Extracting detailed metabolic information and connections from mammalian gut microbiomes via metaproteomics

A diverse community of bacteria populates the mammalian gastrointestinal tract. These populations exist in a balance with the host assisting with key functions, particularly metabolism of intractable fibers and immune modulation. Disruption of this balance can lead to diseases such as infection, inflammatory bowel syndrome, and obesity. Common symptoms include chronic pain, chronic inflammation, and altered metabolism. Several taxonomic classifications of bacteria have been associated with these diseases, but Recent studies have indicated that these finding are not always statistically valid. An explanation for this is that microbial communities between individuals and even across time can vary substantially even when the individuals have a similar health status. Microbial function, however, is a promising arena to study disease scenarios. Omics methods, which measure the entire gene content of a community are a particularly powerful set of techniques with which to analyze the potential and active function of microbiome communities. Metaproteomics detects and quantifies proteins directly from environmental samples and can be used to measure gut microbiome functional activity. This dissertation applied the use of LC-MS/MS based metaproteomics and metagenomic sequencing to study gut microbiome function in adult humans with Crohn’s disease, preterm infants with necrotizing enterocolitis, and obese and morphine treated mice. Intense variation across time and individuals was observed at the discrete protein sequence level; however, specific functions such as reactions and metabolic modules were shown to be more conserved. Fully connected metabolic networks and pathways were reconstructed from these metaproteomes, and specific metabolic functions are shown to be affected by necrotizing colitis, diet induced obesity, and morphine. This dissertation makes a major step forward by showing discrete metabolic reactions can be effectively analyzed using metaproteomic data.