no. vol. pp. chemistry the of journal in u.s.a. · pdf filethe best characterized biological...

THE JOURNAL OF BIOLWICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular B i o l o g y , Inc

Vol. 269, No. 16, Issue of April 22. pp. 12129-12136, 1994 Printed in U.S.A.

Photoaffinity Labeling of the A h Receptor with 3-[3H]Methylcholanthrene and Formation of a 165-kDa Complex between the Ligand-binding Subunit and a Novel Cytosolic Protein*

(Received for publication, June 29, 1993, and in revised form, December 9, 1993)

Sonia M. F. de Morais, John V. Giannone, and Allan B. OkeyS From the Department of Pharmacology, Medical Sciences Building, University of Toronto,

~~

Toronto, Ontario M5S lA8,Canada

The aromatic hydrocarbon (Ah) receptor is a cytosolic protein that binds halogenated ligands such as 2,3,7,8- tetrachlorodibenzo-p-dioxin (TCDD) and nonhalogenated ligands such as 3-methylcholanthrene (MC) and benzo[a]pyrene. The best characterized biological response mediated by the Ah receptor is induction of cy- tochrome P4501A1 (CYPlAl). Photoaffinity labeling of the Ah receptor has been reported only with halogenated ligands such as TCDD and some of its iodinated derivatives. In this study, photolabeling of the Ah receptor was achieved with the nonhalogenated aromatic hydrocarbon 13H]MC. Sources of Ah receptor were the mouse hepatoma cell line Hepa-lclc9 and the human colon adenocarcinoma line LSlSO. Cytosolic fractions either were used in a crude form or were enriched by glycerol density gradient centrifugation. These then were incubated with i3H]MC, irradiated with W light (1300 nm), precipitated with acetone, and analyzed by SDS-polyacrylamide gel electrophoresis. The yield of photoadduct formation was lower with 13H]MC (-1%) compared with [3HlTCDD (3.5%) in Hepa-lclc9 cells. The same was true in LSl80 cells, Le. the yield was 0.2% for ['HIMC uersus 5.48 0.26% for [3HlTCDD. The relative molecular mass of the [3HlMC-labeled receptor estimated by SDS-polyacrylamide gel electrophoresis was 94,600 2,400 (mean * S.E.) for Hepa-lclc9 cells and 113,600 2 3,200 for LSl80 cells; these are the same molecular masses as determined by photolabeling with f3H1TCDD. In velocity sedimentation assays of mouse cytosol, L3H1MC binds specifically to two cytosolic proteins: the 4 S carcinogen-binding protein and the Ah receptor (9 S ) . However, no photolabeling of the 4 S protein was detected in our experiments. ['HIMC photolabeling of the human Ah receptor from LS180 cells was detected only in experiments using enriched cytosolic preparations. In addition to the 95-kDa ligand-binding subunit, a specifically radiolabeled protein of 164,900 * 5,800 kDa was also detected in Hepa-lclc9 cytosol photolabeled with [3H]MC, suggesting cross-linking, by MC, of another subunit of the multimeric Ah receptor com-

* This work was supported by a grant (to A. B. 0. and Patricia A. Harper) from the National Cancer Institute of Canada with funds from the Canadian Cancer Society and by a grant (to A. B. 0.) from the Medical Research Council of Canada. Preliminary reports of this work were presented at the annual meetings of the Canadian Federation of Biological Societies (de Morais, S. M. F., Giannone, J. V., and Okey,A. B. (1991) Proc. Can. Fed. Bzol. SOC. 34, 112) and the American Society for Pharmacology and Experimental Therapeutics (de Morais, S. M. F., Giannone, J. V., and Okey, A. B. (1991) Pharmacologist 33, 164). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "adver- tisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$- To whom correspondence should be addressed. Tel.: 416-978-2723; Fax: 416-978-6395.

plex to the ligand-binding subunit. Immunochemical analysis showed that the ligand-binding subunit of the Ah receptor is one component of the 165-kDa complex. The other protein in the complex could not be identified with antibodies to the heat shock proteins hsp90 or hsp70 or with antibodies to the p59 protein or A h receptor nuclear translocator protein. The identity and func- tion of the protein that becomes cross-linked to the ligand-binding subunit require further investigation.

The aromatic hydrocarbon (Ah)' receptor is a cytosolic protein that mediates the biochemical and toxic effects of 2,3,7,8- tetrachlorodibenzo-p-dioxin (TCDD) and its related halogenated aromatic analogues (Landers and Bunce, 1991). Nonhalogenated ligands for the Ah receptor include carcino- gens such as 3-methylcholanthrene (MC) and benzo[alpyrene (Okey et al., 1984, 1988). Both classes of ligand bind to the Ah receptor with high affinity (Poland and Glover, 1974; Poland et al., 1976; Okey and Vella, 1982); in fact, the affinity with which MC binds to the cytosolic Ah receptor is nearly as high as the affinity of the prototype agonist, TCDD; moreover, MC is as effective as TCDD at transforming the receptor to the DNA- binding state (Riddick et al., 1994). The Ah receptor has been detected in several animal species (Denison et al., 19861, and more recently, i t has been identified in human tissues and cell lines (Roberts et al., 1986; Manchester et al., 1987; Harper et al., 1988, 1991; Lorenzen and Okey, 1991; Waithe et al., 1991). The best characterized biological response regulated by the Ah receptor, following ligand binding, is induction of CYPlAl gene expression (Nebert and Gonzalez, 1987; Okey, 1990). Following initial ligand binding to the receptor in the cytoplasm, the ligand-Ah receptor complex translocates to the nucleus and interacts with specific regulatory DNA sequences located up- stream (5') of the coding region of the CYPlAl gene (Whitlock, 19901, thereby enhancing transcription of the gene.

Until recently, physicochemical characterization of the Ah receptor had to be performed under nondenaturing conditions due to the reversible nature of the ligand binding and to lability of the receptor protein. This problem was overcome by the development of the photoaffinity labeling technique, which has been used extensively in the characterization of enzymes (Ruoho et al., 1973) and hormone receptors (Horwitz and Fran- cis, 1988). Photoaffinity labeling consists of covalently linking a ligand to a receptor by exposure to ultraviolet light, thereby

The abbreviations used are: Ah, aromatic hydrocarbon; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; MC, 3-methylcholanthrene; CBP, carcinogen-binding protein; TCDF, 2,3,7,8-tetrachlorodibenzofuran;

liquid chromatography; ARNT, Ah receptor nuclear translocator; ALBS, PAGE, polyacrylamide gel electrophoresis; HPLC, high performance

Ah receptor ligand-binding subunit.

12129

12130 Photolabeling of the Ah Receptor making the ligand-receptor complex amenable to physicochemical analysis. The ideal photolabeling ligand has photo- labile groups, such as an azido group, or has functional groups that are reactive under UV light (Bayley and Knowles, 1977). Photolabeling of the Ah receptor has been reported using iodinated derivatives of TCDD such as the azido derivative, 2-azido-3-['251]iodo-7,8-dibromodibenzo-p-dioxin (Poland et al., 1986; Perdew, 19921, and 7-[12511iodo-2,3,-dibromodibenzo-p-dioxin (Wang et al., 1991). r3H]TCDD has been shown to photolabel the nuclear Ah receptor (Landers et al., 1989) and has been useful in demonstrating that the ligand-binding subunit of the mouse Ah receptor has the same molecular mass in cytosol as in the nuclear form (Prokipcak and Okey, 1990); the same is true of the human Ah receptor (Harper et al., 1991). Using the photoaffinity labeling technique as an initial step, Bradfield et al. (1991) were able to purify enough mouse Ah receptor to obtain an amino-terminal sequence and to produce polyclonal antibodies to the Ah receptor ligand-binding subunit (Poland et al., 1991). Knowledge of the amino-terminal sequence was crucial in the design of probes to isolate cDNAs for the Ah receptor ligand-binding subunit (Ema et al., 1992; Bur- bach et al., 1992). This group has also reported the use of photoaffinity labeling to detect multiple forms of the receptor (Poland and Glover, 1987) and interstrain variations in the molecular mass of the ligand-binding subunit of the Ah receptor in mice (Poland and Glover, 1990). A combination of photoaffinity labeling and chemical cross-linking was used by Perdew (1992) to detect multimeric forms of the Ah receptor in cytosolic and nuclear fractions from mouse hepatoma Hepa-1 cells.

Photoaffinity labeling using nonhalogenated ligands for the Ah receptor is of interest since this class of ligands includes several compounds that are highly toxic or carcinogenic. In rodent cytosols, these ligands bind specifically and with high affinity to two cytosolic proteins: the Ah receptor itself (which sediments at 9 S in sucrose density gradients under conditions of low ionic strength) and the carcinogen-binding protein (CBP; which sediments at 4 S). A derivative of MC, l-ox0-3-[~H]methylcholanthrene (Arnold et al., 1987), has been reported to photolabel CBP. In that procedure, CBP was separated from other proteins, including the Ah receptor, by column chromatography prior to the photolabeling procedure; and hence, photolabeling of the Ah receptor could not have been assessed. A pyrene derivative, 4-[3H]azidopyrene, also has been used to photolabel and purify CBP (Collins and Marletta, 1986), and although in that study proteins were not separated by column chromatography prior to photolabeling, photoaffinity labeling of the Ah receptor was not detected.

The objectives of this study were to determine if L3H]MC could photolabel the mouse and human Ah receptors and, if so, whether the properties and identities of the proteinsb) photolabeled with r3H]MC would be the same as those of the proteins labeled with r3H]TCDD.

MATERIALS AND METHODS Chemi~als-[~H]MC (generally labeled; 23 Ci/mmol) was obtained

from Amersham Canada Ltd. (Oakville, Ontario, Canada). ['HITCDD (33 Ci/mmol) and nonradioactive 2,3,7,8-tetrachlorodibenzofuran (TCDF) were generous gifts from Dr. S. Safe (Texas A & M University). TCDD and TCDF are extremely toxic substances and should be handled with care, as described by Poland and Glover (1975). SDS-PAGE sup- plies and equipment were obtained from Bio-Rad. Sucrose (density gradient-grade) was obtained from Beckman Instruments. Nonradioactive MC was obtained from Aldrich, and molybdate (sodium salt), dithiothreitol, dexamethasone, bovine serum albumin, Ponceau S stain, and catalase were obtained from Sigma. Dimethyl sulfoxide, glycerol, char- coal (Norit A), and EDTA were obtained from Fisher.

Purification of 3-fH&"thyZ~holanthrene-[~H]MC was purified by reverse-phase high performance liquid chromatography (HPLC) using the following conditions: mobile phase, 100% methanol; flow rate, 1

mumin; ultraviolet detection a t 230 nm; and a 5-pm particle size Spherisorb C,, column (Johns Scientific, Toronto, Ontario, Canada). The retention time for the MC peak was -9 min. The fraction containing pure VHIMC was collected, dried under N, in the dark, and redissolved in toluene. Purity (297%) was assessed by HPLC using a radio- isotope detector (Beckman Instruments). Purified 13H]MC was stored as a toluene solution in the dark a t -20 "C. Working solutions in dimethyl sulfoxide were prepared as needed, and unused portions were dis- carded.

Cell Culture and Cytosolic Preparation-The mouse hepatoma cell line Hepa-lclc9, kindly provided by Dr. M. Dufresne (University of Windsor, Windsor, Ontario, Canada), and the human colon adenocarcinoma line LS180 (American Type Culture Collection, Rockville, MD) were grown as monolayer cultures under standard conditions (Harper et al., 1991). Hepa-lclc9 cytosolic fractions were prepared in HEGD buffer (25 mM HEPES, 3 mM EDTA, 10% (vh) glycerol, 2 mM dithiothreitol, pH 7.4). LS180 cytosolic fractions were prepared in HEGD buffer with the addition of 20 m sodium molybdate. Cytosols were labeled in vitro with either 13HlMC or L3H1TCDD for 1 h a t 4 "C in the absence or presence of a 100-fold higher concentration of a nonradioactive competing ligand (TCDF or nonradioactive MC) as indicated in the figure legends. In a separate experiment, cytosol was incubated with [3HlMC for 4 h a t 30 "C to convert the receptor to the 6 S DNA-binding form as described by Riddick et aZ. (1994). After incubation with L3HIMC, cytosol was separated on preparative glycerol gradients, and the 6 S and 9 S regions were collected separately, photoactivated by UV irradiation, and analyzed by gel slicing or Western blotting. Protein determinations were performed using the method of Bradford (1976).

Enrichment of Cytosol by Glycerol Density Centrifugation- Undiluted cytosol (- 10 mg of proteidml) was incubated with 'H-labeled ligand for 1 h a t 4 "C, layered onto glycerol density gradients (13-ml tubes), and centrifuged a t 372,000 x g,, for 2 h. The fractions corresponding to the Ah receptor (9 S) and the 4 S carcinogen-binding protein (Lesca et al., 1987) were collected and submitted to the photolabeling procedure as indicated in the figure legends.

Photoaffinity Labeling of Ah Receptor and Separation by SDS- PAGE-Cytosols labeled with 3H-labeled ligand or enriched cytosolic preparations were placed in 60-mm tissue culture dishes (without cov- ers) on ice. Samples were placed 2 cm under a 450-watt UV lamp (Hanovia, Ace Glass) and irradiated for 2 min. After irradiation, proteins were precipitated with cold acetone and stored a t -20 "C overnight. Precipitates were centrifuged, dissolved in sample buffer (60 mM Tris, 10% glycerol, 5% SDS, and 10 mM dithiothreitol, pH 6.81, and boiled for 5 min. Nonradiolabeled standard proteins (Bio-Rad) included myosin (200 kDa), @-galactosidase (116 kDa), phosphorylase b (97 kDa), bovine serum albumin (66 kDa), ovalbumin (46 kDa), and carbonic anhydrase (30 kDa). I4C-Labeled standards (Amersham Canada Ltd.) included myosin (200 kDa), phosphorylase b (doublet of 100 and 92.5 kDa), bovine serum albumin (69 kDa), ovalbumin (46 kDa), and carbonic anhydrase (30 kDa). Photoaffinity-labeled samples were analyzed by electrophoresis in the presence of SDS on 20-cm-long gels by the method of Laemmli (1970).

Analysis by Slicing and Fluorography-The slicing technique was performed either on gels or on nitrocellulose paper after electrophoretic transfer. Gels were fixed and stained with Coomassie Brilliant Blue R-250, followed by destaining in 45% methanol and 7% acetic acid and rehydration in 5% methanol and 7% acetic acid. Each sample lane was cut into 3-mm slices, and each slice was digested with 700 pl of Protosol (DuPont NEN) overnight at room temperature. After digestion, 10 ml of scintillation mixture (Ready-Value, Beckman Canada, Mississauga, Ontario, Canada) containing 100 pl of glacial acetic acid was added to each vial prior to determination of total radioactivity by liquid scintillation counting. For proteins transferred to nitrocellulose paper, staining was performed with Ponceau s, each sample lane was sliced into 3-mm slices, and total radioactivity in each slice was determined by liquid scintillation counting. The yield of photolabeling was calculated by adding the total radioactivity (disintegrationdminute) in the receptor peak (minus base-line background) and expressing it as a percentage of total receptor present in the initial nonphotolabeled cytosol, as previously determined by sucrose density gradient analysis. The relative molecular mass for the receptor was determined from a linear plot of log(re1ative molecular mass) uersus distance migrated by the standard proteins.

For analysis by fluorography, gels were stained with Coomassie Bril- liant Blue and incubated in Enlightening (DuPont NEN) before drying. Dried gels were placed in contact with Kodak X-Omat AR film (Eastman Kodak Co.) a t -70 "C for 6-10 weeks before photographic development.

Zmmunoblotting and Immunochemical Detection-Proteins were

Photolabeling of the A h Receptor 12131

separated by SDS-PAGE, transferred to nitrocellulose membranes, and stained with Ponceau S. The identification of the following proteins was performed using immunodetection: 1) ARNT protein, using a polyclonal anti-ARNT serum (kindly provided Dr. Oliver Hankinson, University of California, Los Angeles) at 1:5,000 dilution and anti-rabbit IgG horseradish peroxidase at 1:50,000 dilution (Amersham Canada Ltd.) as secondary antibody; 2) hsp90, using the AC88 monoclonal antibody at 1:5,000 dilution and anti-mouse I g G horseradish peroxidase at 1:50,000 dilution (Amersham Canada Ltd.); 3) hsp70, using the N27F3-4 monoclonal antibody at 1:5,000 dilution and anti-mouse IgG horseradish peroxidase at 1:50,000 dilution (Amersham Canada Ltd.); and 4) p59 protein, using the JP-1 polyclonal serum (a gift from Dr. Lee Faber, Medical College of Ohio) at 1:500 dilution and anti-goat IgG horseradish peroxidase at 1:25,000 dilution. Polyclonal antiserum to a synthetic peptide corresponding to the amino terminus of the ligand-binding subunit (Poland et al., 1991) was prepared and characterized in our labo- ratory.’

Blots were incubated in Blotto (3% skim milk powder in TNT buffer (0.1 m~ Tris, 150 m~ NaC1,0.2% Tween 20, pH 8.0)) overnight to block nonspecific sites and then washed twice with TNT buffer. The primary antibody solution in Blotto was then added and incubated for 3 h at room temperature, followed by three washes with TNT buffer. The blots were then incubated with the horseradish peroxidase-linked secondary antibody solution (in Blotto) for 1 h at room temperature and washed three times with TNT buffer. Detection of peroxidase activity was performed by adding chemiluminescence detection reagents (ECL detection system, Amersham Canada Ltd.) and incubating for 1 min. Visu- alization was performed by exposing the blot to a light-sensitive film (Hyperfilm-ECL, Amersham Canada Ltd.) until optimum exposure (1-30 min depending upon the antibody used).

Immunoprecipitation with Antibodies to ARNT Protein and Ligand- binding Subunit-Hepa-lclc9 cytosol (1 mg of proteidml, 3-ml total volume) was incubated with 10 n~ t3HIMC for 1 h and photolabeled for 3 min. A control (nonphotolabeled) sample was run in parallel through- out the procedure. Anti-ARNT serum (100 pl) or anti-ALBS serum (100 pl) was added to the incubations, and these were left to incubate overnight at 4 “C. Approximately 200 pl of protein A-Sepharose CL-4B (Pharmacia Canada Inc., Dorval, Quebec, Canada) was added to the incubations and left at room temperature for 2 h with mixing. Nonspe- cifically bound proteins were washed away with HEGD buffer containing 150 m~ NaCl and 1% Nonidet P-40. Sample loading buffer for SDS-PAGE (500 pl) was added, and the mixture was left at room temperature for 2 h. Sepharose beads were separated by centrifugation.

RESULTS

Irradiation of Hepa Cytosol-A radiolabeled protein of -95 kDa was observed when Hepa-lclc9 cytosolic extracts were incubated with l3H1MC, irradiated with W light at >300 nm for 2 min, and analyzed on 10% polyacrylamide gels. No radiolabeled protein was detected in a similar incubation that did not undergo irradiation (dark control) (Fig. 1). In a time course for the photoaffinity labeling, aliquots of a [3H]MC/Hepa-lclc9 cytosol incubation where exposed to UV light for increasing lengths of time. The maximal yield of photolabeling was achieved in -2 min, and further exposure to W light did not improve photolabeling yield (Fig. 2). Sucrose density gradient analysis of the photolability of the Ah receptor in the presence of ligand showed that after 2 min of exposure to W light, -70% of the Ah receptor loses its ability to bind ligand reversibly (Fig. 3). The reversible ligand binding ability of 4 S CBP was unaf- fected by exposure to W light (Fig. 3).

Specificity of Photolabeling-The specificity of the [3H]MC- photolabeled Ah receptor peak for Ah receptor agonists was indicated by the absence of radioactive peaks when an excess concentration of competitor ligand (TCDF or nonradiolabeled MC) was used in the incubation (Fig. 4). In addition to the Ah receptor that had a molecular mass of 94,600 2 2,400 kDa (mean 2 S.E., n = lo), a protein of higher molecular mass (164, 900 2 5800 kDa, n = 5) was also specifically photolabeled with l3H1MC (Fig. 4). In this experiment, the separating gel con-

’ J. V. Giannone, A. B. Okey, and P. A. Harper, Can. J. Physiol. Phar- macol., submitted for publication.

h 400 E a

200 1 1 6 9 7 6 6 4 3 3 1

3 v v v v Y Y

u) [3H]MC, NON-IRRADIATED

a 2 200-

5 100-

> t 2 U

6 0

$ 04 I 0 4 0 1 2

DISTANCE (crn) FIG. 1. Photoaffinity labeling of cytosolic Ah receptor from

mouse Hepa-lclc9 cells with 3-[SHlmethylcholanthrene. Cytosol (4 ml, 4 mg/ml protein) was incubated in vitro with 5 n~ 3-[3Hlmethyl- cholanthrene. The incubation volume was divided into two aliquots, and one of them was irradiated for 2 min. Both samples were precipitated with acetone, and the pellets were redissolved in sample loading buffer. Aliquots (50 pl, -700 pg of protein) were electrophoretically separated

these were digested and analyzed for total radioactivity. Migrated dis- on a 10% polyacrylamide gel. Lanes were sliced into 3-mm slices, and

tance was measured from the top of the separating gel. Arrows denote molecular masses (in kilodaltons) and positions of protein standards.

0 5 1 0 1 5 2 0

TIME FIG. 2. Time course for photoaffinity labeling of cytosolic Ah

receptor from mouse Hepa-lclc9 cells with S-[sHlmethylcholan- threne. Cytosol (12 ml, 4 mg/ml protein) was incubated in vitro with 5 n~ 3-[3Hlmethylcholanthrene. The incubation volume was divided into four aliquots and irradiated for 0, 2, 10, or 20 min. Samples were prepared for electrophoresis as described for Fig. 1 and under “Materi- als and Methods.” The total amount of radioactivity (disintegrations/ minute) in the Ah receptor peak was plotted against time.

tained 7.5% polyacrylamide to resolve proteins in the higher molecular mass range (Fig. 4). In further experiments, the polyacrylamide composition was found to be optimal at 6%, in which case the Ah receptor migrated half the length of the gel (Fig. 51, and the 165-kDa protein was at least 4 cm away from the start of the separating gel (Fig. 6). Initial experiments (Fig. 1) using 10% acrylamide gels were aimed at allowing detection of both the Ah receptor and 4 S CBP; the latter has been reported to have molecular masses in the range of 30-40 kDa. However, no radiolabeled protein was detected in that molecular mass range in any of our experiments with [3HlMC (Figs. 1, 4, and 6). The background radioactivity in our experiments was low ( ~ 1 5 0 dpm), and no major peaks due to nonspecific labeling or receptor proteolysis were detected. Unincorporated radiolabeled ligand migrated with the dye front.

Enrichment by Glycerol Density Gradient Centrifugation- The cytosolic receptor from a human colon carcinoma cell line, LS180, failed to show photoaffinity labeling with l3H1MC despite the ability of this receptor to be photolabeled with L3H1TCDD (data not shown). To enhance detection of photola-

12132 Photolabeling of the Ah Receptor

Sucrose Gradient Profile E E.

0 F 0

a

12000

2 8000 U

W

0

I

a 4000

5 0, 0

0 5 10 15 20 25 FRACTION NUMBER

FIG. 3. Sucrose density gradient profiles showing photolysis of Ah receptor from mouse Hepa-lclc9 cells after 2-min exposure to UV light. Cytosol (2 ml, 3 mg/ml protein) was incubated in vitro with 10 nM 3-[3Hlmethylcholanthrene. The incubation volume was divided into two aliquots, and one of them was irradiated for 2 min. Each of the aliquots was analyzed by velocity sedimentation on sucrose gradients. [14ClFormaldehyde-labeled bovine serum albumin (BSA; 4.4 S) and [‘4Clformaldehyde-labeled catalase (CAT; 11.3 S ) were added to the gradients as internal sedimentation markers (arrows).

200 116 9 7 66 43 - E, v v v v v E 200-

0

0 2 4 6 6 10 12 14

DISTANCE (cm)

FIG. 4. Specificity of photoaffinity labeling of cytosolic Ah receptor from mouse Hepa-lclc9 cells with 3-[3H]methylcholanthrene. Cytosol (1.5 ml, 3 mg/ml protein) was incubated in vitro with 10 nM 3-[3Hlmethylcholanthrene in the absence of competitor or in the presence of a 100-fold higher concentration of TCDF or nonradioactive 3-methylcholanthrene. Samples were irradiated for 2 min and precipitated with acetone, and the pellets were redissolved in sample loading buffer. Aliquots (200 pl, 1.5 mg of protein) were electrophoretically

nitrocellulose paper, and each lane was sliced and analyzed for total separated on a 7.5% polyacrylamide gel. Proteins were transferred to

radioactivity as described under “Materials and Methods.” Migrated distance was measured from the top of the separating gel. Arrows denote molecular masses (in kilodaltons) and positions of protein standards. Peaks at -13 cm coincide with the dye front.

beled proteins, we used preparative glycerol gradients to enrich the 9 S fraction of human cytosol; this achieved a 3-5-fold enrichment of the Ah receptor. After enrichment, a [3HlMC- photolabeled protein of -115 kDa was detected upon analysis by SDS-PAGE (Fig. 7).

Photoaffinity labeling of enriched mouse Hepa-lclc9 cytosol was also performed and achieved a 10-fold increase in the radioactive peak for the rodent Ah receptor (Fig. 6, upper panel) compared with the nonenriched samples (Fig. 1). The fractions of the glycerol gradient corresponding to 4 S CBP were also pooled and irradiated in the presence of [3HlMC. Despite the -2-fold enrichment of the 4 S protein, no photoaffinity labeling by [3H]MC was detected (Fig. 6, lower panel 1. Cytosolic preparations of Hepa-1 cytosol containing the “transformed” 6 S receptor were separated from the 9 S receptor complex on preparative glycerol gradients. Photolabeled products of -95 and

6000 - 200 V

116 9 7 6 6 V V v P

40001 1 1-j

20001 il

0 3 6 9 12 15

DISTANCE (cm) FIG. 5. Human LS180 cells: photoaffinity labeling of cytosolic

Ah receptor with [’HITCDD and 3-[3H]methylcholanthrene using enriched cytosolic preparations. Cytosol (1.5 ml, -10 mg/ml protein) was incubated in vitro with 30 nM 3-[”Hlmethylcholanthrene or 30 nM c3H1TCDD. Samples were loaded onto 13-ml glycerol gradients and centrifuged at 372,000 x g,, for 2 h. The fractions corresponding to the 9 S receptor peak were collected, irradiated for 2 min, and precipitated with acetone, and the pellets were redissolved in sample loading buffer. Aliquots (200 pl, -1 mg of protein) were electrophoretically separated on a 6% polyacrylamide gel. Proteins were transferred to nitrocellulose paper, and each lane was sliced and analyzed for total radioactivity as described under “Materials and Methods.” Migrated distance was measured from the top of the separating gel. Arrows denote molecular masses (in kilodaltons) and positions of protein standards. Peaks at - 13 cm coincide with the dye front.

-165 kDa were obtained from the photolabeled 9 S gradient fractions, but no photolabeled products were detectable in samples from the 6 S region of the gradient, and ALBS was not detected in any region of the gel when the 6 S fraction was analyzed by Western blotting (data not shown).

Immunochemical Analyses-In an attempt to identify the components of the -165-kDa band, we photolabeled Hepa- lclc9 cytosol with [3HlMC, separated the proteins by SDS- PAGE, and performed Western blotting with antibodies to several candidate proteins. In the lane probed with an antiserum to ALBS (Fig. 8), two bands (-95 and -165 kDa) co-migrated with the radioactive peaks from photolabeled cytosol observed by gel slicing, confirming that the ligand-binding subunit is a component of the -165-kDa band. Lanes probed with antibodies to hsp90, hsp70, p59, and ARNT proteins exhibited single bands at positions corresponding to the expected molecular masses of the monomeric forms of each of these candidate proteins; however, none of these antibodies produced a band in the 160-170-kDa region (Fig. 8), indicating that although hsp90, hsp70, p59, and ARNT proteins are present in Hepa-lclc9 cytosol, none of these proteins is cross-linked to the ligand-binding subunit during photolabeling with I3H1MC. Cytosol that had been photolabeled with [3H]MC was immunoprecipitated with anti-ALBS antibody; both a -95-kDa band and a -165- kDa band were detected by Western blotting with the precipitate, but no bands were detected in the supernatant fraction remaining after immunoprecipitation (Fig. 9, lanes 13).

ARNT protein is the candidate protein most likely to be cross-linked to ALBS. It is possible that the direct Western

Photolabeling of the Ah Receptor 12133

1 2 0 0 1 1 6 9 7 * * * 2000 {

n 2 0 0 1 1 6 9 7 6 6

450 * * * * 300 1

I o : ' , , , . , . , . ,

0 3 6 9 1 2 1 5

DISTANCE (cm) FIG. 6. Mouse Hepa-lclcS cells: photoaffinity labeling of cyto-

solic Ah receptor and absence of labeling of 4 S protein with 3-['Hlmethylcholanthrene using enriched cytosolic preparations. Cytosol (1.5 ml, -10 mg/ml protein) was incubated in ctitro with 30 nM 3-I"HImethylcholanthrene. The sample was loaded onto 13-mI glycerol gradients and centrifuged a t 372,000 x g , , for 2 h. The fractions corresponding to the 4 S protein and the 9 S receptor peak were collected, irradiated for 2 min, and precipitated with acetone, and the pellets were redissolved in sample loading buffer. Aliquots (200 pl, -1 mg of protein) were electrophoretically separated on a 6% polyacrylamide gel. Proteins were transferred to nitrocellulose paper, and each lane was sliced and analyzed for total radioactivity as described under "Ma- terials and Methods." Migrated distance was measured from the top of the separating gel. Arrows denote molecular masses (in kilodaltons) and positions of protein standards. Peaks a t -13 cm coincide with the dye front.

blots of photolabeled cytosol might not be sensitive enough to detect ARNT if it were cross-linked to ALBS in a 1:l stoichi- ometry. Therefore, we also performed immunoprecipitation experiments to test rigorously for the possible presence of ARNT in the -165-kDa band. Cytosol photolabeled with ['HIMC was immunoprecipitated using either anti-ALBS or anti-ARNT antiserum. Supernatant and pelleted fractions resulting from the immunoprecipitation with each antiserum were analyzed by Western blotting with anti-ALBS and anti-ARNT antisera (Fig. 9). After immunoprecipitation with anti-ARNT and staining with anti-ALBS, no ALBS band was detectable in proteins from the resolubilized pellet (Fig. 9, lune 2 ) either near 95 kDa or in the 165-kDa region. In raw cytosol or in the supernatant remaining after immunoprecipitation with anti-ARNT, anti- ALBS antiserum did detect ALBS both a t -95 kDa and in the 165-kDa region (Fig. 9, lanes 4 and 6 ) , indicating that although ALBS was present in cytosol, it was not coprecipitated with ARNT when the precipitation was done with anti-ARNT antiserum. In the reciprocal experiment (in which immunoprecipitation was done with anti-ALBS antibody), staining with anti- ARNT did not detect ARNT in the immunoprecipitated pellet in either the 86- or 165-kDa region; monomeric ARNT protein was detected with anti-ARNT serum a t -86 kDa in raw cytosol or in the supernatant remaining after immunoprecipitation with anti-ALBS. The results of immunoprecipitation with both anti- ARNT and anti-ALBS antisera further corroborate the evidence that ARNT is not part of the 165-kDa complex.

Analysis by Fluorography-Analysis by fluorography con-

2 0 0 "

1 0 0 "

69-0

I I

LS180 HEPA-1 CYTOSOL CYTOSOL

FIG. 7. Photoaffinity labeling of cytosolic Ah receptor from mouse Hepa-lclcS and human LS180 cells with 3-['H]methylcholanthrene and ['HITCDD using enriched cytosolic preparations: analysis by fluorography. Cytosol (1.5 ml, - 10 mg/ml protein) was incubated in oitro with 30 nv 3-I.'Hlmethylcholanthrene or 30 ny 1:'HITCDD. Samples were loaded onto 13-ml glycerol gradients and centrifuged a t 372,000 x gz,v for 2 h. The fractions corresponding to the 9 S receptor peak were collected, irradiated for 2 min, and precipitated

Aliquots (200 1.11, -1 mg of protein) were electrophoretically separated with acetone, and the pellets were redissolved in sample loading buffer.

on a 6% polyacrylamide gel, which was subsequently stained, dried, and analyzed by fluorography as described under "Materials and Methods." Al l lanes had equivalent amounts of protein as judged by Coomassie Blue staining of the gel. The positions of l"C-labeled protein markers are shown by the arrows; molecular masses for the markers are given in kilodaltons

c

-66- a

If 4

tosols photolabeled with 3-[3Hlmethylcholanthrene. Cytosol (0.5 FIG. 8. Immunochemical analyses of mouse Hepa-lclcS cell cy-

ml, 3 mg/ml protein) was photolabeled as described for Fig. 4. Dilutions

als and Methods." The line plot on the left was obtained by slicing and for the primary and secondary antibodies are described under "Materi-

counting nitrocellulose paper to which the ['HIMC-photolabeled cytosol, run on SDS-PAGE, had been transferred. Molecular mass markers on the sliced blot and on the Western blots were aligned for direct com-

ALBS is represented by the most intense stain a t -95 kDa and the stain parison. In the lane stained with anti-ALBS antiserum (far right),

at -165 kDa; the 165-kDa band is present only in cytosols that have been photolabeled (data not shown). The identity of the band at a mass slightly higher than that of 95-kDa ALBS is not known.

firmed the results obtained using the slicing technique. The band detected after photoaffinity labeling of the human Ah receptor by L3HIMC was very faint in the fluorogram (Fig. 71, although a peak was detected by slicing the nitrocellulose pa-

12134 1 2 3

160 kDa- complex

LBS 4-0

Photolabeling of the Ah Receptor 4 5 6 7 0 9

0 + ARNT

I l l C P S

Anti-LBS; ppt with

Anl i -LBS s l a i n

I I I I I I C P S C P S

Anll -ARNT; Anl l -LBS; ppt with ppl with

Anl l -LBS Anti -ARNT s l a i n s l a i n

mouse Hepa-lclc9 cells photolabeled with 3-['H]methylcholan- FIG. 9. Immunoprecipitation experiments using cytosol from

threne. Cytosol (0.5 ml, 3 mg/ml protein) was photolabeled as described for Fig. 4. The immunoprecipitation protocol is described under "Mate- rials and Methods." Control lanes ( C ) represent cytosols that were processed in a similar manner to those of the immunoprecipitated samples, with the exception that no antisera were added. After immunoprecipitation, the pelleted ( P ) and supernatant ( S ) products were separated by SDS-PAGE, and Western blotting was performed with anti-ARNT and anti-ALRS antisera as indicated. The photograph for lanes 7-9 (anti-ARNT stain) was sliced to align the ARNT bands with molecular mass markers and with lanes 1-6; no bands were visible with anti-ARNT staining in the 160-170-kDa region in the original film or in the photographic print.

per (Fig. 5, bottom panel ). The fluorogram also shows that the human Ah receptor was efficiently photolabeled with ["HITCDD. An equivalent molecular mass of -115 kDa was observed for the human Ah receptor when it was photolabeled with either ["HIMC or ["HITCDD. The mouse Hepa-lclc9 Ah receptor was efficiently photolabeled with both ["HIMC and ["HITCDD, although a stronger band can be observed using ["H ITCDD. However, only ["H IMC showed photolabeling of the higher molecular mass protein (-165 kDa) (Fig. 7), which also had been detected using the slicing technique. The mass of the mouse Ah receptor was -95 kDa using migration values obtained with either ligand.

Yield of Photocovalent Adducts-The yield of "H-labeled ligand-Ah receptor covalent adducts formed upon exposure to UV light was calculated using the radioactivity from the sliced nitrocellulose paper to which the proteins had been transferred. The radioactivity (disintegrations/minute) in the receptor peak was summed, and after subtracting the background radioactivity, it was expressed as a percentage of total receptor applied to the gel. The total Ah receptor (100%) was estimated for each sample in a separate analysis by sucrose density gradient assay. Using this estimation, the yield of photoadduct formation with ['HIMC (0.86 2 O.ll%, n = 9) was consistently lower than that with ["HITCDD (3.5%3, n = 2) in mouse Hepa- lc lc9 cells. In human LS180 cells, the same pattern was ob-

served, i.e. the yield was 0.2% for ['HIMC uersus 5.48 t 0.26% ( n = 4) for ["H ITCDD.

DISCUSSION

This is the first report of photoaffinity labeling of the Ah receptor with a nonhalogenated ligand, ["HIMC. ["HJMC photolabeled the cytosolic Ah receptor from both mice (Hepa-lclc9 cell line) and humans (LS180 cell line). The photolabeling was specific, as evidenced by the competition experiments using TCDF or nonradiolabeled MC. The relative molecular mass for the mouse Hepa-lclc9 Ah receptor of -95 kDa was in agreement with determinations that used ['HITCDD as photoligand (Landers et al. 1989; Prokipcak and Okey, 19901, suggesting that the receptor proteins photolabeled by ["HIMC and [.''H]TCDD are the same. The efficiency of photolabeling was consistently lower with ["HJMC than with ["HlTCDD in both mouse and human cell lines, even though the binding affinities of these two ligands for the receptor are approximately the same (only %fold higher IC,, for MC; Riddick et al., 1994). The reason for the lower photolabeling efficiency with ["HIMC is unknown, as is the mechanism by which ["HIMC is photoactivated; ["HITCDD may be more photoreactive than ["]TCDD. The mechanism of photoactivation of TCDD has been postulated to involve dechlorination, with the remaining free radical attacking certain amino acid residues. The 13HIMC molecule does not have obvious photoactivable groups, and the mechanism for the labeling is unclear. The time required for photoaffinity labeling is on the order of 1-2 min with ["HIMC; longer exposure does not increase the amount of photoadducts formed. The Ah receptor protein showed high photolability as seen by the 70% loss of binding after exposure to UV light for 2 min. This could explain why there is no increase in photoadduct formation with increased exposure to UV light since the Ah receptor molecules would no longer be capable of noncovalently binding the ligand.

Photoafinity labeling of the mouse Ah receptor with ['HIMC was more efficient than with the human receptor, despite their equal binding affinities for ['HIMC and ["HITCDD and despite equal concentrations of both receptors when analyzed by sucrose gradient assays (Harper et al., 1991). The mass of the human Ah receptor (-115 kDa) photolabeled with ['HIMC agrees with values reported when using ["]TCDD as photoligand (Harper et al., 1991). ['HITCDD photoaffinity labeling of another human cell line, the squamous cell carcinoma line A431, also confirms the same molecular mass for the human Ah receptor.3

["HIMC did not photolabel 4 S CBP, which previously has been shown to bind [:3H]MC reversibly by sucrose gradient analysis. This is interesting since an MC derivative, 1-oxo-3- methylcholanthrene, has been shown to photoaffinity label 4 S CBP (Arnold et al., 1987). One possibility is that this derivative has a higher affinity for the 4 S protein than MC itself, and this hypothesis could be assessed by comparing the kinetics of reversible binding of these two ligands. 1-Azidopyrene also has been shown to photolabel 4 S CBP (Collins and Marletta, 1986), but not the Ah receptor, and this also could be due to a selec- tively higher affinity of this ligand for 4 S CBP, compared to its affinity for the Ah receptor.

A 70-kDa proteolytic fragment of the Hepa-lclc9 Ah receptor has been previously reported when using an iodinated derivative of TCDD to photolabel the receptor (Poland and Glover, 1988). Despite the high degree of photoaffinity labeling b-5000 dpm) in our experiments, no breakdown products of the Ah receptor were observed. The pattern of photolabeling with both ['HIMC and ["]TCDD was very clean, and no extra major

' S. M. F. de Morais, unpublished observations.

Photolabeling of the Ah Receptor 12135

radiolabeled bands or peaks due either to nonspecific labeling or to proteolytic degradation were observed in either rodent or human cytosols by gel slicing or fluorography.

The -165-kDa band observed in photolabeled samples prob- ably was produced by cross-linking of the ligand-binding subunit of the Ah receptor to another protein (or proteins). ALBS was clearly detected as a component of the - 165-kDa complex by Western blotting using polyclonal antiserum to ALBS. Ad- ditional immunoprecipitation experiments using anti-ALBS antiserum provided further evidence that ALBS is present both in the 95- and 165-kDa bands since both bands were detected in the precipitate from immunoprecipitation with anti-ALBS and no bands were observed in the supernatant from the anti-ALBS immunoprecipitation. The strategy we used to attempt to identify initially the cross-linked partner primarily was to test, immunochemically, for several proteins that have been shown or have been postulated to be associated with the Ah receptor or receptors for steroid hormones. These include ( a ) heat shock proteins such as hsp90, which has been shown to be associated with the cytosolic ALBS, (Perdew, 1988; Denis et al., 1988) and hsp70, a protein with a mass that would, when combined with the 95-kDa ALBS, produce a complex with a mass of 165 kDa; ( b ) p59 protein, which is present in several steroid hormone receptor complexes (Tai et al. 1986, 1992), but which has not been detectable in Ah receptor complexes (Prokipcak et al., 1989); and ( c ) ARNT protein, a "nuclear translocation" factor that has been shown to associate with the nuclear Ah receptor (Hoffman et al., 1991; Reyes et al., 1992). None of these candidate proteins could be detected immunochemically in the 165- kDa complex, even after optimization by immunoprecipitation. I t cannot be completely ruled out that, even though the immunochemical assays are very sensitive, they still might not be suffciently sensitive to identify one of these candidate proteins.

Recently, several multimeric forms of the Ah receptor were detected by photolabeling with an iodinated derivative of TCDD and chemical cross-linking (Perdew, 1992). None of the multimeric forms detected by Perdew was in the 160-170-kDa range, in which we detected the complex after photolabeling with [SH]MC. Interestingly, the author postulated the presence of a -50-kDa protein in a 327-kDa species produced by cross- linking; this could be the component cross-linked to ALBS in our studies, but the combined mass ofALBS (-95 kDa) plus the 50-kDa component would be smaller than that of the complex detected in our studies (164 5 6 kDa). It might be possible that the - 165-kDa band we detected after photolabeling represents a homodimer of ligand-binding subunits; in this case, the pre- dicted mass ( - 180 kDa) would be significantly higher than that of the -165-kDa complex we detected.

Immunoprecipitation experiments were performed to increase further the likelihood of detecting ARNT protein, the most likely candidate to be a component of the - 165-kDa complex. Since no band was detected in the anti-ARNT immuno- precipitate stained with anti-ALBS and since the 165-kDa complex remained in the supernatant after immunoprecipitation with anti-ARNT, it is highly unlikely that ARNT is a component of the -165-kDa complex. Whitelaw et al. (1993) did report coprecipitation of ARNT with ALBS in the presence of the ligand TCDD; however, in their study, ARNT protein was added back in vitro t o cytosol prepared from a mutant Hepa-1 cell line that does not express ARNT, and their study may not be com- parable with our experiments, in which ARNT was constitu- tively expressed in Hepa-lclc9 cells.

We used another approach to try to determine if ARNT was part of the 165-kDa protein complex by photolabeling fractions from glycerol gradients containing the enriched transformed receptor. Transformation was induced by incubating cytosol

with [3H]MC at 30 "C for 4 h rather than at 4 "C for 1 h, as had been done in most experiments. Since ARNT has been shown to be associated with the transformed nuclear Ah receptor, it would be expected that the 30 "C 4-h incubation would provide a higher proportion of photolabeled 165-kDa protein complex than was obtained at 4 "C. However, no photolabeled proteins were detected in the 6 S fractions, indicating either that ARNT is not associated with the ligand-binding subunit in cytosol or that not enough of the Ah receptor was in the transformed state to be detected. Our evidence that ARNT is not part of the cytosolic Ah receptor complex is in agreement with a recent publication by Probst et al. (1993), who demonstrated by immunoprecipitation techniques that ARNT protein is not associated with the Ah receptor in cytosol from Hepa-1 cells.

There was a faint band at the same relative molecular mass (165 kDa) when T3H1TCDD was used as a photoligand, suggesting that the cross-linking also occurs with TCDD, albeit to a much lesser extent. Since the dimensions for these two molecules are different, with MC having a larger size (1.4 x 0.7 nm) than TCDD (0.3 x 1.0 nm), the MC molecule possibly could reach further amino acid residues of a neighboring subunit and facilitate cross-linking.

In summary, r3H]MC can be used as a photoaffnity ligand for both rodent and human cytosolic Ah receptors, although enriched preparations are required for detection of the photolabeled human receptor. This technique has the potential to allow characterization of the differences between the binding sites for MC and TCDD and between the human and rodent receptors using purified preparations of the Ah receptor. Photoaffinity labeling with L3H]MC could be used therefore to improve our understanding of the mechanisms that underlie the biological effects mediated by the Ah receptor, principally the differences in the mechanisms involving nonhalogenated and halogenated ligands. Further studies will be required to identify the protein or proteins that become cross-linked to the ligand-binding subunit and to discover the functional importance of the non-ALBS partner in the -165-kDa complex. Analysis of mutants of Hepa-1 cells indicates that at least one more protein in addition to ALBS and ARNT proteins is required for activity of the receptor complex (Hankinson, 19931, and it is possible that this is the unknown protein in the 165-kDa cross-linked complex.

Acknowledgments-We are grateful to Dr. Oliver Hankinson for sup- plying anti-ARNT serum and to Dr. Lee Faber for providing the JP-1 polyclonal serum to p59 protein. We thank Dr. David S. Riddick and Dr. Patricia A. Harper (University of Toronto) for helpful discussions.

REFERENCES

Arnold, P. S., Garner, R. C., and Tierney, B. (1987) Biochem. J . 242, 375-381 Bayley, H., and Knowles, J. R. (1977) Methods Enzymol. 46, 69-114 Bradfield, C. A., Glover, E., and Poland, A. (19911 Mol. Pharmacol. 39, 13-19 Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 Burhach, K. M., Poland, A., and Bradfield, C. A. (1992) Proc. Natl. Acad. Sci.

Collins, S., and Marletta, M. A. (1986) Biochemistry 25,4322-4329 Denis, M.. Cuthill, S., Wikstrom,A.-C., Poellinger, L.. and Gustafsson, J:A. (19881

Bimhem. Biophys. Res. Commun. 155,801-807 Denison, M. S., Wilkinson, C . F., and Okey, A. B. 11986) Chemosphere 15, 1665-

1672 Ema. M., Sogawa, K., Watanahe, N., Chujoh,Y., Matsushita, N., Gotoh, O., Funae.

Y., and Fujii-Kuriyama, Y. (1992) Biochem. Biophys. Res. Commun. 184, 246- 253

U. S. A. 89, 8185-8189

Hankinson, 0. 119931 Arch. Biochem. Blophys. 300, 1-5 Harper, P. A., Golas, C. L., and Okey. A. B. (19881 Cancer Res. 48, 2388-2395 Harper, P. A., Prokipcak, R. D., Bush, L. E., Golas, C. L., and Okey, A. B. (19911

Hoffman, E. C . , Reyes, H., Chu. F.-F., Sander, F., Conley, L. H., Brooks, B. A., and Arch. Biochem. Biophys. 290, 27-36

Horowitz, K. B., and Francis, M. D. (19881 in Affinity Labelling and Cloning of Hankinson, 0. (1991) Science 252, 954-958

Steroid Hormone Receptors IGronemeyer, H., ed) pp. 186-198, Ellis Horwood Ltd., Chichester, England

Laemmli, U. K. (1970) Nature 227, 680-685 Landers, J. P., and Bunce, N. J. (1991) Biochem. J. 276,273-287 Landers, J. P., Piskorska-Pliszczynska, J., Zacharewski, T., Bunce, N. J., and Safe, S. 119891 J. Bid . Chem. 264, 18463-18471

12136 Photolabeling of the Ah Receptor Lesca, P., Fernandez, N., and Roy, M. (1987) J. Biol. Chem. 262,48274835

Manchester, D. K.. Gordon, S. K., Golas, C. L., Roberts, E. A.. and Okey, A. B. Lorenzen, A,, and Okey, A. B. (1991) Toxicol. Appl. Pharmacol. 107, 203-214

Nebert, D. W., and Gonzalez, F. J. (1987)Annu. Rev. Biochem. 56,945-993 Okey, A. B. (1990) Pharmacol. & Thez 45,241-298 Okey, A. B., and Vella, L. M. (1982) Euz J. Biochem. 127 ,3947 Okey, A. B., Dube, A. W., and Vella, L. M. (1984) Cancer Res. 44, 14261432 Okey, A. B., Denison, M. S., Harper, P. A., and Prokipcak, R. D. (1988) in Micro-

somes and Drug Oxidation (Miners, J. O., Birkett, D. J., Drew, R., and McMa- nus, M., eds) pp. 31-38, Taylor & Francis Ltd., London

(1987) Cancer Res. 47,48614868

Perdew, G. H. (1988) J. Biol. Chem. 263, 13802-13805 Perdew, G. H. (1992) Biochem. Biophys. Res. Commun. 182, 55-62 Poland, A., and Glover, E. (1974) Mol. Pharmacol. 10,349-359 Poland, A., and Glover, E. (1975) Mol. Pharmacol. 11, 389-398 Poland, A., and Glover, E. (1987) Biochem. Biophys. Res. Commun. 146, 1439-

Poland, A,, and Glover, E. (1988)Arch. Biochem. Biophys. 261, 103-111 Poland, A.. and Glover, E. (1990) Mol. Pharmacol. 38,30&312 Poland, A., Glover, E., and Kende, A. S. (1976) J. Biol. Chem. 251,493&4946 Poland, A,, Glover, E., Ebetino, F., and Kende, A. S. (1986) J. Biol. Chem. 261,

Poland, A,, Glover, E., and Bradfield, C. A. (1991) Mol. Pharmacol. 39, 20-36

1449

6352-6365

Probst, M. R., Reisz-Porszasz, S . , Agbunag, R. V., Ong, M. S., and Hankinson, 0.

Prokipcak, R. D., and Okey, A. B. (1990) Biochem. Biophys. Res. Commun. 172,

Prokipcak, R. D., Faber, L. E., and Okey, A. B. (1989)Arch. Biochem. Biophys. 274,

Reyes, H., Reisz-Porszasz, S., and Hankinson, 0. (1992) Science 266, 1193-1195 Riddick, D. S., Huang, Y., Harper, P. A,, and Okey, A. B. (1994) J. Biol. Chem. 269,

Roberts, E. A,, Golas, C. L., and Okey, A. B. (1986) Cancer Res. 46,37394743 Ruobo. A. E.. Kiefer, H.. Roeder. P. E., and Singer. S. J. (1973) Proc. Natl. Acad. Sci.

(1993) Mol. Pharrnacol. 44,511-518

698-704

648-658

12118-12128

u. s. A. io, 2567-2571 Tai, P.-K. K., Maeda, Y., Nakao, K., Wakirn, N. G., Duhring, J. L., and Faber, L. E.

Tai, P.-K. K., Albers, M. W., Chang, H., Faber, L. E., and Schreiber, S. L. (1992) (1986) Biochemistv 26,5269-5275

Waithe, W. L., Michaud, M., Harper, P. A., Okey, A. B., and Anderson, A. (1991)

Wang, X., Narasimhan, T. R., Morrison, V, and Safe, S. (1991) Arch. Biochem.

Whitelaw, M., Pongratz, I., Wilhelmsson, A., Gustafsson, J.-A,, and Poellinger, L.

Whitlock, J. P. (1990)Annu. Rev. Pharmacol. Toricol. 30, 251-277

Science 256, 1315-1318

Biochem. Pharrnacol. 41,85-92

Biophys. 287, 186194

(1993) Mol. Cell. Biol. 13, 2504-2514

no. vol. pp. chemistry the of journal in u.s.a. · pdf filethe best characterized biological...

Documents