Biochemical and structural studies of NOD-like receptors and their inhibition by small molecule inhibitors

Abstract

NOD-like receptors (NLRs) are innate immune sensors that can form inflammasomes driving pyroptosis. NLRP3 is the best studied member of the NLRs to date, and its dysregulation has been linked to human diseases relevant to our contemporary aging society. NLRP9 and NLRP12 are less well studied, but they have been described to form inflammasomes during infections. However, the existence of both inflammasomes has never been biochemically confirmed, and the mode of action of small molecule inhibitors is poorly understood. In this thesis, I show that recombinant human NLRP9 forms a defined and stable monomer in solution that is expected to adopt an ADP-bound and inactive conformation. Overexpression of NLRP9 in cells is not sufficient to nucleate ASC specks, which contradicts inflammasome formation and is in great contrast to NLRP3 or NLRP12. In line, the NLRP9 Pyrin domain (PYD) does not polymerize into filaments or interact with ASC. Based on a 1.95 Å crystal structure of NLRP9-PYD, these observations can be explained by finding several mismatches in residues that would otherwise form interfaces in a filament. <br /> Recombinant human NLRP12 associates with tubulin superfamily proteins, suggesting a role of microtubules in inflammasome formation. In addition, the detergent CHAPS can abrogate ATPase activity, which might indicate NLRP12 activation at lipid membranes. However, NLRP12^PYD does not polymerize into filaments or interact with ASC, contradicting inflammasome formation. Based on a previously determined crystal structure of NLRP12^PYD, this discrepancy was investigated but did not yield a clear explanation. Since recombinant NLRP12^NACHT can assemble into oligomers, it is supposed that the NACHT domain acts as a scaffold for PYD polymerization. Crystals of NLRP12-^NACHT have been generated under various conditions, but have not yet diffracted sufficiently for structure determination. <br /> The development of an optimized purification protocol allowed the determination of a 2.48 Å crystal structure of NLRP3 in complex with the prototypic inhibitor CRID3. NLRP3^NACHT adopts an ADP-bound and inactive conformation stabilized by three intramolecular interdomain interfaces, each containing a disease-relevant mutation site. CRID3 binds NLRP3 with nanomolar affinity and inhibits ATPase activity. The binding site is formed by a cleft located on the backside of the Walker A motif and is also required to adopt the active state. In this way, CRID3 glues four subdomains of the NACHT domain together with the transition LRR and locks NLRP3 in the inactive conformation. Binding experiments demonstrated that CRID3 can be extended at the eastern side without loss of interaction, and that substitution of the furan moiety could be an option for the development of advanced NLRP3 inhibitors with lower hepatotoxicity. I anticipate that these data could pave the way for rational and structure-guided drug optimization in the future not only for NLRP3 but for all NOD-like receptors.