The astounding complexity of microbial communities limits the ability to study the role of genomic diversity in shaping the community composition at the species level. With the advancement and increased affordability of high-throughput sequencing methods, it is increasingly recognized that genomic diversity at the sub-species level plays an important role in selection during microbial community succession. Recent studies using the cpn60 universal target (UT) have shown that it is a high-resolution tool that provides superior resolution in comparison to 16S rRNA based tools and can predict genome relatedness. However, studies to characterize the nature and degree of genome content differences predicted by cpn60 UT sequence relationships have not been conducted. In this work, we focused on the Enterococcus community obtained from feces of healthy pigs. Enterococci are both accessible with selective culture, and interesting since the genus includes members that are significant human pathogens and others that are used as probiotics. Previous work has shown that cpn60 UT sequences can resolve pig fecal Enterococcus faecalis and E. hirae into phylogenetically and phenotypically distinct ecotypes. The utility of cpn60 UT sequences for resolution of Enterococcus species was first established in the year 2000, and this demonstration included 17 Enterococcus species. We have expanded the analysis to include all currently recognized Enterococcus species and confirmed that cpn60 UT sequences provide higher resolution than 16S rRNA sequences for identification of Enterococcus species. In addition, we showed that cpn60 UT sequences could resolve sub-groups in E. faecium consistent with results obtained from comparison of whole genome sequences. GTG rep PCR based clusters for E. faecalis and E. hirae isolates were generally consistent with the cpn60 defined Enterococcus ecotypes observed in our previous study, suggesting that cpn60 UT sequences predict overall genomic relationships. Results from analysis of CRISPR sequences provided insights into the extensive networking and transfer of genetic material that takes place within the intestinal Enterococcus community. We conducted whole genome sequencing of representative isolates to characterize further the genomic diversity in cpn60 defined E. hirae ecotypes to determine the nature and degree of genome content differences predicted by cpn60 UT sequences. Differences in phosphotransferase systems, amino acid metabolism pathways for glutamine, proline and selenocystiene, potassium-transporting ATPases, copper homeostasis systems and putative prophage associated sequences, CRISPRs and antibiotic resistance genes were observed. Results from in vitro growth competition assays showed that isolates from E. hirae-1 and E. hirae-2 were able to out-compete isolates from the E. hirae-3 ecotype, consistent with the relatively low abundance of E. hirae-3 relative to E. hirae-1 and E. hirae-2 previously observed in the pig fecal microbiome, and with observed gene content differences between the ecotypes. Results presented in this thesis provide a genomic basis for the definition of ecotypes within E. hirae and confirm the utility of the cpn60 UT sequence for high resolution profiling of complex microbial communities.

Microbial succession, composition and ecological distribution within the gastro-intestinal tract are critical areas of study since commensal bacteria have been shown to affect animal health and development. A series of experiments were conducted to determine whether altered microbial succession in neonatal animals would modulate the development and health of pigs later in life. An initial experiment in conventional pigs was conducted to establish the early postnatal microbial succession profile and to identify early colonizing bacterial species. Culture-independent analysis of digesta and mucosal microbiota showed distinct variation between the proximal and distal gastro-intestinal tract (GIT) indicating that fecal or distal gut profiles cannot be used to predict succession in the upper GIT. Temporally, Clostridium spp. were found to be most prevalent in the GIT microbiota of the neonatal pig up to 0.5 d of age, accompanied by a high abundance of Escherichia and Shigella spp. These genera were transiently displaced by Streptococcus spp. followed by a preponderance of Lactobacillus spp. between 3 and 20 d of age. Subsequently, a “snatch-farrow” model was employed to modulate early postnatal microbial succession and investigate the effects on postweaning microbial composition. Pigs were collected into sterile towels directly from the vaginal canal and transferred to a sterile isolator environment for the first 4 days. Pigs were either inoculated with sow feces or not at 1 d of age resulting in significant differences in fecal microbial profile at 4 days of age, prior to removal from isolators. Analysis using terminal restriction fragment length polymorphisms (TRFLP) of intestinal microbiota at 28 d of age did not show significant clustering or variation in diversity indices for either group during the 4-d postnatal isolator phase. However, enumeration of selected taxa using quantitative PCR did indicate significant treatment differences in postweaning microbiota. Despite these results, this approach was rejected for further use as the protocol provided only moderate control of early postnatal colonization and variation and unpredictability of the timing of natural farrowing contributed to significant litter effects. Finally, a gnotobiotic monoassociation model was used investigate the effects of modulating early postnatal microbial succession on postweaning physiology, microbial composition and mucosal gene expression. Twenty-four cesarean-section derived piglets were monoassociated for the first 4 days of life with either L. mucosae (L), S. infantarius (S), C. perfringens (C) or E. coli (E). Pigs from treatments E and L animals showed the highest growth rate during the conventional rearing period (7-28 d of age). Monoassociation with different bacterial species during the first 4 d of life resulted in significant changes in postweaning microbial composition in small intestine and colon as assessed by quantitative PCR, although TRFLP did not identify unique clustering by treatment or variation in diversity. L. mucosae was the only inoculant species with significant variation, with a reduction in the colonic mucosa at 28 days of age. Monoassociation with L. mucosae was also associated with increased nutrition related gene expression in small intestine. Pigs monoassociated with E. coli had low expression of microbial sensing (TLR2 and 4), NFkappaB complex genes and mucins at 28 d of age. This study clearly showed that controlled early microbial succession in neonatal pigs altered post-weaning commensal microbiota composition, postweaning physiology and host gene expression in small and large intestine. The findings suggest the importance of peri-natal management and feeding strategies in promoting postweaning health and performance.