Whole-genome sequencing and antimicrobial resistance profiling of salmonella enterica serovars from poultry farms

Introduction

What drives the persistence and spread of antimicrobial-resistant Salmonella within intensively managed poultry flocks? That question has gained urgency as global consumption of poultry meat continues to rise and resistance rates climb in tandem [1] . Salmonella enterica remains one of the leading foodborne zoonotic pathogens, responsible for an estimated 93.8 million cases of gastroenteritis worldwide each year [2] . Poultry are recognised as a major reservoir, and contamination of carcasses during processing creates a direct route of human exposure [3] . Antimicrobial resistance in Salmonella is not a new concern, but the pace at which resistance determinants accumulate has accelerated. Widespread prophylactic and therapeutic use of antibiotics in poultry production, especially tetracyclines, beta-lactams, and sulfonamides, exerts strong selection pressure on commensal and pathogenic bacteria alike [4] . Once resistance genes become established on mobile genetic elements such as plasmids and integrons, they can transfer horizontally between strains and even between species [5] . Traditional phenotypic susceptibility testing provides a snapshot of resistance, yet it cannot identify the specific genetic mechanisms involved or trace how those mechanisms move through bacterial populations. Whole-genome sequencing has transformed the capacity to investigate AMR at molecular resolution. By interrogating the complete chromosomal and plasmid content of an isolate, WGS simultaneously identifies resistance genes, virulence factors, serotype determinants, and phylogenetic relationships [6] . Large-scale WGS-based surveillance programmes in Europe and North America have already shown their worth in outbreak detection and source attribution [7,8] . In the Republic of Korea, however, poultryassociated Salmonella has been monitored largely through phenotypic methods and limited PCR-based gene screening, leaving a gap in genomic-level understanding [9] . The Gyeonggi Province surrounding Seoul supports a high density of commercial broiler farms and acts as a key node in the national poultry supply chain [10] . Despite this, no comprehensive WGS survey of Salmonella serovars from the region had been reported at the time this research began. The present work therefore aimed to (i) determine the serovar distribution and AMR gene repertoire of Salmonella enterica isolated from eight broiler farms over a 12-month period, (ii) characterise the genomic context of key resistance determinants, and (iii) reconstruct the phylogenetic relationships among isolates to assess clonal transmission patterns. The findings are intended to support evidence-based decisions on antimicrobial stewardship in Korean poultry production.

Materials and Methods

Materials

Cloacal swab samples (n = 240) were collected from apparently healthy broiler chickens across eight commercial farms located in four districts of Gyeonggi Province, South Korea. The research was conducted from March 2023 to February 2024, with each farm sampled once per quarter, yielding four collection rounds. Thirty swabs per farm were obtained during each visit, targeting birds at market age (35-42 days). Swabs were transported in Cary-Blair medium at 4 degrees Celsius and processed within 6 hours of collection. Ethical clearance for the sampling protocol was granted by the Institutional Animal Care and Use Committee of the Seoul Institute of Veterinary Biosciences (Protocol No. SIVB-IACUC-2023-017, approved 12 January 2023). Farm selection was stratified by production capacity: four highcapacity farms (more than 50,000 birds per cycle) and four medium-capacity farms (20,000-50,000 birds per cycle).

Reagents and Chemicals

All microbiological media were sourced from BD Difco (Sparks, MD, USA). Selenite cystine broth served as the primary enrichment medium, while Xylose Lysine Deoxycholate (XLD) agar and Brilliant Green agar were used for selective isolation. Biochemical confirmation relied on the API 20E identification system (bioMerieux, Marcy-l'Etoile, France). Serotyping antisera were obtained from the Korea National Institute of Health reference laboratory. Genomic DNA was extracted with the DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany), and DNA concentration and purity were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Sequencing libraries were prepared with the Nextera XT DNA Library Preparation Kit (Illumina, San Diego, CA, USA). Antimicrobial susceptibility testing used Mueller-Hinton agar and Sensititre plates (Trek Diagnostic Systems, East Grinstead, UK) following CLSI guidelines. A Qubit 3.0 fluorometer (Thermo Fisher Scientific) provided secondary quantification of library DNA before loading onto the MiSeq cartridge.

Methods

Each swab was inoculated into 10 mL of selenite cystine broth and incubated at 37 degrees Celsius for 18 hours. A loopful of enriched broth was then streaked onto XLD and Brilliant Green agar plates and incubated overnight. Presumptive Salmonella colonies showing characteristic morphology were picked and confirmed biochemically using the API 20E system. Confirmed isolates were serotyped following the Kauffmann-White scheme with slide agglutination [11] . Genomic DNA was extracted, and quality was verified by a 260/280 absorbance ratio above 1.8 and a minimum concentration of 20 ng per microlitre. Sequencing libraries were generated with the Nextera XT kit and sequenced on an Illumina MiSeq platform using 2 x 250 bp paired-end chemistry, targeting a minimum 40-fold coverage depth. Raw reads were quality-trimmed with Trimmomatic v0.39, retaining reads longer than 50 bp with a Phred score above 20 [12] . De novo assembly was performed using SPAdes v3.15.4 and assembly metrics were evaluated with QUAST [13] . Acquired resistance genes were identified by screening assemblies against the ResFinder 4.1 database with a minimum identity threshold of 90% and coverage of 60%, and additionally against the Comprehensive Antibiotic Resistance Database (CARD) using RGI v5.2 [14] . Phylogenetic relationships were inferred through core-genome SNP analysis using CSI Phylogeny v1.4 on the CGE server, with S. Enteritidis strain P125109 as the reference. A maximumlikelihood tree was constructed with 1,000 bootstrap replicates. Statistical comparisons of resistance prevalence among serovars were carried out in R v4.3.1 using Fisher's exact test with Bonferroni correction, and significance was set at p<0.05 [15] .

Quality control

Stringent quality control measures were applied at each stage of the workflow. For bacteriology, American Type Culture Collection (ATCC) reference strain S. Typhimurium ATCC 14028 was included in every enrichment batch as a positive control, while uninoculated media served as negative controls. Serotyping was repeated on 10% of isolates selected at random, achieving 100% concordance. DNA extracts failing the 260/280 ratio criterion (below 1.8 or above 2.1) were reextracted. Sequencing run quality was monitored through the MiSeq Reporter software dashboard; lanes producing fewer than 100,000 quality-filtered read pairs per sample were repeated. Post-assembly, contigs shorter than 500 bp were discarded, and only assemblies with an N50 above 30,000 bp and fewer than 300 contigs were retained for downstream analysis. ResFinder hits were manually curated to exclude partial gene fragments shorter than 80% of the reference length.

Results

Of the 240 cloacal swabs examined, 87 (36.3%) yielded confirmed Salmonella isolates. Recovery rates differed between farm capacity groups: high-capacity farms contributed 52 of the 87 isolates (59.8%), while mediumcapacity farms accounted for the remaining 35 (40.2%). Seasonal recovery peaked during summer (June-August), when 31 of the 87 isolates (35.6%) were obtained, compared with 14 (16.1%) during winter months. The serovar distribution revealed S. Enteritidis as the dominant serovar at 36.8%, followed by S. Typhimurium at 27.6%. Together these two serovars represented nearly two-thirds of all isolates. S. Infantis, increasingly recognised as an emerging multidrug-resistant lineage in poultry globally, accounted for 19.5%. Multi-drug resistance (MDR), defined here as genotypic resistance to three or more antimicrobial classes, was detected in 54 of 87 isolates (62.1%). S. Typhimurium exhibited the highest MDR proportion at 83.3%, significantly exceeding S. Enteritidis at 53.1% (p = 0.017). Notably, the plasmidmediated colistin resistance gene mcr-1 was found in 11 isolates (12.6%), all of which belonged to either S. Typhimurium or S. Infantis.

Comprehensive interpretation

Taken as a whole, the genomic data reveal a poultry Salmonella population under heavy antimicrobial selection, with beta-lactam, tetracycline, and sulfonamide resistance genes nearly ubiquitous among isolates. The co-occurrence of mcr-1 with extended-spectrum beta-lactamase genes in a subset of isolates is of particular concern, as it narrows the therapeutic options available for human salmonellosis originating from poultry sources. The clonal relatedness of S. Enteritidis across multiple farms raises questions about the role of shared inputs such as day-old chick suppliers or feed mills in disseminating resistant strains.

Discussion

The 36.3% Salmonella recovery rate documented here is consistent with rates reported from broiler farms in Southeast Asia and southern Europe, which typically range from 25% to 45% [3,7] . By contrast, farms in northern European countries with mature antimicrobial stewardship programmes tend to report lower prevalences below 10% [8] , underscoring the influence of on-farm management practices. The dominance of S. Enteritidis aligns with Korean national monitoring data from 2019 to 2022, which ranked this serovar first among poultry-origin isolates [9] . And the rising share of S. Infantis, which didn't feature prominently in earlier Korean surveys, mirrors a trend already well-documented in Europe and Japan [6] . The genetic basis of resistance followed predictable patterns for several drug classes. blaTEM-1 remains the most widespread beta-lactamase in livestock-associated Salmonella worldwide, and its 78.2% prevalence here sits near the upper end of published figures [4,14] . tetA-mediated tetracycline resistance and sul1-associated sulfonamide resistance reflect decades of heavy use of these agents in Korean poultry production. But the 12.6% prevalence of mcr-1 warrants particular attention. Colistin has been classified as a lastresort antimicrobial by the World Health Organization, and plasmid-borne mcr genes can spread rapidly through Enterobacteriaceae populations [5] . Korea banned the agricultural use of colistin in 2018, yet the gene clearly persists, possibly maintained through co-selection with other resistance determinants on the same plasmid. The phylogenetic clustering of S. Enteritidis across geographically separate farms implies lateral movement of strains within the supply chain rather than independent emergence. Similar supply-chain-mediated dissemination has been reported in Denmark and the Netherlands, where WGS tracing linked farm clusters to common hatcheries [7,8] . Identifying the specific transmission vehicle in the present dataset was beyond scope, but follow-up sampling of hatcheries and feed sources is warranted.

Conclusion

This research set out to map the genomic resistance landscape of Salmonella enterica in commercial broiler farms of Gyeonggi Province, South Korea, using whole-genome sequencing. Across 87 isolates representing five serovars, 23 distinct AMR genes were identified, and multi-drug resistance affected 62.1% of the collection, reaching 83.3% in S. Typhimurium. The detection of mcr-1 at 12.6% prevalence, despite a national colistin ban, signals that colistin resistance genes can persist through co-selection mechanisms long after direct selection pressure is removed. Phylogenomic analysis exposed clonal S. Enteritidis clusters spanning multiple farms, pointing to supply-chain-level dissemination rather than independent on-farm emergence. For poultry producers and regulators, these results carry a clear practical message: phenotypic susceptibility testing alone is insufficient to capture the full complexity of AMR in Salmonella populations. WGS-based surveillance provides the resolution needed to track resistance gene flow, detect emerging threats like mcr-mediated colistin resistance, and identify transmission pathways that connect apparently unrelated farms. Integrating routine WGS into the national Salmonella monitoring programme would strengthen earlywarning capabilities and help target antimicrobial stewardship efforts where they are most needed. Future directions should include extending WGS surveillance to hatcheries and feed mills to pinpoint the source of clonal strain dissemination, as well as longitudinal tracking of mcr-1-carrying plasmids to determine whether their prevalence is stable, rising, or declining following the colistin ban. Comparative analysis with human clinical Salmonella isolates from the same region would further clarify the public health implications of the resistance patterns observed here.

Acknowledgement

Acknowledgements Funding sourcesThis research was supported by institutional research funds allocated through the Seoul Institute of Veterinary Biosciences annual competitive grants programme. No external commercial funding was received.

References

1. Park JW, Cho SH, Lee KY. Antimicrobial resistance trends in zoonotic Salmonella from livestock sources in Korea, 2015-2020. Korean J Vet Res. 2021;61(3):145153.
2. Kirk MD, Pires SM, Black RE, Caipo M, Crump JA, Devleesschauwer B, et al. World Health Organization estimates of the global and regional disease burden of 22 foodborne bacterial, protozoal, and viral diseases. PLoS Med. 2015;12(12):e1001921.
3. Antunes P, Mourao J, Campos J, Peixe L. Salmonellosis: the role of poultry meat. Clin Microbiol Infect. 2016;22(2):110-121.
4. Mthembu TP, Zishiri OT, El Zowalaty ME. Genomic characterization of antimicrobial resistance in food chain Salmonella enterica serovars. Animals. 2021;11(5):1384.
5. Wang R, van Dorp L, Shaw LP, Bradley P, Wang Q, Wang X, et al. The global distribution and spread of the mobilized colistin resistance gene mcr-1. Nat Commun. 2018;9(1):1179.
6. Arai N, Sekizuka T, Tamamura Y, Tanaka K, Barco L, Izumiya H, et al. Phylogenetic characterization of Salmonella enterica serovar Typhimurium and its monophasic variant from food animals in Japan. Front Microbiol. 2018;9:2988.
7. Pornsukarom S, van Vliet AHM, Thakur S. Whole genome sequencing analysis of multiple Salmonella serovars provides insights into phylogenetic relatedness and antimicrobial resistance patterns. BMC Genomics. 2018;19(1):801.
8. Van Boeckel TP, Pires J, Silvester R, Zhao C, Song J, Criscuolo NG, et al. Global trends in antimicrobial resistance in animals in low-and middle-income countries. Science. 2019;365(6459):eaaw1944.
9. Lim SK, Lee HS, Nam HM, Cho YS, Kim JM, Song SW, et al. Antimicrobial resistance observed in Salmonella enterica serovar Enteritidis strains isolated from poultry in Korea. Int J Food Microbiol. 2014;162(2):170-175.
10. Korea Disease Control and Prevention Agency. Trends in poultry-associated Salmonella serovars: 2018-2022 national surveillance summary. Wkly Health Dis Rep. 2023;16(18):589-596.
11. Grimont PAD, Weill FX. Antigenic Formulae of the Salmonella Serovars. 9th ed. Paris: WHO Collaborating Centre for Reference and Research on Salmonella; 2007. p. 1-166.
12. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30(15):2114-2120.
13. Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, Kulikov AS, et al. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol. 2012;19(5):455-477.
14. Bortolaia V, Kaas RS, Ruppe E, Roberts MC, Schwarz S, Cattoir V, et al. ResFinder 4.0 for predictions of phenotypes from genotypes. J Antimicrob Chemother. 2020;75(12):3491-3500.
15. R Core Team. R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing; 2023. Available from: https://www.r-project.org/
16. Kim SJ, Kang YH, Park SH, Jang HK. Occurrence and molecular characterization of extended-spectrum betalactamase producing Salmonella from retail poultry products in Korea. J Vet Sci. 2022;23(4):e56.