Physics (Internal)

Modeling Collective Behavior in Dictyostelium Discoideum

This event is part of the Preliminary Oral Exam.

Examining Committee: Pankaj Mehta, Daniel Segre, Rama Bansil and Martin Schmaltz Abstract: Dictyostelium discoideum are a soil-dwelling amoebae, that transition from unicellular to multicellular behavior in response to starvation. Once starved, cells within a population secrete pulses of small signaling molecule cyclic AMP (cAMP) into the extracellular environment. The simultaneous secretion and detection of cAMP by individual cells gives rise to collective synchronous oscillations of cytosolic cAMP across the population. For large populations, these oscillations take the form of spatially extended spiral waves. These waves guide the movement of cells toward a common location, usually the spiral center, and allow cells to aggregate into multicellular slugs. Such complex behavior coupled with the ease of genetic manipulation in  D. discoideum makes it an ideal organism for the study of collective behavior in eukaryotic systems. Furthermore, recent experimental advances in microfluidic devices and fluorescent tagging have provided us with an opportunity to study these cells in a quantitative fashion. The intricate behavior of this organism is the result of a complex biological network that has only been partially elucidated. The goal of our work is to model the networks that give rise to these collective behaviors. Conventionally, this is done by modeling biological structures using bottom-up approaches where the proteins and biomolecules involved in the network are explicitly incorporated in a large model. However these models often have several deficiencies due to the incomplete knowledge of the full biological network as well as the large number of unknown parameters. This problem is especially acute in eukaryotes, such as D. Discoideum, where bottom-up approaches are hard to implement due to complexity of underlying circuit.   In contrast to this bottom-up approach, here we present a top-down approach based on the idea of universality from statistical physics and dynamical systems. Contrary to usual bottom-up methods, this approach allows us to build a model which produces desired qualitative behavior in a wide range of parameter ranges. Since our model is designed to capture qualitative behavior of the system, there is no need for parameter tuning as is the case in most model-building procedures. To test the predictions of our model, we rely on the experimental data of single cells provided by our collaborators in the lab of Thomas Gregor.   By analyzing the experimental data we propose a simple two-variable model for the dynamics of the Dictyostelium communication network. The model is similar to the Fitzhugh-Nagumo model, commonly used for modeling excitable media. This model consists of a strong positive feedback loop that activates the production of the signaling molecule cAMP and a slower negative feedback that turns off the production. This model makes several predictions that have been verified by our collaborators. Furthermore we have extended the model to the the level of cellular populations using a mean-field approach. The model explains several already existing experimental results including the transition to collective oscillations as a function of population density and signal degradation. As a next step we have incorporated space into our model by assuming diffusion of cAMP into the extracellular environment. The resulting model is capable of producing spiral waves similar to what is observed experimentally.   As the next step, using our simulations, we will study different features of spiral patterns such as the dynamics of spiral center density, overall chirality, front width and speed and will compare them with experimental results for wild-type and mutant cells. This will allow us to determine the kinetic parameter for mutant strains and help elucidate the role of different proteins in the signaling pathway. This work will also allows us to design future experiments that relate our simple model to the the genes and proteins in the Dictyostelium signaling pathway.