Identification of multidrug-resistant enterobacteriaceae in fecal samples from infants residing in Talara, Piura, Peru

ABSTRACT Fecal colonization by antimicrobial-resistant bacteria in infants is a potential risk for future antibiotic therapy. We aimed to determine the sociodemographic characteristics and frequency of infants who were fecal carriers of ciprofloxacin-resistant enterobacteriaceae (FCCRE) and their associated resistance genes. We analyzed fecal samples from 41 infants from the district of Talara, Piura, Peru in 2019. The presence of 3 quinolone resistance genes was evaluated: aac(6’)-Ib-cr, qnrB and oqxA as well as of 2 beta-lactamase genes: bla CTX-M,bla PER-2. We found that 68% of infants were FCCRE, Escherichia coli (83.3%) was the most frequent bacteria. The genotypic analysis detected: oqxA (41.1%), qnrB (26.7%), aac(6’)-Ib-cr (20%) and the bla CTX-M gene (93.3%) of the isolates with beta-lactamases. The high frequency of FCCRE alerts us of the potential risk of this antibiotic family becoming less useful over time.


INTRODUCTION
Antimicrobial resistance (AMR) is a serious public health problem.The World Health Organization (WHO) has estimated that by 2050 all antibiotics will be ineffective.The rapid, early, and widespread dissemination of AMR genes is considered the main reason for this (1) .
Colonization with drug-resistant bacteria is a major health risk due to the potential transfer of resistance genes to pathogenic bacteria and their easy dissemination between individuals.
The intestinal microbiota has been reported to be a major source of urinary tract, respiratory, and bloodstream infection (2) .
Quinolones are a broad family of antibiotics whose resistance increased rapidly in the 1990s (3) .Mutations in the quinolone resistance determinant region (QRDR) are the main mechanism that confers a high level of resistance.Plasmid-mediated quinolone resistance markers (PMQR), such as Qnr proteins, the AAC(6')-Ib-cr enzyme and the OqxAB pump, exert a low level of resistance.However, they play an important role in the selection of chromosomal mutants in QRDR (3) .ß-lactams are the main family of antibiotics, and extended-spectrum ß-lactamases (ESBL) represent the main mechanism of resistance to these antibiotics (4) .CTX-M enzymes, predomhttps://doi.org/10.17843/rpmesp.2022.394.11870Motivation for the study: infants, due to restricted nutritional intake, limited motor capacity and low antibiotic exposure, are a population protected from multidrug-resistant bacteria.
Main findings: in this study, 68% of the infants were colonized by quinolone-resistant bacteria, mainly by E. coli (83.3%).The gene mainly associated with quinolone resistance was oqxA (41.4%).Also, half of the isolates were ESBL producers; and resistance was caused by the bla CTX-M gene in 93.3% of the isolates.
Implications: these findings alert us about the high presence of antimicrobial-resistant bacteria in a vulnerable population; which shows the potential risk of this antibiotic family becoming less useful over time.
In addition to being related to the use and abuse of antimicrobials, AMR is also related, although not directly, to the socioeconomic status of the family and the economic and healthcare development of a country (5) .The study of populations potentially protected from colonization by AMR bacteria offers an opportunity to gather necessary evidence on the microbial quality of the environment.Infants are such a population, due to their limited motor capacity, low exposure to antimicrobials and limited nutritional variety (6) .
This study aimed to determine the frequency and sociodemographic characteristics of infants who are fecal carriers of ciprofloxacin-resistant enterobacteriaceae (FCCRE) and the presence of PMQR and ESBL genes in the district of Talara, Piura region, Peru.

THE STUDY
A descriptive cross-sectional study was carried out.We enrolled 41 infants between 3 and 12 months of age from the district of Talara, Piura region, Peru.Of these, 28 infants were enrolled from the Talara Baja area based on information provided by the healthcare center and 13 infants from the Talara Alta area by identifying infants in the community who met the selection criteria.
Samples were collected between September and December 2019.The sample size was not calculated for this study.Sample selection was non-probabilistic by convenience under the following selection criteria: having informed consent from the mother for the child's participation, having provided a stool sample, having a frequency of breastfeeding greater than four times per day, and having been born vaginally.We excluded infants whose mothers had consumed antibiotics 15 days prior to stool sample collection.

Search and identification of quinolone-resistant enterobacteria
Stool samples were transported in Cary-Blair medium to the Microbiology and Immunology Laboratory (LMI) of the University of Piura in Lima for processing.Stool samples were placed on MacConkey agar supplemented with 2 mg/L ciprofloxacin for presumptive isolation of quinolone-resistant enterobacteria.
Identification and antimicrobial sensitivity profiling was carried out with the Vitek 2 compact automated system (Biomeriux, France); the interpretation process followed the Clinical and Laboratory Standards Institute (CLSI) recommendations (7) .Phenotypic detection of ESBL was carried out by the double-disk method according to CLSI recommendations (7) .

Detection of quinolone and beta-lactam resistance genes.
Bacterial DNA was extracted using DNA Purification kit GeneJetGenomic (ThermoScientific), following the manufacturer's recommendations.The presence of 3 PMQR genes: aac(6')-Ib-cr, qnrB, and oqxA was determined using primers previously described in the literature (8) .In addition, the presence of two ESBL-associated resistance genes bla CTX-M and bla PER-2 was studied among the isolates with evidence of ESBL (8) .We searched for CTX-M groups 1, 2 and 9 in the isolates in which the bla CTX-M gene was detected (8) .

Clonal relationship
The clonal relationship between ESBL-producing isolates was determined using the ERIC-PCR technique, as described in the literature (9) .The Past software version 4.0 was used to integrate, in a dendogram, the results through the UPGMA algorithm.Isolates showing more than 90% identity were considered clonally related.

Survey
A structured questionnaire was administered to 37 mothers, 11 with non-FCCRE children and 26 with PFRC children (the questionnaire was not administered to 2 mothers with FCCRE children and 2 mothers with ciprofloxacin-sensitive children).The questionnaire was based on the questions in Chapter 1 "Household and population characteristics", module "Characteristics of dwellings and households" of the Demographic and Family Health Survey (ENDES) -2014 (10) .In addition, information was collected on the age and sex of infants.

Statistical analysis
SPSS Statistics for Windows, Version 25.0.(Armonk, NY: IBM Corp) was used for the statistical analysis.Qualitative variables were described using frequency graphs; and quantitative variables were described using tables.Fisher's exact test was applied to evaluate differences between groups carrying quinolone-resistant and quinolone-sensitive enterobacteria and the Student's t-test was used to compare means.P values <0.05 were considered significant.

Ethical considerations
Ethical approval for the main study "Association of postnatal stress with the quality of the breast milk microbiome and its relationship with iron deficiency anemia in infants" (Project code: PI2008-UDEP) was granted by the Institutional Research Ethics Committee of the Faculty of Human Medicine of the San Martín de Porras University.Informed consent and assent were obtained from the mothers of each infant prior to enrollment.

Demographic and economic analysis
No statistical difference was found between the infants who were fecal carriers of ciprofloxacin-sensitive and resistant enterobacteria with regard to sex.However, a statistical difference was found according to age (p = 0.007).Table 1 shows the results for other household variables.Finally, no statistical difference was found between the economic status and any of the evaluated variables.

DISCUSSION
Intestinal colonization with AMR bacteria reflects the extent of the spread of bacterial resistance to antimicrobials (11) .In this study, more than half of the infants were FCCRE (68% -2016 (12) .In Peru, Pons et al. reported 12.1% of FCCRE in 222 healthy children aged 2 to 12 months, from peri-urban localities of Lima between 2006 -2007 (13) .
There are several phenomena that may explain the high frequency of FCCRE.In Peru, the consumption of antibiotics without prescription is frequent.Rojas-Adrianzen et al. in 2016, identified that 53.4% of apothecary/pharmacy users incur in this practice (14) .In addition, environmental contamination with AMR bacteria, mainly due to overcrowding and inadequate excreta management, which are important sources of dissemination of bacterial resistance in developing countries (15) .
Regarding AMR in other antibiotic families, the high levels of resistance to AMP and SXT stand out with 93.3% and 66.7%, respectively.A study by Kalter et al., on 145 healthy children between 3 and 12 months of age, reported AMP and SXT resistance levels of 60% and 57.7%, respectively (22) .
Regarding PMQR-associated genes, our results are similar to those reported by Pons et al., who evaluated the frequency of fecal carriers of quinolone-resistant E. coli in children younger than 12 months and found aac(6')-Ib-cr and qnrB frequencies of 20% and 6%, respectively (17) .Zhao et al, in a 2018 analysis of 736 healthy children aged 3 to 6 years, reported that 8.8% and 1.8% of the 113 ciprofloxacin-resistant isolates carried the aac(6')-Ib-cr and qnrB genes, respectively (18) .In our study, the oqxA gene was identified in 41.1% of the isolates.The oqxA gene was first reported in Latin America in 2017 by Saba et al. in one isolate within 101 cephalosporin-resistant Enterobacteriaceae of clinical origin (8) .
Fifty percent of the isolates were ESBL producers, which suggests a potential association of resistance transmission between these two antibiotic families, as evidenced in a previous study (19) ..This result is consistent with our findings regarding the predominance of CTX-M-group 1.
On the other hand, ten clonal groups were found in ES-BL-producing E. coli.Three isolates (15A, 62A, 49A) showed a strong clonal relationship in one of them, even though isolate 15A did not carry the bla CTX-M gene.This finding suggests to us that, despite the close phylogenetic relationship, the acquisition of resistance mechanisms has diverse origins (4) .Furthermore, the high heterogeneity of ESBL-producing E. coli reflects a strong influence of the familial environment for the acquisition of these bacteria.This high degree of phylogenetic diversity has been reported by other studies and shows, due to the degree of dissemination, the public health risk that this resistance mechanism represents (20) .
The economic analysis of the households showed a homogeneous distribution among the groups of infants, which leads us to think of possible conditioning factors that were not evaluated.Likewise, recent studies have shown the relevance of the transmission of AMR bacteria through breast milk to the infant's intestine, which could be an important factor in the colonization process (11) .On the other hand, we found that older infants had a higher frequency of being FC-CRE, which is related to a higher degree of interaction with the environment, as has been reported by other studies (13) .
One of the limitations of this study is the limited sample size which may not reflect the overall situation of the studied population.On the other hand, although this study evaluated some genes associated with PMQR, resistance to https://doi.org/10.17843/rpmesp.2022.394.11870quinolones mainly results from mutations in the QRDR (8) .
Therefore, it would be important to analyze the other PM-QR-associated genes that were not evaluated in this research.

Figure 2 .
Figure 2. Distribution of ESBL and PMQR type resistance genes according to phenotype and bacterial genus., : Presence; :absence.

Finally, the studied
population showed a high frequency of fecal commensal bacteria resistant to quinolones, partly associated by ESBL, bla CTX-M type.This is worrisome because it reflects the extent of the dissemination of AMR bacteria and their possible role in limiting the therapeutic use of antimicrobials in this population.Therefore, these findings alert us to the high presence of AMR bacteria in a population with low direct exposure to antimicrobials, which shows the potential risk of this antibiotic family becoming less useful over time.

Table 1 .
Sociodemographic analysis between infants who are fecal carriers of ciprofloxacin-resistant enterobacteriaceae (FCCRE) and non-FCCRE infants.Dendrogram based on Euclidean genetic distance using the unpaired method of analysis at the arithmetic mean (UPGMA), constructed by presence-absence analysis of consensus intergenic regions by PCR (ERIC-PCR), for ESBL E. coli isolates.
), mainly due to E. coli (83.3%).The oqxA gene was the most frequent (41.4%) among the PMQR genes.In addition, half of the isolates were ESBL producers, and 93.3% of these were carriers of the bla CTX-M gene.https://doi.org/10.17843/rpmesp.2022.394.11870*Thequestionnaire was not administered to 2 mothers with FCCRE infants and 2 mothers with non-FCCRE infants, a Fisher's exact test with bilateral contrast, b Student's t-test.SD=standard deviation https://doi.org/10.17843/rpmesp.2022.394.11870Figure 3.In contrast to the high frequency of FCCRE found in this research, Purohit et al. reported 12.2% of FCCRE in a sample with children from rural India during the period 2014