Saposins are lipid transfer proteins required for the degradation of sphingolipids in the lysosome. These small proteins bind lipids by transitioning from a closed, monomeric state to an open conformation exposing a hydrophobic surface that binds and shields hydrophobic lipid tails from the aqueous environment. Saposins form a range of multimeric assemblies to encompass these bound lipids and present them to hydrolases in the lysosome. This lipid-binding property of human saposin A has been exploited to form lipoprotein nanodiscs suitable for structural studies of membrane proteins. Here we present the crystal structure of a unique tetrameric assembly of murine saposin A produced serendipitously, following modifications of published protocols for making lipoprotein nanodiscs. The structure of this new saposin oligomer highlights the diversity of tertiary arrangement that can be adopted by these important lipid transfer proteins.

Background: Lipid antigens are presented on the surface of cells by the CD1 family of glycoproteins, which have structural and functional similarity to MHC class I molecules. The hydrophobic lipid antigens are embedded in membranes and inaccessible to the lumenal lipid-binding domain of CD1 molecules. Therefore, CD1 molecules require lipid transfer proteins for lipid loading and editing. CD1d is loaded with lipids in late endocytic compartments, and lipid transfer proteins of the saposin family have been shown to play a crucial role in this process. However, the mechanism by which saposins facilitate lipid binding to CD1 molecules is not known and is thought to involve transient interactions between protein components to ensure CD1-lipid complexes can be efficiently trafficked to the plasma membrane for antigen presentation. Of the four saposin proteins, the importance of Saposin B (SapB) for loading of CD1d is the most well-characterised. However, a direct interaction between CD1d and SapB has yet to be described. Methods: In order to determine how SapB might load lipids onto CD1d, we used purified, recombinant CD1d and SapB and carried out a series of highly sensitive binding assays to monitor direct interactions. We performed equilibrium binding analysis, chemical cross-linking and co-crystallisation experiments, under a range of different conditions. Results: We could not demonstrate a direct interaction between SapB and CD1d using any of these binding assays. Conclusions: This work strongly indicates that the role of SapB in lipid loading does not involve direct binding to CD1d. We discuss the implication of this for our understanding of lipid loading of CD1d and propose several factors that may influence this process.