1. ABI Innovation: A new computational framework for the prediction of microbiome dynamics
Source: NSF-ABI (DBI-1458347)
Start – End: 09/01/2015 – 08/31/2019
Abstract: Dynamics of polymicrobial communities play a fundamental role in the functioning of many natural, engineered and host-associated systems. Even though the application of DNA sequencing technologies has allowed profiling the response of these communities to external perturbations, the resulting knowledge stems from descriptive and correlation-based analysis of these data. This strongly limits our understanding of the ecology (e.g. how the microbes interact) of these systems and more importantly does not allow us to make predictions. In this proposal we will overcome this limitation. By leveraging on our recently published method for inference of ecological structure from time-series abundance data, we will deliver the first computational method that allows for simulating and predicting microbiome dynamics consistent with metagenomics observations. This will be achieved by the development of novel regularization-based algorithms for model parameters inference and by the application of tools for metagenome reconstruction based on model predictions and phylotypes with genomes that have been fully sequenced. Method testing and validation will represent major part of the proposed studies. Validation against data from simple in silico and in vitro microbial ecosystem of known ecological structure will allow determining accuracy of the proposed approaches not only in predicting temporal dynamics but also in recovering the correct interaction microbial interaction network and response to external perturbation. Method application (and validation) on different metagenomics datasets from our collaborators at Memorial Sloan-Kettering Cancer Center will allow for testing fundamental hypothesis about the role of the intestinal microbiome in resisting foreign pathogen colonization and in shaping host immunity.
2. Pipeline for probiotics discovery, engineering and manufacturing
Source: UMass President Sci. Tech.
Start – End: 07/01/2017 – 01/31/2019
Abstract: PIs at UMass Dartmouth and UMass Medical School will design and produce novel proprietary probiotic cocktails with inducible and tunable gene expression to improve the health of humans, animals and plants, as well as to produce new forms of energy or value-added products. There are millions of bacteria present in each gram of a backyard soil, in a few milliliters of ocean water, and covering the inside and the outside of the entire human body. It is now known that the output of the environment-host-microbe interaction can be exploited to improve health, produce new forms of energy and make new value-added products. To date, very few laboratories are able to construct processes and achieve scientific milestones needed to rationally design, produce and optimize bacterial consortia for these types of applications. Failures have occurred because current probiotic discovery performed in the private sector lacks computational, synthetic and systems biology tools (and their ad hoc integration) to finely optimize these products. Companies are still currently focused on “more achievable” aims, such as finding novel uses of well-studied organisms or ways to manipulate the natural microbiome. Our proposed Probiotic Pipeline will combine microbiome discovery approaches with genetic engineering and synthetic biology to provide private companies with lyophilized bacterial assemblies whose behavior in a system can be finely tuned by an end user. We will leverage the multi-year expertise in bacterial genetics, microbiome biology and microbial bioengineering from three established laboratories at UMass Dartmouth (Bucci, Silby and Brigham) combined with the expertise of our clinical collaborators at UMass Medical School (McCormick, Haran, Ward), to propose a first of its kind “Probiotic Discovery, Engineering, and Manufacturing Pipeline” (PDEMP) that will build this capacity. (See conceptual diagram in Fig. 1). Laboratories at UMass Dartmouth will be responsible for bacterial discovery (Bucci/Silby/Brigham), genetics (Bucci/Silby/Brigham), in vitro (Silby/Brigham), and in silico (Bucci) optimization, while the UMass Medical collaborators will be responsible for testing in vivo (McCormick/Ward) and in the clinics (Haran). For this S&T proposal, we propose to launch two pilot projects aimed at curing (P1) Enterobacteriaceae spp. and (P2) Clostridium difficile infections in the inflamed intestines.
3. Exploring the potential impacts of bioactive compounds from cranberries on colon and gut health
Source: Albert Charitable Trust Fund & match from Massachusetts Department of Public Health
Role: Co-PI (Neto)
Start – End: 01/01/2018 – 12/31/2018
Abstract: Cranberries are a native crop to Massachusetts, widely grown across the Northern US and Canada, and considered both food and a botanical with potential health benefits including urinary tract health associated with its antibacterial properties. Multiple compounds in cranberries have been reported to reduce tumor cell growth and proliferation and alter signaling processes in cells. Colon cancer is the third most common cancer type, and its development has been linked to lifestyle and dietary habits. Inflammation such as that associated with colitis and Crohn’s disease is also a risk factor. Consuming foods high in anti-inflammatory and antioxidative compounds such as polyphenols may therefore provide a chemo-preventative strategy to reduce risk of developing these conditions. Cranberries are rich in polyphenols that interact with a wide variety of bacteria including gut microbes that cause UTIs and other health conditions by reducing adhesion, biofilm, co-aggregation. Preliminary studies through the UMass Cranberry Health Research Center demonstrated that in a mouse model of inflammatory colon cancer/colitis, feeding cranberry powder reduced tumor number and size, as well as reducing tissue inflammation. Because cranberry is thought to alter gut microbiome, such interactions could have far-reaching effects on many health conditions and involve a variety of interactions between metabolites, microbes, and the immune system. In the studies of this project we propose to: (1) establish firmer links between bioactive properties and phytochemical composition of cranberries, understand synergistic effects, metabolites, factors influencing cranberry quality, (2) establish strong data on how cranberry influences the gut – microbial community, inflammatory processes in the tissues, the role of the immune system, and tumorigenesis using established cellular, animal models., and (3, long-term) Use this information to design clinical studies on cranberry’s potential health roles beyond UTIs, including health conditions such as colon cancer, colitis, and Crohn’s disease.
4. Mechanisms of phage resistance evolution and microbial community engineering in bacterial biofilms
Role: Co-PI (Nadell, Dartmouth College)
Start – End: 09/01/2018 – 08/31/2022
Amount: $247,521 to UMass Dartmouth
Abstract: Bacteria mostly live in densely-populated sessile communities, termed biofilms. Biofilms are a fundamental feature of natural microbial life, as well as a root cause of virulence and stress tolerance among bacterial pathogens. In their natural environments, bacteria are frequently attacked by bacteriophages (phages), which are viruses that are a major driver of bacterial death and horizontal gene transfer. Though biofilms have been studied for several decades, and phages have been studied for a century, the interaction between these two ubiquitous features of bacterial life is largely unexplored at the cell-length spatial scales on which phage infection occurs and spreads through bacterial populations. Thus, we know very little of how phage epidemics occur in bacterial biofilms, a gap in our understanding of microbial ecology, evolution, and community engineering. Here we aim to fill gap, using two unique tools we have recently developed: spatial simulations that capture the key features of biofilm growth and phage propagation/diffusion, and a novel experimental system in which a synthetic approach is used to track phage infection in live biofilms in space and time at high resolution. These approaches will be combined to yield new fundamental understanding of phage-bacteria interaction and coevolution in biofilms, with impacts ranging from synthetic microbial community engineering to development of novel antibiotic-alternative therapeutics.
1. Mathematical modeling from metagenomics – minimizing enteric infections risk
Source: NIH-R15 (1R15AI112985-01A1)
Start – End: 02/01/2015 – 06/31/2018
Abstract: Enteric infections represent a critical issue in today’s healthcare. Recent analysis of DNA sequencing data has demonstrated that such infections are associated with the prophylactic treatment with broad-spectrum antibiotics. This is due to their role in killing the native intestinal microbiota, which normally antagonizes pathogens. Computational analysis of these data should facilitate the optimization of antibiotic and fecal transplantation strategies. This is not yet possible because the currently used methods are based on correlations. The main goal of this project is to combine recently developed and novel mathematical modeling tools with metabolic pathways inference and experimentation to predict the risk of enteric diseases and to prototype rationally designed fecal transplantation therapies to minimize it. Leveraging on preliminary work, the PI and collaborators propose to: predict all the stable microbiota profiles mediating colonization by clinically–relevant enteric pathogens using 16S rRNA-constrained mathematical models; combine hazard regression modeling with microbiota dynamics predictions to evaluate the risk of enteric infections in hospitalized patients; determine microbial metabolic pathways associated with the interactions between native intestinal bacteria and enteric pathogens; prototype modeling-based fecal transplantation strategies by experimental validation of modeling predictions. The design of rational therapies minimizing the incidence of enteric diseases depends on our understanding of the dynamics regulating the intestinal microbiota. For this reason, the proposed research is timely and relevant to the mission of the NIAID. The application of new predictive models to DNA sequencing data from a large population of hospitalized patients will allow identifying microbiota states with probiotic (and dysbiotic) properties to be targeted by therapies. The forecasting of microbial dynamics, combined with novel statistical models based on risk analysis, will deliver the first computational tool for monitoring the risk of enteric diseases in quasi-real time. The application of metabolic reconstruction methods to the mathematical modeling predictions will provide new insights about potential metabolic mechanisms regulating and responsible for the predicted stability of pathogen-refractory and compatible stable steady states. The experimental validation of the modeling predictions, not only will allow evaluating the predictive power of the developed mathematical frameworks, but also will provide the opportunity to test the efficacy of the proposed rationally designed fecal transplantation strategies.
2. Propane synthesis by engineered microbial consortia
Source: UMass Dartmouth MSFP
Role: Co-PI (Brigham)
Start – End: 01/01/2017 – 06/30/2017
Abstract: The work proposed here is a study of the metabolic engineering of a versatile bacterium, R. eutropha, to produce propane a novel value-added product. The tasks in this proposal will be performed with the goal of establishing a cradle-to-gate process of microbial propane synthesis using waste carbon feedstocks. Another main objective of the proposed study is the creation of opportunities for undergraduate STEM education. Many of the tasks listed below are ideal for the creation of undergraduate research projects. In this proposed study, the integration of biology, chemistry, and engineering presents opportunities for undergraduate and graduate students to work as part of a multidisciplinary team.
3. Gut microbiome modeling to predict Clostridium difficile colonization in hospitalized patients
Source: UMCCTS Life Sciences Moment Fund Award
Role: Co-PI (Haran)
Start – End: 01/01/2017 – 12/31/2017
Abstract: We will conduct a longitudinal study of 100 patients ages 50 years and older hospitalized for treatment of a non-intestinal bacterial infection with 6 common antibiotic regimens and obtain serial stool samples on days 0 (the first day of antibiotics), 3 and 6 (during therapy) and 1 day after discharge from the hospital to determine C. difficile colonization status. From this cohort we anticipate that 25 patients will be colonized with C. difficile during the course of hospital treatment. We will test our previously-developed Clostridium difficile Risk Assessment Method for Prevention Study (CRAMPS) tool in predicting which patients in our cohort will become colonized with C. difficile. CRAMPS will be trained on 16S rRNA and qPCR data obtained from Illumina sequencing of serial fecal samples. C. difficile density will be quantified using standard rtPCR as previously described.17 We hypothesize that CRAMPS will accurately predict short-term changes in the microbiome in order to identify patients at risk for C. difficile colonization.
4. Leveraging evolution to develop therapeutics against enteropathogenic E. coli
Source: UMass Dartmouth MSFP
Start – End: 01/01/2016 – 06/30/2016
Abstract: Enteropathogenic Escherichia coli (EPEC) is a highly virulent bacterial pathogen, responsible for life-threatening complications upon ingestion of contaminated food or water. Critical to the success of this pathogen is the ability to colonize the host, which includes passing through a very acidic stomach, and being able to compete against the resident flora. Antibiotic treatment to clear EPEC is under debate as they increase the risk of clinical complications. Therefore, a primary objective is to develop therapeutics alternative to antibiotics targeting pathogen’s ability to colonize the host. Direct alteration of genes involved in colonization is a possible approach. However, as this involves a large reduction in pathogen’s fitness it also increases the likelihood of fixation of compensatory mutations. Building on a set of recently-collected preliminary data from experimental evolution assays, whole-genome sequencing, microbial competition experiments and protein modeling, we propose to solve this problem by leveraging those ecological and evolutionary traits that reduce this pathogen’s host colonization ability but that are simultaneously under selection in non-host secondary environments. Three major reasons make this project highly innovative. First, it will provide novel evolutionary-based mechanisms to be targeted in the development of therapeutics against EPEC. Second, this research will deliver a framework where genetic variations are mapped to defined phenotypes and can be compared to isolates after environmental outbreaks. Third, the methodology here applied to study EPEC can be generalized also to other pathogenic species and lead the way for the development of new evolutionary biology-based therapies.
5. Model reduction techniques for patient-specific data inference for gut microbiota dynamics
Source: UMass Dartmouth MSFP
Start – End: 01/01/2014 – 06/30/2014
Abstract: We propose to develop a computational approach based on the reduced basis framework to handle the computationally prohibitive task of simulating the gut microbiome and immune system dynamics. An efficient and accurate computational approach developed in this work will allow us to better understand the dynamics of the gut microbiome at the level of detail of high-throughput sequencing data. Our methods will build on our recent previous research on inference and prediction from high-throughput sequencing data and enable us to gain a better understanding of the dynamics of the commensal microbiota. Specifically, it will extend ridge-regression based approaches for the inference from time-series sequencing data to determine individual-species rather than ensemble network of microbial interactions and response to antibiotic effects.
6. Understanding microbial dynamics to improve biotechnological applications
Source: UMass Dartmouth MSFP
Understanding microbial dynamics to improve biotechnological applications
Role: Co-PI (Silby)
Start – End: 01/01/2014 – 06/30/2014
Abstract: The understanding of microbial interaction networks and developing the ability to predict the resulting dynamics will allow improved monitoring soil health as well as rectifying possible problems with microbial populations before they emerge. An extension of this concept is the rational design of communities of bacteria which make a desirable product such as a biofuel. To date, approaches to understanding microbial systems have been limited to descriptions of the components, but have not provided predictive power. Our new algorithm has the power to generate predictions on system responses based on quantitative data of populations with input regarding perturbation. The proposed research will develop an interacting network of bacteria which simulates a simple ecological system. We will modify a group of closely related Pseudomonas fluorescens strains to interact and use this group to test our model and improve it.
7. Mathematical modeling of the intestinal microbiome to determine anti-inflammatory microbial consortia
Source: Sponsored Research Agreement – Vedanta Biosciences
Start – End: 01/01/2014 – 06/30/2014
Description of Funded Research: The goal of the Principal Investigator (PI) contribution to the research is to perform computational and statistical data analysis of time-series of 17 regulatory T cell-inducing Clostridium strains previously identified by Vedanta (“VE202”), and variants of VE202. Vedanta and its collaborators in disjunction from the PI will have the data generated experimentally. The data will then be delivered to the Principal Investigator electronically in text file format. The Principal Investigator will analyze the data using the statistical approach that he recently published in his publication [….]. The focus of the analysis will be to use fast computational methods to individuate subsets of these 17 strains which are compositionally stable based on the temporal dynamics observed in the data and analyzed by the PI. The PI will produce a report where the results of the data analysis will be reported, which the company may use for follow-up applications