The Oct-2 protein binds cooperatively to adjacent octamer sites

An intrinsically disordered transcription activation domain increases the DNA binding affinity and reduces the specificity of NFκB p50/RelA

Many transcription factors contain intrinsically disordered transcription activation domains (TADs), which mediate interactions with coactivators to activate transcription. Historically, DNA-binding domains and TADs have been considered as modular units, but recent studies have shown that TADs can influence DNA binding. Whether these results can be generalized to more TADs is not clear. Here, we biophysically characterized the NFκB p50/RelA heterodimer including the RelA TAD and investigated the TAD's influence on NFκB-DNA interactions. In solution, we show the RelA TAD is disordered but compact, with helical tendency in two regions that interact with coactivators. We determined that the presence of the TAD increased the stoichiometry of NFκB-DNA complexes containing promoter DNA sequences with tandem κB recognition motifs by promoting the binding of NFκB dimers in excess of the number of κB sites. In addition, we measured the binding affinity of p50/RelA for DNA containing tandem κB sites and single κB sites. While the presence of the TAD enhanced the binding affinity of p50/RelA for all κB sequences tested, it also increased the affinity for nonspecific DNA sequences by over 10-fold, leading to an overall decrease in specificity for κB DNA sequences. In contrast, previous studies have generally reported that TADs decrease DNA-binding affinity and increase sequence specificity. Our results reveal a novel function of the RelA TAD in promoting binding to nonconsensus DNA, which sheds light on previous observations of extensive nonconsensus DNA binding by NFκB in vivo in response to strong inflammatory signals.

Diversity among POU transcription factors in chromatin recognition and cell fate reprogramming

The POU (Pit-Oct-Unc) protein family is an evolutionary ancient group of transcription factors (TFs) that bind specific DNA sequences to direct gene expression programs. The fundamental importance of POU TFs to orchestrate embryonic development and to direct cellular fate decisions is well established, but the molecular basis for this activity is insufficiently understood. POU TFs possess a bipartite 'two-in-one' DNA binding domain consisting of two independently folding structural units connected by a poorly conserved and flexible linker. Therefore, they represent a paradigmatic example to study the molecular basis for the functional versatility of TFs. Their modular architecture endows POU TFs with the capacity to accommodate alternative composite DNA sequences by adopting different quaternary structures. Moreover, associations with partner proteins crucially influence the selection of their DNA binding sites. The plentitude of DNA binding modes confers the ability to POU TFs to regulate distinct genes in the context of different cellular environments. Likewise, different binding modes of POU proteins to DNA could trigger alternative regulatory responses in the context of different genomic locations of the same cell. Prominent POU TFs such as Oct4, Brn2, Oct6 and Brn4 are not only essential regulators of development but have also been successfully employed to reprogram somatic cells to pluripotency and neural lineages. Here we review biochemical, structural, genomic and cellular reprogramming studies to examine how the ability of POU TFs to select regulatory DNA, alone or with partner factors, is tied to their capacity to epigenetically remodel chromatin and drive specific regulatory programs that give cells their identities.

A long-range flexible billboard model of gene activation

Gene regulation is fundamentally important for the coordination of diverse biologic processes including homeostasis and responses to developmental and environmental stimuli. Transcription factor (TF) binding sites are one of the major functional subunits of gene regulation. They are arranged in cis-regulatory modules (CRMs) that can be more active than the sum of their individual effects. Recently, we described a mechanism of glucocorticoid (GC)-induced gene regulation in which the glucocorticoid receptor (GR) binds coordinately to multiple CRMs that are 10s of kilobases apart in the genome. In those results, the minority of GR binding sites appear to involve direct TF:DNA interactions. Meanwhile, other GR binding sites in a cluster interact with those direct binding sites to tune their gene regulatory activity. Here, we consider the implications of those and related results in the context of existing models of gene regulation. Based on our analyses, we propose that the billboard and regulatory grammar models of cis-regulatory element activity be expanded to consider the influence of long-range interactions between cis-regulatory modules.

Assembly of the RFX complex on the MHCII promoter: role of RFXAP and RFXB in relieving autoinhibition of RFX5

The RFX complex is key component of a multi-protein complex that regulates the expression of the Major Histocompatibility Class II (MHCII) genes, whose products are essential for the initiation and development of the adaptive immune response. The RFX complex is comprised of three proteins--RFX5, RFXAP, and RFXB--all of which are required for expression of MHCII genes. We have used electrophoretic mobility shift assays to characterize the DNA binding of RFX5 and the complexes it forms with RFXB and RFXAP, to the proximal regulatory region of the MHCII promoter. DNA binding of RFX5 is inhibited by domains flanking its DNA binding domain, and both RFXAP and RFXB are required to overcome the inhibition of both domains. We provide evidence that a single RFX complex binds to the proximal regulatory region of the MHCII promoter and identify regions of the DNA that are important for high affinity binding of the RFX complex. Together, our results provide the most detailed view to date of the assembly of the RFX complex on the MHCII promoter and how its DNA binding is regulated.

Oct transcription factors mediate t(14;18) lymphoma cell survival by directly regulating bcl-2 expression

Oct-1 and Oct-2 are members of the POU homeodomain family of transcriptional regulators and are critical for normal embryonic development. Gene-targeting studies showed that Oct-1 and Oct-2 are largely dispensable for B-cell development and immunoglobulin production, although both Oct-2 and Bob-1 are required for a proper immune response and germinal center formation. In these studies, we investigated the role of Oct factors in B-cell lymphomas. Recent investigations have shown increased expression of Oct-2 and Bob-1 in lymphomas, and we observed greatly increased levels of Oct-2 in lymphoma cells with the t(14;18) translocation. Decreased expression of Oct-1, Oct-2, or Bob-1 by RNA interference resulted in apoptosis and down-regulation of bcl-2 expression. Furthermore, Oct-2 induced bcl-2 promoter activity and mediated this effect through three regions in the bcl-2 P2 promoter. Although these regions did not contain canonical octamer motifs, we observed the direct interaction of Oct-2 with all three sites both in vitro by EMSA and in vivo by chromatin immunoprecipitation assay. Moreover, by mutation analysis we found that the ability of Oct-2 to activate bcl-2 required C/EBP, Cdx, and TATA-binding sites. Oct-2, therefore, acts as a cell survival factor in t(14;18) lymphoma cells by directly activating the antiapoptotic gene bcl-2.

Mouse lymphoid cell line selected to have high immunoglobulin promoter activity

Immunoglobulin variable region promoters are predominantly B-cell specific, but the molecular basis for this specificity has not been elucidated. To further understand how B-cell-specific immunoglobulin promoter expression is mediated, the murine lymphoid cell line 2017 was engineered to express the green fluorescent protein under the control of an immunoglobulin heavy chain promoter and selected for high activity using multiple rounds of fluorescence-activated cell sorting. Rare clones with intense and stable immunoglobulin promoter activity were isolated. Transient transfection experiments demonstrated that two different immunoglobulin promoters and two other B-cell-specific promoters have higher activities in the selected cell lines relative to the parental line and to the non-cell-type-specific histone H2B promoter. The increased immunoglobulin activity required nucleotide residues downstream of the transcription initiation site which were also important for maximal activity in B cells and which were conserved in other B-cell-specific promoters. Unlike the unselected cells, the 2017 variants also showed activation of their endogenous immunoglobulin heavy chain variable regions.

Optimal Oct-2 affinity for an extended DNA site and the effect of GST fusion on site preference

The regulator of immunoglobulin expression Oct-2 and the related widely expressed transcription factor Oct-1 have been shown to interact with DNA sequences containing an "octamer" motif, ATGC(A/T)AAT. To better understand Oct-2 function we have used random oligonucleotide selection and competition assays to define the optimal recognition site for this protein. The selected site contains an extended sequence that is remarkably similar to octamer-heptamer sequences found in immunoglobulin heavy-chain regulatory sequences, and the affinity of Oct-2 for this site is at least 50-fold greater than for sites containing the octamer motif alone. Fusion to glutathione S-transferase, a widely used model for protein-DNA and protein-protein interaction, does not alter the optimal Oct-2 recognition site, but inhibits Oct-2 POU-domain dimerization, slows the dissociation rate of the GST-Oct-2/DNA complex, and increases the relative importance of the heptamer domain for Oct-2 binding. These data advance our ability to identify in vivo targets of POU-factor regulation and also suggest that GST-fusion proteins should be used with caution in DNA-binding studies.

Atypical binding of the neuronal POU protein N-Oct3 to noncanonical DNA targets. Implications for heterodimerization with HNF-3 beta

The capacity of POU proteins to recognize different DNA sequences and to bind target DNA in the form of monomers, cooperative dimers or heterodimers is important in relation to their transcriptional regulatory properties. The N-Oct3 neuron-specific protein binds to an octamer-like sequence (AAATAATGC) within the (-102/-72) neuronal promoter region of the human aromatic L-amino acid decarboxylase (AADC) gene. In this atypical case the POUh and POUs tetrameric subsites are spaced one nucleotide apart and in switched order as compared with the consensus octamer. Moreover this POU binding motif overlaps the hepatocyte nuclear factor HNF-3 beta binding site (TGCTCAGTAAA) which itself contains a heptamer-like sequence (CTCAGTA). Using the isolated DNA binding domains (DBD) of the two proteins, it is shown that, when binding to this unusual recognition sequence, N-Oct3 either exhibits noncooperative homodimerization or allows the simultaneous binding of the second transcription activator HNF-3 beta. CD studies indicate that the binding of N-Oct3 monomers/dimers and N-Oct3-HNF-3 beta heterodimers to the DNA induces conformational changes of both protein and DNA. Partial proteolysis/MALDI-MS was used in conjunction with molecular modelling to show that the protein conformational change resulting from binary N-Oct3/DNA complex formation occurs within the linker peptide joining the POUs and POUh subdomains. Furthermore, modelling the N-Oct3/HNF-3 beta/DNA ternary complex predicts a nucleotide rearrangement in the overlap region and an interaction between both transcription factors. In the light of our findings, which illustrate both site-dependent and site-independent protein and DNA conformational changes, general implications for the allosteric function of DNA response elements in transcriptional regulation are discussed.

POU domain factors in the neuroendocrine system: lessons from developmental biology provide insights into human disease

POU domain factors are transcriptional regulators characterized by a highly conserved DNA-binding domain referred to as the POU domain. The structure of the POU domain has been solved, facilitating the understanding of how these proteins bind to DNA and regulate transcription via complex protein-protein interactions. Several members of the POU domain family have been implicated in the control of development and function of the neuroendocrine system. Such roles have been most clearly established for Pit-1, which is required for formation of somatotropes, lactotropes, and thyrotropes in the anterior pituitary gland, and for Brn-2, which is critical for formation of magnocellular and parvocellular neurons in the paraventricular and supraoptic nuclei of the hypothalamus. While genetic evidence is lacking, molecular biology experiments have implicated several other POU factors in the regulation of gene expression in the hypothalamus and pituitary gland. Pit-1 mutations in humans cause combined pituitary hormone deficiency similar to that found in mice deleted for the Pit-1 gene, providing a striking example of how basic developmental biology studies have provided important insights into human disease.

Conserved gene structure and transcription factor sites in the human and mouse deoxycytidine kinase genes

Deoxycytidine kinase (dCK) phosphorylates several anti-cancer and anti-viral nucleoside analogs. The enzyme is predominantly expressed in lymphoid tissues regulated by an unknown mechanism. We have cloned and sequenced the 20 kbp mouse dCK gene and approximately 1.7 kbp of the 5' flanking regions of both the human and mouse dCK genes. Five major inter-species conserved motifs were identified in the 5' region including the transcription initiation region, an SP1 site and two closely located putative octamer transcription factor sites. Luciferase reporter experiments showed that the human dCK 5' region efficiently initiated transcription but no tissue regulatory element could be identified.

Synergism with the coactivator OBF-1 (OCA-B, BOB-1) is mediated by a specific POU dimer configuration

POU domain proteins contain a bipartite DNA binding domain divided by a flexible linker that enables them to adopt various monomer configurations on DNA. The versatility of POU protein operation is additionally conferred at the dimerization level. The POU dimer formed on the PORE (ATTTGAAATGCAAAT) can recruit the transcriptional coactivator OBF-1, whereas POU dimers formed on the consensus MORE (ATGCATATGCAT) or on MOREs from immunoglobulin heavy chain promoters (AT[G/A][C/A]ATATGCAA) fail to interact. An interaction with OBF-1 is precluded since the same Oct-1 residues that form the MORE dimerization interface are also used for OBF-1/Oct-1 interactions on the PORE. Our findings provide a paradigm of how specific POU dimer assemblies can differentially recruit a coregulatory activity with distinct transcriptional readouts.

Requirement of LIM domains for LIM1 function in mouse head development

Lim1, also known as Lhx1, encodes a LIM homeodomain transcription factor that is essential for head development in the mouse. As with other LIM homeodomain proteins, LIM1 has two LIM domains located N-terminal to the homeodomain, with each LIM domain containing two zinc finger motifs. LIM domains can physically interact with other proteins to form protein complexes that regulate transcription. Previous studies have suggested that LIM domains negatively regulate the transcriptional activity of their associated homeodomains. To investigate the requirement of LIM domains for LIM1 activity, we have mutated the Lim1 gene to alter the conserved amino acid residues that are required for zinc finger structure within both of the LIM domains. Although mice homozygous for this Lim1 allele express the mutant mRNA and protein appropriately, they are a phenocopy for Lim1-null mice. These results suggest that the integrity of the LIM domains is essential for LIM1 activity in mouse head development. genesis 27:12-21, 2000.

Regulation of human involucrin promoter activity by a protein kinase C, Ras, MEKK1, MEK3, p38/RK, AP1 signal transduction pathway

Involucrin is a marker of keratinocyte terminal differentiation. Our previous studies show that involucrin mRNA levels are increased by the keratinocyte differentiating agent, 12-O-tetradecanoylphorbol-13-acetate (TPA) (Welter, J. F., Crish, J. F., Agarwal, C., and Eckert, R. L. (1995) J. Biol. Chem. 270, 12614-12622). We now study the signaling cascade responsible for this regulation. Protein kinase C and tyrosine kinase inhibitors inhibit both the TPA-dependent mRNA increase and the TPA-dependent increase in hINV promoter activity. The relevant response element is located within the promoter proximal regulatory region and includes an AP1 site, AP1-1. Co-transfection of the hINV promoter with dominant negative forms of Ras, MEKK1, MEK1, MEK7, MEK3, p38/RK, and c-Jun inhibit the TPA-dependent increase. Wild type MEKK1 enhances promoter activity and the activity can be inhibited by dominant negative MEKK1, MEK1, MEK7, MEK3, p38/RK, and c-Jun. In contrast, wild type Raf-1, ERK1, ERK2, MEK4, or JNK1 produced no change in activity and the dominant negative forms of these kinases failed to suppress TPA-dependent transcription. Treatment with an S6 kinase (S6K) inhibitor, or transfection with constitutively active S6K produced relatively minor changes in promoter activity, ruling out a regulatory role for S6K. These results suggest that activation of involucrin transcription involves a pathway that includes protein kinase C, Ras, MEKK1, MEK3, and p38/RK. Additional pathways that transfer MEKK1 activation via MEK1 and MEK7 also may function, but the downstream targets of these kinases need to be identified. AP1 transcription factors appear to be the ultimate target of this regulation.

DNA bending is essential for the site-specific recognition of DNA response elements by the DNA binding domain of the tumor suppressor protein p53

We have used circular permutation assays to determine the extent and location of the DNA bend induced by the DNA binding domain of human wild type p53 (p53DBD) upon binding to several naturally occurring DNA response elements. We have found that p53DBD binding induces axial bending in all of the response elements investigated. In particular, response elements having a d(CATG) sequence at the junction of two consensus pentamers in each half-site favor highly bent complexes (bending angle is approximately 50 degrees ), whereas response elements having d(CTTG) bases at this position are less bent (bending angles from approximately 37 to approximately 25 degrees ). Quantitative electrophoretic mobility shift assays of different complexes show a direct correlation between the DNA bending angle and the binding affinity of the p53DBD with the response elements, i.e. the greater the stability of the complex, the more the DNA is bent by p53DBD binding. The study provides evidence that the energetics of DNA bending, as determined by the presence or absence of flexible sites in the response elements, may contribute significantly to the overall binding affinity of the p53DBD for different sequences. The results therefore suggest that both the structure and the stability of the p53-DNA complex may vary with different response elements. This variability may be correlated with variability in p53 function.

Structure of Pit-1 POU domain bound to DNA as a dimer: unexpected arrangement and flexibility

Pit-1, a member of the POU domain family of transcription factors, characterized by a bipartite DNA-binding domain, serves critical developmental functions based on binding to diverse DNA elements in its target genes. Here we report a high resolution X-ray analysis of the Pit-1 POU domain bound to a DNA element as a homodimer. This analysis reveals that Pit-1 subdomains bind to perpendicular faces of the DNA, rather than opposite faces of the DNA as in Oct-1. This is accomplished by different spacing and orientation of the POU-specific domain. Contrary to previous predictions, the dimerization interface involves the carboxyl terminus of the DNA recognition helix of the homeodomain, which in an extended conformation interacts with specific residues at the amino terminus of helix alpha1 and in the loop between helices alpha3 and alpha4 of the POU-specific domain of the symmetry related monomer. These features suggest the molecular basis of disease-causing mutations in Pit-1 and provide potential basis for the flexible allostery between protein domains and DNA sites in the activation of target genes.

Eukaryotic transcription factor-DNA complexes

Eukaryotes have three distinct RNA polymerases that catalyze transcription of nuclear genes. RNA polymerase II is responsible for transcribing nuclear genes encoding the messenger RNAs and several small nuclear RNAs. Like RNA polymerases I and III, polymerase II cannot recognize its target promoter directly and initiate transcription without accessory factors. Instead, this large multisubunit enzyme relies on general transcription factors and transcriptional activators and coactivators to regulate transcription from class II promoters. X-ray crystallography and nuclear magnetic resonance spectroscopy have been used to study complexes of general transcription factors and transcriptional activators with their specific DNA targets. This work has provided important structural insights into transcription initiation by polymerase II and the more general problem of DNA sequence recognition.

Regulation of human involucrin promoter activity by POU domain proteins

POU domain transcription factors are expressed in the epidermis and are thought to be important regulators of keratinocyte gene expression. In the present article we demonstrate that POU transcription factors suppress transcription of the human involucrin (hINV) promoter. Cotransfection of pINV-2473, a construct containing 2473 base pairs of hINV upstream sequence linked to luciferase, with POU homeodomain transcription factors Oct1, Oct2, Brn4, SCIP, Skn1a or Skn1i, results in a strong suppression of basal promoter activity. The hINV upstream region includes a consensus POU transcription factor binding site, 5'-ATGCAAAT-3', centered around nucleotide -1277. Although this site interacts with POU factors, assays of promoter activity for a series of progressive 5' end truncations demonstrate that this site is not required for POU factor-dependent transcriptional suppression. Suppression is observed with the shortest truncation construct tested, pINV-41, suggesting that this inhibition may be mediated by effects on TATA box proteins. SCIP mutants that lack transactivation or DNA binding domains were shown to suppress transcription, suggesting that the DNA binding and transactivation domains are not required for suppression. Moreover, cotransfection of the pINV-2473 with pKSM13(+)OCT, which contains a single consensus OCT binding site, results in an increase in basal promoter activity, suggesting that endogenous POU factors suppress hINV promoter activity. In addition to inhibiting basal transcription, POU transcription factors also suppress phorbol ester-stimulated hINV promoter activity. These studies suggest that suppression of hINV promoter activity does not require the amino-terminal segment of the POU factor or direct POU factor interaction with DNA and suggest that the suppression may be via indirect interaction with other proteins in the vicinity of the TATA box. Thus, involucrin joins the ranks of a small set of genes that are regulated by POU factors in an octamer binding site-independent manner.

The Oct-1 POU-specific domain can stimulate small nuclear RNA gene transcription by stabilizing the basal transcription complex SNAPc

The RNA polymerase II and III human small nuclear RNA promoters have a common basal element, the proximal sequence element, which binds the TATA box-binding protein-containing complex SNAPc. They also contain an enhancer characterized by a highly conserved octamer sequence, which constitutes a binding site for the broadly expressed POU domain transcription factor Oct-1. The POU domain is a bipartite DNA-binding domain consisting of a POU-homeo (POUH) domain and a POU-specific (POUs) domain joined by a flexible linker. Here, we show that the Oct-1 POU domain but not the related Pit-1 POU domain can facilitate the binding of SNAPc to the proximal sequence element, and activate transcription. The effect is probably mediated by protein-protein contacts, and 1 of 30 amino acid differences between the Oct-1 and Pit-1 POUs domains is the key determinant for the differential interaction with SNAPc and the ability to activate transcription. These results show that a function that is the hallmark of activation domains, namely, recruitment of a basal transcription complex resulting in activation of transcription, can be performed by a DNA-binding domain. In this case, subtle changes between activator DNA-binding domains, as subtle as a single amino acid difference, can profoundly affect interaction with the basal transcription machinery.

Footprint analysis of estrogen receptor binding to adjacent estrogen response elements

Quantitative DNase I footprinting assays were employed to simultaneously measure the amount of estrogen receptor (ER) bound to each site in constructs containing multiple estrogen response elements (EREs). These assays revealed identical, high affinity ER-ERE binding, Kd of approximately 0.25 nM, for estradiol-liganded ER (E2-ER), 4-hydroxytamoxifen liganded ER (4-OHT-ER), tamoxifen aziridine liganded ER (TAz-ER), and unliganded dimeric ER, for each ERE in constructs containing up to four tandem EREs. Increasing concentrations of ER resulted in the same pattern of occupancy for each ERE, whether or not the site was located near other EREs. Similarly, the presence or absence of E2, 4-OHT, or TAz ligand did not change ER-ERE interaction. Since activated ER-ERE binding affinity is identical, whether ER is liganded or unliganded, ligand cannot regulate ER-ERE binding affinity. These results support the hypothesis that ligand-dependent conformational changes primarily determine how ER interacts with components of the transcription initiation complex that mediate gene transactivation. In addition, footprint assays revealed that, following ER binding, an AT-rich site adjacent to the ERE becomes hypersensitive to DNase I digestion. This sequence may be easily or intrinsically bent, assisting in recruiting ER to ERE sites.

Transcription of the lung-specific surfactant protein C gene is mediated by thyroid transcription factor 1

Surfactant protein C (SP-C) is expressed in alveolar Type II epithelial cells of the lung. In order to determine the mechanism(s) that regulate gene transcription, we have analyzed the activation of the murine SP-C promoter in mouse lung epithelial cells (MLE cells) and in HeLa cells after co-transfection with a vector expressing rat thyroid transcription factor-1 (TTF-1). TTF-1 transactivated SP-C-chloramphenicol acetyltransferase constructs containing -13 kilobase pairs to -320 base pairs (bp) of the 5 flanking region of the SP-C gene. Essential cis-acting elements were functionally localized to between -320 and -180 bp from the start of transcription by transfection analysis. Five DNase-protected regions, indicating multiple protein-DNA interactions within the -320 bp TTF-1-responsive region of the SP-C gene, were identified by DNase footprint analysis. A 40-bp segment of SP-C DNA from -197 to -158 linked to a heterologous promoter-chloramphenicol acetyltransferase construct activated expression after co-transfection with CMV-TTF-1 in HeLa and MLE cells. The -197 to -158 segment contained two consensus TTF-1 sites, which were specifically identified as TTF-1 binding sites by gel retardation and antibody supershift with MLE cell nuclear extracts and purified TTF-1 homeodomain protein. Site-specific mutagenesis of either of the TTF-1 binding sites completely blocked activation by TTF-1, indicating both sites are required for TTF stimulation of SP-C transcription.

Crystallization and preliminary X-ray analysis of Pit-1 POU domain complexed to a 28 base pair DNA element

The POU domain, representing an approximately 150 amino acid conserved region, serves as the DNA-recognition domain for a large number of eukaryotic transcription factors. Bipartite in nature, the POU domain is comprised of a N-terminal POU-specific domain connected by a linker of variable length to a C-terminal homeodomain. We report here co-crystals of pituitary-specific factor Pit-1 POU domain bound as a dimer to a 28 bp DNA fragment. The crystals diffract to at least 2.3 angstroms in resolution and belong to space group P1 with unit cell dimensions of a = 42.5 angstroms, b = 50.1 angstroms, c = 55.8 angstroms, alpha = 76.7 degrees, beta = 79.3 degrees, and gamma = 67.2 degrees.

Extended theory of the electrophoretic mobility-shift analysis of nonspecific protein-DNA complexes, featuring cooperativity

The simulated electrophoretic mobility-shift behavior of a model system, in which the nonspecific binding of a protein to a DNA fragment is cooperative, was compared with the experimental behavior of the DNA: histone-like bacterial protein (HU) system. It was concluded that the binding of HU to an 88 bp DNA fragment is, at least, not highly cooperative. The theory of mobility-shift analysis was extended even further to encompass high affinity sequence-specific binding of protein to a DNA fragment followed by weak nonspecific binding, the latter governed by conditional probabilities. In addition to featuring a ladder of incremental protein-DNA complexes, the computed mobility-shift patterns placed emphasis upon stabilization of weak, nonspecific complexes in gel cages.

A compilation of composite regulatory elements affecting gene transcription in vertebrates

Over the past years, evidence has been accumulating for a fundamental role of protein-protein interactions between transcription factors in gene-specific transcription regulation. Many of these interactions run within composite elements containing binding sites for several factors. We have selected 101 composite regulatory elements identified experimentally in the regulatory regions of 64 genes of vertebrates and of their viruses and briefly described them in a compilation. Of these, 82 composite elements are of the synergistic type and 19 of the antagonistic type. Within the synergistic type composite elements, transcription factors bind to the corresponding sites simultaneously, thus cooperatively activating transcription. The factors, binding to their target sites within antagonistic type composite elements, produce opposing effects on transcription. The nucleotide sequence and localization in the genes, the names and brief description of transcription factors, are provided for each composite element, including a representation of experimental data on its functioning. Most of the composite elements (3/4) fall between -250 bp and the transcription start site. The distance between the binding sites within the composite elements described varies from complete overlapping to 80 bp. The compilation of composite elements is presented in the database COMPEL which is electronically accessible by anonymous ftp via internet.

Upstream enhancer activity in the human surfactant protein B gene is mediated by thyroid transcription factor 1

Surfactant protein B (SP-B) is selectively expressed in bronchiolar and alveolar epithelial cells of the lung. We identified an upstream enhancer located in the 5'-flanking region of the human SP-B gene (-439 to -331 base pair, hSP-B(-439/-331)) by deletion analysis of SP-B-luciferase constructs assessed in transfection assays in vitro. The element cis-activated the expression of an SV40 promoter-luciferase reporter gene in a human pulmonary adenocarcinoma cell line (H441-4). Three distinct binding sites for the nuclear transcription protein, thyroid transcription factor 1 (TTF-1), were identified, and the purified TTF-1 homeodomain was bound to bhe region of hSP-B(-439/-331). Co-transfection of H441-4 cells with the expression vector pCMV-TTF-1 trans-activated the native human SP-B promoter and the SV40 promoter fused with the SP-B enhancer. Mutations of the TTF-1 binding sites in the upstream enhancer blocked TTF-1 binding and transactivation activity. In summary, TTF-1 interacts with distinct proximal (-80 to -110) and distal (-439 to -331) cis-acting elements than regulate lung epithelial cell-specific transcription of the human SP-B gene.

PkpA, a novel Phycomyces blakesleeanus serine/threonine protein kinase

This work reports the cloning and sequencing of pkpA, a gene of the filamentous fungus Phycomyces blakesleeanus, whose expression seems to be coupled to vegetative growth. This gene encodes a putative serine/threonine-specific protein kinase, whose sequence is related to that of the yeast protein STE20, involved in pheromone-response pathways, and to a number of MAPK kinase proteins. However, detailed analysis of the kinase sequence suggests that PkpA is a novel serine/threonine protein kinase that probably participates as an intermediate in an intracellular system controlling nuclear proliferation in P. blakesleeanus.

The essential DNA-binding protein sap1 of Schizosaccharomyces pombe contains two independent oligomerization interfaces that dictate the relative orientation of the DNA-binding domain

The sap1 gene from Schizosaccharomyces pombe, which is essential for mating-type switching and for growth, encodes a sequence-specific DNA-binding protein with no homology to other known proteins. We have used a reiterative selection procedure to isolate binding sites for sap1, using a bacterially expressed protein and randomized double-strand oligonucleotides. The sap1 homodimer preferentially selects a pentameric motif, TA(A/G)CG, organized as a direct repeat and spaced by 5 nucleotides. Removal of a C-terminal dimerization domain abolishes recognition of the direct repeat and creates a new specificity for a DNA sequence containing the same pentameric motif but organized as an inverted repeat. We present evidence that the orientation of the DNA-binding domain is controlled by two independent oligomerization interfaces. The C-terminal dimerization domain allows a head-to-tail organization of the DNA-binding domains in solution, while an N-terminal domain is involved in a cooperative interaction on the DNA target between pairs of dimers.

Involvement of the Ets family factor PU.1 in the activation of immunoglobulin promoters

The B cell-specific expression of immunoglobulin (Ig) genes is controlled by the concerted action of variable (V) region promoters and intronic or 3' enhancers, all of which are active in a lymphoid-specific manner. A crucial highly conserved element of the V region promoters is the octamer site -ATTTGCAT-, which can be bound by ubiquitous (Oct-1) as well as B cell-specific (Oct-2) factors. Another less conserved element found in many Ig promoters is pyrimidine-rich and has been shown to be functionally important, in particular for those Ig promoters that have only an imperfect octamer site. In this study we have analyzed the factors binding specifically to the pyrimidine-rich motif of the V kappa 19 promoter, a light chain gene promoter with an imperfect octamer site. Using nuclear extracts prepared from B cells, we detected two sets of specific complexes in electrophoretic mobility shift experiments. One complex appears to be ubiquitous but enriched in lymphoid cells and represents the binding of a potentially novel factor with an apparent molecular mass of approximately 50 kDa. The other complex was found only with extracts from pre-B or B cells as well as from a macrophage cell line and appears to be caused by the binding of PU.1, a factor of the Ets family. We show that on this Ig promoter Oct factors (Oct-1 or Oct-2) and PU.1 can bind concomitantly but without synergism. By transfection experiments in non-B cells we demonstrate that PU.1 is indeed able to activate this promoter in concert with Oct-2. Furthermore, we show that PU.1 can bind with varying affinities to the pyrimidine-rich elements of several other Ig promoters. These data suggest a more general role for PU.1 or other members of the Ets family in the activation of Ig promoters.

Brain and testis selective expression of the glutathione S-transferase Yb3 subunit is governed by tandem direct repeat octamer motifs in the 5'-flanking region of its gene

To gain insight into mechanisms of cell type-specific transcription of class mu-glutathione S-transferase genes, the gene encoding the Yb3 subunit was cloned. Yb3 subunits are selectively expressed at high levels in rat brain and testis but not in liver or kidney. The Yb3 subunit gene spans over 6 kb and consists of 8 exons and 7 introns and a sequence consisting of tandem direct repeat consensus octamer DNA binding motifs separated by a 6 base pair (bp) spacer was identified in its 5'-flanking region. Gel shift assays with a 40 bp segment of DNA containing the two consensus octamer sequences, revealed the presence of specific binding proteins in nuclear extracts of rat brain, testis and C6 glioma cells. DNA binding activity was greatly reduced in liver, kidney and HTC cells. Reporter vectors carrying segments of the 5'-flanking region of the Yb3 subunit gene fused to a luciferase gene were introduced into C6 glioma cells which express high levels of Yb3 subunits, and into HTC cells which do not. The plasmids consisting of the Yb3 gene promoter up to, but not including, the octamer motifs did not support luciferase transcription in the C6 glioma cells, but larger fragments that included the octamer repeat sequences, effectively directed transcription in the C6 glioma cells. With mutated octameric sequences transcriptional activity was greatly reduced, and none of the same Yb3 constructs directed substantial luciferase transcription in the HTC cells. The results show that octamer motifs in the 5'-flanking region of the Yb3 subunit gene are functional and are the principal cis-acting elements that account for its discrete cell type-selective expression. This gene is one of the few known targets for octamer DNA binding transcription factors in brain.

Homeodomain proteins. Cooperating to be different

The product of the Drosophila extradenticle gene interacts cooperatively with homeodomain proteins encoded by homeotic selector genes, and may account in part for their distinct regulatory properties.

Functional differences between the Oct2 transactivation domains determine the transactivation potential of individual Oct2 isoforms

The lymphocyte specific transcription factor Oct2 is involved in mediating the B-cell specific transcriptional activity of the octamer motif. Mutational analyses in the context of the complete Oct2 protein had indicated that Oct2 contains two transactivation domains. These two domains appeared to be redundant for activation from a promoter proximal position, whereas stimulation from a remote enhancer position specifically required the C-terminal transactivation domain and an additional B-cell restricted activity. We have generated fusion proteins between the DNA binding domain of the yeast Gal4 transcription factor and individual Oct2 protein domains to analyze their transactivation potential separately. We show that both N- and C-terminal domains can stimulate transcription from a promoter proximal position independently. However, only the C-terminal transactivation domain activates from a distance and it can only do so in B-cells. The C-terminal transactivation domain represents a composite transactivation domain. Whereas removal of just 9 aminoacids from the extreme C-terminus lead to a complete inactivation of this domain deletions from the other side resulted in a gradual loss of activity. We also characterized the transactivation potential of different N-terminal regions of Oct2 generated by alternative splicing. We show that the N-terminus of one of the isoforms, Oct2.3, contains a negative regulatory domain (NRD), which can inactivate the neighbouring glutamine-rich transactivation in cis. The presence of this NRD affects the overall phosphorylation state of the Oct2 protein. This result suggests that the mechanism of inactivation might involve differential protein phosphorylation.

The human papillomavirus type 16 E2 transcription factor binds with low cooperativity to two flanking sites and represses the E6 promoter through displacement of Sp1 and TFIID

The E6 promoters of all genital human papillomaviruses have a characteristic alignment of transcription factor binding sites. Activation of the basic transcription complex at the TATA box depends upon a sequence-aberrant Sp1 site. Repression of E6 promoters is achieved by two binding sites for the viral E2 protein positioned between the Sp1 site and the TATA box. We have purified the human papillomavirus type 16 E2 protein after expression in Escherichia coli and studied its binding and repression properties with oligonucleotides representing the homologous promoter sequences. A Kd value of 3 x 10(-10) M indicated binding properties expected for a native protein. We found low cooperativity in the binding of two E2 dimers to flanking sites, both when these sites were separated by 3 nucleotides, as in the natural promoter, and when they were further apart. E2 protein, bound close to the distal Sp1 site, displaced the Sp1 factor even when the aberrant sequence was replaced by a typical Sp1 core recognition site. The high affinity of E2 protein for its binding site even led to Sp1 displacement at concentrations of E2 protein nearly 2 orders of magnitude lower than those of Sp1. Functional analyses of mutated E6 promoter sequences showed repression by this distal E2 binding site in the complete absence of binding to the proximal E2 binding site. From our findings and observations published by others, we conclude that each of the E2 binding sites in the E6 promoter of genital human papillomaviruses plays a separate role by displacing the transcription factors Sp1 and TFIID.

Reconstitution of transcriptional activation domains by reiteration of short peptide segments reveals the modular organization of a glutamine-rich activation domain

The POU domain activator Oct-2 contains an N-terminal glutamine-rich transcriptional activation domain. An 18-amino-acid segment (Q18III) from this region reconstituted a fully functional activation domain when tandemly reiterated and fused to either the Oct-2 or GAL4 DNA-binding domain. A minimal transcriptional activation domain likely requires three tandem Q18III segments, because one or two tandem Q18III segments displayed little activity, whereas three to five tandem segments were active and displayed increasing activity with increasing copy number. As with natural Oct-2 activation domains, in our assay a reiterated activation domain required a second homologous or heterologous activation domain to stimulate transcription effectively when fused to the Oct-2 POU domain. These results suggest that there are different levels of synergy within and among activation domains. Analysis of reiterated activation domains containing mutated Q18III segments revealed that leucines and glutamines, but not serines or threonines, are critical for activity in vivo. Curiously, several reiterated activation domains that were inactive in vivo were active in vitro, suggesting that there are significant functional differences in our in vivo and in vitro assays. Reiteration of a second 18-amino-acid segment from the Oct-2 glutamine-rich activation domain (Q18II) was also active, but its activity was DNA-binding domain specific, because it was active when fused to the GAL4 than to the Oct-2 DNA-binding domain. The ability of separate short peptide segments derived from a single transcriptional activation domain to activate transcription after tandem reiteration emphasizes the flexible and modular nature of a transcriptional activation domain.

Reconstitution of transcriptional activation domains by reiteration of short peptide segments reveals the modular organization of a glutamine-rich activation domain

The POU domain activator Oct-2 contains an N-terminal glutamine-rich transcriptional activation domain. An 18-amino-acid segment (Q18III) from this region reconstituted a fully functional activation domain when tandemly reiterated and fused to either the Oct-2 or GAL4 DNA-binding domain. A minimal transcriptional activation domain likely requires three tandem Q18III segments, because one or two tandem Q18III segments displayed little activity, whereas three to five tandem segments were active and displayed increasing activity with increasing copy number. As with natural Oct-2 activation domains, in our assay a reiterated activation domain required a second homologous or heterologous activation domain to stimulate transcription effectively when fused to the Oct-2 POU domain. These results suggest that there are different levels of synergy within and among activation domains. Analysis of reiterated activation domains containing mutated Q18III segments revealed that leucines and glutamines, but not serines or threonines, are critical for activity in vivo. Curiously, several reiterated activation domains that were inactive in vivo were active in vitro, suggesting that there are significant functional differences in our in vivo and in vitro assays. Reiteration of a second 18-amino-acid segment from the Oct-2 glutamine-rich activation domain (Q18II) was also active, but its activity was DNA-binding domain specific, because it was active when fused to the GAL4 than to the Oct-2 DNA-binding domain. The ability of separate short peptide segments derived from a single transcriptional activation domain to activate transcription after tandem reiteration emphasizes the flexible and modular nature of a transcriptional activation domain.

A critical role for chromatin in mounting a synergistic transcriptional response to GAL4-VP16

The role of chromatin in mounting a synergistic transcriptional response to GAL4-VP16 was investigated. Strong synergy was observed when chromatin templates were used in vitro. The synergy was severely reduced when naked DNA templates were transcribed. In vivo synergy was strong when nonreplicating templates were used. However, the use of replicating templates, which involved transient disruptions of chromatin, led to strong reductions in synergy. In both of these low-synergy responses, transcription levels were high. We infer that strong synergy has a requirement for chromatin that may be understood in terms of the competition between multiple activator molecules and histone cores for promoter DNA.

A critical role for chromatin in mounting a synergistic transcriptional response to GAL4-VP16

The role of chromatin in mounting a synergistic transcriptional response to GAL4-VP16 was investigated. Strong synergy was observed when chromatin templates were used in vitro. The synergy was severely reduced when naked DNA templates were transcribed. In vivo synergy was strong when nonreplicating templates were used. However, the use of replicating templates, which involved transient disruptions of chromatin, led to strong reductions in synergy. In both of these low-synergy responses, transcription levels were high. We infer that strong synergy has a requirement for chromatin that may be understood in terms of the competition between multiple activator molecules and histone cores for promoter DNA.

Structure and genomic organization of VH gene segments in the channel catfish: members of different VH gene families are interspersed and closely linked

To determine the structure and organization of germline VH gene segments in the channel catfish, genomic lambda libraries were screened with cDNA probes representing different catfish VH gene families. Thirty-six VH positive genomic clones were isolated and four of these were characterized by restriction mapping and Southern blot analysis with probes specific for each known VH gene family. The four clones, representing about 65 kb of DNA, contained 21 VH segments. The average distance between segments was about 3 kb and gene segments representing different VH gene families were interspersed with each other. Dot-blot hybridization analysis of all 36 genomic clones (average insert size 16-18 kb) indicated that the average clone contained gene segments representing four different VH families. In addition, these analyses indicated that VH segments representing each VH family could be found closely linked to gene segments representing each of the other VH families. Genomic restriction fragments containing a VH segment of each gene family were sequenced. These analyses showed that the general structure of VH segments is conserved in catfish. These structural features include the presence of a leader sequence split by a short intron, an uninterrupted open reading frame encoding readily identified framework and complementarity determining regions, and a downstream recombination signal sequence represented by a consensus heptamer, a 22-24 bp spacer, and an A-rich nonamer. Upstream of the VH segments was an octamer sequence. These analyses indicate that the organization and structure of VH segments typically associated with VH loci of higher vertebrates evolved early in phylogeny at the level of the bony fishes.

The product of the adenovirus intermediate gene IVa2 is a transcriptional activator of the major late promoter

During the course of lytic infection, the adenovirus major late promoter (MLP) is induced to high levels after replication of viral DNA has started. We had previously shown that sequence elements located downstream of the MLP start site were implicated in this late-specific transcriptional activation (DE1, between +85 and +98; DE2, between +100 and +120). Two positive transcription factors involved in this activation have been detected. DEF-A, which specifically binds to DE1 and also to the 3' portion of DE2 (DE2a), and DEF-B, which interacts with the 5' part of DE2 (DE2b). When present together, these two proteins cooperatively assemble onto the DE2 element. We now report the purification of DEF-B and show that it is identical to the product of the adenovirus IVa2 gene product. This conclusion is based on microsequence analysis of DEF-B as well as on the inhibitory effect of antibodies against IVa2 on the DNA-binding activity of DEF-B and also on DE-dependent in vitro transcription. In addition, we show that bacterially synthesized IVa2 protein binds to the DE sequences with the same specificity as DEF-B. Finally, in transfected cells, a recombinant IVa2 protein stimulates MLP activity in a DE-dependent fashion. The physiological implications of these findings are discussed.

Oct-6: a regulator of keratinocyte gene expression in stratified squamous epithelia

POU domain proteins have been implicated as regulators of differentiation and development, particularly in early embryogenesis and in neural morphogenesis. Given that neural and epidermal lineages originate from a common precursor (ectodermal) cell, we explored the possibility that POU proteins are involved in epidermal differentiation. Using reverse transcription-PCR and degenerate oligonucleotides, we generated several POU domain cDNAs from cultured human epidermal mRNAs. One of these encoded a sequence identical to the rodent Tst-1/SCIP/Oct-6 POU domain. Subsequently, we isolated a cDNA encoding a 45.3-kDa protein with 98% sequence identity to rat Tst-1/SCIP and 94% identity to mouse Oct-6. This protein bound specifically to the canonical octamer motif, warranting its designation as human Oct-6. By RNase protection assays, by PCR, and by immunoblot analysis, Oct-6 was expressed in cultured epidermal keratinocytes. By in situ hybridization, Oct-6 mRNA was detected not only in epidermis but also a variety of other stratified squamous epithelia and with greater signals than testis, the tissue in which this POU protein was originally discovered. Moreover, Oct-6 exerted a marked and specific negative influence on expression of the K5 and K14 genes, abundantly expressed in most dividing stratified squamous epithelial cells and downregulated as cells commit to terminally differentiate. The repressive effect was complex, but it was not observed with Oct-1, nor was it seen with a truncated Oct-6 missing the POU domain. Taken together, our studies suggest that Oct-6 may play an important role in controlling gene expression in stratified squamous epithelia, including epidermis.

Mapping critical residues in eukaryotic DNA-binding proteins: a plasmid-based genetic selection strategy with application to the Oct-2 POU motif

Discrimination between allowed and disallowed amino acid substitutions provides a powerful method for analysis of protein structure and function. Site-directed mutagenesis allows specific hypotheses to be tested, but its systematic application to entire structural motifs is inefficient. This limitation may be overcome by genetic selection, which allows rapid scoring of thousands of randomly (or pseudorandomly) generated mutants. To facilitate structural dissection of DNA-binding proteins, we have designed two generally applicable bacterial selection systems: pPLUS selects for the ability of a protein to bind to a user-defined DNA sequence, whereas pMINUS selects against such binding. Complementary positive and negative selections allow rapid mapping of critical residues. To test and calibrate the systems, we have investigated the bipartite POU domain of the human B-cell-specific transcription factor Oct-2. (i) An invariant residue (Asn347) in the DNA-recognition helix of the POU-specific homeodomain (POUHD) was substituted by each of the 19 other possible amino acids. The mutant proteins each exhibited decreased specific DNA binding as defined in vivo by genetic selection and in vitro by gel retardation; relative affinities correlate with phenotypes in the positive and negative selection systems. An essential role for Asn347 in wild-type POUHD-DNA recognition is consistent with homologous Asn-adenine interactions in cocrystal structures of canonical homeodomains. (ii) Extension of pPLUS/pMINUS selection to the POU-specific subdomain (POUs) is demonstrated by analysis of mutations in its putative helix-turn-helix (HTH) element.(ABSTRACT TRUNCATED AT 250 WORDS)

Specificity determinants for the interaction of lambda repressor and P22 repressor dimers

The related phage lambda and phage P22 repressors each bind cooperatively to adjacent and separated operator sites, an interaction that involves a pair of repressor dimers. The specificities of these interactions differ: Each dimer interacts with its own type but not with dimers of the heterologous repressor. The two repressors exhibit significant amino acid sequence homology in their carboxy-terminal domains, which are responsible for both dimer formation and the dimer-dimer interaction. Here, we identify a collection of amino acid substitutions that disrupt the protein-protein interaction of DNA-bound lambda repressor dimers and show that several of these substitutions have the same effect when introduced at the corresponding positions of P22 repressor. We use this information to construct a variant of the lambda repressor bearing only six non-wild-type amino acids that has a switched specificity; that is, it binds cooperatively with P22 repressor, but not with wild-type lambda repressor. These results identify a series of residues that determine the specificities of the two interactions.

The POU domains of the Oct1 and Oct2 transcription factors mediate specific interaction with TBP

We had previously shown that the ubiquitous Oct1 and the lymphoid-specific Oct2 transcription factors stimulate transcription at the level of stable preinitiation complex formation. We have therefore investigated whether the octamer binding proteins might physically interact with TBP, the TATA box binding protein component of the TFIID factor. By using several different experimental systems we show that TBP efficiently associates with Oct1 and Oct2. The interaction is direct and does not depend on the presence of DNA or additional proteins. N- and C-terminal deletions of the different proteins were used to localize the domains involved in the interaction. We show that the POU homeodomain of Oct2 and the evolutionarily conserved C-terminal core domain of TBP are both required and sufficient for the interaction. The Oct1 POU domain, which is highly homologous to the Oct2 POU domain, likewise mediates interaction with TBP. The interaction can also be observed in vivo, as TBP can be co-precipitated with Oct2 from co-transfected Cos1 cells and TBP co-immunoprecipitates with the endogenous Oct1 from HeLa cells. Co-transfection of human TBP and Oct2 expression vectors into B cells resulted in a synergistic activation of an octamer motif containing promoter.

Evolution of gene networks by gene duplications: a mathematical model and its implications on genome organization

Networks of transcriptional regulators have key roles in metazoan development. Important forces in the evolution of these networks are gene duplications and gene deletions, events that may change the spatiotemporal expression pattern of network genes. A measure for the probability of such changes after gene-duplication events is proposed. This measure is based on a simple mathematical model that describes such networks as dynamical systems and on properties of ensembles of these dynamical systems. It is predicted that this probability depends only on the fraction of genes duplicated in a single event and that it is largest if approximately 40% of the genes in a network are duplicated. This property is robust with respect to variations in model parameters. On these grounds, it is argued that (i) evolution of gene networks should preferentially occur either by duplication of single genes or by duplication of all genes involved in a network, and that (ii) tight linkage ("clustering") or strong dispersal are the two evolutionarily most favorable forms of genomic organization of genes forming such networks.

Oct-6: a regulator of keratinocyte gene expression in stratified squamous epithelia

POU domain proteins have been implicated as regulators of differentiation and development, particularly in early embryogenesis and in neural morphogenesis. Given that neural and epidermal lineages originate from a common precursor (ectodermal) cell, we explored the possibility that POU proteins are involved in epidermal differentiation. Using reverse transcription-PCR and degenerate oligonucleotides, we generated several POU domain cDNAs from cultured human epidermal mRNAs. One of these encoded a sequence identical to the rodent Tst-1/SCIP/Oct-6 POU domain. Subsequently, we isolated a cDNA encoding a 45.3-kDa protein with 98% sequence identity to rat Tst-1/SCIP and 94% identity to mouse Oct-6. This protein bound specifically to the canonical octamer motif, warranting its designation as human Oct-6. By RNase protection assays, by PCR, and by immunoblot analysis, Oct-6 was expressed in cultured epidermal keratinocytes. By in situ hybridization, Oct-6 mRNA was detected not only in epidermis but also a variety of other stratified squamous epithelia and with greater signals than testis, the tissue in which this POU protein was originally discovered. Moreover, Oct-6 exerted a marked and specific negative influence on expression of the K5 and K14 genes, abundantly expressed in most dividing stratified squamous epithelial cells and downregulated as cells commit to terminally differentiate. The repressive effect was complex, but it was not observed with Oct-1, nor was it seen with a truncated Oct-6 missing the POU domain. Taken together, our studies suggest that Oct-6 may play an important role in controlling gene expression in stratified squamous epithelia, including epidermis.

Bidirectional transcription from the human immunoglobulin VH6 gene promoter

The human immunoglobulin (Ig) heavy chain VH6 gene promoter contains an imperfect octamer (AgGCAAAT) and is not dependent on the Ig heavy chain enhancer for activity; reporter constructs containing this promoter are very active in non-B cells. In experiments designed to characterize regions upstream of the transcriptional start site that are important for promoter function, we produced a series of deletion constructs, including one containing sequences between -74 and -146. Surprisingly, this fragment had promoter activity in both orientations. Inspection of the VH6 promoter sequence indicated that there was a possible TATA box in the proper orientation upstream of the imperfect octamer. The -74 to -146 fragment functioned as a promoter in the reverse orientation in three B cell lines and in non-B (HeLa) cells, with a much higher level of activity seen in the HeLa cells. To determine if the promoter could work in both directions simultaneously, reporter genes were positioned up- and downstream of a VH6 promoter fragment. Reporter gene activity was found for both genes in B cells and HeLa cells. Using a reverse transcriptase-polymerase chain reaction procedure (RT-PCR), we found a transcript corresponding to sequences upstream of the VH6 promoter in RNA from both the lymphoblastoid cell line ML-1, which actively transcribes the VH6 promoter, and the REH cell line, which does not. No transcripts were found in the KB epithelial cell line. Two or three mRNA 5' ends were found that mapped between -137 to -143 from the authentic VH6 transcription site, 31-37 nucleotides upstream of the putative TATA box. Inspection of the sequence upstream of the VH6 promoter demonstrated the presence of an open reading frame capable of coding for 96 amino acids. The VH6 promoter represents the second Ig promoter with bidirectional activity.

Analysis of the imperfect octamer-containing human immunoglobulin VH6 gene promoter

The octamer sequence ATGCAAAT is highly conserved in the promoter of immunoglobulin heavy and light chain genes and is one of the sequence motifs involved in the control of transcription of these genes. The promoter region of an human immunoglobulin heavy chain variable gene, the sole member of the VH6 gene family, was found to differ from other VH gene promoters: it contains neither the conserved octamer motif nor a heptamer sequence, and generally bears little resemblance to other VH gene transcriptional control regions. An imperfect octamer sequence with a single nucleotide substitution (AgGCAAAT) is located 108 bp upstream of the ATG translation start site, and 81 bp upstream of the transcription initiation site. We sought to determine which sequence elements within the VH6 promoter were responsible for transcription initiation by creating progressive deletions of a 1 kb fragment from this region and testing their ability to function as promoter elements in B and non-B cells (HeLa). The minimum fragment required for full promoter function was 110 bp, but a fragment with only 65 bp retained 30-50% activity in B cells. Similar levels of transcription were seen when the -146 bp promoter containing two point mutations in the imperfect octamer was tested. Mutation of a possible pyrimidine box sequence located downstream of the TATA box was shown to have only a minor effect (10-30%) on transcription when three nucleotides were changed. Surprisingly, CAT activity was not B cell-specific, as all constructs had virtually the same activity in several B cell lines and in HeLa cells. Removal of the TATA box led to a 50% reduction in CAT activity, and the region upstream of the TATA box functioned as a promoter in both orientations. The transcriptional activity of the VH6 promoter was virtually enhancer independent: only a minor increase was observed when the immunoglobulin or SV40 enhancer was added to the promoter construct. Electrophoretic mobility shift assays of transcription factor binding to the region around the imperfect octamer indicated that binding was weak when nuclear extracts from either B cells or HeLa cells were used. The amount of complex shifted was increased by mutating the imperfect octamer to a perfect one. Chimeras produced between the VH6 promoter and a B cell-specific promoter from a member of the human VH2 gene family demonstrated that the lack of tissue specificity was due to the absence of a repressor of non-B cell transcription in the VH6 promoter. These results indicate that the VH6 promoter is relatively simple, requiring little more than the TATA element and the imperfect octamer, and transcription from this promoter lacks B cell specificity and is not dependent on the enhancer element.

Mouse Brn-3 family of POU transcription factors: a new aminoterminal domain is crucial for the oncogenic activity of Brn-3a

The class IV POU domain genes Brn-3a, -b and -c are differentially expressed during neural development and at least Brn-3a also in neuroectodermal tumors. In contrast to Brn-3b and Brn-3c, Brn-3a encodes two protein variants: Brn-3a(l) and Brn-3a(s). Brn-3a(s) lacks 84 aminoterminal residues but is otherwise identical to Brn-3a(l). Outside the well conserved carboxyterminal POU domains all three Brn-3 proteins (-a, -b and -c) diverge until the aminoterminal end where a new domain of about 100 amino acids is identified. This domain is conserved only between Brn-3 proteins and other class IV POU factors. Brn-3a(l) that contains the complete domain but not Brn-3a(s) that lacks 84 amino acids of it is able to tumorigenically transform primary fibroblasts. Brn-3b that lacks 40 amino acids of the new domain does itself not transform, but abolishes the oncogenic potential of Brn-3a(l) when transfected together. This demonstrates not only that Brn3-a(l) is a proto-oncogene and may well be causally involved in the generation of neuroectodermal tumors but also suggests that the intactness of the new aminoterminal domain described here is crucial for oncogenic activity. In addition, our data indicate that Brn-3b acts as an inhibitor of Brn-3a(l) activity.