Liver Cytosol Corticosteroid Binder IB, a New Binding Protein*

A new binding protein named corticosteroid Binder IB elutes just after ligandin in DEAE-Sephadex chromatograms. It has been partially purified to about 2500-fold over cytosol proteins. Calculation of the number of steroid binding sites, assuming one site per molecule of Binder IB fraction after DEAE-Sephadex chromatography, suggests a concentration of the binding protein of about 0.0004% of cytosol proteins. Its p1 value is judged to be 7.5 to 8 its elution on DEAE-Sephadex chromatograms. of 30,500 f 10’70 by filtration and a Stokes radius of 20 A. Binder cortisol, and corticosterone in oitro with estimated K, values of 1, 13. and 25 nM, specificity

A new binding protein named corticosteroid Binder IB elutes just after ligandin in DEAE-Sephadex chromatograms.
It has been partially purified to about 2500-fold over cytosol proteins. Calculation of the number of steroid binding sites, assuming one site per molecule of Binder IB fraction after DEAE-Sephadex chromatography, suggests a concentration of the binding protein of about 0.0004% of cytosol proteins. Its p1 value is judged to be 7.5 to 8 from its elution position on DEAE-Sephadex chromatograms. IB has an apparent molecular weight of 30,500 f 10'70 by gel filtration and a Stokes radius of 20 A. Binder IB binds radioactive dexamethasone, cortisol, and corticosterone in oitro with estimated K, values of 1, 13. and 25 nM, respectively. Saturation curves are abnormal, showing two phases. The saturation curves within the physiological range of concentrations of steroid are abnormal and suggestive of cooperativity. The second phase, at concentrations of glucocorticoids above saturation and physiological levels, shows extensive binding. After fractionation from other steroid binding proteins, the specificity of binding from competition studies in vitro is dexamethasone 2 cortisol = corticosterone = estradiol-17P > deoxycorticosterone = dihydrotestosterone > aldosterone = cortexolone > testosterone. Other steroids tested are less efficient ligands. The binding is probably noncovalent, but strong; and the complex becomes more dissociable as purification proceeds, suggesting a conformational change in the protein. Storage and rebinding with steroid are possible throughout the purification process, although extensive ligand dissociation and denaturation of the protein occur after the final purification step. Binding in uitro is temperature-sensitive and binding is sharply pH dependent with an optimum at 7.5. The ligand is the unmetabolized steroid as judged by extraction of steroid-IB complex with methylene chloride and subsequent thin layer chromatography.
The physiological function of this protein is unknown at present and purification of the major corticosteroid hormone receptor to homogeneity may be required before the function of Binder IB is fully understood.
Previously, we have reported the presence of four corticosteroid binding proteins in adrenalectomized rat liver cytosol separable after intraperitoneal administration of radioactive corticosteroid (1,2). Recently, we also noted the occurrence of another binding protein, designated corticosteroid binder IB, which elutes just after ligandin (Binder I) in DEAE-Sephadex A-50 column chromatograms (3. 4 Steroid Extraction-To extract the steroid from the in uiuo-labeled IB from DEAE-Sephadex A-50, equal volumes of cold methylene chloride and sample were added to a conical, glass-stoppered test tube. This mixture was mixed in a vortex for 5 min, centrifuged for 5 min at 700 x g, and the organic layer was removed. The sample was extracted twice more and the pooled extracts were concentrated in a stream of N, to about 0.5 ml. An aliquot was applied to a precoated silica gel plate and developed in benzene/acetone (l/l), and in a second determination, methylene chloride/acetone/methanol (7/2/l). A-50 was obtained using potassium phosphate and Tris-HCI buffers at concentrations of 50 mM. The routine binding assay was followed as described above except that 5-ml aliquots of Binder IB pool were incubated with 1 ml of the appropriate buffer at the prescribed pH. All incubations were carried out at 4O. Buffers of the same pH were used to elute the column. Protein Concentrations-These were determined from absorbance at 280 nm and by the method of Lowry et al. (10).  Fig. 1. Binder IB is resolved from ligandin (Binder I) and is eluted after it. This procedure was used for the purification of the steroid-protein complex and for preparation of uncomplexed binder for in vitro experiments.

Molarity
By virtue of its elution position, we conclude that the p1 value of IB is between pH 7 to 8, since ligandin has a p1 value of 8.9 determined by electrofocusing (14,15 in Fig. 2. The protein concentration in this pool is usually 1 to 1.6 mg/ml. The results of specificity of binding are summarized in Table I. Dexamethasone appears to be the favored steroid of those tested, followed closely by the natural glucocorticoids.
Curiously, 17@-estradiol competes with radio- active corticosterone binding as well as the glucocorticoids, and dihydrotestosterone is a more efficient ligand than testosterone. This specificity of binding is somewhat similar to that of the hormone receptor (Binder II) (3). It also appears to be different from transcortin (16). Although the specificity of IB towards the functional glucocorticoids might appear to be lower than expected is when the ability of anti-glucocorticoids, such as estradiol, testosterone, and cortisone is examined, the secondary increase in the saturation of binding curves (Fig. 3 would be expected not only to compete with the higherspecificity, low-capacity sites, but would also occupy some of the sites of the second phase of the saturation curve, making it appear that the potent glucocorticoids are only slightly more active than steroids which are anti-inducers.  Fig. 3. The initial portions of the saturation curves appear to be sigmoidal for the three steroids.
As the levels of the glucocorticoids are increased, the curves level off and at still higher levels the binding is greatly increased.
The large increases in binding at high levels of steroids may probably indicate a second set of low affinity, poorly saturable sites (17). The dissociation constants can be estimated roughly from the saturation phase of each curve to be: dexamethasone, about 1 nM; corticosterone, 25 nM, and cortisol about 13 nM. The complexity of these curves represents a significant difference from those observed with the hormone receptor (Binder II) (3), although it is possible that sigmoidicity might be a function of protein concentration and that supplementation with a nonsteroid binding protein might alter apparent sigmoidal behavior, as in the case with the chick oviduct progesterone receptor (18). Protein concentrations used here were between 1 to 4 mg/ml and were in the same range as similar studies with the hormone receptor which did not give sigmoidal curves (3). However, when ovalbumin, a protein which does not bind steroid (18), was added to aliquots of binder IB so that the total protein concentration was raised from 4 to 10 mg, the sigmoidal nature of the saturation curve with radioactive corticosterone was unchanged so that sigmoidicity is not a function of protein concentration within this range of protein concentration.
The data of Fig. 3 suggest, in addition to the data in Table I, that the same binding site is involved in the binding of the three steroids since, at saturation, 3 to 4 pmol of steroid are bound per mg of protein in each case.
Effects of Temperature and pH on Binding-Unlabeled Binder IB pool was used directly for temperature and pH studies as described above. To determine the effects of temperature, 5-ml portions of the IB pool were incubated with 25 nM [3H]corticosterone at the temperatures indicated in Fig.  4 for 90 min. Thereafter the incubation mixtures were rapidly cooled in ice and binding was measured. The resulting binding is obviously temperature-dependent as the inverse linear relationship between binding and increase of temperature of incubation shows in Fig. 4 between 4 and 45". At the highest temperature the association of steroid and protein is virtually nil.
The results of incubating Binder IB and radioactive corticosterone at different pH values are shown in Fig. 5. Five milliliters of Binder IB pool from DEAE-Sephadex chromatography were incubated with 25 nM [3H]corticosterone together with phosphate or Tris buffer so that the final buffer concentration was 50 mM at the pH specified in Fig. 5 graphed on a Sephadex G-100 column calibrated with proteins of known molecular weight to estimate apparent molecular weight. These data are shown in Fig. 6 where the apparent molecular weight is 30,500 * 10%. This experiment has been repeated 16 times with very good agreement with this value. In addition, the most highly purified preparation to be described below has been examined by gel filtration and the apparent molecular weight agrees with the value from the sample fractionated through the DEAE-Sephadex column step. By analysis of the gel filtration data (19), the Stokes radius of Binder IB is 20 A.
State of Ligand-At the stage of the DEAE-Sephadex step, only about 20% of the radioactivity from [3H]corticosterone is readily extracted into 4 equal successive volumes of cold methylene chloride. The radioactivity is shown to be identical with unmetabolized steroid, as shown in Fig. 7, using thin layer chromatography.
As the purification of IB proceeds, the complexed radioactivity strips on columns, so that the binding is judged to be noncovalent but strong. It appears, in terms of dissociation during column chromatography, to be a much stronger binding than that of the hormone receptor. This conclusion that IB binds the unmetabolized hormone aligns well with the studies presented in Table I G-25 column ("Materials and Methods").
The bound pool is chromatographed on a column of DEAE-Sephadex A-50 and the labeled pool eluting just after ligandin is collected (Fig. 1). This pool is lyophilized to achieve a 4-to 6-fold concentration of protein. During thawing a precipitate of nonspecific protein forms and is centrifuged off. The supernatant is chromatographed on a column of Sephadex G-100 (Fig. 8). The pool from Sephadex G-100 is chromatographed on a column of CM-Sephadex C-50 (Fig. 9) as a final step. The purification sequence is summarized in Table II. Since there are four other binding proteins in cytosol, specific radioactivity cannot be used as an index to purification of IB until it has been resolved from the other binding proteins. This is achieved at the DEAE-Sephadex chromatographic step and we assume quantitative recovery of bound steroid up to that point. In terms of protein fractionated away there is a 22-fold purification after the DEAE-Sephadex step. Precipitation of nonspecific protein after lyophilization and thawing, and chromatography on a column of Sephadex G-100 gives additional l7-fold purification. Subsequent chromatography on CM-Sephadex gives a further purification of 6-to 7-fold to give an over-all purification of about 2500.fold. If IB were homogeneous at this point, its concentration would be about 0.04% of cytosol proteins. Examination of purity by disc gel electrophoresis at each major fractionation step is shown in Fig. 10. The disc gel pattern of the 2500-fold purified material shows one major band and a faint minor band suggesting that IB is not yet homogeneous.
Similar results were obtained in six separate experiments.
Although the reduction in the number of bands in the disc gels as purification proceeds corresponds to the loss of protein during each purification step, the single major band in the final disc gel electrophoresis represents, undoubtedly, a major contaminant rather than IB. This is likely since at this stage direct calculation of the number of steroid binding sites in the IB pool from DEAE-Sephadex indicates that IB is about 0.0004% of the cytosol proteins under these conditions.
One 6803 tons) has been assumed, although the suggestion of cooperativity in the saturation curves may subject this assumption to later revision. This would represent about 50 the concentration of the hormone receptor (Binder II). Unfortunately, we have been unable to apply disc gel electrophoresis or isoelectrofocusing in sucrose gradients to this protein and maintain the steroid-protein complex up to the present time. The binder may be frozen after any of the chromatographic steps and still retain its binding capacity for steroid. The purified binder (CM-Sephadex fraction) is stable and may be stored frozen for several weeks without loss of radioactivity or binding capacity. Fractionation steps subsequent to CM-Sephadex reveal extensive dissociation of the steroid-protein complex and increased lability of the protein. The behavior of IB described above is very different from our experiences with the hormone receptor (3). DISCUSSION The physiological role of Binder IB is unknown and we may not have insight into this problem until the hormone receptor,  itself, is homogeneous.
However, in view of its apparent basicity, its elution position close to that of ligandin in DEAE-Sephadex chromatograms (Fig. 1) and its ability to bind dexamethasone in common with ligandin as well as the hormone receptor (Binder II), it seems possible that Binder IB may be related in some way to either ligandin or the receptor. This possibility will be explored directly at a later time when it may be possible to carry out very large scale experiments to provide sufficient homogenous IB for preparation of antibody and cross-reaction with the other binding macromolecules. Because ligandin (Binder I in Fig. 1) has recently been proposed to be identical with GSH S-alkyl transferase B, one of a group of glutathione transferases which catalyzes the transfer of GS-to the methyl group of methyl iodide or to other cosubstrates (20), IB could be an enzyme. Should it have such an activity, the effects of nonsubstrate ligands upon enzymatic reaction would be of utmost interest. Militating against this possibility, however, is the fact that these transferases uniformly appear to have molecular weights of 45,000 with two equal size subunits, whereas IB is about 30,000. It would be difficult to reconcile this property with ligandin or the GSH transferase enzymes. There is the possibility that IB may be one of two equally sized subunits of the hormone receptor (67,000 daltons) because of its apparent molecular weight, binding specificity, and apparent K, values for binding of dexamethasone, corticosterone, and cortisol. We have observed a second corticosteroid-bound macromolecule, in addition to receptor, in the nucleus in in vivo experiments with a p1 value by isoelectrofocusing of 7.5 (3). This opens the possibility that this second macromolecule may be IB and that it occurs in the nucleus because it may have some steroid receptor activity in its own right or it may derive from the receptor. Up to now, the progesterone receptor from chick oviduct has been shown to contain a steroid binding subunit which can be dissociated from the receptor by calcium ions (18). We have repeated similar experiments with Binder II which did not dissociate a steroid binding subunit either as measured by ion exchange chromatography or by gel filtration (Sephadex G-100) chromatography. This leads to the conclusion that if IB is the steroid binding subunit of the receptor, it is much more tightly bound than is the case with the chick oviduct progesterone receptor. Of interest to the association between steroid and IB is our observation that radioactive steroid becomes increasingly extractable into cold methylene chloride with increasing purity of IB. This can be interpreted as either progressive denaturation or a conformational change in the protein. The latter possibility seems most likely since rebinding with radioactive steroid can be accomplished at all stages of purification. This explanation is supported by the complex saturation curves for steroid binding (Fig. 3). At the final stage of purification here, there is a great increase in dissociation of steroid upon application of further purification procedures. Table II shows that as much as 25% of bound corticosterone is associated with this protein fraction in vivo. With such a low level of protein in the cell, extremely large scale experiments will be required to prepare a homogeneous protein. However, even at low protein concentrations in the cell, IB might store small amounts of hormone, releasing steroid slowly as the cellular concentration continued to fall, if the cooperative phenomena observed in vitro are operative in the physiological context. This mechanism could be especially important in view of the growth-promoting effects of adrenal corticosteroids (21), but it remains for future investigation to delineate the physiological role of the binding protein.
Achnowledgnent-We thank Dr. M. H. Cake for critical reading of the manuscript.