BMB reports

ZAS3 represses NFκB-dependent transcription by direct competition for DNA binding

Joung-Woo Hong1 & Lai-Chu Wu2,3,* 1The Graduate School of East-West Medical Science, Kyung Hee University, Yongin 446-701, Korea, Departments 2Molecular and Cellular Biochemistry, 3Internal Medicine, College of Medicine and Public Health, The Ohio State University, Columbus, OH, 43210, USA

NFκB and ZAS3 are transcription factors that control important scription of target genes including anti-apoptotic genes. Thus, cellular processes including immunity, cell survival and apop- mutations in components of the NFκB signaling pathway that tosis. Although both proteins bind the κB-motif, they produce make NFκB constitutively expressed are often implicated in opposite physiological consequences; NFκB activates tran- certain types of cancers (4). scription, promotes cell growth and is often found to be con- The ZAS family of large, separated-paired-C2H2 stitutively expressed in cancer cells, while ZAS3 generally re- proteins also regulates transcription through binding κB-motif presses transcription, inhibits cell proliferation and is down- (5, 6). Unlike NFκB that promotes tumorigenesis, the ZAS pro- regulated in some cancers. Here, we show that ZAS3 inhibits teins are most likely tumor suppressors. Down-regulation of NFκB-dependent transcription by competing with NFκB for ZAS2 (HIVEP2) is often found in breast cancer (7), and func- the κB-motif. Transient transfection studies show that N-termi- tional loss of ZAS1 and ZAS2 is associated with poor prognosis nal 645 amino acids is sufficient to repress transcription acti- for chronic lymphocytic leukemia (8). Furthermore, ZAS3 defi- vated by NFκB, and that the identical region also possesses in- ciency has resulted in tumor formation, cell immortalization trinsic repression activity to inhibit basal transcription from a and growth acceleration (9). Hence, NFκB and ZAS, the two promoter. Finally, in vitro DNA-protein interaction analysis major families of cellular κB-binding proteins, appear to have shows that ZAS3 is able to displace NFκB by competing with opposing effects on the regulation of transcription and growth. NFκB for the κB-motif. It is conceivable that ZAS3 has ther- A recent study shows that ZAS3 inhibits NFκB by directly asso- apeutic potential for controlling aberrant activation of NFκB in ciating with TNF receptor-associated factor 2 (TRAF2), thereby various diseases. [BMB reports 2010; 43(12): 807-812] blocking nuclear translocation of NFκB (10). Consistent with these results, NFκB is constitutively expressed in ZAS3 knock- out cells (11). INTRODUCTION Although NFκB and ZAS3 proteins are able to bind similar DNA binding sites, how they cooperatively regulate tran- NFκB controls key cellular processes, including immunity, cell scription of downstream target genes and thereby how they survival and apoptosis by activating transcription of target control the switch between cell survival and death remains a genes through the κB motif (GGGRNNYYCC) (1, 2). NFκB is mystery. ubiquitously expressed in the cell, and its cellular activity is In the present study, we show that the N-terminal 645 ami- subjected to tight post-translational control. In most cell types, no acids of ZAS3 compete efficiently with NFκB for κB-motif NFκB exists as an inactive form in the cytoplasm by associat- and abolish NFκB-dependent transcription. In addition, the ing with an inhibitor, IκB. Many stimuli, such as TNFα and lip- N-terminal region possesses intrinsic repression activity. Fur- opolysaccharide (LPS), activate cellular signaling pathways that thermore, ZAS3 does not prevent overexpressed NFκB from trigger the assembly of the IκB kinase (IKK) complex. IKK phos- translocating into the nucleus. Together, these results suggest phorylates IκB and marks it for ubiquitination and degradation that DNA competition by ZAS3 alone is sufficient to repress (3). NFκB then translocates into the nucleus (at which time it is transcription activated by NFκB and that ZAS3 is an endoge- referred to as activated) and induces NFκB-dependent tran- nous κB-motif competitor with intrinsic repression activity.

*Corresponding author. Tel: 1-614-293-3042; Fax: 1-614-293-5631; RESULTS E-mail: [email protected] DOI 10.5483/BMBRep.2010.43.12.807 ZAS3 represses NFκB-activated transcription by binding the κB-motif Received 2 November 2010, Accepted 6 November 2010 Previously, ZAS3 was shown to bind the κB-motif and mediate transcriptional repression (11). To extend our understanding of Keywords: DNA competition, κB-motif-binding proteins, NFκB, Transcriptional repression, ZAS3 ZAS3 function, we asked if ZAS3 could repress transcription http://bmbreports.org BMB reports 807 ZAS3 inhibits NFκB by DNA competition Joung-Woo Hong and Lai-Chu Wu

activated by NFκB by binding to the κB-motif. Increasing on luciferase activity. Taken together, these data indicate that amounts of expression vectors encoding ZAS protein (amino ZAS3 represses transcription that is activated by NFκB and this acid 106 to 2013) were co-transfected into HEK293 cells to- repression requires binding of ZAS3 to the κB-motif. gether with a constant amount of expression vector encoding p65 and p50, and a reporter plasmid (p3xκB-Fos-Luc, see Fig. The N-terminal region of ZAS3 is required for repression of 1A top) containing three consecutive copies of κB-motifs and a NFκB-activated transcription c-fos minimal promoter (−65 to +109, relative to tran- To determine which part(s) of the ZAS3 protein are necessary scription start site as +1) immediately upstream of the firefly for transcriptional repression, a number of systemic truncations luciferase gene. As expected, ectopic expression of p65 and were made by deleting different structural components, one at p50 increased luciferase activity ∼10-fold more than basal a time. To assay repression, p65 and p50 were used as an acti- level. Co-expression of ZAS3 repressed the activity of p65 and vator and p3xκB-Fos-Luc as the reporter. First, either the N- or p50 up to almost the basal level in a dose-dependent manner C-terminal half was deleted from full-length ZAS3, and its re- (Fig 1A). An essentially identical result was obtained even pression activity was examined (Fig. 2A). Deleting the C-termi- when luciferase expression was activated up to ∼100-fold by nal half containing the ZASC domain had little effect, if any, increased expression of p65 and p50 (data not shown). The fact that ZAS3 expression could significantly inhibit luciferase activity induced by p65 and p50 expression suggests that ZAS represses NFκB-activated transcription. To determine whether the repression activity of ZAS is exerted through binding the κB-motif, a similar co-transfection experiment was performed with a c-fos minimal promoter containing a reporter plasmid (pFos-Luc, see Fig. 1B top), but no κB-motif. In the absence of the κB-motif, neither p65/p50 nor ZAS3 had a negative effect

Fig. 2. The N-terminal portion of ZAS3 is sufficient to repress Fig. 1. ZAS3 represses NFκB-activated transcription. The indicated NFκB-activated transcription. (A) The indicated truncation deriva- luciferase reporter plasmids, diagramed above each graph, were tives of ZAS3 were tested for their ability to repress transcription cotransfected with the NFκB (p65 and p50) and FLAG-tagged activated by NFκB. Like the FLAG-tagged full-length ZAS3 protein full-length ZAS3 (amino acid 106-2013) expression vectors into used in Fig. 1, all ZAS3 proteins were tagged with octa-peptide HEK293 cells. Relative luciferase activities, which are 100 X Luc FLAG. The schematic of the FLAG-ZAS3 fusion proteins and the experiment/Lucactivated (A) or 100 X Lucexperiment/Lucbasal (B), are shown. corresponding relative luciferase activities are shown in the left Lucactivated is the normalized luciferase activity in the presence of and right, respectively. The numbers beside each diagram indicate NFκB expression vectors and in the absence of ZAS3; Lucbasal is the first and last amino acid of each protein relative to the amino the normalized luciferase activity in the absence of both NFκB acid sequences of wild-type ZAS3 (Top, left panel). Ten nano- and ZAS expression vectors. (A) ZAS3 represses luciferase expre- grams of p3xκB-Luc and 10 ng of pCH110 were cotransfected ssion activated by overexpressed NFκB. Ten nanograms of p3xκB- with or without 0.1 μg of each p65 and p50 expression plasmid Fos-Luc and 10 ng of pCH110 were cotransfected with or with- and/or 0.2 μg of each ZAS3 expression vector into HEK293 cells. out 0.1 μg of each p65 and p50 expression vector and/or in- Relative luciferase activities were determined as described in Fig. creasing (0.1, 0.2 and 0.4 μg) amount of ZAS3 expression plas- 1. (B) All overexpressed ZAS3 proteins were localized in the nu- mid into HEK293 cells. (B) ZAS3 does not affect basal transcrip- clei of transfected HEK293 cells. Four micrograms of each ZAS3 tion in the absence of the κB-motif. Ten nanograms of pFos-Luc expression plasmid was transfected into HEK293 cells. Nuclear and 10 ng of pCH110 were cotransfected with or without 0.1 μg and cytoplasmic extracts were prepared and western blot analysis of p65 and p50 expression plasmid and/or 0.4 μg of ZAS3 ex- was performed with antibodies for FLAG, hsp90 and histone H1. pression plasmid into HEK293 cells. N; nuclear extract, C; cytoplasmic extract.

808 BMB reports http://bmbreports.org ZAS3 inhibits NFκB by DNA competition Joung-Woo Hong and Lai-Chu Wu

on the repression activity of ZAS3 (106-1186), but deleting the DNA binding domain (DBD) and a reporter plasmid contain- N-terminal half containing the ZASN domain resulted in a sig- ing GAL4-binding sites (p4xG-SV-Luc) (13, 14) were used (Fig. nificant loss of its repression activity (1186-2153). Next, we 3A). GAL4-ZAS3 expression vectors were co-transfected into deleted the zinc fingers in the ZASN domain, the nuclear lo- HEK293 cells with a reporter plasmid containing four consec- calization signal (NLS) region or the entire ZASN domain from utive GAL4-binding sites. Transcription from this SV40 basal ZAS106-1186 (261-1092, 106-750 and 750-1186, respectively). promoter of p4xG-SV-Luc was weak, but readily detectable, None of the two ZAS3 vectors, 261-1092 or 750-1185, gave and could be quantified accurately (Fig. 3A). All GAL4 DBD- rise to strong repression. However, a ZAS3 vector containing ZAS3 fusion proteins containing a ZASN domain (GLA4106- the ZASN domain and zinc finger 3, 107-750, did display the 2013, GAL4106-1186, and GAL4106-750) repressed basal strongest repression activity among all ZAS3 constructs tested. transcription from the SV40 minimal promoter (up to ∼20- Although repression activity of more the truncated form of the fold) (Fig. 3A), whereas a GAL4-ZAS3 protein containing a C-terminal half, 1497-2012, was also measured, it was inactive ZASC domain (GAL41186-2153) failed to repress the transcri- as a repressor and resulted in weak activation. These findings ption in the absence of activators. Since there are no endoge- indicate that a 1.9 kb region around the ZASN domain is suffi- nous transcriptional regulators that bind to the GAL4-binding cient for the full repression activity of ZAS3. sites (15), our data support the hypothesis that ZAS3 can also Recently, it was reported that ZAS3 inhibits NFκB activity induced by TNF signaling pathway through direct interaction with TRAF2 in the cytoplasm (10). This finding made it hard to exclude the possibility that the observed repression activity of ZAS3 is due to direct interaction with other proteins such as TRAF2 involved in transcription and/or signaling pathway. To measure the expression and localization of the ZAS3 proteins, all ZAS3 proteins were tagged with a FLAG octapetide (D-Y- K-D-D-D-D-K) following a methionine (M), and nuclear and cytoplasmic extracts were prepared after transfection and west- ern blot analysis with a monoclonal anti-FLAG antibody (M2, Sigma) were performed (Fig. 2B). These results make several points. First, while there were reproducible differences in the levels of ZAS3 protein accumulation in nuclear extracts, there was no correlation between the amount of protein detected and its repression activity. A second point is that repression ac- tivity was also not correlated with protein size. Finally, regard- less of NLS, all proteins were localized in the nucleus and none were detectable in the cytoplasm. This result strongly supports the previous finding that ZAS3 acts as mainly in the nucleus, but not in the cytoplasm (11), al- though this notion is not consistent with the previous ob- servation that ZAS3 blocks translocation of NFκB into the nu- Fig. 3. ZAS3 also possesses intrinsic activity of transcription cleus by interacting with TRAF2 in the cytoplasm (10). The repression. (A) The GAL4 DNA-binding domain (DBD) (amino acid 1-147)-fused ZAS3 proteins were tested for their ability to re- separated nuclear and cytoplasmic extracts were verified by press basal transcription from the SV40 promoter. Schematic of measuring nuclear and cytoplasmic marker proteins, histone FLAG-GAL4 DBD-ZAS3 fusion proteins and the corresponding - H1 and hsp90 (Fig. 2B, bottom two panels). These findings in- ative luciferase activities are shown in the left and right, respectively. dicate that ZAS3 is a nuclear protein and the N-terminal region The numbers beside each diagram indicate the first and last ami- no acid of each protein relative to the amino acid sequence of containing the ZASN domain is required for the repression ac- wild-type ZAS3. The GAL4 DBD-ZAS3 proteins containing ZASN tivity of ZAS3, suggesting that ZAS3 binding to DNA is essen- domain repress basal transcription from the SV40 promoter. Ten tial for its repression activity. nanograms of p4xG-SV-Luc and 10 ng of pCH110 were cotrans- fected with 0.2 μg of each ZAS3 expression vector or the equal amount of expression vector containing GAL4 DBD only into ZAS3 represses basal transcription HEK293 cells. Relative luciferase activities are shown in which the In our co-transfection experiments (Fig. 1A, last lane), ZAS3 percent activity in the absence of ZAS3 expression vector was strongly blocked luciferase activity in the absence of p65 and considered to be 100%. (B) All GAL4 DBD-ZAS3 proteins were p50 proteins, providing evidence that ZAS3 represses basal localized in nuclei of HEK293 cells. Four micrograms of each ZAS3 expression plasmid was transfected into HEK293 cells, nu- transcription in the absence of activators, as in a direct re- clear and cytoplasmic extract were prepared and Western blot pression model (12). To determine if ZAS3 could function as a analysis was performed with antibodies for GAL4 DBD, hsp 90 direct repressor, GAL4-ZAS3 fusion proteins containing GAL4 and Histone H1. http://bmbreports.org BMB reports 809 ZAS3 inhibits NFκB by DNA competition Joung-Woo Hong and Lai-Chu Wu

Fig. 4. The competing effect of ZAS3 on binding of p65/p50 to 32P-κB-DNA. (A) Electrophoretic mobility shift assay (EMSA) of 32P-κB- DNA and recombinant proteins, as indicated at the top of the lanes. Competitors were added in 150-fold excess of unlabeled κB-DNA. (B) DNA competition EMSA. GST-tagged recombinant p65 and p50 (GST-p65/p50) were mixed with 32P-κB-DNA and ZAS3 fusion proteins, and then the mixture was incubated for 10 minutes before the addition of excess unlabeled κB-DNA. The final mixture including the ex- cess amount of unlabeled κB-DNA was incubated for another 10 minutes before gel loading. (C) Protein competition assays showing that ZAS3 competes with NFκB for κB-DNA. EMSA of GST-p65/p50 (10 ng), varying amounts of ZAS3 fusion proteins, GST106-750 or GST1497-2012, and 32P-κB-DNA. Control GST (10 ng) did not form DNA-protein complexes.

directly interfere with the general transcription machinery by tive affinity of GST106-750 was higher than that of GST1497- using the intrinsic repression activity in addition to its binding 2012, indicating that there is a correlation between relatively affinity to the κB motif. Consistent with the results of previous strong DNA-binding affinity and repression activity. Namely, western blot analysis (Fig. 2B), all GAL4-ZAS3 fusion proteins the repression activity of ZAS3 is proportional to the strength were also located in the nucleus (Fig. 3B). Although there was of the DNA-binding affinity. This result supports a notion that a again some variation in protein expression level, there was no competition model is the most likely mechanism underlying obvious correlation between expression and repression activity. the observed repression activity of ZAS3 against NFκB-acti- For instance, the strong repressor, 106-1186, appears to be vated transcription. present in the lowest abundance (Fig. 3B, lane 3). Collectively, We also assessed whether the GST-ZAS3 fusion proteins it is conceivable that while NFκB remains idle in the cyto- could compete with NFκB for binding the κB-motif. To ad- plasm due to the lack of stimulated signaling, ZAS3 occupies dress this, we pre-incubated GST-p65/p50 with one of the κB-motifs and strictly prevent promoters of NFκB target genes GST-ZAS3s and 32P-labeled κB-motif probe, and then we add- from transcribing in an uncontrolled manner. ed the indicated amount of unlabeled κB-motif probe (cold κB-motif-DNA) to the mixture. GST-p65/p50 and GST106-750 ZAS3 displaces NFκB from κB-motif. remained to bind κB-motif, even with an excess amount of the Given the fact that ZAS3 and NFκB have the ability to bind the unlabeled κB-motif, whereas the smallest amount of the un- κB-motif and ZAS3 represses transcription activated by NFκB, labeled κB-motif was sufficient to displace GST1497-2012 it is possible that ZAS3 represses NFκB-activated transcription from the preformed complex. Furthermore, when increasing through a competition mechanism. To determine whether ZAS amounts of one of the GST-ZAS3s were added to the mixture, competes with NFκB for binding the κB-motif, an electropho- where the constant amount of GST-p65/p50 had been pre-in- retic mobility shift assay (EMSA) was performed with bacte- cubated with the labeled κB-motif, only GST106-750, but not rially expressed ZAS3 and p65/p50. The N- and C-terminal GST1497-2012, efficiently abrogated the interaction between 645 and 516 amino acids of ZAS3 (106-750 and 1497-2012), κB-motif and GST-p65/p50. Together, the above results in- which presented the strongest and the weakest repression ac- dicate that the N-terminal 645 amino acids of ZAS3 can dis- tivities, were fused with glutathione-S-transferase (GST). The place NFκB from the κB-motif, suggesting that competition be- GST106-750 contains the ZASN domain, zinc finger 3 and the tween ZAS3 and NFκB for a common DNA site is mainly re- first proline-rich region, and the GST1497-2012 contains the sponsible for the repression activity of ZAS3. ZASC domain and the linker region that is conserved among ZAS proteins (6). DISCUSSION In EMSA, both GST-p65/p50 and GST-ZAS3, but not GST alone, formed a complex with 32P-labeled κB-motif with in- Here, we provide the first evidence that the large zinc finger stantly distinguishable gel mobility and the specificity of bind- protein, ZAS3, represses NFκB-activated transcription in a κB- ing was verified with unlabeled competitors (Fig. 4A). The - motif-dependent manner. The repression activity of ZAS3 ap-

810 BMB reports http://bmbreports.org ZAS3 inhibits NFκB by DNA competition Joung-Woo Hong and Lai-Chu Wu

pears to result from the combination of competition with repression domain of ZAS3 provides a future framework for NFκB for the κB-motif and its intrinsic repression ability. developing a specific inhibitor to block aberrant expression of Based on the concept of a genetic switch proposed by NFκB target genes. Ptashne (16), we propose a “dual-switch” model, whereby An accumulating number of studies have shown that NFκB ZAS3 and NFκB can independently bind one or multiple cop- activates transcription of numerous genes involved in cell sur- ies of κB-motifs to regulate transcription of NFκB-target genes. vival and death in addition to immunity, development and The transcription activity of the target gene depends on the oc- diseases. Although, to date, several proteins that synergistically cupancy of ZAS3 and NFκB on the κB-motifs. At least one enhance or attenuate transcriptional activation competency of mechanism can be envisioned for the model. In the resting NFκB have been identified (22-24), a transcription factor that state, ZAS3 not only masks the κB-motif, but it also prevents represses NFκB-activated transcription via binding κB-motif NFκB target genes from unnecessary expression by using its di- has not been reported. As presented here, the ability of ZAS3 rect repression activity. A recent study demonstrated that to repress transcription of NFκB-target genes by binding a NFκB does not constantly reside in the cytoplasm, but rather, κB-motif indicates that the transcription of the NFκB-target shuttles continuously in and out of the nucleus in the unsti- genes does not rely solely on NFκB, but also on the coopera- mulated state (17), indicating the possible existence of a small tion with at least one more factor, ZAS3. Based on the ex- amount of NFκB in the nucleus. In this condition, a relatively istence of several κB-motif-binding proteins in addition to large amount of ZAS3 competes with the residual NFκB for NFκB [reviewed in (11)], however, it is not likely that NFκB κB-motifs and occupies the κB-motifs, resulting in repression and ZAS3 are the only proteins that are able to bind to of NFκB target genes. A quenching repression model (12) is κB-motif. At least other ZAS family members, ZAS1 and ZAS2, the least likely scenario to explain the repression ability of have been demonstrated to specifically interact with the ZAS3 on the NFκB-dependent transcription, because of the ob- κB-motif (reviewed in (6)). Therefore, a combinational effect of servation that when ZAS3 and NFκB are competing for the ZAS1, ZAS2, ZAS3 and NFκB on the κB-motif should be ex- κB-motif, each forms only one DNA-protein complex in vitro amined to fully understand the mechanism of NFκB-dependent (Fig. 4B, C). This result suggests that the two proteins do not transcription. directly interact with each other. However, the possible in- volvement of a cofactor protein(s) between ZAS3 and NFκB as MATERIALS AND METHODS well as the physical interaction of ZAS3 with other Rel family members, aside from p65 and p50, still remains to be tested. Detailed information is described in Supplementary Material. A functional domain(s) that possesses the intrinsic repression ability of ZAS3 is located in the N-terminal region of ZAS3. Acknowledgements The 645 amino acids surrounding the ZASN domain did dis- We would like to give the deepest thank to Dr. Denis C. play repression activity that is stronger than or at least equiv- Guttridge for pFos-Luc, p3xkB-Fos-Luc, p65 and p50. This alent to that of full-length ZAS3, whereas the C-terminal half work was supported by a grant from the National Cancer containing the ZASC domain had little, if any, effect on the re- Institute (p30CA15058). pression activity and actually produced weak activation. Func- tional repression domains have been shown to require general REFERENCES hydrophobicity (i.e. proline and alanine) (18, 19). Additio- nally, the abundance of serine, threonine and acidic residues 1. Pierce, J. W., Lenardo, M. and Baltimore, D. (1988) is a common feature of transcriptional activation domains (20, Oligonucleotide that binds nuclear factor NF-κB acts as a 21). The existence of serine, threonine and an acidic stretch, as lymphoid-specific and inducible enhancer element. Proc. well as the lack of hydrophobic abundance in the N-terminal Natl. Acad. Sci. U.S.A. 85, 1482-1486. 2. Oeckinghaus, A. and Ghosh, S. (2009) The NF-kB Family ZAS3, were unexpected. However, the N-terminal 645 amino of Transcription Factors and its regulation. Cold Spring acids strongly repressed both basal and activated transcription Harb. Perspect. Biol. 1, a000034. and the result was reproducible, although there was a minor 3. Chen, G. and Goeddel, D. V. (2002) TNF-R1 signaling: a variation in the strength. It is not clear how this seemingly par- beautiful pathway. Science 296, 1634-1635. adoxical observation could explain the repression ability of the 4. Gilmore, T., Gapuzan, M. E., Kalaitzidis, D. and Starczy- N-terminal region. The simplest explanation is that other N-ter- nowski, D. (2002) Rel/NF-κB/I κB signal transduction in minal regions, aside from the ZASN domain, may have the the generation and treatment of human cancer. Cancer novel class of a repression domain, or that an unknown com- Lett. 181, 1-9. 5. Allen, C. E., Mak, C. H. and Wu, L. C. (2002) The κB tran- bination of known features, such as serine/threonine stretch scriptional enhancer motif and signal sequences of V(D)J and acidic residues, may produce novel repression domain in recombination are targets for the zinc finger protein HIVEP3/ the N-terminal 643 amino acids. The possibility is speculative KRC: a site selection amplification binding study. BMC and is only one of several that might be consistent with our Immunol. 3, 10. currently available results. Nonetheless, defining the minimal 6. Wu, L. C. (2002) ZAS: C2H2 zinc finger proteins involved http://bmbreports.org BMB reports 811 ZAS3 inhibits NFκB by DNA competition Joung-Woo Hong and Lai-Chu Wu

in growth and development. Gene Expr. 10, 137-152. 15. Kakidani, H. and Ptashne, M. (1988) GAL4 activates gene 7. Fujii, H., Gabrielson, E., Takagaki, T., Ohtsuji, M., Ohtsuji, expression in mammalian cells. Cell 52, 161-167. N. and Hino, O. (2005) Frequent down-regulation of 16. Ptashne, M. (1986) A genetic switch. Cell Press and Black- HIVEP2 in human breast cancer. Breast Cancer Res. Treat. well Scientific Publications, Cambridge, Massachusetts. 91, 103-112. 17. Ghosh, S. and Karin, M. (2002) Missing pieces in the 8. Aalto, Y., El-Rifa, W., Vilpo, L., Ollila, J., Nagy, B., Vihi- NF-κB puzzle. Cell 109(Suppl), S81-96. nen, M., Vilpo, J. and Knuutila, S. (2001) Distinct gene ex- 18. Han, K. and Manley, J. L. (1993) Transcriptional re- pression profiling in chronic lymphocytic leukemia with pression by the Drosophila even-skipped protein: defi- 11q23 deletion. Leukemia 15, 1721-1728. nition of a minimal repression domain. Genes Dev. 7, 9. Allen, C. E., Muthusamy, N., Weisbrode, S. E., Hong, J. 491-503. W. and Wu, L. C. (2002) Developmental anomalies and 19. Han, K. and Manley, J. L. (1993) Functional domains of neoplasia in animals and cells deficient in the large zinc the Drosophila Engrailed protein. EMBO J. 12, 2723- finger protein KRC. Genes Chromosomes Cancer 35, 287- 2733. 298. 20. Sigler, P. B. (1988) Transcriptional activation. Acid blobs 10. Oukka, M., Kim, S. T., Lugo, G., Sun, J., Wu, L. C. and and negative noodles. Nature 333, 210-212. Glimcher, L. H. (2002) A mammalian homolog of Droso- 21. Ptashne, M. (1988) How eukaryotic transcriptional activa- phila schnurri, KRC, regulates TNF receptor-driven re- tors work. Nature 335, 683-689. sponses and interacts with TRAF2. Mol. Cell 9, 121-131. 22. Nissen, R. M. and Yamamoto, K. R. (2000) The glucocorti- 11. Hong, J. W., Allen, C. E. and Wu, L. C. (2003) Inhibition coid receptor inhibits NFκB by interfering with serine-2 of NF-κB by ZAS3, a zinc-finger protein that also binds to phosphorylation of the RNA polymerase II carboxy-termi- the κB motif. Proc. Natl. Acad. Sci. U.S.A. 100, 12301- nal domain. Genes Dev. 14, 2314-2329. 12306. 23. Gerritsen, M. E., Williams, A. J., Neish, A. S., Moore, S., 12. Levine, M. and Manley, J. L. (1989) Transcriptional re- Shi, Y. and Collins, T. (1997) CREB-binding protein/p300 pression of eukaryotic promoters. Cell 59, 405-408. are transcriptional coactivators of p65. Proc. Natl. Acad. 13. Ayer, D. E., Laherty, C. D., Lawrence, Q. A., Armstrong, Sci. U.S.A. 94, 2927-2932. A. P. and Eisenman, R. N. (1996) Mad proteins contain a 24. Papin, S., Cazeneuve, C., Duquesnoy, P., Jeru, I., Sahali, dominant transcription repression domain. Mol. Cell Biol. D. and Amselem, S. (2003) The tumor necrosis factor α- 16, 5772-5781. dependent activation of the human mediterranean fever 14. Fleischer, T. C., Yun, U. J. and Ayer, D. E. (2003) Identifi- (MEFV) promoter is mediated by a synergistic interaction cation and characterization of three new components of between C/EBPβ and NFκB p65. J. Biol. Chem. 278, the mSin3A corepressor complex. Mol. Cell Biol. 23, 48839-48847. 3456-3467.

812 BMB reports http://bmbreports.org