Studying Human-associated Microbial communities, in health and disease

We are interested in understanding how individual and persistent human-associated microbial and viral communities affect health. Infection with a bacterial pathogen, vaccination, immune development and even taking a Tylenol does not occur in a vacuum. Dynamic microbial and viral communities constantly inhabit our bodies, encoding the majority of the unique genes that alter these processes. Resident microbial and viral community composition is unique to each human, with strong similarities across families and also ethnicity.
Around the time I finished my PhD in biochemistry, high-throughput sequencing technology emerged, and I developed a strong interest in applying microbial ecology approaches to the study of human-associated microbial and viral communities.

We use a combination of omics techniques to characterize the genetic potential of human-associated microbial and viral communities, and connect this with the active chemical processes in the community. Discoveries based on longitudinal and cross-sectional studies of human samples can be tested in model cultures in the lab. The hypothesis underlying our work is that the unique, persistent microbial communities inhabiting the niches that comprise each human undergo recognizable changes in many human disease states. The microbial community structure and metabolic profiles of altered microbial communities are detectable, and can be used to diagnose and monitor infections and lifestyle disease, and tailor treatments to specific microbial states. I have used culture-independent microbial DNA sequencing in my research to demonstrate the stability and persistence of individual oral microbial communities in healthy people, and to examine the origin of a poorly understood infection common in malnourished children (NOMA). Currently, we are monitoring metabolites from Cystic Fibrosis (CF) breath samples and linking them with metagenomic and metatranscriptomic data from paired sputum samples to elucidate the physiology of microbial and viral communities in the CF lung over time.

In a sense, metabolomics is an ancient approach. For example, multiple accounts starting around 1500 BC used taste or insects to detect excess glucose in urine and diagnose diabetes. Modern versions of these approaches using dogs or bees to detect conditions such as cancer and tuberculosis all rely on the premise that unique patterns of metabolites emerge from disease. Now we have an array of tools that detect the potential and actual processes in an environment, from high-throughput sequencing of DNA and RNA to metabolite detection with chromatography and mass spectroscopy. A large number of genes and potential metabolic processes come to humans from their resident microbial and viral communities. Metagenomic sequencing provides a profile for this potential, while metatranscriptomes reveal the actively transcribed genes. Finally, metabolomes reveal the products of the chemical reactions occurring in a sample, many of which are unique to a microbial community member or metabolic pathway. In addition to potential identification of an infecting microbe, intercepting chemical communication signals between microbes and the human host is essentially eavesdropping on the system, to learn about the active processes in the community. In combination with well-planned study design, this under-explored territory will yield biomarker signals that can be used to diagnose and monitor health conditions at a greater level of clarity (e.g. earlier diagnosis, or more carefully targeted treatment).

Microbial communities and metabolic signatures in Cystic Fibrosis airways

We are studying how microbes and viruses persist in long term infections in the lungs of people with Cystic Fibrosis. The airways of CF patients are characterized by complex polymicrobial infections that are not fully characterized by standard microbial assessments. CF patients experience intermittent pulmonary exacerbations that correlate with accelerated loss of lung function. A primary goal of therapy is to reduce the frequency and severity of these exacerbations, yet clinicians lack biomarkers to detect oncoming disease flares before the symptoms and irreversible lung damage occur. Antibiotic treatment is not typically administered until irreversible damage from inflammation has occurred. We are using a combination of microbial and viral DNA sequencing (metagenomics), community RNA sequencing (metatranscriptomics) and metabolite profiling with GC-MS and LC-MS (metabolomics) to characterize CF lung infections over time. Our long term goal is to find molecules in breath gas samples that are specific to an individual CF patient or disease state (i.e. stable vs exacerbated) that could be used as biomarkers to monitor infection and direct antibiotic treatment. By linking microbial metagenomic sequencing data with volatile metabolites from breath samples (in collaboration with Dr. Simone Meinardi and Dr. Don Blake at UC Irvine), we aim to identify biomarkers of active microbial physiology to detect infections earlier and treat them more specifically. Most of the ~100 biomarkers used in medical diagnostics and disease monitoring are metabolites, or small molecules (<1500 Da) that are produced in the course of metabolism. Metabolites that are specific to particular strains of bacteria, or even interactions between bacteria, are observable with chromatography and mass spectrometry. These molecules may enable detection of specific microbial metabolisms in a particular microbial community and disease state. Linking metabolite data with microbial sequence data will enable us to understand the identity and physiology of the most active microbes. Ultimately, we may unveil mechanisms that explain how bacteria persist in the CF lung. This information could allow doctors to diagnose and treat infections specifically.

Fecal Dark Matter (i.e. characterizing phage genes of unknown function in lean and obese twins)

Most predicted genes in viral metagenomes have diverged so extensively that they cannot be characterized with current bioinformatics. However, they are actually homologs (or functional analogs) of previously characterized proteins, and we can use existing methods to elucidate their function. The goal of this project is to characterize the structure and function of abundant unknown phage genes that are likely to have metabolic functions that are important for viral-host interactions using a combination of computational, structural and physiological approaches.

Structure of Human Oral Microbial Communities

Healthy human oral microbial communities are unique to each person and stable over time; we are using high throughput sequencing and ecological statistics to understand the distribution and composition of bacterial communities across people, families and disease states, such as the facial-gangrene inducing condition that occurs in malnourished humans, called NOMA.

Analysis of a novel bacterial genus and species using high throughput sequencing

A novel bacteria discovered in the lungs of a petshop parakeet in Basel, Switzerland was sequenced and assembled, leading to the designation of a novel bacterial genus and species, Basilia psittacopulmonis DSM24701. Annotation and analysis of this genome, containing a large number of unknown sequences and lacking close neighbors that have been sequenced, led to the exploration of several themes related to how bacterial genomes evolve.

Molecular Basis of Human Disease Book

Conducting interviews with people affected by human diseases, and incorporating them into a book that will be published by Garland Science. The book roughly follows the structure of the course I developed in UC Irvine’s Molecular Biology and Biochemistry department. Syllabus