First report of a carbapenem-resistant Serratia sarumanii clinical strain co-harboring bla_KPC-2 and bla_NDM-1 genes in China

Highlights

• •First report of bla_KPC-2 and bla_NDM-1 co-occurrence in Serratia sarumanii from China.
• •bla_KPC-2 on novel IncFII/IncFIB plasmid; bla_NDM-1 on conjugative IncX3 plasmid.
• •bla_KPC-2 plasmid showed full stability; bla_NDM-1 region excised after passage.
• •Strain shows MDR but retains susceptibility to tigecycline and aminoglycosides.
• •Findings highlight urgent need for CRE surveillance in emerging Serratia species.

Abstract

Carbapenem-resistant Enterobacteriaceae (CRE), particularly those co-harboring multiple carbapenemase genes, pose a significant global health threat. However, the coexistence of bla_KPC-2 and bla_NDM-1 in Serratia sarumanii has not been previously reported. This study aims to report and characterize the first carbapenem-resistant S. sarumanii (CRSS) clinical strain MAS3954 in China co-harboring bla_KPC-2 and bla_NDM-1, focusing on its genetic characteristics, plasmid stability, and transfer potential. Whole-genome analysis revealed that the bla_KPC-2 and bla_NDM-1 were located on two distinct plasmids. Plasmid pMAS3954-KPC (113,856 bp, IncFII/IncFIB) exhibited low similarity (<66%) to known plasmids, indicating a novel fusion event between pKPC-h2 and S. marcescens chromosome, while pMAS3954-NDM (55,235 bp, IncX3) was highly conserved (100% identity/coverage). Conjugation experiments showed that the bla_NDM-1 was transferable, while bla_KPC-2 was not. During 10 days of continuous passage, the genetic context of bla_NDM-1 was gradually excised from the plasmids after the 8th day, whereas they maintained 100% retention for bla_KPC-2. S. sarumanii MAS3954 was multidrug-resistant (MDR), including carbapenems, β-lactams, β-lactam/β-lactamase inhibitors, trimethoprim-sulfamethoxazole, tetracycline, nitrofurantoin, colistin, and fluoroquinolones, but remained susceptible to certain aminoglycosides and tigecycline. Phylogenomic analysis identified a distinct clade for S. sarumanii MAS3954, diverging notably from other strains. Comparison of resistance genes further highlighted the unique co-harboring of bla_KPC-2 and bla_NDM-1 in MAS3954, absent in other strains. To our knowledge, this study represents the first characterization of clinical S. sarumanii strain co-harboring bla_KPC-2 and bla_NDM-1. The findings highlight the need for enhanced surveillance and infection control to prevent the spread of these MDR strains in healthcare settings.

Graphical abstract

1. Introduction

The Serratia marcescens complex (SMC) comprises a group of phenotypically similar yet genetically distinct species within the Serratia genus, including S. marcescens, S. nematodiphila, S. bockelmannii, S. ureilytica, and S. nevei (Ono et al., 2022). Although a formal taxonomic definition of the SMC has not yet been established by the National Center for Biotechnology Information (NCBI), recent genomic studies have proposed these species as its core members (Aracil-Gisbert et al., 2024). The taxonomic position of S. sarumanii, a newly proposed species isolated from clinical samples, remains uncertain and is not currently included within the established SMC framework (Klages et al., 2024).

S. sarumanii is an emerging Gram-negative, rod-shaped, urease-negative bacterium that has been isolated from clinical specimens such as wounds and urine (Klages et al., 2024). Sharing pathogenic characteristics with S. marcescens, S. sarumanii is implicated in infections of the wound, urinary tract, and potentially the bloodstream (Bes et al., 2021; Cosimato et al., 2024; Klages et al., 2024; Tuttobene et al., 2024). Its environmental resilience and capacity to persist on surfaces enable it to spread within hospital settings through contaminated medical devices and other surfaces, highlighting the urgent need for stringent infection control measures (Klages et al., 2024). Notably, S. sarumanii exhibits significant resistance to antimicrobials such as penicillins, cephalosporins, and certain extended-spectrum β-lactams, a trait attributed to unique resistance genes such as bla_SRT-2 (Klages et al., 2024). The similarities between S. sarumanii and S. marcescens in nosocomial infection patterns, transmission mechanisms, and resistance profiles emphasize the broader implications for infection control and treatment strategies in healthcare-associated infections.

Carbapenems, a critical class of β-lactam antimicrobials, are considered last-line treatments for infections caused by multidrug-resistant (MDR) Gram-negative bacteria (Tacconelli et al., 2018). However, the increasing prevalence of carbapenemase-encoding genes, notably bla_NDM-1 (New Delhi metallo-β-lactamase-1) and bla_KPC-2 (Klebsiella pneumoniae carbapenemase-2), poses a substantial threat to their efficacy (Wang et al., 2018; Bush and Bradford, 2020). These genes enable the hydrolysis of carbapenems, rendering them ineffective and facilitating the rapid spread of resistant strains through horizontal gene transfer mechanisms (Nordmann et al., 2011; Cantón et al., 2012; Fu et al., 2024; Li et al., 2024). The emergence of carbapenem-resistant clinical strains has become a global health challenge, particularly when multiple carbapenemase genes co-exist in a single strain, further limiting treatment options (Bush and Bradford, 2020; Gao et al., 2020). Among carbapenemases, KPC and NDM are the most frequently identified worldwide (Bush and Bradford, 2020). However, strains co-harboring both bla_KPC-2 and bla_NDM-1 remain relatively rare (Wu et al., 2015; Wang et al., 2017; Bes et al., 2021; Qiao et al., 2023; Yuan et al., 2023; Zhang et al., 2023; Chen et al., 2024; Zhang et al., 2024). Notably, the presence of strains co-harboring these genes in S. sarumanii has not been previously reported.

In this study, we identified a carbapenem-resistant S. sarumanii strain co-harboring bla_KPC-2 and bla_NDM-1 from a sputum specimen. The bla_KPC-2 and bla_NDM-1 genes were located on two distinct plasmids, and an analysis of their genetic environments, in conjunction with NCBI data, revealed potential mechanisms of dissemination. Furthermore, we assessed the genetic stability of these resistance genes, offering valuable insights into their persistence and dissemination. In addition, a phylogenomic analysis was conducted to assess the evolutionary relationship of S. sarumanii MAS3954 with other strains. This highlighted its distinct resistance gene profile and phylogenetic divergence, emphasizing its clinical and epidemiological significance as a high-risk carbapenemase-producing lineage within the species.

2. Materials and methods

2.1. Bacterial strain and initial identification

S. sarumanii MAS3954 was isolated from the sputum of a patient in Ma'anshan People's Hospital, Anhui, China, in 2023, as part of routine diagnostic and public health surveillance. Initial identification was performed using Vitek 2 (GN cards), matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS), and 16S rRNA gene sequencing. PCR and Sanger sequencing using previously reported primers further confirmed that this strain carries both bla_KPC and bla_NDM genes (Poirel et al., 2011). Sequence analysis was conducted using the Basic Local Alignment Search Tool (BLAST) (https://blast.ncbi.nlm.nih.gov/, accessed on 8 June 2024). The S. sarumanii MAS3954 strain has been deposited in the Center for Human Pathogenic Culture Collection at the China CDC.

2.2. Antimicrobial susceptibility testing (AST)

The minimum inhibitory concentrations (MICs) were determined using Sensititre^TM Gram Negative GN4F AST Plate (Thermo Fisher Scientific, Waltham, MA) with the broth microdilution method. MICs for colistin and ceftazidime-avibactam were also assessed using the broth microdilution method. A total of 26 antimicrobials were tested, including ampicillin, piperacillin, ampicillin-sulbactam, ceftazidime-avibactam, ticarcillin-clavulanic acid, piperacillin-tazobactam, aztreonam, cefazolin, cefepime, ceftazidime, ceftriaxone, amikacin, gentamicin, tobramycin, ciprofloxacin, levofloxacin, doripenem, ertapenem, imipenem, meropenem, nitrofurantoin, tetracycline, minocycline, tigecycline, trimethoprim-sulfamethoxazole, and colistin. The tigecycline results were interpreted according to the U.S. Food and Drug Administration (FDA) guidelines (Marchaim et al., 2014). Other antimicrobial agents were interpreted following the standards of the Clinical and Laboratory Standards Institute (CLSI, 2024, https://clsi.org) (CLSI, 2024). Escherichia coli ATCC 25922 was used as the quality control strain.

2.3. Whole-genome sequencing (WGS) and bioinformatics analysis

Genomic DNA was extracted using the Wizard Genomic DNA Extraction Kit (Promega, Madison, WI, USA) and sequenced on the Illumina PE150 and PacBio Sequel IIe platforms. Hybrid assembly was performed using Canu v2.2 (Koren et al., 2017) after quality control. Assembly quality was assessed using BUSCO v5.5.2 (Manni et al., 2021). Bacterial strain identities were confirmed through Average Nucleotide Identity (ANI) calculations using FastANI v1.334 (Jain et al., 2018) and the Genome-to-Genome Blast Distance Phylogeny (GBDP) algorithm, which simulates digital DNA-DNA hybridization (dDDH) for species delineation (Meier-Kolthoff et al., 2013). Species boundaries were defined with a minimal ANI cutoff of 95-96% and a dDDH cutoff of 70% (Goris et al., 2007; Richter and Rosselló-Móra, 2009; Meier-Kolthoff et al., 2013; Chun et al., 2018). Further strain identification at the genus and species levels was performed using the TYGS platform (https://tygs.dsmz.de/, accessed on 27 August 2024) (Meier-Kolthoff and Göker, 2019). Genome annotation was conducted using Prokka v1.14.6 (Seemann, 2014), Prodigal v2.6.3 (Hyatt et al., 2010), and RAST 2.0 (Aziz et al., 2008) (https://rast.nmpdr.org/, accessed on 2 August 2024), in combination with the Pfam 37.0 (Mistry et al., 2021) and NCBI-NR databases. Insertion sequences were identified through the ISfinder database (Siguier et al., 2006). Antimicrobial resistance genes were predicted using Resfinder v4.3.1 (Clausen et al., 2018; Bortolaia et al., 2020) and the Comprehensive Antibiotic Resistance Database (CARD, http://arpcard.mcmaster.ca) (Alcock et al., 2023). The Resistance Gene Identifier (RGI) v6.0.3 was used with parameters set to an E-value of 1e-5, identity ≥80%, and coverage ≥80%. PlasmidFinder v2.1.6 (Carattoli et al., 2014) was used to identify plasmid replicons. The in silico prediction of transferability of bla_KPC-2 and bla_NDM-1-harboring plasmids was evaluated using oriTfinder (https://bioinfo-mml.sjtu.edu.cn/oriTfinder/, accessed on 6 September 2024 and again on 6 May 2025) (Li et al., 2018). BLAST in the NCBI database was employed to identify sequences with high plasmid similarity. Sequence comparisons and map generation were performed using Easyfig 2.2.5 (Sullivan et al., 2011) and BLAST Ring Image Generator (BRIG) 0.95 (Alikhan et al., 2011).

2.4. Plasmid conjugation assay and plasmid stability testing

Conjugation experiments were conducted to assess the transferability of bla_KPC-2 and bla_NDM-1-harboring plasmids of S. sarumanii MAS3954 using filter mating (Shintani et al., 2019). Streptomycin-resistant E. coli EC600 was used as the recipient strain. Transconjugants were selected on MacConkey agar supplemented with streptomycin (1500 µg/mL) and meropenem (2 µg/mL), and confirmed by MALDI-TOF MS and PCR. AST of both the recipient strain and transconjugants was performed as described above. The conjugation frequency (CF) was calculated as follows: CF = (number of transconjugant bacterial colonies × 10y) / (number of recipient bacterial colonies × 10x), where x and y represent the dilution factors. Plasmid stability testing was performed to evaluate the stability of bla_KPC-2 and bla_NDM-1-harboring plasmids as described previously (Zhang et al., 2023), with slight modifications. Briefly, strains were cultured in Luria-Bertani (LB) broth at 37 °C with shaking (200 rpm), and transferred every 24 hours into fresh LB broth at a 1:1000 dilution for 10 consecutive days (approximately 200 generations). On days 2, 4, 6, 8, and 10, serial dilutions of the cultures were plated on LB agar, and approximately 60 colonies were randomly selected and screened by PCR for the presence of bla_KPC-2 and bla_NDM-1 (Poirel et al., 2011). All experiments were performed in triplicate at each time point.

2.5. Fitness cost assessment

Fitness cost was assessed through growth curve analysis. The growth kinetics of the strains were measured using a Bioscreen fully automated microbial growth curve analyzer at 37°C and OD600. The experiment was performed in triplicate. Growth curves were compared using Tukey’s multiple comparison test following a one-way analysis of variance (ANOVA).

2.6. Phylogenomic analysis

A total of ten S. sarumanii genomes, including MAS3954 and nine publicly available assemblies retrieved from the NCBI GenBank database (https://www.ncbi.nlm.nih.gov/, accessed on 4 May 2025), were included in this analysis (Table S2). Prokka v1.14.6 was used to generate GFF3-formatted output files for each strain (Seemann, 2014). Core genome analysis was conducted with Roary v3.13.0, using a 95% identity threshold to identify conserved genes across the dataset (Page et al., 2015). SNP-sites v2.5.1 (Page et al., 2016) was used to generate a SNP matrix from the core gene alignment, followed by recombination removal using Gubbins v3.3.1 (Croucher et al., 2015). A maximum-likelihood phylogenetic tree was constructed based on the core-genome SNP alignment using IQ-TREE v2.3.4 with the GTR+G substitution model and 1,000 bootstrap replicates for branch support (Minh et al., 2020). The final tree was visualized using iTOL v7 (https://itol.embl.de).

2.7. Statistical analyses

Data were analyzed using GraphPad Prism 9.5.0. One-way ANOVA was applied to assess significant differences, with a P-value of < 0.05 considered statistically significant.

2.8. Data availability

The complete genome sequence of strain MAS3954 sequenced in this study has been deposited at GenBank under BioProject ID PRJNA1261644. The 16S rRNA gene sequence of strain MAS3954 has been deposited in GenBank under accession number PV597925.

3. Results

3.1. Initial identification and genome-wide comparisons

Clinical MALDI-TOF MS analysis identified strain MAS3954 as S. marcescens, with a score above 2.28, indicating “high-confidence identification”. dDDH analysis based on the complete genome sequence revealed that strain MAS3954 was closely related to several type strains, including S. sarumanii K-M0706^T, S. marcescens ATCC 1388^T, S. nevei S15^T, S. bockelmannii S3^T, and S. ureilytica JCM 16474^T (Table 1). Strain MAS3954 exhibited the highest dDDH value (93.0%) with S. sarumanii, confirming that it is the closest relative. In contrast, S. sarumanii shared a maximum dDDH value of only 69.7% with S. nevei, the closest match among other Serratia type strains. Based on the dDDH thresholds of 70% for species delineation and 79% for subspecies classification (Meier-Kolthoff and Göker, 2019), strain MAS3954 was classified as a member of S. sarumanii. To further support this classification, the ANI was also calculated for the type strains (Table S1).

Table 1.

Digital DNA-DNA Hybridization (dDDH) analysis of strain MAS3954 and related Serratia species (%).

Strains	MAS3954	S. sarumanii K-M0706T	S. marcescens ATCC 13880T	S. nevei LMG 31536T	S. bockelmannii LMG 31535T	S.ureilytica JCM 16474T
MAS3954	100	93.0	63.7	69.7	63.1	58.4
S. sarumanii K-M0706T	93.0	100	63.8	69.7	63.7	58.4
S. marcescens ATCC 13880T	63.7	63.8	100	62.1	66.6	62.3
S. nevei LMG 31536T	69.7	69.7	62.1	100	62.8	58.1
S. bockelmannii LMG 31535T	63.1	63.7	66.6	62.8	100	62.7
S.ureilytica JCM 16474T	58.4	58.4	62.3	58.1	62.7	100

3.2. General features of S. sarumanii MAS3954, plasmid pMAS3954-KPC, and plasmid pMAS3954-NDM

The genome of S. sarumanii MAS3954 comprised a complete circular chromosome (5,214,053 bp, 59.97% GC content) and two plasmids. The bla_KPC-2 and bla_NDM-1 genes were located on plasmid pMAS3954-KPC (113,856 bp, 54.79% GC content) and pMAS3954-NDM (55,235 bp, 49.12% GC content), respectively. These plasmids belonged to the IncFII/IncFIB and IncX3 incompatibility groups (Table 2). S. sarumanii MAS3954 harbored multiple acquired resistance determinants to β-lactams, including bla_KPC-2, bla_NDM-1, bla_SRT-2, bla_TEM-1B, ble_MBL, and bla_SHV-12. Notably, antimicrobial resistance (AMR) genes such as bla_SRT-2, aac(6′)-Ic, rsmA, msbA, emrR, EF-Tu, CRP, and KpnH were located on the chromosome (Table 2).

Table 2.

Genetic characteristics of S. sarumanii MAS3954.

Genetic characteristics	MAS3954 chromosome	pMAS3954-KPC	pMAS3954-NDM
Size	5,214,053 bp	113,856 bp	55,235 bp
GC Content (%)	59.97	54.79	49.12
CDSs	4,754	134	69
rRNA	22	-	-
sRNA	109	-	-
tRNA	95	-	-
β-lactamase genes	blaSRT-2	blaTEM-1B, blaKPC-2	blaSHV-12, blaNDM-1, bleMBL
Other resistance genes	aac(6′)-Ic, rsmA, msbA, emrR, EF-Tu, CRP, KpnH	qnrS1	-
blaKPC-2 gene location	-	IncFII/IncFIIB plasmid	-
blaNDM-1 gene location	-	-	IncX3 plasmid

Note: -, none.

The complete nucleotide sequences and sizes of the bla_KPC-2 and bla_NDM-1-harboring plasmids are shown in Fig. 1A and Fig. 1B. Plasmid pMAS3954-KPC exhibited low coverage (<66.0%) compared to plasmids available in the NCBI database, suggesting that it represents a previously unreported fusion plasmid. Comparative analysis revealed that pMAS3954-KPC may have originated from the fusion of the plasmid pKPC-h2 (MT550691.1) from a K. pneumoniae strain and the chromosome of S. marcescens CM5 (AP028487.1). The plasmid encoded a predicted origin of transfer (oriT) region and a relaxase gene, as well as a partial type IV secretion system (T4SS) cluster, but lacked a type IV coupling protein (T4CP), which is essential for conjugation (Fig. 1A).

Fig. 1.

Comparative analysis of plasmids pMAS3954-KPC and pMAS3954-NDM. (A) Comparison of the pMAS3954-KPC with pKPC-h2 (MT550691.1) and S. marcescens CM5 chromosome (AP028487.1). (B) Comparison of the pMAS3954-NDM with pF3517-NDM (CP137176.1), pNDM1_090351 (CP046892.1), p128379-NDM (MF344560.1), and pZHDC33 (KX094555.1). The diagram also shows the locations of predicted conjugation elements based on oriTfinder analysis.

Plasmid pMAS3954-NDM carried genes associated with plasmid stability (parA, parB, and H-NS), regulation (tnpR, deoR, and mobP1), and conjugation transfer, including kikA, relaxase, the T4CP such as virD4, and T4SS components virB1-11. Additionally, pMAS3954-NDM harbored critical AMR genes, including bla_SHV-12 and bla_NDM-1. Comparative analysis revealed that pMAS3954-NDM shared 100% coverage and 100% identity with ten plasmids in the NCBI database, including pF3517-NDM (CP137176.1) from C. freundii F3517, pNDM1_090351 (CP046892.1) from K. pneumoniae SCKP090351, p128379-NDM (MF344560.1) from E. hormaechei 128379, and pZHDC33 (KX094555.1) from E. coli ZHDC33 (Fig. 1B).

3.3. Genetic environment of bla_KPC-2 and bla_NDM-1 genes

The genetic context of bla_KPC-2 on plasmid pMAS3954-KPC was defined by a distinctive arrangement spanning 21,872 bp. This configuration included the sequence: ISKpn19-tnpR-qnrS1-tnpR-IS3-bla_TEM-1B-tnpR-Tn3-IS26-ISKpn27-IS26-bla_KPC-2-ΔISKpn6-korC-klcA-tnpR-TnAs1-ISKpn19 (Fig. 2A). The core genetic environment of bla_KPC-2, comprising IS26-ISKpn27-IS26-bla_KPC-2-ΔISKpn6-korC-klcA, was classified as a bla_KPC-bearing non-Tn4401 elements (NTE_KPC-I) (Chen et al., 2014). Compared to plasmid p3024-KPC, pMAS3954-KPC contained ΔTn6296, but the internal composition and arrangement of its elements were distinct. This included retaining key mobile genetic elements (MGEs), such as IS26, ISKpn27, TnAs1, and ISKpn19, while also incorporating the additional bla_TEM-1B gene. In contrast to p1-1613258639, the bla_KPC-2 region of pMAS3954-KPC was conserved, but significant differences were observed in the surrounding elements, including the truncation of Tn3 and ΔISKpn6. Additionally, the presence of bla_TEM-1B and qnrS1 indicated that this plasmid carried MDR determinants.

Fig. 2.

Genetic characteristics of bla_KPC-2 and bla_NDM-1 gene contexts in S. sarumanii MAS3954. (A) Linear alignment of the genetic context of bla_KPC-2 gene in plasmid p3024-KPC (CP047683.1), pMAS3954-KPC, and p1-165238639 (CP083056.1). (B) Linear alignment of the gene context of bla_NDM-1 in plasmid pEC9-NDM-1 (CP060952.1), pMAS3954-NDM, p128379-NDM (MF344560.1), pF3517-NDM (CP137176.1), and pA575-NDM (MH917283.1). Antimicrobial resistance (AMR) genes and insertion sequences are highlighted with red and green arrows, respectively.

The bla_NDM-1 gene on pMAS3954-NDM resided within a composite structure of ∆Tn3000, ∆Tn125, and ΔTn3, with a total length of 14,549 bp, represented as ISCR27-groEL-ΔgroES-cutA-dsbD-trpF-ble_MBL-bla_NDM-IS5-ΔISAba125-IS5-IS3000-ΔTn3 (Fig. 2B). This structure was frequently identified in bla_NDM-1-harboring Enterobacteriaceae in China. Upstream of this composite structure was MGE based on IS26, including insertion sequences IS26, ISKox3, and AMR genes such as bla_SHV-12. Sequence comparisons revealed minor variations in the immediate genetic context of bla_NDM-1 across five plasmids (pMAS3954-NDM, p128379-NDM, pF3517-NDM, pEC9-NDM-1, and pA575-NDM).

3.4. Transferability, plasmid stability, and fitness cost of bla_KPC-2 and bla_NDM-1 co-harboring CRSS strain

The transferability of bla_KPC-2 and bla_NDM-1 was assessed through conjugation assays. The bla_NDM-1 gene was successfully transferred into E. coli EC600 with a conjugation frequency of 2.2 × 10^-2 (Table 3), while the transfer of the bla_KPC-2 was not transferable. The successful transfer of bla_NDM-1 and the absence of bla_KPC-2 transfer were confirmed by PCR. Plasmid stability testing demonstrated that bla_KPC-2 and bla_NDM-1-harboring plasmids remained stable in the CRSS over 200 generations. During a 10-day continuous passage, the genetic environment surrounding bla_NDM-1 in S. sarumanii MAS3954 was gradually excised from the plasmid after the 8th day. In contrast, bla_KPC-2 was stably maintained, with 100% retention. (Fig. 3A). Significant growth differences (P < 0.001) were observed among strain MAS3954, the bla_NDM-1-positive E. coli transconjugant MAS3954T(NDM), and the E. coli EC600 recipient strain (Fig. 3B). Compared to E. coli EC600, MAS3954T(NDM) displayed a markedly delayed exponential phase and a reduced overall growth rate, indicating impaired bacterial fitness (P < 0.001). The slower growth of MAS3954T(NDM) reflects the metabolic burden of maintaining the bla_NDM-1-harboring plasmid, thereby demonstrating a measurable fitness cost in the absence of antimicrobial selection pressure.

Table 3.

Conjugation transfer rate of strain MAS3954 and E. coli EC600.

Donor	Recipient	Number of bacterial colonies (dilution factor)		CF
		MacConkey (SM)	MacConkey (MEM+SM)
MAS3954	EC600	403 (x = 10-1)	89 (y = 1)	2.2 × 10-2

Note: SM, streptomycin; MEM, meropenem. CF, conjugation frequency; CF = (number of transconjugant bacterial colonies × 10y) / (number of recipient bacterial colonies × 10x), where x and y represent the dilution factors.

Fig. 3.

Stability of bla_KPC-2 or bla_NDM-1 plasmids in S. sarumanii MAS3954 and growth curve of donor, recipient, and transconjugant strains. (A) Stability of bla_KPC-2 and bla_NDM-1 plasmids in S. sarumanii MAS3954 along 10-day serial passage. (B) Growth curve of the donor strain MAS3954, the recipient strain E. coli EC600, and the transconjugant MAS3954T(NDM) over 30 h.

3.5. Phenotypic and genotypic characterization of resistance

AST results for S. sarumanii MAS3954, the recipient strain E. coli EC600, and the transconjugant MAS3954T(NDM) were shown in Table 4. S. sarumanii MAS3954 exhibited resistance to a broad range of antimicrobials, including carbapenems (doripenem, ertapenem, imipenem, and meropenem), β-lactams (ampicillin, piperacillin, cefazolin, cefepime, ceftazidime, ceftriaxone, and aztreonam), β-lactam/β-lactamase inhibitor combinations (ampicillin-sulbactam, ceftazidime-avibactam, ticarcillin-clavulanic acid, and piperacillin-tazobactam), fluoroquinolones (ciprofloxacin and levofloxacin), trimethoprim-sulfamethoxazole, tetracycline, nitrofurantoin, and colistin. However, the strain remained susceptible to amikacin, gentamicin, tobramycin, and tigecycline, offering limited available treatment options. WGS confirmed the presence of bla_KPC-2 and bla_NDM-1 genes, consistent with the observed resistance phenotypes (Table 2). In the transconjugant MAS3954T(NDM), resistance to β-lactams and most carbapenems was preserved. However, MIC values for imipenem (4 mg/L) were reduced compared to the donor strain (>8 mg/L). Susceptibility to colistin, nitrofurantoin, and fluoroquinolones was restored, whereas sensitivity to aminoglycosides and tigecycline remained unchanged.

Table 4.

Minimal inhibitory concentrations (MICs) and antimicrobial susceptibility profiles of S. sarumanii MAS3954, its bla_NDM-1-positive transconjugant: MAS3954T(NDM), and E. coli EC600.

Antimicrobial Class/Agent	MICs (µg/mL) [Antimicrobial susceptibility]
	MAS3954	MAS3954T(NDM)	E. coli EC600
Aminoglycosides
Amikacin	<8[S]	<8[S]	<8[S]
Gentamicin	<2[S]	<2[S]	<2[S]
Tobramycin	<2[S]	<2[S]	<2[S]
Penicillins
Ampicillin	>16[R]	>16[R]	<8[S]
Piperacillin	>64[R]	>64[R]	<16[S]
Monobactams
Aztreonam	>16[R]	>16[R]	<1[S]
Cephalosporins
Cefazolin	>16[R]	>16[R]	2[S]
Cefepime	>32[R]	32[R]	<4[S]
Ceftazidime	>16[R]	>16[R]	<1[S]
Ceftriaxone	>32[R]	>32[R]	<0.5[S]
Carbapenems
Doripenem	>4[R]	>4[R]	<0.5[S]
Ertapenem	>8[R]	8[R]	<0.25[S]
Imipenem	>8[R]	4[R]	<0.5[S]
Meropenem	>8[R]	8[R]	<0.5[S]
β-lactam/β-lactamase inhibitors
Ampicillin-Sulbactam	>16/8[R]	>16/8[R]	<4/2[S]
Ceftazidime-Avibactam	>16/4[R]	>16/4[R]	<0.5/4[S]
Ticarcillin-Clavulanic acid	>64/2[R]	>64/2[R]	<8/2[S]
Piperacillin-Tazobactam	>128/4[R]	>128/4[R]	<8/4[S]
Fluoroquinolones
Ciprofloxacin	>2[R]	<0.5[S]	<0.5[S]
Levofloxacin	>8[R]	<1[S]	<1[S]
Sulfonamides
Trimethoprim-Sulfamethoxazole	>4/76[R]	<2/38[S]	<2/38[S]
Tetracyclines
Tetracycline	>8[R]	>8[R]	<4[S]
Minocycline	8[I]	2[S]	<1[S]
Tigecycline	<1[S]	<1[S]	<1[S]
Polymyxins
Colistin	>8[R]	1[S]	1[S]
Nitrofurans
Nitrofurantoin	>64[R]	<32[S]	<32[S]

Note: R, resistant; I, intermediate; S, susceptible.

3.6. Phylogenomic relationship and resistance gene comparison

To investigate the evolutionary position of S. sarumanii MAS3954, we performed a phylogenomic analysis including MAS3954 and nine publicly available S. sarumanii genomes retrieved from the NCBI database (Table S2). A maximum-likelihood phylogenetic tree was generated based on a core-genome SNP alignment. The resulting tree revealed that MAS3954 formed a distinct clade, separate from other strains, indicating notable phylogenetic divergence (Fig. 4). Metadata analysis showed that the ten strains were isolated between 2021 and 2023 from both clinical and environmental sources in China, Germany, and the Philippines.

Fig. 4.

Phylogenomic relationship and resistance gene profiles of ten S. sarumanii strains. The maximum-likelihood phylogenetic tree was constructed based on core-genome SNPs from MAS3954 and nine publicly available genomes. The heatmap on the right shows the presence (red) or absence (white) of acquired antimicrobial resistance genes in each strain. Metadata annotations include country of origin (1), isolation type (2), and collection date (3).

In parallel, we conducted a comparative analysis of acquired AMR genes among these S. sarumanii strains (Fig. 4). Most strains harbored conserved chromosomally encoded resistance determinants, including aac(6′)-Ic, rsmA, msbA, EF-Tu, CRP, KpnH, bla_SRT-2 and emrR. However, MAS3954 uniquely harbored both bla_KPC-2 and bla_NDM-1, two clinically significant carbapenemase genes that were not present in any other strains. Additionally, MAS3954 carried several other acquired resistance genes, such as bla_SHV-12, bla_TEM-1B, and qnrS1, contributing to its extensive multidrug resistance phenotype. This distinct resistance gene profile highlights the MDR nature of MAS3954 and further reinforces its clinical and epidemiological relevance as a potentially high-risk, carbapenemase-producing lineage within the S. sarumanii species.

4. Discussion

The presence of carbapenemase-encoding genes such as bla_KPC-2 and bla_NDM-1, which confer resistance to critical classes of antimicrobials, has been widely reported in several bacterial species (Wu et al., 2015; Wang et al., 2017; Bes et al., 2021; Qiao et al., 2023; Yuan et al., 2023; Zhang et al., 2023; Chen et al., 2024; Zhang et al., 2024). However, the concurrent harboring of both genes in S. sarumanii has not been previously documented. In this study, we report the first clinical strain of S. sarumanii MAS3954 co-harboring bla_KPC-2 and bla_NDM-1 in Eastern China. This novel finding not only contributes to our understanding of the genetic diversity of MDR organisms but also underscores the urgent need for enhanced genomic surveillance, particularly in under-monitored regions, where the emergence of AMR is increasingly becoming a significant public health threat (Iskandar et al., 2021).

Traditional diagnostic tools, such as MALDI-TOF MS, are commonly used for rapid microbial identification. However, they have limitations when distinguishing closely related Serratia species due to high genomic similarity (Aracil-Gisbert et al., 2024; Klages et al., 2024). In our study, S. sarumanii MAS3954 was initially misidentified as S. marcescens, but was later accurately identified as S. sarumanii using high-resolution genomic techniques, such as dDDH and ANI (Goris et al., 2007; Jain et al., 2018). This underscores the critical need for integrating genomic techniques into diagnostic workflows, particularly for emerging or undercharacterized taxa within the Serratia genus.

A significant finding of this study is the identification of a chimeric plasmid, pMAS3954-KPC, which exhibited substantial sequence homology with both the K. pneumoniae plasmid pKPC-h2 and the S. marcescens CM5 chromosome. This suggests a potential plasmid-chromosome recombination event, which may have contributed to the formation of this novel plasmid structure. Although no MGEs were identified at the fusion junctions, the mosaic structure of pMAS3954-KPC provides indirect evidence of a historical recombination event. To our knowledge, this represents the first report of such a fusion plasmid in S. sarumanii, highlighting the dynamic nature of plasmid evolution and its role in the spread of antimicrobial resistance. In contrast, the pMAS3954-NDM plasmid belongs to the well-characterized IncX3 plasmid lineage, which has been widely disseminated across various Enterobacteriaceae species (Guo et al., 2022). The conserved backbone of this plasmid, along with the co-localization of bla_NDM-1 with other resistance determinants, such as bla_SHV-12 and ble_MBL, further reinforce the critical role of IncX3 plasmids in the global spread of carbapenem resistance.

Genetic analysis of the regions surrounding the bla_KPC-2 and bla_NDM-1 genes in S. sarumanii MAS3954 revealed that the plasmids were highly mobile, which could facilitate gene dissemination across bacterial species (Ochman et al., 2000). The pMAS3954-KPC, although containing several MGEs, did not transfer to an E. coli recipient in our study, suggesting that despite its mobility, the plasmid’s transferability may be limited by factors such as the absence of a complete conjugation system or host adaptability (Thomas and Nielsen, 2005). In contrast, the pMAS3954-NDM plasmid, which exhibited a composite structure containing elements like ∆Tn3000, ∆Tn125, and ∆Tn3, showed high transferability and was successfully conjugated into E. coli. This plasmid’s ability to transfer efficiently is consistent with previous findings indicating that bla_NDM-1 plasmids typically exhibit strong transferability, particularly when transposable elements with conjugative potential are present (Zhang et al., 2023; Yao et al., 2024). Although bla_NDM-1 plasmids demonstrated higher transferability, the transconjugants showed a significant decrease in growth rate, reflecting a fitness cost associated with carrying the plasmid. This fitness cost may reduce the competitiveness of the strain in environments without antimicrobial selection pressure (Liu et al., 2021). However, in environments with stronger antimicrobial selection pressure, this fitness cost may be offset, as suggested by previous studies (Ahmad et al., 2023; Element et al., 2023).The MAS3954 strain exhibited MDR, being resistant to nearly all β-lactams, carbapenems, fluoroquinolones, colistin, and other commonly used antimicrobials. This left tigecycline, amikacin, and tobramycin as the only viable treatment options. Such a resistance profile poses significant therapeutic challenges, with MDR organisms associated with higher treatment failure and mortality (Barrasa-Villar et al., 2017). Early detection, rigorous infection control, and genomic surveillance are critical to managing and preventing the spread of these high-risk strains, thereby improving clinical outcomes and treatment effectiveness (Landman et al., 2024).

In traditional Serratia species, core chromosomally encoded resistance genes, such as aac(6′)-Ic, are commonly present and are conserved across species (Shaw et al., 1992; Tavares-Carreon et al., 2023). However, the concurrent presence of bla_KPC-2 and bla_NDM-1 remains unprecedented in S. sarumanii and is rarely reported in S. marcescens (Bes et al., 2021; Ymaña et al., 2022; Zhang et al., 2024). This highlights the unique and concerning nature of this strain, as the coexistence of these two carbapenemase genes is typically found in more clinically aggressive pathogens, such as K. pneumoniae and Enterobacter species (Pereira et al., 2015; Bes et al., 2021; Zhang et al., 2023). The relatively limited reports of bla_KPC-2 and bla_NDM-1 in S. marcescens suggest that the presence of both genes in S. sarumanii may represent an emergent and evolving resistance profile that warrants further investigation.

5. Conclusions

In conclusion, this study reports, to our knowledge, the first clinical strain of S. sarumanii harboring both bla_KPC-2 and bla_NDM-1. Our findings emphasize the urgent need for enhanced surveillance and infection control measures to mitigate the spread of MDR strains in healthcare settings. Furthermore, this study underscores the pivotal role of plasmids in the transmission of resistance genes and the emergence of highly resistant pathogens. These findings call for intensified genomic surveillance, especially in under-monitored regions, to effectively manage and prevent the spread of such high-risk pathogens.

Funding

This work was supported by the National Science and Technology Fundamental Resources Investigation Program of China (2021FY100904), the screening of Streptococcus agalactiae capsular polysaccharide vaccine production strain (HX2025-001), and the construction of protein fingerprints of pathogens (KFYJ2024003) funded by the National Institute for Communicable Disease Control and Prevention, China CDC.

CRediT authorship contribution statement

Xiao Liu: Conceptualization, Methodology, Formal analysis, Investigation, Visualization, Writing – original draft. Zhen Liu: Methodology, Visualization, Writing – review & editing. Xuemei Bai: Methodology, Investigation, Writing – review & editing. He Gao: Methodology, Writing – review & editing. Zhiwen Sun: Formal analysis, Writing – review & editing. Duochun Wang: Conceptualization, Methodology, Project administration, Resources, Supervision, Validation, Funding acquisition, Writing – original draft, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.