An overview of carbapenem-resistant organisms from food-producing animals, seafood, aquaculture, companion animals, and wildlife

Carbapenem resistance (CR) is a major global health concern. CR is a growing challenge in clinical settings due to its rapid dissemination and low treatment options. The characterization of its molecular mechanisms and epidemiology are highly studied. Nevertheless, little is known about the spread of CR in food-producing animals, seafood, aquaculture, wildlife, their environment, or the health risks associated with CR in humans. In this review, we discuss the detection of carbapenem-resistant organisms and their mechanisms of action in pigs, cattle, poultry, seafood products, companion animals, and wildlife. We also pointed out the One Health approach as a strategy to attempt the emergency and dispersion of carbapenem-resistance in this sector and to determine the role of carbapenem-producing bacteria in animals among human public health risk. A higher occurrence of carbapenem enzymes in poultry and swine has been previously reported. Studies related to poultry have highlighted P. mirabilis, E. coli, and K. pneumoniae as NDM-5- and NDM-1-producing bacteria, which lead to carbapenem resistance. OXA-181, IMP-27, and VIM-1 have also been detected in pigs. Carbapenem resistance is rare in cattle. However, OXA- and NDM-producing bacteria, mainly E. coli and A. baumannii, are cattle's leading causes of carbapenem resistance. A high prevalence of carbapenem enzymes has been reported in wildlife and companion animals, suggesting their role in the cross-species transmission of carbapenem-resistant genes. Antibiotic-resistant organisms in aquatic environments should be considered because they may act as reservoirs for carbapenem-resistant genes. It is urgent to implement the One Health approach worldwide to make an effort to contain the dissemination of carbapenem resistance.

/fvets. . have a structure defined by a carbapenem coupled with a β-lactam ring. In addition, these antibiotics contain a carbon instead of a sulfone in the fourth position of the thiazolidine moiety β-lactam ring, which confers protection against most β-lactamases (1). The widespread use of these antibiotics has increased to a worldwide emergence of carbapenem-resistant organisms (CROs), which constitute a critical growing public health threat, mainly in hospital settings, as their prescription has escalated in recent years and used for treating life-threatening infections. Carbapenemresistant Enterobacteriaceae (CRE) [i.e., Carbapenem-resistant Klebsiella pneumoniae (CR-Kp), Carbapenem-resistant Escherichia coli (CREc), Enterobacter spp., Serratia spp., and Proteus spp.] are some of the most critical CROs because they are associated with infections that lead to high mortality and have the potential to spread carbapenem resistance via mobile genetic elements (2). In addition, non-fermenting bacteria such as carbapenemresistant Acinetobacter baumannii (CRAB) and carbapenemresistant Pseudomonas aeruginosa (CRPA) have also emerged as critical CROs (3)(4)(5).
The four significant carbapenem mechanisms of resistance include the presence of β-lactamase enzymes called carbapenemases, which hydrolyze carbapenem antibiotics encoded on chromosomal or plasmid genes, the synergistic effect of other β-lactamases with bacterial cell membrane permeability due to alterations or mutations in porins, the low affinity of penicillinbinding proteins (PBPs) in different species, and the increased efflux pumps (6,7).
Class D serine oxacillinases (OXAs) have been commonly detected worldwide. They have hydrolytic activity against βlactams, high activity against penicillin, and weak activity against extended-spectrum cephalosporins and carbapenems. OXA-23 and OXA-48 are variants widely dispersed globally. OXA-23 is almost strictly restricted to Acinetobacter spp., wily OXA-48 was found among K. pneumoniae and Enterobacterales (20).
Carbapenems such as imipenem, meropenem, ertapenem, doripenem, biapenem, faropenem, and panipenem have been approved for use in human clinical settings (6). However, carbapenems are not licensed in livestock or veterinary fields; therefore, carbapenem resistance is not commonly tested in animals. However, extended-spectrum beta-lactamases (ESBL), including ceftiofur, cefquinome, cefpodoxime, cefoperazone, and cefovecin, are commonly found in this sector. Additionally, ceftiofur, a third-generation cephalosporin, is the main cephalosporin used in veterinary fields and has been approved for treating bacterial infections in food-producing animals (i.e., pneumonia, arthritis, septicemia, meningitis, metritis, and polyserositis) (25). For this reason, these antibiotics could provide selection pressure that favors the expression of carbapenem-resistant (CR) strains.
Indeed, the author's statement in the Scientific Opinion on carbapenem resistance in food-animal ecosystems that "diagnostic isolates of veterinary origin classified as microbiologically resistant to third-and fourth-generation cephalosporins based on epidemiological cutoff values, should be subjected to phenotypic testing for carbapenem resistance and carbapenemase production and subsequent molecular identification and characterization of carbapenemase production genes because they favor the emergence of carbapenem-resistant isolates" (26).
Many studies have reported CROs in livestock, seafood, companion animals, wildlife, and their environments (7,(27)(28)(29)(30). As animals have been identified as a relevant source of multidrug-resistant (MDR) bacteria, they can serve as reservoirs for carbapenem-resistant bacteria, and foodborne routes and emission in the environment through the excreta and the subsequent exposure of humans via the environment serve as transmission pathways for carbapenem-resistance genes from animals to humans and vice versa. Furthermore, their incidence may be underestimated because there is usually no surveillance, which dismisses the potential risks to human health. This review aimed to summarize the occurrence and molecular mechanisms of resistance in carbapenem-resistant organisms in food-producing animals, seafood, companion animals, and wildlife to address the importance of antimicrobial resistance surveillance in these sectors, resistance dissemination to the environment and humans, and potential public health risks.
. Carbapenem-resistant from seafood and aquaculture Carbapenem-resistance genes have been identified in isolates from aquatic environments. Generally, resistance to carbapenems in bacteria from aquatic systems such as Vibrio and Shewanella spp., LMB-1 (Linz metallo-β-lactamase) (13) as well as Enterobacterales, is mainly mediated by the production of carbapenemases encoded by chromosomal genes or by plasmids. However, carbapenemases are variable, with Class B enzymes and the enzymes described only on aquatic species such as VMB-1 from Vibrio alginolyticus (17,31) and VMB-2 (32) from Vibrio diabolicus on shrimps; as well as the Class A, VCC-1 from non-toxigenic Vibrio cholerae on shrimp (8).
In 2017, in Canada, Brouwer et al. isolated the E. cloacae complex from shrimp (Litopenaeus vannamei) originating in India. The isolated had a ST previously described in companion animals in Japan, ST813, and was positive for the bla IMI−2 gene, which is located in a plasmid p3443-IMI2, which is closely related to IncFII plasmids and pIMI-6, which was described in an E. cloacae complex clinical isolate from Canada and carries the carbapenemase bla IMI−6 (33). The same strain displayed the plasmid p3442-FLC-1 that carries the gene encoding a novel class A carbapenemase FLC-1 with close sequence similarity to bla FRI−1, previously described in imipenem-resistant E. cloacae recovered from a clinical patient in France (33). In 2013, OXA-23producing A. baumannii, on fish Pagellus acarne harvested in the Mediterranean Sea in Algeria, was reported. The isolate belonged to the widespread sequence type 2 (ST2)/international clone II and harbored aminoglycoside-modifying enzymes [aac(6 ′ )-Ib and aac(3 ′ )-I genes] as well as the naturally occurring  (31), and bla VMB−2 identified on a plasmid-borne composite transposon ISShfr9-ISCR1-bla VMB−2 -bla CARB−12 -aadA1-ISShfr9, where ISShfr9 was found to be disseminated in multidrug-resistant (MDR) pathogens (32); as well as producing the enzymes VMB-1 encoded by a gene bla VMB−1 located in an integron bearing, highly transmissible IncC type plasmid, namely pVB1796 (15), and VAM-1 located in a conjugative plasmid, namely, pC1579 (17). Other studies in Korea have reported carbapenem-resistant Vibrio spp., isolates from shrimp (41), cockles (42), or hard-shell mussels harboring bla OXA genes (43). In Europe, in 2018 and 2017, Italy reported V. cholerae resistant to meropenem (44) and V. vulnificus isolated from shellfish resistant to imipenem and meropenem (45). KPC-3-producing E. coli in mussels (Mytilus galloprovincialis) and OXA-23-producing A. baumannii ST2 isolates from mussels and oysters (Crassostrea gigas) were also reported in Tunisia in 2016 (46,47). The bla KPC−3 gene was identified on an ∼180 kb IncFII plasmid carrying Tn4401d transposon and belonged to the ST167 phylogroup A of the ST10 complex. Interestingly, the authors mention that the predominance of bla KPC−3 in Portugal was also associated with the spread of an IncF plasmid carrying Tn4401d. The ST10 complex was reported previously in a hospital from the US to spread bla KPC genes. In 2018, France reported the isolation of NDM-1-producing V. parahaemolyticus ST864 from a shelled shrimp tail imported from Vietnam, which harbored the epidemic plasmid IncA/C (48). In South America and Ecuador in 2015, Vibrio spp., resistant to imipenem, was isolated from shrimp from seawater (49). In 2020, E. coli, Enterobacter cloacae complex, and K. pneumoniae were found in tilapia fish from Egyptian fish farms carrying bla KPC , bla OXA−48 , and bla NDM (50). In 2018, in Taiwan, carbapenem-resistant Shewanella algae were isolated from small abalone (Haliotis diversicolor) harboring genes encoding OXA-55 and multiple antibiotic-resistance genes including dfrA3 (trimethoprim resistance), tet (35) (tetracycline resistance), and qnrA3 (quinolone resistance); and the pmrCAB operon, which has been shown to mediate resistance to colistin (51). In Taiwan in 2019, carbapenem-resistant S. algae carrying bla OXA−55 with multiple genes encoding efflux pumps was detected in Asian hard clam (Meretrix lusoria) (52). In 2020 in Italy, OXA-55-like producing S. algae was found (53). All the previous information demonstrated the presence of CROs in seafood and aquaculture. Even when bacteria in aquatic environments are mainly non-pathogenic, their occurrence highlights the relevance of the food production chain in the global spread of antibiotic-resistance genes. Moreover, it is crucial to consider the seafood market, where countries can consume a specific product imported from a separate region by another country, which could have different regulations for antimicrobial resistance surveillance in food. Canada is one example of this since shrimp consumed by Canadians are imported from Asian countries. However, carbapenemase genes have been detected in these products. Examples include the isolation of Enterobacter cloacae or Enterobacter aerogenes harboring bla IMI−1 , bla IMI−2 , or bla NDM−1 genes in retail seafood (39, 54).
In addition, it is remarkably the high amount of antibiotics used and their multiple classes in aquaculture for prophylactic purposes or metaphylactic treatment. Quinolones, tetracyclines, amphenicols, and sulfonamides are the most commonly used classes (55); however, they also include aminoglycosides, antimycobacterial (rifampin), beta-lactams (aminopenicillins and cephalosporins), and polymyxins. This factor could promote selective pressure for the emergence of antimicrobial resistance and the selection of multidrug-resistant organisms among seafood animals. Indeed, global antimicrobial consumption is estimated to reach 13,600 tons by 2030 (55). Additionally, antibiotic residues are difficult to eliminate by water treatment plants and can be discharged into water flows, thus being a source of antimicrobial resistance genes.
Moreover, fish do not effectively metabolize antibiotics; thus, the active substance passes into the environment in the feces. Indeed, it has been suggested that ∼70-80% of the antibiotics applied in aquaculture are dispersed into water systems (37), which might provide a selection and enrichment mechanism for resistant bacteria (22). Studies reporting the occurrence of carbapenem-resistant genes worldwide in seafood and aquaculture are summarized in Table 2.
. Occurrence of carbapenem-producing bacteria in terrestrial food-producing animals Intensive farming has frequently been associated with the excessive use of antimicrobials and drug-resistant microorganisms isolated from food-producing animals that can be transmitted to humans via direct contact with animals or ingestion of derived food products (9). By 2030, global antimicrobial use from human, terrestrial, and aquatic food-producing animal sectors will reach 236,757 tons annually, with an estimated proportion of terrestrial food-producing animal use of 174,549 tons, representing 73.7% of the global consumption of antimicrobials (55). This intensive use of antibiotics creates selective pressure for the emergence of antimicrobial resistance among farmers and the environment. In addition, antibiotics continue to be used in livestock production as prophylaxis and animal growth support (9), which may lead to the emergence of antimicrobial resistance.
The transmission of antimicrobial resistance between foodproducing animals and humans can occur via the food chain, by consuming food products contaminated with antimicrobial resistance genes or antimicrobial-resistant bacteria, through direct contact between humans and animals, or through shared environmental sources such as contaminated water (56,57). In addition, resistance can be transmitted to livestock from environmental sources (i.e., hospital sewage, wastewater treatment plants contaminating water and soil, surface water flow, and wildlife) and biological vectors, such as flies and wild birds (21, 58). Moreover, livestock growth promotion antibiotics may interact with the animal gut microbiota and introduce increased variation in antimicrobial resistance genes (ARGs) in the gut (59), thus increasing their dispersion. Studies reporting the occurrence of carbapenem-resistant genes worldwide in terrestrial foodproducing animals are summarized in Table 3.

. . Carbapenem resistance in swine
In pigs, carbapenem resistance has been observed in microorganisms from different bacterial species, including E. coli, Salmonella, P. aeruginosa, and A. baumannii. In 2011, VIM-1-producing E. coli ST88 and Salmonella Infantis harboring bla VIM−1 IncHI2 plasmids were reported in Germany (60,61). ST88 was also previously identified among chickens, cattle, and humans in Germany. Moreover, class 1 integron harbored by an IncH12 plasmid was found in human strains (60). In addition, during the sampling period in the same year (2011), 35 isolates were positive for bla VIM−1 , indicating that carbapenemaseproducing bacteria may persist in livestock farms. Another study reported VIM-1-producing Salmonella Infantis in Germany in 2017 (59). The encoded genes bla VIM−1 and bla NDM−5 have reported resistance to third-generation cephalosporins used in animal husbandry. Moreover, Salmonella Infantis is one of the leading causes of human salmonellosis in Europe and a zoonotic pathogen commonly transferred via contaminated food products (59). In 2017, Pulss et al. (30) reported a porcine E. coli isolated carrying OXA-181 carbapenemase and the coexistence of mcr-1 (mobilized colistin resistance gene) and acquired carbapenemase gene bla OXA−48−like on isolates originated from Italy farms. bla OXA−181 gene was located on al IncX3 plasmid (pEcIHIT31346-OXA-181), which presented high nucleotide similarity >99% to previously published plasmids from human sources (plasmid pOXA181_14828 of an E. coli isolated from a human patient in China) and also carried qnrS1 (plasmid-media quinolone resistance gene), thus providing evidence of the possible link between human-and animal-derived carbapenem resistance. In 2015, bla IMP−27 was detected in Proteus mirabilis, Morganella morganii, Providencia rettgeri, Proteus vulgaris, Enterobacter cancerogenus, Citrobacter braakii, Enterobacter cloacae, Citrobacter spp., Citrobacter farmeri, Citrobacter koseri, and Klebsiella oxytoca in the United States, within an IncQ1 plasmid recovered from the nursery and farrowing barns of a swine production system (57). In 2013 in China, A. baumannii harboring bla NDM−1 genes isolated from lung samples of pigs with pneumonia and sepsis were identified (62). Meropenem-resistant P. aeruginosa strains carrying bla OXA−486 , bla OXA−396 , bla OXA−50 , and bla PAO were also found in Italy in 2018, as were meropenem-resistant isolates of Pseudomonas oryzihabitans and P. aeruginosa (9). Three other isolates of P. aeruginosa carrying bla PAO , bla OXA−50 , bla OXA−486 , and bla OXA−488 were detected in animals reared on different farms (85). Interestingly, two isolates of P. aeruginosa ST938 carrying bla PAO and bla OXA−396 and the resistance genes to aminoglycosides [aph(3′)-IIb], fosfomycin (fosA4), and chloramphenicol (catB7) were detected, one in a pig and another one in 83-year-old patients. However, no epidemiological links were demonstrable between the animal and the patient. Other sequence types found were ST274, ST782, and ST885. The presence of bla OXA−50 is concerning because this variant confers a decreased susceptibility to ampicillin, ticarcillin, and meropenem. In addition, the OXA-50 family also comprises bla OXA−396 , bla OXA−486 , and bla OXA−488 genes (9). bla OXA48 -like contained no plasmid, and bla OXA−181 -carrying IncX plasmid has also been reported in E. coli isolated from Italian fattening pigs (64). In the study, the authors recovered samples from fattening pigs, cattle, and workers from slaughterhouses. Twenty-four isolates were positive for bla OXA−181 and one for bla OXA−48 . The isolates presented high ST diversity within ST5229 with higher prevalence. Different plasmid replicons were present in the isolates, with IncX1 and IncX3, and IncF types being the most represented. OXA-48-producing isolates did not contain any plasmid replicon. Furthermore, the authors detected an OXA-181-producing E. coli belonging to ST410 isolated in two fecal samples from fattening pigs, described as a high-risk clone associated with bla OXA−181 in human patients. Moreover, in China .

. . Carbapenem resistance in poultry
Among broiler farms, an increased prevalence of CRE has been shown, mainly in E. coli, K. pneumoniae, and P. mirabilis, harboring bla NDM genes, with bla NDM−1 and bla NDM−5 being the most predominant in chicken farm environments. Specifically, farming presented a higher prevalence of CRO among the studies, mainly attributed to the heavy use of antimicrobials on farms, transportation activities, and inadequate farm disinfection and management. Similarly, bla NDM genes are usually carried by the IncX3 plasmid, which is clinically significant because it contributes to disseminating various bla NDM variant genes (86).
In 2010, carbapenem-resistant isolates were detected in eight chicken farms, six duck farms, and one pig slaughterhouse in China. One of these isolates, Acinetobacter lwoffii, was identified as positive for bla NDM−1 , particularly on a 270 kb plasmid (65). In 2021, NDM-producing P. mirabilis was reported in broiler chickens (68). The isolate harbors a plasmid named pSNYG35, a pPrY2001-like plasmid that shares high nucleotide identity with pHFK418-NDM and an NDM-1-encoding plasmid from clinical P. mirabilis. Recently, Su et al. (87) reported isolation rates of 3.57% for carbapenem-resistant E. coli, 10% for carbapenem-resistant P. mirabilis, and 3.03% for carbapenem-resistant K. pneumoniae in six broiler fattening farms in China. Among carbapenem-resistant isolates, six E. coli carried class I integron, one carried class II integron, four P. mirabilis carried class I or II integrons, and one K. pneumoniae carried class 1 integron. All of these isolates harbor bla NDM and bla OXA genes. In 2016 in Egypt, carbapenemproducing K. pneumoniae (CR-Kp) in broiler poultry farming was reported. The authors found that 42% of the isolates from poultry samples carried bla NDM (11 isolates carried bla KPC , bla OXA−48 , and bla NDM ; four isolates carried bla KPC , bla NDM or bla OXA−48 and one isolate carried bla NDM alone) (70). Interestingly, the authors collected 49 fecal samples from workers and veterinarians working in the poultry farm; 56% of the samples were CR-Kppositive, with all strains carrying the three carbapenemase genes bla KPC , bla OXA−48 , and bla NDM , and 5% of them displayed all the carbapenemase-encoding genes at the same time. Furthermore, the prevalence was higher in farm workers (67%) compared to veterinarians (33%), indicating that transmission could be facilitated by close contact between broilers and humans since the workers are in continuous contact with the animals and lived on the farm during the fattening program. However, the study did not compare clones or plasmids; non-genetic relationships between humans and chickens were found Morganellaceae, Alcaligenes faecalis, and Pseudomonas putida, with the variants NDM-5, NDM-1, and NDM-9 as well as a novel NDM-like-metallo-β-lactamase (NLM) within IncX3, IncA/C2, and IncFII as major bla NDM -carrying plasmid types among isolates. Moreover, they found the coexistence of mcr-1 or mcr-8 on K. pneumoniae positive for bla NDM−1 . The authors identified 14 sequence types among the E. coli isolates, with ST6751 being the most prevalent. ST6716, ST156, ST69, ST48, and ST10 were also found. STs 6751, 10, 125, and 746 were recovered from chicken and environmental samples (sewage trenches, corridor floors, drooping boards, nipple drinkers, and air). Most of the K. pneumoniae isolates were ST37, followed by ST3410 and ST726. Additionally, in China, Shi et al. (71) reported the presence of the resistance genes bla NDM , bla VIM , bla KPC , and bla OXA−like on broiler, layer, and pig farms with a significative higher relative abundance on bla OXA−like genes from 2016 to 2019. The authors detected a prevalence of 20-30% for bla KPC and bla VIM genes, respectively, and a prevalence of 75% for bla NDM , reflecting the great incidence of carbapenemase-producing genes in farming. Moreover, the study also found the coexistence of colistin resistance gene mcr-1 and bla NDM with pig and chicken farms displaying high prevalence. In 2020 in Egypt, 155 meropenem-resistant isolates were obtained from retail chicken meat, indicating that carbapenem-producing

. . Carbapenem resistance in cattle
Carbapenem-resistant bacteria are rare in cattle. However, since 2012, more studies have reported CROs in cattle with OXAand NDM-producing bacteria leading to carbapenem resistance. In 2010, nine OXA-23-producing Acinetobacter genomospecies 15TU were reported in France, with a Tn2008 as a vehicle for the spread (72). In 2016 in the United States, a novel bla OXA−497 gene was detected in A. baumannii, which is part of the OXA-51like enzyme group and displays resistance to ertapenem; however, these enzymes are naturally occurring in A. baumannii (73, 88). In 2017, bla OXA−23 harboring Acinetobacter indicus-like strains that displayed imipenem, meropenem, and doripenem resistance were isolated from nasal swabs of two calves in Germany. bla OXA−23 was localized on the chromosome and surrounded by interrupted Tn2008 transposon structures. In addition, genetic relatedness between bovine isolates and Acinetobacter indicus type strains A648 T and human clinical A. indicus isolates were found (77). In 2022, 27.7% of CRAB bacteria in Pakistan were reported to harbor bla OXA−23 and bla OXA−51 within 17 isolates carrying bla IMP and one isolate carrying bla NDM−1 . The typical sequence types found were ST642 and the international clone ST2 (83). In Egypt in 2019, carbapenem-resistant P. aeruginosa (CRPA) was reported in buffaloes and cattle with a prevalence of 60 and 59% (50 total samples) within isolates harboring bla KPC , bla OXA−48 , and bla NDM . The authors also found carbapenem-resistance genes from drinking water within 67% of prevalence and from stool human samples within 80% of prevalence. Additionally, phylogenetic analysis showed that cattle and water sequences were in one cluster and more related to each other than to human isolates (80). Similarly, in Egypt in 2014, five E. coli carrying bla OXA−48, and one E. coli carrying bla OXA−181 were reported in dairy cattle (29). In South Africa, 28-42% of carbapenem resistance was found in isolates such as E. coli, K. pneumoniae, P. mirabilis, and Salmonella spp., carrying bla KPC , bla NDM , bla OXA−23 , bla VIM , bla OXA−48 and bla GES with different prevalence (81). In Italy, in 2021, the EU harmonized antimicrobial resistance (AMR) monitoring program reported that units of fattening pigs (21/301) and bovines (4/310) were positive for OXA-48-like E. coli (n = 24 OXA-181, n = OXA-48) (64). Most recently, in Tunisia, one isolate of ESBL-producing E. coli from calves with diarrhea carrying bla OXA−48 and bla IMP were reported (84).
Antimicrobial resistance mediated through NDM enzymes is present in cattle. In 2013 in India, E. coli harboring the bla NDM−5 gene was detected in milk samples of dairy cattle suffering mastitis (74). NDM-5-producing E. coli isolates from raw milk collected in a dairy farm in Algeria and India in 2016 were again found (75, 89). bla NDM−1 gene in E. coli isolated from cattle, carried in an IncC plasmid, was reported in 2022 in Spain. The IncC plasmid also carried genes for aminoglycoside, sulphonamide, and trimethoprim resistance (82). K. pneumoniae carrying bla NDM−5 located on IncX3 plasmid was isolated from dairy cows in China in 2017 (76). The presence of IncX3 plasmid is highly relevant since it mediates the spread of genes encoding resistance to clinically relevant antibiotics. It has been reported to encode qnrB7, qnrS, bla CTX−M−3 , bla SHV−12 , bla KPC−2 , bla KPC−3 , bla NDM−1 , bla NDM−4 , bla NDM−5 , bla NDM−7 , bla NDM−13 , bla NDM−17 , and bla OXA−181 (86). The authors found that the K. pneumoniae bla NDM−5 positive belonged to five STs, within ST1661 and ST2108, which were the most prevalent. The bla NDM−5 gene was located on the ∼46 kb IncX3 plasmid. The plasmid shared a similar genetic context and was nearly identical to the human K. pneumoniae plasmid (pNDM-MGR194) previously reported in India. Among beef cattle in the United States in 2018, isolates of K. pneumoniae carrying bla KPC−2 from feces were detected in 72 samples (78). bla VIM gene located on an Incl1 plasmid of a novel sequence type (ST 297) from E. coli was well-detected among calves from India in 2019 (79).

. Occurrence of carbapenem resistance in wildlife
Antimicrobial resistance genes can colonize wild animals following contact with sewage, human waste, or animal manure (90). Human feces and manure runoff are the primary sources of AMR in wild animals, as intake water polluted with feces could directly or indirectly contaminate other animals and the surrounding environment. Moreover, AMR genes, such as carbapenemase genes, could originate from environmental bacteria, such as the OXA-48 family of enzymes, which occurs naturally in Shewanella spp., a genus that inhabits lake sediments (91,92), and OXA-23 enzymes, which are almost entirely restricted to A. baumannii and originate from the environmental species Acinetobacter radioresistens (93). Chickens have been proposed as a source of carbapenemase-producing Salmonella enterica in livestock (85).
On the other hand, rivers and water flow are also an environment from the emergence of CROs. For example, Pseudomonas fluorescens was recovered from the Seine River (Paris, France) in 2010, which expressed PF-1, a novel Ambler class A carbapenemase (109). In 2022 in Poland, 301 carbapenem-resistant Acinetobacter strains were isolated from municipal wastewater and river water (110). In 2005, carbapenem-resistant bacteria were reported on water bodies in the United States (111, 112). In 2019, carbapenem-resistant bacteria on water bodies were isolated, including Enterobacter asburiae, Aeromonas veronii, Cupriavidus gilardii, Pseudomonas, and Stenotrophomonas spp. This study found that most strains were carbapenemase producers, and all the isolates of Enterobacter asburiae carried the bla IMP−2 gene (111). Other studies have also reported the presence of carbapenemresistant strains in seawater, stormwater, and surface runoff water at Costa locations in Sydney, Australia, in 2020 (113). Therefore, water environments are an important reservoir of bacteria resistant to carbapenems and other antibiotics, including bacteria carrying intrinsic and acquired carbapenemase genes.

. Occurrence of carbapenem resistance in companion animals
Carpabenemase-producing Enterobacterales (CPE) and nonfermenting bacteria have also been reported in companion animals. As we pointed out before, carbapenems are not approved for veterinary use. The prescription is restricted to treating urinary tract infections and respiratory tract infections in dogs and cats originating from multidrug-resistant (MDR) E. coli, K. pneumoniae, and P. aeruginosa bacteria. Additionally, the treatment must be supported by a veterinarian specializing in infectious disease, and by a pharmacologist (114), even though . /fvets. . the continuous evidence of carbapenem-producing bacteria in companion animals has been increasing. Companion animals can acquire carbapenemase-producing bacteria through direct contact with colonized hosts and the through contaminated environments such as veterinary hospitals (115,116). In this regard, the human-pet bond has favored the silent transmission of carbapenem-producing bacteria to companion animals by a reverse zoonotic route called zooanthroponosis (115)(116)(117). Indeed, in Finland in 2015, identically isolates from dogs (with a long history of recurrent otitis externa without carbapenem prescription) and human family members with NDM-5-producing multidrug-resistant ST167 E. coli were reported. In addition, the same family carried an identical extended-spectrum beta-lactamase (ESBL) CTX-M-group 9 E. coli ST69, indicating interspecies transmission (118). In 2018 in Brazil, six VIM-2 carbapenemaseproducing P. aeruginosa ST233 isolates were recovered from an infected dog, its owner (with a history of hospitalization), and its domestic environment (sofa, balcony, and water cooler) (116). ST233 has been reported as an international high-risk clone associated with carbapenemase production with resistance to all antimicrobial drugs. It has generally been restricted to human hospital settings (119-121), suggesting a zooanthroponotic transmission of this clone after the patient's hospital discharge. More recently, in 2022 in Guangzhou, China, a large-scale investigation on the prevalence of bla NDM -positive E. coli isolates from companion animals and their healthcare providers in clinical veterinary settings revealed the clonal spread bla NDM -positive ST453 E. coli isolates between both species (122). In France in 2022, OXA-48-producing K. pneumoniae were isolated from companion animals (dogs, cats, horses, cattle, and birds) with 56.2% (59/105 isolates) of the isolates belonging to the human-associated MDR ST11, ST15, and ST307 lineages, suggesting that numerous humanassociated clones could infect the animal host (123).
Among carbapenemases on companion animals, NDM-5 and OXA-48-like carbapenemases are the most frequently described enzymes, with E. coli and K. pneumoniae being the main carbapenem-producing Enterobacterales, along with the nonfermenting bacteria A. baumannii (124). OXA-48 has been identified in Enterobacteriaceae from dogs and cats in different countries, such as Germany (2013) and the United States (2009-2013) (27,125), as well as in an ST38 E. coli isolated from fowl (Gallus domesticus) in 2015 in Lebanon (28). In 2012, in Belgium, two OXA-23-producing Acinetobacter spp. were detected in fecal samples from 20 hospitalized horses, both resistant to imipenem and presented resistance to tetracyclines, sulfonamides, trimethoprim, and gentamicin but were still susceptible to colistin (126).
On the other hand, the KPC enzyme has also been reported. In 2018 in Brazil, in K. pneumoniae and E. coli from dogs, the bla KPC−2 gene was found in Tn4401 transposons contained in IncN plasmids, which also carried bla CTX−M−15 , and other clinically significant resistance determinants conferring resistance to aminoglycosides (aadA5), quinolines (qnrS1), macrolides [mph(A) and erm(B)], sulfonamides (sul1), tetracycline [tet(B)], and trimethoprim (dfrA17), and point of mutation conferring quinolone resistance (127). In Brazil 2021, the KPC-2-producing K. pneumoniae belonging to the high-risk international clone ST11/CG258 in a dog with urinary tract infection carrying the IncN plasmid assigned to ST15 was reported (128). The bla KPC−4 gene was detected in 2016 in Ohio, US, in an IncHI2 plasmid in the context of the Tn4401b transposon in Enterobacter xiangfangensis isolated from a clinical dog sample with ST171, which has been responsible for major clusters of human CRE infections in the northeastern and upper-midwestern of the United States (129, 130). IMP-4 has been reported in Salmonella enterica serovar Typhimurium isolated from cats in Australia in 2016 (131). NDM-1 was isolated in 2013 in the United States from dogs and cats from E. coli that also carried bla CTX−M−15 and belonged to ST167 (132), as well as in China, with Acinetobacter species carrying bla NDM−1 and bla OXA−23 (133), and in Italy from A. radioresistens (134). Recently, in 2022 in China, five bla NDM−5 harboring E. coli were reported in dogs and cats, all of them multidrogo resistant. The bla NDM−5 gene was located on 46 kb IncX3 plasmids in the five strains. Additionally, one strais coharbored bla NDM−5 -encoding-IncX3 plasmid along with an mcr-1-IncX4 hybrid plasmid (135). OXA-48 has mainly been described in dogs, cats, and horses and mostly from infections such as urinary tract infections (UTIs) isolates from E. coli, K. pneumoniae, E. cloacae, and K. oxytoca (136,137). VIM-1 and VIM-2 were also reported in dogs infected with K. pneumoniae and P. aeruginosa in Spain and Korea in 2016 and 2018, respectively (138, 139), as well as OXA-181producing extra-intestinal pathogenic E. coli ST410 from a dog in Portugal in 2020 (140), OXA-23-mediated carbapenem-resistance A. baumannii ST2 from a cat (141), and OXA-66-producing A. baumannii isolated from cats (124, 142). For livestock production, bla NDM was detected in 73.8% (n = 45) and 62.2% (n = 28) of pig and chicken farms, respectively. Furthermore, they found that human NDM-positive E. coli isolates shared 15 (n = 111), 11 (n = 90), and 10 (n = 96) STs with those from chickens, pigs, and flies, respectively. NDM-positive isolates belonging to ST167, ST206, ST10, and ST48 were recovered from all four origins. Furthermore, the authors found that large proportions of bla NDM genes (>70%) were associated with IncX3 plasmid in both animals (pig, chicken, and fly isolates) and humans. The authors also predicted the origins of 463 NDM-positive isolates. They found that 19% (n = 24), 8.1% (n = 10), and 1.6% (n = 2) of chicken NDM-positive E. coli isolates (n = 123) were predicted to originate from humans, pigs, and flies, respectively. In contrast, 27.3% (27/99) of pig NDM-positive E. coli isolates were predicted to originate from humans. Similarly, 53.8% (n = 105) and 14.9% (n = 29) of human isolates were predicted to have originated from chickens and pigs. All fly-derived isolates (n = 46) were predicted to have originated from humans (n = 5, 10.9%), chickens (n = 22, 47.8%), and pigs (n = 19, 41.3%). These results indicated .

. Transmission between animals and humans
/fvets. . positive associations and transmission of CREc between animals and humans. Indeed, the authors hypothesize that "CREc first arose in clinical settings and was then introduced into livestock animals, which are favorable hosts for the persistence of CREc. This led to the circulation of CREc between humans and animals, either via the food chain or through environmental vectors". In 2019, Li et al. (144) sampled 12 villages in China used as pig production farms [using the household as a single surveillance unit (resident and their backyard animals, including farm and companion animals)] and two commercial pig farms near the villages. The authors collected flies, fecal samples from humans, pigs, chickens, cattle, goats, ducks, one donkey, dogs, and cats across the villages, and additional fecal samples from pigs and farm workers at the two commercial farms. They obtained 88 CREC isolates that contained the bla NDM carbapenemase gene, 17 from humans, 44 from pigs, 12 from chickens, 12 from flies, two from dogs, and one from cattle. No CREC isolates were recovered from workers of pigs at the two nearest commercial pig farms. The authors detected bla NDM−5 , bla NDM−1 , and bla NDM−9 , with most of these bla NDM -genes likely located on IncX3-type plasmids. Indeed, the bla NDM -carrying regions/plasmid (IncX3) in CRE isolates from humans exhibited >99% nucleotide sequence identity to those in isolates from backyard animals and flies. MLST showed that six human CRE-NDM-positive isolates displayed ST48, ST10, ST1114, or ST6910 shared by animal isolates. ST48 was the most prevalent and was associated with isolates from pigs, humans, chickens, and flies. Furthermore, they found that two human isolates displayed only three single-nucleotide polymorphisms (SNPs) with two pig isolates from the same village. They also reported that CREC isolates from flies have human and dog origins, while chicken isolates had a predominant origin from pigs and dogs. In addition, the single cattle-derived isolate was clustered with the chicken isolates. Therefore, many CRE isolates from humans, backyard animals, and flies originated from hosts other than those included in the study.
The studies above showed a positive association between livestock production and human CREC infections since they identified a close relationship between the genomic profile of carbapenem-resistant isolates from humans and animals. High similarities between isolates from different sources were found. However, the studies only focused on NDM-positive isolates and not on other carbapenemases-encoding genes. Further studies are needed to elucidate the link between humans and animals.
. One health approach for antimicrobial resistance Different strategies to combat antimicrobial resistance have been developed, including the One Health approach, the EU Harmonized AMR Monitoring Program conducted in Italy in 2021, the National Action Plan for Combating Antibiotic-Resistant Bacteria (CARB) by the US (2020-2025), and the implementation of antimicrobial risk assessment.
One Health approach is a term recognized in the EU in 2016 by the United Nations Political Declaration on Antimicrobial Resistance (AMR), which states that human health, animal health, and the environment are interconnected and that disease is transmitted from humans to animals and vice versa. Furthermore, the environment could be a potential source of new resistant microorganisms; therefore, AMR should be addressed in all scenarios (25,146,147).
Strictly, One Health is defined as "a collaborative, multisectoral, and trans-disciplinary approach-working at local, regional, national, and global levels-to achieve optimal health (and wellbeing) outcomes recognizing the interconnections between people, animals, plants and their shared environment" (146). Therefore, a multidisciplinary approach is required to prevent the spread and emergence of antimicrobial resistance. Antimicrobial resistance (AMR) surveillance using the One Health approach has been implemented in Europe, the UK, and the US to mitigate the crisis. However, the lack of implementation in most developing countries resulted in the underestimation of the burden of AMR on terrestrial and aquatic animals and the environment (147).
Improper management of antimicrobials, such as inadequate control of infection, use of antimicrobials as growth promoters (long-term, low-dose mass medication), prophylaxis in livestock, farmed fish in aquaculture systems, agricultural debris, environmental pollutants from sewage, pharmaceutical industry waste, manure runoff from farms, use of heavy metals, use of disinfectants, and migration of people and animals infected with resistant bacteria, facilitate the spread of resistance between humans and animals (25,148). Consequently, the One Health approach is fundamental since it is a multidisciplinary approach that tries to prevent, predict, detect, and respond to AMR (25,148).
Critical strategies for addressing AMR from the One Health perspective includes: (1) conduct a global campaign to raise awareness of antimicrobial resistance and the damage caused by the overuse and misuse of antibiotics, (2) improve hygiene measures and prevent the spread of infections (i.e., decrease missing of . /fvets. . animals from different sources, stress of transport, unsanitary or crowded conditions), (3) reduce the use of antimicrobials in agriculture and their dissemination to the environment (including third-generation cephalosporins, fluoroquinolones, colistin, tetracyclines, and macrolides; i.e., growth promoters such as colistin has been banned in Europe, Canada, Denmark, United States, and other countries) (24), (4) improve global surveillance of drug resistance in order to understand and clarify the new mechanisms of resistance acquisition and predict future threats, (5) promote new and rapid clinical diagnoses, (6) promote the development and use of vaccines and alternatives to antibiotics (i.e., phage therapy, probiotics, antibodies, lysins, among others), (7) improve the number of studies in the field, (8) generated a global innovation fund for early-stage research on new treatments, (9) promote investment in new drugs and in the improvement of existing drugs, and finally, (10) build a global coalition for real action against AMR (25,148,149). All previous studies have highlighted the urgent need to establish a One Health AMR surveillance system to understand the magnitude of the AMR problem, specifically the carbapenemresistance problem, identify trends, and determine how all scenarios are linked and establish settings to content the widespread carbapenem-resistant organisms and genes. This approach requires the integration of human healthcare, livestock, aquaculture, and the environment, as well as other variety of disciplines and fields (149). Furthermore, the role of infections caused by antimicrobialresistant organisms in wildlife may also have to be addressed, along with resistant organisms from aquatic environments, as they could possess intrinsic resistance and the possibility of being transmitted horizontally. In addition, carbapenem-resistant in companion animals has to take seriously since the human-pet bond might favor the silent transmission of clinically significant multidrug-resistant bacteria through zooanthroponosis (148).
Moreover, antimicrobial residues in fish products can persist in aquatic environments through excreta. For example, testing foodstuffs for carbapenem-resistant bacteria is not a legal requirement in any country; however, even a low prevalence of carbapenem-resistant genes has been detected in imported shrimp and salmon. In addition, studies have shown that aquaculture and terrestrial farms exhibit significant differences in drug consumption, with the aquaculture sector exhibiting the lowest. However, commensal bacterial flora can act as reservoirs of AMR genes, which may be transferred to microorganisms capable of causing human and animal diseases. Furthermore, it has been documented that animals excrete a significant percentage (75-90%) of antimicrobials without being metabolized and dispersed into the environment (145), which could be taken up by wild animals and function as a reservoir for antimicrobial resistance genes (149).

. Discussion
The reports on carbapenem-resistant organisms published from seafood and aquaculture are still low. Most of these reports must include information on potential sources or transmission between humans, animals, and their environment. Similarly, some of the microorganisms found are of clinical importance. Some examples are carbapenem-producing Vibrio alginolyticus, which causes vibriosis, wound infection, and ear infection; Vibrio cholerae, which causes cholera; Vibrio parahaemolyticus, that cause acute, self-limiting gastroenteritis (150); Shewanella algae, a potential foodborne zoonotic agent in humans that causes necrotizing fasciitis, discitis, meningitis, biliary infection, pneumonia, and endocarditis (151); and Enterobacter cloacae complex that is common in nosocomial settings and capable of producing several infections such as pneumonia, urinary tract infections, and septicemia (152).
On livestock, carbapenem resistance has been observed in microorganisms from different bacterial species, critical in human settings, and associated with significant public health concerns worldwide, including E. coli, Salmonella, P. mirabilis, P. aeruginosa, and A. baumannii. In swine and poultry settings, VIM-1-producing Salmonella Infantis was found. This bacterium is a zoonotic pathogen commonly transferred via contaminated food products that have been implicated in human salmonellosis and foodborne outbreaks associated with egg and chicken meat (153). For instance, A. baumannii is an important opportunistic pathogen for hospital-acquired infections commonly associated with multidrug resistance. The mortality rate of A. baumannii infection has been estimated to be over 50% (154). In livestock such as cattle and pigs, A. baumannii causes mastitis, pneumonia, and sepsis. In companion, animals cause urinary tract infections (155)(156)(157).
Whether the presence of OXA-23-producing A. baumannii poses a substantial public health threat is unclear, but the presence of NDM-1 producers in A. lwoffii and A. baumannii isolated from poultry, swine, and cattle, which are clinically relevant to humans, is worrying. Similarly, NDM-producing E. coli and K. pneumoniae, isolated from poultry and cattle, are worrisome since they have been reported in clinical isolates worldwide.
The carbapenem-resistance determinants in wild animals need to be better understood. Wild animals may act as potential environmental reservoirs for bacterial resistance. VIM, NDM, OXA-48, and KPC-producing K. pneumoniae are the more frequent carbapenemases reported, followed by IMP, and NDMproducing E. coli. Contaminated food and water are the main routes of transmission of carbapenem-resistance bacteria to wild animals (158). However, anthropogenic pressure plays an essential role in the emergence of resistance, particularly in this setting. Interestingly, migrating birds (i.e., gulls) have been proposed to serve as a vehicle for disseminating carbapenem-resistance genes (159).
Among food-producing animals, the link between farming practices, animal health, carbapenem-resistant organisms' development and spread to farmers, and the presence of carbapenem-resistant organisms in foodstuffs requires much more investigation. Three studies positively associated livestock production with human CREC infections (143)(144)(145). These associations are mainly based on the observation that bla NDMcarrying IncX3 plasmid isolated from humans exhibited between 75 and 99% nucleotide sequence identity to those in isolates from other sources, including chicken, pigs, and fly isolates. Moreover, one article identified a close relationship between the core-genome sequences of NDM-positive E. coli from humans and animals. The source-tracing analysis revealed indistinct boundaries between Frontiers in Veterinary Science frontiersin.org . /fvets. .
human-and animal-derived NDM-positive E. coli (143). Several studies have found genetic similarities between carbapenemproducing bacteria from animal and human sources, including the detection of a porcine E. coli isolated carrying OXA-181 carbapenemase located on al IncX3 plasmid with high nucleotide similarity (99%) to previously published plasmid from human sources (30); E. cloacae IMI-1 positive isolated from clam and human-source E. cloacae ST373 isolate harboring bla IMI−1 sharing >75% similarity (8), and the detection of varying levels of genetic similarity at discrete genetic loci between E. coli carrying KPC-2 and K. pneumoniae containing KPC-3 isolated from gulls and humans in Alaska (104).
These examples demonstrated that direct anthropozoonotic or zooanthroponotic transmission might be possible for CRE. However, to estimate the public health relevance of this transmission, more studies are needed to elucidate the problem. The addition of high-throughput technology, such as wholegenome sequencing and next-generation sequencing (NGS), has been permitted to determine the genetic relationship among CRO from different species at gene, plasmid, and strain levels. Similarly, introducing discriminant analysis of principal components (DAPCs) is helpful for tracing carbapenem-producing strains' potential origins.

. Conclusion
The data presented in this review confirm the widespread of carbapenemase-producing bacteria and encoding genes in foodproducing animals, seafood, aquaculture, companion animals, and wildlife as a cause of representing a severe problem for human and animal health.
Several studies have shown genetic similarities between human and animal carbapenem-resistance isolates, thus, demonstrating the possible cross-species transmission. Nonetheless, epidemiologic and genotypic analysis studies are needed to understand better the dynamics of antimicrobial drug resistance transmission between humans, animals, and the environment. In addition, the presence of CROs in the food chain compromises food safety and security and increases the chance of cross-border transmission of these bacteria.
One Health approach can help to implement global monitoring programs and establish antimicrobial risk assessments for the zoonotic and environmental sectors to address AMR emergencies. It is essential to identify and share best practices and policies globally. Collaboration between governments is needed to address cross-border health threats of AMR.

Author contributions
AG-B conceived and designed the idea for the manuscript. FR-C and AG-B wrote the manuscript. AG-B, FR-C, and FA-G reviewed the manuscript. All authors read and approved the final manuscript.
Funding FR-C received a research grant from CONACYT (grant number: 373176). This study was supported by the internal research project PIB19-3 of the Autonomous University of Aguascalientes.