Antimicrobial resistance and associated risk factors in Escherichia coli isolated from Peruvian dogs: A focus on extended-spectrum β-lactamases and colistin

Background and Aim: Established antimicrobial resistance (AMR) surveillance in companion animals is lacking, particularly in low-middle-income countries. The aim of this study was to analyze AMR and its risk factors in Escherichia coli isolated from dogs at two veterinary centers in Lima (Peru). Materials and Methods: Ninety dogs were included in the study. Antimicrobial susceptibility was established by disk diffusion, whereas microdilution was used to determine colistin susceptibility. Mechanisms related to extended-spectrum β-lactamases (ESBL) and colistin resistance were determined by polymerase chain reaction. Clonal relationships of colistin-resistant isolates were assessed by XbaI-pulsed-field gel electrophoresis. Results: Thirty-five E. coli strains were isolated. High levels of resistance to ampicillin (57.1%), nalidixic acid (54.3%), tetracycline (48.6%), and azithromycin (25.7%) were detected. Cephalosporin resistance levels were ≥20% and those for colistin were 14.3%. Twelve (34.2%) isolates were ESBL producers; of these, six blaCTX-M-55 (50.0%), 2 (16.6%) blaCTX-M-15, and 2 (16.6%) blaCTX-M-8-like genes were found. The five colistin-resistant isolates were clonally unrelated, with four of them presenting amino acid codon substitutions in the mgrB gene (V8A) or mutations in the mgrB promoter (a12g, g98t, and c89t). Furthermore, dog age, <6 years (p = 0.027) and raw diet (p = 0.054) were associated with resistance to a greater number of antibiotic families. Conclusion: Despite small number of samples included, the study found that dogs studied were carriers of multidrug-resistant E. coli, including last-resort antimicrobials, representing a public health problem due to close contact between dogs and humans. This issue suggests the need for larger studies addressed to design strategies to prevent the spread of resistant micro-organisms in small animal clinics and domestic settings,


Introduction
Antimicrobial resistance (AMR) is a significant public health threat worldwide, highlighting the health crisis linked to the worldwide increase in infections caused by pathogens resistant to last-line antibiotics [1].In this sense, companion animals are often exposed to antibiotics due to the treatment of infections or prophylaxis [2] or through the consumption of food products [3], increasing selective pressure and leading to the development of multidrug resistance in their commensal bacteria [4].Thus, pets can be reservoirs and hosts for resistant bacteria, raising concerns about the risks to both human and animal health [5].
Escherichia coli is a common member of the intestinal microbiota of humans and companion animals and is involved in many intestinal and extraintestinal infections [6,7].Because of its ubiquity and high facility for acquiring AMR genes, it is a good marker of antibiotic resistance [6][7][8].In recent years, the emergence and rapid dissemination of E. coli producers of extended-spectrum β-lactamases (ESBL) [9,10], carbapenemases [11,12], and even the harboring of genes of resistance to last-resort antibiotics such as colistin [13] have become a paramount concern for animal health.
While a great variety of ESBLs have been described in the literature, the most widely distributed Copyright: Ventura, et al.Open Access.This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.The Creative Commons Public Domain Dedication waiver (http:// creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Available at www.veterinaryworld.org/Vol.17/April-2024/17.pdfgenes conferring resistance to broad-spectrum cephalosporins are the bla CTX-M genes.Among these, bla C- TX-M-15 is the most common ESBL in most regions of the world, whereas others, such as bla CTX-M-27 and bla CTX-M-55 , are particularly relevant in specific areas [14].Furthermore, regarding carbapenemases, reports of bla NDM , bla KPC , and bla OXA-48 have increased among enterobacterial in recent years, with this increase also being reported in veterinary microorganisms [15,16].Finally, regarding colistin resistance mechanisms, plasmid-mediated colistin resistance mcr genes and mutations in chromosomal genes have been frequently reported [17][18][19].
In small animals, antibiotic resistance and its associated factors have important implications for veterinary and public health [20].However, scientific literature on this critical problem is scarce.In Peru, studies on the frequency of antibiotic-resistant microorganisms and their risk factors in companion animals are limited [8].Thus, the aim of this study was to determine AMR levels and its risk factors in dogs from veterinary clinics in Lima, Peru.

Ethical approval and Informed consent
This study was approved by the Institutional Ethics Committee in Research with Animals and Biodiversity of the Universidad Científica del Sur (452-2020-POS99).In addition, all the owners of the canine companion animals provided written informed consent for sampling their animals.

Study period and location
A cross-sectional study was conducted from January to July 2021 on clinically healthy dogs (i.e.Dogs attending for bathing or vaccination services) at two veterinary clinics in districts with different socioeconomic characteristics (Surco and Villa El Salvador) in Lima, Peru.Surco is one of the largest high-income residential districts of Lima (population ~400,000), whereas Villa El Salvador is one of the largest low-income residential districts of Lima (population ~2,400,000).All dogs that had consumed antibiotics within the 3 months before sampling were excluded from the study.

Sampling
Convenience sampling was performed.Rectal swab samples were placed in Cary Blair preservation and maintained at 4-8°C during transportation to the Universidad Científica del Sur laboratories according to standardized protocols [5].

Epidemiologic information
The owners provided epidemiological information on the canine companion animals.The canine data collected included sex, age (classified into two groups: <6 years and ≥6 years), diet (raw meat diet/ prepared dry diet/mixed/home diet), owner's history of antimicrobial use, and contact with other companion animals (yes/no).

E. coli isolation
Samples were homogenized and seeded on MacConkey plates (Oxoid, Basingstoke, UK) or in McConkey agar containing disks of ceftazidime (30 µg) at one end of the plate and incubated at 37°C for 24 h.Then, isolated colonies compatible with E. coli growing within the ceftazidime-treated halos or without antibiotic pressure were selected and confirmed by polymerase chain reaction (PCR) amplifying the E. coli-specific uidA gene [21].

Antibiotic resistance genes
In ESBL-positive phenotype isolates, the presence of bla CTX-M (including bla CTX-M1 , bla CTX-M2 , bla C- TX-M8 , and bla CTX-M9 ) and bla SHV genes was determined by PCR [25].The amplified products were recovered and purified using the EZNA Gel Extraction Kit (Omega Bio Tek, Norcross, GA) and sequenced (Macrogen, Seoul, South Korea) to determine the specific gene.Meanwhile, the genes involved in carbapenem resistance were amplified by PCR: bla VIM , bla IMP , bla KPC , bla NDM , and bla OXA-48 [26].In the case of colistin-resistant isolates, the presence of transferable mcr-1 to mcr-5 genes was analyzed by PCR [27].In addition, the presence of point mutations in chromosomal genes, such as mgrB, was assessed by PCR [28] and subsequent sequencing (Macrogen, Seoul, South Korea).

Phylogenetic and clonal analyses
The phylogenetic groups of the colistin-resistant isolates were established according to the Clermont extended scheme [29].Clonal relationships among the colistin-resistant isolates were determined by pulsed-field gel electrophoresis (PFGE) of the total DNA digested with 40 U of XbaI restriction enzyme, as previously described by Gautom [30].DNA fragments were resolved in 1 % agarose gels using the CHEF-DRIII system (Bio-Rad Laboratories Inc, Hercules, USA).The PFGE conditions used were at 6 V cm -2 at 14°C and with a pulse time ranging from 1 s to 30 s for 19 h.Normalization and processing of the PFGE images and construction of the dendrogram were performed using GelJ v2.0 software (https://sourceforge. net/projects/gelj/) [31].A UPGMA (unweighted pair group method with arithmetic mean) tree was constructed using Dice similarity indices, complete linkage, 1% optimization, and 1.3 % position tolerance.Isolates were considered clonally related with a Dice similarity index ≥80% [32].

Statistical analysis
A descriptive analysis was performed by summarizing the qualitative variables according to their absolute and relative frequencies and the quantitative variables according to their mean and standard deviation.During this analysis, intermediate-resistant phenotypes were considered to be resistant.In addition, AMR-associated factors were assessed by modeling the number of resistant antimicrobial families using a Poisson regression model.ESBL and multidrug resistance were also evaluated separately as secondary outcomes of interest.Finally, we estimated the prevalence ratio (PR) as the magnitude of the association of interest.During these analyses, diet categories included dry, raw meat, mixed, and homemade; however, for the regression analysis, we converted this variable to a dichotomous variable (raw vs. no raw diet).All statistical analyses were performed using the STATA MP v14.0 statistical package (Stata Corp LP, College Station, Texas), and each estimate was reported with their respective 95% confidence intervals (CI).
Regarding the owner's history of antimicrobial use, 44.4% of the animal owners had consumed antibiotics within the last month.

E. coli isolation and identification
Of the 90 fecal analyzed samples, 35 (38.9%) were E. coli-positive (of these, eight strains were isolated with the presence of CAZ and 27 from the medium without antibiotic disk).The characteristics of the dogs are shown in Table-1.When comparing the characteristics of the animals with E. coli-positive versus E. coli-negative fecal samples, there was a significantly frequent history of owners who had consumed antibiotics in the last month among the samples positive for E. coli (65.7% vs. 30.9%,p = 0.001) (Table-1).

ESBL characterization
Twelve ESBL-producing E. coli isolates were detected (13.3% of the tested samples and 34.3% of the E. coli-positive samples).The bla CTX-M-1G gene was amplified in 8 (66.7%) individuals, whereas no bla SHV was detected.The variant bla CTX-M-55 was detected in six isolates, and bla CTX-M-15 was identified in the remaining two isolates.In addition, in two isolates bla CTX-M-8G was amplified, and three isolates were positive for universal bla CTX-M , but no variant was determined.Finally, none of the ESBL-encoding genes sought were amplified in the two remaining isolates.

Carbapenemase characterization
All isolates presented an intermediate phenotype to carbapenems.No positive results were detected among the carbapenemase genes assessed.

Colistin characterization
Five isolates were colistin-resistant, and mutations in the mgrB gene were found in four isolates.Three of these isolates presented alterations in the promoter region, including two possessing the base changes a-12g and g-98t and the remaining isolate showing the change c-89t.Meanwhile, one isolate presented an amino acid codon substitution V8A (Figure -2).In addition, no mcr genes were detected.

Phylogeny and clonal relationship between colistin-resistant isolates
The five colistin-resistant E. coli isolates were clonally unrelated.Three isolates belonged to group B1, and the other two belonged to groups D and B2, respectively (Figure -2).

Factors associated with AMR
Regression analysis showed that animals aged 6 years presented E. coli isolates with resistance to a lower number of antimicrobial families (PR = 0.41; 95% CI: 0.19-0.90;p = 0.027).Although marginally significant (p = 0.054), dogs consuming raw food presented E. coli resistant to a higher number of antimicrobial families.The study did not have sufficient power to perform a multivariate Poisson regression model; likewise, no factor associated with ESBL or multidrug resistance was identified (Table -2).

Discussion
The present results demonstrate that companion animals, such as dogs, have antimicrobial-resistant E. coli that may be transmitted between animals and humans due to direct or indirect contact.Significantly high levels of antibiotic resistance were observed in canine E. coli isolates, with over half of the isolates being resistant to ampicillin and over half resistant to nalidixic acid and tetracycline.Moreover, approximately 40% of the isolates were cephalosporine-resistant and levofloxacin.Resistance levels to these antibiotics are important because these drugs are mainly used in human and veterinary medicine.Furthermore, this study highlights the presence of E. coli resistant to carbapenems and polymyxins, which is of concern because these antibiotic families are considered last resort antimicrobial treatment in human infections.In addition, we observed an association between dog age and the type of diet and the presence of E. coli resistance to a greater number of antibiotic families.

Antibiotic pattern
No. of isolates AMC, CFS, ATM, TET, Q, AMP, AZM 1 CFS, ATM, SXT, TET, Q, AMP, AZM 1 CFS, ATM, SXT, TET, GEN, Q, AMP 1 AMP = Ampicillin, AZM = Azithromycin, Q = Quinolones, CFS = Cephalosporines, COL = Colistin, TET = Tetracycline, SXT = Sulfamethoxazole/ trimethoprim, ATM = Aztreonam, GEN = Gentamicin Previous studies characterizing resistance levels and mechanisms in E. coli recovered from companion animals in Peru are scarce [8].However, studies conducted in Cajamarca (Northern Peru) [33] showed resistance levels of 61% for streptomycin and 51% for ampicillin, the antibiotics most commonly used in daily clinical practice in the region.The levels of antibiotic resistance in the present study were similar to those reported from Minas Gerais, a populous state in Brazil [3], although they did not report colistin resistance.
Our results are consistent with the increase in ESBL-producing E. coli reported in dogs worldwide, mainly related to the high use of β-lactams in clinical practice in veterinary medicine [34,35].The percentage of samples possessing ESBL-producing E. coli was 13.3%, similar to other countries in the region, such as Ecuador [36].In the comparison among isolates, we found 34.3% of ESBL producers, which is a lower percentage than that reported in Thailand [10].
Regarding CTX-M variants, various ESBLs were found in the study, with bla CTX-M-55 being the most frequent.In addition, CTX-M-15 and CTX-M-8G were found.It is of note that CTX-M-55 has been previously reported in Peru in studies including samples of diverse origins (healthy human and clinical samples, marketed food, drinking water, and animals including wild mammals, poultry, livestock, and companion animals) [37][38][39].Moreover, in ESBL studies in this region, CTX-M-55 has been the most frequently reported [3].It should also be noted that CTX-M-15, the dominant variant worldwide [40], was not very frequently identified in this study.Similarly, other studies in the area have described the same finding, suggesting that CTX-M-15 is not the predominant ESBL in Peru [37,41].CTX-M-8G, a rare ESBL group, has been previously detected in dog feces in Northern Peru [38].It is worth mentioning that some of the ESBL-positive strains did not amplify CTX-M, suggesting that other mechanisms may be involved, such as TEM or other less frequent ESBLs, a scenario that has already been reported in other cases in the country [25].
Colistin is used as a last resort for human clinical use against multiresistant bacteria, and it has not been marketed for veterinary use in small animals in the country since 2019 [42].However, the results of our study showed that 14.3% of E. coli isolates were resistant to colistin, even in E. coli isolated from animals considered to be clinically healthy.In previous studies, mcr-1 carriage was associated with colistin resistance in E. coli from dogs and their owners [35]; however, resistance to colistin in the present study was probably related to chromosomal mutations, which were reported as a major mechanism in other Enterobacteria such as Klebsiella pneumoniae [43].In this sense, three of five colistin-resistant isolates possessed mutations in the promoter region of mgrB, and another isolate possessed the amino acid substitution V8A.While no data about the effect of mutations in the promoter region of mgrB have been found in the literature, alterations in the promoter region of genes have the potential to affect (increasing, decreasing, or avoiding) the final expression levels of subsequently encoded peptides [44].However, while the role of these alterations in promoter regions in final colistin MIC levels cannot be ruled out, they are outside the PhoP box and do not affect the -10 region or transcription start sites.Nevertheless, further studies are necessary to elucidate the effects of these alterations [45].Meanwhile, the amino acid substitution V8A has been described in both colistin-resistant and colistin-susceptible isolates, suggesting a polymorphism or little effect on final colistin MIC levels [46].Furthermore, PFGE and phylogenetic analysis demonstrated that these strains were clonally  unrelated.Thus, exposure to colistin, which has been used in the avian sector for many years as a possible source of exposure through the food chain, can be considered [47].
When exploring the factors associated with AMR, we observed that diet appears to be a critical factor in resistance to certain families of antimicrobials.Although a dry diet based on different types of proteins, carbohydrates, grains, and other components that undergo dehydration was reported to be consumed by most of the dogs in our study, we found an association between consumption of a raw diet and resistance to a higher number of antimicrobial families.Raw food as the basis of a dog diet has gained interest recently because of studies or cases of cardiac pathologies related to different types of preservatives or ingredients used in dry foods [48].However, the consumption of raw meat has been associated with a high risk of antibiotic-resistant E. coli and other food-borne bacteria, such as Salmonella [49].It has also been associated with MDR E. coli in dogs with a raw meat-based diet in Brazil [3], emphasizing the importance of surveillance and control of this type of food, the consumption of which has increased in recent years.Moreover, this study found younger age to be a risk factor for AMR, despite age being related to longer exposure time to antimicrobials [50].This fact could be related to the use of antimicrobials for treating diarrhea associated with dietary changes in puppies.
Close contact between pets, their owners, and the environment has led to the sharing of bacteria, including resistant bacteria [51].Thus, the spread of bacteria from humans to animals and vice versa poses a risk for treating non-human infections and may compromise the treatment of infections in veterinary medicine.Therefore, the high consumption of antimicrobials by animal owners is noteworthy in our study, with more than 40% of the owners having consumed antibiotics in the previous month.

Limitation
One of the limitations of this study was the lack of data related to the clinical history of the dogs from which the samples were collected, thereby precluding the analysis of the association between antibiotic resistance and underlying pathologies.Another limitation was that while the study power was sufficient to recognize the strongest associated factors, it was not sufficient to adjust a multivariate regression model.

Conclusion
Our study demonstrates that dogs can be an important source of antibiotic resistance genes, including ESBLs and those involved in resistance to antibiotics of last resort, such as colistin.This is likely related to the high veterinary and dietary use of antibiotics, despite the prohibition of their use.The results of this study provide a better understanding of the presence of antibiotic-resistant bacteria in companion animals.They also highlight the need for strategies to prevent the spread of resistant strains in small animal clinics and domestic settings.

Table - 1
: Epidemiological characteristics of the sampled dogs (total) and of those Escherichia coli positive.

Table - S1
: Antibiotic resistance patterns observed in the Escherichia coli isolates from fecal samples of dogs.

Table - 2
: Factors associated with antimicrobial family resistance.