Modeling signal transduction process in cell population application to transfer of drug resistance

Che-Chi Shu, Purdue University

Abstract

Enterococcus faecalis is a highly antibiotic-resistant nosocomial pathogen using horizontal gene transfer to spread drug resistance from donors, resistance-bearing cells, to recipients, resistance-deficient cells. One ways to reduce nosocomial infection of E. faecalis can be achieved by ceasing horizontal gene transfer. However, this is a complex process with more than one kind of signaling molecule which allows cells detecting not only the population of recipients but also donors. For such a complicated system, the aid of mathematical model to guide the experiment is undoubtedly more efficient. In this study, a deterministic cell model based on mechanism from literature has been established and shows qualitatively consistence results with experimental observations. With this model, we understand that gene regulation of the sense and anti-sense interaction is essential for a bistable switch behavior and that cells are capable of sensing the donor concentration to switch from induced state back to un-induced state in order to keep both donors and recipients in a population. From the views of evolution, it is a much better strategy for cells to allow different sub-populations with varying kinds of drug resistance than incurring a heavy metabolic burden of carrying all drug resistant genes in each cell. Next, we incorporate stochastically into the population balance model to further account for random behavior of individual cells. Remarkably, this newly developed model reveals that the interaction of cells through a shared environment and stochastic effects are both important. It is demonstrated that the neglect of cell interaction may lead to qualitative bias and the absence of stochasticity may lead to under-prediction of induction. One such outcome from neglecting cell interaction is that bistability leads to bimodal distributions of protein expression in the population. Although this kind of coexistence of bistability and bimodality can be broadly observed in nature, our more comprehensive theoretical approach shows that exceptions may be more general. Furthermore, we use the biofilm circumstance to demonstrate that lack of considering intracellular stochasticity may lead to quantitative difference in prediction. Finally, the model development in this dissertation goes farther towards analyzing the actual process of conjugation. Different kinds of population have been simultaneously tracked in the model. Overall, our study provides better understanding of the phenomena underlying the transfer of drug resistance among microbial species and serves as the foundation for the application of population balance to biological systems.

Degree

Ph.D.

Advisors

Ramkrishna, Purdue University.

Subject Area

Chemical engineering|Pharmacy sciences

COinS