Résumé : Plasma membrane transport proteins play a crucial role in all cells by conferring to the cell surface a selective permeability to a wide range of ions and small molecules. The activity of these transporters is often regulated by controlling their amount at the plasma membrane, via intracellular trafficking. The recent boom in the numbers of crystallized transporters shows that many of them that belong to different functional families with little sequence similarity adopt the same structural fold implying a conserved transport mechanism. These proteins belong to the APC (Amino acid-Polyamine-organoCation) superfamily and their fold is typified by the bacterial leucine transporter LeuT. This LeuT fold is characterized by inverted structural repeats of 5 transmembrane domains that harbor the central substrate-binding site and a pseudo-symmetry axis parallel to the membrane. The yeast Saccharomyces cerevisiae possesses about 16 amino acid permeases (yAAPs) that belong to the APC superfamily and that display various substrate specificity ranges and affinities. Topological, mutational analysis and in silico data indicate that yAAPS adopt the LeuT fold.

In this work we combined computational modeling and yeast genetics to study substrate binding by yAAPs and the endocytosis of these transporters in response to substrate transport. In the first part of this work, we analyzed the selective recognition of arginine by the yeast specific arginine permease, Can1. We constructed three-dimensional models of Can1 using as a template the recently resolved structure of AdiC, the bacterial arginine:agmatine antiporter, which is also a member of the APC superfamily. By comparison of the binding pockets of Can1 and Lyp1, the yeast specific lysine permease, we identified key residues that are involved in the recognition of the main and side chains of arginine. We first showed that the network of interactions of arginine in Can1 is similar to that of AdiC, and that the selective recognition of arginine is mediated by two residues: Asn 176 and Thr 456. Substituting these residues by their corresponding residues in Lyp1 converted Can1 into a specific lysine permease. In the second part of this work, we studied the regulation of two permeases, Can1 and the yeast general amino acid permease, Gap1. In the presence of their substrates, Gap1 and Can1 undergo ubiquitin-dependent endocytosis and targeting to the vacuolar lumen for degradation. We showed that this downregulation is not due to intracellular accumulation of the transported amino acids but to transport catalysis itself. By permease structural modeling, mutagenesis, and kinetic parameter analysis, we showed that Gap1 and Can1 need to switch to an intermediary conformational state and persist a minimal time in this state after binding the substrate to trigger their endocytosis. This down-regulation depends on the Rsp5 ubiquitin ligase and involves the recruitment of arrestin-like adaptors, resulting in the ubiquitylation and endocytosis of the permease.

Our work shows the importance of the structural analysis of yAAPs to get further insight into the different aspects of their function and regulation. We validate the use of a bacterial APC transporter, AdiC, to construct three-dimensional models of yAAPs that can be used to guide experimental analyses and to provide a molecular framework for data interpretation. Our results contribute to a better understating of the recognition mode of amino acids by their permeases, and the regulation of this transport in response to substrate binding.