MUTATIONS IN STAT3 ASSOCIATED WITH HUMAN HYPER IgE SYNDROME ENHANCE NFκB AND MAPK-MEDIATED GENE EXPRESSION

Hyper IgE syndrome (HIES) is a multisystem disorder resulting in bone and immune system abnormalities. It is associated with mutations in STAT3, which disrupt protein domains responsible for transcriptional function. Patients with HIES display osteoporosis and enhanced inflammatory cytokine production similar to hematopoietic Stat3-deficient mice. Since osteoclast and inflammatory cytokine genes are NFκB targets, these observations indicate a possible deregulation of NFκB signaling in both mice and humans with STAT3-deficiency. Here, we sought to examine the role of STAT3 in the regulation of NFκB-mediated gene expression through analysis of three HIES STAT3 point mutations in both hematopoietic and non- hematopoietic cells. We found that IL-6-induced tyrosine phosphorylation of STAT3 was partially or completely abrogated by HIES mutations in the transactivation domain (V713L) or SH2 domain (V637M), respectively, in both hematopoietic and non- hematopoietic cells. By contrast, IL-6-induced tyrosine phosphorylation of an HIES mutant in the STAT3 DNA-binding domain (R382W) was intact. The R382W and V713L mutants significantly reduced IL-6-dependent STAT3 transcriptional activity in reporter gene assays. Moreover, the R382W and V637M mutants significantly diminished IL-6-responsive expression of the endogenous STAT3 target gene, Socs3, as assessed by quantitative real-time PCR (qPCR) in the RAW macrophage cell line. These observations indicate the HIES mutants dominantly suppress the transcriptional activity of wild type STAT3, albeit to varying degrees. All three HIES mutants enhanced LPS-induced expression of the NFκB target genes IL6 (IL-6), Cxcl10 (IP- 10), and Tnf (TNFα) in RAW cells, as indicated by qPCR. Furthermore, overexpression of wild type STAT3 in Stat3-deficient murine embryonic fibroblasts significantlyreduced LPS-stimulated expression of IL6, Cxcl10, and IL12p35. In addition, in aprimary murine osteoclast differentiation assay, a STAT3-specific SH2 domain inhibitor led to significantly increased levels of osteoclast-specific gene expression. These results suggest that STAT3 serves as a negative regulator of NFκB-mediated gene expression, and furthermore imply that STAT3 mutations associated with HIES contribute to the osteopenia and inflammation observed in HIES patients.

{\it cis\/} -acting elements and {\it trans\/} -acting factors that control serum amyloid A gene expression during inflammation

To understand how the serum amyloid A (SAA) genes are regulated, the cis-acting elements and trans-acting factors involved in the regulation of mouse SAA3 and rat SAA1 genes expression during inflammation were analyzed.^ To identify DNA sequences involved in the liver-specific expression of the mouse SAA3 gene, the 5$\sp\prime$ flanking region of this gene was analyzed by transient transfection studies. Results suggest that C/EBP, a liver-enriched transcription factor, plays an important role for the enhanced expression of the mouse SAA3 gene in hepatocytes.^ Transfection studies of the regulation of the expression of rat SAA1 gene indicated that a 322 bp fragment ($-$304 to +18) of the gene contains sufficient information for cytokine-induced expression of the reporter gene in a liver cell-specific manner. Further functional analysis of the 5$\sp\prime$ flanking region of the rat SAA1 gene demonstrated that a 65 bp DNA fragment ($-$138/$-$73) can confer cytokine-inducibility onto a heterologous promoter both in liver and nonliver cells. DNase I footprint and gel retardation assays identified five putative cis-regulatory elements within the 5$\sp\prime$ flanking region of the gene: one inducible element, a NF$\kappa$B binding site and four constitutive elements. Two constitutive elements, footprint regions I and III, were identified as C/EBP binding sites with region III having over a 10-fold higher affinity for C/EBP binding than region I. Functional analysis of the cis-elements indicated that C/EBP(I) and C/EBP(III) confer liver cell-specific activation onto a heterologous promoter, while sequences corresponding to the NF$\kappa$B element and C/EBP(I) impart cytokine responsiveness onto the heterologous promoter. These results suggest that C/EBP(I) possesses two functions: liver-specific activation and cytokine responsiveness. The identification of two cytokine responsive elements (NF$\kappa$B and C/EBP(I)), and two liver-specific elements (C/EBP(I) and C/EBP(III)) implies that multiple cis-acting elements are involved in the regulation of the expression of the rat SAA1 gene. The tissue-specific and cytokine-induced expression of rat SAA1 gene is likely the result of the interactions of these cis-acting elements with their cognate trans-acting factors as well as the interplay between the different cis-acting elements and their binding factors. (Abstract shortened with permission of author.) ^

Control of {\it recA\/} expression in {\it Escherichia coli}

The recA gene is essential for SOS response induction, for inducible DNA repair and for homologous recombination in E. coli. The level of recA expression is significant for these functions. A basal level of about 1000 molecules of RecA protein is sufficient for homologous recombination of the cell and is essential for the induction of the SOS response. Based on previous observations, two models regarding the origin of the basal RecA protein were postulated. One was that it comes from the leaky expression of the LexA repressed promoter. The other was that it is from another weak but constitutive promoter. The first part of this thesis is to study these possibilities. An $\Omega$ cartridge containing the transcription terminator of gene 32 of T4 phage was exploited to define a second promoter for recA expression. Insertion of this $\Omega$ cartridge downstream of the known promoter gave rise to only minor expression. Purification and N-terminus sequencing of the RecA protein from the insertion mutant did not support the existence of a second promoter. To determine whether the basal RecA is due to the leaky expression of the known LexA repressed promoter, recA expression of a SOS induction minus strain (basal level expression of recA) was compared with that of a recA promoter down mutation recA1270. The result demonstrated that there is leaky expression from the LexA repressed promoter. All the evidence supports the conclusion that there is only one promoter for both basal and induced expression levels of recA.^ Several translation enhancer sequences which are complementary to different regions of the 16S rRNA were found to exist in recA mRNA. The leader sequence of recA mRNA is highly complementary to a region of the 16S rRNA. Thus it appeared that recA expression could be regulated at post-transcriptional levels. The second part of this thesis is focused on the study of the post-transcriptional control of recA expression. Deletions of the complementary regions were created to examine their effect on recA expression. The results indicated that all of the complementary regions were important for the normal expression of recA and their effects were post-transcriptional. RNA secondary structures of wild type recA mRNA was inspected and a stem-loop structure was revealed. The expression down mutations at codon 10 and 11 were found to stabilize this structure. The conclusions of the second part of this thesis are that there is post-transcriptional control for recA expression and the leader sequence of recA mRNA plays more than one role in the control of recA expression. ^

Transcriptional repression of serum amyloid A1 by AP -2 contributes to its liver -specific expression

Studies on the transcriptional regulation of serum amyloid A1 (SAA1) gene, a liver specific acute-phase gene, identified a regulatory element in its promoter that functioned to repress (SAA1) gene transcription in nonliver cells. This silencer element interacts with a nuclear protein that is detectable in HeLa cells, fibroblasts and placental tissues but not in liver or liver-derived cells. As the expression pattern of this repressor is consistent with its potential regulatory role in repressing SAA1 expression, and that many other liver gene promoters also contain this repressor binding site, we sought to investigate whether this repressor may have a broader functional role in repressing liver genes. ^ We have utilized protein purification, cell culture, transient and stable gene transfection, and molecular biology approaches to identify this protein and investigate its possible function in the regulation of (SAA1) and other liver genes. Analyses of amino acid sequence of the purified nuclear protein, and western blot and gel shift studies identified the repressor as transcription factor AP-2 or AP-2-like protein. Using transient transfection of DNA into cultured cells, we demonstrate that AP-2 can indeed function as a repressor to inhibit transcription of SAA1 gene promoter. This conclusion is supported by the following experimental results: (1) overexpression of AP-2 in hepatoma cells inhibits conditioned medium (CM)-induced expression of SAA1 promoter; (2) binding of AP-2 to the SAA1 promoter is required for AP-2 repression function; (3) one mechanism by which AP-2 inhibits SAA1 may be by antagonizing the activation function of the strong transactivator NFκB; (4) mutation of AP-2 binding sites results in derepression of SAM promoter in HeLa cells; and (5) inhibition of endogenous AP-2 activity by a dominant-negative mutant abolishes AP-2's inhibitory effect on SAM promoter in HeLa cells. In addition to the SAM promoter, AP-2 also can bind to the promoter regions of six other liver genes tested, suggesting that it may have a broad functional role in restricting the expression of many liver genes in nonliver cells. Consistent with this notion, ectopic expression of AP-2 also represses CM-mediated activation of human third component of complement 3 promoter. Finally, in AP-2-expressing stable hepatoma cell lines, AP-2 inhibits not only the expression of endogenous SAA, but also the expression of several other endogenous liver genes including albumin, α-fetoprotein. ^ Our findings that AP-2 has the ability to repress the expression of liver genes in nonliver cells opens a new avenue of investigation of negative regulation of gene transcription, and should improve our understanding of tissue-specific expression of liver genes. In summary, our data provide evidence suggesting a novel role of AP-2 as a repressor, inhibiting the expression of liver genes in nonliver cells. Thus, the tissue-specific expression of AP-2 may constitute an important mechanism contributing to the liver-specific expression of liver genes. ^

Rsk-2 activity is necessary for epidermal growth factor-induced phosphorylation of CREB protein and transcription of c-fos gene

Activation by growth factors of the Ras-dependent signaling cascade results in the induction of p90 ribosomal S6 kinases (p90rsk). These are translocated into the nucleus upon phosphorylation by mitogen-activated protein kinases, with which p90rsk are physically associated in the cytoplasm. In humans there are three isoforms of the p90rsk family, Rsk-1, Rsk-2, and Rsk-3, which are products of distinct genes. Although these isoforms are structurally very similar, little is known about their functional specificity. Recently, mutations in the Rsk-2 gene have been associated with the Coffin–Lowry syndrome (CLS). We have studied a fibroblast cell line established from a CLS patient that bears a nonfunctional Rsk-2. Here we document that in CLS fibroblasts there is a drastic attenuation in the induced Ser-133 phosphorylation of transcription factor CREB (cAMP response element-binding protein) in response to epidermal growth factor stimulation. The effect is specific, since response to serum, cAMP, and UV light is unaltered. Furthermore, epidermal growth factor-induced expression of c-fos is severely impaired in CLS fibroblasts despite normal phosphorylation of serum response factor and Elk-1. Finally, coexpression of Rsk-2 in transfected cells results in the activation of the c-fos promoter via the cAMP-responsive element. Thus, we establish a link in the transduction of a specific growth factor signal to changes in gene expression via the phosphorylation of CREB by Rsk-2.

Proteasome regulation of activation-induced T cell death

Lactacystin, a microbial metabolite that inhibits protease activity only in the proteasome, was used to study the role of the proteasome in the activation-induced cell death (AICD) of T cells. Lactacystin induces DNA fragmentation and apoptosis in a T cell hybridoma (DO.11.10) in a dose-dependent manner. Between 1 and 10 μM, the mildly cytotoxic lactacystin inhibited the AICD of DO.11.10 cells cultured in anti-CD3-coated wells. Degradation of IκBβ and the translocation of the NF-κB (p50/RelA) into the nucleus, which occurred at 1.5 hr after anti-CD3 activation, were inhibited by lactacystin. Lactacystin did not inhibit the expression of nuclear transcription factor Oct-1. The activation-induced expression of the immediate–early gene, Nur77, and the T cell death genes, CD95 (Fas) and CD95 ligand (FasL), were inhibited. Functional expression of FasL cytotoxicity and the increase of cell surface Fas were also inhibited. Lactacystin must be added within 2 hr of activation to efficiently block AICD. In addition, lactacystin failed to inhibit the killing of DO.11.10 by FasL-expressing allo-specific cytotoxic effector cells. These observations strongly suggest a direct link between the proteasome-dependent degradation of IκBβ and the AICD that occurs through activation of the FasL gene and up-regulation of the Fas gene.

Deregulated expression of c-Myc in a translocation-negative plasmacytoma on extrachromosomal elements that carry IgH and myc genes

The induced expression of c-Myc in plasmacytomas in BALB/c mice is regularly associated with nonrandom chromosomal translocations that juxtapose the c-myc gene to one of the Ig loci on chromosome 12 (IgH), 6 (IgK), or 16 (IgL). The DCPC21 plasmacytoma belongs to a small group of plasmacytomas that are unusual in that they appear to be translocation-negative. In this paper, we show the absence of any c-myc-activating chromosomal translocation for the DCPC21 by using fluorescent in situ hybridization, chromosome painting, and spectral karyotyping. We find that DCPC21 harbors c-myc and IgH genes on extrachromosomal elements (EEs) from which c-myc is transcribed, as shown by c-myc mRNA tracks and extrachromosomal gene transfer experiments. The transcriptional activity of these EEs is supported further by the presence of the transcription-associated phosphorylation of histone H3 (H3P) on the EEs. Thus, our data suggest that in this plasmacytoma, c-Myc expression is achieved by an alternative mechanism. The expression of the c-Myc oncoprotein is initiated outside the chromosomal locations of the c-myc gene, i.e., from EEs, which can be considered functional genetic units. Our data also imply that other “translocation-negative” experimental and human tumors with fusion transcripts or oncogenic activation may indeed carry translocation(s), however, in an extrachromosomal form.

Estrogen alters behavior and forebrain c-fos expression in ovariectomized rats subjected to the forced swim test

Estrogen has been implicated in brain functions related to affective state, including hormone-related affective disorders in women. Although some reports suggest that estrogen appears to decrease vulnerability to affective disorders in certain cases, the mechanisms involved are unknown. We used the forced swim test (FST), a paradigm used to test the efficacy of antidepressants, and addressed the hypotheses that estrogen alters behavior of ovariectomized rats in the FST and the FST-induced expression of c-fos, a marker for neuronal activity, in the rat forebrain. The behaviors displayed included struggling, swimming, and immobility. One hour after the beginning of the test on day 2, the animals were perfused, and the brains were processed for c-fos immunocytochemistry. On day 1, the estradiol benzoate-treated animals spent significantly less time struggling and virtually no time in immobility and spent most of the time swimming. Control rats spent significantly more time struggling or being immobile during a comparable period. On day 2, similar behavioral patterns with still more pronounced differences were observed between estradiol benzoate and ovariectomized control groups in struggling, immobility, and swimming. Analysis of the mean number of c-fos immunoreactive cell nuclei showed a significant reduction in the estradiol benzoate versus control groups in areas of the forebrain relating to sensory, contextual, and integrative processing. Our results suggest that estrogen-induced neurochemical changes in forebrain neurons may translate into an altered behavioral output in the affective domain.

Identification of a Human VPF/VEGF 3′ Untranslated Region Mediating Hypoxia-induced mRNA Stability

Hypoxia is a prominent feature of malignant tumors that are characterized by angiogenesis and vascular hyperpermeability. Vascular permeability factor/vascular endothelial growth factor (VPF/VEGF) has been shown to be up-regulated in the vicinity of necrotic tumor areas, and hypoxia potently induces VPF/VEGF expression in several tumor cell lines in vitro. Here we report that hypoxia-induced VPF/VEGF expression is mediated by increased transcription and mRNA stability in human M21 melanoma cells. RNA-binding/electrophoretic mobility shift assays identified a single 125-bp AU-rich element in the 3′ untranslated region that formed hypoxia-inducible RNA-protein complexes. Hypoxia-induced expression of chimeric luciferase reporter constructs containing this 125-bp AU-rich hypoxia stability region were significantly higher than constructs containing an adjacent 3′ untranslated region element without RNA-binding activity. Using UV-cross-linking studies, we have identified a series of hypoxia-induced proteins of 90/88 kDa, 72 kDa, 60 kDa, 56 kDa, and 46 kDa that bound to the hypoxia stability region element. The 90/88-kDa and 60-kDa species were specifically competed by excess hypoxia stability region RNA. Thus, increased VPF/VEGF mRNA stability induced by hypoxia is mediated, at least in part, by specific interactions between a defined mRNA stability sequence in the 3′ untranslated region and distinct mRNA-binding proteins in human tumor cells.

Mechanism for fetal globin gene expression: Role of the soluble guanylate cyclase–cGMP-dependent protein kinase pathway

Despite considerable concerns with pharmacological stimulation of fetal hemoglobin (Hb F) as a therapeutic option for the β-globin disorders, the molecular basis of action of Hb F-inducing agents remains unclear. Here we show that an intracellular pathway including soluble guanylate cyclase (sGC) and cGMP-dependent protein kinase (PKG) plays a role in induced expression of the γ-globin gene. sGC, an obligate heterodimer of α- and β-subunits, participates in a variety of physiological processes by converting GTP to cGMP. Northern blot analyses with erythroid cell lines expressing different β-like globin genes showed that, whereas the β-subunit is expressed at similar levels, high-level expression of the α-subunit is preferentially observed in erythroid cells expressing γ-globin but not those expressing β-globin. Also, the levels of expression of the γ-globin gene correlate to those of the α-subunit. sGC activators or cGMP analogs increased expression of the γ-globin gene in erythroleukemic cells as well as in primary erythroblasts from normal subjects and patients with β-thalassemia. Nuclear run-off assays showed that the sGC activator protoporphyrin IX stimulates transcription of the γ-globin gene. Furthermore, increased expression of the γ-globin gene by well known Hb F-inducers such as hemin and butyrate was abolished by inhibiting sGC or PKG activity. Taken together, these results strongly suggest that the sGC–PKG pathway constitutes a mechanism that regulates expression of the γ-globin gene. Further characterization of this pathway should permit us to develop new therapeutics for the β-globin disorders.

Studying Early Nodulin Gene ENOD40 Expression and Induction by Nodulation Factor and Cytokinin in Transgenic Alfalfa1

ENOD40, an early nodulin gene, is expressed following inoculation with Rhizobium meliloti or by adding R. meliloti-produced nodulation (Nod) factors or the plant hormone cytokinin to uninoculated roots. We isolated two MsENOD40 clones, designated MsENOD40–1 and MsENOD40–2, with distinct promoters from an alfalfa (Medicago sativa cv Chief) genomic library. The promoters were fused to the reporter gene uidA (gus), and the constructs were introduced into alfalfa. We observed that the MsENOD40–1 construct was expressed almost exclusively under symbiotic conditions. The MsENOD40–2 construct was transcribed under both symbiotic and nonsymbiotic conditions and in nonnodular and nodular tissues. Both MsENOD40 promoter-gus constructs were similarly expressed as nodules developed, and both were expressed in roots treated with 6-benzylaminopurine or purified Nod factor. However, no blue color was detected in nodule-like structures induced by the auxin transport inhibitor N-1-(naphthyl)phthalamic acid on roots of plants containing the MsENOD40–1 promoter construct, whereas pseudonodules from plants containing the MsENOD40–2 promoter construct stained blue. A 616-bp region at the distal 5′ end of the promoter is important for proper spatial expression of MsENOD40 in nodules and also for Nod-factor and cytokinin-induced expression.

Immunoregulatory properties of ISG15, an interferon-induced cytokine.

ISG15 is a 15-kDa protein of unique primary amino acid sequence, which is transcriptionally regulated by interferon (IFN) alpha and IFN-beta. Because it is synthesized in many cell types and secreted from human monocytes and lymphocytes, we postulated that ISG15 might act to modulate immune cell function. ISG15 stimulated B-depleted lymphocyte proliferation in a dose-dependent manner with significant proliferation induced by amounts of ISG15 as low as 1 ng/ml (58 pM). Maximal stimulation of [3H]thymidine incorporation by B-depleted lymphocytes occurred at 6-7 days. Immunophenotyping of ISG15-treated B-depleted lymphocyte cultures indicated a 26-fold expansion of natural killer (NK) cells (CD56+). In cytotoxicity assays, ISG15 was a potent inducer of cytolytic activity directed against both K562 (100 lytic units per 10(6) cells) and Daudi (80 lytic units per 10(6) cells) tumor cell targets, indicating that ISG15 enhanced lymphokine-activated killer-like activity. ISG15-induced NK cell proliferation required coculturing of T and NK cells, suggesting that soluble factor(s) were required. Measurement of ISG15-treated cell culture supernatants for cytokines indicated production of IFN-gamma (> 700 units/ml). No interleukin 2 or interleukin 12 was detected. IFN-gamma itself failed to stimulate lymphocyte proliferation and lymphokine-activated killer cell activation. Further, induced expression of IFN-gamma mRNA was detected by reverse transcription-PCR in T lymphocytes after ISG15 treatment but not in NK cells. Enhancement of NK cell proliferation, augmentation of non-major histocompatibility complex-restricted cytotoxicity, and induction of IFN-gamma from T cells identify ISG15 as a member of the cytokine cascade and suggest that it may be responsible for amplifying and directing some of the immunomodulatory effects of IFN-alpha or IFN-beta.

Involvement of the double-stranded-RNA-dependent kinase PKR in interferon expression and interferon-mediated antiviral activity.

The signaling mechanisms responsible for the induced expression of interferon (IFN) genes by viral infection or double-stranded RNA (dsRNA) are not well understood. Here we investigate the role of the interferon-induced dsRNA-dependent protein kinase PKR in the regulation of IFN induction. Biological activities attributed to PKR include regulating protein synthesis, mediating IFN actions, and functioning as a possible tumor suppressor. Since binding of dsRNA is required for its activation, PKR has been considered as a candidate signal transducer for regulating IFN expression. To examine this role of PKR, loss-of-function phenotypes in stable transformants of promonocytic U-937 cells were achieved by two different strategies, overexpression of an antisense PKR transcript or a dominant negative PKR mutant gene. Both types of PKR-deficient cells were more permissive for viral replication than the control U-937 cells. As the result of PKR loss, they also showed impaired induction of IFN-alpha and IFN-beta genes in response to several inducers--specifically, encephalomyocarditis virus, lipopolysaccharide, and phorbol 12-myristate 13-acetate. Interestingly, while IFN-alpha induction by dsRNA was impaired in PKR-deficient cells, IFN-beta induction remained intact. Loss of PKR function also resulted in decreased antiviral activity as elicited by IFN-alpha and, to a greater extent, by IFN-gamma. These results implicate PKR in the regulation of several antiviral activities.

A heme-protein-based oxygen-sensing mechanism controls the expression and suppression of multiple proteins in anoxia-tolerant turtle hepatocytes.

The O2 sensitivity of protein expression was assessed in hepatocytes from the western painted turtle. Anoxic cells consistently expressed proteins of 83.0, 70.4, 42.5, 35.3, and 16.1 kDa and suppressed proteins of 63.7, 48.2, 36.9, 29.5, and 17.7 kDa. Except for the 70.4-kDa protein, this pattern was absent during aerobic incubation with 2 mM NaCN, suggesting a specific requirement for O2. Aerobic incubation with Co2+ or Ni2+ increased expression of the 42.5-, 35.3-, and 16.1-kDa protein bands which was diminished with the heme synthesis inhibitor 4,6-dioxoheptanoic acid. Proteins suppressed in anoxia were also suppressed during aerobic incubation with Co2+ or Ni2+ but this was not relieved by 4,6-dioxoheptanoic acid. The anoxia- and Co2+/Ni2+-induced expression of the 42.5-, 35.3-, and 16.1-kDa protein bands was antagonized by 10% CO; however, with the exception of the 17.7-kDa protein, this was not found for any of the O2- or Co2+/Ni2+-suppressed proteins. Anoxia-induced proteins were compared with proteins expressed during heat shock. Heat shock proteins appeared at 90.2, 74.8, 63.4, 25, and 15.5 kDa and were of distinct molecular masses compared with the anoxia-induced proteins. These results suggest that O2-sensing mechanisms are active in the control of protein expression and suppression during anoxia and that, in the case of the 42.5-, 35.3-, 17.7-, and 16.1-kDa proteins, a conformational change in a ferro-heme protein is involved in transducing the O2 signal.

Antitumor promotion by phenolic antioxidants: inhibition of AP-1 activity through induction of Fra expression.

Induction of phase 2 detoxification enzymes by phenolic antioxidants can account for prevention of tumor initiation but cannot explain why these compounds inhibit tumor promotion. Phase 2 genes are induced through an antioxidant response element (ARE). Although the ARE resembles an AP-1 binding site, we show that the major ARE binding and activating protein is not AP-1. Interestingly, AP-1 DNA binding activity was induced by the phenolic antioxidant tert-butylhydroquinone (BHQ), but the induction of AP-1 transcriptional activity by the tumor promoter 12-O-tetradecanoylphorbol 13-acetate (TPA) was inhibited by this compound. BHQ induced expression of c-jun, junB, fra-1, and fra-2, which encode AP-1 components, but was a poor inducer of c-fos and had no effect on fosB. Like c-Fos and FosB, the Fra proteins heterodimerize with Jun proteins to form stable AP-1 complexes. However, Fra-containing AP-1 complexes have low transactivation potential. Furthermore, Fra-1 repressed AP-1 activity induced by either TPA or expression of c-Jun and c-Fos. We therefore conclude that inhibitory AP-1 complexes composed of Jun-Fra heterodimers, induced by BHQ, antagonize the transcriptional effects of the tumor promoter TPA, which are mediated by Jun-Fos heterodimers. Since AP-1 is an important mediator of tumor promoter action, these findings may explain the anti-tumor-promoting activity of phenolic antioxidants.