New Delhi Metallo-β-Lactamase Plasmidomes of Intensive Care Unit Clinical and Environmental Enterobacterales Isolates Authors: Florencia Martino, Denise De Belder, Renata Monge, Agustina Forastiero, Molly Patrick, Kara A Moser, Alejandra Corso, Fernando Pasteran, Matthew Robinson, Valeria Fabre Published: 2026 ABSTRACT AbstractBackgroundThere are limited data on the molecular epidemiology of New Delhi metallo-β-lactamase (NDM)–producing carbapenem-resistant Enterobacterales (CRE) in Latin America, despite the high burden of NDM-CRE in the region. This study characterized blaNDM-harboring plasmids from CRE from patients and the environment of an Argentine hospital.MethodWe performed whole-genome sequencing of 40 environmental isolates (Klebsiella complex, n = 11; Enterobacter cloacae, complex n = 16; Citrobacter freundii, n = 10; Serratia marcescens, n = 3) and 4 clinical isolates (Klebsiella pneumoniae, n = 4) recovered from the environment and rectal surveillance samples of patients in the same intensive care units (ICUs) at a tertiary care center in Argentina, July 2023–February 2024, respectively. Representative genomic groups underwent Oxford Nanopore sequencing and hybrid genome assembly for complete plasmid characterization. Resolved blaNDM-harboring plasmids were compared across species and publicly available reference sequences to characterize plasmid relatedness and dissemination patterns.ResultsTen representative genomic groups were identified across species, including 7 defined sequence types and 3 assigned in this study. blaNDM-1 and blaNDM-5 were present in 72.5% (29/40) and 27.5% (11/40) of environmental isolates, respectively (all 4 clinical isolates were blaNDM-5). Hybrid assembly of the representative genomic groups revealed blaNDM-1 was located on highly conserved plasmids belonging to the IncC incompatibility group across multiple CRE species. blaNDM-5 was restricted to Klebsiella spp. and was associated with structurally diverse plasmids including IncF- and IncHI1B/IncFIB-type backbones in both environmental and patient isolates.ConclusionsThese findings highlight the ICU environment as a reservoir for blaNDM plasmids. New Delhi metallo-β-lactamase (NDM)-harboring carbapenem-resistant organisms (CROs) are increasing globally [1–3], with a sustained increase in blaNDM prevalence observed in Latin America and the Caribbean [4], including in Argentina, where this study was conducted [5–10]. Hospitalization is a risk factor for carbapenem-resistant Enterobacterales (CRE) colonization [11]. Transmission of blaNDM CRE in the healthcare setting can occur through multiple mechanisms, including via contaminated hands and medical equipment, or the environment [12]. In a previous study, we documented a substantial burden of blaNDM CROs in the ICU environment of hospitals in resource-limited settings, particularly in sinks [13]. Furthermore, we found evidence of blaNDM CRO transfer between the environment and patients based on genomic and epidemiological analysis [13–16]. bla NDM is most frequently carried by conjugative plasmids capable of transferring across Gram-negative bacterial species, allowing the gene to remain in environmental reservoirs over time [17, 18]. The increasing prevalence of blaNDM-carrying CRE, their impact on patient outcomes, and the potential for horizontal transfer of resistance determinants within patient- and environment-associated microbial communities in healthcare settings highlight the need to better understand the epidemiology and plasmidome of blaNDM-carrying CRE in the hospital setting. This study aimed to characterize the blaNDM plasmidome of environmental and patient isolates to assess plasmid sharing and resistance modules. METHODS Study Design and Setting We performed a retrospective genomic analysis of blaNDM CREs recovered from samples collected from the ICU environment of a tertiary center in Buenos Aires, Argentina, between July 2023 and February 2024, or from patients hospitalized in participating ICUs between September 2023 and January 2024. Details regarding specimen collection and processing, organism isolation and identification methods, characterization of environmental reservoirs, and cluster determination were recently published [13]. Briefly, environmental samples were collected from high-touch surfaces near patients and wet surfaces in sinks using composite samples with the sponge method. Sink drains in patient bathrooms and common areas were sampled using E-swabs. Clinical samples were collected using rectal swabs as part of routine weekly surveillance in the ICU. CROs were cultured using CHROMagar plates and identified through MALDI-TOF (Bruker). Whole-Genome Sequencing and Bioinformatics Analysis Our protocol for whole-genome sequencing (WGS) has been previously described [13]. Briefly, genomic DNA was extracted using the DNeasy Blood & Tissue Kit (Qiagen) following the manufacturer's protocol. Short-read sequencing was performed on Illumina NovaSeq 6000 (150 bp paired-end reads). A subset of 10 isolates was selected for long-read sequencing and hybrid assembly, including isolates from previously identified single nucleotide variant (SNVs)-based clusters. Specifically, we selected isolates from a SNV cluster with a different blaNDM allele and/or a different plasmid replicon profile or different blaNDM-associated resistance combinations. Long-read sequencing was performed on an Oxford Nanopore MinION (Mk1B) using R10.4 flow cells and the Rapid Barcoding Kit (SQK-RBK114). Hybrid assemblies were generated using Unicycler v0.4.8 [19]. Resistance genes, plasmid incompatibility groups, and insertion sequences were identified with PHoeNIx (https://github.com/CDCgov/phoenix) [20], which integrates the AMRFinderPlus, ARG-ANNOT, and ResFinder databases [21, 22]. Plasmid relaxase families, mating pair formation systems, oriT signatures, predicted mobility (conjugative, mobilizable, non-mobilizable), and host range were inferred using MOB-suite with default parameters [23]. Sequence types (STs) were assigned with MLST 2.0, and novel alleles and allelic profiles identified during this study were submitted to the respective BIGSdb/PubMLST and BIGSdb-Pasteur databases for official allele and ST assignment. Genomic comparisons were performed with BLAST and Artemis Comparative Tool [24]. Plasmids backbones resolved and characterized in this study were queried against the NCBI nucleotide database using BLASTn (as of February 2026) to contextualize their genetic relatedness globally as well as for regional dissemination. Matches were retained using thresholds of ≥90% query coverage and ≥90% nucleotide identity. RESULTS General Features and Distribution of blaNDM Genes in Recovered CRE Isolates A total of 44 blaNDM-producing CRE isolates were included in this study: 40 environmental and 4 patient (rectal surveillance samples) isolates. The blaNDM-1 allele was detected in 29/44 (66%) and the blaNDM-5 allele in 15/44 (34%) isolates. All patient isolates carried the blaNDM-5 gene. Genomic and epidemiological information are shown in Figures 1, 2, and 3 and Supplementary Figures 3, 4, and 5. Genomic and epidemiological information of blaNDM-carrying plasmids recovered from clinical and environmental Enterobacterales isolates. The diagram summarizes the distribution of 10 blaNDM-positive plasmids selected for Oxford Nanopore–Illumina hybrid assembly and the corresponding closed plasmids carrying blaNDM. For each isolate, the host species, MLST, collection date, intensive care unit (ICU), sampling area, plasmid incompatibility group, and antimicrobial resistance genes (ARGs) encoded on the blaNDM-carrying plasmid are shown. Six isolates carried blaNDM-1 on highly similar IncC plasmids recovered from multiple Enterobacterales species and environmental reservoirs, whereas 4 isolates carried blaNDM-5 on distinct plasmid backbones (IncP6, IncFIA(HI1)/IncR, IncFII, and IncHI1B(pNDM-MAR)/IncFIB(pNDM-Mar)). Filled squares indicate the presence of a gene or replicon, while empty squares indicate absence or unavailable information. Expanded genomic and epidemiological information of all blaNDM-1-positive isolates recovered during the sampling period. All blaNDM-1-positive isolates are organized by host species, isolation date, intensive care unit (ICU), sampling area, isolate ID, and single nucleotide variant (SNV) cluster. Species are indicated by color, and environmental sampling sources are indicated by icons for sink in common area, sink in patient bathroom, and sink drain in patient bathroom. SNV clusters were defined as isolates belonging to the same species and sequence type with ≤10 pairwise SNV differences except for Serratia marcescens (≤24 pairwise SNV differences). A subset of isolates was selected for Oxford Nanopore–Illumina hybrid assembly and long-read plasmid resolution. Solid lines indicate isolates with long-read-confirmed blaNDM-1-carrying IncC plasmid structures; dashed lines indicate isolates for which the blaNDM-1 plasmid structure could not be fully resolved and was inferred from short-read data, antimicrobial resistance gene (ARG) content, and genomic relatedness. Filled squares indicate gene presence; empty squares indicate gene absence. ARGs and plasmid incompatibility groups were identified using PHoeNIx, which integrates the AMRFinderPlus, ARG-ANNOT, and ResFinder databases. ARGs located on the blaNDM-1-carrying plasmid, or inferred from short-read data, are displayed together with the blaNDM-1 plasmid incompatibility group, plasmid name, and plasmid size when resolved by hybrid assembly. Sequence types (STs) were assigned using MLST 2.0; novel allele or allelic profiles (E. asburiae ST3544 and E. mori ST3545) were submitted to the corresponding BIGSdb/PubMLST for official ST assignment. Expanded genomic and epidemiological information of all blaNDM-5-positive Klebsiella quasipneumoniae and Klebsiella pneumoniae isolates recovered during the sampling period. Isolates are organized by host species, isolation date, intensive care unit (ICU), sampling area, isolate ID, and single nucleotide variant (SNV) cluster. Species and blaNDM-5 plasmid incompatibility groups are indicated by color. Sampling sources are indicated by icons for patient, high-touch surfaces (HTS)-patient area, sink in patient bathroom, and sink drain in patient bathroom. All patient isolates were obtained from unique individuals. SNV clusters were defined as isolates belonging to the same species and sequence type with ≤10 pairwise SNV differences. A subset of isolates was selected for Oxford Nanopore–Illumina hybrid assembly and long-read plasmid resolution. Solid lines indicate isolates with long-read-confirmed blaNDM-5-carrying plasmid structures; dashed lines indicate isolates for which plasmid structure could not be fully resolved and was inferred from short-read data. Filled squares indicate gene presence; empty squares indicate gene absence, except for isolates with unresolved plasmid structures, for which associated ARGs could not be inferred from short-read data. Antimicrobial resistance genes (ARGs) located on the blaNDM-5-carrying plasmid, or inferred from short-read data, are displayed together with the plasmid incompatibility group, and plasmid name and plasmid size when resolved by hybrid assembly. Resistance genes and incompatibility groups were identified using PHoeNIx, which integrates the AMRFinderPlus, ARG-ANNOT, and ResFinder databases. There were 5 clonal clusters among environmental blaNDM-1-producing Enterobacterales, including Citrobacter freundii ST8 and ST257, Enterobacter cloacae ST25, Enterobacter hormaechei ST90, and Enterobacter asburiae ST3544 (assigned in this study). These environmental isolates were mostly recovered from sink basins and drains. The clusters involved both ICUs and spanned over several months (Figure 2). Additionally, we detected several blaNDM-1 CRE in sinks that were not part of a cluster, including Enterobacter soli ST2935, Enterobacter mori ST3545 (assigned in this study), C. freundii ST22 and ST564, as well as Serratia marcescens ST356 (Figure 2 and Supplementary Figure 1). In contrast to blaNDM-1 distribution, blaNDM-5 was restricted to Klebsiella spp., including K. pneumoniae ST11, ST5994, ST469, ST9976 (assigned in this study) and K. quasipneumoniae ST3520. Two ST11 clonal clusters included 2 patient isolates (Figure 3). Environmental blaNDM-5-positive Klebsiella isolates were mostly recovered from high-touch surfaces (Figure 3, Supplementary Figure 2). Genomic Features of blaNDM-1 Across the 6 representative hybrid-WGS blaNDM-1-positive isolates selected for hybrid WGS, blaNDM-1 was located on IncC-type plasmids. The blaNDM-1-harboring IncC plasmids in blaNDM-1 E. cloacae complex ST25, E. hormaechei ST90, and E. asburiae ST3544 were 70–170 kb. Similar IncC plasmids were identified in blaNDM-1 C. freundii, including 138 and 158 kb IncC plasmids in C. freundii ST22 and ST257, respectively, with high nucleotide identity and coverage to those observed in the E. cloacae complex. Similarly, S. marcescens carried blaNDM-1 on a 132 kb IncC plasmid that was largely colinear with the IncC plasmids identified in Enterobacter and Citrobacter, with variation restricted to small accessory regions (Figure 4). Pairwise analysis of blaNDM-1-carrying IncC plasmids from intensive care unit (ICU) environmental isolates. Each horizontal line represents one complete IncC plasmid sequence obtained by hybrid assembly. The first line corresponds to the IncC plasmid from an Enterobacter cloacae complex isolate (JH144Ecl; ICU B; sink in common area; October 2023; 71 kb). The second line shows a single representative sequence because the IncC plasmids recovered from E. cloacae complex (JH289Ecl; ICU B; sink in common area; November 2023; 138 kb) and Enterobacter soli (JH419Ecl; ICU B; sink in patient bathroom; January 2024; 138 kb) were nearly identical to each other and to the previously published Argentinian IncC plasmid pSMA17468 from Serratia marcescens (NCBI accession: MK123268). The third line corresponds to the IncC plasmid from Citrobacter freundii (JH3Cfr; ICU A; bathroom sink drain; July 2023; 158 kb). The last line corresponds to the IncC plasmid from Serratia marcescens (JH279Sma; ICU B; sink splash zone; November 2023; 132 kb). Regions of high nucleotide identity (>98%) are shown as shaded blocks connecting the sequences. Arrows indicate coding sequences; antimicrobial resistance genes are highlighted. All blaNDM-1-harboring IncC plasmids contained blaNDM-1 adjacent to the same resistance region, within an otherwise conserved IncC backbone, but differed in nearby accessory resistance content across host species (Figure 4). The main difference was a localized insertion in the E. cloacae complex IncC plasmid, indicating a localized accessory insertion (Figure 4). Genomic Features of blaNDM-5 The blaNDM-5 allele was detected exclusively in Klebsiella spp., and the hybrid assemblies showed that blaNDM-5 was carried on structurally diverse plasmid backbones both within and across Klebsiella species (Figures 3 and 5). In Klebsiella quasipneumoniae ST3520, blaNDM-5 was located on a small 39 kb IncP6 plasmid (Figure 5). The same strain also carried a second larger plasmid of 239 kb with the IncHI1B(pNDM-MAR)/IncFIB(pNDM-Mar) replicons, which did not harbor blaNDM-5. In contrast, in Klebsiella pneumoniae ST5994, blaNDM-5 was integrated into a related approximately 250 kb IncHI1B(pNDM-MAR)/IncFIB(pNDM-Mar) hybrid plasmid. Comparison of these large IncHI1B/IncFIB hybrid plasmids showed a conserved backbone between the 250 kb plasmid in K. pneumoniae ST5994 and the 239 kb plasmid in K. quasipneumoniae ST3520, with the key difference being the insertion of the blaNDM-5-containing region in the K. pneumoniae plasmid, shared with other antimicrobial resistance genes. This region was highly similar to the blaNDM-5 region carried by the 39 kb IncP6 plasmid in K. quasipneumoniae ST3520, supporting exchange of a conserved blaNDM-5-associated resistance segment between distinct plasmid backbones (Figure 5). Pairwise analysis of blaNDM-5-associated plasmids from Klebsiella isolates. Each horizontal line represents one complete plasmid sequence obtained by hybrid assembly. The first line corresponds to the IncHI1B(pNDM-MAR)/IncFIB(pNDM-Mar) plasmid from K. quasipneumoniae ST3520 (pJH17Kqpn_239; 239 kb), which does not carry blaNDM-5. The second line corresponds to the IncHI1B(pNDM-MAR)/IncFIB(pNDM-Mar) plasmid from K. pneumoniae ST5994 (pJH-p-15Kpn_250; 250 kb), which carries blaNDM-5 and is 100% identical at the nucleotide level to pJH17Kqpn_239 across the shared backbone, excluding the insertion region which contains blaNDM-5 resistance region, which is highly similar to that carried by the IncP6 plasmid from K. quasipneumoniae ST3520 (pJH17Kqpn_39; 39 kb) represented as the last line. Regions of high nucleotide identity (>98%) are shown as shaded blocks connecting the sequences. Arrows indicate coding sequences; antimicrobial resistance genes are highlighted. In K. pneumoniae ST11, hybrid assembly identified blaNDM-5 on 2 distinct plasmid backbones: an 89 kb IncFIA(HI1)/IncR plasmid and a 104 kb IncFII plasmid. Although these plasmids differed in replicon content, their blaNDM-5-associated resistance regions shared 100% nucleotide identity across the complete aligned region (14 476 bp). bla NDM-Positive Short-Read Assemblies Short-read assemblies of blaNDM-positive CRE showed plasmid replicons and blaNDM-associated resistance-gene profiles matching those observed in the hybrid assembly isolates (Figures 2 and 3 and Supplementary Figures 3, 4, and 5). This included IncC-associated contigs carrying the blaNDM-1 region in multiple Enterobacterales isolates, as well as the same blaNDM-5-associated replicon markers and resistance-gene sets observed in the hybrid Klebsiella representatives. BLASTn comparison against the hybrid-resolved IncC plasmid references had ≥99% coverage and ≥99% nucleotide identity to one or more closed IncC reference plasmids. Antimicrobial Resistome Beyond blaNDM IncC plasmids carrying blaNDM-1 uniformly contained a conserved accessory region (CMY-type β-lactamases, aminoglycoside-modifying enzymes, sulfonamide-resistance genes, and bleO). We found additional antimicrobial resistance genes on separate genetic elements (Supplementary Figures 3 and 4). For example, a C. freundii ST22 isolate harbored blaKPC-2, and E. soli ST2935 uniquely carried a qnrE1-associated multidrug-resistance plasmid, independent of the IncC backbone, that also included blaPER-2, aac(6')-Ib-cr5, blaOXA-1, catB3, arr-3, sul1, mph(A), and blaTEM-1. For blaNDM-5-harboring Klebsiella isolates, we found accessory regions that were more heterogeneous and had broad antimicrobial resistance (Supplementary Figure 5). The 39 kb IncP6 plasmid in K. quasipneumoniae ST3520 carried blaNDM-5 with dfrA12, aadA2, sul1, bleO, rmtB1, and blaTEM-1. The same isolate also carried a separate 239 kb IncHI1B(pNDM-MAR)/IncFIB(pNDM-Mar) plasmid encoding additional resistance genes, including sul2, qnrS1, blaCTX-M-15, blaOXA-1, catB3, and mph(A). This distribution contrasted with K. pneumoniae ST5994, a related IncHI1B/IncFIB plasmid that had acquired the same blaNDM-5 region with a broader accessory resistance region that shared with the 39 kb IncP6 plasmid (Figure 5). Additional plasmid-associated resistance elements included an IncFIB(K) plasmid carrying tet(D) and blaSHV-1 in the ST5994 isolate and a small ColpVC plasmid encoding qnrB19 in other ST11 isolates (Supplementary Figure 5). Comparative Analysis of Plasmid Mobility and Structural Relatedness of Current and Historical Isolates Analysis of blaNDM-1-carrying IncC plasmids identified the core genetic machinery for conjugation (including a relaxase and a mating pair formation system), consistent with the potential for dissemination across a broad range of bacterial hosts. Furthermore, all blaNDM-1-carrying IncC plasmids recovered from Enterobacter isolates from this study were 100% identical to publicly available IncC plasmids previously reported in clinical Enterobacterales in Argentina (Figure 4) [8, 25]. In addition, we identified highly similar IncC plasmids in Enterobacterales from multiple countries in public databases. Regarding blaNDM-5 plasmids, large IncHI1B/IncFIB- and IncF-type plasmid backbones were predicted to be conjugative or mobilizable, although in K. quasipneumoniae, blaNDM-5 was detected on a small IncP6 plasmid lacking detectable conjugation machinery and was therefore predicted to be non-mobilizable. The plasmid harbored by K. pneumoniae ST5994 in the present study closely matched a IncFIB/IncHI1B blaNDM-5 plasmid identified in a K. pneumoniae ST307 in 2021 in Argentina (CP166600) with 96% sequence coverage and 100% nucleotide identity. This blaNDM-5 plasmid matched publicly available K. pneumoniae plasmids from other countries, whereas the other 2 blaNDM-5 plasmids did not show BLASTn matches meeting the ≥90% coverage and ≥90% nucleotide identity thresholds. DISCUSSION In this study, we analyzed the plasmids associated with blaNDM-1 and blaNDM-5 of CRE recovered from ICU patients and their hospital environment. We found 2 clearly distinct plasmid-mediated dissemination patterns. blaNDM-1 was consistently found on a highly conserved IncC plasmid backbone across multiple Enterobacterales genera, whereas blaNDM-5 was restricted to Klebsiella spp. and was associated with more structurally diverse plasmid backbones. Several of the isolates from this study were genetically related based on SNV analysis, supporting a role of the ICU environment as a vector for healthcare transmission [13]. In our study, we identified several clusters related to sinks. Sinks have been associated with sustained transmission of CRE in healthcare settings through biofilm formation in common plumbing [26]. Although previous studies have shown that the NDM variants (eg, NDM-1 and NDM-5) may influence the susceptibility profile [14–16], it remains unclear if they are associated with different advantages for survival in the environment. Chromosomal relatedness, captured by ST and SNV distances of bacterial genomes, is the first layer to assess transmission because it reflects vertical spread through clonal bacterial expansion. Characterization of the plasmidomes improves the understanding of AMR gene dissemination within transmission networks at the host and pathogen level. IncC plasmids are well-recognized carriers of blaNDM-1 [12, 27]. In our study, blaNDM-1 was consistently embedded within a highly conserved IncC backbone circulating across E. cloacae complex, C. freundii, and S. marcescens. The hybrid-assembled IncC plasmids shared an almost identical core structure, including the same resistance genes, and were nearly indistinguishable from a validated conjugative IncC plasmid recovered from a clinical specimen in 2014 [8] and in 2020 and 2021 [25] in Argentina. Closely related blaNDM-1-carrying IncC plasmids have been described in clinical isolates from multiple countries, supporting the notion that this IncC backbone is broadly distributed among diverse Enterobacterales hosts [27–36]. These data suggest IncC plasmids offer a stable and low-plasticity backbone that may support interspecies maintenance and spread of blaNDM-1. In our study, blaNDM-5 was confined to Klebsiella spp. and was associated with structurally diverse and compositionally dynamic plasmids, including IncF, IncHI1B/IncFIB, and IncP6 replicons. Our findings suggest high plasmid plasticity, evidenced by blaNDM-5 plasmid segment exchange and recombination of replication modules with substantial variation in accessory resistance genes among blaNDM-5-positive Klebsiella isolates. The coexistence of small plasmid elements such as ColpVC with larger conjugative plasmids carrying extended resistance determinants illustrates the capacity of these plasmids to accumulate and reshuffle resistance genes within a single host lineage. Our results support previous findings that IncF-type plasmids facilitate the emergence and diversification of blaNDM-5 and may contribute to broader host-range dissemination [12, 27]. Hybrid assembly of clinical and environmental K. pneumoniae ST11 isolates showed that blaNDM-5 was carried on plasmids that differed in replicon composition and accessory resistance regions. Although specific ST11 clusters were defined using a ≤10-SNV threshold based on whole-genome pairwise comparisons corrected for putative recombination or acquisition events, all ST11 isolates analyzed remained closely related, with pairwise distances ranging from 0 to 39 SNVs. By contrast, their blaNDM-5-carrying plasmids differed in replicon composition and accessory resistance content, indicating that most detectable diversity within this lineage resides primarily in the plasmidome rather than in the chromosome. In addition to blaNDM, several isolates carried additional resistance determinants that broadened the functional load of their plasmids. Accessory genes within IncC plasmids were largely uniform, whereas IncF and IncHI1B/IncFIB plasmids displayed wider and more heterogeneous profiles, incorporating ESBLs, PMQR genes, and 16S methylases. We also detected qnrE1 on an Enterobacter plasmid, a configuration consistent with ISEcp1-mediated mobilization from the chromosome as proposed by Albornoz and colleagues [37]. This highlights the role of insertion sequences in capturing, mobilizing, and reshuffling antimicrobial resistance genes across plasmids and chromosomes. There are several limitations to this study. Samples were obtained from a single center, limiting the generalizability of results; however, blaNDM is the dominant carbapenemase in Argentina [5, 38], and our analysis of historical isolates makes these data relevant. Due to cost constraints, we performed hybrid assembly on a limited number of isolates; however, several short-read isolates belonged to the same species/ST clusters as the hybrid-assembly representatives with extremely small SNV distances, supporting similar mechanisms of resistance as those detected through hybrid assembly. While this study identified distinct plasmid-associated patterns for blaNDM-1 and blaNDM-5 dynamics, it did not directly evaluate whether these differences influence environmental persistence, transmission efficiency, or dissemination in the environment or among bacterial species. Finally, we included a few clinical isolates, which were all Klebsiella spp. Future studies including a larger sample are needed to better characterize NDM-related plasmidomes in humans. To our knowledge, this is the first study characterizing NDM-harboring plasmids from environmental CRE recovered from a healthcare setting in Argentina. The study provides insight into blaNDM dissemination and may inform future studies to optimize strategies to minimize the burden of AMR in the hospital setting. Supplementary Material