Annual Report of Activities CNC 2008 - Center for Neuroscience and ...

Molecular Systems Biology | Head: Armindo SalvadorObjectivesMain objectives of our group are:1. Discovering and understanding the naturallyevolved principles of quantitative design of themost prevalent elementary circuits in metabolicnetworks. These design principles are rulesassociating quantitative design (e. g., relationshipsamong enzyme kinetic parameters or among theseand effector concentrations) to function.2. Kinetic modeling and systems analysis of thebiogenesis, fate and regulation of reactivemetabolic intermediates. Reactive intermediates areinvolved in many pathologies and, much for thesame reasons that make them toxic, also participatein important regulatory mechanisms. We seek abetter grasp of the trade‐offs involved in themetabolic processes that generate these species andto understand how these trade‐offs inpend on theregulatory design of these processes.3. Modeling the permeation of the blood‐brainbarrier by amphiphiles. The aim is to understandhow the interactions of amphiphilic drugs with thevarious relevant aqueous, lipid, and glycocalyxcompartments to which they partition in the bloodstream and while crossing the blood‐brain barrieraffect the efficacy of their delivery to brain tissues.4. Developing a method for profiling mitotic‐cycledependentmetabolism without having tosynchronize cells.5. Developing an Internet‐based platform fordistributed colaboration in kinetic modeling ofbiochemical processes. Current solutions for archivaland communication of kinetic models just store“frozen” versions of the models and do not promotediscussion and further development. This is a majorlimitation beacuse model development should beviewed as a dynamic process reflecting the evolvingknowledge about biochemical processes. We seek todevelop a platform — WikiModels — fordeveloping models as a comunity activity throughconstant open peer‐review of modeling decisions,recording successive states of a model and trackingcredit for contributions.Main AchievementsWe tested if the design principles we derived formoiety‐transfer cycles [Coelho PMBM, Salvador A,Savageau MA (2009). PLoS Comput. Biol.5:e1000319, and paper in preparation] applyextensively. Tests so far focused mainly redoxcycles and were based on both detailed analyses ofwell‐characterized instances and broad surveys ofmetabolite concentrations and enzyme‐kineticparameters. We examined in detail glutathione,NADPH, cyt b5 and FAD cycles in humanerythrocytes and found them to adhere to thepredicted design principles. A set of KMs for >50enzymes involved in NADPH redox cycles in E.coli and S. cerevisiae indicates that the predicteddesign principles hold extensively.We developped a mathematical approach forsystematically constructing a “dimensionallycompressed”design space and partitioning this intodiscrete regions of qualitatively different performance[Savageau MA, Coelho PMBM, Fasani RA, Tolla DA,Salvador A (2009). Proc. Natl. Acad. Sci. USA 106:6435].Activation of uncoupling protein 2 (UCP2) by 4‐hydroxynonenal (HNE) decreases the protonmotivepotential across the mitochondrial inner membrane,with ensuing decrease in superoxide (SO) formationby Complex I. Because SO is one of the maininitiators of lipid peroxidation, of which HNE is anend‐product, the overall process is a negativefeedback loop on SO production. We have investigatedpossible functional advantages of this indirectfeedback loop versus direct UCP2 activation by SO,which would in principle permit faster regulation.The analysis indicates that the HNE‐mediateddesign is better at managing the trade‐off betweenefficiency of energy production and minimizationof the generation of reactive intermediates.We developped a mathematical model for theredistribution of amphiphiles among lipid compartmentsin human blood and crossing of the blood‐brainbarrier. The model permitted identifying limitingsteps in the transport of amphiphiles frombloodstream to the brain tissue.36Fig.1. Design of the coupled NADPH and GSH redox cycles (a) in humanerythrocytes. (b) Good performance requires orchestration amongparameters. Each parameter is represented in a radial axis extending from 0(inner tip) to 1 (outer tip). A polygon joining values in each axis indi‐catesthe parameter set for a given realization of the circuit. The red polygonsindicate the realizations that warrant a level of performance matching orexceeding that in vivo. The dashed blue polygon indicates the parameter setthat holds in normal human erythrocytes.