Host-specific differences in the contribution of an ESBL IncI1 plasmid to intestinal colonization by Escherichia coli O104:H4

Host-specific differences in the contribution of an ESBL IncI1 plasmid to intestinal colonization... Abstract Objectives To assess stability and contribution of a large ESBL-encoding IncI1 plasmid to intestinal colonization by Escherichia coli O104:H4 in two different mammalian hosts. Methods Specific-pathogen-free 3–4-day-old New Zealand White rabbits and conventionally reared 6-week-old weaned lambs were orally infected with WT E. coli O104:H4 or the ESBL-plasmid-cured derivative, and the recovery of bacteria in intestinal homogenates and faeces monitored over time. Results Carriage of the ESBL plasmid had differing impacts on E. coli O104:H4 colonization of the two experimental hosts. The plasmid-cured strain was recovered at significantly higher levels than WT during late-stage colonization of rabbits, but at lower levels than WT in sheep. Regardless of the animal host, the ESBL plasmid was stably maintained in virtually all in vivo passaged bacteria that were examined. Conclusions These findings suggest that carriage of ESBL plasmids has distinct effects on the host bacterium depending upon the animal species it encounters and demonstrates that, as for E. coli O157:H7, ruminants could represent a potential transmission reservoir. Introduction In 2011, Escherichia coli O104:H4 caused a large outbreak of haemolytic uraemic syndrome centred in northern Germany, in which more than 50 people died.1 Subsequent genotypic and phenotypic analyses demonstrated that this strain was unusual, with features of both enterohaemorrhagic (EHEC) and enteroaggregative E. coli (EAEC) pathotypes.2,3 Typical of all EHEC, the strain contained genes coding for Shiga toxin (Stx), specifically Stx2a, which is associated with severe clinical outcome in patients.4 However, it also adhered to tissue culture cells in a characteristic ‘stacked-brick’ manner, a trait that defines the EAEC pathotype and is associated with aggregative adherence fimbriae, the genes for which are carried on a plasmid (pAA). Unusual for both pathotypes, E. coli O104:H4 harboured a large plasmid carrying genes coding for ESBLs. While several studies have focused on the contribution of pAA to E. coli O104:H4 pathogenesis,5–7 none has considered the significance of the carriage of the ESBL plasmid beyond its immediate impact of limiting antibiotic options. ESBLs are enzymes that confer resistance to many frontline β-lactam antibiotics and are a dominant mechanism of antimicrobial resistance in Gram-negative bacteria.8 Many types of ESBL have been described, but those belonging to the CTX-M type have become widespread in the UK and elsewhere.9–12 Genes encoding ESBLs are most often located on large, conjugative plasmids, in part providing a mechanistic explanation for their spread.8 Bacterial carriage of resistance plasmids, such as those encoding ESBLs, is believed to confer a biological fitness cost to the host bacterium, although experimental evidence to support this is conflicting. In vitro-based growth assays have shown both detrimental13 and no fitness cost14,15 associated with carriage of CTX-M-containing plasmids. However, Schaufler et al.16 concluded that the carriage of the ESBL plasmid was only associated with a fitness cost when bacteria were grown on surfaces rather than as planktonic cultures. In their study, surface-associated plasmid-cured variants of various pathogenic E. coli clonal lineages exhibited changes in fimbriae production, an ability to form biofilms or be motile; surface-associated attributes that may contribute to colonization and persistence in the mammalian intestine. Given the widespread prevalence of ESBL-producing organisms that are also pathogens of animals and/or humans, it is perhaps surprising that more studies to examine their impact in the context of the host have not been performed. Moreover, evidence to indicate whether ruminants act as a reservoir for EAEC17 or Shiga toxin-producing EAEC isolates, such as E. coli O104:H4,18 is lacking, although carriage following experimental infection of calves has recently been demonstrated.19 Herein, we describe the impact of pESBL, the IncI1 plasmid of E. coli O104:H4, on the capacity of the pathogen to colonize the intestine of two different mammalian species. Plasmid carriage hindered late-stage colonization of the intestine of infant rabbits, a model of EHEC-mediated intestinal disease in humans.20 In contrast, E. coli O104:H4 persisted in the intestine and could be detected in faeces of weaned sheep for up to 4 weeks, independent of pESBL carriage. These studies indicate that carriage of a large ESBL plasmid mediated host-specific differences in the persistence of the host bacterium and demonstrate that, as for E. coli O157:H7,21,22 ruminants could represent a potential transmission source. Materials and methods Strains and culture conditions The strains and plasmids used in the study are listed in Table S1 (available as Supplementary data at JAC Online). BL211, an Stx2 deletion mutant of E. coli O104:H4 strain C227–11,5 was used in this study to enable the animal experiments to be performed in the available facilities and reduce any risk of serious infection to research personnel. Bacteria were routinely grown in LB medium or on LB agar plates containing the appropriate antibiotics: 10 mg/L gentamicin; 10 mg/L tetracycline; or 2 mg/L cefotaxime. Construction of the pESBL-cured derivative Strain BL211 harbours three plasmids including a large 88.5 kb ESBL-encoding plasmid that belongs to incompatibility group I1 (IncI1) and carries both the blaTEM-1 and blaCTX-M-15 genes2 (subsequently annotated as blaCTX-M-323). Incompatibility based curing13 was used to rid strain BL211 of this plasmid (originally referred to as pESBL-EA11 and herein called pESBL). Briefly, plasmid pIFM27, a sacB-containing plasmid, which encodes the IncI1 plasmid replication down-regulator RNAI that directly interferes with the replication of IncI plasmids, was introduced into BL211 by electroporation. Transformants were selected on LB agar supplemented with kanamycin (50 mg/L) and were subsequently screened for an inability to grow on cefotaxime-containing LB medium. Curing of plasmid pIFM27 from cefotaxime-susceptible colonies was accomplished by recovery of colonies on LB agar supplemented with 5% sucrose. Subsequently, sucrose-resistant, cefotaxime-susceptible colonies were screened for loss of pESBL and pIFM27 by PCR and plasmid profiling (Figure S1A and B). Loss of pESBL did not impact the ability of the strain (herein called BL320) to grow in vitro in LB medium in single strain growth assays (Figure S1C). PCR analysis of colonies Multiplex PCR was used to assess the chromosomal and plasmid gene content of BL211, BL320 and randomly selected colonies recovered from rabbits or sheep infected with the parental strain BL211. Primer sequences, expected product sizes and reaction conditions are reported in Table S1. Infant rabbit studies All experimental protocols were approved by the local Animal Welfare Ethical Review Body and carried out in accordance with the UK Animals (Scientific Procedures) Act 1986. Time-mated specific pathogen-free adult New Zealand White rabbits were purchased from Harlan Laboratories (Derby, UK) at 2–3 weeks gestation. Following parturition, mixed sex litters were kept together in a nesting box with the lactating doe and housed under standard conditions. Infant rabbit infections were performed on 3–4-day-old pups essentially as described previously.5 Briefly, rabbits were administered ranitidine intraperitoneally (5 mg/kg body weight) and 2 h later ∼1 × 109 cfu bacteria re-suspended in sodium bicarbonate solution (2.5 g of NaHCO3 in 100 mL of H2O) was given via oral gavage. Following challenge, animals were monitored twice daily for signs of intestinal disease. Given that all bacterial strains lacked Stx, clinical signs were scored with the following revised scale: none (no adherent faecal material on fur and intestines appear normal with hard, formed digesta in the distal colon), intestinal disease (no adherent faecal material on fur, but colon contains soft, poorly formed digesta) and diarrhoea (adherent faecal material on fur and colon contains liquid or unformed digesta). Rabbits were euthanized by a Schedule 1 method at either day 3 or 7 post-infection. Tissues were aseptically sampled post mortem for bacterial enumeration. Samples from the distal small intestine, caecum, mid colon, distal colon and stools were homogenized in PBS, serially diluted and plated on to LB agar supplemented with gentamicin and tetracycline. Where no colonies were detected following plating of undiluted tissue homogenates, the number of bacteria recovered was set using the lower limit of detection as a value. Spread plates, which contained well-spaced colonies, were chosen for replica plating to LB agar with and without cefotaxime. Colonies failing to grow in the presence of the antibiotic were subject to multiplex PCR to confirm the loss of blaCTX-M. All infections were performed in at least two independent litters to limit any litter-specific effects. Sheep colonization studies All experimental protocols were approved by the local Animal Welfare Ethical Review Body and carried out in accordance with the UK Animals (Scientific Procedures) Act 1986. Experimentally inoculated, weaned sheep were used as a model of natural ruminant infection as previously described for E. coli O157:H7.24,25 Conventionally reared 6-week-old cross-bred commercial lambs were divided into mixed sex groups of eight animals and housed under biosecure conditions. Prior to challenge, individual sheep were confirmed as free of E. coli O104:H4 by screening faecal samples with an in-house E. coli O104-specific immunomagnetic separation capture assay and an agglutination assay.26 After 1 week acclimation, sheep were orally inoculated with ∼5 × 109 cfu of bacteria (BL211 or BL320) delivered in a volume of 11 mL using a worming gun (Novartis, UK). The inocula were prepared from 16 h aerobically incubated cultures, which were grown in LB broth, pelleted by centrifugation and finally re-suspended in PBS. Faecal samples were collected per rectum from all animals on days 1–14 post-infection and twice weekly thereafter until day 39. On days 4 and 39 post-infection, 3 and 5 animals, respectively, from each group were euthanized and tissue samples (1 g) were collected from the ileum, caecum, spiral colon, rectum and recto-anal junction. Prior to microbiological analysis, faecal and tissue samples were homogenized in buffered peptone water (BPW) at a ratio of 1:10 (w/v) using a vortex (faeces) or an Ystral D-79282 homogenizer (tissues). Ten-fold serial dilutions of the homogenized samples were plated directly on to sorbitol MacConkey agar plates supplemented with tetracycline and gentamicin. If no colonies were observed after overnight incubation, samples were enriched by incubating the BPW homogenates at 37°C for 18 h followed by re-plating to provide a qualitative result. Selected colonies from all faecal and tissue samples were screened by multiplex PCR as described above. Statistical analysis The presence or absence of disease in rabbits was expressed in a contingency table and analysed using Fisher’s exact test. Bacterial count data (cfu/g) were log transformed and differences in the number of WT or cured cells recovered in each intestinal section compared using Student’s t-test. In the sheep infection studies, bacterial count data were log transformed and the total cfu shed over days 1–4 (8 animals per group) or days 5–39 (5 animals per group) were calculated using AUC following the trapezoidal rule (GraphPad Prism, version 5). Differences in the AUC of strains were compared using Student’s t-test. All statistical analysis was performed using GraphPad Prism (version 5). Results Previously, the E. coli O104:H4 outbreak strain was shown to colonize the infant rabbit intestine and cause diarrhoea in a manner that was dependent on Stx, but independent of pAA, the plasmid responsible for mediating aggregative adherence on cultured cells.5 The contribution of the 88.5 kb β-lactamase-encoding plasmid (pESBL) to E. coli O104:H4 pathobiology was not explored. Stable maintenance of pESBL during in vitro growth23 may indicate that pESBL plays an important role in the organism’s biology. To investigate this further, we cured pESBL from the Shiga toxin negative derivative of E. coli O104:H4 and examined its contribution to colonization of rabbit and sheep intestines. pESBL hinders E. coli O104:H4 long-term colonization of infant rabbits Consistent with earlier findings,5 oral infection of infant rabbits with the Shiga-toxin-negative derivative caused few visible signs of disease. Loose stools were detected in 18% (3 of 17) and 11% (2 of 19) of animals infected with the WT and plasmid-cured strain, respectively, between days 2 and 3 post-infection (Table S2). As the majority of animals did not exhibit any manifestations of diarrhoea, we focused on the role of the plasmid in bacterial colonization of the rabbit intestine. The distribution and number of challenge E. coli present in the intestine of rabbits infected with BL211 or BL320 were determined at days 3 and 7 post-infection (Figure 1a–d). Regardless of the infecting strain, there were no differences in the number of BL211 or BL320 cfu recovered from these regions at day 3 post-infection. In contrast, by day 7 post-infection, 1–2 logs fewer BL211 than BL320 were recovered in all regions of the intestine. Specifically, colonization by the parent strain BL211 was significantly reduced compared with the pESBL-cured strain in the ileum (160-fold; P < 0.01) and caecum (215-fold; P < 0.01) and was lower, but did not reach statistical significance, in the colon [mid colon 95-fold (P = 0.07) and distal colon 50-fold (P = 0.13)] of infected rabbits. These findings suggest that carriage of pESBL hinders the longer-term persistence of E. coli O104:H4 particularly in the upper regions of the rabbit intestine. Figure 1. View largeDownload slide Recovery of bacteria in infant rabbits orally infected with E. coli O104:H4 or a derivative lacking pESBL. Concentration (cfu/g) of bacteria recovered at 3 and 7 days post-infection in intestinal homogenates of rabbits infected with the indicated strain (WT BL211 and pESBL-cured strain, BL320). Data points represent individual animals (at day 3: BL211, n = 19 and BL320, n = 17; at day 7: BL211, n = 6 and BL320, n = 9) and the bar represents the geometric mean. Statistical analysis was performed using Student’s t-test with P ≤ 0.05 deemed significant. Figure 1. View largeDownload slide Recovery of bacteria in infant rabbits orally infected with E. coli O104:H4 or a derivative lacking pESBL. Concentration (cfu/g) of bacteria recovered at 3 and 7 days post-infection in intestinal homogenates of rabbits infected with the indicated strain (WT BL211 and pESBL-cured strain, BL320). Data points represent individual animals (at day 3: BL211, n = 19 and BL320, n = 17; at day 7: BL211, n = 6 and BL320, n = 9) and the bar represents the geometric mean. Statistical analysis was performed using Student’s t-test with P ≤ 0.05 deemed significant. To investigate whether pESBL was stably maintained in the WT strain during infection, representative colonies recovered from each animal were replica-plated on to medium supplemented with/without cefotaxime. Loss of ESBL activity was rarely found, even after 7 days growth in the intestine. In all, ∼2.5 × 103 colonies recovered at either day 3 or 7 post-infection were screened for growth on cefotaxime-containing medium and only one colony (recovered at day 3) failed to grow on the antibiotic-containing medium. Loss of blaCTX-M in this colony was confirmed by PCR (see Figure S1A). Together these observations suggest that while pESBL hinders the ability of the host bacterium to persist in the rabbit intestine, the plasmid is stably maintained in the cell. pESBL aids persistence in the ruminant intestine To investigate the role of pESBL in the colonization of the ruminant intestine, we orally infected groups of 6-week-old conventional lambs with BL211 or BL320, and monitored the presence of bacteria up until day 39 post-infection. Owing to the lower number of bacteria usually recovered following E. coli O157:H7 challenge of sheep,27 an additional enrichment step was included when necessary during sample processing in these experiments. As expected, none of the infected lambs showed gross signs of disease after challenge and at post mortem all intestinal tissues appeared normal. Regardless of the infecting strain, most animals shed high numbers of cells (>107 cfu/g) the day after challenge, declining thereafter (Figure 2a). Within each group, some animals continuously shed the challenge bacteria whereas others showed intermittent shedding or only shed for a few days. However, the magnitude and duration of faecal shedding in BL211-infected animals was generally greater than for BL320-infected animals (Figure 2a). The median time before the number of shed bacteria fell below experimental detection limits in two consecutive samples collected from the same animal was 35 (range 4–35) and 18 (range 10–21) days for BL211 and BL320, respectively (Figure 2a, inset). As a result, higher numbers of bacteria were recovered from animals infected with BL211 compared with BL320 in the later stages of the infection. Moreover, bacteria were more likely to be recovered by direct plating rather than following overnight enrichment of faecal samples from animals infected with BL211 versus BL320 (Figure 2b). Despite these trends, the magnitude and duration of shedding when expressed as the AUC for each strain approached, but did not reach, statistical significance (P = 0.08). Thus, the presence of pESBL appears to prolong the duration of E. coli O104:H4 shedding in sheep, albeit with high inter-animal variability. Figure 2. View largeDownload slide Faecal shedding of E. coli O104:H4 or the pESBL-cured derivative from orally infected 6-week-old conventional weaned lambs. Mixed sex lambs (n = 8 per group) were group-housed and individual faecal samples collected per rectum. Samples were homogenized in BPW and plated directly, or after 18 h enrichment, on SMAC medium supplemented with appropriate antibiotics. Count data were log transformed and the mean (±standard error) number of bacteria recovered for each strain was expressed over time (a). Inset shows boxplots representing the median, lower and upper quartiles, and the minimum and maximum duration of shedding for each strain. DL = experimental limit of detection based on average weight of tissue. Proportion of faecal samples in which bacteria were not found or recovered directly or following enrichment (b). Figure 2. View largeDownload slide Faecal shedding of E. coli O104:H4 or the pESBL-cured derivative from orally infected 6-week-old conventional weaned lambs. Mixed sex lambs (n = 8 per group) were group-housed and individual faecal samples collected per rectum. Samples were homogenized in BPW and plated directly, or after 18 h enrichment, on SMAC medium supplemented with appropriate antibiotics. Count data were log transformed and the mean (±standard error) number of bacteria recovered for each strain was expressed over time (a). Inset shows boxplots representing the median, lower and upper quartiles, and the minimum and maximum duration of shedding for each strain. DL = experimental limit of detection based on average weight of tissue. Proportion of faecal samples in which bacteria were not found or recovered directly or following enrichment (b). While E. coli O157:H7 preferentially colonizes the mid to lower intestinal tract of sheep,24,27 the site of E. coli O104:H4 colonization is less well-defined. Thus, samples from different regions of the intestine (ileum, caecum, colon, rectum and anal-rectal junction) were collected from infected animals at day 4 (n = 3) and day 39 (n = 5) post-infection. At day 4 post-infection, low levels of challenge bacteria were recovered from all five sites of the intestine and all animals, regardless of the infecting strain, yielded bacteria from at least two intestinal sites. While more bacteria tended to be recovered from animals infected with the parent strain BL211, the tissue distribution did not differ markedly between the two strains (Table S3) or from E. coli O157:H7.24 At day 39 post-infection, challenge bacteria could no longer be recovered, even with enrichment, from any of the intestinal samples collected, even though the organisms could still be detected in the animals’ faeces. As found during the rabbit challenge experiments, the ESBL-producing plasmid was stably maintained in colonies recovered from the sheep. Consistent with reports from human infection,7 at least one colony was found to lack aggR indicative of pAA loss. Discussion While a considerable amount of work has been performed elucidating the genetic and biochemical basis of ESBL resistance, fewer studies have attempted to uncover the contribution of ESBL-containing plasmids to pathogen biology, particularly in the context of bacterial survival and carriage in animal hosts. By generating a plasmid-cured derivative of Stx-negative ESBL-producing E. coli O104:H4, we were able to assess the impact of the carriage of the ESBL plasmid in two complementary animal hosts: infant rabbits, which are used as a model of Stx-mediated intestinal disease,5,20 and weaned sheep, which are a potential ruminant reservoir host of Stx-producing isolates.28 We found that carriage of pESBL affected the fitness of the bacterium in the intestine of the two experimental hosts, with the cured strain being recovered at higher levels than WT in rabbits, but at lower levels (than WT) in sheep. Our findings challenge the idea that fitness costs are consistent across different assays as reported previously.29 While Vogwill and MacLean29 found a significant correlation between fitness scores assessed using in vitro and in vivo assays, the impact of the host was not fully ascertained in their analyses. All the in vivo studies examined in their study were performed in a single host species, mice. Our findings highlight the importance of the host context when considering the impact of resistance plasmids on the fitness of the bacterium during intestinal carriage. Species-specific differences in the host can mitigate the requirement for particular bacterial factors and may explain the differing impact of pESBL on the host bacterium in this study. For example, gltA, encoding a citrate synthase and mtlD, encoding a mannitol metabolic protein, are important for effective Vibrio cholerae colonization of the infant rabbit intestine, but not the infant mouse intestine.30 Differences in the availability of carbon and energy sources within the two animal intestines were hypothesized to explain the relative necessity of these genes. Likewise, differences between the rabbit and sheep gastrointestinal tracts could impact the contribution of plasmid-borne factors on BL211 fitness. In addition to the ESBL genes, which are unlikely to confer a direct selective advantage in our experiments, pESBL contains 95 genes coding for factors of unknown and known function. One of these is Hha, a haemolysin expression-modulating protein that appears to affect a myriad of surface-associated cellular phenotypes in E. coli including bacterial motility, cell aggregation and biofilm formation.31 Loss of Hha in laboratory K12 and some ESBL-producing E. coli isolates led to increased swimming activity and reduced biofilm formation compared with their parental strains;16 phenotypes that may alter the ability of the organism to colonize and persist in the intestine. The late-stage colonization advantage of the pESBL-cured strain was not evident in sheep, a ruminant host. Instead, it appeared that loss of pESBL resulted in more rapid clearance of the cured strain from the intestine than WT (Figure 2). Only detailed molecular analyses involving deletion mutants and further in vivo experimentation will uncover the factors and/or mechanisms that explain these fitness outcomes. Like the prototypical E. coli O157:H7 serotype, we found that E. coli O104:H4 was able to persist in the ruminant intestine for at least 4 weeks. These findings are consistent with those recently reported following experimental infection of E. coli O104:H4 in weaned calves.19 Finally, we found that pESBL was stably maintained during infection, as most cells recovered from the rabbit or sheep intestine retained the ability to grow on cefotaxime-containing medium. Similar observations were reported for calves infected with E. coli O104:H4,19 for pigs infected with E. coli carrying an IncI/ST12 blaCTX-M-encoding plasmid32 and, most recently, in streptomycin-treated mice infected with E. coli carrying a large non-conjugative virulence plasmid.33 As noted by others, why these large plasmids are stably maintained in the host bacterium in the absence of obvious selection pressure is intriguing. Yamaichi et al.23 identified six regions of pESBL as essential for pESBL replication or segregation. Intriguingly one of these regions mapped to hha. Thus, as well as hha potentially affecting the surface-expressed properties of the host bacterium, its presence may also help to ensure that pESBL is stably maintained in the cell during intestinal growth. Unlike some other IncI ESBL plasmids circulating in the UK,10 pESBL also appears to contain a recognizable plasmid addiction system (PndAC). While not identified as playing a role in plasmid maintenance in E. coli O104:H4,23 PndAC has been found to play a role in the maintenance of other plasmids.34,35 Collectively our studies suggest that the impact of plasmids bearing antibiotic resistance determinants on pathogen biology needs to be understood in the context of the host. For zoonotic food-borne pathogens, this should include intestinal environments that are healthy (i.e. reservoir hosts) or diseased, as host factors, such as inflammatory cells,36 are known to affect antimicrobial resistance (AMR) transmission and pathogen survival. Finally, our studies show that ruminants, such as sheep, can act as reservoir hosts for E. coli O104:H4 and thus they should be considered as a potential source of transmission to humans. Acknowledgements We would like to acknowledge the assistance of the Animal Services Unit staff at the Animal and Plant Health Agency (Weybridge, UK) and the Experimental Biology Unit staff at the University of Surrey (Guildford, UK). We also thank Irene Freire Martin for providing the IncI curing vector. Funding This work was supported by the European Union 7th Framework Programme grant entitled ‘Anticipating the global onset of novel epidemics’ (project number 278976). Transparency declarations None to declare. Supplementary data Tables S1–S3 and Figure S1 are available as Supplementary data at JAC Online. References 1 Buchholz U , Bernard H , Werber D et al. German outbreak of Escherichia coli O104:H4 associated with sprouts . N Engl J Med 2011 ; 365 : 1763 – 70 . Google Scholar CrossRef Search ADS PubMed 2 Rasko DA , Webster DR , Sahl JW et al. Origins of the E. coli strain causing an outbreak of hemolytic-uremic syndrome in Germany . N Engl J Med 2011 ; 365 : 709 – 17 . Google Scholar CrossRef Search ADS PubMed 3 Bielaszewska M , Mellmann A , Zhang W et al. Characterisation of the Escherichia coli strain associated with an outbreak of haemolytic uraemic syndrome in Germany, 2011: a microbiological study . Lancet Infect Dis 2011 ; 11 : 671 – 6 . Google Scholar CrossRef Search ADS PubMed 4 Persson S , Olsen KE , Ethelberg S et al. Subtyping method for Escherichia coli shiga toxin (verocytotoxin) 2 variants and correlations to clinical manifestations . J Clin Microbiol 2007 ; 45 : 2020 – 4 . Google Scholar CrossRef Search ADS PubMed 5 Munera D , Ritchie JM , Hatzios SK et al. Autotransporters but not pAA are critical for rabbit colonization by Shiga toxin-producing Escherichia coli O104:H4 . Nat Commun 2014 ; 5 : 3080 . Google Scholar CrossRef Search ADS PubMed 6 Boisen N , Hansen AM , Melton-Celsa AR et al. The presence of the pAA plasmid in the German O104:H4 Shiga toxin type 2a (Stx2a)-producing enteroaggregative Escherichia coli strain promotes the translocation of Stx2a across an epithelial cell monolayer . J Infect Dis 2014 ; 210 : 1909 – 19 . Google Scholar CrossRef Search ADS PubMed 7 Zhang W , Bielaszewska M , Kunsmann L et al. Lability of the pAA virulence plasmid in Escherichia coli O104:H4: implications for virulence in humans . PLoS One 2013 ; 8 : e66717 . Google Scholar CrossRef Search ADS PubMed 8 Blair JM , Webber MA , Baylay AJ et al. Molecular mechanisms of antibiotic resistance . Nat Rev Microbiol 2015 ; 13 : 42 – 51 . Google Scholar CrossRef Search ADS PubMed 9 Day MJ , Rodríguez I , van Essen-Zandbergen A et al. Diversity of STs, plasmids and ESBL genes among Escherichia coli from humans, animals and food in Germany, the Netherlands and the UK . J Antimicrob Chemother 2016 ; 71 : 1178 – 82 . Google Scholar CrossRef Search ADS PubMed 10 Doumith M , Dhanji H , Ellington MJ et al. Characterization of plasmids encoding extended-spectrum β-lactamases and their addiction systems circulating among Escherichia coli clinical isolates in the UK . J Antimicrob Chemother 2012 ; 67 : 878 – 85 . Google Scholar CrossRef Search ADS PubMed 11 Fischer J , Rodríguez I , Baumann B et al. blaCTX-M-15-carrying Escherichia coli and Salmonella isolates from livestock and food in Germany . J Antimicrob Chemother 2014 ; 69 : 2951 – 8 . Google Scholar CrossRef Search ADS PubMed 12 Livermore DM , Canton R , Gniadkowski M et al. CTX-M: changing the face of ESBLs in Europe . J Antimicrob Chemother 2007 ; 59 : 165 – 74 . Google Scholar CrossRef Search ADS PubMed 13 Freire MI , Thomas CM , Laing E et al. Curing vector for IncI1 plasmids and its use to provide evidence for a metabolic burden of IncI1 CTX-M-1 plasmid pIFM3791 on Klebsiella pneumoniae . J Med Microbiol 2016 ; 65 : 611 – 8 . Google Scholar CrossRef Search ADS PubMed 14 Fischer EA , Dierikx CM , van Essen-Zandbergen A et al. The IncI1 plasmid carrying the blaCTX-M-1 gene persists in in vitro culture of a Escherichia coli strain from broilers . BMC Microbiol 2014 ; 14 : 77 . Google Scholar CrossRef Search ADS PubMed 15 Cottell JL , Webber MA , Piddock LJ. Persistence of transferable extended-spectrum-β-lactamase resistance in the absence of antibiotic pressure . Antimicrob Agents Chemother 2012 ; 56 : 4703 – 6 . Google Scholar CrossRef Search ADS PubMed 16 Schaufler K , Semmler T , Pickard DJ et al. Carriage of extended-spectrum β-lactamase-plasmids does not reduce fitness but enhances virulence in some strains of pandemic E. coli lineages . Front Microbiol 2016 ; 7 : 336 . Google Scholar CrossRef Search ADS PubMed 17 Cassar CA , Ottaway M , Paiba GA et al. Absence of enteroaggregative Escherichia coli in farmed animals in Great Britain . Vet Rec 2004 ; 154 : 237 – 9 . Google Scholar CrossRef Search ADS PubMed 18 Wieler LH , Semmler T , Eichhorn I et al. No evidence of the Shiga toxin-producing E. coli O104:H4 outbreak strain or enteroaggregative E. coli (EAEC) found in cattle faeces in northern Germany, the hotspot of the 2011 HUS outbreak area . Gut Pathog 2011 ; 3 : 17 . Google Scholar CrossRef Search ADS PubMed 19 Hamm K , Barth SA , Stalb S et al. Experimental infection of calves with Escherichia coli O104:H4 outbreak strain . Sci Rep 2016 ; 6 : 32812 . Google Scholar CrossRef Search ADS PubMed 20 Ritchie JM , Thorpe CM , Rogers AB et al. Critical roles for stx2, eae, and tir in enterohemorrhagic Escherichia coli-induced diarrhea and intestinal inflammation in infant rabbits . Infect Immun 2003 ; 71 : 7129 – 39 . Google Scholar CrossRef Search ADS PubMed 21 Fegan N , Gobius KS. Pathogenic Escherichia coli and One Health implications . Curr Top Microbiol Immunol 2013 ; 366 : 49 – 62 . Google Scholar PubMed 22 Persad AK , LeJeune JT. Animal reservoirs of shiga toxin-producing Escherichia coli . Microbiol Spectr 2014 ; 2 : EHEC-0027-2014. 23 Yamaichi Y , Chao MC , Sasabe J et al. High-resolution genetic analysis of the requirements for horizontal transmission of the ESBL plasmid from Escherichia coli O104:H4 . Nucleic Acids Res 2015 ; 43 : 348 – 60 . Google Scholar CrossRef Search ADS PubMed 24 Woodward M , Best A , Sprigings K et al. Non-toxigenic Escherichia coli O157:H7 strain NCTC12900 causes attaching-effacing lesions and eae-dependent persistence in weaned sheep . Int J Med Microbiol 2003 ; 293 : 299 – 308 . Google Scholar CrossRef Search ADS PubMed 25 La Ragione RM , Best A , Clifford D et al. Influence of colostrum deprivation and concurrent Cryptosporidium parvum infection on the colonization and persistence of Escherichia coli O157:H7 in young lambs . J Med Microbiol 2006 ; 55 : 819 – 28 . Google Scholar CrossRef Search ADS PubMed 26 OIE World Organisation for Animal Health . Verocytotoxigenic Escherichia coli. In: OIE , ed. Manual of Diagnostic Tests and Vaccines for Terrestrial Animals . OIE , 2016 ; Chapt 2.9.10. 27 Grauke LJ , Kudva IT , Yoon JW et al. Gastrointestinal tract location of Escherichia coli O157:H7 in ruminants . Appl Environ Microbiol 2002 ; 68 : 2269 – 77 . Google Scholar CrossRef Search ADS PubMed 28 La Ragione RM , Best A , Woodward MJ et al. Escherichia coli O157:H7 colonization in small domestic ruminants . FEMS Microbiol Rev 2009 ; 33 : 394 – 410 . Google Scholar CrossRef Search ADS PubMed 29 Vogwill T , MacLean RC. The genetic basis of the fitness costs of antimicrobial resistance: a meta-analysis approach . Evol Appl 2015 ; 8 : 284 – 95 . Google Scholar CrossRef Search ADS PubMed 30 Kamp HD , Patimalla-Dipali B , Lazinski DW et al. Gene fitness landscapes of Vibrio cholerae at important stages of its life cycle . PLoS Pathog 2013 ; 9 : e1003800 . Google Scholar CrossRef Search ADS PubMed 31 Barrios AF , Zuo R , Ren D et al. Hha, YbaJ, and OmpA regulate Escherichia coli K12 biofilm formation and conjugation plasmids abolish motility . Biotechnol Bioeng 2006 ; 93 : 188 – 200 . Google Scholar CrossRef Search ADS PubMed 32 Mourand G , Touzain F , Jouy E et al. Rare spontaneous loss of multiresistance gene carrying IncI/ST12 plasmid in Escherichia coli in pig microbiota . Antimicrob Agents Chemother 2016 ; 60 : 6046 – 9 . Google Scholar CrossRef Search ADS PubMed 33 Porse A , Gumpert H , Kubicek-Sutherland JZ et al. Genome dynamics of Escherichia coli during antibiotic treatment: transfer, loss, and persistence of genetic elements in situ of the infant gut . Front Cell Infect Microbiol 2017 ; 7 : 126 . Google Scholar CrossRef Search ADS PubMed 34 Cottell JL , Saw HT , Webber MA et al. Functional genomics to identify the factors contributing to successful persistence and global spread of an antibiotic resistance plasmid . BMC Microbiol 2014 ; 14 : 168. Google Scholar CrossRef Search ADS PubMed 35 Furuya N , Komano T. Nucleotide sequence and characterization of the trbABC region of the IncI1 plasmid R64: existence of the pnd gene for plasmid maintenance within the transfer region . J Bacteriol 1996 ; 178 : 1491 – 7 . Google Scholar CrossRef Search ADS PubMed 36 Nedialkova LP , Denzler R , Koeppel MB et al. Inflammation fuels colicin Ib-dependent competition of Salmonella serovar Typhimurium and E. coli in enterobacterial blooms . PLoS Pathog 2014 ; 10 : e1003844 . Google Scholar CrossRef Search ADS PubMed 37 Yatsuyanagi J , Saito S , Sato H et al. Characterization of enteropathogenic and enteroaggregative Escherichia coli isolated from diarrheal outbreaks . J Clin Microbiol 2002 ; 40 : 294 – 7 . Google Scholar CrossRef Search ADS PubMed 38 Fang H , Ataker F , Hedin G et al. Molecular epidemiology of extended-spectrum β-lactamases among Escherichia coli isolates collected in a Swedish hospital and its associated health care facilities from 2001 to 2006 . J Clin Microbiol 2008 ; 46 : 707 – 12 . Google Scholar CrossRef Search ADS PubMed 39 Xu L , Ensor V , Gossain S et al. Rapid and simple detection of blaCTX-M genes by multiplex PCR assay . J Med Microbiol 2005 ; 54 : 1183 – 7 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please email: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Antimicrobial Chemotherapy Oxford University Press

Host-specific differences in the contribution of an ESBL IncI1 plasmid to intestinal colonization by Escherichia coli O104:H4

Loading next page...
 
/lp/ou_press/host-specific-differences-in-the-contribution-of-an-esbl-inci1-plasmid-YbIE6EUTZU
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please email: journals.permissions@oup.com.
ISSN
0305-7453
eISSN
1460-2091
D.O.I.
10.1093/jac/dky037
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives To assess stability and contribution of a large ESBL-encoding IncI1 plasmid to intestinal colonization by Escherichia coli O104:H4 in two different mammalian hosts. Methods Specific-pathogen-free 3–4-day-old New Zealand White rabbits and conventionally reared 6-week-old weaned lambs were orally infected with WT E. coli O104:H4 or the ESBL-plasmid-cured derivative, and the recovery of bacteria in intestinal homogenates and faeces monitored over time. Results Carriage of the ESBL plasmid had differing impacts on E. coli O104:H4 colonization of the two experimental hosts. The plasmid-cured strain was recovered at significantly higher levels than WT during late-stage colonization of rabbits, but at lower levels than WT in sheep. Regardless of the animal host, the ESBL plasmid was stably maintained in virtually all in vivo passaged bacteria that were examined. Conclusions These findings suggest that carriage of ESBL plasmids has distinct effects on the host bacterium depending upon the animal species it encounters and demonstrates that, as for E. coli O157:H7, ruminants could represent a potential transmission reservoir. Introduction In 2011, Escherichia coli O104:H4 caused a large outbreak of haemolytic uraemic syndrome centred in northern Germany, in which more than 50 people died.1 Subsequent genotypic and phenotypic analyses demonstrated that this strain was unusual, with features of both enterohaemorrhagic (EHEC) and enteroaggregative E. coli (EAEC) pathotypes.2,3 Typical of all EHEC, the strain contained genes coding for Shiga toxin (Stx), specifically Stx2a, which is associated with severe clinical outcome in patients.4 However, it also adhered to tissue culture cells in a characteristic ‘stacked-brick’ manner, a trait that defines the EAEC pathotype and is associated with aggregative adherence fimbriae, the genes for which are carried on a plasmid (pAA). Unusual for both pathotypes, E. coli O104:H4 harboured a large plasmid carrying genes coding for ESBLs. While several studies have focused on the contribution of pAA to E. coli O104:H4 pathogenesis,5–7 none has considered the significance of the carriage of the ESBL plasmid beyond its immediate impact of limiting antibiotic options. ESBLs are enzymes that confer resistance to many frontline β-lactam antibiotics and are a dominant mechanism of antimicrobial resistance in Gram-negative bacteria.8 Many types of ESBL have been described, but those belonging to the CTX-M type have become widespread in the UK and elsewhere.9–12 Genes encoding ESBLs are most often located on large, conjugative plasmids, in part providing a mechanistic explanation for their spread.8 Bacterial carriage of resistance plasmids, such as those encoding ESBLs, is believed to confer a biological fitness cost to the host bacterium, although experimental evidence to support this is conflicting. In vitro-based growth assays have shown both detrimental13 and no fitness cost14,15 associated with carriage of CTX-M-containing plasmids. However, Schaufler et al.16 concluded that the carriage of the ESBL plasmid was only associated with a fitness cost when bacteria were grown on surfaces rather than as planktonic cultures. In their study, surface-associated plasmid-cured variants of various pathogenic E. coli clonal lineages exhibited changes in fimbriae production, an ability to form biofilms or be motile; surface-associated attributes that may contribute to colonization and persistence in the mammalian intestine. Given the widespread prevalence of ESBL-producing organisms that are also pathogens of animals and/or humans, it is perhaps surprising that more studies to examine their impact in the context of the host have not been performed. Moreover, evidence to indicate whether ruminants act as a reservoir for EAEC17 or Shiga toxin-producing EAEC isolates, such as E. coli O104:H4,18 is lacking, although carriage following experimental infection of calves has recently been demonstrated.19 Herein, we describe the impact of pESBL, the IncI1 plasmid of E. coli O104:H4, on the capacity of the pathogen to colonize the intestine of two different mammalian species. Plasmid carriage hindered late-stage colonization of the intestine of infant rabbits, a model of EHEC-mediated intestinal disease in humans.20 In contrast, E. coli O104:H4 persisted in the intestine and could be detected in faeces of weaned sheep for up to 4 weeks, independent of pESBL carriage. These studies indicate that carriage of a large ESBL plasmid mediated host-specific differences in the persistence of the host bacterium and demonstrate that, as for E. coli O157:H7,21,22 ruminants could represent a potential transmission source. Materials and methods Strains and culture conditions The strains and plasmids used in the study are listed in Table S1 (available as Supplementary data at JAC Online). BL211, an Stx2 deletion mutant of E. coli O104:H4 strain C227–11,5 was used in this study to enable the animal experiments to be performed in the available facilities and reduce any risk of serious infection to research personnel. Bacteria were routinely grown in LB medium or on LB agar plates containing the appropriate antibiotics: 10 mg/L gentamicin; 10 mg/L tetracycline; or 2 mg/L cefotaxime. Construction of the pESBL-cured derivative Strain BL211 harbours three plasmids including a large 88.5 kb ESBL-encoding plasmid that belongs to incompatibility group I1 (IncI1) and carries both the blaTEM-1 and blaCTX-M-15 genes2 (subsequently annotated as blaCTX-M-323). Incompatibility based curing13 was used to rid strain BL211 of this plasmid (originally referred to as pESBL-EA11 and herein called pESBL). Briefly, plasmid pIFM27, a sacB-containing plasmid, which encodes the IncI1 plasmid replication down-regulator RNAI that directly interferes with the replication of IncI plasmids, was introduced into BL211 by electroporation. Transformants were selected on LB agar supplemented with kanamycin (50 mg/L) and were subsequently screened for an inability to grow on cefotaxime-containing LB medium. Curing of plasmid pIFM27 from cefotaxime-susceptible colonies was accomplished by recovery of colonies on LB agar supplemented with 5% sucrose. Subsequently, sucrose-resistant, cefotaxime-susceptible colonies were screened for loss of pESBL and pIFM27 by PCR and plasmid profiling (Figure S1A and B). Loss of pESBL did not impact the ability of the strain (herein called BL320) to grow in vitro in LB medium in single strain growth assays (Figure S1C). PCR analysis of colonies Multiplex PCR was used to assess the chromosomal and plasmid gene content of BL211, BL320 and randomly selected colonies recovered from rabbits or sheep infected with the parental strain BL211. Primer sequences, expected product sizes and reaction conditions are reported in Table S1. Infant rabbit studies All experimental protocols were approved by the local Animal Welfare Ethical Review Body and carried out in accordance with the UK Animals (Scientific Procedures) Act 1986. Time-mated specific pathogen-free adult New Zealand White rabbits were purchased from Harlan Laboratories (Derby, UK) at 2–3 weeks gestation. Following parturition, mixed sex litters were kept together in a nesting box with the lactating doe and housed under standard conditions. Infant rabbit infections were performed on 3–4-day-old pups essentially as described previously.5 Briefly, rabbits were administered ranitidine intraperitoneally (5 mg/kg body weight) and 2 h later ∼1 × 109 cfu bacteria re-suspended in sodium bicarbonate solution (2.5 g of NaHCO3 in 100 mL of H2O) was given via oral gavage. Following challenge, animals were monitored twice daily for signs of intestinal disease. Given that all bacterial strains lacked Stx, clinical signs were scored with the following revised scale: none (no adherent faecal material on fur and intestines appear normal with hard, formed digesta in the distal colon), intestinal disease (no adherent faecal material on fur, but colon contains soft, poorly formed digesta) and diarrhoea (adherent faecal material on fur and colon contains liquid or unformed digesta). Rabbits were euthanized by a Schedule 1 method at either day 3 or 7 post-infection. Tissues were aseptically sampled post mortem for bacterial enumeration. Samples from the distal small intestine, caecum, mid colon, distal colon and stools were homogenized in PBS, serially diluted and plated on to LB agar supplemented with gentamicin and tetracycline. Where no colonies were detected following plating of undiluted tissue homogenates, the number of bacteria recovered was set using the lower limit of detection as a value. Spread plates, which contained well-spaced colonies, were chosen for replica plating to LB agar with and without cefotaxime. Colonies failing to grow in the presence of the antibiotic were subject to multiplex PCR to confirm the loss of blaCTX-M. All infections were performed in at least two independent litters to limit any litter-specific effects. Sheep colonization studies All experimental protocols were approved by the local Animal Welfare Ethical Review Body and carried out in accordance with the UK Animals (Scientific Procedures) Act 1986. Experimentally inoculated, weaned sheep were used as a model of natural ruminant infection as previously described for E. coli O157:H7.24,25 Conventionally reared 6-week-old cross-bred commercial lambs were divided into mixed sex groups of eight animals and housed under biosecure conditions. Prior to challenge, individual sheep were confirmed as free of E. coli O104:H4 by screening faecal samples with an in-house E. coli O104-specific immunomagnetic separation capture assay and an agglutination assay.26 After 1 week acclimation, sheep were orally inoculated with ∼5 × 109 cfu of bacteria (BL211 or BL320) delivered in a volume of 11 mL using a worming gun (Novartis, UK). The inocula were prepared from 16 h aerobically incubated cultures, which were grown in LB broth, pelleted by centrifugation and finally re-suspended in PBS. Faecal samples were collected per rectum from all animals on days 1–14 post-infection and twice weekly thereafter until day 39. On days 4 and 39 post-infection, 3 and 5 animals, respectively, from each group were euthanized and tissue samples (1 g) were collected from the ileum, caecum, spiral colon, rectum and recto-anal junction. Prior to microbiological analysis, faecal and tissue samples were homogenized in buffered peptone water (BPW) at a ratio of 1:10 (w/v) using a vortex (faeces) or an Ystral D-79282 homogenizer (tissues). Ten-fold serial dilutions of the homogenized samples were plated directly on to sorbitol MacConkey agar plates supplemented with tetracycline and gentamicin. If no colonies were observed after overnight incubation, samples were enriched by incubating the BPW homogenates at 37°C for 18 h followed by re-plating to provide a qualitative result. Selected colonies from all faecal and tissue samples were screened by multiplex PCR as described above. Statistical analysis The presence or absence of disease in rabbits was expressed in a contingency table and analysed using Fisher’s exact test. Bacterial count data (cfu/g) were log transformed and differences in the number of WT or cured cells recovered in each intestinal section compared using Student’s t-test. In the sheep infection studies, bacterial count data were log transformed and the total cfu shed over days 1–4 (8 animals per group) or days 5–39 (5 animals per group) were calculated using AUC following the trapezoidal rule (GraphPad Prism, version 5). Differences in the AUC of strains were compared using Student’s t-test. All statistical analysis was performed using GraphPad Prism (version 5). Results Previously, the E. coli O104:H4 outbreak strain was shown to colonize the infant rabbit intestine and cause diarrhoea in a manner that was dependent on Stx, but independent of pAA, the plasmid responsible for mediating aggregative adherence on cultured cells.5 The contribution of the 88.5 kb β-lactamase-encoding plasmid (pESBL) to E. coli O104:H4 pathobiology was not explored. Stable maintenance of pESBL during in vitro growth23 may indicate that pESBL plays an important role in the organism’s biology. To investigate this further, we cured pESBL from the Shiga toxin negative derivative of E. coli O104:H4 and examined its contribution to colonization of rabbit and sheep intestines. pESBL hinders E. coli O104:H4 long-term colonization of infant rabbits Consistent with earlier findings,5 oral infection of infant rabbits with the Shiga-toxin-negative derivative caused few visible signs of disease. Loose stools were detected in 18% (3 of 17) and 11% (2 of 19) of animals infected with the WT and plasmid-cured strain, respectively, between days 2 and 3 post-infection (Table S2). As the majority of animals did not exhibit any manifestations of diarrhoea, we focused on the role of the plasmid in bacterial colonization of the rabbit intestine. The distribution and number of challenge E. coli present in the intestine of rabbits infected with BL211 or BL320 were determined at days 3 and 7 post-infection (Figure 1a–d). Regardless of the infecting strain, there were no differences in the number of BL211 or BL320 cfu recovered from these regions at day 3 post-infection. In contrast, by day 7 post-infection, 1–2 logs fewer BL211 than BL320 were recovered in all regions of the intestine. Specifically, colonization by the parent strain BL211 was significantly reduced compared with the pESBL-cured strain in the ileum (160-fold; P < 0.01) and caecum (215-fold; P < 0.01) and was lower, but did not reach statistical significance, in the colon [mid colon 95-fold (P = 0.07) and distal colon 50-fold (P = 0.13)] of infected rabbits. These findings suggest that carriage of pESBL hinders the longer-term persistence of E. coli O104:H4 particularly in the upper regions of the rabbit intestine. Figure 1. View largeDownload slide Recovery of bacteria in infant rabbits orally infected with E. coli O104:H4 or a derivative lacking pESBL. Concentration (cfu/g) of bacteria recovered at 3 and 7 days post-infection in intestinal homogenates of rabbits infected with the indicated strain (WT BL211 and pESBL-cured strain, BL320). Data points represent individual animals (at day 3: BL211, n = 19 and BL320, n = 17; at day 7: BL211, n = 6 and BL320, n = 9) and the bar represents the geometric mean. Statistical analysis was performed using Student’s t-test with P ≤ 0.05 deemed significant. Figure 1. View largeDownload slide Recovery of bacteria in infant rabbits orally infected with E. coli O104:H4 or a derivative lacking pESBL. Concentration (cfu/g) of bacteria recovered at 3 and 7 days post-infection in intestinal homogenates of rabbits infected with the indicated strain (WT BL211 and pESBL-cured strain, BL320). Data points represent individual animals (at day 3: BL211, n = 19 and BL320, n = 17; at day 7: BL211, n = 6 and BL320, n = 9) and the bar represents the geometric mean. Statistical analysis was performed using Student’s t-test with P ≤ 0.05 deemed significant. To investigate whether pESBL was stably maintained in the WT strain during infection, representative colonies recovered from each animal were replica-plated on to medium supplemented with/without cefotaxime. Loss of ESBL activity was rarely found, even after 7 days growth in the intestine. In all, ∼2.5 × 103 colonies recovered at either day 3 or 7 post-infection were screened for growth on cefotaxime-containing medium and only one colony (recovered at day 3) failed to grow on the antibiotic-containing medium. Loss of blaCTX-M in this colony was confirmed by PCR (see Figure S1A). Together these observations suggest that while pESBL hinders the ability of the host bacterium to persist in the rabbit intestine, the plasmid is stably maintained in the cell. pESBL aids persistence in the ruminant intestine To investigate the role of pESBL in the colonization of the ruminant intestine, we orally infected groups of 6-week-old conventional lambs with BL211 or BL320, and monitored the presence of bacteria up until day 39 post-infection. Owing to the lower number of bacteria usually recovered following E. coli O157:H7 challenge of sheep,27 an additional enrichment step was included when necessary during sample processing in these experiments. As expected, none of the infected lambs showed gross signs of disease after challenge and at post mortem all intestinal tissues appeared normal. Regardless of the infecting strain, most animals shed high numbers of cells (>107 cfu/g) the day after challenge, declining thereafter (Figure 2a). Within each group, some animals continuously shed the challenge bacteria whereas others showed intermittent shedding or only shed for a few days. However, the magnitude and duration of faecal shedding in BL211-infected animals was generally greater than for BL320-infected animals (Figure 2a). The median time before the number of shed bacteria fell below experimental detection limits in two consecutive samples collected from the same animal was 35 (range 4–35) and 18 (range 10–21) days for BL211 and BL320, respectively (Figure 2a, inset). As a result, higher numbers of bacteria were recovered from animals infected with BL211 compared with BL320 in the later stages of the infection. Moreover, bacteria were more likely to be recovered by direct plating rather than following overnight enrichment of faecal samples from animals infected with BL211 versus BL320 (Figure 2b). Despite these trends, the magnitude and duration of shedding when expressed as the AUC for each strain approached, but did not reach, statistical significance (P = 0.08). Thus, the presence of pESBL appears to prolong the duration of E. coli O104:H4 shedding in sheep, albeit with high inter-animal variability. Figure 2. View largeDownload slide Faecal shedding of E. coli O104:H4 or the pESBL-cured derivative from orally infected 6-week-old conventional weaned lambs. Mixed sex lambs (n = 8 per group) were group-housed and individual faecal samples collected per rectum. Samples were homogenized in BPW and plated directly, or after 18 h enrichment, on SMAC medium supplemented with appropriate antibiotics. Count data were log transformed and the mean (±standard error) number of bacteria recovered for each strain was expressed over time (a). Inset shows boxplots representing the median, lower and upper quartiles, and the minimum and maximum duration of shedding for each strain. DL = experimental limit of detection based on average weight of tissue. Proportion of faecal samples in which bacteria were not found or recovered directly or following enrichment (b). Figure 2. View largeDownload slide Faecal shedding of E. coli O104:H4 or the pESBL-cured derivative from orally infected 6-week-old conventional weaned lambs. Mixed sex lambs (n = 8 per group) were group-housed and individual faecal samples collected per rectum. Samples were homogenized in BPW and plated directly, or after 18 h enrichment, on SMAC medium supplemented with appropriate antibiotics. Count data were log transformed and the mean (±standard error) number of bacteria recovered for each strain was expressed over time (a). Inset shows boxplots representing the median, lower and upper quartiles, and the minimum and maximum duration of shedding for each strain. DL = experimental limit of detection based on average weight of tissue. Proportion of faecal samples in which bacteria were not found or recovered directly or following enrichment (b). While E. coli O157:H7 preferentially colonizes the mid to lower intestinal tract of sheep,24,27 the site of E. coli O104:H4 colonization is less well-defined. Thus, samples from different regions of the intestine (ileum, caecum, colon, rectum and anal-rectal junction) were collected from infected animals at day 4 (n = 3) and day 39 (n = 5) post-infection. At day 4 post-infection, low levels of challenge bacteria were recovered from all five sites of the intestine and all animals, regardless of the infecting strain, yielded bacteria from at least two intestinal sites. While more bacteria tended to be recovered from animals infected with the parent strain BL211, the tissue distribution did not differ markedly between the two strains (Table S3) or from E. coli O157:H7.24 At day 39 post-infection, challenge bacteria could no longer be recovered, even with enrichment, from any of the intestinal samples collected, even though the organisms could still be detected in the animals’ faeces. As found during the rabbit challenge experiments, the ESBL-producing plasmid was stably maintained in colonies recovered from the sheep. Consistent with reports from human infection,7 at least one colony was found to lack aggR indicative of pAA loss. Discussion While a considerable amount of work has been performed elucidating the genetic and biochemical basis of ESBL resistance, fewer studies have attempted to uncover the contribution of ESBL-containing plasmids to pathogen biology, particularly in the context of bacterial survival and carriage in animal hosts. By generating a plasmid-cured derivative of Stx-negative ESBL-producing E. coli O104:H4, we were able to assess the impact of the carriage of the ESBL plasmid in two complementary animal hosts: infant rabbits, which are used as a model of Stx-mediated intestinal disease,5,20 and weaned sheep, which are a potential ruminant reservoir host of Stx-producing isolates.28 We found that carriage of pESBL affected the fitness of the bacterium in the intestine of the two experimental hosts, with the cured strain being recovered at higher levels than WT in rabbits, but at lower levels (than WT) in sheep. Our findings challenge the idea that fitness costs are consistent across different assays as reported previously.29 While Vogwill and MacLean29 found a significant correlation between fitness scores assessed using in vitro and in vivo assays, the impact of the host was not fully ascertained in their analyses. All the in vivo studies examined in their study were performed in a single host species, mice. Our findings highlight the importance of the host context when considering the impact of resistance plasmids on the fitness of the bacterium during intestinal carriage. Species-specific differences in the host can mitigate the requirement for particular bacterial factors and may explain the differing impact of pESBL on the host bacterium in this study. For example, gltA, encoding a citrate synthase and mtlD, encoding a mannitol metabolic protein, are important for effective Vibrio cholerae colonization of the infant rabbit intestine, but not the infant mouse intestine.30 Differences in the availability of carbon and energy sources within the two animal intestines were hypothesized to explain the relative necessity of these genes. Likewise, differences between the rabbit and sheep gastrointestinal tracts could impact the contribution of plasmid-borne factors on BL211 fitness. In addition to the ESBL genes, which are unlikely to confer a direct selective advantage in our experiments, pESBL contains 95 genes coding for factors of unknown and known function. One of these is Hha, a haemolysin expression-modulating protein that appears to affect a myriad of surface-associated cellular phenotypes in E. coli including bacterial motility, cell aggregation and biofilm formation.31 Loss of Hha in laboratory K12 and some ESBL-producing E. coli isolates led to increased swimming activity and reduced biofilm formation compared with their parental strains;16 phenotypes that may alter the ability of the organism to colonize and persist in the intestine. The late-stage colonization advantage of the pESBL-cured strain was not evident in sheep, a ruminant host. Instead, it appeared that loss of pESBL resulted in more rapid clearance of the cured strain from the intestine than WT (Figure 2). Only detailed molecular analyses involving deletion mutants and further in vivo experimentation will uncover the factors and/or mechanisms that explain these fitness outcomes. Like the prototypical E. coli O157:H7 serotype, we found that E. coli O104:H4 was able to persist in the ruminant intestine for at least 4 weeks. These findings are consistent with those recently reported following experimental infection of E. coli O104:H4 in weaned calves.19 Finally, we found that pESBL was stably maintained during infection, as most cells recovered from the rabbit or sheep intestine retained the ability to grow on cefotaxime-containing medium. Similar observations were reported for calves infected with E. coli O104:H4,19 for pigs infected with E. coli carrying an IncI/ST12 blaCTX-M-encoding plasmid32 and, most recently, in streptomycin-treated mice infected with E. coli carrying a large non-conjugative virulence plasmid.33 As noted by others, why these large plasmids are stably maintained in the host bacterium in the absence of obvious selection pressure is intriguing. Yamaichi et al.23 identified six regions of pESBL as essential for pESBL replication or segregation. Intriguingly one of these regions mapped to hha. Thus, as well as hha potentially affecting the surface-expressed properties of the host bacterium, its presence may also help to ensure that pESBL is stably maintained in the cell during intestinal growth. Unlike some other IncI ESBL plasmids circulating in the UK,10 pESBL also appears to contain a recognizable plasmid addiction system (PndAC). While not identified as playing a role in plasmid maintenance in E. coli O104:H4,23 PndAC has been found to play a role in the maintenance of other plasmids.34,35 Collectively our studies suggest that the impact of plasmids bearing antibiotic resistance determinants on pathogen biology needs to be understood in the context of the host. For zoonotic food-borne pathogens, this should include intestinal environments that are healthy (i.e. reservoir hosts) or diseased, as host factors, such as inflammatory cells,36 are known to affect antimicrobial resistance (AMR) transmission and pathogen survival. Finally, our studies show that ruminants, such as sheep, can act as reservoir hosts for E. coli O104:H4 and thus they should be considered as a potential source of transmission to humans. Acknowledgements We would like to acknowledge the assistance of the Animal Services Unit staff at the Animal and Plant Health Agency (Weybridge, UK) and the Experimental Biology Unit staff at the University of Surrey (Guildford, UK). We also thank Irene Freire Martin for providing the IncI curing vector. Funding This work was supported by the European Union 7th Framework Programme grant entitled ‘Anticipating the global onset of novel epidemics’ (project number 278976). Transparency declarations None to declare. Supplementary data Tables S1–S3 and Figure S1 are available as Supplementary data at JAC Online. References 1 Buchholz U , Bernard H , Werber D et al. German outbreak of Escherichia coli O104:H4 associated with sprouts . N Engl J Med 2011 ; 365 : 1763 – 70 . Google Scholar CrossRef Search ADS PubMed 2 Rasko DA , Webster DR , Sahl JW et al. Origins of the E. coli strain causing an outbreak of hemolytic-uremic syndrome in Germany . N Engl J Med 2011 ; 365 : 709 – 17 . Google Scholar CrossRef Search ADS PubMed 3 Bielaszewska M , Mellmann A , Zhang W et al. Characterisation of the Escherichia coli strain associated with an outbreak of haemolytic uraemic syndrome in Germany, 2011: a microbiological study . Lancet Infect Dis 2011 ; 11 : 671 – 6 . Google Scholar CrossRef Search ADS PubMed 4 Persson S , Olsen KE , Ethelberg S et al. Subtyping method for Escherichia coli shiga toxin (verocytotoxin) 2 variants and correlations to clinical manifestations . J Clin Microbiol 2007 ; 45 : 2020 – 4 . Google Scholar CrossRef Search ADS PubMed 5 Munera D , Ritchie JM , Hatzios SK et al. Autotransporters but not pAA are critical for rabbit colonization by Shiga toxin-producing Escherichia coli O104:H4 . Nat Commun 2014 ; 5 : 3080 . Google Scholar CrossRef Search ADS PubMed 6 Boisen N , Hansen AM , Melton-Celsa AR et al. The presence of the pAA plasmid in the German O104:H4 Shiga toxin type 2a (Stx2a)-producing enteroaggregative Escherichia coli strain promotes the translocation of Stx2a across an epithelial cell monolayer . J Infect Dis 2014 ; 210 : 1909 – 19 . Google Scholar CrossRef Search ADS PubMed 7 Zhang W , Bielaszewska M , Kunsmann L et al. Lability of the pAA virulence plasmid in Escherichia coli O104:H4: implications for virulence in humans . PLoS One 2013 ; 8 : e66717 . Google Scholar CrossRef Search ADS PubMed 8 Blair JM , Webber MA , Baylay AJ et al. Molecular mechanisms of antibiotic resistance . Nat Rev Microbiol 2015 ; 13 : 42 – 51 . Google Scholar CrossRef Search ADS PubMed 9 Day MJ , Rodríguez I , van Essen-Zandbergen A et al. Diversity of STs, plasmids and ESBL genes among Escherichia coli from humans, animals and food in Germany, the Netherlands and the UK . J Antimicrob Chemother 2016 ; 71 : 1178 – 82 . Google Scholar CrossRef Search ADS PubMed 10 Doumith M , Dhanji H , Ellington MJ et al. Characterization of plasmids encoding extended-spectrum β-lactamases and their addiction systems circulating among Escherichia coli clinical isolates in the UK . J Antimicrob Chemother 2012 ; 67 : 878 – 85 . Google Scholar CrossRef Search ADS PubMed 11 Fischer J , Rodríguez I , Baumann B et al. blaCTX-M-15-carrying Escherichia coli and Salmonella isolates from livestock and food in Germany . J Antimicrob Chemother 2014 ; 69 : 2951 – 8 . Google Scholar CrossRef Search ADS PubMed 12 Livermore DM , Canton R , Gniadkowski M et al. CTX-M: changing the face of ESBLs in Europe . J Antimicrob Chemother 2007 ; 59 : 165 – 74 . Google Scholar CrossRef Search ADS PubMed 13 Freire MI , Thomas CM , Laing E et al. Curing vector for IncI1 plasmids and its use to provide evidence for a metabolic burden of IncI1 CTX-M-1 plasmid pIFM3791 on Klebsiella pneumoniae . J Med Microbiol 2016 ; 65 : 611 – 8 . Google Scholar CrossRef Search ADS PubMed 14 Fischer EA , Dierikx CM , van Essen-Zandbergen A et al. The IncI1 plasmid carrying the blaCTX-M-1 gene persists in in vitro culture of a Escherichia coli strain from broilers . BMC Microbiol 2014 ; 14 : 77 . Google Scholar CrossRef Search ADS PubMed 15 Cottell JL , Webber MA , Piddock LJ. Persistence of transferable extended-spectrum-β-lactamase resistance in the absence of antibiotic pressure . Antimicrob Agents Chemother 2012 ; 56 : 4703 – 6 . Google Scholar CrossRef Search ADS PubMed 16 Schaufler K , Semmler T , Pickard DJ et al. Carriage of extended-spectrum β-lactamase-plasmids does not reduce fitness but enhances virulence in some strains of pandemic E. coli lineages . Front Microbiol 2016 ; 7 : 336 . Google Scholar CrossRef Search ADS PubMed 17 Cassar CA , Ottaway M , Paiba GA et al. Absence of enteroaggregative Escherichia coli in farmed animals in Great Britain . Vet Rec 2004 ; 154 : 237 – 9 . Google Scholar CrossRef Search ADS PubMed 18 Wieler LH , Semmler T , Eichhorn I et al. No evidence of the Shiga toxin-producing E. coli O104:H4 outbreak strain or enteroaggregative E. coli (EAEC) found in cattle faeces in northern Germany, the hotspot of the 2011 HUS outbreak area . Gut Pathog 2011 ; 3 : 17 . Google Scholar CrossRef Search ADS PubMed 19 Hamm K , Barth SA , Stalb S et al. Experimental infection of calves with Escherichia coli O104:H4 outbreak strain . Sci Rep 2016 ; 6 : 32812 . Google Scholar CrossRef Search ADS PubMed 20 Ritchie JM , Thorpe CM , Rogers AB et al. Critical roles for stx2, eae, and tir in enterohemorrhagic Escherichia coli-induced diarrhea and intestinal inflammation in infant rabbits . Infect Immun 2003 ; 71 : 7129 – 39 . Google Scholar CrossRef Search ADS PubMed 21 Fegan N , Gobius KS. Pathogenic Escherichia coli and One Health implications . Curr Top Microbiol Immunol 2013 ; 366 : 49 – 62 . Google Scholar PubMed 22 Persad AK , LeJeune JT. Animal reservoirs of shiga toxin-producing Escherichia coli . Microbiol Spectr 2014 ; 2 : EHEC-0027-2014. 23 Yamaichi Y , Chao MC , Sasabe J et al. High-resolution genetic analysis of the requirements for horizontal transmission of the ESBL plasmid from Escherichia coli O104:H4 . Nucleic Acids Res 2015 ; 43 : 348 – 60 . Google Scholar CrossRef Search ADS PubMed 24 Woodward M , Best A , Sprigings K et al. Non-toxigenic Escherichia coli O157:H7 strain NCTC12900 causes attaching-effacing lesions and eae-dependent persistence in weaned sheep . Int J Med Microbiol 2003 ; 293 : 299 – 308 . Google Scholar CrossRef Search ADS PubMed 25 La Ragione RM , Best A , Clifford D et al. Influence of colostrum deprivation and concurrent Cryptosporidium parvum infection on the colonization and persistence of Escherichia coli O157:H7 in young lambs . J Med Microbiol 2006 ; 55 : 819 – 28 . Google Scholar CrossRef Search ADS PubMed 26 OIE World Organisation for Animal Health . Verocytotoxigenic Escherichia coli. In: OIE , ed. Manual of Diagnostic Tests and Vaccines for Terrestrial Animals . OIE , 2016 ; Chapt 2.9.10. 27 Grauke LJ , Kudva IT , Yoon JW et al. Gastrointestinal tract location of Escherichia coli O157:H7 in ruminants . Appl Environ Microbiol 2002 ; 68 : 2269 – 77 . Google Scholar CrossRef Search ADS PubMed 28 La Ragione RM , Best A , Woodward MJ et al. Escherichia coli O157:H7 colonization in small domestic ruminants . FEMS Microbiol Rev 2009 ; 33 : 394 – 410 . Google Scholar CrossRef Search ADS PubMed 29 Vogwill T , MacLean RC. The genetic basis of the fitness costs of antimicrobial resistance: a meta-analysis approach . Evol Appl 2015 ; 8 : 284 – 95 . Google Scholar CrossRef Search ADS PubMed 30 Kamp HD , Patimalla-Dipali B , Lazinski DW et al. Gene fitness landscapes of Vibrio cholerae at important stages of its life cycle . PLoS Pathog 2013 ; 9 : e1003800 . Google Scholar CrossRef Search ADS PubMed 31 Barrios AF , Zuo R , Ren D et al. Hha, YbaJ, and OmpA regulate Escherichia coli K12 biofilm formation and conjugation plasmids abolish motility . Biotechnol Bioeng 2006 ; 93 : 188 – 200 . Google Scholar CrossRef Search ADS PubMed 32 Mourand G , Touzain F , Jouy E et al. Rare spontaneous loss of multiresistance gene carrying IncI/ST12 plasmid in Escherichia coli in pig microbiota . Antimicrob Agents Chemother 2016 ; 60 : 6046 – 9 . Google Scholar CrossRef Search ADS PubMed 33 Porse A , Gumpert H , Kubicek-Sutherland JZ et al. Genome dynamics of Escherichia coli during antibiotic treatment: transfer, loss, and persistence of genetic elements in situ of the infant gut . Front Cell Infect Microbiol 2017 ; 7 : 126 . Google Scholar CrossRef Search ADS PubMed 34 Cottell JL , Saw HT , Webber MA et al. Functional genomics to identify the factors contributing to successful persistence and global spread of an antibiotic resistance plasmid . BMC Microbiol 2014 ; 14 : 168. Google Scholar CrossRef Search ADS PubMed 35 Furuya N , Komano T. Nucleotide sequence and characterization of the trbABC region of the IncI1 plasmid R64: existence of the pnd gene for plasmid maintenance within the transfer region . J Bacteriol 1996 ; 178 : 1491 – 7 . Google Scholar CrossRef Search ADS PubMed 36 Nedialkova LP , Denzler R , Koeppel MB et al. Inflammation fuels colicin Ib-dependent competition of Salmonella serovar Typhimurium and E. coli in enterobacterial blooms . PLoS Pathog 2014 ; 10 : e1003844 . Google Scholar CrossRef Search ADS PubMed 37 Yatsuyanagi J , Saito S , Sato H et al. Characterization of enteropathogenic and enteroaggregative Escherichia coli isolated from diarrheal outbreaks . J Clin Microbiol 2002 ; 40 : 294 – 7 . Google Scholar CrossRef Search ADS PubMed 38 Fang H , Ataker F , Hedin G et al. Molecular epidemiology of extended-spectrum β-lactamases among Escherichia coli isolates collected in a Swedish hospital and its associated health care facilities from 2001 to 2006 . J Clin Microbiol 2008 ; 46 : 707 – 12 . Google Scholar CrossRef Search ADS PubMed 39 Xu L , Ensor V , Gossain S et al. Rapid and simple detection of blaCTX-M genes by multiplex PCR assay . J Med Microbiol 2005 ; 54 : 1183 – 7 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please email: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of Antimicrobial ChemotherapyOxford University Press

Published: Feb 28, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off