Genomic Epidemiology of Global Carbapenemase-Producing Escherichia coli, 2015–2017

The E. coli population is dominated by diverse sequence types with varied geographic distributions, warranting ongoing genomic surveillance.

C arbapenems are effective options available for treating serious infections caused by multidrugresistant (MDR) Enterobacterales bacteria (1). The emergence of carbapenem resistance is a major public health concern, and the World Health Organization has identified carbapenem-resistant Enterobacterales as critical-priority bacteria (2).
Because E. coli is mainly responsible for human community-associated infections (6), it evades conventional hospital-based infection-prevention measures (7). E. coli is an important One Health (i.e., human, animal, environmental health) reservoir for antimicrobial resistance (AMR) genes (8). Tracking global mobile genetic elements and E. coli clones associated with carbapenemase genes is a public health priority (9) and aids in designing management and prevention strategies.
Comprehensive epidemiology data about carbapenemase-producing E. coli is limited to institutional, regional, or countrywide surveys (10). We used short-read whole-genome sequencing (WGS) to describe the molecular characteristics and international distribution of carbapenemase-producing E. coli. We describe the geographic distribution of different carbapenemase genes (including their associations with dominant sequence types [STs], clades and underlying mobile genetic elements), other β-lactamases, AMR genes, and virulence factors.
We defined major STs as representing >10% and minor STs as representing 5%-10% of the total E. coli carbapenemase population (12). Dominant STs were both major and minor STs.

Statistical Analyses
We conducted all analyses in R 3.6.1 (27). Initially, we attempted to fit generalized linear mixed models with country-level random effects to summarize comparisons between dominant STs with respect to antimicrobial and virulence genes. Most models failed to converge, possibly because of the low number of isolates   for some STs and the large number of countries involved. Thereafter, we attempted to use exact logistic regression models for clustered data, as previously described (28). Similarly, most models failed to converge, particularly for comparisons involving ST1284 where all isolates were obtained from a single country. We then used Fisher exact tests to perform pairwise comparisons of antimicrobial and virulence genes among dominant STs. We used Mann-Whitney tests for comparison of virulence scores between dominant STs. We adjusted p values for multiple comparisons within each outcome by using the false discovery rate (29). We defined statistical significance as p>0.05.

Discussion
A World Health Organization report showed the lack of adequate surveillance programs in many parts of the world, especially from lower-and middle-income countries (LMICs) (38). That report identified bacteria, including carbapenem-resistant E. coli, where global surveillance data are urgently required. LMICs bear a considerable share of the disease burden attributable to MDR E. coli but lack adequate genomic surveillance systems (39). Our study aimed to describe the global molecular epidemiology of 229 carbapenemase-producing E. coli obtained from 36 countries (including 20 LMICs) during 2015-2017. Isolates with multiple AMR genes dominated the population. The most common carbapenemase group was the OXA-48-like carbapenemases (44%), followed by NDMs (32%), KPCs (21%), VIMs (2%), and IMPs (1%). OXA-48-like carbapenemases were numerous in Egypt, Jordan, and Turkey; NDMs were numerous in Egypt, Thailand, and Vietnam, and KPCs were numerous in Colombia, Italy, and the United States.
A recent survey of global carbapenemase-producing E. coli for the period 2002-2017 included 343 carbapenem-resistant isolates obtained mainly from the United States (40). KPC (16%), NDM (16%), and OXA-48-like (13%) carbapenemases were common. The study screened for different E. coli phylogroups and certain STs (ST131, ST648, and ST405). Phylogroup B2 isolates were common, and phylogroup A was dominant in Asia. Global ST131 with bla KPC s was the most common ST, followed by ST648 with bla OXA-48-like and ST405 with bla NDM s.
Molecular-based surveillance studies have shown that OXA-48-like enzymes are common among global carbapenemase-producing Enterobacterales (4,5). OXA-48 is currently the most common OXA-48-like derivative and OXA-181 the second most common derivative (44). OXA-48 is endemic in North Africa, Middle East, and Turkey (44). E. coli with bla OXA-48 is linked to various STs (44). In our study, OXA-48 was identified among 18 STs from 15 countries. E. coli with bla OXA-48 was common in Turkey, where it was linked with ST11260.
Molecular-based surveillance studies have shown that NDMs are often the most common carbapenemase in certain regions (e.g., the Indian subcontinent) (4,5). NDM-1 is the most frequent NDM enzyme and associated with various STs within diverse plasmid platforms (45). In our survey, NDM-1 was identified among 14 different STs obtained from 10 countries. E. coli with bla NDM-1 was not linked with a specific ST and was evenly distributed among the different countries.
E. coli with bla KPC is associated with ST131 (47) on diverse plasmid platforms (48). E. coli ST131 is global MDR high-risk clone associated with fluoroquinolone resistance and bla CTX-M s (49). ST131 belongs to clades A/H41, B/H22, and C/H30 (50). C/H30 is divided into subclades C0, C1_nonM27, C1_M27, and C2. In our survey, KPC genes were found among 26 different STs from 11 countries. E. coli with bla KPC was common in Colombia linked with various STs. ST131 was responsible for 32% of KPC isolates and obtained from Italy, Israel, Guatemala, Puerto Rico, and the United States (including Puerto Rico). ST131 with bla KPC was dominated by the C1_nonM27 subclade. This dominance is different from that observed by Johnson et al. study (40), where the C2 subclade was common. The ST131-C1_M27 subclade in our survey was positive for bla NDM-1 and bla OXA-232 .
Among this study's strengths is that it included a large global collection of recent isolates representing multiple LMICs. We characterized all isolates using short-read WGS and provided novel information regarding the geographic distribution and MDR determinants of dominant STs and their respective clades and subclades (e.g., global ST410 was linked with bla OXA-181 , ST131 with bla KPCs , ST167 with bla NDM-5 , and ST405 with various carbapenemases).
We showed that the underlying molecular epidemiology within the same carbapenemase groups were very different (e.g., NDM-1 was linked with various STs, including ST131-C2/H30, whereas NDM-5 was linked with ST167-B and ST410B4/H24Rx). The geographic distribution of isolates with NDM-1 and NDM-5 was different (e.g., NDM-1 showed global distribution whereas those with NDM-5 were numerous in Egypt, Thailand, and Vietnam). Similar differences were described for isolates with OXA-48 (various STs) and OXA-181 (linked with ST410B4/H24Rx). Future genomic surveys should use methodologies that characterize individual carbapenemases.
We also showed that global bla OXA-181 was harbored on near identical IncX3 plasmids (irrespective of the ST or geographic location). This finding suggests that highly similar IncX3 plasmids were mainly responsible for the global distribution of OXA-181 genes, the most common carbapenemase in this collection. The control of such IncX3 plasmids should be a public health priority.
Limitations of this study include the fact that flanking regions and plasmids harboring carbapenemases were not fully reconstructed because of the limitations of short-read sequencing (30). The characterization of plasmids is vital to fully comprehend the molecular epidemiology of global carbapenemase-producing E. coli, and a follow-up study using long-read sequencing is under way. Several countries included only few isolates (Table) and therefore may not be fully representative of what carbapenemaseproducing E. coli dominates in that region.
In summary, the global carbapenemase-producing E. coli population is dominated by diverse STs with different characteristics and varied geographic distributions. This characterization was especially apparent within certain carbapenemases groups (i.e., NDM-1 vs. NDM-5 or OXA-48 vs. OXA-181). Ongoing genomic surveillance to characterize individual carbapenemases will assist in designing management and prevention strategies to help curtail the spread of AMR bacteria.

About the Author
Dr. Peirano is a research associate at Alberta Precision Laboratories and the University of Calgary. Her research interests include the molecular epidemiology of antimicrobial-resistant organisms. from Egypt). One ST131-C2 with NDM-1 was also positive for VIM-1.