Analysis of intrahospital and global dissemination and resistome dynamics of NDM-1-producing ST773 Pseudomonas aeruginosa high-risk clone

Abstract

Objectives

To analyse the intrahospital and global dissemination and resistome dynamics of the concerning NDM-1 MBL-producing ST773 P. aeruginosa high-risk clone.

Methods

A total of 17 NDM-1-producing P. aeruginosa isolates recovered in 2022–24 from 10 patients at Hospital Clinic of Barcelona (HCB), Spain, were studied through susceptibility testing and WGS. Expression of resistance genes was analysed through quantitative (real-time) RT–PCR. Forty ST773 genomes from isolates recovered worldwide were also incorporated in the phylogenetic and resistome analysis.

Results

All HCB NDM-1-producing isolates were assigned to ST773 except one (ST357 additionally producing VEB-9 and linked epidemiologically to India). The index ST773 case was a 41-year-old woman admitted to the oncology ward in February 2022 after breast cancer surgery in Ukraine. These isolates were closely related and the bla_NDM-1 gene was located in the same 117 kb integrative conjugative element. All ST773-NDM-1 producers from HCB and the 40 worldwide isolates shared the same acquired resistance determinants [aadA11-like, rmtB4, qnrVC1 and tet(G)], as well as some of the antibiotic resistance mutations (mexZ, mexT, gyrA and parC). Other specific mutations such as an oprD deletion were shared only with isolates from Ukrainian patients transferred to Madrid or the Netherlands. Lastly, HCB isolates evolved further resistome mutations during intrahospital dissemination, including regulators of AmpC (mpl) and MexAB-OprM (nalD), linked to the acquisition of aztreonam/avibactam resistance, and thus remaining only susceptible to cefiderocol and colistin.

Conclusions

This work evidences the transborder spread and intrahospital dissemination and evolution of the emerging ST773-NDM-1 P. aeruginosa high-risk clone.

Introduction

The global spread of antibiotic-resistant bacteria poses a significant challenge to public health, with Pseudomonas aeruginosa representing a critical threat due to its intrinsic antibiotic resistance and its outstanding ability to develop further resistance by the acquisition and selection of chromosomic mutations and/or the acquisition of additional resistance genes.¹

A nationwide genomic epidemiological study recently conducted in Spain showed a generalized decrease in P. aeruginosa resistance in the last 5 years but an increase in the proportion of XDR strains producing carbapenemases in association with the spread of the high-risk clone ST235.² Likewise, other studies have also suggested that the production of acquired carbapenemases appears to be rising in frequency worldwide.³ Among these enzymes, NDM is particularly concerning as it confers resistance to a wide range of β-lactam antibiotics, including carbapenems and newer β-lactam/β-lactamase inhibitor combinations, often considered last-resort treatments for MDR infections.⁴

The emergence of NDM-1-producing P. aeruginosa strains, especially in high-risk clones like ST773, exacerbates this threat, as these clones are highly transmissible within healthcare settings and across borders. NDM-1-producing ST773 isolates have already been detected in various countries worldwide, reflecting its potential for widespread dissemination.^5–19 Understanding both the local and global dissemination patterns, as well as the resistome dynamics of NDM-1-producing ST773 P. aeruginosa, is crucial for developing effective infection control strategies and to mitigate the impact of these emerging antibiotic-resistant pathogens.

In this study, we investigated the introduction, dissemination and persistence of an NDM-1-producing P. aeruginosa ST773 clone within a tertiary Spanish hospital over a 2 year period. By combining phenotypic and whole-genome approaches we aimed to provide insights into the genetic relatedness and evolution of this high-risk clone worldwide in order to enhance the understanding of the mechanisms driving the global spread of the NDM-1-producing ST773 P. aeruginosa high-risk clone.

Materials and methods

P. aeruginosa strains and antibiotic susceptibility testing

From April 2022 to February 2024, NDM-producing P. aeruginosa isolates were obtained from 10 different patients admitted to the Hospital Clinic of Barcelona (HCB), an 850-bed tertiary hospital, in both clinical samples and surveillance rectal swabs.

MICs of piperacillin/tazobactam, ceftazidime, cefepime, ceftolozane/tazobactam, ceftazidime/avibactam, imipenem, meropenem, ciprofloxacin, tobramycin, amikacin and colistin were determined by broth microdilution using Sensititre^™ panels (plate code: FRCNRP2, Thermo Fisher Diagnostics, S.LU.). To determine the MICs of aztreonam, aztreonam/avibactam and cefiderocol, an in-house broth microdilution was performed according to ISO-EUCAST guidelines (http://www.eucast.org) using CAMHB for aztreonam and aztreonam/avibactam, and ID-CAMHB for cefiderocol. EUCAST clinical breakpoints (v14) were used for interpretation of SIR categories (except for aztreonam/avibactam). The reference strain PAO1 and its knockout mutant PAOΔpiuC were included in the susceptibility testing experiments as quality controls.

Since most of the isolates were identified as ST773, an emerging high-risk clone of P. aeruginosa reported in various countries, we conducted a thorough search of the PubMed database to compare our isolates with those described globally. The search was based on the following criteria: ST773 [OR] NDM [AND] P. aeruginosa and, when publicly available, reads and/or draft genomes were downloaded for further analysis. (ST773-NDM-1 producers: ERR3181579, ERR3181597, ERR3181683, GCA_017292115, GCA_023995335, GCF_003725635, GCF_003954355, GCF_003954525, GCF_009664165, GCF_009791355, GCF_013255565, GCF_018448985, GCF_028595405, GCF_028595445, GCF_028595535, GCF_028595585, GCF_028595705, GCF_028595795, GCF_036321415, GCF_036321435, GCF_036321965, GCF_036321985, GCF_036322045, GCF_036418295, SRR23955892, SRR23955894;^5–10,12,16 ST773-NDM-1 non-producers: GCF_000790805, GCF_000791735, GCF_000796095, GCF_002411915, GCF_003585175, GCF_003836135, GCF_007559065, GCF_017693465, GCF_017693745, GCF_017693755, GCF_021245785, GCF_021245865, GCF_028595895, GCF_900636975.^9,12,20

WGS and resistome analysis

Genomic DNA was obtained with a commercially available extraction kit (High Pure PCR Template Preparation Kit, Roche Diagnostics) and indexed paired-end libraries were prepared with the Illumina DNA Prep Library Preparation Kit (Illumina) and sequenced on an Illumina MiSeq^® benchtop sequencer using a MiSeq reagent kit v3 600 cycles (Illumina).

Short reads were de novo assembled with SPAdes v3.15.5 (−careful) and the de novo assemblies were used to infer the ST by using MLST v2.09 according to PubMLST typing schemes,²¹ to explore the presence of horizontally acquired antimicrobial resistance genes (http://genepi.food.dtu.dk/resfinder) and to study the structural integrity of the OprD porin. In addition, a variant calling analysis was performed to study the P. aeruginosa mutational resistome; for this purpose, Snippy software v4.6.0 (https://github.com/tseemann/snippy) was used using the P. aeruginosa PAO1 genome (NC_002516.2) as reference. SNPs and short insertions and deletions (InDels) found in the P. aeruginosa mutational resistome genes were extracted and polymorphisms were filtered.²² The levels of expression, in comparison with the P. aeruginosa reference strain PAO1, of the genes encoding the chromosomal β-lactamase AmpC (ampC) and the three P. aeruginosa efflux pump components [mexB (MexAB-OprM), mexD (MexCD-OprJ) and mexY (MexXY-OprM)], were determined in selected isolates by quantitative (real-time) RT–PCR (RT–qPCR) with an Eco real-time PCR system (Illumina), according to previously described protocols.²³

Genomic characterization of the bla_NDM-1 genetic environments

Representative isolates (HCB22-1-0337, HCB22-1-0461 and HCB23-7-1060) were subjected to Oxford Nanopore long-read sequencing to further investigate the bla_NDM-1 genetic environments. For this, genomic DNA was obtained with the Monarch^® Genomic DNA Purification Kit (New England Biolabs) and libraries were prepared as per the manufacturer’s protocol using the SQK-LSK114 Ligation Sequencing Kit v14 and the SQK-NBD114.24 Native Barcoding Kit and loaded onto a MinION flow cell vR10.4.1 and sequenced in a GridION^TM device (Oxford Nanopore Technologies).

Complete hybrid genomes were obtained combining short and long reads using the de novo hybrid assembler hybracter v0.7.3²⁴ and annotated with Bakta v1.9.4.²⁵ Finally, the bla_NDM-1 genetic environments were visualized and plotted using Proksee (https://proksee.ca.)²⁶

Phylogenetic analysis and comparison with worldwide isolates of P. aeruginosa ST773

To infer the genetic relatedness and possible transmission events among patients, a core SNP-based maximum-likelihood tree was performed using Parsnp v1.2 from the Harvest Suite package, forcing the inclusion (-c) of all genomes. In addition, a core-genome MLST (cgMLST) was performed using the open-source algorithm chewBBACCA;²⁷ allele matrix distances were obtained and a minimum spanning tree was constructed with GrapeTree.²⁸

Results and discussion

Antibiotic susceptibility profiles and resistome characterization of NDM-1-producing P. aeruginosa isolates detected at HCB

A total of 17 NDM-1-producing P. aeruginosa isolates obtained from 10 different patients were studied (Table 1). All isolates were assigned to the emerging high-risk clone ST773, with the single exception of isolate HCB23-6-1048 (Patient 6), which was assigned to ST357.

The cgMLST analysis of the ST773 isolates showed that they were closely related (range: 0–13 allele differences), overlapping the intrapatient minimum and maximum allele differences (0–4) with the interpatient differences (0–7).

The first ST773 NDM-1-producing P. aeruginosa isolate was detected in a 41-year-old woman, who was admitted to the oncology ward in February 2022. She was originally from Ukraine, where she had undergone breast cancer surgery before being relocated to Spain. Next, an NDM-1-producing P. aeruginosa isolate was detected in October 2022 in a 72-year-old woman (Patient 3) who was admitted to the hepatology ward without previous travel history. As these patients did not overlap in time, environmental reservoirs such as high-touch surfaces and sinks were investigated; however, none was identified. Later, 13 more NDM-1-producing P. aeruginosa isolates were detected in 8 additional patients admitted to the traumatology, hepatology and oncology wards. The complete spatio-temporal distribution of isolates (n = 17) and patient admissions are presented in Figure S1 (available as Supplementary data at JAC-AMR Online).

In the ST773-HCB isolates, the bla_NDM-1 gene was chromosomally encoded, along with a bleomycin resistance gene, both located between two IS91-family transposases within an ∼117 kb integrative conjugative element (ICE) (Figure S2). Further characterization revealed that this ICE was identical to a bla_NDM-1 ICE recently described in a NDM-1-producing P. aeruginosa isolate (NDM-Pa-1, OQ806931.1) obtained from an Ukrainian patient attending a hospital in Madrid (>99.99% identity, 116996/116997 bp).¹⁷ This ICE was flanked by 23 bp attL and attR sequences (GTCTCGTTTCCCGCTCCAAACAT) and additionally contained two different aminoglycoside resistance genes (aadA11-like and rmtB4), a quinolone resistance pentapeptide repeat protein (QnrVC1) and a tetracycline efflux protein (TetG). Likewise, genes involved in the mobilization, integration and maintenance of pathogenicity genomic islands were identified.

HCB-ST773 isolates were resistant to all antibiotics tested except colistin and cefiderocol and displayed variable resistance to aztreonam and aztreonam/avibactam. To completely explain the resistance profiles, we further explored the presence of chromosomal resistance mutations. All ST773 isolates lacked the oprD gene due to a 2084 bp deletion (Δnt1043814–1045898, NC_002516.2), which had likely occurred as a result of a homologous recombination between a repetitive sequence of 13 bp (CAAGCTGGGGCGG) located upstream and downstream of the deleted region. Antibiotic resistance mutations were also detected in the genes coding for the glycerol-3-phosphate transporter (glpT-W413X), the DNA gyrase subunit A (gyrA-T83I) and the DNA topoisomerase IV subunit A (parC-S87L), as well as in MexXY (mexZ-aa80Δ3) and MexEF-OprN (mexT-R40H) efflux pump regulators. Furthermore, with the exception of the first isolate from Patient 1 (HCB22-1-0337), all isolates showed additional mutations either in mpl (nt111insC, n = 2), which participates in the regulation of the chromosomal cephalosporinase AmpC, or in the MexAB-OprM efflux-pump regulator nalD (T11P, n = 14) (Table 1). The impact of the mpl and nalD mutations on the increased expression of ampC (25-fold) or mexB (4-fold) was confirmed by RT–qPCR in two representative isolates showing each type of mutation (Table 2). As shown in Table 1, the presence of the nalD mutation leading to MexAB-OprM overexpression was clearly associated with the acquisition of aztreonam and aztreonam/avibactam resistance. Likewise, the mexZ mutation was confirmed to lead to MexXY overexpression but, in contrast, the detected mutation in mexT does not drive MexEF-OprN overexpression and is therefore likely a natural polymorphism (Table 2).

Finally, it should be mentioned that in the ST357 isolate (HCB23-6-1048), the bla_NDM-1 gene was accompanied by a bla_VEB-9 (also designated bla_VEB-1a) ESBL gene, several genes encoding for different aminoglycoside resistance determinants (aph(6)-Id, aph(3′′)-Ib, aac(3)-Id, aac(6′)-Il, rmtB4), a quinolone resistance pentapeptide repeat protein (qnrVC1) and other acquired antibiotic resistance genes (Table 1). The presence of all these acquired resistance determinants correlated well with the antibiotic susceptibility profile, as this isolate was resistant to all antibiotics except colistin and cefiderocol. Recently, Rossell et al.²⁹ investigated the origin of carbapenemase-producing P. aeruginosa ST357 in the Netherlands, comparing the resistome of their isolates with international ones, and concluded that this clone was introduced to the Netherlands via repatriation of critically ill patients from Kenya. The acquired resistome of our isolate (HCB23-6-1048) was identical to an Indian subclade (SNP cluster PDS000019917) of the international collection investigated in that work. Interestingly, HC23-6-1048 was isolated in a 69-year-old man admitted to the traumatology ward who had been previously admitted to an ICU in a hospital in Jaipur (India) with sepsis. We then compared the genetic environment of the bla_NDM-1 in our isolate with a Dutch representative isolate of the Kenyan subclade (isolate RIVM_C050529); we found that while in the Dutch isolate the bla_NDM-1 gene was located within a resistance island containing other resistance determinants such as bla_VEB-9, in our isolate the bla_NDM-1 gene was alone (Figure S2), the bla_VEB-9 gene being located distantly in the chromosome. Thus, both subclades had apparently acquired the bla_NDM-1 and bla_VEB-9 genes through different events, highlighting the ability of this high-risk clone to acquired foreign resistance determinants.

Global analysis of the genomic epidemiology and resistome of ST773 P. aeruginosa

To infer the potential geographical origin of the NDM-1-producing ST773 HCB P. aeruginosa isolates, 40 additional ST773 P. aeruginosa genomes obtained between 2004 and 2022 from patients in different countries worldwide were downloaded from public databases for phylogenetic and resistome comparisons. This international ST773 collection included 26 NDM-1 producers but also included 14 isolates that did not harbour a bla_NDM-1 gene, some of them containing a different carbapenemase-encoding gene such as bla_IMP-45 or bla_VIM-2 (Figure 1).

The genetic relatedness of the HCB isolates and the international ST773 P. aeruginosa genomes was explored, combining both SNP (Figure 1) and gene-by-gene approaches (Figure 2). As shown in Figure 1, NDM-1 producers were demonstrated to be closely related and clustered together, forming a distinct branch separated from non-NDM-1 producers. By cgMLST analysis, a median allele difference within the entire collection of 47 (range 0–143) was determined to be higher among the non-NDM-1 isolates compared with the NDM-1 ones (73 versus 29) (Figure 2). Allele differences confirmed that the HCB isolates were closely related to one another and also to the two isolates included in the international collection from Madrid (range: 0–16). Notably, the isolates from Patient 1 (HCB-1) and the two isolates from Madrid were the most genetically distant, allele differences ranging from 9 to 16 compared with the 0–7 allele differences documented among the other HCB isolates. Moreover, we found that among these patients, intrapatient minimum and maximum allele differences (0–4) overlapped with the interpatient range (0–7). Altogether, these findings pointed to a common geographical origin for the introduction of this NDM-1 ST773 clone in both Spanish hospitals (Ukraine) followed by sustained local interpatient transmission events after its introduction into the HCB.

As shown, with the single exception of the Hungarian isolate (GCF_003725635), all NDM-1 producers had the same acquired resistance determinants as those detected in the HCB isolates [aadA11-like, rmtB4, qnrVC1 and tet(G)], as well as the antibiotic resistance mutations found in mexZ, mexT, gyrA and parC chromosomal genes. However, other HCB genetic markers such as the inactivating mutation in glpT (W413X) and the OprD deletion, were only found in the two NDM-1-producing isolates obtained from two Ukrainian patients attending a hospital in Madrid, genomic features that along with epidemiological data suggest a common origin of these isolates. Moreover, the characterization of the bla_NDM-1 genetic environment in isolate HCB22-1-0461 (Figure S2) additionally supports this hypothesis, as it was shown to be located in the same 117 kb ICE (>99.99% identity, 116996/116997) and inserted within exactly the same chromosomal position as isolate NDM-Pa-1 from Madrid (OQ806931.1).¹⁶

On the other hand, non-NDM-1 producers were less clonal and exhibited more variable resistomes (Figure 1). Among them, just the mutation in mexT was conserved, suggesting that the acquisition of this mutation occurred a long time ago and before the acquisition of other antibiotic resistance genes/mutations.

Conclusions

We report the concomitant detection of NDM-1-producing P. aeruginosa isolates belonging to two distinct high-risk clones, ST357 and ST773, in a tertiary Spanish hospital. Using a combination of molecular and WGS approaches, we have characterized their resistomes in depth, compared them with those of international collections and determined by phylogenetic analysis their most likely geographical origins and routes of transmission.

NDM-1-producing ST773 P. aeruginosa isolates detected worldwide showed a higher degree of similarity. Genomic features and phylogenetic analysis of the HCB ST773 isolates strongly linked them with those recently reported in two Ukrainian patients from a hospital in Madrid, isolates that have been linked to NDM-1-producing ST773 strains found in Ukrainian patients in the Netherlands.¹⁷ Thus, our findings provide further evidence that the ongoing war in Ukraine has been a key factor driving the clonal expansion of this NDM-1-ST773 P. aeruginosa strain across Europe. In addition, we confirmed sustained local interpatient transmission events following the initial introduction of this clone into a tertiary hospital in Barcelona. Finally, our detailed resistome analysis identified common elements to all NDM-1 ST773 isolates and specific features of the lineages disseminated in Europe linked to the Ukrainian war, as well as genomic markers of resistance evolution during its intrahospital dissemination.

In summary, this work evidences the transborder spread and intrahospital dissemination and evolution of the emerging ST773-NDM-1 P. aeruginosa high-risk clone. Likewise, this work underscores the importance of surveillance to prevent the spread of successful high-risk clones and antibiotic resistance determinants worldwide, positing genomic epidemiology and resistome analysis as essential tools to understand their behaviour and transmission patterns.

Funding

This work was supported by the Instituto de Salud Carlos III, Ministerio de Ciencia e Innovación and Unión Europea—NextGenerationEU through grants PI21/00017, PI24/00010 and Personalized and precision medicine grant (MePRAM Project, PMP22/00092). Researchers were also supported by the I+D+i grant PID2021-127402OB-I00, funded by MCIN/AEI/10.13039/501100011033, co-financed by the European Development Regional Fund ‘A way to achieve Europe’ and grant 2021 SGR 01569 from the Departament d’Universitats, Recerca i Societat de la Informació of the Generalitat de Catalunya. We acknowledge support from the Spanish Ministry of Science, Innovation and Universities through the ‘Centro de Excelencia Severo Ochoa 2019–2023’ Program (CEX2018-000806-S), and from the Generalitat de Catalunya ‘CERCA Program’.

Transparency declarations

A.O. has received speaker fees and research grant funding from Shionogi, MSD and Pfizer. All other authors: none to declare.

Data availability

Read files for isolates sequenced in this study have been deposited in the European Nucleotide Archive under project number PRJEB75515.

Supplementary data

Figures S1 and S2 are available as Supplementary data at JAC-AMR Online.

References

Author notes