Following the white rabbit : release and regulation of HMGB1 in inflammation

Abstract

Dysregulated inflammatory responses are characterized by excessive release of endogenous pro-inflammatory molecules, danger-associated molecular patterns (DAMPs). A prototypical DAMP, High Mobility Group Box 1 protein (HMGB1), is primarily bound to DNA in the nucleus of most eukaryotic cells. However, when HMGB1 exits the cell, it acquires novel functions and can trigger immune activation. The aim of this thesis was to study the release of HMGB1, follow its fate at the site of inflammation and therapeutically block its activity.

Multiple studies have reported that cells release HMGB1 in either a passive or an active manner. In order to understand the possible impact of cell death in inflammatory conditions, we performed well-controlled in vitro analyses of HMGB1 release under different cell death modes. Necrotic cells instantly released high levels of HMGB1, while apoptotic cells retained HMGB1 unless they advanced to secondary necrosis. The controlled immunogenic cell death types, such as pyroptosis and necroptosis, secreted HMGB1 along with other pro-inflammatory molecules in comparable or higher levels than cells activated with lipopolysaccharide (LPS) and interferon gamma (IFNγ).

Next, we investigated HMGB1 proteolytic regulation at the site of inflammation. Human neutrophil elastase (HNE), cathepsin G (CG) and matrix metalloproteinase 3 (MMP-3), endoproteases present in arthritis-affected joints, cleaved HMGB1 into smaller peptides in vitro. While HNE and MMP-3 truncated the acidic C-terminal tail domain of HMGB1, CG completely degraded the protein. We showed that the C-terminal tail truncation negatively regulated HMGB1 binding to Toll-like receptor 2 (TLR2). Only the protein form devoid the acidic tail domain bound TLR2 in vitro. Thus, truncation of HMGB1 by endoproteases might improve its affinity to receptors. TLR2 and HMGB1 interaction did not result in cytokine induction. However, both full-length and truncated HMGB1 formed complexes with potent TLR2 ligand peptidoglycan (PGN) to potentiate the inflammatory response in peripheral blood mononuclear cells (PBMCs). Lastly, we developed a novel tool that could enable anti-HMGB1 therapy in the clinic by engineering a chimeric, partially humanized, monoclonal antibody (h2G7). The antibody displayed anti-inflammatory effects when tested in a mouse model of paracetamol-induced liver toxicity. Experiments with mutated variants, non-Fc receptor binding and non-complement binding variants, of the novel antibody suggested HMGB1 neutralization as the mode of action.

Taken together, results presented in my thesis increased the knowledge about HMGB1 biology regarding its release during immunogenic cell death, proteolytic regulation at the site of inflammation and interaction with its receptor TLR2. In addition, the generation of a partially humanized monoclonal antibody could promote the introduction of anti-HMGB1 therapy in the clinic.