Both the mSTM ST34 isolates from poultry (Sm78, Sm103) and the STM ST19 strain SL1344 were obtained from our lab’s collection. mSTM ST34 strain SSH006 was isolated from a female patient who showed resistance to imipenem and cephalosporin, two of the most often tested antimicrobial agents (Yang et al., 2017).

To achieve the objectives of this study, searches were conducted in the Enterobase databases to retrieve the genomes of STM and mSTM using the serovar option (“Typhimurium” OR “i:” OR “monophasic”) (Zhou et al., 2020). Complete genome sequences of Salmonella were obtained from National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov/). Additional data were sourced from the China National GeneBank Database under BioProject accession CNP0004753.The search strategy was followed to ensure that the released genome around the globe before February 2023 related to the prevalence of mSTM ST34 should be collected and included in this systematic research for further processing and screening. The complete information was summarized in Supplementary Table S1 .

Additionally, from 2019 to 2022, the Centers for Disease Control and Prevention and general hospitals collaborated on a clinical surveillance program for Salmonella infections in patients from Zhejiang Province, China. A total of 1,270 patients, aged 0 to 92 years, were clinically confirmed with a Salmonella infection diagnosis. Fecal samples were cultured according to the recommended protocol (Mikoleit, 2010). Diarrhea was defined as having three or more abnormally loose stools in the previous 24 hours. All isolated strains were serotyped by slide agglutination using commercial antiserum (SSI Diagnostica, Denmark). No sex preference was considered when recruiting patients. Demographic information, medical history, and both typical and atypical symptoms of Salmonella infection were provided by participants upon recruitment, and clinical data during hospitalization were collected from electronic medical records. Detailed information is provided in Supplementary Table S2 .

The Gram-negative bacterial genomic DNA extraction kit (DR0306050; Easy-Do) was used to extract DNA from isolates. Sequencing was performed using an Illumina Novaseq 6000 platform in 150 bp paired-end mode. De novo assembly was conducted by SPAdes v.3.12.0, as previously reported (Bankevich et al., 2012). All detected genomes have been deposited in the China National GeneBank DataBase under BioProject Accession CNP0004753.

All isolates were confirmed by in silico serotyping conducted by SISTR v.1.1.1 (Yoshida et al., 2016). Subsequently, multi-locus sequence typing (MLST) was performed using MLST v.2.19.0 (https://github.com/tseemann/mlst). The antimicrobial resistance genes were identified using ResFinder v.4.1 (>90% identity and >80% coverage) (Bortolaia et al., 2020). The plasmids were identified using MOB-suite (Robertson and Nash, 2018).

SNPs in the core genome were identified by Snippy v.4.4.5 (https://github.com/tseemann/snippy). The recombination regions were filtered from the core genome alignment by Gubbins v.2.3.4 (Croucher et al., 2015). The maximum-likelihood phylogenetic tree was generated using the best model (TVM+F) by IQ-TREE v.1.6.12 (Nguyen et al., 2015). TreeTime provides routines for ancestral sequence reconstruction and inference of molecular-clock phylogenies (Sagulenko et al., 2018).

In addition, the invasiveness index was assessed according to the method described previously (Wheeler et al., 2018). Briefly, we used the random forest model trained by the former method and R v.4.1.2 to calculate the invasiveness index. This model showed the applicability of calculating the invasiveness index for different species of Salmonella. Specifically, the DeltaBS metric was used to identify mutations in protein-coding genes in the whole genome sequence of the strains used for model building, from which 196 top predicted genes were obtained for measuring invasiveness in S. enterica.

Antimicrobial susceptibility testing was performed using the broth dilution method. The minimal inhibitory concentrations (MIC) of eight antimicrobial agents—ampicillin, amoxicillin/clavulanic acid, gentamicin, tetracycline, spectinomycin, streptomycin, trimethoprim-sulfamethoxazole and meropenem—were determined. The results were interpreted according to Clinical and Laboratory Standards Institute guidelines (Patel, 2016). Escherichia coli ATCC 25922 was used as a quality control. Isolates that showed resistance to at least three classes of antimicrobials were defined as MDR.

A 1:100 dilution of a saturated culture was incubated overnight in LB or M9 medium (6 g/L Na₂HPO₄, 3 g/L KH₂PO₄, 1 g/L NH₄Cl, 0.5 g/L NaCl, 2 g/L glucose, 246.5 mg/L MgSO₄·7H₂O, 1 mg/L thiamine·HCl, 14.7 mg/L CaCl₂, pH 7 ± 0.2 at 25°C) with shaking at 180 rpm at 37°C. The OD₆₀₀ was measured at different time points. To rescue SSH006 growth, casamino acids (CAS: 65072-00-6;C304284; Aladdin; Shanghai, China) were added to the culture at 4 h.

The overnight culture was diluted to 0.06–0.08 of OD₆₀₀. Three microliters of the bacterial suspension were added to the center of LB medium plate containing 0.3% agar. The plates were incubated statically at 37°C for 6 h under aerobic and anaerobic conditions. The displacement was measured via a vernier caliper at 6-hour intervals. Three replications were performed for each group.

The strain was washed and resuspended in 1 × phosphate-buffered saline (PBS). A bacterial suspension equivalent to ~10⁷ colony-forming unit (CFU)/ml was added to PBS and incubated at 180 rpm at 37°C. The number of colonies and the OD₆₀₀ were measured every day for nine days. Relative survival was calculated as the ratio of CFU at each time point to CFU at time zero. At least three independent trials were performed.

Specific pathogen-free Babcock chicken embryos were obtained from Shennong Chick Farm. The 19-day-old embryos were hatched under biosecurity conditions in a 37.2°C incubator. Chicks were fed under aseptic conditions until 1 day of age. Water and food were withheld for 3 hours before and after gavage with mSTM ST34 isolates or SL1344 at a concentration of 10⁸ CFU/mL. Chicken survival (n=17~37 per group) was monitored daily for 7 days. At 2 and 5 days post-infection (dpi), the heart, liver, and spleen (n=3, per group) were weighed and homogenized in PBS. Samples were serially diluted and dispensed onto xylose lysine deoxycholate (XLD) agar to count the number. Feces were collected during the 7-day monitoring period.

Six to eight-week-old female C57BL/6N mice were purchased from Vital River Laboratory Animal Technology (Beijing, China). The mice were maintained under a 12-hour light/dark cycle and fed a standard chow diet in a specific pathogen-free (SPF) facility at the Laboratory Animal Center of Zhejiang University, following the National Institutes of Health Guide for the Care and Use of Laboratory Animals. The streptomycin-pretreated mouse colitis model was described in a previous study (Barthel et al., 2003). Water and food were withheld for 4 hours before gavage with 20 mg of streptomycin, after which they were reintroduced. Twenty hours later, water and food were again withheld for 4 hours. Single colonies were inoculated to an OD600 of ~0.1 (~10^8 CFU/mL) and then diluted 100-fold. The mice were infected intragastrically with 100 μL of the bacterial suspension. Water and food were provided 1 hour post-infection. At 3 and 5 dpi, the infected mice were sacrificed, and tissue samples were collected. Tissues were homogenized in PBS buffer, and CFU were determined by serial dilution plating on XLD solid plates containing 100 μg/mL streptomycin or 100 μg/mL streptomycin plus 50 μg/mL ampicillin. Samples that fell below the detection limit were recorded as 0 CFU.

For histopathology staining, the tissue samples fixed in 4% paraformaldehyde (BL539A; biosharp; Hefei, China) were embedded in paraffin, sliced (4 μm thickness), dehydrated and then sectioned for hematoxylin and eosin (HE) staining.

The Rdar morphology was determined as previously described (Singletary et al., 2016). The 3 μL of ~100 CFU/ml bacterial solution was dripped in lysogeny broth (LB) plates without salt and supplemented with 40 μg/mL Congo red (CAS: 573-58-0; Aladdin; Shanghai, China) and 20 μg/mL Coomassie blue (CAS: 6104-58-1; Aladdin; Shanghai, China). Plates were incubated at 37°C for 5 days without inversion. To ensure accurate results, duplicates of every experiment were conducted.

Ethical approval was granted by the Ethics Committee of the Zhejiang Centers for Disease Control and Prevention (2019-014). No potentially identifiable images or data were included in this article, and the data cannot be used to identify individuals. Written informed consent for participation was not required for this study in accordance with national legislation and institutional requirements. All experimental protocols (ZJU20190093; ZJU20220295) involving animals were conducted according to the ethical policies and procedures approved by the Review Committee of the College of Animal Sciences at Zhejiang University.

Quantitative data, exponential (Malthusian) growth, and Pearson correlation were supported by GraphPad Prism 9. All statistical details of experiments, including statistical tests, the exact value of n, what n represents, the p-value and the means with standard deviations, were provided in the corresponding figure legends. Figure layout and proper annotation were accomplished by Adobe Illustrator 2021 (https://www.adobe.com). Schematics were created using BioRender (https://biorender.com).

To project a global picture of mSTM ST34, we included the analysis of a total of 44,282 isolates from lab collections and various databases ( Supplementary Table S1 ). Most mSTM ST34 isolates are from humans, followed by poultry and livestock ( Figure 1A ). We further examined the mSTM ST34 isolates from humans. Over the past two decades, the prevalence of mSTM ST34 showed a significant increase from 1.7% in 2001-2005 to 44.5% in 2016-February 2023 ( Figure 1B ). This growth was most noticeable in Europe. STM ST19 persisted among human isolates in the 1990s, but from 2001 onward, the overall growth rate of STM ST19 was lower than that of mSTM ST34 ( Figure 1C ). Spatiotemporal evolution analysis showed that the ST34 strain formed a distinct branch, separating it from ST19, and mSTM ST34 derived from STM ST34 introduced into different continents and set off a global pandemic that affected livestock, poultry, and humans ( Figure 1D ). Overall, the widespread distribution of mSTM ST34 poses a significant challenge for livestock and poultry production as well as clinical treatment.

The whole genome sequences were screened for the detection of antimicrobial determinants encoding resistance to various antimicrobial categories. Our findings revealed a steady high prevalence of MDR mSTM ST34 worldwide during the past two decades, with 80% to 100% of the isolates showing MDR ( Figure 2A ). Compared to STM ST19, mSTM ST34 isolates showed higher resistance to several classes of antibiotics, including aminoglycosides (tobramycin, 99.7%; streptomycin, 83.8%), β-lactams (amoxicillin, 85.3%; ampicillin, 85.6%; piperacillin, 85.2%; ticarcillin, 85.2%), sulfonamides (sulfamethoxazole, 83.1%), and tetracyclines (doxycycline, 86.2%; tetracycline, 86.6%) ( Figure 2B ). Additionally, we revealed that the mSTM ST34 population contains 116 conjugative plasmid types ( Figure 2C ). Notably, conjugative IncQ1 showed a high correlation with the MDR pattern of mSTM ST34 (Pearson correlation, 0.781) ( Figure 2D ).

From April 2019 to September 2022, 1270 patients infected with Salmonella were identified from hospitals in Zhejiang Province, China ( Figure 3A ). The main serovars included mSTM (344 isolates), Typhimurium (276), Enteritidis (196) ( Figure 3B ). The dominant subtype of mSTM isolates was ST34 (339/344), which was included in this study cohort ( Figure 3C ). The complete information about these patients was summarized in Supplementary Table S2 . The surveillance results showed that 153 children (≤12 years old) and 20 elderly (≥65 years old) were infected with MDR mSTM ST34. One outbreak affected seven patients (24~45 years old) in a hospital in Quzhou City. Notably, genomic analysis revealed that some isolates are resistant to clinical antibiotics, including ciprofloxacin (37.8%), third-generation cephalosporins [cefixime (0.3%); cefotaxime (23.3%); ceftazidime (23.3%); ceftriaxone (22.1%)], fourth-generation cephalosporin [cefepime (27.1%)], azithromycin (4.1%) and colistin (0.6%) ( Figure 3D ).

Consistent with global isolates, compared with highly invasive STM ST313, these mSTM ST34 isolates have a lower invasiveness index ( Figure 3E ), indicating that mSTM ST34 mainly maintains an intestinal lifestyle. This corresponds to most patients having diarrhea symptoms (295/339) instead of septicemia ( Supplementary Table S2 ).

In this study, we compared the traits of mSTM ST34 and STM ST19, represented by multi-drug resistant mSTM ST34 isolates Sm78, Sm103, SSH006 ( Figure 4A ), alongside the STM ST19 isolate SL1344. The four isolates grew rapidly in LB medium and did not show any difference ( Figure 4B ). However, in the nutrient-poor M9 medium, SSH006 was barely able to grow in the M9 medium ( Figure 4C ). The complement of casamino acids rescued the growth of SSH006. Despite the poor growth in M9 medium, SSH006 was able to persist in a nutrient-poor PBS solution, and its growth condition was recovered on day 9 ( Figure 4D ). mSTM ST34 only expressed fliC-encoded flagella, but had no defect in motility ( Figure 4E ). In addition, SSH006 was more motile under anaerobic conditions than under aerobic conditions.

The virulence capacity of the mSTM ST34 isolates was first assessed by in vivo chick infection ( Figure 5A ). The chicks infected with mSTM ST34 isolates showed the lower survival rate (Sm103, 54.5%; SSH006, 61.8%; Sm78, 79.5%), compared to SL1344 (80.4%) ( Figure 5B ). The infection feature assessed at 2 and 5 dpi showed a higher bacterial burden in the mSTM ST34-infected groups ( Figure 5C ). Additionally, continuous shedding of mSTM ST34 was observed throughout the 7-day monitoring period” to “observed throughout the 7-day monitoring period ( Figure 5D ).

We further utilized a streptomycin-pretreated mouse model to evaluate infection phenotypes ( Figure 6A ). The bacterial loads in the liver, spleen, colon, and cecum were determined at 3 and 5 dpi ( Figure 6B ). In comparison to the systemic typhoid-like infection SL1344 group, mSTM ST34 isolates exhibited a preference for intestinal colonization. Specifically, mSTM ST34 isolates were not detected in some spleen samples (8 out of 15), and the loads of mSTM ST34 in the colon and cecum were significantly higher than those of SL1344 at 3 dpi. Pathological sections also indicated that the cecum, rather than the liver and spleen, suffered more severe damage in the mSTM ST34 isolate-infected groups at 3 dpi ( Figure 6C ).

Notably, mSTM ST34 SSH006 isolated from a patient showed substantial biofilm formation ( Figure 6D ). The mature biofilm on the Congo red plate had a distinct three-dimensional structure, with elongated cells expressing curli and cellulose. Biofilm formation protects mSTM ST34 from antibiotics and complicates clinical treatment. Overall, mSTM ST34 is a zoonotic pathogen that causes high mortality in poultry and leads to intestinal diseases, posing significant challenges for poultry production and clinical treatment ( Figure 7 ). We observed considerable strain-specific dynamic in virulence and other phenotype.