The process of Raf-1 activation is complex. Ras recruits Raf to the plasma membrane, where a series of activation steps that are yet to be fully understood are required for full Raf-1 activation. To investigate Raf-1 activation, an assay was developed to measure Raf-1 specific activity. Using this assay, it was found that the Raf- 1 zinc-finger in the Raf-1 CRD was not necessary for membrane recruitment of Raf-1 by Ras, but was essential for full activation of Raf-1 at the plasma membrane. Membrane targeting did not compensate for the absence of the zinc finger. In addition, membrane targeting of constitutively activated Raf-1 incremented activity, but full activation required co-expression with Ras G12V. This sensitivity to regulation by Ras at the plasma membrane was abrogated by mutations in the Raf zinc finger but was unaffected by mutation of the minimal Ras binding domain. Thus, Ras has a role in the activation of Raf-1 via a mechanism that requires the Raf zinc finger, as well as a role in Raf-1 membrane recruitment through an interaction with the minimal Ras binding domain.
14-3-3 proteins complex with many signalling molecules, including the Raf-1 kinase. The role and nature of this complex was investigated. It was found that 14-3-3 bound to Raf-1 in the cytosol, but was totally displaced when Raf-1 was recruited to the membrane. When serum-starved cells were stimulated with EGF, some recruitment of 14- 3-3 to the membrane was evident. Interestingly, this 14-3-3 recruitment correlated with Raf-1 dissociation and inactivation, not with Raf-1 recruitment. In vivo, overexpression of 14-3-3 potentiated the specific activity of membrane-recruited Raf-1, but caused a reduction in the amount of Raf-1 bound to the membrane. In vitro experiments also demonstrated that formation of a 14-3-3/Raf-l complex was critical for Raf-1 recruitment and activation by Ras, since the removal of 14-3-3 from Raf-1 through peptide competition resulted in a Raf-1 protein that could not be activated. These results demonstrate that the interaction of 14-3-3 with Raf-1 is permissive for recruitment and activation by Ras, that 14-3-3 is displaced upon membrane recruitment, and that 14-3-3 may recycle Raf-1 from the membrane to the cytosol.
The critical, initial step in Raf-1 activation is recruitment to the plasma membrane by Ras. Ras proteins are highly conserved, but differ significantly in their postranslational processing and mechanism of membrane anchorage. The second part of my thesis addressed whether biological differences between Ras isoforms could be due to differences in their membrane microlocalisation. Caveolae are a plasma membrane microdomain that have been implicated both in cholesterol homeostasis and in signal transduction. Two dominant negative amino-terminal truncation mutants of caveolin, the major structural protein of caveolae were assayed for effects on Ras function. It was found that one of these mutants, CavDGV, completely blocked H-Ras-mediated Raf-1 activation, but had no effect on K-Ras-mediated Raf-1 activation. Strikingly, the inhibitory effect of CavDGV on H-Ras signalling was completely reversed by replenishing cell membranes with cholesterol and was mimicked by cyclodextrin treatment, which depletes membrane cholesterol. These results provide a crucial link between the cholesterol-trafficking role of caveolin and its postulated role in signal transduction. They also provide direct evidence that H-Ras and K-Ras, which are targeted to the plasma membrane by different carboxy-terminal anchors, operate in functionally distinct microdomains within the plasma membrane.
Finally, the effect of endocytosis on Ras signalling was investigated. Dynamin and Rab5 are involved in different aspects of endocytosis. Co-expression of wild type Rab5 potentiated H-Ras mediated Raf-1 activation, while mutant Rab5Q79L or wild type dynamin co-expression abrogated H-Ras mediated Raf-1 activation. In contrast, K-Ras mediated Raf-1 activation was not affected by either wild type or mutant dynamin co-expression, but was potentiated by both wild type and mutant Rab5Q79L co-expression. Further examination of the cellular localisation of the activated form of H-Ras revealed that it was significantly mislocalised by wild type or K44A mutant dynamin, by Rab5Q79L mutant co-expression, and to a lesser extent, by wild type Rab5 coexpression. K-Ras localisation was not affected by wild type or mutant Rab5 or dynamin. Taken together, these results suggest that H-Ras signalling is sensitive to perturbations in the endocytic pathway, whereas K-Ras signalling is not.