The GTPase of immunity-associated proteins (GIMAP) is expressed in eukaryotic phyla including a subset of molluscs, vertebrates, and some protists. It is predominantly ex- pressed in the lymphoid organs of mammals and other vertebrates where it plays roles in the homeostasis of the immune system. My study focuses on human (h) GIMAP6, a cytosolic member of the GIMAP family that is widely expressed across the lymphoid lineages. Studies within our group have uncovered a highly specific interaction between GIMAP6 and hGABARAPL2 (gamma- aminobutyric acid receptor-associated protein-like 2), a mammalian homologue of the autophagy-related protein 8 (Atg8). Using bacterially expressed GABARAPL2 and GIMA- P6 I have shown that the interaction between the two proteins is direct. My studies have attempted to gain an understanding of the molecular requirements for this interaction using site-directed mutagenesis and pull-down assays. Mutational analyses, including point mutations within, and truncations of, GIMAP6 and GABARAPL2 have revealed a number of things. Close to its N-terminus GIMAP6 carries a sequence correspond- ing to a canonical Atg8 interacting motif (AIM), a motif frequently found in proteins that interact with the Atg8 family. My studies indicate, however, that these residues do not play a role in the interaction. Using GTP-agarose, I was able to demonstrate that GIMAP6 could bind GDP and GTP and by mutating key residues within its GTP bind- ing domain, I could disrupt the interaction of GIMAP6 with GABARAPL2. I have also shown that the C-terminal 10 amino acids of GIMAP6 are necessary for the interaction. Interestingly, variants of GIMAP6 that were unable to bind GTP-agarose were also un- able to interact with GABARAPL2, hinting at a crucial role for nucleotide binding in the GIMAP6-GABARAPL2 interaction. Within GABARAPL2, deletion of the N-terminal α-helix resulted in loss of the interaction. A chimeric protein in which the correspond- ing region in MAP1LC3B, a protein unable to interact with GIMAP6, was replaced by GABARAPL2’s N-terminal α-helix reproduced the interaction suggesting that this region

is critical for the interaction. Studies in our group have shown that GIMAP6 relocalises to autophagosomes on induction of autophagy. I have shown that variants of GIMAP6 unable to interact with GABARAPL2 fail to display a similar relocalisation. Finally, recent research has demonstrated that members of the GIMAP family can homo- and hetero-dimerise. I have shown that GIMAP6 can interact with itself and intriguingly, also shows a specific interaction with GIMAP7. Contrary to what was observed for the GIMAP6-GABARAPL2 interaction, truncating the N-terminus of GIMAP6 abrogated the interaction with GIMAP7. These findings evoke the possibility that the GIMAP GTPases function together in an interacting network.