Year

2005

Degree Name

Doctor of Philosophy

Department

Department of Biological Sciences - Faculty of Science

Abstract

Exploiting the concept that pathogenic and commensal Escherichia coli strains can be distinguished on the basis of housekeeping enzymes by MLEE, a panel of 58 virulence genes was used to analyse the phylogenetic relationship between commensal E. coli and pathogenic E. coli from clinical cases of porcine neonatal and post-weaning diarrhoea (PWD). Of the 58 virulence genes examined, 17 (STb, iha, hlyA, STa, aidA, F18, east1, LT, fimH, iroNE.coli, aah, F4, traT, F41, F5, saa and stx2) were able to differentiate between the groups of isolates based on Chi-squared analysis. Application of two other statistical methods; an agglomerative hierarchical algorithm and principle coordinate analysis (PCO) to the isolates resulted in a relational depiction consisting of a branching tree dendrogram and point-clustering respectively. In both situations, member clones designated as commensals and pathogens were separated from each other. However, commensal and neonatal diarrhoea (ND) isolates tended to be more similar to each other and together, more different from post-weaning diarrhoea clones. The observation that individual post-weaning and commensal clones could be segregated suggested that a similar approach based on an analysis of many clones from healthy and scouring weaners may be able to distinguish pigs on their health status. This was successfully tested by analysis of virulence gene profiles of representative subset populations of E. coli recovered from individual animal samples using a replicating machine to entrap bacteria on hydrophobic grid membrane filters (HGMF). This method referred to as population based microbial analysis (PoMA) provided a reliable tool to distinguish between clinical, subclinical and healthy pigs. PoMA was used to determine the population virulence gene profile, generating a host profile for each individual pig. Using the 58 virulence genes analysed, 13 (LT, aah, aidA, cvaC, STb, STa, bmaE, papC, iroNE.coli, papG allele I’, ehxA, cdt & F18) could be used to distinguish between the two groups of piglets (based on Chi-squared analysis). Further statistical analysis to generate a cluster dendrogram, and principle coordinate (PCO) analysis revealed segregation of isolates based on health status (healthy and scouring separated into different groupings). When extended to a group of subclinical animals, the method also successfully clustered these in between the healthy and scouring weaners by principle coordinate analysis PoMA has the hallmarks of a statistically reliable diagnostic tool for the analysis of clinical samples from individual animals, pools of animals and even, pools of samples from different farms for comparative analysis. When applied to newborn piglets with neonatal diarrhoea, PoMA replication on CHROMAgar Orientation medium successfully identified the presence of an unreported population of enterococci that dominated on culture media intended to recover primarily E. coli. Selection of single clones followed by enumeration revealed elevated numbers of enterococci in scouring neonates compared to healthy piglets with E. faecium present at a similar ratio to E. faecalis whilst none of the latter was present in the enterococci isolated from healthy piglets. Since antibiotic resistance of enterococci has recently gained eminence as a threat to human health and because microbial contamination from pig meat can also enter the human food chain, an analysis of porcine enterococci isolates was carried out to compare their virulence genes and antibiotic resistance gene (ARG) profiles to clinical human enterococci isolates. The porcine strains had similar virulence genes to human isolates but fortuitously, none of the isolates examined so far in this thesis contained the vanA or vanB genes for vancomycin resistance. A role for enterococci in neonatal diarrhoea remains to be elucidated.

Download

Share

COinS
 

Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.