Regulatory Gene Product of the Ah Complex COMPARISON OF 2,3,7,8-TETRACHLORODIBENZO-p”IOXIN AND 3-METHYLCHOLANTHRENE BINDING TO SEVERAL MOIETIES IN MOUSE LIVER CYTOSOL*

The major regulatory gene product of the murine Ah complex appears to be a cytosolic receptor. 2,3,7,8-[1,6-3H]Tetrachlorodibenzo-p-dioxin or [3H]3-methylchol-anthrene binding to the Ah receptor and other moieties in hepatic cytosol was examined by gel permeation chromatography, velocity sedimentation (sucrose density gradient centrifugation), dextran-charcoal adsorp- tion, and anion exchange chromatography. In the liver of Ah-responsive C57BL/6N and the Ahb/Ahd hetero- zygote, both radioligands bind to three major components: peak I, a large aggregate which is eluted in the void volume of Sephacryl 5-300 columns and which sediments as a residue to the bottom of sucrose density gradients; peak 11, an asymmetric protein (Mr E 245,000) with a Stokes radius of about 75 A; and peak 111, a globular protein (Mr a 87,000) with an estimated Stokes radius of 40 A. The peak I aggregate is not adsorbed by dextran-coated charcoal and therefore represents the large proportion of nonsaturable radio- ligand binding measured by dextran-charcoal adsorption. The peak I1 protein has a size of about 9.0 S in low ionic strength and 7.5 S in high ionic strength, high affinity for 2,3,7,8-tetrachlorodibenzo-p-dioxin

The major regulatory gene product of the murine Ah complex appears to be a cytosolic receptor. 2,3,7,8-[1,6-3H]Tetrachlorodibenzo-p-dioxin or [3H]3-methylcholanthrene binding to the Ah receptor and other moieties in hepatic cytosol was examined by gel permeation chromatography, velocity sedimentation (sucrose density gradient centrifugation), dextran-charcoal adsorption, and anion exchange chromatography. In the liver of Ah-responsive C57BL/6N and the Ahb/Ahd heterozygote, both radioligands bind to three major components: peak I, a large aggregate which is eluted in the void volume of Sephacryl 5-300 columns and which sediments as a residue to the bottom of sucrose density gradients; peak 11, an asymmetric protein (Mr E 245,000) with a Stokes radius of about 75 A; and peak 111, a globular protein (Mr a 87,000) with an estimated Stokes radius of 40 A. The peak I aggregate is not adsorbed by dextran-coated charcoal and therefore represents the large proportion of nonsaturable radioligand binding measured by dextran-charcoal adsorption. The peak I1 protein has a size of about 9.0 S in low ionic strength and 7.5 S in high ionic strength, high affinity for 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), and saturability at TCDD concentrations greater than about 1.0 m. The peak II protein is not detectable in the liver of Ah-nonresponsive DBA/2N and the Ahd/ Ahd homozygote and therefore represents the Ah receptor. The peak I11 protein has an estimated size of 5.0 S, is not saturable with either TCDD or 3-methylcholanthrene under the conditions of these experiments, and is not associated with the Ahb allele. 3-Methylcholanthrene binds to the peak 111 protein to a greater extent than TCDD.
These data explain the discrepancies between the dextran-charcoal adsorption and sucrose density gradient assays. Any further studies of the function of the Ah receptor and these other ligand-binding moieties (e.g. nuclear translocation) should include gel permeation chromatography in order to distinguish among the various binding components.
The murine Ah locus controls the induction (by polycyclic aromatic compounds such as 3-methylcholanthrene, benzo-[alpyrene, P-naphthoflavone, and TCDD') of many drug-me-* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. tabolizing enzyme activities in virtually all tissues examined (reviewed in Ref. 1). The Ah complex is believed to comprise regulatory, structural, and probably temporal genes which may or may not be linked (2). Multiple forms of cytochrome P-450 are believed to be among the many structural gene products "turned on" during the sequence of events following exposure of the animal to the polycyclic aromatic inducer (3,

4).
A cytosolic receptor for these inducers in genetically "responsive" mice was postulated (5,6) and is now believed to be the major product of the Ah regulatory genes. Experimental evidence in support of this hypothesis has been reported, with the use of a dextran-coated charcoal adsorption assay (7-9), isoelectric focusing following trypsin treatment (10, 11), sucrose density gradient following dextran-charcoal adsorption (12,13), and a detergent-washing procedure with purified nuclei (14). There is no detectable receptor in "Ah-nonresponsive" strains of mice (8,12). The cytosolic receptor with specifically bound inducer appears to translocate into the nucleus during cytochrome P1-450 induction by polycyclic aromatic compounds in the intact liver of Ah-responsive mice (12), in cytosolic and nuclear preparations in vitro (14), and in cell cultures (13). The cytosolic location of the Ah receptor (7)(8)(9)(10)(11)(12)(13)(14) and the temperature-dependent translocation of the inducer-receptor complex (13) are very similar to the properties of the steroid hormone receptors.
Estimates of TCDD-specific binding sites (8,12,13) have ranged between about 10 and 90 fmol/mg of cytosolic protein (-900 to 8,100 sites/cell), with an apparent Kd for ["HITCDD of approximately 0.7 nM (8,10,12). The cytosolic location of a 3-methylcholanthrene-specific binding protein and apparent nuclear uptake of this complex were recently reported; estimates of 3-methylcholanthrene-specific binding sites and apparent Kd were 770 fmol/mg of cytosolic protein and 2.8 m, respectively (15). In the current study we use differences at the Ah locus and four independent separatory methods to demonstrate that mouse liver cytosol contains at least two distinct proteins that bind TCDD and 3-methylcholanthrene. Only one, however, is shown to be the Ah receptor, because this peak is absent in Ah-nonresponsive (Ahd/Ahd) mice.

EXPERIMENTAL PROCEDURES
Chemi~als-[~H]TCDD (55 Ci/mmol) was purchased from KOR Isotopes (Cambridge, MA); gas chromatographic analysis by the manufacturer indicated that the product might contain as much as 20% tri-, penta-, and hexachlorodibenzo-p-dioxins. Nonlabeled TCDD was a generous gift of Dow Chemical Company (Midlands, MI).
Animals-Ah-responsive B6 (Ahb/Ahb) and Ah-nonresponsive D2 (Ahd/Ahd) mice were obtained from the Veterinary Resources Branch, National Institutes of Health. Breeding to obtain responsive heterozygotes (Ahb/Ahd) and nonresponsive homozygotes (Ahd/Ahd) from the B6D2Fl X D2 backcross was carried out in the Developmental Pharmacology Branch mouse colony, National Institute of Child Health and Human Development. Phenotyping progeny of the B6D2Fl X D2 backcross, with respect to the Ah locus, was carried out by the zoxazolamine paralysis test (16) on weanlings (either sex) between 3 and 4 weeks of age. The rigid environment and free access to food and water were described previously (13).
Preparation of CytosoGMice were N e d by cervical fracture. The portal vein was cut and the livers were perfused in situ with HEDG buffer containing 0.1 M NaCl. All subsequent operations were carried out at 4 "C. The livers were removed, minced, and then homogenized in 2 volumes of HEDG buffer (with 0.1 M NaCl) using a Teflon-glass homogenizer. The homogenate was centrifuged at 10,000 X g for 15 min and the supernatant fraction was centrifuged at 100,000 X g for 1 h. The supernatant (cytosolic) fractions were quickly frozen in 2-ml aliquots and stored in liquid nitrogen. No losses of [3H]TCDD or [3H]3-methylcholanthrene binding capacity were noted during storage in liquid nitrogen for 3 months.
Treatment of Cytosol with Radioligand In Vitro-Usually 1 ml of cytosol was treated with the desired concentration of [3H]TCDD or ["H]3-methylcholanthene for 1 h at 0-4 "C. Unless otherwise specified, 15 mg of cytosolic protein/ml were used in all experiments, except sucrose density gradient centrifugation, in which case 5 mg of protein/ml was used. In some samples, nonlabeled TCDD or 3-methylcholanthrene (in 100-fold excess) was added to the radioligand before treating the cytosol. As noted before (13), p-dioxane was the solvent for TCDD. A minimal amount of acetone was the solvent for 3-methylcholanthrene. After treatment at 4 "C for 1 h, the cytosol was examined by one of four independent techniques.
Gel Permeation Chromatography-Sephacryl S-300 gel was equilibrated in HEDG buffer containing 0.5 M NaCl, and columns with dimensions of 35 X 0.9 cm were prepared. The chromatography of 1 ml of the radioligand-treated cytosol was performed by gravity flow (approximately 5 ml/h). Proteins and blue dextran were used as markers to calibrate the columns.
Dextran-coated Charcoal Adsorption-The radioligand-treated cytosol was added to a dextran-charcoal pellet (10 mg of charcoal/mg of dextran, pelleted from HEDG buffer). Dextran-charcoal was resuspended on a Vortex mixer, and the sample was incubated at 4 "C for 15 min before the dextran-charcoal was removed by centrifugation at 4000 X g for 15 min. Aliquots of cytosol were taken both before and after dextran-charcoal treatment for determination of "total" and "bound" radioactivity (8).
Velocity Sedimentation-Following dextran-charcoal adsorption, 300 yl of cytosol was layered onto linear (5 to 20%) sucrose density gradients prepared in HEDG buffer containing 0.1 M NaCl. Gradients were centrifuged at 2 "C for 16 h at 235,000 X g in a Beckman SW 60 Ti rotor. The gradients were then separated in 0.2-ml fractions with an ISCO model 640 gradient fractionater. ["CIFormaldehyde-labeled bovine serum albumin and other protein standards were sometimes used as internal sedimentation markers (12).
Anion-exchange Chromatography-DEAE-52 gel was equilibrated in HEDG buffer containing 0.05 M NaCl, and columns with dimensions of 10 X 1.5 cm were prepared. Chromatography of 1 ml of the radioligand-treated cytosol was performed by gravity flow (approximately 10 ml/h), with a linear NaCl elution.
Determination of Radioactivity-Samples of 0.2 ml or less were dissolved in Aquasol. Tritium was measured at 40 to 50% efficiency, and I4C at 90% efficiency, with the use of a Mark 111 Tracor spectrometer.

RESULTS
Comparison of Gel Permeation Chromatography with Velocity Sedimentation: L3H/TCDD Binding to B6 a n d 0 2 Cytosol-The sucrose density gradient assay is designed to resolve macromolecules with sedimentation coefficients of about 15 S or less. In order to detect larger components, we therefore used gel permeation chromatography (Figs. 1 and 2).
[3H]TCDD binding in the Ah-responsive B6 mouse was observed in three regions (Fig. 1A): peak I eluting in the void fractions; peak I1 eluting in fractions 23 to 26 (Stokes radius P 75 A); and peak I11 eluting in fractions 30 to 33 (Stokes radius e 40 A). Peak I appears to represent a very large complex or aggregation. The saturability of components in these peaks was examined by incubating B6 cytosol with the radioligand plus a 100-fold excess of nonlabeled TCDD, and only peak I1 exhibited saturability (Fig. 1A).
In order to relate these peaks to those observed on sucrose density gradients, we applied samples from the peak I, peak 11, and peak 111 regions of the Sephacryl S-300 eluate directly to sucrose gradients (Figs. 1B and 2B). Peak I resulted in a "trailing" of [3H]TCDD in the upper fractions of the gradient and a large amount on the bottom of the centrifuge tube. Peak I1 resulted in a peak with a maximum at gradient fractions 10 to 12 (approximately 9 S). Peak I11 resulted in a peak at gradient fractions 6 and 7 (about 5 S). Fig. 2 illustrates that the Ah-nonresponsive D2 mouse does not have the saturable component in peak I1 but does possess peaks I and 111. These data indicate that peak I1 represents the Ah receptor and that the B6 liver cytosol has at least two other TCDD-binding moieties that are nonsaturable, not associated with the Ah locus, and present also in D2 liver Comparison of Gel Permeation Chromatography with Dextran-Charcoal Adsorption-Sephacryl S-300 and other chromatographic matrices quantitatively adsorb free [3H]TCDD which can be released only by organic solvents such as pdioxane (12). The r3H]TCDD eluted from the Sephacryl S-300 column (Figs. L4 and ZA) therefore represents radiolabel bound to macromolecules. More than 45% of the total ["HI-TCDD eluted from the column was associated with the nonsaturable peak I (Table I).
[3H]TCDD associated with peak I was not adsorbed to dextran-coated charcoal (data not illustrated). Comparing the two methods (Table I), we found approximately the same amount of bound radioligand in the dextran-charcoal adsorption assay as by gel permeation chromatography (the sum of peaks I, 11, and 111).
Comparison of Velocity Sedimentation with Dextran-Charcoal Adsorption-The dextran-coated charcoal adsorption assay (7-9) provides a rapid method for separating free from macromolecular-bound [3H]TCDD; however, all three components of binding are included in this measurement. The addition of excess nonlabeled TCDD as a competitor has variable effects on the nonsaturable components. As shown in Figs. 1 and 2, [3H]TCDD binding to peak I is often decreased; however, at higher protein and radioligand concentrations, [3H]TCDD binding to peak I is increased by addition of competitor (data not shown). These effects make it extremely difficult to use dextran-coated charcoal adsorption alone as the basis for measuring ["]TCDD binding to the peak I1 moiety. In sucrose density gradients, peak I material sediments to the bottom of the centrifuge tube (Figs. 1B and 2B); hence, peak I does not interfere with the detection of peak 11.
Effect of Protein Concentration-There are certain situations (e.g. estimation of "free" TCDD concentrations when cytosol.

TABLE I
Comparison of gel permeation chromatography with dextrancharcoal adsorption Following treatment of B6 liver cytosol with C3H]TCDD alone or in the presence of a 100-fold excess of nonlabeled TCDD, aliquots were examined by dextran-coated charcoal adsorption. The remainder of the sample was examined by Sephacryl S-300 chromatography and represents the data illustrated in Fig. L4. The chromatogram was divided into three regions: fractions 16 to 22 (peak I); fractions 23 to 27 (peak 11); and fractions 28 to 35 (peak 111). The values shown reflect the total radioactivity in these regions, and no corrections are made for possible overlap of components. Peak I material is not totally separable from peak I1 material (Fig. IA), so that quantitation of Deak I1 bv gel permeation chromatography should require estimation of the degree of contribution by peak I radioactivity. attempting to perform Scatchard analysis) in which it becomes essential to know the amount of C3H]TCDD associated with peak 1. Peak I represents approximately one-fourth of the total [3H]TCDD added in vitro to cytosol concentrations of 5 and 10 mg of protein/ml (Fig. 3A). At 15 mg of protein/ ml, however, peak I comprised almost 70% of the total [3H]-TCDD added in vitro. Although peak I1 increased in proportion to protein concentration, peak I11 was markedly diminished at 15 mg/ml. Under conditions when peak I is greatest (i.e. high protein concentration) and therefore "free" [3H]-TCDD would be lowest, relatively little binding to the peak I11 moiety occurred. It is thus concluded that peak I11 material has lower affinity for TCDD than peak I1 material. These data illustrate why there has been so much difficulty with Scatchard analysis by either the dextran-charcoal adsorption method (8, 9) or the sucrose density gradient method (12).
[3H]TCDD binding in the 5.0 S region of sucrose density gradients (peak III material) was quite low at both high and low cytosolic protein concentrations (Fig. 3B), in contrast to what was seen for peak I11 with gel permeation chromatography. Both the gel permeation chromatography and velocity sedimentation methods measure ligand binding under nonequilibrium conditions. The time required for analysis on sucrose density gradients (about 18 h) is considerably longer than that for gel permeation chromatography (about 6 h).
The quantitative differences in peak I11 in Fig. 3 between A and B are therefore probably related to dissociation of ligand ("TCDD off-rate") from the peak 111 macromolecule(s) as a function of time after the in vitro [3H]TCDD treatment. The large proportion of radioligand associated with peak I has made it impossible for us to study accurately the kinetics of [3H]TCDD binding to the peak I1 and peak I11 moieties. The partial saturability of peak 111 may reflect heterogeneity, as well as a large number of binding sites. Estimation ofSize-The approximate molecular weights of the peak I1 and peak 111 moieties were calculated from their respective Stokes radii and sedimentation coeffkients ( Table   11). The peak I1 component, believed to represent the Ah receptor, appears to be an asymmetric molecule ( M , 2 245,000, f/fo P 1.6). The peak I11 material is presumably more symmetric (M, = 87,000; f/fo = 1.2).
Comparison of Gel Permeation Chromatography with Velocity Sedimentation: [3HJ3-MethyZchoZanthrene Binding to B6 and 0 2 Cytosol-Various properties of the peak 111 material in mouse liver caused us to wonder if this were analogous to the 3-methylcholanthrene-binding protein in rat liver recently reported by Fuller et al. (18) and Tierney et al. (15). Three major peaks were detected in B6 cytosol by gel permeation chromatography (Fig. 4A), and the positions of elution of the [3H]3-methylcholanthrene-binding peaks were essentially the same as those of the ["HITCDD-binding peaks.
With sucrose density gradient analysis (Fig. 5) ['H]3-methylcholanthrene eluted in the positions of the peak I1 and peak 1 1 1 material, as had been seen for r3H]TCDD (Fig. IS). At either 1 or 10 nM [3H]3-methylcholanthrene, 100-fold excess concentrations of nonlabeled 3-methylcholanthrene decreased peak I1 almost completely while decreasing peak I11 less than 50%. These data suggest that the saturable binding for ["H]3methylcholanthrene is principally in peak I1 but that some saturable binding sites may exist in the peak 111 material.
Effect of Trypsin and Detergents-Peak I1 was entirely destroyed by trypsin treatment under conditions that did not affect peak I or peak I11 to any substantial degree (Fig. 6).
The detergents sodium deoxycholate and NP-40 abolished binding to peak 11, but the [3H]TCDD was then eluted in the void fractions of the Sephacryl S-300 column after detergent treatment, z.e. as an increase in peak I (data not illustrated).
TCDD and other polycyclic aromatic inducers are extremely hydrophobic and insoluble, and it seems plausible that cyto-  solic components such as proteins or lipids could form large micellar structures that might include these ligands.

Genetic Analysis of [3H]3-Methylcholanthrene and [3H]-TCDD
Binding to Liver Cytosol-Ah-responsive B6 (Ahb/ Ahh) and the heterozygote (Ahb/Ahd) possess levels of receptor that are sufficient for the induction of cytochrome PI-450 and its associated aryl hydrocarbon hydroxylase activity, whereas Ah-nonresponsive D2 (Ahd/Ahd) and the Ahd/Ahd homozygotes as children of the B6D2F1 X D2 backcross do not exhibit detectable levels of cytosolic receptor (1,12). Because B6 and D2 mice represent very different inbred strains, we chose to study the relationship of peak I1 and peak I11 [3H]3-methylcholanthrene binding as a function of the Ahb allele.
Individual offspring from the B6D2F1 X D2 backcross were phenotyped 10 days earlier by the zoxazolamine paralysis test (16) and therefore classified as Ah-responsive Ahb/Ahd or Ah-nonresponsive Ahd/Ahd individuals (Fig. 7). The responsive heterozygote exhibited both peak I1 and peak I11 [3H]3methylcholanthrene binding, whereas the nonresponsive homozygote demonstrated large peak I11 [3H]3-methylcholanthrene binding only. With [3H]TCDD, peak I1 was prominent in the Ahb/Ahd mouse and absent in the Ahd/Ahd mouse, and peak 111 was negligible in both; this result is consistent with that previously reported (12). We therefore conclude that this [3H]3-methylcholanthrene binding moiety of peak I11 is not associated with the Ah receptor, whereas the peak I1 binding correlates well.
Analysis of Peak I1 and Peak III f3H]3-MethylchoZanthrene Binding by Anion-Exchange Chromatography-In all the above experiments, peak I1 and peak I11 have been separated on the basis of their hydrodynamic properties. Anionexchange chromatography also can resolve [3H]TCDD and After the column was washed with 100 ml of HEDG buffer containing 50 mM NaC1, the proteins were eluted with a linear NaCl gradient and 1-ml fractions were analyzed for radioactivity, protein content (Am), and conductivity. Further details are described under "Experimental Procedures." [3H]3-methylcholanthrene binding into distinct components. Okey et al. (12) previously showed that anion-exchange chromatography resolved [3H]TCDD binding into a single major component that eluted a t concentrations between 0.16 and 0.20 M NaC1; those experiments were carried out under conditions (high cytosolic protein concentrations), however, that would lead to minimal [3H]TCDD binding to peak I11 material. With changes in the experimental conditions (Fig. 8), we were able to separate clearly peak I1 and peak I11 [3H]3methylcholanthrene binding: peak I1 was eluted at concentrations of 0.16 to 0.20 M NaCl and was found only in the Ahresponsive B6 cytosol; peak I11 was eluted at lower concentrations of NaCl (about 0.10 M NaC1) and was found to be equally large in both the Ah-responsive B6 and the Ah-nonresponsive D2 cytosol.

DISCUSSION
With the use of gel permeation chromatography, we have resolved the apparent discrepancies between the dextrancharcoal adsorption assay (7)(8)(9) and the sucrose density gradient assay (12,13) for the Ah receptor. If saturating concentrations of [3H]TCDD are used, the sucrose density gradient assay provides an accurate measure of [3H]TCDD binding to the peak I1 moiety. Using this assay, we have shown that binding is saturable and proportional to protein concentration over a wide range of experimental conditions. The aggregated material (peak I) does not correlate with the Ah locus. With the sucrose density gradient assay, this material sediments to the bottom of the tube and therefore does not interfere with determination of Ah receptor levels (peak 11). We have also shown in this report that the major 3-methylcholanthrene binding moiety (peak I11 in this report) described in rat liver cytosol (15,18) is not associated with the Ah receptor and the cytochrome P1-450 induction process.
Gel permeation chromatography and velocity sedimentation techniques both involve significant departures from equilibrium binding conditions; these methods can therefore be used only for estimates of parameters such as dissociation constants and numbers of binding sites. Both these techniques, however, are able to separate distinctly the major TCDD-and 3-methylcholanthrene-binding components. Peak I1 appears to represent the Ah receptor, with a size of about 9.0 S in low ionic strength, high affinity for TCDD, and saturability at TCDD concentrations greater than 1 nM. Peak I11 is much more prominent when 3-methylcholanthrene rather than TCDD is used as the ligand, is not associated with the Ah receptor, is not saturable, and represents at least 100 times more binding sites than peak 11. Although peak I11 is not destroyed by mild trypsin treatment that destroys peak I1 (Fig. 6), Bresnick and co-workers (15) have demonstrated that this moiety is destroyed by prolonged Pronase treatment. The binding reflected in both peak I1 and I11 must therefore represent protein. Peak I is some sort of large aggregate. At low cytosolic protein concentrations, the proportion of peak I binding is considerably less (Fig. 3) explaining the success with dextran-charcoal adsorption under certain experimental conditions (7)(8)(9).
TCDD, as the most potent inducer, exhibits a large peak I1 and small peak 111, whereas 3-methylcholanthrene binding is much more prominent in peak I11 than in peak 11. Benzo[a]pyrene, an inducer less potent than TCDD and perhaps equivalent in potency to 3-methylcholanthrene, showed binding characteristics similar to those of 3-methylcholanthrene on sucrose density gradients (12): a small but detectable peak in the Ah receptor range (peak 11) and a very large binding component of lower molecular weight (peak 111).
Polycyclic aromatic inducers are able to enhance aminopyrine N-demethylase (19), aniline hydroxylase, d-benzphet-amine N-demethylase, chlorcycline N-demethylase, ethylmorphine N-demethylase, and pentobarbital hydroxylase (20) activities in Ah-nonresponsive mice (usually 40% to 2-fold enhancement). Small increases in electrophoretic bands are detectable by sodium dodecyl sulfate-polyacrylamide gels of liver microsomes from Ah-nonresponsive mice.' Hepatic glutathione transferase induction (with l-chloro-2,4-dinitrobenzene dinitrobenzene as substrate) by 3-methylcholanthrene is also not associated with the Ah complex and occurs in some nonresponsive mice having no detectable Ah receptor (21). It is therefore possible that we are technically unable to measure the presence of some Ah receptor in these nonresponsive mice. Alternatively, some other receptor or a "nonreceptor" mech-anism3 might be responsible for this induction process by polycyclic aromatic compounds in Ah-nonresponsive mice. Also, peak 111 could be heterogeneous and might contain a component responsible for this process in Ah nonresponsive mice.
Although Ah-nonresponsive mice have no detectable Ah receptor, there is significant aryl hydrocarbon hydroxylase induction in mouse fetal cell cultures exposed to benzo[a]anthracene (23) and in numerous tissues in vivo when nonresponsive mice receive sufficient doses of TCDD (5). This difference in sensitivity (Ah-nonresponsive mice requiring 12 to 18 times the dose of inducer given to Ah-responsive mice in order to attain the same response (6)) implies that the Ah structural genes are intact and that some defect lies in the Ah regulatory genes, presumably the Ah receptor. This 12to 18fold difference in sensitivity cannot be readily explained by the binding data in this report, however, which show a complete absence of peak I1 in liver cytosol from Ah-nonresponsive mice. It is possible that a 12-to 18-fold decrease in affinity would result in dissociation of the inducer-receptor complex during gel permeation chromatography (which takes about 6 h) or other lengthy methods of analysis. It would be useful to be able to measure ligand binding under equilibrium conditions, but this does not appear to be possible in crude cytosol because of the large proportion of nonsaturable binding of inducer to the peak I aggregate.
Several recent studies have dealt with the interesting possibility that the TCDD-or 3-methylcholanthrene-binding moieties undergo translocation to the cell nucleus during the process of induction. Okey et al. (12)  Unknown nonreceptor mechanisms leading to increases in enzyme activity might include: (i) direct binding of chemicals to cytochromes P-450 or other membrane moieties; (ii) perturbation of various membrane components (e.g. changes in membrane fluidity); (iii) interaction of nonmetabolized parent drug, or its metabolites, directly with nucleic acids or proteins in the nucleus without the need of a cytosolic receptor; (iv) activation of pre-existing inactive enzyme protein; and (v) inhibition of degradation. In cultured hepatocytes from fetal rat liver (22), for example, aryl hydrocarbon hydroxylase activity can be stabilized, and even enhanced, in the presence of the inducer benz-[alanthracene or phenobarbital and in the almost complete absence of RNA and/or protein synthesis. with the use of sucrose density gradients, have been unable to c o n f i i , however, the in vitro nuclear translocation demonstrated by Greenlee and Poland (14), who used the dextrancharcoal adsorption method. Tierney et a2. (15) demonstrated that the "peak B" 3-methylcholanthrene-binding component (equivalent to peak I11 in this report) undergoes temperaturedependent translocation to the nucleus.
We believe the best approach to studying nuclear translocation and all other aspects of any polycyclic aromatic binding moiety is to separate and characterize the various components contributing to the overall binding. In this report we have demonstrated the usefulness of gel permeation chromatography and anion-exchange chromatography, combined with an appreciation of the best protein and salt concentrations to obtain maximal resolution.