Elucidating Biophysical and Structural Mechanisms of Multivalent Lectin-Glycan Binding using Glyconanoparticle Probes

Multivalent lectin-glycan interactions (MLGIs) are widespread and vital for biology, and also hold the key to many therapeutic applications. However, the underlying structural and biophysical mechanisms for many MLGIs remain poorly understood, limiting our ability to design glycoconjugates that can potently target specific MLGIs for therapeutic intervention. Glycosylated nanoparticles (NPs) have recently emerged as powerful biophysical probes for studying MLGIs, revealing key information regarding the binding behaviours of different lectins, which can subsequently inform their therapeutic potential. Despite these advances, however, the structural and biophysical mechanisms behind the specific binding behaviours of different lectins towards glycan-NPs is largely unknown. This research focuses on furthering our understanding of these mechanisms by exploiting the unique multifunctional properties of glycan-conjugated quantum dots (glycan-QDs) and gold nanoparticles (glycan-GNPs), as well as other nanoscale tools, to improve our understanding of the structural, thermodynamic and kinetic rationale behind the MLGIs of two closely related and immunologically interesting tetrameric lectins, DC-SIGN and DC-SIGNR.

DC-SIGN and DC-SIGNR (collectively, DC-SIGN/R) display near-identical mannose-binding motifs and have both been identified to facilitate viral infection. Despite this, these lectins have been shown to have very different affinities to some of the same viral glycoproteins or other glycoconjugates, both in solution and on cell surfaces. This has been attributed to their differences in multivalent spatial specificity, induced by differences in their tetrameric structures. For example, our group previously demonstrated that upon incubation with the same glycan-QDs, DC-SIGN bound simultaneously to individual QDs, whereas DC-SIGNR formed much weaker crosslinking interactions. Here, a QD-FRET technique for determining the thermodynamic and kinetic contributions of MLGIs in solution has been developed, and has revealed that, though both DC-SIGN/R display similar enthalpically driven MLGIs with mannosylated QDs (4× that of monovalent binding), DC SIGNR incurs a greater entropic penalty and slower kinetics which can be attributed to its less favourable crosslinking binding mode (Chapter 3). Furthermore, a 16 amino acid C-terminal (not present in DC SIGNR) has been shown to contribute to the specificity of DC SIGN MLGIs; the removal of which completely alters the thermodynamics towards entropically driven binding, and enables the possibility of crosslinking. In addition to these assays, DC-SIGN/R-functionalised supported lipid bilayers have been developed to better replicate the native environment of lectins on cell membranes (Chapter 5). Here, QCM-D and cryo-EM studies with mannosylated GNPs show that these different binding modes can still be observed on membrane surfaces, where DC-SIGNR⋅glycan-GNP binding is highly dependent upon the lectin surface density, whereas DC-SIGN binding shows no such dependency. This may therefore contribute to the observed differences between DC-SIGN/R MLGIs in both solution-phase and cell surface assays.

The crosslinking behaviour of DC-SIGNR can be attributed to the inability of its four binding sites to bridge the glycan display of an individual glycan-NP. Here, a single molecule FRET technique has been developed, using a new tetramer FRET pair labelling strategy, in conjugation with MD simulations to provide an estimate of the inter-binding site distances of DC SIGN/R (Chapter 6). These results demonstrate a broader tetrameric model for DC-SIGNR which may explain its inability to tetravalently bind to glycan-NPs of these sizes. However, this has been demonstrated to be overcome using glycosylated quantum rods (glycan-QRs), whereby the reduced surface curvature of the cylindrical middle section of these nanomaterials is able to encourage tetravalent binding of all four binding sites of DC-SIGNR, reducing the amount of crosslinking and thus resulting in stronger MLGIs (Chapter 4).

Additionally, understanding of these interactions has also paved the foundation of other innovative nano-therapeutics which apply these principles. Here, DC SIGN-conjugated NPs have been developed to target the glycosylation sites of the SARS-CoV-2 S protein, showing to effectively neutralise a range of pseudotyped and authentic SARS-CoV-2 variants (Chapter 8).

Overall, the findings presented in this thesis reveal crucial information regarding the binding of key MLGIs which can be used to help inform the rational design of more potent and specific lectin or glycan targeting agents for potential therapeutic applications.