Association of Escherichia coli O157:H7 with Houseflies on a Cattle Farm

Abstract

The ecology of Escherichia coli O157:H7 is not well understood. The aims of this study were to determine the prevalence of and characterize E. coli O157:H7 associated with houseflies (HF). Musca domestica L. HF (n = 3,440) were collected from two sites on a cattle farm over a 4-month period and processed individually for E. coli O157:H7 isolation and quantification. The prevalence of E. coli O157:H7 was 2.9 and 1.4% in HF collected from feed bunks and a cattle feed storage shed, respectively. E. coli O157:H7 counts ranged from 3.0 × 10¹ to 1.5 × 10⁵ CFU among the positive HF. PCR analysis of the E. coli O157:H7 isolates revealed that 90.4, 99.2, 99.2, and 100% of them (n = 125) possessed the stx1, stx2, eaeA, and fliC genes, respectively. Large populations of HF on cattle farms may play a role in the dissemination of E. coli O157:H7 among animals and to the surrounding environment.

Enterohemorrhagic Escherichia coli O157:H7 is a well-known causative agent of hemorrhagic colitis and hemolytic uremic syndrome in humans. Outbreaks of the food-borne illness caused by E. coli O157:H7 have been reported throughout the northern hemisphere, most frequently in the United States, Canada, Japan, and the United Kingdom. In the United States alone, E. coli O157:H7 causes more than 73,000 cases of human infection every year (11). E. coli O157:H7 strains commonly carry verotoxins (encoded by the stx1 and stx2 genes) and factors for the attachment to the host mucosa, including intimin (encoded by the eaeA gene) (11). The low infectious dose and high virulence of E. coli O157:H7 make infections severe and life-threatening, particularly for young children, the elderly, and those with weakened immune systems (11). The main reservoir for E. coli O157:H7 is the intestinal tracts of healthy cattle. Individual cattle are transiently colonized and shed E. coli O157:H7 in their feces (1). The sources of E. coli O157:H7, which colonizes cattle, are not well understood, and little is known about the ecology of E. coli O157:H7 in the environment (1). Additionally, the high variability in the prevalence of E. coli O157:H7 among cattle suggests the possibility of a reservoir of E. coli O157:H7 external to cattle. However, other than the detection of E. coli O157:H7 in nonbovine animals, including sheep, horses, dogs, and wild birds (1), the ecology of this pathogen has not been extensively studied.

One of the potential modes of dissemination of this pathogen in the environment is by insects that are associated with animal feces and manure, primarily houseflies (HF; Musca domestica L.). HF larvae develop in animal feces, including cattle manure. Consequently, HF commonly build up very large populations on cattle farms and in other animal facilities. Previously, a laboratory-based study demonstrated that E. coli O157:H7 ingested by HF remained viable in the fly excreta and that the HF were able to carry and disseminate E. coli for several days (9). In Japan, HF were implicated in the transmission of E. coli O157:H7 from reservoir animals to other animals and humans (10).

The objectives of this study were (i) to assess the temporal prevalence and concentration of E. coli O157:H7 associated with HF collected on a beef cattle farm, (ii) to characterize E. coli O157:H7 isolates by screening for the virulence factors genes and antibiotic resistance, and (iii) to determine the concentration of fecal coliforms carried by HF in the cattle farm environment as an indicator of fecal contamination.

HF were collected by sweep net from two sites (feed bunks and a storage shed containing steam-flaked corn) on a cattle feedlot in northeastern Kansas four times per week from mid-June until the end of October 2003. From each collection, 50 HF were randomly selected and processed for E. coli O157 isolation and quantification. Individual HF (n = 3,440) were homogenized in 1 ml of phosphate-buffered saline (pH 7.2; ICN Biomedicals) and serially diluted. Dilutions were plated by a direct drop plate technique onto sorbitol MacConkey agar (Becton Dickinson) with cefixime (25 μg/liter) and tellurite (1.25 mg/liter). Plates were incubated overnight at 37°C, and sorbitol-negative colonies with a morphology characteristic of E. coli O157 were tested for the O157 antigen by the latex agglutination assay (Oxoid Limited, Basingstoke, England). Colonies positive for the O157 serogroup were counted, and up to three positive colonies per sample were subcultured on Trypticase soy agar (Becton Dickinson) and identified by the API Rapid 20E test (Biomerieux Vitek). Isolated E. coli O157 strains were tested for the virulence genes (stx1, stx2, and eaeA) and the flagellar H7 gene (fliC) by PCR as described previously (4, 6). The sensitivity of E. coli O157:H7 isolates (n = 108) to 10 antibiotics (ampicillin, apramycin, ceftiofur, chlortetracycline, enrofloxacin, gentamicin, neomycin, oxytetracycline, spectinomycin, and trimethoprim-sulfamethoxazole) was measured by determination of the MIC in accordance with the National Committee for Clinical Laboratory Standards (NCCLS) guidelines (12) with a diagnostic system (TREK Diagnostic Systems, Cleveland, Ohio) with the Sensititre AutoInoculator, sensitivity antibiotic plates, and the SensiTouch plate reader. Fecal coliform counts in randomly selected individual fly homogenates (n = 350) were determined by a drop plate technique with mFC (membrane fecal coliform) agar (Oxoid Limited) incubated at 44.5°C for 18 to 24 h. Fecal coliform colonies were counted from dilutions and calculated per fly.

The prevalence of E. coli O157:H7 associated with HF was 2.9 and 1.4% in feed bunks and a steam-flaked corn storage shed, respectively, with peaks in June and July (Table 1). E. coli O157:H7-positive HF were detected every week throughout the screening period. E. coli O157:H7 counts ranged from 3.0 × 10¹ to 1.5 × 10⁵ CFU per fly (Table 1). The high concentration of E. coli O157:H7 associated with individual HF is an important factor in potential transmission. Cattle manure and other components of the cattle environment usually do not contain a high number of E. coli O157:H7 bacteria (10⁵ CFU) per gram of sample (5, 13). Interestingly, about 30% of the positive HF (n = 125) were collected from a storage shed containing flaked corn. This indicates that HF carry E. coli O157:H7 in different parts of a farm and potentially contaminate various substrates on which they feed or rest. Johnson et al. (8) outlined various habitats of E. coli O157:H7 in the farm environment, including manure heaps, ponds, dams and wells, barns, calf hutches, straw and other beddings, feed and feed troughs, water, soil, and farm equipment. They also noted that, once present in the environment, this pathogen can be transferred to other sites by rainwater, wind, and spreading of manure, as well as by animals and humans. Unfortunately, most studies report only qualitative data on E. coli O157:H7. Our results indicate that HF can be an important vector for E. coli O157:H7 in the cattle farm environment.

TABLE 1.

Prevalence and concentration of E. coli O157:H7 associated with HF on a beef cattle farm

Mo (2003)	Feed bunk		Corn storage
	No. positive/no. tested (%)	E. coli O157:H7 count range (CFU/fly)	No. positive/no. tested (%)	E. coli O157:H7 count range (CFU/fly)
June	3/65 (4.6)	8.0 × 101-3.0 × 102	2/75 (2.7)	3.0 × 102-1.2 × 104
July	32/525 (6.1)	3.0 × 101-9.0 × 103	9/375 (2.4)	3.0 × 101-1.5 × 105
August	9/425 (2.1)	3.0 × 101-4.8 × 104	3/375 (0.8)	3.0 × 101-1.5 × 102
September	4/450 (0.9)	3.0 × 101-1.2 × 103	4/450 (0.9)	3.0 × 101-1.2 × 102
October	4/350 (1.1)	3.0 × 101-1.2 × 102	5/350 (1.4)	3.0 × 101-3.0 × 102
Total	52/1,815 (2.9)	3.0 × 101-4.8 × 104	23/1,625 (1.4)	3.0 × 101-1.5 × 105

PCR analyses of E. coli O157:H7 isolates from HF (n = 125) revealed that 90.4, 99.2, 99.2, and 100% of the latex agglutination-positive isolates (O157) possessed the stx1, stx2, eaeA, and fliC genes, respectively (Table 2). Most of the isolates (90.4%) possessed both toxin genes and the intimin gene, indicating that these strains are highly virulent. Resistance of E. coli O157:H7 to ampicillin (4.6%), ceftiofur (10.2%), chlortetracycline (97.2%), gentamicin (17.6%), neomycin (2.7%), oxytetracycline (100%), spectinomycin (12%), and trimethoprim-sulfamethoxazole (17.6%) was detected. No resistance to apramycin or enrofloxacin was detected. Twenty-six percent of the isolates showed multidrug (three or more antibiotics) resistance.

TABLE 2.

Characteristics of E. coli O157:H7 isolates and the number of fecal coliforms associated with HF collected from feed bunks and a corn storage shed

Site	No. of isolates positive for virulence genes detected in E. coli O157:H7 isolates/no. tested (%)				Fecal coliform count per fly (n = 334)
	stx1	stx2	eaeA	fliC	Mean	Median	Range
Feed bunks	84/95 (88.4)	94/95 (98.9)	94/95 (98.9)	95/95 (100)	2.4 × 105	3.6 × 104	3.0 × 101-3 × 106
Corn storage	29/30 (96.7)	30/30 (100)	30/30 (100)	30/30 (100)	1.8 × 105	1.6 × 104	3.0 × 101-3 × 106
Total	113/125 (90.4)	124/125 (99.2)	124/125 (99.2)	125/125 (100)	2.1 × 105	2.4 × 104	3.0 × 101-3 × 106

Little is known about the association of E. coli O157:H7 with insects, including HF. Previously, E. coli O157:H7 was reported from HF in two studies from Japan (7, 10). Although both studies implicated HF in the transmission of this pathogen, the conclusions reached were based on a very limited data set. Sasaki et al. (14) addressed the role of regurgitation and excretion by HF in the dissemination of E. coli O157:H7 and suggested that these processes enhance the potential transmission of the pathogen. In addition, Kobayashi et al. (9) reported that HF may not be simple mechanical vectors of E. coli O157:H7 and that this pathogen likely multiplies in the gastrointestinal tract of HF. The high numbers of E. coli O157:H7 bacteria detected in our HF samples strongly support this hypothesis.

Fecal coliforms were detected in 334 (95.4%) out of 350 HF screened. Counts of fecal coliforms ranged from 3.0 × 10¹ to 3 × 10⁶ CFU/fly with a mean count of 2.1 × 10⁵ CFU/fly and a median count of 2.4 × 10⁴ CFU/fly (Table 2). The large number of fecal coliforms in HF indicates a potential to harbor other pathogens.

The very large populations, suitable habitats (feces, manure), unrestricted movement, and mode of feeding (regurgitation) of HF represent a likely mechanism for the spread of E. coli O157:H7 on cattle farms and into the surrounding environment. The infectious dose of E. coli O157:H7 is very low, probably fewer than 100 cells (17). Humans can become infected through the consumption of contaminated food, particularly inadequately cooked ground beef (often in the form of beef burgers) and milk (nonpasteurized or contaminated postpasteurization). Other sources of infection include contaminated nonpasteurized apple cider, water (drinking and swimming), vegetables, mayonnaise, lamb, venison, deer jerky, cured salami, and direct contact (animal to person or person to person) (2, 15, 16). Because of the attraction of HF to places where food is prepared, stored, and consumed, these insects present a potential route for the contamination of food and drinks. The dispersal range of HF is usually 0.5 to 2 miles, although distances as great as 10 to 20 miles have been reported (3). Human infections associated with E. coli O157:H7 are more common during summer months, with the majority of cases occurring from May to September. This correlates with an increased shedding of E. coli O157:H7 by cattle, as well as with very large populations of HF on cattle farms and in other animal operations.

In summary, our study demonstrated that HF carry virulent E. coli O157:H7 in the farm environment primarily during the summer and may play an important role in the ecology and transmission of this pathogen among individual cattle and potentially to the surrounding farm and urban environment. Information on the association of E. coli O157:H7 with HF will assist in developing more comprehensive and quantitative risk assessments, as well as formulating E. coli O157:H7 intervention strategies that should include an effective HF management program.