Low pH-induced conformational change and dimerization of sortilin triggers endocytosed ligand release

Abstract

Low pH-induced ligand release and receptor recycling are important steps for endocytosis. The transmembrane protein sortilin, a β-propeller containing endocytosis receptor, internalizes a diverse set of ligands with roles in cell differentiation and homeostasis. The molecular mechanisms of pH-mediated ligand release and sortilin recycling are unresolved. Here we present crystal structures that show the sortilin luminal segment (s-sortilin) undergoes a conformational change and dimerizes at low pH. The conformational change, within all three sortilin luminal domains, provides an altered surface and the dimers sterically shield a large interface while bringing the two s-sortilin C-termini into close proximity. Biophysical and cell-based assays show that members of two different ligand families, (pro)neurotrophins and neurotensin, preferentially bind the sortilin monomer. This indicates that sortilin dimerization and conformational change discharges ligands and triggers recycling. More generally, this work may reveal a double mechanism for low pH-induced ligand release by endocytosis receptors.

Crystal structure of the ligand‐free form of the Vps10 ectodomain of dimerized Sortilin at acidic pH

“…Therefore, the pH at the cellular compartments containing Sortilin dramatically changes during Sortilin trafficking in cells. Consistent with the findings of the most recent studies , we also demonstrated that sSortilin dimerizes under a low pH buffer condition. In addition, SEC analysis revealed that the A495E mutant in mouse Sortilin, which is located in the vicinity of F128 at the hydrophobic loop, abolishes the potency of pH‐dependent dimerization .…”

“…Consequently, our structure could clearly represent the contribution of E455 to the hydrophobic interaction with F137 in Sortilin dimerization at acidic pH. The conformational changes that were observed upon dimerization, as revealed through structural comparison with human sSortilin, were also confirmed independently in these recent studies , indicating the importance of conformational changes for dimerization of sSortilin at acidic pH.…”

“…Of note, although the resolution of the structure determined at acidic pH by Leloup et al . was higher than that of our determined structure, the residues (448‐ATA‐450) in the vicinity of the dimer interface were disordered in this structure, resulting in an unclear electron density map in the vicinity of the loop at residues 448–455 (Fig. S1C).…”

“…During preparation of this paper, the crystal structures of the human and mouse sSortilin dimer were reported by two independent groups ; however, the crystallization conditions employed in these studies were quite different from those in the present study. Consequently, the cell dimensions of their crystals also differed from those of our crystals.…”

“…Consequently, the cell dimensions of their crystals also differed from those of our crystals. Structural alignment revealed that our determined sSortilin dimer well resembled both mouse sSortilin dimer (PDB ID: http://www.rcsb.org/pdb/search/structidSearch.do?structureId=5NMT) (the Cα‐Cα RSMD is 0.93 Å) and human sSortilin dimer (PDB ID: http://www.rcsb.org/pdb/search/structidSearch.do?structureId=6EHO) (the Cα‐Cα RSMD is 0.91 Å) (Fig. S1A, B).…”