Signal transduction and transcriptional regulation by glucocorticoid receptor-LexA fusion proteins

Modular structure of glucocorticoid receptor domains is not equivalent to functional independence. Stability and activity of the steroid binding domain are controlled by sequences in separate domains

A long-standing conundrum of glucocorticoid receptors has been why the steroid binding domain is active in hybrid proteins but not in isolation. For this reason, the precise boundaries of the steroid binding domain have not been defined. These questions have now been systematically examined with a variety of receptor deletion constructs. Plasmids encoding amino acids 537-673 and 537-795 of the rat receptor did not yield stable proteins, while the fusion of receptor or non-receptor sequences upstream of 537-673 afforded stable proteins that did not bind steroid. Wild type steroid binding affinity could be obtained, however, when proteins such as beta-galactosidase or dihydrofolate reductase were fused upstream of receptor amino acids 537-795. Studies of a series of dhfr/receptor constructs with deletions at the amino- and carboxyl-terminal ends of the receptor sequence localized the boundaries of the steroid binding domain to 550-795. The absence of steroid binding upon deletion of sequences in the carboxyl-terminal half of this domain was consistent with improperly folded receptor sequences. This conclusion was supported by analyses of the proteolysis and thermal stability of the mutant receptors. Thus, three independent regions appear to be required for the generation of the steroid binding form of receptors: 1) a protein sequence upstream of the steroid binding domain, which conveys stability to the steroid binding domain, 2) sequences of the carboxyl-terminal amino acids (674-795), which are required for the correct folding of the steroid binding domain, and 3) amino-terminal sequences (550-673), which may be sufficient for steroid binding after the entire steroid binding domain is properly folded. These results establish that the steroid binding domain of glucocorticoid receptors is not independently functional and illustrate the importance of both protein stability and protein folding when constructing mutant proteins.

The peroxisome proliferator activated receptors (PPARS) and their effects on lipid metabolism and adipocyte differentiation

The three types of peroxisome proliferator activated receptor (PPAR), alpha, beta (or delta), and gamma, each with a specific tissue distribution, compose a subfamily of the nuclear hormone receptor gene family. Although peroxisome proliferators, including fibrates and fatty acids, activate the transcriptional activity of these receptors, only prostaglandin J2 derivatives have been identified as natural ligands of the PPAR gamma subtype, which also binds thiazolidinedione antidiabetic agents with high affinity. Activated PPARs heterodimerize with RXR and alter the transcription of target genes after binding to specific response elements or PPREs, consisting of a direct repeat of the nuclear receptor hexameric DNA core recognition motif spaced by one nucleotide. The different PPARs can be considered key messengers responsible for the translation of nutritional, pharmacological and metabolic stimuli into changes in the expression of genes, more specifically those genes involved in lipid metabolism. PPAR alpha is involved in stimulating beta-oxidation of fatty acids. In rodents, a PPAR alpha-mediated change in the expression of genes involved in fatty acid metabolism lies at the basis of the phenomenon of peroxisome proliferation, a pleiotropic cellular response, mainly limited to liver and kidney and which can lead to hepatocarcinogenesis. In addition to their role in peroxisome proliferation in rodents, PPAR is also involved in the control of HDL cholesterol levels by fibrates and fatty acids in rodents and humans. This effect is, at least partially, based on a PPAR-mediated transcriptional regulation of the major HDL apolipoproteins, apo A-I and apo A-II. The hypotriglyceridemic action of fibrates and fatty acids also involves PPARs and can be summarized as follows: (1) an increased lipolysis and clearance of remnant particles, due to changes in LPL and apo C-III levels, (2) a stimulation of cellular fatty acid uptake and their conversion to acyl-CoA derivatives by the induction of FAT, FATP and ACS activity, (3) an induction of fatty acid beta-oxidation pathways, (4) a reduction in fatty acid and triglyceride synthesis, and finally (5) a decrease in VLDL production. Hence, both enhanced catabolism of triglyceride-rich particles as well as reduced secretion of VLDL particles are mechanisms that contribute to the hypolipidemic effect of fibrates and FFAs. Whereas for PPAR beta no function so far has been identified, PPAR gamma triggers adipocyte differentiation by inducing the expression of several genes critical for adipogenesis.

Contextual dependence of steroid receptor function on an androgen-responsive enhancer

The enhancer of the mouse sex-limited protein (Slp) gene includes a consensus hormone response element (HRE) that interacts with several auxiliary elements for steroid induction. The 160-bp fragment. C' delta 2, confers response to androgen or glucocorticoid in transfection, while a 120-bp subfragment, C' delta 9, is activated only by androgen in some cells. Site-directed mutants were tested to identify elements affecting differential response of androgen or glucocorticoid receptors (AR, GR). While most mutations of C' delta 2 affected induction by either steroid similarly, disruptions of the consensus HRE or an octamer-like sequence were more severe for GR than AR activity. An HRE half-site was critical to androgen-specific induction of C' delta 9 but had little impact in the nonspecific C' delta 2 context. In DNase I footprinting, full-length AR and GR bound similarly to the consensus HRE but dissimilarly to nonconsensus sites. Intriguingly, NF-kappa B bound the region of C' delta 2 absent from C' delta 9. Expression of I kappa B decreased response of C' delta 2, but not C' delta 9, confirming a permissive role of NF-kappa B in steroid activation. In this case, different factors may associate with receptors in the presence of NF-kappa B than those that confer androgen specificity in NF-kappa B's absence, suggesting that exclusion of some factors from a specific transcription complex is as crucial as inclusion of others. This dissection of C' delta 2 and C' delta 9 in vitro reveals subtle distinctions in AR and GR interactions that may underlie specific hormonal response in vivo.

Moloney murine leukemia virus activates NF-kappa B

Nonacutely transforming retroviruses, such as Moloney murine leukemia virus (M-MuLV), differ from transforming viruses in their mechanisms of tumor induction. While the transforming viruses cause tumors by transduction of oncogenes, the leukemia retroviruses, lacking oncogenes, employ other mechanisms, including promoter insertion and enhancer activation. Although these two mechanisms occur in many tumors induced by leukemia viruses, a substantial proportion of such tumors do not show site-specific proviral insertions. Thus, other, unidentified virus-driven mechanisms may participate in tumorigenesis. In these studies, we show that infection of cells by M-MuLV activates expression of Rel family transcription factors. In murine cells chronically infected with M-MuLV, gel shift analyses with kappaB DNA-binding motifs from the murine immunoglobulin kappa light chain enhancer demonstrated induction of at least two distinct kappaB enhancer-binding complexes. Supershifting and immunoblotting analyses defined p50, p52, RelB, and c-Rel subunits as constituents of these virus-induced protein complexes. Transient transfections performed with kappaB-dependent reporter plasmids showed transcriptional activation in M-MuLV-infected cells relative to uninfected cells. Induction of Rel/NF-kappaB transcription factor activity by M-MuLV infection may prove relevant to the mechanism of M-MuLV-induced leukemia.

Intracellular receptors use a common mechanism to interpret signaling information at response elements

The glucocorticoid receptor (GR) activates transcription in certain glucocorticoid response element (GRE) contexts, and represses or displays no activity in others. We isolated point mutations in one GRE, plfG, at which GR activated transcription under conditions in which the wild-type element was inactive or conferred repression, implying that GREs may carry signals that are interpreted by bound receptors. Consistent with this notion, we identified a mutant rat GR, K461A, which activated transcription in all GRE contexts tested, implying that this residue is important in interpretation of GRE signals. In a yeast screen of 60,000 GR mutants for strong activation from plfG, all 13 mutants isolated contained substitutions at K461. This lysine residue is highly conserved in the zinc-binding region (ZBR) of the intracellular receptor (IR) superfamily; when it was mutated in MR and RARbeta, the resulting receptors similarly activated transcription at response elements that their wild-type counterparts repressed or were inactive. We suggest that IR response elements serve in part as signaling components, and that a critical lysine residue serves as an allosteric "lock" that restricts IRs to inactive or repressing configurations except in response element contexts that signal their conversion to transcriptional activators. Therefore, mutation of this residue produces altered receptors that activate in many or all response element contexts.

Agonist-free transformation of the glucocorticoid receptor in human B-lymphoma cells

Nuclear translocation of activated glucocorticoid receptors (GRs) is a necessary step in the signal transduction by these GC hormones. Although in vitro activation of GRs can occur in the absence of a functional ligand, it is generally assumed that binding of a cognate hormone is required for activation of the intracellular GR. By indirect immunocytochemistry and Western-blot analysis, it was found that, in spontaneously aggregated human lymphoma DoHH2 cells, hormone-free GRs are located in the nucleus. Disruption of the aggregates redistributed GRs to a predominantly cytosolic location. Upon spontaneous re-aggregation the GR again became localized to the nucleus. Intracellular cross-linking of the heteromeric receptor complex was applied to investigate the protein composition of cytoplasmic and nuclear receptors. Untransformed, cytosolic GRs could be demonstrated by [3H]dexamethasone binding capacity and hsp90 co-immunoprecipitation, whereas absence of these characteristics suggested an activated conformation of the nuclear GRs. These observations suggest that cell-cell interactions are capable of transforming GRs in the absence of a ligand.

A 10-amino-acid sequence in the N-terminal A/B domain of thyroid hormone receptor alpha is essential for transcriptional activation and interaction with the general transcription factor TFIIB

The effects of the thyroid hormone (3,5,3'-triiodo-L-thyronine [T3]) on gene transcription are mediated by nuclear T3 receptors (T3Rs). alpha- and beta-isoform T3Rs (T3R alpha and -beta) are expressed from different genes and are members of a superfamily of ligand-dependent transcription factors that also includes the receptors for steroid hormones, vitamin D, and retinoids. Although T3 activates transcription by mediating a conformational change in the C-terminal approximately 220-amino-acid ligand-binding domain (LBD), the fundamental mechanisms of T3R-mediated transcriptional activation remain to be determined. We found that deletion of the 50-amino-acid N-terminal A/B domain of chicken T3R alpha (cT3R alpha) decreases T3-dependent stimulation of genes regulated by native thyroid hormone response elements about 10- to 20-fold. The requirement of the A/B region for transcriptional activation was mapped to amino acids 21 to 30, which contain a cluster of five basic amino acids. The A/B region of cT3R alpha is not required for T3 binding or for DNA binding of the receptor as a heterodimer with retinoid X receptor. In vitro binding studies indicate that the N-terminal region of cT3R alpha interacts efficiently with TFIIB and that this interaction requires amino acids 21 to 30 of the A/B region. In contrast, the LBD interacts poorly with TFIIB. The region of TFIIB primarily involved in the binding of cT3R alpha includes an amphipathic alpha helix contained within residues 178 to 201. Analysis using a fusion protein containing the DNA-binding domain of GAL4 and the entire A/B region of cT3R alpha suggests that this region does not contain an intrinsic activation domain. These and other studies indicate that cT3R alpha mediates at least some of its effects through TFIIB in vivo and that the N-terminal region of DNA-bound cT3R alpha acts to recruit and/or stabilize the binding of TFIIB to the transcription complex. T3 stimulation could then result from ligand-mediated changes in the LBD which may lead to the interaction of other factors with cT3R alpha, TFIIB, and/or other components involved in the initiation of transcription.

Role of acidic and phosphorylated residues in gene activation by the glucocorticoid receptor

To investigate the role of acidic and phosphorylated amino acids in the function of the major transactivation domain (tau 1) of the glucocorticoid receptor, we have performed a mutagenesis study. Aspartic and glutamic acid residues were neutralized in clusters of 2 to 4 amino acids throughout the tau 1 domain. The activity of the mutant proteins was determined using transactivation assays in yeast and mammalian cells. Some acidic residues in the core region of tau 1 appear to play a minor role in tau 1 activity, but, generally, individual acidic residues are not critical for activity. Mutagenesis of five serine residues that are phosphorylated in the mouse glucocorticoid receptor and which are conserved in the human receptor did not affect the transactivation activity of the tau 1 domain in yeast. As in mouse cells, these serine residues are the predominant sites of phosphorylation for ectopically expressed receptor in yeast, since the mutant protein lacking all five sites had a severely reduced phosphorylation level. Mutant proteins in which larger numbers of acidic residues are neutralized show a progressive decrease in activity indicating that acidity in general is important for tau 1 function. However, our results are not consistent with the "acid blob" theory of transactivator function that has been suggested for some other activator proteins. Other putative roles for the acidity of tau 1 are discussed.

Ligand-regulated site-specific recombination

Site-specific recombination offers a potential way to alter a living genome by design in a precise and stable manner. This potential requires strategies which can be used to regulate the recombination event. We describe a strategy to regulate FLP recombinase activity which relies on expressing FLP as a fusion protein with steroid hormone receptor ligand binding domains (LBDs). In the absence of a ligand cognate to the LBD, the recombinase activity of the fusion protein is extremely low. Upon ligand administration, recombinase activity is rapidly induced. These results outline the basis for inducible expression or disruption strategies based on inducible recombination. Additionally, we have exploited the conditional nature of FLP-LBD fusion proteins to direct integration of a plasmid into a specific genomic site at frequencies approaching the frequency of random integration.

A dopamine-responsive domain in the N-terminal sequence of Pit-1. Transcriptional inhibition in endocrine cell types

The POU transcription factor Pit-1 activates the prolactin gene in pituitary lactotrophs and may integrate responses of the gene to external signals. To study the role of Pit-1 in dopaminergic inhibition of the prolactin gene, we transiently transfected Pit-1 and dopamine D2 receptor vectors into a series of heterologous cell lines and examined dopamine regulation of the prolactin gene promoter. Regulation was Pit-1-dependent in all cell lines tested. Moreover, dopamine responsiveness was cell type-specific: stimulatory in fibroblasts (COS-7) and muscle-type cells (P19/Me2SO-induced) and inhibitory in pancreatic endocrine (RIN, InR1-G9) and neural-like (P19/retinoic acid-induced) cells. Because dopaminergic responses in Pit-1-transfected RIN cells paralleled those in pituitary GH4 cells, the islet cell line was used to test for sequences in Pit-1 that mediate negative hormone signals. Dopamine responsiveness of the Pit-1 transactivation domain (residues 8-80) was examined using a chimeric LexA construct. LxPit-1, LxSp1, and Lx-glucocorticoid receptor fusions all activated basal transcription, but only LxPit-1 was regulated by dopamine. Regulatory responses of LxPit-1 and full-length Pit-1 were quantitatively similar. In addition, gain-of-function G alpha mutants that inhibit Pit-1-dependent promoters in GH4 cells also suppressed selectively Pit-1- or LxPit-1-dependent promoters in RIN cells. This demonstrates that Pit-1 can function as a specific target for distinct inhibitory G protein signals. Interestingly, Pit-1 sequences N-terminal to the DNA-binding POU domain appear to be sufficient in mediating regulation by these pathways.

Homotypic interactions of chicken GATA-1 can mediate transcriptional activation

We used a one-hybrid system to replace precisely the finger II chicken GATA-1 DNA-binding domain with the binding domain of bacterial repressor protein LexA. The LexA DNA-binding domain lacks amino acids that function for transcriptional activation, nuclear localization, or protein dimerization. This allowed us to analyze activities of GATA-1 sequences distinct from DNA binding. We found that strong transcriptional activating sequences that function independently of finger II are present in GATA-1. Sequences including finger I contain an independent nuclear localizing function. Our data are consistent with cooperative binding of two LexA-GATA-1 hybrid proteins on a palindromic operator. The sensitivity of our transcription assay provides the first evidence that GATA-1 can make homotypic interactions in vivo. The ability of a non-DNA-binding form of GATA-1 to activate gene expression by targeting to a bound GATA-1 derivative further supports the notion that GATA-1-GATA-1 interactions may have functional consequences. A coimmunoprecipitation assay was used to demonstrate that GATA-1 multimeric complexes form in solution by protein-protein interaction. The novel ability of GATA-1 to interact homotypically may be important for the formation of higher-order structures among distant regulatory elements that share binding sites for this transcription factor. We also used the system to test the ability of GATA-1 to interact heterotypically with other activators.

The human peroxisome proliferator-activated receptor (PPAR) subtype NUC1 represses the activation of hPPAR alpha and thyroid hormone receptors

We have cloned two human peroxisome proliferator-activated receptor (PPAR) subtypes, hPPAR alpha and hNUC1. hPPAR alpha is activated by clofibric acid and other PPAR activators. hNUC1 is not activated by these compounds acting instead as a repressor of hPPAR alpha and human thyroid hormone receptor transcriptional activation. Repression is specific since hNUC1 does not significantly repress activation by the progesterone or retinoic acid receptors. We demonstrate co-operative binding of hNUC1 and hRXR alpha to a PPAR-responsive element and show that in the presence of hRXR alpha, the affinity of hNUC1 for the peroxisome proliferator is comparable to that of hPPAR alpha. Furthermore, repression of hPPAR alpha can be overcome by transfecting excess hPPAR alpha. We propose that hNUC1 represses the activity of hPPAR alpha by titrating out a factor required for activation. Our data further suggests convergence of thyroid hormone- and peroxisome-mediated fatty acid metabolism pathways. Overcoming hNUC1 repression could be a means of increasing the activity of these receptors.

A transcriptional repressor obtained by alternative translation of a trinucleotide repeat

Triplet nucleotide repeats are ubiquitous and rapidly evolving sequences in eucaryotic genomes. They are sporadically found in coding regions of transcription regulators where they become translated in different homopolymeric aminoacid (HPAA) stretches, depending on the local frame. Poly(CAG) yields three different HPAAs (poly Gln, Ser or Ala). Current sequence databases indicate a clear bias in the size and frequency of these HPAAs according to the rule: (Gln)n > (Ser)n >> (Ala)n. Aiming to understand the reasons of this bias, we changed the translational reading frame of the highly polymorphic CAG-repeat that normally encodes poly-Gln in the N-terminal portion of the rat glucocorticoid receptor (GR). The GR mutant in which the CAG repeat is translated to poly-Ala (called GR[Ala]) is incapable of transactivation, but maintains competence for hormone binding, nuclear translocation and specific DNA binding. We show that GR desactivation is obtained only when a very precise threshold length of the repeat is reached. GR[Ala] displays a strong negative dominance when tested for transcriptional activation in vivo and may become useful for selective competition of receptor dependent activities in tissue culture cells and transgenic animals. We discuss the implications of our findings for the understanding of the evolutionary behaviour of trinucleotide repeats in coding sequences.

Cortisol resistance and the guinea pig glucocorticoid receptor

The guinea pig has been employed as a model to study the structure/function relationships of the glucocorticoid receptor (GR), and to determine the regions of the receptor important for binding hormone and antihormone. Guinea pigs have high levels of circulating cortisol and GR with a approximately 20-fold lower affinity for dexamethasone than mouse GR. Cloning and sequencing of guinea pig GR has identified 24 amino acid changes in the ligand-binding domain (LBD) compared to the human GR. By substituting the guinea pig GR LBD for the human LBD in a human GR expression vector we have shown in cotransfection studies that guinea pig GR LBD confers glucocorticoid resistance as observed in vivo. In initial studies guinea pig GR LBD appeared constitutively active; in subsequent studies to determine which of the 24 amino acid changes present in the guinea pig GR LBD conferred resistance, it became apparent that the guinea pig LBD (LBD delta), amplified by PCR for subcloning into the human GR expression vector, contained a single adenine deletion in the hinge region within ten bases of the PCR primer. This single base deletion resulted in a frameshift bringing a stop codon into frame one codon after the deletion. While this now clearly accounts for the observed constitutive activity, since it is known that C-terminally truncated steroid receptors exhibit constitutive activation such a truncation is more difficult to reconcile with the repeatedly demonstrable hormone dose-response curves obtained with this guinea pig GR LBD delta.(ABSTRACT TRUNCATED AT 250 WORDS)

The tau 4 activation domain of the thyroid hormone receptor is required for release of a putative corepressor(s) necessary for transcriptional silencing

The C terminus of nuclear hormone receptors is a complex structure that contains multiple functions. We are interested in the mechanism by which thyroid hormone converts its receptor from a transcriptional silencer to an activator of transcription. Both regulatory functions are localized in the ligand binding domain of this receptor superfamily member. In this study, we have identified and characterized several functional domains within the ligand binding domain of the human thyroid hormone receptor (TR beta) conferring transactivation. Interestingly, these domains are localized adjacent to hormone binding sites. One activation domain, designated tau 4, is only 17 amino acids in length and is localized at the extreme C terminus of TR. Deletion of six amino acids of tau 4 resulted in a receptor that could still bind hormone but acted as a constitutive silencer, indicating that tau 4 is required for both transactivation and relief of the silencing functions. In addition, we performed in vivo competition experiments, the results of which suggest that in the absence of tau 4 or hormone, TR is bound by a corepressor protein(s) and that one role of hormone is to release corepressor from the receptor. We propose a general model in which the role of hormone is to induce a conformational change in the receptor that subsequently affects the action of tau 4, leading to both relief of silencing and transcriptional activation.

Influence of a steroid receptor DNA-binding domain on transcriptional regulatory functions

We have isolated two independent mutations in the DNA-binding domain of the rat glucocorticoid receptor, P493R and S459A, that implicate DNA binding in the control of attached transcriptional activation domains, either that of the receptor itself or of VP16. The mutants are capable of activating transcription normally, but unlike wild-type receptors, they interfere with particular transcriptional activators in yeast and mammalian cells, and inhibit growth when overexpressed in yeast. The mutant residues reside at positions within the three-dimensional structure of the receptor that could, in principle, transduce structural changes from the DNA-binding surface of the receptor to other functional domains. These findings, together with the salt dependence of specific and nonspecific DNA binding by these receptors, suggest that specific DNA acts as an allosteric effector that directs the functional interaction of the receptor with targets of transcriptional activation and that the P493R and S459A mutants mimic the allosteric effect of specific DNA, allowing the receptor to interact with regulatory targets even in the absence of specific DNA binding.

Functional evidence for ligand-dependent dissociation of thyroid hormone and retinoic acid receptors from an inhibitory cellular factor

The ligand-binding domains of thyroid hormone (L-triiodothyronine [T3]) receptors (T3Rs), all-trans retinoic acid (RA) receptors (RARs), and 9-cis RA receptors (RARs and RXRs) contain a series of heptad motifs thought to be important for dimeric interactions. Using a chimera containing amino acids 120 to 392 of chicken T3R alpha (cT3R alpha) positioned between the DNA-binding domain of the yeast GAL4 protein and the potent 90-amino-acid transactivating domain of the herpes simplex virus VP16 protein (GAL4-T3R-VP16), we provide functional evidence that binding of ligand releases T3Rs and RARs from an inhibitory cellular factor. GAL4-T3R-VP16 does not bind T3 and does not activate transcription from a GAL4 reporter when expressed alone but is able to activate transcription when coexpressed with unliganded T3R or RAR. This activation is reversed by T3 or RA, suggesting that these receptors compete with GAL4-T3R-VP16 for a cellular inhibitor and that ligand reverses this effect by dissociating T3R or RAR from the inhibitor. A chimera containing the entire ligand-binding domain of cT3R alpha (amino acids 120 to 408) linked to VP16 [GAL4-T3R(408)-VP16] is activated by unliganded receptor as well as by T3. In contrast, GAL4-T3R containing the amino acid 120 to 408 ligand-binding region without the VP16 domain is activated only by T3. The highly conserved ninth heptad, which is involved in heterodimerization, appears to participate in the receptor-inhibitor interaction, suggesting that the inhibitor is a related member of the receptor gene family. In striking contrast to T3R and RAR, RXR activates GAL4-T3R-VP16 only with its ligand, 9-cis RA, but unliganded RXR does not appear to be the inhibitor suggested by these studies. Further evidence that an orphan receptor may be the inhibitor comes from our finding that COUP-TF inhibits activation of GAL4-T3R-VP16 by unliganded T3R and the activation of GAL4-T3R by T3. These and other results suggest that an inhibitory factor suppresses transactivation by the T3Rs and RARs while these receptors are bound to DNA and that ligands act, in part, by inactivating or promoting dissociation of a receptor-inhibitor complex.

Functional evidence for ligand-dependent dissociation of thyroid hormone and retinoic acid receptors from an inhibitory cellular factor

The ligand-binding domains of thyroid hormone (L-triiodothyronine [T3]) receptors (T3Rs), all-trans retinoic acid (RA) receptors (RARs), and 9-cis RA receptors (RARs and RXRs) contain a series of heptad motifs thought to be important for dimeric interactions. Using a chimera containing amino acids 120 to 392 of chicken T3R alpha (cT3R alpha) positioned between the DNA-binding domain of the yeast GAL4 protein and the potent 90-amino-acid transactivating domain of the herpes simplex virus VP16 protein (GAL4-T3R-VP16), we provide functional evidence that binding of ligand releases T3Rs and RARs from an inhibitory cellular factor. GAL4-T3R-VP16 does not bind T3 and does not activate transcription from a GAL4 reporter when expressed alone but is able to activate transcription when coexpressed with unliganded T3R or RAR. This activation is reversed by T3 or RA, suggesting that these receptors compete with GAL4-T3R-VP16 for a cellular inhibitor and that ligand reverses this effect by dissociating T3R or RAR from the inhibitor. A chimera containing the entire ligand-binding domain of cT3R alpha (amino acids 120 to 408) linked to VP16 [GAL4-T3R(408)-VP16] is activated by unliganded receptor as well as by T3. In contrast, GAL4-T3R containing the amino acid 120 to 408 ligand-binding region without the VP16 domain is activated only by T3. The highly conserved ninth heptad, which is involved in heterodimerization, appears to participate in the receptor-inhibitor interaction, suggesting that the inhibitor is a related member of the receptor gene family. In striking contrast to T3R and RAR, RXR activates GAL4-T3R-VP16 only with its ligand, 9-cis RA, but unliganded RXR does not appear to be the inhibitor suggested by these studies. Further evidence that an orphan receptor may be the inhibitor comes from our finding that COUP-TF inhibits activation of GAL4-T3R-VP16 by unliganded T3R and the activation of GAL4-T3R by T3. These and other results suggest that an inhibitory factor suppresses transactivation by the T3Rs and RARs while these receptors are bound to DNA and that ligands act, in part, by inactivating or promoting dissociation of a receptor-inhibitor complex.

The length and location of CAG trinucleotide repeats in the androgen receptor N-terminal domain affect transactivation function

Some transcription factors contain stretches of polyglutamine encoded by repeats of the trinucleotide CAG. Expansion of the CAG repeat in the androgen receptor (AR) has been correlated with the incidence and severity of X-linked spinal and bulbar muscular atrophy (Kennedy's disease). In order to understand the relationship of this mutation to AR function, we constructed ARs that varied in the position and size of the polyglutamine tract, and assayed for the abilities of these mutant receptors to bind androgen and to activate transcription of several different AR-responsive reporter genes. Elimination of the tract in both human and rat AR resulted in elevated transcriptional activation activity, strongly suggesting that the presence of the polyglutamine tract is inhibitory to transactivation. Progressive expansion of the CAG repeat in human AR caused a linear decrease of transactivation function. Importantly, expansion of the tract did not completely eliminate AR activity. We postulate that this residual AR activity may be sufficient for development of male primary and secondary sex characteristics, but may fall below a threshold level of activity necessary for normal maintenance of motor neuron function. This functional abnormality may be representative of other genetic diseases that are associated with CAG expansion mutations in open reading frames, such as spinocerebellar ataxia type I and Huntington's disease.

Expression of the glucocorticoid receptor gene is regulated during early embryogenesis of Xenopus laevis

To study the possible role of the glucocorticoid receptor (GR) in early embryogenesis, we isolated a Xenopus glucocorticoid receptor cDNA from an embryonic stage 17 cDNA library. Overexpression of this Xenopus GR in COS cells confers the ability to transactivate a GRE-tk CAT promoter construct in a ligand dependent manner. Expression of the Xenopus GR gene at the RNA level was analyzed by Northern blot hybridization. Transcripts of 4 and 6 kb are present in oocytes. The 4 kb mRNA is abundant and is degraded together with the 6 kb mRNA during cleavage stages of early development. Between stages 17 and 24, GR messengers are extremely rare. From stage 32 onwards, both GR transcripts start to be expressed again at intermediate levels. These results provide the first evidence that expression of the GR gene is regulated during early embryonic development.

Human small cell lung cancer cell lines expressing the proopiomelanocortin gene have aberrant glucocorticoid receptor function

Some human small cell lung carcinomas (SCLC) secrete proopiomelanocortin (POMC) derived peptides, but in contrast to the pituitary, glucocorticoids fail to inhibit this hormone production. We have previously described an in vitro model using human SCLC cell lines that express POMC and are resistant to glucocorticoids. We have now identified the glucocorticoid receptor (GR) in the SCLC cell line COR L24 using a whole cell ligand binding assay (Kd = 5.7 nM; Bmax = 11 fmol/million cells), while another cell line, DMS 79, lacked significant glucocorticoid binding. To analyze GR function both positive (GMCO) and negative (TRE)3-tkCAT), glucocorticoid-regulated reporter gene constructs were transfected into COR L24 cells. In the SCLC cell line, neither hydrocortisone nor dexamethasone (500-2,000 nM) significantly induced chloramphenicol acetyltransferase expression from GMCO; in addition, they did not suppress chloramphenicol acetyltransferase expression from (TRE)3-tkCAT. Similar results were obtained with two other POMC-expressing SCLC cell lines. Expression of wild type GR in COR L24 cells restored glucocorticoid signaling, with marked induction of GMCO reporter gene expression by dexamethasone (9,100 +/- 910%; n = 3), and an estimated EC50 of 10 nM. This failure of the GR explains the resistance of the POMC gene to glucocorticoid inhibition and may have implications for cell growth in SCLC.

Delineation of a small region within the major transactivation domain of the human glucocorticoid receptor that mediates transactivation of gene expression

Previous deletion analysis localized the major transactivation function of the human glucocorticoid receptor to a 185-amino acid segment close to the N terminus of the receptor protein. This region was named tau 1 [Hollenberg, S. M. & Evans, R. M. (1988) Cell 55, 899-906]. To delineate the smallest active region within tau 1, we have systematically tested the transactivation capacity of deletion derivatives of the tau 1 domain, fused to the glucocorticoid receptor DNA-binding domain, in yeast cells. Internal scanning deletions suggested that residues near the C terminus of tau 1 are most important for activity. Deletions of N-terminal and C-terminal sequences identified a 41-amino acid "core" region near the C terminus of tau 1 that is crucial for tau 1 function. Small peptide fragments containing the tau 1 core region are competent for transactivation, while regions outside the tau 1 core are not active. We have previously demonstrated that the intact tau 1 domain squelches the activity of a minimal promoter in vivo and in vitro, suggesting involvement of interactions with a component/components of the basal transcription machinery in the mechanism of transactivation. This activity was maintained in the tau 1 core-containing segments.

Active, interactive, and inactive steroid receptor mutants

In transient co-transfection assays, there is extensive cross-interaction between glucocorticoid receptor (GR) domains. For example, mutation of the conserved Ile residue at position 484 (rat GR map) to cysteine allows a net separation of transactivation and DNA binding. We also observed that the ligand binding domain plays a key role in cooperative transactivation. Furthermore, some carboxy-located mutations markedly alter the response of GR to agonists and antagonists. Finally, different reading frames of the CAG repeat that normally produces an amino-located poly-Gln repeat profoundly affect GR transactivation without altering DNA or ligand binding. This trans-dominant negative phenotype, seen when the CAG repeat yields a poly-Ala stretch, may turn out to be an excellent tool for functional analysis of GR in transgenic organisms.

A novel cis element mediating ligand-independent activation by c-ErbA: implications for hormonal regulation

A novel type of hormone-responsive element (HRE) is described. Unlike classical HREs, this element, RSV-T3RE (found in Rous sarcoma virus-long terminal repeat), mediates strong activation by the c-ErbA alpha thyroid hormone (T3) receptor in the absence of T3, and addition of T3 reverses this response. Whereas both c-ErbA alpha and v-ErbA are potent ligand-independent activators through the RSV-T3RE, c-ErbA beta is not. The RSV-T3RE is recognized and activated by either c-ErbA alpha homodimers or c-ErbA alpha/retinoid X receptor (RXR) heterodimers. Ligand-independent activation by c-ErbA alpha depends on a unique N-terminal activation domain, while the C-terminal activation domain is not absolutely required. Ligand-dependent activation, on the other hand, requires the C-terminal but not the N-terminal activation domain. Upon binding to the RSV-T3RE, c-ErbA alpha assumes a different conformation than when bound to a classical T3RE. c-ErbA alpha is therefore capable of selective deployment of activation domains, dictated both by the HRE with which it interacts and by T3 binding.

Conditional transformation of cells and rapid activation of the mitogen-activated protein kinase cascade by an estradiol-dependent human raf-1 protein kinase

We report a strategy for regulating the activity of a cytoplasmic signaling molecule, the protein kinase encoded by raf-1. Retroviruses encoding a gene fusion between an oncogenic form of human p74raf-1 and the hormone-binding domain of the human estrogen receptor (hrafER) were constructed. The fusion protein was nontransforming in the absence of estradiol but could be reversibly activated by the addition or removal of estradiol from the growth media. Activation of hrafER was accompanied in C7 3T3 cells by the rapid, protein synthesis-independent activation of both mitogen-activated protein (MAP) kinase kinase and p42/p44 MAP kinase and by phosphorylation of the resident p74raf-1 protein as demonstrated by decreased electrophoretic mobility. The phosphorylation of p74raf-1 had no effect on the kinase activity of the protein, indicating that mobility shift is an unreliable indicator of p74raf-1 enzymatic activity. Removal of estradiol from the growth media led to a rapid inactivation of the MAP kinase cascade. These results demonstrate that Raf-1 can activate the MAP kinase cascade in vivo, independent of other "upstream" signaling components. Parallel experiments performed with rat1a cells conditionally transformed by hrafER demonstrated activation of MAP kinase kinase in response to estradiol but no subsequent activation of p42/p44 MAP kinases or phosphorylation of p74raf-1. This result suggests that in rat1a cells, p42/p44 MAP kinase activation is not required for Raf-1-mediated oncogenic transformation. Estradiol-dependent activation of p42/p44 MAP kinases and phosphorylation of p74raf-1 was, however, observed in rat1a cells expressing hrafER when the cells were pretreated with okadaic acid. This result suggests that the level of protein phosphatase activity may play a crucial role in the regulation of the MAP kinase cascade. Our results provide the first example of a cytosolic signal transducer being harnessed by fusion to the hormone-binding domain of the estrogen receptor. This conditional system not only will aid the elucidation of the function of Raf-1 but also may be more broadly useful for the construction of conditional forms of other kinases and signaling molecules.

Conditional transformation of cells and rapid activation of the mitogen-activated protein kinase cascade by an estradiol-dependent human raf-1 protein kinase

We report a strategy for regulating the activity of a cytoplasmic signaling molecule, the protein kinase encoded by raf-1. Retroviruses encoding a gene fusion between an oncogenic form of human p74raf-1 and the hormone-binding domain of the human estrogen receptor (hrafER) were constructed. The fusion protein was nontransforming in the absence of estradiol but could be reversibly activated by the addition or removal of estradiol from the growth media. Activation of hrafER was accompanied in C7 3T3 cells by the rapid, protein synthesis-independent activation of both mitogen-activated protein (MAP) kinase kinase and p42/p44 MAP kinase and by phosphorylation of the resident p74raf-1 protein as demonstrated by decreased electrophoretic mobility. The phosphorylation of p74raf-1 had no effect on the kinase activity of the protein, indicating that mobility shift is an unreliable indicator of p74raf-1 enzymatic activity. Removal of estradiol from the growth media led to a rapid inactivation of the MAP kinase cascade. These results demonstrate that Raf-1 can activate the MAP kinase cascade in vivo, independent of other "upstream" signaling components. Parallel experiments performed with rat1a cells conditionally transformed by hrafER demonstrated activation of MAP kinase kinase in response to estradiol but no subsequent activation of p42/p44 MAP kinases or phosphorylation of p74raf-1. This result suggests that in rat1a cells, p42/p44 MAP kinase activation is not required for Raf-1-mediated oncogenic transformation. Estradiol-dependent activation of p42/p44 MAP kinases and phosphorylation of p74raf-1 was, however, observed in rat1a cells expressing hrafER when the cells were pretreated with okadaic acid. This result suggests that the level of protein phosphatase activity may play a crucial role in the regulation of the MAP kinase cascade. Our results provide the first example of a cytosolic signal transducer being harnessed by fusion to the hormone-binding domain of the estrogen receptor. This conditional system not only will aid the elucidation of the function of Raf-1 but also may be more broadly useful for the construction of conditional forms of other kinases and signaling molecules.

In vitro analysis of Ah receptor domains involved in ligand-activated DNA recognition

The Ah receptor (AHR) is a basic helix-loop-helix protein that mediates the effects of 2,3,7,8-tetrachloro-dibenzo-p-dioxin. In this report, we describe a rabbit reticulocyte system that allows functional expression of both the AHR and its dimeric partner, the AHR nuclear translocator protein (ARNT). By using this in vitro system, we were able to reconstitute agonist binding to the AHR and agonist-induced AHR-ARNT recognition of a cognate DNA enhancer sequence. Expression of AHR deletion mutants revealed the location of N-terminal domains responsible for ligand and DNA recognition and C-terminal domains that play roles in agonist-induced DNA recognition.

Functional domains of the simian foamy virus type 1 transcriptional transactivator (Taf)

The genome of simian foamy virus type 1 encodes a transcriptional transactivator (Taf) that dramatically elevates gene expression directed by the viral long terminal repeat. In this report, we describe the functional domains of simian foamy virus type 1 Taf. Several taf mutants and fusion proteins of Taf and the DNA-binding domain of the Saccharomyces cerevisiae transcriptional transactivator GAL4 were used in this study. Taf contains two potent activation domains. One of the activation domains is located at the amino terminus (positions 1 to 48, with position 1 representing the initiator amino acid methionine) and contains several acidic amino acids. The second activation domain was mapped to a region at the carboxy terminus (positions 277 to 300). These two domains activate gene expression directed by the viral long terminal repeat independently of each other. No significant amino acid sequence homology between the activation domains is noted. Thus, Taf belongs in part to the family of acidic transcriptional transactivators. The activation domain at the carboxy terminus is conserved among foamy virus transactivators but is not related to other known transcriptional activators. Therefore, the mechanism of gene activation by the carboxy terminus of Taf may be novel. In addition, a potential binding domain rich in basic amino acids (positions 179 to 222) and a highly conserved sequence among foamy virus transactivators (positions 93 to 109) were found to be critical for Taf activity.

Prokaryotic enhancer-binding proteins reflect eukaryote-like modularity: the puzzle of nitrogen regulatory protein C

Images

Functional domains of the AraC protein

The AraC protein, which regulates the L-arabinose operons in Escherichia coli, was dissected into two domains that function in chimeric proteins. One provides a dimerization capability and binds the ligand arabinose, and the other provides a site-specific DNA-binding capability and activates transcription. In vivo and in vitro experiments showed that a fusion protein consisting of the N-terminal half of the AraC protein and the DNA-binding domain of the LexA repressor dimerizes, binds well to a LexA operator, and represses expression of a LexA operator-beta-galactosidase fusion gene in an arabinose-responsive manner. In vivo and in vitro experiments also showed that a fusion protein consisting of the C-terminal half of the AraC protein and the leucine zipper dimerization domain from the C/EBP transcriptional activator binds to araI and activates transcription from a PBAD promoter-beta-galactosidase fusion gene. Dimerization was necessary for occupancy and activation of the wild-type AraC binding site.

The glucocorticoid receptor hormone binding domain mediates transcriptional activation in vitro in the absence of ligand

We show that recombinant rat glucocorticoid receptor (vvGR) expressed using vaccinia virus is indistinguishable from authentic GR with respect to DNA and hormone binding. In the absence of hormone, vvGR is mainly found in the cytoplasm in a complex with heat shock protein 90. Upon incubation with ligand, vvGR is released from this complex and translocated to the nucleus. Thus, the ligand binding domain displays the known biochemical properties. However, in vitro, transcription from a synthetic promoter and from the mouse mammary tumor virus (MMTV) promoter is enhanced by recombinant GR in a ligand independent manner. Both transactivation domains contribute to the transcriptional activity, additively on a synthetic promoter and cooperatively on the MMTV promoter. We thus provide the first evidence that in vitro the hormone binding domain has a transcriptional activity even in the absence of ligand.

Structure and function of the estrogen receptor

The hormone binding domain of the estrogen receptor is required not only for binding estradiol but also to form stable homodimers of the protein and mediate transcriptional activation by the receptor. Residues that are essential for estrogen binding are also involved in dimerization, suggesting that the hormone-binding pocket is at or near the dimer interface. Distinct hydrophobic and charged residues are essential for hormone-dependent transcriptional activation, and these appear to be conserved by other members of the nuclear receptor family. We have found that the pure antiestrogens ICI 164384 and ICI 182780 increase the turnover of the receptor compared with that in the presence of estradiol. Because it is likely that the pure antiestrogens bind to a similar region of the receptor as that of estradiol, we propose that they inhibit receptor dimerization by means of their 7 alpha alkyl-amide extension. It appears that as a consequence nuclear uptake is inhibited and the receptor more rapidly degraded in the cytoplasm.

The positive-acting sulfur regulatory protein CYS3 of Neurospora crassa: nuclear localization, autogenous control, and regions required for transcriptional activation

The positive-acting global sulfur regulatory protein, CYS3, of Neurospora crassa turns on the expression of a family of unlinked structural genes that encode enzymes of sulfur catabolism. CYS3 contains a leucine zipper and an adjacent basic region (b-zip), which together constitute a bipartite sequence-specific DNA-binding domain. Specific anti-CYS3 antibodies detected a protein of the expected size in nuclear extracts of wild-type Neurospora under conditions in which the sulfur circuit is activated. The CYS3 protein was not observed in cys-3 mutants. Nuclear extracts of wild type, but not cys-3 mutants, also showed specific DNA-binding activity identical to that obtained with a CYS3 protein expressed in Escherichia coli. A truncated CYS3 protein that contains primarily the b-zip domain binds to DNA with high specificity and affinity in vitro, yet fails to activate gene expression in vivo, and instead inhibits the function of the wild-type CYS3 protein. Amino-terminal, carboxyterminal, and internal deletions as well as alanine scanning mutagenesis were employed to identify regions of the CYS3 protein that are required for its trans-activation function. Regions of CYS3 carboxy terminal to the b-zip motif are not completely essential for function although loss of an alanine-rich region results in decreased activity. All deletions amino terminal to the b-zip motif led to a complete loss of CYS3 function. Alanine scanning mutagenesis demonstrated that an unusual prolinerich domain of CYS3 appears to be very important for function and is presumed to constitute an activation domain. It is concluded that CYS3 displays nuclear localization and positive autogenous control in Neurospora and functions as a trans-acting DNA-binding protein.

Mineralocorticoid and glucocorticoid receptor activities distinguished by nonreceptor factors at a composite response element

Mineralocorticoid and glucocorticoid hormones elicit distinct physiologic responses, yet the mineralocorticoid receptor (MR) and glucocorticoid receptor (GR) bind to and activate transcription similarly from a consensus simple hormone response element (HRE). The activities of GR and MR at plfG, a 25-base pair composite response element to which both the steroid receptors and transcription factor AP1 can bind, are analyzed here. Under conditions in which GR represses AP1-stimulated transcription from plfG, MR was inactive. With the use of MR-GR chimeras, a segment of the NH2-terminal region of GR (amino acids 105 to 440) was shown to be required for this repression. Thus, the distinct physiologic effects mediated by MR and GR may be determined by differential interactions of nonreceptor factors with specific receptor domains at composite response elements.

On the mechanism of DNA binding by nuclear hormone receptors: a structural and functional perspective

The nuclear hormone receptor DNA-binding domain consists of two zinc finger-like modules whose amino acids are highly conserved among the members of the receptor superfamily. In this review, we describe the various genetic, biochemical, and structural experiments that have been carried out primarily for the DNA-binding domains of the glucocorticoid and estrogen receptors. We describe how the structural and functional information have permitted us to predict properties of the DNA-binding domains of other nuclear receptors. We postulate how receptors discriminate closely related response elements through sequence-specific contacts and distinguish symmetry of target sites through protein-protein interactions. This mechanism explains in part how the receptors regulate diverse sets of genes from a limited repertoire of core response elements. Lastly, we describe the stereochemical basis of nuclear receptor dysfunction in certain clinical disorders.

Genetic dissection of the signaling domain of a mammalian steroid receptor in yeast

The mechanism of signal transduction by steroid receptor proteins is complex and not yet understood. We describe here a facile genetic strategy for dissection of the rat glucocorticoid receptor "signaling domain," a region of the protein that binds and transduces the hormonal signal. We found that the characteristics of signal transduction by the receptor expressed in yeast were similar to those of endogenous receptors in mammalian cells. Interestingly, the rank order of particular ligands differed between species with respect to receptor binding and biological efficacy. This suggests that factors in addition to the receptor alone must determine or influence ligand efficacy in vivo. To obtain a collection of receptors with distinct defects in signal transduction, we screened in yeast an extensive series of random point mutations introduced in that region in vitro. Three phenotypic classes were obtained: one group failed to bind hormone, a second displayed altered ligand specificity, and a third bound hormone but lacked regulatory activity. Our results demonstrate that analysis of glucocorticoid receptor action in yeast provides a general approach for analyzing the mechanism of signaling by the nuclear receptor family and may facilitate identification of non-receptor factors that participate in this process.

Involvement of the SIN4 global transcriptional regulator in the chromatin structure of Saccharomyces cerevisiae

We have cloned and sequenced the SIN4 gene and determined that SIN4 is identical to TSF3, identified as a negative regulator of GAL1 gene transcription (S. Chen, R.W. West, Jr., S.L. Johnson, H. Gans, and J. Ma, submitted for publication). Yeast strains bearing a sin4 delta null mutation have been constructed and are temperature sensitive for growth and display defects in both negative and positive regulation of transcription. Transcription of the CTS1 gene is reduced in sin4 delta mutants, suggesting that Sin4 functions as a positive transcriptional regulator. Additionally, a Sin4-LexA fusion protein activates transcription from test promoters containing LexA binding sites. The sin4 delta mutant also shows phenotypes common to histone and spt mutants, including suppression of delta insertion mutations in the HIS4 and LYS2 promoters, expression of promoters lacking upstream activation sequence elements, and decreased superhelical density of circular DNA molecules. These results suggest that the sin4 delta mutation may alter the structure of chromatin, and these changes in chromatin structure may affect transcriptional regulation.

Involvement of the SIN4 global transcriptional regulator in the chromatin structure of Saccharomyces cerevisiae

We have cloned and sequenced the SIN4 gene and determined that SIN4 is identical to TSF3, identified as a negative regulator of GAL1 gene transcription (S. Chen, R.W. West, Jr., S.L. Johnson, H. Gans, and J. Ma, submitted for publication). Yeast strains bearing a sin4 delta null mutation have been constructed and are temperature sensitive for growth and display defects in both negative and positive regulation of transcription. Transcription of the CTS1 gene is reduced in sin4 delta mutants, suggesting that Sin4 functions as a positive transcriptional regulator. Additionally, a Sin4-LexA fusion protein activates transcription from test promoters containing LexA binding sites. The sin4 delta mutant also shows phenotypes common to histone and spt mutants, including suppression of delta insertion mutations in the HIS4 and LYS2 promoters, expression of promoters lacking upstream activation sequence elements, and decreased superhelical density of circular DNA molecules. These results suggest that the sin4 delta mutation may alter the structure of chromatin, and these changes in chromatin structure may affect transcriptional regulation.

Fused protein domains inhibit DNA binding by LexA

Many studies of transcription activation employ fusions of activation domains to DNA binding domains derived from the bacterial repressor LexA and the yeast activator GAL4. Such studies often implicitly assume that DNA binding by the chimeric proteins is equivalent to that of the protein donating the DNA binding moiety. To directly investigate this issue, we compared operator binding by a series of LexA-derivative proteins to operator binding by native LexA, by using both in vivo and in vitro assays. We show that operator binding by many proteins such as LexA-Myc, LexA-Fos, and LexA-Bicoid is severely impaired, while binding of other LexA-derivative proteins, such as those that carry bacterially encoded acidic sequences ("acid blobs"), is not. Our results also show that DNA binding by LexA derivatives that contain the LexA carboxy-terminal dimerization domain (amino acids 88 to 202) is considerably stronger than binding by fusions that lack it and that heterologous dimerization motifs cannot substitute for the LexA88-202 function. These results suggest the need to reevaluate some previous studies of activation that employed LexA derivatives and modifications to recent experimental approaches that use LexA and GAL4 derivatives to detect and study protein-protein interactions.