Identification of ERF-1 as a member of the AP2 transcription factor family.

The ERF-1 transcription factor was previously shown to be involved in the regulation of estrogen receptor (ER) gene transcription in hormonally responsive breast and endometrial carcinomas. In this study we sought to identify the gene for ERF-1. ERF-1 activates ER gene transcription by binding to the imperfect palindrome CCCTGCGGGG within the promoter of the ER gene. ERF-1 protein was purified from the ER-positive breast carcinoma cell line, MCF7, utilizing ion exchange and DNA affinity chromatography. Peptide sequence analysis was used to isolate a 2.7 kb cDNA clone from an MCF7 cDNA library. This cDNA encodes a protein of 48 kDa previously identified as the AP2gamma transcription factor. By gel-shift analysis, in vitro synthesized ERF-1 comigrates with MCF7 native ERF-1 complex and demonstrates identical sequence binding specificity as native ERF-1. In addition, AP2 polyclonal antisera supershifts both in vitro synthesized and native ERF-1 complexes. These results show that ERF-1 is a member of the AP2 family of developmentally regulated transcription factors. Given the central role of ER expression in breast carcinoma biology, ERF-1 is likely to regulate expression of a set of genes characteristic of the hormonally-responsive breast cancer phenotype.

Activate NF-kappa B or die?

Activation of the transcription factor NF-kappaB has been linked to apoptosis, with the factor playing either an anti-apoptotic or a pro-apoptotic role, depending on the type of cell in which it is expressed.

Systematic mutational analysis of the death domain of the tumor necrosis factor receptor 1-associated protein TRADD.

Tumor necrosis factor receptor 1 (TNF-R1) mediates most of the biological properties of TNF including activation of the transcription factor NF-kappaB and programmed cell death. An approximately 80-amino acid region within the intracellular domain of the receptor, termed the death domain, is required for signaling NF-kappaB activation and cytotoxicity. A TNF-R1-associated protein TRADD has been discovered that interacts with the death domain of the receptor. Elevated expression of TRADD in cells triggers both NF-kappaB activation and programmed cell death pathways. The biological activities of TRADD have been mapped to a 111-amino acid region within the carboxyl-terminal half of the protein. This region shows sequence similarity to the death domain of TNF-R1 and can self-associate and bind to the TNF-R1 death domain. We have performed an alanine scanning mutagenesis of TRADD's death domain to explore the relationship among its various functional properties. Mutations affecting the different activities of TRADD do not map to discrete regions but rather are spread over the entire death domain, suggesting that the death domain is a multifunctional unit. A mutant that separates cell killing from NF-kappaB activation by the TRADD death domain has been identified indicating that these two signaling pathways diverge with TRADD. Additionally, one of the TRADD mutants that fails to activate NF-kappaB was found to act as dominant negative mutant capable of preventing induction of NF-kappaB by TNFalpha. Such observations provide evidence that TRADD performs an obligate role in TNF-induced NF-kappaB activation.

Sp1 is a component of the cytokine-inducible enhancer in the promoter of vascular cell adhesion molecule-1.

Transcription of the vascular cell adhesion molecule-1 (VCAM-1) gene in endothelial cells is induced by the inflammatory cytokines interleukin-1 beta, tumor necrosis factor-alpha, and lipopolysaccharide. Previous studies demonstrated that the cytokine-response region in the VCAM1 promoter contains binding sites for the transcription factors nuclear factor-kappa B (NF-kappa B) and interferon regulatory factor-1. Using a saturation mutagenesis approach, we report that the cytokine-inducible enhancer consists of these previously characterized elements and a novel region located 3' of the NF-kappa B sites. Electrophoretic mobility shift assays and DNase I footprint studies with endothelial nuclear extracts and recombinant protein revealed that the transcriptional activator Sp1 interacts with this novel element in a specific manner. Transient transfection assays using vascular endothelial cells revealed that site-directed mutations in the Sp1 binding element decreased tumor necrosis factor-alpha-induced activity of the VCAM1 promoter. The cytokine-induced enhancer of the VCAM1 gene requires constitutively bound Sp1 and induced heterodimeric NF-kappa B for maximal promoter activity.

Induction of the gene encoding mucosal vascular addressin cell adhesion molecule 1 by tumor necrosis factor alpha is mediated by NF-kappa B proteins.

Mucosal vascular addressin cell adhesion molecule 1 (MAdCAM-1) is involved in trafficking of lymphocytes to mucosal endothelium. Expression of MAdCAM-1 is induced in the murine endothelial cell line bEnd.3 by tumor necrosis factor alpha (TNF-alpha), interleukin 1, and bacterial lipopolysaccharide. Here we show that TNF-alpha enhances expression of a firefly luciferase reporter directed by the MAdCAM-1 promoter, confirming transcriptional regulation of MAdCAM-1. Mutational analysis of the promoter indicates that a DNA fragment extending from nt -132 to nt +6 of the gene is sufficient for TNF-alpha inducibility. Two regulatory sites critical for TNF-alpha induction were identified in this region. DNA-binding experiments demonstrate that NF-kappa B proteins from nuclear extracts of TNF-alpha-stimulated bEnd.3 cells bind to these sites, and transfection assays with promoter mutants of the MAdCAM-1 gene indicate that occupancy of both sites is essential for promoter function. The predominant NF-kappa B binding activity detected with these nuclear extracts is a p65 homodimer. These findings establish that, as with other endothelial cell adhesion molecules, transcriptional induction of MAdCAM-1 by TNF-alpha requires activated NF-kappa B proteins.

Three NF-kappa B binding sites in the human E-selectin gene required for maximal tumor necrosis factor alpha-induced expression.

Transcription of the gene encoding the endothelial cell-leukocyte adhesion molecule (ELAM-1; E-selectin) is induced in response to various cytokines, including tumor necrosis factor-alpha (TNF-alpha) and interleukin-1. A DNase I-hypersensitive site in the 5' proximal promoter region of the E-selectin gene is observed in human umbilical vein endothelial cells only following TNF-alpha treatment, suggesting the presence of a TNF-alpha-inducible element close to the transcriptional start site. Transient transfection studies in endothelial cells demonstrated that 170 bp of upstream sequences is sufficient to confer TNF-alpha inducibility. Systematic site-directed mutagenesis of this region revealed two regulatory elements (-129 to -110 and -99 to -80) that are essential for maximal promoter activity following cytokine treatment. Protein binding studies with crude nuclear extracts and recombinant proteins revealed that the two elements correspond to three NF-kappa B binding sites (site 1, -126; site 2, 116; and site 3, -94). All three sites can be bound by NF-kappa B when used as independent oligonucleotides in mobility shift assays. However, within the context of a larger promoter fragment, sites 2 and 3 are preferentially occupied over site 1. These data are consistent with results obtained in transfection studies demonstrating that mutations in sites 2 and 3 are more detrimental than mutations within site 1. Hence, inducibility of the E-selectin gene requires the interaction of NF-kappa B proteins bound to multiple regulatory elements.

Transcription factor based therapeutics: drugs of the future?

During the past decade there has been an explosion of information relating to our understanding of eukaryotic gene expression. The identification of transcription factors as the key regulatory molecules in this process, and the analysis of their structure and function have revealed that these proteins are potential targets for therapeutic intervention.

The cell-type-specific activator region of c-Jun juxtaposes constitutive and negatively regulated domains.

Dissection of the cell-type-specific activation region in c-Jun reveals two functionally separable regulatory subdomains. One subdomain (a1) functions as a transcriptional activator; adjacent to it is a newly identified domain (epsilon) which, together with the previously defined delta region, interacts with a cellular factor that modulates the action of a1. Mutants that lack epsilon are constitutively active and do not interact with the cell-type-specific repressor, whereas mutants that have sustained changes in a1 exhibit a reduced trans-activation potential but retain the ability to interact with the repressor. This bipartite and modular organization of the a1/epsilon domain is further established by demonstrating that a1 can be replaced by the heterologous acidic activator of VP16 and retain proper negative regulation by the cell-specific c-Jun inhibitor along with epsilon and delta. Repression of Jun activity by the inhibitor is not caused by a change in stability, nuclear localization, or DNA-binding activity of the protein. Instead the inhibitor apparently regulates transcriptional activation by interacting directly with delta/epsilon and perhaps masking the a1 domain. These studies suggest that multifunctional activation domains, which are structurally complex, may play an important role in the mechanisms that govern inducible tissue-specific gene expression.

v-Src and EJ Ras alleviate repression of c-Jun by a cell-specific inhibitor.

The AP-1 family of transcription factors, which includes the proto-oncogene products c-Jun and c-Fos, controls the stimulation of cellular genes by growth factors and the expression of oncogenes, including src and ras. Transcriptional activation by c-Jun is regulated by a cell-type-specific inhibitor that represses the activity of a transcriptional activation domain (A1) of c-Jun by operating through the adjacent negative regulatory region (delta). Here we show that cotransfection of the src or ras oncogene enhances the transcriptional activity of a GAL4:c-Jun hybrid that includes the delta-A1 region of c-Jun, suggesting that the DNA binding and dimerization domain of c-Jun is not required for stimulation by Src or Ras. Moreover, induction of c-Jun activity by Src and Ras occurs in cell lines containing the c-Jun inhibitor but not in a cell line lacking it. The region in c-Jun essential for the stimulatory action of these oncogenes maps to domain A1. These findings suggest the existence of signal-transduction pathways that result in an increase in transcriptional activity of c-Jun and AP-1 by disrupting the c-Jun:inhibitor interaction.

Control of c-Jun activity by interaction of a cell-specific inhibitor with regulatory domain delta: differences between v- and c-Jun.

Analysis of transcriptional activation properties of c-Jun chimeras in different cell lines suggests that it contains an activator domain (A1) that is negatively regulated by a cell type-specific inhibitor. A regulatory domain of c-Jun, delta, previously identified by in vitro experiments, also regulates transcriptional activation by c-Jun in vivo. The delta domain facilitates or stabilizes the interaction of the cellular inhibitor with A1. v-Jun, which lacks delta, is a stronger transcriptional activator than c-Jun, since its activity is not efficiently repressed by the cellular inhibitor. In vitro transcription with chimeric Jun proteins and extracts from different cell types confirms that the A1 and delta domains are repressed in a cell type-specific manner. These findings implicate a specific cellular factor in the negative regulation of c-Jun activity and suggest a molecular basis for the observed difference in transcriptional properties between v-Jun and c-Jun.

The transforming domain alone of the latent membrane protein of Epstein-Barr virus is toxic to cells when expressed at high levels.

A previously unrecognized activity has been associated with the product of the BNLF-1 gene of Epstein-Barr virus. This gene encodes the latent membrane protein of Epstein-Barr virus. When the gene was expressed at high levels, it was toxic to all cell lines tested, which included six human B-lymphoid lines as well as BALB/3T3, 143/EBNA-1, and HEp-2 cells. The BNLF-1 gene was previously shown to induce anchorage-independent and tumorigenic growth in Rat-1 and BALB/3T3 cells. We demonstrate here that only those mutations in the BNLF-1 gene that score positively in the anchorage-independent growth assay were cytotoxic when expressed at high levels. It is therefore possible that the same activities of the latent membrane protein that are necessary to induce anchorage-independent growth of some rodent cell lines also confer toxicity to many cell lines when expressed at high levels.

The multiple membrane-spanning segments of the BNLF-1 oncogene from Epstein-Barr virus are required for transformation.

The BNLF-1 gene from Epstein-Barr virus (EBV) induces anchorage-independent and tumorigenic growth in rodent cell lines. The BNLF-1 protein (also termed LMP) is a membrane protein, and its predicted amino acid sequence indicates that the protein has six membrane-spanning segments in addition to a short amino-terminal (approximately 25 amino acids) and a long carboxyl-terminal (approximately 200 amino acids) cytoplasmic domain. To identify the regions of the protein that are essential for its transforming activity, we have constructed deletion mutants of the BNLF-1 gene and tested them for transforming activity. Surprisingly, the entire carboxyl-terminal cytoplasmic domain is dispensable for transforming activity, whereas the putative membrane-spanning segments are essential. These observations indicate that BNLF-1 has a novel function that is distinct from the functions associated with other membrane-associated viral transforming proteins. We speculate that BNLF-1 is a receptor for a growth-promoting agent, with its trans-membrane domain involved in ligand binding, and its amino-terminal domain or cytoplasmic loops involved in coupling BNLF-1 to effector molecules in the cell, a situation analogous to the rhodopsin group of receptors.

Transformation of Balb 3T3 cells by the BNLF-1 gene of Epstein-Barr virus.

The BNLF-1 protein is the only non-nuclear Epstein-Barr virus (EBV) encoded protein that has been detected in B-lymphocytes immortalized by EBV. We demonstrate that the BNLF-1 gene induces anchorage-independent growth and tumorigenic transformation of the murine cell-line, Balb/3T3. This demonstration extends the earlier observation that the BNLF-1 gene can transform Rat-1 cells. In addition we find that the BNLF-1 protein is located in the particulate fraction of cells, is phosphorylated, and is turned over with a half-life of 2.0 to 3.5 h in the BNLF-1 transformed Balb/3T3 cells, just as it is in EBV-genome-positive B-cell-lines.

Posttranslational processing of an Epstein-Barr virus-encoded membrane protein expressed in cells transformed by Epstein-Barr virus.

The BamHI Nhet fragment of the B958 strain of Epstein-Barr virus (EBV) encodes a membrane protein (BNLF-1) that is present in cells transformed by EBV. We made a hybrid protein in which a polypeptide sequence from the carboxyl-terminal part of BNLF-1 is fused to Escherichia coli beta-galactosidase. This hybrid protein was used to immunize rabbits, and the resulting antiserum was purified by immunoaffinity chromatography. The antiserum was able to immunoprecipitate BNLF-1 from cell lysates. We found that BNLF-1 is phosphorylated at serines in EBV genome-positive B-cell lines. Pulse-chase analyses with [35S]methionine indicated that BNLF-1 is turned over in lymphoblasts with a half-life of approximately 5 h. Protein immunoblots of EBV genome-positive B-cell lines revealed both a 62,000-molecular-mass band corresponding to BNLF-1 and a myriad of lower-molecular-mass bands. We postulate that these lower-molecular-mass bands are degradation products resulting from the turnover of BNLF-1 in cells. The BNLF-1 gene was expressed in COS cells, and the protein was both phosphorylated and turned over in these cells.

Leucine biosynthesis in yeast : Identification of two genes (LEU4, LEU5) that affect α-Isopropylmalate synthase activity and evidence that LEU1 and LEU2 gene expression is controlled by α-Isopropylmalate and the product of a regulatory gene.

Tetrad analysis indicates that α-isopropylmalate synthase activity of yeast is determined by two separate genes, designated LEU4 and LEU5. LEU4 is identified as a structural gene. LEU5 either encodes another α-isopropylmalate synthase activity by itself or provides some function needed for the expression of a second structural gene. The properties of mutants affecting the biosynthesis of leucine and its regulation suggest that the expression of LEU1 and LEU2 (structural genes encoding isopropylmalate isomerase and ß-isopropylmalate dehydrogenase, respectively) is controlled by a complex of a-isopropylmalate and a regulatory element (the LEU3 gene product). Similarities and differences between yeast and Neurospora crassa with respect to leucine biosynthesis are discussed.