Viral M45 and necroptosis-associated proteins form heteromeric amyloid assemblies.

The murine cytomegalovirus protein M45 protects infected mouse cells from necroptotic death and, when heterologously expressed, can protect human cells from necroptosis induced by tumour necrosis factor receptor (TNFR) activation. Here, we show that the N-terminal 90 residues of the M45 protein, which contain a RIP homotypic interaction motif (RHIM), are sufficient to confer protection against TNFR-induced necroptosis. This N-terminal region of M45 drives rapid self-assembly into homo-oligomeric amyloid fibrils and interacts with the RHIMs of the human kinases RIPK1 and RIPK3, and the Z-DNA binding protein 1 (ZBP1), to form heteromeric amyloid fibrils in vitro Mutation of the tetrad residues in the M45 RHIM attenuates homo- and hetero-amyloid assembly by M45, suggesting that the amyloidogenic nature of the M45 RHIM underlies its biological activity. The M45 RHIM preferentially interacts with RIPK3 and ZBP1 over RIPK1 and alters the properties of the host RHIM protein assemblies. Our results indicate that M45 mimics the interactions made by RIPK1 or ZBP1 with RIPK3, thereby forming heteromeric amyloid structures, which may explain its ability to inhibit necroptosis.

Pathological mutations differentially affect the self-assembly and polymerisation of the innate immune system signalling adaptor molecule MyD88.

BACKGROUND: Higher-order self-assembly of proteins, or "prion-like" polymerisation, is now emerging as a simple and robust mechanism for signal amplification, in particular within the innate immune system, where the recognition of pathogens or danger-associated molecular patterns needs to trigger a strong, binary response within cells. MyD88, an important adaptor protein downstream of TLRs, is one of the most recent candidates for involvement in signalling by higher order self-assembly. In this new light, we set out to re-interpret the role of polymerisation in MyD88-related diseases and study the impact of disease-associated point mutations L93P, R196C, and L252P/L265P at the molecular level. RESULTS: We first developed new in vitro strategies to characterise the behaviour of polymerising, full-length MyD88 at physiological levels. To this end, we used single-molecule fluorescence fluctuation spectroscopy coupled to a eukaryotic cell-free protein expression system. We were then able to explore the polymerisation propensity of full-length MyD88, at low protein concentration and without purification, and compare it to the behaviours of the isolated TIR domain and death domain that have been shown to have self-assembly properties on their own. These experiments demonstrate that the presence of both domains is required to cooperatively lead to efficient polymerisation of the protein. We then characterised three pathological mutants of MyD88. CONCLUSION: We discovered that all mutations block the ability of MyD88 to polymerise fully. Interestingly, we show that, in contrast to L93P and R196C, L252P is a gain-of-function mutation, which allows the MyD88 mutant to form extremely stable oligomers, even at low nanomolar concentrations. Thus, our results shed new light on the digital "all-or-none" responses by the myddosomes and the behaviour of the oncogenic mutations of MyD88.