Skip to main content
NCBI home page
Microbiology Spectrum logoLink to Microbiology Spectrum
. 2025 Apr 15;13(5):e03115-24. doi: 10.1128/spectrum.03115-24

Genomic analysis of carbapenemase-encoding plasmids and antibiotic resistance in carbapenem-resistant Klebsiella pneumoniae isolates from Vietnam, 2021

Lisa Göpel 1,#, Le Thi Kieu Linh 2,3,#, Bui Tien Sy 2,4,#, Sébastien Boutin 1,5, Simone Weikert-Asbeck 1, Elias Eger 6, Susanne Hauswaldt 1, Truong Nhat My 2,4, Kaan Kocer 1,5, Nguyen Trong The 2,4, Jan Rupp 1,5,7, Le Huu Song 2,3,4, Katharina Schaufler 6,8, Thirumalaisamy P Velavan 2,3,9,10,#, Dennis Nurjadi 1,2,5,✉,#
Editor: Ahmed Babiker11
PMCID: PMC12054150  PMID: 40231682

ABSTRACT

Carbapenem resistance in gram-negative rods is increasing in low- and middle-income countries. We conducted a single-center study to identify carbapenemase-encoding plasmids in carbapenem-resistant Klebsiella pneumoniae isolates causing human infections in Vietnam. The secondary objective was to investigate the prevalence of multidrug-resistant (MDR) and hypervirulent K. pneumoniae in this setting. Our genomic analysis study characterized 105 of 245 clinical K. pneumoniae isolates at the 108 Military Hospital in Hanoi, Vietnam, collected from intensive care unit and regular wards between 1 January 2021 and 31 December 2021. All isolates were characterized using long- and short-read sequencing, followed by hybrid assembly. Comprehensive genomic analysis was performed to identify carbapenemase-encoding plasmids, complemented by extended antibiotic susceptibility testing for commonly used and novel antibiotics. We observed a high prevalence of NDM-4-related carbapenem resistance (30.5%, 32/105) mostly carried by a specific 83-kb IncFII plasmid co-carrying the blaTEM-1 (46.9%, 15/32). The genomic content of the blaNDM-4-harboring plasmids is highly variable. While blaOXA-181 and blaOXA-48 were predominantly located on an IncX3 and an IncL plasmid, respectively, the majority of plasmids harboring blaKPC-2 were not related to any named Inc-type. All isolates exhibited the MDR phenotype; however, the majority remained susceptible to the siderophore-cephalosporin cefiderocol (79%, 83/105). All isolates were susceptible to aztreonam/avibactam. In addition, we identified a hypervirulent, carbapenem-resistant K. pneumoniae ST23 strain, confirmed through both in vitro and in vivo experiments. Our study provides insights into plasmids harboring the carbapenemases New Delhi metallo-β-lactamase, oxacillinase-48 like, and K. pneumoniae carbapenemase-2 circulating in Vietnam.

IMPORTANCE

Carbapenem resistance in Klebsiella pneumoniae is a major public health threat, especially in low- and middle-income countries. This study examined resistant strains from a hospital in Vietnam to understand how they spread and which antibiotics might still work. We found that a significant number of these bacteria carried resistance genes on different types of plasmids. Despite their resistance to many antibiotics, most strains remained susceptible to newer substances like cefiderocol and aztreonam/avibactam. Alarmingly, we also identified a hypervirulent strain that is carbapenem resistant, potentially posing an even greater risk to patients. This research provides insight into the epidemiology of the carbapenemase gene-harboring plasmids in a Vietnamese hospital. Understanding these resistance patterns can help guide antibiotic use and policy decisions to combat the growing threat of multidrug-resistant infections in Vietnam.

KEYWORDS: NDM-4, KPC-2, OXA-48, carbapenemase-producing Klebsiella pneumoniae, hypervirulent Klebsiella pneumoniae, Vietnam

INTRODUCTION

Klebsiella pneumoniae, as a member of the Enterobacterales, is considered a commensal of the gut and can colonize healthy individuals without causing infection (1). However, under certain circumstances, K. pneumoniae can transition from an asymptomatic colonizer to a pathogen leading to infections, such as pneumonia, urinary tract infections, bloodstream infections, and wound infections (2). In recent years, K. pneumoniae infections have become a major clinical challenge due to the ability of K. pneumoniae to develop or acquire resistance to various antimicrobial agents, leading to the emergence of multidrug-resistant (MDR) phenotypes (3).

The emergence of antimicrobial resistance in K. pneumoniae has been attributed to multiple factors, including antibiotic overuse in both clinical and agricultural sectors. The global success of K. pneumoniae can generally be traced back to specific clonal lineages, sometimes associated with multidrug resistance and the convergence of (hyper-)virulence and resistance (4–7). In Asia, the emergence of carbapenem-resistant K. pneumoniae has poses a significant challenge for public health and clinical practice (6, 8). Carbapenem resistance in K. pneumoniae is often conferred by the production of carbapenemases, such as Klebsiella pneumoniae carbapenemase-2 (KPC-2), New Delhi metallo-β-lactamase (NDM), or oxacillinase (OXA)-48 like. The localization of carbapenemase genes on various plasmids and their spread through horizontal gene transfer in hospitals has been reported, with plasmid-mediated transmission accounting for 50% of carbapenemase-producing Enterobacterales (9). A recent study by Pham et al. revealed the presence of blaKPC-2 (57.1%, 184/322) and blaNDM (50%, 179/322) genes in meropenem-resistant K. pneumoniae strains isolated in 2017 and 2018 from patient and environmental samples in the intensive care units (ICUs) of two hospitals in Vietnam (8). The authors identified a putative self-transmissible IncN plasmid harboring blaKPC-2 in the majority of blaKPC-2-positive isolates (97.8%, 180/184). We recently reported the co-resistance of carbapenem-resistant K. pneumoniae isolates to colistin (37%, 39/105) (10). These isolates were collected in 2021 from the ICU and regular wards of the 108 Military Hospital in Hanoi, Vietnam. blaNDM was the most prevalent carbapenemase (48/103, 46.6%), followed by blaOXA-48 like (43/103, 41.7%) and blaKPC-2 (43/103, 41.7%). Although 30.1% (31/103) of the carbapenemase producers carried two different carbapenemase-encoding genes and most carbapenemases were identified on plasmids, no additional analysis of carbapenemase-encoding plasmids was performed. Data on the localization of the carbapenemase genes blaKPC-2, blaNDM, and blaOXA-48 like on plasmids in K. pneumoniae isolates from hospital patients in Vietnam and the characterization of these plasmids are scarce.

With limited therapeutic options for infections caused by carbapenem-resistant K. pneumoniae, novel treatment strategies are urgently needed. In particular, the treatment of metallo-β-lactamase (MBL) producers, such as NDM, is especially challenging since this potent β-lactamase is able to degrade almost all β-lactam drugs, and there are no effective β-lactamase inhibitors, which can reliably inhibit this enzyme approved for clinical use. Several novel substances or combinations of drugs have been identified as showing great potential, including the siderophore-conjugated cephalosporin, cefiderocol, and the combination of β-lactams and β-lactamase inhibitors, such as aztreonam/avibactam (11). However, considering the local molecular epidemiology of carbapenem resistance, data on the efficacy of these substances are important in clinical practice.

The objective of the present study was to investigate and characterize carbapenemase-encoding plasmids in carbapenemase-resistant K. pneumoniae isolates from a previously described collection of 105 strains obtained during routine microbiological diagnostics in a tertiary hospital in Hanoi, Vietnam. As a secondary objective, this study aimed to investigate the prevalence of MDR and hypervirulent K. pneumoniae in this setting and their susceptibility to novel antibiotics.

MATERIALS AND METHODS

Study setting and study population

This study is based on a collection of 105 carbapenem-resistant K. pneumoniae isolates that we have previously described (10). Briefly, the representative samples used were collected from hospitalized patients from 22 clinical departments at the 108 Military Central Hospital in Hanoi, Vietnam, including ICU and normal wards, between 1 January 2021 and 31 December 2021. Of the 524 K. pneumoniae isolates from clinical samples obtained from various clinical specimens (respiratory, blood, urine, and body fluid samples) detected in the routine microbiological diagnostic in 2021, 245 (47%) were carbapenem-resistant (MIC imipenem >4 mg/L and/or MIC meropenem >8 mg/L). We performed extended susceptibility testing and whole-genome sequencing (WGS) on 105 (43%) non-duplicate strains from 101 patients (4 patients have 2 morphologically different strains). The definition of non-duplicate strains isolated from a single patient is either morphologically different or diverging antibiotic susceptibility profiles.

Bacterial identification and antibiotic susceptibility testing

The bacterial culture was carried out in accordance with standard procedures meeting ISO 15189:2012 standards in the routine microbiological diagnostic laboratory. Species identification was performed by matrix-assisted laser desorption ionization time-of-flight mass spectrometry [matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF), Vitek MS, BioMérieux]. The initial antibiotic susceptibility testing (AST) was conducted as part of the routine microbiological diagnostic procedures in Vietnam. This testing was performed using the Vitek2 Compact (BioMérieux) automated AST system, on the AST-N204 card. Furthermore, AST was performed by broth microdilution (BMD; Sensititre EUMDRXXF, Thermo Fisher, Germany). The MICs of cefiderocol were determined by utilizing the broth microdilution method with iron-depleted cation-adjusted Mueller-Hinton broth (CA-MHB) (12), while aztreonam/avibactam MICs were determined by MIC gradient test (Etest) (bestbion dx GmbH, Germany). For cefiderocol-resistant isolates, BMD was performed using iron-depleted CA-MHB, with the addition of 100 mg/L dipicolinic acid (DPA). This chemical compound is capable of chelating metal ions, thereby inhibiting the activity of MBL, and is useful to determine whether the cefiderocol resistance was due to overexpression or hydrolysis activity of the NDM enzyme. Antibiotic susceptibility was interpreted according to the EUCAST clinical breakpoints v13. Escherichia coli ATCC25922 was used as a quality control strain.

Whole-genome sequencing and bioinformatic analysis

Whole-genome sequencing and bioinformatic analysis were performed as previously described (10). The assembled genomes were screened for plasmid and typed using the MOB-suite (command mob_typer with default parameter) (13, 14). Only the plasmid with complete sequence and typed as plasmid by the MOB-suite were used for the subsequent analysis. Complete plasmids were compared using average nucleotide identity (ANI) with ANIclustermap (v1.1.0). The tree was used for the visualization of the plasmid primary and secondary clusters from the MOB-suite, and clusters based on ANI (ANI >99.9%) were visualized using Clinker (15) and gggenome R package. In order to identify any mutations that could potentially contribute to cefiderocol resistance despite the addition of DPA, a comparative analysis between resistant isolates and all susceptible isolates (including the resistant isolates, which become sensitive with the addition of DPA) was conducted. Initial mapping was performed against the corresponding wild-type reference strain KP21315 (GenBank accession number: CP124702) to facilitate the annotation of the single nucleotide polymorphisms (SNPs), and SNP calling was performed using snippy (https://github.com/tseemann/snippy). All the SNPs found in the sensitive strains or in the resistant isolates, which become sensitive with the addition of DPA, were considered as polymorphisms non-related to cefiderocol resistance.

Phenotypic confirmation of hypervirulence

To evaluate the hypervirulent phenotype, Kp084 was challenged in a variety of in vitro and in vivo virulence assays, including determining hypermucoviscosity, siderophore secretion, resistance to complement-mediated killing, survival in bile salts, and the ability to cause mortality in larvae of the greater wax moth larvae (Galleria mellonella). All phenotypic assays were based on previously published protocols (16, 17), with minor modifications as described below. For reference, we used a well-characterized hypervirulent K. pneumoniae ST420 (PBIO2030 [16]) and E. coli K-12 W3110. Further details are included in the Supplementary Material.

Data analysis

Descriptive statistical analysis was performed using STATA/BE v18.0 (StataCorp, USA). Statistics for the validation of the hypervirulence phenotype were performed using GraphPad Prism 8.0 (https://www.graphpad.com/). The non-parametric Kruskal-Wallis test was applied to compare bacterial groups using median values. A P-value of <0.05 was considered statistically significant.

RESULTS

In 2021, a total of 524 non-duplicate K. pneumoniae isolates were collected from various clinical specimens in the routine microbiological diagnostic, of which 245 (47%) were phenotypically resistant to meropenem and/or imipenem. Of the 245 carbapenem-resistant isolates, every second isolate was selected (122, 50%) for WGS. However, 17 isolates could not be recovered, and only 105 (43%) isolates from 101 patients (4 patients had 2 morphologically distinct isolates) underwent WGS. Information regarding quality control of the sequencing of 105 isolates, as well as antimicrobial resistance genes, virulence genes, and carbapenemase-encoding plasmids, can be found in the Supplementary Data Set and was already reported in part recently (10).

Carbapenemase-encoding plasmids

A total of 79.1% (106/134) of carbapenemase genes were identified on plasmids in 86 of the 105 analyzed isolates. The plasmids carrying blaNDM-4 could be clustered based on average nucleotide identity into 14 variants. The most abundant variant (A: 15/29) was an 83-kb IncFII plasmid, also carrying bleMBL, blaTEM-1, and rmtB1 genes, which were all identified in isolates belonging to ST16 (Fig. 1). The blaNDM-4-carrying plasmid positive for replicon-type repB(R1701) was identified in 11 isolates from 7 distinct multi-locus sequence types (ST15, ST273, ST307, ST438, ST580-1LV, ST656, and ST789), while only 1 isolate was found to harbor blaNDM-4 on an IncFIB plasmid (ST438). Further analysis revealed identical MOB-suite primary and secondary clusters for 15/16 IncFII plasmids and 9/11 repB(R1701) plasmids (Supplementary Data Set; Fig. S1 and S2). Plasmids that are assigned to the same secondary MOB cluster may be regarded as sufficiently related to be potentially part of an outbreak.

Fig 1.

Circular phylogenetic tree presents sequence clustering organized by MLST types and primary clusters, including ST15, ST16, ST273, ST307, ST438, ST580-1L5, ST656, ST789, AA276, AA450, AA588, and AE583, with strain names arranged around the structure.

Open in a new tab

ANI-based dendrogram and genomic content of the blaNDM-4-carrying plasmids. Complete plasmids (n = 29) were compared using average nucleotide identity with ANIclustermap (v.1.1.0) and plasmid cluster. Four sections of the circle (separated by white gaps) represent, from inner to outer: (i) ST and primary cluster-ID color code as in the legend, (ii) replicon types (presence of the Inc-type in green, absence in gray), (iii) insertion sequence elements (presence of the IS in orange, absence in gray), and (iv) antimicrobial resistance genes present on the plasmids (presence of the AMR gene in black, absence in gray).

The plasmids carrying blaNDM-1 were less diverse with only four variants. The most prevalent variant (variant A, 4/9) was an 86-kb IncFII plasmid co-carrying blaCTX-M-15, blaTEM-1 (two copies), fosA3, rmtB1, and bleMBL (Fig. 2). MOB-suite analysis revealed the same primary (AA327) and secondary cluster (AI227) for all IncFII plasmids (7/9), which were identified in seven isolates belonging to ST16 (Supplementary Data Set; Fig. S1 and S3).

Fig 2.

Circular phylogenetic tree presents sequence clustering organized by MLST types and primary clusters, including ST16, ST517, AA035, AA327, and AA450, with strain names arranged along the branching structure.

Open in a new tab

ANI-based dendrogram and genomic content of the blaNDM-1-carrying plasmids. Complete plasmids (n = 9) were compared using average nucleotide identity with ANIclustermap (v.1.1.0) and plasmid cluster. The four sections of the circle (separated by white gaps) represent, from inner to outer: (i) ST and primary cluster-ID color code as in the legend, (ii) replicon types (presence of the Inc-type in green, absence in gray), (iii) insertion sequence elements (presence of the IS in orange, absence in gray), and (iv) antimicrobial resistance genes present on the plasmids (presence of the AMR gene in black, absence in gray).MLST, multi-locus sequence type.

All 16 IncX3 plasmids carrying blaOXA-181 exhibited a high degree of conservation across the entire population (ANI: 99.8%–100%, identical MOB-suite primary and secondary cluster) and were found to carry the qnrS1 gene. These plasmids were identified in isolates belonging to ST16 (15 isolates) and ST23 (1 isolate). The same high level of similarity was observed in 15/16 blaOXA-48-carrying plasmids, which all harbored only blaOXA-48 and belonged to the IncL type (Supplementary Data Set; Fig. S4 and S5). One isolate (ST16) was found to carry a plasmid containing two replicon types (IncL and IncR) and multiple antimicrobial resistance genes (Fig. 3).

Fig 3.

Circular phylogenetic tree presents sequence clustering organized by MLST types and primary clusters, including ST16, ST16L5, ST231, AA002, and AA452, with strain names arranged along the branching structure.

Open in a new tab

ANI-based dendrogram and genomic content of the blaOXA-48-carrying plasmids. Complete plasmids (n = 16) were compared using average nucleotide identity with ANIclustermap (v.1.1.0) and plasmid cluster. The four sections of the circle (separated by white gaps) represent, from inner to outer: (i) ST and primary cluster-ID color code as in the legend, (ii) replicon types (presence of the Inc-type in green, absence in gray), (iii) insertion sequence elements (presence of the IS in orange, absence in gray), and (iv) antimicrobial resistance genes present on the plasmids (presence of the AMR gene in black, absence in gray).

A total of 36 blaKPC-2-carrying plasmids were identified and categorized into 14 variants (ANI) and 5 different primary clusters (MOB-suite). The majority of plasmids (22/36) carrying blaKPC-2 were not related to any named Inc-type and carry qnrS1, blaTEM-1, Mrx, and mph(A) in addition to blaKPC-2. These were identified in ST16 (16 isolates) and ST11 (6 isolates) (Fig. 4). Furthermore, seven blaKPC-2-carrying IncN plasmids were detected and primarily found in isolates belonging to ST15 (five isolates). The remaining plasmids carried the repB(R1701) replicon type and were identified in ST16 (six isolates) and ST16-1LV (one isolate). While the primary and secondary MOB clusters were found to be identical for IncN and repB(R1701), respectively, the analysis of ANI revealed clustering in three variants for both plasmid types (Supplementary Data Set; Fig. S4 and S6).

Fig 4.

Circular phylogenetic tree presents sequence clustering by MLST types, ST1, ST11, ST15, ST16, and ST16L5, and primary clusters, AA036, AA038, AA040, AA449, and AA552, with strain names arranged along the branching structure.

Open in a new tab

ANI-based dendrogram and genomic content of the blaKPC-2-carrying plasmids. Complete plasmids (n = 36) were compared using average nucleotide identity with ANIclustermap (v.1.1.0) and plasmid cluster. The four sections of the circle (separated by white gaps) represent, from inner to outer: (i) ST and primary cluster-ID color code as in the legend, (ii) replicon types (presence of the Inc-type in green, absence in gray), (iii) insertion sequence elements (presence of the IS in orange, absence in gray), and (iv) antimicrobial resistance genes present on the plasmids (presence of the AMR gene in black, absence in gray). MLST, multi-locus sequence type.

A total of 20 carbapenemase-producing K. pneumoniae isolates (all belonging to ST16) carrying 2 different carbapenemase-encoding plasmids were identified. The blaNDM-4-encoding IncFII plasmid (cluster A) and the IncX3-harboring blaOXA-181 plasmid were present in 14 isolates. Six isolates harbored different plasmid cluster combinations of blaKPC-2-carrying plasmids and blaNDM-1-encoding plasmids IncFII (Supplementary Data Set). Further visualization and comparison of the overall plasmid structures of carbapenemase-encoding plasmids from this study can be found in the Supplementary Material.

Phenotypic resistance to β-lactam antibiotics

All isolates resistant to three or more antimicrobial substance classes were considered to be MDR. The majority of isolates exhibited resistance to aztreonam and cefepime (98.1% and 96.2%, respectively), while resistance rates for imipenem and meropenem ranged from 81% to 85.7%, respectively. Two isolates (Kp052 and Kp072) showed susceptibility to meropenem despite the presence of carbapenemase-encoding genes blaNDM-4 and blaOXA-48, respectively. Seven of the nine imipenem-susceptible isolates harbored at least one carbapenemase gene, specifically blaOXA-48 (three isolates), blaKPC-2 (two isolates), blaNDM-4 (one isolate), and a combination of blaKPC-2 and blaNDM-1 (one isolate). The AST results also revealed high resistance rates for piperacillin/tazobactam (105/105, 100%) and ceftolozane/tazobactam (101/105, 96.2%), while the combination of ceftazidime/avibactam exhibited a 44.8% (47/105) resistance rate among isolates. The combination of carbapenem antibiotics with β-lactamase inhibitors demonstrated a lower resistance rate for imipenem/relebactam (63/105, 60%) and meropenem/vaborbactam (61/105, 58.1%) compared to single imipenem and meropenem exposure. All isolates were susceptible to aztreonam in combination with avibactam (MIC ≤4 mg/L). Concerning non-β-lactam antibiotics, the resistance rate to tobramycin was high (91/105, 86.7%) but lower for amikacin (54/105, 51.4%), eravacycline (54/105, 51.4%), and tigecycline (51/105, 48.6%) (Supplementary Data Set).

Twenty-two of 105 (21%) tested isolates were resistant to cefiderocol (MIC ≥4 mg/L). Of these, 10 isolates were positive for KPC-2, and 9 isolates were positive for NDM. When adding 100 µg/mL of MBL inhibitor, DPA, into the iron-depleted CA-MHB, 15/22 (68.2%) of the initially cefiderocol-resistant isolates became susceptible to cefiderocol (including all nine NDM-producing isolates). For the remaining seven cefiderocol-resistant isolates (all KPC-2-producing K. pneumoniae), specific mutations observed only in these strains were detected in the ftsI gene, which encodes for the penicillin-binding protein 3 and is one of the known target proteins of cefiderocol (five of seven isolates) as well as genes coding for iron binding and transport (fhuA, fepA, fepD, and nqrF; six of seven isolates). Furthermore, an SNP in the diacylglycerol kinase, which is associated with metal binding, was identified in one isolate (Supplementary Data Set).

Hypervirulent ST23 K. pneumoniae

In one isolate of K. pneumoniae ST23 (Kp084), obtained from the tracheal aspirate of a neurosurgical patient experiencing post-operative respiratory infection and carrying both blaKPC2 and blaOXA-48, biomarkers linked to hypervirulent K. pneumoniae were detected. The yersiniabactin operon and the colibactin operon were chromosomally localized. The colibactin operon was inserted via a transposable element (insertion sequence IS102, (IS5 family, group IS903) upstream of the yersiniabactin locus. The salmochelin operon and the aerobactin operon were located on a 223-kb plasmid carrying two replicon types IncHI1B(pNDM-MAR)/repB and several cation efflux operons (cus and fec) or metal resistance gene (Fig. 5a). However, Kp084 did not exhibit a hypermucoid phenotype despite encoding hypermucoviscosity-modulating genes (e.g., rmpA and rmpA2). The hypervirulent trait of Kp084 was assessed using in vitro siderophore assays and in vivo G. mellonella survival assays with results comparable to the reference hypervirulent K. pneumoniae PBIO2030 (Fig. 5b through f), a previously characterized strain belonging to ST420 (17).

Fig 5.

Genome maps present colibactin and yersiniabactin clusters, circular chromosome presents coding regions, bar charts compare siderophore secretion, hypermucoviscosity, serum resistance, and bile salt survival, and line graph presents infection survival.

Open in a new tab

Loci of virulence determinant conferring the hypervirulence phenotype in isolate Kp084. (a) The gene clusters conferring colibactin and yersiniabactin production are located within the chromosome. The colibactin operon was inserted by an IS5 family transposable element (IS102) upstream of the yersiniabactin operon. The aerobactin operon (iucABCD iutA genes) and the salmochelin operon (iroBCDN genes), indicated by red font, are located on an ~220-kb IncHI1b plasmid. There were no resistance genes detected on this plasmid. (b) The amount of siderophore secreted is expressed as the mean of the siderophore concentration in the supernatant of the bacterial culture and the standard error (n = 4). (c) Determination of hypermucoviscosity by sedimentation assay (n = 3). Results are expressed as the mean ratio of the OD600 of the supernatant after centrifugation at 1,000 × g for 5 min to the total OD600 and the standard error. (d) Survival in 50% human serum (n = 3). Results are expressed as mean and standard error of log2 fold change in CFU/mL after 4 h of incubation in the presence of human serum. (e) Resilience against 50 mg/mL bile salts (n = 3). Results are shown as mean percent survival rates and standard error. (f) Kaplan-Meier plot of mortality in the G. mellonella larvae infection model (n = 3). Results are expressed as mean percent mortality after the injection of 105 CFU per larva. The well-characterized PBIO2030 and the E. coli K12 W3110 were used as hypervirulent K. pneumoniae reference and negative control (NC), respectively. For all results, Kp084 was statistically compared to PBIO2030 using the Kruskal-Wallis test, and the following indicates the level of significance (P value): ns, not significant (P ≥ 0.05); *, P < 0.05; ****, P < 0.0001.

DISCUSSION

The majority of carbapenemase genes identified were located on plasmids (106/134), which were predominantly of the IncX3 type for blaOXA-181 and the IncL type for blaOXA-48. Plasmids carrying blaNDM-4 and blaNDM-1 exhibited a higher degree of diversity, with the majority of these plasmids belonging to the IncFII type. These carbapenemase-encoding plasmid types have been identified in Enterobacteriaceae from multiple countries across several continents (18, 19). In contrast, the majority of plasmids harboring blaKPC-2 were not related to any named Inc-type. In the present study, only 19.4% (7/36) blaKPC-2 genes were identified on an IncN plasmid. Previous data from K. pneumoniae isolates collected in 2017 and 2018 from two hospitals in Vietnam demonstrated a much higher detection rate of blaKPC-2 on one IncN plasmid, with a percentage of 97.8% (180/184) (8). The authors also reported the high prevalence of co-carriage of blaKPC-2 and blaNDM-1 in 164 out of 357 isolates. However, no plasmids were identified as harboring blaNDM-1. Here, we report on the co-carriage of one blaNDM-4-encoding IncFII plasmid and one IncX3 plasmid harboring blaOXA-181 in 14 isolates, while 6 isolates carried both blaKPC-2 and blaNDM-1 on different plasmids. Although the clinical implications and significance of dual carbapenemase producers remain unclear, the presence of two β-lactamases, belonging to different classes, e.g., metallo- and serine β-lactamase, may further limit the clinical use of novel antibiotics by rendering them ineffective. Therefore, the increasing presence of dual carbapenemase producers and the localization of carbapenemase genes on plasmids should be closely monitored.

Due to the high prevalence of NDM producers in our study isolates, resistance to newer β-lactam and β-lactam/β-lactamase inhibitors was expected. Although cefiderocol is not yet used in Vietnamese hospitals, 21% of the carbapenem-resistant K. pneumoniae showed reduced susceptibility (increased MIC) to cefiderocol. A possible explanation for this phenomenon is the low hydrolytic activity of NDM toward cefiderocol or overexpression of NDM (20, 21). Indeed, the addition of DPA to inhibit NDM activity resulted in the reduction of the cefiderocol MIC to susceptible levels in all nine of the cefiderocol-resistant isolates, indicating that the MIC increase was mediated by the NDM activity. In those isolates that remained cefiderocol resistant despite the addition of DPA, several mutations were detected in genes associated with iron binding and transport. Resistance to cefiderocol in Enterobacterales has been attributed to mutations in the TonB-dependent siderophore receptors (22, 23). On a positive note, there was no resistance to the combination of aztreonam/avibactam in this study cohort. Although avibactam cannot withstand hydrolysis by NDM, the combination partner aztreonam is considered to be stable to NDM but not to other β-lactamases, and the combination with avibactam is thought to avoid degradation of aztreonam by other enzymes, such as AmpC and CTX-M (24).

We detected a hypervirulent and carbapenem-resistant K. pneumoniae ST23 (convergent pathotype). Although the rmpA and rmpA2 genes associated with hypermucoviscosity (25) were present, our isolate did not exhibit a hypermucoid phenotype. Since we only detected a single case of hypervirulent K. pneumoniae in this study, we could not elucidate whether the lack of hypermucoviscosity is a natural evolutionary trajectory of this particular clone or a one-off deviation. The loss of hypermucoviscosity may pose a challenge for the identification of the hypervirulent phenotype in routine microbiological diagnostics as the hypervirulent phenotype is commonly associated with increased capsule production. In contrast to our study, previous studies from Vietnam found no hypervirulent K. pneumoniae (8, 26). However, in these studies, the isolates were collected before 2018, while our K. pneumoniae isolates were collected between January and December 2021, suggesting that the hypervirulent K. pneumoniae ST23 may be emerging in Vietnam. Convergent pathotypes, such as K. pneumoniae ST307, have been described to cause outbreaks, and the spread of this clone should be closely monitored (4).

This study has several limitations. First, the present study was conducted exclusively with the clinical isolates from hospitalized patients from a single center. A total of 245 carbapenem-resistant K. pneumoniae were collected during the designated sampling period, of which a random selection of 105 isolates was subjected to further investigation via whole-genome sequence analysis. The strength of our study was the application of hybrid (long- and short-read) sequencing on all isolates, coupled with comprehensive phenotypic resistance testing. Furthermore, we found evidence for the occurrence of convergent pathotypes of K. pneumoniae that simultaneously combine resistance and enhanced virulence. This finding was validated by phenotypic testing and comparison with a well-characterized reference strain. However, we did not perform the mouse infection model, which is considered to be a more accurate in vivo assay for assessing hypervirulence compared to the Galleria mellonella assay (27). Despite this, we successfully demonstrated that the isolate identified in this study possesses the genomic characteristics associated with a hypervirulent phenotype.

In conclusion, our data demonstrate the localization of carbapenemase genes on different plasmids in carbapenemase-producing K. pneumoniae in Vietnam. Comprehensive antimicrobial susceptibility testing revealed high rates of resistance to the majority of the antibiotics tested. However, most isolates were susceptible to cefiderocol, and all isolates were susceptible to aztreonam-avibactam, suggesting that these drugs are promising for treating infections caused by MDR and extensively resistant Enterobacterales. Furthermore, the detection of MDR and hypervirulent K. pneumoniae ST23 in the hospital setting is alarming and should be closely monitored. Our data also highlight the importance of incorporating WGS to understand the transmission dynamics of MDR bacteria in Vietnam and to develop tailored and effective infection prevention and control measures.

Supplementary Material

Reviewer comments
reviewer-comments.pdf (106.9KB, pdf)

ACKNOWLEDGMENTS

K.S. received a grant from the Federal Ministry of Education and Research (BMBF, Germany) entitled “Disarming pathogens as a different strategy to fight antimicrobial-resistant gram negatives” (01KI2015). T.P.V. received PAN-ASEAN Coalition for Epidemic and Outbreak Preparedness (PACE-UP; DAAD Project ID: 57592343) and staff support through JPIAMR I-CRECT (BMBF-01KI2207).

The study was funded through grants from the PAN-ASEAN Coalition for Epidemic and Outbreak Preparedness (PACE-UP; DAAD Project ID: 57592343) and staff support through JPIAMR I-CRECT.

D.N. has received speaker’s honoraria from Shionogi and Cepheid outside the scope of this work. All other authors reported no conflict of interest.

Contributor Information

Dennis Nurjadi, Email: dennis.nurjadi@uni-luebeck.de.

Ahmed Babiker, Rush University Medical Center, Chicago, Illinois, USA.

ETHICS APPROVAL

All participants gave informed consent before taking part in the study. The study was approved by the Institutional Review Board of the 108 Military Central Hospital, Hanoi, Vietnam (108MCH/RES/MENTNGITIS-V-D3-25-04-2017) and (108MCH/RES/I-CRECT-V1.1-D2-06-04-2021).

DATA AVAILABILITY

The draft genomes presented in this study can be found in the NCBI Genbank repositories under the Bioproject PRJNA1043438.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/spectrum.03115-24.

Supplemental data set. spectrum.03115-24-s0001.xlsx.

Overview of complete carbapenemase-encoding plasmids analyzed in this study.

spectrum.03115-24-s0001.xlsx (253.5KB, xlsx)
DOI: 10.1128/spectrum.03115-24.SuF1
Supplemental material. spectrum.03115-24-s0002.pdf.

Supplemental Methods; Fig. S1 to S6.

spectrum.03115-24-s0002.pdf (1.9MB, pdf)
DOI: 10.1128/spectrum.03115-24.SuF2
OPEN PEER REVIEW. reviewer-comments.pdf.

An accounting of the reviewer comments and feedback.

reviewer-comments.pdf (106.9KB, pdf)
DOI: 10.1128/spectrum.03115-24.SuF3

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.

REFERENCES

  • 1. Cocker D, Chidziwisano K, Mphasa M, Mwapasa T, Lewis JM, Rowlingson B, Sammarro M, Bakali W, Salifu C, Zuza A, et al. 2023. Investigating One Health risks for human colonisation with extended spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae in Malawian households: a longitudinal cohort study. Lancet Microbe 4:e534–e543. doi: 10.1016/S2666-5247(23)00062-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Gorrie CL, Mirčeta M, Wick RR, Judd LM, Lam MMC, Gomi R, Abbott IJ, Thomson NR, Strugnell RA, Pratt NF, Garlick JS, Watson KM, Hunter PC, Pilcher DV, McGloughlin SA, Spelman DW, Wyres KL, Jenney AWJ, Holt KE. 2022. Genomic dissection of Klebsiella pneumoniae infections in hospital patients reveals insights into an opportunistic pathogen. Nat Commun 13:3017. doi: 10.1038/s41467-022-30717-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Antimicrobial Resistance Collaborators . 2022. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet 399:629–655. doi: 10.1016/S0140-6736(21)02724-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Heiden SE, Hübner N-O, Bohnert JA, Heidecke C-D, Kramer A, Balau V, Gierer W, Schaefer S, Eckmanns T, Gatermann S, Eger E, Guenther S, Becker K, Schaufler K. 2020. A Klebsiella pneumoniae ST307 outbreak clone from Germany demonstrates features of extensive drug resistance, hypermucoviscosity, and enhanced iron acquisition. Genome Med 12:113. doi: 10.1186/s13073-020-00814-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Shaidullina ER, Schwabe M, Rohde T, Shapovalova VV, Dyachkova MS, Matsvay AD, Savochkina YA, Shelenkov AA, Mikhaylova YV, Sydow K, et al. 2023. Genomic analysis of the international high-risk clonal lineage Klebsiella pneumoniae sequence type 395. Genome Med 15:9. doi: 10.1186/s13073-023-01159-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Wyres KL, Nguyen TNT, Lam MMC, Judd LM, van Vinh Chau N, Dance DAB, Ip M, Karkey A, Ling CL, Miliya T, Newton PN, Lan NPH, Sengduangphachanh A, Turner P, Veeraraghavan B, Vinh PV, Vongsouvath M, Thomson NR, Baker S, Holt KE. 2020. Genomic surveillance for hypervirulence and multi-drug resistance in invasive Klebsiella pneumoniae from South and Southeast Asia. Genome Med 12:11. doi: 10.1186/s13073-019-0706-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Lee C-R, Lee JH, Park KS, Kim YB, Jeong BC, Lee SH. 2016. Global dissemination of carbapenemase-producing Klebsiella pneumoniae: epidemiology, genetic context, treatment options, and detection methods. Front Microbiol 7:895. doi: 10.3389/fmicb.2016.00895 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Pham MH, Hoi LT, Beale MA, Khokhar FA, Hoa NT, Musicha P, Blackwell GA, Long HB, Huong DT, Binh NG, Co DX, Giang T, Bui C, Tran HN, Bryan J, Herrick A, Feltwell T, Nadjm B, Parkhill J, van Doorn HR, Trung NV, Van Kinh N, Török ME, Thomson NR. 2023. Evidence of widespread endemic populations of highly multidrug resistant Klebsiella pneumoniae in hospital settings in Hanoi, Vietnam: a prospective cohort study. Lancet Microbe 4:e255–e263. doi: 10.1016/S2666-5247(22)00338-X [DOI] [PubMed] [Google Scholar]
  • 9. Marimuthu K, Venkatachalam I, Koh V, Harbarth S, Perencevich E, Cherng BPZ, Fong RKC, Pada SK, Ooi ST, Smitasin N, et al. 2022. Whole genome sequencing reveals hidden transmission of carbapenemase-producing Enterobacterales. Nat Commun 13:3052. doi: 10.1038/s41467-022-30637-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Sy BT, Boutin S, Kieu Linh LT, Weikert-Asbeck S, Eger E, Hauswaldt S, Nhat My T, The NT, Rupp J, Song LH, Schaufler K, Velavan TP, Nurjadi D. 2024. Heterogeneity of colistin resistance mechanism in clonal populations of carbapenem-resistant Klebsiella pneumoniae in Vietnam. Lancet Reg Health West Pac 51:101204. doi: 10.1016/j.lanwpc.2024.101204 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Ageevets V, Sulian O, Avdeeva A, Chulkova P, Ageevets I, Gostev V, Alieva K, Golikova M, Sidorenko S. 2024. Minimum inhibitory concentrations of aztreonam-avibactam, ceftazidime-avibactam and meropenem in clinical isolates of Klebsiella pneumoniae harboring carbapenemase genes. J Antibiot 77:706–710. doi: 10.1038/s41429-024-00758-8 [DOI] [PubMed] [Google Scholar]
  • 12. Hackel MA, Tsuji M, Yamano Y, Echols R, Karlowsky JA, Sahm DF. 2019. Reproducibility of broth microdilution MICs for the novel siderophore cephalosporin, cefiderocol, determined using iron-depleted cation-adjusted Mueller-Hinton broth. Diagn Microbiol Infect Dis 94:321–325. doi: 10.1016/j.diagmicrobio.2019.03.003 [DOI] [PubMed] [Google Scholar]
  • 13. Robertson J, Bessonov K, Schonfeld J, Nash JHE. 2020. Universal whole-sequence-based plasmid typing and its utility to prediction of host range and epidemiological surveillance. Microb Genom 6:mgen000435. doi: 10.1099/mgen.0.000435 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Robertson J, Nash JHE. 2018. MOB-suite: software tools for clustering, reconstruction and typing of plasmids from draft assemblies. Microb Genom 4:e000206. doi: 10.1099/mgen.0.000206 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Gilchrist CLM, Chooi YH. 2021. clinker & clustermap.js: automatic generation of gene cluster comparison figures. Bioinformatics 37:2473–2475. doi: 10.1093/bioinformatics/btab007 [DOI] [PubMed] [Google Scholar]
  • 16. Eger E, Heiden SE, Becker K, Rau A, Geisenhainer K, Idelevich EA, Schaufler K. 2021. Hypervirulent Klebsiella pneumoniae sequence type 420 with a chromosomally inserted virulence plasmid. Int J Mol Sci 22:9196. doi: 10.3390/ijms22179196 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Eger E, Schwabe M, Schulig L, Hübner N-O, Bohnert JA, Bornscheuer UT, Heiden SE, Müller JU, Adnan F, Becker K, Correa-Martinez CL, Guenther S, Idelevich EA, Baecker D, Schaufler K. 2022. Extensively drug-resistant Klebsiella pneumoniae counteracts fitness and virulence costs that accompanied ceftazidime-avibactam resistance acquisition. Microbiol Spectr 10:e0014822. doi: 10.1128/spectrum.00148-22 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Pitout JDD, Peirano G, Kock MM, Strydom KA, Matsumura Y. 2019. The global ascendency of OXA-48-type carbapenemases. Clin Microbiol Rev 33:e00102-19. doi: 10.1128/CMR.00102-19 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Wu W, Feng Y, Tang G, Qiao F, McNally A, Zong Z. 2019. NDM metallo-β-lactamases and their bacterial producers in health care settings. Clin Microbiol Rev 32:e00115-18. doi: 10.1128/CMR.00115-18 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Hornsey M, Phee L, Wareham DW. 2011. A novel variant, NDM-5, of the New Delhi metallo-β-lactamase in a multidrug-resistant Escherichia coli ST648 isolate recovered from a patient in the United Kingdom. Antimicrob Agents Chemother 55:5952–5954. doi: 10.1128/AAC.05108-11 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Simner PJ, Mostafa HH, Bergman Y, Ante M, Tekle T, Adebayo A, Beisken S, Dzintars K, Tamma PD. 2022. Progressive development of cefiderocol resistance in Escherichia coli during therapy is associated with an increase in blaNDM-5 copy number and gene expression. Clin Infect Dis 75:47–54. doi: 10.1093/cid/ciab888 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Klein S, Boutin S, Kocer K, Fiedler MO, Störzinger D, Weigand MA, Tan B, Richter D, Rupp C, Mieth M, Mehrabi A, Hackert T, Zimmermann S, Heeg K, Nurjadi D. 2022. Rapid development of cefiderocol resistance in carbapenem-resistant Enterobacter cloacae during therapy is associated with heterogeneous mutations in the catecholate siderophore receptor cirA. Clin Infect Dis 74:905–908. doi: 10.1093/cid/ciab511 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. McElheny CL, Fowler EL, Iovleva A, Shields RK, Doi Y. 2021. In vitro evolution of cefiderocol resistance in an NDM-producing Klebsiella pneumoniae due to functional loss of CirA. Microbiol Spectr 9:e0177921. doi: 10.1128/Spectrum.01779-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Karlowsky JA, Kazmierczak KM, de Jonge BLM, Hackel MA, Sahm DF, Bradford PA. 2017. In vitro activity of aztreonam-avibactam against Enterobacteriaceae and Pseudomonas aeruginosa isolated by clinical laboratories in 40 countries from 2012 to 2015. Antimicrob Agents Chemother 61:e00472-17. doi: 10.1128/AAC.00472-17 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Gonzalez-Ferrer S, Peñaloza HF, Budnick JA, Bain WG, Nordstrom HR, Lee JS, Van Tyne D. 2021. Finding order in the chaos: outstanding questions in Klebsiella pneumoniae pathogenesis. Infect Immun 89:e00693-20. doi: 10.1128/IAI.00693-20 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Linh TD, Thu NH, Shibayama K, Suzuki M, Yoshida L, Thai PD, Anh DD, Duong TN, Trinh HS, Thom VP, Nga LTV, Phuong NTK, Thuyet BT, Walsh TR, Thanh LV, Bañuls A-L, van Doorn HR, Van Anh T, Hoang TH. 2021. Expansion of KPC-producing Enterobacterales in four large hospitals in Hanoi, Vietnam. J Glob Antimicrob Resist 27:200–211. doi: 10.1016/j.jgar.2021.09.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Russo TA, MacDonald U. 2020. The Galleria mellonella infection model does not accurately differentiate between hypervirulent and classical Klebsiella pneumoniae. mSphere 5:e00850-19. doi: 10.1128/mSphere.00850-19 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Reviewer comments
reviewer-comments.pdf (106.9KB, pdf)
Supplemental data set. spectrum.03115-24-s0001.xlsx.

Overview of complete carbapenemase-encoding plasmids analyzed in this study.

spectrum.03115-24-s0001.xlsx (253.5KB, xlsx)
DOI: 10.1128/spectrum.03115-24.SuF1
Supplemental material. spectrum.03115-24-s0002.pdf.

Supplemental Methods; Fig. S1 to S6.

spectrum.03115-24-s0002.pdf (1.9MB, pdf)
DOI: 10.1128/spectrum.03115-24.SuF2
OPEN PEER REVIEW. reviewer-comments.pdf.

An accounting of the reviewer comments and feedback.

reviewer-comments.pdf (106.9KB, pdf)
DOI: 10.1128/spectrum.03115-24.SuF3

Data Availability Statement

The draft genomes presented in this study can be found in the NCBI Genbank repositories under the Bioproject PRJNA1043438.

Articles from Microbiology Spectrum are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES