METAGENOMIC METHODS TO INVESTIGATE MOBILE ELEMENT CONTEXT AND NASCENT TRANSCRIPTION IN THE HUMAN GUT MICROBIOME

116 pagesMicrobial systems are in continuous flux. Considering the human gut microbiome, changes in community composition are associated with differences in host immune regulation, metabolic function, and a litany of physiological metrics. A growing number of diseases can be etiologically explained by the presence of some strains or the absence of others, though myriad conditions associated with the microbiome cannot be cleanly attributed to the actions of single organisms. Indeed, the true nature of the human microbiome is analogized best as a tangled web of ecological co-dependency, a careful balance between resource antagonism and symbiotic negotiations at the Maginot Line of the gut epithelium. Short-read sequencing is an invaluable tool for examining the nucleic acid content of microbiome samples, and the tools of metagenomic assembly are getting ever-better at partitioning reads into near-complete pseudo genomes. However, many important microbial genes are found on genetic constructs that are readily shared between bacterial cells. These constructs, called mobile genetic elements (MGEs), are difficult to assemble, and their promiscuity confounds reference-based mapping of taxa to functions. Getting to the ground truth of gene-taxa pairings requires that we extend classic metagenomic sequencing to retain information about in situ MGE context. Of course, the carriage of a particular gene does not tell us to what extent a gene is expressed in the gut niche, so metagenomic techniques must be paired with tailored transcriptomics methods to ultimately draw causal links from genes to bacteria to human cells. In this dissertation, I present the application of two sequencing techniques, Hi-C and PRO-seq, to human microbiome samples, with the goal of contributing a partial framework for gaining greater insight into the tripartite interaction between bacterial cells, mobile genetic elements, and the human that encapsulates it all. The human gut microbiome is a reservoir of antibiotic resistance genes (ARGs) that can be accessed by pathogens via horizontal gene transfer, leading to multidrug-resistant infections. Metagenomic short-read sequencing can reveal community composition and the presence of ARGs, but assembly alone is insufficient to link ARGs on extrachromosomal elements with their host strains, and culture of ARG-containing gut microbes is complicated by specific nutritive requirements and low oxygen tolerance. In Chapter 2 of this dissertation, I will discuss the application of metagenomic proximity ligation to probe the microbiomes of neutropenic patients with hematologic malignancies. Broadly, we observe individualistic networks of mobile gene carriage and increased exchange of antibiotic resistance genes in the guts of hospitalized patients, with implications for understanding the emergence of multi-drug resistant Enterobacteriaceae. Transcriptional analyses of mixed bacterial communities can give valuable insights into the ecological and metabolic interactions of neighboring species. However, RNAseq of microbiomes is confounded by variable efficiency of ribosomal RNA depletion across organisms and the short half-lives of most bacterial mRNAs, meaning that only robust transcriptional changes are typically observed in bulk meta-transcriptomic experiments. In Chapter 3, I discuss the application of a minimally modified precision run-on sequencing protocol (PRO-seq) for run-on transcription from engaged prokaryotic RNA polymerase, allowing for the biotinylation and capture of nascent bacterial transcripts. We show that PRO-seq is replicable in both E. coli and diverse members of the human gut microbiome, and that PRO-seq gives information beyond that of RNAseq concerning RNA polymerase dynamics at metagenomic loci