Comparison of Diversities of Escherichia coli O157 Shed from a Cohort of Spring-Born Beef Calves at Pasture and in Housing

Abstract

A cohort of spring-born beef calves demonstrated limited genetic and phenotypic diversity of Escherichia coli O157 when kept in a state of isolation. Despite this, there was a difference in the pulsed-field gel electrophoresis and phage types of isolates shed by cattle at pasture compared with those shed by the same cattle when weaned and housed.

Escherichia coli O157 is an important human pathogen that has been reported around the world (1, 22), and cattle are an important reservoir of this pathogen (4, 13). The presence and prolonged survival of E. coli O157 on pasture and in the farm environment are strongly related to the presence of cattle feces (10, 26). Understanding the epidemiology of E. coli O157 in animals kept at different sites on the farm is useful for controlling the risk factors that cause the infection to spread.

This study examined the differences between the prevalence of shedding, the genetic and phenotypic diversity, and antibiotic susceptibility among E. coli O157 isolates shed by a cohort of spring-born calves at pasture and while housed.

Study group and sampling.

Sixty-nine calves on a Scottish beef cattle and sheep farm were monitored for shedding of E. coli O157 from June 2001 until April 2002 (Table 1). The calves were born between 2 February 2001 and 24 May 2001 and were at pasture with their dams until housed on 25 October 2001, when they were separated from their dams and split into two groups, group N with 32 calves and group S with 37 calves. Each group was housed on straw within separate pens in an open-sided shed. There was no physical contact between cattle in the two groups, and each pen had separate feed and water troughs. Housed calves were fed home-grown silage, barley, peas, and commercial mineral supplement. At no time did the calves have direct contact with sheep on the farm.

TABLE 1.

Sampling date, location of cattle and group size, sampling procedure, and number of samples taken

Sampling date (day/mo/yr)	Location (no. of cattle)	Sampling procedure	No. of samples	No. of positive samples	PFGE pattern(s) (no. of isolates)
5/06/2001	Pasture (69)	Rectal	69	0
17/08/2001	Pasture (69)	Rectal	69	23	D (11), E (10), H (1), I (1)
22/08/2001	Pasture (69)	Rectal	69	28	D (13), E (12), H (2), I (1)
29/08/2001	Pasture (69)	Rectal	69	16	A (1), D (9), E (6)
3/09/2001	Pasture (69)	Rectal	69	12	A (1), D (6), E (4), I (1)
10/09/2001	Pasture (69)	Rectal	69	1	E (1)
1/10/2001	Pasture (69)	Rectal	69	2	E (2)
17/10/2001	Pasture (69)	Rectal	69	1	E (1)
30/10/2001	Group N (32)	Fecal pat	10	0
	Group S (37)	Fecal pat	10	7	A (7)
5/11/2001	Group N (32)	Fecal pat	10	0
	Group S (37)	Fecal pat	10	9	A (9)
12/11/2001	Group N (32)	Fecal pat	10	0
	Group S (37)	Fecal pat	10	7	A (6), B (1)
19/11/2001	Group N (32)	Fecal pat	30	0
	Group S (37)	Fecal pat	10	3	A (3)
27/11/2001	Group N (32)	Fecal pat	10	0
	Group S (37)	Fecal pat	30	4	A (3), B (1)
2/12/2001	Group N (32)	Fecal pat	10	0
	Group S (37)	Fecal pat	30	0
10/12/2001	Group N (32)	Fecal pat	20	0
	Group S (37)	Fecal pat	20	1	A (1)
4/02/2002	Group N (32)	Fecal pat	20	17	A (17)
	Group S (37)	Fecal pat	20	0
11/02/2002	Group N (32)	Fecal pat	20	17	A (16), C (1)
	Group S (37)	Fecal pat	20	1	A (1)
20/02/2002	Group N (32)	Fecal pat	20	7	A (7)
	Group S (37)	Fecal pat	20	7	A (7)
25/02/2002	Group N (32)	Fecal pat	20	0
	Group S (37)	Fecal pat	20	0
4/03/2002	Group N (32)	Fecal pat	20	0
	Group S (37)	Fecal pat	20	6	A (5), F (1)
18/03/2002	Group N (32)	Fecal pat	20	13	A (13)
	Group S (36)	Rectal	36	16	A (15), B (1)
25/03/2002	Group N (32)	Fecal pat	20	4	A (4)
	Group S (36)	Fecal pat	20	7	A (6), K (1)
8/04/2002	Group N (29)	Fecal pat	20	3	A (3)
	Group S (12)	Fecal pat	16	7	A (7)
14/04/2002	Group N (10)	Fecal pat	20	1	A (1)
	Group S (13)	Fecal pat	20	1	A (1)

All fecal samples were refrigerated within 2 h of sampling. Within 48 h immunomagnetic separation was performed (24). All isolates were submitted to the Scottish E. coli O157 Reference Laboratory for phage typing (1) and tested for the presence of genes encoding Shiga toxin and intimin by using multiplex PCR (23, 28).

Antibiotic sensitivity testing.

Breakpoint sensitivity testing was performed by a standard agar dilution method (2) for ampicillin, cefaclor, cefotaxime, cefuroxime, chloramphenicol, gentamicin, nalidixic acid, streptomycin, tetracycline, trimethoprim, and sulfamethoxazole. British Society for Antimicrobial Chemotherapy (2) guidelines with E. coli NCTC 10418 standard cutoffs were used for interpreting results.

PFGE.

Pulsed-field gel electrophoresis (PFGE) (3, 14, 24) was performed with 50 U of XbaI or 30 U of NotI (Promega) or 10 U of AvrII (New England BioLabs). For XbaI, a linearly ramped switching time from 2.1 to 54.2 s for 22 h at 14°C at 6.0 V cm⁻¹ (200 V) was used, for NotI 10 to 30 s for 18 h was used, and for AvrII 5 to 50 s for 22 h was used. Digested profiles were compared using BioNumerics (version 3.0) software (Applied Maths, Ghent, Belgium).

Statistical analysis.

Statistical analysis was performed using SAS v8.2 (SAS Institute Inc., Cary, N.C.). The age distribution of calves in groups N and S was tested using a runs test (29), and the shedding rates of calves at pasture and when housed were compared using the Mann-Whitney U test (29).

Prevalence of shedding E. coli O157.

Sampling details are given in Table 1. A total of 1,144 rectal and fecal pat samples were taken. When housed, the mean age of calves was 30.4 (standard deviation, 2.3) weeks in group N and 28.4 (standard deviation, 3.1) weeks in group S. Calves were randomly distributed between pens N and S by age (runs test: n₁ = 37, n₂ = 32, u = 38; P = 0.5962).

Overall, there was no demonstrable difference in shedding between calves when housed and at pasture. However, when shedding occurred, the rate of shedding was greater among calves in pen S (0.025 < P < 0.05) and pen N (0.05 < P ≤ 0.10) than when at pasture (Fig. 1). This difference may be related to the density of animals, environment, or diet. Housed cattle were fed barley, which has been shown to increase shedding of E. coli O157 (5). Shedding of E. coli O157 occurred in waves characterized by a sharp increase followed by a gradual decline. Like others (6, 15-17) we also noted a rise in the prevalence of shedding in late summer and early autumn. While calves were at pasture, shedding occurred in cattle subsequently housed in pens N and S. However, no detectable shedding occurred among cattle held in pen N for at least 7 weeks after housing while, among cattle held in pen S, the prevalence of shedding rose sharply to 90.0% within 2 weeks of housing and then steadily declined to a very low level 7 weeks after housing (Fig. 1). Feeding and management of cattle in pens N and S were identical and are unlikely to have contributed to the differential shedding patterns.

FIG. 1.

Prevalence of E. coli O157 shedding among group N and group S calves at pasture and when housed. Cattle to be housed in pen N and cattle to be housed in pen S are indicated by light gray and dark gray, respectively. ns, not sampled.

PFGE patterns of isolates.

Nine XbaI-PFGE profiles (Fig. 2) were identified among the 221 isolates available for typing. Eighty-three isolates were recovered from cattle at pasture comprising five PFGE profiles: A, D, E, H, and I. Most of these 83 (90.4%) isolates were closely related, either PFGE profile D (47.0%) or E (43.4%). PFGE type D was not isolated after 3 September 2001, but PFGE type E was detected every time that E. coli O157 was recovered from cattle at pasture. With the exception of type A, none of the other profiles isolated from cattle at pasture was detected after the cattle were housed. After housing five PFGE types were detected: A, B, C, F, and K. The dominant type was A—a type not closely related to either PFGE type D or type E—representing 61 of 62 (98.4%) isolates from cattle in pen N and 71 of 76 (93.4%) isolates from cattle in pen S. Such dominance by particular PFGE types among E. coli O157 shed by cattle has been reported before (21).

FIG. 2.

Phylogenic analysis of E. coli O157 isolates, by PFGE gel pattern following restriction with XbaI. Dice similarities were subjected to cluster analysis as unweighted matched-pair groups.

Verocytotoxins and intimin.

All E. coli O157 isolates in this study were devoid of vtx₁, but almost all carried vtx₂ and eae (Table 2). This is consistent with studies elsewhere in Europe (6, 7, 19, 20), but in the United States, the proportion of bovine E. coli O157 isolates with vtx₂ as the only vtx gene is 10 to 20% (8, 21, 26). Carriage of vtx₂ within PFGE types D and E was inconsistent, and differential carriage of vtx₂ within a particular PFGE type has been reported before (24).

TABLE 2.

Phenotypic and genotypic profile of E. coli O157 isolates

PFGE type	PT	vtx1	vtx2	eae	Frequency (no.)
A	21/28	0	1	0	2
A	21/28	0	1	1	132
B	21/28	0	1	1	3
C	21/28	0	1	1	1
D	32	0	0	1	1
D	32	0	1	1	38
E	21/28	0	1	1	1
E	32	0	0	1	1
E	32	0	1	1	34
F	21/28	0	1	1	1
H	32	0	1	1	3
I	32	0	1	1	3
K	21/28	0	1	1	1

PT.

Isolate phage type (PT) is summarized in Table 2. Of 221 isolates 141 (63.8%) isolates were PT 21/28 and 80 (36.2%) were PT 32. Prior to housing all but three isolates (96.4%) were PT 32; the remaining three isolates were PT 21/28, of which two were PFGE type A and one was PFGE type E. The two PFGE pattern A, PT 21/28 isolates were shed by a single ox on 20 August 2001 and 3 September 2001. This animal was subsequently housed in pen N. After housing, all isolates were PT 21/28. Although commonly shed by housed cattle, E. coli O157 with PFGE type A and/or PT 21/28 was not commonly shed by cattle at pasture. This is notable, as shedding of E. coli O157 with PFGE type A and PT 21/28 by cattle in pen S was evident at least 7 weeks before such shedding in pen N. Calves in pen N may have been indirectly contaminated (e.g., via animal handlers, contaminated feeds, vermin, or flies) from those in pen S, thus explaining the delayed shedding in pen N and shared PFGE profile.

With the exception of PFGE pattern E, there was no variation in PT within each PFGE type (Table 2). All XbaI-PFGE type E isolates were PT 32, except for one isolate that was PT 21/28 which was detected before housing. E. coli O157 isolates with identical XbaI-PFGE patterns but different PTs have been previously reported (3, 25). Further PFGE analysis of the isolate with PFGE type E, PT 21/28 with NotI and AvrII did not reveal a difference from other vtx₂-positive PFGE type E isolates.

Antibiotic sensitivity.

All 211 isolates that we tested were resistant to sulfamethoxazole but sensitive to the other antibiotics in our test panel, unlike findings from other studies in which E. coli O157 isolates were resistant to a range of antibiotics (9, 11, 18, 27).

We demonstrated limited genetic and phenotypic diversity of E. coli O157 shed by cattle kept in isolation. Nevertheless, there was a marked difference between the PFGE patterns and PTs of E. coli O157 isolated from cattle when at pasture and the same cattle when housed, although the pattern of vtx₁, vtx₂, and eae carriage and antibiotic susceptibility showed little variation. Our results imply that different E. coli O157 strains are shed by the same cattle under different environmental conditions and diets which affect bovine gastrointestinal ecology (12) and suggest that populations of E. coli O157 are responsive to feeding and management practices which could be exploited to develop strategies for controlling E. coli O157 shedding by cattle.