Photoaffinity labelling of the Ah receptor

Fd Chem. Toxic. Vol. 24, No. 6/7, pp. 781-787, 1986

0278-6915/86 $3.00+ 0.00 Pergamon Journals Ltd

Printed in Great Britain

P H O T O A F F I N I T Y L A B E L L I N G OF THE Ah RECEPTOR A. POLAND and E. GLOVER McArdle Laboratory for Cancer Research, University of Wisconsin, Madison, WI and H. EBETINO and A. KENDE Department of Chemistry, University of Rochester, Rochester, NY, USA Abstraet--A series of halodibenzo-p-dioxins bearing the arylazide photolabile functional group were synthesized and tested as photoaffinity labels for the Ah receptor. 2-Azido-3-iodo-7,8dibromodibenzo-p-dioxin (KD=0.76x 10-9M) w a s selected for radiosynthesis. Analysis of the 1251-photoaffinity-labelledproteins in mouse-liver cytosol by denaturing gel electrophoresis revealed two peptides which had apparent molecular masses of 95,000 and 70,000 daltons respectively, were labelled in an approximately 1: 1 ratio and were selectively labelled at low concentrations of the photoatfinity ligand (0.05 K D = 0.04 × 10 -9 M). In addition, their labelling was inhibited by co-incubation with an excess of unlabelled ligand. On chromatographic separation under non-denaturing conditions, these two peptides co-migrated. These studies suggest that the Ah receptor in mouse liver cytosol is a heterodimer composed of two non-covalently bound peptides (95 K and 70 K) which each have a ligand binding site.

Introduction The biochemical and toxic responses produced by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and related halogenated aromatic hydrocarbons are mediated by the reversible stereospecific binding of these compounds to a soluble protein, the Ah receptor, and the ensuing co-ordinate gene expression initiated by this receptor-ligand complex (Poland & Kimbrough, 1984; Poland & Knutson, 1982). This Ah receptor regulates the induction of cytochrome P : 4 5 0 and the concomitant increase in associated monooxygenase activity and other co-ordinately expressed drugmetabolizing enzymes. It also regulates the morphological (i.e. 'toxic') changes, many of which involve cellular proliferation and/or differentiation in epithelial tissues, a response that is more limited in expression, depending on tissue type and animal species. The Ah receptor has been identified and characterized in the cytosol fraction (100,000-g supernatant) in a variety of tissues from mammalian and other vertebrate species. Following the administration of [3H]TCDD, the receptor-ligand complex is associated with the nuclear fraction and may be extracted by a high salt concentration (Greenlee & Poland, 1979; Okey, Bondy, Mason et al. 1979). The model of the Ah receptor in which the unoccupied receptor is present in the cytosol and, upon ligand binding, the ligand-receptor complex translocates into the nucleus has been called into question recently by Whitlock & Galeazzi (1984). These investigators presented evidence in Hepa-I cells (a mouse hepatoma cell line) indicating that the unoccupied Ah receptor normally

resides in the nucleus and that its cytoplasmic location is an artefact of cellular disruption in large volumes of buffer. At present, the identification and characterization of the Ah receptor is dependent on high-affinity reversible binding with a radioligand of high specific activity (e.g. [3H]methylcholanthrene). Using reversible radioligand binding has permitted the physicochemical characterization of the receptor and its chromatographic behaviour (Okey et al. 1979; Poellinger, Lund, Gillner et al. 1983). The relatively little progress reported on the purification of the receptor is attributable to two main factors--the trace concentration of this protein in most tissues (the highest concentration in rodent liver is 100--150 fmol/mg soluble protein or 1-2 #g/g tissue), necessitating an approximately 105-fold purification to homogeneity, and the dependence on reversible radioligand binding for identification. The latter has restricted the use of classical protein purification agents such as chaeotropic agents and detergents, since their use would result in a loss of receptor ligand binding. One approach to avoid the lability of reversible ligand binding to receptor is to use a radiolabelled photoaffinity ligand, which upon irradiation forms a covalent attachment to the Ah receptor. In this paper we outline our preliminary experiments on photoaffinity labelling.

Experimental

Chemical synthesis. The synthesis of the photoaffinity ligands (Fig. 1) and the radiosynthesis of 2-azido-3-[~25I]iodo-7,8-dibromodibenzo-p-dioxin will be presented more fully elsewhere. Briefly, comAbbreviations: EDTA = ethylenediaminetetraacetic acid; KD = equilibrium binding dissociation constant; pounds 1--4 (Fig. 1) were synthesized from the corresNMR = nuclear magnetic resonance; 1-MeNH: ponding 1-amino- and 2-aminohalodibenzo-p-dioxin TCDD = l-methylamino-2,3,7,8-tetrachlorodibenzo-p- by formation of the diazonium salt with NOBF4 in methylene chloride and conversion to the corredioxin; MOPS = 4-morpholinepropanesulphonic acid.

781

A. POLAND et al.

782

K D[nM)

, c, .Fo.

o44

The 100,000-g supernatant fraction was aspirated and was stored at - 7 0 ° C or further processed to the protamine sulphate precipitate.

Protamine sulphate precipitation of hepatic cytosol.

2 c,

oT

'

21

0.49

4 Br Br~ / ' ~

o

.,

0 ~.~-

o.~6 I

,o, rA-~N3 CH2-NH-CJ~

6

13.0

o

CH2 - N H - C - ~ N3

8.1

Determination of the equilibrium dissociation binding constants of the photoaffinity ligands

o CH2-NH-(~-~/~ N3

8

0

113.0

H

CH2 -NH-C~(CH2}5-1~1- - ~ N

To hepatic cytosol (5 mg protein/ml) in MEN or in MEN plus 10% glycerol, at 0°C, was added a solution of protamine sulphate (final concentration 0.2 mg/ml), and after standing for 15 minutes the precipitate was collected by centrifugation at 5000 g for 10 minutes and stored at - 7 0 ° C until use. This procedure precipitated virtually all the Ah receptor, about 15% of the cytosol protein and most of the nucleic acid, providing about a sevenfold enrichment in the receptor per mg of protein. On the day of use, the protamine sulphate pellet (15 mg protein, equivalent to I g liver) was dissolved in 250 gl MEN buffer plus 2 M-NaC1, diluted with buffer to lower the NaCI to 0.5 M and passed over a small column of CMSepharose (Pharmacia, Inc., Piscataway, N J) to remove the protamine sulphate.

3

11.0

Fig. 1. Photoatiinity ligands and their equilibrium dissociation binding constants.

sponding azides by sodium azide. Compounds 5-8 were formed by coupling 1-aminomethyl-2,3,7,8tetrachlorodibenzo-p-dioxin to the N-hydroxysuccinimidyl esters of azidophenyl acids, which are commercially available (Pierce Chemical Co., Rockford, IL). Chemical identification and purity were confirmed by N M R or mass spectrometry.

2- Azido- 3- [125i] iodo- 7,8- dibromodibenzo- p-dioxin. 2-Amino-7,8-dibromodibenzo-p-dioxin was iodinated with cartier-free Na125I (NEZ-0334, New England Nuclear, Boston, MA) and Chloramine T (Sigma Chemical Co., St Louis, MO) in methanol/ HEO/H2SO4 . The reaction mixture was extracted with methylene chloride and the extract was dried over MgSO4 and reacted with NOBF 4 to form the diazonium salt. Finally sodium azide in acetonitrile was added to form the azide. The product was purified by HPLC, using an isocratic mode (methanol-H20, 85:15, v/v) on a C-18 reverse-phase column. Overall yield was 15-20% of the Na~25I incorporated in the product. Preparation of hepatic cytosol. C57BL/6J mice were killed by cervical dislocation and the livers were removed, homogenized in 3 vols MEN buffer (MOPS, 25mM; EDTA, l mM; NAN3, 0.02%; pH 7.5 at 0°C) plus 10% glycerol and centrifuged at 10,000 g for 20 minutes. The supernatant fraction was carefully removed by aspiration to avoid surface lipid, and was centrifuged at 100,000 g for 1 hour.

Hepatic cytosol (2mg protein/ml) in MEN buffer plus 10mM-fl-mercaptoethanol and 10% glycerol was incubated with 1 × 10-gM-[3H]TCDD (c. 30 Ci/mmol) in the presence or absence of a 200-fold molar excess of unlabelled 2,3,7,8tetrachlorodibenzofuran (TCDBF) or with varying concentrations of the photoaffinity ligands for 30 minutes at 20°C. One half the volume of a suspension of charcoal/dextran (final concentration 1/0.1%) was added and after 5 minutes the mixture was centrifuged. The radioactivity in the supernatant was then determined by liquid scintillation spectrometry. Specific binding was calculated as the total b i n d i n g - non-specific binding (binding in the presence of an excess of TCDBF). From a plot of specific binding v. the log of competing ligand concentration, the concentration of competing ligand that reduced specific binding by 50% of the initial binding was estimated (designated the ECs0). The equilibrium dissociation binding constant (Ko) for the competing ligand was calculated from KD=ECs0/1 + [A]/KA where [A] = the concentration of [3H]TCDD and KA = the equilibrium dissociation constant for TCDD (=0.3 x 10-9M). Photolysis. The conditions for photoaffinity ligand binding and photolysis for each experiment are presented in the legends to Figs 2 & 3 and the footnote to Table 1. Gel electrophoresis and autoradiography. Following incubation of the protamine sulphate-precipitated fraction of cytosol with [125I]-labelled photoaffinity ligand and UV irradiation, the protein was precipitated with 9 vols cold acetone and the pellet was washed with acetone and then dissolved in electrophoresis sample buffer (which contained 2% lithium dodecyl sulphate and 20 mM-dithiothreitol). The samples were subjected to denaturing electrophoresis on a discontinuous polyacrylamide slab gel (stacking gel T = 4%, separating gel T = 7.5%) and the gels were stained with Coomassie Blue and dried. The dried gel was exposed to a sheet of XAR-5 film (Eastman Kodak Co., Rochester, NY) backed with an in-

B6 (Ah ~) 1.5 K D

0.SK D

B 6 ( A h a)

0.15K o

O.05K o

t.5K o

Fig. 2. Covalent binding of tzsl-labelled photoaffinity ligand to the protamine sulphate fraction of mouse liver cytosol. The protamine sulphate precipitate of mouse liver cytosol (C57BL/6J mice, i.e. B6(Ah b) and C57BL/6J mice congenic for the Ah d gene, which determines the low affinity Ah receptor B6(Aha)) was prepared and resuspended as described in Experimental. Approximately 300/~g protein in 1 ml was incubated with varying concns of the ~25I-labelled photoaffinity ligand (0.05, 0.15, 0.5 and 1.5 × KD; 1.5Ko= 1.14 x 10-9ra=5.5 × 106 cpm/ml) in the presence or absence of an excess of unlabelled 2,3,7,8-tetrachlorodibenzofuran for 30 rain at 20°C and then for 5 rain at 0°C. An equal volume of charcoal/dextran suspension (final concn 1%/0.1%) was added for 5rain at 0°C and the mixture was centrifuged and then irradiated (sunlamps, 2 > 300 nm, 80 watts at 4 cm for 20 sec). The proteins were precipitated with 10 vols acetone and resuspended in electrophoresis sample buffer, 100/~g protein was subjected to electrophoresis on a 7.5% polyacrylamide denaturing slab gel, and the labelled bands were visualized by autoradiography.

TCDD (nM) 0

0.19

0.38

0.76

1.50

3.00

6.10

15.2

0

Fig. 3. Covalent binding of 2-azido-3-[125I]iodo-7,8-dibromodibenzo-p-dioxin and its inhibition by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). The protamine sulphate-precipitated fraction of C57BL/6J mouse liver cytosol was prepared and resuspended as described in Experimental and 2 ml of this preparation, containing approximately 330/~g protein was incubated with the ~25I-labelled photoaffinity ligand (0.76 x 10 -gM= 1 x K D=3.7 × 106dpm/ml) along with varying concns of TCDD (~15.2× 10-gM) for 30rain at 20°C. The incubation mixture was cooled to 0°C for 5min, charcoal/dextran (final concn I/0.1%) was added to adsorb unbound ligand, and the mixture was centrifuged. The samples were irradiated (sunlamps, 40 watts at 8 cm for 2 rain) and the protein was precipitated by cold acetone and subjected to electrophoresis and visualization by autoradiography. 783

Photoaffinity labelling of the Ah receptor

785

Table 1. Assay for receptor inactivation by photoaffinity ligands: loss of specific binding of [3H]TCDD Specific binding In dark

Incubation

dpm

Control (no additions) Detergent/charcoal control Compound No. 3: 3 × KD 10 × Kt~ 30 × D o Compound No. 8: 3 × KD 10 × KI~ 30 × K D

11,176

With UV irradiation

% of control

dpm

% of control

Dark-light difference in % values

10,011

10,553

100

11,363

100

0

11,517 I 1,027 9928

109 105 94

9264 6918 6572

82 61 58

27 44 36

10,379 9077

99 86 67

7972 7318 5895

70 64 52

29 27 15

7022

Mouse liver cytosol (c. 2 mg protein/ml) was incubated with the photoaffinity ligand (at concentrations of 3, 10 and 30 × KD) for 30min at 20°C. One sample was kept in the dark and the other was irradiated (sunlamp, 40 watts, 8 cm away for 2 rain). The ligand or photoproducts that were not covalently bound were removed by the addition of Triton X-100 (0.5%) for 15 rain followed by the addition of Biobead SM-2 resin (1 g/ml) for 20 min, filtered through a GF/A filter to remove the resin, incubated with charcoal/dextran (final conch I/0.1%) and centrifuged. The cytosol was incubated with 2,3,7,8-tetrachloro[3H]dibenzo-p-dioxin ( 1 . 3 n ~ ) _ a 200-fold excess of unlabelled 2,3,7,8-tetrachlorodibenzofuran. After 30 rain at 20°C, the unbound ligand was removed by the addition of charcoal/dextran, and the bound radioactivity was determined. The specific binding of cytosol treated with detergent/charcoal additions is taken as 100% and as equivalent to that of untreated cytosol and cytosol irradiated in the absence of a photoaffinity ligand.

tensifying screen (Cronex, Dupont Inc.) and held at - 7 0 ° C until film development (Laskey & Mills, 1977). Results

We synthesized a series of halodibenzo-p-dioxins with photolabile arylazide groups as potential photoaffinity labels for the receptor, and evaluated them for their reversible binding to the receptor, their UV absorption spectrum and photolability (data not presented) and their capacity to inactivate the receptor on photolysis. As shown in Fig. 1, one group of compounds was based on the coupling of 1-aminomethyl-2,3,7,8 - tetrachlorodibenzo-p-dioxin (1-NH2Me-TCDD) to azidobenzoic acids or azidobenzamides and a second group was based on converting 1- or 2-aminohalodibenzo-p-dioxins to their corresponding azides. For 1-NH2Me-TCDD the equilibrium dissociation constant, KD, is 2 riM, and the attachment of the bulky phenylazide derivatives to the l-aminomethyl group (compounds 5, 6 and 8) reduced the receptor affinity to 8-13 riM, while for compound 7, which incorporated a phenolic group, the K o was >100nM. In contrast, compounds 1-4, with the azido group attached directly to the dibenzop-dioxin ring, had much high receptor affinities, the K D range being 0.4-2.1 riM. The capacity Of these non-radiolabelled photoaffinity ligands to bind covalently to the receptor was evaluated by an indirect method namely the determination of the reduction in reversible binding of [3H]TCDD after irradiation with the photoaffinity ligand and the removal of the photoproducts. This approach is modelled on an exchange assay used by Katzenellenbogen, Johnson, Carlson & Myers (1974) for examining the binding of photolabile oestrogens to the oestrogen receptor.

The photoaffinity ligands (at concentrations of 3, 10 and 30 x KD) were incubated with liver cytosol, and one sample was irradiated while a second identical sample was not irradiated. Following procedures to remove non-covalently bound ligand and photoproducts, the receptor binding of [3H]TCDD was determined for each sample. The difference in specific binding between the non-irradiated and irradiated sample is a measure of receptor inactivation. Photoaffinity labelling and pseudoaffinity labelling (Ruoho, Kiefer, Roeder & Singer, 1973) can cause receptor inactivation, and thus this assay serves only as a maximal estimate of this parameter. As shown in Table 1, the specific binding of [3H]TCDD to cytosol treated with detergent/ hydrophobic resin and charcoal/dextran is similar to that of control cytosol, and is therefore considered to represent 100% binding. UV irradiation of cytosol (in the absence of photoaffinity labels) has no effect on binding. Cytosol incubated with 2-azido-3,7,8trichlorodibenzo-p-dioxin (compound 3; Fig. 1) but not irradiated has nearly control levels of specific binding, but irradiated samples containing this ligand display a concentration-related decrease in specific binding of [3H]TCDD. The addition of a photoaffinity ligand with lower receptor affinity to cytosol (compound 8) produces a concentrationdependent reduction in subsequent [3H]TCDD specific binding in the absence of irradiation, indicating that the adsorption/extraction procedures are not adequate to remove the ligand completely. Irradiation of these samples produces a greater reduction in specific binding. Despite the limitations of this indirect assay, it permitted the identification of promising photoaffinity labels for radiosynthesis. We selected 2-azido-3-iodo7,8-dibromodibenzo-pdioxin (compound 4; Fig. 1) for radiolabelling with ~25I on the basis of its high reversible binding to the

786

A. POLAND et al.

receptor, its capacity for irradiation-mediated receptor inactivation and the ease of chemical synthesis. A partially purified preparation of Ah receptor from C57BL/6J mouse liver cytosol (protamine sulphateprecipitated fraction) was incubated with varying concentrations of this ~25I-labelled photoaffinity ligand (1.5KD (=l.14mM), 0.5Ko, 0.15K D and 0.05 KD), in the presence or absence of excess unlabelled TCDBF, and irradiated. The protein was precipitated and washed with acetone, the pellet was redissolved in electrophoresis sample buffer and subjected to electrophoresis, and the gel was stained and visualized by autoradiography. As shown in Fig. 2, the irradiated t25I-labelled photoaffinity ligand (in the absence of unlabelled TCDBF) labels several protein bands. Two bands are of particular interest because they are blocked by the presence of an excess of unlabelled TCDBF (lanes 1 v. 2, 3 v. 4, 5 v. 6, 7 v. 8), and they are selectively labelled by low concentrations of the photoaffinity ligand (lanes 5 and 7), indicating a high binding affinity. The last two -lai~es,of.Fig. 3 show the results of labelling experiments on a partially purified preparation of liver cytosol from congenic C57BL/6J mice carrying the Ah d gene which determines a lower affinity Ah receptor. Comparison of these last two lanes shows no selectively labelled bands (blocked by unlabelled TCDBF). The apparent molecular masses of these two labelled protein bands were determined by autoradiography of a 5-20% gradient polyacrylamide long denaturing gel using 125I-labelled marker proteins and were found to be 95,000 and 70,000 daitons (data not shown). The protamine sulphate-precipitated fraction of C57Bl/6J mouse liver cytosol was incubated with the ~25I-labelled photoaffinity ligand (1 x KD.= 0.76 nM) in the presence of varying concentrations of TCDD (0-15.2 nM) and the irradiated proteins were subjected to electrophoresis with visualization of the labelled bands by autoradiography. As shown in Fig. 3, TCDD produces a concentration-related decrease in the intensity of the two bands with apparent Mr of 95 K and 70 K. When these bands were cut from the gel and counted in a gamma counter (Table 2), the radioactivity was equivalent in both and the coincubation with TCDD produced an equivalent reduction of labelling in both.

Table 2. Covalent binding of 2-azido-3-[~2~l]iodo-7,8-dibenzo-pdioxin to the 95 K and 70 K peptides and inhibition by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) Covalent binding (cpm) Concn of TCDD(nM) 0 ~ontrol; n = 2) 0.19 0.38 0.76 1.5 3.0 6.1 15.2

95Kband

70Kband

1421 824 671 464 356 227 206 176

1448 822 738 460 370 270 261 174

The 95 K and 70 K bands from the gel in Fig. 2 were cut out and counted in a gamma counter.

Further characterization of the 95 K and 70 K photoaffinity labelled peptides and their relationship to the Ah receptor (detailed data not shown) may be summarized as follows: (1) [f25I]-photoaffinity-labelled cytosol was subjected to high performance liquid chromatography (HPLC), using a size exclusion column (TSKG4000SW) and an anion exchange column (MonoQ), and the fractions were analysed by denaturing gel electrophoresis and autoradiography. The 95 K and 75 K labelled peptides co-migrate on both columns, and the elution peak corresponds to that of cytosol reversibly labelled with [3H]TCDD. (2) The electrophoretic behaviour of the ~25I-photoaffinity-labelled bands in denaturing gels is the same in the presence or absence of a reducing agent, suggesting that the peptides are not linked by dithiol bonds. (3) Storage of J25I-photoaffinity-labelled cytosol after irradiation for 0-5 days prior to acetone precipitation and electrophoresis produces no change in the labelling of the 95 K and 70 K bands and no formation of new bands. This observation, coupled with the stoichiometric labelling of the two bands and their co-migration of HPLC, suggests that the 70 K band is not a proteolytic degradation product of the 95 K band. (4) The labelled 95 K and 70 K bands were cut out of a denaturing gel, the peptides were eluted and incubated with varying concentrations of proteolytic enzymes, and the proteolytic fragments were subjected to electrophoresis and autoradiography. The patterns of proteolytic fragments in the 95 K and 70 K peptides were nearly identical, indicating substantial homology.

Discussion

We have examined a series of arylazide photoaffinity ligands for the Ah receptor and, using one of these, radiolabelled 2-azido-3-[125I]iodo-7,8 dibromodibenzo-p-dioxin, have characterized the receptor in mouse liver cytosol. Analysis of the photoaffinity labelled products in cytosol by denaturing gel electrophoresis revealed selective labelling of two peptides with apparent M r values of 95 K and 70 K. These two peptides show a similar affinity for the photoaffinity ligand (estimated by the competition with unlabelled ligands), are labelled in an approximately 1 : 1 ratio, co-migrate on nondenaturing size exclusion and anion exchange chromatography, and give a similar pattern of proteolytic fragments. Hannah, Nebert & Eisen (1981) have reported that the Ah receptor from C57BL/6 mouse liver has a Stokes radius of 75 ,~ and an Mr of 245,000 daltons, while Poel!inger et al. (1983) reported a Stokes radius of 61 A and an M~ of 115,000 daltons. Our data on the mouse hepatic Ah receptor suggest that the receptor is a heterodimer with an apparent molecular mass of 165,000, composed of two noncovalently linked peptides, of 95,000 and 70,000. Each of these has a ligand binding site and they have considerable structural homology.

Photoaffinity labelling of the Ah receptor REFERENCES

Greenlee W. G. & Poland A. (1979). Nuclear uptake of 2,3,7,8-tetrachlorodibenzo-p-dioxin in C57BL/6J and DBA/2J mice. J. biol. Chem. 254, 9814. Hannah R. B., Nebert D. W. & Eisen H. J. (1981). Regulatory gene product of the Ah complex. J. biol. Chem. 256, 4584. Katzenellenbogen J. A., Johnson H. J., Carlson K. E. & Myers H. (1984). Photoreactivity of some light-sensitive estrogen derivatives. Use of an exchange assay to determine their photointeraction with rat uterine estrogen binding protein. Biochemistry, N.Y. 13, 2986. Laskey R. A. & Mills A. D. (1977). Enhanced autoradiographic detection of 32p and 125I using intensifying screens and hypersensitized film. FEBS Left. 82, 314. Okey A. B., Bondy G. P., Mason M. E., Kahl G. F., Eisen H. J., Guenthner T. M. & Nebert D. W. (1979). Regulatory gene product for the Ah locus. J. biol. Chem. 254, 11636.

787

Poellinger L., Lund J,, Gillner M., Hansson L.-A. & Gustafsson J.-A. (1983). Physicocbemical characterization of specific and nonspecific polyaromatic hydrocarbon binders in rat and mouse liver cytosol. J. biol. Chem. 258, 13535. Poland A. & Kimbrough R. (Editors) (1984). Biological Mechanisms o f Dioxin Action. Banbury Report 18. p. 500. Cold Spring Harbor Laboratory, New York. Poland A. & Knutson J. C. (1982). 2,3,7,8-Tetrachlorodibenzo-p-dioxin and related halogenated aromatic hydrocarbons: examination of the mechanism of toxicity. A. Rev. Pharmac. Toxic. 22, 517. Ruoho A. E., Kiefer H., Roeder P. E. & Singer S. J. (1973). The mechanism of photoaflinity labeling. Proc. natn. Acad. Sci. U.S.A. 70, 2567. Whitlock J. P. & Galeazzi D. R. (1984). 2,3,7,8-Tetrachlorodibenzo-p-dioxin receptors in wild type and variant mouse hepatoma cells. J. biol. Chem. 259, 980.