Overexpression of Batf induces an apoptotic defect and an associated lymphoproliferative disorder in mice.

Activator protein-1 (AP-1) is a dimeric transcription factor composed of the Jun, Fos and Atf families of proteins. Batf is expressed in the immune system and participates in AP-1 dimers that modulate gene expression in response to a variety of stimuli. Transgenic (Tg) mice overexpressing human BATF in T cells were generated using the human CD2 promoter (CD2-HA (hemagglutinin antigen) - BATF). By 1 year of age, over 90% of the mice developed a lymphoproliferative disorder (LPD). The enlarged lymph nodes characteristic of this LPD contain a polyclonal accumulation of T cells with a CD4(+) bias, yet efforts to propagate these tumor cells in vitro demonstrate that they do not proliferate as well as wild-type CD4(+) T cells. Instead, the accumulation of these cells is likely due to an apoptotic defect as CD2-HA-BATF Tg T cells challenged by trophic factor withdrawal in vitro resist apoptosis and display a pro-survival pattern of Bcl-2 family protein expression. As elevated levels of Batf expression are a feature of lymphoid tumors in both humans and mice, these observations support the use of CD2-HA-BATF mice as a model for investigating the molecular details of apoptotic dysregulation in LPD.

Characterization of murine BATF: a negative regulator of activator protein-1 activity in the thymus.

BATF belongs to the AP-1/ATF superfamily of transcription factors and forms heterodimers with Jun proteins to bind AP-1 consensus DNA. Unlike Fos/Jun heterodimers which stimulate gene transcription, BATF/Jun heterodimers are transcriptionally inert and inhibit biological processes that are associated with the overstimulation of AP-1 activity. Here, we describe the murine BATF cDNA and genomic clones and map the BATF locus to chromosome 12 D2-3. Using in situ hybridization of BATF mRNA, we show that BATF gene expression is highly restricted, with the most prominent signals detected in the thymus. BATF mRNA levels are regulated differentially during discrete stages of T cell development and are up-regulated following activation of T cells in the periphery. To demonstrate the impact of BATF on AP-1 activity in vivo, AP-1 luciferase reporter mice were crossed to transgenic mice overexpressing BATF exclusively in thymic T cells. Results show that elevated levels of BATF protein correlate with reduced transactivation by AP-1. Since the differential regulation of AP-1 activity is linked to key transitions in the developing immune system, our observations support a critical role for BATF in determining the overall level of AP-1 activity, and thus AP-1 target gene expression, in specific T cell subtypes.

Differential effects of Ras signaling through NFkappaB on skeletal myogenesis.

Oncogenic Ras (H-Ras G12V) inhibits skeletal myogenesis through multiple signaling pathways. Previously, we demonstrated that the major downstream effectors of Ras (i.e., MEK/MAPK, RalGDS and Rac/Rho) play a minor, if any, role in the differentiation-defective phenotype of Ras myoblasts. Recently, NFkappaB, another Ras signaling target, has been shown to inhibit myogenesis presumably by stimulating cyclin D1 accumulation and cell cycle progression. In this study, we address the involvement of NFkappaB activation in the Ras-induced inhibition of myogenesis. Using H-Ras G12V and three G12V effector-loop variants, we detect high levels of NFkappaB transcriptional activity in C3H10T1/2-MyoD cells treated with differentiation medium. Myogenesis is blocked by all Ras proteins tested, yet only in the case of H-Ras G12V are cyclin D1 levels increased and cell cycle progression maintained. Expression of IkappaBalpha SR, an inhibitor of NFkappaB, does not reverse the differentiation-defective phenotype of Ras expressing cultures, but does induce differentiation in cultures treated with tumor necrosis factor (TNFalpha) or in cultures expressing the RelA/p65 subunit of NFkappaB. These data confirm that NFkappaB is a target of Ras and suggest that the cellular actions of NFkappaB require additional signals that are discriminated by the Ras effector-loop variants. Results with IkappaBalpha SR convincingly demonstrate that H-Ras G12V does not rely on NFkappaB activity to block myogenesis, an observation that continues to implicate another unidentified signaling pathway(s) in the inhibition of skeletal myogenesis by Ras.

B-ATF functions as a negative regulator of AP-1 mediated transcription and blocks cellular transformation by Ras and Fos.

B-ATF is a nuclear basic leucine zipper protein that belongs to the AP-1/ATF superfamily of transcription factors. Northern blot analysis reveals that the human B-ATF gene is expressed most highly in hematopoietic tissues. Interaction studies in vitro and in vivo show that the leucine zipper of B-ATF mediates dimerization with members of the Jun family of proteins. Chimeric proteins consisting of portions of B-ATF and the DNA binding domain of the yeast activator GAL4 do not stimulate reporter gene expression in mammalian cells, indicating that B-ATF does not contain a conventional transcription activation domain. Jun/B-ATF dimers display similar DNA binding profiles as Jun/Fos dimers, with a bias toward binding TRE (12-O-tetradecanolyphorbol-13-acetate-response element) over CRE (cyclic AMP-response element) DNA sites. B-ATF inhibits transcriptional activation of a reporter gene containing TRE sites in a dose-dependent manner, presumably by competing with Fos for Jun and forming transcriptionally inert Jun/B-ATF heterodimers. Stable expression of B-ATF in C3H10T1/2 cells does not reduce cell viability, but does result in a reduced cellular growth rate when compared to controls. This effect is dominant in the presence of the growth promoting effects of the H-Ras or the v-Fos oncoproteins, since expression of B-ATF restricts the efficiency of focus formation by these transforming agents. These findings demonstrate that B-ATF is a tissue-specific transcription factor with the potential to function as a dominant-negative to AP-1.

The Adrenomedullin gene is a target for negative regulation by the Myc transcription complex.

The Myc family of transcription factors plays a central role in vertebrate growth and development although relatively few genetic targets of the Myc transcription complex have been identified. In this study, we used mRNA differential display to investigate gene expression changes induced by the overexpression of the MC29 v-Myc oncoprotein in C3H10T1/2 mouse fibroblasts. We identified the transcript of the adrenomedullin gene (AM) as an mRNA that is specifically down-regulated in v-Myc overexpressing C3H10T1/2 cell lines as well as in a Rat 1a cell line inducible for c-Myc. Nucleotide sequence analysis of the mouse AM promoter reveals the presence of consensus CAAT and TATA boxes as well as an initiator element (INR) with significant sequence similarity to the INR responsible for Myc-mediated repression of the adenovirus major late promoter (AdMLP). Reporter gene assays confirm that the region of the AM promoter containing the INR is the target of Myc-mediated repression. Exogenous application of AM peptide to quiescent C3H10T1/2 cultures does not stimulate growth, and constitutive expression of AM mRNA in C3H10T1/2 cells correlates with a reduced potential of the cells to be cotransformed by v-Myc and oncogenic Ras p21. Additional studies showing that AM mRNA is underrepresented in C3H10T1/2 cell lines stably transformed by Ras p21 or adenovirus E1A suggest that AM gene expression is incompatible with deregulated growth in this cell line. We propose a model in which the repression of AM gene expression by Myc is important to the role of this oncoprotein as a potentiator of cellular transformation in C3H10T1/2 and perhaps other cell lines.

A cytoplasmic structure resembling large protein aggregates induced by interferons.

IFP 35 is an interferon (IFN)-regulated leucine zipper protein, expression of which is observed in a variety of cell types including monocytes/macrophages, epithelial cells and fibroblasts. Using immunofluorescence studies, we demonstrate that IFP 35 is found in characteristic punctate cytoplasmic structures after IFN treatment. Co-localization experiments using double immunofluorescence and confocal laser scanning microscopy failed to show association of IFP 35 with known organelles (mitochondria, peroxisomes, endoplasmic reticulum, lysosomes, endosomes, Golgi complex), ribosomes, or actin filaments. Subcellular fractionation to separate membrane-associated from cytoplasmic proteins demonstrated that IFP 35 localizes to the cytoplasm. Separation of postnuclear supernatant from HeLa cells by gel filtration revealed that IFP 35 eluted at a molecular mass of 200-440 kD, suggesting that IFP 35 is part of protein complexes. Electron microscopic studies showed cytoplasmic clusters of a few aggregates of IFP 35 in IFN-treated cells which were neither associated with nor surrounded by a membrane. A combination of immunoprecipitation and immunofluorescence studies of cells transfected with a hemagglutinin epitope-tagged IFP 35 expression construct demonstrated complex formation and co-localization of endogenous and transfected IFP 35. Taken together, our studies demonstrate that IFP 35 associates with unique cytoplasmic structures that are distinct from known organelles and resemble large protein aggregates.

Target genes and cellular regulators of the Myc transcription complex.

Precisely how the Myc proteins promote cellular growth and transformation remains an enigma. Structural and functional studies unequivocally demonstrate that Myc proteins have the potential to function as regulators of gene transcription in vivo. Therefore, the current view is that Myc-induced cellular growth is orchestrated by products of the genes regulated by Myc and that efforts to curtail Myc function are best directed toward the accessory proteins that interact with Myc to modulate its function as a transcription factor. This review is focused on recent work designed to identify the genetic targets of Myc activity in cells and to characterize Myc-interacting proteins. These studies have provided an intriguing set of molecular tools that can be exploited to generate new information on the role of Myc in cell growth and oncogenic transformation.

A role for RalGDS and a novel Ras effector in the Ras-mediated inhibition of skeletal myogenesis.

Oncogenic Ras inhibits the differentiation of skeletal muscle cells through the activation of multiple downstream signaling pathways, including a Raf-dependent, mitogen-activated or extracellular signal-regulated kinase kinase/mitogen-activated protein kinase (MEK/MAPK)-independent pathway. Here we report that a non-Raf binding Ras effector-loop variant (H-Ras G12V,E37G), which retains interaction with the Ral guanine nucleotide dissociation stimulator (RalGDS), inhibits the conversion of MyoD-expressing C3H10T1/2 mouse fibroblasts to skeletal muscle. We show that H-Ras G12V,E37G, RalGDS, and the membrane-localized RalGDS CAAX protein inhibit the activity of alpha-actin-Luc, a muscle-specific reporter gene containing a necessary E-box and serum response factor (SRF) binding site, while a RalGDS protein defective for Ras interaction has no effect on alpha-actin-Luc transcription. H-Ras G12V,E37G does not activate endogenous MAPK, but does increase SRF-dependent transcription. Interestingly, RalGDS, RalGDS CAAX, and RalA G23V inhibit H-Ras G12V, E37G-induced expression of an SRF-regulated reporter gene, demonstrating that signaling through RalGDS does not duplicate the action of H-Ras G12V,E37G in this system. As additional evidence for this, we show that H-Ras G12V,E37G inhibits the expression of troponin I-Luc, an SRF-independent muscle-specific reporter gene, whereas RalGDS and RalGDS CAAX do not. Although our studies show that signaling through RalGDS can interfere with the expression of reporter genes dependent on SRF activity (including alpha-actin-Luc), our studies also provide strong evidence that an additional signaling molecule(s) activated by H-Ras G12V,E37G is required to achieve the complete inhibition of the myogenic differentiation program.

A general method to design dominant negatives to B-HLHZip proteins that abolish DNA binding.

We describe a method to design dominant-negative proteins (D-N) to the basic helix-loop-helix-leucine zipper (B-HLHZip) family of sequence-specific DNA binding transcription factors. The D-Ns specifically heterodimerize with the B-HLHZip dimerization domain of the transcription factors and abolish DNA binding in an equimolar competition. Thermal denaturation studies indicate that a heterodimer between a Myc B-HLHZip domain and a D-N consisting of a 12-amino acid sequence appended onto the Max dimerization domain (A-Max) is -6.3 kcal.mol-1 more stable than the Myc:Max heterodimer. One molar equivalent of A-Max can totally abolish the DNA binding activity of a Myc:Max heterodimer. This acidic extension also has been appended onto the dimerization domain of the B-HLHZip protein Mitf, a member of the transcription factor enhancer binding subfamily, to produce A-Mitf. The heterodimer between A-Mitf and the B-HLHZip domain of Mitf is -3.7 kcal.mol-1 more stable than the Mitf homodimer. Cell culture studies show that A-Mitf can inhibit Mitf-dependent transactivation both in acidic extension and in a dimerization-dependent manner. A-Max can inhibit Myc-dependent foci formation twice as well as the Max dimerization domain (HLHZip). This strategy of producing D-Ns may be applicable to other B-HLHZip or B-HLH proteins because it provides a method to inhibit the DNA binding of these transcription factors in a dimerization-specific manner.

Signaling through mitogen-activated protein kinase and Rac/Rho does not duplicate the effects of activated Ras on skeletal myogenesis.

The ability of basic helix-loop-helix muscle regulatory factors (MRFs), such as MyoD, to convert nonmuscle cells to a myogenic lineage is regulated by numerous growth factor and oncoprotein signaling pathways. Previous studies have shown that H-Ras 12V inhibits differentiation to a skeletal muscle lineage by disrupting MRF function via a mechanism that is independent of the dimerization, DNA binding, and inherent transcriptional activation properties of the proteins. To investigate the intracellular signaling pathway(s) that mediates the inhibition of MRF-induced myogenesis by oncogenic Ras, we tested two transformation-defective H-Ras 12V effector domain variants for their ability to alter terminal differentiation. H-Ras 12V,35S retains the ability to activate the Raf/MEK/mitogen-activated protein (MAP) kinase cascade, whereas H-Ras 12V,40C is unable to interact directly with Raf-1 yet still influences other signaling intermediates, including Rac and Rho. Expression of each H-Ras 12V variant in C3H10T1/2 cells abrogates MyoD-induced activation of the complete myogenic program, suggesting that MAP kinase-dependent and -independent Ras signaling pathways individually block myogenesis in this model system. However, additional studies with constitutively activated Rac1 and RhoA proteins revealed no negative effects on MyoD-induced myogenesis. Similarly, treatment of Ras-inhibited myoblasts with the MEK1 inhibitor PD98059 revealed that elevated MAP kinase activity is not a significant contributor to the H-Ras 12V effect. These data suggest that an additional Ras pathway, distinct from the well-characterized MAP kinase and Rac/Rho pathways known to be important for the transforming function of activated Ras, is primarily responsible for the inhibition of myogenesis by H-Ras 12V.

A minimal regulatory region maintains constitutive expression of the max gene.

Max is a basic helix-loop-helix/leucine zipper protein that forms heterodimers with the Myc family of proteins to promote cell growth and with the Mad/Mxi1 family of proteins to inhibit cell growth. The role of Max as the obligate binding partner for these two protein families necessitates the observed constitutive expression and relatively long half-life of the max mRNA under a variety of growth conditions. In this study, we have used the chicken max gene to map DNA elements maintaining max gene expression in vertebrate cells. We have identified a minimal regulatory region (MRR) that resides within 115 bp of the max translation initiation site and that possesses an overall structure typical of TATA-less promoters. Within the MRR are two consensus binding sites for Sp1, a ubiquitously expressed transcription factor that plays a role in the expression of many constitutive genes. Interestingly, we show that direct binding by Sp1 to these sites is not required for MRR-mediated transcription. Instead, the integrity of a 20-bp DNA element in the MRR is required for transcriptional activity, as is the interaction of this DNA element with a 90-kDa cellular protein. Our data suggest that it is the persistence of this 90-kDa protein in vertebrate cells which drives max gene expression, insulates the max promoter from the dramatic changes in transcription that accompany cell growth and development, and ensures that adequate levels of Max will be available to facilitate the function of the Myc, Mad, and Mxi1 families of proteins.

Distinct signaling pathways regulate transformation and inhibition of skeletal muscle differentiation by oncogenic Ras.

Expression of oncogenic Ras in 23A2 skeletal myoblasts is sufficient to induce both a transformed phenotype and a differentiation-defective phenotype, but the signaling pathways activated by oncogenic Ras in these cells and their respective contribution to each phenotype have not been explored. In this study, we investigated MAP kinase activity in control 23A2 myoblasts and in 23A2 myoblasts rendered differentiation-defective by the stable expression of an oncogenic (G12V)Ha-ras gene (Ras9 cells). The MAP kinase immunoprecipitated from Ras9 cells was 30-40% more active than that from control 23A2 cells. To establish if this elevated MAP kinase activity is essential to the maintenance of the oncogenic Ras-induced phenotype, we utilized the selective MAP kinase kinase 1 (MEK1) inhibitor, PD 098059. PD 098059 decreased the MAP kinase activity in Ras9 cells to the level found in 23A2 cells. PD 098059 did not affect the ability of 23A2 myoblasts to differentiate. PD 098059 reverted the transformed morphology of Ras9 cells but did not restore the ability of these cells to express the muscle-specific myosin heavy chain gene or to form muscle fibers. Treatment with PD 098059 also did not affect the ability of oncogenic Ha-Ras to establish a non-myogenic phenotype in C3H10T1/2 cells co-expressing MyoD. These results demonstrate that the activation of MAP kinase is necessary for the transformed morphology of Ras9 cells but is not required by oncogenic Ras to establish or to maintain a differentiation-defective phenotype. While these data do not rule out the possibility that constitutive signaling by MEK1 or MAP kinase could inhibit myoblast differentiation, they clearly demonstrate that other pathways activated by oncogenic Ras are sufficient to inhibit differentiation.

Assignment of the chicken MAX gene to chromosome 5p by fluorescence in situ hybridization.

It has been shown that the protein encoded by the MAX gene plays an important role in the physiological activity of Myc oncoproteins. In this study, we determined the chromosome location of the chicken MAX gene via fluorescence in situ hybridization. Hybridization of two biotinylated cloned fragments of 5.7 kb and 12.0 kb derived from the chicken MAX locus localized the gene to chromosome 5p. It is the third gene marker to be assigned to this telocentric macrochromosome. Since the MAX sequence is highly conserved both at the nucleotide and at the amino acid level in a wide range of vertebrate species, our data may provide evidence for the existence of a segmental homology between human and chicken chromosomes.

IFP 35 forms complexes with B-ATF, a member of the AP1 family of transcription factors.

Multicellular organisms achieve the spatial and temporal regulation of genes during growth and development through the differential expression of transcription factors that associate in various combinations. In this paper, we report the physical association of B-ATF, a member of the AP1 family of basic leucine zipper transcription factors, with IFP 35, a leucine zipper protein that is translocated to the nucleus following the treatment of cells with interferons and for which no binding partners previously have been described. Our data suggest that the formation of B-ATF:IFP 35 heterodimers is an interferon-inducible event in specialized cell types expressing both proteins and that changes in AP1 mediated gene transcription likely play a role in the response of these cells to interferons.

Casein kinase II increases the transcriptional activities of MRF4 and MyoD independently of their direct phosphorylation.

The myogenic regulatory factors (MRFs) are a subclass of a much larger group of basic helix-loop-helix transcription factors which includes members of the E protein such as E47, E2-2, and HEB. Although the MRFs are unique in their ability to confer a myogenic phenotype on nonmuscle cells, they require E protein partners to form a MRF-E protein heterodimer, which represents the functional myogenesis-inducing complex. The mechanisms controlling homodimer and heterodimer formation in vivo remain largely unknown, although it is likely that posttranslational modification of one or both basic helix-loop-helix partners is critical to this regulatory event. In this respect, MyoD and MRF4, both members of the MRF family, exist in vivo as phosphoproteins and contains multiple consensus phosphorylation sites, including sites for casein kinase II (CKII) phosphorylation. In this study, we demonstrate that overexpression of CKII increases the transcriptional activities of MRF4 and MyoD in vivo. Interestingly, mutation of the individual CKII sites within MRF4 and MyoF does not alter the ability of CKII to enhance MRF transcriptional activity, suggesting that the effect of CKII expression on the MRFs is indirect. Given that the MRFs require dimerization with E protein partners to activate muscle-specific transcription, the effects of CKII expression on E protein function also were examined. Our studies show that E47 serves as an in vitro substrate for CKII and that CKII-phosphorylated E-47 proteins no longer bind to DNA. These observations were confirmed by in vivo experiments showing that overexpressing of CKII produces a dramatic reduction in E47 homodimer-directed transcription. We conclude from these studies that CKII may act as a positive regulator of myogenesis by preventing E protein homodimers from binding to muscle gene regulatory elements.

B-ATF: a novel human bZIP protein that associates with members of the AP-1 transcription factor family.

A new member of the ATF/CREB family of transcription factors, called B-ATF, has been isolated from a cDNA library prepared from Epstein-Barr virus stimulated human B cells. B-ATF is a 125 amino acid nuclear protein possessing a basic leucine zipper domain that is most similar to the basic leucine zipper of ATF-3. Northern blot analysis of polyadenylated mRNA isolated from a variety of human tissues and established cell lines indicates that the 1.0 kilobase B-ATF mRNA is expressed differentially, with the strongest hybridization detected in lung and in Raji Burkitt's lymphoma. Efficient homodimerization of the B-ATF protein cannot be detected using the yeast two hybrid system or using in vitro binding assays with glutathione-s-transferase-B-ATF and maltose binding protein-B-ATF fusion proteins produced in E. coli. However, a yeast two hybrid library screen has identified the human oncoprotein JunB as a specific binding partner for B-ATF. Glutathione-s-transferase-B-ATF heterodimerizes efficiently with in vitro translated JunB, c-Jun, and JunD, but only weakly associates with c-Fos. In addition, electrophoretic mobility shift assays demonstrate that a B-ATF/c-Jun protein complex can interact with DNA containing a consensus binding site for AP-1, suggesting that B-ATF functions as a tissue-specific modulator of the AP-1 transcription complex in human cells.

Ras p21Val inhibits myogenesis without altering the DNA binding or transcriptional activities of the myogenic basic helix-loop-helix factors.

MRF4, MyoD, myogenin, and Myf-5 are muscle-specific basic helix-loop-helix transcription factors that share the ability to activate the expression of skeletal muscle genes such as those encoding alpha-actin, myosin heavy chain, and the acetylcholine receptor subunits. The muscle regulatory factors (MRFs) also exhibit the unique capacity to initiate the myogenic program when ectopically expressed in a variety of nonmuscle cell types, most notably C3H10T1/2 fibroblasts (10T1/2 cells). The commitment of myoblasts to terminal differentiation, although positively regulated by the MRFs, also is controlled negatively by a variety of agents, including several growth factors and oncoproteins such as fibroblast growth factor (FGF-2), transforming growth factor beta 1 (TGF-beta 1), and Ras p21Val. The molecular mechanisms by which these varied agents alter myogenic terminal differentiation events remain unclear. In an effort to establish whether Ras p21Val represses MRF activity by directly targeting the MRF proteins, we examined the DNA binding and transcription activation potentials of MRF4 and MyoD when expressed in 10T1/2 cells or in 10T1/2 cells expressing Ras p21Val. Our results demonstrate that Ras p21Val inhibits terminal differentiation events by targeting the basic domain of the MRFs, and yet the mechanism underlying this inhibition does not involve altering the DNA binding or the inherent transcriptional activity of these regulatory factors. In contrast, FGF-2 and TGF-beta 1 block terminal differentiation by repressing the transcriptional activity of the MRFs. We conclude that the Ras p21Val block in differentiation operates via an intracellular signaling pathway that is distinct from the FGF-2 and TGF-beta 1 pathways.