DNA binding analysis of glucocorticoid receptor specificity mutants.

The glucocorticoid receptor (GR) DNA binding domain consists of several conserved amino acids and folds into two zinc finger-like structures. Previous transactivation experiments indicated that three amino acids residing in this region, Gly, Ser and Val, appear to be critical for target-site discrimination. Based on the solved crystal structure, these residues are at the beginning of an amphipathic alpha-helix that interacts with the DNA's major groove; of these, only valine, however, contacts DNA. In order to examine their functional role directly, we have substituted these residues for the corresponding amino acids from the estrogen receptor (ER), overexpressed and purified the mutant proteins, and assayed their binding specificity and affinity by gel mobility shifts using glucocorticoid or estrogen response elements (GRE or ERE, respectively) as DNA probes. We find that all three residues are indeed required to fully switch GR's specificity to an ERE. The contacting valine in GR is of primary importance. The corresponding residue in ER, alanine, is less important for specificity, while glutamic acid, four amino acids towards the N-terminus, is most critical for ER discrimination. Finally, we show that the GR DNA binding domain carrying all three ER-specific mutations has a significantly higher affinity for an ERE than the ER DNA binding domain itself. We interpret these results in the context of both the data presented here and the crystal structure of the GR DNA binding domain complexed to a GRE.     

Characterization and sequence-specific binding to mouse mammary tumor virus DNA of purified activated human glucocorticoid receptor

Activated glucocorticoid receptor (GR) from the human cell line HeLa S3 was purified by differential chromatography on DNA-cellulose followed by DEAE-Sepharose chromatography to 50-60% homogeneity according to sodium dodecyl sulfate gel electrophoresis and densitometric scanning of silver-stained gels. These gels routinely demonstrated a main band of Mr 94,000 (94K band) and two minor bands of Mr 79,000 (79K band) and 39,000 (39K band), respectively. Photoaffinity labeling indicated that the hormone was bound to the 94K and 79K components. In some preparations, a 72K band was observed. Further characterization of the purified receptor by gel permeation chromatography on Sephadex G-200 revealed a receptor complex with a Stokes radius of 5.8 nm. The sedimentation coefficient of the purified receptor was 4.4 Sw. In analogy to the rat hepatic GR, limited proteolysis of the purified GR with trypsin or alpha-chymotrypsin led to degradation of the 94K and 79K components and appearance of 28K and 39K fragments, respectively. In addition, no difference in the protease digestion pattern using Staphylococcus aureus V8 protease was observed. Immunoblotting using a monoclonal antibody raised against the 94K GR from rat liver demonstrated cross-reactivity with the human 94K and 79K proteins from HeLa S3 cells, indicating similar antigenic characteristics between rat and human GR. In our study, five out of nine tested monoclonal antibodies against the rat liver GR cross-reacted with human GR. DNase I and exonuclease III protection experiments demonstrated binding of the purified human GR to specific GR binding regions in mouse mammary tumor virus DNA.(ABSTRACT TRUNCATED AT 250 WORDS)     

Mechanisms of glucocorticoid function in human leukemic cells: analysis of receptor gene mutants of the activation-labile type using the covalent affinity ligand dexamethasone mesylate

In the cultured acute lymphoblastic leukemic (ALL) cell line, clones of sensitive cells are killed by receptor-occupying concentrations of glucocorticoids. In addition, several types of resistance have been identified. The types of resistance are r- (glucocorticoid binding site loss), ract/l (activation labile receptors) and r+ly- (defective lysis mechanism). The two types of receptor mutants have been examined for the presence and expression of the glucocorticoid receptor (GR) gene. Southern blot analysis, using a full-length cDNA probe for human GR, shows that the gene in both is grossly intact. Examination of the expression of the gene by Northern blots reveals the presence of normal, 7-kb message in both types of receptor mutants, though in amounts somewhat reduced from wild-type. This report focuses on the activation labile mutants. Since characterization of these mutants suggests that they can bind ligand but not retain it during activation, we hypothesized that they would respond normally to a ligand that could not be lost during activation. This seems to be the case. When the covalent affinity ligand dexamethasone mesylate, itself a partial glucocorticoid agonist/antagonist, is used, the ract/l cells are killed to an extent corresponding to that evoked by a sub-optimal concentration of the full agonist dexamethasone. We conclude: (1) that the ract/l receptors can function to kill cells if provided a ligand that they do not lose during activation; (2) that the partial agonist activity of dexamethasone mesylate for cell killing is not due to release of a small amount of free dexamethasone; (3) that the poor agonist activity of dexamethasone mesylate receptor complexes suggests that the role of steroid is strictly to participate in conversion of the receptor to its DNA binding form, after which presence of the steroid actually interferes with proper receptor action.     

Structure and specific DNA binding of the rat liver glucocorticoid receptor

During recent years major advances have been made in our understanding of glucocorticoid mechanism of action. This progress has been made possible by access to purified glucocorticoid receptor in significant amounts as well as by application of hybrid DNA technology within the field of glucocorticoid control of gene expression. Especially the mammary tumour virus genome has turned out to be a convenient experimental system suitable for such investigations. This paper summarizes some of the work carried out in our own laboratory, partially in collaboration with Dr Keith Yamamoto and his associates at the Department of Biochemistry and Biophysics, University of California, San Francisco, U.S.A.     

DNA binding specificity of mutant glucocorticoid receptor DNA-binding domains

Mutation of a small number of amino acids in the DNA-binding domain of the estrogen receptor to the corresponding sequence of the glucocorticoid receptor switches the specificity of the receptor in transactivation assays (Mader, S., Kumar, V., de Verneuil, H., and Chambon, P. (1989) Nature 338, 271-274). We have made the corresponding reciprocal mutations in the context of the glucocorticoid receptor DNA-binding domain and studied the binding of wild type and mutant purified proteins to palindromic glucocorticoid and estrogen response elements as well as to elements of intermediate sequence, using gel mobility shift assays. We show here that a protein with two altered amino acids binds glucocorticoid and estrogen response elements with a low but equal affinity, whereas a protein with an additional changed residue has a high affinity for estrogen response elements but still retains a considerable affinity for glucocorticoid response elements. Using binding sites of intermediate sequence we have further characterized the interaction with DNA. The in vitro DNA binding results are confirmed by in vivo transactivation assays in yeast. Finally we suggest a testable model for amino acid/base pair interactions involved in recognition by the glucocorticoid receptor DNA-binding domain of its target sequence.     

Molybdate-sensitive and molybdate-resistant activation-labile glucocorticoid-receptor mutants of the human lymphoid cell line CEM-C7

The basis for the glucocorticoid resistance of three (3R7, 3R43 and 4R4) genetically independent mutants derived from the glucocorticoid-sensitive human lymphoid cell line CEM-C7 was examined. Each mutant contained significant, albeit reduced, amounts of steroid binding activity measured in both whole and broken cell assays. These activities were of similar high affinity and specificity to that seen in the sensitive parent, suggesting that the steroid binding portion of the receptors in these mutants was normal. However, nuclear translocation of steroid-receptor (SR) complexes was defective in all three clones. Analysis of SR complex activation by DEAE-cellulose chromatography established that receptors from the mutant clones were extremely labile under conditions which would activate normal SR complexes. Such lability was not exhibited by the unoccupied or occupied but unactivated forms of these receptors, indicating that all three were "activation labile" (act1). SR complexes of clones 3R7 and 4R4 were completely protected by the inclusion of molybdate during attempted activation and were classified as act1:molybdate-sensitive. However, SR complexes of clone 3R43 were unstable during attempted activation, even in the presence of 50 mM molybdate, and were thus classified as act1:molybdate-resistant. Our results suggest that while the act1 phenotype may predominate among spontaneously derived glucocorticoid-resistant mutants derived from CEM-C7, this phenotype may be the consequence of at least two different mutations.     

Unspecific DNA binding of the DNA binding domain of the glucocorticoid receptor studied with flow linear dichroism

The unspecific interaction between the DNA-binding domain of the human glucocorticoid receptor and DNA was studied using linear dichroism (LD) and circular dichroism (CD) spectroscopy. The amplitude of the LD signal was found to increase upon addition of protein at ionic strengths less than 60 nM Na+, indicating an increased persistence length of the complex compared to uncomplexed DNA. Analysis of the LD spectrum suggests that the binding does not involve intercalation of tyrosine residues. Evidence of saturation is found at a binding stoichiometry of approximately 5 DNA base pairs per protein monomer.     

Thermodynamics of steroid binding to the human glucocorticoid receptor.

The thermodynamics of the interaction of glucocorticoids with their receptor were studied in cytosol from human lymphoblastoid cells. The rate and affinity constants of dexamethasone and cortisol between 0 degree and 25 degrees C were calculated by curve-fitting from time-course and equilibrium kinetics. The data were consistent with a simple reversible bimolecular interaction. Arrhenius and Van't Hoff plots were curvilinear for both steroids. At equilibrium, the solution for the equation delta G = delta H - T X delta S (eqn. 1) was (in kJ X mol-1) -47 = 36 - 83 (dexamethasone) and -42 = -9 - 33 (cortisol) at 0 degree C. Enthalpy and entropy changes decreased quasi-linearly with temperature such that, at 25 degrees C, the respective values were -50 = -75 + 25 and -43 = -48 + 5. Thus, for both steroids, the interaction was entropy-driven at low temperature and became entirely enthalpy-driven at 20 degrees C. Thermodynamic values for the transition state were calculated from the rate constants. For the forward reaction, eqn. (1) gave 45 = 84 - 39 (dexamethasone) and 46 = 60 - 14 (cortisol) at 0 degree C, and 44 = 24 + 20 (dexamethasone) and 46 = 28 + 18 (cortisol) at 25 degrees C. These data fit quite well with a two-step model [Ross & Subramanian (1981) Biochemistry 20, 3096-3102] proposed for ligand-protein interactions, which involves a partial immobilization of the reacting species governed by hydrophobic forces, followed by stabilization of the complex by short-range interactions. On the basis of this model, an analysis of the transition-state thermodynamics led to the conclusion that no more than half of the steroid molecular area is engaged in the binding process.     

Modulation of DNA binding of glucocorticoid receptor by aurintricarboxylic acid

Effects of aurintricarboxylic acid (ATA) were examined on the DNA binding properties of rat liver glucocorticoid-receptor complex. The DNA-cellulose binding capacity of the glucocorticoid-receptor complex was completely abolished by a pretreatment of receptor preparation with 0.1-0.5 mM ATA at 4 degrees C. The half-maximal inhibition (i.d.50) in the DNA binding of [3H]triamcinolone acetonide-receptor complex [( 3H]TARc) was observed at 130- and 40 microM ATA depending upon whether the inhibitor was added prior to or following the receptor activation. The entire DNA-cellulose bound [3H]TARc could be extracted in a concentration-dependent manner by incubation with 2-100 microns ATA. The [3H]TARc remained intact under the above conditions, the receptor in both control and ATA-treated preparations sedimented in the same region in salt-containing 5-20% sucrose gradients. The action of ATA appeared to be on the receptor and not on DNA-cellulose. The DNA-binding capacity of ATA-treated receptor preparations could be recovered upon exhaustive dialysis. The treatment with ATA did not appear to change the ionic behavior of heat activated GRc; the receptor in both control and the ATA-treated preparations showed similar elution profiles. Therefore, ATA appears to alter the binding to and dissociation of glucocorticoid-receptor complex from DNA. The use of ATA should offer a good chemical probe for analysis of the DNA binding domain(s) of the glucocorticoid receptor.     

Protein-protein interactions facilitate DNA binding by the glucocorticoid receptor DNA-binding domain

We have studied the interaction of the DNA-binding domain of the glucocorticoid receptor with a glucocorticoid response element from the tyrosine aminotransferase gene. This response element consists of two binding sites (half-sites) for the glucocorticoid receptor DNA-binding domain. The sequences of these two half-sites are not identical, and we have previously shown that binding occurs preferentially to one of the half-sites (Tsai, S.-Y., Carlstedt-Duke, J., Weigel, N. L., Dahlman, K., Gustafsson, J.-A., Tsai, M.-J., and O'Malley, B. W. (1988) Cell 55, 361-369). We show here that binding to the low affinity half-site is dependent on previous occupancy of the high affinity half-site. This facilitated binding is dependent on the distance between the two half-sites and their relative orientation but is not dependent on the integrity of the DNA backbone. This is consistent with a model where DNA binding is not only dependent on interactions between the protein and its DNA target sequence but is also influenced by interactions between the protein molecules bound.     

Mechanism of high-mobility group protein B enhancement of progesterone receptor sequence-specific DNA binding

The DNA-binding domain (DBD) of progesterone receptor (PR) is bipartite containing a zinc module core that interacts with progesterone response elements (PRE), and a short flexible carboxyl terminal extension (CTE) that interacts with the minor groove flanking the PRE. The chromosomal high-mobility group B proteins (HMGB), defined as DNA architectural proteins capable of bending DNA, also function as auxiliary factors that increase the DNA-binding affinity of PR and other steroid receptors by mechanisms that are not well defined. Here we show that the CTE of PR contains a specific binding site for HMGB that is required for stimulation of PR-PRE binding, whereas the DNA architectural properties of HMGB are dispensable. Specific PRE DNA inhibited HMGB binding to the CTE, indicating that DNA and HMGB–CTE interactions are mutually exclusive. Exogenous CTE peptide increased PR-binding affinity for PRE as did deletion of the CTE. In a PR-binding site selection assay, A/T sequences flanking the PRE were enriched by HMGB, indicating that PR DNA-binding specificity is also altered by HMGB. We conclude that a transient HMGB–CTE interaction alters a repressive conformation of the flexible CTE enabling it to bind to preferred sequences flanking the PRE.     

Protein-protein contacts in the glucocorticoid receptor homodimer influence its DNA binding properties

We have investigated the influence of the N-terminal domain of the 94-kDa glucocorticoid receptor on the DNA:receptor interaction. An alpha-chymotrypsin-induced 39-kDa receptor fragment, containing the hormone and DNA binding domains, binds DNA with a reduced specificity compared to the intact 94-kDa receptor. Various footprinting assays did not reveal any qualitative differences when comparing the DNA contact points made by the two different receptor entities. Like the intact receptor, the 39-kDa receptor fragment binds as a dimer to DNA. Glutaraldehyde cross-linking demonstrated a difference in the protein:protein contacts of the two homodimers. Furthermore, the dimeric 94-kDa receptor did not recognize a half-DNA site, while the dissociated 94-kDa receptor dimer and the dimeric 39-kDa receptor fragment allowed binding to such a site. These results suggest that the loss of the N-terminal domain of the receptor affects the steric arrangement and/or rigidity of the two DNA binding domains of the receptor homodimer, resulting in a decreased DNA binding specificity of the 39-kDa receptor fragment.     

Isolation and sequence-specific DNA binding of the Antennapedia homeodomain.

The homeodomain encoded by the Antennapedia (Antp) gene of Drosophila was overproduced in a T7 expression vector in Escherichia coli. The corresponding polypeptide of 68 amino acids was purified to homogeneity. The homeodomain was analysed by ultracentrifugation and assayed for DNA binding. The secondary structure of the isolated homeodomain was determined by nuclear magnetic resonance spectroscopy. DNA-binding studies indicate that the isolated homeodomain binds to DNA in vitro. It selectively binds to the same sites as a longer Antp polypeptide and a full-length fushi tarazu (ftz) protein. Therefore, the homeodomain represents the DNA-binding domain of the homeotic proteins.     

Large scale purification of the nuclear thyroid hormone receptor from rat liver and sequence-specific binding of the receptor to DNA

Methodology is reported for extracting thyroid hormone receptors from rat liver nuclei and for purifying these such that certain receptor properties can be examined. The extraction technique resulted in 1700 pmol of receptor/2 kg of liver and bypasses centrifugation in dense sucrose. The receptor was then purified by sequential heparin-Sepharose, DEAE-Sepharose, and phospho-Ultrogel chromatography and size exclusion and hydrophobic interaction high performance liquid chromatography. These steps yielded 23-35 micrograms of receptor at 0.7-1.5% purity from two 2-kg liver preparations. The cross-linkers disuccinimidyl suberate and N-succinimidyl-6-(4-azido-2-nitrophenylamino)hexanoate were employed to covalently attach 125I-labeled 3,5,3'-triiodo-L-thyronine (T3) to the purified receptor. Autoradiography after denaturing polyacrylamide gel electrophoresis revealed major 49,000 Mr and minor 58,000 Mr specific T3-binding proteins. The purified receptors exhibited high affinity (Kd = 100 pM) single site T3-binding activity. Because of the high affinity and specificity of [125I]T3 for the receptor, it was possible to uniquely identify the receptor containing DNA-protein complexes in a gel retardation assay and thus directly demonstrate for the first time that the receptor can specifically recognize sequences in the 5'-flanking DNA of the rat growth hormone gene. [125I]T3-labeled receptor migrated at the same position as the major gel-retarded 32P-labeled DNA band. Specific DNA competed for the binding much more strongly than nonspecific DNA. Thus, the purification procedure results in relatively large quantities of receptor at a purity sufficient for detecting and studying a number of its properties including specific DNA binding activity.     

Domains of the human androgen receptor and glucocorticoid receptor involved in binding to the nuclear matrix

Steroid receptors have been reported to bind to the nuclear matrix. The nuclear matrix is operationally defined as the residual nuclear structure that remains after extraction of most of the chromatin and all soluble and loosely bound componnets. To obtain insight in the molecular mechanism of the interaction of steroid receptors with the nuclear matrix, we studied the binding of several deletion mutants of the human androgen receptor (hAR) and the human glucocorticoid receptor (hGR) to the nuclear matrix. Receptor binding was tested for two different nuclear matrix preparations: complete matrices, in which most matrix proteins are retained during the isolation procedure, and depleted matrices, which consist of only a subset of these proteins. The results show that the C-terminal domain of the hAR binds tightly to both depleted and complete matrices. In addition, at least one other domain of the hAR binds to complete matrices but not to depleted matrices. In contrast to the hAR, the hGR binds only to complete matrices. For this interaction both the DNA-binding domain and the C-terminal domain of the hGR are required, whereas the N-terminal domain is not. We conclude that specific protein domains of the hAR and the hGR are involved in binding to the nuclear matrix. In addition, our results indicate that the hAR and the hGR are attached to the nuclear matrix through different molecular interactions.     

Thermodynamics of sequence-specific binding of PNA to DNA

For further characterization of the hybridization properties of peptide nucleic acids (PNAs), the thermodynamics of hybridization of mixed sequence PNA-DNA duplexes have been studied. We have characterized the binding of PNA to DNA in terms of binding affinity (perfectly matched duplexes) and sequence specificity of binding (singly mismatched duplexes) using mainly absorption hypochromicity melting curves and isothermal titration calorimetry. For perfectly sequence-matched duplexes of varying lengths (6-20 bp), the average free energy of binding (DeltaG degrees ) was determined to be -6.5+/-0.3 kJ mol(-1) bp(-1), corresponding to a microscopic binding constant of about 14 M(-1) bp(-1). A variety of single mismatches were introduced in 9- and 12-mer PNA-DNA duplexes. Melting temperatures (T(m)) of 9- and 12-mer PNA-DNA duplexes with a single mismatch dropped typically 15-20 degrees C relative to that of the perfectly matched sequence with a corresponding free energy penalty of about 15 kJ mol(-1) bp(-1). The average cost of a single mismatch is therefore estimated to be on the order of or larger than the gain of two matched base pairs, resulting in an apparent binding constant of only 0.02 M(-1) per mismatch. The impact of a mismatch was found to be dependent on the neighboring base pairs. To a first approximation, increasing the stability of the surrounding region, i.e., the distribution of A.T and G.C base pairs, decreases the effect of the introduced mismatch.     

Cadmium-113 NMR studies of the DNA binding domain of the mammalian glucocorticoid receptor

The DNA binding domain of the mammalian glucocorticoid hormone receptor (GR) contains nine highly conserved cysteine residues, a conservation shared by the superfamily of steroid and thyroid hormone receptors. A fragment [150 amino acids (AA) in length] consisting of GR residues 407-556, containing within it the entire DNA binding domain (residues 440-525), has been overexpressed and purified from Escherichia coli previously. This fragment has been shown to contain 2.3 +/- 0.2 mol of Zn(II) per mole of protein [Freedman, L. P., Luisi, B. F., Korszun, Z. R., Basavappa, R., Sigler, P. B., & Yamamoto, K. R. (1988) Nature 334, 543]. Zn(II) [or Cd(II) substitution] has been shown to be essential for specific DNA binding. 113Cd NMR of a cloned construct containing the minimal DNA binding domain of 86 AA residues [denoted GR(440-525)] with 113Cd(II) substituted for Zn(II) identifies 2 Cd(II) binding sites by the presence of 2 113Cd NMR signals each of which integrates to 1 113Cd nucleus. The chemical shifts of these two sites, 704 and 710 ppm, suggest that each 113Cd(II) is coordinated to four isolated -S- ligands. Shared -S- ligands connecting the two 113Cd(II) ions do not appear to be present, since their T1s differ by 10-fold, 0.2 and 2.0 s, respectively. Addition of a third 113Cd(II) or Zn(II) to 113Cd2GR(440-525) results in occupancy of a third site, which introduces exchange modulation of the two original 113Cd NMR signals causing them to disappear. Addition of EDTA to the protein restores the original two signals.(ABSTRACT TRUNCATED AT 250 WORDS)