Disulfide engineering of immunostimulatory antibodies for augmented receptor agonism and immune cell stimulation

Abstract

Immunostimulatory antibodies represent a promising strategy for cancer immunotherapy. By binding and activating co-stimulatory molecules, such as certain tumour necrosis factor receptors (TNFRs), expressed on immune cells, immunostimulatory antibodies can enhance the immune response towards tumours, resulting in powerful anti-cancer effects. Antibody binding aims to induce clustering of receptors on the cell surface, which leads to receptor activation and downstream immune signalling, resulting in CD8+ T cell activation and tumour cell killing. However, despite strong activity in pre-clinical models, with several antibodies targeting TNFRs reaching clinical trials, translation into the clinic has been largely unsuccessful, due to weak agonism or dose-limiting toxicity. Suboptimal clinical outcomes are attributed, in part, to the requirement of co-stimulatory receptors to undergo higher-order clustering to achieve efficient signalling, with differences in bivalent antibody geometry and valency compared to the natural ligand limiting clustering capacity.
The hIgG1 isotype (often used in clinical candidates) is generally a poor agonist, relying on FcγR-mediated cross-linking to induce receptor oligomerisation and downstream signal transduction. In contrast, the hIgG2 isotype can deliver strong FcγR-independent agonistic activity via increased receptor clustering, due to its unique hinge disulfide arrangement, which provides conformational restriction and rigidity. Furthermore, for a clinically-relevant anti-hCD40 antibody, ChiLob7/4, modification of hIgG2 hinge disulfides, through cysteine-to-serine (C/S) exchange mutations, was shown to influence immune-stimulating activity via changes in conformational flexibility.
In this thesis, the importance of conformation and rigidity in immunostimulatory antibodies is further explored, using structure-guided disulfide engineering strategies to enhance receptor agonism and immune cell stimulation.
First, the hIgG2 C/S exchange mutations, previously evaluated in ChiLob7/4, were introduced into several additional antibodies targeting the co-stimulatory TNFRs CD40, 4-1BB and OX40. The impact on binding, activity and conformation was assessed using a range of cellular, biophysical and structural techniques,
demonstrating that the principles of increasing agonism by restricting antibody conformation through disulfide modification were applicable across different TNFRs.
To investigate this concept further, and move beyond nature, structure-guided approaches were used to design hIgG2 antibody variants in ChiLob7/4, based upon the framework of the most agonistic C/S (disulfide ‘cross-over’) variant, with additional disulfides introduced between F(ab) arms, both within and outside of the hinge. Using an integrative approach combining X-ray crystallography with anomalous scattering, small angle X-ray scattering (SAXS), molecular dynamics (MD) simulations and cellular immunoassays, subtle increases in rigidity were found to deliver significant improvements in FcγR-independent immunostimulatory activity.
However, these engineering approaches rely on the specific amino acid sequence of the hIgG2 hinge region, limiting translational utility to other hIgG isotypes. Therefore, next a set of X/C exchange mutations were evaluated, outside the context of the hIgG2 hinge disulfide cross-over, in hIgG1, hIgG2 and hIgG4 using ChiLob7/4 as proof of concept. The introduction of a single disulfide between opposing heavy chains in the CH1 domain above the hinge region was shown by X-ray crystallography with sulfur single wavelength anomalous scattering experiments, to constrain F(ab) arms in close proximity. This compact conformation was validated using negative stain electron microscopy, small angle X-ray scattering and MD simulations, indicating a reduction in conformational diversity compared to the native antibodies. The unique constrained conformation translated to significant increases in FcγR-independent agonism in NF-κB GFP reporter assays and human B cell assays (proliferation, activation and homotypic adhesion), concomitant with enhanced receptor clustering assessed by DMD-based confocal microscopy. This disulfide engineering approach was implemented into multiple antibodies of the hIgG1 isotype targeting four different TNFRs, evoking improved receptor agonism associated with increased compactness.
Finally, the unexpected dimerisation of one of the antibodies, upon introduction of hIgG1 engineered disulfides, was briefly explored using a variety of biophysical and structural approaches, with the light chain complementarity determining region 2 identified as the likely determinant for low affinity homotypic self-association. This led to the formation of dimeric ring-shaped molecules in solution, mediating higher levels of receptor agonism.
Together these disulfide engineering strategies, validated through a range of orthogonal and complementary integrative methods, provide a rational means to modify antibody conformation and rigidity, enabling the tuning of immunostimulatory agonistic activity for more effective anti-cancer therapeutics.

More information

Identifiers

ORCID for Isabel Grace Elliott: orcid.org/0000-0003-4535-5932

ORCID for Jonathan Essex: orcid.org/0000-0003-2639-2746

ORCID for Mark Cragg: orcid.org/0000-0003-2077-089X

ORCID for Ivo Tews: orcid.org/0000-0002-4704-1139

Catalogue record

Export record

Altmetrics