Date of Award

Winter 12-15-2019

Author's School

Graduate School of Arts and Sciences

Author's Department

Biology & Biomedical Sciences (Molecular Microbiology & Microbial Pathogenesis)

Degree Name

Doctor of Philosophy (PhD)

Degree Type



Our H. sapiens genome does not encode enzymes required to degrade most of the complex plant-derived polysaccharides present in our diets (e.g., hemicelluloses, xylans, resistant starches, pectins). Our gut microbiomes, which contain ~100 times more genes than our human genome, encode a vast repertoire of carbohydrate-active enzymes that catalyze the metabolism of these dietary glycans to products that benefit members of the microbiota and their human host. Western diets are characterized by a paucity of fruits and vegetables and by extension, dietary fiber. Low-fiber diets, in turn, have been associated with a loss of diversity in the microbiota. This loss of diversity is a characteristic of the microbial communities of obese individuals.Members of the lab have conducted a screen of 34 food-grade fibers in gnotobiotic mice harboring a defined human gut microbial community composed of cultured sequenced bacterial strains and fed a representative USA diet high in saturated fats and low in fruits and vegetables (HiSF-LoFV). This screen, which involved testing the effects of almost 144 systematically manipulated diets on the fitness of community members, was designed to identify components of fibers that targeted ‘beneficial’ saccharolytic Bacteroides species underrepresented in the gut communities of obese (Ob) co-twin members of obesity-discordant twin pairs. Bioactive components of fibers that selectively increase the representation of these taxa, and expression of their beneficial metabolic features, are candidates for incorporation into microbiota-directed food prototypes designed to ameliorate the obesity and metabolic dysfunction phenotypes that are transmitted by Ob co-twin microbiota to recipient germ-free mice and abrogated by the introduction of the targeted Bacteroides.My thesis represents a partnership between food science and microbiome science. It focuses on an analysis of the mechanisms by bioactive components of lead plant fibers emerging from this screen incorporated into food prototypes that affect the structural and functional properties of microbial communities from obese co-twins transplanted into gnotobiotic mice and whether results from these mouse models translate to humans.I colonized groups of germ-free mice with intact uncultured fecal microbiota from obese co-twins in obesity concordant or discordant pairs. I subjected mice colonized with a given donor microbiota to a diet oscillation. The diet oscillation consisted of feeding mice a HiSF-LoFV diet followed by the same diet + one lead fiber, then by a return to the unsupplemented diet, followed by HiSF-LoFV + a second lead fiber, then by a return to the unsupplemented diet, and finished by supplementation with a third lead fiber. The effects of fiber consumption on (i) body weight and composition (the latter defined by quantitative magnetic resonance), (ii) microbiota/ microbiome configuration (sequencing V4-bacterial 16S rDNA amplicons and whole community DNA), and (iii) community metabolism (short-chain fatty acids, monosaccharide, and linkage-analysis of carbohydrates) were defined. I found that supplementation with lead fibers (i) significantly reduced body weight gain, (ii) significantly increase the abundance of targeted Bacteroides plus CAZymes and mcSEED metabolic pathways/modules involved in the utilization of bioactive components of the lead fibers. Higher-order singular value decomposition (HOSVD) analysis of microbiome datasets revealed temporal projections associated with fiber consumption and identified robust responsive, and hypo-responsive donor microbiota and the microbiome features driving the variance.I subsequently tested the effects of one of the lead fibers in two separate pilot clinical studies, one involving female twins discordant or concordant for obesity, including those whose microbiota had been selected for the preclinical gnotobiotic mouse experiments, and the other involving overweight and obese singleton adults from the St. Louis metropolitan region. The study with twins was done supplementing their regular diets with a snack food prototype enriched with the lead-fiber. The other study involved controlled feeding with a diet high in saturated fats and low in fruits and vegetables, and supplemented with the same fiber snacks. HOSVD plus linear models disclosed significant relationships between fiber-snack consumption, changes in microbiota/microbiome structural and functional configurations, and changes in the plasma proteome. Understanding the transformation capacity of the bioactive components of fibers by the gut microbiota and their effects on host health will be critical in the development of microbiota-directed foods (MDFs).


English (en)

Chair and Committee

Jeffrey I. Gordon

Committee Members

Daniel E. Goldberg, Lora L. Iannotti, Samuel Klein, Amanda L. Lewis,


Permanent URL: https://doi.org/10.7936/56v8-q108

Available for download on Friday, December 15, 2119