| dc.description.abstract |
<p>The human gut is home to trillions of microbes that interact intimately with the
host and its diet. An important emergent phenotype by these microbes is colonization
resistance, the process by which a microbial community resists colonization of an
exogenous microbe. This resistance barrier is critical for protecting humans from
infectious enteric pathogens. However, it is detrimental to the deliberate engraftment
of probiotics which are live microbes beneficial to the host. The mechanisms behind
these barriers have been studied extensively, but the microbiome is a network of many
biological nodes and ecological edges that can interact with an invader in numerous
ways. Therefore, defining the precise mechanisms for resistance of a specific pathogen
or probiotic is challenging, due in large to this combinatorial challenge. In the
second chapter of this dissertation, I demonstrate a novel approach that can suggest
key taxa or host factors associated with clinical outcomes of interest including colonization
resistance. In particular, I leveraged a rare prospective cohort study with machine
learning methods for identifying gut bacterial signatures associated with susceptibility
to cholera. I demonstrated that the human gut microbiota can predict the susceptibility
of its host to the diarrheal disease. One of the predictive gut microbes identified
by my model, Paracoccus aminovorans, facilitated the growth of Vibrio cholera, the
etiologic agent of cholera, in vitro. My model also linked gut microbiota structure,
clinical outcomes, and age. This integrative approach suggested that gastrointestinal
immaturity of the host and its gut microbiome may be crucial for resisting colonization
of enteric pathogens. The predictive model was also over-represented with members
of the Bacteroidetes phylum, including several Bacteroides species. These taxa belong
to a genus that is dominant in the gut of human on a western diet. Genomic, biochemical,
and metabolic studies have vastly studied the traits of these Bacteroides species
and how they interact with the host. As obligate anaerobes that are stably colonized
in the human gut, the Bacteroides is an ideal model genus for studying ecological
mechanisms of colonization resistance. In the third chapter of this dissertation,
I developed a high-throughput rapid approach for inferring the relative abundance
of several Bacteroides species in a mixed community grown on single carbon substrates.
I validated the utility of this method by investigating whether Bacteroides species
cooperate or compete when carbon resources are limited. By profiling the growth of
mixed cultures on single carbohydrates, I show that Bacteroides exhibit both patterns
of resource cooperation and competition. Together, these chapters show that development
and application of novel computational and experimental tools can shed light on the
intimate interactions between diet, microbiome, and the host in the context of colonization
resistance.</p>
|
|
