Community and genomic analysis of the human small intestine microbiota

Our intestinal tract is densely populated by different microbes, collectively called microbiota, of which the majority are bacteria. Research focusing on the intestinal microbiota often use fecal samples as a representative of the bacteria that inhabit the end of the large intestine. These studies revealed that the intestinal bacteria contribute to our health, which has stimulated the interest in understanding their dynamics and activities. However, bacterial communities in fecal samples are different compared to microbial communities at other locations in the intestinal tract, such as the small intestine. Despite that the small intestine is the first region where our food and intestinal microbiota meet, we know little about the bacteria in the small intestine and how they influence our overall well-being. This is mainly attributable to difficulties in obtaining samples with the small intestine being located between the stomach and the large intestine. Therefore, the work in this thesis aimed at providing a better understanding of the composition and dynamics of the human small intestinal microbiota and to provide insight in the metabolic potential as well as immunomodulatory properties of some of its typical commensal inhabitants. Small intestinal samples used in the work described in this thesis were collected from ileostomy subjects, individuals that had their large intestine surgically removed and the end of the small intestine connected to an abdominal stoma, providing access to luminal content of the small intestine. Considering the importance of molecular techniques in contemporary ecological surveys of microbial communities, first of all, two technologies, barcoded pyrosequencing and phylogenetic microarray analysis were compared in terms of their capacity to determine the bacterial composition in fecal and small intestinal samples from human individuals. As PCR remains a crucial step in sample preparation for both techniques, the use of different primer pairs in the amplification step was assessed in terms of its impact on the outcome of microbial profiling. The analyses revealed that the different primer pairs and the two profiling technologies provide overall similar results for samples of fecal and terminal ileum origin. In contrast, the microbial profiles obtained for small intestinal samples by barcoded pyrosequencing and phylogenetic microarray analysesdiffered considerably. This is most likely attributable to the constraints that are intrinsic to the use of the microarray to enable the detection of predefined microbiota members only, which is due to the probe design that is largely based on large intestinal microbiota communities. However, the pyrosequencing technology also allows for identification of bacteria that were not in advance known to inhabit our intestinal tract. The pyrosequencing technology was used as the method of choice to study the total and active small intestinal communities in ileostoma effluent samples from four different subjects through sequencing the 16S ribosomal RNA gene (rDNA) and ribosomal RNA (rRNA) contentcombined with metatranscriptome analysis by Illumina sequencing of cDNA derived from enriched mRNAof the same sample set to investigate the activities of the small intestinal bacteria. The composition of the small intestinal bacterial communities as assessed from rDNA, rRNA, and mRNA patterns appeared to be similar, indicating that the dominant bacteria in the small intestine are also highly active in this ecosystem. Streptococcusspp. were among the bacterial species that were detected in each ileostoma effluent sample, albeit that their abundance varied greatly between samples from the same subject as well as samples from different subjects. Veillonellaspp. frequently co-occurred with Streptococcus spp., indicating that the Streptococcusand Veillonellapopulations play a prominent role in the human small intestine ecosystem and their co-occurrence suggests a metabolic relation between these genera. Therefore, cultivation and molecular typing methodologies were employed to zoom-in on these groups, which revealed that the richness of the small intestinal streptococci strongly exceeded the diversity that could be estimated on basis of 16S rRNA analyses, and could be extended to the genomic lineage level (anticipated to resemble strain-level). From ileostoma samples 3 different Streptococcusspecies were recovered belonging to the S. mitisgroup, S. bovisgroup, and S. salivariusgroup, which could be further divided in 7 genomic lineages. Notably, the Streptococcuslineages that were isolated displayed distinct carbohydrate utilization capacities, which may imply that their growth and relative community composition may respond quite strongly to differences in the dietary intake of simple carbohydrates over time. This notion is in good agreement with the observation that the Streptococcuslineage populations fluctuated in time with only one Streptococcuslineage being cultivated from both ileostoma samples collected in a one-year time frame. Conversely, the cultivated Veillonellaisolates from samples during that same time-interval consistently encompassed a single lineage. Furthermore, this Veillonellalineage could be isolated from both the oral cavity as well as the ileostoma effluent. Analogously, three Streptococcuslineages that belong to a single phylotype also appeared to be present in bacterial communities from the oral cavity as well as the small intestine. These observations suggest the representatives of the Veillonellaand Streptococcusgenera that are encountered in the oral and small intestinal microbial ecosystems are closely related and indicate that the oral microbiota may serve as an inoculum for the upper GI tract. The metabolic capacity of 6 small intestinal Streptococcus lineages, that were obtained from a single ileostoma effluent sample, was further investigated through the determination of genomic sequences of these lineages. The small-intestinal Streptococcusgenomes were found to encode different carbohydrate transporters and the necessary enzymes to metabolize different sugars, which was in excellent agreement with what carbohydrates could be used by representative strains of the Streptococcuslineages. To further our understanding how the different streptococci as representatives of the dominant small intestinal bacterial populations may influence our immune system, human dendritic cells were stimulated with strains of the different Streptococcuslineages to study their immunomodulatory properties. The Streptococcuslineages differed significantly in their capacity to modulate cytokine responses of blood-monocyte derived immature dendritic cells. As Streptococcusand Veillonellafrequently co-occur in the small intestinal ecosystem, pair-wise combinations of strains of these two species were also tested for their combined immunomodulatory properties. This resulted in considerably different cytokine responses as those that could be predicted from the stimulations with either Streptococcusor Veillonella, indicating that it is not trivial to predict gut mucosal associated immune responses and thatthe composition of the intestinal microbiota as a whole may have a distinct influence on an individual’s immune status. In conclusion, the work described this thesis provides an expansion to the accumulating knowledge on the human intestine microbiota. Whereas most studies focus on the microbiota present in the distal regions of the intestinal tract, this study targeted the microbiota of the poorly proximal regions of the intestine and also addressed its capacity to interact with the local mucosal tissue. The data presented here can be exploited to guide the design of future studies that aim to elucidate the interplay between diet, microbiota and the mucosal tissues in the human small intestinal tract.