Structures of Marburgvirus glycoprotein and its complex with NPC1 receptor
Summary
This Nature study presents the first cryo-EM structures of Marburgvirus (RAVV) glycoprotein (GP) in three states: unbound GPcl (3.17 Å), GPcl bound to human NPC1-C (3.53 Å) and GP-ΔM bound to a neutralising nanobody Nanosota-MB1 (2.98 Å). The authors show that MBV GP mediates cell entry far more efficiently than Ebola GP, reveal a partially constrained glycan-cap loop that regulates receptor access, describe a distinct, higher-affinity NPC1 binding mode for MBV, and identify a nanobody that neutralises MBV by mimicking NPC1 binding.
Key Points
- First high-resolution cryo-EM structures of Marburgvirus GP (RAVV) in apo form and bound to NPC1-C and a neutralising nanobody.
- RAVV GP-driven entry into human cells is substantially more efficient than EBOV GP (reported up to ~300-fold in some cell types under matched expression conditions).
- A nine-residue glycan-cap loop (loop 207) occupies the receptor-binding site in ligand-free GPcl, indicating partial constraint rather than full flexibility.
- NPC1-C binds RAVV GPcl in a rotated orientation using three loops (including a newly engaged loop 3), producing ~11-fold stronger affinity than EBOV GPcl.
- NPC1 binding loosens GP protomer packing and triggers conformational changes that favour the pre-fusion-to-post-fusion transition.
- Nanosota-MB1 is the first reported neutralising nanobody against MBVs; it mimics NPC1, binds the RBS with ultra-high affinity and neutralises multiple MBV strains (IC50 ≈ 0.5 µg ml−1).
- Structural data and deposited models (PDB: 9NPS, 9NPR, 9NPT; EMDB: EMD-49631, EMD-49630, EMD-49632) provide concrete targets for antivirals and antibody design.
Content summary
The team expressed and purified RAVV GP constructs (including GP-ΔM and GPcl) and human NPC1-C, solved three cryo-EM structures, and complemented structural work with surface plasmon resonance, mutagenesis and pseudovirus entry/neutralisation assays. Key mechanistic findings are:
1) A glycan-cap-derived loop remains bound in the receptor-binding site of GPcl despite proteolytic removal of the cap, showing the glycan cap can partially shield the RBS yet remain constrained.
2) NPC1-C contacts the RAVV RBS with three loops (not two as in EBOV), rotating its approach and creating extra anchoring that correlates with an ~11-fold stronger KD for RAVV GPcl versus EBOV GPcl.
3) NPC1 binding expands distances between GP protomers and reduces buried interfaces, likely lowering the barrier to the fusion-active state.
4) The nanobody Nanosota-MB1 occupies the same hydrophobic pocket as NPC1 loop 2, competitively blocks receptor binding and effectively neutralises pseudoviruses bearing RAVV and MARV GPs.
Context and relevance
Marburgviruses (MARV and RAVV) are highly lethal filoviruses with a ~73% average case fatality rate versus ~44% for Ebola. There are no licensed vaccines or therapeutics for MBVs. Understanding GP-mediated entry is therefore critical for pandemic preparedness. These structures reveal why MBV entry may be more efficient than Ebola — a combination of glycan-cap dynamics, a reshaped receptor-binding pocket and stronger NPC1 engagement — and highlight the GPcl–NPC1 interface as a priority drug and antibody target.
Why should I read this?
Quick take: if you work on antivirals, antibodies, or filovirus biology — read it. The paper explains, in structural detail, why Marburgvirus might be so good at getting into human cells and gives actionable targets (including a potent nanobody) that could be developed into therapeutics. If you’re short on time, the key bits are the NPC1 binding differences and the nanobody neutralisation data.
Author style
Punchy: this is a high-impact, must-read structural paper. The authors not only solve first-of-their-kind structures but link them to function and neutralisation — so the work goes straight from molecular mechanism to clear translational leads.