From April 15th to June 17th, 2024, a total of 114 samples of retail chicken meats were collected and tested, all purchased from retail markets in Ulaanbaatar, Mongolia. These included domestic Mongolian whole chickens and imported Chinese and Russian sliced chicken meats. Retail chicken meats were aseptically placed in a sterile plastic bag containing 400 mL of buffered peptone water broth (BPW, Difco, Detroit, MI, USA) and shaken for 2 mins. From this initial rinse, a 20 mL aliquot of the BPW-sample mixture was transferred, vortex-mixed with fresh 20 mL of BPW for 15 s, then incubated at 36 ± 1°C for 18–24 h. Subsequently, 100 μL of the incubated culture was vortex-mixed for 15 s in 10 mL of Rappaport-Vassiliadis (RV) broth, then incubated at 41.5 ± 1°C for 20–24 h. Cultures exhibiting a color change were streaked onto Salmonella ChromoSelect agar (Sigma-Aldrich, St. Louis, MO, USA) and incubated at 37°C for 24 h. Pink colonies presumptively identified as Salmonella spp. on the agar were confirmed using the VITEK-MS system (bioMérieux, Marcy-L’Etoile, France). Confirmed isolates were stored at −80°C in glycerol. All Salmonella spp. isolates obtained during the study period were included for the further studies, following their transport to Konkuk University, Seoul, Korea.

Genomic DNA was extracted from overnight cultures of Salmonella spp. using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions. The purity and concentration of the extracted DNA were measured with the NanoDrop™ spectrophotometer (Thermo Fisher Scientific) and the Quantus™ fluorometer (Promega, Madison, WI, USA), respectively. Library preparation was performed with the TruSeq Nano DNA Library Prep kit (Illumina, San Diego, CA, USA) following the manufacturer’s instructions. The library was then sequenced on a NovaSeq 6,000 platform (Illumina, San Diego, CA, USA), generating 150 bp paired-end reads. Read quality was assessed using fastp v.0.20.1 (13). Low-quality bases and adapter sequences were trimmed using Trimmomatic v0.39 (14). Trimmed reads were assembled using SPAdes v4.1.0 (15) with the “—careful” option. Serotype prediction was performed using SeqSero v1.3.1 (16) on the assembled contigs, and only isolates identified as serovar Enteritidis were included in further analyses. The whole-genome sequencing data generated in this study has been deposited in the NCBI BioProject database under accession number PRJNA1236736.

Antimicrobial susceptibility testing was determined using the Sensititre™ ARIS HiQ AST system (Thermo Fisher Scientific, Waltham, MA, USA) with the Sensititre™ panel (KRCDC2F; Thermo Fisher Scientific, Waltham, MA, USA). The following antimicrobials were tested: ciprofloxacin (CIP, 0.03–0.5 μg), nalidixic acid (NAL, 2–128 μg), imipenem (IMI, 1–8 μg), colistin (COL, 2–16 μg), ampicillin (AMP, 2–64 μg), tetracycline (TET, 2–128 μg), chloramphenicol (CHL, 2–32 μg), azithromycin (AZI, 2–32 μg), gentamicin (GEN, 1–64 μg), streptomycin (STR, 2–128 μg; Ref. National Antimicrobial Resistance Monitoring System, 2014) (17), amikacin (AMI, 4–64 μg), trimethoprim/sulfamethoxazole (SXT, 1/19–16/304), cefotaxime (FOT, 1–32 μg), ceftriaxone (AXO, 1–32 μg), cefoxitin (FOX, 4–32 μg), and ceftazidime (TAZ, 1–16 μg).

Purified bacterial isolates were suspended in distilled pure water and adjusted to a 0.5 McFarland standard using a nephelometer. A 10 μL aliquot of the adjusted suspension was inoculated into 11 mL of cation-adjusted Mueller-Hinton (MH) medium and dispensed into the KRCDC2F Sensititre™ panel using the Sensititre AIM automated dispensing system. The inoculated microplates were sealed with a microplate-sealing film and incubated at 37°C for 18 h in the Sensititre ARIS HiQ automated analysis system.

After incubation, a fluorescence auto-read system determined minimum inhibitory concentration (MIC) values. These MIC values were determined according to Clinical and Laboratory Standards Institute (CLSI) M100-2018 28th edition guidelines (Wayne, PA, USA) (18), except for Streptomycin, which followed the National Antimicrobial Resistance Monitoring System (NARMS) criteria (17). Escherichia coli ATCC25922 was used as a quality control standard. Salmonella isolates resistant to more than three classes and more than one antimicrobial in a single class were designated MDR strains, following standard criteria (19).

Whole-genome single nucleotide polymorphisms (wgSNPs) were identified among S. Enteritidis (n = 45) in this study using Snippy v4.6.0¹ (20), with S. Enteritidis strain P125109 as the reference for SNP calling (GenBank accession no. AM933172.1, NCBI RefSeq accession no.NC_011294.1). Prophage regions, repetitive genomic regions, and recombination regions in the reference genome were identified and excluded from the analysis. Specifically, prophage regions were identified using PHASTEST (21), repetitive genomic regions were determined by self-alignment of the reference genome using the nucmer tool from the Mummer package v.4.0.1 (22) using the --maxmatch and --nosimplify options, and recombination regions were filtered out using Gubbins v.3.4 (23). A maximum likelihood phylogenetic tree was constructed from the cleaned alignment using a TVMe model selected by ModelFinder (24) in IQ-TREE (25) under 1,000 bootstrap replicates. The phylogenetic tree was visualized using the interactive Tree of Life version 5 (iTOLv7)². Genotype clustering was performed using Bayesian Analysis of Population Structure (BAPS) through the “FastBAPS” R package.

For phylogenetic analysis with Chinese isolates, genomic sequences of S. Enteritidis isolates (n = 816) from China were downloaded from EnteroBase³ (26). A phylogenetic tree was constructed using Mashtree v.1.4.6 (27) with bootstrapping using Mash distances. To reduce redundancy and computational complexity, Treemmer v0.3 (28) was applied with the -lmc option, pruning the tree to maintain its overall topology while preserving all Mongolian S. Enteritidis isolates and selectively reducing the number of Chinese isolates by removing leaves that contribute minimally to the tree’s diversity.

The final dataset included a total of 106 isolates comprising 45 Mongolian isolates from this study and 61 Chinese isolates (detailed in Supplementary Table S3). These isolates were analyzed using wgSNP, followed the construction of a maximum-likelihood phylogenetic tree using a TVM + F + ASC model, selected by ModelFinder (24) in IQ-TREE (25) under 1,000 bootstrap replicates and determination of population structure with FastBAPS as detailed above (29).

All isolates were screened for antimicrobial resistance genes (ARGs) using a pipeline available at github.com/maxlcummins/pipelord. Gene screening was performed using abricate v1.0.1⁴ in conjunction with the following databases: ResFinder (30) and PlasmidFinder (31). Genes were marked as present when detected by abricate and filtered using a custom R script⁵ to have 90% identity and 90% coverage.

A total of 56 Salmonella spp. isolates were recovered from 114 retail chicken meat samples: 48 isolates (63.2%, 48/76) from Mongolian domestic whole chickens and eight isolates (23.5%, 8/34) from Chinese imported sliced chicken meats. No Salmonella spp. was detected in Russian imported sliced chicken meat (n = 4). Sampling metadata for all Salmonella spp. are provided in Supplementary Table S1.

Whole-genome sequencing (WGS) analysis identified four distinct serovars among the 56 studied Salmonella isolates (Table 1). The predominant serovar was S. Enteritidis (80.4% 45/56), with 41 isolates recovered from Mongolian domestic chicken and four from Chinese imported chicken. S. Typhimurium (12.5% 7/56) was detected exclusively in Mongolian domestic chicken, while S. Agona (3.6% 2/56) and S. Indiana (3.6% 2/56) were only identified in imported samples from China.

The antimicrobial resistance profiles of 45 S. Enteritidis isolates were determined using the broth dilution method against 16 antimicrobial agents, and the results are presented in Table 2 (Supplementary Table S2). The highest resistance was observed against nalidixic acid (100% 45/45), followed by ampicillin (93.3% 42/45), and streptomycin (88.9% 40/45). Resistance to third-generation cephalosporins (cefotaxime and ceftriaxone) was observed in 4 isolates (8.9%), all of which concurrently resistant to ampicillin, nalidixic acid, streptomycin, tetracycline, and trimethoprim/sulfamethoxazole (Table 2). Colistin resistance was also detected in four isolates (8.9%), two originating from Mongolia domestic chicken and the other two from imported Chinese chicken. The remaining 41 isolates (91.1%) were classified as intermediate to colistin.

The WGS analysis results for the 45 S. Enteritidis isolates in this study are summarized in Table 3 (Supplementary Table S4). The most frequently detected resistance genes were aac(6′)-Iaa (100% 45/45), blaTEM-1B (93.3% 42/45), aph(3″)-Ib (88.9% 40/45), and sul2 (88.9% 40/45), which confer resistance to aminoglycosides, penicillins, and sulfonamides, respectively. The blaCTX-M-14 gene, which confers resistance to third-generation cephalosporins, was detected in 4 isolates (8.9%). These isolates are harbored aadA5, dfrA17, fosA3, qnrS1, and tet(A).

In addition to antimicrobial resistance gene detection, chromosomal mutations associated with quinolone resistance were also investigated in Table 3. All isolates carried at least one non-synonymous point mutation in the gyrA gene including gyrA D87Y (Aspartic acid to Tyrosine) in 93.3% (42/45) of isolates, gyrA D87G (Aspartic acid to Glycine) in one isolate (2.2%), gyrA D87N (Aspartic acid to Asparagine) and gyrA S83Y (Serine to Tyrosine) in two isolates (4.4%), respectively. None of the isolates had a point mutation in the gyrB, parC, and parE genes (data not shown). Although four isolates exhibited phenotypic resistance to colistin in antimicrobial susceptibility testing, the mobile colistin resistance (mcr) gene was not detected in any of them. A single chromosomal mutation, pmrB L14F, which is associated with colistin resistance, was identified in one isolate; However, this mutation did not correspond with the observed phenotypic resistance.

Two dominant AMR profiles were identified among the S. Enteritidis isolates. The AMP-NAL-STR profile, observed in the majority of isolates (77.8%, 35/45),was associated with the presence of blaTEM-1B, aac(6′)-Iaa, aph(3″)-Ib, and sul2, as well as the gyrA D87Y chromosomal mutation. The AMP-FOX-AXO-NAL-STR-TET-STX profile, which conferred resistance to the highest number of antimicrobial classes, was found in 8.9% (4/45) of isolates, and was characterized by blaTEM-1B, blaCTX-M-14, fosA3, qnrS1, tet(A), sul2, and dfrA17.

WGS analysis of 45 S. Enteritidis isolates in this study revealed four plasmid incompatibility (Inc) types: IncFIB(S) and IncFII were the most common, detected in 91.1% (41/45), followed by IncX1_4 in 88.9% (32/45), and IncI1-α in 11.1% (5/45). Seven distinct Inc. group profiles were identified, as summarized in Table 4 (Supplementary Table S1). The predominant profile -IncFIB(S), IncFII(S), and IncX1- was observed in 71.1% (32/45), followed by IncFIB(S) and IncFII(S) in 8.9% (4/45), and three profiles each accounting for 6.7% (3/45): IncFIB(s), IncFII(S), and IncI1-α; IncFII(S), IncX1; and IncX1 alone. The remaining profiles were each identified in a single isolate, reflecting a diverse plasmid landscape among the isolates.

A midpoint-rooted maximum-likelihood (ML) phylogenetic tree was constructed based on whole-genome single nucleotide polymorphism (wgSNP) analysis of 396 SNP sites to assess the genetic diversity among 45 S. Enteritidis isolates obtained in this study (Figure 1). Based on the SNP alignment and phylogenetic structure, FastBAPS analysis grouped the 45 S. Enteritidis isolates into two BAPS clusters: BAPS 1, consisting of Chinese imported isolates, and BAPS 2, comprising Mongolian domestic isolates.

Mongolian domestic S. Enteritidis isolates exhibited distinct clustering patterns in the phylogenetic analysis. To further investigate the genetic basis of this separation, AMR gene profiles and plasmid incompatibility (Inc) groups were analyzed, revealing three genetically distinct clusters. Notably, isolates within BAPS 2 (Mongolian domestic group), were further subdivided into a major and minor subgroup based on difference in AMR gene content and Inc. group composition.

The analysis of AMR genes showed that isolates in BAPS 1 harbored fewer AMR genes than those in BAPS 2, with the aac(6′)-Iaa gene consistently present in both clusters. Most isolates (35/45) in BAPS 2 group carried the key antimicrobial resistance genes, including aac(6′)-Iaa, aph(3″)-IIa, aph(6)-Id, blaTEM-1B, and sul2. In contrast, the minor domestic subgroup harbored additional AMR genes, including clinically significant antimicrobial resistance genes like the extended-spectrum beta-lactamase (ESBL) genes blaCTX-M-14, as well as plasmid-mediated quinolone resistance (PMQR) genes qnrS1.

Plasmid analysis revealed a strong association between Inc. group composition and the distribution of AMR genes. In the BAPS 1, most isolates harbored only IncF plasmid. In contrast, the major Mongolian domestic subgroup primarily exhibited a combination of IncF and IncX1 plasmids, except for three isolates that contained solely IncX1. The minor domestic subgroup also showed a distinct profile characterized by the presence of IncF, IncI1-α, and IncX1 plasmids. This variation suggests a potential role of IncI1-α in facilitating the acquisition and dissemination of AMR genes, warranting further investigation.

The distinct clustering profile also highlighted a notable genetic divergence between Mongolian domestic S. Enteritidis isolates and those sourced from China, as further supported by a comparative analysis with 61 publicly available Chinese S. Enteritidis genomes (Figure 2). A midpoint-rooted ML phylogenetic tree was constructed based on wgSNP analysis of 1,502 SNP sites. FastBAPS grouped the combined dataset into four clusters (BAPS A to D), clearly separating Mongolian isolates (BAPS 2) from Chinese ones (BAPS 1), consistent with their AMR gene profiles and Inc. group compositions.

Notably, all five Chinese imported S. Enteritidis isolates clustered closely with isolates from China, indicating high genetic similarity. In BAPS A, one Mongolian domestic isolate (KUSA_046) grouped with four Chinese isolates obtained from animal feed (n = 1), aquatic animal (n = 1), human (n = 1), and poultry (n = 2). BAPS B comprised two Chinese imported isolates (KUSA_048 and KUSA_049), which clustered with three Chinese isolates from humans (n = 2) and wild animals (n = 1), as well as one Mongolian domestic isolate associated with samples from the environment (n = 1), human (n = 7), wild animal (n = 1), and poultry (n = 6).

BAPS C included one Chinese imported isolate that clustered with 34 Chinese isolates from various sources. In contrast, BAPS D consisted exclusively of Mongolian domestic isolates and formed a distinct phylogenetic cluster. Despite this separation, BAPS C and BAPS D shared similar AMR gene profiles, including aac(6′)-Iaa, aph(3″)-Ib, aph(6)-Id, blaTEM-1B, and sul2, which are associated with resistance to aminoglycosides, β-lactams, and sulfonamides.