INTRODUCTION

Bacterial urinary tract infections (UTIs) are a common cause of disease in dogs. The most frequently isolated bacteria in both humans and dogs are Escherichia coli, followed by Staphylococcus, Proteus, and Klebsiella species (Byron 2019; ^{Gómez Beltrán et al. 2020}; ^{Sierra González and Arango Uribe 2017}). UTIs and other bacterial infections are frequently treated with ß-lactam antibiotics, however resistance to this antibiotic group has increased in the last decade. The main mechanism of resistance to ß-lactam in Enterobacteriaceae is the production of extended-spectrum ß-lactamases (ESBL). These enzymes confer resistance to penicillins, cephalosporins (including oximinocephalosporins), and aztreonam. On the contrary, ESBL are unable to hydrolyze cefamycins (cefoxitin and cefotetan) and carbapenems (^{Paterson and Bonomo 2005}).

Most ESBLs can be classified into three groups: (i) TEM and (ii) SHV-derived from gene mutations in the classic narrow-spectrum ß-lactamases TEM-1, TEM-2, and SHV-1-, and (iii) CTX-M group which predominates over the other ESBL-types globally. CTX-M enzymes possess preferential hydrolytic activity over cefotaxime and have been classified into five groups (CTX-M-1, CTX-M-2, CTX-M-8, CTX-M-9, and CTX-M-25), based on amino acid identities (^{Bonnet 2004}). ESBL-type enzymes are usually encoded in plasmids, facilitating their rapid dissemination between different species through horizontal transfer by conjugation (^{Morejón García 2013}). In addition, the same plasmid can carry genes that encode resistance to quinolones, aminoglycosides, tetracyclines, and trimethoprim/sulfamethoxazole, which contribute to antibiotic multi-resistance (^{Tafur et al. 2011}). Due to their impact on morbidity, mortality, and frequent resistance and limited therapeutic options, carbapenem-resistant and ESBL-producing Enterobacteriaceae are included in the priority pathogens list on antibiotic resistance of the World Health Organization (^{Tacconelli et al. 2017}).

Close contact between humans (owners, vets, animal caretakers, etc.) and pets, as well as direct contact with animal secretions during the performance of routine actions related to their care (feeding, grooming, cleaning living spaces, petting), implies a potential transmission risk of pathogens between both species, as well as the transfer of resistance genes. Moreover, since it is very common that dogs urinate in public areas, the contamination of the soil and water sources could represent a problem for the community and the environmental health.

In Colombia, in the clinical practice in companion animals the use of antibiotics that are also used in human medicine is ent (^{Cabrera García 2010}), mainly ß-lactams (^{Astaiza Martínez et al. 2016}; ^{Gutiérrez et al. 2002}). In addition, it is common in daily practice the recognition of pathologies based on presumptive diagnoses and the scarce use of diagnostic aids, such as culture with antibiogram, resulting in the establishment of empirical antibiotic therapies is common in the daily practice (^{Gómez Beltrán et al. 2021}; ^{Sánchez et al. 2015}).

Estimating the use of antimicrobials in companion animal species in Colombia is also difficult. Usually, its magnitude is assessed in reference to the sales reported by the manufacturing laboratories, without considering the dispensing of authorized antimicrobials for human use, as well as injectable antimicrobials designed for food-producing animals that are also used in companion ones (^{Cabrera García 2010}).

The widespread and excessive use of antibiotics, mainly third-generation cephalosporins have usually been attributed as a selection pressure that drives ESBL evolution (^{Paterson and Bonomo 2005}). Thus, inappropriate use of antibiotics in animals can contribute to the spread of resistant bacteria that cause complicated infections in both animals and humans, and that are associated with higher morbidity, mortality, and treatment costs in humane medicine (^{Alós 2015}). Therefore, understanding the phenomenon of antimicrobial resistance from a veterinary medicine perspective is important to determine its impact on public health.

Although urinary infections in dogs are very common, as well as the use of antibiotics in companion animals is high and contaminated urine can be discharged into the environment, information about the phenomenon of antimicrobial resistance in companion animals in Colombia is limited. Therefore, this study aimed to characterize isolates of ESBLs producing E. coli and K. pneumoniae, obtained from urine samples of dogs from the metropolitan area of Valle del Aburrá (Antioquia, Colombia).

MATERIALS AND METHODS

DISCUSSION

In this study, a general frequency of 11,6% (22/190) of E. coli and K. pneumo-niae producing ESBL was found. When compared to similar studies where both bacteria were analyzed in dog urine, this frequency was slightly higher than the one found in Brazil of 9,6% (2/62) (^{Melo et al. 2018}) but lower than that reported in Switzerland of 25% (39/156) (^{Zogg et al. 2018}). However, if we consider the difference in the length of the study of those researches (two and four years respectively) compared with this one-year study, our findings may represent a major concern by showing a greater number of diseased dogs carrying E. coli and K. pneumoniae producing ESBL in a shorter period of time.

In the same way, the frequency of isolates ESBL producing E. coli and K. pneumoniae observed is also remarkably higher than those obtained from diseased companion animals between 2008 and 2010 in a European surveillance program (^{Bogaerts et al. 2015}) where only five E. coli and one K. pneumoniae were isolated. Such a difference can probably be explained by the fact that all chronically diseased animals, as well as those who had been recently treated with antibiotics, were excluded in that study, while the majority of the animals in our study had a history of hospitalization and/or use of antibiotics in the previous year.

Other studies have also evaluated the presence ofESBL-Ec in dogs suffering from UTI worldwide. Among the isolates of E. coli in this study, the observed frequency of 6,6% (11/22) ofESBL producers is higher but not so far than that found previously in the United States (4%), Switzerland (3,4%) and China (3,8%) ( ^{Huber et al. 2013}; ^{Li et al. 2017}; ^{O'Keefe et al. 2010}), although data from those studies were single-institution based and may not be entirely comparable.

As it has been mentioned, research - particularly in animals- has been more focused on studying ESBLs produced by E. coli, yet the predominance of ESBL production in this study was higher in K. pneumoniae isolates than in E. coli. Some studies in Japan, Italy, Portugal, Germany, and other European countries have also reported a high number of multidrug resistance and third-generation cephalosporin resistant K. pneumoniae strains among companion animals (^{Donati et al. 2014}; E^{wers et al. 2012}; ^{Harada et al. 2012}; Marques et al. 2018). The latest suggests that ESBL-Kp is frequent in clinical samples from companion animals and that it is necessary to continue monitoring resistance mechanisms in this species.

The predominant gene class detected in our study was CTX-M. The CTX-M family of enzymes is grouped based on the similarities in amino acid sequences; for example, within the group CTX-M-1 are the enzymes CTX-M-1, CTX-M-3, CTX-M-10, CTX-M-12, CTX-M-15, and CTX-M-32 (^{Galvis and Moreno 2019}). Several studies suggest that CTX-M-1 is the most frequent genotype, not only in companion animals, but also in food-producing animals, food, wild animals, and veterinarians that work with dogs and cats (^{Abbas et al. 2019}; ^{Baede et al. 2015}; ^{Cormier et al. 2019}; ^{Darwich et al. 2019}; E^{wers et al., 2012}; ^{Irrgang et al. 2018}).

In Colombia, a few studies had analyzed the presence of bla _{CTX M} genes. From the poultry chain, ^{Castellanos et al. (2017}) identified Escherichia coli carrying bla _{CTX M} ,blaCTX-M-8 ^and bla_CTX-M-₁₅ then in 2018, bla _{CTX M 2} ´group was found to be the most prevalent ESBL gene among Salmonella enterica isolates (^{Castellanos et al. 2018}). In addition, CTX-M group 1 was detected in enterobacteriaceae isolated from bulk tank milk samples (^{Vásquez Jaramillo et al. 2017}).

In addition, the predominant circulation of the CTX-M group 1 in dogs described here is consistent with the epidemiology of CTX-M in E. coli and K. pneumoniae causing UTI in humans in Colombia, both at hospital and community settings (^{Blanco et al. 2016}; ^{Galvis and Moreno 2019}; ^{Leal et al. 2013}; ^{Martínez et al. 2012}; ^{Rada et al. 2019}). They all reported the presence of the CTX-M group 1, with bla_CTX-M-1, bla_CTX-M-3, ba_CTX-M-15 and bla_{CTX-M 32} being the mainly found enzymes. We also detected TEM and SHV-like genes in some isolates, but it is important to note that these genes could also encode narrow-spectrum beta-lactamases and confirmation of ESBL-type should be performed by gene sequencing.

The detection of CTX-M type enzymes in dog's uropathogens increases the risk of clinical failure because of the use of inappropriate empirical treatments and therefore results in complicated urinary infections. In our study, the highest frequency of resistance was found to enrofloxacin and ciprofloxacin, while the lowest resistance was observed against nitrofurantoin among the E. coli isolates, in addition to resistance to penicillins and cephalosporins.

Regarding K. pneumoniae isolates, it was observed that the vast majority were multidrug-resistant, highlighting the resistance to nitrofurantoin, trimetho-prim/sulfamethoxazole, enrofloxacin, and ciprofloxacin. These findings agreed with a previous retrospective study carried out in a veterinary diagnostic laboratory in Medellin, Colombia, which evaluated the susceptibility to antimicrobials of bacterial isolates from dogs and reported that in general, the level of resistance was high (20-50%) to at least 6 or more antimicrobials. Specifically for urinary tract infections caused by E. coli or en-terobacteriaceae (Klebsiella spp., Proteus spp.), amikacin and florfenicol were the only drugs that demonstrated 100% in vitro efficacy (^{Gómez Beltrán et al. 2020}).

The resistance of ESBL-producing strains to other classes of antimicrobial agents can be explained by the fact that plasmids harboring bla genes frequently carry genes that confer resistance to fluoroquinolones, aminoglycosides, and SXT among other antibiotics (Coque et al. 2018; ^{Wieler et al. 2011} ), as already reported in E. coli and K. pneumoniae isolates from urine of dogs with UTIs (^{Huber et al. 2013}; ^{Zogg et al. 2018}). This suggests that veterinarians are encouraged to improve antibiotic prescriptions, based on susceptibility testing in order to reduce the selection pressure on the pathogens.

Surveillance studies implementing a One Health approach are required to assess the level of genetic relatedness among resistant bacteria from animal and human origins as well as to help elucidate the complex epidemiology of transmission and the role of companion animals in the spread of multi-resistant bacteria. Additionally, these findings contribute to the design of policies related to the use of antibiotics in both human and veterinary medicine, as well as other factors such as hygiene procedures, that may be associated with the emergence of the antimicrobial resistance phenomenon in the country.

This study has some limitations. First, it was difficult to obtain accurate data from the patients' clinical record, such as the type of UTI and details on previous antimicrobial treatment. Because veterinarians request urine culture and antibiogram more frequently from complicated than uncomplicated cases our results likely correspond to those of complicated UTI patients. Second, the selection of antibiotics to be evaluated in each patient was defined by the treating veterinarian rather to being standardized, which makes the comparison of susceptibility results in all the isolates difficult. Finally, it will be important to include in future studies additional molecular techniques to investigate the predominant ESBL-positive clones causing UTI in dogs.

CONCLUSIONS

An overall frequency of 11,6% (22/190) ESBL-producing E. coli and K. pneumoniae isolates in urine samples from dogs with UTI. CTX-M group 1 was the main ESBL detected, and a multidrug resistance profile was identified in Kpneumoniae. The timely identification of ESBL-producing strains of enterobacteriaceae is essential for evaluating the development of antimicrobial resistance and for establishing successful antibiotic therapies in the small-species daily practice, as well as to strengthen the implementation of infection control measures and the rational administration of antibiotics for the protection of public health.