|Home | About | Journals | Submit | Contact Us | Français|
The inhibitor of kB kinase (IKK) complex, a critical positive regulator of nuclear factor κB (NF-κB) signaling and immune function, can be activated in vitro by polyubiquitin chains. The modification of signaling intermediates by polyubiquitin and their association with ubiquitin-binding domain-containing proteins within multimolecular signaling complexes therefore represent a potential mechanism for activation of IKK and NF-κB in vivo.
The transcription factor nuclear factor κB (NF-κB) controls the expression of genes that mediate immunity and inflammation. Because the outcomes of the NF-κB inflammatory program are extensive, its activation is maintained under tight cellular control. A paradigm for signaling to NF-κB has emerged through two decades of research. NF-κB dimers constitutively associate with the inhibitors of κB, IκBα, IκBβ and IκBε in resting cells. Association with IκB masks the nuclear localization signal on NF-κB and retains NF-κB dimers in the cytosol, segregated from target genes. Stimulation of the lymphocyte antigen receptor (AgR), Toll-like receptors (TLRs) and receptors to inflammatory cytokines activate the inhibitor of kB kinase (IKK) complex to phosphorylate IκB molecules and promote their degradation. NF-κB dimers are thereby freed to enter into the nucleus and activate transcription of target genes. Although NF-κB activation has been studied in great detail, the mechanism by which the IKK complex is activated still remains somewhat unclear: unlike signaling driven by cascades of protein phosphorylation, the protein associations and post-translational modifications that direct the activation of IKK are still being elucidated.
A recent report by Zhijian Chen's group in Nature (Xia et al., 2009) provides new insight into how IKK is activated. They demonstrate that tumor-necrosis factor-activated kinase 1 (TAK1), a critical regulator of IKK, and the IKK complex itself can be activated by unanchored polyubiquitin chains in a cell-free system. Although it was known that protein ubiquitination influences the activation of IKK, this study provides the first evidence that polyubiquitin chains can directly activate protein kinases and therefore stimulate the NF-κB pathway. Protein activation by polyubiquitin chains is mediated by specialized protein ubiquitin-binding domains (UBDs). Transactivation of TAK1 was shown to require the UBD of the TAK-1-associated TAK1 binding protein (TAB2), whereas the UBD for IKK activation is located in the NF-κB essential modulator (NEMO), a regulatory component of the IKK complex (Figure 1). Additionally, the authors make the novel observation that polyubiquitin chains generated in vitro can be unanchored from their protein targets, and that free ubiquitin polymers containing unique K63 linkages support enhanced activation of IKK and TAK1.
This report represents the first evidence of direct activation of a protein kinase by polyubiquitin molecules. Ubiquitin-mediated kinase activation not only suggests a mechanism for IKK activation in vivo, it also emphasizes the potential importance of polyubiquitin modifications in various components of the NF-κB signaling pathway. The diversity of polyubiquitin chains generated by multiple ubiquitin ligases along with differences in their associations with protein targets further helps explain how UBD-containing kinases and regulatory proteins might be selectively activated.
This and other reports suggesting a key role for ubiquitin in cell signaling, however, also raise a biological conundrum: how can a molecule as ubiquitous and abundant as ubiquitin specifically activate signaling molecules and enzymes, especially in a system that must be as tightly controlled as NF-κB activation. Perhaps the assembly of particular complexes of upstream and accessory signaling molecules may allow for specific recruitment of the kinases that are to be activated. Downstream of the AgR, a multimolecular complex of polyubiquitinated signaling intermediates including B cell lymphoma protein Bcl10, mucosa-associated lymphoid tissue lymphoma translocation gene 1 (Malt1) and tumor necrosis factor receptor-associated factor (TRAF6) are thought to recruit IKK to the plasma membrane (Schulze-Luehrmann and Ghosh, 2006). Similarly, following engagement of TLRs and cytokine receptors, TRAF2, receptor interacting protein (RIP-I) and interleukin-1 receptor-associated kinases (IRAK) are known to undergo modification by ubiquitin (Hayden and Ghosh, 2008; Bhoj and Chen, 2009). Although it was known that ubiquitination of these signaling molecules contributed to NF-κB activation, the mechanism remained unclear. Some reports proposed that polyubiquitinated signaling proteins provided a scaffold for protein–protein interactions, whereas other reports suggested ubiquitination triggered protein oligomerization, a necessary step in assembly of signaling complexes. These functions likely contribute to proximal signaling, as the recruitment of the IKK complex to multimolecular surface complexes is required for efficient activation. However, the current paper suggests that ubiquitin polymers, besides functioning in protein complex formation and oligomerization, may promote TAK1 and IKK activation by direct interaction.
As the importance of ubiquitin for cell signaling becomes better established, the nature of the polyubiquitin chains that participate in cellular activation is also gathering attention. Conventional polyubiquitin chains that promote protein degradation via the proteasome are conjugated through lysine K48 residues on ubiquitin molecules. However, many polyubiquitin chains detected on signaling adapters in the NF-κB pathway contain K63 linkages, known as regulatory ubiquitin. Xia et al. convincingly demonstrate that only K63-linked polyubiquitin chains are capable of activating TAK1, as mutation of the K63 residue abrogates TAK1 activation in their cell-free system. In contrast, optimal activation of the IKK complex was achieved using ubiquitin polymers containing both K48 and K63 linkages. Interestingly, linear polyubiquitin chains that were recently implicated in NEMO binding and IKK activation were not able to support IKK activation in this system, perhaps due to a missing cofactor or adapter molecule (Rahighi et al., 2009). The association between polyubiquitin chains and target protein also varies depending on the nature of the chain. Unlike polyubiquitin chains covalently attached to TRAF6 or IRAK, TAB2 and NEMO-associated ubiquitin chains were found to be unanchored and susceptible to N-terminal ubiquitin cleavage. Interestingly, UBDs display some specificity for ubiquitin chains in vitro, as K63-linked polyubiquitin chains non-covalently associated with TRAF6 preferentially activated TAK1. K63 chains are formed by UBC13, which is the best-studied E2 enzyme in the NF-κB pathway (Yamamoto et al., 2006a). The authors also examined polyubiquitination by UBCH5C, which was observed to catalyze formation of polyubiquitin chains containing both K48 and K63 linkages. Activation of the IKK complex was more efficient with UBCH5C-generated polyubiquitin chains than chains made by UBC13, suggesting an inherent preference of the NEMO UBD for such chains. Such diversity of polyubiquitin chains and the UBDs through which they affect cell signaling suggests a possible mechanism for how ubiquitin drives distinct and specific outputs of cell signaling.
The multiplicity of E2 enzymes and polyubiquitin chains also offers an explanation for the lack of genetic studies supporting an essential role for ubiquitination in vivo. UBC13 has long been considered the critical E2 enzymes organizing NF-κB activation. Yet a UBC13-deficient mouse failed to demonstrate a clear requirement of this molecule for NF-κB activation downstream of antigen-receptors and TLRs (Yamamoto et al., 2006b). Likewise, promiscuity among sites of protein ubiquitination complicates attempts to demonstrate that a given lysine residue is essential for signaling to IKK and NF-κB using genetic models.
Given the challenges to genetic interrogation of E2 enzymes and ubiquitin-modified lysines, the study of ubiquitin-regulated proteins is imperative. Xia et al demonstrate that the UBD domain of TAB2 is critical for TAK1 trans-phosphorylation and activation. Yet although TAK1 is an established regulator of IKK activation, the genetic evidence for a role for TAB2 remains unclear: TAB2-deficient mice fail to manifest defects in NF-κB activation (Shim et al., 2005). Although compensation by its homolog TAB3 might mask a requirement for TAB2, TAB2/3 double knockout animals have yet to be reported, and hence definitive evidence for the involvement of TAB proteins in NF-κB activation is not yet available. Similar to TAB2/TAK1, Xia et al. found the UBD of NEMO to be required for IKK activation by ubiquitin. Yet although work in cell lines has suggested the NEMO UBD to be important for IKK activation, an in vivo requirement for its UBD has yet to be demonstrated. And although in vitro experiments in Xia et al. suggest that IKK may be activated independently of TAK1, further analysis of the UBD of NEMO and IKK auto-phosphorylation is required to understand how polyubiquitin might promote IKK activation in this manner.
More extensive structural characterization of ubiquitin-regulated protein complexes will complement genetic studies of ubiquitin signaling. It would be interesting to know whether the TAB2 UBD can distinguish polyubiquitin presented in different peptide contexts to selectively promote TAK1 activation. Structural studies coupled with mutational analyses of protein UBDs will clarify how polyubiquitin chains can specifically activate protein kinases during cell signaling. Finally, multimerization of ubiquitin-modified signaling molecules including Bcl10, TRAF6 and NEMO has been shown to modulate signaling to NF-κB (Hayden and Ghosh, 2008). Further characterization of polyubiquitin-modified proteins within multimeric complexes, and characterization of both the kinetics of their formation and their localization within the cell will help construct a more complete picture of polyubiquitin-mediated signaling in intact cells in vivo.
Canonical signaling to NF-κB through IKK occurs downstream of various cell surface receptors and within different cellular contexts. It remains to be determined whether direct activation of IKK by unanchored polyubiquitin chains operates in any or all of them. Polyubiquitin-mediated protein activation may represent an intrinsic mechanism of IKK activation, or it may serve a modulatory function to tweak signaling outputs under certain cellular conditions. The findings of Xia et al., and the novel mechanism of signal transduction they represent, offer a new and intriguing option to better understand cell signaling and activation during immunity and inflammation.