Kinetic models for association of 2,3,7,8-tetrachlorodibenzo-p-dioxin with the Ah receptor

ARCHIVES Vol.

267,

OF BIOCHEMISTRY No.

1,

November

AND

BIOPHYSICS

15, pp. 384-397,1988

Kinetic Models for Association of 2,3,7,8-Tetrachlorodibenzo-p-dioxin with the Ah Receptor’ NIGEL *Guelph-

J. BUNCE,*,‘JAMES

P. LANDERS,*

Waterloo Centre for Graduate Work in Chemistry, 2 WI; and fDepartment of Vetem’nary Physiology College Stat&, Received

AND

S. H. SAFE?

University of Guelph, Guelph,, Ontario, Canada and Pharmacology, Texas A&M University, Texas 778.@

N1G

June 8,1988

Saturation binding studies of the interaction between 2,3,7,8-tetrachlorodibenzo-pdioxin (TCDD) and the Ah receptor obtained from the hepatic cytosol of Wistar rats have been carried out. The conventional Scatchard analysis for determination of the equilibrium constant for ligand-receptor binding has been shown to be inappropriate due to thermal inactivation of the unoccupied receptor. Simulation models of the receptor-ligand binding kinetics which take into account receptor degradation have been developed and the results are consistent with two alternative kinetic models. In Model 1, reversible 2,3,7,8-TCDD-receptor binding occurs in parallel with inactivation of the unbound receptor; analysis of the observed data using this model suggests that the previously determined equilibrium constants (K,,,) for association of the ligand with the receptor are orders of magnitude too low and the total initial receptor concentrations are somewhat underestimated. In Model 2, the unbound receptor is converted unimolecularly to an activated state which then undergoes competitive degradation or entrapment by ligand. Experiments have been carried out over the temperature range 4-37”C, enabling activation parameters to be obtained. According to Scheme 1, the activation enthalpies for association of receptor with ligand and for thermal inactivation of the unoccupied receptor are high, and numerically almost identical (AH* ca 125 kJ mall’). These reactions are strongly entropically driven and this is consistent with association being accompanied by a conformational change in the receptor protein, and the previously postulated binding of the ligand to a hydrophobic pocket. According to Scheme 2, there is only one enthalpy of activation because both inactivation and entrapment by 2,3,7,8TCDD are fast processes which follow the same slow activation step. On the basis of this latter model, a lo-’ M concentration of 2,3,7,8-TCDD is sufficient to trap roughly twothirds of the activated receptors. o 1988 Academic press, IX

diPolychlorinated dibenzo-p-dioxins, benzofurans, and biphenyls are members of a chemical family (polyhalogenated aromatics) which elicit a number of common

toxic and biological responses (l-6). The toxic responses observed in several animal species include dermal toxicity, reproductive problems, body weight loss, teratogenicity, hepatotoxicity, gastric lesions, lymphoid involution, immunotoxicity, and carcinogenicity. The complete spectrum of toxicity is not observed in any single anima1 species; however, body weight loss, lymphoid involution and/or immunotoxicity are the most frequent responses ob-

1 This work was supported in part by grants from the Natural Sciences and Engineering Research Council of Canada, the National Institutes of Health (ES03554), and the Environmental Protection Agency. a To whom correspondence should be addressed.

0003-9861/88 Copyright All rights

$3.00

0 1988 by Academic Press, Inc. of reproduction in any form reserved.

384

Ah

RECEPTOR-2,3,7,8-TETRACHLORODIBENZO-p-DIOXIN

served in laboratory animals. Polyhalogenated aromatics also induce several enzyme systems in mammals and mammalian cells in culture including microsomal benzo[a]pyrene hydroxylase (aryl hydrocarbon hydroxylase, AHH)3 and associated cytochrome P-450 isozymes (7-12). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) is the most potent known inducer of this enzyme system; because of its extremely high toxicity, it has been used extensively as a prototype for investigating the mechanism of action of these toxins. [“HI-2,3,7,8-TCDD of high specific activity (52.5 Ci/mmol) was initially utilized to identify a specific binding protein present in the hepatic cytosol of responsive C57BL/6 mice (13) and the presence of this protein in hepatic and extrahepatic tissues of diverse species and mammalian cells in culture has been reported (14-24). It has been proposed that like the steroid hormones, 2,3,7,8-TCDD and related compounds elicit their responses via initial binding to this cytosolic aryl hydrocarbon (Ah) receptor protein present in target tissue (1, 3). Support for the role of the Ah receptor in the mechanism of action of halogenated aromatics is derived from several lines of evidence which include (i) saturable binding of [3H]-2,3,7,8TCDD and other aryl hydrocarbons to the receptor protein which is present in low concentrations in target tissues (cl00 fmol/mg cytosolic protein) (1, 17-20,

25,26); (ii) high affinity TCDD)-Ah receptor 3.0 nM) (13,21,27,28);

ligand (i.e., 2,3,7,8interactions (K. 0.13-

’ Abbreviations used: AHH, aryl hydrocarbon hydroxylase; TCDD, tetrachlorodibenzo-p-dioxin; Ah, aryl hydrocarbon; TCDF, 2,3,7,8 - tetrachlorodibenzofuran; EDTA, ethylenediaminetetraacetic acid; DMSO, dimethyl sulfoxide; PPO, 2,5-diphenyloxazole; POPOP, 1,4-bis-2-(2-(5-phenyloxazolyl))-benzene; Mops, 3-(N-morpholino)propanesulfonic acid; DTE, dithioerythritol; DTT, dithiothreitol; MEGD, buffer containing Mops, EDTA, glycerol, and DTE; HAP, hydroxylapatite; Hepes, 4-(2-hydroxyethyl)-lpiperazineethanesulfonic acid; HEGD, buffer containing Hepes, EDTA, glycerol, and DTE.

INTERACTIONS

(iii) stereoselective ligand-Ah interactions (l-5,29-32);

385 receptor

(iv) response intensity to 2,3,7,8-TCDD and related compounds which is in part related to cellular receptor levels and sustained nuclear receptor occupancy (l-3,

33-35). Several studies in genetically inbred mice and mammalian cells in culture (1,3, 6,34, 35) have clearly illustrated the role of the bound Ah receptor-2,3,7,8-TCDD complex in the mechanism of cytochrome PI-450 induction; however, the mechanisms of toxic actions of the class of compounds are not well understood. The molecular properties of the crude and partially purified Ah receptor and its interaction with 2,3,7,8-TCDD have been investigated by several groups (13, 17, 18, 23, 36-42). Like many of the steroid hormone receptors, the Ah receptor complex is possibly a protein oligomer which contains a hydrophobic binding site and interacts with DNA and polyanions. A recent study has shown that the bound 9 S 2,3,7,8TCDD-receptor protein complex from rat and mouse hepatic cytosol disaggregates in high salt (0.4 M KCl) to give a 4 S complex; however, the rate of the 4-9 S conversion was more rapid for the rat than the mouse receptor complex (22). Kester and Gasiewicz (36) have found that the unbound rat hepatic cytosolic receptor is unstable and this can result in an underestimation of cellular receptor concentrations. However, the conventional Scatchard analysis of 2,3,7,8-TCDD receptor binding isotherms assumes that for a simple bimolecular interaction of receptor (R) plus ligand (L) to form a receptor-ligand (RL) complex (Eq. [l]), there is no significant degradation of R, L, or RL. R+L=RL Beck and Goren (43) have shown that Scatchard plot analyses of bimolecular equilibrium reactions in which any of the components is subject to degradation yield incorrect equilibrium constants and they have proposed kinetic simulation models to reevaluate such types of data. In our

386

BUNCE,

LANDERS.

case, where it is the unliganded receptor that is thermally labile, such a kinetic model is shown as Scheme 1:

AND

SAFE

specific binding components added the day of use.

Preparation R+L$RL

r

R 4 Inactivation

PI

The major objective of this in vitro study is focused on developing appropriate simulation models to further understand the TCDD-receptor binding process. One could argue that such in vitro investigations should utilize purified Ah receptor; however, presently this is not possible. Furthermore, the pure protein may not reflect the situation in vivo, as previously noted with steroid hormone receptors. In this paper we investigate the kinetics of binding of 2,3,7,8-TCDD to the receptor at pH 7.4 over a temperature range which includes 37%, the operating temperature for the protein in vivo. MATERIALS

AND

METHODS

Materials Glycerol, disodium ethylenediaminetetraacetate (EDTA), toluene (Scintanalyzed), dimethyl sulfoxide (DMSO), 2,5-diphenyloxazole (PPO), 1,4-bis(2-(5-phenyloxazolyl)benzene (POPOP), sucrose, and sodium chloride were purchased from Fisher Scientific Co. (Toronto, Ontario). 3-[N-Morpholinolpropanesulfonic acid (Mops), 4-(2-hydroxyethyl)-I-piperazine ethanesulfonic acid (Hepes), dithioerythritol (DTE), dithiothreitol (DTT), hydroxylapatite, and dextran were purchased from Sigma Chemical Co. (St. Louis, MO). Triton X-100 was purchased from Terochem Ltd. (Edmonton, Canada). Charcoal was purchased from BDH (Toronto, Ontario). Tritiated 2,3,7,8-tetrachlorodibenzo-p-dioxin ([3H]TCDD) (sp act 30 Ci/ mmol) was prepared in our laboratory and purified by HPLC (purity >98%). The synthesis of 2,3,7,8-TCDD and 2,3,7,8-tetrachlorodibenzofuran (TCDF) with a purity of greater than 99% have been described elsewhere (29). Male Wistar rats were obtained from Charles River (Canada).

Buffer The buffer, MEGD, contained 23 mM Mops, 1 mM EDTA, 10% (v/v) glycerol, and 1 IIIM DTE at pH 7.4. For maximum stabilization of the hepatic cytosolic

for

TCDD,

of Hepatic

DTE

was

Cytosol

Male Wistar rats (one month of age; average weight, 100 g) were housed in wire cages and allowed free access to Purina certified rodent chow, No. 5002, and water and maintained on diurnal cycle of 13 h of light/l1 h of darkness. The animals were sacrificed by cervical dislocation and the livers were immediately perfused in situ by the hepatic portal vein with icecold isotonic saline (25 ml supplemented with EDTA (0.1 IIIM)). The excised livers were rinsed once with 15 ml of fresh ice-cold MEGD buffer and finely minced. The minced livers were rinsed again with buffer and homogenized in ice-cold MEGD buffer using a Teflonglass Potter-Elvehjem tissue homogenizer. The homogenate was centrifuged at 10,OOOg for 20 min at 4°C in a Sorvall Superspeed RC2-B centrifuge and the resulting supernatant then centrifuged at 100,000~ for 60 min at 4°C in a Beckman L8-55 ultracentrifuge. Surface lipid was removed by aspiration after each centrifugation. The supernatant was collected by aspiration, the protein concentration was determined immediately (44), and the cytosol was stored in small volumes at -70°C in a Revco freezer until used. Control experiments showed that specific binding of TCDD to the Ah receptor was not diminished by freezing over a period of at least 10 weeks and then thawing; see control experiments, below.

Hydroxylapatite

(HAP)

Assay

The procedure is based on that of Gasiewicz and Neal (26). A mixture containing hepatic cytosol, 1 nM [3H]-2,3,7,8-TCDD and, where appropriate, a 500-fold molar excess of unlabeled 2,3,7,8-TCDF was incubated at 20°C for 2 h. Identical results were obtained when unlabeled 2,3,7,8-TCDD was utilized, however, 2,3,7,8-TCDF was routinely used in the studies because of its greater solubility. Temperature was maintained using a circulating water bath (Haake Model D1 circulator) with accuracy of ?O.l”C. Identical experiments were carried out with unlabeled 2,3,7,8-TCDD in place of unlabeled 2,3,7,8-TCDF and gave the same results (data not shown). After this time, 0.20 ml of this incubation mixture was added to 0.25 ml of a hydroxylapatite slurry (5/12, w/v, in icecold MEGD buffer), mixed, and incubated on ice for 35-40 min with the tubes gently agitated every 10 min. After 40 min, 1.0 ml of MEGD buffer containing 1% Triton X-100 was added; the contents of the tubes were thoroughly mixed and then centrifuged at 2000g for 2 min. The supernatant was poured off and the washing procedure was repeated twice more. One milliliter of 95% ethanol was added to the washed

Ah

RECEPTOR-2,3,7,8-TETRACHLORODIBENZO-p-DIOXIN

pellets and the resulting slurry was transferred quantitatively to a plastic 20-ml scintillation vial, 10 ml of scintillation cocktail (25 mM PPO, 0.32 mM POPOP, 33% (v/v) Triton X-100 in Scintanalyzed toluene) was added, and the radioactivity was determined by liquid scintillation counting (50-60% efficiency) with a Packard scintillation counter 4000 series. Specific binding was determined as the difference between the total binding (mixtures not containing unlabeled 2,3,7&TCDF) and nonspecific binding (mixtures with excess unlabeled 2,3,7&TCDF).

Control 1.

of the Ah Receptor in the Preparation Buffers

MEGD vs HEGD &&rer. Work on the Ah receptor has commonly utilized the Hepes-based buffer system HEGD (pH 7.6) for preparation of hepatic cytosol(1, 6,18,29). A Mops-based buffer system MEGD (pH 7.4) has been used in this study. MEGD is a necessary component in an ion-exchange chromatography technique presently under investigation for partial purification of the Ah receptor. As a result, this buffer was used in the work described here, with the specific aim that the kinetic studies be comparable. Specific binding of [aH]-2,3,7&TCDD (1 nM) to hepatic cytosol(2 mg/ml) in MEGD and HEGD buffers as determined with hydroxylapatite binding assay was not significantly different as shown below:

HEGD MEGD

Specific binding + SD (fmol/mg protein)

(pH 7.6) (pH 7.4)

56.3 +- 3.5 53.0 f 2.4

DTE vs DTT as stabilizer. As reducing agents for stabilization of protein mixtures such as hepatic cytosol, DTT and dithioerythritol DTE are essentially equal in their stabilizing capacity. Although DTE is slightly less soluble, it was chosen for use because it is generally less hygroscopic than DTT and thus is more amenable to use in small quantities and long storage. Specific binding of [3H]-2,3,7,8-TCDD (1 nM) to hepatic cytosol (2 mg/ml) prepared in MEGD buffer in the presence of each of these reducing agents when determined with the hydroxylapatite binding assay is essentially the same as indicated below:

MEG +DTT +DTE

buffer (1 (1

mM) mM)

2. E#ect of the Freeze/Thaw The results tabulated below freeze-thaw procedure has very specific binding of [3H]-2,3,7,8-TCDD cytosol (2 mg/ml) as determined apatite binding assay.

Protocol indicate that the little effect on the (1 nM) to hepatic with the hydroxyl-

Specific binding + SD (fmol/mg protein) Control cytosol Freeze/thaw cytosol

52.5 f 1.6 49.7 + 3.4

Experiments

Stability

Buffer

387

INTERACTIONS

Specific binding + SD (fmol/mg protein) 36.8 + 1.3 37.2 f 1.0

3. Stability

of 2,3,7,&TCDD

in Hepatic

cytosoz To hepatic cytosol (5 ml, 2 mg of protein/ml) was added 2,3,7,8-TCDD in DMSO such that the final concentration of ligand was 10 nM. One set of duplicate samples was frozen immediately upon addition of 2,3,7,8-TCDD to -70°C with an ethanol/dry bath while a second set of duplicate samples was allowed to incubate at 23°C for 2 h before freezing to -70°C. As a control, 2,3,7,8-TCDD was added to 5 ml of MEGD buffer (final concentration of 2,3,7,8-TCDD, 1 nM) and frozen immediately to -70°C. The samples were removed from the freezer and allowed to thaw just to room temperature. They were immediately transferred to larger (40 ml) vials to which 10 ml of concentrated hydrochloric acid was added. Each sample was then spiked with [‘3C12]2,3,7,8-tetrachlorodibenzo-p-dioxin (50 ng). This “internal standard” was added directly to the sample using a microliter syringe. The vials were then allowed to stand (with occasional swirling) for 1 h. The samples were then extracted with pentane (3 X 10 ml); the original sample bottles were also rinsed with pentane and these rinses were added to the extracts. The extracts were dried over anhydrous sodium sulfate and then evaporated just to dryness. The residues remaining were then purified using a multilayer column procedure (anhydrous NazSOl/siiica/H,S04+silica/silica/NaOH+silica/silica, eluting with 10%) v/v, CH& in pentane) followed by a carbon column (Carbopak C/carbon eluting successively with hexane, cyclohexane + CHzCl,, CH2Clz + CHa + toluene, and toluene). The final eluates from the carbon column purification were evaporated to dryness and the residues were transferred to microvials using dichloromethane rinses. The rinses were then blown down to dryness using Nz and finally redissolved in 25 ~1 of isooctane for GUMS analysis, using a Hewlett-Packard Model 5890 gas chromatograph equipped with Model 5970 mass selective detector operated at 70 eV, and calibrated using perfluorotributylamine. The 12.5-m

388

BUNCE.

LANDERS,

AND

SAFE

HP column was operated with splitless injection using a temperature program of 90°C isothermal for 5 min, followed by a temperature ramp of 15°C per minute to 320°C. GC/MS Analysis. 2,3,7,8-TCDD was monitored through the molecular ion cluster at m/e 320,322 and 324, while the internal standard[i3C1z-TCDD was monitored at m/e 332, 334, and 336. The results are given below:

InM

2,3,‘7,8-TCDD MEGD

buffer

Incubation time (min) 0

Cytosol

0

cytoso1

120

Native 2,3,7,8-TCDD (corrected for recovery of internal standard) (+l ng) 9.0 8.7 7.6 6.0 7.0 6.0

4. Evidence for Noncovalent Binding 2,3,7,8-TCDD to the Ah Receptor

of

Hepatic cytosol(6 mg/ml, 10ml) was incubated for 1 h on ice with 10 nM [3H]-2,3,7,8-TCDD in the presence and absence of 5 mM nonlabeled 2,3,7,8-TCDF. Dextran-coated charcoal (acid washed) was added to and incubated with the cytosol for 15 min at 2°C and then removed by centrifugation at 2OOOg for 2 min. The supernatant (3.0 ml) was layered on 5-20% sucrose density gradients in 40-ml heat-sealable tubes and centrifuged in a Ti60 rotor for 12 h at 2°C at 54,000 rpm (Beckman L8-55 centrifuge). The gradients were then fractionated using an Isco Model 640 density gradient fractionator and the location of the specific binding peak was determined by liquid scintillation counting. The fractions containing the control peak (tubes incubated with [3H]-2,3,7,8-TCDD only) were pooled and the radioactivity was determined. Aliquots of the pooled fraction sample were treated with ice-cold acetone (acetone:sample, 9:l; 16 h; -20°C) and the precipitate was washed three times with ice-cold acetone. A fourth acetone washing did not reduce the protein-associated radioactivity further. The precipitated protein was then redissolved in buffer and the amount of protein-associated radioactivity was determined by liquid scintillation counting. The results are described below: Protein-associated Pooled fractions (3.0 ml) radioactivity (dpm) of the control peak (at 95% confidence level) No treatment Acetone treated

59,733 + 1194 38 f 14

FIG.

1. (A)

Time

course for specific binding of 1.0 to the Ah receptor (0). All data refer to 20°C. (B) Specific binding of [aHI-2,3,7,8TCDD to the Ah receptor which has been incubated with [3H]-2,3,7,8-TCDD at time zero and competed with 500 nM nonradiolabeled TCDF at times indicated (0). Due to the rapid rate of specific binding of [3H]TCDD to proteins in the hepatic cytosol, addition of the ligand to the cytosolic mixture with immediate freezing to -70°C (20-30 s) consistently showed a baseline specific binding of 5-8 fmol/mg cytosolic protein. Consequently, curves A and B consistently cross at slightly more than 50% of the maximum. nM rH]-2,3,7,8-TCDD

Ligand

Displacement

Studies

To hepatic cytosol (1 ml, 2 mg/ml protein) was added [3H]-2,3,7,8-TCDD (10 ~1, 10m7 M) and incubated at 20°C for convenient times. Excess unlabeled 2,3,7,8-TCDF (10 ~1, 5 X 10m5 M) was then added, the solutions were mixed thoroughly, and a further 2-h incubation was carried out. The samples were then analyzed using the hydroxylapatite binding assay as described above. Typical results are given in Fig. 1.

Kinetics

of Ligand Binding Temperatures

at Different

To hepatic cytosol (1.0 ml, 2 mg/ml) in duplicate tubes was added [3H]-2,3,7,8-TCDD (10 ~1, 10e7 M in DMSO) and either unlabeled 2,3,7,8-TCDF (10 ~1, 5 x 10m5 M in DMSO) for nonspecific binding determination or DMSO (10 ~1) for determination of total binding. After thorough mixing the cytosol-ligand mixture was immediately incubated in a water bath and maintained at constant temperature with an accuracy of +-O.l”C. Samples were removed from the constant temperature bath at predetermined times and rapidly frozen to -70°C with ethanol/dry ice. When each mixture had been incubated for the appropriate time, the frozen samples were thawed individually in a 2°C water bath, 0.20 ml was added to 0.25 ml hydroxylapatite slurry, and the remainder of the hydroxylapatite binding assay was carried out as de-

Ah

RECEPTOR-2,3,7,8-TETRACHLORODIBENZO-p-DIOXIN

scribed earlier. Identical experiments were carried out over a temperature range 4-37°C. The second-order rate constants are summarized in Table IA.

Kinetics

of Receptor

Inactivation

Unliganded hepatic cytosol (1.0 ml, 2 mg/ml) was equilibrated in a constant temperature water bath maintained with an accuracy of kO.l”C. After incubation for predetermined times, samples were removed and immediately frozen in a bath of ethanol/dry ice at -70°C. When all samples were run, [3H]-2,3,7,8TCDD (10 ~1, 10 7 M in DMSO) in the presence and absence of a 500-fold molar excess of nonradiolabeled 2,3,7,%TCDF was added to the samples while still frozen. All samples were then thawed simultaneously in a water bath at 2”C, thoroughly mixed, and incubated for 2 h at 20°C. After this incubation the remainder of the hydroxylapatite assay was carried out. Identical experiments were carried out over a temperature range4-37°C. The first-order rate constants are given in Table IB.

Simulation For Scheme y

-dlRl

-

dt F

Studies

1, the differential

equations

= kJRl[L]

- k,[RL]

= kJRl[L]

+ k,[R]

are

- k,[RL]

= kJR]

where P represents the inactivated product. By setting starting conditions [RL] = 0, [P] = 0, and by assuming values for Ro, kf, kd, and k,, a numerical solution to the problem can be obtained using a fourth-order Runge-Kutta procedure (43). Values of kf and kd were measured experimentally; those of k, and R. were adjusted until the simulated and experimental binding curves matched as closely as possible. For Scheme 2, with b 6 kD + k,[L], an analytical solution is possible (Discussion, Eq. [3]). The simulated binding was generated using trial values of k,, and R. and again the simulated and experimental curves were matched as closely as possible. RESULTS

[3H]-2,3,7,8-TCDD-Receptor Binding Association and Dissociation Figure 1 (curve A) illustrates the time course binding of a saturating concentration of [3H]-2,3,‘7,8-TCDD (lo-’ M) and rat hepatic cytosol(2 mg/ml) at 20°C. Satura-

INTERACTIONS

389

tion (defined throughout this paper as 90% of the asymptotic value) was observed within 2 h. In a parallel series of experiments the amount of ligand which could be displaced from the complex was determined by incubating the hepatic cytosol with [3H]-2,3,7,8-TCDD and adding a 500fold molar excess of unlabeled 2,3,7,8TCDF at selected times. The apparent displaceability of the radioligand from the complex was time dependent (Fig. 1, curve B). Since curves A and B intersect at a value close to 50% of maximum binding, it was concluded that the “displaceable” [3H]-2,3,7,8-TCDD at any particular time actually represents the fraction of Ah receptor protein that has not yet bound to the radioligand. Moreover, it is evident that at longer time points, minimal displacement of the bound [3H]-2,3,7,8-TCDD by excess unlabeled competitor is observed. This nearly irreversible binding of [3H]-2,3,7,8-TCDD with rodent hepatic cytosol has previously been observed (36,45) and contrasts with the more facile displacement of steroids (e.g., estradiol) from their receptor protein complexes (46). It appears that binding of 2,3,7,8TCDD to the receptor stabilizes the protein, and that the bound complex is stable for many hours under conditions whereas the free receptor is inactivated rapidly. The term inactivation of the receptor is used to denote loss of Ah receptor-binding activity. Control experiments showed that this long-term binding of 2,3,7,8-TCDD to the receptor in buffer at pH 7.4 is not the result of a permanent covalent attachment: under denaturing conditions (acetone at -20°C for 16 h) the [3H]-2,3,7,8-TCDD is released from the protein matrix. Another set of controls showed that 2,3,7,8-TCDD is not chemically modified as a result of incubation with the cytosolic preparations; and is quantitatively recovered unchanged by extraction with an organic solvent. Temperature-Dependent Saturation Binding of [3H]-2,3,7,8-TCDD and Inactivation of the Unbound Rat Hepatic Cytosolic Receptor Protein Figure 2 summarizes binding and inactivation

the saturation of the unbound

390

BUNCE,

LANDERS,

a

AND

SAFE

b

---)

,0-O iIT-A+,. 50 f,. I:: 20 / l ” 4oi,

-7

,--19

100

.-. .-.

200 300

i

,/A

.

30

.-.

/

f

m-m .-. .-.

,= ,u

10 0i 0

e-m2

1. 1

/ 4

6 Time

FIG. 2. (M), 20°C TCDD to Symbols

100 200 300

a

10

12

14

6 Time

I

6

10

*a

(hours)

B. Receptor inactivation

formation kf mol-’ 2.9 3.0 7.2 2.5 3.6 5.2 9.1 7.9 1.4 1.1 3.4 3.1

x x X X X x x x x x x x

ki (mini)

mini) lo6 lo6 lo6 lo7 10’ lo7 lo7 lo7 lo* lo8 lo* loa

9.85 10.35 20.30 20.45 30.15 30.50 36.70

2.7 3.4 1.5 2.3 8.9 1.2 1.4

2,3,7&TCDD; kfwas obtained as the initial slope from a plot of specific binding vs time at known ligand concentrations.4 These rate constants were shown to be independent of changes in the initial concentrations of either C3H]-2,3,7,8-TCDD or the protein in t,he cytosol (data not shown). Simulation of Association Curves of [3H]2,3,7,8-TCDD-Rat Hepatic Cytosolic Receptor Protein Interactions A computer simulation program was adapted from that of Beck and Goren (43) and the [3H]-2,3,7,8-TCDD-receptor binding data (Fig. 3) were fitted with the aid of an IBM-XT compatible microcomputer to a simulation based on Scheme 1. The error bars on the experimental data points in

EXPERIMENTALRATECONSTANTSFORLIGAND BINDINGANDRECEPTORINACTIVATION ATDIFFERENTTEMPERATURES

9.85 10.30 16.00 20.40 20.70 25.50 30.25 30.30 31.20 31.20 37.00 37.00

4

f--

(a) Time course for specific binding of 1 nM [aHI-2,3,‘7&TCDD to the Ah receptor at 10°C (o), and 30°C (A). (b) Time course of receptor degradation as judged by binding[3H]-2,3,7,8the receptor after preincubation of cytosol in the absence of ligand for the times indicated. as in (a).

TABLE

(liters

2

(hours)

rat hepatic cytosolic receptor protein at 10, 20, and 30°C. At 2O”C, for example, over 50% of the unbound receptor inactivates within 40 min. More limited data were also obtained at 37 and 4°C (data not shown). The experimental rate constants are given in Table I: kd was obtained from a first-order plot of the amount of receptor remaining following inactivation in the absence of

A. Complex

OC 0

x x x X X x x

4 The validity of this conclusion Scheme 1 at short reaction times 1o-3 1o-3 lo-* lo-’ 10-a 10-l 10-i

- 9 Therefore, y

= kfIRHL]

R = R,@d = k/IRIL]

+ ‘P

is as follows. From when [RL] = 0:

lil

+ kd[R]. ,

= kflL][R&%

+ ‘Pl)t.

When t is small, the expansion of the exponential may be truncated at the first term: -A WI At

= kALxRoJ

in the limit

t + 0.

[ii]

Ah

i; aw EZ T;: &E E

RECEPTOR-2,3,7,8-TETRACHLORODIBENZO-p-DIOXIN

391

INTERACTIONS

6 5 4

2

0 0

200

400

600

TIME (minutes)

FIG. 3. Fit of the observed experimental binding of [3H]-2,3,7,8-TCDD to the Ah receptor at 20°C (0) to the computer-simulated binding curve (A) according to Scheme 1. Curve B is a simulation of the time course of binding of [3H]-2,3,‘7,8-TCDD-Ah receptor at 20°C using values of Ro, kf, and kd from Table I, but with k, chosen to give kD = 2.5 nM.

Figs. 3 and 5 include the statistical uncertainty in the scintillation counting and in the replication (duplicate) of the experiment. Since kd and kfcould be obtained experimentally, there were only two parameters to be fixed: R,, the initial receptor concentration (prior to inactivation) and k,. Of these, R,, was fixed by adjusting its value until the experimental saturation binding concentration was matched by the simulation. The kinetic experiments demonstrate that [RL] dissociation was minimal since the saturating binding concentration of this complex was maintained for at least 10 times the length of time required to reach saturation of the receptor. The values of k, are therefore very small and may approach zero. The k, values shown in Table I were obtained by assuming that over a period of 10 saturation lifetimes the maximum amount of dissociation would be 10%. On this basis, the k, values quoted are maximal. DISCUSSION

The Scatchard receptor-ligand

analysis of bimolecular complexation reactions

assumes that the [RI, [L], and [RL] values measured at saturation represent the equilibrium concentrations of these species. As shown in Fig. 2b [and also by Kester and Gasiewicz (36)], the unbound receptor rapidly loses its capacity to bind 2,3,7&TCDD. Therefore the concentrations of R, L, and RL at saturation will not represent equilibrium concentrations and can lead to inaccurate estimation of K,,, and the initial concentration (R,) of receptor binding sites. From the experimental values for kf and kd, we estimated the values of k, and R. by matching the experimental results with the computer simulated curve (Fig. 3). The values, normalized to 10,20, and 3O”C, are summarized in Table II. Since k, is a maximum estimate for the dissociation constant, the K,,, (= kf/k,) values are minimum estimates; and it is not possible to assign any meaningful uncertainty to them. The K,,, minima (3 X 10” to 3 X 10” Mm’) were at least 2 orders of magnitude greater than those corresponding to the published values of KD (13,21,27,28). A recent report by Bradfield and Poland (47) shows that in the case of the Ah receptor from the

392

BUNCE,

LANDERS, TABLE

SUMMARY

OF RATE

AND

SAFE

II

CONSTANTS NORMALIZED TO 10,20, AND 30°C

T(“C)

9.85

k,” (liter mol-’ mm’) kd" (mm’) k,b,"(mm’) Kassbsd (liter mol-‘) Saturation level” (mol liter

2.99 x 10” 0.0027 1.0 x 1om4 3.0 x 1o’O (6.0 X lo-“)@ 1.22 x lo-” 700 0.0030 0.0050

-1 )

[%I” Saturation time (min)” k&l (min-‘) kJ(min-‘)

20.40 2.54 X lo7 0.019 1.5 x 1o-4 1.7 x 10” 9.3 x lo-” 1.68 X 10-l' 100 0.025 0.033

30.30 7.85 X lo7 0.102 2.5 X W4 3.1 x 10" 9.3 x lo-” 2.20 x 1o-'o 30 0.079 0.14

a Experimental value. b Fitted parameter. ’ Maximum estimate. d Minimum estimate. ’ Saturation may not have been complete. f Scheme 2.

C57BL/6J mouse the conventional Scatchard analysis is also inappropriate due to ligand depletion. These authors estimate KD = 6 X 10-l’ M at infinite protein dilution. The results suggest that K,,, has been underestimated in the past because if kf and kd had the values determined experimentally in our study and if k, had a magnitude consistent with the literature values of KD (i.e., k, = KDkf), the complex should readily dissociate with time. This effect, which is simulated in Fig. 3B, was not observed experimentally. Previous studies (36) allude to the “stickiness” and irreversibility of the [3H]-2,3,7,8-TCDDAh receptor protein complex compared with reversible dissociation of some steroid receptor-hormone complexes; these revised equilibrium constants more accurately reflect the unique properties of the 2,3,7,8-TCDD-Ah receptor complex and are also consistent with the unusually high potency and persistence of the in vivo effects of 2,3,‘7,8-TCDD. The activation parameters for the Ah receptor protein-2,3,7,8-TCDD complexation reaction and for the inactivation of the uncomplexed receptor protein were calculated from the temperature dependence of kf and kd over the temperature range lo37°C. Less reliance was placed on the ki-

netic data obtained at 37°C because the reactions were sufficiently fast that few data points could be obtained in each run. The values of the activation parameters are given in Table III. The “graphical” data are those obtained from conventional Arrhenius plots (Fig. 4); the “computational” results are based on a new one-step methodology (48) which permits calculation of activation parameters directly from the raw experimental measurements without having to obtain the rate constants as an intermediate step. The new method allows the uncertainty in the activation enthalpy, expressed as a 95% confidence interval, to be obtained unambiguously. The activation energy for thermal inactivation of the hepatic Ah receptor from the Wistar rat is comparable with that (146 kJ mol-‘) recently reported by Kester and Gasiewicz (36) for the Sprague-Dawley rat. It is also

TABLE

III

ACTIVATION PARAMETER COMPARISON Graphical Complexation Inactivation

127 128

AH* (kJ/mol) Computational 140 f 19 122* 9

AS’ (kJImo1.K) 86.2 34.8

Ah

SC32

---

RECEPTOR-2,3,7,8-TETRACHLORODIBENZO-p-DIOXIN

33

Temperature-1

FIG. 4. (a) Arrhenius ius plot (In kfvs l/T)

34

4 ( K-l,

393

INTERACTIONS

36

x lo3

plot (In k,vs for complexation

l/T)

)

Temperature-1

for degradation of [3H]-2,3,7,8-TCDD

similar to the values obtained for inactivation of the mouse androgen receptor (110 kJ mall’ (49)) and of glucocorticoid receptors (80-130 kJ mall’ (50)). The inactivation parameters are consistent with considerable conformational changes in the protein attendant upon both complexation and inactivation. They fall short of the ca. 300 kJ mall’ values (51) expected for complete unfolding of the protein. The large activation energy (enthalpy) for complex formation is consistent with the high affinity of the Ah receptor for associating with nonpolar ligands of defined molecular dimensions (5,15, 29, 52). The gross conformational change is reflected in the large positive AH:, indicating that at the transition state many hydrogen bonds are in the process of reorganization, while the very large positive AS! suggests that the driving force in this hydrophobic binding is solvent reorganization as the molecule changes its conformation. Thus the increase in solvent entropy is sufficient to allow binding to be moderately fast at 20°C (and very fast at 37°C) despite the large enthalpic activation barrier that has to be overcome. Biologically, Scheme 1 represents a rather conventional mechanism for ligand association, with equilibration being complicated somewhat by receptor inactivation. The fit to the data is not perfect, as shown in Fig. 3; the early part of the association occurs faster than predicted, and the later part is slower. This effect is seen

( K-l,

of the unoccupied Ah receptor. to the Ah receptor.

x 103

)

(b) Arrhen-

consistently from run to run. Kinetically, this would be consistent with two independent sites for binding, either if the Ah receptor were an aggregate (53) or if the protein underwent cleavage distant from the binding site (e.g. (54)), and the uncleaved and cleaved protein bound 2,3,‘7&TCDD at different rates. The results of these more complex models have not been reported due to the difficulty of assigning physical meaning to the derived rate constants as the complexity of the kinetic model and the number of fitted parameters increases. A completely different explanation of the binding kinetics is suggested by the work of Matsumura and colleagues (5557), who reported that 2,3,7&TCDD is capable of altering cell membranes. In the case of the epidermal growth factor receptor (56), 2,3,7&TCDD stimulates its phosphorylation. In steroid hormone receptors the rate of ligand binding is substantially different depending upon whether the receptor is phosphorylated (58). Gasiewicz and Bauman (59) have recently speculated upon a rather analogous proposal, namely that the Ah receptor may be activated by 2,3,7,8-TCDD-induced change of the unoccupied receptor to a less charged, possibly dephosphorylated, form. As yet, there is no evidence that the extent of phosphorylation of the Ah receptor is affected by 2,3,7&TCDD; nevertheless, the idea suggests that a mechanism for 2,3,7,%TCDD/ Ah receptor binding involving receptor activation prior to binding would be worth

394

BUNCE.

LANDERS,

considering. The simplest form of this mechanism is shown as Scheme 2, in which the receptor is activated to an intermediate, I, followed by competitive degradation of the unoccupied receptor and formation of the RL complex, both these latter steps being fast compared with the first: ko Rszr

%I c

kp

PI

degradation RL

According to Table III, rate constants kf and kd have the same temperature dependence, i.e., they have the same activation energies. If Scheme 1 or a variant is the correct association mechanism, this would be simply a coincidence; if Scheme 2 were correct, they would be the same because the rates of both processes depend only upon the rate constant ko. The similar activation enthalpies (energies) are thus consistent with Scheme 2. The conclusion that ligand binding and receptor degradation both involve a conformational reorganization of the protein remains unchanged. Many variants of Scheme 2 are possible; however, the simplest model, which will be discussed here, has none of the steps considered reversible and has k,, representing the rate-limiting step.5 The time dependence of RL formation is then given by Eq. [3] which is a simple exponential function:

[ELI = [&I (kF~~~k~

(1 - emkot).[31

The experimentally measured rate of receptor inactivation in the absence of 2,3,7,8-TCDD leads to the rate constant k. according to this model. The initial rate of complex formation is given by Eq. [4], in which the factor a = kF[L]/(kF[L] + kD) is the fraction of intermediates I that are trapped by the ligand.

5 A scheme analogous to Eq. [2], hut having reverse steps corresponding to S and kF, has also been examined. This can also be fitted to the saturation binding curve, but because there are five unknown rate constants as well as [I?,,] to be fitted, the fit is not unique. Consequently, it has not been reported.

AND

SAFE

y

= k,JRJ.

[41

At a single temperature and in the absence of ligand, kd of Scheme 1 is associated with & of Scheme 2. In order to compare kf of Scheme 1 with an appropriate parameter in Scheme 2 we have, from the initial rate experiments, kJL] = &a. The values of kJL] are given in Table II and when [2,3,7,%TCDD] = lo-’ M, a value in the range 0.58-0.75 will accommodate the kinetic data over the temperature interval studied. Thus provided that Scheme 2 is operative, this concentration of 2,3,7,8TCDD will trap roughly two-thirds of all the activated receptor molecules. Since only a fraction (01) of the initial receptor concentration (R,,) becomes ligand bound, then the model represented by Scheme 2 indicates that once again R. is underestimated at the saturation limit. Figure 5 shows a fit of the experimental saturation binding data to Scheme 2 at 2O”C, using [Ro] = 1.22 X 10-l’ mol liter-‘, k. = 0.033 min-‘, and CY= 0.75. The fit of the experimental data to both Schemes 1 and 2 (Figs. 3 and 6) was comparable. Again this implies that the simplest version of Scheme 2 may not fully represent the reaction mechanism. It is important to note that two different mechanisms (Schemes 1 and 2) fit the experimental data equally well, and both mechanisms have different biological implications which will be addressed in future studies. In conclusion it has been shown that a simple association model does not account for the binding of 2,3,7&TCDD to the Ah receptor. Conclusions obtained by a Scatchard analysis of experimental ligand association data underestimate both the K,,, and the total concentration of receptor sites. The binding of 2,3,7,8-TCDD to the Ah receptor is thus qualitatively and quantitatively different from the binding of steroid hormones to their receptors in several ways. First, receptor inactivation occurs competitively with binding over the whole range of temperatures from 4 to 37°C; second, there is as yet no direct evidence for complex dissociation; and third,

Ah

RECEPTOR-2,3,7,8-TETRACHLORODIBENZO-p-DIOXIN

INTERACTIONS

i

0

200

400 TIME (minutes)

FIG. 5. Fit of the experimental binding of [3H]-2,3,7,8-TCDD computer-simulated binding curve (-) according to Scheme

binding of the ligand stabilizes the receptor to an extent much greater than observed with the corresponding steroid hormone receptors. Two different kinetic models have been used to rationalize the experimental data. In Scheme 1, association is viewed as being in competition with receptor degradation. According to this model, previous estimates of the association constant K,,, are orders of magnitude too low (if reversal occurs at all), and the initial cytosolic receptor concentration has been underestimated by a factor of 2-3. In Scheme 2, we postulate the slow conversion of the Ah receptor to an activated state, and this species is either inactivated or is competitively trapped by 2,3,‘7,%TCDD. In the simplest version of this model, none of the steps are considered as reversible but the model can readily be extended to accommodate reversibility. The coincidence of Al$ and AIIZ is consistent with Scheme 2, but before this mechanism can be seriously considered, it will be imperative to show that the same coincidence of A$ and A& is found for a

to the Ah receptor

at 20°C (9)

to the

2.

series of halogenated aryl hydrocarbon radioligands, rather than just one. Current studies in our laboratory, utilizing new radioligands and cytosolic receptor protein from different animal species, are in progress in order to provide a better understanding of the interaction between the Ah receptor and members of the PCDD and PCDF families. The long-term goal is to understand better what influence the interplay of the kinetic parameters has upon the toxicity of PCDD/PCDFs considering both different congeners in one species and a common congener in different species.

ACKNOWLEDGMENTS We owe particular thanks to Dr. J. S. Beck, University of Calgary, for discussions on curve fitting and for kindly making available to us a copy of his receptor-ligand binding program and to Dr. Uwe Oehler for adapting Dr. Beck’s program to run on our equipment. We also thank Brock Chittim and Judie Sparling of Wellington Environmental Consultants for GC/MS quantitation of 2,3,‘7&TCDD in hepatic cytosol. We also thank Dr. Lorna Safe for the synthesis of [3H]-2,3,7,8-TCDD.

396

BUNCE,

LANDERS,

REFERENCES 1. POLAND, A., GREENLEE, W. F., AND KENDE, A. S. (1979) Ann. N. y. Acad Sci 320,214-230. 2. POLAND, A., AND KNUTSON, J. C. (1982) Annu. Rev. Pharmacol. ToxicoL 22,517-554. 3. POLAND, A., KNUTSON, J., AND GLOVER, E. (1983) in Human and Environmental Risks of Chlorinated Dioxins and Related Compounds (Tucker, R. E., Young, A. L. and Gray, A. P., Eds.), pp. 539-559, Plenum, New York. 4. SAFE, S. (1984) CRCCrit. Rev. To&coL 13,319-395. 5. SAFE, S. (1986) Annu. Rev. PharmacoL ToxicoL 26, 371-399. 6. NEBERT, D. W., EISEN, H. J., NEGISHI, M., LANG, M. A., HJELMELAND, L. M., OKEY, A. B. (1981) Annu. Rev. Pharmacol. ToxicoL 21,431-462. 7. EISEN, H. J., HANNAH, R. R., LEGRAVEREND, C., OKEY, A. B., AND NEBERT, D. W. (1983) in Biochemical Actions of Hormones (Litwack, G., Ed.), Vol. X, pp. 227-257, Academic Press, New York. 8. ISRAEL, D. J., AND WHITLOCK, J. P., JR. (1983) J. BioL Chem. 258,10390-10394. 9. JONES, P. B. C., GALEAZZI, D. R., FISHER, J. M., AND WHITLOCK, J. P., JR. (1985) Science 227, 1499-1502. 10. HANKINSON, O., ANDERSEN, R. D., BIREN, B. W., SANDER, F., NEGISHI, M., AND NEBERT, D. W. (1985) J. BioL Chem. 260,1790-1795. 11. TUKEY, R. H., HANNAH, R. R., NEGISHI, M., NEBERT, D. W., AND EISEN, H. J. (1982) Cell 31,275284. 12. TUKEY, R. H., NEGISHI, M., AND NEBERT, D. W. (1982) Mol. Pharmacol. 22,779-786. 13. POLAND, A., GLOVER, E., AND KENDE, A. S. (1976) J. Biol. Chem. 251,4436-4446. 14. ROBERTS, E. A., SHEAR, N. H., OKEY, A. B., AND MANCHESTER, D. K. (1985) Chemosphere 14, 661-674. 15. SAFE, S., FUJITA, T., ROMKES, M., PISKORSKAPLISZCZYNSKA, J., HOMONKO, K., AND DENOMME, M. A. (1986) Chemosphere 15, 16571664. 16. OKEY, A. B. (1983) in Human and Environmental Risks of Chlorinated Dioxins and Related Compounds (Tucker, R. E., Young, A. L., and Gray, A. P., Eds.), pp. 423-440, Plenum, New York. 17. OKEY, A. B., BONDY, G. P., MASON, M. E., KAHL, G. F., EISEN, H. T., GUENTHNER, T. M., AND NEBERT, D. W. (1979) J. Biol. Chem. 254,1163611648. 18. OKEY, A. B., BONDY, G. P., MASON, M. E., NEBERT, D. W., FORSTER-GIBSON, C. J., MUNCAN, J., AND DUFRESNE, M. J. (1980) J. BioL Chem. 255, 11415-11422.

AND

SAFE

19. OKEY, A. B., MASON, M. E., AND VELLA, L. M. (1983) in Extrahepatic Drug Metabolism and Chemical Carcinogenesis (Rydstrom, J., Montelius, J., and Bengtsson, M., Eds.), pp. 389-399, Elsevier Science, Amsterdam. 20. OKEY, A. B., VELLA, L. M., AND IVERSON, F. (1984) Canad. J. Physiol. Pharmacol. 62,1292-1295. 21. GASIEWICZ, T. A., AND RUCCI, G. (1984) Mol. PharmacoL 26,90-98. 22. DENISON, M. S., VELLA, L. M., AND OKEY, A. B. (1986) J. BioL Chem. 261,3987-3995. 23. DENISON, M. S., WILKINSON, C. F., AND OKEY, A. B. (1986) Chemosphere 15,1665-1672. 24. DENISON, M. S., AND WILKINSON, C. F. (1985) Eur. J. B&hem. 147,429-435. 25. CARLSTEDT-DUKE, J., ELFSTROM, G., SNOCHOWSKI, M., HOGBERG, B., AND GUSTAFSSON, J. A. (1978) ToxicoL Lett. 2,365-373. 26. GASIEWICZ, T. A., AND NEAL, R. A. (1982) Anal. B&hem. 124,1-11. 27. DENISON, M. S., VELLA, L. M., AND OKEY, A. B. (1986) J. BioL Chem. 261,10189-10195. 28. OKEY, A. B., AND VELLA, L. M. (1982) Eur. J. Bip them. 127,39-47. 29. BANDIERA, A., SAWYER, T., ROMKES, M., ZMUDZKA, B., SAFE, L., MASON, G., KEYS, B., AND SAFE, S. (1984) Toxicology 32,131-144. 30. BANDIERA, S., SAFE, S., AND OKEY, A. B. (1982) Chem. BioL Interact. 9,259-278. 31. MASON, G., SAWYER, T., KEYS, B., BANDIERA, S., ROMKES, M., PISKORSKA-PLISZCZYNSKA, J., Z~IIUDZKA, B., AND SAFE, S. (1985) Toxicology 37,1-12. 32. MASON, G., FARRELL, K., KEYS, B., PISKORSKAPLISZCZYNSKA, J., SAFE, L., AND SAFE, S. (1986) Toxicology 41,21-31. 33. HUDSON, L. G., SHAIKH, R., TOSCANO, W. A., AND GREENLEE, W. F. (1983) Biochem. Biophys. Res. Commun. 115,611-617. 34. POLAND, A., AND GLOVER, E. (1975) Mol. Pharmacol. 11,389-398. 35. POLAND, A., AND GLOVER, E. (1980) Mol. Pharmacol. 17,86-94. 36. KESTER, J. E., AND GASIEWICZ, T. A. (1987) Arch. Biochem. Biophys. 252,606-625. 37. NEBERT, D. W., EISEN, H. J., AND HANKINSON, 0. (1984) Biochem. PharmacoL 33,917-924. 38. WHITLOCK, J. P., JR., AND GALEAZZI, D. R. (1984) J. BioL Chem. 259,980-985. 39. POLAND, A., GLOVER, E., EBETTINO, F. H., AND KENDE, A. S. (1986) J. BioL Chem. 261, 63526365. 40. POELLINGER, L., AND GULLBERG, D. (1985) MoL Pharmacol. 27,271-276. 41. POELLINGER, L., LUND, J., GILLNER, M., HANSSON, L.-A., AND GUSTAFSSON, J.-A. (1983) J. BioL Chem. 258,13535-13542.

Ah

RECEPTOR-2,3,7,8-TETRACHLORODIBENZO-p-DIOXIN

42. HANNAH, R. R., LUND, J., POELLINGER, L., GILLNER, M., AND GUSTAFSSON, J.-A. (1986) Eur. J. B&hem. 156,237-242. 43. BECK, J. S., AND GOREN, H. J. (1983) J. Recept. Res. 3,561-577. 44. LOWRY, 0. H., ROSEBROUGH, N. J., BARR, A. L., AND RANDALL, R. J. (1951) J. Biol. Chem. 193, 265-2’75. 45. FARRELL, K., AND SAFE, S. (1987) Biochem. J. 244, 539-546. 46. SASSON, S., AND NOTIDES, A. C. (1983) J. Biol. Chem. 258,8113-8117. 47. BRADFIELD, C. A., AND POLAND, A. (1988) Fed. Proc., abstract 8720. 48. BUNCE, N. J., FORBER, C. L., HUTSON, J. M., AND MCINNES, C. (1988) J. Chem. Sot. Perkin Trans. 2.363-368. 49. WRIGHT, W. W., CHAN, K. C., AND BARDIN, C. W. (1981) Endocrinology 108,2210-2216. 50. SCHAUMBERG, B. P. (1972) Biochim. Biophys. Acta 261,219-235. 51. TANFORD, C. (1968) Adv. Protein Chem. 23, 121282.

INTERACTIONS

397

52. DENOMME, M. A., HOMONKO, K., FIJJITA, T., SAWYER, T., AND SAFE, S. (1985) Mol. Pharmacol. 27,656-661. 53. LUND, J., KURL, R. N., POELLINGER, L., AND GusTAFSSON, J. A. (1982) Biochim. Biophys. Acta 716,16-23. 54. POLAND, A., GLOVER, E. (1987) Biochem. Biophys. Res. Commun. 146,1439-1449. 55. BREWSTER, D. W., MADHUKAR, B. V., AND MATSUMURA, F. (1982) Biochem. Biophys. Res. Commm. 107,68-74. 56. MADHUKAR, B. V., BREWSTER, D. W., AND MATSUMURA, F. (1984) Proc. Natl. Acad. Sci. USA 81, 7407-7411. 57. BOMBICK, D. W., MADHUKAR, B. V., BREWSTER, D. W., AND MATSUMURA, F. (1985) Biochem. Biophys. Res. Commun. 127,296-302. 58. RAO, V. S. K., AND FOX, C. F. (1987) in Recent Advances in Steroid Hormone Action (Moudgil, V. K., Ed.), pp. 339-366, de Gruyter, Berlin/ New York. 59. GASIEWICZ, T. A., AND BAUMAN, P. A. (1987) J. Biol. Chem. 262,2116-2120.