[Cell Cycle 3:7, 869-872; July 2004]; ©2004 Landes Bioscience
Reprogramming RelA Perspectives
*Correspondence to: Neil D. Perkins; School of Life Sciences; Division of Gene Regulation and Expression; MSI/WTB Complex; Dow Street; University of Dundee; Dundee, DD1 5EH Scotland, United Kingdom; Tel.: +44.1382.345.606; Fax: +44.1382.348.072; Email:
[email protected] Received 05/03/04; Accepted 05/04/04
Different NF-κB subunits have both unique and sometimes overlapping characteristics and can form a wide variety of DNA-binding homo and heterodimers.1 One of the most studied subunits is the RelA (or p65) subunit, which is expressed ubiquitously and is often found as a heterodimer with the p50 subunit.1 RelA has diverse functions but is often associated with conferring resistance to programmed cell death.2 In most non-stimulated, untransformed cells, RelA containing NF-κB complexes are typically held in an inactive cytoplasmic form, bound to one of a family of inhibitor proteins, the IκBs.1 Induction of DNA-binding generally results from activation of the IκB-kinase (IKK) signalosome complex, which phosphorylates one of the IκBs, promoting their ubiquitination and degradation by the proteasome.3 This releases the RelA complex, allowing it to translocate to the nucleus. Aberrant activation of nuclear localized RelA is observed with many human diseases, although this is rarely associated with genetic alterations to the rela gene itself .4 Since NF-κB regulates many components of the inflammatory response, such as cytokines, chemokines and their receptors,5 this is particularly true of diseases resulting from chronic inflammation or which have an inflammatory component.6 In addition, aberrant activation of RelA as well as other NF-κB subunits is also associated with many human cancers and the anti-apoptotic function of RelA has been found to reduce the efficacy of many common cancer therapies.4 For these reasons, many companies and researchers are in the process or designing or isolating specific inhibitors of the NF-κB pathway, and IKK activity in particular, for use as therapeutic drugs.7 But will inhibition of NF-κB and/or IKK always have the same effect under all circumstances and would such treatments always prove beneficial? In this article we shall discuss some of the issues arising from recent work from our laboratory and other research groups that suggests such approaches to NF-κB based therapy should be taken with care.
ce.
NF-κB, p65, histone deacetylase, ARF, p53, transcription, repression, chemotherapy, ultraviolet, UV-C
sci
en
KEY WORDS
ACKNOWLEDGEMENTS
ibu
Do
Previously published online as a Cell Cycle E-Publication: http://www.landesbioscience.com/journals/cc/abstract.php?id=955
The diversity of activators and targets of the NF-κB transcription factor family demands that there be regulatory mechanisms in place to control the specificity with which genes under their control are induced. In part this can be achieved through selective induction of different NF-κB subunits and through co-operative interactions with heterologous DNAbinding proteins and co-activators. Recent work from our laboratory indicates another critical mechanism regulating NF-κB. We find that the RelA(p65) NF-κB subunit does not always function as an inducer of gene expression, but under certain circumstances can be programmed to actively repress these same target genes. This repressor form of NF-κB appears to be induced by distinct, atypical pathways of activation and also through the action of some tumor suppressors. The identification of these pathways not only allows a reinterpretation of NF-κB function in normal cells and during tumorigenesis but could also have implications for both traditional and NF-κB based cancer therapy.
No tD ist r
School of Life Sciences; Division of Gene Regulation and Expression; MSI/WTB Complex; University of Dundee; Dundee, Scotland
te
ABSTRACT
Kirsteen J. Campbell Neil D. Perkins*
©
20
04
La
nd
es
Bio
We would like to thank all the members of the N.D.P. laboratory for their help and assistance. N.D.P. is funded by a Royal Society University Fellowship and K.J.C. is a member of the four year Wellcome Trust Ph.D. studentship programme at the University of Dundee.
κB? THE ROLE OF IKK IN TYPICAL AND ATYPICAL ACTIVATION OF NF-κ
Some of the best-characterized activators of RelA containing NF-κB complexes are inflammatory cytokines such as tumor necrosis factor α (TNF) and interleukin 1 (IL-1), although NF-κB DNA-binding is induced by numerous other stimuli, many of which are relevant to tumor initiation or therapy. Such inducers include ultraviolet (UV-C and UV-B) light, ionizing radiation, phorbol esters, hypoxia and the chemotherapeutic agents daunorubicin, doxorubicin (adriamycin), etoposide, camptothecin, vincristine, vinblastine and cisplatin.4,5 There are distinct differences between different inducers of RelA DNA-binding, however, which we divide into two broad categories. Typical activators, such as TNF and IL-1, induce rapid RelA complex nuclear localization and DNA binding www.landesbioscience.com
Cell Cycle
869
REPROGRAMMING RelA
within minutes of the stimulus being applied.3 Other inducers Table 1 SOME CIRCUMSTANCES WHEN RELA HAS BEEN DESCRIBED AS FACILITATING APOPTOSIS of RelA containing NF-κB Stimuli Cell type Reference complexes such as UV-C, doxorubicin and daunorubicin, Inducible expression of RelA Pro-B cells Sheehy and Schlissel 1999 result in DNA binding with Over-expression of RelA Breast cancer cells Ricca, et al. 2001 much slower kinetics, peaking UV-B, etoposide and tenoposide T-lymphocytes Kasibhatla, et al. 1998 after around 4 hours.8-10 We Inducible expression of p53, Osteosarcoma, colon carcinoma Ryan, et al. 2000 refer to these as atypical activaactinomycin-D and adriamycin and mouse fibroblast cells tors. Camptothecin Neurons Aleyasin, et al. 2004 Typical inducers of NF-κB Hydrogen peroxide and pervanadate Cervical carcinoma cells Kaltschmidt, et al. 2000 culminate in activation of the Inducible expression of ARF Osteosarcoma cells Rocha, et al. 2003 IKK signalosome, which is composed of two catalytic subunits, IKKα (IKK1) and IKKβ (IKK2) together with a regulatory subunit IKKγ (also known target genes.10 This was observed with transiently transfected as NEMO).3 In response to typical inducers, IKKβ phosphorylates reporter plasmids and analysis of endogenous genes. Importantly, IκB proteins at two conserved serine residues in the N-terminus.3 A the use of siRNAs and rela null cells demonstrated that RelA was more complicated situation is seen with atypical inducers, where required for repression to occur. The inhibitory effects of UV-C and both IKK-dependent and independent mechanisms of activation daunorubicin on NF-κB mediated gene expression were dominant have been proposed. Both UV-C and doxorubicin can activate NF- over activation resulting from TNF stimulation. This provided further κB in an IKK independent manner.8,9 In contrast, other reports indi- evidence that the inhibition of gene expression we were observing cate that UV-C, camptothecin, doxorubicin and etoposide stimulated did not result from a passive mechanism, where the RelA complexes NF-κB DNA binding is IKK dependent11,12 but with, at least in induced were merely defective in their ability to induce transcription. some cases, a requirement for the zinc finger domain of IKKγ not seen Rather, and in contrast to the conventional role ascribed to it as an with TNF.11 Furthermore after genotoxic stress IKKγ becomes activator, RelA acquired a dominant transcriptional repressor funcsumoylated, whereupon it translocates to the nucleus, becoming tion. Furthermore, as a result of repressing rather than inducing the ubiquitinated in a step dependent upon the ATM checkpoint kinase expression of anti-apoptotic genes, this provided an explanation for before being exported back to the cytoplasm where it mediates how RelA could facilitate rather than inhibit apoptosis. Whether this induction of NF-κB.13 Thus, while the relative importance of IKK- repression extends to the majority or just a subset of NF-κB target dependent and independent mechanisms of NF-κB induction have genes was not determined. There are reports of NF-κB inducing yet to be resolved in all cases, it is clear that atypical induction of pro-apoptotic genes, such as Fas or Fas ligand,2,5,16 although in our NF-κB displays numerous mechanistic differences to that seen with study induction of Fas by UV-C and daunorubicin was not RelA typical inducers. Are these differences reflected in different functions dependent. However, we can envisage a situation where NF-κB for RelA induced under these circumstances? complexes repress anti-apoptotic pathways while also inducing pro-apoptotic genes, thus tipping the balance between life and death decisively towards cellular mortality. RelA AND APOPTOSIS In trying to determine the mechanism behind this repression, our A large number of studies in recent years have led to the general attention was drawn to a report by Ashburner et al., which demonperception that NF-κB, and the RelA subunit in particular, mediates strated that RelA association with histone deactetylases (HDACs) resistance to programmed cell death induced by many stimuli, controlled basal expression of NF-κB target genes.21 We wondered including TNF, chemotherapy agents and ionizing radiation, whether such interactions between RelA and HDAC complexes through inducing the expression of a wide variety of anti-apoptotic could be regulated and account for the effects we had seen. Consistent genes.2,4,5 Although there is heavy bias in the literature towards these with this hypothesis we found that UV-C and daunorubicin induced reports describing anti-apoptotic functions for NF-κB, many studies association of RelA with HDACs, resulting in deacetylation of 10 also suggest that under some circumstances the opposite might occur histones at RelA targeted promoters. Accepted wisdom holds that activation of RelA DNA binding and RelA might help induce or facilitate cell death (Table 1).14-20 In represents activation of gene expression. We challenge this view and particular, we noted that many inducers which could be classified as suggest that dependent on the nature of the stimuli, RelA can either atypical inducers of NF-κB DNA-binding had been associated with activate or repress gene expression. We hypothesize, that as a result a pro-apoptotic function. of different post-translational modifications, RelA can be programmed to either function as an activator or repressor of transcription (Fig. 1). ACTIVE REPRESSION OF TRANSCRIPTION BY RelA A number of phosphorylation events leading to transcriptionally We noticed that the transcriptional activity of NF-κB induced by active RelA have been described.22 These modifications have either atypical inducers had not been explored in the same level of detail as been shown to, or are inferred to, assist in the recruitment of co-actiwith other activators. Given that it was probable that NF-κB had vator complexes such as p300/CBP. In contrast, we propose that different functions under these circumstances, we investigated NF-κB stimulation with atypical inducers such as UV-C, daunorubicin or regulated gene expression following induction by the atypical activators doxorubicin differentially modifies RelA, resulting in the recruitment UV-C and the highly related chemotherapeutic agents daunorubicin of HDAC containing co-repressor complexes leading to the inhibiand doxorubicin. Surprisingly, we found that in response to these tion of anti-apoptotic gene expression. Consistent with this model, stimuli, RelA repressed rather than activated anti-apoptotic NF-κB our preliminary, unpublished, data suggests that RelA is phosphorylated 870
Cell Cycle
2004; Vol. 3 Issue 7
REPROGRAMMING RelA
Figure 1. Programming RelA to either repress or activate gene expression. In this model, phosphorylation of RelA in response to certain inducers, such as TNF, results in the association of NF-κB complexes with coactivator proteins and the induction of gene expression. We propose that differential modification, such as phosphorylation at other sites on RelA in response to atypical inducers of NF-κB or the ARF tumour suppressor, results in the association of RelA with co-repressor, histone deacetylase containing, complexes. This leads to repression of certain NF-κB target genes. Our data suggests that modifications leading to this repressor form of RelA are dominant over modifications leading to activation. Therefore, compounds that specifically induce this switch from activator to repressor could have therapeutic potential and offer an alternative to global inhibition of NF-κB activity using IKK inhibitors.
in response to these atypical inducers and that these modifications occur at sites distinct to those seen with TNF. We cannot rule out other forms of modification having an important role, however. It is also possible that induction of heterologous DNA-binding proteins, functioning co-operatively with RelA, might also play a role in targeting anti-apoptotic genes for transcriptional repression.
HOW COMMON IS ACTIVE REPRESSION OF TRANSCRIPTION BY RelA?
When we first started investigating the effects of UV-C on NF-κB, our initial impression was that we were seeing transcriptionally inactive complexes rather than active repression of transcription. It was only when we looked closer at the data and performed further experiments that we realized a more dynamic effect was occurring. It is tempting to speculate therefore, that other examples in the literature of apparently “transcriptionally inactive” NF-κB might actually be circumstances where RelA has assumed a repressor function. It is possible that active repression of gene expression may be a more widely utilized function of RelA than is described in our paper. Possible examples of such circumstances include non-steroidal anti-inflammatory drugs (NSAIDs) and glucocorticoids that have been shown to induce RelA DNA binding with no transcriptional activity. Although the inhibitory affect on NF-κB has often been explained by prevention of IκB degradation, a number of reports reveal the situation can be more complex. For example, induction of transcriptionally inactive NF-κB complexes was observed in apoptotic HT-29 colon cancer cells in response to the COX-2 selective NSAID, NS-398.23 Furthermore, in a study with many parallels to the dominant effect of UV-C and daunorubicin over TNF that we reported, Egan et al. found that the aminosalicylate mesalamine dominantly inhibited interleukin 1 (IL-1) stimulated NF-κB dependent transcription and prevented IL-1 induced phosphorylation of the RelA subunit.24 Additionally, investigation of the action of glucocorticoids by the Haegeman and Adcock laboratories has shown that without altering RelA DNA binding, glucocorticoids prevent transactivation by disrupting RelA’s interaction with the basal transcription machinery25 by repressing RelA activated histone acetyltransferase activity and through inducing recruitment of HDAC2.26 www.landesbioscience.com
In addition to these examples, our findings with atypical activators are reminiscent of other results from our laboratory. We have also reported that the ARF tumor suppressor induces the association of RelA with HDAC1, resulting in the repression of the Bcl-xL promoter and gene.20 This appears to be mechanistically distinct to the effects seen with UV-C and daunorubicin: ARF mediated repression is dependent upon the RelA threonine 505 residue and the ATR checkpoint kinase, something not seen with the atypcial activators.10 Furthermore, ARF expression did not result in repression of the X-IAP gene, something seen with both UV-C and daunorubicin10,20 (Rocha S, Perkins N, unpublished observation). These observations suggest that there are multiple pathways leading to transcriptional repressor forms of RelA and that these result in differential repression of transcription. These differences might have distinct functional consequences in vivo.
CAN RelA MEDIATED REPRESSION BE HARNESSED FOR THERAPEUTIC PURPOSES?
There are circumstances where total inhibition of NF-κB could have serious consequences. In addition to the effects of ARF and p53 on NF-κB that we and others have reported, a tumor suppressor function for RelA has been previously described. For example, inhibition of NF-κB in the skin leads to squamous cell carcinoma.27,28 Furthermore, Gapuzan et al. found that immortalized rela null mouse fibroblasts can display a weakly transformed phenotype and form tumors in SCID mice.29 These, and other studies, suggest that in some cell types, under some circumstances, RelA might function as a tumor suppressor rather than a tumor promoter.30 We propose that active repression of anti-apoptotic genes by RelA provides, at least in part, a mechanistic explanation for how NF-κB could actually help inhibit tumor growth. Moreover, our data suggests that with some cancer treatments, in some tumors, NF-κB might in fact contribute to the effectiveness of the therapy rather than inhibit it, as is usually thought. In addition, active repression of NF-κB dependent gene expression could contribute to anti-inflammatory processes and also the resolution of inflammation, which has been demonstrated to be NF-κB dependent.31 Overly efficient inhibition of these processes using IKK inhibitors could have unfortunate side effects. Since the effects of UV-C and daunorubicin on RelA are dominant over those of TNF, it is possible that agents capable of switching RelA to its active repressor form, even in a background of constitutive activation, could be used to counter the pathogenic effects of NF-κB’s anti-apoptotic activities (Fig. 1). Derivatives of daunorubicin or doxorubicin that maintain the ability to switch RelA to an active repressor of anti-apoptotic target genes without the side effects of
Cell Cycle
871
REPROGRAMMING RelA
conventional chemotherapies, or other compounds that mimic the effects of UV-C or ARF on RelA, could be powerful anti-tumor or anti-inflammatory agents. Due to promoter specific effects (see above) it might be expected that these repressor forms of NF-κB might be more effective at inhibiting the expression of some genes while leaving others untouched, possibly resulting in fewer side effects. Thus while IKK inhibitors have an excellent chance of having real clinical value, targeting other aspects of NF-κB function could provide an alternative strategy for exploiting NF-κB function for therapeutic drug development. References 1. Ghosh S, May MJ, Kopp EB. NF-κB and rel proteins: Evolutionarily conserved mediators of immune responses. Ann Rev Immunol 1998; 16:225-60. 2. Kucharczak J, Simmons MJ, Fan YJ, Gelinas, C. To be, or not to be: NF-κB is the answer —role of Rel/NF-κB in the regulation of apoptosis. Oncogene 2003; 22:8961-82. 3. Karin M, Ben-Neriah Y. Phosphorylation meets ubiquitination: The control of NF-κB activity. Ann Rev Immunol 2000; 18:621-3. 4. Baldwin AS. Control of oncogenesis and cancer therapy resistance by the transcription factor NF-κB. J Clin Invest 2001; 107:241-6. 5. Pahl HL. Activators and target genes of Rel/NF-κB transcription factors. Oncogene 1999; 18:6853-66. 6. Ali S, Mann DA. Signal transduction via the NF-κB pathway: a targeted treatment modality for infection, inflammation and repair. Cell Biochem Funct 2004; 22:67-79. 7. Haefner B. NF-κB: Arresting a major culprit in cancer. Drug Discov Today 2002; 7:653-63. 8. Tergaonkar V, Bottero V, Ikawa M, Li QT, Verma, IM. IκB kinase-independent IκBa degradation pathway: Functional NF-κB activity and implications for cancer therapy. Mol Cell Biol 2003; 23:8070-83. 9. Kato T, Delhase M, Hoffmann A, Karin M. CK2 is a C-terminal IκB kinase responsible for NF-κB activation during the UV response. Mol Cell 2003; 12:829-39. 10. Campbell KJ, Rocha S, Perkins ND. Active repression of antiapoptotic gene expression by ReIA(p65) NF-κB. Mol Cell 2004; 13:853-65. 11. Huang TT, Feinberg SL, Suryanarayanan S, Miyamoto, S. The zinc finger domain of NEMO is selectively required for NF-κB activation by UV radiation and topoisomerase inhibitors. Mol Cell Biol 2002; 22:5813-25. 12. Bottero V, Busuttil V, Loubat A, Magne N, Fischel JL, Milano G, Peyron, JF. Activation of nuclear factor kB through the IKK complex by the topoisomerase poisons SN38 and doxorubicin: A brake to apoptosis in HeLa human carcinoma cells. Cancer Res 2001; 61:7785-91. 13. Huang TT, Wuerzbrger-Davis SM, Wu ZH, Miyamoto S. Sequential modification of NEMO/IKKγ by SUMO-1 and ubiquitin mediates NF-κB activation by genotoxic stress. Cell 2003; 115:565-76. 14. Ryan KM, Ernst MK, Rice NR, Vousden KH. Role of NF-κB in p53-mediated programmed cell death. Nature 2000; 404:892-7. 15. Sheehy AM, Schlissel, MS. Overexpression of RelA causes G1 arrest and apoptosis in a proB cell line. J Biol Chem 1999; 274:8708-16. 16. Kasibhatla S, Brunner T, Genestier L, Echeverri F, Mahboubi A, Green DR. DNA damaging agents induce expression of Fas ligand and subsequent apoptosis in T lymphocytes via the activation of NF-κB and AP-1. Mol Cell 1998; 1:543-51. 17. Ricca A, Biroccio A, Trisciuoglio D, Cippitelli M, Zupi G, Del Bufalo D. RelA over-expression reduces tumorigenicity and activates apoptosis in human cancer cells. Brit J Cancer 2001; 85:1914-21. 18. Kaltschmidt B, Kaltschmidt C, Hofmann TG, Hehner SP, Droge W, Schmitz ML. The pro- or anti-apoptotic function of NF-κB is determined by the nature of the apoptotic stimulus. Eur J Biochem 2000; 267:3828-35. 19. Aleyasin H, Cregan SP, Iyirhiaro G, O’Hare MJ, Callaghan SM, Slack RS, Park DS. Nuclear factor-κB modulates the p53 response in neurons exposed to DNA damage. J Neurosci 2004; 24:2963-73. 20. Rocha S, Campbell KJ, Perkins ND. p53- and Mdm2-independent repression of NF-κB transactivation by the ARF tumor suppressor. Mol Cell 2003; 12:15-25. 21. Ashburner BP, Westerheide SD, Baldwin AS. 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 2001; 21:7065-77. 22. Vermeulen L, De Wilde G, Notebaert S, Vanden Berghe W, Haegeman G. Regulation of the transcriptional activity of the Nuclear Factor-κB p65 subunit. Biochem Pharmacol 2002; 64:963-70. 23. Smartt HJM, Elder DJE, Hicks DJ, Williams NA, Paraskeva C. Increased NF-κB DNA binding but not transcriptional activity during apoptosis induced by the COX-2-selective inhibitor NS-398 in colorectal carcinoma cells. Brit J Cancer 2003; 89:1358-65. 24. Egan LJ, Mays DC, Huntoon CJ, Bell MP, Pike MG, Sandborn WJ, Lipsky JJ, McKean DJ. Inhibition of interleukin-1-stimulated NF-κB RelA/p65 phosphorylation by mesalamine is accompanied by decreased transcriptional activity. J Biol Chem 1999; 274:26448-53.
872
25. De Bosscher K, Berghe WV, Vermeulen L, Plaisance S, Boone E, Haegeman G. Glucocorticoids repress NF-κB-driven genes by disturbing the interaction of p65 with the basal transcription machinery, irrespective of coactivator levels in the cell. Proc Natl Acad Sci USA 2000; 97:3919-24. 26. Ito K, Jazrawi E, Cosio BJ, Barnes PJ, Adcock IM. p65-activated histone acetyltransferase activity is repressed by glucocorticoids. J Biol Chem 2001; 276:30208-15. 27. Dajee M, Lazarov M, Zhang JY, Cai T, Green CL, Russell AJ, et al. NF-κB blockade and oncogenic Ras trigger invasive human epidermal neoplasia. Nature 2003; 421:639-43. 28. van Hogerlinden M, Rozell BL, Ahrlund-Richter L, Toftgard R. Squamous cell carcinomas and increased apoptosis in skin with inhibited Rel/nuclear factor-κB signaling. Cancer Res 1999; 59:3299-303. 29. Gapuzan MER, Yufit PV, Gilmore TD. Immortalized embryonic mouse fibroblasts lacking the RelA subunit of transcription factor NF-κB have a malignantly transformed phenotype. Oncogene 2002; 21:2484-92. 30. Perkins ND. NF-κB: tumor promoter or suppressor? Trends Cell Biol 2004; 14:64-9. 31. Lawrence T, Gilroy DW, Colville-Nash PR, Willoughby DA. Possible new role for NF-κB in the resolution of inflammation. Nature Med 2001; 7:1291-7.
Cell Cycle
2004; Vol. 3 Issue 7
