A novel repression module, an extensive activation domain, and a bipartite nuclear localization signal defined in the immediate-early transcription factor Egr-1

Role of EGR1 in regulation of stimulus-dependent CD44 transcription in B lymphocytes

The immediate-early gene egr-1 encodes a transcription factor (EGR1) that links B-cell antigen receptor (BCR) signals to downstream activation events through the regulation of previously unidentified target genes. Here we identify the gene encoding the lymphocyte homing and migration protein CD44 as a target of EGR1 regulation in B cells. BCR-induced increases in CD44 mRNA expression and transcription levels are shown to occur in EGR1-expressing but not in nonexpressing subclones of the B-cell line WEHI-231. Kinetics of egr-1 transcription and the appearance of nuclear EGR1 protein precede CD44 induction and occur within 30 min after stimulation in the EGR1-expressing subclone. A single EGR1 binding motif is demonstrated at bp -301 of the human CD44 promoter. Cotransfection of a CD44 promoter-chloramphenicol acetyltransferase reporter construct with an egr-1 expression vector resulted in a 6.5- to 8.5-fold induction of transcriptional activity relative to an empty expression vector. The EGR1 binding motif was shown to be necessary for stimulus-induced expression of a CD44 promoter-chloramphenicol acetyltransferase reporter construct in nontransformed B lymphocytes and was required for transactivation by an EGR1 expression vector in a B-cell line. These studies identify EGR1 as an intermediary linking BCR-derived signals to the induction of CD44. The relevance of these molecular events to BCR signal transduction and antigen-stimulated B-cell-mediated immune responses is discussed.

Transcriptional regulation of the Icam-1 gene in antigen receptor- and phorbol ester-stimulated B lymphocytes: role for transcription factor EGR1

Intercellular adhesion molecule (ICAM) 1/CD54 plays an important role in T cell dependent B cell activation and for function of B lymphocytes as antigen-presenting cells. ICAM-1 expression is upregulated as a consequence of B lymphocyte antigen receptor (BCR) signaling, thereby serving to render antigen-stimulated B cells more receptive to T cell-mediated costimulatory signals. We have investigated BCR-induced expression of the Icam-1 gene in primary B cells and B cell lines and have found it to be dependent on BCR-induced expression of the transcription factor EGR1. Icam-1 transcription, induced by BCR cross-linking or bypassing the BCR with phorbol ester, is absent in a B cell line in which the EGR1-encoding gene (egr-1) is methylated and not expressed. A potential EGR1-binding site was located at -701 bp upstream of the murine Icam-1 gene transcription start site and shown by electrophoretic mobility shift assay to bind to murine EGR1. Mutation of this site in the context of 1.1 kb of the Icam-1 promoter significantly abrogated transcriptional induction by phorbol ester and anti-mu stimulation in primary B cells. A direct effect of EGR1 on the Icam-1 promoter is suggested by the ability of EGR1 expressed from an SV40-driven expression vector transactivate the wild-type Icam-1 promoter, whereas mutation of the EGR1 mutation of the EGR1 binding motif at -701 bp markedly compromises this induction. These data identify EGR1 as a signaling intermediate in BCR-stimulated B cell functional responses, specifically linking BCR signal transduction to induction of the Icam-1 gene. Furthermore, similar findings for BCR-induced CD44 gene induction (Maltzman, J.S., J.A. Carman, and J.G. Monroe. 1996. Role of EGR1 in regulation of stimulus-dependent CD44 transcription in B lymphocytes. Mol. Cell. Biol. In press) suggest that EGR1 may be an important signaling molecule for regulating levels of migration and adhesion molecules during humoral immune responses.

Functional domains in the Mig1 repressor

Mig1 is a zinc finger protein that mediates glucose repression in the yeast Saccharomyces cerevisiae. It is related to the mammalian Krox/Egr, Wilms' tumor, and Sp1 proteins and binds to a GC-rich motif that resembles the GC boxes recognized by these proteins. We have performed deletion mapping in order to identify functional domains in Mig1. We found that a small C-terminal domain comprising the last 24 amino acids mediates Mig1-dependent repression of a reporter gene. This effector domain contains several leucine-proline dipeptide repeats. We further found that inhibition of Mig1 activity in the absence of glucose is mediated by two internal elements in the Mig1 protein. A Mig1-VP16 hybrid activator was used to further investigate how Mig1 is regulated. Mig1-VP16 can activate transcription from promoters containing Mig1-binding sites and suppresses the inability of Snf1-deficient cells to grow on certain carbon sources. We found that a deletion of the SNF1 gene increases the activity of Mig1-VP16 fivefold under derepressing conditions but not in the presence of glucose. This shows that the hybrid activator is under negative control by the Snf1 protein kinase. Deletion mapping within Mig1-VP16 revealed that regulation of its activity by Snf1 is conferred by the same internal elements in the Mig1 sequence that mediate inhibition of Mig1 activity in the absence of glucose.

12-O-tetradecanoylphorbol-13-acetate activation of the MDR1 promoter is mediated by EGR1

P-glycoprotein, the product of the MDR1 gene (multidrug resistance gene 1), is an energy-dependent efflux pump associated with treatment failure in some hematopoietic malignancies. Its expression is regulated during normal hematopoietic differentiation, although its function in normal hematopoietic cells is unknown. To identify cellular factors that regulate the expression of MDR1 in hematopoietic cells, we characterized the cis- and trans-acting factors mediating 12-O-tetradecanoylphorbol-13-acetate (TPA) activation of the MDR1 promoter in K562 cells. Transient-transfection assays demonstrated that an MDR1 promoter construct containing nucleotides -69 to +20 conferred a TPA response equal to that of a construct containing nucleotides -434 to +105. TPA induced EGR1 binding to the -69/+20 promoter sequences over a time course which correlated with increased MDR1 promoter activity and increased steady-state MDR1 RNA levels. The -69/+20 promoter region contains an overlapping SP1/EGR site. The TPA-responsive element was localized to the overlapping SP1/EGR site by using a synthetic reporter construct. A mutation in this site that inhibited EGR protein binding blocked the -69/+20 MDR1 promoter response to TPA. The expression of a dominant negative EGR protein also blocked the TPA response of the -69/+20 promoter construct. Finally, the expression of EGR1 was sufficient to activate a construct containing tandem MDR1 promoter SP1/EGR sites. These data suggest a role for EGR1 in modulating MDR1 promoter activity in hematopoietic cells.

Role of EGR-1 in thapsigargin-inducible apoptosis in the melanoma cell line A375-C6

Induction of apoptosis by diverse exogenous signals is dependent on elevation of intracellular Ca2+. This process of cell death can be blocked by actinomycin D, indicating that it requires gene transcription events. To identify genes that are required for apoptosis, we used thapsigargin (TG), which inhibits endoplasmic reticulum-dependent Ca(2+)-ATPase and thereby increases cytosolic Ca2+. Exposure to TG led to induction of the zinc finger transcription factor, EGR-1, and apoptosis in human melanoma cells, A375-C6. To determine the functional relevance of EGR-1 expression in TG-inducible apoptosis, we employed a dominant negative mutant which functionally competes with EGR-1 in these cells. Interestingly, the dominant negative mutant inhibited TG-inducible apoptosis. Consistent with this observation, an antisense oligomer directed against Egr-1 also led to a diminution of the number of cells that undergo TG-inducible apoptosis. These results suggest a novel regulatory role for EGR-1 in mediating apoptosis that is induced by intracellular Ca2+ elevation. We have previously shown that in these melanoma cells, EGR-1 acts to inhibit the growth arresting action of interleukin-1. Together, these results imply that EGR-1 plays inducer-specific roles in growth control.

Identification of NAB1, a repressor of NGFI-A- and Krox20-mediated transcription

NGFI-A (also called Egr1, Zif268, or Krox24) and the closely related proteins Krox20, NGFI-C, and Egr3 are zinc-finger transcription factors encoded by immediate-early genes which are induced by a wide variety of extracellular stimuli. NGFI-A has been implicated in cell proliferation, macrophage differentiation, synaptic activation, and long-term potentiation, whereas Krox20 is critical for proper hindbrain segmentation and peripheral nerve myelination. In previous work, a structure/function analysis of NGFI-A revealed a 34-aa inhibitory domain that was hypothesized to be the target of a cellular factor that represses NGFI-A transcriptional activity. Using the yeast two-hybrid system, we have isolated a cDNA clone which encodes a protein that interacts with this inhibitory domain and inhibits the ability of NGFI-A to activate transcription. This NGFI-A-binding protein, NAB1, is a 570-aa nuclear protein that bears no obvious sequence homology to known proteins. NAB1 also represses Krox20 activity, but it does not influence Egr3 or NGFI-G, thus providing a mechanism for the differential regulation of this family of immediate-early transcription factors.

Splinkerettes--improved vectorettes for greater efficiency in PCR walking

Images

Repression by SSN6-TUP1 is directed by MIG1, a repressor/activator protein

The SSN6-TUP1 protein complex represses transcription of diversely regulated genes in the yeast Saccharomyces cerevisiae. Here we present evidence that MIG1, a zinc-finger protein in the EGR1/Zif268 family, recruits SSN6-TUP1 to glucose-repressed promoters. DNA-bound LexA-MIG1 represses transcription of a target gene in glucose-grown cells, and repression requires SSN6 and TUP1. We also show that MIG1 and SSN6 fusion proteins interact in the two-hybrid system. Unexpectedly, we found that LexA-MIG1 activates transcription strongly in an ssn6 mutant and weakly in a tup1 mutant. Finally, LexA-MIG1 does not repress transcription in glucose-deprived cells, and MIG1 is differentially phosphorylated in response to glucose availability. We suggest a role for phosphorylation in regulating repression.

The zinc finger transcription factor EGR-1 impedes interleukin-1-inducible tumor growth arrest

Interleukin-1 (IL-1) is a growth arrest signal for diverse human tumor cell lines. We report here that the action of this cytokine in melanoma cells is associated with induction of EGR-1, a zinc finger protein that activates gene transcription. Both growth arrest and EGR-1 are induced via the type I receptor of IL-1. To determine the role of EGR-1 in IL-1 action in melanoma cells, we used a chimera expressing the transrepression domain of the Wilm's tumor gene, WT1, and the DNA binding domain of Egr-1. This chimera competitively inhibited EGR-1-dependent transactivation via the GC-rich DNA binding sequence, indicating that it acted as a functional dominant negative mutant of Egr-1. Melanoma cell lines stably transfected with the dominant negative mutant construct were supersensitive to IL-1 and showed accelerated G0/G1 growth arrest compared with the parental cell line. The effect of the dominant negative mutant construct was mimicked by addition of an antisense Egr-1 oligomer to the culture medium of the parental cells: the oligomer inhibited EGR-1 expression and accelerated the growth-inhibitory response to IL-1. These data imply that EGR-1 acts to delay IL-1-mediated tumor growth arrest.

The serum response factor nuclear localization signal: general implications for cyclic AMP-dependent protein kinase activity in control of nuclear translocation

We have identified a basic sequence in the N-terminal region of the 67-kDa serum response factor (p67SRF or SRF) responsible for its nuclear localization. A peptide containing this nuclear localization signal (NLS) translocates rabbit immunoglobulin G (IgG) into the nucleus as efficiently as a peptide encoding the simian virus 40 NLS. This effect is abolished by substituting any two of the four basic residues in this NLS. Overexpression of a modified form of SRF in which these basic residues have been mutated confirms the absolute requirement for this sequence, and not the other basic amino acid sequences adjacent to it, in the nuclear localization of SRF. Since this NLS is in close proximity to potential phosphorylation sites for the cAMP-dependent protein kinase (A-kinase), we further investigated if A-kinase plays a role in the nuclear location of SRF. The nuclear transport of SRF proteins requires basal A-kinase activity, since inhibition of A-kinase by using either the specific inhibitory peptide PKIm or type II regulatory subunits (RII) completely prevents the nuclear localization of plasmid-expressed tagged SRF or an SRF-NLS-IgG conjugate. Direct phosphorylation of SRF by A-kinase can be discounted in this effect, since mutation of the putative phosphorylation sites in either the NLS peptide or the encoded full-length SRF protein had no effect on nuclear transport of the mutants. Finally, in support of an implication of A-kinase-dependent phosphorylation in a more general mechanism affecting nuclear import, we show that the nuclear transport of a simian virus 40-NLS-conjugated IgG or purified cyclin A protein is also blocked by inhibition of A-kinase, even though neither contains any potential sites for phosphorylation by A-kinase or can be phosphorylated by A-kinase in vitro.

The Ikaros gene encodes a family of functionally diverse zinc finger DNA-binding proteins

We previously described the lymphocyte-restricted Ikaros gene encoding a zinc finger DNA-binding protein as a potential regulator of lymphocyte commitment and differentiation. Here, we report the isolation of four additional Ikaros transcripts, products of alternate splicing that encode functionally diverse proteins. The Ikaros proteins contain unique combinations of zinc finger modules that dictate their overall sequence specificity and affinity. The Ik-1 and Ik-2 proteins can both bind, albeit with different affinities, to the same recognition sequences present in a number of lymphocyte-specific regulatory elements. The Ik-3 and the Ik-4 proteins interact only with a subset of these motifs. The Ik-1 and Ik-2 proteins can strongly stimulate transcription, whereas Ik-3 and Ik-4 are weak activators. Significantly, the transcription activation potential of the Ikaros proteins correlates with their subcellular localization. Upon ectopic expression of the Ikaros isoforms in nonlymphoid cells, Ik-1 and Ik-2 localize to the nucleus, whereas Ik-3 and Ik-4 are predominantly found in the cytoplasm. The Ikaros isoforms are expressed differentially in lymphocytes: Ik-1 and Ik-2 mRNAs are the predominating forms, and Ik-4 is present in significant amounts only in early T-cell progenitors, whereas Ik-3 and Ik-5 transcripts are expressed at relatively low levels throughout lymphocyte development. The ability of the Ikaros gene to generate functionally diverse proteins that may participate in distinct regulatory pathways substantiates its role as a master regulator during lymphocyte development.

The Ikaros gene encodes a family of functionally diverse zinc finger DNA-binding proteins

We previously described the lymphocyte-restricted Ikaros gene encoding a zinc finger DNA-binding protein as a potential regulator of lymphocyte commitment and differentiation. Here, we report the isolation of four additional Ikaros transcripts, products of alternate splicing that encode functionally diverse proteins. The Ikaros proteins contain unique combinations of zinc finger modules that dictate their overall sequence specificity and affinity. The Ik-1 and Ik-2 proteins can both bind, albeit with different affinities, to the same recognition sequences present in a number of lymphocyte-specific regulatory elements. The Ik-3 and the Ik-4 proteins interact only with a subset of these motifs. The Ik-1 and Ik-2 proteins can strongly stimulate transcription, whereas Ik-3 and Ik-4 are weak activators. Significantly, the transcription activation potential of the Ikaros proteins correlates with their subcellular localization. Upon ectopic expression of the Ikaros isoforms in nonlymphoid cells, Ik-1 and Ik-2 localize to the nucleus, whereas Ik-3 and Ik-4 are predominantly found in the cytoplasm. The Ikaros isoforms are expressed differentially in lymphocytes: Ik-1 and Ik-2 mRNAs are the predominating forms, and Ik-4 is present in significant amounts only in early T-cell progenitors, whereas Ik-3 and Ik-5 transcripts are expressed at relatively low levels throughout lymphocyte development. The ability of the Ikaros gene to generate functionally diverse proteins that may participate in distinct regulatory pathways substantiates its role as a master regulator during lymphocyte development.

11q23 translocations split the "AT-hook" cruciform DNA-binding region and the transcriptional repression domain from the activation domain of the mixed-lineage leukemia (MLL) gene

Translocations involving chromosome band 11q23, found in acute lymphoid and myeloid leukemias, disrupt the MLL gene. This gene encodes a putative transcription factor with homology to the zinc fingers and other domains of the Drosophila trithorax gene product and to the "AT-hook" motif of high mobility group proteins. To map potential transcriptional activation or repression domains of the MLL protein, yeast GAL4 DNA-binding domain and MLL hybrid protein-expressing plasmids were cotransfected with chloramphenicol acetyltransferase reporter plasmids in a transient transfection system. We found that MLL contains a strong activation domain and a repression domain. The former, located telomeric (3') to the breakpoint region, activated transcription 18-fold to > 200-fold, depending on the promoter and cell line used for transfection. A repression domain that repressed transcription 4-fold was located centromeric (5') to the breakpoint region of MLL. The MLL AT-hook domain protein was expressed in bacteria and was utilized in a gel mobility shift assay to assess DNA-binding activity. The MLL AT-hook domain could bind cruciform DNA, recognizing structure rather than sequence of the target DNA. In translocations involving MLL, loss of an activation domain with retention of a repression domain and a DNA-binding domain on the der(11) chromosome could alter the expression of downstream target genes, suggesting a potential mechanism of action for MLL in leukemia.

The immediate-early gene Egr-1 regulates the activity of the thymidine kinase promoter at the G0-to-G1 transition of the cell cycle

Production of thymidine kinase (TK) protein parallels the onset of DNA synthesis during the cell cycle. This process is regulated at transcriptional, posttranscriptional, and translational levels to cause a 40- to 50-fold increase in cytosolic enzymatic activity as cells progress from G1 to S phase. Transcriptional activation of the mouse TK gene through the cell cycle is dependent upon previously characterized cis elements of the proximal promoter, called MT1, MT2, and MT3, that bind at least two different complexes: TKE during the transition of cells from quiescence (G0) to G1, and Yi later at the G1/S boundary. To identify the transcription factors involved in this regulation, we screened a mouse fibroblast cDNA expression library with a labeled MT3 oligonucleotide probe and isolated a clone that encodes Egr-1, an immediate-early transcription factor, whose expression is regulated by serum or growth factors during the G0-to-G1 transition when cells reenter the cell cycle. Electrophoretic mobility shift assays demonstrate that Egr-1 is involved in the TKE complex that binds to the MT3 element and that expression of Egr-1 induces transcription of a mouse TK-chloramphenicol acetyltransferase reporter in transient transfections. These results suggest a role for Egr-1 in regulating expression of the TK gene at the G0-to-G1 transition.

The immediate-early gene Egr-1 regulates the activity of the thymidine kinase promoter at the G0-to-G1 transition of the cell cycle

Production of thymidine kinase (TK) protein parallels the onset of DNA synthesis during the cell cycle. This process is regulated at transcriptional, posttranscriptional, and translational levels to cause a 40- to 50-fold increase in cytosolic enzymatic activity as cells progress from G1 to S phase. Transcriptional activation of the mouse TK gene through the cell cycle is dependent upon previously characterized cis elements of the proximal promoter, called MT1, MT2, and MT3, that bind at least two different complexes: TKE during the transition of cells from quiescence (G0) to G1, and Yi later at the G1/S boundary. To identify the transcription factors involved in this regulation, we screened a mouse fibroblast cDNA expression library with a labeled MT3 oligonucleotide probe and isolated a clone that encodes Egr-1, an immediate-early transcription factor, whose expression is regulated by serum or growth factors during the G0-to-G1 transition when cells reenter the cell cycle. Electrophoretic mobility shift assays demonstrate that Egr-1 is involved in the TKE complex that binds to the MT3 element and that expression of Egr-1 induces transcription of a mouse TK-chloramphenicol acetyltransferase reporter in transient transfections. These results suggest a role for Egr-1 in regulating expression of the TK gene at the G0-to-G1 transition.

Mapping and mutagenesis of the amino-terminal transcriptional repression domain of the Drosophila Krüppel protein

We previously demonstrated that the Drosophila Krüppel protein is a transcriptional repressor with separable DNA-binding and transcriptional repression activities. In this study, the minimal amino (N)-terminal repression region of the Krüppel protein was defined by transferring regions of the Krüppel protein to a heterologous DNA-binding protein, the lacI protein. Fusion of a predicted alpha-helical region from amino acids 62 to 92 in the N terminus of the Krüppel protein was sufficient to transfer repression activity. This putative alpha-helix has several hydrophobic surfaces, as well as a glutamine-rich surface. Mutants containing multiple amino acid substitutions of the glutamine residues demonstrated that this putative alpha-helical region is essential for repression activity of a Krüppel protein containing the entire N-terminal and DNA-binding regions. Furthermore, one point mutant with only a single glutamine on this surface altered to lysine abolished the ability of the Krüppel protein to repress, indicating the importance of the amino acid at residue 86 for repression. The N terminus also contained an adjacent activation region localized between amino acids 86 and 117. Finally, in accordance with predictions from primary amino acid sequence similarity, a repression region from the Drosophila even-skipped protein, which was six times more potent than that of the Krüppel protein in the mammalian cells, was characterized. This segment included a hydrophobic stretch of 11 consecutive alanine residues and a proline-rich region.

Mapping and mutagenesis of the amino-terminal transcriptional repression domain of the Drosophila Krüppel protein

We previously demonstrated that the Drosophila Krüppel protein is a transcriptional repressor with separable DNA-binding and transcriptional repression activities. In this study, the minimal amino (N)-terminal repression region of the Krüppel protein was defined by transferring regions of the Krüppel protein to a heterologous DNA-binding protein, the lacI protein. Fusion of a predicted alpha-helical region from amino acids 62 to 92 in the N terminus of the Krüppel protein was sufficient to transfer repression activity. This putative alpha-helix has several hydrophobic surfaces, as well as a glutamine-rich surface. Mutants containing multiple amino acid substitutions of the glutamine residues demonstrated that this putative alpha-helical region is essential for repression activity of a Krüppel protein containing the entire N-terminal and DNA-binding regions. Furthermore, one point mutant with only a single glutamine on this surface altered to lysine abolished the ability of the Krüppel protein to repress, indicating the importance of the amino acid at residue 86 for repression. The N terminus also contained an adjacent activation region localized between amino acids 86 and 117. Finally, in accordance with predictions from primary amino acid sequence similarity, a repression region from the Drosophila even-skipped protein, which was six times more potent than that of the Krüppel protein in the mammalian cells, was characterized. This segment included a hydrophobic stretch of 11 consecutive alanine residues and a proline-rich region.

The Krüppel-associated box-A (KRAB-A) domain of zinc finger proteins mediates transcriptional repression

We have previously reported the cloning, sequencing, and partial characterization of Kid-1, a zinc finger-encoding cDNA from the rat kidney. The Kid-1 protein and approximately one-third of all other zinc finger proteins contain a highly conserved region of approximately 75 amino acids at their NH2 terminus named Krüppel-associated box (KRAB), which is subdivided into A and B domains. The evolutionary conservation, wide distribution, and genomic organization of the KRAB domains suggest an important role of this region in the transcriptional regulatory function of zinc finger proteins. The functional significance of the KRAB domain was evaluated by studying transcriptional activities of yeast GAL4-rat Kid-1 fusion proteins containing various regions of the non-zinc-finger domain of Kid-1. Transcriptional repressor activity of GAL4-Kid-1 fusion proteins maps to the KRAB-A domain. The KRAB-A domain of another zinc finger protein, ZNF2, also has repressor activity. Site-directed mutagenesis of conserved amino acids in this motif results in decreased repressor activity. Thus, we have established a functional significance for the KRAB-A domain, a consensus sequence common in zinc finger proteins.

Krüppel-associated boxes are potent transcriptional repression domains

The Krüppel-associated box (KRAB) is a highly conserved, 75-aa region containing two predicted amphipathic alpha-helices. The KRAB domain is present in the amino-terminal regions of more than one-third of all Krüppel-class Cys2His2 zinc finger proteins and is conserved from yeast to man; however, its function is unknown. Here it is shown that the KRAB domain functions as a DNA binding-dependent transcriptional repressor when fused to a heterologous DNA-binding domain from the yeast GAL4 protein. A 45-aa segment containing one of the predicted KRAB amphipathic helices was necessary and sufficient for repression. Amino acid substitutions in the predicted helix abolished the repression function. These results assign a function, transcriptional repression, to the highly conserved KRAB box and define a minimal repression domain which may aid in identifying mechanisms of repression.