Immunologic Research 1999;19(2-3): 183-189

Regulation of Inducible Gene Expression by the Transcription Factor NF- B

Section of Immunobiology, Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, CT 06526

Abstract

Key Words

The transcription factor NF-KB plays a critical role in regulating inducible gene expression in immune responses. The activation of NF-~cB is regulated at multiple levels, probably reflecting a need to maintain a tight control of its activity. We have recently discovered that direct phosphorylation of NF-~cB itself is essential for its transcriptional activity and that phosphorylation acts as a switch to determine association of nuclear NF-~B with histone acetylases vs deacetylases.

NF-~;B Gene expression Phosphorylation Histone deacetylation CBP/p300

Introduction The ability of an organism to react rapidly to changes in its environment is a crucial feature of immune responses. These responses include the synthesis and secretion of effector molecules, such as cytokines and adhesion molecules, and allow the recruitment of lymphocytes, macrophages, and natural killer cells to the sites of infection. The events that lead to the initiation of these immune responses, e.g., bacterial infection, cause the upregulation of genes encoding different effector molecules, mainly through the activation of inducible transcription factors. One such inducible transcription factor is NF-~B, which plays a critical role in both innate and adaptive immunity (1,2). NF-~cB belongs to a

Sankar Ghosh Yale University School of Medicine New Haven, CT 06510 E-mail: [email protected]

9 1999 Humana Press Inc. 0257-277X/99/ 19:183-189/$11.75

family of transcription factors known as the Rel family, which shares an approx 300 amino acid-long region of homology known as the Rel homology domain (RHD), and contains sequences required for DNA binding and dimerization (1). In mammalian cells, there are five known members of the Rel family, p50/p105, p52/p100, p65, c-Rel, and RelB. In keeping with the ancient origins of innate i m m u n e responses, transcription factors belonging to the Rel family are also present in invertebrates, such as Drosophila, where they are important regulators of primordial h o s t - d e f e n s e m e c h a n i s m s (3). There are three known Rel proteins in Drosophila, dorsal, dif, and Relish (1). Genetic analysis has revealed that both dorsal and dif play very

183

important roles in host-defense, primarily by controlling the expression of a large number of antimicrobial peptides (3). Interestingly, the similarity between Rel proteins in invertebrates and vertebrates also extends to the pathways that lead to their activation (3,4). The research in our laboratory is primarily focused toward understanding the regulation of the Rel/NF-~zB transcription factors in immune responses and in the development of the immune system. NF-~cB is normally retained in an inactive form in the cytoplasm through interactions with a class of anchoring molecules known as boBs, which mask the nuclear localization signals on NF-~cB proteins. There are three major IK:B isoforms in mammalian cells, I~:B-or bzB-[3, and bzB-E (1). Exposure of cells to stimuli that indicate infection, stress, or injury leads to the activation of intracellular signaling pathways that cause the phosphorylation of bzB proteins, followed by their ubiquitination and degradation (1). Degradation of b:B proteins unmasks the nuclear localizing signals on NF-~cB proteins that are then rapidly translocated to the nucleus. Studies carried out over the past few years has focused intensely on the mechanism by which cellsurface signals are transmitted to the cytosolic NF-~cB:bcB complexes. A major step forward in understanding these pathways has been the identification and cloning of the specific protein kinases that phosphorylate I~:B proteins (5-10). These novel protein kinases have been found to be components of a high-mol-wt complex that consists of two catalytic subunits, I~cB-kinase 1 (IKK1) and IKK2, along with other scaffolding and adapter subunits. The identity of the other components in the IKK complex has not been fully determined. Two adapter/scaftblding subunits have been characterized, NEMO (IKKT) and IKAP (11,12). The exact role these subunits play in the activity of the kinase complex is unclear at present. However, genetic studies indicate

184

that absence of NEMO completely abrogates signaling to NF-~zB (12). Therefore, it appears that assembly of a complex of proteins is essential for the biological function of the IKK complex. Current research in numerous laboratories, including ours, is directed toward trying to understand how the various components of the IKK complex function together to allow interaction with upstream components of signaling pathways, and to understand the biochemical changes that underlie the activation of the IKK complex. Activation of the IKK complex causes the phosphorylation of I~B proteins at two N-terminal serine residues (serines 32 and 36 in I~;B-cr and serine 19 and 23 in I~B-13) and leads to their ubiquitination and degradation (1,5). An important step in the degradation of bzB proteins is the ligation of ubiquitin molecules to the phosphorylated boB protein. The specific ubiquitin ligase complex responsible for specifically recognizing and ubiquitinating phosphorylated bzB proteins has been recently identified to be a member of an emerging class of ubiquitin ligases (13). The receptor in the ligase complex is [3-TrCP, a protein that contains F-box and WD repeats. The WD repeats specifically recognize and bind phosphorylated I~:B-cr Transfection of a mutant [~-TrCP protein lacking the F-box completely abrogates signal-mediated degradation of phosphorylated bcB-c~, which can then accumulate to significantly high levels in stimulated cells (13). Although ~-TrCP is clearly the protein that binds to phosphorylated bcB-or it is not clear if it is also the ubiquitin ligase. Homologs of [3-TrCP in other species exist in a complex with two other proteins, Skp- 1 and Cul- 1, and mammalian ]3-TrCP also forms a similar complex (13). It is possible that [3-TrCP provides the recognition capability and the other components have the ligase activity. It is, however, unclear how specific ]3-TrCP is for boB vs other proteins, since it also acts as the specific receptor for an HIV protein, VpU,

Ghosh

which has a sequence similar to that phosphorylated in bcB-c~ (14). Therefore it is possible that ~3-TrCP can recognize any substrate that can undergo phosphorylation at a site similar to that in I~:B-ot. Another important question is whether ]3-TrCP can recognize I~;B-]3 and I~:B-E, the other IKB isoforms that also undergo signal-dependent phosphorylation and degradation. An intriguing aspect about the signal-induced degradation of these other I~B isoforms is their slower kinetics of degradation (15,16). It will be interesting to see if I~:B-13/~ are recognized with a lower affinity by ~-TrCP thus explaining the slower rate of degradation. These issues and other aspects of the regulation of NF-~zB by different I~;B isoforms remain to be fully understood and are under investigation in our laboratory. The importance of NF-~zB in coordinating numerous aspects of both innate and adaptive immunity is indisputable (1,3). It is widely believed that targeted inhibition of NF-~B activity would be the key to the development of highly effective antiinflammatory therapeutics. In fact, two of the most widely used classes ofantiinflammatory drugs, salicylates and steroids, act by inhibiting the activity of NF-~cB (17-19). Therefore a thorough understanding of the mechanism by which NF-~zB functions as a transcription factor will be crucial in developing novel antiinflammatory therapeutics. We have also discovered recently that NF-~cB plays an important role in lymphocyte development, in particular as a survival factor in response to preT or preB cell receptor signaling (Voll and Ghosh, manuscript in preparation). This highlights the important role of NF-vd3 as an antiapoptotic factor and probably also helps to explain the oncogenic properties of different Rel family proteins. The wide range of biological systems that are affected by NF-~cB makes it an attractive subject of study, and we believe our analysis of the regulation of NF-~cB will allow us to address broader biological questions. In this

Regulation of Inducible Gene Expression

article, however, we will focus on and describe in detail one particular facet about the regulation of NF-~cB activity.

Regulation of the Transcriptional Activity of NF-~B Through Phosphorylation of the p65 Subunit by Protein Kinase A The novelty of the pathways that lead to the signal-induced degradation of I~;B-c~ has resulted in most of the attention in the field being focused on events in the cytoplasm leading to the phosphorylation and degradation of I~:B-c~. However, other inducible transcription factors, including those that translocate from the cytoplasm to the nucleus, are regulated through direct modification of the transcription factor itself. The most common form of modification is phosphorylation, which either enhances or inhibits the activity of the transcription factor in the nucleus. An accidental observation made in our laboratory has revealed that the activity of NF-~zB is also controlled by a fascinating and novel pathway that regulates the activity of nuclear NF-~cB and helps to ensure that only NF-~B induced in response to external signals is able to activate transcription. During the purification of I•B-]3, one of the two major boB isoforms that was cloned and characterized in our laboratory, we observed a copurifying polypeptide o f - 4 2 kDa (15). The copurification of this 42-kDa protein through multiple column chromatography steps led us to suspect that it was not simply a contaminating protein. Sequencing of the 42-kDa protein revealed that it was the catalytic subunit of PKA (20). This was an surprising observation, since we could not detect the presence of the regulatory subunit of PKA in the same purified fraction. PKA is generally thought to exist exclusively as a tetrameric complex of the catalytic (C) and regulatory (R) subunits (C2R2)(21). Elevation of cAMP levels inside the cell in response to external signals leads to the binding of cAMP

[ 85

to the PKA regulatory subunit. The consequent conformational change leads to the dissociation of the complex and release of active protein kinase A. Because we failed to detect PKA activity in the purified I~;B-[3 preparation, we assumed that the binding of PKA to 11
186

from the cytoplasm to the nucleus to phosphorylate CREB. Accumulation of free PKAc in the cytoplasm precedes its gradual nuclear translocation. In contrast, the PKAc released from NF-~B-IKB complexes is resequestered rapidly by free PKAr and therefore, does not remain in an active form for a sufficiently long period to allow nuclear translocation. Therefore, the effect of the two forms of active PKAc on gene expression is quite distinct, PKAc activated by elevation of cAMP activates the transcription factor CREB, whereas the PKAc released as a consequence of degradation of I~B regulates the transcriptional activity of NF-~B by phosphorylating the p65 subunit (Fig. 1).

Role of CBP/p300 in Regulating the Transcriptional Activity of NF-~B The next question that we wanted to answer was how phosphorylation of NF-~;B by PKAc affected its transcriptional activity. We began our analysis by using the well-characterized regulation of CREB transcriptional activity by PKA as a model (23). Phosphorylation of CREB at serine 133 by PKA allows interaction with a nuclear coactivator, known as CREB binding protein (CBP) (22). Recent studies have revealed that CBP is involved in the regulation of a large number of transcription factors, including NF-~B (24,25). However, the mechanism by which the interaction between CBP and NF-}cB was regulated remained unclear. Immunoprecipitation experiments indicated that association of p65 and CBP in stimulated cells was dependent on phosphorylation of p65 (26). To understand how phosphorylation regulated the interaction between these proteins, we began by identifying the different regions of interaction in both p65 and CBP. Such analysis revealed that CBP (and its homolog p300) interacted with NF-szB p65 through a novel bivalent interaction. One set of interaction was dependent on phosphorylation of p65 at

Ghosh

Fig. I. A schematic depiction of how phosphorylation acts as a switch to recruit histone acetylase vs deacetylases to nuclear NF-~B. In unstimulated cells, any nuclear NF-~B (either p50 homodimers or p50:p65 heterodimers) is unphosphorylated and is bound to HDAC. In contrast, the NF-~B that enters the nucleus from the cytoplasm in response to signaling is phosphorylated and specifically associates with CBP/p300. It is only the CBP/p300 associated NF-~;B that is transcriptionally active.

serine 276, whereas a second interaction occurred in the absence of phosphorylation (26). This posed a quandary, since the ability of p65 and CBP to interact in the absence of phosphorylation made it difficult to understand why phosphorylation was obligatory for CBP-dependent stimulation of NF-~B transcriptional activity. The answer lay in the manner in which the two CBP interaction sites

Regulation of Inducible Gene Expression

existed in the unphosphorylated p65 protein. The p65 protein can be broadly divided into two halves, the N-terminal -300 amino acids comprising the Rel homology domain that includes DNA binding, dimerization, and the site for PKA phosphorylation, and the remaining C-terminal tail of -250 amino acids. We have found that in unphosphorylated p65, the C-terminal tail folds back in a manner such

187

that it masks the N-terminus of p65. A consequence of this intramolecular interaction is the masking of both the CBP-interaction sites, thus preventing unphosphorylated p65 from binding to CBP. Phosphorylation of p65 weakens this intramolecular interaction between the N- and C-terminus of p65 creating a novel phosphorylation-dependent site of interaction and also unmasking the phosphorylation-independent site of interaction with CBP.

Phosphorylation of NF-~B and Its Interaction with Histone Deacetylase ! (HDAC-I) The findings described above provided an explanation for why phosphorylation of p65 stimulated the transcriptional activity of NF-~zB. However, they failed to explain the necessity of having yet another layer of regulation in the overall functioning of NF-~cB. Our recent studies have begun to shed some light on this particular question. It is known that in most unstimulated cells, a small amount of NF-~cB can be detected in the nucleus, the more abundant form being p50 homodimers (27). It is known that p50 dimers are incapable of driving transcription in transfected cells, yet in in vitro assays, p50 can support transcription from NF-~zB-dependent promoters (28). We wanted to characterize the nuclear NF-~zB proteins in unstimulated cells to understand why they were incapable of driving transcription, It is known that CBP/ p300 possesses endogenous histone acetylase activity, and therefore, we wanted to determine whether the acetylation property of CBP/HDAC was important in regulating the transcriptional activity of NF-~zB (29). The likely target for acetylation is the core histones, and it is believed that acetylation of histone tails helps in loosening the chromatin structure, thereby facilitating transcription. We therefore used a mutant form of CBP that is unable to acetylate histones (30). The

188

mutant protein is approx 50% less effective in stimulating NF-~cB transcriptional activity compared to wild-type CBP/p300, suggesting acetylase function of CBP along with its other properties, e.g., its ability to act as a coactivator, is important in stimulating transcription by NF-~cB (29). We then tested whether histone deacetylases had the reverse effect on transcription by NF-~zB. It is known that histone deacetylases are associated with repressive forms of different transcription factors, e.g., thyroid hormone receptors (29). We found that transfection of increasing amounts of HDAC- 1 inhibited NF-~zB-dependent transcription in a dose-dependent manner. The contribution of deacetylase activity on the inhibitory effect of HDAC-1 was further supported by experiments where treatment of cells with trichostatin A (TSA), an inhibitor of histone deacetylases, significantly enhanced transcription from a NF-~cB-dependent reporter in unstimulated cells. Immunoprecipitation experiments showed that nuclear NF-~zB in unstimulated cells was associated with HDAC-1. Therefore, it appears that nuclear NF-~B in unstimulated cells, which is unphosphorylated, is associated with HDAC-1, whereas in stimulated cells, the NF-~cB that enters the nucleus is phosphorylated on its p65 subunit by PKA, and preferentially associates with CBP/p300. Therefore phosphorylation of p65 by PKA acts as a switch that determines whether the nuclear NF-~B would be bound to HDAC-1 (and therefore be inactive or repressive) or CBP (thus being transcriptionally active) (Fig. 1). Such a regulatory scheme would therefore help to ensure that only NF-~cB that enters the nucleus in response to external stimuli would be able to activate transcription, whereas NF-~zB that is in the nucleus for any other reason is transcriptionally silent. This mechanism is unique among the various inducible transcription factors, since it imposes an additional layer of control on the ability of

Ghosh

NF-KB to activate transcription. Most likely the existence of multiple layers of regulation reflects the necessity of maintaining NF-~:B as a true inducible transcription factor at all times.

Summary The results described in this article have led us to conclude that the activity of NF-KB is regulated at multiple levels (Fig. 1). Signaldependent degradation of boB, followed by nuclear translocation, is but one part of its regulation, and phosphorylation by PKA as a consequence of signaling helps to ensure that

only NF-~zB that enters the nucleus from the cytoplasm is able to activate transcription. In addition, the mechanism used to inactivate the nuclear NF-KB in unstimulated cells is through association with HDAC- 1. We believe that this additional level of regulation provides another target for designing inhibitors of NF-KB activity and thus help in developing novel antiinflammatory drugs.

Acknowledgments The research in the author's laboratory is funded by HHMI and by a grant from NIH.

References 1 GhoshS, May MJ, KoppEB: NF-r,B and Rel proteins: evolutionariIy conserved mediators of immune responses. Annu Rev Immunol 1998;16:225-60. 2 Kopp E, Ghosh S: NF-KB and rel proteins in innate immunity. Adv Immunol 1995;58:1-27. 3 Medzhitov R, Janeway CA, Jr.: An ancient system of host defense. Curr Opinion Immunol 1998;10: 12-15. 4 Medzhitov R, Preston-Hurlburt P, Kopp E, Stadlen A, Chen C, Ghosh S, et al.: MyD88 is an adaptor protein in the hToll/lL-1 receptor family signaling pathways. Mol Cell 1998;2:253-258. 5 May MJ, Ghosh S: Signal transduction through NF-KB.Immunol Today 1998; 19:80-88. 6 DiDonatoJA,HayakawaM, Rothwarf DM, Zandi E, Karin M: A cytokine-responsive boB kinase that activates the transcription )actor NF-~B Nature. 1997;388: 548-554. 7 Mercurio F, Zhu H, Murray BW, Shevchenko A, Bennett BL, Li J, et al.: IKK- 1and IKK-2: cytokineactivated I~r kinases essential for NF-lcB activation. Science 1997; 278:860-866. 8 Regnier CH, Song HY, Gao X, Goeddel DV, Cao Z, Rothe M:

Regulation of Inducible Gene Expression

9

10

I1

12

13

14

Identification and characterization of an I~B kinase. Cell 1997; 90:373-383. Woronicz JD. Gao X, Cao Z, Rothe M, Goeddel DV: hoB kinase-beta: NF-~B activation and complex formation with bob kinase-a and NIK. Science 1997;278:866-869. Zandi E, Rothwarf DM, Delhase M, Hayakawa M, Karin M: The I~:B kinase complex (IKK) contains two kinase subunits, 1KKa and IKKb, necessary for boB phosphorylation and NF-KBactivation. Cell 1997;91:243-252. Cohen L, Henzel WJ, Baeuerle PA: IKAP is a scaffold protein of the IKB kinase complex. Nature 1998;395:292-296. Yamaoka S, Courtois G, Bessia C, Whiteside ST, Well R, Agou F, et al.: Complementation cloning of NEMO, a component of the I~B kinase complex essential for NF-KB activation. Cell 1998; 93:1231-1240. Yaron A, Hatzubai A, Davis M, Lavon I, Amit S, Manning AM, et al.: Identification of the receptor component of the I~r ligase. Nature 1998;396:590-594. Margottin F, Bour SP, Durand H, Selig L, Benichou S, Richard V, et al.: A novel human WD protein, h-lgTrCp,that interacts with HIV- 1

15

16

17

18

19

20

Vpu connects CD4 to the ER degradation pathway through an F box motif. Mol Cell 1998;1:565-574. ThompsonJ, PhillipsR, ErdjumentBromage H, Tempst P, Ghosh S: I~;B-b regulates the persistent response in a biphasic activation of NF-~B Cell 1995;80:573-582. Whiteside ST, Epinat JC, Rice NR, Israel A: l~Be, a novel member of the b;B family, controls RelA and cRel NF-~B activity. EMBO J 1997;16:1413-1426. Auphan N, DiDonato JA, Rosette C, Helmberg A, Karin M: lmmunosuppression by glucocorticoids: inhibition of NF-KBactivity through induction of IKB synthesis. Science 1995;270:286-290. Kopp E, Ghosh S: Inhibition of NF-~cB by sodium salicylate and aspirin. Science 1994;265:956-959. Scheinman RI, Cogswell PC, Lofquist AK, Baldwin AS, Jr.: Role of transcriptional activation of IKB~ in mediation of immunosuppression by glucocorticoids. Science 1995;270:283-286. Zhong H, SuYang H, ErdjumentBromage H, Tempst P, Ghosh S: The transcriptional activity of NF-KB is regulated by the IKBassociated PKAc subunit through a cyclic AMP-independent mechanism. Cell 1997;89:4134.24.

189

21 Scott JD: Cyclic nucleotide-dependent protein kinases. Pharmacol Yher 1991;50:123-145. 22 Parker D, Ferreri K, Nakajima T, LaMorte V J, Evand R, Koerber SC, et al.: Phosphorylation of CREB at ser- 133 induces complex formation with CREB-binding protein via a direct mechanism. Mol Cell Biol 1996;16:694-703. 23 Kwok RPS, Lundblad JR, Chrivia JC, Ricbards JP, Bachinger HP, Brennan RG, et al.: Nuclear protein CBP is a coactivator for the transcription factor CREB Nature 1994;370:223-225. 24 Gerritsen ME, Williams AJ, Neish AS, Moore S, Shi Y, Collins T:

190

CREB-binding protein/p300 are transcriptional coactivators of p65. Proc Natl Acad Sci USA 1997;94:2927-2932. 25 Perkins ND, Felzien LK, Betts JC, Leung K, Beach DH, Nabel GJ: Regulation of NF-KB by cyclindependent kinases associated with the p300 coactivator. Science 1997;275:523-527. 26 Zbong H, Voll RE, Ghosh S: Phosphorylation of NF-lcBp65 by PKA stimulates transcriptional activity by promotinga novel bivalent interaction with the coactivator CBP/ p300. Mol Cell 1998;1:661-671. 27 Kang SM, Tran AC, Grilli M, Lenardo MJ: NF-~B subunit regu-

lation in nontransformed CD4+ T lympbocytes. Science 1992;256: 1452-1456. 28 FujitaT, Nolan GP, Ghosh S, Baltimore D: Independent modes of transcriptional activation by the p50 and p65 subunits of NF-KB. Genes Dev 1992;6:775-787. 29 Struhl K: Histone acetylation and transcriptional regulatory mechanisms. Genes Dev 1998; 12:599-606. 30 Korzus E, Torchia J, Rose DW, Xu L, Kurokawa R, McInerney EM, et al.: Transcription factor-specific requirements for coactivators and their acetyltransferase functions. Science 1998;279:703-707.

Ghosh