Selenite Brilliant Green (SBG) enrichment solution (#HB8606, Qingdao Haibo), Blood plate (#CP10002, Shanghai Kemagar), Salmonella Shigella (SS) agar medium (#HB4089, Qingdao Haibo), Xylose lysine deoxycholate (XLD) agar medium (#HB4105, Qingdao Haibo), MacConkey agar medium (#HB6238-9, Qingdao Haibo), three-sugar iron agar slant (#HB4088, Qingdao Haibo), Gram staining solution (#HB8278, Qingdao Haibo), Gram-negative bacteria identification card (#21341, Merieux, France), Salmonella typing diagnostic serum (#882116,#152,116,#332,106, Senyan, Japan), Antimicrobial Susceptibility Testing (AST) panel for aerobic Gram-negative bacilli (#B3226B, Thermo Fisher, America), DNA extraction reagents (#51304, QIAGEN); all reagents were used within their expiry dates.

Constant temperature incubator (MIR-H263L-PC, PHCBI), Optical microscope (CX21FS1, Olympus), Automatic microbial identification and drug sensitivity analysis system (VITEK 2 COMPACT, Merieux, France), Turbidimeter (DensiCHEK plus, Merieux, France), Microbial susceptibility instrument (Vizion^®, Thermo), High-throughput sequencer (model: 550, Illumina) were used for sequencing bacterial genomes (2nd generation), etc.

The S. enterica isolates 2023JS33 and 2023JS34 were extracted from the stool sample of a healthy female, 52 years old, located in southeast China, without typhoid fever or any gastrointestinal complaints. The studies involving humans were approved by Committee of Zhejiang Provincial Center for Disease Control and Prevention. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided.

An appropriate amount of feces was inoculated in SBG enrichment broth and incubated at 36 °C for 24 h. Then, broth containing bacteria was inoculated on SS agar medium and MacConkey agar medium by drawing lines in sections and incubated at 36 °C for 24 h. After that, a single colony was selected and inoculated with Triple Sugar Iron (TSI) and incubated at 36 °C for 24 h. Pick the interested bacterial species for microscopic examination with Gram stain and subsequently inoculate into blood plates and incubate at 36°C for 24 h. The purified bacterial species were identified by automatic biochemical identification and examined with the serum agglutination test.

One to two single colonies were picked by inoculation rings and emulsified in sterile water. Adjust solution to 0.5 McFarland turbidity for biochemical identification with Gram-negative bacteria identification cards that had been rewarmed in advance. The sterilized saline was used as agglutination control and the agglutination phenomenon was observed within 2 min.

An appropriate amount of Salmonella serum was dropped on a clean slide and mixed with the bacterial moss, picked out by an inoculation ring, and sterilized saline thoroughly. If no agglutination was observed, other commercial serums were used to conduct serum agglutination tests one by one, according to the instructions of reagents.

The antimicrobial susceptibility of the isolates was determined by microdilution broth assay. In detail, tested isolates (2023JS33,2023JS34) and quality control strain (ATCC25922) were streaked and inoculated on blood agar plates and incubated at 36°C for 24 h. Individual colonies were picked with an inoculation ring seeded again on blood agar plates and incubated for 24 h at 36 °C. One or two colonies were picked from freshly prepared blood agar plates and emulsified in sterile water. Adjusted the solution to 0.5 McFarland turbidity and mixed thoroughly. Then, the bacterial suspension prepared above 10 μL was added to a test tube containing 11 mL cation-adjusted Mueller-Hinton broth (CAMHB) and mixed well. The mixture should be used within 15 min. Replace the test tube cover with a Sensititre^® disposable sampling head and add the sample to the CHNENF drug sensitivity test plate according to AIM^® instructions. Remove the test tube/sampling head combination from AIM^® within 30 s after completion of sample loading in the drug susceptibility plate.

After the inoculation of the drug sensitivity plate, the purity of the final culture solution was checked, and all micropores were covered with a sealing film. After the incubation at 36°C for 24 h, all samples were read with a microbial susceptibility instrument, Vizion^®. The minimum inhibitory concentration (MIC) of the drugs that naked eye could see was recorded and defined as sensitive (S), moderately sensitive (I) and resistant (R) according to the standard of Clinical and Laboratory Standards Institute (CLSI) (2023). The quality control strain was Escherichia coli ATCC25922. As CLSI does not provide streptomycin resistance breakpoint, it was determined according to the National Antimicrobial Resistance Monitoring System (NARMS) MIC criteria [Centers for Disease Control and Prevention (CDC), 2018].

Two isolates (2023JS33 and 2023JS34) were sent to the genetic testing laboratory of Zhejiang Tianke High-tech Development Co., Ltd. for deep sequencing (3rd generation sequencing). Whole genomic DNA was extracted by Gentra Puregene Yeast/Bact Kit (Qiagen, Valencia, CA) and sequenced using the GridION X5 platform (Oxford Nanopore Technology).

Colorless, translucent, and smooth round colonies were observed on MacConkey agar medium (Figure 1A), supporting that JS33 and JS34 belonged to Salmonella (Farhoudi Moghaddam et al., 1988). In addition, JS33 and JS34 formed round, moist, smooth, translucent colonies that became lighter in color on the SS agar medium (Figure 1B). However, JS33 was a colony with a black center, which differed from JS34 (Figure 1B). These two bacterial isolates with inconsistent morphology on the SS agar medium were selected and inoculated on the three-sugar iron agar slant (Figure 1C). It showed that both isolates fermented glucose and produced acid and gas but did not ferment lactose and sucrose, as these two gas-produced (+) isolates showed acid (K) on the slant and alkali (A) on the bottom (Figure 1C).

Interestingly, JS33 (K/A++) generated H₂S (+) (black), while JS34 (K/A + -) did not produce H₂S (−) (Figure 1C). Gram-stained microscopic examination revealed both isolates as Gram-negative bacilli due to the appearance of a loosely distributed red color (Figure 1D). Gram-negative bacterial identification cards also identified high similarity in JS33 and JS34 except for the production of H₂S (Table 1, bold text). The serotypes of the two suspected Salmonella isolates were 6, 8: r: l, w, which could be identified as Salmonella enterica serovar Goldcoast according to the White-Kauffmann-Le Minor antigenic table. The 2nd generation sequencing also identified JS33 and JS34 as S. Goldcoast and there were structural variations in the JS34 assembled genome and some of the original reads, compared with the reference genome (Figure 1E).

Based on the latest version of CLSI breakpoints, JS33 and JS34 were evaluated with MDR to ampicillin (AMP), compound sulfamethoxazole (or selectrin, SXT), chloramphenicol (CHL), tetracycline (TET), and streptomycin (STR) (Table 2). They were both intermediate-resistant to ampicillin/sulbactam (AMS), colistin (CT), and polymyxin (BPOL). Besides, JS33 was sensitive to cefazolin (CFZ), while JS34 was intermediate-resistant (Table 2). Both JS33 and JS34 were sensitive to azithromycin (AZM), ciprofloxacin (CIP), nalidixic acid (NAL), and gentamicin (GEN), as well as cefotaxime (CTX), ceftazidime (CAZ), cefoxitin (CFX), imipenem (IPM), amoxicillin/clavulanic acid (AMC), cefuroxime (CXM), cefepime (CPM), ceftazidime-avibactam (CZA), meropenem (MEM), ertapenem (ETP), tigecycline (TGC), and amikacin (AMI).

To elucidate the genetic background, we extracted DNA samples from purified JS33 and JS34 isolates and subjected them to deep sequencing using GridION (Oxford Nanopore Technology). This process yielded a comparable annotated sequence number in Non-Redundant (NR), Swiss-port, Kyoto Encyclopedia of Genes and Genomes (KEGG), and Clusters of Orthologous Genes (COG) databases in JS33 and JS34, respectively (Figure 2A). Ten protein-coding genes were exclusively expressed in either JS33 or JS34 (Figures 2A,B). Among them, the thiosulfate reductase (K08352), nitrate reductase (K02567), β-lactamase (K18698), and clavulanate-9-aldehyde reductase (K12677) were unique to JS33. Other proteins, such as fibronectin-binding autotransporter adhesin (K19231), DNA (cytosine-5)-methyltransferase 1(K00558), and REP-associated tyrosine transposase (K07491), were expressed in both JS33 and JS34, but with different gene numbers (Figure 2B).

Despite the distinct genome, the function classification identified high genetic similarities between JS33 and JS34. The COG function classification was consistent in both isolates, with only minor variations in the number of genes (Figure 2C). Amino acid transport and metabolism, carbohydrate transport and metabolism, transcription, cell wall/membrane/envelope biogenesis, and energy production and conversion were the top five gene-enriched functions, highlighting the shared roles of JS33 and JS34. The KEGG analysis consistently emphasized the roles of JS33 and JS34 in regulating metabolism and participating in genetic information processing (Supplementary Figure S1).

Deep sequencing also revealed that JS33 and JS34 were closely related to infectious disease and drug resistance, while one more gene was identified in JS33, which encoded β-lactamase class A (Figure 2B; Supplementary Figure S1). β-lactamases are the most common reason resulting in resistance to β-lactam antibiotics in Gram-negative bacteria (Bush and Bradford, 2019). By combing the detailed KEGG classification with gene identification, we found that the exclusively expressed K12677 and K18698 participated in the biosynthesis of secondary metabolites, butanoate metabolism, β-lactam resistance, and clavulanic acid biosynthesis. They were responsible for mild differences in drug response between JS33 and JS34 (Figure 3A). K02567 and K08352 participated in energy metabolism by regulating nitrogen and sulfur metabolism, respectively. Particularly, K08352 played an essential role in H₂S production (Figure 3B).

Moreover, the resistance gene identifier (RGI) identified 55 resistance genes (42 in config1 and 13 in config2, >50% identities, E-value<0.00001), while 53 ARGs were common in JS33 and JS34, and 45 of 55 ARGs showed more than 90% identities (Figure 4). Based on the analysis of virulence factors in pathogenic bacteria, we found that the gene encoded K19231 in JS33 was linked to the upaH gene, which regulates the AIDA-I type autotransporter protein, a rarely glycosylated protein. The gene encoded K02567 was associated with the nuoG gene and functioned as an anti-apoptosis factor. The gene encoded K08352 was related to narG and was involved in anaerobic respiration. However, the differentially expressed genes did not correlate with bacterial virulence in JS34.

Furthermore, the subcellular localization of secretory proteins in JS33 and JS34 were similar based on PSORTb analysis¹ (Yu et al., 2010). Most secretory proteins were located at cytoplasmic and fewer were in cytoplasmic membrane, while a few were in periplasmic. The Prophage prediction based on PHAge Search Tool Enhanced Release software (PHASTER) also showed high similarity in contig_1 and relatively less similarity in contig_2 between JS33 and JS34² (Arndt et al., 2016). Genomic island prediction based on Island Viewer and Crispr-Cas prediction based on CRISPR finder³ were the same as each other (Couvin et al., 2018). No difference was observed in JS33 and JS34 based on the carbohydrate-active enzymes database.