UC Irvine

As the world continues to industrialize, there has been a decreased incidence of infectious diseases and an increased incidence in non-communicable chronic diseases, such as obesity and cancer. One potential causal factor is the human microbiome, which refers to the collection of commensal bacteria, fungi, viruses, archaea, and other microorganisms that inhabit our bodies. The human microbiome has a collective genome which outnumbers human genes by 150-to-1, and it produces an expansive repertoire of metabolites which affect health. As such, the gut microbiome has major implications in digestion, educating the host immune system, preventing the colonization of pathogenic bacteria by occupying intestinal niches, and more.

Several components of the industrialized lifestyle have been known to alter the gut microbiome, leading to negative health consequences. The result is a microbiome that promotes inflammation and is associated with non-communicable chronic diseases. As these diseases become more prevalent in industrialized countries, there is an urgency to understand the role of the microbiota in human health. To investigate this issue, my research seeks to answer the following three questions: (1) What is the role of the industrialized gut microbiome in the development of colorectal cancer? (2) How does diet affect the gut microbiome and inflammation in people with myeloproliferative neoplasms? (3) How is dietary fiber utilized by the gut microbiota of US and Moroccan individuals in vitro?

To answer the first question, samples were collected at the UCI Medical Center from 140 subjects during and after colonoscopy to characterize the microbiome associated with colorectal polyps. I used a combination of amplicon and shotgun metagenomic sequencing approaches to understand the effect of sampling method on microbial composition, and to describe the microbiome associated with healthy tissue and two types of colorectal polyps. I discovered that sampling method significantly explained 10-15% of the variation observed in microbiome composition. Additionally, using samples obtained from the colon mucosa, I was able to find associations with microbiome composition and colorectal polyps derived from the serrated pathway of colorectal carcinogenesis, such as a depletion in the lignan-degrading microbe, Eggerthella lenta. Lastly, I was able to use the microbiome to inform machine learning classifiers to accurately distinguish between healthy and polyp-bearing samples (Area under curve = 0.87-0.99).

To address the second question, we collaborated with Dr. Angela G. Fleischman, who conducted a dietary intervention in subjects afflicted with myeloproliferative neoplasms, a class of blood cancer. In this 15-week clinical trial, 28 individuals were assigned to receive dietary counseling following either a Mediterranean style eating pattern or one following U.S. guidelines. Blood and fecal samples were collected to examine inflammation and the gut microbiome, respectively. Using shotgun metagenomic sequencing, I discovered no significant alterations in gut microbiome diversity and composition due to a Mediterranean diet. I did find a significant association between the gut microbiome and myeloproliferative neoplasm subtype, explaining approximately 6% of the variance in microbiome composition. Lastly, I found several significant correlations between microbial species, function, and cytokine concentrations.

My first two chapters suggested a link between the microbiome and dietary fiber, thus, in my final chapter I explored the effect of dietary fiber on microbial growth in vitro. In these experiments, I took the feces of 15 healthy individuals from the US and 15 Moroccans and cultured them anaerobically for 24 hours in the absence and presence of inulin, pectin, or psyllium husk. I found significant cohort effects on the microbiome, with US samples becoming dominated with Bifidobacterium and Moroccan samples becoming dominated with Clostridia. Furthermore, I demonstrate that pectin and psyllium husk perform differential enrichment of microbes and their associated carbohydrate-active enzymes.