Identification of NF-Κ B-Regulated Genes Induced by TNF Α Utilizing

Identiﬁcation of NF-jB-regulated genes induced by TNFa utilizing expression proﬁling and RNA interference

Anwu Zhou1, Shane Scoggin1, Richard B Gaynor 1,2 and Noelle Sevilir Williams*,1

Division of Hematology-Oncology, Department of Medicine, Harold Simmons Cancer Center, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390-8594, USA

Tumor necrosis factor a (TNFa) is a proinflammatory major groups of transcription factors, activator protein- cytokine with important roles in regulating inflammatory 1 (AP-1) and nuclear factor kB (NF-kB) (Barnes, 1997; responses as well as cell cycle proliferation and apoptosis. Karin et al., 1997). The genes induced by AP-1 and NF- Although TNFa stimulates apoptosis, it also activates the kB, such as IL-6, Rantes, 12-lipoxygenase, and IL-8, are transcription factor NF-jB, and studies have shown that involved in chronic and acute inflammatory responses, inhibition of NF-jB potentiates the cytotoxicity of TNFa. but some, such as cIAP2, Bcl-xL, and 14-3-3Z, also act Since several chemotherapy agents act like TNFa to both to suppress TNFa-induced apoptosis (Leong and promote apoptosis and activate NF-jB, understanding the Karsan, 2000; Manos and Jones, 2001). These data are role of NF-jB in suppressing apoptosis may have in accord with studies demonstrating that inhibition of significant clinical applications. To understand the effects NF-kB potentiates the cytotoxicity of TNFa. Transfec- of stimulation with TNFa and the role of NF-jBin tion of several human tumor lines with a dominant- regulating this response, a 23k human cDNA microarray negative form of the NF-kB inhibitor IkBa rendered was used to screen TNFa-inducible genes in HeLa cells. them each more susceptible to the apoptotic effects of Real-time PCR verified expression changes in 16 of these TNFa (Van Antwerp et al., 1996; Wang et al., 1996). In genes and revealed three distinct temporal patterns of addition, sulindac, an NSAID inhibitor of NF-kB expression after TNFa stimulation. Using RNA inter- activation, was shown to enhance significantly TNFa- ference to disrupt expression of the p65 subunit of NF-jB, mediated apoptosis of several lung cancer cell lines all but two of the genes were shown to depend on this (Berman et al., 2002). Since a number of standard transcription factor for their expression, which correlated chemotherapy agents act like TNFa to both promote well with the existence of NF-jB binding sites in most of apoptosis and activate NF-kB, understanding the role of their promoters. Inflammatory, proapoptotic, and anti- NF-kB in suppressing apoptosis may have significant apoptotic genes were all shown to be regulated by NF-jB, clinical applications. demonstrating the wide variety of targets activated by NF-kB transcription complexes are comprised of a NF-jB signaling and the necessity of differentiating variety of homo- and hetrodimers formed by p50, p52, among these genes for therapeutic purposes. p65 (RelA), RelB, and c-Rel subunits. The most Oncogene (2003) 22, 2034–2044. doi:10.1038/sj.onc.1206262 common form of NF-kB consists of a p50 and p65 heterodimer (Baeuerle and Henkel, 1994; Thanos and Keywords: TNFa;NF-kB; cDNA microarrays; RNA Maniatis, 1995). This complex is present predominantly interference; gene expression profiles; apoptosis in the cytoplasm, where it is bound to a group of inhibitory proteins known as inhibitors of NF-kB(IkBs) (Verma et al., 1995). In response to a variety of stimuli, such as the cytokines TNFa and interleukin-1 (IL-1), the Introduction IkB proteins are phosphorylated by the IkB kinase (IKK) signalosome, are ubiquitinated, and are subse- ONCOGENOMICS Tumor necrosis factor a (TNFa) is a potent proin- quently degraded by the proteasome (Chen et al., 1995; flammatory cytokine with important roles in control of Karin and Ben-Neriah, 2000). This process results in immune and inflammatory responses as well as cell cycle translocation of NF-kB to the nucleus, where it binds to proliferation and apoptosis (Tracey and Cerami, 1993). kB sites located in the promoter region of a variety of TNFa stimulation can result in the activation of a genes and activates their expression. caspase cascade leading to apoptosis (Chang and Yang, In this study, we systematically investigated TNFa- 2000), but more commonly causes activation of two induced gene expression in HeLa cells using cDNA microarrays. In order to identify NF-kB-regulated genes *Correspondence: NS Williams, involved in the antiapoptotic and proinflammatory E-mail: [email protected] responses to TNFa, NF-kB expression was disrupted 2Current address: Lilly Corporate Center, Eli Lilly and Company, Indianapolis, IN 46285, USA using small interfering RNA (siRNA) directed against Received 16 October 2002; revised 20 November 2002; accepted 21 its p65 subunit. TNFa-inducible expression profiles November 2002 in the presence and absence of NF-kB were then Identification of NF-jB-dependent-genes induced by TNFa A Zhou et al 2055 compared. All results were validated using real-time, TNFa induced gene expression profiles quantitative PCR. A number of known TNFa-inducible, NF-kB-dependent genes were identified in addition to Real-time PCR provides an effective and accurate several novel targets of NF-kB. As both pro- and method to quantify messenger RNA levels (Bustin, antiapoptotic genes were regulated by NF-kB, these 2000). Figure 1 shows the short-term (30 min) and long- studies demonstrate that targeting specific downstream term (4 h) effects of TNFa on the mRNA levels of the 16 genes of NF-kB may be more effective than targeting genes observed by microarray analysis to be upregulated NF-kB itself for augmenting apoptosis in response to by TNFa stimulation. Three patterns of gene expression TNFa. were noted. In the first pattern, expression of the gene was rapidly elevated at 30 min, then declined at 4 h (Figure 1a). This group included seven genes: IL-6, GRO-1, Egr-1, TIEG, ETR-101, MKP-1, and JunB. In the second pattern, expression of the gene increased Results more slowly, with elevated levels noted only at 4 h (Figure 1b). This group included MCP-1, B94, SOD-2, Microarray analysis of gene expression induced by TNFa NK4, and PGES. In the third pattern, the expression To investigate systematically genes that are induced by of the gene increased rapidly and maintained a high TNFa, a custom-made 23k human cDNA microarray level of expression relative to unstimulated cells was used. HeLa cells were stimulated with TNFa for (Figure 1c). This group included A20, IkBa, COX-2, either 30 min or 4 h. Cells were harvested and total RNA and syndecan-4. was extracted and reverse transcribed into cDNA using Cy5 (unstimulated)- and Cy3(stimulated)-labeled dCTP NF-kB dependence of TNF-a-induced genes for two-color microarray analysis. Two independent experiments were performed, and each experiment was To analyse the role of NF-kB in the TNFa-induced hybridized to two microarrays, for a total four arrays expression of these 16 genes, RNA interference (RNAi) per sample pair. The array data were subject to a simple was utilized to disrupt expression of the p65 subunit of algorithm (see Materials and methods) to set a lower the NF-kB transcription factor. RNAi is evolutionarily bound threshold and to normalize for unequal incor- conserved in plants, worms, and flies, and recently has poration of label. Then, average ratios and their been demonstrated to function in mammalian cells standard deviations (s.d.) were calculated. Genes with (Elbashir et al., 2001). The use of RNAi in mammalian an s.d. less than the average ratio were sorted in cells involves the transfection of an annealed 21-mer of descending order of the average ratios. Two upregulated sense and antisense RNA oligonucleotides (siRNAs) gene lists were obtained for each time point (30 min and corresponding to a portion of a gene of interest. These 4 h after TNFa stimulation). The first list identified 59 RNAs then specifically bind to the cellular RNA and genes with an average ratio of at least threefold activate a process that leads to degradation of the upregulation after 30 min of TNFa stimulation mRNA and a subsequent 80–90% decrease in the levels (Table 1). The second list showed 61 genes with an of the corresponding protein (Caplen et al., 2001; average ratio of at least 2.5-fold upregulation after 4 h of Harborth et al., 2001; Dudley et al., 2002). This analysis TNFa stimulation (Table 2). There were no genes at was performed because NF-kB is one of the two major either time point that were downregulated by TNFa by transcription factors activated by TNFa and is the Xtwofold. primary regulator of the antiapoptotic response induced From these two lists, a number of genes were chosen by TNFa (Krikos et al., 1992; Wu et al., 1998; Baud and for verification and further analysis. Using the MarC-V Karin, 2001). HeLa cells were evenly divided into two software, only those genes showing a confidence score of groups. One group was transfected with annealed, 35% or higher were considered. This excludes genes double-stranded RNA oligonucleotides (siRNA) direc- whose mean intensity was low and thus had a higher ted against NF-kB p65, while the other group was level of uncertainty associated with them. In addition, a treated with only oligofectamine as a control. Previous cluster program developed by Michael Eisen (Eisen studies showed that unrelated siRNAs, including those et al., 1998) was used to give a gene priority if it directed against the HTLV-I tax gene and the HIV-1 clustered together with known TNFa-induced genes reverse transcriptase gene, do not alter the TNFa (data not shown). Each gene was first sequenced to response and give equivalent results to oligofectamine confirm its identity and then the array data were alone. In addition, expression of over 14 genes, validated using SYBR Green quantitative, real-time including p-65, IKKa, IKKb, cyclin D1, APC, and c- PCR. A total of 16 confirmed genes were induced by Jun, has been examined in the presence of siRNA TNFa at either 30 min or 4 h, or at both time points, as directed against tax or with oligofectamine alone, and shown in Table 3. These genes included GRO-1, IL-6, there is no difference in expression levels (Takaesu and MCP-1, Egr-1, JunB, MKP-1, IkBa, A20, B94, synde- Gaynor, unpublished data). Thus, cells transfected with can-4, SOD-2, COX-2, PGES, TIEG, NK4, and siRNA against p65 or with oligofectamine alone were ETR101. While most of the genes are known to be treated with TNFa for 30 min and 4 h. Figure 2 shows induced by TNFa, the last three have not previously that the expression of p65 was blocked in the p65 RNAi- been reported to be TNFa inducible. treated cells and that its expression was not affected by

Oncogene Oncogene 2056

Table 1 Genes upregulated after 30 min stimulation of HeLa cells with TNFaa GenBank Gene name Average fold change GenBank Gene name Average fold change accession # TNFa stimulated accession # TNFa stimulated versus unstimulated versus unstimulated W46900 GROI oncogene (melanoma growth-stimulating 7.1 AA486533 Early growth response 1 (Egr-1) 3.6

activity, alpha) (GRO-1) NF- of Identiﬁcation AA486533 Early growth response 1 (Egr-1) 6.2 T52830 Homo sapiens, clone IMAGE 4183312, mRNA, 3.5 partial cds W55872 Nuclear factor of kappa light polypeptide gene 6.1 AA629591 Testis enhanced gene transcript (BAX inhibitor 1) 3.5 enhancer in B-cells inhibitor, alpha AA149095 Dual speciﬁcity phosphatase 1 (MKP-1) 5.7 AA157813Interferon, alpha-inducible protein 27 3.5 T67270 Ribosomal protein L10 5.3 AA252318 TATA element modulatory factor 1 3.5 j

N98591 Interleukin 6 (IL-6) 5.2 N94468 junB proto-oncogene (JunB) 3.5 TNF by induced B-dependent-genes W47485 Sigma receptor (SR31747 binding protein 1) 5.2 T97593 Heterogeneous nuclear ribonucleoprotein U 3.5 (scaffold attachment factor A) AA457697 High-density lipoprotein binding protein (vigilin) 5.0 W44860 Cell death-regulatory protein GRIM19 3.5 W56300 Nuclear factor of kappa light polypeptide gene 5.0 AA464627 Proteolipid protein 2 (colonic epithelium-enriched) 3.4 enhancer in B-cells inhibitor, alpha (IjBa) AA045731 TGFB-inducible early growth response (TIEG) 4.9 AA427688 Protein phosphatase 2 (formerly 2A). regulatory 3.4 subunit A (PR 65), alpha isoform AA682905 Huntingtin-associated protein interacting protein 4.8 AA600217 Activating transcription factor 4 (tax-responsive 3.4 (duo) enhancer element B67) Zhou A R11526 Parathymosin 4.7 AA457092 Serologically deﬁned breast cancer antigen 84 3.4

H69335 Pirin 4.5 AA406027 CD5 antigen (p56–62) 3.4 al et

H44051 Similar to keratin, type 1 cytoskeletal 14 4.4 AA497020 Heat shock protein 75 3.3 a AA64421 Cyclooxygenase 2 4.2 AA115310 ESTs. Highly similar to 2209333A protein disulfide 3.3 isomerase [H. sapiens] AA453577 Solute carrier family 39 (zinc transporter), member 3 4.2 AA633757 Splicing factor 3b, subunit 2, 145 kDa 3.3 W56300 Nuclear factor of kappa light polypeptide gene 4.2 AA461476 Putative nucleolar RNA helicase 3.3 enhancer in B-cells inhibitor, alpha (IjBa) AA496025 MCM7 minichromosome maintenance deficient 7 4.1 H08564 Transgelin 2 3.2 (Saccharomyces cervisiae) AA496359 Immediate early protein (ETR101) 3.9 AA489343 Proteasome (prosome, macropain) subunit, beta 3.2 type 7 H68845 Peroxiredoxin 2 3.9 AA454963 Mitochondrial ribosomal protein L36 3.2 N53172 G protein-coupled receptor 3.8 AA455126 ATP synthase, H+ transporting, mitochondrial F0 3.2 complex, subunit c (subunit 9) isoform 2 AA476272 TNFa-induced protein 3 (A20) 3.8 W55872 Nuclear factor of kappa light polypeptide gene 3.2 enhancer in B-cells inhibitor, alpha N91311 Peroxiredoxin 5 3.7 AA448289 D123 gene product 3.1 AA055656 Prohibition 3.7 A1014487 Cysteine-rich, angiogenic inducer 61 3.1 AA442092 Unknown 3.7 R44740 Mitogen-activated protein kinase kinase 1 3.1 A1000256 Tubulin, beta, 2 3.7 AA448261 High-mobility group (nonhistone chromosomal) 3.1 protein isoforms I and Y AA180742 Tubulin, alpha, 1 (testis specific) 3.7 N70878 Heat shock 27 kDa protein I 3.0 AA005153 PDZ domain protein (Drosophila in a D-like) 3.6 AA625651 Thyroid receptor interacting protein 15 3.0 R43766 Eukaryotic translation elongation factor 2 3.6 AA485913 Chloride intracellular channel I 3.0 T67270 Ribosomal protein L10 3.6

aGenes reliably upregulated by Xthreefold using microarray analysis are shown. Boldface genes were validated by quantitative PCR Table 2 Genes upregulated after 4 h stimulation of HeLa cells with TNFaa GenBank Gene name Average fold change GenBank Gene name Average fold change accession # TNFa stimulated accession # TNFa stimulated versus unstimulated versus unstimulated AA425102 Small inducible cytokine A2 5.0 AA147214 Growth arrest and DNA-damage-inducible. alpha 2.8 (monocyte chemotactic protein 1) (MCP-1) AA451716 Nuclear factor of kappa light 4.7 W46900 GRO1 oncogene (melanoma growth stimulating activity, 2.8 polypeptide gene enhancer in B-cells 1 (p 105) alpha) (GRO-1) AA682905 Huntingtin-associated protein 4.4 W68403ESTs. Highly similar to IJHULM leukocyte adhesion 2.8 interacting protein (duo) protein beta chain [H. sapiens] AA489343 Proteasome (prosome, macropain) 4.2 AA775803Peroxredoxin 1 2.7 subunit, beta type 7 W55872 Nuclear factor of kappa light 4.0 AA425102 Small inducible cytokine A2 (monocyte chemotactic 2.7 polypeptide gene enhancer in B-cells protein 1) (MCP-1) inhibitor, alpha T67270 Ribosomal protein L10 3.9 W67174 Integrin beta 1 (fibronectin receptor, beta polypeptide, 2.7 antigen CD29 includes MDF2. MSK12) AA634261 Chloride intracellular channel 4 3.7 H20743 Cell division cycle 34 2.7 AA457114 Tumor necrosis factor, 3.7 AA180742 Tubulin, alpha 1 (testis specific) alpha-induced protein 2 (B94) T67270 Ribosomal protein L10 3.6 AA280514 Translocated promoter region (to activated MET 2.7 oncogene) AA284492 SWI/SNF Tetraspan related, 33.5AA599120 matrix associated, actin-dependent 2.7 regulator of chromatin, subfamily e member 1 W56300 Nuclear factor of kappa 3.5 T64483Nef-associated factor 1 2.7 light polypeptide gene enhancer in B-cells inhibitor, alpha (IjBa) AA457114 Tumor necrosis factor, alpha-induced 3.3 AA488084 Superoxide dismutase 2, mitochondrial (SOD-2) 2.7 protein 2 (B94) AA458965 Natural killer cell transcript 4 (NK4) 3.3 H85464 Deleted in split-hand/split-foot 1 region 2.7 AA181300 Proteasome (prosome, macropain) subunit beta AA451716 Nuclear factor of kappa light polypeptide gene enhancer 2.6 type, 8 (large multifunctional protease 7) in B-cells 1 (p105) AA683050 Ribosomal protein S8 3.2 AA476294 Nucleolin 2.6 Zhou A NF- of Identification R17811 Splicing factor 3a, subunit 3, 3.2 A1954996 Myosin regulatory light chain 2.6 60 kDa R92806 GDP dissociation inhibitor 2 3.0 N20798 V-kit Hardy–Zuckerman 4 feline sarcoma viral onco- al et gene homolog H70017 Mannosidase, alpha, class 2A, 3.0 W56300 Nuclear factor of kappa light polypeptide gene enhancer 2.6 member 1 in B-cells inhibitor, alpha (IjBa) AA488346 Myosin, light polypeptide 6, alkali, 3.0 AA487608 Homolog of yeast long chain polyunsaturated fatty acid 2.6 j

smooth muscle and nonmuscle elongation enzyme 2 TNF by induced B-dependent-genes AA459956 Putative ribonuclease III 3.0 AA487466 ESTs. Highly similar to 2102231B Orn decarboxylase 2.6 antizyme [H. sapiens] AA629801 UMP-CMP kinase 2.9 AA629796 H. sapiens cDNA FLJ22128 ﬁs, clone HEP195432.6 AA418811 Fibrillin I (Marfan syndrome) 2.9 T67270 Ribosomal protein L10 2.6 AA961752 HSPC037 protein 2.9 H62473 Transforming growth factor, beta receptor III (betagly- 2.5 can, 300 kDa) A1540460 Prostaglandin E synthase (PGES) 2.9 R20379 Eukaryotic translation elongation factor 2 2.5 T75522 Protein phosphatase 1, 2.8 AA011415 Kell blood group 2.5 regulatory subunit 10 H60549 CD59 antigen p18-20 2.8 N67038 Emopamil binding protein (sterol isomerase) 2.5 (antigen identiﬁed by monoclonal antibodies 16 3A5, EJ16, EJ30, EL32 and G344) AA464627 Proteolipid protein 2 2.8 AA488433 Chromosome 20 open-reading frame 45 2.5

(colonic epithelium-enriched) a AA599178 Ribosomal protein L27a 2.8 AA451686 Cyclin C 2.5 AW070893DNA segment on chromosome X 2.8 T68518 Protein-l-isoaspartale (d-asparate) O-methyltransferase 2.5 (unique) 9928 expressed sequence A1014487 Cysteine-rich angiogenic inducer, 61 2.8 AA634008 ESTs. highly similar to 40S ribosomal protein S23 [H. 2.5 sapiens] H08548 ATP citrate lyase 2.8

aGenes upregulated by X2.5-fold using microarray analysis are shown. Boldface genes were validated by quantitative PCR Oncogene 2057 Identification of NF-jB-dependent-genes induced by TNFa A Zhou et al 2058 Table 3 Genes identified by microarrays and confirmed by real-time, quantitative PCR to be upregulated by TNFa GenBank Gene name Abbreviation Function Reference accession # AA045731 TGFB-inducible early growth TIEG Zinc-finger transcription factor, None response proapoptotic AA148736 Syndecan-4a Syndecan-4 Endothelial cell growth Zhang et al., 1999 AA149095 MAP kinase phosphatase 1 MKP-1 Antiapoptotic Guo et al., 1998 AA425102 Small inducible cytokine A2 MCP-1 Monocyte chemoattractant protein Kakizaki et al., 1995 AA457114 TNFa-induced protein 2 B94 TNFa-induced primary response gene Sarma et al., 1992 AA458965 Natural killer cell transcript 4 NK4 IL-2-induced, unknown function None AA476272 TNFa-induced protein 3A20 Zinc-finger protein, antiapoptosis Opipari et al., 1992 AA486533 Early growth response 1 Egr-1 Transcription factor, proapoptosis Chaudhary et al., 1996; Grimbacher et al., 1998 AA488084 Superoxide dismutase 2 SOD-2 Antioxidative, antiapoptosis Antras-Ferry et al., 1997; Maehara et al., 2000 AA496359 Human transcription factor ETR 101 Immediate early gene None ETR101 AA64421 Cyclooxygenase 2 COX-2 Proinflammatory process Yamamoto et al., 1995 A1540460 Prostaglandin E synthase PGES Proinflammatory process Stichtenoth et al., 2001 N94468 JunB proto-oncogene JunB Immediate early gene Cornelius et al., 1990; Stephens et al., 1992 N98591 Interleukin 6 IL-6 Proinflammatory process Stephens et al., 1992 W46900 GRO-1 oncogene GRO-1 Proinflammatory process Haskill et al., 1990 W56300 IkBa IkBa NF-kB inhibitor de Martin et al., 1993

aAlthough syndecan-4 was upregulated at 30 min by othreefold by microarray analysis, cluster analysis demonstrated that it behaved similar to other TNFa-induced genes, and quantitative PCR validated that it was signiﬁcantly induced by TNFa

TNFa in the control group. Real-time PCR was then melanomas and also promotes chemotaxis of granulo- used to quantify accurately differences in the expression cytes and endothelia (Dong et al., 1999). Although levels of the 16 TNFa-stimulated genes after treatment increases in GRO-1 mRNA levels were noted within with siRNA directed against p65. Figure 3shows the 30 min of TNFa stimulation, GRO-1 protein levels were relative amount of mRNA from the 16 genes. The elevated signiﬁcantly only at 4 h of stimulation. This is results indicate that the expression levels for 12 genes likely because of the time necessary for translation, were signiﬁcantly blocked by p65 siRNA at both the 30- secretion, and accumulation of this protein. As shown in min and 4-h points, while expression levels of PGES and Figure 4b, p65 RNAi effectively blocked GRO-1 JunB were only inhibited at the 4-h time point. expression, consistent with the dependence of GRO-1 Expressions of Egr-1 and MKP-1 were not affected at on NF-kB as noted by the quantitative real-time PCR either time point by p65 siRNA. These results indicated analysis. that the majority of the TNFa-inducible genes were regulated by NF-kB. Promoter analysis of TNFa-inducible genes

TNFa-induced, NF-kB-dependent protein expression NF-kB regulates gene expression through binding to promoter elements found in a variety of cellular genes Western blot analysis and ELISA were used to confirm with the DNA sequence GGGYNNRRCC (Le Bail that changes in gene expression for three of the genes et al., 1993). An attempt was made to correlate the noted by real-time PCR to be elevated by TNFa and existence of NF-kB binding sites in the promoter region dependent on NF-kB were consistent with changes at of each of the genes induced by TNFa with their the protein level. It is known that in normal cells, IkBa regulation by siRNA directed against p65. As listed in protein is degraded immediately after TNFa stimulation Table 4, the promoter regions of most genes, for which and then is transcribed in response to NF-kB activation, siRNA directed against p65 showed an inhibitory effect, leading to increased protein levels. However, as shown are known to contain one or more functional NF-kB in Figure 4a, after the initial decrease, IkBa protein binding sites. These genes include A20, COX-2, GRO-1, levels never increased in the cells treated with siRNA IkBa, IL-6, MCP-1, SOD-2, and Syndecan-4. Several directed against p65. This confirms that re-expression of genes, including B94, ETR101, NK4, and TIEG, have IkBa after TNFa stimulation is dependent on the NF- putative NF-kB binding sites in their 50 flanking region, kB pathway. COX-2 protein levels were also examined as determined by a search of each of these sequences by Western blot. COX-2 protein levels were elevated in with the GeneRunner software. As siRNA results response to TNFa, and these increases were partially indicated that these genes were dependent on NF-kB blocked in cells treated with siRNA directed against for their expression, the putative NF-kB promoter p65, consistent with the real-time PCR data. Finally, elements are likely functional. Although JunB contains ELISA was used to quantify the levels of GRO-1, which several NF-kB sites in a downstream enhancer region is a secreted protein that serves as a growth factor for (Brown et al., 1995), it appeared to be minimally

Oncogene Zhou A NF- of Identiﬁcation tal et j -eedn-ee nue yTNF by induced B-dependent-genes a

Figure 1 TNFa-inducible gene expression profiles in HeLa cells. HeLa cells were stimulated with 20 ng/ml TNFa for 30 min or 4 h, and RNA was extracted. This RNA was used to prepare cDNA that was subject to real-time PCR to quantify the mRNA level of each gene. Each sample was run in triplicate, and the relative amount of mRNA was normalized to the 18S RNA content in each sample. Normalized expression values (specific gene expression levels/18s expression levels) are presented. Column 1: unstimulated cells; column 2: 30 min after TNFa stimulation; column 3: 4 h after TNFa stimulation. Data points are represented by the mean7s.d. of triplicate wells. Each RNA sample represents a pool of RNA samples from two independent biological experiments Oncogene 2059 Identification of NF-jB-dependent-genes induced by TNFa A Zhou et al 2060 dependent on NF-kB for its expression. No NF-kB Discussion binding sites were found in the sense orientation in the upstream region of the PGES gene, despite the fact that TNFa, an inflammatory cytokine with cytotoxic proper- this gene appeared to be dependent on NF-kB expres- ties, acts like many standard chemotherapeutic agents to sion at the 4-h time point. It is possible that like JunB, it both promote apoptosis and activate the transcription contains a downstream enhancer that binds NF-kBor factor NF-kB. Since several studies have shown that that its dependence on NF-kB is indirect. The upstream inhibition of NF-kB potentiates the cytotoxicity of TNFa, regions of both Egr-1 and MKP-1 contain NF-kB sites understanding the mechanism by which NF-kB carries but the siRNA analysis did not indicate a dependence out this function may be of clinical significance. If on NF-kB for their expression. It is possible that these sites are not functional or that they are not necessary for TNFa-induced expression of these two genes. Thus, these data show that NF-kB dependence, as revealed by the RNAi experiments, correlates well in cases in which a gene has been shown to have a functional NF-kB promoter element (i.e. A20, COX-2, GRO-1, IkBa, IL-6, MCP-1, SOD-2, and Syncean-4). For several additional genes, our sequence and p65 RNAi data provide the first evidence of the existence of functional NF-kB elements in their promoters.

Figure 3 siRNA directed against p65 alters TNFa-induced gene expression. HeLa cells were transfected with 50 nm annealed, double-stranded p65 siRNA oligonucleotides or oliogofectamine alone for 72 h. The cells were then stimulated with 20 ng/ml TNFa for 30 min and 4 h. The horizontal dashed line indicates the level of expression of the gene after TNFa stimulation and in the presence Figure 2 RNA interference inhibits p65 expression in HeLa cells. of p65, which was arbitrarily set to 1. The two columns for each HeLa cells were transfected with 50 nm annealed, double-stranded gene represent expression levels of each mRNA following TNFa p65 siRNA oligonucleotides or oligofectamine alone for 72 h. The stimulation for either 30 min or 4 h in the presence of siRNA cells were then stimulated with 20 ng/ml TNFa for 0, 0.5, or 4 h. directed against p65, relative to expression levels in control-treated The cells were lysed with PD buffer, and 100 mg of protein was cells. Stars indicate a signiﬁcant inhibitory effect of the absence of loaded in each lane. SPT5 was used as an internal control p65 (Po0.05) relative to the controls, according to Student’s t-test

Figure 4 siRNA directed against p65 alters TNFa-induced gene expression at the protein level. (a). HeLa cells were transfected with 50 nm annealed, double-stranded p65 siRNA oligonucleotides or oliogofectamine alone for 72 h. The cells were then stimulated with 20 ng/ml TNFa for 0, 0.5, 1, or 4 h. The cells were lysed with PD buffer, and 100 mg of protein was loaded in each lane with b-tubulin used as an internal control. (b) HeLa cells were transfered with 50 nm annealed, double-stranded p65, siRNA oligonucleotides or oligofectamine alone for 72 h and were stimulated by 20 ng/ml TNFa for 0, 0.5, 1, and 4 h. Cell culture media (200 ml) from each condition were used to measure the concentration of GRO-1 protein following protocols provided with the kit. All samples along with recombinant human GRO-1 standards were run in duplicate. The mean7s.d. from two separate experiments is shown

Oncogene Identification of NF-jB-dependent-genes induced by TNFa A Zhou et al 2061 Table 4 Summary of NF-kB binding sites in gene promoter regions Gene name Known elements in gene promoter NF-kB elementa Clone Reference accession #b A20 2 NF-kB binding sites Functional N/A Krikos et al., 1992 COX-2 CRE, ATF-1, NF-IL6, AP2, NF-kB Functional N/A Yamamoto et al., 1995 GRO-1 NF-kB, IUR Functional N/A Wood et al., 1995 IkBa NF-kB binding site Functional N/A Le Bail et al., 1993 IL-6 Ap-1, CRE, C/EBP, NF-kB Functional N/A Vanden Berghe et al., 2000 MCP-1 Two NF-kB binding site Functional N/A Ueda et al., 1997 SOD-2 Egr-1, Sp1, NF-kB binding site Functional N/A Das et al., 1995; Tanaka et al., 1999; Maehara et al., 2001; Syndecan-4 NF-kB binding site Functional N/A Zhang et al., 1999 JunB NF-Y, ETS, STAT, APRF/NF-kB Functional enhancer N/A Brown et al., 1995; Coffer et al., 1995; Finch et al., 2001 B94 Promoter sequence unknown À2362*, À2708*, À2718 AL161669 None ETR101 SRE À313 AC011446 Chung et al., 2000 NK4 Promoter sequence unknown À497*, À508, À850*, À860 AC108134 None TIEG Sp1, JunB À339*, À1849*, À1859 AP002851 Fautsch et al., 1998 Egr-1 CRE, SRE À1408*, À1418, À1498 AC113403 Sakamoto et al., 1994; Tsai et al., 2000 MKP-1 CRE, AP1, AP2, Sp1 À1307*, À1449* AC008654 Kwak et al., 1994 PGES GC boxes, tandem barbie boxes, À1259* AL592219 Forsberg et al., 2000 AHR element aIf no NF-kB element was previously described, the gene was subject to a computer motify search, and the distance of the identified site from the ATG start codon is identified. A star indicates the element was found on the reverse strand. bAccession numbers of the chromosome clone that contains the upstream sequence of the gene used for the computer motif search if applicable (N/A: not applicable) specific downstream targets of NF-kB were identified NF-kB regulates the expression of genes whose function that were critical for its antiapoptotic effects, specific includes inflammation (COX-2, IL-6, GRO-1, PGES, chemotherapeutic agents that targeted these genes could and MCP-1), apoptosis or decreased cell growth (TIEG, be given in concert with standard agents for more JunB), survival or cell growth (Syndecan-4, A20, SOD- effective induction of apoptosis in the tumor. To better 2, and COX-2), and a variety of other functions (B94, understand the downstream effects of stimulation with NK4, IkBa, and ETR101), including hematopoiesis and TNFa and the role of NF-kB in regulating the response feedback regulation of the NF-kB pathway. Since genes to TNFa a 23k human cDNA microarray was used to with opposing functions are activated, in order to systematically screen TNFa-inducible genes in HeLa augment apoptosis in response to TNFa, it may be cells. Quantitative, real-time PCR verified the expression more effective to target the downstream antiapoptotic changes in 16 of these genes. Using RNA interference to genes controlled by NF-kB rather than NF-kB itself. In disrupt expression of the p65 subunit of the transcrip- fact, in these HeLa cells, disruption of NF-kB expres- tion factor NF-kB, all but two of the genes were shown sion by siRNA directed against p65 did not significantly to depend on NF-kB for their expression. Inflammatory, enhance apoptosis in response to TNFa (data not proapoptotic, and antiapoptotic genes were dependent shown). Experiments examining the role of the identified on NF-kB, demonstrating the wide variety of targets of antiapoptotic genes in regulating TNFa-induced apop- the NF-kB signaling pathway and the necessity of tosis are now under way. Second, all genes with a differentiating among these genes for therapeutic proposed role in feedback regulation of the NF-kB purposes. pathway, such as A20 and IkBa, are induced early. Three patterns of gene expression were noted. In the Third, the delayed kinetics of expression of SOD-2 after first pattern, expression of the gene was rapidly elevated TNFa stimulation may be due to a dependence on Egr-1 at 30 min following TNFa stimulation and then declined for its transcription as is seen in other systems. For by 4 h post-treatment. This group included seven genes: example, after stimulation with PDGF, it is known that IL-6, GRO-1. Egr-1, TIEG, ETR101, MKP-1, and SOD-2 expression is regulated by Egr-1 (Maehara et al., JunB. In the second pattern, expression of genes 2001). Finally, the kinetics of COX-2 and PGES increased more slowly with elevated levels noted only expression correlate well with their known roles in at 4 h (Figure 1b). This group included MCP-1, B94, prostaglandin synthesis. COX-2, which is expressed SOD-2, NK4, and PGES. In the third pattern of TNFa- earlier than PGES, is needed for an earlier step in inducible gene expression, there was rapid and sustained prostaglandin synthesis than is PGES. upregulation of the gene after stimulation. This group Overall, the results obtained in this study are twofold. included A20, IkBa, syndecan-4, and COX-2. First, this study systematically screened TNFa-induced Although no clear pattern emerges from an analysis genes using cDNA microarrays, and confirmed, using of the function of these genes and the timing of their real-time PCR, 16 genes including three novel targets expression, there are a few points worth noting. First, that are upregulated following TNFa stimulation. A

Oncogene Identification of NF-jB-dependent-genes induced by TNFa A Zhou et al 2062 number of genes identified here agree with the results of Array hybridization and scanning a recent study examining TNFa-induced genes in human Total RNA (25 mg) from sample/control was reverse tran- vein endothelial cells (Murakami et al., 2000), including scribed into cDNA using Superscript II (Invitrogen) with oligo A20, B94, IkB, and members of the GRO and dT primers (Invitrogen), and Cy3-(sample)- and Cy5-(control)- interleukin families. Second, by using siRNA directed labled dCTP (Amersham, Piscataway, NJ, USA). After against p65, the dependence of these genes on the purification through an AutoSeq G-50 column (Amersham), transcription factor NF-kB after TNFa induction was and concentration with a Microcon centrifugal filter (Milli- investigated. The results correlated well with the pore, Bedford, MA, USA), the cDNA’s were combined and existence of NF-kB binding sites in the promoter regions hybridized to the cDNA array in a solution containing 10 mg of these genes. As both pro- and antiapoptotic genes are Cot-1 DNA (Invitrogen), 32 mg tRNA (Sigma, St Louis, MO, regulated by NF-kB, these studies demonstrate that USA), 10 mg polyA RNA (Sigma), 4 Â SSC and 0.25% SDS. After an overnight hybridization at 621C, slides were washed, targeting specific downstream genes of NF-kB may be scanned using a GenePix 4000B Scanner (Axon Instruments, more effective than targeting NF-kB itself for augment- Union City, CA, USA) and analysed by GenePix Pro software ing apoptosis in response to TNFa. bundled with the scanner. Data were normalized using spreadsheet-based software (MarC-V) developed at UT South- western Medical Center (Schageman et al., 2002) to control for uneven incorporation of dyes. In brief, a lower-bound thresh- Materials and methods olding calculation is first performed such that no background subtracted intensity value is less than a threshold value. This SiRNAi transfection and RNA preparation threshold is calculated for both the Cy3and Cy5 channels HeLa cells used for siRNA transfection were cultured in based on many internally replicated negative controls that DMEM (Gibco BRL, Rockville, MD, USA), supplemented consist of SSC solvent alone. Once the threshold has been with 10% fetal bovine serum (FBS) (Gibco BRL), 2 mm applied, ratios are computed for each element and log l-glutamine, 50 U/ml of penicillin, and 50 mg/ml of strepto- transformed (base 10). Finally, a ratio normalization step is mycin. Approximately one million cells were plated per performed. Each log ratio is normalized by multiplying each six-well plate in DMEM with 10% FBS and 1% glutamine. denominator by a normalization coefficient defined as 10 raised The following day, siRNAs were transfected using oligofecta- to the mean log ratio of the entire array. Average ratios from mine (Invitrogen, Carlsbad, CA, USA) at a final RNA duplicated arrays and their standard deviations (s.d.) were concentration of 50 nm per well. The sense p65 oligonucleotide calculated. Genes with an s.d. less than the average were sorted (50-GCCCUAUCCCUUUACGUCdTdT-30) along with its in descending order of the average ratios. Further verification corresponding antisense oligonucleotide was synthesized and analysis was then focused on the resulting genes. and annealed according to the manufacturer’s instructions (Dhamacon Research, Lafayette, CO, USA). After 72 h, the cells were stimulated with 20 ng/ml TNFa Quantitative real-time PCR (Roche, Indianapolis, IN, USA) for various amounts of For validation of array results, cDNA was prepared from a time. The cells were harvested, total RNA was extracted pool of total RNA of two independent experiments, using a using the RNeasy kit (Qiagen, Valenica, CA, USA) and combination of oligo-dT and random primers and Superscript treated with the DNA-free kit (Ambion, Austin, TX, USA) II (all from Invitrogen). PCR primers for each gene were according to the manufacturer’s instructions to remove designed using Primer Express software (Applied Biosystems, residual genomic DNA. Foster City, CA, USA), with a melting temperature at 58–601C and a resulting product of approximatley 100 bp. Each PCR was carried out in triplicate in a 20 ml volume using SYBR cDNA microarrays Green Master Mix (Applied Biosystems) for 15 min at 951C for Three clone sets consisting of approximately 23000 cDNA initial denaturing, followed by 40 cycles of 951C for 30 s and clones were purchased from Research Genetics (Huntsville, 601C for 30 s in the ABI Prism 7700 sequence Detection AL, USA). The first set contained 9592 individual clones from System. cDNA prepared from Universal RNA (Stratagene, La Research Genetics’ Human Sequence Verified Set plates 1- Jolla, CA, USA) was used to construct a standard curve for 100. The second set consisted of approximately 5700 cancer- each gene. Values for each gene were normalized to expression related genes chosen by our group from the National Cancer levels of 18S RNA. The primers used for PCR analysis were: Institute Cancer Genome Anatomy Project (CGAP) website. 18S forward, 50-AGGAATTGACGGAAGGGCAC-30, re- The third set consisted of the 7500-gene Sequence Veri- verse, 50-GGACATCTAAGGGCATCACA-3; A20 forward, fied Human Known Gene Set from Research Gentics. Plas- 50-CTGCCCAGGAATGCTACAGATAC-30, reverse, 50-GT mid DNA was prepared by using the Real Prep 96 kit GGAACAGCTCGGATTTCAG-30; COX-2 forward, 50-CAC (Qiagen). Clone inserts were amplified by PCR using CCATGTCAAAACCGAGG-30, reverse, 50-CCGGTGTTGA vector-specific primers (universal forward 50-CTGCAAGGC- GCAGTTTTCTC-30; GRO-1 forward, 50-AGGAAGCTC GATTAAGTTGGGTAAC-30, universal reverse 50-GTGAG- ACTGGTGGCTG-30, reverse, 50-TAGGCACAATCCAGG CGGATAACAATTTCACACAGGAAACAGC-30, pCMV- TGGC-30;IkBa forward, 50-GATCCGCCAGGTGAAGG SPORT6 forward 50-ACGGCCAGTGCCTAGCTTAT-30, G-30, reverse, 50-GCAATTTCTGGCTGGTTGG-30; IL-6 for- and pCMV-SPORT6 reverse 50-CACAGGAAACAGCTAT- ward, 50-AATTCGGTACATCCTCGACGG-30, reverse, 50- GACCA-30). PCR products were precipitated with isopropa- GGTTGTTTTCTGCCAGTGCC-30; JunB forward, 50-GG nol, resuspended in 3 Â SSC, and printed on polylysine-coated AACAGCCCTTCTACCACG-30, reverse, 50-GGCTCGGTT glass slides using a BioRobotics (Woburn, MA, USA) TCAGGAGTTTG-30; MCP-1 forward, 50-TCGCCTCCAG- MicroGrid II arrayer. Printed slides were crosslinked by UV CATGAAAGTC-30, reverse, 50-GGCATTGATTGCATCT irradiation and postprocessed by conventional methods GGC-30; SOD-2 forward, 50-TGCACTGAAGTTCAATG (DeRisi et al., 1997). GTGG-30, reverse, 50-CTTCCAGCAACTCCCCTTTG-30;

Oncogene Identification of NF-jB-dependent-genes induced by TNFa A Zhou et al 2063 Syndecan-4, forward, 50-GGGTGGTTGAGTGAGTGAATT sham) and developed using ECL (Amersham). A human TTC-30, reverse, 50-ACCTCAACTATTCCAGCCCCA-30; GRO-1 ELISA kit was purchased from R&D Systems Egr-1 forward, 50-CAGCAGCACCTTCAACCCTC-30, re- (Minneapolis, MN, USA) and used according to the manufac- verse, 50-CCAGCACCTTCTCGTTGTTCA-30; MKP-1 turer’s instructions. forward, 50-AGCAGAGGCGAAGCATCATC-30, reverse, 50- CCCAGCCTCTGCCGAAC-30; B94 forward, 50-TCTCAC Identification of NF-kB promoter binding sites TGTTGACCCTTTGGC-30, reverse, 50-TGACCCGCAGAA CTGGAAG-30; ETR101 forward, 50-TCCCTCAGACACA For each gene, the full-length mRNA sequence was blasted CGGACAC-30, reverse, 50-AGCTCAGCCTTATGCCTTT against genomic sequences on the National Center for CC-30; NK4 forward, 50-AAAATGCAAAATGCAGAAT- Biotechnology (NCBI) website to identify a genomic clone CAGG-30, reverse, 50-TAAGCCGCCACTGTCTCCAG-30; containing upstream sequences. A 3kb upstream sequence for PGES forward, 50-CCCCCAGTATTGCAGGAGTG-30, re- each gene was analysed with Gene Runner software (Hastings verse, 50-AGACGAAGCCCAGGAAAAGG-30; TIEG for- Software, Hastings on Hudson, NY, USA) to search for the ward, 50-CCAAAGCTCAGGCAACAAGTG-30, reverse, 50- kB motif (GGGRNNYYCC) (Le Bail et al., 1993). TTGGGCAGGTCTGGTGGT-30. Statistical analysis Western blot and ELISA analysis A Student’s t-test was used to compare the statistical Cells were lysed in PD buffer (40 mm Tris-HCl, pH 8.0, 500 mm significance of differences in gene expression levels in the sodum chloride, 0.1% NP-40, 6 nm EDTA, 6mm EGTA, 5mm presence and absence of siRNA directed against p65. A P- sodium fluoride, 1 mm sodium orthovanadate, and 5 mm b- value less than 0.05 was considered to be statistically glycerophosphate). Western blot analysis was performed significant. following conventional protocols using antibodies for p65 (sc-372, Santa Cruz Biotechnology, Santa Cruz, CA, USA), IkBa (sc-371, Santa Cruz Biotechnology), COX-2 (sc-1745, Santa Cruz Biotechnology), and b-tubulin (T 4026, Sigma). A Acknowledgements rabbit polyclonal antibody was generated against SPT5 We like to thank Dr Rama M Surabhi for the development (Ivanov et al., 2000). All antibody dilutions used were of the p65 siRNA oligonucleotides, Denise Tavana for 1 : 1000 except for COX-2, which was 1 : 100 and SPT5, which growing the HeLa cells, Udit Verma for helpful sug- was 1 : 500. After extensive washing, the membranes were gestions, and Alejandra Herrera for help with preparation incubated with either a 1 : 10 000 dilution of anti-goat horse- of the figures. This work was supported by The Harold radish peroxidase (HRP) (sc-2020, Santa Cruz Biotechnology), Simmons Cancer Center, The Grant Dove Foundation, The a 1 : 1000 dilution of anti-rabbit HRP (NA934V, Amersham), National Women’s Cancer Research Alliance, and NIH grant or a 1 : 1000 dilution of anti-mouse HRP (9A931V, Amer- CA74128.

References

Antras-Ferry J, Maheo K, Morel F, Guillouzo A, Cillard P de Martin R, Vanhove B, Cheng Q, Hofer E, Csizmadia V, and Cillard J. (1997). FEBS Lett., 403, 100–104. Winker H and Bach FH. (1993). EMBO J., 12, 2773–2779. Baeuerle PA and Henkel T. (1994). Annu. Rev. Immunol., 12, DeRisi JL, Iyer VR and Brown PO. (1997). Science, 278, 141–179. 680–686. Barnes PJ. (1997). Int. J. Biochem. Cell. Biol., 29, 867–870. Dong G, Chen Z, Kato T and Van Waes C. (1999). Cancer Baud V and Karin M. (2001). Trends Cell Biol., 11, 372–377. Res., 59, 3495–3504. Berman KS, Verma UN, Harburg G, Minna JD, Cobb MH Dudley NR, Labbe JC and Goldstein B. (2002). Proc. Natl. and Gaynor RB. (2002). Clin. Cancer Res., 8, 354–360. Acad. Sci. USA, 99, 4191–4196. Brown RT, Ades IZ and Nordan RP. (1995). J. Biol. Chem., Eisen MB, Spellman PT, Brown PO and Botstein D. (1998). 270, 31129–31135. Proc. Natl. Acad. Sci. USA, 95, 14863–14868. Bustin SA. (2000). J. Mol. Endocrinol., 25, 169–193. Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K Caplen NJ, Parrish S, Imani F, Fire A and Morgan RA. and Tuschl T. (2001). Nature, 411, 494–498. (2001). Proc. Natl. Acad. Sci. USA, 98, 9742–9747. Fautsch MP, Vrabel A, Subramaniam M, Hefferen TE, Chang HY and Yang X. (2000). Microbiol. Mol. Rev., 64, Spelsberg TC and Wieben ED. (1998) Genomics, 51, 821–846. 408–416. Chaudhary LR, Cheng SL and Avioli LV. (1996). Mol. Cell. Finch JS, Rosenberger SF, Martinex JD and Bowden GT. Biochem., 156, 69–77. (2001). Gene, 267, 135–144. Chen Z, Hagler J, Palombella VJ, Melandri F, Scherer D, Forsberg L, Leeb L, Thoren S, Morgenstern R and Jakobsson Ballard D and Maniatis T. (1995). Genes Dev., 9, 1586– P. (2000). FEBS Lett., 471, 78–82. 1597. Grimbacher B, Aicher WK, Peter HH and Eibel H. (1998). Chung KC, Kim SM, Rhang S, Lau LF, Gomes I and Ahn Rheumatol. Int., 17, 185–192. YS. (2000). Eur. J. Biochem., 267, 4676–4684. Guo YL, Baysal K, Kang B, Yang LJ and Williamson JR. Coffer P, Lutticken C, van Puijenbroek A, Klop-de Jonge M, (1998). J. Biol. Chem., 273, 4027–4034. Horn F and Kruijer W. (1995). Oncogene, 10, 985–994. Harborth J, Elbashir SM, Bechert K, Tuschl T and Weber K. Cornelius P, Marlowe M, Lee MD and Pekala PH. (1990). (2001). J. Cell. Sci., 114, 4557–4565. J. Biol. Chem., 265, 20506–20516. Haskill S, Peace A, Morris J, Sporn SA, Anisowicz A, Lee SW, Das KC, Lewis-Molock Y and White CW. (1995). Am. Smith T, Martin G, Ralph P and Sager R. (1990). Proc. Natl. J. Physiol., 269, L588–L602. Acad. Sci. USA, 87, 7732–7736.

Oncogene Identiﬁcation of NF-jB-dependent-genes induced by TNFa A Zhou et al 2064 Ivanov D, Kwak YT, Guo J and Gaynor RB. (2000). Mol. Stephens JM, Butts MD and Pekala PH. (1992). J. Mol. Cell. Biol., 20, 2970–2983. Endocrinol., 9, 61–72. Kakizaki Y, Waga S, Sugimoto K, Tanaka H, Nukii K, Stichtenoth DO, Thoren S, Bian H, Peters-Golden M, Takeya M, Yoshimura T and Yokoyama M. (1995). Kidney Jakobsson PJ and Crofford LJ. (2001). J. Immunol., 167, Int., 48, 1866–1874. 469–474. Karin M and Ben-Neriah Y. (2000). Annu. Rev. Immunol., 18, Tanaka M, Fuentes ME, Yamaguchi K, Durnin MH, 621–663. Dalrymple SA, Hardy KL and Goeddel DV. (1999). Karin M, Liu Z and Zandi E. (1997). Curr. Opin. Cell. Biol., 9, Immunity, 10, 421–429. 240–246. Thanos D and Maniatis T. (1995). Cell, 80, 529–532. Krikos A, Laherty CD and Dixit VM. (1992). J. Biol. Chem., 267, 17971–17976. Tracey KJ and Cerami A. (1993). Annu. Rev. Cell Biol., 9, Kwak SP, Hakes DJ, Martell KJ and Dixon JE. (1994). J. Biol. 317–343. Chem., 269, 3596–3604. Tsai JC, Liu L, Cooley BC, DiChiara MR, Topper JN and Le Bail O, Schmidt-Ullrich R and Israel A. (1993). Embo J., Aird WC (2000). FASEB J., 14, 1870–1872. 12, 5043–5049. Ueda A, Ishigatsubo Y, Okubo T and Yoshimura T. (1997). J. Leong KG and Karsan A. (2000). Histol. Histopathol., 15, Biol. Chem., 272, 31092–31099. 1303–1325. Van Antwerp DJ, Martin SJ, Kafri T, Green DR and Verma Maehara K, Hasegawa T and Isobe KI. (2001). J. Cell. IM. (1996). Science, 274, 787–789. Biochem., 77, 474–486. Vanden Berghe W, Vermeulen L, De Wilde G, De Bosscher K, Maehara K, Oh-Hashi K and Isobe KI. (2001). FASEB J., 15, Boone E and Haegeman G. (2000). Biochem. Pharmacol., 60, 2025–2026. 1185–1195. Manos EJ and Jones DA. (2001). Cancer Res., 61, 433–438. Verma IM, Stevenson JK, Schwartz EM, Van Antwerp D and Murakami T, Mataki C, Nagao C, Umetani M, Wada Y, Ishii Miyamoto S. (1995). Genes Dev., 9, 2723–2735. M, Tsutsumi S, Kohro T, Saiura A, Aburatani H, Wang CY, Mayo MW and Baldwin ASJ. (1996). Science, 274, Hamakubo T and Kodama T. (2000). J. Atheroscler. 784–787. Thromb., 7, 39–44. Opipari Jr AW, Boguski MS and Dixit VM. (1990). J. Biol. Wood LD, Farmer AA and Richmond A. (1995). Nucleic Chem., 265, 14705–14708. Acids Res., 23, 4210–4219. Sakamoto KM, Fraser JK, Lee HJ, Lehman E and Gasson JC. Wu MX, Ao A, Prasad KV, Wu R and Schlossman SF. (1998). (1994). Mol. Cell. Biol., 14, 5975–5985. Science, 281, 998–1001. Sarma V, Wolf FW, Marks RM, Shows TB and Dixit VM. Yamamoto K, Arakawa T, Ueda N and Yamamoto S. (1995). (1992). J. Immunol., 148, 3302–3312. J. Biol. Chem., 270, 31315–31320. Schageman JJ, Basit M, Gallardo TD, Garner HR and Shohet Zhang Y, Pasparakis M, Kollias G and Simons M. (1999). J. RV. (2002). Bio Techniques, 32, 338–344. Biol. Chem., 274, 14786–14790.

Oncogene