WHO Critical Priority Escherichia coli as One Health Challenge for a Post-Pandemic Scenario: Genomic Surveillance and Analysis of Current Trends in Brazil

ABSTRACT The dissemination of carbapenem-resistant and third generation cephalosporin-resistant pathogens is a critical issue that is no longer restricted to hospital settings. The rapid spread of critical priority pathogens in Brazil is notably worrying, considering its continental dimension, the diversity of international trade, livestock production, and human travel. We conducted a nationwide genomic investigation under a One Health perspective that included Escherichia coli strains isolated from humans and nonhuman sources, over 45 years (1974–2019). One hundred sixty-seven genomes were analyzed extracting clinically relevant information (i.e., resistome, virulome, mobilome, sequence types [STs], and phylogenomic). The endemic status of extended-spectrum β-lactamase (ESBL)-positive strains carrying a wide diversity of blaCTX-M variants, and the growing number of colistin-resistant isolates carrying mcr-type genes was associated with the successful expansion of international ST10, ST38, ST115, ST131, ST354, ST410, ST648, ST517, and ST711 clones; phylogenetically related and shared between human and nonhuman hosts, and polluted aquatic environments. Otherwise, carbapenem-resistant ST48, ST90, ST155, ST167, ST224, ST349, ST457, ST648, ST707, ST744, ST774, and ST2509 clones from human host harbored blaKPC-2 and blaNDM-1 genes. A broad resistome to other clinically relevant antibiotics, hazardous heavy metals, disinfectants, and pesticides was further predicted. Wide virulome associated with invasion/adherence, exotoxin and siderophore production was related to phylogroup B2. The convergence of wide resistome and virulome has contributed to the persistence and rapid spread of international high-risk clones of critical priority E. coli at the human-animal-environmental interface, which must be considered a One Health challenge for a post-pandemic scenario. IMPORTANCE A One Health approach for antimicrobial resistance must integrate whole-genome sequencing surveillance data of critical priority pathogens from human, animal and environmental sources to track hot spots and routes of transmission and developing effective prevention and control strategies. As part of the Grand Challenges Explorations: New Approaches to Characterize the Global Burden of Antimicrobial Resistance Program, we present genomic data of WHO critical priority carbapenemase-resistant, ESBL-producing, and/or colistin-resistant Escherichia coli strains isolated from humans and nonhuman sources in Brazil, a country with continental proportions and high levels of antimicrobial resistance. The present study provided evidence of epidemiological and clinical interest, highlighting that the convergence of wide virulome and resistome has contributed to the persistence and rapid spread of international high-risk clones of E. coli at the human-animal-environmental interface, which must be considered a One Health threat that requires coordinated actions to reduce its incidence in humans and nonhuman hosts.

and siderophore production was related to phylogroup B2. The convergence of wide resistome and virulome has contributed to the persistence and rapid spread of international high-risk clones of critical priority E. coli at the human-animal-environmental interface, which must be considered a One Health challenge for a post-pandemic scenario. IMPORTANCE A One Health approach for antimicrobial resistance must integrate wholegenome sequencing surveillance data of critical priority pathogens from human, animal and environmental sources to track hot spots and routes of transmission and developing effective prevention and control strategies. As part of the Grand Challenges Explorations: New Approaches to Characterize the Global Burden of Antimicrobial Resistance Program, we present genomic data of WHO critical priority carbapenemase-resistant, ESBL-producing, and/or colistin-resistant Escherichia coli strains isolated from humans and nonhuman sources in Brazil, a country with continental proportions and high levels of antimicrobial resistance. The present study provided evidence of epidemiological and clinical interest, highlighting that the convergence of wide virulome and resistome has contributed to the persistence and rapid spread of international high-risk clones of E. coli at the humananimal-environmental interface, which must be considered a One Health threat that requires coordinated actions to reduce its incidence in humans and nonhuman hosts. KEYWORDS Enterobacterales, resistome, virulome, high-risk clones, One Health, multidrug resistance, carbapenems, colistin, carbapenemases, ESBL, MCR, genomic surveillance, South America A ntibiotic resistance is an ever-growing threat that contributes to serious adverse consequences, such as therapeutic failure, economic burden, and increased mortality rates worldwide (1,2). In this regard, the multisectoral (human, animal, and agriculture) overuse of important antimicrobial drugs has been considered the main driver of the multidrug resistance phenomenon (3). In an attempt to overcome this global public health challenge, the World Health Organization (WHO) has provided a red alert for "Highest Priority Critically Important Antimicrobials'' (i.e., broad-spectrum cephalosporins, carbapenems, and polymyxins), as well as insights about emergent critical-priority pathogens, including Escherichia coli (2,4).
To mitigate these challenges, an interdisciplinary One Health approach has been encouraged to prevent and combat the emergence and dissemination of antibiotic resistance interlinking humans, animals, and their shared environments (10). Herein, we performed a One Health surveillance using microbiological methods and whole-genome sequencing (WGS) to provide an up-to-date scenario of the epidemiology, antimicrobial-resistance profile, genomic features, phylogenetic relationship, and patterns of E. coli recovered from the human-animal-environmental interface in Brazil, the largest country in Latin America.

RESULTS
One Health background of E. coli strains. E. coli genomes investigated included 167 strains from human-animal-environment interface circulating in 40 Brazilian cities, and 16 states with highest population density (North: Pará and Tocantins states; Northeast: Paraíba, Pernambuco, Ceará, Sergipe, Rio Grande do Norte and Bahia states; Midwest: Goiás and the Federal District states; Southeast: Minas Gerais, São Paulo, and Rio de Janeiro states; South: Paraná, Santa Catarina, and Rio Grande do Sul states).
The 54 E. coli strains screened in this study displayed resistance to clinically relevant fluoroquinolones, aminoglycosides, colistin, broad-spectrum cephalosporins and/or carbapenems, of which 42 were classified as multidrug-resistant strains (Fig. 1).
E. coli clones belonging to pandemic sequence types (STs). Among 167 genomes of E. coli strains circulating at the human-animal-environmental-food interface (from humans; pets: dog and cat; livestock animals: cattle, chicken/poultry, horse, and turkey; wild animals: anteater, elephant, fish, ocelot, owl, penguin, and vulture; meat: chicken; crustacean: shrimp; seafood: mussels and oysters; vegetables: spinach and cabbage; freshwater; seawater; soil; drainage; and directional residue) in all regions of Brazil ( Fig. 2A), we identified 69 different multilocus sequence types, in addition to 4 novels STs. The dissemination of international clones ST10, ST38, ST117, ST131, ST224, ST354, ST410, ST457, ST648, and ST744 was confirmed (Fig. 2B). Most common STs associated with human hosts were ST131, ST410, and ST354, whereas those associated with animal hosts were ST131 and ST648. On the other hand, ST38 and ST10 were the most predominant among food and environmental strains. Dominant overlaps in STs shared by all reservoirs/hosts (human, animal, food, and environmental) were associated with ST131 and ST648, which represent 11.4% (n = 19) and 5.4% (n = 9) of all strains, respectively.
Phylogenomics and evolutionary dynamics of human and nonhuman E. coli clones. The SNP maximum likelihood tree of human and nonhuman E. coli genomes, constructed using RAxML-NG (100 bts), is presented in Fig. 6. Strikingly, E. coli lineages from distinct sources (human, animal, environmental and food) and geographic regions, in Brazil, were closely related in the phylogenetic tree, being grouped in clades. Pangenome and SNP matrix data are quoted in Fig. S1 and, Table S2 and S3. A zoom-in on the representative clades of important pandemic clones depicted in the phylogenetic tree is shown in Fig. 7. In this regard, E. coli strains belonging to the ST10, collected in the Southeast and Midwest regions of Brazil, were nested within a human, animal, and environmental clade. ST10 has been a persistent One Health clone, present in this country since at least 1989. The ST10 clade included strains carrying bla CTX-M-8 (ONE69 and ONE88), bla CTX-M-1 and mcr-1.1 (ICBEC7P and ICBEC13AM), and bla CMY-2 (ONE39) resistance genes ( Fig. 7A and Table S3).
For the ST38 clone, phylogenetic tree revealed a clade that included strains isolated from food (mussels, chicken meat, cabbage), birds of prey (vulture), and environmental (seawater) samples (Fig. 7B). ST38 has been present since at least 2016, at the Southeast For the international ST131 clade, most strains presented a broad and critical resistome that included bla CTX-M-8 , bla CTX-M-9 , bla CTX-M-15 and bla CTX-M-27 ESBL genes. These strains have been identified from 2010, highlighting the rapid dissemination of this international high-risk clone over the last decade (Fig. 7C, Table S3). A multiplicity of origins and geographic features among these strains denoted One Health implications. All strains displayed a virulome contributing to adherence, invasion, toxins production, iron uptake, and protectins/serum resistance.
The phylogenomic analyses of ST354 showed a clade with strains presenting the broadest set of virulence determinants, which have been circulating at the humanenvironmental interface from 2016 ( Fig. 7D and Table S3). From the resistome perspective, all human E. coli ST354 strains carried the mcr-1.1 gene; with ONE46, EW239 and ONE78 strains being more closely related (.99% identity).
Lineages of ST410 have been identified at the human-animal interface in North, Northeast and Southeast regions of Brazil, from 2014. Phylogenomics revealed that while ONE50, ONE51 and ONE52 human strains shared 99.9% identity (Fig. 7E and Table S3) and an identical resistome and virulome, A240 and A241 (animal hosts) and BR12-DEC (human host) strains shared .93% identity.
Phylogenomic investigation also confirmed that the high-risk ST648 clone has been another closely related pandemic lineage with One Health implications in this country ). The phylogeny was tested against 100 bootstrap replications and the image was visualized using iTOL version 5.6.1 (https://itol.embl.de). The panel on the right represents the geographic region and the resistome and virulome context. (Fig. 7F), from at least 2013, when the first bla KPC-2 -positive E. coli strain (1326F) belonging to this clone was isolated from a human host in Southern Brazil (Table S3). In the phylogenetic three, the 1326F strain was closely related to the ONE111 bla CTX-M-15 -positive E. coli strain isolated from a food sample collected in 2016, in the Southeast region. Moreover, from core genome SNPs analysis, WHO critical priority EW222 (bla CTX-M-15 ), 7469 (bla CTX-M-15 , and bla IMP-1 ) and ONE16 (bla CTX-M-15 , and bla KPC-2 ) strains isolated from environmental and human samples collected in the Northeast, Southeast and South regions, shared .99% identity (Fig. 7F and Table S3).
Finally, E. coli strains belonging to ST744 were clustered within a human-animal clade that included stains circulating in different geographical regions, since at least 2014 ( Fig. 7G and Table S3). In this regard, the first strain belonging to this lineage was isolated from a human host and carried bla KPC-2 and mcr1.1 genes, being closely related to the ONE41 (bla CTX-M-1 -and bla CMY-2 -positive) human strain identified in 2015, at the Southeast region. Interestingly, ONE27 (bla NDM-1 -positive) and ONE82 (bla CTX-M-2 -positive) strains were isolated from human and a fish samples at the South and Southeast regions, respectively; sharing 98.1% identity in the phylogenetic three (Fig. 7G).

DISCUSSION
Brazil is the largest and most populated country in South America, with a wide geographical area that comprises different ecosystems with a remarkable biodiversity (11). Unfortunately, in most metropolitan areas of this country antimicrobial resistance has been a critical issue, not restricted to hospital settings (5). In fact, in the last years, there has been growing concern that the nationwide propagation of antibiotic resistance is also associated with environmental reservoirs that are linked to anthropogenic activities such as animal husbandry, agribusiness activities and wastewater treatment. In this study, we performed a genomic-based surveillance using a One Health approach, elucidating the genomic background of E. coli strains recovered from the human-animal-food-environmental interface over the past 45 years. Our findings highlight (i) a broad antimicrobial resistome, (ii) convergence of virulence and resistance genes, (iii) successful expansion of international high-risk clones, and (iv) phylogenomic diversity with strains isolated from human and nonhuman samples clustering together.
The resistome suggests that ESBL endemicity has been associated with b-lactamases conferring resistance to cefoxitin, where CTX-M-positive E. coli have been identified in humans, vegetables, chicken meat, wild and food-producing animals, pets, Amazonian fish, and aquatic environments (12)(13)(14)(15)(16)(17)(18)(19). In our study, we found the predominance of bla CTX-M-15 , followed by bla CTX-M-2 , and bla CTX-M-8 ESBL genes. Interestingly, food samples were identified as potential reservoirs associated with bla CTX-M -type genes, whereas it is well documented that food-producing animals play an important role as a reservoir of MDR pathogens (20). In line with this, the hypothesis has been suggested that the commercial chicken meat could be a reservoir of E. coli strains coharboring bla CTX-M and colistin resistance mcr-1 genes (21).
The overuse of colistin as a growth promoter in food-producing animals and its clinical usage to treat carbapenem-resistant infections has been a red alert to global health (22)(23)(24). We found several strains from all sources (human, animal, food, and environmental) surveyed harboring mcr-type genes, including those belonging to international sequence types ST10, ST131, ST354, ST393, ST410, and ST744. Main plasmids that have driven the spread of the mcr-1 genes have belonged to IncX4, IncI2, and IncHI2 incompatibility groups (23,25). Here, the presence of IncX4 plasmid replicon gene was confirmed in 82.8% of mcr-positive strains, which reinforces the endemicity of this plasmid type within a One Health perspective.
The rapid spread of bla KPC-2 -positive carbapenem-resistant E. coli strains during the last years is another epidemiological data observed in this study, which also requires closer attention, particularly due to its association with international high-risk clones ST90, ST224, ST648 and ST744 (5, 7, 26). Although, there has been concern around bla KPC in E. coli of ST131, another globally disseminated and clinically successful clone (27), here this lineage was only associated with the presence of bla CTX-M-8 , bla CTX-M-9 , bla CTX-M-15 , bla CTX-M-27 and bla CMY-2 b-lactamase genes.
Another point that deserves attention is the rate of quinolone resistance related to chromosomal mutations and/or PMQR genes. Quinolone resistance has not been limited to health care-associated infections. In this study, approximately 37% of strains, including those from human, animal, food, and environmental sources/host presented one or more quinolone resistance genes, whereas 64% of the total had at least one mutation, confirming a worldwide trend, where the presence of quinolone-resistant E. coli strains has been found in a wide range of human and nonhuman reservoirs (5,7,(27)(28)(29)(30)(31)(32).
Our findings also reveal that the resistome of the screened E. coli strains extends to heavy metals, biocides, and pesticides, which could contribute for the development of coresistance to antibiotics and other antimicrobial agents (33). This should be considered a critical problem since biocides are agents for disinfection commonly used in domiciliary, veterinary and hospital settings (33). Common active ingredients in disinfectants are quaternary ammonium compounds (QACs), of which benzalkonium chloride is a concern because of its widespread use combined with environmental impacts (34). Disinfectants containing quaternary ammonium compounds, including benzalkonium chloride (BAC), act mainly by disturbing the integrity and function of the cell membrane that leads to cell death. Since bacterial resistance to BAC compounds is based on overexpression of efflux pumps, QAC-induced overexpression of efflux pumps can lead to: i) cross-resistance for clinically relevant antibacterial agents, including fluoroquinolones; ii) stress response facilitating mutation in the Quinolone Resistance Determining Region; iii) biofilm formation increasing the risk of transfer of mobile genetic elements carrying fluoroquinolone or QAC resistance determinants (35). Once benzalkonium chloride enters the environment, it can be lethal to aquatic organisms contributing to antimicrobial resistance. As well as for antibiotic residues, environmental entry of QACs is through wastewater effluent and sewage. Therefore, the dissemination of antibiotic resistance and qac genes among E. coli strains circulating across clinical boundaries could be a direct consequence of anthropogenic activities (36).
Convergence of resistence and virulence in E. coli lineages is another issue that has worried the medical community (37). Indeed, virulence potential of ESBL-and/or AmpC-b-lactamase-producing E. coli strains from healthy food animals from Europe and South America has been recently documented (38)(39). In our study, virulence factors involved in adherence and biofilm (e.g., afaA, faeC, fimC, focX, hra, iha, pap, and/or sfa genes) were remarkably shared, mainly by the international high-risk clones identified. Previous analysis on the distribution of adherence/biofilm genes in E. coli lineages has highlighted the presence of these genes in strains belonging to ST131 and ST648 clones (40).
Regarding protectins/serum resistance genes, most E. coli strains (59.9%) were positive for traT, omp, and iss genes. The iss gene plays a role on microbial survival to serum due to the phagocytosis protection factor; omp (outer membrane protein) gene is associated with evasion of the body's defense allowing intracellular survival, whereas traT gene is involved in the inhibition of complement system activity (41).
Siderophores (high-affinity iron chelating molecules) have been received special attention as essential virulence factors of bacteria, acting as a toxin and/or modulator of the immune system of the host (42)(43)(44)(45). Strains that can produce more siderophores are considered highly virulent (45). In this study, more than 40% E. coli strains also carried siderophore (aerobactin, yersiniabactin, salmochelin, and enterobactin)-encoding genes. Additionally, we predicted clinically relevant virulence genes encoding hemolysin, enteroaggregative heat-stable toxin, secreted autotransporter toxin, cytotoxic necrotizing factor 1, serine protease autotransporter, and vacuolating autotransporter toxin (41,46). E. coli have been grouped into eight major phylogroups (i.e., A, B1, B2, C, D, E, F and G) based on genetic analyses. While A and B1 phylogroups are widely associated with commensal lineages, B2 has been a predominat pathogenic lineage. Additionally, strains from phylogroup B1 have been found to persist longer and to tolerate lower temperatures than the remaining phylogroups, resulting in higher ability to colonize aquatic ecosystems than A and B2 phylogroups, which have been linked to an animal-associated lifestyle (47). Many recognized virulence factors are enriched among lineages belonging to specific phylogroups of E. coli that are evolutionarily quite old, globally distributed, and commonly inhabiting healthy human guts, such as phylogroup B2. In this study, Clermont phylotyping analysis confirmed that strains carrying the colibactin gene clbB belong to phylogroup B2, as previously reported (48). On the other hand, some lineages of E. coli B2 carried the broadest virulome. In this way, it has been suggested that strains B2 have been commonly responsible for extraintestinal infections and possess numerous virulence genes (17,37,41). The occurrence of clinically relevant resistance genes (e.g., bla CTX-M and bla KPC ) was observed in strains belonging to A and B1 phylogroups, which presented a smaller set of virulence genes, as previously reported (37,41,49,50).
The establishment of high-risk clones overlapping human and nonhuman sources offers the opportunity to successfully disseminate resistance genes that makes difficult the control this situation (9,10,60). Under this perspective, other factors that drive antibiotic resistance deserve attention, such as overuse of antibiotics in multiple sectors to treat infections or for agriculture purposes, poor implementation of hygiene and sanitization actions, and the environmental contamination by inappropriate waste treatment (3). These aspects provide a snapshot of the need to enforce a comprehensive One Health strategy that considers the connection of human, animal, and environmental health in order to preserve the effectiveness of currently available antibiotics (3).
In general, over the last decades, there has been a significant increase in antibiotic prescribing and consumption leading bacterial resistance to the point of becoming a global priority (66). During COVID-19 pandemic use of antibiotics and biocide has grown even more rapidly (66,67), which can strongly favor the selection and dissemination of WHO critical priority resistant pathogens at the human-animal interface (68). In fact, an increase in the incidence of antimicrobial resistance has been documented during COVID-19 pandemic (69,70), with a rapid increase in multidrug-resistant organisms, including extended-spectrum b-lactamase (ESBL)-producing and/or carbapenem-resistant NDM-producing Enterobacterales, A. baumannii, and methicillin-resistant Staphylococcus aureus (MRSA). The cause has been multifactorial and is particularly related to high rates of antimicrobial agent utilization in COVID-19 patients with a relatively low rate of co-or secondary infection (71). In this regard, household transmission of carbapenemase-producing organisms has been linked to hospital discharge (72), becoming a potential way for transmission of such bacteria to humans and companion animals during the pandemic period (73); since COVID-19 pandemic has increased relationships and interactions between family members, and between humans and pets, supported by widespread social distancing and isolation measures. In Brazil, these facts are particularly worrying, since the prevalence of critical priority ESBL and carbapenemase producers has been higher than reported in other countries (74,75). Additionally, in October 2021 an epidemiological alert on emergence and increase of new combinations of carbapenemases in Enterobacterales, triggered by the increased use of broad-spectrum antibiotics in patients with COVID-19, in Latin America and the Caribbean was announced by the Pan American Health Organization (PHAO), emphasizing the importance of appropriate microbiological diagnosis and the effective and articulated implementation of infection prevention and control programs (76).
As a limitation of this study, since short-read sequencing technology was used, it could influence the number of contigs and lead to the missing of resistance, virulence, and plasmid replicon genes in the genomes analyses. In addition, draft genomes may overestimate the number of accessory genes in Roary. We also did not perform a temporal analysis since it was not possible to retrieve strains from every year. Furthermore, we recognized that our study showed an uneven geographic distribution of the analyzed strains because most of them are from the Southeastern region.
In conclusion, we have investigated the genomic background of critical priority E. coli strains circulating at the human-animal-environment interface in Brazil, documenting the successful spread of international high-risk clones with a broad antimicrobial resistome, with bla CTX-M ESBL and mcr-1 genes being endemic, and the rapid and worrisome expansion of bla KPC-2 and bla NDM-1 carbapenemase genes. We also found clinically relevant virulomes among E. coli strains, which together with the broad resistomes could contribute to the pathogenicity. This genetic background of E. coli must be a key factor that has contributed to adaptation and dissemination of critical priority clones in human and nonhuman hosts, which is a serious problem that needs urgent actions that includes both stricter surveillance and more judicious use of antimicrobials, under a One Health perspective. Finally, these observations alert us to the worsening of the antimicrobial resistance problem in Brazil, after the COVID-19 pandemic.

MATERIALS AND METHODS
Sequenced isolates and metadata. During a multicentric surveillance study (One Health Brazilian Resistance [OneBR] project, http://www.onehealthbr.com/) conducted to characterize the burden of antimicrobial resistance associated with WHO critical priority pathogens in Brazil, we collected 104 E. coli strains isolated from different sources (humans, food-producing animals, companion animals, wildlife, polluted environments, and food), over a 10-year period (2010 to 2019). The isolates were obtained from all geographic regions (North, Northeast, Midwest, Southeast and South). From this collection, 54 E. coli strains displaying resistance to broad-spectrum cephalosporins, carbapenems, colistin and/or fluoroquinolones, were selected for WGS and investigated in this study.
For comparative genomic analysis, we used 36 E. coli genomes previously published and deposited in our platform OneBR (http://www.onehealthbr.com/). Additionally, we included 77 publicly available Brazilian E. coli genomes deposited at the GenBank database (https://www.ncbi.nlm.nih.gov/genbank/). These genomes were chosen based on the resistance profile of each strain, and complete epidemiological information (i.e., source, origin, city, and date of collection). In total, 167 genomes of E. coli strains circulating at the human-animal-environment interface, in Brazil, were analyzed in this study. Detailed information is presented in Table S1.
The genomes were clustered in four groups according to their origin (human, animal, environmental, and food). The human group included samples from blood, bone fragment, catheter, feces, urine, fistula fluid, ileum mucosa, prosthesis, rectal swab, and other body fluids (e.g., foot, pelvic, or tracheal secretion). Genomes of E. coli isolated from pets (dog and cat), livestock animals (cattle, chicken/poultry, horse, and turkey), and wild animals (anteater, elephant, fish, ocelot, owl, penguin, and vulture) have been obtained from rectal/cloacal sample or different sites of infection. The food group included strains from meat (chicken), crustacean (shrimp), seafood (mussels and oysters), and vegetables (spinach and cabbage); whereas E. coli strains recovered from freshwater, seawater, soil, drainage, and directional residue, were included in the environmental group.
To assess the phylogenomic relationship of E. coli strains under a One Health approach, 167 draft genomes from human and nonhumans strains, collected in several regions of Brazil, were analyzed. The analysis of the E. coli pangenome was accessed by Roary pipeline version 3.13.0 (90) using annotated draft assemblies in GFF3 format produced by Prokka (91). The gene presence/absence output from Roary was used to construct the pangenome gene presence-absence matrix that was visualized using Roary.plots.py program. The core gene alignment was used to construct the phylogenetic structure of E. coli strains. SNPs (single nucleotide polymorphism) were extracted from the core gene alignment using SNP-sites (92) and a maximum likelihood tree based on SNP alignment was constructed using RAxML-NG version 0.9.0 under the generalized time-reversible model with gamma-distributed rate heterogeneity (93). The phylogeny was tested against 100 bootstrap replications and the resulting tree was visualized with iTOL version 5.6.1 (94). The percent identity was automatically calculated from the core gene alignment by SNP-sites and a table generated from the Geneious software. Metadata of E. coli strains have been made freely publicly available for interactive exploration through Microreact (https:// microreact.org/project/noM6Wi46mnpdYSzENnWmKX) (95). In addition to the general analysis of the strains (n = 167), the most represented pandemic clones (ST10, ST38, ST131, ST354, ST410, ST648, and ST744) were also analyzed separately.
Data availability. The data sets analyzed during the current study are available in the GenBank (https://www.ncbi.nlm.nih.gov/genbank) and corresponding access numbers are described in supplementary information (Table S1). Additionally, data are also available at the OneBR platform (http:// onehealthbr.com).