High Biofilm-Forming Ability and Clonal Dissemination among Colistin-Resistant Escherichia coli Isolates Recovered from Cows with Mastitis, Diarrheic Calves, and Chickens with Colibacillosis in Tunisia

Background: Escherichia coli (E. coli) is one of the main etiological agents responsible for bovine mastitis (BM), neonatal calf diarrhea (NCD), and avian colibacillosis (AC). This study aimed to assess resistance and virulence genes content, biofilm-forming ability, phylogenetic groups, and genetic relatedness in E. coli isolates recovered from clinical cases of BM, NCD, and AC. Materials/Methods: A total of 120 samples including samples of milk (n = 70) and feces (n = 50) from cows with BM and calves with NCD, respectively, were collected from different farms in Northern Tunisia. Bacterial isolation and identification were performed. Then, E. coli isolates were examined by disk diffusion and broth microdilution method for their antimicrobial susceptibility and biofilm-forming ability. PCR was used to detect antimicrobial resistance genes (ARGs), virulence genes (VGs), phylogenetic groups, and Enterobacterial repetitive intergenic consensus PCR (ERIC-PCR) for their clonal relationship. Results: Among the 120 samples, 67 E. coli isolates (25 from BM, 22 from AC, and 20 from NCD) were collected. Overall, 83.6% of isolates were multidrug resistant. Thirty-six (53.73%) isolates were phenotypically colistin-resistant (CREC), 28.3% (19/67) were ESBL producers (ESBL-EC), and forty-nine (73.1%) formed biofilm. The blaTEM gene was found in 73.7% (14/19) of isolates from the three diseases, whilst the blaCTXM-g-1 gene was detected in 47.3% (9/19) of isolates, all from AC. The most common VG was the fimA gene (26/36, 72.2%), followed by aer (12/36, 33.3%), cnf1 (6/36, 16.6%), papC (4/36, 11.1%), and stx1 and stx2 genes (2/36; 5.5% for each). Phylogenetic analysis showed that isolates belonged to three groups: A (20/36; 55.5%), B2 (7/36; 19.4%), and D (6/36; 16.6%). Molecular typing by ERIC-PCR showed high genetic diversity of CREC and ESBL E. coli isolates from the three animal diseases and gave evidence of their clonal dissemination within farms in Tunisia. Conclusion: The present study sheds new light on the biofilm-forming ability and clonality within CREC and ESBL-EC isolated from three different animal diseases in Tunisian farm animals.


Introduction
Escherichia coli (E. coli) is a highly diverse group of Gram-negative bacteria with the ability to colonize and persist in humans, warm-blooded animals, and abiotic environments [1,2]. However, some pathovars of E. coli are responsible for severe gastrointestinal

Sampling and Sample Collection
In the period from February to April 2016, a total of 120 samples were collected from cows with mastitis and diarrheic calves in three adjacent farms (FIV, FV, and FVI) located in Bizerte and Ariana governorates in Northern Tunisia. Animals showed repetitive episodes of illness without death and were subjected to clinical examination by veterinarians. Following oral consent from animals' owners, samples including mastitis milk and feces were collected from cows (n = 70) and calves (n = 50), respectively. Farms included in the present survey were characterized by a number of animals ranging from 15 to 32 and were not under control by official veterinary services. According to animals' owners, the most commonly used antibiotics in treating diseased animals in these farms included β-lactams, aminoglycosides, fluoroquinolones, and tetracyclines. Cows and calves included in the present investigation shared the grazing environment, food, and water sources. In addition, cow's milk was used to feed young calves. All samples were transported to the laboratory within a few hours of collection in refrigerated boxes and processed immediately. Twentytwo E. coli isolates previously identified from chickens who died of colibacillosis in three different farms (I, II and III) located in Nabeul, Ben Arous, and Zaghouane governorates in Northeast Tunisia were included in the present study for further analysis [20].

Isolation and Bacteria Identification
One hundred microliters from each sample was placed in brain heart infusion broth (Oxoid Ltd., Basingstoke, UK) and incubated aerobically at 37 • C for 24 h. Then, 10 µL of culture suspensions were seeded onto MacConkey agar (Merck, Darmstadt, Germany) plates and incubated overnight at 37 • C. Isolates with typical E. coli morphology were selected and seeded onto Endo agar (Merck) and incubated overnight at 37 • C. One presumptive colony per sample was selected and identified by conventional methods including Gram staining and biochemical tests (oxidase, catalase, urea-Indole, lactose, and glucose fermentation gas production ability in Kigler-Hajna agar) and by an API 20E system (BioMerieux, Marcy l'Etoile, France). Bacterial DNA for polymerase chain reaction (PCR) analysis was prepared by boiling a loopful of bacteria in 400 µL of TE buffer (10 mMTris-2 mM EDTA) for 10 min, followed by centrifugation for 15 min at 10,000× g. Subsequently, isolates were confirmed as E. coli using species-specific PCR targeting the uidA gene encoding for β-glucuronidase structural protein [24] (Table 1).

Antimicrobial Susceptibility Testing and Screening for ESBL Production
Antimicrobial susceptibility of all E. coli isolates was determined using the disc diffusion method and interpreted according to the Clinical and Laboratory Standards Institute [25] and the European Committee on Antimicrobial Susceptibility Testing guidelines [26]. The following antibiotics (Oxoid) were used (µg/disk): Ampicillin (AMP, 10 µg),  (FEP,30 µg) in the proximity of amoxicillin-clavulanic acid (AMC, 20/10 µg) was used for the screening of ESBL production. E. coli ATCC25922 and Klebsiella pneumonia ATCC700603 were used as ESBL-negative and positive control strains, respectively. The isolates were defined as multidrug-resistant (MDR) if they exhibited resistance to at least one agent belonging to three or more antimicrobial families [27].

Colistin Susceptibility Testing and Screening of Colistin Resistance Genes
The minimum inhibitory concentration (MIC) of colistin was determined using the broth microdilution method (BMD) according to the CLSI guidelines [28]. Dilution methods were performed using colistin sulfate (Sigma-Aldrich, Merck, Darmstadt, Germany) tested over a range from 0.25 to 128 µg/mL. All experiments were performed in triplicate. E. coli ATCC 25922 was used as a quality control strain. The mcr-1, mcr-2, mcr-3, and mcr-4 genes encoding for colistin resistance were investigated by PCR in all isolates with MIC ≥2 µg/mL as described elsewhere [18,[29][30][31].

Biofilm Formation Assay
The biofilm formation ability of the 67 E. coli isolates was performed in 96-well microtiter plates [34]. Briefly, an overnight culture was diluted (1:100) in TSB containing 1% glucose and inoculated onto microtiter plates at 37 • C for 18 h without aeration. The free-floating planktonic bacteria were removed and washed, dried for 60 min at 60 • C, and stained with 0.06% crystal violet. The biofilm was quantified in duplicate, after adding 200 µL of 95% ethanol using a microtiter plate reader by an enzyme-linked immunosorbent assay plate reader at 570 nm (BioRad). Each strain was tested in triplicate and each assay was performed in duplicate. E. coli ATCC25922 and S. epidermidis strain ATCC12228 were used as positive and negative controls, respectively. The isolates were classified as strong biofilm producer: 4 × OD C < OD; moderate biofilm producer: 2 × ODc < OD ≤ 4 × ODc; weak biofilm producer: ODc < OD < 2 × ODc; and no biofilm producer: OD ≤ ODc [35]. The cut-off value (ODc) is defined as three standard deviations (SD) above the mean OD of the negative control (TSB plus 1% glucose, without bacterial cells) [35].

E.coli Molecular Typing by ERIC-PCR
CREC isolates were fingerprinted by ERIC-PCR as described by Bilung et al. (2018) [40] ( Table 1) and different ERIC-PCR profiles were analyzed visually and numerically according to Tenover et al. (1998) [41]. Then, the phylogenetic tree was established using MVSP 3.2 software. The comparison between ERIC-PCR profiles was conducted using the Jaccard coefficient, and a dendrogram was constructed using the unweighted pair group method with arithmetic mean (UPGMA).

Statistical Analysis
Statistical analysis was performed in IBM SPSS 22.0. A chi-squared test χ 2 using the Pearson Chi-square test was employed to estimate differences between colistin resistance, virulence genes, biofilm formation, and ESBL production rates in E. coli from the three animal diseases, whereby a probability of less than 0.05 was considered statistically significant.

Collected E. coli Isolates
A total of 45 (37.5%) out of 120 bovine samples displayed a positive culture for E. coli (25 isolates from BM and 20 from NCD). In addition, 22 E. coli isolates recovered from chickens that died of AC, which were previously identified, were added to the collection for further analysis. Overall, 67 E. coli isolates were included in the present investigation.

Antimicrobial Susceptibility Testing and Screening for ESBL Production
The highest rates of antibiotic resistance in the 67 E. coli isolates were found for cefsulodine (67/67; 100%), followed by ceftazidime (    Avian isolates showed the most important antibiotic resistance rates for the majority of antibiotics tested except for cefotaxime, ceftazidime, and colistin, for which mastitis isolates displayed the highest resistance rates as shown in Table 2. Although diarrheal isolates showed the lowest resistance rates to the majority of antibiotics, they revealed higher resistance frequencies for ceftazidime, cefotaxime, cefoxitin, and gentamicin than those in avian isolates. Of the 67 E. coli isolates, only 19 (28.3%) were ESBL producers. Among them, nine (47.3%) originated from chickens with colibacillosis, seven (36.8%) from cows with mastitis, and three (15.8%) from calves with diarrhea (Table 3). Table 3. Phenotypic resistance patterns, minimal inhibitory concentrations of colistin, ESBL production, and biofilm-forming ability in 67 E. coli isolates recovered from bovine mastitis, calves' diarrhea, and avian colibacillosis in farms in Tunisia.

Detection Genes Encoding ESBL Enzymes and Other Resistance Markers
The bla TEM gene encoding for ESBL production was found in 73.7% (14/19) of isolates (seven from cows, four from chickens, and three from calves). However, the bla CTXM-g-1 gene was only detected in nine (9/19; 47.3%) phenotypically ESBL-EC strains. All bla CTXM-g-1-carrying isolates were from chickens with colibacillosis, whilst none of the mastitis and diarrhea isolates carried this gene. The bla SHV gene was detected in a single isolate of avian origin (1/19; 5.2%), whilst all isolates were free of the bla CTX-M-g-8 and bla CTX-M-g-9 genes (Table 4). A statistically significant relationship was found between ESBL production and the type of animal diseases (p < 0.05) ( Table 5).

Detection of Virulence and Biofilm Encoding Genes in CREC Isolates
The fimA gene was detected in 27 of the 36 CREC isolates (75%). Among these strains, 17 (63%) were from cows with mastitis and 9 (33.3%) from chickens with colibacillosis. In chickens, the stx1 and stx2 genes were observed in two isolates (9% for each), whilst isolates from mastitis and diarrhea were free of these genes. The papC gene was observed in three and one isolates from mastitis and chickens, respectively. The aer gene encoding for aerobactin was detected in 7 out of the 22 chicken isolates (31.8%) and 5 out of the 25 mastitis isolates (20%). In addition, the cnf l gene was found in three isolates from both chickens and mastitis (13.6% and 12%, respectively). Mastitis isolates harboring VGs contained either the fimA gene alone or in combination with aer, cnf 1, and papC genes. However, none of the diarrheal isolates was positive for the analyzed genes.

CREC Molecular Typing by ERIC-PCR
Genetic relationship analysis for the 36 CREC isolates using ERIC-PCR showed 14 different ERIC types (ETs). Identical ETs were allocated letters from A to G, while unique ETs were assigned letters from H to N (Figure 1). The UPGMA method indicated seven different clones (A, B, C, D, E, F, G). The most prevalent ET was the type A (7/36; 19.4%) identified in the mastitis isolates, followed by the type B (5; 13.8%) found in isolates from diarrhea and mastitis, then the types F and G (4/36; 11.1%, for each) detected in avian isolates. The ETs C, D, and E (3/36; 8.3%, for each) were found in either mastitis or avian isolates. While ETs C and E were observed in mastitis and avian isolates, the type D was only detected in mastitis isolates (Table 4, Figure 1). The dendrogram analysis indicates that the highest genetic diversity was observed in the eighteen CREC isolates collected from bovine mastitis with the presence of eight different ETs (A, B, C, D, E, M, L, and E). Among these ETs, the types A and D detected in mastitis isolates were found in different farms (FIV, FV, and FVI; FIV). Likewise, the ETs C and E including mastitis and avian isolates were found in three different farms, (FI, FIII, FIV) and (FI, FIV, FVI), respectively. Type B was detected in isolates collected from mastitis, diarrhea, and mastitis belonging to FIV, FV, and FVI. However, the ETs F and G included only avian isolates from three different farms (FI, FII, FII) (Table 4, Figure 1).
In calves, the same trend in resistance rates was observed regarding cefsulodine, ceftazidime (100% for each), and cefotaxime (85%). The lowest resistance rate was found for streptomycin (70%) and cefoxitin (75% for each), whilst the highest resistance rate to tetracyclines (75%) was recorded compared to mastitis isolates. These resistance rates are higher than those found by Srivani et al. (2017) [43].
In chickens, the highest rate of antibiotic resistance was found regarding cefsulodine (100%), streptomycin and nalidixic acid (95.4% for each), and tetracyclines (86.3%). The resistance rates for streptomycin (95.4%) and tetracyclines (86.3%) were higher than those found by Wang et al., 2021 [44] in chickens with colibacillosis in China. However, the resistance rate for gentamicin (4.5%) was lower compared to what was recorded in that study [44].
Interestingly, all mcr-1-positive E. coli isolates were multidrug-resistant, exhibiting resistance to common antimicrobials. This finding is in agreement with those found by Liu et al. (2020) [46].
In our study, E. coli isolates from the three different origins were free of the mcr-2, mcr-3, or mcr-4 genes. This result may be explained by the possession of other mcr gene variants or chromosomal mutation(s) [48]. Subsequently, further molecular investigations are needed to identify genes involved in colistin resistance in these isolates.
In the present investigation, a statistically significant association between colistin resistance, mcr-1 carriage, and biofilm formation ability was found supporting previous studies [49].
Strains isolated from chickens showed an important ESBL production rate (9/22, 40.90%) that was lower than that found by Parvin et al. (2020) (86%) [50] and higher than that observed in the study of Johar et al., 2021(3.8%) [45] in E. coli isolated from chickens. In cows with mastitis, 28% (7/25) of the E. coli isolates were phenotypically ESBL producers. This frequency is close to that obtained by Liu et al. (2020) [46], who recorded only 20% of ESBL-producing E. coli among 249 strains isolated from milk from cows with mastitis. The lowest rate of ESBL-producing isolates (3/20, 15%) was recorded in isolates from calves' diarrhea.
The bla CTXM-g-1 was detected only in ESBL-producing avian isolates, whilst the bla TEM gene was found in isolates from the three animal pathologies. This result is consistent with previous studies in which the bla TEM gene was the most predominant ESBL encoding gene in E. coli isolated from avian colibacillosis, diarrheic calves, and bovine mastitis [20,51,52].
A total of 33 out of the 36 colistin-resistant E. coli isolates contained at least one of the following genes: tetA, sul1, mcr-1, aadA, floR, strB, dfrAI cmlA, strA, and strB, demonstrating the important antibiotic resistance pool in CREC isolates recovered from the three animal pathologies. Previous studies have demonstrated the detection of the same aforementioned genes in E. coli from cows with mastitis [42], avian colibacillosis [44], and diarrheic calves [51]. In the present study, antibiotics were widely used in farm animals either for treating mastitis, colibacillosis, and diarrhea or even to enhance their productivity. This practice promotes the dissemination of ARB that could reach food products, causing serious public health issues [53,54].
A total of 49 of the 67 isolates (49/67; 73.1%) formed biofilm. In cows, the percentage of biofilm-forming E. coli isolates (20/25, 80%) was lower than that found by [55], who reported 100% of biofilm-forming E. coli isolates from acute clinical environmental bovine mastitis in Brazil. In chickens, 77.3% (17/22) of isolates showed biofilm-forming ability. This frequency was markedly higher than that reported in the study of [49], in which only 45% of the APEC strains showed biofilm formation ability. The lowest rate (12/20, 60%) of biofilm formation was observed in calves when compared to the other origins. These results are higher than those reported in the study of [47], in which 45% of E. coli isolated from calves with diarrhea in Uruguay formed biofilm. Based on the observations of the biofilm formation assay, the study suggested that mastitis isolates were more biofilm producers than those of avian and calf origins. Our findings show that bovine mastitis, avian colibacillosis, and neonatal calf diarrhea may be biofilm-related diseases as biofilm-forming bacteria can be resilient to the immune system, antibiotics, and other treatments [17,49,56]. In addition, biofilm plays a key role in horizontal gene transfer (HGT) facilitated by highly dense cells nearby [57] which smooth the movement of RGs and virulence factors, especially under the selective pressure of antibiotics [58,59]. In the biofilm formation process, the key event is the attachment to the surface leading to subsequent aggregation and mature biofilm formation. This increases the stability of bacteria to cause diseases and enhances their drug resistance capacity [14].
The co-occurrence of various VGs encoding for Shiga toxins known as diarrhea genic (stx1/2), aerobactin synthesis (aer), fimbria type I (fimA), and P-fimbriae involved in septicemia (papC) emphasizes the fact that these avian isolates might be incriminated in the morbidity of chickens. Diagnosing APEC based on virulence genes is difficult since there is no specific set of virulence genes systemically associated with APEC [7]. However, based on the presence of specific genes, three isolates could be categorized as Shiga toxin-producing E. coli (STEC), containing stx1, stx2, or stx1 and stx2 genes [62]. STEC represents a public health threat if transmitted to humans as they can adhere to host epithelial cells and cause damage [63]. Furthermore, previous studies have provided evidence of potential zoonotic transmission of STEC isolates recovered from diarrheic cattle and their food products to humans, representing an emerging public health threat [61].
Mastitis isolates harbored fimA, aer, cnf 1, and papC. However, the presence of these genes could not determine whether these isolates are ExPEC or not. Other virulence genes associated with ExPEC such as traT, fyuA, and iutA genes were found in E. coli from bovine mastitis in previous studies [10]. The cnf 1 genes were detected in 24%of mastitis isolates, which is contradictory to the study of Suojala et al., 2011 [12] from Finland, in which all E. coli isolates from bovine mastitis were free of this gene. However, the stx1, stx2, and hlyA genes were not detected in bovine isolates in our study following a previous study [12].
Although diarrheal isolates were free of the analyzed VGs, they might be reservoirs for other VGs such as fimH and csgA [44]. Animals and humans in contact with calves may become infected through their feces serving as reservoirs for antibiotic and virulence genes.
The results of the assessment of VGs showed high genetic heterogenicity among isolates as shown in other studies [47]. This heterogeneity might be the result of the acquisition and/or the deletion of genetic elements and localization of many virulenceassociated genes on bacteriophages, plasmids, transposons, and pathogenicity islands contributing to either the gain or loss of these pathogenic attributes [64]. Although many VGs were not detected in E. coli isolates, the biofilm-forming ability might be due to the presence of pathogenicity islands and the expression of other virulence determinants.
It is important to take into account that given the small number of isolates, particularly those originating from calves, it is difficult to draw a clear conclusion about their virulence patterns. Thus, further molecular characterization of VRs in all isolates would be of great relevance to better elucidate the virulence background of E. coli incriminated in the three animal pathologies. Phylogroup distribution showed that most of the CREC isolates (20/36; 55.5%) were allotted to phylogroup A. In contrast to colibacillosis isolates that belonged mostly to the B2 phylogroup (8/12), mastitis and diarrheal isolates were mostly of the A phylogroup (13/18 and 5/6, respectively). This finding might be due to the difference in the origin of samples and the health status of animals. Indeed, avian E. coli isolates were recovered from dead chickens, whilst mastitis and diarrhea were isolated from diseased animals. The phylogroup results found in this study demonstrate the high pathogenicity of avian isolates compared to those from cows and calves. In previous studies, the virulent B2 group was frequently detected in ExPEC incriminated in severe human infections [12], which demonstrates the high zoonotic potential of avian isolates. Mastitis isolates were of phylogroups A1, D1, or B2. In contrast to phylogenetic type A, phylogroups D and B2 were considered virulent by Clermont et al. 2000 [38].
In this study, most of the mastitis isolates (72.2%) belonged to phylogroup A, a finding that agrees with similar studies that revealed the predominance of A and D phylogroups in E. coli isolates from cows with mastitis [65]. Contrarily, lower amounts of isolates were allotted to the D and B2 phylogroups, corroborating previous studies [65].
In the present investigation, the ERIC-PCR-based genotyping analysis of the 36 CREC isolates recovered from BM, NCD, and AC showed an important level of genetic diversity (14 ERIC profiles). The ERIC band patterns ranged from 1 to 10 bands with a size range from 100 to 2000 bp, comparable with reports from Egypt [61]. The twenty-two mastitis isolates showed the most identical ERIC profiles. Among these isolates, seven displayed the same ET (A), and three belonged to the ET (D), demonstrating clonal dissemination of E. coli among cows with mastitis. This result is consistent with that of Nüesch-Inderbinen et al., 2019 [10], who found high genetic diversity in E. coli isolates collected from cases of bovine mastitis. The six strains isolated from calves' diarrhea showed four different profiles (groups B, H, K, and N); among them, ET (B) included three isolates. This result is in agreement with the study conducted by Gharieb et al. (2019) [66], who observed seven clusters in E. coli isolates recovered from diarrheic calves.
It is worth noting that the study of genetic relatedness using ERIC-PCR in avian isolates showed concordance with phylogenetic analysis results found in a previous study using the technique of pulsed-field electrophoresis (PFGE), considered as a reference technique for the molecular typing of bacteria [20]. In the present investigation, ERIC-PCR genotyping showed that some mastitis and colibacillosis isolates belonged to the same ET (groups C and E). This result is different from that of Grami et al. (2014) [67], who found no clonal relationship between strains from colibacillosis and bovine mastitis in Tunisia. In addition, ET (B) included strains from mastitis (n = 2) and calf diarrhea (n = 3), whilst no identical ETs were observed in chickens and calves' isolates. ERIC-PCR analysis showed not only identical profiles but also unrelated patterns among CREC recovered from the three animal pathologies, which may reflect the diversity of CREC clones incriminated in these animal diseases in Tunisia.
The ETs (A) and (D) detected in mastitis isolates were found in different farms (FIV, FV, and FVI and FIV and FVI, respectively). This result indicates the dissemination of two different CREC clones in farms included in this study. This could be explained by the movement of animals between farms and their sharing of grazing and water sources. In addition, the (C) and (E) ETs including mastitis and avian isolates were found in three different farms (FI, FIII, and FIV and FI, FIV, and FVI, respectively). This finding demonstrates the involvement of the same CREC clones in avian colibacillosis and bovine mastitis in Tunisia. The ET (B) was detected in isolates collected from diarrhea and mastitis belonging to FIV, FV, and FVI, indicating that this clone disseminated among diseased calves and cows from different farms in Tunisia. However, the ETs (F) and (G) found in CREC of avian origin were circulating in three different farms (FI, FII, FII) but were absent in mastitis and calves' isolates. There were no identical CREC clones between diarrheal and colibacillosis isolates. However, some diarrheal isolates (D22 and D43) showed close ETs with avian isolates clustered in the ETs (F) and (C), respectively.
ESBL and non-ESBL-producing CREC isolates from the three animal pathologies in the different farms were found to be related by ERIC genotyping. Moreover, ERIC-PCR has revealed a clonal relationship between E. coli biofilm-producer isolates from cows with BM following previous investigations in Brazil [68].
In the present study, the combinations of identical ETs, biofilm-forming ability, phylogenetic groups, virulence, and resistance profiles among some of the CREC isolates from the same or different animal pathology highlights the potential intra-species cross-transmission of these isolates and/or their genes in the study region.
On combining data, the majority of strong biofilm-producing CREC isolates were of either mastitis or colibacillosis origin and were allotted to the A, B2, and D phylogroups. These isolates displayed seven ETs that were circulating between cows, calves, and chickens, suggesting clonal dissemination of strong biofilm-producing CREC isolates with clinically relevant phylogroups in farms from Tunisia.

Conclusions
The present study showed a high prevalence of MDR E. coli (83.6%) isolated from BM, NCD, and AC. CREC and ESBL-EC isolates were shown from the three different origins. E. coli isolates harbored a combination of resistance, virulence, and β-lactamaseencoding genes and were assigned to the A, B, and D phylogroups. This is the first report of the biofilm formation ability in E. coli isolated from clinical cases of bovine mastitis, avian colibacillosis, and neonatal calves' diarrhea in Tunisia. Our study revealed a high propensity of E. coli isolates recovered from diseased animals to produce biofilm, suggesting the importance of biofilm-forming ability in the pathogenesis process. Further, this paper sheds new light on the diversity and the clonality observed within CREC and ESBL-EC isolates from three different animal diseases in farms in Northern Tunisia.

Data Availability Statement:
The data presented in this study are available on request from Dr Ramzi Boubaker Elandoulsi.