Research Interests

Our research interests are focused on the elucidation of bacterial interactions with environment and other living cells using a variety of systems biology tools. Current directions include:

1) Role of intestinal bacteria in human health and in gastrointestinal diseases such as IBS, colon cancer, and obesity

2) Host-microbial interactions and stress responses of pathogenic and commensal Enterobacteria

3) Use of genetic engineering and mathematical modeling to study biology principles

Below we describe current research directions and projects in more detail

In the healthy adult, there are 10¹¹-10¹⁴ bacteria colonizing the intestine. This outnumbers the total tissue cells in the body by at least an order of magnitude. The composition and activity of this complex microbial system (called microflora or microbiota) have a major influence on health and disease. Commensal microbiota contribute to trophic functions of the gut (producing fermentation products and vitamins that can be used by intestinal epithelial cells), stimulate immune function of the gastrointestinal tract, transform or excrete toxic substances, protect the host against invasion by pathogenic species, and modulate gut motility. At the same time, changes in intestinal microbiota are associated with a variety of GI disorders such as irritable bowel syndrome, inflammatory bowel disease, and colon cancer.

Unfortunately, our knowledge of human intestinal microbiota is far from complete. The relative lack of data has, to a large extent, been due to the difficulty of using standard microbiological techniques to study this complex community. The development of molecular taxonomy tools based on 16S rRNA sequence interrogation has allowed a phylogenetic classification of intestinal bacterial species, and presents an opportunity to design methods for the detection of non-culturable species.

In a recently completed project we have designed, developed, and validated a custom high-throughput phylogenetic Microbiota microarray containing probes to 775 different microbial species (OTUs) of human gut microbiota. The designed microarray has been rigorously validated and we have now shown that this array can be used successfully to interrogate human intestinal microbiota. We have shown that the microarray can be used not only to examine population structure of intestinal microbiota by profiling total genomic DNA isolated from fecal samples and biopsies, but it can also provide quantitative assessment of relative metabolic activities of individual community members through the interrogation of total community RNA.

This Microbiota Array is now used to study the composition of human gut microbiota in a variety of recently completed and ongoing projects. The goal of most of these studies is to test whether gut microbial communities are changed in individuals diagnosed with different gastrointestinal diseases. If such differences are found, this will allow us to make putative associations between disease state and specific microbiota changes which can eventually lead to the development of novel drugs such as pro- and prebiotics.

For example, using PCA statistical analysis of the abundance of different bacteria in the human gut, we can separate fecal samples between healthy adults (aHLT) and children (kHLT), and we can easily distinguish samples obtained from children suffering from IBD (kIBD).

Since each microarray measures the hybridization level of every interrogated sequence, it provides quantitative signal for each examined bacterial species (OTU) in each sample. This feature allows not only direct comparisons among samples but it also provides an opportunity to assess if a particular species is detected in all or most samples. A set of such species can be considered to form a core microbiome of members present in every community of particular human microbiota, which can potentially be attributed to an important role of these species in inter-species and host-microbial interactions. Using Microbiota Array, we have recently profiled 60 human fecal samples among healthy adults and adolescents and among adolescents diagnosed with obesity and diarrhea-predominant IBS. Among all these samples, we defined a robust core of 44 microbial species. Most of the gut microbiota species belonged to the so-called “shared” group, which we defined as those present in multiple but not all samples. As can be observed from the figure, many individual microbiomes contained unique microbiota species that were detected only in that particular sample.

In addition to the phylogenetic analysis of microbial communities, in collaboration with Dr. Nicholas Reo we have also initiated an analysis of metabolic activity of gut microbial communities by using proton NMR measurements of fecal extracts. Similar to our microarray findings, stool samples from patients with a disease can be separated from healthy ones based on the metabolite levels in their stool samples. Since many of these metabolites are produced by resident microbiota, combining microarray and NMR analyses on the same samples will allow us to putatively link metabolite production / consumption and the presence and abundance of different bacterial members.

Finally, to complement our clinical studies we have also established a multi-compartment gut fermentor system. This system consists of three consecutively linked fermentor vessels each aimed to simulate an anaerobic environment of the human ascending, transverse, and descending colon, respectively. Because all parameters of such system can be presicely controlled (such as nutrient availability, transfer rate, and pH of the environment), this lets us design and carry out mechanistic experiments that study microbial ecology or human intestinal communities. This system also allows us to carry out experiments that are not possible to do in humans, such as adding probiotic and other bacterial species to the community and studying effects in real time.

Escherichia coli is a species of enteric bacteria usually living as commensals in human intestinal tract. Though most of E. coli strains are harmless and even beneficial to their human hosts, a number of different pathogenic strains are also known. Pathogenic E. coli are broadly divided into two major categories: the enteric pathogens (mostly agents of diarrhea and colitis) and extraintestinal pathogens. Among the former are enterohemorrhagic E.coli (EHEC) that are associated with bloody diarrhea and hemolytic uremic syndrome; an example of extraintestinal pathogenic E. coli are uropathogenic strains (collectively called UPEC) that are the predominant group causing urinary tract infections and their extensions such as pyelonephiritis.

Most bacteria including E. coli possess ensembles of transcriptionally regulated genes, commonly called stress or shock response systems, that enable them to adapt to changes in the chemical or physical aspects of their environment, including water activity (osmotic pressure), pH, temperature, and oxygen concentration. Adaptation to most environmental shifts generally involves two stages: a transient or acute phase (shock response) that consists of rapid responses needed to initiate the adaptation to the new conditions, and a continuous or chronic phase (stress response) that consists of responses that are needed to support exponential growth, possibly at a new growth rate, in the altered environment.

We have recently profiled the transcriptional response of commensal E. coli K12 to osmotic and heat stresses. Several new findings were uncovered with the most interesting one being up-regulation of many genes of oxidative stress regulon by both the heat and an increased medium osmolarity. This can explain the previously noted cross-protection of sub-lethal osmotic stress against further oxidative and heat stresses.

Our current goal is to study osmotic regulation in uropathogenic E. coli. Urinary tract infections usually develop via ascending route, where cells enter the urinary tract through urethra, and move up to bladder and then kidneys. During this path bacterial cells are subjected to significant osmotic pressure from human urine, and UPEC ability to survive and adapt to such environmental stress is important for infection development. Urine contains significant amounts of both urea and inorganic ions, but we do not know to which extent bacterial cells respond differently to the presence of these compounds in the environment. Though uropathogenic E. coli possess a number of known osmotically regulated systems, these have not yet been shown to be necessary for cell survival in human urine, and further studies are needed. Therefore, we utilize gene expression microarrays, qRT-PCR, and computational approaches to delineate UPEC adapatation to high concentration of urea and inorganic salts in growth environment.

Many biological systems are inherently complex and consist of many hundrends of members or relationships. Comprehensive experimental study of such systems is often impossible or prohibitive. In such situations, mathematical modeling of the system based on limited amount of experimental data can provide insight into the structure and function of these complex systems that otherwise would be difficult to obtain. In collaboration with Dr. Brent Foy we are interested in utilizing mathematical biomodeling approach to gain understanding of microbial systems and processes.

For example, we have developed a mathematical model aimed to simulate multi-template PCR amplification of 16S ribosomal DNA sample (obtained from microbial community) and subsequent detection of these amplified 16S rDNA species by phylogenetic microarray. Many current studies of complex microbial communities rely on the isolation of community genomic DNA, amplification of 16S ribosomal RNA genes, and subsequent examination of community structure through interrogation of the amplified 16S rDNA pool by high-throughput sequencing, phylogenetic microarrays, or quantitative PCR. To build the model we used parameters estimated from the experimental results obtained with Microbiota Array in this model, and focused our attention on how microbial species detection and the accuracy of species abundance estimates depend on the number of PCR cycles used to amplify 16S rDNA. Figure shows the results of our simulations: species detection initially improved with each additional PCR cycle and reached optimum between 15-20 cycles of amplification. The use of more than 20 cycles of PCR amplification and/or more than 50 ng of starting genomic DNA template was, however, detrimental to both the fraction of detected community members and the accuracy of abundance estimates. Overall, the outcomes of the model simulations matched well available experimental data.