J Mol Med (2003) 81:549–557 DOI 10.1007/s00109-003-0469-0

REVIEW

Lin-Feng Chen · Warner C. Greene

Regulation of distinct biological activities of the NF-κB transcription factor complex by acetylation Received: 23 April 2003 / Accepted: 7 July 2003 / Published online: 15 August 2003 © Springer-Verlag 2003

LIN-FENG CHEN received his Ph.D. degree in Cellular and Molecular Biology in 1999 from Kyoto University, Japan. He is currently a research scientist in the Gladstone Institute of Virology and Immunology, University of California, San Francisco. His research interests include regulation of nuclear function of transcription factor NF-κB.

WARNER C. GREENE received his M.D. and Ph.D. from the Washington University School of Medicine. He is founding director of the Gladstone Institute of Virology and Immunology and professor of medicine and of microbiology and immunology at the University of California, San Francisco. His laboratory studies the molecular basis for HIV and HTLV-I retroviral pathogenesis and the mechanisms of the NF-κB/Rel family of eukaryotic transcription factors.

Abstract Although the proximal cytoplasmic signaling events that control the activation of the NF-κB transcription factor are understood in considerable detail, the subsequent intranuclear events that regulate the strength and duration of the NF-κB-mediated transcriptional response remain poorly defined. Recent studies have revealed that NF-κB is subject to reversible acetylation and that this posttranslational modification functions as an intranuclear molecular switch to control NF-κB action. In this review, we summarize this new and fascinating mechanism through which the pleiotropic effects of NF-κB are regulated within the cells. NF-κB is a heterodimer composed of p50 and RelA subunits. Both subunits are acetylated at multiple lysine residues with the p300/CBP acetyltransferases playing a major role in this process in vivo. Further, the acetylation of different lysines regulates different functions of NF-κB, including transcriptional activation, DNA binding affinity, IκBα assembly, and subcellular localization. Acetylated forms RelA are subject to deacetylation by histone deacetylase 3 (HDAC3). This selective action of HDAC3 promotes IκBα binding and rapid CRM1-dependent nuclear export of the deacetylated NFκB complex, which terminates the NF-κB response and replenishes the cytoplasmic pool of latent NF-κB/IκBα complexes. This readies the cell for the next NF-κB-inducing stimulus. Thus, reversible acetylation of RelA serves as an important intranuclear regulatory mechanism that further provides for dynamic control of NF-κB action. Keywords Acetylation · Deacetylation · p300 · HDAC3 · NF-κB · RelA

L.-F. Chen · W. C. Greene (✉) Gladstone Institute of Virology and Immunology, University of California, P.O. Box 419100, San Francisco, CA 94141-9100, USA e-mail: [email protected] Tel.: +1-415-8263800, Fax: +1-415-8261817 W. C. Greene Departments of Medicine and of Microbiology and Immunology, University of California, San Francisco, California, USA

Abbreviations NF-κB: Nuclear factor κB · RHD: Rel homology domain · IKK: IκB kinase complex · TSA: Trichostatin A · HDAC: Histone deacetylase · HAT: Histone acetyltransferase · TNF-α: Tumor necrosis factor α · NCoR: Nuclear receptor corepressor · SMRT: Silencing mediator for retinoid and thyroid hormone receptors · SRC-1, 3: Steroid receptor coactivator 1, 3 · NIK: NF-κB-inducing kinase · MEF: Mouse embryo fibroblast · CRM-1: Chromosomal region maintenance-1

550 Fig. 1 Schematic diagram of the activation of NF-κB by two pathways. In the canonical NF-κB activation pathway, stimulus-induced phosphorylation of two N-terminal serines in the IκBs is mediated by the IKKs. After ubiquitination and degradation of IκBα by the 26S proteasome complex, the liberated NF-κB heterodimer (p50:p65) translocates into the nucleus and activates the transcription of target genes. In the alterative pathway, IKKα is activated by different members of the TNF-α family (for example, BAFF and CD40). Along with NIK, IKKα induces the phosphorylation-dependent processing of p100 and generation of p52:RelB heterodimers. p52:RelB translocates to the nucleus and activates its target genes

Introduction The inducible NF-κB transcription factor complex plays a central role in regulating the inflammatory, immune, and anti-apoptotic responses in mammals [1]. The NF-B signaling pathway is evolutionarily conserved. In mammals, five Rel family members have been identified: RelA/p65, RelB, c-RelA, p50/p105, and p52/p100 [1, 2], which form various NF-κB homo- and heterodimers. All of these proteins contain an N-terminal Rel homology domain (RHD) consisting of approximately 300 amino acids. The N-terminal portion of the RHD mediates both backbone and sequence-specific contacts with the major groove of DNA. The C-terminal portion of this domain also contributes to backbone contacts but additionally controls dimerization with other Rel family members and interaction with the NF-κB inhibitors IκBα [2, 3]. While all of the Rel proteins bind DNA, only RelA, cRel, and RelB contain potently active C-terminal transcriptional activation domains [2]. Activation of the RelA family of transcription factors is controlled by two distinct pathways (Fig. 1). In the canonical activation pathway, the prototypical NF-κB complex, a heterodimer of p50 and RelA subunits, is chiefly sequestered in the cytoplasm through its assembly with a member of the IκB family of inhibitory proteins [1]. Stimulus-induced phosphorylation of two N-terminal serines in the IκBs is mediated by the macromolecular IκB kinase complex (IKK) termed the signalsome that contains IKKα/IKK1, IKKβ/IKK2, and the regulatory subunit IKKγ/NEMO [4]. This response in turn triggers the rapid ubiquitination and subsequent degradation of IκB by the 26S proteasome complex. The newly liberat-

ed NF-κB heterodimer then rapidly translocates into the nucleus, where it engages cognate κB enhancer elements and activates gene expression [1, 2]. The second activation pathway involves inducible proteolytic processing of NF-κB2/p100 Rel protein (Fig. 1). Different members of the TNF receptor superfamily (for example, BAFF and CD40) selectively activate the NF-κB-inducing kinase (NIK) and IKK1, leading to the phosphorylation of p100 and the processing of p100 to p52 by the proteasome. This pathway principally generates p52:RelB heterodimers, which also translocate into the nucleus and regulate various target genes [5, 6, 7, 8]. NF-κB activates a variety of cellular genes that control diverse biological functions, including the inflammatory and immune responses and apoptosis. In addition, NF-κB induces the expression of specific genes that encode negative regulators of NF-κB [9, 10], thus providing a negative feedback loop that ensures tight regulation of NF-κB action. One of the negative regulators induced by NF-κB is IκBα [11, 12, 13]. Newly synthesized IκBα proteins can shuttle in and out of the nucleus and can physically remove NF-κB from DNA. Subsequent nuclear export of IκBα promotes the return of the now inactive NF-κB–IκBα complex to the cytoplasm and serves to terminate the NF-κB transcriptional response (Fig. 1) [14, 15]. The NF-κB-mediated transcriptional response importantly involves the participation of various coactivator and corepressor proteins (Table 1). Initially, the C-terminal transcriptional activation domain of RelA was described as interacting with the amino-terminal region of p300, and further specific cyclin-dependent kinases regulate transcriptional activation by NF-κB through inter-

551 Table 1 Coactivators and corepressors that participate in the regulation of NF-κB activation Target proteins

Function

Reference

Coactivators p300/CBP PCAF SRC-1 SRC-3 Tip60

RelA, p50 RelA, p50 p50 RelA Bcl3

Enhance NF-κB activation, acetylate RelA and p50 Enhance NF-κB activation, acetylate RelA Enhance NF-κB activation Enhance NF-κB activation Bridging factor for p50 and coactivators

[16, 17, 18, 19, 29, 39, 41] [18, 40] [18, 21] [22, 23] [24, 25]

Corepressors HDAC1,2 HDAC3 SMRT/NCoR

RelA, p50 RelA RelA, p50

Inhibit NF-κB activation Deacetylate RelA, inhibit NF-κB activation Inhibit transactivation of NF-κB

[27, 28] [29, 40] [26]

actions with the coactivator p300 [16]. Later studies have revealed that both the N-terminal RHD and the Cterminal transcriptional activation domain of RelA interact with the amino-terminal CH1 domain of p300 [17, 18]. More recent studies indicate that the phosphorylation status of RelA regulates its interaction with CBP [19]. This interaction appears biologically relevant since transactivation mediated by NF-κB is consistently increased by coexpression of p300/CBP coactivators [17, 18, 20]. Coactivators other than p300/CBP may also participate as modifiers of the transcriptional activity of NFκB. For example, SRC-1/N-CoA-1, a nuclear receptor coactivator within the p160 family, can function as a coactivator with NF-κB [18, 21]. A second p160 family member, SRC-3, may similarly exert coactivating effects with NF-κB. Of note, SRC-3 is phosphorylated by the IKKs, and this posttranscriptional modification may promote nuclear import of SRC-3 [22, 23]. These findings reveal an unexpected linkage of SRC-3 to proximal signaling components of NF-κB pathway that may enable the subsequent coactivator function of SRC-3 to the NFκB. Both SRC-1 and SRC-3 increase NF-κB-mediated transactivation in a dose-dependent manner [21, 22]. Finally, another acetyltransferase, Tip60, can also enhance NF-κB activation [24, 25]. As might be expected by the participation of coactivators, NF-κB activity is also regulated by various corepressors. The nuclear SMRT and NCoR corepressors both interact with p50 and RelA and inhibit transactivation mediated by NF-κB [26]. Specific histone deacetylases (HDACs) also interact with NF-κB and regulate its biological activities [27, 28, 29]. Most notably, the Nterminal region of the RelA subunit of NF-κB selectively interacts with HDAC3, which promotes the deacetylation of RelA [29]. The association of NF-κB with the corepressor proteins HDAC1 and HDAC2 also blocks the expression of NF-κB-regulated genes, including the interleukin-8 (IL-8) gene. However, in contrast to the action of HDAC3 on RelA, HDAC1 and HDAC2 appear to exert their effects by modifying the state of histone tail acetylation within the surrounding chromatin [27, 28]. Both p300 and CBP contain a histone acetyltransferase (HAT) enzymatic activity that regulates gene expres-

sion in part by acetylating the N-terminal tails of histones [30]. Acetylation and deacetylation are reversible reactions. Acetylated histones are associated with transcriptionally active regions in the genome, while deacetylated histones accumulate in transcriptionally repressed regions [31, 32, 33]. However, histones are not the only substrate of the HAT. HATs such as p300/CBP also target various non-histone proteins for acetylation. The first such non-histone protein to be regulated by acetylation was the tumor suppressor p53 [34]. Acetylation of p53 markedly stimulates its sequence-specific DNA-binding activity [34]. Additional non-histone proteins, most notably transcription factors like GATA-1, MyoD, HNF-4, TFIIEβ, E2F, and Smad7, have been identified as biological targets of various acetyltransferases. Acetylation of these factors leads to various alterations, including changes in DNA binding, transcriptional activity, protein–protein interactions, nuclear export, and protein stability [35, 36, 37, 38]. Of note, the NF-κB transcription factor is similarly regulated by posttranscriptional acetylation, which lead to increased DNA binding, enhanced transactivation, and resistance to assembly with IκBα [29, 39, 40, 41]. The mechanism through which p300/CBP and other coactivators enhance, while HDACs and corepressors inhibit NF-κB transcriptional activity is almost certainly multifactorial, involving direct effect on this transcriptional factor and changes in chromatin structure [27, 28, 29, 41]. In this review, we shall focus on biological effects induced by directed acetylation and deacetylation of RelA subunit of the NF-κB complex.

RelA is acetylated by p300/CBP and deacetylated by HDAC3 Acetylation of RelA was first demonstrated in vivo by [3H]-acetate radiolabeling, which showed that overexpressed RelA is acetylated in vivo by endogenous HATs [29]. Additional studies revealed that endogenous RelA is also acetylated following stimulation of cells with TNF-α or PMA. These finding indicate that acetylation is detectable under physiological conditions and that this

552

modification is induced in a signal-coupled manner [29, 40]. In vivo acetylation of RelA can also be detected using various antibodies that react with acetylated lysine provided cells are cotransfected with expression vectors encoding a relevant HAT. Both p300 and CBP, but not PCAF, acetylate RelA under these conditions. Not surprisingly, mutation of the acetyltransferase domain within p300 prevents the acetylation of RelA [41]. Further, coexpression of either a HAT-deficient mutant of p300 or the adenovirus E1A gene product (an inhibitor of endogenous p300/CBP activity) [42, 43] blocks the acetylation of RelA induced by endogenous HATs. These findings argue that p300/CBP plays a key role in the regulation of RelA acetylation in vivo. Although p300 acetylates RelA in vivo, recombinant p300 alone fails to acetylate RelA in vitro despite the use of conditions where acetylation of p53 or histones is readily apparent [18, 19, 44]. However, RelA can be acetylated in vitro using anti-p300 immunoprecipitates from 293T cells as efficiently as p53, arguing that an unknown cofactor that coimmunoprecipitates with p300 is likely required for efficient acetylation of RelA in vitro (Chen et al. unpublished data, 2002). RelA is also subject to deacetylation by HDAC3 both in vivo and in vitro. Overexpression of HDAC3, but not HDAC1 or 2, deacetylates RelA and markedly impairs κB enhancer dependent gene expression [29, 40]. When RelA is coexpressed with HDAC3, the level of acetylated RelA decreases in in vivo [3H]-acetate radiolabeling assay [29, 40]. Deacetylation of RelA by HDAC3 can also be detected through diminished reactivity with antiacetylated lysine antibodies [29]. Acetylated RelA is deacetylated by HDAC3 in vitro as well [40]. These effects of HDAC3 are blocked by trichostatin A (TSA), a general inhibitor of HDAC enzymatic activity, indicating that the deacetylase function of HDAC3 is required for these effects [29, 40]. Of note, HDAC3 alone has little or no enzymatic activity, but is activated after forming a stable complex with SMRT/NCoR [45, 46, 47]. SMRT interacts with RelA and p50 directly and represses RelA-mediated transactivation [26]. Activation of NF-κB also alters the cytoplasmic shuttling of SMRT/NCoR [48, 49]. Together, these findings suggest that SMRT/NCoR plays a key role in HDAC-3-mediated deacetylation of RelA. Deletional and site-specific mutagenesis of RelA indicates that acetylation occurs at three major sites: lysines 218, 221, and 310. While preserving the overall charge, substitution of all three of these lysines with arginine (designated RelA-KR) markedly blocks acetylation, although a low-level residual signal is present in [3H]-acetate radiolabeling assays. Thus, other lysine residues within RelA may also be modified by acetylation. In this regard, lysines 122 and 123 in RelA may also be targets for acetylation by p300 and PCAF [40]; however, whether the remaining acetylation signal in the RelA-KR mutant derives from acetylation of these two lysines is not yet clear. Sequence alignment of all of the mammalian Rel proteins showed that lysines 218 and 221 are highly conserved as well as lysines 122 and 123, whereas lysine 310 is present only in RelA. Of note, the corresponding

p50 lysines to K218 and K221 in RelA are also targeted for acetylation by p300. However, the biological consequence of these modifications of p50 remains unclear (Chen et al. unpublished data, 2002).

RelA acetylation modulates the transcriptional activity of NF-κB Acetylation can clearly alter the activity of many transcription factors [30, 37]. A potential role for acetylation in the regulation of NF-κB-mediated transactivation first emerged with the finding that TSA enhances κB-luciferase reporter gene expression in the presence of TNF-α [29]. TSA also potentiates TNF-α-mediated activation of the IL-6 promoter and the HIV long terminal repeat [20, 50]. NF-κB activation is also increased by coexpression of wild-type p300/CBP. However, these effects are not observed when a HAT-deficient mutant of p300 is coexpressed [17, 18, 41]. Instead, HAT-deficient p300 diminishes TNF-α-induced NF-κB activation [41]. In addition, the RelA-KR fails to display enhanced transcriptional activity when coexpressed with p300/CBP, suggesting that p300/CBP enhances RelA-mediated transcription, at least in part, by promoting direct acetylation of RelA. Conversely, expression of HDAC3, but not of HDAC1, 2, 4, 5, and 6, inhibits TNF-α-mediated activation of NFκB. HDAC3 does not display these inhibitory effects in the presence of TSA, indicating that its deacetylase activity is required for the observed biological effects [29]. Mutagenesis studies suggest that the acetylation of lysines 221 and 310 is important in regulating the overall transcriptional activity of NF-κB [41]. In luciferase reporter assays, both the K310R and RelA-KR mutants display greatly impaired transcriptional function (>85%), whereas the K221R mutant had modestly impaired function (~50%). However, replacing lysine 310 with glutamine, which may approximate the physical changes associated with acetylation, enhances the transcriptional activity of RelA (Chen et al. unpublished data, 2002). Together, these findings suggest that acetylation at lysine 310, and to a lesser extent at lysine 221, may be required for the full transactivation function of RelA (Fig. 2). As discussed below, acetylation of lysine 221 also plays a central role in the regulation of RelA DNA binding and assembly with IκBα. Such changes in DNA binding and IκBα assembly properties of RelA-K221R likely contribute to the observed declined in transcriptional activity. In contrast, mutation of lysine 310 to arginine does not alter the DNA binding or IκBα assembly properties of the protein, but it markedly inhibits transcriptional activity, indicating that acetylated lysine 310 may form a platform for the recruitment of an as yet unidentified factor, possibly one containing a bromodomain, that is required for the full RelA transcriptional response. Bromodomain-containing proteins specifically interact with acetylated lysine residues [51, 52, 53, 54]. For example, the bromodomain of PCAF specifically interacts with acetylated lysine 8 in histone H4 and with ace-

553

Fig. 2 Acetylation of specific lysine residues regulates distinct functions of RelA. Acetylation of lysine 221 increases DNA binding affinity for the κB-enhancer and in combination with the acetylated lysine 218 prevents the association of RelA with IκB. Acetylation of lysine 310 likely controls the association of an as yet unknown factor that is required for full transactivation by RelA

tylated lysine 50 in HIV Tat [51, 53]. Similarly, the bromodomain of p300/CBP interacts with acetylated MyoD [54]. By analogy, acetylated lysine 310 may mediate the recruitment of a bromodomain-containing protein that plays a key role in the NF-κB transcriptional response. However, this bromodomain-containing factor does not appear to correspond to p300, since wild-type RelA and RelA K310R bind similar amounts of p300 in coimmunoprecipitation assays [41]. In contrast to RelA, the biological consequences of p50 acetylation are less clear. Lysines 431, 440, and 441 in p50 are targets for acetylation [39]. p50 acetylation can also be detected in resting human foreskin fibroblasts and is increased after stimulation with TNF-α or lipopolysaccharide. As with RelA, overexpression of p300 augments p50 acetylation and enhanced p50 acetylation correlates with increased p50 binding to the COX2 promoter and transcriptional activation of this gene [55]. To date, p300 has only been reported to acetylate p50 in the presence of HIV Tat protein in vitro; however, in vivo p300 can acetylate p50 in the absence of HIV Tat, suggesting that acetylation of p50 may involve the participation of bridging molecules [39, 55]. In streptavidin/biotin pull-down assays with labeled NF-κB oligonucleotides, acetylation of p50 increased DNA binding to, and increased NF-κB-mediated transcriptional effects on, the HIV long terminal repeat, suggesting that acetylation of this NF-κB subunit may also enhance its functional properties [39].

Acetylation of lysine 221 of RelA regulates the DNA binding activity of NF-κB Acetylation of transcription factors can lead to contradictory effects on the intrinsic DNA binding properties of these proteins. For example, acetylation enhances the affinity of p53 or GATA-1 for DNA [34, 56] but impairs DNA binding of CDP/cut or HMGI (Y) [44, 57]. Blocking HDAC with TSA increases TNF-α-induced binding of NF-κB to DNA, suggesting that acetylation of RelA

may positively regulate its DNA binding properties [41, 50]. Analysis of lysine-to-arginine RelA substitution mutants further supports this notion. Specially, mutation of lysine 221 of RelA induces a sharp decline in overall DNA binding activity of RelA homodimers, whereas mutation of the two other major acetylation sites in RelA, lysines 218 and 310, does not lead to a similar effect. The defective DNA binding of the lysine 221 mutant is not simply due to an inability to form homodimers, since all of these mutants dimerize as efficiently as wild-type RelA [41]. When these RelA mutants are tested in the context of p50/RelA NF-κB complexes, steadystate DNA binding activity measured in electrophoretic mobility shift assays is not markedly changed, reflecting the strong DNA binding properties of p50. However, when unlabeled κB enhancer DNA is added as a competitor after assembly of the NF-κB/κB enhancer complex, the off-rate of DNA binding to the κB enhancer is accelerated when acetylation is blocked either by the argininefor-lysine mutation at residue 221 or by coexpression of the dominant-negative p300 HAT-deficient mutant. When biotinylated forms of the κB enhancer are used to capture bound RelA, acetylated forms of RelA are also identified, indicating that assembly is with the κB enhancer DNA [41]. Together, these results indicate that the acetylation of RelA at lysine 221 enhances the overall binding affinity of the NF-κB complex for the κB enhancer (Fig. 2). Inspection of the crystal structure of the RelA-p50-DNA complex reveals that lysine 221 participates in direct contact with the DNA backbone [58]. Thus, acetylation of lysine 221 may result in a conformational change in the protein that enhances its binding to the κB enhancer [41]. A similar mechanism has been hypothesized for GATA-1, where acetylation in the DNA binding domain enhances the DNA binding activity of this transcription factor [56].

RelA acetylation regulates the assembly of NF-κB with IκBα Acetylation also regulates protein–protein interactions. For example, acetylation of adenovirus E1A at lysine 239 blocks its interaction with CtBP [59], while acetylation of MyoD enhances its interaction with p300/CBP [54]. Similarly, acetylation of RelA plays a key role in regulating its assembly with it natural inhibitor, IκBα. In GST-IκBα pull-down studies, the binding of RelA to GST-IκBα is markedly diminished by p300-mediated acetylation of RelA and is increased by HDAC3-mediated deacetylation of RelA [29]. Consistent with this finding, wild-type p300, but not the HAT-deficient p300 mutant, blocks the interaction of RelA with IκBα in a mammalian one-hybrid assay system. Acetylation of RelA by wild-type p300 prevents the cytoplasmic sequestration of RelA by IκBα. Similar effects are not observed with RelA-KR which is poorly acetylated by p300 [29]. Studies of lysine-to-arginine substitution mutants of RelA have revealed that the acetylation of lysine 221 plays a

554

key role in IκBα assembly [41]. In support of this conclusion, the addition of recombinant GST-IκBα significantly diminished DNA binding with the mutant of lysine 221. Substituting arginines for lysines 218 and 221 or for all three sites (RelA-KR) produced even greater levels of inhibition of DNA binding in the presence of GST-IκBα [41]. These findings support the notion that acetylation of RelA at lysine 221, alone or in combination with lysine 218, blocks the ability of IκBα to assembly with RelA (Fig. 2). In the crystal structure of RelA-p50-IκBα [60, 61], lysine 221 directly interacts with methionine 279 in the sixth ankyrin repeat of IκBα. Thus, acetylation of this lysine could cause a conformational changes in RelA that increases its affinity for the κB enhancer and disables its interaction with IκBα (Fig. 2). Alternatively, the more rapid off-rate of κB enhancer DNA binding displayed by RelA-K221R may derive in part from its greater affinity for IκBα.

ginine substitutions at the three major acetylation sites (GFP-RelA-KR) principally localizes in the cytoplasm. However, when expressed in the IκBα-deficient MEFs, this protein is found predominantly in the nucleus. Further, the cytoplasmic expression pattern of RelA-KR mutant appears to be dependent on the nuclear export signal present in IκBα. Expression of a mutant IκBα that lacks the nuclear export signal produces a predominantly nuclear pattern of expression for the RelA-KR in IκBα-deficient MEFs. In contrast, wild-type IκBα promotes efficient translocation of GFP-RelA-KR proteins to the cytoplasm in these cells [41]. These various findings thus provide strong support for the hypothesis that deacetylation of lysine 221 in the RelA-KR mutant enhances the interaction with IκBα and the hypoacetylated RelA exported from the nucleus in a manner that is principally dependent on the nuclear export signal of IκBα.

Conclusions and future perspectives RelA acetylation regulates nuclear export of NF-κB Acetylation has been shown to regulate the nuclear export of some transcription factors. For example, acetylation of lysine residues within the nuclear localization signals of CIIAT and HNF-4 leads to an increased nuclear accumulation of both factors [62, 63] likely reflecting impaired nuclear export. Acetylation of RelA also regulates the nuclear export of NF-κB. When overexpressed in HeLa cells, GFP-RelA fusion proteins are principally localized in the nucleus. However, when coexpressed with HDAC3, these proteins are found exclusively in the cytoplasm. This HDAC3-induced translocation of nuclear RelA into the cytoplasm is due principally to the deacetylation of RelA, which promotes IκBα assembly and IκBα-dependent nuclear export of RelA. Consistent with this model, HDAC3 fails to promote RelA translocation into the cytoplasm in IκBα-deficient MEFs [29]. In addition, a GFP-RelA fusion protein containing lysine-to-arFig. 3 Schematic model for the role of HDAC3-mediated deacetylation of RelA as an intranuclear molecular switch promoting IκBα binding and IκBα-dependent nuclear export of the NF-κB complex. This deacetylation-controlled response leads to the termination of the NF-κB transcriptional response and aids in reestablishing latent cytoplasmic forms of NF-κB bound to IκBα, thereby preparing the cell to respond to the next NF-κB-inducing stimulus

The studies described here have revealed key regulatory roles for acetylation and deacetylation in the regulation of NF-κB activity. This posttranslational modification affects many different biological properties of the RelA subunit of NF-κB, including DNA binding, transcriptional activity, assembly with the IκB inhibitors, and subcellular localization of the RelA-containing complexes. RelA is acetylated by p300/CBP upon TNF-α stimulation, and acetylated forms of RelA are deacetylated through specific interaction with HDAC3. Deacetylation of RelA promotes its effective binding to IκBα and leads to IκBα-dependent nuclear export of the NF-κB complex by a CRM-1-dependent pathway. Reversible acetylation of RelA thus functions as an intranuclear molecular switch that both controls the duration of the NF-κB transcriptional response and contributes to replenishment of the depleted cytoplasmic pool of latent NF-κB/IκBα complexes, thereby readying the cell for the NF-κB-inducing signal (Fig. 3).

555

p300/CBP plays an important role in the acetylation of NF-κB in vivo and in vitro but is probably not the only HAT that mediates acetylation of RelA. It is quite possible that other NF-κB coactivators with HAT activity, such as SRC-1 and SRC-3, also play a role, perhaps in conjunction with p300/CBP. Further investigation is needed to determine whether the coactivator function of these proteins reflects their ability to acetylate NF-κB directly or to modify chromatin structure. Overexpression of HDAC3 induces the deacetylation of RelA in vivo, and other HDACs appear to lack this activity [29]. However, how the physical interplay between RelA and HDAC3 is regulated remains unclear. For example, it is not known whether the HDAC3 activity that deacetylates RelA is also signal-coupled or alternatively is constitutively active. Future studies will help to clarify further the relationship between these intriguing reactions of acetylation and deacetylation and how they contribute to the overall regulation of NF-κB action. Additionally, while TNF-α stimulates acetylation of RelA, the stoichiometry of the acetylation reaction remains unknown. Is a small percentage of RelA acetylated, and is it this form that is most relevant to the ensuring biological effects? Alternatively, acetylation may involve most RelA proteins and current systems underestimate the extent to which this modification occurs. In addition to acetylation, NF-κB can undergo other posttranslational modifications, such as phosphorylation. RelA is phosphorylated at multiple sites by different kinases [64, 65, 66, 67]. Phosphorylation of serine 276 by the catalytic subunit of PKA in response to lipopolysaccharide promotes the efficient recruitment of CBP [19]. Phosphorylation at one or more of these sites could thus trigger the recruitment of p300/CBP, leading to enhanced RelA acetylation. Precedence for contingent coupling of phosphorylation and acetylation is provided by the p53 and retinoblastoma tumor suppressor protein [68, 69, 70]. We are just beginning to appreciate the dynamic role of reversible acetylation as a new regulatory mechanism that allows greater control of NF-κB. It should be noted that many of these findings require/involve the overexpression of proteins and the use of NF-κB mutant proteins. Thus more experiments focusing on the endogenous NF-κB protein are required. Using specific antiacetylated NF-κB antibodies and chromatin immunoprecipitation assays, it is likely that future studies will help define the significance of NF-κB acetylation in a more biologically relevant environment. Regulation of the NF-κB action by reversible acetylation could provide insights into other key questions in NF-κB biology. For example, do glucocorticoids inhibit the acetylation of RelA? If so, to what extent does this contribute to their anti-inflammatory effects? Finally, it is intriguing to consider the possibility that selective interference with RelA acetylation or enhancement of RelA deacetylation by small moleculars could lead to a new class of anti-inflammatory or immune suppression drugs.

Acknowledgements This work was supported in part by a National Institutes of Health grant (RO1 CA89001–02) to W.C.G., a National Institutes of Health training grant (T32 AI07305) to L.F.C., and by funds from the J. David Gladstone Institutes, and Pfizer,, and benefited from core facilities provided through the UCSF-GIVI Center for AIDS Research (National Institutes of Health Grant P30 MH59037). The authors thank R. Givens and S. Cammack for manuscript preparation, J. Carroll and C. Goodfellow for graphics, and G. Howard and S. Ordway for editorial assistance.

References 1. Baldwin AS Jr (1996) The NF-κB and IκB proteins: new discoveries and insights. Annu Rev Immunol 14:649–683 2. Ghosh S, May MJ, Kopp EB (1998) NF-κB and Rel proteins: evolutionarily conserved mediators of immune responses. Annu Rev Immunol 16:225–260 3. Baeuerle PA (1998) IκB-NF-κB structures: at the interface of inflammation control. Cell 95:729–731 4. Karin M (1999) How NF-κB is activated: the role of the IκB kinase (IKK) complex. Oncogene 18:6867–6874 5. Xiao G, Harhaj EW, Sun SC (2001) NF-κB-inducing kinase regulates the processing of NF-κB2 p100. Mol Cell 7:401–409 6. Senftleben U, Cao Y, Xiao G, Greten FR, Krahn G, Bonizzi G, Chen Y, Hu Y, Fong A, Sun SC, et al (2001) Activation by IKKα of a second, evolutionary conserved, NF-κB signaling pathway. Science 293:1495–1499 7. Claudio E, Brown K, Park S, Wang H, Siebenlist U (2002) BAFF-induced NEMO-independent processing of NF-κB2 in maturing B cells. Nat Immunol 3:958–965 8. Coope HJ, Atkinson PG, Huhse B, Belich M, Janzen J, Holman MJ, Klaus GG, Johnston LH, Ley SC (2002) CD40 regulates the processing of NF-κB2 p100 to p52. EMBO J 21:5375–5385 9. Karin M, Cao Y, Greten FR, Li ZW (2002) NF-κB in cancer: from innocent bystander to major culprit. Nat Rev Cancer 2:301–310 10. Senftleben U, Karin M (2002) The IKK/NF-κB pathway. Crit Care Med 30:S18–S26 11. Beg AA, Finco TS, Nantermet PV, Baldwin AS Jr (1993) Tumor necrosis factor and interleukin-1 lead to phosphorylation and loss of IκBα: a mechanism for NF-κB activation. Mol Cell Biol 13:3301–3310 12. Brown K, Park S, Kanno T, Franzoso G, Siebenlist U (1993) Mutual regulation of the transcriptional activator NF-κB and its inhibitor, IκBα. Proc Natl Acad Sci U S A 90:2532–2536 13. Sun SC, Ganchi PA, Ballard DW, Greene WC (1993) NF-κB controls expression of inhibitor IκBα: evidence for an inducible autoregulatory pathway. Science 259:1912–1915 14. Arenzana-Seisdedos F, Thompson J, Rodriguez MS, Bachelerie F, Thomas D, Hay RT (1995) Inducible nuclear expression of newly synthesized IκBα negatively regulates DNA-binding and transcriptional activities of NF-κB. Mol Cell Biol 15:2689–2696 15. Arenzana-Seisdedos F, Turpin P, Rodriguez M, Thomas D, Hay RT, Virelizier JL, Dargemont C (1997) Nuclear localization of IκBα promotes active transport of NF-κB from the nucleus to the cytoplasm. J Cell Sci 110:369–378 16. Perkins ND, Felzien LK, Betts JC, Leung K, Beach DH, Nabel GJ (1997) Regulation of NF-κB by cyclin-dependent kinases associated with the p300 coactivator. Science 275:523–527 17. Gerritsen ME, Williams AJ, Neish AS, Moore S, Shi Y, Collins T (1997) CREB-binding protein/p300 are transcriptional coactivators of p65. Proc Natl Acad Sci USA 94:2927– 2932 18. Sheppard KA, Rose DW, Haque ZK, Kurokawa R, McInerney E, Westin S, Thanos D, Rosenfeld MG, Glass CK, Collins T (1999) Transcriptional activation by NF-κB requires multiple coactivators. Mol Cell Biol 19:6367–6378

556 19. Zhong H, Voll RE, Ghosh S (1998) Phosphorylation of NF-κB p65 by PKA stimulates transcriptional activity by promoting a novel bivalent interaction with the coactivator CBP/p300. Mol Cell 1:661–671 20. Vanden Berghe W, De Bosscher K, Boone E, Plaisance S, Haegeman G (1999) The NF-κB engages CBP/p300 and histone acetyltransferase activity for transcriptional activation of the interleukin-6 gene promoter. J Biol Chem 274:32091– 32098 21. Na SY, Lee SK, Han SJ, Choi HS, Im SY, Lee JW (1998) Steroid receptor coactivator-1 interacts with the p50 subunit and coactivates NF-κB-mediated transactivations. J Biol Chem 273:10831–10834 22. Werbajh S, Nojek I, Lanz R, Costas MA (2000) RAC-3 is a NF-κB coactivator. FEBS Lett 485:195–199 23. Wu RC, Qin J, Hashimoto Y, Wong J, Xu J, Tsai SY, Tsai MJ, O’Malley BW (2002) Regulation of SRC-3 (pCIP/ACTR/AIB-1/ RAC-3/TRAM-1) coactivator activity by IκB kinase. Mol Cell Biol 22:3549–3561 24. Dechend R, Hirano F, Lehmann K, Heissmeyer V, Ansieau S, Wulczyn FG, Scheidereit C, Leutz A (1999) The Bcl-3 oncoprotein acts as a bridging factor between NF-κB/Rel and nuclear co-regulators. Oncogene 18:3316–3323 25. Baek SH, Ohgi KA, Rose DW, Koo EH, Glass CK, Rosenfeld MG (2002) Exchange of N-CoR corepressor and Tip60 coactivator complexes links gene expression by NF-κB and β-amyloid precursor protein. Cell 110:55–67 26. Lee SK, Kim JH, Lee YC, Cheong J, Lee JW (2000) Silencing mediator of retinoic acid and thyroid hormone receptors, as a novel transcriptional corepressor molecule of activating protein-1, NF-κB, and serum response factor. J Biol Chem 275: 12470–12474 27. Zhong H, May MJ, Jimi E, Ghosh S (2002) The phosphorylation status of nuclear NF-κB determines its association with CBP/p300 or HDAC-1. Mol Cell 9:625–636 28. Ashburner BP, Westerheide SD, Baldwin AS Jr (2001) The p65 (RelA) subunit of NF-κB interacts with the histone deacetylase (HDAC) corepressors HDAC1 and HDAC2 to negatively regulate gene expression. Mol Cell Biol 21:7065– 7077 29. Chen LF, Fischle W, Verdin E, Greene WC (2001) Duration of nuclear NF-κB action regulated by reversible acetylation. Science 293:1653–1657 30. Berger SL (1999) Gene activation by histone and factor acetyltransferases. Curr Opin Cell Biol 11:336–341 31. Imhof A, Yang XJ, Ogryzko VV, Nakatani Y, Wolffe AP, Ge H (1997) Acetylation of general transcription factors by histone acetyltransferases. Curr Biol 7:689–692 32. Kuo MH, Allis CD (1998) Roles of histone acetyltransferases and deacetylases in gene regulation. Bioessays 20:615–626 33. Cheung WL, Briggs SD, Allis CD (2000) Acetylation and chromosomal functions. Curr Opin Cell Biol 12:326–333 34. Gu W, Roeder RG (1997) Activation of p53 sequence-specific DNA binding by acetylation of the p53 C-terminal domain. Cell 90:595–606 35. Sterner DE, Berger SL (2000) Acetylation of histones and transcription-related factors. Microbiol Mol Biol Rev 64:435– 459 36. Martínez-Balbás MA, Bauer UM, Nielsen SJ, Brehm A, Kouzarides T (2000) Regulation of E2F1 activity by acetylation. EMBO J 19:662–671 37. Chen H, Tini M, Evans RM (2001) HATs on and beyond chromatin. Curr Opin Cell Biol 13:218–224 38. Gronroos E, Hellman U, Heldin CH, Ericsson J (2002) Control of Smad7 stability by competition between acetylation and ubiquitination. Mol Cell 10:483–493 39. Furia B, Deng L, Wu K, Baylor S, Kehn K, Li H, Donnelly R, Coleman T, Kashanchi F (2002) Enhancement of NF-κB acetylation by coactivator p300 and HIV-1 Tat proteins. J Biol Chem 277:4973–4980 40. Kiernan R, Bres V, Ng RW, Coudart MP, El Messaoudi S, Sardet C, Jin DY, Emiliani S, Benkirane M (2003) Post-activa-

41. 42.

43.

44. 45.

46. 47.

48.

49.

50.

51.

52.

53. 54.

55. 56. 57.

58.

tion turn-off of NF-κB-dependent transcription is regulated by acetylation of p65. J Biol Chem 278:2758–2766 Chen LF, Mu Y, Greene WC (2002) Acetylation of RelA at discrete sites regulates distinct nuclear functions of NF-κB. EMBO J 21:6539–6548 Hamamori Y, Sartorelli V, Ogryzko V, Puri PL, Wu HY, Wang JY, Nakatani Y, Kedes L (1999) Regulation of histone acetyltransferases p300 and PCAF by the bHLH protein twist and adenoviral oncoprotein E1A. Cell 96:405–413 Chakravarti D, Ogryzko V, Kao HY, Nash A, Chen H, Nakatani Y, Evans RM (1999) A viral mechanism for inhibition of p300 and PCAF acetyltransferase activity. Cell 96:393– 403 Munshi N, Merika M, Yie J, Senger K, Chen G, Thanos D (1998) Acetylation of HMG I(Y) by CBP turns off IFNβ expression by disrupting the enhanceosome. Mol Cell 2:457–467 Wen YD, Perissi V, Staszewski LM, Yang WM, Krones A, Glass CK, Rosenfeld MG, Seto E (2000) The histone deacetylase-3 complex contains nuclear receptor corepressors. Proc Natl Acad Sci U S A 97:7202–7207 Guenther MG, Barak O, Lazar MA (2001) The SMRT and N-CoR corepressors are activating cofactors for histone deacetylase 3. Mol Cell Biol 21:6091–6101 Fischle W, Dequiedt F, Hendzel MJ, Guenther MG, Lazar MA, Voelter W, Verdin E (2002) Enzymatic activity associated with class II HDACs is dependent on a multiprotein complex containing HDAC3 and SMRT/N-CoR. Mol Cell 9:45–57 Espinosa L, Ingles-Esteve J, Robert-Moreno A, Bigas A (2003) IκBα and p65 regulate the cytoplasmic shuttling of nuclear corepressors: cross-talk between notch and NF-κB pathways. Mol Biol Cell 14:491–502 Espinosa L, Santos S, Ingles-Esteve J, Munoz-Canoves P, Bigas A (2002) p65-NF-κB synergizes with notch to activate transcription by triggering cytoplasmic translocation of the nuclear receptor corepressor N-CoR. J Cell Sci 115:1295– 1303 Quivy V, Adam E, Collette Y, Demonte D, Chariot A, Vanhulle C, Berkhout B, Castellano R, de Launoit Y, Burny A, et al (2002) Synergistic activation of human immunodeficiency virus type 1 promoter activity by NF-κB and inhibitors of deacetylases: potential perspectives for the development of therapeutic strategies. J Virol 76:11091–11103 Mujtaba S, He Y, Zeng L, Farooq A, Carlson JE, Ott M, Verdin E, Zhou MM (2002) Structural basis of lysine-acetylated HIV-1 Tat recognition by PCAF bromodomain. Mol Cell 9:575–586 Owen DJ, Ornaghi P, Yang JC, Lowe N, Evans PR, Ballario P, Neuhaus D, Filetici P, Travers AA (2000) The structural basis for the recognition of acetylated histone H4 by the bromodomain of histone acetyltransferase gcn5p. EMBO J 19:6141– 6149 Dhalluin C, Carlson JE, Zeng L, He C, Aggarwal AK, Zhou MM (1999) Structure and ligand of a histone acetyltransferase bromodomain. Nature 399:491–496 Polesskaya A, Naguibneva I, Duquet A, Bengal E, Robin P, Harel-Bellan A (2001) Interaction between acetylated MyoD and the bromodomain of CBP and/or p300. Mol Cell Biol 21:5312–5320 Deng WG, Zhu Y, Wu KK (2003) Up-regulation of p300 binding and p50 acetylation in TNF-α-induced cyclooxygenase-2 promoter activation. J Biol Chem 278:4770–4777 Boyes J, Byfield P, Nakatani Y, Ogryzko V (1998) Regulation of activity of the transcription factor GATA-1 by acetylation. Nature 396:594–598 Li S, Aufiero B, Schiltz RL, Walsh MJ (2000) Regulation of the homeodomain CCAAT displacement/cut protein function by histone acetyltransferases p300/CREB-binding protein (CBP)-associated factor and CBP. Proc Natl Acad Sci USA 97:7166–7171 Chen FE, Huang DB, Chen YQ, Ghosh G (1998) Crystal structure of p50/p65 heterodimer of transcription factor NFκB bound to DNA. Nature 391:410–413

557 59. Zhang Q, Yao H, Vo N, Goodman RH (2000) Acetylation of adenovirus E1A regulates binding of the transcriptional corepressor CtBP. Proc Natl Acad Sci USA 97:14323– 14328 60. Huxford T, Huang DB, Malek S, Ghosh G (1998) The crystal structure of the IκBα/NF-κB complex reveals mechanisms of NF-κB inactivation. Cell 95:759–770 61. Jacobs MD, Harrison SC (1998) Structure of an IκBα/NF-κB complex. Cell 95:749–758 62. Soutoglou E, Viollet B, Vaxillaire M, Yaniv M, Pontoglio M, Talianidis I (2001) Transcription factor-dependent regulation of CBP and P/CAF histone acetyltransferase activity. EMBO J 20:1984–1992 63. Spilianakis C, Papamatheakis J, Kretsovali A (2000) Acetylation by PCAF enhances CIITA nuclear accumulation and transactivation of major histocompatibility complex class II genes. Mol Cell Biol 20:8489–8498 64. Vermeulen L, De Wilde G, Damme PV, Vanden Berghe W, Haegeman G (2003) Transcriptional activation of the NF-κB p65 subunit by mitogen- and stress-activated protein kinase-1 (MSK1). EMBO J 22:1313–1324

65. Zhong H, SuYang H, Erdjument-Bromage H, Tempst P, Ghosh S (1997) The transcriptional activity of NF-κB is regulated by the IκB-associated PKAc subunit through a cyclic AMP-independent mechanism. Cell 89:413–424 66. Sakurai H, Chiba H, Miyoshi H, Sugita T, Toriumi W (1999) IκB kinases phosphorylate NF-κB p65 subunit on serine 536 in the transactivation domain. J Biol Chem 274:30353–30356 67. Wang D, Westerheide SD, Hanson JL, Baldwin AS Jr (2000) TNF-α-induced phosphorylation of RelA/p65 on Ser 529 is controlled by casein kinase II. J Biol Chem 275:32592–32597 68. Sakaguchi K, Herrera JE, Saito S, Miki T, Bustin M, Vassilev A, Anderson CW, Appella E (1998) DNA damage activates p53 through a phosphorylation-acetylation cascade. Genes Dev 12:2831–2841 69. Liu L, Scolnick DM, Trievel RC, Zhang HB, Marmorstein R, Halazonetis TD, Berger SL (1999) p53 sites acetylated in vitro by PCAF and p300 are acetylated in vivo in response to DNA damage. Mol Cell Biol 19:1202–1209 70. Chan HM, Krstic-Demonacos M, Smith L, Demonacos C, La Thangue NB (2001) Acetylation control of the retinoblastoma tumour-suppressor protein. Nat Cell Biol 3:667–674