Degree Name

Doctor of Philosophy


School of Biological Sciences - Faculty of Science


The mechanism of action of probiotics is based on competitive exclusion and immune modulation. However, the literature is scant on supporting data because of the failure to adopt a systems approach to probiotic functionality. This has been partially addressed in this thesis by taking into consideration the tripartite interaction between bacteria and bacteria in the enteric community; between bacteria and the host animal and finally, between the host immune response (innate or acquired) on the plethora of microbes that inhabit the gastrointestinal tract. A trial involving newly inducted cattle in a feedlot, formed the basis of initial attempts to assess the benefits of a commercial probiotic formulation - Protexin on intestinal health by enumeration of a select subset of cultivable bacteria species and by assessment of immune modulation. The results failed to demonstrate a significant change in the population dynamics of cultured faecal microbes but did show that Protexin stimulated immune responsiveness in T cells. Carcass analysis demonstrated a significant reduction in marbling or intramuscular fat deposition. In the course of examining the faecal microflora from feedlot cattle, the presence of high levels of Bacillus spores suggested that one possible reason for the lack of a growth benefit may be attributed to a high endogenous level of bacilli. Since there were no reliable methodologies for identifying Bacillus species, an alternative procedure was developed involving amplified ribosomal DNA restriction analysis (ARDRA). With this protocol, we were able to show that cattle faeces contained large numbers of Bacillus spores representing different mesophilic species, where B. subtilis, B. licheniformis and B. clausii dominated. The presence of a stable population of coliforms in cattle faeces that was not altered by probiotic feeding highlighted the importance of developing better techniques to characterise diversity in E. coli, a potential food-borne pathogen of economic significance to the cattle industry. The use of virulence genes to genotype coliforms provided a method for differentiating between pathogenic, clinical and commensal isolates of E. coli. Altogether, a combination of uni- and multiplex PCR assays was developed to screen for 50 virulence genes (VGs) from 8 pathotypes of E. coli. There was a significant association between phylogroupings and VG ownership. This result showed clearly that the lack of or possession of VGs in member isolates of each phylogenetic group can be used to assess diversity and potential pathogenesis of E. coli. To understand better the importance of pathogenic enteric coliforms, an alternative animal model involving pigs with post-weaning diarrhoea was used to investigate the relationship between pathogenicity and commensalism by VG profiling. Porcine enterotoxigenic E. coli (ETEC) were found to carry VGs identified in E. coli that cause extraintestinal infection. Furthermore, by using the appropriate methods of statistical analysis, VG profiling had the capacity to predict the pathogenic and commensal status of individual clones. By developing the capacity to rapidly characterise and genotype virulence and commensalism in E. coli, it is now feasible to examine how probiotic feeding can modulate the population dynamics of different community members in pigs with enteric disease, as well as changes in the coliform populations. Finally, another arm of the tripartite interaction involving bacteria and host interaction was modelled in vitro by examining the primary signalling events between bacteria and intestinal epithelial cells. These investigations focused on the judicious selection of T84 as the reporter intestinal epithelial cell line because of low level expression of inflammatory transcripts from 6 other epithelial cell lines. Using a panel of coliforms genotyped for virulence or lack of virulence, the signalling events that followed on from the primary interaction between bacterium and cell, showed there was a lack of correlation between VGs and gene activation. Nonetheless, all the coliform strains tested varied in their capacity to signal transduce T84, confirming that this differential bioactivity can be exploited in the ranking of candidate probiotic strains. The differential responses seen with different E. coli strains and the lower and more consistent activation patterns recorded by LABs for both cytokine and chemokine gene activation, demonstrate that a semi-quantitative ranking of microbial bioactivity can be obtained. Such an approach if adopted in conjunction with an even wider panel of genes in a standardised in vitro environment can provide invaluable information on the selection of appropriate strains to be further tested in vivo.

02Whole.pdf (2457 kB)