Progesterone-binding Components of Chick Oviduct I. PRELIMINARY CHARACTERIZATION OE’ CYTOPLASMIC COMPONENTS*

Abstract Induction of the synthesis of the specific protein avidin by a single administration of progesterone has been demonstrated previously in chicks in vivo and in tissue minces and monolayer cultures of chick oviduct. To investigate the mechanism of induction, macromolecular components of oviduct cytoplasm which bind 3H-progesterone in vitro were isolated and characterized by sucrose gradient centrifugation, polyacrylamide gel electrophoresis, enzymatic digestion, and gel filtration on Agarose (Bio-Rad Laboratories, Richmond, California). The radioactive steroid in the isolated complex was identified as progesterone, not a metabolite, by paper chromatography. The interaction with 3H-progesterone has an apparent dissociation constant kd ≃ 8 x 10-10 m in 0.3 m KCl at 1° and is reversed by mild heating and by unlabeled progesterone g testosterone g 20α-hydroxy-4-pregnene-3-one g 17β-estradiol g cortisol g estrone g androstenedione. The participation of protein in the steroid-binding site was inferred from the destruction of the complex by 10-3 m p-hydroxymercuribenzoate and by Pronase, but not by ribo- or deoxyribonucleases. The apparent number and size of the cytoplasmic binding components vary with the concentration of KCl and the technique of isolation and detection. In the absence of KCl, the major components are characterized by sedimentation coefficients, s020,w, of about 5 S and 8 S, molecular weights of about 1.0 and 3.6 x 105 (estimated from the variation of electrophoretic mobility with gel concentration), and sufficiently large effective radii to be eluted in, or near, the void volume of columns of Agarose A-0.5m. Under the same conditions the corticosteroid-binding globulin (CBG) of chick plasma behaves as a single component with s020,w ≃ 3.7 S and mol wt ≃ 6.0 x 104. In solutions containing 0.3 m KCl, the cytoplasmic components and CBG all sediment at about the same rate but may be distinguished from each other by the distribution coefficients on Agarose A-0.5m. From the latter results, molecular Stokes radii of 55 and 63 A can be calculated for the cytoplasmic components, compared with 37 A for CBG. The progesterone-binding components of chick oviduct cytoplasm are thus distinguishable from CBG by all physicochemical methods tested. A functional role for these components in the induction of avidin synthesis by progesterone is supported by (a) the parallel order of effectiveness of various steroids in competing with progesterone binding and in potency as inducers, and (b) the analogous effects of treating the chicks with diethylstilbestrol on progesterone-binding activity and on avidin induction.

bution coefficients on Agarose A-0.5m. From the latter results, molecular Stokes radii of 55 and 63 A can be calculated for the cytoplasmic components, compared with 37 A for CBG. The progesterone-binding components of chick oviduct cytoplasm are thus distinguishable from CBG by all physicochemical methods tested.
A functional role for these components in the induction of avidin synthesis by progesterone is supported by (a) the parallel order of effectiveness of various steroids in competing with progesterone binding and in potency as inducers, and (b) the analogous effects of treating the chicks with diethylstilbestrol on progesterone-binding activity and on avidin induction.
Although the sequence of events which mediate the effect of a steroid hormone on its target tissue is not yet known, an early step is thought to be the interaction between the hormone and a specific macromolecular component of the target tissue (3)(4)(5). The distribution of such "receptor" molecules could determine tissue specificity, and the hormone-receptor complex could participate in the regulation of RNA and protein metabolism in the target tissue.
The receptor hypothesis is supported by recent results from several laboratories.
These studies include preliminary characterizations of receptor molecules for estrogens (5-S), aldosterone (9, lo), and androgens (11,12). In each case the target tissue has been shown to contain proteins that are unique to these tissues, are present in small concentrations, and are recognizable by the ability to bind the respective steroid specifically and with high affinity.
Estrogen binding to uterine target cells has been the subject of intensive study. The uptake and distribution of radioactively labeled estradiol by the uterus has been followed by cell fractionation (13) and by radioautographic techniques (8,14). The hormone appears to be bound to cytoplasmic macromolecules initially and then transported to the nucleus, where stimulation of RNA synthesis takes place (8). The affinities of the cyto-plasmic receptor for e&radio1 and several analogues correlate well with the respective biological activities (13,15). It therefore appears quite probable that the receptors play a central role in the mechanism of action of estrogens in the uterus.
Less is known about the binding and localization of progestins in target tissues. A progesterone-and cortisol-binding protein has been isolated from the uterus of castrated rats, but the investigators did not demonstrate its distinction from the corticosteroid-binding globulin of plasma or its intracellular origin, although simple plasma contamination was excluded (16). A substance capable of binding pregnenolone and progesterone has also been reperted to be present in the cytoplasm of rat prostate, but the progestins serve no known function in this gland (17). More recently, a specific progesterone-binding macromolecule has been demonstrated in the supernatant fraction of the corpus luteum of the pregnant cow, an organ which synthesizes and secretes progesterone, as distinguished from a target organ (18).
Since the study of steroid-receptor interactions may be an effective and necessary approach to the mechanism of hormone action, experiments were undertaken in the chick oviduct, a hormone-responsive system which is simpler and more specific than many now available.
In this organ, progesterone induces de nouo synthesis of a single, specific protein, avidin (19-21). We have reported elsewhere the subcellular distribution and metabolic fate of 3H-progesterone in the chick oviduct (22). After injection in viva, a major fraction of labeled steroid was detected in the cytoplasmic supernatant fraction as a macromolecular complex that did not dissociate on passage through a Sephadex G-200 column (22). It was thought that a cell-free system in vitro would permit more precise and easily controlled studies on progesterone binding.
In the experiments reported here, the hormone wa,s added directly to an aqueous extract of chick oviduct. MATERIALS Ah-D METHODS Preparation of Plasma and Cytoplasmic Supernata,nt Fractions Female Rhode Island Red chicks had received either no hormone treatment for 20 days, starting at 4 days old (unstimulated), or 5 mg of diethylstilbestrol in sesame oil subcutaneously for 20 days (on diethylstilbestrol), or 20 days of diethylstilbestrol injections, followed by 10 days without treatment (off diethylstilbestrol).
Chick plasma was prepared by cardiac puncture into heparinized syringes and centrifugation at 1,000 x g for 15 min. Excised organs (oviduct, lung, and spleen) were washed 10 min in NaCl solution, weighed, sliced, and homogenized (glass barrel, Teflon pestle) in batches of 2 to 2.4 g in 6 ml of 0.25 M sucrose, 0.01 M Tris, 1.5 mM EDTA, pH 7.4, at temperatures close to 0". Centrifugation for 15 min at 1,000 X g to remove the nuclear-endoplasmic reticulum-myofibrillar fraction was followed by centrifugation for 1 hour at 105,000 x g at 1" to obtain the cytoplasmic supernatant fraction (cytosol).
Protein concentrations in the cytosol preparations were determined by the Lowry method (23).
The samples of human corticosteroid-binding globulin consisted of unfractionated plasma from pregnant women who had been treated for 3 days with 8 mg of dexamethasone.
Enzymes and protein standards were obtained commercially with the exception of human fibrinogen (courtesy of Dr. John Finlayson, National Institutes of Health), human thyroglobulin Counting efficiency under these conditions was 41 y0 for nonaqueous samples and 34y0 for l-ml aqueous samples. The radioactive steroid extracted from the macromolecular complex was identified by descending paper chromatography in ligroin-methanol-water (100:90:10, v/v) (25). Sucrose Gradient Centrifugation (d6)-Linear gradients of 5 to 20% sucrose in various buffers were prepared with a Buchler triple outlet gradient mixer and Polystaltic pump.
Incubation of plasma or oviduct cytosol with the radioactive hormone for 20 to 45 min was found adequate, since binding also occurred during the first few hours of centrifugation.
Samples were diluted 1:3 immediately before layering onto cold gradients. Centrifugation was performed at 1" in the Spinco model L2-65B ultracentrifuge using the Spinco SW-65 rotor at 60,000 rpm (average force 257,000 x g) or the Spinco SW-56 rotor at 55,000 rpm (average force 297,000 x g). Fractions were collected from the gradient bottom by means of the Buchler piercing unit and drop counter.
Radioactivity was determined on O.l-to 0.2-ml aliquots of undiluted fractions or on l-ml aliquots of diluted fractions after measurement of absorbance in a Zeiss spectrophotometer.
Quenching of radioactivity did not vary significantly with the concentration of sucrose in the fractions. Low recoveries of radioactivity in certain experiments were shown to be due to failure to recover highly aggregated forms of the labeled receptor complex from the bottom of the centrifuge tube and to adsorption of steroids to the Tygon tubing in the gradient fractionation setup. Sedimentation coefficients of steroid-binding components were estimated by comparison with those of ovalbumin, human CBG, and catalase.
Heat and Enzymatic Release Studies-The complex formed by incubation of 3H-progesterone (4.7 X 10e8 M) with oviduct cytosol from chicks on diethylstilbestrol was isolated by centrifugation for 21 hours at 257,000 X g in sucrose gradients containing 0.3 M KCI. Gradient fractions comprising the major peak of bound counts were pooled.
illiquots of 0.4 ml were incubated for 30 min at 0" with (a) no addition, (b) 0.5 mg of Pronase, (c) 0.5 mg of bovine RNase, or (d) 0.5 mg of DNase plus 2 ~1 per ml of 1 M magnesium acetate; all enzymes were obtained from Worthington.
Another aliquot was incubated 30 min at 37". Treated samples were layered onto 4-ml columns of Sephadex G-50, eluted with 0.01 M Tris, 1.5 mM EDTA, pH 7.4, and collected into l-ml fractions, all at 4". The labeled steroid complex was eluted immediately after the void volume, whereas the steroid released by the treatment was eluted in the total volume of the column. Recovery of radioactivity through the procedure was about 80%. Polyacrylamide Gel Electrophoresis-Separation gels (4.2 x 0.6 cm) with total acrylamide concentrations cf 4 to 11% were prepared from acrylamide monomer containing 2% of the cross-linking agent methylene bis-acrylamide.
Photopolymerization (Buchler "Polyprep") was initiated by riboflavin with tetramethylethylenediamine as accelerator (27). Stacking gels (1.2 X 0.6 cm) containing a total acrylamide concentration of 3.125%, 20% methylene bis-acrylamide, were prepared, layered onto the separation gels, and photopolymerized similarly. The multiphasic buffer system was a modification to 0" and pH 10.2 of the Tris system of Davis (28). Duplicate samples containing 50 to 100 ~1 of cytosol or 5 ~1 of plasma were incubated with labeled hormone, mixed with an equal volume of 50% sucrose containing bromphenol blue dye, and layered onto the stacking gels. Following electrophoresis at 2 ma per gel at 0" (Buchler "Polyanalyst"), the separation gels were removed from the glass tubing, chilled to -2O", and sectioned transversely into l.l-mm slices with an "egg slicer" type of device (Earl Sandbek Specialized Medical Instrumentation, Baltimore, Maryland). The radioactive hormone in each slice was measured after extraction for 1 hour into a toluene-Liquifluor scintillation fluid (960:40, v/v). The molecular weight and net charge of each labeled complex was estimated from plots of the log of the mobilit'y of the complex relative to the dye front versus acrylamide concentration in the gel (29). Under the conditions used, bromphenol blue had a relative mobility, RF, of 1.0 so that the Rp of the standard and steroid-binding proteins could be calculated with reference to this dye (27).
Agarose Gel Filtration-All chromatographic procedures were performed at 14".
Agarose beads for gel filtration (Bio-Gel A-0.5m from Bio-Rad, Richmond, California) were packed over a shallow base of fine glass beads in columns (1.27 X 110 cm) and washed extensively with 0.01 M Tris, 1.5 IIIM EDTA, pH 7.4, containing no salt or 0.3 M KCl. Samples of plasma (0.2 ml) or cytosol (0.4 to 2.0 ml) were generally placed on a column after incubat,ion with 1 to 5 ~1 of lop5 M labeled steriod at 0" for 90 or 120 min. The optical density of eluate was monitored by a Beckman DB-G spectrophotometer and 10 inch recorder enroute to a Gilson fraction collector.
The flow rate was maintained at 12.2 to 13.4 ml per hour, and radioactivity was measured on aliquots of 5-min fractions.
Residual radioactivity was removed by replacing Tygon tubing and washing the columns with buffer at room temperature.
Among the chromatographic systems tested, Agarose A-0.5m permitted superior resolution among binding components and higher flow rates than Sephadex G-200.
Filtration experiments on polyacrylamide gels (Bio-Rad P-150) were abandoned because of considerable holdup of steroid beyond the elution volume of 3H-water. low concentration of steroid required (lOmE M), and a slow rate of release of the steroid at 1" is inferred from the persistence of binding when the complex sediments away from the free steroid at the meniscus.

Sucrose
The association reaction occurs either instantaneously or during the early hours of centrifugation; preliminary incubation of cytosol with labeled hormone does not increase the fraction of radioactivity bound.
In gradients containing 0.3 M KCl, the rate of sedimentation of the progesterone and cortisol bound by chick oviduct cytosol is comparable to that of human CBGl (Fig. 1). The reversibility of the steroid binding and the relative affinities of various steroids for the cytoplasmic components were established by a series of experiments like those shown in Fig. 2. A large sample of cytosol was incubated 2 hours with 3H-progesterone (2 to 5 X lo-* M), then split into aliquots which were reincubated with the competing steroid (4 x 1OV M) and centrifuged in 5 to 20% sucrose gradients containing 0.3 M KCI. The relative decrease in the peak of bound radioactivity suggests the following order of affinity under these conditions: progesterone > testosterone > 17~ethynyl-lQ-nortestosterone > 20oL-hydroxy-4-pregnene-3-one > 170.estradiol > cortisol > estrone > androstenedione. the free steroid concentration, kd is the dissociation constant, and (P) is the total concentration of binding sites in the cytosol. This equation is rigorously applicable only to equilibria between a ligand and a single class of noninteracting binding sites. Since the separation of free from bound steroid favors dissociation of the complex, and since gradient centrifugation in 0.3 M KC1 at available speeds does not resolve the several binding components revealed by other techniques, the following values serve only as first approximations: The assumption that the binding component is a protein with a molecular weight of 1.5 X lo5 (see Table III) and a single site of interaction with steroid leads to a calculated weight concentration of 2.8 pg per ml of cytosol. The binding molecules thus represent about 0.02'$& of the total protein (15 mg per ml) in this cytosol preparation.
This value for the relative concentration of the binding components is the same as that determined for the uterine estrogen receptors (6).
The concentration of KC1 has been shown to influence dra- k,, represents an average of the dissociat,ion constants of all binding sites in the 3.7 S peak.
matically the molecular size of the estrogen-binding components of rat and rabbit uterus (30) and calf endometrium (31). The analogous sensitivity of the progesterone-binding components of chick oviduct cytosol is apparent in the gradient centrifugation results in Fig. 4. In the absence of salt, there is significant binding to macromolecules with sedimentation coefficients, s$~, of about 5 S and 8 S. In the short period of centrifugation used in the illustrated experiment, the smaller component in cytosol is not clearly distinguished from CBG (s!,,, of human and presumably chick CBG = 3.79 S (32)).
Its sedimentation coefficient was determined to be 5 S by longer centrifugationof samples to which crystalline ovalbumin (s~o,~ = 3.67 S (33)) was added as an internal marker and was visualized as a peak of absorbance. The value of .s&,~ for the faster peak was determined by shorter centrifugation with catalase standards (s&~ = 11.2 S (34)) run in parallel gradients not containing EDTA.
The 8 S peak appears from its shape to encompass more than one component.
Since there is no progesterone-binding component of comparable size in the plasma of the same diethylstilbestroltreated chicks, the 8 S peak must represent one or more intracellular progesterone-binding components. In the presence of 0.3 M KCl, the 3H-progesterone bound by oviduct cytosol sediments as a single peak at the same rate as CBG and as crystalline ovalbumin added as an internal marker The sedimentation coefficient of the radioactive complex in 0.3 M KC1 is therefore given as 3.7 S, with the following reservations.
(a) The 105,000 x g supernatant of cytoplasm is a crude preparation, in which the sedimentation rate of the progesterone complex may be altered by chemical or hydrodynamic interactions with other constituents of the mixture; (b) calculation of s&J,~ by comparison with other macromolecules requires that the unknown and standards have the same partial specific volume (ti).
To ascertain whether fi of the progesterone-binding components differs significantly from that of ovalbumin, centrifugation components of oviduct cytosol and plasma. Centrifugation at 297,000 X g was stopped after 15 hours for gradients containing KCl, and after 10 hours for gradients without KCl, to reveal both peaks of steroid bound by cytosol (~20,~ N_ 5 and 8 S). Crystalline ovalbumin (absorbance peak) was added to one sample as an internal reference for the sedimentation coefficient.
of labeled cytosol mixed with crystalline ovalbumin was performed in gradients of various densities.
For example, 5 to 20% sucrose gradients were prepared in buffered solutions containing 25% (w/v) NaBr, to obtain gradients with densities (p) at 1" of 1.2024 to 1.2515 g per cm3 (35). These conditions were selected so that ovalbumin, having fi = 0.748 cm3 per g (36) or p = 1.336 g per cm3, would sediment slowly to the bottom of the tube, while even the most dense of lipoproteins (p = 1.21 g per cm3 (37)) would sediment through only the upper part of the gradient.
In gradients containing 40% NaBr (p = 1.3134 to 1.3658 g per cm3) ovalbumin would sediment through part of the gradient while all lipoproteins would float.
The experimental results for three gradients of widely separated densities are shown in Fig. 5. The similar amount of retardation of the binding components and ovalbumin (peak of absorbance) in each gradient indicates the similarity of their partial specific volumes.
This result justifies the use of ovalbumin as a standard for the evaluation of the sedimentation coefficient of the binding components and excludes the possibility that they contain significant amounts of lipid.
It should The indicated amounts of NaBr were included in 5 to 20% sucrose gradients in 0.3 M KC1 to provide linear gradients of density at 1" of 1.0585 to 1.1137, 1.2024 to 1.2515, and 1.3134 to 1.3658, respectively.
The similar amount of retardation of the binding components and ovalbumin (absorbance peak) indicates the similarity of their densities and the absence of significant lipid content in the binding components.
also be noted that the progesterone-binding components are neither precipitated nor denatured by these extremely high salt concentrations.
Other gradient centrifugation experiments showed that oviduct cytosol from chicks receiving diethylstilbestrol continuously and from chicks withdrawn from treatment for 10 days had similar patterns of progesterone binding; the concentration of sites in the cytosol from chicks off diethylstilbestrol has not been determined.
The binding activity was also shown to be diminished but readily detectable in cytosol preparations stored frozen for 5 months.
Polyacrylamide Gel ElectrophoresisAs in sucrose gradient centrifugation in the absence of KC1 (Fig. 4), two major peaks of bound 3H-progesterone are revealed by electrophoresis of oviduct cytosol on polyacrylamide gels at pH 10.2 and 0" (Fig. 6). The major progesterone-binding molecules of cytosol (I and 11) migrate more slowly under these conditions than the minor progesterone-binding component which also is the major cortisolbinding component in the cytosol (III).
The latter peak is identical in relative mobility, RF, with the progesterone-and cortisol-binding component of chick plasma, and presumably represents a small amount of CBG in the cytosol preparations (cf. Fig. 9). Unlabeled progesterone (lo-' M) completely abolished the binding of 10mg M 3H-progesterone.
The same concentration of unlabeled cortisol eliminated only Peak III from the progesterone-labeled electrophoretic pattern. To distinguish between the contributions of net charge and molecular size to the electrophoretic mobility, experiments similar to those in Fig. 6 were carried out in gels of different total acrylamide concentration all cross-linked to the same extent (38). Fig. 7 shows the variation with total acrylamide concentration of log RF for the major progesterone-binding components of oviduct cytosol and plasma.
From these results, estimates of the molecular weight and net charge of each labeled complex were obtained by applying the method and computer programs of Rodbard and Chrambach (27,29).
The first step was to calculate the retardation coefficient (Es), which is the negative slope of the weighted least square linear  Fig. 6).
In this plot the slopes depend only on molecular size, whereas intercepts at 0% gel reflect the free electrophoretic mobility (27).
regression through the data for log RF versus total acrylamide concentration, from the results in Fig. 7   CBG is consistent with the values obtained by Agarose gel filtration (see below). The next step in the analysis of the data in Fig. 7 was to extrapolate the lines for log RF versus total acrylamide concentrations to 0% gel concentration to obtain the free solution mobility and hence the net charge on each component under the given conditions (27). The results summarized in Table I confirm the  distinction between the cytoplasmic binding components and CBG with respect to both molecular weight and net charge to mass ratio under these conditions.
The observations that the charge to mass ratio for Components I and II is the same (within the experimental uncertainty) and that the binding components tend to aggregate under other conditions suggest that Component I is an aggregate of Component II. Given the computed values of molecular weight (Table  I) and the knowledge that most regulatory proteins have the structures of symmetrical oligomers, we would tentatively propose that the correct molecular weight of Component II is about 9 x lo4 and that Component I is a tetramer of II. Furthermore, comparison of these estimates with the sedimentation coefficients of the binding components in the absence of KC1 (but at pH 7. The relative amounts of A1 and Az are constant for a given pool of cytosol but vary slightly among different preparations. The A components are eluted in a region of low optical density at 280 rnp, indicative of a larger molecular size than most of the cytoplasmic constituents (see "Physical Parameters of the Binding Components in Presence of 0.3 M KCl").
The next peak of binding activity (B) is tentatively identified as chick CBG by the superposition of chromatographs of progesterone-labeled oviduct cytosol and chick plasma in Fig. 9. By contrast, the possibility that the A components are artifactual aggregates of CBG is negated by the complete absence of bound progesterone in the A region after filtration of labeled plasma or refiltration of the isolated B peak.
Treatment of the chicks with the synthetic estrogen diethylstilbestrol affects both the amount of 3H-progesterone bound and the distribution of proteins in the oviduct cytosol. As shown in Fig. 10, the labeled macromolecular complex is just detectable in gel filtration Fractions 55 to 70 of cytosol from unstimulated chicks (no diethylstilbestrol), compared with extensive binding in these fractions of cytosol from chicks on diethylstilbestrol. This effect of continuous estrogen treatment may reflect an increased number of binding molecules, analogous to the increase of ovalbumin (absorbance peak in Fraction 82). Alternatively, diethylstilbestrol~ treatment may increase the affinity of these components for progesterone.
Withdrawal of estrogen for 10 days before sacrifice (chicks off diethylstilbestrol) does not significantly change the pattern of progesterone binding (cf. Fig. 8) although the optical density pattern of the eluate reverts toward that of cytosol from unstimulated chicks. The concentration of sites in cytosol from chicks off diethylstilbestrol has not be determined.
If the progesterone-binding molecules of oviduct cytoplasm participate in the specific response of the organ to the hormone, the concentration of these components in nontarget organs is expected to be low or negligible.
This prediction is borne out  by the gel filtration patterns of progesterone-labeled lung and spleen cytosols from chicks on diethylstilbestrol (Fig. 11). In both organs the interaction of the hormone with materials which are excluded from the gel exceeds the binding in the region of the major oviduct components (cf. Fig. 8 of lung cytosol, is identical in elution position with chick CBG (cf. Fig. 9). The change in molecular size of the binding components as a function of KC1 concentration in centrifugation experiments (Fig. 4) was corroborated by the gel filtration experiments in Fig. 12 Other experiments were designed to determine whether the overlapping peaks (A1 and AZ) observed in the presence of KC1 represent interconvertible components which would re-equilibrate after isolation.
From the chromatogram shown in Fig. 12  Identification of Bound Radioactivity-In earlier experiments on the incubation of chick oviduct minces with 3Hprogesterone at 37", after 1 hour more than half the radioactivity in the cytosol was present as metabolites of progesterone, including testosterone and pregnenediol (22). It was therefore important to analyze the steroid in the radioactive complex formed under the present conditions: incubation of the cytosol fraction of homogenized oviduct with 3H-progesterone at 0". Following filtration of labeled cytosol on Agarose in 0.3 M KC1 buffer, the eluate fractions containing the major binding component (Peak A, Fig. 8) were pooled and extracted with ether. The extracted steroid was chromatographed on paper with the Bush A-l solvent system (25) which resolves progesterone from the expected metabolites, including 20a-and 20&hydroxy-4pregnene-3-one, testosterone, 170c-hydroxyprogesterone, and pregnenediol (22). Of the recovered counts, 92% co-chromatographed with the progesterone standard, indicating that very little metabolism of bound progesterone occurred during the present, incubation and isolation procedures.
Protein Nature of Steroid-binding Site-Among the treatments used to explore the chemical nature of the progesterone-binding components were (a) brief heating of the cytosol before addition There is little binding in spleen and significant binding activity in lung only in Fractions 74 to 81, where the plasma component is eluted (cf. Fig. 9).  Fig. 4).
of the hormone, (b) inclusion of p-chloromercuribenzoate during incubation with the hormone, (c) mild heating of the isolated macromolecular hormone complex, and (d) enzymatic digestion of the complex.
Heating of the cytosol to 60" for 2 min prevented subsequent binding of 3H-progesterone except for limited interaction with aggregates excluded from Agarose A-0.5m in the presence of 0.3 M KCl.
Addition of p-chloromercuribenzoate to a final concentration of 1 mM in cytosol likewise eliminated all binding except barely detectable amounts to material in the excluded volume of the column.
This result indicates not only the protein nature of the hormone-binding site, but also the involvement of sulfhydryl groups in either the interaction with the ligand or the maintenance of the active structure of the binding component.
The integrity of the macromolecular steroid complex isolated by sucrose gradient centrifugation and collection in the cold is demonstrable by gel filtration on Agarose (Fig. 13) or on Sephadex G-50. The pattern of elution from Sephadex G-50 (not illustrated) contains the bound hormone in the excluded volume and the free steroid in the total column volume.
This sharp resolution of bound from free hormone by short Sephadex columns was used to measure the 3H-progesterone released from the isolated complex by mild heat, Pronase, and nucleases. The results in Table II  Molecular size may also be estimated from the Agarose gel filtration data. Distribution coefficients of the peaks of bound radioactivity, K. were calculated according to the standard formula where 'CT, is the elution volume of the component, V, is the void volume of the column, and Vt is the total liquid volume, here approximated by the elution volume of free steroid. Apparent molecular weights of steroid-binding components were estimated by assuming a linear dependence of mol w@ on Kg1'3) using as standards, ovalbumin, bovine serum albumin monomer, and BSA dimer (39). Results for the major progesterone-binding components (A1 and AJ and Component B (cf. Figs. 8, 12) are given in the last column of Table III.
These values of apparent molecular weights correspond to true molecular weights only if the binding molecules have the same 8, shape, and degree of solvation as the globular protein standards.
Since the latter two factors are unknown for the cytosol components, the more valid parameter which may be determined from the chromatograms is the effective molecular radius a (Stokes raidus).
For the binding components listed in Table III  The effective pore radius (r) of the Agarose was calculated to be 192 A from the observed distribution coefficient and the published radius of BSA monomer.
The Stokes radii of the binding components were used in turn to calculate the diffusion coefficients given in Table III, from D = KT/6rva in which K is the Boltzmann constant, T is absolute temperature, and 7 is the solvent viscosity.
The apparent discrepancy between the slow sedimentation rate (e) Different states of aggregation of the same components are being observed.
The first hypothesis is ruled out by the results of centrifugation experiments in high density gradients (Fig. 5). The alteration of the molecular size by sucrose was disproven by equilibrating the column and eluting with buffered 0.3 M KC1 containing 5% sucrose. The amount and elution positions of bound 3H-progesterone were unaffected by the sucrose.
The third hypothesis was pursued by combining the results for the sedimentation coefficient (s) and the Stokes radius (a) to calculate the molecular weight (1M) and the corresponding frictional ratio, j/j0 (41,42) : where N is Avogadro's number, p is the solvent density in grams per cm3, and 6 denotes the solvation. When ti and 6 were assigned values which are typical for proteins (6 = 0.734 cm3 per g, 6 = 0.2 g of solvent per g of protein (38,42)), the estimates of f/f0 and M in Table III were obtained. For Component B the value found for j/j0 is normal for globular proteins, and the estimates of M from the graph of K&3 versus Ml/z and from the combination of s and a are in reasonable agreement (Table  III) and are wit.hin the range reported for rat CBG (43). The results for j/j0 of Components A1 and Aa correspond to prolate ellipsoids with axial ratios of 18 and 14, respectively (44). Still higher axial ratios are calculated (22 for Al, 17 for Az) if solvation is neglected.
Alternatively, the frictional ratio of the A components may be assumed to be in the normal range, say 1.18, and the amounts of solvation required to account for the large Stokes radii may be calculated.
The resultant value for Al of 6 = 3.2 g of solvent per g of solute is an order of magnitude higher than those usually observed and excludes this explanation.
To ascertain whether the same binding components are in fact revealed by centrifugation and gel filtration in 0.3 M KCl, the fractions of a sucrose gradient containing the peak of bound 3H-progeste.rone were chromatographed on the Agarose column. The recording of optical density in Fig. 13 shows that most of the protein in the 3.7 S peak is eluted from the column in Fractions 88 to 96, as expected for globular proteins with molecular weights of 42,000 to 48,000. By contrast, the macromolecular complex containing the radioactive steroid displays the characteristic elution pattern of the A components, corresponding to much larger molecules. This demonstration of the A peak in chromatograms of the isolated 3.7 S peak excludes the possibility that the components observed by the two techniques are unrelated.
The ensemble of results is compatible with the identity of the major components detected by centrifugation and gel filtration in the presence of KCl, if the components are rather asymmetric, or with the assumption that the A peak of the chromatograms represents specific aggregates of the 3.7 S components.
Regardless of its molecular basis, this unique combination of a low sedimentation coefficient and large molecular size should greatly facilitate the isolation of the A components from other cytosol constituents.

DISCUSSION
The presence of specific progesterone-binding components in oviduct cytoplasm frcm estrogen-treated chicks has been demonstrated by ultracentrifugal, electrophoretic, and chromatographic techniques. These cytoplasmic components are readily distinguished from the progesterone-binding component in the plasma of the same chicks by all techniques utilized.
The cytoplasmic components have significantly higher affinity for progesterone than for cortisol, with an estimated kd for progesterone of about 8 X lo-i0 M in 0.3 M KC1 at 1-4". This value is remarkably close to that determined by the same technique for the interaction of estradiol with its uterine receptors in rat, kd N 7 X lo-i0 M (6). It is likely, however, that the initial estimate of kd for the progest,erone-binding component will be revised by equilibrium studies on less crude preparations. While several other workers have reported data which implicate progesterone-binding to macromolecules in animal cells, the components described previously do not seem to fulfill the predictions for a specific target tissue progesterone "receptor". The molecule studied by Milgrom and Baulieu (16)  The pregnenolone-and progesteronebinding molecule described by Karsznia et al. (17) was isolated from the prostate gland, in which the progestins serve no known function, and differs from other reported receptors in its stability and enhanced affinity for steroids at 60". Leymarie and Gueriguian (18) have achieved the chromatographic separation of a specific progesterone-binding component from all cortisol-binding activity in the corpus luteum of the pregnant cow. However, the function of the binding components in an organ which synthesizes and secretes the hormone is likely to differ from the "receptor" function in a target organ. The data reported here thus represent the initial isolation and partial characterization of specific target tissue receptors for progesterone which are not present in the plasma or nontarget tissues.
The progesterone-binding components of chick oviduct cytoplasm exhibit various states of aggregation or asymmetry as a function of the KC1 concentration and the technique of isolation and detection.
In the absence of KCI, the major components are characterized by sedimentation coefficients of about 5 S and 8 S, which for globular proteins correspond to molecular weights of 7.0 to 8.6 x lo4 and 1.5 to 1.8 X 105, respectively, suggesting a monomer-dimer relationship (34). From the variation of electrophoretic mobility with acrylamide gel concentration and the extrapolated value for the free solution mobility of the major components, computer analysis gave molecular weights of 1.0 (or 0.9) x lo5 and 3.6 X 106, respectively, and the same charge to mass ratio for both components, consistent with a monomer-tetramer relationship. Still higher states of aggregation are inferred from the elution pattern from Agarose A-0.5m in solutions of low ionic strength.
By contrast, in solutions containing 0.3 M KCl, the macromolecular progesterone complex appears to sediment as a single peak with s&,~ N 3.7 S, corresponding to mo1ecuIar weight of 4.2 to 4.8 x lo4 for globular proteins, or about half the weight of the smaller component observed in the absence of KCl. Agarose gel filtration in the same solvent of the isolated 3.7 S peak partially resolves two major components of bound progesterone with Stokes radii of 55 and 63 A, respectively.
As in the studies without KCI, the apparent size on gel filtration is larger than that expected from the centrifugation results. These observations may be reconciled by postulating that specific aggregates of the steroid-macromolecular complex are formed during gel filtration and gel electrophoresis or that the binding components are extremely asymmetric, with axial ratios in the range of 14 to 22. Further experiments are required to test the proposed relationships among the forms of the components detected in these preliminary studies.
The abilities of various steroids to compete with the binding of 3H-progesterone in the presence of KC1 correlate well with their respective capacities to induce avidin synthesis in Z&O (measured as micrograms of avidin per g of oviduct, 20 hours after injection of 5 mg of steroid (22)). The most active competitors of binding, 20a-hydroxy-4-pregnene-3-one, testosterone, and 17or-ethynyl-19.nortestosterone, are all active inducers. Conversely, cortisol and androstenedione, -which are poor binding competitors, show no inductive activity.
These results, coupled with the parallel effects of diethylstilbestrol treatment on progesterone-binding activity and on avidin induction, suggest that the binding components play a functional role as the receptor in the induction process. At present, we have no chemical data upon which to construct a solid hypothesis for the sequence of molecular events in steroid hormone action.
An initial interaction of a steroid with a cytoplasmic receptor, followed by activation of the transcriptional apparatus to produce new RNA and finally new protein, is a popular concept (22,45,46). However, not only is there no consensus on whether the important regulatory step is at the level of transcription or translation (46), but there is no definitive evidence for a functional role for the target tissue receptors. Jensen et al. (47) have attributed the difference between the 5 S form of the estrogen receptor from uterine nuclei and the 4 S form of the cytoplasmic receptor in 0.3 M KC1 to the addition of a nuclear constituent related to the regulatory function of the molecule.
Raynaud-Jammet and Baulieu (48) have reported that incubation of endometrial nuclei with uterine cytoplasm permitted an estrogen-mediated stimulation of CTP incorporation into nuclear RNA, possibly through a functional cytoplasmic estrogen-receptor complex. The independent confirmation of these provocative experiments using the progesterone-binding components of oviduct cytoplasm and nuclei (49) may allow us to delineate a functional role for the receptor and define the biochemistry of the interaction of the hormonereceptor complex with the cell nucleus.

Acknozoledgments-We
are grateful for the invaluable cooperation of Dr. Karl Piez on the chromatographic techniques, Dr. Joram Piatigorsky on the gradient centrifugation experiments, and Drs. Andreas Chrambach and David Rodbard on the analysis by disc gel electrophoresis.
Mrs. Patricia Middleton and Mrs. Catherine Sullivan provided capable technical assistance.