In situ architecture of the lipid transport protein VPS13C at ER-lysosomes membrane contacts

Abstract

VPS13 is a eukaryotic lipid transport protein localized at membrane contact sites. Previous crystallographic and cryo-EM studies of VPS13 fragments suggested that it may transfer lipids between adjacent bilayers by a bridge-like mechanism. Direct evidence for this hypothesis from a full-length structure and from in situ studies, however, is still missing. Here we have capitalized on AlphaFold predictions to complement the structural information already acquired and to generate a model of full-length human VPS13C, the Parkinson's disease-linked VPS13 paralog localized at contacts between the ER and endo/lysosomes. Such model predicts a ~30-nm rod-shaped structure with a hydrophobic groove that extends throughout its length. We further investigated whether such a structure can be observed in situ at ER-endo/lysosome contacts. To this aim, we combined genetic approaches with cryo-focused-ion-beam (cryo-FIB) milling and cryo-electron tomography (cryo-ET) in HeLa cells overexpressing this protein (either full length or with an internal truncation) along with VAP, its anchoring binding partner at the ER. Using these methods we identified rod-like densities that span the space separating the two adjacent membranes and that fit the predicted sizes of full lengths VPS13C or of its shorter truncated mutant. Only a subset of them appeared to directly contact the ER bilayer in spite of being anchored to the ER membrane by VAP. Collectively these findings provide strong evidence for a bridge-model of lipid transport, and also suggest that direct contact of VPS13C with the ER bilayer to mediate lipid transport may be regulated.

A partnership of the lipid scramblase XK and of the lipid transfer protein VPS13A at the plasma membrane

“…1A-C and Movie S1). The rod harbors a groove that runs along its entire length and has a hydrophobic floor, supporting the hypothesis that VPS13A (like other VPS13 proteins), bridges two lipid bilayers to allow bulk flow of lipids between them(3, 4, 23, 24). VPS13A is anchored to the ER via a VAP binding FFAT motif(14, 16, 25) found at about ∼6nm from the N-terminal end of the rod, which comprises the so-called chorein motif.…”

“…5E). A similar segregation of a VPS13 family member (VPS13C) and another membrane contact site protein (PDZD8) was observed at ER-endolysosome contacts, possibly reflecting, at least in part, size exclusion due to the length of VPS13 proteins(23). VPS13A is ∼22 nm long, whereas contacts formed by junctophilins in muscle cells has been reported to be around 12nm(47).…”

“…Future studies should address a potential role of VPS13A and XK in PtdSer exposure in human medium spiny neurons, where the expression of the two proteins is enriched and highly correlated. In the context of neurodegeneration, it is of interest that a WWE domain is present within the structure of VPS13A, as well as of VPS13C (23), whose mutations also results in neurodegeneration (Parkinson's disease) (57). WWE domains were shown to bind Poly-ADP-Ribose (PAR) (28), a molecule implicated in glutamate excitotoxicity-mediated neuronal death (also termed PARthanathos, due to its dependence on PAR) (58).…”

“…As in other VPS13 isoforms, the modules that surround the C-terminal region of VPS13A are the arc-like VPS13 Adaptor Binding (VAB), a bundle of --helices (called ATG2-C domain) that have similarity to a similar bundle found in ATG2-C (26), and a PH domain (27). An additional previously unnoticed small folded module is the WWE domain, which is also present in VPS13C (23), the closest human paralogue of VPS13A and in the metazoan orthologs of VPS13A/C. WWE domains in other proteins were shown to bind Poly-ADP-Ribose (PAR) (28).…”

Supramolecular architecture of the ER-mitochondria encounter structure in its native environment

“…In our final MDFF model, the cavities of the SMP domains do not form a continuous conduit, as they are too narrow for lipid transport at the Mdm12-Mdm34 and Mmm1-Mdm12 interfaces (Figures 4C and Supplementary Figure S8). This observation suggests that unlike other tunnel-like lipid transfer proteins such as the VPS13 family, which form large unrestrained conduits (30,31), the lipid pathway in ERMES could be restrained by bottlenecks at the interfaces between domains. Furthermore, in our model, the cavity of Mdm34 is near the OMM surface, and the predicted interaction between Mdm34 and the OMM protein Mdm10 occurs close to residues previously found to drive Mdm34-Mdm10 complex formation (20) (Figure 4D).…”