Skip to main content

Deciphering the ubiquitin code of the TNF receptor signalling complex and its functional role in inflammation and immunity

Final Report Summary - UB-DECODED (Deciphering the ubiquitin code of the TNF receptor signalling complex and its functional role in inflammation and immunity)

TNF is a pro-inflammatory cytokine which plays a crucial role in response to infections, auto-immunity and cancer related-inflammation. Poly-ubiquitination has emerged as a master-regulator of signalling pathways. Eight types of poly-ubiquitin chains have been described and are designated depending on the residue targeted for the attachment of the incoming ubiquitin onto the target ubiquitin (i.e K6, K11, K27, K29, K33, K48, K63 and M1 chains). M1 chains are a crucial modulator of immune receptor signalling, including that of TNFR1. Thus, we have focused on LUBAC (composed of the components HOIP, HOIL-1 and SHARPIN), the only known E3-ligase identified so far that is capable of catalysing the de-novo generation of M1-linked ubiquitin chains under native conditions. Upon binding of TNF to TNFR1, a first complex is formed called the TNF-Receptor Signalling Complex (TNF-RSC) or complex I which is responsible for the induction of gene-activatory pathways. A cytosolic complex is also formed, secondarily, upon TNF stimulation, likely owing to the deubiquitination of TNF-RSC components and hence to the destabilisation of complex I. This second complex, termed complex II, accounts for the death-inducing properties of TNF. In the context of TNF signalling, we had previously shown that LUBAC, by ubiquitinating RIPK1 and NEMO, stabilizes complex I and promotes TNF-induced gene-activatory signalling whilst restricting the formation of complex II and thus preventing TNF-induced cell death.

During the duration of this ERC project, we have aimed at precisely defining the role of LUBAC and linear ubiquitination events in TNF signalling and their relevance towards some of the physiological and pathological roles of this cytokine. Our efforts have led to several important findings:

- We have generated novel ubiquitin-binding tools allowing for enrichment in total ubiquitin, K63-chains or M1-chains with previously unreached sensitivity and specificity. The use of these tools enabled us to identify TNFR1 and TRADD as novel LUBAC substrates in the context of TNF signalling (Draber et al. Cell Reports, 2015) and has contributed to the discovery LUBAC’s role in additional signalling pathways triggered by other receptors capable of inducing cell death and gene activation (Draber et al. Cell Reports, 2015; Zinngrebe et al., Journal of Experimental Medicine 2016; Lafont et al., EMBO Journal 2017).

- We have identified HOIP, independently of its activity, as being required for the recruitment of the deubiquitinase CYLD to the TNF-RSC. Importantly, this is not limited to TNF signalling but likely constitutes a general modality of CYLD recruitment, as we also observed it in NOD2 and TRAILR1/2-driven signalling activation (Draber et al. Cell Reports, 2015; Lafont et al., EMBO Journal 2017). CYLD, which we demonstrated to have dual specificity for K63- and M1-chains in signalling complexes, is a crucial mediator of TNF-induced death. Importantly, we demonstrate that, within the TNF-RSC, CYLD removes linear ubiquitin chains from TNFR1, TRADD and RIPK1, hence limiting TNF-induced death (Draber et al., Cell Reports 2015). Moreover, we also demonstrated that CYLD does not interact directly with HOIP but that it does so via the adaptor protein SPATA2, an additional, previously unrecognised factor in TNF signalling that is, consequently, also required for TNF-induced cell death (Kupka et al., Cell Reports 2016).

- We have shown that, in addition to CYLD, also A20 is recruited to the TNF-RSC in a LUBAC-dependent manner. However, in the case of A20 the recruitment relies on the specific binding of A20 to linear chains via its Znf7 (Draber et al., Cell Reports 2015). Importantly, our work demonstrates that the DUB activity of A20 is entirely dispensable for its role in restricting TNF-induced death and gene-activatory signalling (Draber et al., Cell Reports 2015). Moreover, A20 stabilizes linear chains within the TNF-RSC, preventing the dismantling of this complex and likely competing with gene-activatory molecules (e.g. NEMO).

- We have generated constitutively HOIP-deficient mice as well as constitutively HOIL-1-deficient, skin-specific HOIP- and HOIL-1-deficient and liver-specific HOIP-deficient mice. This has allowed us to address the physiological role of LUBAC. The first study showed that LUBAC deficiency leads to embryonic lethality (Peltzer et al., Cell Reports 2014). The study regarding liver-specific HOIP deficiency demonstrated that LUBAC acts as a liver tumour suppressor (Shimizu et al., Hepatology 2017). The results of the other studies are currently under revision and considered for publication.

We discovered a pro-tumourigenic role for endogenous TRAIL/TRAIL-R-induced signalling in KRAS-driven lung and pancreatic cancer (von Karstedt et al., Cancer Cell 2015) and demonstrated that TRAIL-induced cytokine secretion promotes a tumour supportive immune microenvironment (Hartwig et al., Molecular Cell 2017).