Development of therapeutic glycopolymers for the treatment of peripheral neuropathies and viral infections

Immune mediated neuropathies are a group of heterogenous disorders affecting the peripheral nervous system. Characteristic for these diseases is the involvement of an immune response against autoantigens on axonal membranes or surrounding myelin. In a variety of peripheral neuropathies autoantibodies target glycan or peptide epitopes in the nodal and paranodal region. These polyneuropathies can have chronic or acute manifestations but generally respond to immunotherapies. Most treatment options however target the immune system unspecifically and can cause serious adverse effects. In some cases, patients do not respond to treatment or even show deterioration. In anti-MAG neuropathy and multifocal motor neuropathy the antigen-specific pathogenic autoantibodies have been well described and therapeutic intervention is aimed at reducing antibody levels or interfering with its effector functions.
A recently developed glycopolymer-based therapeutic approach for the treatment of anti-MAG neuropathy (PPSGG) specifically targets these disease-causing autoantibodies by presenting multiple carbohydrate mimetic copies of its antigen on a biodegradable poly-L-lysine (PLL) scaffold.
In this thesis we review the importance of a reduction of autoantibody levels for clinical improvement in anti-MAG neuropathy, discuss the mechanism of action of PPSGG, evaluate its safety profile, and develop new glycopolymers to treat related peripheral neuropathies and viral infections.
Despite clinical evidence for the pathogenicity of anti-MAG IgM autoantibodies, the significance of antibody titers as a predictive factor for response to therapy remains controversial. Current literature does not provide conclusive evidence on the association between reduced anti-MAG IgM titers and clinical improvement of neuropathic symptoms. We performed a retrospective study to test our hypothesis that changes in antibody titers are correlated with clinical response in anti-MAG neuropathy patients. We included 50 studies involving 410 anti-MAG neuropathy patients undergoing immunotherapy in our analysis and characterized relative change in anti-MAG IgM titers, paraprotein levels, and total IgM prior, during, or post-treatment. Patients were categorized according to their response to treatment into “responders”, “non-responders”, or “acute deteriorating” and the studies were qualified as “supportive” or “not supportive”. 40 studies supported the hypothesis that “responders” showed relative reduction of anti-MAG IgM titers and “non-responders” did not show significant change in antibody titers, confirming our hypothesis.
In an experimental study we further investigated the pharmacodynamic properties of PPSGG and tested its inhibitory potential on peripheral nerves. The polymer selectively prevented binding of anti-MAG IgM from patient sera to non-human primate sciatic nerves. In an immunological mouse model, PPSGG showed superiority in removing anti-MAG IgM antibodies compared to B-cell depletion with an anti-CD20 antibody (analogous to Rituximab). Safety evaluation with human and murine peripheral blood mononuclear cells (PBMC) showed no interaction. Furthermore, no increase in systemic inflammatory markers was observed in mice or in human PBMC ex vivo after treatment.
We investigated the binding characteristics of PPSGG and anti-MAG IgM as well as pharmacokinetic properties to understand the mode of action of the glycopolymer. First, physicochemical and morphological characteristics of the polymer were investigated. The linear rod-shaped PPSGG showed an approximate length of 100 nm, a hydrodynamic radius of around 60 nm, and a highly negative charge ( 46 mV ). Because of its high negative charge density, it was readily taken up by liver-and spleen-resident macrophages through scavenger receptors, explaining the previously observed short half-life in mice (approx. 17 min). No large aggregate formation in vitro and no immune complex deposition in murine liver, spleen, kidney, or brain was observed. Despite the extensive uptake in Kupffer cells in the liver, it did not exhibit hepatotoxic effects in a human hepatic tissue ex vivo. In the presence of anti-MAG IgM, PPSGG preferentially formed complexes in a 1:1 or 1:2 stoichiometry in vitro, supporting previously reported dose titration experiments with an immunological mouse model.
Analogous to the approach for the treatment of anti-MAG neuropathy, we developed a glycopolymer displaying GM1 carbohydrate mimetics on a PLL backbone to treat anti-GM1 mediated peripheral neuropathies. A series of ten mimetics was synthesized and conjugated to the PLL backbone for testing with MMN patient sera. The specificity and inhibitory potential of the glycopolymers was assessed by competitive ELSIA. After screening of 22 MMN patient samples we identified three interesting candidates for further evaluation in functional assays. One of most active glycopolymers was carrying the natural GM1 carbohydrate epitope and was excluded for further selection, because the aim of the project was to develop simplified glycomimetcs that retain or increase binding affinity while simultaneously reducing synthetic complexity. One of the two remaining candidates showed high temperature sensitivity and could not inhibit anti-GM1 antibody binding to GM1 at physiological temperatures. The remaining glycopolymer, composed of the natural GM1 core epitope with only a replacement of the terminal glucose by a tyramine moiety, prevented anti-GM1 antibody binding to terminal axonal networks in vitro and ex vivo using an animal model for acute motor axonal neuropathy.
Being confronted with the surge of the COVID-19 pandemic and the rapid progress in scientific discoveries and understanding of SARS-CoV-2, we applied our expertise in glycoplymer development to tackle SARS-CoV-2 infections. Increasing evidence was pointing at the involvement of DC-SIGN in the pathogenesis of COVID-19. Interaction of viral spike protein with DC-SIGN was reported to lead to internalization of the virus by immune cells, thus presenting an alternative entry receptor for SARS-CoV-2, independent of ACE2. Infection of immune cells is involved in exaggerated immune response in severe COVID-19. It was later reported that DC-SIGN levels were increased in severe COVID-19 patients showing elevated levels of proinflammatory macrophages, inflammatory cytokines and chemokines. Inhibition of the interaction between DC-SIGN and spike protein might serve as a strategy to prevent these severe disease courses. We demonstrated that mannose-functionalized PLL glycopolymers efficiently inhibit the attachment of spike protein to DC-SIGN presenting cells with picomolar affinity in a competitive setting. Pre-treatment of the cells lead to prolonged receptor internalization and protected the cells for up to 6 h from virus binding. Moreover, DC-SIGN acts as a receptor for multiple viruses and we could additionally demonstrate effective inhibition of HIV and Ebola viral glycoprotein biding to DC-SIGN presenting cells. This host-directed approach might not only be applicable for multiple unrelated viral infections but could be unaffected by the rapidly mutating variants of SARS-CoV-2.
In a follow-up study we reported the discovery of a new class of potent glycomometic DC SIGN ligands from a library of triazole-based mannose analogs. Structure-based optimization yielded a glycomimetic ligand with over 100-fold improved binding affinity compared to methyl α D mannopyranoside. Multivalent display of the ligand on PLL was able to inhibit SARS-CoV-2 spike glycoprotein binding to DC-SIGN expressing cells, as well as DC-SIGN mediated trans-infection of ACE2 expressing cells by SARS-CoV-2 spike protein expressing viruses in nanomolar concentrations.