Enteropathogenic Escherichia coli Prevalence in Laboratory Rabbits

Abstract

Rabbit-origin enteropathogenic E. coli (EPEC) causes substantial diarrhea-associated morbidity and has zoonotic potential. A culture-based survey was undertaken to ascertain its prevalence. EPEC was isolated from 6/141 (4.3%) commercially-acquired laboratory rabbits. Three of these did not have diarrhea or EPEC-typical intestinal lesions; they instead had background plasmacytic intestinal inflammation. Asymptomatically infected rabbits may function as EPEC reservoirs.

1. Introduction

Enteropathogenic E. coli (EPEC) is an important cause of diarrhea in both animals and humans (Abba et al., 2009; Garcia et al., 2010; Hill et al., 1991; Nguyen et al., 2006). Rabbits infected with rabbit- and human-origin strains have also been used as experimental models of human E. coli infection (Cantey & Blake, 1977; Garcia et al., 2002, 2006; Moon et al., 1983; Shringi et al., 2012). In a recent report, we described 10.5% EPEC-associated morbidity and 1.44% mortality in a large laboratory rabbit cohort (Swennes et al., 2012). Disease presentation coincided with recent shipment, and the O145:H2 strain responsible was sensitive to the second-generation fluoroquinolone enrofloxacin. That strain’s high associated morbidity, high subclinical infection prevalence (20%), and zoonotic potential prompted us to investigate laboratory rabbit EPEC prevalence.

2. Materials and Methods

2.1 Bacterial culture

Fecal samples were obtained from adult 141 New Zealand white rabbits acquired from 5 commercial vendors and housed at 5 research institutions. All rabbits were clinically healthy except for 3 that had diarrhea. Samples were collected in Brucella broth containing 10% glycerol, homogenized, streaked on blood and MacConkey agar plates, and grown at 37 °C under aerobic conditions. Lactose-positive colonies were sub-cultured based on morphology and speciated using API 20E identification test strips (bioMérieux, Marcy l’Etoile, France). All isolates were tested for antibiotic sensitivity using the Kirby-Bauer method as previously described (Swennes et al., 2012).

2.2 Molecular characterization

DNA was extracted from pure E. coli cultures using the High Pure PCR Template Preparation Kit (Roche Applied Science, Indianapolis, IN). Isolates were differentiated using repetitive sequence-based PCR (REP-PCR) and PCR-tested for eae, stx1, and stx2 as previously described (Swennes et al., 2012). Representative isolates were serotyped by the Pennsylvania State University E. coli Reference Center (University Park, PA).

2.3 Histopathology

Pieces of duodenum, jejunum, ileum, cecum, ampulla, transverse colon, and descending colon were fixed in 10% neutral buffered formalin, paraffin embedded, cut into 5 μm sections, mounted on glass slides, and stained with hematoxylin and eosin. Slides were examined by a board-certified veterinary pathologist.

3. Results

E. coli isolates (n = 163) were obtained from 86 (61%) of the rabbits sampled. Of the 163 isolates, 13% were sensitive to ampicillin, 56% to amoxicillin-clavulanic acid, 4% to ephalothin, and 41% to gentamicin. All isolates were sensitive to trimethoprim-sulfamethoxazole and enrofloxacin. Isolates were differentiated by repetitive sequence-based PCR (REP-PCR), revealing 13 distinct banding patterns each corresponding to a single serotype (Table 1). Virulence factor-based PCR testing was performed for eae, the gene encoding the adhesin intimin that is involved in host epithelial association, as well as stx1 and stx2, which encode the shiga-like toxins which have been previously identified in rabbit-origin E. coli strains (Garcia & Fox, 2003; Garcia et al., 2002). Isolates from 6 rabbits (4.3%) tested eae-positive and stx-negative, the EPEC-characteristic genetic profile (Table 1). Two eae-positive strains were isolated from these rabbits, including an O145:H2 strain previously associated with rabbit diarrhea and a novel OM:H1 (multiple O-reactive) strain (Swennes et al., 2012). EPEC-positive rabbits were obtained from 2 vendors, each harboring either EPEC strain. Of rabbits presenting with diarrhea, 3 of 3 (100%) harbored the O145:H2 strain, while 3 other EPEC-positive rabbits harbored either the O145:H2 or OM:H1 strain and had no associated medical history. The diarrheic potential of the OM:H1 strain merits further investigation.

Table 1.

E. coli isolates of each REP-PCR genotype and corresponding serotype obtained.

REP-PCR genotype	Serotype	eae PCR	n1
1	O2:H1	−	3
2	OM:H1	+	1
3	O-:H7	−	1
4	O7:H7	−	14
5	O8:H10	−	3
6	O8:H49	−	1
7	O103:H7	−	8
8	O145:H2	+	5
9	O170:H5	−	14
10	O174:H28	−	41
11	O13:H4	−	1
12	O103:H16	−	1
13	O18:H7	−	1

¹Number of rabbits harboring the indicated REP-PCR genotype/serotype.

Four of 6 EPEC-positive rabbits, 2 diarrheic and 2 asymptomatic, were available for necropsy following the initial survey. In all cases, the infecting EPEC strain was re-isolated from the cecal contents postmortem. Grossly, the 2 diarrheic rabbits had perineal, tail, and hindlimb accumulation of foul-smelling fecal material. Cecal and colonic serosal reddening suggestive of hemorrhage was evident, and the large intestine was distended with gas and watery feces. Histologically, EPEC-typical intestinal epithelial ulceration with mixed heterophilic/eosinophilic and plasma cell infiltrates was present (Heczko et al., 2000; Peeters et al., 1984; Swennes et al., 2012). The 2 non-diarrheic EPEC-infected rabbits, which harbored either the O145:H2 or OM:H1 strain, displayed no gross lesions or EPEC-typical histologic changes. Instead, an increase in mucosal lamina propria cellularity was noted between intestinal crypts and consisted primarily of plasma cells and few heterophils (Figure 1). Relatively few lymphocytes were found primarily in aggregates. Apoptotic cells were also noted within the mucosal epithelium.

Figure 1.

Ileum from an EPEC-positive rabbit with no clinical signs. A diffuse increase in cellularity in the mucosal lamina propria, comprised primarily of plasma cells, was noted. Occasional lymphoid aggregates are also seen. Hematoxylin and eosin stain. Bar = 100 μm.

4. Conclusions

The survey performed here indicates that diverse E. coli strains are frequently found in the clinically healthy rabbits’ normal intestinal flora. Of the rabbits surveyed, 4.3% harbored EPEC, indicating that these bacteria persist at low levels in commercially-acquired rabbits. Of the infected rabbits, half did not have diarrhea or intestinal histologic changes consistent with EPEC-associated disease. Instead, they had intestinal plasmacytic infiltration similar to lymphocytic-plasmacytic inflammatory bowel disease of dogs and cats (Day et al., 2008). A prior study identified comparable intestinal plasmacytic infiltration in 53 of 102 (52%) rabbits used for various experimental purposes (Li et al., 1996). It is likely that these rabbits had underlying immune-mediated intestinal disease with secondary EPEC infection. This data infers that laboratory rabbits may frequently have idiopathic chronic intestinal inflammation and display no overt clinical signs.

Pathogenic E. coli have been associated with chronic intestinal inflammatory disease in a variety of species. In one report, cotton-top tamarins with EPEC-positive fecal cultures were 2.7 times more likely to have active colitis (Mansfield et al., 2001). Additionally, adherent/invasive E. coli, a heterogeneous pathotype possessing few conventional virulence genes, have been isolated from the ileal mucosa of Crohn’s disease patients (Darfeuille-Michaud et al., 2004) and from the colonic mucosa of Boxer dogs with granulomatous colitis (Simpson et al., 2006). Population-level analyses also indicate that Crohn’s disease patients’ ileal mucosa contains numerous E. coli expressing various virulence genes, including eae (Baumgart et al., 2007).

These reports suggest that the inflamed intestine presents a favorable niche for pathogenic E. coli such as EPEC. In this context, laboratory rabbits with intestinal inflammatory conditions may become infected and remain undetected within rabbit populations until other host or environmental factors, such as Lawsonia intracellularis (Schauer et al., 1998) or rotavirus (Thouless et al., 1996) infection, low dietary fiber intake (Gidenne & Licois, 2005), shipment (Swennes et al., 2012), or increased stress level (Everest, 2007) potentiate fulminant disease. High subclinical EPEC prevalence (20%) was noted in our recent investigation (Swennes et al., 2012). However, this report suggests that without an inciting cause to promote fecal shedding, baseline laboratory rabbit infection prevalence is likely much lower, i.e. 4.3%. Subclinical EPEC infection is at least 5.9% prevalent in children (Afset et al., 2004; Knutton et al., 2001) and may be 2.5% prevalent in the general human population (Robins-Browne et al., 2004).

Laboratory rabbit enzootic EPEC infection is also significant because these strains are potentially zoonotic. Animal- and human-origin EPECs possess similar virulence factors and share similar clonal origins (Moura et al., 2009). In addition to rabbits, EPEC O145:H2 strains have been identified in non-diarrheic sheep (Vettorato et al., 2009), diarrheic human infants (Gonzalez et al., 1997), and wastewater samples (Doughari et al., 2011). Also, piglets and rabbits are routinely experimentally infected with EPEC and enterohemorrhagic E. coli (EHEC) strains from humans and other animals, indicating that infection is not host species-dependent (Garcia et al., 2006; Moon et al., 1983). EHEC transmission from cattle to wild rabbits to humans has also been associated with human hemorrhagic diarrhea and hemolytic uremic syndrome cases (Bailey et al., 2002; Pritchard et al., 2001). Our findings indicate that non-diarrheic rabbits function as reservoirs that potentiate EPEC-associated diarrheal disease and zoonotic transmission. Laboratory rabbit users and suppliers should monitor for and exclude potentially pathogenic E. coli strains.