Article Text

Download PDFPDF

Genes associated with Escherichia coli isolates from calves with diarrhoea and/or septicaemia
  1. G. Wu, BSc, MSc, PhD1,
  2. M. Mafura, BSc1,
  3. B. Carter, BSc, PhD2,
  4. K. Lynch, BSc, MSc1,
  5. M. F. Anjum, BSc, PhD1,
  6. M. J. Woodward, BSc, PhD1 and
  7. G. C. Pritchard, BSc, BVM&S, DVM&S3
  1. 1 Veterinary Laboratories Agency (VLA) - Weybridge, Woodham Lane, New Haw, Addlestone, Surrey KT15 3NB
  2. 2 South East Wales Trials Unit, School of Medicine, Cardiff University, Cardiff CF14 4XN
  3. 3 FRCVS, VLA - Bury St Edmunds, Rougham Hill, Bury St Edmunds, Suffolk IP33 2RX
  1. E-mail for correspondence: g.wu{at}vla.defra.gsi.gov.uk

Statistics from Altmetric.com

NEONATAL calf diseases have important economic impacts on the livestock industry worldwide. In the USA, for example, the National Animal Health Monitoring System has estimated the preweaning mortality of dairy calves to be 10.8 per cent, with diarrhoea accounting for 52.2 per cent of the losses (McGuirk and Ruegg 2008). Escherichia coli septicaemia, with or without diarrhoea, is also an important cause of death in young calves, associated with colostrum deprivation (Fey 1971) and other management and husbandry factors. A range of pathogenic microorganisms have been identified as aetiological agents of diarrhoea and septicaemia in calves; E coli has been recognised as an important cause of these diseases for over 100 years (Sojka 1965). Enteric colibacillosis (‘white scours’) is usually associated with enterotoxigenic E coli (ETEC) (Smith and Halls 1967), but can also occasionally be associated with verocytotoxigenic E coli (VTEC) (Chanter and others 1984, Milnes and others 2006). This short communication describes an investigation of the genetic characteristics of E coli isolates from calves with diarrhoea and/or septicaemia in the UK.

Diagnostic samples (carcases, blood, viscera or faeces) from diseased calves were submitted by private veterinary surgeons to the regional laboratories of the Veterinary Laboratories Agency (VLA) for a range of examinations, including tests for ETEC as appropriate. Under an additional enhanced screening programme, E coli isolates implicated in the presenting disease were forwarded to a central reference laboratory (VLA – Weybridge) for further analysis. During 2006 and 2007, a total of 260 E coli isolates associated with calf diarrhoea were serogrouped according to the established criteria (Sojka 1965). The common serogroups were O101 (24 per cent), O8 (12 per cent), O9 (7.7 per cent), O26 (5 per cent), O2 (3.5 per cent) and untypable (15.4 per cent); other serogroups were much less commonly encountered. During the same period, a total of 60 E coli isolates associated with calf septicaemia were also serogrouped; these were mainly O8 (10 per cent), O153 (6 per cent), O78 (5 per cent) and untypable (28 per cent), and other serogroups were less common.

The virulence profiles of a representative cross-section of isolates from the two disease categories, excluding ETEC, were analysed using a miniaturised microarray containing 56 virulence-related genes, including those for fimbriae, adhesins, toxins, microcins and iron acquisition; the details of the genes and the experimental procedures have been described previously by (Anjum and others 2007). It is well established that the presence of certain virulence genes defines the pathotype of an isolate (for example, stx=VTEC). For virulence typing by the array, only isolates with clear clinical histories were included: non-haemorrhagic diarrhoea (22 strains), haemorrhagic diarrhoea (24 strains) and extraintestinal infections (colisepticaemia, 17 strains) of calves, and representing the diversity of serogroups. In addition, three enteropathogenic E coli (EPEC), three VTEC and four ETEC were used as control strains.

The results were used to generate a hierarchical clustering demonstrating the distance between similar genes and isolates, shown in Fig 1 (Pearson's correlation with an average linkage algorithm). Two of the three VTEC and three EPEC strains clustered together, agreeing with the authors' expectation, since both groups are known to contain a similar adhesin (eae) gene, with the VTEC group also harbouring verocytotoxin genes. Another eae-negative VTEC was not within this cluster. The four ETEC strains harboured a different set of adhesin and toxin genes from those in the EPEC and VTEC groups, and were clustered together. Nine strains harboured genes for cytotoxic necrotising factor (cnf1) and are defined as necrotoxigenic E coli (NTEC) in Fig 1. This group of strains also contained genes for cytolethal distending toxin (cdtB), six of the nine strains also harboured f17G and two of those contained the fimbrial gene f17A as well. Among them, some strains also harboured microcin genes cma, cba, mcmA, the iron acquisition gene iroN or the P-related fimbriae gene prfB. This group was composed of strains isolated from both cases of diarrhoea and extraintestinal infections. Four strains harboured f17G with or without f17A. Another group contained genes for pfrB, iron acquisition (iroN and ireA) or microcin (mch). No virulence factors were detected in three strains. For many strains studied here (Fig 1) it was not possible to distinguish between isolates from diarrhoea and isolates from extraintestinal infections on the basis of carriage of any specific genes on the array. Additionally, the carriage of virulence genes varied within the same serogroup.

FIG 1

Virulence gene profiles of isolates from diseased calves from England and Wales. Both strain (dendrogram on the left) and genes (dendrogram on the top) were clustered with Pearson's correlation and the average linkage algorithm. Yellow represents a gene that is present and blue represents a gene that is absent. Multiple copies of genes are shown with brighter colours and genes with partial homology are shown in darker colours. Multiple probes were used for some genes, for example, eae. EPEC Enteropathogenic Escherichia coli, ETEC Enterotoxigenic E coli, ExPEC Extraintestinal pathogenic E coli that were isolated from extraintestinal locations, VTEC Verocytotoxigenic E coli

It is interesting to note the similarities between strains from extraintestinal infections and diarrhoea cases, which is consistent with an earlier report (Van Bost and others 2001). At the moment, gene probes that can differentiate them are not available, and further studies to characterise the virulence determinants of these isolates are needed in order to achieve better differentiation.

Fisher's exact test was used to test the association between genes detected in the strains and the associated diseases: haemorrhagic diarrhoea versus non-haemorrhagic diarrhoea (group 1), and diarrhoea versus extraintestinal infection (group 2). Examples of the results are presented in Table 1. The iroN and mchF genes were found to have a significant association with non-haemorrhagic diarrhoea (P=0.0037 and P=0.021, respectively). The iroBCDEN gene cluster has been described for extraintestinal strains, with iroN being orthologous to a catecholate siderophore receptor gene identified in Salmonella species (Baumler and others 1998, Russo and others 1999). IroN expression was shown to be regulated by the ferric uptake regulator (Fur) and increased by incubating the respective E coli strains in human urine, ascitic fluid, or blood (Russo and others 1999, 2002). Russo and others (2002)demonstrated that iroN enables the uptake of the catecholate siderophore enterobactin and contributed significantly to the virulence of E coli using a mouse infection model of ascending urinary tract infection. mchF forms part of a type I secretion apparatus together with mchE for the export of microcin MccH47 (Poey and others 2006). The role of iroN and mchF in non-haemorrhagic diarrhoea has not been described and is worth further study. No significant difference in terms of virulence gene carriage was found among strains from diarrhoea and from extraintestinal infections.

In conclusion, isolates from diarrhoeic and septicaemic calves belonged to diverse serogroups and the carriage of virulence genes varied within the same serogroup, supporting the view that serogroup cannot be used to infer pathotype (Butler and Clarke 1994). NTEC strains and strains harbouring genes for microcins and iron acquisition or P-related fimbriae were found among E coli associated with both diarrhoeic and septicaemic disease. It is interesting to note the significantly biased distribution of the iroN and mchF genes among E coli associated with non-haemorrhagic diarrhoea. Further studies are required to elucidate the pathogenic mechanism of these organisms.

Acknowledgements

The authors thank the VLA regional laboratories and surveillance centres for submitting the E coli isolates for further analysis. This work was funded by Defra through the Endemic Diseases and Welfare programme (ED1000) and the non-statutory zoonoses project (FZ2100).

References

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.