Ras G-proteins are key players in tumorigenesis. Human cells contain four Ras proteins, each of which is activated by many guanine nucleotide exchange factors (GEFs), and each of which in turn activates a growing list of effectors. It is not known how a given Ras protein is networked with a given GEF and a proper effector. The overall objective of this proposal is thus to study how Ras signaling selectivity is maintained in the cell. Our studies with mammalian cells have shown that Ras proteins localize to both the plasma membrane and the endomembrane, so that Ras proteins may control different effectors in different compartments. However, it is not known which effectors can be activated by Ras proteins at a given compartment, and whether Ras signaling selectivity can be regulated by other means. We turned to a much simpler system to address these issues. The fission yeast S. pombe contains just a single Ras protein, Ras1, which is clearly compartmentalized to control two distinct functions-in the plasma membrane, Ras1 activates a MAP kinase pathway to mediate mating pheromone signaling, while in the endomembrane Ras1 activates Cdc42 to control cell polarity (shape) and mitosis. We have also discovered that each of the two Ras1 pathways is controlled by its own GEF, and these two GEFs are not functionally interchangeable even if they are overexpressed. These results support the hypothesis that Ras signaling selectivity can be regulated by (1) binding to a particular GEF and (2) localizing to a specific cell compartment. Furthermore, we hypothesize that endomembrane-restricted Ras can activate Cdc42, which in mammalian cells has already been shown to act downstream of Ras during cell transformation. The first goal of this project is to elucidate how GEFs selectively regulate Ras pathways using S. pombe, with its relatively simple and well characterized Ras system and its ease of genetic manipulation. We will investigate: 1A) whether Ras GEFs, like Ras itself, must spatially segregate; 1B) whether specific GEFs select specific Ras lipidation (and hence localization) states, and mediate specific Ras/effector binding; 1C) which conserved GEF residues mediate signaling selectivity; and 1D) where Ras1 interacts with Cdc42 and how Cdc42 localization is controlled. Our second goal is to apply these findings to the mammalian Ras system, focusing in particular on Ras-induced transformation. We will determine: 2A) whether cell transformation by Ras proteins requires an interaction with Cdc42 at the endomembrane; 2B) whether endomembrane-restricted Ras and/or Cdc42 can transform human non-malignant epithelial cells; and 2C) whether the endomembrane-localized GEF RasGrp1 controls this Ras/Cdc42 interaction via the mechanisms revealed in Aim 1.
Showing the most recent 10 out of 15 publications
