According to previous method (Xia et al., 2009), the S. enterica SC2020597 was isolated from a hospital in Guangyuan, Sichuan province, China in 2020. Then, pure cultured SC2020597 was identified as S. enterica using matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) (Bruker, Bremen, Germany). Meanwhile, the species of isolated Salmonella was further confirmed by an online rMLST analysis (Jolley et al., 2018). Subsequently, the minimum inhibitory concentrations (MICs) of SC2020597 was tested by broth microdilution according to Clinical and Laboratory Standards Institute (CLSI) guidelines (https://clsi.org/). E. coli ATCC25922 was used for the quality control.

In order to investigate the transfer ability of sul4, both conjugation assays and electroporation experiments were performed. For the conjugation assay, we used SC2020597 as donor strains and E. coli C600 as recipients. The donor and recipient strains were cultured into the logarithmic growth phase with an OD600 value of 0.4 in LB broth, then mixed at a ratio of 1:1 and cultured overnight on LB agar plates. The transconjugants were screened on LB agar plates containing rifampin (300 mg/L) and trimethoprim/sulfamethoxazole (4/76 mg/L). For the electroporation experiment, the genomic DNA of SC2020597 was used as donor DNA, and electrocompetent cells of S. enterica ATCC13076 were used as recipients. Electroporation conditions were 200 Ω, 1.8 kV and 25 uF. The transconjugants were screened on LB agar plates containing trimethoprim/sulfamethoxazole (4/76 mg/L). Then, all transconjugants were confirmed by PCR methods targeted at sul4 and 16S rDNA genes.

The genomic DNA of SC2020597 was extracted using FastPure Bacteria DNA Isolation Mini Kit (Vazyme™, China) following the protocol descripting in the manufacturer. The purity and quality of extracted genomic DNA were evaluated using NanoDrop (Thermo ScientificTM) and using a dsDNA High Sensitivity (HS) Assay kit on the Qubit 4 Fluorometer, respectively. Then, 200 μg genomic DNA was sent to GENEWIZ (Suzhou China) to subject short-read sequencing with PE150 strategy on Illumina Hiseq 2,500 platform. Meanwhile, long-read genomic sequencing of SC2020597 was conducted at Oxford Nanopore Technologies MinION platform in our laboratory. Briefly, the long-read sequencing library was prepared using the SQK-RBK109 1D Rapid Barcoding genomic DNA kit according to the user handbook. Then, the prepared library was sequenced with R9.4 flow cells on MinION and the sequencing process was managed with MinKNOW.

The cyclic plasmid form of the chromosomally integrated plasmid was detected using the inverse PCR method with primers cir_F: TTCAGACGGACTGGACATCG and cir_R: GCGGAATTGTTCAGGGGGTA. The genomic DNA of SC2020597 was used as template DNA. The master mix for application in long fragments was used for PCR amplification reaction system.

The short-read raw reads were filtered to remove low-quality base and adapters using fastp with default parameters (Chen et al., 2018). Then, the complete genome of SC2020597 was generated with a hybrid assembly strategy combining of clean short-read data and long-read data using Unicycler (Wick et al., 2017). Functional annotation of the complete genome was performed by a web-based RAST annotation engine (Aziz et al., 2008; Overbeek et al., 2014; Brettin et al., 2015). Antibiotic resistance genes (ARGs), plasmid replicon genes and insertion sequences (ISs) were identified using abricate tool (https://github.com/tseemann/abricate) based on AMRFinderPlus (Feldgarden et al., 2019), PlasmidFinder (Carattoli et al., 2014) and ISFinder (Siguier et al., 2006) databases, respectively. Multilocus sequence typing (MLST) of the complete bacterial genome was performed using mlst tool (https://github.com/tseemann/mlst). Using Salmonella genomes as input data, the web-based application SISTR was used to identify the Salmonella serovar (Yoshida et al., 2016). Plasmid comparisons and genetic context comparisons visualization were performed with BRIG (Alikhan et al., 2011) and Easyfig (Sullivan et al., 2011) tools.

To confirm the resistance function of the mutated sul4, TA-cloning was performed using a 5 min TA/Blunt-Zero Cloning Kit developed by Vazyme (Vazyme, China). Briefly, the sul4 gene and its predicted promoter were amplified by PCR using primers sul4_F: TGCCTGCAGGTCGACTCTAGAACCCAAAAGTCTGTAGCCCAAA, sul4_R: ACGGCCAGTGAATTGAGCTCTGGTCTAGTICAAAATCGATCATGT, and then cloned into pUC19 vector. Meanwhile, in order to verify the effect of the base mutation on the function of sul4, an unmutated sul4 recombinant expression plasmid was constructed. Subsequently, the recombinant plasmids were introduced chemically into E. coli DH5α. At last, we tested the resistance phenotype of the transconjugants using broth microdilution.

The genome sequences of SC2020597 were deposited into the National Center for Biotechnology information (NCBI) under BioProject PRJNA946266.

The isolate SC2020597 was recovered from a clinical patient. We identified it as S. enterica using MALDI-TOF-MS and confirmed by Ribosomal Multilocus Sequence Typing (rMLST) analysis. MLST analysis showed that isolate SC2020597 belonged to ST26 S. enterica. Serovar analysis classified the isolate as S. enterica subsp. enterica serovar Thompson, which is one of the most frequent Salmonella serovars involved in human infection (Eun et al., 2019). According to previous investigations, the prevalence of S. Thompson in clinical patients and food in China was 3.9 and 5.4%, respectively (Wang et al., 2017; Fan et al., 2020). The serovar has the potential to cause outbreaks of Salmonella infection. Antimicrobial susceptibility testing showed that the isolate was resistant to multiple antibiotics including kanamycin, ampicillin, tetracycline, sulfamethoxazole and trimethoprim/sulfomethoxazole, but sensitive to aztreonam, meropenem, ciprofloxacin, colistin and enrofloxacin (Supplementary Table S1).

The draft genome of SC2020597 was obtained by short-read genome assembly. Many ARGs, plasmid replicon genes and ISs were identified in the draft genome. Of note, we found a sul4 gene in isolate SC2020597, which had previously only been found in metagenomes (Razavi et al., 2017; Marathe et al., 2019; Hutinel et al., 2022). The sul4 gene in SC2020597 showed 100% coverage and 99.88% nucleic acid identity to sul4 (NG_056174). The one nucleic acid substitution of sul4 causes one amino acid change (W120R). To verify the function of the novel sul4, the intact sul4 gene and its promoter were cloned into pUC19 vector and introduced into E. coli DH5α. The sul4 positive transconjugants were resistant to trimethoprim/sulfomethoxazole and had a MIC for trimethoprim/sulfomethoxazole that was more than 16-fold higher than E. coli DH5 with an empty vector (≤ 1/19 mg/L to >32/608 mg/L). Meanwhile, unmutated sul4-bearing E. coli DH5 also showed full resistance to trimethoprim/sulfomethoxazole (>32/608 mg/L). This demonstrated that W120R substitution had no effect to the function of sul4.

To decipher the genomic structure feature of SC2020597, long-read sequencing was performed. Then, the complete genome of SC2020597 was generated by a hybrid assembly strategy using short-read and long-read data. The isolate harbored one chromosome with a length of 5, 035, 375 bp and two plasmids, pSC2020597_48k with a length of 48, 530 bp, and pSC2020597_5k with a length of 5, 754 bp. Plasmid replicon analysis showed that pSC2020597_48k was an IncFII type plasmid and pSC2020597_5k was a Col type small plasmid. Of note, no ARG was found in the two plasmids. Online blastn analysis showed that many plasmids from Salmonella were similar to pSC2020597_48k, indicating that such plasmids were common in Salmonella. Plasmid pSC2020597_5k belonged to a group of small plasmids with a broad host range that were widely distributed in Enterobacteriaceae.

A total of 22 ARGs, including sul4, were detected in SC2020597, and all of them were located on chromosome (Supplementary Table S2). In addition, three plasmid replicon genes were discovered on the chromosome of SC2020597. Further analysis found that a multi-drugs resistant plasmid was integrated into chromosome of SC2020597. The chromosomally integrated plasmid was 330, 232 bp in length and was designated as PSC2020597-sul4-330 k-c (Figure 1A). The phenomenon that plasmid was integrated into chromosome have been frequently reported (Muthuirulandi Sethuvel et al., 2021; Shigemura et al., 2021; Chang et al., 2022), especially in Salmonella, and is commonly mediated by homologous recombination of ISs. We found that PSC2020597-sul4-330 k-c was flanked by IS1F on chromosome of SC2020597. Meanwhile, a 7 bp direct repeat sequence was found around PSC2020597-sul4-330 k-c (Figure 1B). These findings provided strong evidences that PSC2020597-sul4-330 k-c was integrated into chromosome via the homologous recombination of IS1R. The precursor of PSC2020597-sul4-330 k-c was a hybrid plasmid with complex structure. Half of PSC2020597-sul4-330 k-c was composed by a typical IncHI2/HI2A plasmid. Another half was made up of flexible genetic arrays. Blastn analysis with NCBI nr database showed that PSC2020597-sul4-330 k-c was most closely related to pSIn_quan12 (GenBank: ON960352.1), which had 66% coverage and 99.98% identity to PSC2020597-sul4-330 k-c (Figure 1A). Meanwhile, many other plasmids similar to PSC2020597-sul4-330 k-c were also found in nr database, and the complete structure of PSC2020597-sul4-330 k-c could almost be covered by the genetic array of these plasmids (Figure 1A). Hence, PSC2020597-sul4-330 k-c was most likely formed through the recombination of diversity of plasmids and then integrated into chromosome. The discovery of sul4 in a complex chromosomally integrated plasmid implied that sul4 gene had already appeared and spread in plasmids. We should take notice to monitor the spread of sul4 by plasmids or other mobile genetic elements.

By the genetic analysis, we found that sul4 was located on a chromosomally integrated plasmid. However, a previous study demonstrated that chromosomally integrated plasmids could also be transferred to other recipients by conjugation assay (Chang et al., 2022). Here, we verified that there is a circle form of chromosomally integrated plasmid PSC2020597-sul4-330 k-c in some clones of SC2020597 using the inverse PCR method and confirmed by Sanger sequencing (Supplementary Figure S1). This phenomenon implied that the chromosomally integrated plasmid PSC2020597-sul4-330 k-c had potential to horizontal transfer by conjugation. Subsequently, we used E. coli C600 as recipients to test the transfer of sul4 by conjugation assay. The sul4 gene could not be transferred into E. coli C600 after multiple tries. Then, we attempted to verify the transfer ability of sul4 using electroporation experiments with S. enterica ATCC13076 as recipients. The sul4-positive S. enterica ATCC13076 was successfully screened and showed resistance to trimethoprim/sulfomethoxazole.

Mobile elements played a key role in the dissemination of ARGs (Partridge et al., 2018). Previous research discovered an ISCR family transposase ISCR20-like (GneBank: MG649402.1) downstream of sul4 that could be mobilized along with their adjacent genes via rolling-circle transposition without the assistance of any other transposase protein (Razavi et al., 2017). Therefore, sul4-ISCR20-like was considered to be mobilizable as an entire integrin (Razavi et al., 2017). Subsequently, we investigated the core genetic structure of sul4 in PSC2020597-sul4-330 k-c. We also found an ISCR20-like in the downstream of sul4, which was consistent with previous structure of sul4 in metagenome (GneBank: MG649402.1) (Figure 2). In addition, other sul4-bearing genetic contexts associated with ISCR20-like or truncated ISCR20-like were also found in other bacterial chromosome or plasmid in NCBI nr database (Figure 2). These findings demonstrated that the dissemination of sul4 was probably driven by ISCR20-like. Although the transfer of sul4-ISCR20-like has not been verified by experiment, other similar transfer events have been confirmed, such as tet(X4)-ISCR2 (He et al., 2019). Hence, the mobilization of sul4-ISCR20-like was likely to happen and promoted the dissemination of sul4. Of note, sul4 was found to be integrated into a class 1 integron in the plasmid of Aeromonas sp. FDAARGOS 1402 (Figure 2). It means that novel mobile genetic structure harboring sul4 is emerging, which may accelerate the propagation of sul4. Therefore, we should take more attention to the surveillance of sul4 in order to slow its spread.