The genus Enterobacter of the bacterial order Enterobacterales comprises environmental and clinical species (Davin-Regli et al., 2019). Clinical isolates of Enterobacter spp. are primarily opportunistic pathogens (Sanders and Sanders, 1997), involved mainly in hospital-associated infections of the urinary and respiratory tract as well as bloodstream infections (Kang et al., 2004; Ramirez and Giron, 2021). Due to similarities in the taxonomically relevant characters, isolates of different Enterobacter species have been misidentified as “E. cloacae” or “Enterobacter species” or “Enterobacter cloacae complex (ECC)”, and remained unrecognized with respect to their true species nomenclature (De Florio et al., 2018; Wu et al., 2020; Godmer et al., 2021). With the introduction of the high-resolution tools based on whole-genome sequencing in bacterial taxonomy, the precise delineation of bacterial species became possible (Hugenholtz et al., 2021). The genomically revised taxonomic structure of Enterobacter revealed Enterobacter xiangfangensis (also referred to as Enterobacter hormaechei subspecies xiangfangensis) as a common pathogenic species worldwide (Chavda et al., 2016; Peirano et al., 2018; Sutton et al., 2018; Wu et al., 2020; Cho et al., 2021; Wu et al., 2021).

Enterobacter species are frequently resistant to first-line antibiotics such as third-generation cephalosporin, penicillin, aminoglycosides, and quinolones, and fourth-generation cephalosporin and carbapenems are currently the most attractive therapeutic options (Davin-Regli et al., 2019). However, for the last 15 years, carbapenem resistance has been increasingly reported from Enterobacterales (Potter et al., 2016), with isolates of Enterobacter ranked among the top three in this group (Cerqueira et al., 2017). The most predominant carbapenemase type found in E. xiangfangensis was NDM, followed by VIM, KPC, OXA-48, and IMP (Peirano et al., 2018), while GIM-1, GIM-2, and IMI-9 were observed only sporadically (Wendel and MacKenzie, 2015; Di Luca et al., 2016; Wendel et al., 2018). Mobile genetic elements such as plasmids, particularly insertion sequence (IS) elements, were observed as a major determinant for the spread of these carbapenemase genes (Chavda et al., 2016; Potter et al., 2016).

For the “last-resort antibiotic”, such as colistin, Enterobacter species have been considered to be susceptible (WHO, 2018; WHO, 2021). Nevertheless, certain phylogenomic groups of this species exhibit colistin heteroresistance often leading to treatment failures as such isolates may initially be classified as being susceptible (Guérin et al., 2016; Mushtaq et al., 2020). A recent study shows that PhoPQ-dependent regulation of the arnBCADTEF gene cassette for transfer of 4-amino-4-deoxy-l-arabinose (l-Ara4N) to lipid A underlies colistin heteroresistance and resistance (Kang et al., 2019).

A trend of cumulative mobile resistance genes has been noted among members of Enterobacterales, including isolates of Enterobacter (Wang et al., 2019; Yang et al., 2022). Isolates carrying multiple resistance genes against different classes of antibiotics are increasingly reported (Chavda et al., 2019; Li et al., 2019). These genes may be juxtaposed within a single cassette, in different combinations, or on different plasmids within a single isolate. Co-existence of multiple resistance genes engenders extreme drug resistance and treatment of these infections are protracted with few therapeutic options remaining. Surveillance of such extensive or extreme drug-resistant pathogens and respective resistance genes harboring genetic elements is important to understand the epidemiology of dissemination and develop control strategies.

The SurvCARE study monitors the incidence of carbapenem-resistant Enterobacterales (CRE) in patients admitted in hospitals across the state of Hessen in Germany. During the surveillance (period 2017–2019), we noticed an overall increase in the number of the two-carbapenemase-carrying CRE isolates from 1.3% (1/79) in 2017, to 4.4% (5/113) in 2018, to 5.6% (9/162) in 2019, and detected an extreme drug-resistant (XDR) E. xiangfangensis isolate va18651, which was subsequently found, following whole-genome sequence analysis, to carry two carbapenemases, i.e., KPC-2 and OXA-48, as well as the mobile colistin resistance mcr-9 gene. Further analysis reveals that the extreme drug resistance capability is plasmid-based, attained through the acquisition of four different plasmids, three of which carried 37 different antibiotic resistance genes that included two different carbapenemase genes bla _KPC-2 and bla _OXA-48 as well as the mobile colistin resistance gene mcr-9.1.

In a study between 2017 and 2019, we collected carbapenem-resistant Enterobacterales isolates (CRE) obtained from patients covering the entire state of Hessen in Germany. Among these, isolate va18651 was found to harbour two carbapenemase-encoding genes, bla_KPC-2 and bla_OXA-48, and the mobile colistin resistance mcr-9 gene. This combination was unique and not found in the Enterobacter species collection. We also screened 3,246 genomes of Enterobacter strains available from the public database to examine if the combination of these genes had been previously reported. The analysis revealed only nine isolates that harbored the combination of two carbapenemases and mcr-9; eight of these isolates were classified as members of the species E. xiangfangensis. While the OXA-48 is frequently detected in other Enterobacterales species, such as Klebsiella and E. coli, this is the first time we observed the OXA-48 determinant in the genus Enterobacter.

The isolate va18651 was initially identified as a member of ECC by MALDI TOF. Due to taxonomic conflicts and the low discriminatory power of MALDI TOF, Enterobacter species may not be correctly identified to the species level (Pavlovic et al., 2012; Godmer et al., 2021). Using recommended genome sequence-based approaches (Chun et al., 2018), the isolate va18651 was identified as E. xiangfangensis (also listed as E. hormaechei subspecies xiangfangensis) (Sutton et al., 2018; Wu et al., 2020). Enterobacter isolates have been reported to be present in the cervix of healthy individuals (Larsen and Monif, 2001; Amabebe and Anumba, 2018) and constitute a risk factor for the urinary tract infection and premature births.

The pathogenicity and virulence gene repertoire of Enterobacter are not well understood (Sanders and Sanders, 1997). Very few studies that examined the virulence capabilities of Enterobacter species in a mouse model of infection are available (Paauw et al., 2009; Krzymińska et al., 2010; Pati et al., 2018). Enterobacter species belonging to MLST sequence type 88 (ST88) have been reported from sporadic human and animal clinical cases as well as from outbreaks (Jia et al., 2018; Börjesson et al., 2019; Tian et al., 2020). To compare va18651 to other ST88 isolates distributed worldwide, we retrieved genomes of 21 ST88 isolates. Interestingly, 8 of 21 (38%) ST88 isolates carried mcr-9 genes on an identical IncHI2 plasmid ( Figure 1A ). A total of 10 (47.7%) isolates carried at least one carbapenemase (bla _KPC-2 and bla _VIM-1, bla _KPC-18 and bla _NDM-1). These data indicated that carbapenemases and mcr-9 are relatively common to ST88. These data urge further systematic studies with the clonal type ST88 in order to validate the common association of carbapenemases and mcr-9. The presence of an ST type present in a wide range of hosts and in different environmental settings as reflected by its carriage of carbapenemase-encoding genes of different types such as bla _KPC-2 and bla _OXA-48 suggest the emergence of a successful clone transcending species and environmental barriers.

While it is common to observe plasmids of between 10 kb and 500 kb in individual isolates of Enterobacterales, the plasmidome of va18651 was, at 604,632 bp, unusually large (Stephens et al., 2020; Darphorn et al., 2021). Highly identical bla _KPC-2-harboring IncN plasmids were reported earlier in Germany (Yao et al., 2014; Becker et al., 2018; Yao et al., 2021). The bla _OXA-48 plasmid observed was identical to the OXA-48-encoding IncL plasmid, which is reported globally. IncL/M plasmids are an emerging threat as they represent a current source of class D carbapenemases and are responsible for the worldwide distribution of bla _CTX-M. This bla _OXA-48 IncL plasmid was reported from diverse Enterobacterales, suggesting that the p2-va18651 and p3-va18651 and the respective encoded antimicrobial resistances were acquired by horizontal transmission. The environment surrounding mcr-9.1 associated with the IS903B of p1-va18651 was identical with those identified in both chromosomes and IncHI1 plasmids of previous studies (Tyson et al., 2020), indicating an mcr-9 acquisition mediated by mobile genetic elements. In total, the isolate va18651 carried 26 different antibiotic resistance genes, each present in one to four copies, and was distributed among 10 classes including β-lactams, aminoglycosides, fluoroquinolone, macrolides, phenicol, rifampicin, tetracycline, trimethoprim, and quaternary ammonium compound resistance gene cassettes (qacEΔ_1). These genes mediate resistance to a broad range of antibiotics that correspond to the drug and detergent resistance phenotypes.

There are few reports of co-carriage of two carbapenemase classes such as KPC-2 and OXA-48 in K. pneumoniae ST11 in Taiwan and Egypt (Wang et al., 2019; Yang et al., 2022) and KPC-2 and NDM-1 in the Enterobacter ST88 strain from Colombia (Accession No. SRR3110109) as well as VIM with OXA-48 in E. xiangfangensis isolates (Tyson et al., 2020), but there are currently no reports on the co-existence of KPC-2 and OXA-48 carbapenemases together with a colistin resistance-encoding gene mcr-9.1.

Enterobacter species carry an arnBCADTEF gene cassette, which has shown to be responsible for the colistin heteroresistance (Kang et al., 2019). The presence of this gene cassette on the chromosome of va18651 encouraged us to determine the colistin heteroresistance capability. As heteroresistance is an incidental phenomenon, the colistin resistance may or may not be detected in routine laboratories. Nevertheless, such heteroresistance capability is important from a clinical perspective as it results in the failure of the antibiotic therapy (Band and Weiss, 2019; Band et al., 2021). Previous studies indicate that the mcr-9 gene is not induced naturally, and expressed only from artificial promoters or in presence of colistin (Carroll et al., 2019; Kieffer et al., 2019). Thus, while the mcr-9 gene is functional, the conditions that induce its activity are not known.

While va18651 carried a high number of different ARGs on plasmids (p1-va18651, p2-va18651, and p3-va18651), the co-occurrence of a higher number of heavy metal resistance genes (HGR) on the plasmids p1-va18651 and p4va18651 was remarkable. Such co-occurrence of ARG and HRG has been recognized for a long time (Foster, 1983), but this coexistence is poorly understood (Li et al., 2017). Nevertheless, it is likely to provide it with additional survival and adaptive properties in ecological niches within hospital settings.

The E. xiangfangensis isolate va18651 carries a unique combination of risk factors, i.e., virulence genes on the chromosome together with a large number of antibiotic and heavy metal/metalloid resistance genes on plasmids. The presence of multiple plasmids harboring highly mobilizable genetic platforms also provide a “sink” and reservoir for fueling the accelerated dispersion of multiple ARGs, thereby paving the way in creating antimicrobial resistant (AMR)-hypervirulent vectors that could spread easily, expanding the incidence of hard-to-treat infections with fatal outcomes.

The data presented in this study are available through the Bioproject PRJNA837392 in the NCBI databases.

YY and TC conceptualized the manuscript. YY and SD performed the bioinformatic analysis and wrote the draft of the manuscript. JF, JS, SD and CI performed the phenotyping and WGS genome sequencing. TC, SD and YY prepared the final manuscript with inputs from all Co-authors. All authors contributed to the article and approved the submitted version.

This work was supported by the German Federal Ministry of Education and Research (BMBF) within the German Center for Infection Research (DZIF)/grant numbers, 8032808811, 8032808818, and 8032808820 to TC/CI, and 031L0209B, Deep-iAMR to TC.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.