Resumen de Signal generation and detection in a cellular context
This thesis takes a Systems Biology approach to propose new methods for the study of the molecular interactions underlying fundamental cellular processes. From signal transduction to gene expression, the strongly nonlinear nature of most of these systems make them perfect candidates to be subjected to detailed mathematical modeling, in an effort to find the elements (species, interactions or degradations) in those networks which allow them to exhibit the complex responses which help keeping the whole organism alive. Thus, our analysis is based on that approach, and is able to reveal the interactions present in a specific signal transduction pathway as well as to characterize the effect that the structure of simple modules of biochemical interactions has on their ability to process and propagate signals.
After a brief introduction to the cellular context where all these processes occur, we start describing the results obtained in the study of the IP3-mediated calcium signaling pathway. This pathway presents calcium oscillations upon stimulation of the cell with constant concentration of diverse ligands. Moreover, it seems to somehow multiplex the information encoded in the ligand type and concentration in the temporal pattern exhibited by those oscillations. By means of experiments performed in Hela cells using microfluidic devices, and computer analysis of previously published models describing this system, we are able to unveil the internal structure of the network that gives rise to such a complex behavior. All this analysis is performed without explicitly fitting the models, taking into account cell-to-cell variation, and without direct intervention in the molecular machinery of the network, thus avoiding the uncertainty caused by those techniques. In addition, the obtained description of the molecular machinery allows us to make predictions for the expected cellular responses under specific conditions, what leads us to propose and perform some experiments which further corroborate our model.
In the last chapter of the thesis, we describe the tools developed to study the effect that different interactions have on the dynamical and steady state response of simple gene networks. First, by taking advantage of the modular description of gene networks, we propose a simple three component module as a general platform to test the influence that specific network interactions (feedbacks, feedforwards or autoregulations) have on the response of the module to amplitude and frequency modulated signals, as well as on its ability to deal with the random fluctuations inherent to gene expression. Using this simple model, we find that some network structures (feedbacks and autoregulations) exhibit trade-offs in the detection of both classes of signals, while other are able to overcome that constraint, being capable of improving the propagation of both types of stimuli. At the same time, we observe that different types of circuits deal with noise differently: while some structures have to increase their signal-to-noise ratio to perform a feasible transmission of noisy signals, others are able to filter noise in the frequency domain by separating the range of frequencies where oscillations are best propagated from that of the fluctuations. Finally, using the simple three component network as a model, we explore the connections between the structure of the networks, their signal propagation abilities, and their response to sudden changes in the input concentration. A statistical analysis helps us to obtain answers which do not depend on the specific regime the circuits are operating in, giving us useful information about the effect that different interactions have on the steady state and dynamic response of the circuits, as well as on the connection between this dynamical behavior and the steady state response.
240699, 20703, 220510, 221026, 221029, 22069, 221128
