FINE ARABINOXYLAN STRUCTURE IMPACTS ON HUMAN GUT MICROBIOTA COMPOSITION AND FUNCTION: POTENTIAL SUBSTRATES FOR MANAGEMENT OF MICROBIOME COMPOSITION AND HEALTH BENEFIT

Complex dietary fibers play an essential role in modulation of gut microbiome structure and function, as they serve as the primary energy source for colonic bacteria that convert these substrates into host-beneficial metabolites. To determine how diverse polysaccharide structures would differentially influence gut microbiota, I developed a 7-day sequential in vitro fermentation assay to reveal the core microbial consortia involved in fiber consumption. I found that diverse microbial communities were sustained by complex arabinoxylans, in comparison to other, simpler fiber structures. The high dilution rate sequential passage cultivation introduces bottleneck effects, which enrich for the organisms most capable of rapid growth on a substrate, which in this case entailed retention of hydrolytic capacity required to consume polysaccharides. Rather than selecting for a sole generalist, these sequential fermentations retained a substrate-specific and membership-stable microbial consortium. Based on this model, I performed a series of studies focused on the influence of fine structural differences of arabinoxylans on the composition and function of the selected consortia. First, I tested the hypothesis that two complex polysaccharides with subtle fine structure differences would select distinct consortia across the different individual donors’ microbiota serving as inocula. To do so, I extracted two sorghum arabinoxylans with subtle structural differences and used them as substrates in a 7-day sequential passage fermentation with four fecal inocula. The results exhibited highly deterministic effects on microbial community compositions and metabolic responses, with respect to diversity, phylogeny, and short-chain fatty acid production. Metagenomic sequencing revealed the genomic potential of two distinct arabinoxylan-fermenting consortia and revealed that the key substrate-degrading species in particular, Agathobacter rectalis versus Bacteroides spp., encoded distinct suites of carbohydrate-active enzyme genes. To further investigate how a precise linkage structure changes would impact microbial compositions, I applied an enzymatic debranching process to remove linkages from one sorghum arabinoxylan and, thereafter, compared the effect of different branching densities of arabinoxylans in governing fecal microbiota fermentation. Debranched arabinoxylan selected a Bacteroides ovatus dominated consortium, whereas native sorghum arabinoxylan selected for large populations of A. rectalis, suggesting a strong deterministic effect of carbohydrate linkages in controlling the microbial ecology of gut-derived communities. Highly stringent fiber fermentation responses to fine structure were found regardless of individual fecal inocula even across two distinct “enterotypes” (i.e. [Ruminococcus + Bacteroides] versus Prevotella). My findings suggested that dominance of enterotype species was ephemeral in this system, as the most efficient glycan-consuming species arose under continued selection in consortia.

Differences in an organisms’s competitiveness for a specific fiber can arise either due to differences in genome-encoded potential or in regulation of polysaccharide-degradation genes. Consequently, I used a model polysaccharide degrader, a Bacteroides cellylosilyticus strain (a key generalist consuming complex arabinoxylans to determine how its carbohydrate-active enzymes were expressed in response to fine fiber structural differences. Native corn bran arabinoxylan, a key substrate of B. cellulosilyticus, with a complex and heterogeneous chemical structure and a very highly complex corn arabinoxylan subunit (DBH) were extracted as substrates to test the polysaccharide utilization loci (PULs) expression response by B. cellylosilyticus. Remaining polysaccharide structures were quantified over time to correlate with gene transcription levels. Different prioritization of polysaccharide degradation genes were found amongst the two substrates, suggesting that B. cellylosilyticus expression of hydrolytic genes was altered by the accessible polysaccharide structures. This result implied that dynamic substrate degradation strategies may be chosen by a gut generalist when confronted with different fine fiber structures. This difference in microbial behavior could potentially influence the competitiveness of the key bacteria response to different arabinoxylan fine structures.

Finally, I tested whether continuous feeding of arabinoxylan would have significant impacts on gut microbiome structure and function in vivo and whether co-administration of an arabinoxylan fiber with its adapted core consortium would improve the ability of specialist microbes to colonize the gut. I administered sorghum arabinoxylan to C57BL/6 mice as a prebiotic, a sorghum arabinoxylan-consuming consortium as a probiotic, and a combination of the two as a synbiotic. Further, I treated other groups of mice with antibiotics to determine whether the ability of the human-derived consortium to engraft in the mouse gut depended upon resistance from the native microbiome. My results showed that human-derived microbiota failed to engraft in mouse gut under any tested condition, but I unexpectedly revealed that administration of arabinoxylan and its adapted consortium greatly improved post-antibiotic microbiota resilience in sex-dependent ways. Together, these studies revealed tight interactions among fiber structures and gut microbial community structures and function across individuals and suggested that these relationships might be leveraged in using fine fiber structures and/or adapted microbial consortia to precisely and predictably manipulate the gut microbiome towards health.