Conformational changes induced in HSV gD by receptor binding that lead to interactions with the regulatory complex gHgL.



Doina Atanasiu


Doina Atanasiu, Wan Ting Saw, Tina, M Cairns, Gary, H Cohen
Basic & Translational Sciences, University of Pennsylvania, School of Dental Medicine

 

Introduction

HSV requires four essential glycoproteins, gD, gHgL and gB and a cell receptor for virus entry and cell fusion. Glycoprotein D, the receptor binding protein, interacts with one of two receptors, HVEM or nectin-1. A comparison of free and receptor-bound gD crystal structures revealed that receptor binding domains are located within the N-terminus of gD, but problematically, the C-terminus lies across and occludes these binding sites. Consequentially, the C-terminus must relocate to allow for both receptor binding, and the gD interaction with the regulatory complex gH/gL. Earlier we constructed a di-sulfide bonded (K190C/A277C) protein that locked the C-terminus to the gD core. Importantly, this mutant bound receptor, but failed to trigger fusion, effectively separating receptor binding and gH/gL interaction.

Methods

Biacore 3000 biosensor system (SPR) was used to detect and measure the direct interaction between gD and gH/gL. First, an anti-gD Mab (1D3) was imine coupled to a CM5 SPR chip and was used to capture gD. After gD capture, soluble gH/gL was flowed across the chip surface. gH/gL binding is indicated by the increase in resonance units (RUs). Site-directed mutagenesis was used to generate 14 gD point mutations. The ability of these mutants to trigger cell-cell fusion was determined using the dual split protein assay (SLA) originally developed to study HIV mediated fusion. This assay uses a chimera of split forms of renilla luciferase and green fluorescent protein. Effector cells are co-transfected with glycoproteins and one of the split reporters. The receptor-bearing target cells are transfected with the second reporter. Co-culture results in fusion and restoration of renilla luciferase function, enabling us to monitor initiation and extent of fusion in live cells, in real time.

Results

Here, we show that “unlocking” gD by reduction restored fusion activity confirming the importance of C-terminal movement in triggering the fusion cascade. We characterize these changes showing that: (1) the C-terminus region exposed by unlocking is a gH/gL binding site; (2) it contains a competition community of monoclonal antibodies (Mabs) that block gH/gL binding and cell-cell fusion. We generated 14 mutations within the gD C-terminus to identify residues important for the interaction with gH/gL and the conformational changes involved in fusion. As one example, gD L268N was not only impaired for function in fusion, but the mutation played a major role in the binding of MC14, a Mab that blocks both the gD-gH/gL interaction and fusion. Subsequent experiments showed that although the L268N protein was antigenically correct, it failed to bind gH/gL and had reduced fusion activity, all events associated with the inhibition of C-terminus movement.

Conclusion

We conclude that binding of the nectin-1 receptor or specific neutralizing Mabs to the residues in the closed form of gD (gD306) induce conformational changes that allow for gH/gL binding. The unlocking of the disulfide bond in gD mutant cys2 by treatment with reducing reagents results in a marked gain of fusion function due to unfettered movement of the C-terminus of gD to allow access for gH/gL to all three binding sites. Site directed mutagenesis of the gH/gL binding site 3 (262-282) reveals residues important not only for binding of specific Mabs that were mapped to this region, but defines residues important for gD function and the interaction with gH/gL.