Molecular basis for Ral isoform diversity

The sequences of RalA and RalB are 82% identical, with the majority of the differences in their C-terminal ~25 residues. The extreme C-terminus of both Ral proteins is lipid modified and anchored into the bilayer. This project builds on a considerable body of work from the HRM lab. We have solved the structures of C-terminally truncated forms of free RalA and RalB, and of RalB in complex with its effector RLIP76. We have probed the energetic landscape of the interaction surfaces between RLIP76 and RalA/RalB and found some residues in RLIP76 whose mutation severely reduces binding to only one of the Ral proteins.

Taken together, our data show that subtle differences exist between RalA and RalB that underpin their different functions. We believe that interactions between the C-termini of the Ral protein and its canonical Ras-like G domain are a key feature of these differences. Although the G domain is well understood, much less is known about the C-termini of Ras-like proteins because they are usually removed for structural analysis. Our preliminary investigations of the C-terminal, variable region of RalA showed that it interacts with the main body of the G protein domain differently in the GDP- and GTP-bound forms of the RalA protein. In this project we will develop this work and investigate lipid-modified, full-length Ral proteins in the presence of membrane mimics. This work builds on the considerable expertise of the academic lab in studying proteins and their interactions using NMR. We will utilize coexpression with farnesyltransferases, spin-labelling technology and membrane mimetics, in combination with NMR techniques, to study the interactions between Ral and the membrane and between the G domain and the C-terminus. Such methodology is just starting to be applied to peripheral membrane proteins such as RalA. Once we have identified residues involved in the interaction between the C-terminus and the G domain, we will design panels of mutants to stabilize or disrupt these interactions. The C-terminus appears to interact with one face of the G domain in the GDP-bound form of RalA and we hypothesize that mutants that stabilize these interactions will stabilize the GDP-bound form, while mutants that disrupt them will have the opposite effect. We will test the mutants, using NMR to probe the interactions, fluorescent nucleotide analogues to assess their relative nucleotide affinity and real-time phosphate release assays to test the intrinsic GTPase activity.

Mutants that have an effect on the in vitro properties of RalA will be taken forward into cell-based assays at AZ. This work will be essential in building our understanding of the role of RalA and will represent a first step towards understanding the differences between the Ral isoforms.