Skip to main content
PMC home page
PLOS ONE logoLink to PLOS ONE
. 2019 May 17;14(5):e0217174. doi: 10.1371/journal.pone.0217174

Assessing the occurrence and transfer dynamics of ESBL/pAmpC-producing Escherichia coli across the broiler production pyramid

Ilias Apostolakos 1, Lapo Mughini-Gras 2,3, Luca Fasolato 1, Alessandra Piccirillo 1,*
Editor: Grzegorz Woźniakowski4
PMCID: PMC6524947  PMID: 31100096

Abstract

Extended-spectrum β-lactamase (ESBL)- and plasmid mediated AmpC-type cephalosporinase (pAmpC)-producing Escherichia coli (ESBL/pAmpC E. coli) in food-producing animals is a major public health concern. This study aimed at quantifying ESBL/pAmpC-E. coli occurrence and transfer in Italy’s broiler production pyramid. Three production chains of an integrated broiler company were investigated. Cloacal swabs were taken from parent stock chickens and offspring broiler flocks in four fattening farms per chain. Carcasses from sampled broiler flocks were collected at slaughterhouse. Samples were processed on selective media, and E. coli colonies were screened for ESBL/pAmpC production. ESBL/pAmpC genes and E. coli phylogroups were determined by PCR and sequencing. Average pairwise overlap of ESBL/pAmpC E. coli gene and phylogroup occurrences between subsequent production stages was estimated using the proportional similarity index, modelling uncertainty in a Monte Carlo simulation setting. In total, 820 samples were processed, from which 513 ESBL/pAmpC E. coli isolates were obtained. We found a high prevalence (92.5%, 95%CI 72.1–98.3%) in day-old parent stock chicks, in which blaCMY-2 predominated; prevalence then dropped to 20% (12.9–29.6%) at laying phase. In fattening broilers, prevalence was 69.2% (53.6–81.3%) at the start of production, 54.2% (38.9–68.6%) at slaughter time, and 61.3% (48.1–72.9%) in carcasses. Significantly decreasing and increasing trends for respectively blaCMY-2 and blaCTX-M-1 gene occurrences were found across subsequent production stages. ESBL/pAmpC E. coli genetic background appeared complex and bla-gene/phylogroup associations indicated clonal and horizontal transmission. Modelling revealed that the average transfer of ESBL/pAmpC E. coli genes between subsequent production stages was 47.7% (42.3–53.4%). We concluded that ESBL/pAmpC E. coli in the broiler production pyramid is prevalent, with substantial transfer between subsequent production levels.

Introduction

Extended spectrum cephalosporins (ESCs) are considered critically important antimicrobials (CIAs) in both human and veterinary medicine [1,2]. Resistance to these antimicrobials is therefore a major public health concern due to the risks of therapy failure [3]. Resistance in Enterobacteriaceae, such as Escherichia coli, caused by production of extended-spectrum β-lactamases (ESBLs) or AmpC β-lactamases (pAmpCs) and transferability of resistance mechanisms are of particular importance, as the encoding genes (hereafter ESBL/pAmpC genes) are often located in promiscuous plasmids [4]. This property of EBSL/pAmpC genes enables their exchange between bacteria, including pathogens, and favours transmission between animals and humans [5]. Usage of ESCs in poultry has been restricted since 2012 in the European Union (EU) [6] except for occasional use in (grand)parent hatcheries [7,8]. However, numerous studies have shown high prevalence rates of ESBL/AmpC-producing E. coli (ESBL/pAmpC E. coli) in broiler flocks across Europe [9–11], even in countries with low antimicrobial use in livestock [7,12]. The presence of ESBL/pAmpC E. coli has been evidenced at all levels of the broiler production pyramid from breeding farms [13], hatcheries [14], fattening farms [15], slaughterhouses [16] to retail meat [17]. Whether broiler meat contaminated with ESBL/pAmpC E. coli represents an important source of human infections is debatable [18]. For instance, a recent comparative risk assessment estimated the exposure to ESBL/pAmpC E. coli through consumption of chicken meat to be lower than beef and pork [19], whereas a meta-analysis identified poultry products as the most likely source of human ESC-resistant infections [20]. The zoonotic transmission potential has prompted many investigations, not only on the prevalence of ESBL/pAmpC E. coli, but also on their transmission routes along the broiler production pyramid. A recent review [21] summarised the transfer of ESBL/pAmpC-producing bacteria in four major pathways; vertical transmission from parent to offspring, transmission in the hatchery, horizontal transmission in fattening farms, and horizontal transmission between farms or from the environment. Additionally, the authors stressed the need for more quantitative data on these transmission pathways.

In Italy, data on both prevalence and transmission dynamics of EBSL/pAmpC genes in poultry are lacking. This study aimed to establish a baseline prevalence of ESBL/pAmpC E. coli in the broiler production pyramid and to characterise the genetic background of resistant isolates in order to assess their transfer between subsequent levels of the broiler production pyramid.

Materials and methods

Sample collection

Three production chains (chain A, B, and C) of an integrated broiler production company in Northern Italy (specifically in Veneto, Lombardy, and Friuli-Venezia Giulia regions) were monitored from January 2017 to January 2018. In each of the 40 sampling visits, faecal samples from 20 randomly selected healthy birds were taken by cloacal swabs in the farms, and 20 carcasses were collected at the slaughterhouse. For each chain, sample collection started at the top of the production pyramid by sampling day-old Parent Stock (PS) chicks, which were the progeny of the Grandparent Stock (GPS) flock located elsewhere. Imported day-old PS chicks were sampled upon arrival to the rearing farm within one hour from delivery. Due to time constraints, samples were not collected for PS chicks of chain B. The same flock of PS chickens was sampled again during the laying period (~30-weeks-old) at the production farm (PS breeders), in which they were moved at the age of 21-weeks. Sampled breeder flocks were not combined with other flocks at the production farms. The offspring of the sampled PS breeders were tracked and sampled in four commercial fattening farms per production chain at two time points; at the age of one-day-old (broiler chicks), and around 30th day of age, within the last week prior slaughter (broilers) (S1 Table). Samples in each broiler farm were collected from the same poultry house, which contained only the progeny of the previously sampled PS breeders. Finally, 20 carcasses from the previously sampled broilers were collected at the slaughterhouse after the chilling process. All breeder flocks and all broiler flocks, except three of them, received antibiotic treatment, mainly amoxicillin for therapeutic reasons unrelated to this study (S1 Fig). Broilers from the three monitored chains were all processed in the same slaughterhouse in Veneto region. Cloacal swabs from birds were collected by veterinarians of the private company as part of the routine monitoring of the flock health status and conducted in compliance with good veterinary practices.

Isolation and detection

Cloacal swabs were directly streaked on Eosin Methylene Blue agar (Microbiol, Italy) supplemented with 1mg/L cefotaxime (CTX-EMB) and incubated at 37±0.5 °C for 20±2h. Carcasses were analysed by both a qualitative and quantitative method. For the qualitative method, carcasses were rinsed with Buffer Peptone Water (BPW), rinsates were incubated (37±0.5 °C for 20±2h) and streaked on CTX-EMB. For the quantitative method, rinsates and three serial dilutions (10−1 to 10−3) were plated on CTX-EMB for subsequent enumeration. One to two morphologically typical E. coli colonies on CTX-EMB (metallic green sheen) were isolated from each sample and subjected to species confirmation by combination of indole test and PCRs targeting E. coli housekeeping genes [22]. Confirmed E. coli isolates were screened for ESBL/pAmpC production by double-disk synergy test using cefotaxime (30 μg) and ceftazidime (30 μg) discs with and without clavulanic acid (10 μg) and according to CLSI guidelines [23]. Additionally, a cefoxitin disc (30 μg) was used to detect potential AmpC-producers.

Molecular characterisation

ESBL/pAmpC gene groups [24] and E. coli phylogroups [25] were detected by multiplex PCRs for all phenotypically resistant isolates. For a selection of 119 isolates, ESBL/AmpC genes were sequenced (Macrogen, Spain) after amplification with the primers described by Dierikx et al. [26] to identify gene variants. This selection was done considering the variability of ESBL/pAmpC genes and E. coli groups per sampling (at least one isolate per phylogroup-bla gene combination per sampling). Moreover, isolates with an AmpC phenotype, but negative for pAmpC genes by multiplex PCR, were analysed for chromosomal mutations in the ampC promoter/attenuator (cAmpC) according to Haldorsen et al. [27].

Data analysis

Prevalence of ESBL/pAmpC E. coli for each production stage was calculated based upon faecal samples or carcasses being positive to ESBL/pAmpC E. coli (at least one isolate). Prevalence estimates and corresponding 95% confidence intervals (95% CI) were adjusted for clustering of observations at the chain and farm levels using cluster-robust standard errors. Chi-square statistic for trends in proportions was used to test the significance of trends in the relative frequencies of different ESBL/pAmpC gene groups, gene variants, and phylogroups over subsequent sampling stages across the whole broiler production pyramid. Statistical analysis was performed using STATA (StataCorp, College Station, USA).

To quantify the possible transfer of ESBL/pAmpC E. coli across the broiler production pyramid, the average pairwise overlap of ESBL/pAmpC genes and E. coli gene groups (M1), gene groups plus gene variants (M2), or gene groups plus gene variants plus phylogroups (M3), between subsequent production stages, was estimated based on the proportional similarity index (PSI) [28], accounting for uncertainty in measurements in a Monte Carlo simulation setting. PSI values range from 0 (no similarities) to 1 (total overlap). The PSI is expressed as:

PSI=(1−0.5∑|rj,x,y,k−rj,x,y,k+1|)

where r denoted the relative frequency of ESBL/pAmpC gene group (M1), ESBL/pAmpC gene variant (as additional strata within gene group) (M2), or E. coli phylogroup (as additional strata within the gene variant, which in turn were strata of the respective gene group) (M3) j (with j = 1,…, J; and J = 7 gene groups, J = 10 gene groups plus gene variants, and J = 7 gene groups plus gene variants plus phylogroups), in farm x (with x = 1,…, X; and X = 4) of chain y (with y = 1,…, Y; and Y = 3) at sampling stage (i.e. production level) k (with k = 1,…, K; and K = 5). The average of the PSIs calculated over the three chains and four farms per chain gave the overall measure of overlap between sampling stages, as follows:

1n∑i=1nPSIi=PSI1+PSI2+…+PSInn

Uncertainty was introduced in the frequencies of M1, M2 and M3 by assuming the following probability distribution:

(r1,x,y,k,r2,x,y,k,…,1−∑j=1J−1rj,x,y,k)≈Dirichlet(X1,x,y,k,X2,x,y,k,…,XJ,x,y,k)

Where r is defined as above, and X is the number of detections of ESBL/pAmpC gene group/gene variant/phylogroup j in farm x of chain y at sampling stage k.

This analysis was performed in @RISK (Palisade Corp., USA) by setting 10,000 iterations with the Latin hypercube sampling technique and a seed of 1.

Results

Prevalence of ESBL/pAmpC E. coli

In total, 820 samples (i.e. 20 samples per chain-farm-sampling stage combination) were collected over 40 sampling visits. From these samples, 537 confirmed E. coli isolates from CTX-EMB were screened for phenotypic resistance to ESCs by disk diffusion and 513 (95.5%) thereof were positive. Overall, 60.3% (95%CI 51.3–68.1%) of samples were positive for presence of ESBL/pAmpC E. coli (Fig 1). ESBL/pAmpC E. coli were recovered at all sampling events on farms (cloacal swabs) and slaughterhouse (carcasses). A high prevalence (92.5%, 95%CI 72.1–98.3%) was found in day-old PS chicks, which dropped to 20% (95%CI, 12.9–29.6%) during the laying period. In fattening broilers, prevalence was higher again (69.2%, 95%CI 53.6–81.3%) at the start of the production cycle, and decreased to 54.2% (95%CI 38.9–68.6%) in the last sampling right before slaughter (Fig 1). At the end of the production pyramid, ESBL/pAmpC E. coli was isolated from 61.3% (95% CI 48.1–72.9%) of carcasses (Fig 1). In 29.2% of carcasses (n = 43 samples) positive for presence of ESBL/pAmpC E. coli by the qualitative method, the load was below the quantification limit (1 Log CFU/mL). In samples with countable ESBL/pAmpC E. coli the median load was 1.66 Log CFU/1 mL rinsing water (min 1 Log CFU/mL, max 4.2 Log CFU/mL).

Fig 1. Prevalence of ESBL/pAmpC E. coli in the broiler production pyramid.

Fig 1

Open in a new tab

From left to right: PS chicks, PS breeders, broiler chicks, broilers and carcasses. Bars represent corresponding 95% confidence intervals.

ESBL/pAmpC genes and phylogroup distribution

Out of the 513 ESC-resistant isolates, 333 (64.9%) had the ESBL and 180 (35.1%) the AmpC phenotype. By multiplex PCR, the following ESBL gene groups were identified: blaCTX-M-group-1 (41.9% of all resistant isolates), blaSHV (17.6%), blaCTX-M-group-9 (2.9%), blaCTX-M-group-2 (2.5%) and blaTEM (0.2%). All isolates with an AmpC phenotype were carrying blaCMY (25.6%), except for 18 isolates in which the following mutations were discovered in the ampC gene control region: −88 (C→T), −82 (A→G), −42 (C→T), −18 (G→A), −1 (C→T) and +58 (C→T). No ESBL/pAmpC genes were found for 28 ESC-resistant E. coli. One-hundred-eighty-three isolates with a confirmed ESBL/pAmpC gene had concurrent presence of blaTEM, which upon sequencing (n = 19 isolates) proved to be blaTEM-1b, a broad-spectrum β-lactamase.

Sequencing of ESBL genes revealed variability in blaCTX-M-group-1, which was mainly comprised of blaCTX-M-55 (54.2%) and blaCTX-M-1 (42.4%), whereas only one isolate (1.7%) was found for each of blaCTX-M-15 and blaCTX-M-164 (Fig 2A). Two isolates had concurrent presence of more than one blaCTX-M-group-1 genes and thus unassignable by Sanger sequencing. In contrast, all blaCMY, blaSHV, blaTEM belonged to the blaCMY-2, blaSHV-12 and blaTEM-52 variants, respectively. Moreover, the blaCMY-2 variant was found in all blaCMY-carrying isolates.

Fig 2.

Fig 2

Open in a new tab

Distribution of ESBL/pAmpC genes (A) and E. coli phylogroups (B) in the broiler production pyramid. (A) cAmpC, isolates with chromosomal mutations in the ampC promoter/attenuator. (A) and (B) from left to right: PS chicks, PS breeders, broiler chicks, broilers and carcasses.

Overall, the most prevalent gene variants were blaCTX-M-55 and blaCMY-2 (27% each) followed by blaCTX-M-1 (21%) and blaSHV-12 (14%). Other gene variants were only sporadically present. All (100%) ESBL/pAmpC E. coli in PS chicks carried blaCMY-2 (Fig 2A). The relative frequency of this latter gene variant showed a statistically significant decreasing trend (linear slope = -0.20, SE = 0.017, p<0.0001) over subsequent production stages, but was present throughout the production pyramid. blaCTX-M-55 appeared in PS breeders, remained present in fattening broilers where it peaked at the end of the production cycle (39.6%), and was the second most prevalent gene variant in chicken carcasses (30.7%), after blaCTX-M-1 (35%). The latter was also present in broiler, but not in breeder flocks (Fig 2A). Both blaCTX-M-1 and blaCTX-M-55 showed a significantly increasing trend (linear slope = 0.11, SE = 0.015, p<0.0001, and linear slope = 0.08, SE = 0.017, p<0.0001, respectively) from breeders to carcasses. Furthermore, blaSHV-12-carrying E. coli were isolated from all production stages but PS chicks, and were predominantly isolated from fattening broiler chicks (34.6%).

The most frequent E. coli genotype was phylogroup F (34.9%) followed by A (19.7%), B1 (15.1%) and B2 (12.1%). Phylogroups C, D, and E were each found in less than 10% of isolates. In PS chicks, phylogroup B2 predominated (94.6%), whereas in PS breeders phylogroups F, A, and B1 were equally present (25%) (Fig 2B). In subsequent production levels, phylogroup F was the most prevalent, but other phylogroups were present as well (Fig 2B). Significantly increasing and decreasing trends along the production chain were respectively found for phylogroups F (linear slope = 0.05, SE = 0.01, p = 0.009) and B2 (linear slope = -0.15, SE = 0.01, p<0.0001).

ESBL/pAmpC genes were found in multiple E. coli phylogroups within the same production stage, but within the same sampling stage as well (Fig 3). However, phylogroup F-blaCTX-M-55 was the most frequent (21.1%) isolate profile, especially in fattened broilers (37%) and carcasses (20%). In contrast, phylogroup A- and B2-blaCMY-2 were the most frequent ones in PS chicks (100%) and PS breeders (33.4%) (Fig 3).

Fig 3. Detected isolates by phylogroup/bla-gene combination and corresponding production stages.

Fig 3

Open in a new tab

Size of segments on the right represent the number of isolates with a specific combination. Size of segments on the left represent the number of isolates detected in different production stages. Ribbons connecting left and right segments represent the number of isolates with a specific combination found on the respective production stage. Chord diagram generated with CIRCOS [29].

ESBL/pAmpC E. coli transfer across the production pyramid

Genotyping data were used to calculate PSIs over subsequent sampling stages at three levels of increasing stratification (M1, M2, and M3). The average stepwise transfer of ESBL/pAmpC E. coli over the whole production pyramid was 55.5% (95%CI 48.3–61.3%) for M1, 47.7% (95%CI 42.3–53.4%) for M2, 47.2% (95%CI 44.2–50.3%) for M3. At the top of the production pyramid, 36.4 to 50.4% (depending on the level of genotyping data stratification) of ESBL/pAmpC E. coli found in PS breeders overlapped with (and could therefore be estimated to originate from) PS chicks (Fig 4). The overlap between fattening chicks and fattened broilers varied from 46.5 to 49.1% (M3 to M1), whereas 51.5 to 66.9% (M3 to M1) of ESBL/pAmpC E. coli found in carcasses were likely to originate from broilers sampled before slaughter (Fig 4). In general, the overlap of genotyping data decreased with increasing stratification of these data due to decreased chance of finding one-to-one matches of the same genotype between two sampling stages. Moreover, the overlap increased with decreasing time elapsed between two sampling stages (Fig 4).

Fig 4. Transfer of ESBL/pAmpC E. coli across the production pyramid.

Fig 4

Open in a new tab

BrS1, PS chicks; BrS2, PS breeders; F1-4S1, farm 1–4 broiler chicks; F1-4S2, farm 1–4 broilers; F1-4Sc, farm 1–4 carcasses. Arrows and text boxes indicate the overlap of genotyping data in subsequent sampling stages (bold) at three levels (M1-M3) of increasing stratification. Underlined text indicates the time elapsed between sampling two consecutive stages.

Discussion

This is the first study investigating the occurrence and potential transfer of ESBL/pAmpC E. coli across the whole broiler production pyramid in Italy. Samples predominantly originated from one Italian region, which is however among the most densely populated poultry areas in the country and thus can be considered representative of a large fraction of Italy’s poultry production [30]. Analysis of 820 samples from three broiler production chains resulted in the identification of ESBL/pAmpC E. coli in all farms and batches of carcasses, with an overall prevalence of 60.3%. Day-old PS chicks showed the highest prevalence, comparable with recent findings from the Netherlands [31], but higher than those found in Norway [7]. All ESBL/pAmpC E. coli found in PS chicks had the AmpC phenotype, carried blaCMY-2, and belonged predominantly to phylogroup B2. Therefore, a prevalent E. coli blaCMY-2 lineage seems to have been introduced through the import of day-old PS chicks. However, due to the widespread occurrence of blaCMY-2 in poultry, other sources of entry cannot be excluded [32]. It has been speculated that the underlying reason for high prevalence at this production level is the in ovo use of cephalosporins in supplying (grand)parent hatcheries, which selects for ESC-resistant E. coli in the gut of young hatchlings [8]. This is a problem for Nordic countries supplied by the same breeding stock, as the blaCMY-2 genotype introduced at the top of the production pyramid in these countries spreads clonally to the bottom, is often the only identified ESBL/pAmpC gene [7,33,34]. Similar findings have been described in Denmark [12]. In contrast, we found a significant decrease of blaCMY-2 and a gradual substitution by mainly blaCTX-M-55 and blaCTX-M-1 along the production pyramid, as well as the sharp reduction of prevalence in PS breeders (Fig 1). A similar decrease in the prevalence of blaCMY-2-carrying E. coli from 95% in week 1 to 0% in week 21 has been described by Dame-Korevaar et al. [31] and was attributed to a selective disadvantage of blaCMY-2-carrying plasmids. Although different EBSL/pAmpC genotypes were identified during the laying period, blaCMY-2 was still dominant and the overlap of genotype distributions between the two sampling stages in breeders was substantial despite the relatively long period between them (36.4–50.4%). Persistence of blaCMY-2 has been previously associated with the IncK and IncI1 plasmid types [11].

In broilers at the start of the cycle, 47.9–55.5% of the ESBL/pAmpC E. coli were likely to have been vertically transmitted from their parents. The impact of the hatchery on colonisation of hatchlings with ESC-resistant E. coli was not in the scope of this study. However, the role of the hatchery has proven to be trivial since ESBLs/pAmpCs found in the hatchery’s environment (e.g. dust and incubators) were different than those recovered from the egg surface and hatchlings [20,35]. As Projahn et al. [14] showed, true vertical transmission (infection during egg formation) is rare and hatchlings seem to get colonised through a mode of pseudo-vertical transmission with contaminated eggs that enter the hatchery. Nonetheless, the finding that at least 43.3% of ESBLs/pAmpCs recovered in broiler chicks originated from their parents supports the notion of a combination of vertical transmission and environmental contamination from the farm environment, together with rapid proliferation of resistant bacteria, resulting into high prevalence (69.2%) and diversity of ESBL/pAmpC E. coli in young broilers [36]. The number of positive faecal samples was 15% lower in fattened broilers, which can be explained by the development of immunity to certain (resistant) E. coli genotypes [37]. However, high transmission ratios and repeated shifts in phylogroups of ESBL/pAmpC E. coli during the fattening period, such as the ones observed in our study (Fig 3) lead to persistence of ESC-resistant E. coli in broiler flocks during the fattening period [37,38]. The overall transfer of EBSL/pAmpC-EC from the start to the end of the broiler production cycle was, on average, 46.5–49.1%, which indicates substantial influence of other sources, such as the environment and the effect of previously fattened flocks, on the genotype landscape of ESBL/pAmpC E. coli [15]. This has been documented by, e.g., the finding of blaCTX-M-2 being only present in fattened broilers, but not in fattening chicks (Fig 2A). The high levels of resistance found are difficult to explain since no cephalosporins were administered in the investigated flocks, but the use of amoxicillin, which was frequent in our study (S1 Fig), could have selected for ESBL/pAmpC E. coli [7].

At the end of the production pyramid, a large proportion of carcasses (61.3%) were found to be contaminated with ESBL/pAmpC E. coli, which is comparable with prevalence found in previous studies [16,39,40], although the load of resistant E. coli was lower in our study. This is a worrying finding since contamination of meat with ESBL/pAmpC E. coli is directly linked with human exposure [21]. The contribution of chicken meat as a source of human ESBL-pAmpC E. coli infections is debatable [19,20,41], however, the finding of phylogenetically related strains with the same gene and plasmid profile shared between patients and poultry meat [17,42–44] warrants for measures to prevent the entrance of ESBL/pAmpC E. coli in the lower levels of the food chain. The highest percentage of transfer of ESBL/pAmpC E. coli between fattened broilers and carcasses points to contamination of carcasses during processing at the slaughterhouse [39], but the time elapsed between these two sampling stages was the shortest (Fig 4). However, results at this production stage need to be interpreted with caution as cross-contamination from previously slaughtered flocks via other routes (e.g. scalding water, defeathering machines, transportation crates) is frequent [16,39,45].

Our study proved that the genetic background of ESBL/pAmpC E. coli in broilers is complex in all production stages, except perhaps for PS chicks, which isolates belonged predominantly to phylogroup B2-blaCMY-2 (Figs 2 and 3). In subsequent production stages, we found common (e.g. blaCTX-M-1, blaSHV-12) and less common (e.g. blaCTX-M-65, blaCTX-M-2) poultry-related ESBL/AmpC genes [8] to be associated with more than one phylogroup (Fig 3), even within the same farm (data not shown), suggesting that horizontal gene transfer contributes to the dissemination of resistance genes, as previously discussed [13]. In contrast, persistence of phylogroup F-blaCTX-M-55 in significant proportions of isolates from breeders to the slaughterhouse is indicative of clonal transmission (Fig 3). An unexpected discovery was the occurrence of blaCTX-M-55 in all production stages except for PS chicks, especially in fattened broilers where it was the most dominant gene (Fig 2A). This is an ESBL gene usually found in food-producing animals and humans in Asia [46] and has been rarely described in poultry in Europe [18,32]. Regarding E. coli genotypes, from PS chicks we observed a significant reduction of the predominant phylogroup B2, which was mainly substituted by group F in broilers and carcasses (Fig 2B). Both phylogroups are associated with extra-intestinal pathogenicity, as they often harbour virulence genes not found in other E. coli genotypes [47]. However, isolates belonging to commensal groups (A, B1, and C) were substantially present as well (37% of all isolates) (Fig 2B).

In conclusion, ESBL/pAmpC E. coli were detectable at all levels of the broiler production pyramid. The highest prevalence was observed in imported PS chicks, with almost all samples being positive for ESBL/pAmpC E. coli, particularly to blaCMY-2. Measures to ensure ESBL-free pedigree flocks in supplying countries and avoid introduction of resistant clones by the purchase of (grand)parent stock is therefore warranted [12,18]. Based on the similarity of ESBL/pAmpC E. coli genotype distributions between subsequent production levels, we showed that the transfer of ESBL/pAmpC E. coli across the whole production pyramid is likely to be substantial, with approximately half of the genotypes found in a given production stage being likely to originate from the previous stage, and with the time between sampling stages and discriminatory power of genotyping data having an effect as well. Interventions like competitive exclusion [38] may help to control dissemination of ESBL/pAmpC E. coli. Mitigation strategies should definitely include biosecurity and disinfection measures to avoid colonisation from environmental sources [15,21]. Prevention of cross-contamination in the slaughterhouse processing line is also crucial, as the load of ESBL/pAmpC E. coli on meat defines human exposure.

Supporting information

S1 Table. Information on the three sampled broiler production chains.

(DOCX)

Click here for additional data file. (14.4KB, docx)
S1 Fig. Hierarchical structure of sampled farms.

BrS1, PS chicks; BrS2, PS breeders; F1-4S1, farm 1–4 broiler chicks; F1-4S2, farm 1–4 broilers; F1-4Sc, farm 1–4 carcasses. amx, amoxicillin; enr, enrofloxacin; oxtet, oxytetracycline; dox, doxycycline. Samples were not collected from PS chicks of chain B.

(TIF)

Click here for additional data file. (755.4KB, tif)
S1 File. Raw data file.

Excel file containing the raw data of this study.

(XLSX)

Click here for additional data file. (45.8KB, xlsx)

Acknowledgments

The Authors wish to thank Matteo Cuccato, Jacopo Ferraresso, Roberto Rizzo and Massimo Zago for help in sample collection and processing, and Panagiotis Mallioris for help in data organization.

Data Availability

All relevant data are within the manuscript and its Supporting Information files.

Funding Statement

This work was supported by the University of Padua (Fondi Investimento Strategico di Dipartimento (SID), Anno 2016 - prot. BIRD167540) to AP.

References

  • 1.OIE. List of antimicrobial agents of veterinary importance. Paris: World Organisation for Animal Health; 2015. [Google Scholar]
  • 2.WHO. Critically important antimicrobials for human medicine– 5th rev. Geneva: World Health Organization; 2017. [Google Scholar]
  • 3.Davies J, Davies D. Origins and evolution of antibiotic resistance. Microbiol Mol Biol Rev. 2010;74(3): 417–33. 10.1128/MMBR.00016-10 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.van Duijkeren E, Schink A-K, Roberts MC, Wang Y, Schwarz S. Mechanisms of bacterial resistance to antimicrobial agents. Microbiol Spectr. 2018;6(1): 51–82. 10.1128/microbiolspec.ARBA-0019-2017 [DOI] [PubMed] [Google Scholar]
  • 5.Brolund A, Sandegren L. Characterization of ESBL disseminating plasmids. Infect Dis (Auckl). 2016;48(1): 18–25. 10.3109/23744235.2015.1062536 [DOI] [PubMed] [Google Scholar]
  • 6.European Commission. Commission implementing decision C(2012) 182 of 13 January 2012 following the referral procedure by the European Medicines Agency’s Committee for Medicinal Products for Veterinary Use. 2012;
  • 7.Mo SS, Norström M, Slettemeås JS, Løvland A, Urdahl AM, Sunde M. Emergence of AmpC-producing Escherichia coli in the broiler production chain in a country with a low antimicrobial usage profile. Vet Microbiol. 2014;171(3–4): 315–320. 10.1016/j.vetmic.2014.02.002 [DOI] [PubMed] [Google Scholar]
  • 8.Baron S, Jouy E, Larvor E, Eono F, Bougeard S, Kempf I. Impact of third-generation-cephalosporin administration in hatcheries on fecal Escherichia coli antimicrobial resistance in broilers and layers. Antimicrob Agents Chemother. 2014;58(9): 5428–5434. 10.1128/AAC.03106-14 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Blanc V, Mesa R, Saco M, Lavilla S, Prats G, Miró E, et al. ESBL- and plasmidic class C β-lactamase-producing E. coli strains isolated from poultry, pig and rabbit farms. Vet Microbiol. 2006;118(3–4): 299–304. 10.1016/j.vetmic.2006.08.002 [DOI] [PubMed] [Google Scholar]
  • 10.Laube H, Friese A, von Salviati C, Guerra B, Käsbohrer A, Kreienbrock L, et al. Longitudinal monitoring of extended-spectrum-beta-lactamase/AmpC-producing Escherichia coli at German broiler chicken fattening farms. Appl Environ Microbiol. 2013;79(16): 4815–20. 10.1128/AEM.00856-13 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Dierikx C, van der Goot J, Fabri T, van Essen-Zandbergen A, Smith H, Mevius D. Extended-spectrum-β-lactamase- and AmpC-β-lactamase-producing Escherichia coli in Dutch broilers and broiler farmers. J Antimicrob Chemother. 2013;68(1): 60–7. 10.1093/jac/dks349 [DOI] [PubMed] [Google Scholar]
  • 12.Agersø Y, Jensen JD, Hasman H, Pedersen K. Spread of extended spectrum cephalosporinase-producing Escherichia coli clones and plasmids from parent animals to broilers and to broiler meat in a production without use of cephalosporins. Foodborne Pathog Dis. 2014;11(9): 740–746. 10.1089/fpd.2014.1742 [DOI] [PubMed] [Google Scholar]
  • 13.Dierikx CM, van der Goot JA, Smith HE, Kant A, Mevius DJ. Presence of ESBL/AmpC-producing Escherichia coli in the broiler production pyramid: a descriptive study. PLoS One. 2013;8(11): e79005 10.1371/journal.pone.0079005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Projahn M, Daehre K, Roesler U, Friese A. Extended-spectrum-beta-lactamase- and plasmid-encoded cephamycinase-producing Enterobacteria in the broiler hatchery as a potential mode of pseudo-vertical transmission. Appl Environ Microbiol. 2017;83(1): AEM.02364-16 10.1128/AEM.02364-16 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Daehre K, Projahn M, Semmler T, Roesler U, Friese A. Extended-spectrum beta-lactamase-/AmpC beta-lactamase-producing Enterobacteriaceae in broiler farms: Transmission dynamics at farm level. Microb Drug Resist. 2018;24(4): 511–518. 10.1089/mdr.2017.0150 [DOI] [PubMed] [Google Scholar]
  • 16.von Tippelskirch P, Gölz G, Projahn M, Daehre K, Friese A, Roesler U, et al. Prevalence and quantitative analysis of ESBL and AmpC beta-lactamase producing Enterobacteriaceae in broiler chicken during slaughter in Germany. Int J Food Microbiol. 2018;281(March): 82–89. 10.1016/j.ijfoodmicro.2018.05.022 [DOI] [PubMed] [Google Scholar]
  • 17.Voets GM, Fluit AC, Scharringa J, Schapendonk C, van den Munckhof T, Leverstein-van Hall MA, et al. Identical plasmid AmpC beta-lactamase genes and plasmid types in E. coli isolates from patients and poultry meat in the Netherlands. Int J Food Microbiol. 2013;167(3): 359–362. 10.1016/j.ijfoodmicro.2013.10.001 [DOI] [PubMed] [Google Scholar]
  • 18.Saliu E- M, Vahjen W, Zentek J. Types and prevalence of extended-spectrum beta-lactamase producing Enterobacteriaceae in poultry. Anim Heal Res Rev. 2017;18(1): 46–57. 10.1017/S1466252317000020 [DOI] [PubMed] [Google Scholar]
  • 19.Evers EG, Pielaat A, Smid JH, van Duijkeren E, Vennemann FBC, Wijnands LM, et al. Comparative exposure assessment of ESBL-producing Escherichia coli through meat consumption. PLoS One. 2017;12(1): e0169589 10.1371/journal.pone.0169589 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Lazarus B, Paterson DL, Mollinger JL, Rogers BA. Do human extraintestinal Escherichia coli infections resistant to expanded-spectrum cephalosporins originate from food-producing animals? A systematic review. Clin Infect Dis. 2015;60(3): 439–452. 10.1093/cid/ciu785 [DOI] [PubMed] [Google Scholar]
  • 21.Dame-Korevaar A, Fischer EAJ, van der Goot J, Stegeman A, Mevius D. Transmission routes of ESBL/pAmpC producing bacteria in the broiler production pyramid, a literature review. Prev Vet Med. 2019;162(July 2018): 136–150. 10.1016/j.prevetmed.2018.12.002 [DOI] [PubMed] [Google Scholar]
  • 22.McDaniels AE, Rice EW, Reyes AL, Johnson CH, Haugland RA, Stelma GN. Confirmational identification of Escherichia coli, a comparison of genotypic and phenotypic assays for glutamate decarboxylase and beta-d-Glucuronidase. Appl Environ Microbiol. 1998;64(10): 4113 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.CLSI. Performance standards for antimicrobial susceptibility testing CLSI supplement M100. 27th ed Wayne, PA: Clinical and Laboratory Standards Institute; 2018. [Google Scholar]
  • 24.Dallenne C, Da Costa A, Decré D, Favier C, Arlet G. Development of a set of multiplex PCR assays for the detection of genes encoding important β-lactamases in Enterobacteriaceae. J Antimicrob Chemother. 2010;65(3): 490–495. 10.1093/jac/dkp498 [DOI] [PubMed] [Google Scholar]
  • 25.Clermont O, Christenson JK, Denamur E, Gordon DM. The Clermont Escherichia coli phylo-typing method revisited: improvement of specificity and detection of new phylo-groups. Environ Microbiol Rep. 2013;5(1): 58–65. 10.1111/1758-2229.12019 [DOI] [PubMed] [Google Scholar]
  • 26.Dierikx CM, van Duijkeren E, Schoormans AHW, van Essen-Zandbergen A, Veldman K, Kant A, et al. Occurrence and characteristics of extended-spectrum-β-lactamase- and AmpC-producing clinical isolates derived from companion animals and horses. J Antimicrob Chemother. 2012;67(6): 1368–1374. 10.1093/jac/dks049 [DOI] [PubMed] [Google Scholar]
  • 27.Haldorsen B, Aasnaes B, Dahl KH, Hanssen A-M, Simonsen GS, Walsh TR, et al. The AmpC phenotype in Norwegian clinical isolates of Escherichia coli is associated with an acquired ISEcp1-like ampC element or hyperproduction of the endogenous AmpC. J Antimicrob Chemother. 2008;62(4): 694–702. 10.1093/jac/dkn257 [DOI] [PubMed] [Google Scholar]
  • 28.Garrett N, Devane ML, Hudson JA, Nicol C, Ball A, Klena JD, et al. Statistical comparison of Campylobacter jejuni subtypes from human cases and environmental sources. J Appl Microbiol. 2007;103(6): 2113–21. 10.1111/j.1365-2672.2007.03437.x [DOI] [PubMed] [Google Scholar]
  • 29.Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, et al. Circos: An information aesthetic for comparative genomics. Genome Res. 2009;19(9): 1639–1645. 10.1101/gr.092759.109 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Niero G, Bortolaia V, Vanni M, Intorre L, Guardabassi L, Piccirillo A. High diversity of genes and plasmids encoding resistance to third-generation cephalosporins and quinolones in clinical Escherichia coli from commercial poultry flocks in Italy. Vet Microbiol. 2018;216: 93–98. 10.1016/j.vetmic.2018.02.012 [DOI] [PubMed] [Google Scholar]
  • 31.Dame-Korevaar A, Fischer EAJ, Stegeman A, Mevius D, van Essen-Zandbergen A, Velkers F, et al. Dynamics of CMY-2 producing E. coli in a broiler parent flock. Vet Microbiol. 2017;203: 211–214. 10.1016/j.vetmic.2017.03.024 [DOI] [PubMed] [Google Scholar]
  • 32.Ewers C, Bethe A, Semmler T, Guenther S, Wieler LH. Extended-spectrum β-lactamase-producing and AmpC-producing Escherichia coli from livestock and companion animals, and their putative impact on public health: a global perspective. Clin Microbiol Infect. 2012;18(7): 646–55. 10.1111/j.1469-0691.2012.03850.x [DOI] [PubMed] [Google Scholar]
  • 33.Myrenås M, Slettemeås JS, Thorsteinsdottir TR, Bengtsson B, Börjesson S, Nilsson O, et al. Clonal spread of Escherichia coli resistant to cephalosporins and quinolones in the Nordic broiler production. Vet Microbiol. 2018;213(December): 123–128. 10.1016/j.vetmic.2017.11.015 [DOI] [PubMed] [Google Scholar]
  • 34.Nilsson O, Borjesson S, Landen A, Bengtsson B. Vertical transmission of Escherichia coli carrying plasmid-mediated AmpC (pAmpC) through the broiler production pyramid. J Antimicrob Chemother. 2014;69(6): 1497–1500. 10.1093/jac/dku030 [DOI] [PubMed] [Google Scholar]
  • 35.Osman KM, Kappell AD, Elhadidy M, ElMougy F, El-Ghany WAA, Orabi A, et al. Poultry hatcheries as potential reservoirs for antimicrobial-resistant Escherichia coli: A risk to public health and food safety. Sci Rep. 2018;8(1): 5859 10.1038/s41598-018-23962-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Dierikx CM, van der Goot J, van Essen-Zandbergen A, Mevius DJ. Dynamics of cefotaxime resistant Escherichia coli in broilers in the first week of life. Vet Microbiol. 2018;222(June): 64–68. 10.1016/j.vetmic.2018.07.001 [DOI] [PubMed] [Google Scholar]
  • 37.Huijbers PMC, Graat EAM, van Hoek AHAM, Veenman C, de Jong MCM, van Duijkeren E. Transmission dynamics of extended-spectrum β-lactamase and AmpC β-lactamase-producing Escherichia coli in a broiler flock without antibiotic use. Prev Vet Med. 2016;131: 12–19. 10.1016/j.prevetmed.2016.07.001 [DOI] [PubMed] [Google Scholar]
  • 38.Ceccarelli D, van Essen-Zandbergen A, Smid B, Veldman KT, Boender GJ, Fischer EAJ, et al. Competitive exclusion reduces transmission and excretion of extended-spectrum-β-lactamase-producing Escherichia coli in broilers. Schaffner DW, editor. Appl Environ Microbiol. 2017;83(11): 1–13. 10.1128/AEM.03439-16 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Pacholewicz E, Liakopoulos A, Swart A, Gortemaker B, Dierikx C, Havelaar A, et al. Reduction of extended-spectrum-β-lactamase- and AmpC-β-lactamase-producing Escherichia coli through processing in two broiler chicken slaughterhouses. Int J Food Microbiol. 2015;215: 57–63. 10.1016/j.ijfoodmicro.2015.08.010 [DOI] [PubMed] [Google Scholar]
  • 40.Reich F, Schill F, Atanassova V, Klein G. Quantification of ESBL-Escherichia coli on broiler carcasses after slaughtering in Germany. Food Microbiol. 2016;54: 1–5. 10.1016/j.fm.2015.10.020 [DOI] [Google Scholar]
  • 41.Carmo LP, Nielsen LR, da Costa PM, Alban L. Exposure assessment of extended-spectrum beta-lactamases/AmpC beta-lactamases-producing Escherichia coli in meat in Denmark. Infect Ecol Epidemiol. 2014;4(1): 22924 10.3402/iee.v4.22924 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Overdevest I, Willemsen I, Rijnsburger M, Eustace A, Xu L, Hawkey P, et al. Extended-spectrum β-lactamase genes of Escherichia coli in chicken meat and humans, The Netherlands. Emerg Infect Dis. 2011;17(7): 1216–22. 10.3201/eid1707.110209 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Leverstein-van Hall MA, Dierikx CM, Stuart JC, Voets GM, van den Munckhof MP, van Essen-Zandbergen A, et al. Dutch patients, retail chicken meat and poultry share the same ESBL genes, plasmids and strains. Clin Microbiol Infect. 2011;17(6): 873–880. 10.1111/j.1469-0691.2011.03497.x [DOI] [PubMed] [Google Scholar]
  • 44.Berg ES, Wester AL, Ahrenfeldt J, Mo SS, Slettemeås JS, Steinbakk M, et al. Norwegian patients and retail chicken meat share cephalosporin-resistant Escherichia coli and IncK/blaCMY-2 resistance plasmids. Clin Microbiol Infect. 2017;23(6): 407.e9–407.e15. 10.1016/j.cmi.2016.12.035 [DOI] [PubMed] [Google Scholar]
  • 45.Projahn M, von Tippelskirch P, Semmler T, Guenther S, Alter T, Roesler U. Contamination of chicken meat with extended-spectrum beta-lactamase producing- Klebsiella pneumoniae and Escherichia coli during scalding and defeathering of broiler carcasses. Food Microbiol. 2019;77: 185–191. 10.1016/j.fm.2018.09.010 [DOI] [PubMed] [Google Scholar]
  • 46.Bevan ER, Jones AM, Hawkey PM. Global epidemiology of CTX-M β-lactamases: temporal and geographical shifts in genotype. J Antimicrob Chemother. 2017;72(8): 2145–2155. 10.1093/jac/dkx146 [DOI] [PubMed] [Google Scholar]
  • 47.Johnson JR, Johnston BD, Gordon DM. Rapid and specific detection of the Escherichia coli sequence type 648 complex within phylogroup F. J Clin Microbiol. 2017;55(4): 1116–1121. 10.1128/JCM.01949-16 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Table. Information on the three sampled broiler production chains.

(DOCX)

Click here for additional data file. (14.4KB, docx)
S1 Fig. Hierarchical structure of sampled farms.

BrS1, PS chicks; BrS2, PS breeders; F1-4S1, farm 1–4 broiler chicks; F1-4S2, farm 1–4 broilers; F1-4Sc, farm 1–4 carcasses. amx, amoxicillin; enr, enrofloxacin; oxtet, oxytetracycline; dox, doxycycline. Samples were not collected from PS chicks of chain B.

(TIF)

Click here for additional data file. (755.4KB, tif)
S1 File. Raw data file.

Excel file containing the raw data of this study.

(XLSX)

Click here for additional data file. (45.8KB, xlsx)

Data Availability Statement

All relevant data are within the manuscript and its Supporting Information files.

Articles from PLoS ONE are provided here courtesy of PLOS

RESOURCES