Research: Integrative Theories and Experiments in Living Systems

In the Ray lab we are working to discover new ways of understanding how molecular networks affect biological growth, fitness, and evolution. We use a theory-driven experimental approach, with a foundation in systems biology, chemical and statistical physics, and information theory. Experimental synthetic biology and single-cell measurements provide an empirical foundation for our theoretical and computational work.


One of the major goals of our lab is to provide a generalized theoretical framework for understanding how infectious diseases and cancer cell growth are guided by molecular mechanisms, how they evolve, and how their growth predisposes them to novel treatment strategies. To this end we are developing experimental and theoretical approaches to predict when cells enter antibiotic-tolerant states, characterizing the statistics of how such states are inherited, and determining the selective pressures on cells on the threshold of critical transitions. The figure below shows an example of a recent effort in the lab toward these goals.


<Metabolic Simulations & Experiments>

Simulations and experiments on a metabolic step affecting cellular growth rates in bacteria. a. Stochastic simulations of a metabolic step. Growth feedback from metabolite toxicity traps a subset of trajectories into irreversible metabolite buildup with consequent cellular growth arrest. The parameter ktA affects the rate at which precursor metabolite enters the simulated system. b. Microscopy of Escherichia coli cells (strain B REL606 lacI PlacO1GFP) grown in a microfluidic device with a constant flow of minimal medium containing lactose at the indicated concentration (in mg/ml). Red and yellow cells are stained with propidium iodide, which indicates cell death. When growth-mediated dilution of green fluorescent protein (GFP) is slowed or halted, the cells are brighter. Brighter green cells are slower growing. At high lactose concentrations, cellular growth rates are heterogeneous, consistent with the simulations predicting growth arrest in a subset of cells. c. Emergence of growth arrested cells in high lactose concentrations is correlated with enrichment of antibiotic persister cells in these conditions, as indicated by the higher fraction of long-lived cells surviving in ampicillin.


Research Profiles of Scientists in the Lab
Andrew Hecht: Simulation and Analysis of Metabolic Toxicity
<Andrew H.>

Recent results from the Ray lab have identified that lactose metabolic toxicity causes growth heterogeneity and persister cell formation in some strains of E. coli but not others. Bioinformatic comparison between sensitive and non-sensitive strains reveal a number of potential causes, including a key difference in the lac operon transacetylase gene, lacA. I am using computational modeling and experimental genetics to characterize the role of LacA and other possible mechanisms in sensitivity to lactose toxicity.

Nicole Lama: Chromosome Dynamics During Bacterial Growth Transitions

Technological advances in time-lapse imaging have made possible the visualization of single cells as they undergo phenotypic transitions, such as when cells shift from growing to non-growing. I am using fluorescence microscopy to measure the dynamics of chromosomes during phenotypic transitions.

GW McElfresh: Bacterial Population Heterogeneity
<GW M.>

My work involves using computational and experimental tools to study gene regulatory and cellular developmental programs involved in cellular heterogeneity and inter-generational information transfer. To study bacterial growth transitions, I am using tools from image analysis to more robustly observe Escherichia coli cells as they grow, diversify, and undergo phenotypic changes such as becoming a persister cell. I am also using RNA sequencing data to investigate global gene expression shifts and identifying root causes of phenotypic changes.

Alex Shearin: Kinetics of Protein-Protein Interactions

I am using a simulation approach to analyze how toxin-antitoxin interactions affect levels of fluctuation in free toxin.

Huijing Wang: Information Transfer and Prediction of Phenotypic Transitions
<Huijing W.>

Pathogenic bacteria can cause acute or chronic infections in humans. Acute infection is generally treatable with antibiotic therapy, but immune responses and antibiotics can push bacterial cells into a slow-growing state with increased antibiotic persistence. Many slow-growing infectious bacteria can form biofilms and cause chronic or stubborn diseases that are recalcitrant to treatment, such as Pseudomonas aeruginosa infection in cystic fibrosis patients and urinary tract infections by uropathogenic Escherichia coli.

<Metabolic Trajectory>

Simulated trajectory of a metabolite on the edge of a critical transition.


Can we predict when cells transition to growth arrest?

In ecological and economic studies, imminent transitions often display dynamical signatures that act as early warning signs. Based on this observation, we are testing the hypothesis that molecular networks in cells exhibit similar early warning signs.