Ubiquitination by K63- and K48-linked chains was already known, before the discovery of linear ubiquitin chains, to play an important part in the activation of NF-κB, arguably the most crucial output of TNFR1 signaling. Activation of the TNFR1 pathway occurs when trimeric TNF crosslinks three TNFR1 monomers to initiate formation of the TNFR1 signaling complex (TNF-RSC). As schematically illustrated in Figure , TNFR1 activation results in the induction of gene activation by NF-κB and mitogen-activated protein kinases (MAPKs) and, depending on the strength of these gene-activatory signals, also in cell death, which can be either apoptotic (non-inflammatory) or necroptotic (inflammatory).
Figure 2 Model of TNFR1 signaling with and without LUBAC activity. Binding of trimeric TNF crosslinks the extracellular domains of three TNFR1 molecules and induces the formation of the TNF-RSC (also referred to as complex I). The tripartite LUBAC (ochre) is recruited (more ...)
NF-κB is a central transcriptional regulator in the induction of immune response genes that, in the absence of activating signals, is located in the cytoplasm. Activation of NF-κB occurs through the action of a kinase complex, referred to as the IkB kinase (IKK) complex, which consists of two catalytic subunits, IKKα (IKK1) and IKKβ (IKK2), and a critical regulatory subunit called NEMO (IKKγ). This complex is required to phosphorylate the inhibitor of NF-κB (IκB), thereby inducing its degradation and releasing NF-κB to relocate to the nucleus and bind to the promoters of immune genes. The IKK complex is recruited to the TNF-RSC through NEMO, and this results in activation of the kinase activity of this complex. MAPKs are activated as a result of recruitment of the TAB/TAK complex into the TNF-RSC. Whilst the TAB/TAK complex is currently thought to be recruited exlusively to K63-linked chains within the TNF-RSC, the IKK complex can be recruited to this complex via linear chains and, albeit with lesser affinity, also via K63- and K11-linked chains [33
LUBAC activity was first implicated in signaling from TNFR1 when TNF-mediated NF-κB activation was shown to be impaired in primary hepatocytes from HOIL-1 knockout mice, and LUBAC was shown to form part of the signaling complex that forms on binding of TNF by the receptor, and moreover to be crucial both to the stability of the TNF-RSC and in determining the outcome of TNF signaling [16
]. How LUBAC recruitment to the TNF-RSC influences signaling outcome is not known in detail, but it is known that NEMO, which is the regulatory component of the kinase complex that activates NF-κB, recognizes linear ubiquitin chains through its specialized ubiquitin-binding domain, UBAN (ubiquitin-binding domain present in ABINs and NEMO) [17
]. The UBAN motif is known also to recognize ubiquitin chains with other linkages - in particular K63 chains, which are also present on components of the TNF-RSC, including on RIP1 [19
]; but the UBAN of NEMO binds linear di-ubiquitin with a different topology and about 100-fold higher affinity than it does K63-linked di-ubiquitin. This suggests that the promotion of NF-κB activation by LUBAC following TNF stimulation may be due to linear ubiquitination of a component of the signaling complex whereby NEMO is recruited to, or retained in, the complex more effectively.
LUBAC also linearly ubiquitinates NEMO itself in the native TNF-RSC [19
]. TNF-induced linear ubiquitination of NEMO preferentially occurs on K285 and K309, and in cells expressing a NEMO K285R/K309R mutant, NF-κB activation induced by LUBAC overexpression or by stimulation with IL-1β was reduced [18
]. The mechanism of linear-ubiquitination-induced NF-κB activation has not been solved, but current data indicate that binding of NEMO to linearly linked ubiquitin induces a conformational change in the helical structure of NEMO that may promote the kinase activity of the IKK complex [17
]. Alternatively, recognition of linear chains by NEMO conjugated to the NEMO molecules of other IKK complexes could bring the kinase domains of the respective IKK complexes into close proximity, thereby enabling trans-autophosphorylation [17
], a process similar to the one that occurs between receptor tyrosine kinases when activated by ligand-induced dimerization.
Together, these findings indicate a functional role for linear ubiquitination in full gene activation by the signaling pathways triggered by TNF in vivo
. In the absence of LUBAC components the TNF-RSC still forms and activation of NF-κB still occurs, albeit at significantly reduced levels [20
]. Experiments with HOIP-deficient cells will be needed to strictly corroborate these findings, but it is likely that the NF-κB activation that still occurs in the absence of LUBAC is mediated by K63- and/or K11-linked chains, which are also present in the native TNF-RSC [19
] and can also bind or be attached to NEMO [33
Absence of LUBAC components also renders cells sensitive to TNF-induced cell death [16
]. Intriguigingly, this cell death is not only apoptotic [19
] but also necroptotic [19
]. Importantly, this is also true of primary keratinocytes obtained from young, non-diseased cpdm
mice. These mice, which are genetically deficient in SHARPIN and thus lack functional LUBAC complexes [19
], have played a central part in the discovery of the physiological function of LUBAC. They present with stark immune system developmental abnormalities, and develop a chronic multi-organ inflammatory syndrome with strong manifestation in the skin (hence the name of this mutation: chronic proliferative dermatitis
)) at about 4 to 6 weeks of age [40
]. The inflammatory syndrome that characterizes cpdm
mice is apparently paradoxical, because it is generally thought that aberrantly high TNF-induced gene activation is the source of inflammation induced by this cytokine. Our finding that TNF stimulation results in aberrant death of cpdm
-derived cells, and that this cell death has both an apoptotic and a necroptotic (and thus inflammatory) component [19
], suggested a different explanation: namely, that the inflammation in cpdm
mice could be due to inflammatory cell death consequent on the absence of SHARPIN-requiring LUBAC activity. To investigate this possibility, we crossed cpdm
mice with TNF-deficient mice, and were able to show that even partial genetic ablation of TNF prevented the formation of inflammatory lesions in cpdm
mice, indicating that TNF-induced cell death is indeed causative for the inflammatory phenotype that characterizes these mice [19
]. It is possible that secondary necrosis, which can occur as a consequence of apoptosis, may also contribute to inflammation in cpdm
Hence, linear ubiquitination is implicated in two different physiological processes: the development of the immune system and the prevention of chronic inflammation, where the latter effect is achieved through interference with TNF-induced cell death. Whether the aberrant cell death in the absence of LUBAC is due to reduced gene-inducing capacity of TNF, to a more direct effect of absence of linear ubiquitin chains from the signaling complexes induced by TNF, or perhaps to a combination of both these effects remains to be established. Our current suggestion for the contribution of LUBAC to these pathways is schematically illustrated in Figure .