Oncogene (2006) 25, 6717–6730 & 2006 Nature Publishing Group All rights reserved 0950-9232/06 $30.00 www.nature.com/onc REVIEW Post-translational modifications regulating the activity and function of the nuclear factor kappa B pathway

ND Perkins

Division of Regulation and Expression, University of Dundee, Dundee, Scotland, UK

The diverse cellular and biological functions of the nuclear activation. Nevertheless, regardless of the stimulus, all factor kappa B (NF-jB) pathway, together with the pathways leading to NF-kB activation rely on inducing catastrophic consequences of its aberrant regulation, numerous post-translational modifications of either the demand specific and highly regulated control of its IKKs, the IkBs or the NF-kB subunits themselves. Since activity. As described in this review, regulation of the induction of NF-kB involves its regulated nuclear NF-jB pathway is brought about through multiple post- localization, there are many critical cytoplasmic mod- translational modifications that control the activity of the ifications required to either inactivate the IkB inhibitory core components of NF-jB signaling: the IjB or to activate the NF-kB subunits directly. Of (IKK) complex, the IjB proteins and the NF-jB subunits possibly equal importance, the function of NF-kB themselves. These regulatory modifications, which include complexes is also controlled by many nuclear modifica- , ubiquitination, acetylation, sumoylation tions, which can regulate transcriptional activity, DNA and nitrosylation, can vary, depending on the nature of the binding and target gene specificity. In addition, the tight NF-jB-inducing stimulus. Moreover, they frequently have temporal control of NF-kB activity means that path- distinct, sometimes antagonistic, functional consequences ways exist to limit its time in the nucleus and post- and the same modification can have different effects translational modifications of the subunits can also depending on the context. Given the important role of contribute to the termination of the NF-kB response. NF-jB in human health and disease, understanding these This review discusses these modifications and how they pathways will not only provide valuable insights into contribute to each step of the NF-kB pathway. mechanism and function, but could also lead to new drug targets and the development of diagnostic and prognostic biomarkers for many pathological conditions. Activation of the IKK complex Oncogene (2006) 25, 6717–6730. doi:10.1038/sj.onc.1209937

Keywords: NF-kB; IKK; IkB; phosphorylation; Induction of nuclear NF-kB activity by many stimuli, ubiquitination; modification particularly those involved in the immune and inflam- matory responses as well as some genotoxic agents and cellular stresses, requires the activation of the IKK complex (Hayden and Ghosh, 2004; Perkins and Gilmore, 2006). Although many upstream signaling pathways lead to IKK activation, there are a number of Introduction common features associated with stimulation of IKK catalytic activity. The activity of nuclear factor kappa B (NF-kB) proteins is regulated by hundreds of different stimuli in all cell types, with many functional consequences (Pahl, 1999; Ubiquitination of NEMO Perkins and Gilmore, 2006). The general framework of The first step in IKK activation involves modification of the canonical and noncanonical pathways leading to the regulatory subunit of the IKK complex, NF-kB activation of NF-kB via degradation of IkB and nuclear essential modifier (NEMO, also known as IKKg), through translocation of NF-kB has been reviewed elsewhere K63-linked polyubiquitination (Burns and Martinon, (Hayden and Ghosh, 2004; Gilmore, 2006). While 2004; Chen, 2005; Krappmann and Scheidereit, common themes exist for induction of NF-kB, for 2005) (Figure 1; Table 1). In contrast to K48-linked example, many inducers of NF-kB work through ubiquitination, K63 linkages do not lead to proteasomal induction of the IkB kinase (IKK) complex, there are degradation, but in this case function as a binding site also numerous exceptions and alternative pathways of for other signaling molecules required for activation of the IKK catalytic subunit IKKb (Burns and Martinon, 2004; Chen, 2005; Krappmann and Scheidereit, 2005). Correspondence: Dr ND Perkins, Division of Gene Regulation and Expression, School of Life Sciences, MSI/WTB Complex, University There are multiple NEMO ubiquitin E3 ligases, which of Dundee, Dow Street, Dundee DD1 5EH, Scotland, UK. vary depending upon the activating pathway (Burns E-mail: [email protected] and Martinon, 2004; Chen, 2005; Krappmann and Post-translational modifications of the NF-jB pathway ND Perkins 6718

Figure 1 Post-translational modifications and structure of the IKK subunits. Shown are schematic diagrams of the principal structural motifs of NEMO, IKKa and IKKb in relation to the known post-translational modifications that regulate their activity. It is possible that IKKa may undergo similar modifications to those seen in IKKb, but these have not yet been described in the literature. at the C-terminus of IKKb are indicated as being approximately 10, since the exact number were not definitively established by the original mutagenesis studies (Delhase et al., 1999). Numbering is from the human proteins, although some definitions of where a domain begins and ends might differ between publications. Abbreviations: CC1 and CC2, coiled coil regions 1 and 2; LZ, leucine zipper motif; ZF, zinc-finger domain; HLH, helix-loop-helix domain; NBD, NEMO-binding domain. The code for the different modifications is shown in the figure.

Table 1 Post-translational modifications of the IKK subunits Site(s) Modification Enzyme(s) Effect Reference

NEMO (IKKg) S31, S43 and S376 Phosphorylation IKKb Unclear Prajapati and Gaynor (2002), Carter et al. (2003) S85 Phosphorylation ATM Activation Wu et al. (2006b) K277 and K309 Sumoylation and Not known Activation Huang et al. (2003a) ubiquitination (mono) K285 Ubiquitination NOD2/RIP2 Activation Abbott et al. (2004) (K63 linked) K399 Ubiquitination Bcl10/Malt/TRAF6 Activation Sun et al. (2004), Zhou et al. (2004) (K63 linked) aa397-417 Ubiquitination Not known Activation Tang et al. (2003) (K6 linked)

IKKb S177 and S181 Phosphorylation TAK1 and others Activation Wang et al. (2001) Y188 and Y199 Phosphorylation c-Src dependent Activation Huang et al. (2003b) Approx. 10 S (aa Phosphorylation IKKb Inhibition Delhase et al. (1999) 664–756) K173 Ubiquitination Not known Activation Carter et al. (2005) (mono) C179 S-Nitrosylation N/A Inhibition Reynaert et al. (2004)

IKKa T23 Phosphorylation Akt/PKB Activation Ozes et al. (1999) S176 and S180 Phosphorylation NIK Activation Ling et al. (1998), Senftleben et al. (2001)

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6719 Scheidereit, 2005). In turn, these different E3 ligases can has also been reported to lead to mono-ubiquitination cause ubiquitination of different lysine residues in of IKKb at Lys-163 (Carter et al., 2005). of NEMO. For example, T-cell receptor (TCR) activation Lys-163 impairs subsequent IKKb auto-phosphoryla- results in Bcl10/MALT1-mediated, TNF receptor-asso- tion of further C-terminal residues in cells ciated factor 6 (TRAF6)-dependent, K63 ubiquitination (Delhase et al., 1999; Schomer Miller et al., 2006), but of NEMO at Lys-399 (Sun et al., 2004; Zhou et al., 2004). not in vitro (Carter et al., 2005). TRAF6, as well as other family members such as TRAF2 An alternative pathway leading to IKKb activation and 5, also mediate NEMO polyubiquitination in loop phosphorylation has also been described: this response to tumor necrosis factor a (TNF) and inter- pathway involves the c-Src-dependent phosphorylation leukin-1 (IL-1) receptor signaling. By contrast, bacterial of Tyr-188 and Tyr-199 of IKKb in response to infection can result in nucleotide binding oligomerization TNF and 12-O-tetradecanoylphorbol-13-acetate (TPA) domain 2 (NOD2, also known as CARD15) and receptor stimulation (Huang et al., 2003b). Given the diversity of interacting 2 (RIP2)-dependent K63-linked poly- NF-kB signaling pathways, it is possible that other ubiquitination of NEMO at lysine 285 (Abbott et al., phosphorylation events also lead to IKKb activation. 2004). TNF signaling has also been reported to result in IKKb has also been reported to be a target of the inhibitor of apoptosis protein 1 (c-IAP1)-dependent, constitutive S-nitrosylation at Cys-179, a site within the K6-linked ubiquitination of NEMO within its zinc-finger activation loop (Reynaert et al., 2004). This modifica- region (Tang et al., 2003). tion inhibits IKKb activity, and TNF stimulation results in denitrosylation at Cys-179 (Reynaert et al., 2004). Whether this denitrosylation of Cys-179 is directly IKK subunit modification linked to phosphorylation is not known. NEMO ubiquitination is just one component in a Interestingly, this same cysteine residue is also required complex series of ubiquitin-based modifications of the for some natural product-mediated inhibition of IKKb signaling molecules involved in IKK complex activation (Liang et al., 2006b). (Burns and Martinon, 2004; Chen, 2005; Krappmann A further regulatory event post-IKKb activation has and Scheidereit, 2005). These include K63-linked auto- also been reported. This involves the IKKb-mediated ubiquitination of TRAF2 and TRAF6, as well as phosphorylation of NEMO at Ser-31, Ser-43 and K63-linked polyubiquitination of the RIP1 kinase. Ser-376, and has been observed following TNF stimulation Interestingly, it has recently been shown that NEMO and expression of the Tax viral oncoprotein (Prajapati itself is a K63-linked polyubiquitin-binding protein and can and Gaynor, 2002; Carter et al., 2003). Further NEMO bind ubiquitinated RIP1 (Ea et al., 2006; Wu et al., 2006a). phosphorylations at residues in the center of the protein Therefore, this cascade of modifications may also serve have also been reported but were only observed upon to oligomerize and activate members of the IKK deletion of the C terminus of NEMO (Prajapati and signaling pathway. Polyubiquitination of these signaling Gaynor, 2002). The possible effect of these phospho- proteins also allows targeting of IKK activating rylations on NEMO activity is unclear. They could to the complex. The best characterized of these kinases potentiate IKK activation or alternatively might serve to is TGFb-activated kinase (TAK1), which has been downregulate IKK activity, possibly by targeting reported to be involved in TNF-, IL-1-, Toll-like receptor NEMO to deubiquitinating enzymes such as the tumor (TLR)- and TCR-mediated activation of NF-kB (Chen, suppressor CYLD or A20 (Chen, 2005; Krappmann and 2005; Krappmann and Scheidereit, 2005; Chen et al., Scheidereit, 2005) to NEMO, TRAF6 and RIP. Such 2006). TAK1 is associated with the TAK binding deubiquitinating enzymes have been shown to have an proteins, TAB2 and TAB3, which are themselves K63- important inhibitory effect on NF-kB signaling and to linked polyubiquitin-binding proteins, and therefore be involved in the termination of the NF-kB response TAB2 and TAB3 can be recruited to the IKK complex (Chen, 2005; Krappmann and Scheidereit, 2005). by TRAF2/6, RIP1 and potentially NEMO itself (Chen, 2005; Krappmann and Scheidereit, 2005; Chen et al., 2006). This allows TAK1-mediated phosphorylation of Activation of IKK in response to genotoxic stress IKKb within its activation loop at Ser-177 and Ser-181, An alternative, NEMO- and IKKb-dependent pathway, resulting in an active IKK complex, which is now which bears some similarities but also key differences to competent for IkB phosphorylation and NF-kB path- the classical signaling pathway described above, is way induction (Wang et al., 2001) (Figure 1; Table 1). activated in response to cellular exposure to DNA- Other putative IKKb-activating kinases have been damaging agents. While activation of the classical NF-kB described, such as PDK1, SGK and MEKK3 (Yang pathway is typically rapid, occurring within minutes of et al., 2001; Huang et al., 2004; Lee et al., 2005; Tanaka cellular exposure to a stimulus, genotoxic agents induce et al., 2005; Zhang et al., 2005). However, the NF-kB with delayed kinetics (Huang et al., 2003a; Wu mechanisms by which these other kinases might be et al., 2006b). At least in part, this appears to result from targeted to the IKK complex are not well characterized. an alternative pathway of NEMO modification and In addition, IKKb autophosphorylation, catalysed by regulation. Here, in response to stress-activated pathways oligomerization of the IKK complex, might play a role induced as a consequence of exposure to DNA-damaging in IKK activation under some circumstances. IKKb agents (but not necessarily DNA damage per se), activation loop phosphorylation at Ser-177 and Ser-181 NEMO translocates to the nucleus where it becomes

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6720 sumoylated on Lys-277 and Lys-309 (Huang et al., proteasome (Chen, 2005; Krappmann and Scheidereit, 2003a; Wu et al., 2006b) (Figure 1; Table 1). This results 2005). Ser-32/36 phosphorylation of IkBa results in in phosphorylation of NEMO on Ser-85 by the ATM binding of the SCF-bTrCP E3 ligase and a member of (Ataxia Telangiectasia Mutated) checkpoint kinase (Wu the UBC4/5 E2 ligase family, leading to ubiquitination et al., 2006b). This in turn results in the replacement of of IkBa at lysines 21 and 22 (Chen, 2005; Krappmann NEMO Sumo modification by mono-ubiquitination and Scheidereit, 2005). Degradation of ubiquitinated (Huang et al., 2003a). This mono-ubiquitination leads IkBa occurs rapidly in cells, whereas degradation of to the nuclear export of NEMO as a complex with IkBb and e is delayed, or sometimes does not occur at all ATM, which is subsequently required for the further (Thompson et al., 1995; Weil et al., 1997; Whiteside activation of the IKK complex in a manner that requires et al., 1997; Fischer et al., 1999; Wu and Ghosh, 2003). the IKK-associated protein ELKS (a protein rich in At least in part, this difference in the kinetics of glutamate (E), leucine (L), lysine (K) and serine (S)) (Wu degradation of the various IkBs is the result of the et al., 2006b). Whether this pathway is activated in more efficient phosphorylation of the IkBa serine response to all genotoxic stresses in all cell types is residues by IKKb (Wu and Ghosh, 2003). currently unclear. IkBa can also be inactivated by kinases other than IKKb in the classical activation pathway. For example, Activation of the non-canonical NF-kB pathway a CK2 (casein kinase II)-dependent pathway has been In addition to pathways that culminate in NEMO- described that is activated in response to short dependent activation of IKKb and the phosphorylation wavelength ultraviolet (UV-C) light treatment or and degradation of IkB proteins, there exists a parallel expression of the Her2/Neu oncogene (Romieu-Mourez et al et al signaling cascade, the ‘non-canonical’ NF-kB pathway, ., 2002; Kato ., 2003). Here, phosphorylation of that results in IKKa activation and the processing of the IkBa is not targeted to Ser-32 and Ser-36, which means a p100 NF-kB subunit to p52 (Bonizzi and Karin, 2004). bTrCP-independent mechanism of IkBa protein inacti- Only certain cellular stimuli, such as stimulation of the vation or degradation is required. Indeed, with Her2/ CD40 and lymphotoxin-b receptors, BAFF, LPS, and Neu activation, IkBa degradation is a consequence of the latent membrane protein-1 of Epstein–Barr virus, the action of calpain, rather than the proteasome et al induce NF-kB activity through the non-canonical path- (Romieu-Mourez ., 2002). CK2 phosphorylation in vitro way (Bonizzi and Karin, 2004). These stimuli appear to of IkBa has been mapped to Ser-283, Ser-289 converge upon NF-kB-inducing kinase (NIK) (Xiao and Ser-293 and Thr-291 and Thr-299, all of which lie et al., 2001), which in turn is required for phosphory- within the IkBa PEST domain, a motif associated with et al et al lation of IKKa at Ser-176 and Ser-180 within its protein degradation (Lin ., 1996; McElhinny ., et al et al activation loop (Ling et al., 1998; Senftleben et al., 1996; Schwarz ., 1996; Kato ., 2003) (Figure 2; 2001) (Figure 1; Table 1). In this pathway, IKKa is Table 2). It is likely that CK2-mediated activation of thought to exist as a homodimer, independent of NF-kB is cell type or context specific. Furthermore, NEMO and IKKb (Bonizzi and Karin, 2004). It will whether other kinases can similarly phosphorylate the be interesting to see whether, like the classical pathway, PEST domain and whether this is a common mechanism there is a role for K63-linked polyubiquitination of of IkBa regulation is currently unknown. CK2 also NIK, IKKa or other proteins in the non-canonical phosphorylates the IkBb PEST domain at Ser-313 and pathway. Activation of the phosphatidylinositol-3-OH Ser-315, resulting in the loss of IkBb mediated inhibition et al et al kinase (PI(3)K) pathway has, in some cell types, been of NF-kB complexes (Chu ., 1996; McKinsey ., shown to lead to Akt/Protein Kinase B phosphorylation 1997). of IKKa at Thr-23 and subsequent activation of the IkBa can also be inactivated by kinases in non-canonical pathway (Gustin et al., 2004, 2006; Ozes response to a wide range of stimuli, such as hypoxia/ et al., 1999). reoxygenation, hydrogen peroxide stimulation, treat- ment with nerve growth factor and the tyrosine inhibitor pervanadate (Koong et al., 1994; Imbert et al., 1996; Singh et al., 1996; Mukhopadhyay Modification and degradation of IjB proteins et al., 2000; Schoonbroodt et al., 2000; Bui et al., 2001). In bone marrow macrophages, TNF- induced activation The principal physiological function of activated IKKb of NF-kB has also been reported to occur through this is to phosphorylate one of the three major cytoplasmic IkBa pathway (Abu-Amer inhibitors of NF-kB: IkBa, b or e (Mercurio et al., 1997; et al., 1998). In these situations, stimuli lead to Whiteside et al., 1997; Woronicz et al., 1997; Zandi phosphorylation of IkBa at Tyr-42 and putative Tyr-42 et al., 1997; Wu and Ghosh, 2003; Hayden and Ghosh, kinases include p56-lck, Syk and c-Src (Abu-Amer et al., 2004). All of these IkB proteins contain a conserved 1998; Fan et al., 2003; Mahabeleshwar and Kundu, 2003; motif near their N termini consisting of two serine Takada et al., 2003). In some instances, Tyr-42 residues that are sites of phosphorylation by IKKb and phosphorylation results in IkBa being degraded, whereas one or two inducibly ubiquitinated lysine residues (Amir in other cases, the result is dissociation of IkBa from the et al., 2004; Hayden and Ghosh, 2004) (Figure 2; NF-kB complex (without degradation). Table 2). In contrast to NEMO, IkB ubiquitination is IkBa can also be rendered refractory to degradation K48-linked, which results in proteolytic degradation by the by the classical pathway by SUMO-1 (small ubiquitin-like

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6721 K21 K22 Ub Ub

S32 S36 Y42 S283S289T291 S293T299 S Ub Ub P P P P P P P P

A A A A A A IκBα 1 N N N N N N PEST 317 K K K K K K K6 73 256 264 Ub

S19 S23 P P Ub S313 S315 PP

A A A A A A IκBβ 1 N N N N N N PEST 361 K K K K K K K6 57 298 307 Ub

S18 S22 Ub P P

A A A A A A A IκBε 1 N N N N N N N 500 K K K K K K K 258 473

S394 S398 PP A A A A A A A Bcl-3 1 N N N N N N N 446 K K K K K K K 126 351

Ub K48 linked ubiquitin S Sumoylation P Phosphorylation

Figure 2 Post-translational modifications and structure of the IkB subunits. Shown are schematic diagrams of the principal structural motifs of IkBa,IkBb,IkBe and Bcl-3 in relation to the known post-translational modifications that regulate their activity. Amino acid numbering is according to the human proteins, although some definitions of where a domain begins and ends might differ between publications. Abbreviations: ANK, Ankyrin repeats; PEST, domain rich in proline (P), glutamate (E), serine (S) and (T). The code for the different modifications is shown in the figure.

modifier 1) modification at Lys-21, which has the Cytoplasmic activation and processing of NF-jB subunits effect of inhibiting ubiquitination at the same residue (Desterro et al., 1998). Sumoylation is inhibited by In addition to IkB inactivation, other important Ser-32/Ser-36 phosphorylation of IkBa and so can modification and processing events, involving the be thought of as an antagonistic pathway to NF-kB NF-kB subunits themselves, occur in the cytoplasm of activation. However, pools of sumoylated proteins are the cell. Most notably, there is the phosphorylation, generally found at low steady-state levels within the cell, ubiquitination and degradation of p100 and p105, the suggesting this simple prevention of IkBa ubiquitination precursor proteins for the p52 and p50 NF-kB subunits. is unlikely to be an efficient means of antagonizing Both p100 and p105 contain IkB-like ankyrin repeats in NF-kB induction. Moreover, protein sumoylation fre- their C-terminal halves, and prior to processing seque- quently requires nuclear localization (Rodriguez et al., ster their NF-kB subunit dimerization partners in the 2001). It is possible, therefore, that sumoylation may be cytoplasm (Hayden and Ghosh, 2004). Despite their involved in the shuttling of IkBa to the nucleus, where it obvious structural similarities, p100 and p105 are binds to NF-kB complexes, inactivating them and regulated by different pathways with very different relocating them back to the cytoplasm. functional consequences.

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6722 Table 2 Post-translational modifications of the IkB proteins Site(s) Modification Enzyme(s) Effect Reference

IkBa S32 and S36 Phosphorylation IKKb Ubiquitination Reviewed by Hayden and Ghosh (2004) K21 and K22 Ubiquitination bTrCP Degradation Reviewed by Hayden and (K48 linked) Ghosh (2004) S283, S289, T291, S293 Phosphorylation CK2 Degradation Lin et al. (1996); and T299 McElhinny et al. (1996); Schwarz et al. (1996) Y42 Phosphorylation p56-lck, Syk and Degradation or Koong et al. (1994) c-Src dissociation from NF-kB K21 Sumoylation Not known Stabilization Desterro et al. (1998)

IkBb S19 and 23 Phosphorylation IKKb Ubiquitination Reviewed by Hayden and Ghosh (2004) K6 Ubiquitination bTrCP Degradation Reviewed by Hayden and (K48 linked) Ghosh (2004) S313 and S315 Phosphorylation CK2 NF-kB inhibition Chu et al. (1996)

IkBe S18 and S22 Phosphorylation IKKb Ubiquitination Reviewed by Hayden and Ghosh (2004) K6 Ubiquitination bTrCP Degradation Reviewed by Hayden and (K48 linked) Ghosh (2004)

Bcl-3 S394 and S398 Phosphorylation GSK3b Degradation Viatour et al. (2004)

Processing of p100 they are located within the p52 Rel Homology Domain The non-canonical NF-kB pathway involves the activa- (RHD), they may also play a role regulating p52 DNA tion of NIK and IKKa with the subsequent phospho- binding or dimerization. rylation of p100 within its C-terminal domain at Ser-866 and Ser-870 (Senftleben et al., 2001; Xiao et al., 2001; Processing of p105 Bonizzi and Karin, 2004) (Figure 3; Table 3). These By contrast to p52/p100, the majority of p50 appears to phosphorylations create an SCF-bTrCP binding site in be generated co-translationally in a constitutive manner p100, homologous to those found in IkBs (Amir et al., (Lin et al., 1998). Inducible processing of p105 can 2004; Liang et al., 2006a), and this triggers K48-linked occur, although this frequently results in the complete ubiquitination of p100, for which Lys-856 is the degradation of p105 without p50 generation (Cohen anchoring residue (Amir et al., 2004). Recently, it has et al., 2004; Hayden and Ghosh, 2004). Unlike p100, been shown that phosphorylation of Ser-872 by IKKa IKKb is the p105 kinase and has been shown to might also represent a critical step required for p100 phosphorylate Ser-927 and Ser-932, which form the processing and may precede phosphorylation at Ser-866 SCF-bTrCP binding site (Orian et al., 2000; Lang et al., and Ser-870 (Xiao et al., 2004). Moreover, after being 2003) (Figure 3; Table 3). Phosphorylation of p105 at recruited to p100 in a Ser-866- and Ser-870-dependent these induces ubiquitination at multiple lysine manner, IKKa has been shown to also phosphorylate residues and results in the complete degradation of p105 p100 at Ser-99, Ser-108, Ser-115 and Ser-123 (Xiao (Cohen et al., 2004). IKKb can also induce SCF-bTrCP- et al., 2004). These sites of phosphorylation are also independent processing of p105, which, while distinct required for efficient processing of p100 to p52, but since from co-translational processing, can also lead to p50

Figure 3 Post-translational modifications and structure of the NF-kB subunits. Shown are schematic diagrams of the principal structural motifs of Rel(p65), RelB, c-Rel, p105 (p50) and p100 (p52) in relation to the known post-translational modifications that regulate their activity. All NF-kB subunits may undergo similar modifications to highly conserved amino acids in the RHD. For example, these might include the equivalents of p105 Cys-62 and RelA Ser-276 and Ser-281. Analogous residues are present in all NF-kB subunits except for p52, which lacks the Ser-276 equivalent. p105 Ser-337 and c-Rel 267 are the equivalents of RelA Ser-276 while RelB Ser-368 is the equivalent of RelA Ser-281. RelA Cys-38 and c-Rel Cys 26 are the equivalents of p105 Cys-62. Not shown is RelA nitration at Tyr-66 and Tyr-152. Amino acid numbering corresponds to the human proteins, although some definitions of where a domain begins and ends might differ between publications. In particular, p100 numbering for C-terminal phosphorylations differs between reports as a result of sequencing errors in the original p52/p100 cDNA cloning papers. The numbering adopted here is that used in Liang et al. (2006a). Abbreviations: RHD, Rel Homology Domain; ANK, Ankyrin repeats; DD, region with homology to a Death Domain; TA1 and TA2, subdomains of the RelA transactivation domain (TAD) (in TA2 the approximate position of conserved regions is indicated); LZ, RelB transactivation domain containing a putative leucine zipper-like motif; TAD, transactivation domain; TA I and TA II, conserved sub-domains of the c-RelA TAD. The code for the different modifications is shown in the figure.

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6723

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6724 Table 3 Post-translational modifications of the p100 and p105 NF-kB subunits Site(s) Modification Enzyme(s) Effect Reference

p100/p52 S99, S108, S115, S123, Phosphorylation IKKa Ubiquitination Senftleben et al. (2001); S866, S870 and S872 Xiao et al. (2001) K856 Ubiquitination bTrCP Degradation Amir et al. (2004) Unknown Acetylation P300 Increased processing or Hu and Colburn (2005); enhanced DNA-binding Deng et al. (2006)

p105/p50 C62 S-nitrosylation, N/A Inhibition of DNA-binding Matthews et al. (1996); oxidation Na and Surh (2006) S337 Phosphorylation PKAc Enhanced DNA-binding Hou et al. (2003) K431, K440 and K441 Acetylation P300 Enhanced DNA-binding Furia et al. (2002); Deng et al. (2003) S927 and 932 Phosphorylation IKKb Ubiquitination Lang et al. (2003) S903 and 907 Phosphorylation GSK3b Stabilizes Demarchi et al. (2003) Multiple lysines Ubiquitination bTrCP Degradation Cohen et al. (2004)

formation (Cohen et al., 2004). Although dependent Table 4). p50 and c-Rel are also modified by PKAc at upon the ubiquitin system, this SCF-bTrCP-indepen- Ser-337 and Ser-267, respectively, the equivalent sites to dent processing of p105 does not involve the ubiquitina- RelA Ser-276 (Mosialos et al., 1991; Hou et al., 2003; tion of p105 seen with complete degradation (Cohen Yu et al., 2004; Guan et al., 2005) (Figure 3; Table 3). et al., 2004). In resting cells, p105 is also phosphorylated Since this serine is also conserved in RelB (but not p52), at Ser-903 and Ser-907 by GSK3b. This stabilizes p105 it is probable that similar modification occurs with this but is required for subsequent inducible p105 processing subunit. Ser-276 in RelA is also phosphorylated by the in response to TNF stimulation (Demarchi et al., 2003). MSK1 kinase, although this is a nuclear event (see Even though the inducible processing of p105 does below) (Vermeulen et al., 2003). Phosphorylation at not typically lead to the generation of p50, it does have Ser-276 (and the analogous site in c-Rel) is thought to an important regulatory role. The Tpl2 (Cot) kinase, disrupt an intramolecular interaction between the RelA which regulates MAP kinase pathway signaling, is found N and C termini, allowing DNA binding and p300/CBP bound to a small portion of the cellular pool of p105 coactivator interaction (Mosialos et al., 1991; Zhong (Beinke et al., 2003; Waterfield et al., 2004). In this state, et al., 1998; Zhong et al., 2002; Yu et al., 2004). Tpl2 is catalytically active and phosphorylates p105 at However, the analysis of relaÀ/À MEFs reconstituted an unknown residue, but Tpl2 is unable to activate the with S276A mutant versions of RelA indicates that the MAP kinase pathway until released after p105 degrada- effect of phosphorylation of this site can be very tion (Babu et al., 2006). This pathway provides an promoter specific (Anrather et al., 2005), suggesting interesting mechanism through which the IKK pathway other effects might compensate for lack of phosphoryla- can crosstalk to the MAP kinase signaling pathway. tion at this site (see below) or that Ser-276 phospho- However, only a small proportion of cellular p105 is rylation has a subtler effect on RelA function in cells. bound to Tpl2, suggesting other consequences of p105 For example, Ser-276 phosphorylation may also affect degradation, such as the release of dimerized NF-kB subunit dimerization preference (Gapuzan et al., 2003). subunits. RelA is also inducibly phosphorylated in response to TNF stimulation at Thr-254, by an unknown kinase (Ryo et al., 2003). This creates a phosphorylated pThr- Cytoplasmic modification of RelA Pro motif that allows binding by the peptidyl-prolyl In addition to phosphorylation/ubiquitination-induced isomerase Pin-1. The action of Pin-1 disrupts RelA’s p100 and p105 processing, the RelA NF-kB subunit also interaction with IkBa and induces NF-kB localization to undergoes cytoplasmic modification prior to nuclear the nucleus (Ryo et al., 2003). This Pin-1-dependent translocation. These modifications can also stimulate mechanism also protects RelA from SOCS-1-mediated NF-kB translocation to the nucleus, either in their own ubiquitination between residues 220–335 and subse- right or as separate activation pathways. In addition, quent (Ryo et al., 2003). As Thr-254 is within phosphorylation of RelA in the cytoplasm may act to the RHD, its phosphorylation and subsequent proline program its transcriptional functions in the nucleus. isomerization at Pro-255 could be expected to also affect In response to LPS, and probably other inducers, RelA DNA-binding or transcriptional activity. RelA is phosphorylated at the highly conserved Ser-276 Phosphorylation of RelA at Ser-536, a site within the residue by the catalytic subunit of protein kinase A transactivation domain, has also been described. This (PKAc), which is in a complex with IkBa, suggesting site is the target of many kinases, with multiple coordinate IkB phosphorylation/degradation and RelA functional effects (see also below). Interestingly, one of modification (Zhong et al., 1998, 2002) (Figure 3; the Ser-536 kinases is IKKb (Sakurai et al., 1999; Haller

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6725 Table 4 Post-translational modifications of the RelA, RelB and c-Rel NF-kB subunits Site(s) Modification Enzyme(s) Effect Reference

RelA(p65) C38 Oxidation — Inhibition Liang et al. (2006b); Na and Surh (2006) Y66 and Y152 Nitration Nitric oxide Inhibition Park et al. (2005b) K122 and K123 Acetylation p300/CBP, Inhibition Kiernan et al. (2003) PCAF S205 Phosphorylation Not known Transcriptional activity Anrather et al. (2005) K218 and K221 Acetylation p300/CBP Inhibits IkBa binding, enhances Chen et al. (2002) DNA-binding (K221) T254 Phosphorylation Not known Prolyl isomerization Ryo et al. (2003) Pro255 Proline isomerization Pin-1 Stabilization and nuclear Ryo et al. (2003) localization aa220–335 Ubiquitination SOCS-1 Degradation Ryo et al. (2003) S276 Phosphorylation PKAc, MSK1 Activation/transcriptional activity Zhong et al. (1998); Vermeulen et al. (2003) S281 Phosphorylation Not known Transcriptional activity Anrather et al. (2005) K310 Acetylation p300/CBP Transcriptional activity Chen et al. (2002) S311 Phosphorylation PKCz Transcriptional activity Duran et al. (2003) T435 Phosphorylation Not known Transcriptional activity Yeh et al. (2004) S468 Phosphorylation IKKe,IKKb, Transcriptional activity Buss et al. (2004a); GSK3b Schwabe and Sakurai (2005), Mattioli et al. (2006) T505 Phosphorylation Chk1 Transcriptional activity Rocha et al. (2005); Campbell et al. (2006) S529 Phosphorylation CK2 Transcriptional activity Wang et al. (2000) S536 Phosphorylation IKKb,IKKa, Transcriptional activity, Sakurai et al. (1999); RSK1 nuclear localization, stability Sizemore et al. (2002); Buss et al. (2004b); Lawrence et al. (2005)

RelB T84 and S552 Phosphorylation Not known Degradation Marienfeld et al. (2001) S368 Phosphorylation Not known Dimerization Maier et al. (2003) c-Rel C27 Oxidation/alkylation — Inhibition Glineur et al. (2000); Na and Surh (2006) S267 Phosphorylation PKA Nuclear localization, Mosialos et al. (1991); transcriptional activity Yu et al. (2004) aa427–480 Ubiquitination Not known Degradation Chen et al. (1998) S454 and S460 Phosphorylation Not Known Transcriptional activity Starczynowski et al. (2005) S471 Phosphorylation PKCz?, NIK? Transcriptional activity Martin and Fresno (2000); Sa´ nchez-Valdepen˜ as et al. (2006) et al., 2002; Mattioli et al., 2004; Buss et al., 2004b). proposed. Following activation by TNF, RelA can Whether phosphorylation at Ser-536 by IKKb is part of dynamically oscillate between the nucleus and cyto- the activation mechanism leading to RelA nuclear plasm (Nelson et al., 2004). Moreover, nuclear RelA is translocation or a mechanism of cytoplasmically prim- rapidly dephosphorylated at Ser-536, suggesting that ing RelA’s transactivation potential is unknown. return to the cytoplasm is necessary, at least under some However, phosphorylation of Ser-536 can also result conditions, to reactivate NF-kB during over a sustained in an IkB-independent mechanism of NF-kB activation time course of activation (Nelson et al., 2004). Other (Bohuslav et al., 2004; Sasaki et al., 2005). Specifically, RelA phosphorylations may also require cytoplasmic in some cell types, induction of the tumor reprogramming. For example, Ser-468 phosphorylation suppressor results in RSK1 kinase activation (Bohuslav of RelA by IKKb has also been reported to be a et al., 2004), which in turn can phosphorylate RelA at cytoplasmic event (Schwabe and Sakurai, 2005) (see Ser-536. Since activation of RSK1 is associated with its below). cytoplasmic to nuclear translocation, it is proposed that, in this instance, Ser-536 phosphorylation occurs in the nucleus. This results in RelA nuclear accumulation Regulation of nuclear NF-jB transcriptional activity through disruption of the cytoplasmic/nuclear shuttling of NF-kB/IkB complexes that occurs in unstimulated Although many studies focus on the mechanisms and cells (Bohuslav et al., 2004). pathways that lead to NF-kB nuclear localization, it is Under other circumstances, an opposite effect of now clear that the functions of nuclear NF-kB subunits Ser-536 phosphorylation on RelA activity has been are also highly regulated. Multiple nuclear modifications

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6726 can affect DNA binding, interactions with coactivators cytoplasmic (Schwabe and Sakurai, 2005). Phosphor- and corepressors as well as the termination of the NF-kB ylation of Thr-505, within the RelA TA2 domain, response (Vermeulen et al., 2003; Chen and Greene, inhibits RelA transactivation by inducing increased 2004; Viatour et al., 2005; Perkins and Gilmore, 2006). association with HDAC1 and can result in transcrip- To date, most studies have focused on RelA nuclear tional repression of some NF-kB target , such as modification, and the analysis of nuclear modifications Bcl-xL (Rocha et al., 2005; Campbell et al., 2006). of other NF-kB subunits is extremely limited. However, Phosphorylation of RelA at Thr-505 requires the Chk1 even with RelA, we are only starting to appreciate the checkpoint kinase and is induced by the ARF tumor complexity of its nuclear regulation. It should also be suppressor or treatment of U-2 OS cells with the noted that different studies offer different levels of proof chemotherapeutic drug cisplatin (Rocha et al., 2005; that a specific kinase is modifying a particular site. In Campbell et al., 2006). Other inducers of Chk1 activity part, this results both from this being a relatively new do not appear to induce Thr-505 phosphorylation, area of study in the NF-kB field and also because in vitro suggesting that mechanisms exist to target Chk1 to kinase assays can be notoriously misleading while many RelA under only certain circumstances. Phosphoryla- small-molecule inhibitors of kinases lack specificity. tion at Thr-435 has also been implicated as a negative Furthermore, many modifications are likely to differ regulator of RelA transactivation (Yeh et al., 2004). between cell types and the nature of the NF-kB-inducing Here, the Thr-435 kinase is unknown, but dephos- stimulus. Indeed, as can be seen with RelA phospho- phorylation of this site by protein phosphatase 4 (PP4) rylation at Ser-536 (see above and below), the effect of a has been suggested to occur in response to cisplatin phosphorylation can differ depending upon its context. treatment (Yeh et al., 2004). Interestingly, all of these RelA phosphorylations occur within evolutionarily conserved subdomains of the TA2 domain, suggesting Phosphorylation of RelA that they regulate the association of RelA with Multiple RelA phosphorylation sites and candidate coactivators and corepressors. Indeed, the TA2 domain kinases have now been described (Vermeulen et al., of RelA has been demonstrated to interact with the CBP 2003; Viatour et al., 2005). Broadly speaking, many of coactivator protein (Zhong et al., 2002). these phosphorylations have a modulatory role on Ser-311, which also lies within the RelA transactiva- NF-kB transcriptional activity. For example, phosphory- tion domain, but close to the junction with the RHD, is lation of Ser-529 and Ser-536, which both lie within the inducibly phosphorylated by the atypical protein kinase TA1 subdomain of the RelA transactivation domain, C, PKCz, in response to TNF stimulation (Duran et al., can stimulate RelA transactivation (Vermeulen et al., 2003). Within the RHD, as discussed above, TNF 2003; Viatour et al., 2005) (Figure 3; Table 4). Phos- stimulation leads to Ser-276 phosphorylation in the phorylation at these sites has been detected in response nucleus by MSK1 (Vermeulen et al., 2003). Further- to many inflammatory stimuli (Vermeulen et al., 2003; more, Ser-281 is inducibly phosphorylated in response Viatour et al., 2005). Multiple Ser-536 kinases have been to LPS by an unknown kinase (Anrather et al., 2005). described in addition to IKKb and RSK1, mentioned Ser-276 and Ser-281 form a highly conserved motif that above. These include IKKa (see also below), IKKe (also is also present in all other NF-kB subunits except p52/ known as IKKi) and NF-kB activating kinase (NAK, p100, which lacks its equivalent of Ser-276 (although it also known as TBK1, TANK-binding kinase 1) (Sakurai retains the Ser-281 site) (Anrather et al., 2005). The et al., 1999; Sizemore et al., 2002; Jiang et al., 2003; Buss same study also showed Ser-205 of RelA can be et al., 2004b; Lawrence et al., 2005). By contrast, to date, phosphorylated in response to LPS. Moreover, muta- only CK2 has been described as a Ser-529 kinase (Wang tion of these sites did not totally inactivate RelA. Rather et al., 2000), although this probably represents the lower they had a modulatory role, affecting RelA’s ability to level of investigation that this site has received. It is transactivate some promoters while leaving others noteworthy that IKKb and CK2 have also been essentially unaffected (Anrather et al., 2005). This is described as inducers of NF-kB DNA binding, implying likely to be a theme of many of these modifications of coordinate regulation of IkB degradation and nuclear NF-kB subunits. That is, the effects of these modifica- transactivation potential. The TA1 domain of RelA tions will modulate target gene specificity and timing, as contains many phosphorylatable residues, suggesting opposed to being simple on–off switches for activity and that other regulatory events might occur in this region. inactivity. Indeed, differential modification of RelA is Recently, phosphorylation within the TA2 subdomain likely to underpin the differences in RelA functionality of RelA has also been described. Ser-468 is inducibly observed in response to diverse regulators such as phosphorylated in response to TNF, IL-1b and T-cell genotoxic agents, tumor suppressors and oncogenes, stimulation and has been described as a target for which are known to affect RelA phosphorylation but at GSK3b,IKKe and IKKb (Buss et al., 2004a; Schwabe unknown sights (Perkins and Gilmore, 2006). and Sakurai, 2005; Mattioli et al., 2006). Ser-468 phosphorylation has been described as both stimulating and inhibiting RelA transactivation. Interestingly, while Acetylation of RelA IKKe-mediated phosphorylation at Ser-468 was de- RelA is also inducibly acetylated at a number of sites scribed as a nuclear event (Mattioli et al., 2006), IKKb and these have different effects on its activity (Chen and phosphorylation at this site was reported to be Greene, 2004; Quivy and Van Lint, 2004) (Figure 3;

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6727 Table 4). The transcriptional coactivators p300 and CBP and Colburn, 2005), while another study indicates that can acetylate RelA at Lys-218, -221 and -310. Lys-221 p300-dependent acetylation of p52 enhances its DNA- acetylation enhances RelA DNA binding, and in binding activity (Deng et al., 2006). These differences conjunction with Lys-218 acetylation, inhibits the could reflect acetylation at different sites, but the sites of association of RelA with newly synthesized IkBa, which acetylation on p52/p100 have not yet been mapped. otherwise could act to relocate NF-kB complexes to the Bcl-3 has structural similarity to the IkBs, but Bcl-3 is cytoplasm (Chen et al., 2002). By contrast, Lys-310 primarily nuclear and functions as a transcriptional acetylation stimulates RelA transactivation (Chen et al., coactivator for p50 and p52 (Hayden and Ghosh, 2004). 2002). RelA is also acetylated at Lys-122 and Lys-123, Bcl-3 is highly phosphorylated, but little is known about which has an inhibitory effect (see below) (Kiernan the sites of these modifications or their effects on Bcl-3 et al., 2003). It is interesting to note that Lys-310 lies activity. One study has shown that Bcl-3 is constitutively adjacent to the PKCz phosphorylation site at Ser-311, phosphorylated by GSK-3b at Ser-394 and Ser-398 and implying a possible functional relationship. Indeed, a that this phosphorylation promotes the degradation of link between phosphorylation and acetylation has been Bcl-3 (Viatour et al., 2004). Inhibition of GSK-3b upon demonstrated, with Ser-276 and Ser-536 phosphoryla- cellular stimulation will therefore lead to Bcl-3 stabiliza- tion resulting in increased p300 binding and therefore tion and stimulation of Bcl-3-dependent target gene enhanced RelA acetylation at Lys-310 (Hoberg et al., expression. This is another example of a pathway 2004; Chen et al., 2005). Furthermore, Lys-310 deace- through which GSK-3b acts to suppress NF-kB activity tylation by corepressor complexes also provides a in unstimulated cells (see p105 and RelA phosphoryla- mechanism to inhibit RelA activity (Hoberg et al., tion above). Interestingly, Bcl-3 nuclear localization in 2004; Yeung et al., 2004). keratinocytes has recently been shown to be regulated by K-63-linked ubiquitination at unknown sites (Massoumi et al., 2006). Moreover, loss of the CYLD Other NF-kB subunits tumor suppressor and deubiquitinase results aberrantly Apart from those modifications leading to processing of nuclear Bcl-3 and p50/p52 dependent stimulation of Cyclin p100 and p105 (see above), the post-translational D1 expression (Massoumi et al., 2006). modification of NF-kB subunits other than RelA has received relatively little attention. c-Rel-mediated trans- activation is stimulated by phosphorylation within its Termination of the NF-jB response transactivation domain at Ser-471 by PKCz in response to TNF and possibly also by NIK (Martin and Fresno, Under normal circumstances the duration of the NF-kB 2000; Martin et al., 2001; Sa´ nchez-Valdepen˜ as et al., response is inherently self-limiting and a number of 2006) (Figure 3; Table 4). Other residues in the c-Rel negative feedback mechanisms have been described transactivation domain are also phosphorylated (Martin (Hayden and Ghosh, 2004; Perkins and Gilmore, and Fresno, 2000), and mutation of Ser-454 and Ser-460 2006). Post-translational modifications of the NF-kB has been shown to affect c-Rel function (Starczynowski subunits can also participate in termination of the et al., 2005). In addition, IKKe has also been shown to NF-kB response. phosphorylate the c-Rel transactivation domain at Phosphorylation of RelA at Ser-536 by IKKa has unknown sites and stimulate its nuclear accumulation been shown to stimulate its increased turnover in (Harris et al., 2006). Mouse c-Rel is phosphorylated in macrophages, in a process contributing to the resolution vitro at an ERK kinase consensus site in its transactiva- of inflammation (Lawrence et al., 2005). This study tion domain (Ser-451) (Fognani et al., 2000) but, although provides yet another function for the RelA Ser-536 otherwise conserved, this exact site is not present in the residue, suggesting that other factors, such as phosphor- human or chimpanzee forms of the protein (Starczynowski ylation at other sites or the cellular context, are critical et al., 2005). c-Rel, in common with RelA and p105, is determinants of the consequence of inducing this also inducibly tyrosine phosphorylated at unknown sites modification. Interestingly, ubiquitin-dependent proteo- (Neumann et al., 1992; Druker et al., 1994; Liu and lysis of promoter-bound RelA is also required for the Beller, 2002; Kang et al., 2003; Pellegatta et al., 2003). efficient termination of NF-kB-dependent transcription RelB is phosphorylated at Ser-368, the equivalent of (Saccani et al., 2004), although whether this ubiquitina- Ser-281 in RelA, and this regulates its dimerization with tion is triggered by prior phosphorylation is not known. p100 (Maier et al., 2003). Similarly, as discussed above, It is also not known whether this ubiquitination of RelA p50 is phosphorylated at the RelA Ser-276 equivalent, is at the same site as that mediated by SOCS-1 and that is, Ser-337 (Hou et al., 2003; Guan et al., 2005). regulated by Pin-1, as described above (Ryo et al., 2003). Thus, it is likely that the equivalents of RelA Ser-276 It is possible that inhibitory phosphorylations within the and Ser-281, as well as other conserved sites, can be RelA transactivation domain, such as at Thr-505, also modified in all NF-kB subunits. p50 DNA-binding have a role in shutting down the NF-kB response. IKKa activity can also be stimulated by acetylation at Lys-431, also regulates c-Rel turnover (Lawrence et al., 2005), Lys-440 and Lys-441 (Furia et al., 2002; Deng et al., and c-Rel is also subject to ubiquitination between 2003). Acetylation of p52/p100 has also been reported residues 427 and 480 that promotes c-Rel proteolysis with divergent effects. One study suggests that acetyla- (Chen et al., 1998). The sites of any modifications of tion induces increased processing of p100 to p52 (Hu c-Rel that mediate these events and whether they are

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6728 linked, are not known. RelB is also subject to phos- that enables the fine control of NF-kB function, phorylation-induced proteolytic turnover (Marienfeld which is required to regulate its diverse activities. et al., 2001). This involves phosphorylation at Thr-84/ Moreover, the catastrophic consequences of aberrant Ser-552 and is context-dependent, being seen with TCR NF-kB activation require multiple regulatory pathways or TPA/ionomycin stimulation but not with TNF to ensure only appropriate activation and correct stimulation (Marienfeld et al., 2001). termination of the NF-kB response. Yet, it is The termination of NF-kB-dependent transcription can also striking how much remains to be discovered and also potentially be mediated by acetylation of RelA at it is likely that we have only scratched the surface Lys-122 and Lys-123, which decreases its DNA-binding of the regulatory modifications of the NF-kB subunits. affinity (Kiernan et al., 2003). In common with acetylation These modifications determine the specificity of NF-kB- at lysines 218, 221 and 310 (see above), acetylation of regulated gene expression and the functionality of Lys-122andLys-123canbeperformedbyp300and nuclear NF-kB as a regulator of critical cellular path- CBP, implying a role both in activation and repression of ways such as cell growth, cytokine secretion and the NF-kB response by these transcriptional coactivators. apoptosis. There is also much still to discover about Other modifications can inhibit NF-kB. Nitric oxide the IKK subunits and how they respond to such diverse (NO)-mediated S-nitrosylation can inhibit p50 DNA- stimuli. In particular, the growing list of functions for binding activity through modification of Cys-62 IKKa implies complex regulation of its activity. Under- (Matthews et al., 1996) (Figure 3; Table 3). Indeed, standing the effects of pathways leading to post- modification of reactive cysteine residues provides a translational modifications within the NF-kB signaling mechanism through which several oxidizing agents, pathway will not only provide insights into the basic natural products and natural product derivatives can mechanisms of NF-kB regulation but, given the role of inhibit both IKK and NF-kB subunits (Straus et al., these proteins in human health and disease, could also 2000; Kwok et al., 2001; Garcia-Pineres et al., 2004; Park provide additional drug targets and biomarkers to aid in et al., 2005a; Liang et al., 2006c; Na and Surh, 2006). the diagnosis and prognosis of many pathological Interestingly, epoxyquinone A inhibits RelA DNA- conditions. binding activity through Cys-38 modification, the equivalent of p50 Cys-62 (Liang et al., 2006b). Further- more, Cys-27 oxidation and alkylation of c-Rel prevent Acknowledgements its phosphorylation and inhibit DNA-binding (Glineur In putting together in this review paper, I am aware et al., 2000). Since this redox-sensitive cysteine residue, of how much easier this has been as a result of the located in a DNA-binding loop, is conserved in all excellent and comprehensive reviews on related topics written NF-kB subunits, this might prove a common mechanism by many of my colleagues. Ihave tried to cite all of these of regulating NF-kB subunit function by oxidation, reviews in this manuscript, together with the primary alkylation and other thiol-dependent modifications (Na literature. My apologies to any reports Iwas unable to cite and Surh, 2006). NO can also induce RelA nitration at either due to lack of space or my oversight. Thanks to Tyr-66 and Tyr-152, the effect of which is dissociation Drs Cameron Bracken, Kirsteen Campbell and Sonia Rocha from p50, increased association with IkBa and seques- for their critical reading of the manuscript. Analysis and number- tration in the cytoplasm (Park et al., 2005b). ing of subunit domain structure was aided by online tools provided by the European Bioinformatics Institute (www.ebi. ac.uk) and DNADyNAmo software (www.bluetractorsoftware. co.uk). Research in the author’s laboratory is supported Conclusions by a programme grant from Cancer Research UK (C1443/ A5435) together with funding from the Association of The complexity and diversity of IKK, IkB and NF-kB International Cancer Research (AICR) and the Wellcome protein modifications is striking. It is this complexity Trust.

References

Abbott DW, Wilkins A, Asara JM, Cantley LC. (2004). Curr Bonizzi G, Karin M. (2004). Trends Immunol 25: 280–288. Biol 14: 2217–2227. Bui NT, Livolsi A, Peyron JF, Prehn JH. (2001). J Cell Biol Abu-Amer Y, Ross FP, McHugh KP, Livolsi A, Peyron JF, 152: 753–764. Teitelbaum SL. (1998). J Biol Chem 273: 29417–29423. Burns KA, Martinon F. (2004). Curr Biol 14: R1040–R1042. Amir RE, Haecker H, Karin M, Ciechanover A. (2004). Buss H, Dorrie A, Schmitz ML, Frank R, Livingstone M, Oncogene 23: 2540–2547. Resch K et al. (2004a). J Biol Chem 279: 49571–49574. Anrather J, Racchumi G, Iadecola C. (2005). J Biol Chem 280: Buss H, Dorrie A, Schmitz ML, Hoffmann E, Resch K, 244–252. Kracht M. (2004b). J Biol Chem 279: 55633–55643. Babu GR, Jin W, Norman L, Waterfield M, Chang M, Wu X Campbell KJ, Witty JM, Rocha S, Perkins ND. (2006). Cancer et al. (2006). Biochim Biophys Acta 1763: 174–181. Res 66: 929–935. Beinke S, Deka J, Lang V, Belich MP, Walker PA, Howell S Carter RS, Pennington KN, Arrate P, Oltz EM, Ballard DW. et al. (2003). Mol Cell Biol 23: 4739–4752. (2005). J Biol Chem 280: 43272–43279. Bohuslav J, Chen LF, Kwon H, Mu Y, Greene WC. (2004). Carter RS, Pennington KN, Ungurait BJ, Ballard DW. (2003). J Biol Chem 279: 26115–26125. J Biol Chem 278: 19642–19648.

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6729 Chen ZJ. (2005). Nat Cell Biol 7: 758–765. Huang WC, Chen JJ, Chen CC. (2003b). J Biol Chem 278: Chen E, Hrdlickova R, Nehyba J, Longo DL, Bose Jr HR, 9944–9952. Li C-CH. (1998). J Biol Chem 273: 35201–35207. Imbert V, Rupec RA, Livolsi A, Pahl HL, Traenckner EB, Chen LF, Greene WC. (2004). Nat Rev Mol Cell Biol 5: Mueller-Dieckmann C et al. (1996). Cell 86: 787–798. 392–401. Jiang X, Takahashi N, Matsui N, Tetsuka T, Okamoto T. Chen LF, Mu YJ, Greene WC. (2002). EMBO J 21: (2003). J Biol Chem 278: 919–926. 6539–6548. Kang JL, Lee HS, Pack IS, Hur KC, Castranova V. (2003). Chen LF, Williams SA, Mu Y, Nakano H, Duerr JM, Mol Cell Biochem 248: 17–24. Buckbinder L et al. (2005). Mol Cell Biol 25: 7966–7975. Kato T, Delhase M, Hoffmann A, Karin M. (2003). Mol Cell Chen ZJ, Bhoj V, Seth RB. (2006). Cell Death Differ 13: 12: 829–839. 687–692. Kiernan R, Bres V, Ng RWM, Coudart MP, El Messaoudi S, Chu ZL, McKinsey TA, Liu L, Qi X, Ballard DW. (1996). Mol Sardet C et al. (2003). J Biol Chem 278: 2758–2766. Cell Biol 16: 5974–5984. Koong AC, Chen EY, Giaccia AJ. (1994). Cancer Res 54: Cohen S, Achbert-Weiner H, Ciechanover A. (2004). Mol Cell 1425–1430. Biol 24: 475–486. Krappmann D, Scheidereit C. (2005). EMBO Rep 6: 321–326. Delhase M, Hayakawa M, Chen Y, Karin M. (1999). Science Kwok BH, Koh B, Ndubuisi MI, Elofsson M, Crews CM. 284: 309–313. (2001). Chem Biol 8: 759–766. Demarchi F, Bertoli C, Sandy P, Schneider C. (2003). J Biol Lang V, Janzen J, Fischer GZ, Soneji Y, Beinke S, Salmeron A Chem 278: 39583–39590. et al. (2003). Mol Cell Biol 23: 402–413. Deng WG, Tang ST, Tseng HP, Wu KK. (2006). Blood 108: Lawrence T, Bebien M, Liu GY, Nizet V, Karin M. (2005). 518–524. Nature 434: 1138–1143. Deng WG, Zhu Y, Wu KK. (2003). J Biol Chem 278: Lee KY, D’Acquisto F, Hayden MS, Shim JH, Ghosh S. 4770–4777. (2005). Science 308: 114–118. Desterro JMP, Rodriguez MS, Hay RT. (1998). Mol Cell 2: Liang C, Zhang M, Sun SC. (2006a). Cell Signal 18: 233–239. 1309–1317. Druker BJ, Neumann M, Okuda K, Franza Jr BR, Griffin JD. Liang MC, Bardhan S, Pace EA, Rosman D, Beutler JA, (1994). J Biol Chem 269: 5387–5390. Porco Jr JA et al. (2006b). Biochem Pharmacol 71: Duran A, Diaz-Meco MT, Moscat J. (2003). EMBO J 22: 634–645. 3910–3918. Liang MC, Bardhan S, Porco Jr JA, Gilmore TD. (2006c). Ea CK, Deng L, Xia ZP, Pineda G, Chen ZJ. (2006). Mol Cell Cancer Lett, in press. 22: 245–257. Lin L, DeMartino GN, Greene WC. (1998). Cell 92: 819–828. Fan C, Li Q, Ross D, Engelhardt JF. (2003). J Biol Chem 278: Lin R, Beauparlant P, Makris C, Meloche S, Hiscott J. (1996). 2072–2080. Mol Cell Biol 16: 1401–1409. Fischer C, Page S, Weber M, Eisele T, Neumeier D, Brand K. Ling L, Cao ZD, Goeddel DV. (1998). Proc Natl Acad Sci (1999). J Biol Chem 274: 24625–24632. USA 95: 3792–3797. Fognani C, Rondi R, Romano A, Blasi F. (2000). Oncogene Liu J, Beller D. (2002). J Immunol 169: 581–586. 19: 2224–2232. Mahabeleshwar GH, Kundu GC. (2003). J Biol Chem 278: Furia B, Deng L, Wu K, Baylor S, Kehn K, Li H et al. (2002). 52598–52612. J Biol Chem 277: 4973–4980. Maier HJ, Marienfeld R, Wirth T, Baumann B. (2003). J Biol Gapuzan ME, Pitoc GA, Gilmore TD. (2003). Biochem Chem 278: 39242–39250. Biophys Res Commun 307: 92–99. Marienfeld R, Berberich-Siebelt F, Berberich I, Denk A, Garcia-Pineres AJ, Lindenmeyer MT, Merfort I. (2004). Life Serfling E, Neumann M. (2001). Oncogene 20: 8142–8147. Sci 75: 841–856. Martin AG, Fresno M. (2000). J Biol Chem 275: 24383–24391. Gilmore TD. (2006). Oncogene 25: 6680–6684. Martin AG, San-Antonio B, Fresno M. (2001). J Biol Chem Glineur C, Davioud-Charvet E, Vandenbunder B. (2000). 276: 15840–15849. Biochem J 352(Part 2): 583–591. Massoumi R, Chmielarska K, Hennecke K, Pfeifer A, Guan H, Hou S, Ricciardi RP. (2005). J Biol Chem 280: Fa¨ ssler R. (2006). Cell 125: 665–677. 9957–9962. Matthews JR, Botting CH, Panico M, Morris HR, Hay RT. Gustin JA, Korgaonkar CK, Pincheira R, Li Q, Donner DB. (1996). Nucleic Acids Res 24: 2236–2242. (2006). J Biol Chem 281: 16473–16481. Mattioli I, Geng H, Sebald A, Hodel M, Bucher C, Kracht M Gustin JA, Ozes ON, Akca H, Pincheira R, Mayo LD, et al. (2006). J Biol Chem 281: 6175–6183. Li Q et al. (2004). J Biol Chem 279: 1615–1620. Mattioli I, Sebald A, Bucher C, Charles RP, Nakano H, Doi T Haller D, Russo MP, Sartor RB, Jobin C. (2002). J Biol Chem et al. (2004). J Immunol 172: 6336–6344. 277: 38168–38178. McElhinny JA, Trushin SA, Bren GD, Chester N, Paya CV. Harris J, Oliere S, Sharma S, Sun Q, Lin R, Hiscott J (1996). Mol Cell Biol 16: 899–906. et al. (2006). J Immunol 177: 2527–2535. McKinsey TA, Chu ZL, Ballard DW. (1997). J Biol Chem 272: Hayden MS, Ghosh S. (2004). Genes Dev 18: 2195–2224. 22377–22380. Hoberg JE, Yeung F, Mayo MW. (2004). Mol Cell 16: Mercurio F, Zhu HY, Murray BW, Shevchenko A, Bennett BL, 245–255. Li JW et al. (1997). Science 278: 860–866. Hou S, Guan H, Ricciardi RP. (2003). J Biol Chem 278: Mosialos G, Hamer P, Capobianco AJ, Laursen RA, 45994–45998. Gilmore TD. (1991). Mol Cell Biol 11: 5867–5877. Hu J, Colburn NH. (2005). Mol Cancer Res 3: 100–109. Mukhopadhyay A, Manna SK, Aggarwal BB. (2000). J Biol Huang Q, Yang J, Lin Y, Walker C, Cheng J, Liu ZG et al. Chem 275: 8549–8555. (2004). Nat Immunol 5: 98–103. Na HK, Surh YJ. (2006). Mol Carcinog 45: 368–380. Huang TT, Wuerzbrger-Davis SM, Wu ZH, Miyamoto S. Nelson DE, Ihekwaba AE, Elliott M, Johnson JR, (2003a). Cell 115: 565–576. Gibney CA, Foreman BE et al. (2004). Science 306: 704–708.

Oncogene Post-translational modifications of the NF-jB pathway ND Perkins 6730 Neumann M, Tsapos K, Scheppler JA, Ross J, Franza Jr BR. Sun L, Deng L, Ea CK, Xia ZP, Chen ZJ. (2004). Mol Cell 14: (1992). Oncogene 7: 2095–3104. 289–301. Orian A, Gonen H, Bercovich B, Fajerman I, Eytan E, Israel A Takada Y, Mukhopadhyay A, Kundu GC, Mahabeleshwar GH, et al. (2000). EMBO J 19: 2580–25891. Singh S, Aggarwal BB. (2003). J Biol Chem 278: Ozes ON, Mayo LD, Gustin JA, Pfeffer SR, Pfeffer LM, 24233–24241. Donner DB. (1999). Nature 401: 82–85. Tanaka H, Fujita N, Tsuruo T. (2005). J Biol Chem 280: Pahl HL. (1999). Oncogene 18: 6853–6866. 40965–40973. Park MH, Song HS, Kim KH, Son DJ, Lee SH, Yoon do Y Tang ED, Wang CY, Xiong Y, Guan KL. (2003). J Biol Chem et al. (2005a). Biochemistry 44: 8326–8336. 278: 37297–37305. Park SW, Huq MD, Hu X, Wei LN. (2005b). Mol Cell Thompson JE, Phillips RJ, Erdjumentbromage H, Tempst P, Proteomics 4: 300–309. Ghosh S. (1995). Cell 80: 573–582. Pellegatta F, Bertelli AA, Staels B, Duhem C, Fulgenzi A, Vermeulen L, De Wilde G, Van Damme P, Vanden Berghe W, Ferrero ME. (2003). Am J Clin Nutr 77: 1220–1228. Haegeman G. (2003). EMBO J 22: 1313–1324. Perkins ND, Gilmore TD. (2006). Cell Death Differ 13: 759–772. Viatour P, Dejardin E, Warnier M, Lair F, Claudio E, Prajapati S, Gaynor RB. (2002). J Biol Chem 277: Bureau F et al. (2004). Mol Cell 16: 35–45. 24331–24339. Viatour P, Merville MP, Bours V, Chariot A. (2005). Trends Quivy V, Van Lint C. (2004). Biochem Pharmacol 68: Biochem Sci 30: 43–52. 1221–1229. Wang C, Deng L, Hong M, Akkaraju GR, Inoue J, Chen ZJ. ReynaertNL,CklessK,KornSH,VosN,GualaAS,WoutersEF (2001). Nature 412: 346–351. et al. (2004). Proc Natl Acad Sci USA 101: 8945–8950. Wang D, Westerheide SD, Hanson JL, Baldwin Jr AS. (2000). Rocha S, Garrett MD, Campbell KJ, Schumm K, Perkins ND. J Biol Chem 275: 32592–32597. (2005). EMBO J 24: 1157–1169. Waterfield M, Jin W, Reiley W, Zhang M, Sun S-C. (2004). Rodriguez MS, Dargemont C, Hay RT. (2001). J Biol Chem Mol Cell Biol 24: 6040–6048. 276: 12654–12659. Weil R, Laurent-Winter C, Israe¨ l A. (1997). J Biol Chem 272: Romieu-Mourez R, Landesman-Bollag E, Seldin DC, 9942–9949. Sonenshein GE. (2002). Cancer Res 62: 6770–6778. Whiteside ST, Epinat J-C, Rice NR, Israe¨ l A. (1997). EMBO Ryo A, Suizu F, Yoshida Y, Perrem K, Liou YC, Wulf G et al. J 16: 1413–1426. (2003). Mol Cell 12: 1413–1426. Woronicz JD, Gao X, Cao Z, Rothe M, Goeddel DV. (1997). Saccani S, Marazzi I, Beg AA, Natoli G. (2004). J Exp Med Science 278: 866–869. 200: 107–113. Wu C, Ghosh S. (2003). J Biol Chem 278: 31980–31987. Sakurai H, Chiba H, Miyoshi H, Sugita T, Toriumi W. (1999). Wu CJ, Conze DB, Li T, Srinivasula SM, Ashwell JD. (2006a). J Biol Chem 274: 30353–30356. Nat Cell Biol 8: 398–406. Sa´ nchez-Valdepen˜ as C, Martin AG, Ramakrishnan P, Wu ZH, Shi Y, Tibbetts RS, Miyamoto S. (2006b). Science Wallach D, Fresno M. (2006). J Immunol 176: 4666–4674. 311: 1141–1146. Sasaki CY, Barberi TJ, Ghosh P, Longo DL. (2005). J Biol Xiao G, Fong A, Sun S-C. (2004). J Biol Chem 279: Chem 280: 34538–34547. 30099–30105. Schomer Miller B, Higashimoto T, Lee YK, Zandi E. (2006). Xiao GT, Harhaj EW, Sun S-C. (2001). Mol Cell 7: 401–409. J Biol Chem 281: 15268–15276. Yang J, Lin Y, Guo Z, Cheng J, Huang J, Deng L et al. (2001). Schoonbroodt S, Ferreira V, Best-Belpomme M, Boelaert JR, Nat Immunol 2: 620–624. Legrand-Poels S, Korner M et al. (2000). J Immunol 164: Yeh PY, Yeh KH, Chuang SE, Song YC, Cheng AL. (2004). 4292–4300. J Biol Chem 279: 26143–26148. Schwabe RF, Sakurai H. (2005). FASEB J 19: 1758–1760. Yeung F, Hoberg JE, Ramsey CS, Keller MD, Jones DR, Schwarz EM, Van Antwerp D, Verma IM. (1996). Mol Cell Frye RA et al. (2004). EMBO J 23: 2369–2380. Biol 16: 3554–3559. Yu SH, Chiang WC, Shih HM, Wu KJ. (2004). J Mol Med 82: Senftleben U, Cao YX, Xiao GT, Greten FR, Krahn G, 621–628. Bonizzi G et al. (2001). Science 293: 1495–1499. Zandi E, Rothwarf DM, Delhase M, Hayakawa M, Karin M. Singh S, Darnay BG, Aggarwal BB. (1996). J Biol Chem 271: (1997). Cell 91: 243–252. 31049–31054. Zhang L, Cui R, Cheng X, Du J. (2005). Cancer Res 65: Sizemore N, Lerner N, Dombrowski N, Sakurai H, Stark GR. 457–464. (2002). J Biol Chem 277: 3863–3869. Zhong HH, May MJ, Jimi E, Ghosh S. (2002). Mol Cell 9: Starczynowski DT, Reynolds JG, Gilmore TD. (2005). 625–636. Oncogene 24: 7355–7368. Zhong HH, Voll RE, Ghosh S. (1998). Mol Cell 1: 661–671. Straus DS, Pascual G, Li M, Welch JS, Ricote M, Hsiang CH Zhou H, Wertz I, O’Rourke K, Ultsch M, Seshagiri S, Eby M et al. (2000). Proc Natl Acad Sci USA 97: 4844–4849. et al. (2004). Nature 427: 167–171.

Oncogene