Case Study: Lippincott-Schwartz Lab

The Lippincott-Schwartz lab at the NIH is widely recognized for innovation in fluorescence microscopy and for significant contributions to cell biology. We have been close collaborators since 1997 bringing modern kinetic analysis to projects involving trafficking of GFP-tagged proteins in living cells. The project illustrated here provided strong support for a major paradigm shift in Golgi membrane trafficking. 


Photobleaching and pharmacological perturbations designed to probe the mechanism of Arf1 mediated cargo sorting and transport. 


FRAP of Golgi COP-I-YFP, FRAP of Golgi Arf1-GFP, COP-I-YFP and Arf1-GFP in response to Brefeldin A (BFA).

  • Competing hypotheses, drawn as familiar symbol-and-arrow diagrams, were entered into ProcessDB, which then automatically translated them into kinetic (mathematical) models
  • Investigators’ knowledge of molecular abundances and relative speeds of the processes represented in the diagrams was captured in database tables. For unknowns, ProcessDB used biologically reasonable numbers whose values were tested against the experimental data.
  • Investigators added biological constraints to the models based on their working hypotheses.
  • FRAP and BFA experimental protocols were entered into ProcessDB Forms.
  • ProcessDB merged all 4 experimental protocols with each kinetic model into a combined “Model of Experiments”, which was then simulated and optimized.
  • Initial model based on established theory failed to account for the experimental data. No parameter values could be found that would simultaneously account for all four data sets.
  • The investigators formulated an alternative, even radical, hypothesis (see diagram) in which Arf1 and COP-I follow separate pathways after Arf1 recruits COP-I to Golgi membranes.
  • This new model was quickly assembled in ProcessDB re-using model components assembled from previously built models.
  • ProcessDB maintained a searchable database of all models that accelerated comparison and testing of alternative hypotheses.
  • ProcessDB helped identify Golgi pools of both Arf1 and COP-I that turn over slowly and may correspond to specialized domains involved in protein trafficking.
  • ProcessDB allowed investigators to demonstrate that a controversial hypothesis was quantitatively consistent with the collected experimental data (see graphs), significantly strengthening the case for a paradigm shift.