Robust Network Topologies for Generating Switch-Like Cellular Responses 英文参考文献.docVIP

RobustNetworkTopologiesforGeneratingSwitch-Like CellularResponses NajafA.Shah1,CasimA.Sarkar1,2* 1Graduate Group in Genomics and Computational Biology, School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, United States of America, 2Departments of Bioengineering and Chemical Biomolecular Engineering, School of Engineering and Applied Sciences, University of Pennsylvania, Philadelphia, Pennsylvania,UnitedStatesofAmerica Abstract Signalingnetworksthatconvertgradedstimuliintobinary,all-or-nonecellularresponsesarecriticalinprocessesranging fromcell-cyclecontroltolineagecommitment.Toexhaustivelyenumeratetopologiesthatexhibitthisswitch-likebehavior, we simulated all possible two- and three-component networks on random parameter sets, and assessed the resulting responseprofiles forboth steepness(ultrasensitivity) andextentofmemory(bistability).Simulations wereused tostudy purely enzymatic networks, purely transcriptional networks, and hybrid enzymatic/transcriptional networks, and the topologies in each class were rank ordered by parametric robustness (i.e., the percentage of applied parameter sets exhibitingultrasensitivityorbistability).Resultsrevealthatthedistributionofnetworkrobustnessishighlyskewed,withthe most robust topologies clustering into a small number of motifs. Hybrid networks are the most robust in generating ultrasensitivity(upto28%)andbistability(upto18%);strikingly,apurelytranscriptionalframeworkisthemostfragilein generatingeitherultrasensitive(upto3%)orbistable(upto1%)responses.Thedisparityinrobustnessamongthenetwork classes is due in part to zero-order ultrasensitivity, an enzyme-specific phenomenon, which repeatedly emerges as a particularlyrobustmechanismforgeneratingnonlinearityandcanactasabuildingblockforswitch-likeresponses.Wealso highlightexperimentallystudiedexamplesoftopologiesenablingswitchingbehavior,inbothnativeandsyntheticsystems, thatrankhighlyinoursimulations.Thisunbiasedapproachforidentifyingtopologiescapableof