Molecular Characterization of Extended Spectrum β-Lactamase (ESBL) and Virulence Gene-Factors in Uropathogenic Escherichia coli (UPEC) in Children in Duhok City, Kurdistan Region, Iraq

Background: The presence of extended-spectrum β-lactamase (ESBL)-producing bacteria among uropathogens is significantly increasing in children all over the world. Thus, this research was conducted to investigate the prevalence of E. coli and their antimicrobial susceptibility pattern, and both genes of ESBL-producing E. coli resistant and virulence factor in UTIs patients among children in Duhok Province, Kurdistan, Iraq. Method: a total of 67 E. coli were identified from 260 urine samples of pediatric patients diagnosed with UTIs aged (0–15 years) which were collected from Heevi Pediatric Teaching Hospital, from August 2021 to the end of February 2022. Result: a high proportion of UPEC infections at ages <5 years and the rates among girls (88%) were significantly higher than those among the boys. A wide variety of E. coli are resistant to most antibiotics, such as Amoxicillin, Ampicillin and Tetracycline, and 64% of them were positive for ESBL. Interestingly, the presence of both the ESBL marker genes (blaTEM, and blaCTX-M) as well as both virulence marker genes (pai and hly) were detected in above 90% of E. coli. Conclusion: the data illustrate an alarming increase in UPEC with ESBL production and the emergence of multidrug-resistant drugs in the early age of children. The public health sectors should further monitor the guidelines of using antibiotics in Kurdistan, Iraq.


Introduction
Urinary tract infections (UTIs) are the most common severe bacterial infections seen in clinical settings worldwide [1,2]. It is estimated that annually there are over 150 million occurrences of UTIs, which are medically expensive [3,4]. On yearly basis, the total healthcare cost of UTIs in the USA, whether for its management or treatment, is over 3.5 billion USD [5], and around 630 million USD for pediatrics UTIs [6]. Pediatric UTIs are estimated to impact 2.4 to 3 % of all American children each year, resulting in more than 1.5 million doctor visits [7], of which 5-14% are dedicated to emergency visits [8]. The incidence of UTIs is common during childhood, and affects in both boys and girls equally during the first year of life [9]. UTIs are considered the second most frequent bacterial infection in children throughout the first seven years of life, and girls account for the majority of infections, approximately 8.4% and less than 1.7% for boys [10][11][12].
It is known that Escherichia coli (E. coli) is the most prevalent bacterial pathogen that causes UTIs and is responsible for 80% of UTIs that occur in children [13]. Numerous virulence factors, such as adhesins, toxins, and siderophores, are present in uropathogenic Escherichia coli (UPEC) strains and contribute to the spread of infection [14,15]. The ability

E. coli Isolation with Phenotypic and Genotypic Detection
In this current study, a total of 260 out-and in-patient children (<15 years) were recruited, from August 2021 and February 2022. From these cultures, 150 (57.7%) were phylogenetic positive bacterial cultures. From 150 samples, E. coli was the predominant pathogenic bacteria, and 67 (44.7%) of them were confirmed phenotypically and molecularly by the species-specific gene (uidA). Phenotypically, E. coli was identified on MacConkey and EMB agars as well as Gram stain and biochemical tests (Oxidas, IMViC, TSI). In addition, all isolated E. coli were genotypically confirmed by the species-specific gene (uidA) (Figure 1a).

Antimicrobial Susceptibility Pattern
All the isolated E. coli underwent an antimicrobial susceptibility test, and the results of the 19 antibiotics and their percentage (yellow colors) are shown in Figure 2. Isolated E. coli were resistant to Tetracycline around (97%) and approximately 85% of them were highly resistant to Amoxicillin and Ampicillin. Furthermore, around 70% of isolated E. coli were resistant to Nalidixic acid, Cefotaxime, Cefixime, Trimethoprim/ Sulfamethoxazole, Gentamicin and Ceftriaxone. In addition, around (57%) of these E. coli were resistant to Amikacin, (51%) to Ciprofloxacin and (45%) to Aztreonam. In contrast, over 85% of E. coli were highly susceptible to Chloramphenicol, Nitrofurantoin and Imipenem.

Phenotypic and Molecular Detection of ESBL Production
All E. coli were subjected to detect the presence of the production of ESBLs by the double disc synergy test (DDST), shown in Figure 2 and Appendix A. The results show that 43 (64%) of E. coli were positive for ESBLs production by DDST. They were also tested for whether they were harboring ESBL genes or not by PCR amplifications as shown in Figure 1. Figure 2 presents individual and collective detecting of marker genes. The fragments which produced ESBL genes include TEM, CTX-M, and SHV, confirmed by agarose gels (Figure 1). Among the ESBL marker genes, TEM and CTX-M were more abundant 63 (94%) and 60 (90%), respectively, and the SHV had the lowest abundance of 1 (1.4%). It is interesting to note that only 3 out of the already-decided 24 non-producer ESBLs E. coli turned out to be non-ESBL, the rest were ESBL (Figure 2). Table 1 illustrates the presence of these ESBL marker genes in all of the isolated E. coli strains and 58 of the isolated strains harboring both TEM and CTX-M genes. Furthermore, the TEM gene was found in the four of the E. coli strains while only one strain harbored three TEM, CTX-M, SHV genes and another strain has only the CTX-M gene. Commented column (e.g.

Molecular Detection of Virulence Factors Genes
The isolated E. coli strains were also tested for the presence of the five virulence factor genes and their PCR product size on agarose gels, and these are shown in Figure 1. The total number and prevalence rate of the five virulence factor genes in E. coli isolates are also shown in Figure 2 and Appendix A. Of the 67 E. coli isolates, 63 (94%), 55 (82%), 27 (40%) 9 (13%) and 4 (6%) of isolates possess pathogenicity island (pai), hemolysin (hly), a fimbrial adhesion (afa), S-fimbrial adhesion (sfa), and cytotoxic necrotizing factor-1 (cnf-1), respectively. Table 1 illustrates the details of virulence factor genes that existed in all of the isolated E. coli strains. The main point of interest is that 21 of the isolated E. coli strains had both pai and hly genes, and 18 strains had three genes, pai, hly and afa. Furthermore, five strains possess three genes, pai, hly, and sfa and, the other five strains possess both pai, and afa. Moreover, two of these E. coli strains possess pai, hly, afa and sfa, one possesses pai, hly, sfa and cnf-1, two possess hly and afa, three possess pai, hly and cnf-1, one possesses pai and sfa, five possess pai, and one possesses hly and afa.

Relationship between UTIs by E. coli and Children Age
The total number of children with UTIs is overwhelming, at earlier ages than the pre-and school aged children. The Spearman's correlation coefficient indicates that there was a significant negative correlation between the UTIs with UPEC and age of children; (r = −0.830, p-value = 0.0003) ( Figure 3).   The correlation analysis indicates that the number of UTIs reduces with children's age, and hence this figure uses a number so that the relationship is evident across ages from 0 to 15 years. Each marker denotes a sum observation of UTIs for a particular age of the children. Spearman's correlation was performed by GraphPad Prism, and significance was considered to be established when p < 0.05.

Relationship between Antibiotic Resistance and Marker Genes of E. coli and Children Ages
Having UTIs Figure 4 illustrates the total number of antibiotics resisted and the marker genes (i.e., antibiotics resistance and virulence factors) of E. coli, in pediatric patients of all ages. It highlights the fact that they are significantly higher in early ages than pre-and school age groups. Figure 4a shows the Spearman's correlation coefficient, indicating     Figure 5 illustrates the relationship between the antibiotic-resistant genes and virulence factors of UPEC infection, using Spearman's correlation coefficient. There were highly significant positive correlations between TEM and CTX (r = 0.98, p-value < 0.0001), pai (r = 0.96, p-value < 0.0001) and hly (r = 0.94, p-value < 0.0001), and also between CTX, and pai (r = 0.96, p-value < 0.0001) and hly (r = 0.95, p-value < 0.0001). Furthermore, despite the very low prevalence of cnf-1, it was also found that cnf-1 was a significant positive correlation with TEM (r = 0.60, p-value = 0.023), CTX (r = 0.60, p-value = 0.021), pai (r = 0.63, p-value = 0.016), and hly (r = 0.62, p-value = 0.018).

Discussion
The incidence of pediatric UTIs has increased dramatically in recent years worldwide, which is caused by a variety of pathogens, particularly Gram-negative bacteria. This study revealed that E. coli was the predominant uropathogen in children in our community, where the prevalence of the infected girls was higher than in boys. These findings were in agreement with the recent results obtained by several studies in Iraq, in Duhok city [35], Erbil city [36,37], and Baghdad city [38], as well as a Child Cohort Study in Taiwan with UTIs between 2004 and 2018 [39]. In fact, there are many factors that can affect the interpretation of UTIs in girls, particularly E. coli. Among these factors are the anatomical structure of the urogenital system, and girls have a shorter distance between the urethral and anal opening [40]. It has been previously suggested that there is a relationship between the arising of the dominant fecal flora, coliform bacteria, colonizing the perineum which enters and ascends to the urinary tract in girls [41][42][43]. In fact, studies have demonstrated that the UPEC strains possess specific properties. For instance, both pili and fimbriae can mediate bacterial attachment and invasion to the uroepithelial cell surface, Figure 5. Correlation analysis among the antibiotic-resistant genes with virulence factors of UPEC infection. The heatmap shows the relationship among the antibiotic-resistant genes and virulence factors of UPEC infection, using computed nonparametric Spearman correlation. The white cells are highly positively correlated between these genes, r > 0.90 and significance were considered when the p-value was <0.0001 (***). At the same time, the light gray colors are moderate positively correlated, with r between 0.60-065 and a p-value < 0.05 (*). On the other hand, the darker gray cells are not significant found. Keys: Gene markers include pai: pathogenicity island, hly: hemolysin, sfa: S-fimbrial adhesion, cnf-1: cytotoxic necrotizing factor-1, and afa: a fimbrial adhesion.

Discussion
The incidence of pediatric UTIs has increased dramatically in recent years worldwide, which is caused by a variety of pathogens, particularly Gram-negative bacteria. This study revealed that E. coli was the predominant uropathogen in children in our community, where the prevalence of the infected girls was higher than in boys. These findings were in agreement with the recent results obtained by several studies in Iraq, in Duhok city [35], Erbil city [36,37], and Baghdad city [38], as well as a Child Cohort Study in Taiwan with UTIs between 2004 and 2018 [39]. In fact, there are many factors that can affect the interpretation of UTIs in girls, particularly E. coli. Among these factors are the anatomical structure of the urogenital system, and girls have a shorter distance between the urethral and anal opening [40]. It has been previously suggested that there is a relationship between the arising of the dominant fecal flora, coliform bacteria, colonizing the perineum which enters and ascends to the urinary tract in girls [41][42][43]. In fact, studies have demonstrated that the UPEC strains possess specific properties. For instance, both pili and fimbriae can mediate bacterial attachment and invasion to the uroepithelial cell surface, allowing them to overcome the host defenses [16,42]. In addition, the majority of pathogenic bacteria in UTIs are the commensal bacteria and their pathogenicity might result from microbiota dysbiosis [42].
In this current study, the frequency of UTIs by E. coli was widely reduced in older children. It showed high infection rates in children aged between 0-3 years and during toilet training for 4-5 years, particularly in girls. The increase in the incidence rate of UTIs at these ages may be due to the use of diapers and the wet tissues for infants and during toilet training for 4-5 years [44]. Boys with UTIs have rarely been detected in this study and this is due to the fact that almost all boys have undergone circumcision in their first week after birth [45]. Thus, this finding is dissimilar to Western studies and highlights the notion that uncircumcision in boys is the main reason for the higher UTIs proportion than girls, which is inverse in our communities [46,47].
In this study, the antimicrobial susceptibility testing of E. coli showed that 97% of them were highly resistant to Tetracycline. This finding is higher than previous studies conducted with different child age groups in Iraq: 63% [48], 83% [33] and 80% [49]. High resistance may primarily be attributed to the diversity of genes in various bacteria that regulate the efflux of Tetracyclines from the cell [50]. Furthermore, this study demonstrated around 85% higher resistance to Penicillin agents, including Amoxicillin and Ampicillin. This may be due to the misuse of older antimicrobials that have made it a less acceptable choice, such as Penicillins agent [51]. Although it has been suggested that certain Penicillins, when taken with β-lactamase inhibitors (clavulanic acid), are useful in treating some infections caused by bacteria that produce ESBLs [50], in this study, 64% of E. coli were resistant to Amoxicillin-clavulanic acid. This finding is in agreement with a similar laboratory-based recent study conducted in Turkey [52], while half of it was found in a study conducted in a European country such as Poland [53].
It has been emphasized that Imipenem is considered a stable antibiotic against the enzyme ESBLs produced by Enterobacteriaceae as well as preferred for treating serious infections caused by bacteria producing β-lactamase [54]. In the current study, approximately 88% of E. coli were sensitive to Imipenem. This finding was slightly lower when compared with other results from studies in Iraq, roughly 95.2% in Erbil city [36,37], and 100% in Baghdad [38]. It is surprising that the very recent data estimated that E. coli have not been observed to be resistant to Imipenem in UTIs patients under 12 old in Iran [55], and in the various ages in Iraqi studies, such as those in Duhok city [56] and Erbil city [57], as well as in Saudi Arabia [58]. In other words, E. coli resistance to Imipenem is steadily increasing when compared with studies of UTIs in various age groups, and this is due to the use of over-the-counter antibiotics and the ease of finding and buying antibiotics without a prescription. The public health sector has not verified the guidelines for how to control the selling of antibiotics. In addition, both Nitrofurantoin and chloramphenicol are considered the second most effective antibiotic against these isolates, roughly 85% in this present study. A similar result was found for Nitrofurantoin in a study conducted in Ethiopia for children under 15 years [59], while a dissimilar finding for E. coli resistance to chloramphenicol was recorded in a study in Iran, accounting for about 72% [60].
It has been previously noted that E. coli which produce extended-spectrum β-lactamase (ESBL) are becoming more common in the community worldwide [61]. In this study, all isolates of E. coli were subjected to detect ESBLs phenotypically in a clinical microbiology laboratory and approximately 64% of E. coli were positive for ESBLs production. The prevalence rate ESBL-producing E. coli was higher compared with several studies from Middle Eastern countries, approximately 30% in both Qatar [27] and Iran [28], about 41% in Turkey [29], and also in European countries, about 16.8% in Sweden [31] and around 2% in Spain [30]. According to Rupp [19], the DDST is much more sensitive and specific. The DDST, however, can produce false-positive or false-negative results [22,62]. Therefore, all isolates of E. coli were subjected to PCR using TEM, CTX-M, and SHV-specific primers for the detection of ESBL genes. Approximately 64% of E. coli were positive for the DDST, while 94% and 89% of them were identified as possessing ESBL genes, TEM and CTX-M, respectively. However, a total of 19 E. coli strains have the ESBL marker genes from 24 non-ESBL producer E. coli, and they were identified by DDST. This finding suggests that the detection of ESBL by PCR is more sensitive than the DDST.
In this study, all isolated E. coli were confirmed by the PCR amplification of E. colispecific marker gene (uidA). As is known, this gene is widely used in the identification of E. coli from clinical samples [63][64][65][66]. Evidence demonstrated that the presence of pai in pathogenic strains carries genes encoding more virulence factors, such as adhesins, toxins (cnf-1 and hly), invasions, and iron absorption systems [67,68]. Altogether, these bacterial virulence factors enhance their ability to adhere, invade and colonize a host and increase the pathogenicity [69,70]. These virulence factors also help pathogenic bacteria to become resistant to immune defenses, such as phagocytosis, the complement system, or adaptive immune responses [69][70][71]. Altogether, these gene factors in UPEC contribute to urinary virulence associated with severe UTIs [18]. Thus, in this study, the high prevalence of virulence factors in the UPEC strain are major causative agents for UTIs in children in Iraq, with a high prevalence rate of both pai (94%) and hly (82%) in almost UPEC strains. The hly gene finding was similar (81%) to a very recent study of uropathogenic E. coli in Egyptian patients (age range 23-56) [72]. These findings were dissimilar to several recent studies among various age groups with UTIs, for pai virulence gene, approximately 72% higher when compared to a study in Zakho City [48] and around 20% higher in three different cities in Kurdistan [33], and the hly virulence gene was found to be 60% higher by [48] and 45% higher by [33]. Whereas afa was roughly similar when compared with a study in Duhok City.
In addition, further analysis was conducted to answer whether the virulence genes have a role in increasing antibiotic-resistant genes or not. Only one study was reported in Egypt by Abd El-Baky et al. [72] which found a significant positive correlation between CTX and hly, similar to this study. Thus, further molecular studies of whole genome sequencing are recommended to highlight the relationship between the virulence factors and antibiotic-resistant genes of these pathogenic E. coli.
Regarding the age, it is noteworthy in this study that the negative correlation between the age of pediatrics and UTIs by UPEC with their resistance to antibiotics and both genes of resistance and virulence factors. This finding has not been previously reported in Iraq. A recently reviewed study has highlighted the reduction in UTIs in children, particularly after age 6 [10]. In addition, the high prevalence of UTIs with E. coli in pediatrics depends on other factors as well. These factors together can cause more UTI outcomes in children than in adolescents. A study by Pärnänen et al. [73] reported that although infants were not exposed to antibiotics, their gut microbiome contains a higher abundance of antibiotic resistance genes than adults. Pärnänen and their colleagues found that the features of infants' fecal antibiotic resistance genes and mobile genetic element profiles were inherited from their mother's gut and breast milk microbiota.
The limitation of this study is the lack of molecular characterization, 16s rRNA gene and whole genome sequencing of E. coli for the phylogenetic analysis might help understand the epidemiology and clinical microbial infection of UPEC in our community.

Sample Collection
The current study was conducted in the microbiology Laboratory of Heevi Pediatric Teaching Hospital in Duhok Governorate, Iraq, from August 2021 to the end of February 2022. A total 260 urine samples were collected from in-patient and out-patient children 0-15 years old with clinical suspicion of UTIs and not receiving antimicrobial treatment.

E. coli Isolation and Identification
Collected urine samples were cultured on MacConkey agar and Blood agar (containing 5% human blood) and aerobically incubated overnight at 37 • C. A growth of a single organism with a count of ≥10 5 colony-forming unit (CFU)/mL was considered to represent the UTIs [74][75][76]. The lactose-fermenting colonies were selected and sub-cultured on both MacConkey and Eosin methylene Blue (EMB) agar to obtain a pure culture and to identify E. coli by green metallic sheen on EMB and Pink colonies on MacConkey agar. Based on the morphological and biochemical characteristics of these pure colonies was identified according to study by [77].

Antimicrobial Susceptibility Test and ESBLs Detection
All isolated E. coli were subjected to antibiotic susceptibility test using Kirby-Bauer disk diffusion method by spreading the inoculated sterile swab on Muller-Hinton agar incubated overnight aerobically at 37 • C according to [78]. Nineteen antibiotics were used, supplied by ( [79]. A double-disk synergy test was used to identify E. coli ESBL producer and five antibiotic disks (Cefotaxime, Ceftazidime, Ceftriaxone and Aztreonam were placed 30 mm from an amoxicillin/clavulanate in the center) were used on Muller-Hinton agar according to the method described by [80].

Bacterial DNA Extraction from E. coli
Bacterial DNA was extracted from isolated E. coli using the Quick DNA Extraction kit (Guangzhou Dongsheng Biotech Co., Ltd.) following the manufacturer's protocol. Bacterial DNA quality was achieved by (NanoDrop™ One UV-Vis Spectrophotometer, Thermo Fisher Scientific, Waltham, MA, USA) and then stored at −20 • C for DNA amplification. Different primers were used for amplification of these marker genes (Macrogen, Seoul, Korea) as described in Table 2. Purified DNA was used for PCR amplification of 9 genes (Table 3). A species-specific primer for E. coli, uidA gene. After confirming E. coli, 3 primers were used to detect ESBL genotypes, bla CTX-M , bla TEM and bla SHV β-lactamase genes. In addition, five primers were used for detecting the five virulence genes, including pathogenicity island (pai), hemolysin (hly), S-fimbrial adhesion (sfa), cytotoxic necrotizing factor-1 (cnf-1), and a fimbrial adhesion (afa).   The amplifications of DNA for each gene were carried out in PCR tubes containing master mix and consisting of 7 µL Tag master (Guangzhou Dongsheng Biotech Co., Ltd., Guangzhou, China), 1 µL of each forward and reverse primers (10 pmol/µL), 2 µL pure DNA (25-50 ng/µL), and adding 9µL of free-nuclease water to make a final volume of 20 µL. The amplification conditions of each gene are described in Table 2.
The DNA lengths of each fragment produced with these 9 primers were confirmed by running 7 µL of each PCR products on 1.2 % agarose gels in 1× TAE buffer and added 5 µL of Safe Gel stain Dye (Guangzhou Dongsheng Biotech Co., Ltd., Guangzhou, China) and the electrophoresis was performed at 85 V for 35 min. The agarose gel was visualized under the UV radiation (Cleaver Scientific Ltd., Rugby, UK). The images of DNA bands were captured, and the estimated amplicon size were compared with the 100 pb DNA ladder (Guangzhou Dongsheng Biotech Co., Ltd., Guangzhou, China) and 300 bp (GeneDirex, Inc, Taoyuan, Taiwan).

Statistical Analysis
GraphPad Prism 9.4.1 was used to analyze the data. Spearman's method was used for nonparametric correlation between the children age and the frequency of UPEC, antibiotics resistance patterns and genes of both ESBL production and virulence factors. Significance was considered to be established when p < 0.05.

Ethical Approval
The approval for conducting this study was given by the Ethical Committee of Duhok Directorate General of Health (ethical code n 18082021-8-15) and the Ethical and Protocol Review Committee of the Biological Sciences Committee (BSCZ) at the University of Zakho (ID: "BSCZ/28/7/2021").

Conclusions
In conclusion, the data illustrates that E. coli remain the most predominant bacteria among pediatrics with UTIs, particularly for girls. They also state an alarming increase in E. coli with ESBL production of UTIs in children of ages younger than the pre-and school age group of children. The E. coli ESBL-producer was more resistant to antibiotics than the non-producer. Moreover, Imipenem and Nitrofurantoin are considered to be the most effective antibiotic choices for the treatment of pediatric UPEC infections. UPEC resistance to antibiotics decreases with children of the older age group. As well as this, molecular detections of ESBL-producing E. coli were more accurate than phenotypic identification. In addition, ESBL colonization and infection are increased in the urinary tract because of these genes' products, pai and hly, as well as two β-lactamase genes, bla CTX-M and bla TEM . The public health sectors would be better if they monitored the guidelines for the use of antibiotics and the careful use of antibiotics as recommended by physicians only. Furthermore, doctors need to give more advice about accurate hygiene, especially cleaning toddlers and pre-school children during toilet training to avoid UTIs. Hence, further molecular studies of different clinical specimens and different ages would achieve a confirmed database for ESBL-producing E. coli in Kurdistan, Iraq.

Conflicts of Interest:
The authors declare no conflict of interest.

Appendix A
The total of antibiotic resistance profiles and genes, ESBL production and virulence factor genes of all isolated E. coli in different pediatric ages.