Skip to main content
NCBI home page
JAC-Antimicrobial Resistance logoLink to JAC-Antimicrobial Resistance
. 2022 May 30;4(3):dlac050. doi: 10.1093/jacamr/dlac050

Genomic epidemiology, antimicrobial resistance and virulence factors of Enterobacter cloacae complex causing potential community-onset bloodstream infections in a tertiary care hospital of Nepal

Sulochana Manandhar 1,2,†, Quynh Nguyen 3,†, To Nguyen Thi Nguyen 4, Duy Thanh Pham 5,6, Maia A Rabaa 7,8, Sabina Dongol 9, Buddha Basnyat 10,11, Sameer M Dixit 12, Stephen Baker 13, Abhilasha Karkey 14,15,✉
PMCID: PMC9155248  PMID: 35663828

Abstract

Objectives

Community-onset bloodstream infections (BSIs) caused by carbapenemase-producing Enterobacter cloacae complex (ECC) species are increasing internationally. This observation suggests that ECC are emerging pathogens, requiring for detailed understanding on their genomic epidemiology including transmission dynamics and antimicrobial resistance profiles.

Patients and methods

We performed WGS on 79 Enterobacter spp. isolated from the patients with clinically significant BSIs and admitted to emergency department of a major tertiary hospital in Nepal between April 2016 and October 2017.

Results

We identified 5 species and 13 STs of ECC. Enterobacter xiangfangensis ST171, one of the globally emerging carbapenem resistant ECC clones with epidemic potential, was the most prevalent (42%). Phylogenetic analysis showed a large (>19 400 SNPs) core genome SNP distance across major STs, which was minimal (<30 SNPs) among the isolates of each prevalent ST, suggesting the relatively recent importation of major STs followed by local clonal expansions. Genomic evidence for resistance to all major antimicrobial classes except for colistin and macrolides was detected. A limited number of isolates also carried blaNDM-1 (n = 2) and blaOXA-48 (n = 1) carbapenemase genes. Virulence factors encoding siderophores (24%), T6SSD (25%) and fimbriae (54%) were detected.

Conclusions

Our study highlighted that MDR ECC clones are important pathogens of BSIs in community. Though of low prevalence, carbapenem resistance observed in our ECC isolates raised concern about further community dissemination, underscoring the need for community surveillance to identify MDR ECC clones with epidemic potential.

Introduction

Bacteria belonging to Enterobacter cloacae complex (ECC) are opportunistic Gram-negative bacilli that comprise a part of the healthy gut microbiota of humans and animals. ECC are also members of the ESKAPE group of pathogens and are frequently associated with MDR nosocomial infections.1 ECC bacteria are intrinsically resistant to several β-lactam antimicrobials because of constitutive expression of the chromosomally located AmpC gene.2,3 Further, a rapid emergence and spread of carbapenem-resistant E. cloacae complex (CREC) variants have been widely reported, complicating the therapeutic management of patients infected with MDR infections.4–6

MDR strains of ECC bacteria can colonize high-contact surfaces in ICUs, where exposure to several clinically used broad-spectrum antimicrobials is probable.7,8 Under suboptimal infection control measures in hospitals, the colonized ECC bacteria can cause nosocomial infections in vulnerable patient population.7,9 In communities, specifically of low- and middle-income countries, the rampant use of broad-spectrum antimicrobials may facilitate the selection and colonization of MDR ECC in human gut. The resulting reservoir can serve as an endogenous source of invasive community-onset bloodstream infections (CO-BSIs), specifically in individuals with debilitating health conditions, such as reduced immunity, malignancy, diabetes and severe illnesses requiring prolonged hospitalization and antimicrobial therapy.9

Bloodstream infection (BSI) is one of prevalent diagnoses among patients with acute febrile illness seeking urgent healthcare in emergency department (ER) of hospitals globally.10–12 The prevalence of Enterobacterales is high among ER-attending patients with bacterial BSI.10,13–15 Studies on aetiology of bacterial CO-BSIs among ER-attending patients is limited in Nepal. A retrospective study conducted in a major tertiary hospital reporting on 23 years (1992–2014) of microbiological data constituting >220 000 bacterial isolates from blood samples is one of the comprehensive studies on bacterial BSIs in Nepal.13 The study evidenced an increasing temporal trend in the proportion of MDR non-Salmonella Enterobacterales, particularly those originating from the outpatients. Further, after Salmonella enterica Typhi and Paratyphi A, MDR ECC was identified to be the most prevalent non-Salmonella Enterobacterales causing BSIs, mostly among ER-attending patients.13 The findings underscored a potentially emerging role of ECC in causation of BSIs. Other studies have also highlighted MDR ECC as an increasingly important bacterial aetiology of CO-BSIs, thereby threatening the current antimicrobial therapy.16,17

Genomic investigation permits an exact identification of species of ECC group, genetic determinants of antimicrobial resistance (AMR) and virulence, and bacterial population structure. Such data can be used to determine the emergence and spread of clinically and epidemiologically relevant Enterobacter spp. A WGS-based phylogenetic study conducted on E. cloacae isolates causing neonatal sepsis in a tertiary hospital in Nepal elucidated the transmission dynamics of infection, and also identified an environmental source for one of the outbreaks.7 Genomic studies on clinical isolates of ECC bacteria causing CO-BSIs among ER-attending patients is however lacking in the existing literature. To fulfill this knowledge gap and aiming to characterize the species, AMR genes and phylogenetic relationships of ECC bacteria causing potential CO-BSIs, we performed WGS and comprehensive genomic analyses of ECC isolated from the patients admitted to ER of a major tertiary hospital in Nepal.

Patients and methods

Ethics

Ethical approval for conducting this study was obtained from Nepal Health Research Council (NHRC) under the approval registration number of 120/2012 for the research entitled ‘Prevalence of extended spectrum β-lactamase (ESBL) pathogens and factors responsible for ESBL associated infections within Patan Hospital’.

Study setting and design

This was a retrospective study on routine laboratory results and archived bacterial isolates in the microbiology department at Patan Hospital, one of the largest tertiary hospitals in Nepal. Annually, Patan Hospital provides healthcare services to nearly 320 000 outpatients and 20 000 inpatients. The ER provides acute healthcare services to >100 patients per day, and >36 000 patients per year.

Study definitions and case ascertainment

A CO-BSI was defined as a BSI producing positive blood culture result from a blood sample taken from a patient having undifferentiated fever within 48 h of admission. At the ER, patients stayed for a maximum of 12–24 h post-admission. Therefore, all ER patients presenting with an undifferentiated fever at admission and producing a positive blood culture were considered to have a potential CO-BSI. After collection of a blood sample for microbiological culture, the patient was empirically treated with IV third-generation cephalosporins. Patients showing clinical improvement were discharged with a prescription of oral azithromycin and requested to come to a follow-up examination with blood culture report on the fifth day. Alternatively, patients failing to show clinical improvement or undergoing deterioration were admitted to an appropriate hospital ward. This being a retrospective study, information on clinical care taken by the patients in other healthcare facilities before being admitted to the ER of Patan Hospital could not be recorded.

Blood culture

A blood sample was collected at bedside by an attending nurse after an aseptic skin preparation and following a standard protocol. For adults, 8–10 mL of blood was collected in a BD BACTEC Plus aerobic/F culture vial (Becton Dickinson, Sparks, MD, USA); while for children, 1–3 mL of blood was collected in BD BACTEC PedsPlus/F culture vial. The inoculated vials were incubated in an automated BD BACTEC FX40 continuous culture system, and subjected to routine microbiological processing for bacterial isolation, identification and antimicrobial susceptibility testing (AST). The Gram-negative bacilli that were lactose-fermenting, motile, indole-negative, urease-negative, citrate-positive, hydrogen sulphide non-producing, and giving acid/acid reaction with gas production in a triple sugar iron test were presumptively identified as Enterobacter spp. AST was performed using the modified Kirby–Bauer disc diffusion method where the antimicrobial discs were placed on Mueller–Hinton agar swabbed with 0.5 McFarland units of overnight bacterial suspension to create a confluent bacterial lawn.18 After incubation at 35 ± 2°C for 18 h, zones of inhibition were measured, and the results were interpreted referring to the zone size breakpoints as recommended by CLSI.19 Pure colonies of Enterobacter spp. were suspended in Red Protect Bacterial Bead Preservation System (TSC Technical Service Consultants Ltd, Lancashire, UK), and stored at −80°C for genomic analyses.

Isolate ascertainment

The microbiology laboratory information management database of Patan Hospital was interrogated for Enterobacter spp. isolated from blood samples originating from the patients attending the ER between April 2016 and October 2017. Replicate Enterobacter spp. isolates having an identical AST profile and originating from the same patient were excluded from the analysis. The archived Enterobacter isolates stored at −80°C in the freezer were retrieved, revived and processed for genomic analyses.

Bacterial conjugation

The conjugation experiment was performed between the recipient isolate of Escherichia coli J53 strain (resistant to amikacin and sodium azide) and each of the donor ECC isolates of ST171, ST134 and ST1377 by combining equal volumes (5 mL) of overnight LB cultures. The bacterial strains were conjugated at 37°C for 12 h in LB broth. The resulting E. coli transconjugants were selected on MacConkey agar containing sodium azide (100 mg/L), amikacin (64 mg/L) and ceftriaxone (6 mg/L). The transfer of AMR plasmids was confirmed by plasmid extraction from the E. coli transconjugants using a modified Kado and Liu method,20 followed by gel electrophoresis. AST of E. coli transconjugants was performed for ciprofloxacin, ofloxacin, cefotaxime, ceftriaxone, co-trimoxazole, gentamicin, amikacin, chloramphenicol, meropenem, imipenem, colistin, tetracycline, streptomycin and erythromycin using the modified Kirby–Bauer disc diffusion method.18

WGS and genomic analysis

Genomic DNA of Enterobacter isolates was extracted using the Wizard Genomic DNA Extraction Kit (Promega, WI, USA) by following manufacturer’s recommendations. The dual index-tagged pooled whole genome DNA libraries were prepared using the Nextera XT DNA Library Preparation Kit (Illumina, CA, USA). WGS was performed on a HiSeq X Ten platform (Illumina) to generate 150 bp paired-end reads.

After an initial quality assessment of raw Illumina reads by FastQC v0.11.5,21 the adapters were removed using Trimmomatic v0.36.22 SRST2 v0.2.023 was used to identify the AMR genes using the ARG-ANNOT24 database; virulence factors using BIGSdb25 and VFDB (http://www.mgc.ac.cn/VFs/) databases; plasmid replicons using the PlasmidFinder26 database, and MLST using MLST typing scheme (https://pubmlst.org/organisms/enterobacter-cloacae). SRST2 used Bowtie2 to map the raw reads against the reference databases, and SAMtools to detect the gene variants. Raw reads were de novo assembled using Unicycler v0.4827 to generate the contigs, followed by a genome annotation using Prokka v1.5.28 Bandage29 was used to visualize the assembly graph of plasmid-carrying isolates. Subsequently, BLASTn which was integrated in Bandage, was used to identify and extract the assembled sequences containing the detected AMR genes and plasmid replicons of IncF/IncR. The extracted sequences were used to search for the best-matching plasmids in NCBI database using BLASTn.

The taxonomic assignment of Enterobacter spp. is complex. Recently, whole genome based average nucleotide identity (ANI) (defined as a mean nucleotide identity of all orthologous genes shared between two microbial genomes) has been used to generate an updated taxonomy of the Enterobacter genus.30 Here, we used a fast approximate ANI tool, called MASH v1.1.1,31 to generate a pairwise ANI-based distance matrix between our Enterobacter isolates and the published reference collection of 22 Enterobacter species,30 comprising Enterobacter quasiroggenkampii, Enterobacter quasimori, Enterobacter wuhouensis, Enterobacter quasihormaechei, Enterobacter huaxiensis, Enterobacter chuandaensis, Enterobacter sichuanensis, Enterobacter chengduensis, Enterobacter soli, E. cloacae, Enterobacter mori, Enterobacter bugandensis, Enterobacter ludwigii, Enterobacter cancerogenus, Enterobacter asburiae, Enterobacter hormaechei, Enterobacter xiangfangensis, Enterobacter kobei, Enterobacter timonensis, Enterobacter oligotrophica, Enterobacter dissolvens and Enterobacter roggenkampii. A pairwise ANI with a cut-off threshold of 96% was used to identify the appropriate species.

Pairwise SNP distance

We sought to compare the genetic distance between the three predominant Enterobacter STs (ST171, ST134 and ST1377) in our collection. Initially, Roary32 was used to construct a pan genome for the three STs with a blastp percentage identity of 99% and a core definition of 99%. SNP-sites v2.1.333 was then used to extract the SNP sites from the core gene alignment (1012 genes), resulting in an alignment of 47 601 SNPs. Pairwise genetic distances (difference in the number of SNPs) within and between the three STs were estimated from the SNP alignment using the ape (v4.1) and adegenet (v2.0.1) packages in R (v3.3.2).34

Phylogenetic analysis

A pairwise SNP distance analysis showed significant SNP differences between the three predominant STs (ST171, ST134 and ST1377), but limited SNP differences within individual STs. To obtain a further phylogenetic resolution for each ST, we reconstructed a pan genome for individual STs using Roary32 (blastp percentage identity of 99%, and core definition of 99%). The core gene alignment of each ST was subjected to Gubbins (v2.3.2)35 to remove the regions of recombination, followed by an extraction of SNPs using SNP-sites v2.1.3.33 The number of core genes for each ST was 3833 (ST171), 3542 (ST134) and 3742 (ST1377). The SNP alignment of each ST was used to construct a maximum likelihood (ML) phylogenetic tree using IQ-TREE v1.4.4,36 with the best-fit evolutionary model (K3Pu + ASC) identified based on the Bayesian Information Criterion in jModelTest implemented in this software. The support for the ML tree was assessed via 100 pseudo-replicates. Further, to examine the phylogeny of ECC ST171 in a global context, we combined genomic data from our study together with 93 publicly available assembled genomes of ST171 isolates from NCBI.37 Core gene alignment was re-analysed using Roary v3.9.1 (roary -e –mafft -p 8 *.gff). The number of core genes were 3689 genes, corresponding to an alignment length of 3 602 642 bp (∼75% of genome size). Subsequently, Gubbins v1.4.5 (default settings) was used to identify and to remove the recombinant SNPs. A final alignment of 1217 core SNPs was used to construct the ML phylogenetic tree using IQ-TREE v1.4.4,36 with the K3Pu + ASC model of nucleotide substitution. Support for the ML tree was assessed via 100 pseudo-replicates.

Accession numbers

All raw Illumina sequence reads of 79 Enterobacter spp. generated in this study on HiSeq X Ten platform (Illumina) were deposited in the European Nucleotide Archive under the accession numbers from ERS6485160 to ERS6485239 (Table S1, available as Supplementary data at JAC-AMR Online). The whole genome sequences of two Enterobacter spp. used as reference strains for BLAST Ring Image Generator (BRIG)38 were retrieved from NCBI (accession numbers: CP031569 and AP021911.1).

Results

Demographics

Between April 2016 and October 2017, we recovered a total of 127 Enterobacter spp. isolated from blood cultures across different departments at Patan Hospital. Of these, 79 (62%) isolates originated from the ER-attending patients having clinically suspected BSIs. The median age of these ER-attending patients was 41 years (IQR: 30–59.5), with a male to female ratio of 0.92:1.

Genome-based species and ST identification of Enterobacter spp.

Enterobacter spp. of 5 different species and 13 STs were identified (Figure 1a). E. xiangfangensis (73.4%, 58/79) and E. hormaechei (20.2%, 16/79) were the two most common species, followed by E. hoffmannii (3.8%, 3/79), E. asburiae (1.3%, 1/79), and E. quasihormaechei (1.3%, 1/79). Among the identified STs, ST171 was found to be the most predominant (41.8%, 33/79), followed by ST134 (22.8%, 18/79), ST1377 (15.2%, 12/79), ST not identified (7.5%, 6/8) and several diverse singleton STs (n = 8). The three prevalent STs (ST171, ST134, and ST1377) collectively accounted for 80% of the total isolates. ST171 and ST134 isolates belonged to E. xiangfangensis, and ST1377 isolates were identified as E. hormaechei.

Figure 1.

Figure 1.

Open in a new tab

Species, ST and phenotypic AST profiles of Enterobacter isolates. (a) The species distribution of our Enterobacter isolates and the distribution of STs within each species. (b, c and d) The phenotypic AST profile of all isolates, of E. xiangfangensis and of E. hormaechei, respectively. The y-axis indicates the percentage of isolates. R, resistant; I, intermediate; S, susceptible.

Antimicrobial susceptibility profile

All 79 Enterobacter isolates were resistant to cefotaxime (Figure 1b). A high proportion of isolates were phenotypically resistant to gentamicin (82.3%; 65/79), chloramphenicol (82.3%; 65/79), co-trimoxazole (92.4%; 73/79) and ciprofloxacin (60.8%; 48/79). Alternatively, resistance to carbapenems was low, with 5.1% (4/79) and 7.5% (6/79) of isolates being resistant to meropenem and imipenem, respectively. At species level, in comparison to E. hormaechei isolates (n = 16), E. xiangfangensis isolates (n = 58) were more frequently resistant to gentamicin (95% versus 38%) and chloramphenicol (93% versus 38%), but less resistant to ciprofloxacin (54% versus 81%) (Figure 1c and d). The proportion of resistant isolates of E. xiangfangensis and E. hormaechei were comparable for cefotaxime (100% versus 100%), meropenem (5.2% versus 5.8%) and co-trimoxazole (92% versus 94%).

Genes for AMR and virulence

A high proportion of Enterobacter spp. isolates tested positive for a diverse set of AMR genes conferring resistance to all classes of antimicrobial agents except carbapenems, macrolides and colistin (Figure 2). Of note, E. xiangfangensis isolates had the most extensive array of AMR genes, including blaCTX-M-1, blaACT, qnrB, dfrA5, strA/B, aac(3)-IIa, aac-aad, blaOXA-1, blaTEM-1D, catA, tet(A), sulII and fosA2. Three isolates of E. xiangfangensis additionally had carbapenem resistance genes including blaOXA-48 (ST171, n = 1) and blaNDM-1 (ST134, n = 1; ST231, n = 1). In comparison to E. xiangfangensis, the E. hormaechei isolates had less extensive AMR gene profiles, namely blaACT, qnrB, blaCTX-M-1 and dfrA5. One E. hormaechei isolate (ST528) also carried a blaNDM-1 gene.

Figure 2.

Figure 2.

Open in a new tab

The distribution of AMR genes by Enterobacter species. The results for minor species E. asburiae (n = 1), E. hoffmannii (n = 3) and E. quasihormaechei (n = 1) were grouped together as ‘Others’. The y-axis indicates the percentage of isolates carrying each AMR gene.

By ST, all ST171 and ST134 isolates carried a comparable group of AMR genes, namely aac(3)-IIa, aac-aad, blaCTX-M-15, blaOXA-1, blaTEM-1D, dfrA5, fosA2, qnrB, strAB and tet(A), predicted to confer resistance to aminoglycosides, third-generation cephalosporins, fosfomycin, trimethoprim, quinolones, streptomycin and tetracyclines (Figure 3). Relatively, ST1377 isolates had fewer AMR genes (aac-aad, blaCTX-M-15, blaOXA-1, dfrA5 and qnrB). The distribution of virulence factors was distinct across the three STs. A majority of ST171 isolates had flagella-encoding gene (fliA) and type-3 fimbriae biosynthetic cluster (mrkA-F), whereas ST134 isolates predominantly carried a salmochelin gene cluster, and an hcp/tssD gene associated with type VI secretion system (T6SS), and ST1377 isolates had a siderophore yersiniabactin gene cluster and lpfB-lpfC genes encoding for long polar fimbriae (Figure 3).

Figure 3.

Figure 3.

Open in a new tab

Recombination-filtered core genome-based ML phylogenetic trees for ECC isolates of ST171 (blue dots), ST134 (red dots) and ST1377 (yellow dots). The phylogenetic trees of ST171, ST134 and ST1377 were rooted using genetically closely related outgroups of ST303 (E. xiangfangensis), ST202 (E. hoffmannii) and ST528 (E. hormaechei), respectively. The tip label indicates the bacterial identification number, the species of ECC and the date of isolation. The presence of an array of AMR genes (blue), plasmid replicons (brown) and virulence genes (purple) for each isolate are also shown in coloured square boxes to the right of the trees. Grey shading indicates the absence of the character state. The scale bar indicates nucleotide substitutions per site.

Plasmid replicons

All ST171 isolates carried a large IncFIB plasmid showing a substantial genetic similarity to a non-AMR IncFIB plasmid of a clinical E. hormaechei isolate detected in USA (accession number: CP031569) (coverage: 90%, identity: 99%). The plasmid had self-transfer modules and multiple heavy metal resistance gene clusters for mercury, copper, iron and arsenic (Figure S1). The conjugation experiments confirmed that the IncFIB plasmid was transferable conferring resistance to ceftriaxone, cefotaxime, erythromycin, streptomycin and co-trimoxazole to the otherwise susceptible recipient bacteria. Although ST134 and ST171 displayed an identical AMR gene profile, we were unable to detect any specific plasmid replicon type in our ST134 isolates by plasmid mapping. However, the conjugation experiments determined that ST134 also carried a self-transferable plasmid conferring resistance to ceftriaxone, cefotaxime, erythromycin, streptomycin and co-trimoxazole. Further, we found that all ST1377 isolates had an IncR plasmid with high synteny and similarity to an IncR plasmid from an environmental Klebsiella pneumoniae isolate reported from Japan (accession number: AP021911.1) (coverage 73%, identity 99%, Figure S2). This plasmid was also transmissible and mediated resistance to ceftriaxone, cefotaxime, erythromycin, streptomycin and co-trimoxazole.

Phylogenetic relatedness and spatiotemporal distribution of the three predominant Enterobacter STs

We sought to investigate the phylogenetic relatedness of the three predominant STs (ST171, ST134, and ST1377). The average pairwise SNP distance between ST171 and ST134 (E. xiangfangensis) was 19 429 SNPs, indicating that the two STs were genetically distant though being belonged to the same species. Predictably, ST171 and ST134 were more distantly related to ST1377 (E. hormaechei) than each another, with average pairwise SNP differences of 35 698 SNPs (ST171 versus ST1377) and 34 632 SNPs (ST134 versus ST1377). In contrast, the total number of SNPs identified within each ST was very limited; the ST171 isolates varied by only 27 SNPs, while the isolates within ST134 and ST1377 each differed by only 6 SNPs. The Enterobacter isolates within each ST formed a distinct cluster (Figure 3). We performed further phylogenetic analysis of our ST171 isolates together with a global collection of carbapenem-resistant ST171 isolates. We found that our ST171 isolates belonged to a single cluster (median pairwise SNP distance: 5 SNPs, range: 0–13 SNPs) separate from the global isolates. The nearest neighbour to our cluster was an ST171 isolate identified in Guatemala in 2013. The medium pairwise SNP distances between the Nepalese clade and the outbreak clade of ST171 from the USA was 288 SNPs (IQR: 283–293 SNPs). These data suggest that our Nepalese ST171 isolates are probably resulting from a single importation followed by local transmission events (Figure S3).

We found a notable trend in temporal distribution of the three above STs over the study period. Although both ST171 and ST134 belonged to E. xiangfangensis species, they were circulating in two distinct time periods. ST134 was exclusively found between April and July 2016 and subsequently disappeared, while ST171 was first detected in September 2016 and until then became a predominant E. xiangfangensis genotype circulating in Nepal (Figure S4). Furthermore, almost all ST1377 (E. hormaechei) isolates were identified between September and December 2016, followed by a substantial decline in the following months with only two isolates detected in February and March 2017. We performed spatial mapping and analysis of the three STs from our limitedly available geographical locations, but we did not find any significant clustering (Figure S5).

Discussion

The genomic characterization of BSIs causing Enterobacter spp. is important for understanding the epidemiological landscape and AMR trends in these organisms. Here, in this genomic epidemiological study, we found diverse species and STs of Enterobacter isolates causing potential CO-BSI. E. xiangfangensis (predominantly ST171 and ST134) followed by E. hormaechei (ST1377) were the most prevalent species and STs circulating in this setting. The dominance of these two species of Enterobacter species in clinical samples was in concordance with a genomic study comprising a global collection of 170 CREC isolated between 2008 and 2014.37 Notably, we found that E. xiangfangensis isolates carried more AMR genes than E. hormaechei. Additionally, we observed that the majority of BSI-causing Enterobacter isolates, irrespective of their exact species, were largely susceptible to carbapenem (>94%), but excessively resistant to third-generation cephalosporins (100%). This was an important finding of the study for designing an optimal empirical antimicrobial therapy in our setting.

The prevalence of CO-BSIs caused by MDR ECC bacteria has increased over the last decade, reducing the therapeutic options.13,16 The fact that all ECC isolates from our study were ESBL-positive showing an absolute phenotypic resistance to third-generation cephalosporins, with a few isolates additionally carrying carbapenem-resistance genes, raises a serious concern. A high prevalence of BSIs caused by ESBL-producing ECC may prompt for an increased use of carbapenems in clinical settings, which will consequently lead to an enrichment of CREC variants. Earlier studies have suggested that widespread use of carbapenems has produced an antimicrobial pressure, which served as a key factor for the global emergence and proliferation of ST171 as a dominant CREC clone.39 Routine AMR surveillance with the integration of WGS data rapidly captures the emergence and tracks the transmission of novel Enterobacter species/genotypes, including the carbapenem-resistant variants.

We found limited genetic diversity among each population of the three predominant ECC STs in Nepal. There are several potential explanations for our observations, including a recent strain importation followed by a rapid clonal expansion, a recent population bottleneck, or community outbreaks. Here, we did not find sufficient evidence for community outbreaks with regards to temporal and/or geographical clustering (Figure S5); instead, we found evidence that the ST171 population likely resulted from a single importation and subsequent local transmission. Furthermore, the fact that we observed temporal changes in transmission of the three predominant ECC STs suggests for potential influence of environmental factors, such as seasonal variation. Such a phenomenon has been observed in the transmission dynamics of isolates of S. enterica Typhi, a major waterborne pathogen that also causes BSIs.40 Collectively, we surmise that the strain importation followed by clonal expansion and environmental factors are the probable factors shaping the observed population genetic structure of ECC in our setting.

The factors that drove the transmission of such clones are largely unknown. Several studies have reported that asymptomatic carriage of ECC bacteria in body parts such as gut, anterior nares, skin and upper respiratory tract may act as an endogenous source of invasive infections in susceptible individuals.41 The ability of ECC to colonize humans may play an important role in its transmission and the development of drug-resistant phenotypes.3,42 We also cannot rule out the possible role of environmental transmission given what we observed in temporal changes of ECC genotypes. We must improve our understanding on the sources and transmission chains of these organisms to improve public health control measures.

The presence of a wide repertoire of virulence factors and AMR genes among the ECC isolates may offer important adaptive mechanisms and contribute to their spread, given the sustained use of antimicrobials in the community.39,43,44 However, the natural reservoir of resistance-virulence determinants in ECC is not well described. In this study, we showed that the resistance-virulence attributing plasmids of ECC are transferable between and across the bacterial genus. We hypothesize that bacteria sharing the same ecological niches may be a major reservoir for the horizontal transmission of antimicrobial resistance/virulence genes in ECC bacteria. The interaction and plasmid/gene transfer dynamics between Enterobacter and other Gram-negative organisms warrants further investigation.

Being a retrospective study, complete demographic and clinical data for patients before admission to ER were lacking. We considered that all clinically significant BSIs among patients presenting to ER were potentially CO-BSIs. As we had no prior information on whether these patients recently received any healthcare services from another health institution before presenting to our ER, we could not exclude that some or all of these infections were healthcare-associated being acquired in other hospitals. If the patients had received recent medical care from the same health institution(s), then they may have acquired the Enterobacter infection from the corresponding institution instead from the community. This discrepancy may explain the low genetic diversity of Enterobacter isolates within specific STs, particularly of ST134 and ST1377, which showed fewer than 10 SNP differences. This warrants a further genomic investigation on clinical ECC isolates of the hospitals accompanied by a systematically collected demographic, pre-clinical and clinical data including prior recent exposure to healthcare settings, underlying predisposing factors, use of antimicrobials and final outcome of the patients.

Supplementary Material

dlac050_Supplementary_Data
Click here for additional data file. (1.6MB, zip)

Acknowledgements

We wish to acknowledge all staff in the clinical microbiology laboratory of Patan Hospital and molecular epidemiology group of Oxford University Clinical Research Unit of Vietnam.

Contributor Information

Sulochana Manandhar, Oxford University Clinical Research Unit, Patan Academy of Health Sciences, Kathmandu, Nepal; Centre for Tropical Medicine and Global Health, Medical sciences division, Nuffield Department of Medicine, University of Oxford, Oxford, UK.

Quynh Nguyen, Oxford University Clinical Research Unit, Hospital for tropical diseases, Ho Chi Minh City, Vietnam.

To Nguyen Thi Nguyen, Oxford University Clinical Research Unit, Hospital for tropical diseases, Ho Chi Minh City, Vietnam.

Duy Thanh Pham, Centre for Tropical Medicine and Global Health, Medical sciences division, Nuffield Department of Medicine, University of Oxford, Oxford, UK; Oxford University Clinical Research Unit, Hospital for tropical diseases, Ho Chi Minh City, Vietnam.

Maia A. Rabaa, Centre for Tropical Medicine and Global Health, Medical sciences division, Nuffield Department of Medicine, University of Oxford, Oxford, UK Oxford University Clinical Research Unit, Hospital for tropical diseases, Ho Chi Minh City, Vietnam.

Sabina Dongol, Oxford University Clinical Research Unit, Patan Academy of Health Sciences, Kathmandu, Nepal.

Buddha Basnyat, Oxford University Clinical Research Unit, Patan Academy of Health Sciences, Kathmandu, Nepal; Centre for Tropical Medicine and Global Health, Medical sciences division, Nuffield Department of Medicine, University of Oxford, Oxford, UK.

Sameer M. Dixit, Center for Molecular Dynamics Nepal, Kathmandu, Nepal

Stephen Baker, Department of Medicine, University of Cambridge, School of Clinical Medicine, Cambridge Biomedical Campus, Cambridge, UK.

Abhilasha Karkey, Oxford University Clinical Research Unit, Patan Academy of Health Sciences, Kathmandu, Nepal; Centre for Tropical Medicine and Global Health, Medical sciences division, Nuffield Department of Medicine, University of Oxford, Oxford, UK.

Funding

This work was supported by the Wellcome Trust and the Royal Society (grant number 100087/14/Z) and Oak Foundation (AK OCAY-15-547).

Transparency declarations

None to declare.

Supplementary data

Table S1 and Figures S1 to S5 are available as Supplementary data at JAC-AMR Online.

References

  • 1. Rice  LB. Federal funding for the study of antimicrobial resistance in nosocomial pathogens: no ESKAPE. J Infect Dis  2008; 197: 1079–81. [DOI] [PubMed] [Google Scholar]
  • 2. Jacoby  GA. AmpC β-lactamases. Clin Microbiol Rev  2009; 22: 161–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Davin-Regli  A, Pagès  JM. Enterobacter aerogenes and Enterobacter cloacae; versatile bacterial pathogens confronting antibiotic treatment. Front Microbiol  2015; 6: 392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Gomez-Simmonds  A, Hu  Y, Sullivan  SB  et al.  Evidence from a New York City hospital of rising incidence of genetically diverse carbapenem-resistant Enterobacter cloacae and dominance of ST171, 2007-14. J Antimicrob Chemother  2016; 71: 2351–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Hargreaves  ML, Shaw  KM, Dobbins  G  et al.  Clonal dissemination of Enterobacter cloacae harboring blaKPC-3 in the upper midwestern United States. Antimicrob Agents Chemother  2015; 59: 7723–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Chavda  KD, Chen  L, Fouts  DE  et al.  Comprehensive genome analysis of carbapenemase-producing Enterobacter spp.: new insights into phylogeny, population structure, and resistance mechanisms. mBio  2016; 7: e02093-16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Stoesser  N, Sheppard  AE, Shakya  M  et al.  Dynamics of MDR Enterobacter cloacae outbreaks in a neonatal unit in Nepal: insights using wider sampling frames and next-generation sequencing. J Antimicrob Chemother  2015; 70: 1008–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Davin-Regli  A, Lavigne  JP, Pagès  JM. Enterobacter spp. update on taxonomy, clinical aspects, and emerging antimicrobial resistance. Clin Microbiol Rev  2019; 32: e00002-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Sanders  W, Sanders  C. Enterobacter spp. pathogens poised to flourish at the turn of the century. Clin Microbiol Rev  1997; 10: 220–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Rothe  K, Wantia  N, Spinner  CD  et al.  Antimicrobial resistance of bacteraemia in the emergency department of a German university hospital (2013-2018): potential carbapenem-sparing empiric treatment options in light of the new EUCAST recommendations. BMC Infect Dis  2019; 19: 1091. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Shapiro  NI, Wolfe  RE, Wright  SB  et al.  Who needs a blood culture? A prospectively derived and validated prediction rule. J Emerg Med  2008; 35: 255–64. [DOI] [PubMed] [Google Scholar]
  • 12. Villalon  N, Farzan  N, Freeman  K. Rate of bacteremia in the hemodialysis patient presenting to the emergency department with fever: a retrospective chart review. Int J Emerg Med  2018; 11: 29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Zellweger  RM, Basnyat  B, Shrestha  P  et al.  Changing antimicrobial resistance trends in Kathmandu, Nepal: a 23-year retrospective analysis of bacteraemia. Front Med  2018; 5: 262. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Cheung  Y, Ko  S, Wong  OF  et al.  Clinical experience in management of bloodstream infection in emergency medical ward: a preliminary report. Hong Kong J Emerg Med  2019; 28: 215–26. [Google Scholar]
  • 15. Anderson  DJ, Moehring  RW, Sloane  R  et al.  Bloodstream infections in community hospitals in the 21st century: a multicenter cohort study. PLoS One  2014; 9: e91713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Al-Hasan  MN, Lahr  BD, Eckel-Passow  JE  et al.  Temporal trends in Enterobacter species bloodstream infection: a population-based study from 1998-2007. Clin Microbiol Infect  2011; 17: 539–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Wang  S, Xiao  S-Z, Gu  F-F  et al.  Antimicrobial susceptibility and molecular epidemiology of clinical Enterobacter cloacae bloodstream isolates in Shanghai, China. PLoS One  2017; 12: e0189713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Bauer  A, Kirby  W, Sherris  J  et al.  Antibiotic susceptibility testing by a standardized single disk method. Am J Clin Pathol  1966; 45: 493–496. [PubMed] [Google Scholar]
  • 19. CLSI . Performance Standards for Antimicrobial Susceptibility Testing—Twenty-Seventh Edition: M100. 2017.
  • 20. Kado  CI, Liu  ST. Rapid procedure for detection and isolation of large and small plasmids. J Bacteriol  1981; 145: 1365–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Andrews  S. FastQC: A Quality Control Tool for High Throughput Sequence Data. 2014. https://www.bioinformatics.babraham.ac.uk/projects/fastqc/.
  • 22. Bolger  AM, Lohse  M, Usadel  B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics  2014; 30: 2114–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Inouye  M, Dashnow  H, Raven  L  et al.  SRST2: Rapid genomic surveillance for public health and hospital microbiology labs. Genome Med  2014; 6: 90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Gupta  SK, Padmanabhan  BR, Diene  SM  et al.  ARG-ANNOT, a new bioinformatic tool to discover antibiotic resistance genes in bacterial genomes. Antimicrob Agents Chemother  2014; 58: 212–220. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Jolley  KA, Maiden  MCJ. BIGSdb: scalable analysis of bacterial genome variation at the population level. BMC Bioinformatics  2010; 11: 595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Carattoli  A, Zankari  E, García-Fernández  A  et al.  In silico detection and typing of plasmids using PlasmidFinder and plasmid multilocus sequence typing. Antimicrob Agents Chemother  2014; 58: 3895–903. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Wick  RR, Judd  LM, Gorrie  CL  et al.  Unicycler: resolving bacterial genome assemblies from short and long sequencing reads. PLoS Comput Biol  2017; 13: e1005595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Seemann  T. Prokka: rapid prokaryotic genome annotation. Bioinformatics  2014; 30: 2068–9. [DOI] [PubMed] [Google Scholar]
  • 29. Wick  RR, Schultz  MB, Zobel  J  et al.  Bandage: interactive visualization of de novo genome assemblies. Bioinformatics  2015; 31: 3350–2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Wu  W, Feng  Y, Zong  Z. Precise species identification for Enterobacter : a genome sequence-based study with reporting of two novel species. Enterobacter quasiroggenkampii sp. nov. and Enterobacter quasimori sp. nov. mSystems  2020; 5: e00527-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Ondov  BD, Treangen  TJ, Melsted  P  et al.  Mash: fast genome and metagenome distance estimation using MinHash. Genome Biol  2016; 17: 132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Page  AJ, Cummins  CA, Hunt  M  et al.  Roary: rapid large-scale prokaryote pan genome analysis. Bioinformatics  2015; 31: 3691–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Page  AJ, Taylor  B, Delaney  AJ  et al.  SNP-sites: rapid efficient extraction of SNPs from multi-FASTA alignments. Microb Genom  2016; 2: e000056. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Duy  PT, Vu Thieu  NT, Nguyen To  NT  et al.  The genetic signatures of Salmonella Typhi carriage in the human gallbladder. bioRxiv  2020; 10.1101/2020.06.08.140053. [DOI] [Google Scholar]
  • 35. Croucher  NJ, Page  AJ, Connor  TR  et al.  Rapid phylogenetic analysis of large samples of recombinant bacterial whole genome sequences using Gubbins. Nucleic Acids Res  2015; 43: e15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Nguyen  LT, Schmidt  HA, Von Haeseler  A  et al.  IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Mol Biol Evol  2015; 32: 268–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Peirano  G, Matsumura  Y, Adams  MD  et al.  Genomic epidemiology of global carbapenemase-producing Enterobacter spp., 2008-2014. Emerg Infect Dis  2018; 24: 1010–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Alikhan  NF, Petty  NK, Ben Zakour  NL  et al.  BLAST Ring Image Generator (BRIG): simple prokaryote genome comparisons. BMC Genomics  2011; 12: 402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Annavajhala  MK, Gomez-Simmonds  A, Uhlemann  AC. Multidrug-resistant Enterobacter cloacae complex emerging as a global, diversifying threat. Front Microbiol  2019; 10: 44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Pham Thanh  D, Thompson  CN, Rabaa  MA  et al.  The molecular and spatial epidemiology of typhoid fever in rural Cambodia. PLoS Negl Trop Dis  2016; 10: e0004785. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Flynn  DM, Weinstein  RA, Nathan  C  et al.  Patients’ endogenous flora as the source of “nosocomial” Enterobacter in cardiac surgery. J Infect Dis  1987; 156: 363–8. [DOI] [PubMed] [Google Scholar]
  • 42. Das  P, Singh  AK, Pal  T  et al.  Colonization of the gut with Gram-negative bacilli, its association with neonatal sepsis and its clinical relevance in a developing country. J Med Microbiol  2011; 60: 1651–60. [DOI] [PubMed] [Google Scholar]
  • 43. Zhou  K, Yu  W, Cao  X  et al.  Characterization of the population structure, drug resistance mechanisms and plasmids of the community-associated Enterobacter cloacae complex in China. J Antimicrob Chemother  2018; 73: 66–76. [DOI] [PubMed] [Google Scholar]
  • 44. Sakeena  MHF, Bennett  AA, McLachlan  AJ. Non-prescription sales of antimicrobial agents at community pharmacies in developing countries: a systematic review. Int J Antimicrob Agents  2018; 52: 771–82. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

dlac050_Supplementary_Data
Click here for additional data file. (1.6MB, zip)

Articles from JAC-Antimicrobial Resistance are provided here courtesy of British Society for Antimicrobial Chemotherapy and Oxford University Press

RESOURCES