI: 2.7\u0026ndash;10.6) greater odds of backyard poultry contact, relative to all other JI-groups (Supplementary Table S4). The stark lack of genomes from commercial poultry sources (only JI-D had a single commercial chicken source genome), and the predominance of backyard poultry-associated outbreak genomes in these groups (n\u0026thinsp;=\u0026thinsp;140/191 in JI-D, n\u0026thinsp;=\u0026thinsp;35/40 in JI-E), strongly suggests JI-D and JI-E strains of Hadar are transmitted through animal contact. It is important to note that cgMLST differentiates JI-B and JI-G genomes from JI-D and JI-E (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). Thus, the pangenome analysis performed here provides additional genomic confidence in these attributions.\u003c/p\u003e \u003cp\u003eA handful of small JI-groups contained genomes from humans with limited epidemiological information, but with one or two genomes from a known source. Specifically, both JI-F and JI-J contain a genome from raw dog food (containing duck) obtained from ad hoc pet food sampling (see Methods) (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). JI-L contains two genomes from imported shrimp (Ecuador) isolated in 2022 (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Given the close relatedness of genomes within JI-groups (median average nucleotide identity within JI-groups is \u0026ge;\u0026thinsp;99.95, Supplementary Fig. S6), the presence of the pet food and imported food genomes alongside genomes from human samples is suggestive of an epidemiological connection, though without exposure information reported by these ill people this link cannot be confirmed. Prospectively, the relatedness of additional human cases found to be within the JI-F and JI-J groups could inform which food items to assess during supplementary interviews of ill people included in an outbreak investigation.\u003c/p\u003e \u003cp\u003eAs mentioned above, several pairs of JI-groups differ only by the presence of PTU-I1 (IncI1) plasmids: JI-A and JI-C (plasmid present), JI-D and JI-E (plasmid present), JI-B and JI-G (plasmid present). We further assessed these pairs for epidemiological patterns associated with plasmid presence, including source of isolation, geographic region, and patient demographics (age, sex, site of infection, hospitalization), but no variables were significantly different between paired groups (\u003cem\u003eV\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.3, corrected Cramer\u0026rsquo;s \u003cem\u003eV\u003c/em\u003e; p\u0026thinsp;\u0026gt;\u0026thinsp;0.005, chi-squared). However, PTU-I1 (IncI1) plasmids were independently associated with backyard poultry-related sources (PTU-I1 n\u0026thinsp;=\u0026thinsp;208, no PTU-I1 n\u0026thinsp;=\u0026thinsp;526) when compared with commercial poultry sources (PTU-I1 n\u0026thinsp;=\u0026thinsp;32, no PTU-I1 n\u0026thinsp;=\u0026thinsp;699), and when compared with all other sources (PTU-I1 n\u0026thinsp;=\u0026thinsp;305, no PTU-I1 n\u0026thinsp;=\u0026thinsp;2345) (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e; p\u0026thinsp;\u0026lt;\u0026thinsp;0.00001, chi-squared). Thus, PTU-I1 plasmids have statistical support to serve as a genetic marker to distinguish strains transmitted via backyard poultry contact versus those more likely attributed to another source, which is of particular value for differentiating REPTDK01 strains that can be transmitted via several pathways.\u003c/p\u003e \u003cp\u003eU.S. Hadar pangenome structure reflects a subset of global diversity\u003c/p\u003e \u003cp\u003eA dataset of Hadar genomes (n\u0026thinsp;=\u0026thinsp;1145) from 33 countries other than the U.S., isolated from 1950 through 2023, was used to assess differences in pangenome structure between separate geographical locations (Supplementary Fig. S7, Supplementary Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). The non-U.S. dataset partially overlapped with U.S. genomes: 33% of non-U.S. genomes clustered within JI-groups identified in the U.S. pangenome data, while 47% formed distinct JI-groups not present in the U.S. dataset (Supplementary Fig. S8, Supplementary Table S6). The non-U.S. dataset contained 3095 genes absent from the U.S. pangenome, while the U.S. dataset contained 1628 genes absent from the non-U.S. dataset (Supplementary Fig. S9). Both datasets exhibited moderately open pangenomes (Heaps\u0026rsquo; law γ value\u0026thinsp;~\u0026thinsp;0.2) and shared a core of 4187 genes. Notably, separate analysis of each dataset revealed similar core gene counts, further highlighting the robustness of the core genome across different geographic populations (Supplementary Fig. S9). Furthermore, the non-US dataset displays a larger number of cloud genes, suggesting a higher diversity within its accessory genome.\u003c/p\u003e \u003cp\u003eSeparate analyses of genomes from the United Kingdom (U.K.) (n\u0026thinsp;=\u0026thinsp;484) and France (n\u0026thinsp;=\u0026thinsp;306) were performed since they represented more than half of the non-U.S. genomes. Of 18 JI-groups defined in the U.S. dataset, the U.K. and France datasets shared only seven (170 genomes, 35%) and six (74 genomes, 24%) JI-groups, respectively. Seventeen U.K. JI-groups (228 genomes, 47%) and nine France JI-groups (162 genomes, 53%) were distinct from those isolated in the U.S (Supplementary Fig. S10 and S11, Supplementary Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). While no temporal shift was observed for pangenome groups from U.K. data (Supplementary Fig. S10), a notable increase in genomes belonging to a novel group, JI-S, was observed in the France dataset, beginning in 2019 (Supplementary Fig. S11). JI-S genomes contain a prophage closely related to prophage 1, highlighting an intriguing parallel dynamic to the recent proliferation of prophage 1-containing groups JI-A and JI-C in the U.S. Thus, these analyses suggest Hadar pangenomic diversity is largely geographically defined, with potentially important genetic overlaps that will be further investigated.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIdentifying the molecular mechanisms underlying shifts in bacterial populations is key to understanding the adaptive forces that drive evolution of human bacterial pathogens. Analysis of the Hadar pangenome confirmed known, and revealed unknown, epidemiological and microevolutionary dynamics. Before 2020, two distinct lineages separately dominated in commercial poultry (JI-B and JI-G) and backyard poultry environments (JI-D and JI-E). In 2020, an emergent lineage closely related to previously circulating strains became dominant, displacing the historical commercial poultry lineage. Around the same time, coinciding with a surge in backyard poultry ownership during the COVID-19 pandemic [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e], this same emergent lineage became dominant among backyard poultry-associated human cases \u0026ndash; confirming through high-resolution pangenomic analysis a link between two presumably separate industries. Epidemiological and biological evidence suggest the presence of a novel phage in the emergent lineage may have contributed to its recent expansion. Interestingly, a similar genetic shift underpinned by an emergent phage-containing lineage was seen in the French genomes analyzed here, suggesting this phenomenon is not restricted to the U.S. The adaptive capacity of this prophage in Hadar, and specifically, the putative pathogenic role of the phage-encoded Zot-like protein, is still being evaluated in U.S. Hadar genomes.\u003c/p\u003e \u003cp\u003eThese new findings can be leveraged to mitigate further spread of this emergent strain in a number of ways. First, comparative plasmid analysis revealed a clonal subcluster of this lineage (JI-C1) that points to a reservoir upstream of backyard poultry suppliers and hatcheries, one that likely interfaces with commercial poultry. Backyard poultry hatchery practices, such as drop-shipping and outsourcing to larger commercial hatcheries to meet demand [\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e], could explain this connection. This data can inform conversations between industry and government stakeholders, as it promotes collective action with the goal of eliminating shared reservoirs affecting multiple industries. Second, functional analyses to determine the contribution of prophage 1 to avian gut colonization could inform intervention strategies in both commercial and backyard poultry settings; for example, by minimizing bacterial burden in birds, which is considered a control strategy to reduce risk of transmission to humans [\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e]. Third, this analysis highlighted the importance of known MGE (e.g., PTU-I1 plasmids) and identified previously uncharacterized MGE (e.g., prophage 1) that can potentially be incorporated into source attribution models and molecular case definitions. For example, PTU-I1 plasmids could serve as a genetic marker that distinguishes backyard poultry-related strains from those transmitted via other sources. More accurate prediction of foodborne versus animal contact transmission pathways and refinement of outbreak and REP strain case definitions both contribute to timelier epidemiological traceback, and ultimately, a reduction in human illness [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eMore generally, this analysis enabled high-resolution genomic linking of human cases with potential sources (e.g., pet food, imported shrimp), raising suspicion of specific vehicles to refine supplemental interviews or traceback efforts when exposure information is limited, and no transmission vehicles are otherwise suspected. Additionally, avenues were identified for investigation of ecological dynamics that underpin persistence of Hadar in different environments. For example, PTU-I1 and other large plasmids are associated with backyard poultry rather than commercial poultry environments; and some JI-groups (with unique MGE profiles) display a unique chicken association rather than the more common turkey signal. Along with highlighting the previously unreported role of prophages in Hadar diversification and microevolution, this broad description of MGE in the U.S. Hadar population is foundational information for pathogen risk modeling, especially as it pertains to carriage of AR. The presence of \u0026ldquo;risky\u0026rdquo; MGE related to AR, virulence, or colonization capacity, can be proactively monitored through existing surveillance programs, and any emergent threats addressed before they become systematically disseminated, as has previously occurred with \u003cem\u003eSalmonella\u003c/em\u003e serotypes Infantis [\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e] and Reading [\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eBy nature of the pangenomic approach employed here, exactly when and where this persisting REP strain arose was not determined; however, a molecular clock analysis is underway to explore the rapid rise and subsequent diversification of this lineage. Further, while source of human illnesses with unknown exposures, or those with multiple exposures (e.g., both commercial and backyard poultry), cannot be definitively determined using this approach, the findings from this study will be assessed within ongoing source attribution modeling to estimate the added value of inclusion of accessory genome content. Further, while efforts were made to obtain genomes representing diverse environments (wildlife, imported foods, commercial poultry production, backyard poultry environments, ill humans), several sources are underrepresented (e.g., live animals on farm) or absent (e.g., hatcheries), potentially missing pangenomic groups that are dominant in these spaces. Expanded analyses that include genomes from underrepresented sources, coupled with deeper investigation into the global diversity of Hadar, will fill important gaps in the pangenome landscape described here.\u003c/p\u003e \u003cp\u003eUnraveling pathogen epidemiology and microevolutionary dynamics is highly complex, and the plethora of available data is both an opportunity and a challenge. Leveraging existing genomic data, we demonstrate the value of JI-based pangenomic analysis for delineating a highly clonal serotype and uncover actionable data to mitigate the spread of an emergent, and potentially more pathogenic, lineage of Hadar. We paint a pangenome landscape of this previously understudied serotype, highlighting the importance of known and unknown MGE, and revealing surprising geographic patterns and dynamics. These findings will inform future risk and source attribution modeling, reducing public health burdens and mitigating impacts on implicated food and animal industries.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors would like to acknowledge state and local public health departments and laboratories for isolation and sequencing of Hadar genomes included in this analysis. The authors thank FDA colleagues Olgica Ceric, Beilei Ge, Claudine Kabera, as well as University of Minnesota Professor Timothy Johnson, for their valuable expertise. This work was supported by the Centers for Disease Control and Prevention (Contract No. 75D30123P18303 to FdlC). This work was also supported by the Spanish Ministry of Science and Innovation MCIN/AEI/10.13039/501100011033 (PID2020-117923GB-I00 to FdlC and MPGB).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eKAT, APC, HEW, GSS, MKS, KB, MPGB and FdlC \u0026ndash; Conceptualization.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eKAT, APC, HEW, GSS, ZE, ML, JYK, MS, CL, BH, BRMS, DM, SM, KM, JH, JMW, JMB and KB \u0026ndash; Data collection.\u003c/p\u003e\n\u003cp\u003eKAT, APC, HEW, GSS, MS, CL, BH, BRMS, KB and UD \u0026ndash; Data curation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eKAT, APC, HEW, MKS, SRS, MPGB and FdlC \u0026ndash; Methodology.\u003c/p\u003e\n\u003cp\u003eKAT, APC, HEW, MKS, SRS, and MPGB \u0026ndash; Analysis.\u003c/p\u003e\n\u003cp\u003eKAT, APC, HEW, MKS, SRS, MPGB and FdlC \u0026ndash; Visualization.\u003c/p\u003e\n\u003cp\u003eKAT, APC, HEW, GSS, KB, MPGB and FdlC \u0026ndash; Writing - original draft.\u003c/p\u003e\n\u003cp\u003eKAT, APC, HEW, GSS, ZE, ML, JYK, MS, GT, CL, BH, BRMS, MKS, DM, SM, KM, JH, JMW, CS, JMB, SS, KB, JPF, UD, SRS, MPGB and FdlC \u0026ndash; Writing - review and editing.\u003c/p\u003e\n\u003cp\u003eAll authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDisclaimer\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe views expressed in this article are those of the authors and do not necessarily reflect the official policy of the Agencies within the U.S. Department of Health and Human Services (CDC, FDA) and the U.S. Department of Agriculture (FSIS, APHIS), or the U.S. Government. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Health and Human Services or the U.S. Department of Agriculture.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eArnold BJ, Huang IT, Hanage WP (2021) Horizontal gene transfer and adaptive evolution in bacteria. Nat Rev Microbiol 20:206\u0026ndash;218\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJagadeesan B et al (2019) The use of next generation sequencing for improving food safety: Translation into practice. Food Microbiol 79:96\u0026ndash;115\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeeper MM et al (2023) Evaluation of whole and core genome multilocus sequence typing allele schemes for \u003cem\u003eSalmonella enterica\u003c/em\u003e outbreak detection in a national surveillance network, PulseNet USA. Front Microbiol 14:1254777\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCastillo-Ram\u0026iacute;rez S (2022) Beyond microbial core genomic epidemiology: towards pan genomic epidemiology. Lancet Microbe 3:e244\u0026ndash;e245\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu CM et al (2023) Using source-associated mobile genetic elements to identify zoonotic extraintestinal \u003cem\u003eE. coli\u003c/em\u003e infections. One Health 16:100518\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcnally A et al (2016) Combined analysis of variation in core, accessory and regulatory genome regions provides a super-resolution view into the evolution of bacterial populations. PLoS Genet 12:e1006280\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuillier L et al (2020) AB_SA: Accessory genes-Based Source Attribution \u0026ndash; tracing the source of \u003cem\u003eSalmonella enterica\u003c/em\u003e Typhimurium environmental strains. Microb Genom 6:mgen000366\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBatz MB et al (2021) Recency-weighted statistical modeling approach to attribute illnesses caused by 4 pathogens to food sources using outbreak data, United States. Emerg Infect Dis 27:214\u0026ndash;222\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeshearse E et al (2021) Attribution of illnesses transmitted by food and water to comprehensive transmission pathways using structured expert judgment, United States. Emerg Infect Dis 27:182\u0026ndash;195\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInteragency Food Safety Analytics Collaboration. Foodborne illness source attribution estimates for Salmonella, Escherichia coli O157, and Listeria monocytogenes \u0026ndash; United States (2022) GA and D.C.: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Food and Drug Administration, U.S. Department of Agriculture's Food Safety and Inspection Service. December 13, 2024\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCommichaux S et al (2023) Assessment of plasmids for relating the 2020 \u003cem\u003eSalmonella enterica\u003c/em\u003e serovar Newport onion outbreak to farms implicated by the outbreak investigation. BMC Genomics 24:165\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePan H et al (2022) Comprehensive assessment of subtyping methods for improved surveillance of foodborne \u003cem\u003eSalmonella\u003c/em\u003e. Microbiol Spectr 10:e02479\u0026ndash;e02422\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePe\u0026ntilde;il-Celis A et al (2024) Mobile genetic elements define the non-random structure of the \u003cem\u003eSalmonella enterica\u003c/em\u003e serovar Typhi pangenome. mSystems 9:e0036524\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrandenburg JM et al (2024) \u003cem\u003eSalmonella\u003c/em\u003e Hadar linked to two distinct transmission vehicles highlights challenges to enteric disease outbreak investigations. Epidemiol Infect 152:e86\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStapleton GS et al (2024) Multistate outbreaks of salmonellosis linked to contact with backyard poultry-United States, 2015\u0026ndash;2022. Zoonoses Public Health 71:708\u0026ndash;722\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNichols M et al (2021) Salmonella illness outbreaks linked to backyard poultry purchasing during the COVID-19 pandemic: United States, 2020. Epidemiol Infect 149:e234\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCenters for Disease Control and Prevention (2024) Data Summary: Persistent Strain of \u003cem\u003eSalmonella\u003c/em\u003e Hadar (REPTDK01) Linked to Backyard Poultry and Ground Turkey. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cdc.gov/salmonella/php/data-research/reptdk01.html\u003c/span\u003e\u003cspan address=\"https://www.cdc.gov/salmonella/php/data-research/reptdk01.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e [Accessed December 2nd, 2024]\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRibot EM, Freeman M, Hise KB, Gerner-Smidt P, PulseNet (2019) Entering the age of next-generation sequencing. Foodborne Pathog Dis 16:451\u0026ndash;456\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTolar B et al (2019) An overview of PulseNet USA databases. Foodborne Pathog Dis 16:457\u0026ndash;462\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcmillan EA et al (2019) Antimicrobial resistance genes, cassettes, and plasmids present in \u003cem\u003eSalmonella enterica\u003c/em\u003e associated with United States food animals. Front Microbiol 10:832\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou Z, Alikhan NF, Mohamed K, Fan Y, Achtman M (2020) The EnteroBase user's guide, with case studies on \u003cem\u003eSalmonella\u003c/em\u003e transmissions, \u003cem\u003eYersinia pestis\u003c/em\u003e phylogeny, and \u003cem\u003eEscherichia\u003c/em\u003e core genomic diversity. Genome Res 30:138\u0026ndash;152\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAchtman M et al (2021) Genomic diversity of \u003cem\u003eSalmonella enterica\u003c/em\u003e -The UoWUCC 10K genomes project [version 2; peer review: 2 approved]. Wellcome Open Res 5\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYoshida CE et al (2016) The \u003cem\u003eSalmonella\u003c/em\u003e In Silico Typing Resource (SISTR): An open web-accessible tool for rapidly typing and subtyping draft \u003cem\u003eSalmonella\u003c/em\u003e genome assemblies. PLoS ONE 11:e0147101\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang S et al (2019) SeqSero2: Rapid and improved \u003cem\u003eSalmonella\u003c/em\u003e serotype determination using whole-genome sequencing data. Appl Environ Microbiol 85:e01746\u0026ndash;e01719\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNoll N, Molari M, Shaw LP, Neher RA (2023) PanGraph: scalable bacterial pan-genome graph construction. Microb Genom 9:mgen001034\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarattoli A et al (2014) In silico detection and typing of plasmids using PlasmidFinder and plasmid multilocus sequence typing. Antimicrob Agents Chemother 58:3895\u0026ndash;3903\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarcill\u0026aacute;n-Barcia MP, Redondo-Salvo S, Vielva L (2075) \u0026amp; De La Cruz, F. MOBscan: automated annotation of MOB relaxases. \u003cem\u003eMethods Mol. Biol.\u003c/em\u003e (2020)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRedondo-Salvo S et al (2021) COPLA, a taxonomic classifier of plasmids. BMC Bioinformatics 22:390\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRedondo-Salvo S et al (2020) Pathways for horizontal gene transfer in bacteria revealed by a global map of their plasmids. Nat Commun 11:3602\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchwengers O et al (2021) Bakta: rapid and standardized annotation of bacterial genomes via alignment-free sequence identification. Microb Genom 7:000685\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBortolaia V et al (2020) ResFinder 4.0 for predictions of phenotypes from genotypes. J Antimicrob Chemother 75:3491\u0026ndash;3500\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobertson J, Nash JHE (2018) MOB-suite: software tools for clustering, reconstruction and typing of plasmids from draft assemblies. Microb Genom 4:e000206\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTagg KA et al (2024) Azithromycin-resistant \u003cem\u003emph(A)\u003c/em\u003e-positive \u003cem\u003eSalmonella enterica\u003c/em\u003e serovar Typhi in the United States. J Glob Antimicrob Resist 39:69\u0026ndash;72\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi C et al (2024) The spread of pESI-mediated extended-spectrum cephalosporin resistance in \u003cem\u003eSalmonella\u003c/em\u003e serovars\u0026mdash;Infantis, Senftenberg, and Alachua isolated from food animal sources in the United States. PLoS ONE 19:e0299354\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWebb HE et al (2022) Genome sequences of 18 \u003cem\u003eSalmonella enterica\u003c/em\u003e serotype Hadar strains collected from patients in the United States. Microbiol Resour Announc 11:e0052222\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBastian M, Heymann S, Jacomy M, Gephi (2009) An open source software for exploring and manipulating networks. \u003cem\u003eProc. Internat. AAAI Conf. Web Social Media\u003c/em\u003e 3, 361\u0026ndash;362\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang RH et al (2024) PhageScope: a well-annotated bacteriophage database with automatic analyses and visualizations. Nucleic Acids Res 52:D756\u0026ndash;D761\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWishart DS et al (2023) PHASTEST: faster than PHASTER, better than PHAST. Nucleic Acids Res 51:W443\u0026ndash;W450\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeemann T (2014) Prokka: rapid prokaryotic genome annotation. Bioinformatics 30:2068\u0026ndash;2069\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePage AJ et al (2015) Roary: rapid large-scale prokaryote pan genome analysis. Bioinformatics 31:3691\u0026ndash;3693\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMinh BQ et al (2020) IQ-TREE 2: New models and efficient methods for phylogenetic inference in the genomic era. Mol Biol Evol 37:1530\u0026ndash;1534\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLetunic I, Bork P (2024) Interactive Tree of Life (iTOL) v6: recent updates to the phylogenetic tree display and annotation tool. Nucleic Acids Res 52:W78\u0026ndash;W82\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLanza VF, Baquero F, De La Cruz F, Coque TM (2016) AcCNET (Accessory Genome Constellation Network): comparative genomics software for accessory genome analysis using bipartite networks. Bioinformatics 33:283\u0026ndash;285\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBergsma W (2013) A bias-correction for Cram\u0026eacute;r\u0026rsquo;s V and Tschuprow\u0026rsquo;s T. J Korean Stat Soc 42:323\u0026ndash;328\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThe NCBI Pathogen Detection Project [Internet]. Bethesda (MD): National Library of Medicine (US), National Center for Biotechnology Information (2016) May Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/pathogens/\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/pathogens/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFasano A et al (1991), \u003cem\u003eVibrio cholerae\u003c/em\u003e produces a second enterotoxin, which affects intestinal tight junctions. \u003cem\u003eProc. Natl. Acad. Sci. U S A\u003c/em\u003e 88, 5242\u0026ndash;5246\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu F, Lee H, Lan R, Zhang L (2016) Zonula occludens toxins and their prophages in \u003cem\u003eCampylobacter\u003c/em\u003e species. Gut Pathog 8:43\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMahendran V et al (2016) Examination of the effects of \u003cem\u003eCampylobacter concisus\u003c/em\u003e zonula occludens toxin on intestinal epithelial cells and macrophages. Gut Pathog 8:18\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDi Pierro M et al (2001) Zonula occludens toxin structure-function analysis. Identification of the fragment biologically active on tight junctions and of the zonulin receptor binding domain. J Biol Chem 276:19160\u0026ndash;19165\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRathore AS, Choudhury S, Arora A, Tijare P, Raghava GPS (2024) ToxinPred 3.0: An improved method for predicting the toxicity of peptides. Comput Biol Med 179:108926\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003e2016 Salmonella Outbreak Linked to Live Poultry in Backyard Flocks Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://archive.cdc.gov/#/details?url=https://www.cdc.gov/salmonella/live-poultry-05-16/index.html\u003c/span\u003e\u003cspan address=\"https://archive.cdc.gov/#/details?url=https://www.cdc.gov/salmonella/live-poultry-05-16/index.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e [Accessed December 2nd, 2024]\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaldhone PR et al (2019) Evaluation of incompatibility group I1 (IncI1) plasmid-containing \u003cem\u003eSalmonella enterica\u003c/em\u003e and assessment of the plasmids in bacteriocin production and biofilm development. Front Vet Sci 6:298\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFoley SL, Kaldhone PR, Ricke SC, Han J (2021) Incompatibility group I1 (IncI1) plasmids: Their genetics, biology, and public health relevance. Microbiol Mol Biol Rev 85:e00031\u0026ndash;e00020\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNichols M et al (2018) Preventing human \u003cem\u003eSalmonella\u003c/em\u003e infections resulting from live poultry contact through interventions at retail stores. J Agric Saf Health 24:155\u0026ndash;166\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBehravesh CB, Brinson D, Hopkins BA, Gomez TM (2014) Backyard poultry flocks and salmonellosis: A recurring, yet preventable public health challenge. Clin Infect Dis 58:1432\u0026ndash;1438\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGal\u0026aacute;n-Rela\u0026ntilde;o \u0026Aacute; et al (2023) \u003cem\u003eSalmonella\u003c/em\u003e and salmonellosis: An update on public health implications and control strategies. Animals 13:3666\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCenters for Disease Control and Prevention (2024) Data Summary: Persistent Strain of \u003cem\u003eSalmonella\u003c/em\u003e Infantis (REPJFX01) Linked to Chicken. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cdc.gov/salmonella/php/data-research/repjfx01.html\u003c/span\u003e\u003cspan address=\"https://www.cdc.gov/salmonella/php/data-research/repjfx01.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e [Accessed December 2nd, 2024]\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHassan R, Buuck S, Noveroske D (2019) Multistate outbreak of \u003cem\u003eSalmonella\u003c/em\u003e infections linked to raw turkey products \u0026mdash; United States, 2017\u0026ndash;2019. MMWR Morb Mortal Wkly Rep 68:1045\u0026ndash;1049\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5896627/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5896627/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe bacterial accessory genome, comprised of plasmids, phages, and other mobile elements, underpins the adaptability of bacterial populations. Pangenome (core and accessory) analysis of pathogens can reveal epidemiological relatedness missed by using core-genome methods alone. Employing a \u003cem\u003ek\u003c/em\u003e-mer-based Jaccard Index approach to compute pangenome relatedness, we explored the population structure and epidemiology of \u003cem\u003eSalmonella enterica\u003c/em\u003e serotype Hadar (Hadar), an emerging zoonotic pathogen in the United States (U.S.) linked to both commercial and backyard poultry. Hadar populations underwent substantial shifts between 2019 and 2020 in the U.S., driven by the expansion of a lineage carrying a previously uncommon prophage-like element. Phylogenetic and pangenomic relatedness, coupled with epidemiological data, suggest this lineage emerged from extant populations circulating in commercial poultry, with subsequent dissemination into backyard poultry environments. We demonstrate the utility of pangenomic approaches for mapping vertical and horizontal diversity and informing complex dynamics of zoonotic bacterial pathogens.\u003c/p\u003e","manuscriptTitle":"The Salmonella enterica serotype Hadar pangenome: Population structure and dynamics of a zoonotic pathogen","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-08 12:09:34","doi":"10.21203/rs.3.rs-5896627/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"8e125d96-5be4-429b-bb3d-30259edfbaf9","owner":[],"postedDate":"May 8th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":43541189,"name":"Biological sciences/Microbiology/Microbial genetics/Bacterial genetics"},{"id":43541190,"name":"Biological sciences/Computational biology and bioinformatics/Genome informatics"},{"id":43541191,"name":"Biological sciences/Genetics/Microbial genetics/Bacterial genetics"}],"tags":[],"updatedAt":"2026-02-04T08:12:02+00:00","versionOfRecord":{"articleIdentity":"rs-5896627","link":"https://doi.org/10.1038/s41467-025-68026-3","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2026-01-24 05:00:00","publishedOnDateReadable":"January 24th, 2026"},"versionCreatedAt":"2025-05-08 12:09:34","video":"","vorDoi":"10.1038/s41467-025-68026-3","vorDoiUrl":"https://doi.org/10.1038/s41467-025-68026-3","workflowStages":[]},"version":"v1","identity":"rs-5896627","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5896627","identity":"rs-5896627","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}