Identification of Sulfhydryl-modified Cysteine Residues in the Ligand Binding Pocket of Retinoic Acid Receptor beta *

(Received for publication, July 9, 1996, and in revised form, September 25, 1996)

Christopher L. Wolfgang Dagger §, Zhen-ping Zhang Dagger , Jerome L. Gabriel Dagger , Ronald A. Pieringer Dagger , Kenneth J. Soprano par and Dianne Robert Soprano Dagger **

From the Dagger  Department of Biochemistry,  Fels Institute for Cancer Research and Molecular Biology and par  Department of Microbiology and Immunology, Temple University School of Medicine, Philadelphia, Pennsylvania 19140

ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
Acknowledgments
REFERENCES


ABSTRACT

The diverse biological functions of retinoic acid (RA) are mediated through retinoic acid receptors (RARs) and retinoid X receptors. RARs contain a high affinity binding site for RA which is sensitive to treatment with sulfhydryl modification reagents. In an attempt to identify which Cys residues are important for this loss of binding, we created three site-specific RARbeta mutants: C228A, C258A, and C267A. The affinity for RA of all three mutant receptors was in the range of that of the wild type protein, suggesting that none of these Cys residues are critical for RA binding. Rather, these modified Cys residue(s) function to sterically hinder RA binding; however, the modified Cys residues critical for the inhibition of binding differ depending on the reagent employed. Only modification of Cys228 is necessary to inhibit RA binding when RARbeta is modified by reagents which transfer large bulky groups while both Cys228 and Cys267 must be modified when a small functional group is transferred. These data suggest that both Cys228 and Cys267 but not Cys258 lie in the ligand binding pocket of RARbeta . However, Cys228 lies closer to the opening of the RARbeta ligand binding pocket whereas Cys267 lies more deeply buried.


INTRODUCTION

Retinoic acid is the potent mediator of the biological effects of vitamin A that include growth, differentiation, and morphogenesis (for review, see Ref 1). These actions of retinoic acid are mediated by two evolutionarily distinct groups of nuclear receptors that belong to the multigene family of steroid/thyroid hormone receptors called retinoic acid receptors (RARs)1 and retinoid X receptors (RXRs) (for review, see Ref. 2). The RARs and RXRs are each made up of three receptor types, designated alpha , beta , and gamma  (2, 5, 6, 7, 8, 9, 10, 11). In dimeric form, these proteins function as ligand-dependent transcriptional regulatory factors by binding to DNA sequences located in the promoter of target genes called retinoic acid-responsive elements (RAREs) or retinoid X-responsive elements (RXREs). In vitro binding studies have demonstrated that both all-trans-RA (RA) and 9-cis-RA are ligands for the RARs, whereas only 9-cis-RA has been shown to be a ligand for the RXRs (12, 13).

Like other members of the steroid/thyroid hormone superfamily, RARs and RXRs consist of six functionally distinct domains designated A-F (2). Unique functions have been described for several of these domains. The A and B domains are important for the ligand-independent transactivation function (AF-1). The C domain, which contains two zinc fingers, is important for both DNA binding and receptor dimerization. The E domain is functionally complex. In addition to containing all the information necessary for high affinity ligand binding, it also contains a ligand-dependent transactivation function (AF-2) and accessory dimerization sequences.

Recently there have been major advances in understanding the nature of the ligand binding domain of RARs and RXRs. High resolution crystal structures of the ligand binding domains of apo-RXRalpha and holo-RARgamma have demonstrated that these receptors share a similar overall fold (14, 15). Furthermore, several amino acids within the ligand binding domains of RARs have been reported to be functionally important for high affinity binding of RA. Previous studies from this laboratory have demonstrated that the positively charged amino acid residues, Arg269 and Lys220, are critical for the high affinity binding of RA and retinoid specificity of RARbeta (16, 17). Ostrowski et al. (18) have reported that Ala225/Ile232 of RARbeta and the homologous amino acid residues Ser232/Thr239 of RARalpha are important for discrimination of subtype-specific synthetic ligands, and Lupisella et al. (19) have demonstrated using fluorescence quenching techniques that Trp227 is within the ligand binding site of RARgamma . Finally, Dallery et al. (20) and Sani et al. (21) have demonstrated a loss of RA binding after treatment of the EF domain of RARalpha and chicken skin RARs, respectively, with several sulfhydryl-modifying reagents suggesting that Cys residue(s) may be located in the ligand binding pocket of these receptors. The goal of this work was to identify which Cys residue(s) are important for this loss of RA binding and to address the role that this Cys residue(s) plays in the binding of RA.

In the current report we have shown that, similar to RARalpha , RA binding to both RARbeta and RARgamma is inhibited by the sulfhydryl-specific modifying reagents DTNB and MMTS and that the sulfhydryl group of the reactive Cys residue(s) of RARbeta lies in or near the retinoid binding pocket. These Cys residue(s) alone do not play a critical role in RA binding. Rather, these modified Cys residue(s) of RARbeta function to sterically hinder the binding of RA. The modified Cys residues critical for the inhibition of RA binding differ depending on the reagent employed. Only modification of Cys228 is necessary when RARbeta is treated with reagents that transfer large bulky groups to reactive Cys residues, whereas both Cys228 and Cys267 must be modified when a small functional group is transferred such as thiomethyl or thiocyanide to fully inhibit RA binding. Taken together these data suggest that Cys228 most likely lies closer to the opening of the ligand binding pocket of RARbeta , whereas Cys267 lies more deeply buried within the ligand binding pocket.


MATERIALS AND METHODS

Plasmid Constructs and Site-directed Mutagenesis

Mutants were created according to the PCR site-directed mutagenesis technique described by Higuchi et al. (22). pSG5-RARbeta , a gift from Professor Pierre Chambon, Strasbourg, France, was linearized with XbaI and used as a template for the preparation of each mutant. Both sense (s) and antisense (as) oligonucleotide primers were purchased from Ransom Hill BioScience (La Jolla, CA). The GCT codon was used to encode the mutant Ala residue indicated in bold and underlined in the mutagenic primers.

For the preparation of C228A, two separate PCR fragments were prepared using the primer pairs RARbeta 5'-s (5'GGGAGGGATCCATCGAGGGTAGATTTGACTGTATGGAT 3') plus C228A-as (5'CTTAATAAT<UNL><B>AGC</B></UNL>CTTGGTGGC 3') and RARbeta 3'-as (5' GAAGGAAGCTTTCACTGCAGCAGTGGTGA 3') plus C228A-s (5'GCCACCAAG<UNL><B>GCT</B></UNL>ATTATTAAG 3'), respectively. The two PCR fragments were purified, annealed, and amplified in a second PCR reaction using the RARbeta 5'-s and RARbeta 3'-as primers. Likewise, the C258A and the C267A mutants were constructed using the RARbeta 5'-s and the RARbeta 3'-as primers and the following mutagenic primers: C258A-s (5'AAAGCCGCC<UNL><B>GCT</B></UNL>TTGGATATC3'), C258A-as (5' GATATCCAA<UNL><B>AGC</B></UNL>GGCGGCTTTG3'), C267A-s (5'CTCAGAATT<UNL><B>GCT</B></UNL>ACCAGGTATAC3') and C267A-as (5' ATACCTGGT<UNL><B>AGC</B></UNL>AATTCTGAG3'). The MscI-StuI restriction fragment containing each Ala mutation was exchanged with that of full-length wild type RARbeta previously cloned in frame into the NotI restriction site of pET29a(+) (Novagen) (pET29a-RARbeta ). In all cases, the presence of the specific mutation and the lack of random mutations were verified by DNA sequence analysis (23).

The entire coding sequences of mouse RARalpha and RARgamma (gift from Professor Pierre Chambon) were cloned in frame into the BamHI/HindIII (RARalpha ) or NotI (RARgamma ) restriction sites of pET29a(+).

Expression of RARbeta and Preparation of Receptor Extracts

Each RAR expression construct was transformed into Escherchia coli K12 strain BL21(DE3) cells (Novagen) (24). Ten µl from a frozen glycerol stock was used to inoculate 5 ml of LB medium containing 30 µg/ml kanamycin and shaken at 37 °C until the A600 was 0.6-1.0. This culture was stored overnight at 4 °C, and the following morning 1 ml of cells was used to inoculate 50 ml of LB medium containing 30 µg/ml kanamycin. This culture was incubated with shaking at 37 °C until the A600 reached 0.6 to 1.0. At this time the cells were induced to express the recombinant S-Tag RAR fusion proteins by the addition of isopropyl-1-thio-beta -D-galactopyranoside to a final concentration of 1 mM. Ninety minutes later the cells were harvested by centrifugation at 5000 × g for 15 min.

The receptor extracts were prepared from a 50-ml culture by resuspending the cell pellet in lysis buffer (50 mM Tris, pH 8.0, 0.1% Triton X-100, and 200 µg/ml lysozyme) at a final concentration of 1 ml of lysis buffer/10 ml of induced culture, freezing at -70 °C followed by rapid thawing at 37 °C, sonication until the solution lost its viscosity, and centrifugation at 17,000 × g for 30 min. The supernatant was aliquoted and stored frozen at -80 °C. Total protein concentration of the receptor extracts was determined with the Bio-Rad protein assay kit using crystalline bovine serum albumin as the standard. The production of the recombinant S-Tag wild type and mutant RAR fusion proteins in the receptor extracts was monitored using the S-Tag Western blot Kit (Novagen).

Retinoic Acid Binding Assays

The receptor extracts were diluted in binding buffer (40 mM HEPES, pH 7.9, 120 mM KCl, 10% glycerol, 0.1% (w/v) gelatin, 1 mM EDTA, 4 mM dithiothreitol (DTT), and 5 µg/ml each of the protease inhibitors aprotinin and leupeptin) to a final concentration of 10-30 µg of total protein (0.1-0.3 pmol of wild type). Similar total protein concentrations were used for the mutant proteins. Total RA binding was determined in the diluted protein extracts by adding [3H]RA (1.82-1.92 TBq/mmol or 49.2-52.0 Ci/mmol, DuPont NEN) in the concentration range of 0.1-20 nM and incubating for 3 h at 27 °C. Dilutions of RA were made in ethanol. In a duplicate set of reactions nonspecific binding was determined in the presence of 200-fold molar excess unlabeled RA (generous gift from Hoffmann-LaRoche). The contribution of ethanol to the total volume was the same for each RA concentration and was equal to 2%. Bound RA was separated from free by extraction with 3% (w/v) equal particle size charcoal-dextran prepared as described by Dukoh et al. (25). All steps in the procedure were performed under yellow light. Specific RA binding was determined by subtracting the nonspecific binding from the total binding. The nonspecific binding was always less than 12% of the total binding. No specific RA binding was detected in receptor extracts prepared from cells containing pET29a-RARbeta which were not induced to express RAR with isopropyl-1-thio-beta -D-galactopyranoside or in receptor extracts prepared from cells containing the empty pET29a plasmid and treated with isopropyl-1-thio-beta -D-galactopyranoside. Apparent equilibrium dissociation constants (Kd) were determined for the wild type and each mutant protein by Scatchard analysis (26).

Chemical Modification of Receptors

The receptor extracts were diluted in the same binding buffer as described above for Kd determinations except that DTT was omitted. Stock solutions of modifying or reducing agents were made up as 100 × stocks in the following solvents: 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB) (Sigma) in ethanol; methyl methanethiosulfonate (MMTS) (Sigma) in ethanol; beta -mercaptoethanol (BME) (Fisher) in binding buffer; DTT (Sigma) in binding buffer; sodium arsenite (Fisher) in binding buffer; N-ethylmaleimide (NEM) (Sigma) in ethanol, p-hydroxymercuribenzoate (p-HMB) (Sigma) in binding buffer, tri-n-butyl-phosphine (TBP) (Aldrich) in ethanol; and sodium cyanide (NaCN) in 0.225 M potassium phosphate, pH 7.2. In all cases, the chemical modification reactions were performed on ice for 5 min. For the chemical reversal or termination of a reaction, chemically modified extract was incubated with reducing agent for an additional 5 min on ice. Total RA binding was determined in duplicate reactions for each treatment group by incubating the sample in the presence of a saturating concentration of 15 nM [3H]RA as determined from the saturation binding experiments (Fig. 3) as described above. The specific [3H]RA binding was determined for each treatment group and expressed as the percent of RA binding in the absence of the chemical-modifying reagent for each protein (% of control binding). The amount of specific bound RA at 100% relative binding was in the range of 0.1-0.3 pmol. Nonspecific binding was always less than 12% of the total bound RA.


Fig. 3. Titration of wild type and mutant RARbeta fusion proteins with RA. Specific binding of [3H]RA to wild type (A), C228A (B), C258A (C), and C267A (D) from a representative experiment is shown. The inset for each panel is one of the three Scatchard plots which was used to calculate the apparent Kd values presented in Table I.
[View Larger Version of this Image (29K GIF file)]



RESULTS

The Effect of Sulfhydryl Modification of RARalpha , RARbeta , and RARgamma on RA Binding

Fig. 1, A-F, depicts the results of the treatment of the full-length recombinant RARalpha , RARbeta , and RARgamma fusion proteins with the Cys-modifying reagents DTNB and MMTS. Similar to what has been previously reported (20, 21), full-length RARalpha exhibits a dramatic reduction in RA binding upon treatment with both of the sulfhydryl-modifying reagents (Fig. 1, A and B). Furthermore, this loss of RA binding is highly specific and is not due to the disruption of structural disulfide bonds in the protein since it is reversible to a level comparable with that of the untreated RARalpha upon the addition of the reducing agents BME or DTT. Likewise, RA binding to RARbeta and RARgamma was similarly affected by both DTNB (Fig. 1, C and E) and MMTS (Fig. 1, D and F). In all cases this inhibition of RA binding was also highly specific since it was reversed by reducing agents. Note that RA binding by RARgamma following MMTS treatment was reversible to approximately 50% of that of untreated RARgamma using DTT. However, treatment of the MMTS modified RARgamma with the more hydrophobic reducing reagent TBP (27) restored RA binding to a level comparable with that of the untreated RARgamma .


Fig. 1.

Inhibition of RA binding activity after treatment of RARs with sulfhydryl-modifying reagents. Receptor extracts containing RARalpha (A and B), RARbeta (C and D), and RARgamma (E and F) were treated with the indicated concentrations of either DTNB (A, C, and E) or MMTS (B, D, and F) for 5 min on ice. In a duplicate set of reactions the DTNB modification was reversed with 20 mM BME, and the MMTS modification was reversed with either 10 mM DTT or 10 mM TBP. In all cases, the reducing agent was added at the end of the 5-min incubation with the modifying reagent, and the samples were incubated on ice for an additional 5 min. Following treatment RA binding was determined using 15 nM [3H]RA as described under the "Materials and Methods." The specific RA binding was determined for each treatment group and was expressed as the percent of RA binding in the absence of the chemical modifying reagent for each protein (% of control binding). Values are mean ± S.E. for at least three independent experiments performed in duplicate.


[View Larger Version of this Image (18K GIF file)]


We next wished to determine if any of these receptors contain reactive vicinal dithiols which are responsible for this loss of RA binding. For these experiments we used sodium arsenite which has been demonstrated to cross-link two thiol groups that are spatially close (vicinal dithiols) (28, 29). Treatment of the recombinant RARalpha , RARbeta , and RARgamma fusion proteins with 10 mM sodium arsenite did not result in any significant inhibition of RA binding (data not shown).

In order to determine if the modified Cys residues following DTNB and MMTS treatment are located within the RA binding pocket of RARbeta , we tested the ability of RA to protect critical Cys residues from chemical modification. In this experiment we used the alkylating agent NEM since the modification reaction can be terminated but not reversed by the addition of a reducing agent. Fig. 2 demonstrates that similar to DTNB and MMTS, RA binding to apo-RARbeta is highly sensitive to modification by NEM. Furthermore, Fig. 2, inset, shows that 82% of the RA binding activity was maintained when holo-RARbeta was treated with 1 mM NEM. This demonstrates protection of the critical Cys residue(s) to NEM modification when RARbeta is liganded with RA (holo-RAR) prior to NEM treatment. This result strongly suggests that the target Cys residue(s) in RARbeta , whose modification by sulfhydryl-modifying reagents including NEM, DTNB, and MMTS results in inhibition of RA binding, lie in or near the retinoid binding site.


Fig. 2. Protection of RARbeta by RA from modification with NEM. Samples of receptor extract containing RARbeta were treated with the indicated concentrations of NEM for 5 min on ice. Following treatment with NEM, RA binding was determined as described under the "Materials and Methods." Values are mean ± S.E. for at least three independent experiments performed in duplicate. Inset, samples of receptor extract containing RARbeta were preincubated with 1 µM unlabeled RA for 3 h at 27 °C. Following the 3-h incubation the receptor samples were treated for 5 min on ice with 1 mM NEM (+NEM) or ethanol (-NEM) followed by addition of 10 mM BME and further incubation on ice for 10 min. Unbound RA was removed by charcoal/dextran extraction followed by incubation of the samples at 4 °C for 30 min to allow the dissociation of the bound RA. RA binding was then determined for samples in each of the treatment groups using 15 nM [3H]RA as described in the legend to Fig. 1. Values are mean ± S.E.
[View Larger Version of this Image (36K GIF file)]


Creation of Site-specific Mutants

Chemical modification of critical Cys residues located within the retinoid binding site of RARbeta could result in loss of RA binding by either the obliteration of a direct interaction of one or more Cys residues with RA or steric hindrance resulting in the inability of RA to access the binding pocket. In order to identify the critical Cys residue(s) and to determine which of these mechanisms are important in the inhibition of RA binding to RARbeta following sulfhydryl modification, we have created three site-specific mutants of RARbeta in which Cys228, Cys258, and Cys267 were individually replaced with an Ala residue (C228A, C258A, and C267A). We have chosen to focus on these residues because Dallery et al. (20) have demonstrated that RA binding to the RARalpha EF domains is sensitive to chemical modification of sulfhydryls. The EF domains of RARalpha contain six Cys residues that are absolutely conserved among the three RARs. Of these six residues, only three (Cys228, Cys258, and Cys267 of RARbeta ) lie within the NH2-terminal 100 amino acid portion of the E domain previously demonstrated to be sufficient for high affinity binding of RA (30, 31, 32).

Western blot analysis of receptor extracts containing the wild type and mutant RARbeta fusion proteins demonstrated a major band that migrated at the same position (approximate molecular mass of 55 kDa) along with several smaller molecular weight degradation products also of similar size. In addition, the wild type and mutant RARbeta fusion proteins displayed a similar level of expression (data not shown).

Effect of Site-specific Mutation on RA Binding

Fig. 3 shows representative saturation curves and corresponding Scatchard plots for wild type and each mutant RARbeta fusion protein. Average apparent Kd values were determined for each receptor using at least three separately prepared receptor extracts (Table I). The Kd value of 0.6 nM RA for wild type RARbeta fusion protein is in good agreement with previous reports (16, 17). Both C258A and C267A displayed a similar affinity for RA comparable with that of wild type RARbeta . Interestingly C228A, which has a Kd value of 4.4 nM, displayed a small decrease (approximately 7-fold) in affinity for RA when compared with that of wild type RARbeta .

Table I.

Comparison of the apparent Kd values for RA of wild type and mutant RARbeta s with and without MMTS treatment


RARbeta Kd for RA (nM)a
Untreated MMTS-treated

Wild type 0.6  ± 0.1 NDb
C228A 4.4  ± 0.1 4.9c
C258A 1.9  ± 0.1 >>20c,d
C267A 1.0  ± 0.1 1.0 ± 0.1

a  Mean ± S.E. for three independent measurements.
b  ND, not determined.
c  Single measurement.
d  No specific binding at 20 nM RA.

Effect of Treatment of C228A, C258A, and C267A Mutants with DTNB and p-HMB

Since the three mutant RARbeta s have an affinity for RA in the range of that of the wild type RARbeta , it is unlikely that sulfhydryl modification of RARbeta inhibits RA binding by obliterating a critical direct interaction between the ligand and any one of these Cys residues in the ligand binding domain of the receptor. Therefore, the alternative mechanism of inhibition involving steric blocking of RA entry into the binding pocket upon sulfhydryl modification of RARbeta was explored. We tested this hypothesis by subjecting each of the Cys mutants to sulfhydryl modification with the assumption that if one or more chemically modified Cys residues mediate the inhibition of RA binding, then replacement of the reactive Cys with an unreactive Ala residue should result in a receptor with RA binding activity comparable with that of the unmodified receptor.

Fig. 4A shows the results of DTNB treatment of each of the mutants, plotted along with wild type RARbeta for comparison. The profiles for inactivation of RA binding by C258A and C267A after DTNB treatment are similar to that of wild type RARbeta suggesting that Cys258 and Cys267 are either not accessible to modification by DTNB or that modification of either of these residues by DTNB is not sufficient to interfere with RA binding. On the other hand, RA binding to C228A is not affected by DTNB modification. This strongly suggests that modification of Cys228 alone with DTNB is responsible for the inhibition of RA binding following DTNB treatment.


Fig. 4. The effect of DTNB and p-HMB treatment of wild type and mutant RARbeta on RA binding. Receptor extracts containing wild type, C228A, C258A, and C267A RARbeta fusion proteins were treated with the indicated concentrations of either DTNB (A) or p-HMB (B) for 5 min on ice. Following treatment with the sulfhydryl-modifying reagent RA binding was determined using 15 nM [3H]RA as described in the legend to Fig. 1. The values presented are the mean ± S.E. for at least three independent determinations performed in duplicate.
[View Larger Version of this Image (16K GIF file)]


In order to confirm this conclusion we have also used the sulfhydryl-specific modifying reagent p-HMB to test each mutant in a similar fashion. Modification of a sulfhydryl by p-HMB involves the transfer of a substituted benzyl ring which is similar in size and structure to the thiobis(2-nitrobenzoic acid) moiety which is transferred during DTNB modification. As can be seen in Fig. 4B, the profiles of inactivation of RA binding by the wild type and mutant RARbeta s treated with p-HMB are essentially identical to that of their respective DTNB-modified receptor. These results further demonstrate that modification of Cys228 alone with either DTNB or p-HMB is sufficient to inhibit RA binding.

Effect of Treatment of C228A, C258A, and C267A Mutants with MMTS

In order to determine if the size of the modifying group is important, we performed similar experiments using MMTS. Unlike DTNB and p-HMB, MMTS treatment of a protein results in the transfer of the small thiomethyl-blocking group to reactive sulfhydryl groups (27). Fig. 5A shows the RA binding activity of the wild type and mutant fusion proteins after modification with MMTS. C258A demonstrated a loss of RA binding upon MMTS treatment which was identical to that of wild type RARbeta further demonstrating that Cys258 is not the Cys residue whose modification inhibits the binding of RA. C228A modified with MMTS displayed RA binding which was approximately 85% of that of the unmodified C228A. This is consistent with the findings obtained with DTNB and p-HMB which demonstrate that Cys228 is an important reactive Cys. Interestingly the MMTS-treated C267A RARbeta displayed approximately 60% of the RA binding activity of the unmodified C267A (Fig. 5A). Furthermore, this inhibition of RA binding by the unmodified C267A was not reversible with DTT but was reversible with the hydrophobic reducing agent TBP (Fig. 5B).


Fig. 5. The effect of MMTS treatment of wild type and mutant RARbeta on RA binding. Receptor extracts containing wild type, C228A, C258A, and C267A RARbeta fusion proteins were treated with the indicated concentrations of MMTS for 5 min on ice (A). In a duplicate set of reactions the MMTS modified C267A was treated with either 10 mM DTT or 10 mM TBP as described in the legend to Fig. 1 (B). Following treatment with the sulfhydryl-modifying reagent and/or reducing reagents, RA binding was determined using 15 nM [3H]RA as described in the legend to Fig. 1. The values presented are the mean ± S.E. for at least three independent determinations performed in duplicate.
[View Larger Version of this Image (14K GIF file)]


Since both C228A and C267A retained a considerable amount of RA binding following treatment with MMTS, we next measured the apparent Kd of both the MMTS unmodified and modified forms (treated with 0.25 nM MMTS) of these two receptors along with C258A (Table I). No specific RA binding was observed up to a concentration of 20 nM by the MMTS-modified C258A. This further demonstrates that either Cys228, Cys267, or both are the modified sulfhydryls in RARbeta responsible for the loss of RA binding upon MMTS modification. Interestingly the Kd values of the MMTS-modified forms of both C267A and C228A were very similar to that of their unmodified form (Table I). This suggests that the inactivation of RA binding of RARbeta by MMTS, which involves the transfer of the small thiomethyl-blocking group, requires the modification of both Cys228 and Cys267.

To further examine the effect of modification of only Cys228 with a small functional group compared with a large bulky group on RA binding, we have examined the effect of cyanolysis of DTNB-modified C258A and C267A. As demonstrated above in Fig. 4, inhibition of RA binding of DTNB-modified RARbeta involves the modification of Cys228. Treatment of DTNB-modified Cys residues with NaCN results in the replacement of the bulky thiobis(2-nitrobenzoic acid) group with thiocyanide which is similar in size to the thiomethyl group transferred upon MMTS treatment (33, 34). It is important to note that MMTS-treated C267A and DTNB-modified C267A treated with NaCN are virtually the same except for the presence of a thiomethyl group instead of a thiocyanide group on the modified Cys. Treatment of DTNB-modified C258A and C267A with NaCN results in a dose-dependent restoration of RA binding which plateaued at greater than 70% of the wild type RA binding at a concentration of 50 nM NaCN (data not shown). It is important to note that this level of RA binding is similar to the level of RA binding observed with MMTS-treated C228A and C267A. This suggests that modifications of Cys228 which result in the transfer of small moieties such as thiocyanide and thiomethyl is not sufficient to fully inhibit RA binding. Taken together this suggests that loss of RA binding by MMTS treatment of RARbeta is the result of the modification of both Cys228 and Cys267.


DISCUSSION

In the current report we have extended the findings of Dallery et al. (20) to demonstrate that RA binding to all three RAR types is blocked similarly by specific sulfhydryl-modifying agents. The concentrations of DTNB and MMTS necessary to inhibit RA binding to RARbeta and RARgamma are similar to those which have been reported previously for the ligand binding domain of RARalpha (20). The reversibility of this effect by reducing agents demonstrates the specificity of these reactions to sulfhydryl groups. Interestingly, unlike RARalpha and RARbeta , RA binding by RARgamma following MMTS treatment was only reversible to about 50% of control binding with the polar reducing agent DTT. However RA binding was restored to near unmodified levels using the hydrophobic reducing agent TBP. This suggests that there is at least one critical Cys residue responsible for inhibition of RA binding upon methanethiolation with MMTS which lies within a more hydrophobic region of RARgamma than in RARbeta and RARalpha . This region of greater hydrophobicity in the ligand binding pocket of RARgamma compared with RARbeta and RARalpha is most likely one of several factors responsible for the ability of synthetic retinoids to display receptor selectivity.

The glucocorticoid receptor (GR), another member of the steroid/thyroid hormone superfamily of receptors, has also been shown to be sensitive to sulfhydryl modification (35). GR modified with MMTS exhibits a bimodal response in which inhibition of ligand binding is maximal both at low and high concentrations of the modifying reagent (36, 37). It has been determined that this response is a result of intramolecular disulfide bond formation between two closely spaced Cys residues (vicinal dithiols) located in the ligand binding pocket of GR upon treatment with low concentrations of MMTS (36, 37). We did not observe this biphasic pattern of inhibition of RA binding in any of the RARs treated with MMTS under similar conditions. In addition, our sodium arsenite results further support the conclusion that there are no reactive vicinal dithiols in any of the three RARs responsible for inhibition of RA binding.

The results of our protection-exchange experiment with RARbeta demonstrating that RA binding can protect Cys residues from chemical modification suggests that these critical Cys residue(s) may be located in or near the ligand binding pocket. Protection may result from either the physical blocking of the modifying reagents by RA or a conformational change in the ligand binding domain upon RA binding which blocks the accessibility of the reactive Cys residues to the modification reagents. The latter is more likely based on the work of Renaud et al. (15) showing that the major difference in the crystal structure of apo-RXRalpha and holo-RARgamma ligand binding domains is in the position of helix 12. In the apo-receptor the entrance to the RA binding site is open while in the holo-receptor helix 12 collapses over the entrance to the binding pocket blocking access to the site.

The loss of RA binding following treatment with MMTS and DTNB cannot be explained by the elimination of critical interactions between any of the Cys residues examined and RA since each of the three Cys mutants have an apparent Kd for RA in the range of that of wild type RARbeta . Interestingly, the C228A mutant did display a 7-fold decrease in affinity for RA when compared with wild type RARbeta . This is not surprising since two amino acids of RARbeta , close in the primary sequence of Cys228 (Ala225 and Ile232), have been demonstrated to be important for interaction with type-selective ligands (18). Furthermore, the crystal structure of holo-RARgamma indicates that Cys237, the homologous residue to Cys228 of RARbeta , lies within 4 Å of RA (15). It is possible that substitution of Cys228 with an Ala removes an interaction between this Cys residue and RA which contributes to the overall stability of holo-RARbeta . On the other hand, it is also possible that this mutation has caused a small disruption of the local topography of the ligand binding pocket sufficient to result in this modest decrease in the affinity of C228A for RA.

Although inhibition of RA binding upon sulfhydryl modification of RARbeta is a result of steric blocking of RA binding, the Cys residues whose modification is critical for this inhibition of RA binding differ depending on the reagent employed. Wild type RARbeta displays similar sensitivity to DTNB and MMTS treatment; however, the C228A and C267A mutants represent novel receptors with unique sensitivity to these two sulfhydryl modifying reagents. Our finding of a differential response of these two mutants to reagents that transfer different size blocking groups is not unprecedented. It has been reported that specific sulfhydryl modification of rabbit muscle creatine kinase with blocking groups of increasing size resulted in the incremental decrease in enzymatic activity (38). Similar results have been found for the GR. Modification of critical Cys residues in GR with MMTS results in a greater amount of residual binding of ligand than modification by iodoacetamide which transfers a relatively large blocking group (36).

Our data demonstrate that both Cys228 and Cys267 are located within the ligand binding pocket of RARbeta such that when modified they inhibit the binding of RA. However Cys228 lies closer to the opening of the ligand binding pocket than Cys267 because Cys228 is accessible to modification by both large and small reagents and upon modification results in the inhibition of RA binding, whereas Cys267 is only sensitive to small reagents. We have estimated the size of DTNB and MMTS by measuring the carbon center to carbon center distances on molecular models. The dimensions of MMTS, when viewed as a molecular structure, are 3.0 Å in height and 4.0 Å in width, whereas those of DTNB are 6.0 and 7.4 Å, respectively. Based on our sulfhydryl modification results, these measurements suggest that Cys267 is located in a region of the pocket not accessible to molecules or side chains of molecules much larger than 4.0 Å in diameter. Since RA is no greater than 4.0 Å at its carboxylate end and at least 5.0 Å at the beta -ionone ring, we can conclude that Cys267 of RARbeta is closest to the carboxylate end of RA. This conclusion is supported by the recently published crystal structure of the ligand binding domain of holo-RARgamma which demonstrates that the homologous residue of Cys228 (Cys237) is within 4 Å of carbon 13 of RA (15). In addition, there is only one amino acid residue in the primary sequence between Cys267 and Arg269 of RARbeta (Arg278 of RARgamma ). We have previously shown Arg269 of RARbeta to be important for RA binding and retinoid specificity most likely by interaction with the carboxyl group of RA (16, 17), and the crystal structure of the ligand binding domain of RARgamma places Arg278 within 4 Å of the carboxylate oxygen 22 of RA (15).


FOOTNOTES

*   This work was supported by National Institutes of Health Grant DK44517 (to D.R.S.). The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§   Supported by the M.D./Ph.D. Program of Temple University School of Medicine.
**   Recipient of a Research Career Development Award from the National Institutes of Health (HD01076). To whom correspondence should be addressed: Dept. of Biochemistry, Temple University School of Medicine, 3420 N. Broad St., Philadelphia, PA 19140. Tel.: 215-707-3266; Fax: 215-707-7536.
1    The abbreviations used are: RAR, retinoic acid receptor; RA, all-trans-retinoic acid; RXR, retinoid X receptor; RARE, retinoic acid response element; RXRE, retinoid X response element; DTNB, 5,5'-dithiobis(2-nitrobenzoic acid); MMTS, methyl methanethiosulfonate; PCR, polymerase chain reaction; BME, beta -mercaptoethanol; DTT, dithiothreitol; NEM, N-ethylmaleimide; p-HMB, p-hydroxymercuribenzoate; TBP, tri-n-butyl-phosphine, NaCN, sodium cyanide; GR, glucocorticoid receptor.

Acknowledgments

We thank Professor Pierre Chambon for the pSG5-RAR constructs and F. Hoffmann-LaRoche for the RA used in these studies. We also thank Dr. Charles Grubmeyer for many helpful discussions.


REFERENCES

  1. Gudas, L., Sporn, M. B., and Roberts, A. B. (1994) in The Retinoids (Sporn, M. B., Roberts, A. B., and Goodman, D. S., eds), 2nd Ed., pp. 453-520, Raven Press, Ltd., New York
  2. Mangelsdorf, D. J., Umesono, K., and Evans, R. M. (1994) in The Retinoids (Sporn, M. B., Roberts, A. B., and Goodman, D. S., eds), 2nd Ed., pp. 319-350, Raven Press, Ltd., New York
  3. Giguere, V., Ong, E. S., Segui, P., and Evans, R. M. (1987) Nature 330, 624-629 [CrossRef][Medline] [Order article via Infotrieve]
  4. Petkovich, M., Brand, N. J., Krust, A., and Chambon, P. (1987) Nature 330, 444-450 [CrossRef][Medline] [Order article via Infotrieve]
  5. Benbrook, D., Lernhardt, E., and Pfahl, M. (1988) Nature 333, 444-450
  6. Brand, N., Petkovich, M., Krust, A., Chambon, P., de The, H., Marchio, A., Tiollais, P., and Dejean, A. (1988) Nature 332, 850-853 [CrossRef][Medline] [Order article via Infotrieve]
  7. Krust, A., Kastner, P., Petkovich, M., Zelent, A., and Chambon, P. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 5310-5314 [Abstract]
  8. Zelent, A., Krust, A., Petkovich, M., Kastner, P., and Chambon, P. (1989) Nature 339, 714-717 [CrossRef][Medline] [Order article via Infotrieve]
  9. Kastner, P., Krust, A., Mendelsohn, C., Garnier, J.-M., Zelent, A., Leroy, P., Staub, A., and Chambon, P. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 2700-2704 [Abstract]
  10. Leroy, P., Krust, A., Zelent, A., Mendelsohn, C., Garnier, J.-M., Kastner, P., Dierich, A., and Chambon, P. (1991) EMBO J. 10, 59-69 [Abstract]
  11. Zelent, A., Mendelosohn, C., Kastner, P., Krust, A., Garnier, J.-M., Ruffenach, F., Leroy, P., and Chambon, P. (1991) EMBO J. 10, 71-81 [Abstract]
  12. Heyman, R. A., Mangelsdorg, D. J., Dyck, J. A., Stein, R. B., Eichele, G., Evans, R. M., and Thaller, C. (1992) Cell 68, 397-406 [Medline] [Order article via Infotrieve]
  13. Levin, A. A., Sturzenbecker, L. J., Kazmer, S., Bosakowski, T., Huselton, C., Allenby, G., Speck, J., Kratzeisen, C., Rosenberger, M., Lovey, A., and Grippo, J. (1992) Nature 355, 369-361
  14. Bourguet, W., Ruff, M., Chambon, P., Gronemeyer, H., and Moras, D. (1995) Nature 375, 377-382 [CrossRef][Medline] [Order article via Infotrieve]
  15. Renaud, J.-P., Rochel, N., Ruff, M., Vivat, V., Chambon, P., Gronemeyer, H., and Moras, D. (1995) Nature 378, 681-689 [CrossRef][Medline] [Order article via Infotrieve]
  16. Tairis, N., Gabriel, J. L., Gyda, M., III, Soprano, K. J., and Soprano, D. R. (1994) J. Biol. Chem. 269, 19516-19522 [Abstract/Free Full Text]
  17. Tairis, N., Gabriel, J. L., Soprano, K. J., and Soprano, D. R. (1995) J. Biol. Chem. 270, 18380-18387 [Abstract/Free Full Text]
  18. Ostrowski, J., Hammer, L., Roalsvig, T., Pokornowski, K., and Reczek, P. R. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 1812-1816 [Abstract]
  19. Lupisella, J. A., Driscoll, J. E., Metzler, W. J., and Reczek, P. R. (1995) J. Biol. Chem. 270, 24884-24890 [Abstract/Free Full Text]
  20. Dallery, N., Sablonniere, B., Grillier, I., Formstecher, P., and Dautrevaux, M. (1993) Biochemistry 32, 12428-12436 [Medline] [Order article via Infotrieve]
  21. Sani, B. P., Singh, R. K., Reddy, L. G., and Gaub, M. P. (1990) Arch. Biochem. Biophys. 283, 107-113 [Medline] [Order article via Infotrieve]
  22. Higuchi, R., Krummel, B., and Saiki, R. K. (1988) Nucleic Acids Res. 16, 7351-7367 [Abstract]
  23. Sanger, F., Coulson, A. R., Barrell, B. G., Smith, A. J. H., and Rose, B. A. (1980) Mol. Biol. 143, 161-178
  24. Studier, F. W., Rosenberg, A. H., Dunn, J. J., and Dubendorff, J. W. (1990) Methods Enzymol. 185, 60-89 [Medline] [Order article via Infotrieve]
  25. Dukoh, S., Pike, W. J., Chandler, J. S., Mancini, J. M., and Haussler, M. R. (1981) Anal. Biochem. 116, 211-222 [Medline] [Order article via Infotrieve]
  26. Scatchard, G. (1946) Ann. N. Y. Acad. Sci. 51, 660-672
  27. Nishimura, J. S., Kenyon, G. L., and Smith, D. J. (1975) Arch. Biochem. Biophys. 170, 461-467 [Medline] [Order article via Infotrieve]
  28. Joshi, S., and Hughes, J. B. (1981) J. Biol. Chem. 256, 11112-11116 [Abstract/Free Full Text]
  29. Gaber, B. P., and Fluharty, A. L. (1968) Q. Rep. Sulfur Chem. 3, 317-322
  30. Damm, K., Heyman, R. A., Umesono, K., and Evans, R. M. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 2989-2993 [Abstract]
  31. Robertson, K. A., Emami, B., Mueller, L., and Collins, S. J. (1992) Mol. Cell. Biol. 12, 3743-3749 [Abstract]
  32. Shen, S., van der Saag, P. T., and Kruijer, W. (1993) Mech. Dev. 40, 177-189 [CrossRef][Medline] [Order article via Infotrieve]
  33. Vanaman, T. C., and Stark, G. R. (1970) J. Biol. Chem. 245, 3565-3573 [Abstract/Free Full Text]
  34. Degani, Y., Neumann, H., and Patchornik, A. (1970) J. Am. Chem. Soc. 92, 6969-6971 [Medline] [Order article via Infotrieve]
  35. Harrison, R. W., Woodward, C., and Thompson, E. (1983) Biochim. Biophys. Acta 759, 1-6 [Medline] [Order article via Infotrieve]
  36. Miller, N. R., and Simons, S. S., Jr. (1988) J. Biol. Chem. 263, 15217-15225 [Abstract/Free Full Text]
  37. Stancato, L. F., Hutchinson, K. A., Chakraborti, P. K., Simons, S. S., and Pratt, W. B. (1993) Biochemistry 32, 3729-3736 [Medline] [Order article via Infotrieve]
  38. Smith, D. J., Maggio, E. T., and Kenyon, G. L. (1975) Biochemistry 14, 766-771 [Medline] [Order article via Infotrieve]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.