Realizar más estudios relacionados con el nivel de conocimientos y

Isolation of target bacteria from faecal samples is often more complex compared to other matrices such as carcass swabs or dairy products, due to the high levels

of naturally competitive bacteria in faeces [65]. Multiple E. coli serogroups have

been identified in faecal samples [17,61], highlighting the diversity of E. coli

present in ruminant faeces. In addition, culture-based isolation methods for non- O157 serogroups have been demonstrated to be less effective for mixed cultures [73] including when used against cattle faeces compared to other sample matrices [65]. For example, serogroup O145, in comparison to the other “super six” serogroups, had the lowest recovery rate from cattle faeces [72].

In this study, the serogroup O145 recovery rate from RAMS was low (Chapter 3), and may have been impacted by the use of re-enriched cultures, which due to the freezing and thawing of enrichments, may have reduced the number of viable serogroup O145 cells during resuscitation. Also, the selection criteria to identify STEC O145 positive enrichments relied on the NeoSeek results (Chapter 3)

method may detect free DNA rather than DNA associated with viable cells, resulting in an over-estimation of STEC O145 positive RAMS enrichments. The phenotypic variation within serogroups, in particular colony colour, and some media being overly selective and inhibiting the growth of the target organism, are major issues with culture-based isolation on currently available media [63,64]. The use of two media in parallel, a highly selective media paired with a media that supports the growth of a wide range of STEC, may provide an enhanced isolation rate for the non-O157 serogroups [58,63,64]. Therefore, the use of multiple media in parallel for the isolation of serogroup O145 in this study may have increased the recovery rate.

The use of selective media coupled with serogroup-specific IMS increased the isolation rate in this study (Chapter 3). However, the effectiveness of IMS varies widely between studies. For example, using IMS-based methods, serogroup O145 isolates were obtained from 71% of the spiked faecal samples (n=14) by Conrad et al. [16] and Posse et al. [65] detected serogroup O145 with an isolation efficiency of 84.6% from artificially inoculated dairy and meat matrices. The discordance in IMS between studies may be partially attributed to sample variation coupled with the presence of highly competitive background flora [65], the concentration of the target bacteria in the sample [73], the type of matrix used or competition with other STEC serogroups reducing bead specificity [73]. Kraft

et al. [72] evaluated the efficacy of three IMS bead brands (Dynabeads®_{, Abraxis}

and Romer) to detect the “super six” serogroups in a variety of sample matrices. Serogroup O145 recovery using the three bead brands was consistently low across all matrices which included food and faecal samples [72]. Serogroups O145 (16.7-66.7%) and O111 (0-94.4%) had the lowest recovery from faeces [72]. Improved recovery of serogroup O145 with the Romer beads was observed over all matrices compared to other bead preparations [72]. Abraxis beads were used in this study, but whether enhanced recovery may have been associated with the use of serogroup-specific O145 beads of other manufacturers is unknown.

Only 1 out of 32 serogroup O145 strains isolated in this study were identified as

stx-positive (Chapter 3); 13 of the remaining 21 strains that underwent WGS were

negative O145 in ruminants, STEC O145 may be less abundant. However, the

instability of the stx-encoding bacteriophage during sub-cultivation [191] may

contribute to the low isolation rate of STEC O145 in this study. Noll et al. [17]

identified 6 out of 19 serogroup O145 isolates from cattle faeces as stx-positive

(4 stx1-positive, 2 stx2-positive). Using the same set of calf-faecal enrichments

as this study, Browne et al. [192] identified six serogroup O145 isolates which

were all stx-negative, giving an overall recovery rate of 13.3%. Similarly

Dewsbury et al. [59] identified 18 serogroup O145 strains from cattle with few

virulent isolates (4 stx1-positive, 2 stx2-positive, 6 eae-positive).

Unlike stx1 and stx2, all strains that underwent WGS in this study were eae

positive (n=53), and a high proportion (46 out of 53) were ehxA-positive (Chapters

3 and 5). In comparison to the other “super six” serogroups fewer serogroup O145

isolates were negative for the common STEC-associated virulence factors stx1,

stx2, eae or ehxA; with 17 of 19 O145 both eae and ehxA-positive [17]. Noll et al.

[17] identified one avirulent (lacking stx1, stx2, eae and ehxA) O145 strain from

cattle faeces, indicating that serogroup O145 strains may on rare occasions be

eae-negative. Another study Hofer et al. [189] examined serogroup O145 strains

from chamois, ibex and deer in Switzerland; all eight O145 detected were eae

negative and two ehxA positive [189]. However all eight O145 strains were stx-

positive (2 stx1c and stx2b-positive, 6 strains stx2b-positive) [189] suggesting

that wild ruminants carry different STEC populations, which are usually eae-

positive, compared to cattle and other well-studied ruminants.

Some STEC serotypes such as O157:H7 (), O26:H11 (), O103:H2 (), O111:H8

() and O145:H28 () are characterised by a single eae subtype [40], however

the association of multiple eae subtypes with specific serogroups, such as O103,

has been observed [41]. In this work the eae  subtype was the most common

(83%) including all 32 strains isolated from RAMS enrichments. However

NeoSeek identification of STEC O145 includes the detection of eae subtype 

SNPs [193]. Therefore RAMS enrichments containing STEC O145 of other non-

 eae subtypes will not have been identified as candidate enrichments for further

analysis in this work. Other studies have identified STEC O145 containing the

O145 whole genome sequences suggests that eae subtypes α, β, ε, ι are comparatively rare.