Modeling membrane localization: case study of a Ras signaling model.

Modeling a biological system requires the careful integration of experimental data. It is unclear how best to incorporate rate constants measured in three-dimensional solution for reactions that physiologically occur between reactants confined to the two-dimensional cell membrane. One method adjusts second order rate constants by a factor that is the ratio of the cytoplasmic volume to the volume of a shell which membrane bound proteins can access. The value for this factor has been estimated to be 250. We have previously used this method in our model of the Ras signaling network that made several experimentally confirmed predictions. Here, we investigate if the value of this parameter affects model based predictions. We find that many of our results are robust to the value used. Two predictions appear to be sensitive to the value of the parameter: predicted levels of WT RasGTP after transfection with WT Ras and the experimentally observed increased levels of WT RasGTP when a GTPase Accelerating Protein (GAP) insensitive Ras mutant is present. For these predictions that are sensitive to the value of the membrane localization parameter, we find that the theoretically derived value of 250 results in model predictions that most closely match experimental observations.