Reduced Sulfhydryl Groups Are Required for Activation of Uterine Progesterone Receptor POSSIBLE INVOLVEMENT OF AN INHIBITOR OF ACTIVATION*

The dependence of uterine progesterone receptor activation on sulfhydryl groups was studied using bind- ing to DNA-cellulose as a measure of activated receptor. Dithiothreitol, thioglycerol, or other sulfhydryl re- ducing reagents stimulated receptor activation 2- to 3-fold. This effect was produced under mild conditions, ie. after a 22-h incubation at 0-3 "C with 50 mM dithi- othreitol. Other characteristics of the sulfhydryl-de-pendent stimulation of receptor activation, such as pH-dependence and sensitivity to the sulfhydryl blocking agent iodoacetamide, suggest that the sulfhydryl groups essential for receptor activation are different from those involved in steroid binding. Progesterone receptor activation by sulfhydryl reduction was a reversible process, dependent on addition or removal (by dialysis) of the reducing agent. Optimal receptor activation (up to 75% of total receptor) occurred when cytosol was diluted with buffer in the presence of dithiothreitol, suggesting that dissociation of either a receptor subunit or an inhibitory factor present in cytosol may also be involved in the activation process. This putative inhibitor appears to have a M, 1 30,000 since it is not removed from cytosol by dialysis or gel filtration. These results emphasize the importance of sulfhydryl groups or a disulfide bridge, perhaps associated with the DNA-binding domain of the receptor, in a key regulatory step in the mechanism of steroid hormone action: activation and subsequent binding of the steroid-receptor complex to DNA or chromatin. Activation,

The dependence of uterine progesterone receptor activation on sulfhydryl groups was studied using binding to DNA-cellulose as a measure of activated receptor. Dithiothreitol, thioglycerol, or other sulfhydryl reducing reagents stimulated receptor activation 2-to 3fold. This effect was produced under mild conditions, ie. after a 22-h incubation at 0-3 "C with 50 m M dithiothreitol. Other characteristics of the sulfhydryl-dependent stimulation of receptor activation, such as pHdependence and sensitivity to the sulfhydryl blocking agent iodoacetamide, suggest that the sulfhydryl groups essential for receptor activation are different from those involved in steroid binding. Progesterone receptor activation by sulfhydryl reduction was a reversible process, dependent on addition or removal (by dialysis) of the reducing agent. Optimal receptor activation (up to 75% of total receptor) occurred when cytosol was diluted with buffer in the presence of dithiothreitol, suggesting that dissociation of either a receptor subunit or an inhibitory factor present in cytosol may also be involved in the activation process. This putative inhibitor appears to have a M, 1 30,000 since it is not removed from cytosol by dialysis or gel filtration. These results emphasize the importance of sulfhydryl groups or a disulfide bridge, perhaps associated with the DNA-binding domain of the receptor, in a key regulatory step in the mechanism of steroid hormone action: activation and subsequent binding of the steroid-receptor complex to DNA or chromatin.
Activation, an important step in the mechanism of steroid hormone action, is the process by which the steroid-receptor complex acquires the ability to interact with nuclear chromatin or DNA and thereby to alter gene expression in target cells (1,2). The activated steroid-receptor complex is defined operationally and quantified by its ability to bind in vitro to polyanions such as DNA (3), DNA-cellulose (4, 5), ATP-Sepharose (6) and phosphocellulose (7).
Although the mechanism of receptor activation is still obscure, some knowledge has been gained through the use of chemical reagents which affect this process. Previous studies suggested a role for sulfhydryl groups in the binding of glucocorticoid (8,9) and progesterone (10) receptors to DNA, but * Supported by Grants PCM 77-25630 from the National Science Foundation and CA-23362, HD-15452, and HD-13152 from the United States Public Health Service. 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.
$ T o whom all correspondence and reprint requests should be addressed. the function of sulfhydryl groups in the general mechanism of steroid receptor activation is not clear.
We discovered and report here that sulfhydryl reducing reagents such as dithiothreitol stimulate activation of progesterone receptor in hamster uterine cytosol. Our data demonstrate that reduced sulfhydryl groups are required for activation. These studies provide new insight into the mechanism of receptor activation, since the sulfhydryl groups or disulfide bridge essential for activation may be associated with the DNA-binding domain of the receptor. Furthermore, reversible oxidation and reduction of sulfhydryl groups may regulate the interaction between the unactivated receptor and an activation inhibitor.

Materials-[1,2-3H2]Progesterone
(55 Ci/mmol) was purchased from New England Nuclear. Dithiothreitol, iodoacetamide, reduced glutathione, and p-chloromercuribenzene sulfonic acid were from Sigma. A'-ethylmaleimide and 2-mercaptoethanol were obtained from Eastman. Adult femde golden hamsters were housed and estrous cycles were established in these animals, as previously described (11). All the hamsters for these studies were killed between 0800 and 1100 h on cycle day 4, during the period when the tissue concentration of uterine cytosol Rp' is highest (11).
Preparation of Uterine Cytosol Containing [3H]Progesterone-Receptor Complex-Uterine cytosol was prepared by homogenization as reported before (11,12) at a buffer to tissue ratio of 4:l in phosphate buffer: 50 m~ sodium phosphate (pH 7.5 at 22 "C), 10% glycerol. The homogenate was centrifuged for 1 h at 170,000 X g to obtain the cytosol fraction. The cytosol was then incubated for 18-22 h at 0 "C with 20 n~ [3H]progesterone (added in ethanol, final ethanol content, 2%, v:v) in the presence of 4 p~ unlabeled cortisol. Cortisol was included to block binding of [3H]progesterone to corticosteroid-binding globulin and, at this concentration, does not compete significantly with the labeled steroid for binding to the progesterone receptor (11,12).
Estimation of Specific ['H]Progesterone Binding in Cytosol-Following the 18-22-h incubation, duplicate 0.1-ml aliquots were taken from each cytosol treatment for measurement of specific ['HI progesterone binding. One set of duplicate samples was incubated at 38 "C for 30 min to destroy receptor binding, whereas the second set was kept at 0 "C. Free and loosely-bound steroids were removed from all samples by treatment with dextran-coated charcoal (11,12). Specific C3H]-progesterone binding was calculated by subtracting the amount of bound radioactivity present in the heat-treated samples from that in the parallel samples kept in the cold. For only the experiment shown in Fig. 2, nonspecific [3H]progesterone binding in cytosol was measured by incubating parallel samples with both labeled steroid and 4 p~ unlabeled progesterone. The magnitude of the nonspecific steroid binding component (5-10% of total bound steroid) estimated in these ways was found to be the same. The sulfhydryl-modifying reagents employed in these studies had no detectable effect on the amount of nonspecific binding.
DNA.cellulose Binding Assay-Binding of [3H]progesterone-receptor complex to DNA-cellulose was used to measure the amount of activated receptor in each treated cytosol. Duplicate aliquots of cytosol (one heat-treated, one control kept in the cold) were incubated with DNA-cellulose in a slurry at 3 "C as described before (4, 12).
After this incubation, DNA-bound Rp was extracted from the washed DNA-cellulose by two successive 1-h incubations with 0.5 M KC1 in phosphate buffer. Radioactivity was measured in these extracts and specific L'HIprogesterone interaction with DNA was computed as above.
General Methods-Radioactivity was measured in aqueous samples as previously described (12). Values shown in the figures and tables represent single determinations. All experiments were replicated two or three times.

RESULTS A N D DISCUSSION
Incubation of uterine cytosol for 19 h with dithiothreitol produced a concentration-dependent enhancement of Rp binding to DNA (Fig. 1). Under equivalent conditions of pH and buffer concentration (pH 7.5, 50 m~) , phosphate was superior to HEPES or Tris in supporting dithiothreitol-dependent activation of Rp. Maximal Rp activation (30-45% of total Rp in this experiment) occurred with 30-50 m~ dithiothreitol in all three buffers. These data suggest that the receptor (or possibly some receptor-modifying agent in cytosol) requires reduced sulfhydryl groups for optimal interaction with DNA. In the absence of added sulfhydryl reducing agents, 10-14% of the receptor is capable of DNA binding ( Fig. 1).
Previous work suggested that optimal binding of steroids to their receptors requires reduced sulfhydryl groups (13-15). Conversely, sulfhydryl blocking agents inhibit steroid binding by receptor (8,13,16-21). Our data indicate that dithiothreitol and other sulfhydryl reducing agents (Table I)  steroid binding in the presence of dithiothreitol occurred at pH 8.5 in each buffer. Moreover, 2-3 mM dithiothreitol or 5-10 mM thioglycerol was sufficient to promote maximal ["HI progesterone binding by Rp. These findings suggest that different sulfhydryl groups, either located on the receptor itself or on some interacting molecule, are responsible for these two effects. A similar conclusion was reached by others (8, 10). Table I shows the effects of the sulfhydryl-blocking reagents PCMB, N-ethylmaleimide, and iodoacetamide on steroid binding to and activation of Rp. At a relatively low concentration (0.5 mM), PCMB and N-ethylmaleimide destroyed steroid binding to Rp. However, at a concentration of 0.5 mM, iodoacetamide caused only a mild decrease in steroid binding and a small reduction in the amount of activated Rp. When 5 IIIM dithiothreitol was included in the DNA-cellulose assay after preincubation of C~~Q S O~ with iodoacetamide, a slight increase in DNA binding of Rp occurred. The enhancement by dithiothreitol was much smaller in this case than that produced by dithiothreitol in the absence of iodoacetamide ( Table 1). These data support the hypothesis that different sulfhydryl groups are involved in Rp activation and steroid binding, since iodoacetarnide can act selectively to depress activation while having little effect on steroid binding.
In order to study better the interaction between dithiothreitol and sulfhydryl-blocking reagent effects on Rp, the experiment shown in Table I1 was done. The receptor was first reduced by treatment with dithiothreitol and then the excess reducing agent was removed by dialysis. Note that steroid binding to Rp was exceptionally stable even after prolonged incubation and dialysis. After 5 h of dialysis, a small increase in the amount of activated receptor was evident in the cytosol. Addition of 50 mM dithiothreitol to this dialyzed cytosol caused a slight enhancement of activation, but this is not a significant increase. In contrast to the data in Table I, neither iodoacetamide nor N-ethylmaleimide had a substantial effect on the DNA-binding properties of the reduced-dialyzed receptor, although these compounds did cause a reduction in steroid binding. These findings strongly suggest that the effect of these sulfhydryl-modifying reagents is directly on the activation process itself, rather than on the binding of activated Rp to DNA. When the dithiothreitol-pretreated cytosol is dialyzed for 24 h, a somewhat different picture emerges (Table   TI). During prolonged dialysis, the previously-reduced sulky-dry1 groups on the receptor appear to be oxidized, as is suggested by the marked decrease in the amount of activated Rp in the control cytosol (no additions). Adding back dithiothreitol to this cytosol caused a dramatic recovery of activated receptor, demonstrating that sulfhydryl-dependent activation of Rp is a readily reversible process. As in the experiment shown in Table I, treatment of the 24 h-dialyzed cytosol with iodoacetamide or N-ethylmaleimide reduced the amount of steroid binding to Rp. In this instance, however, the proportion of activated receptor in the cytosol was not changed by these reagents. Collectively, the data in Tables I and I1 demonstrate that the unactivated (oxidized) form of Rp is more sensitive than the activated (reduced) form to the inhibitory effects of sulfhydryl blocking reagents and to the stimulatory effects of sulfhydryl reducing reagents. In particular, iodoacetamide and N-ethylmaleimide have no apparent effect on the ability of previously activated receptor to bind to DNA. These findings suggest that the sulfhydryl groups involved in Rp activation are masked or that alkylation by the sulfhydrylblocking reagents does not interfere with DNA binding once the receptor is activated. This may result from a conformational change in the receptor molecule which translocates the sulfhydryl groups away from the DNA-binding site (e.g. by dissociation of a subunit or inhibitor). Alternatively, if the Effects of different sulfhydryl modifying reagents on ["HJprogesterone binding in uterine cytosol and on binding of cytosolprogesterone receptor to DNA-cellulose Uterine cytosol was prepared in phosphate buffer and incubated ml aliquots of each treated cytosol were assayed for specific ['HI with [3H]progesterone, as described under "Experimental Proce-progesterone binding by dextran-coated charcoal treatment and for dues." After 5.5 h of incubation with steroid, additions were made to binding of receptor to DNA-cellulose. In some cases, further additions 0.5-ml aliquots of cytosol (left column). Additions were made from (indicated in the second column from the left) were made to the concentrated stock solutions in distilled water adjusted to between samples at the time of DNA-cellulose binding assay. pH 7.6 and pH 8.  Effect of dithiothreitol removal by dialysis and subsequent treatment with sulfhydryl modifying reagents on binding of cytosol Progesterone receptor to DNA-cellulose Uterine cytosol prepared in phosphate buffer was labeled with [3H] was divided into 0.5-ml aliquots which were incubated for 3 h at 3 "C progesterone and incubated at 3 "C for the times shown, with or with the indicated sulfhydryl modifying reagents. These samples were without 50 m~ dithiothreitol. A portion of the dithiothreitol-treated assayed for specific [3H]progesterone binding and for binding of cytosol was dialyzed against 250 volumes of phosphate buffer for 5 or progesterone receptor to DNA-cellulose in parallel with undialyzed 24 h. During the first 5 h of dialysis, the buffer was changed hourly cytosol samples. DNA-binding site were locked in the unactivated, oxidized receptor by a disulfide bond on the DNA-binding domain, reduction of this bond might permit the molecule to "relax," exposing the DNA-binding site and rendering the molecule activated. The latter situation would be analogous to the case of progesterone binding to uteroglobin (22). In contrast, the sulfhydryl groups associated with steroid binding are available to these reagents in both the activated and unactivated forms of Rp. All four of the thiol-reducing agents we tested enhanced Rp activation when they were included at high concentrations in cytosol during the incubation with steroid (Table I). These reagents (except for GSH) also stimulated Rp activation when they were added to cytosol only during the 1-h incubation with DNA-cellulose (Table I). At concentrations one-tenth as high as those used during the preincubation period (compensating for the ten-fold cytosol dilution in the assay), activation was stimulated by these three compounds to nearly the same extent as that in the corresponding pretreated cytosols. These findings indicate that some activation of receptor may be occurring during the 1-h DNA-cellulose binding assay.
The time course of binding of dithiothreitol-pretreated and untreated Rp preparations to DNA-cellulose was studied in order to understand better what happens during the assay period (Fig. 2). Binding of Rp to DNA-cellulose occurs in two distinct phases. In phase I, representing the fist 4-10 min of cytosol exposure to DNA-cellulose, rapid DNA binding of previously activated Rp takes place. In t h e slower, more prolonged phPse 11, further activation of Rp occurs during the assay. In phase 11, the DNA-binding step is not rate-limiting, and binding of Rp to DNA occurs as activated Rp is produced. The above interpretation is borne out by examination of the binding curves in Fig. 2. Treatment of cytosol with 50 mM dithiothreitol prior to the assay increases the level of preactivated receptor (phase I) 2-to 3-fold over that in untreated cytosol (measured as the difference in Rp binding to DNA between dithiothreitol-pretreated and untreated cytosols at the earliest time, 5 min after cytosol addition to DNA-cellulose). We believe this difference represents a dithiothreitolinduced shift in equilibrium between activated and unactivated forms of Rp in cytosol, presumably through reduction During phase 1 1 , Rp binding to DNA-cellulose increased steadily in the cytosol pretreated with 50 mM dithiothreitol (top two curves of Fig. 2 ) . Inclusion of 5 mM dithiothreitol in the assay (in addition to that contributed by the cytosol) did not increase the rate or final amount of DNA-bound Rp achieved in phase 11. This suggests that the sulfhydryl group(s) necessary for receptor activation had been fully reduced during the 18-h preincubation and that sulfhydryl reduction is probably not rate-limiting for Rp binding to DNA in these cases. On the other hand, sulfhydryl group reduction is most certainly the rate-limiting step in the binding of Rp to DNA-cellulose in the absence of dithiothreitol (bottom curve of Fig. 2). Binding of receptor to DNA in phase I1 is very slow under these conditions, emphasizing the critical importance of reduced sulfhydryl groups for DNA binding. When dithiothreitol (5 mM) is added to untreated cytosol at, the beginning of the DNA-cellulose assay (Fig. 2, center curve), Rp is predominantly unactivated early in the assay and the amount of DNA-bound Rp is the same as that in dithiothreitol-untreated cytosol. As dithiothreitol-promoted sulfhydryl reduction progresses, the amount of activated Rp increases. Eventually, of sulfhydryl groups. this curve converges (between 3 and 4 h) with those of the dithiothreitol-pretreated cytosols. Thus, a change in rate-limiting step appears to occur during this incubation.

Sulfhydryl-dependent Activation of Progesterone Receptor
It is possible that the ten-fold dilution of cytosol which occurs at the beginning of the DNA-cellulose binding assay results in dissociation of a subunit or an inhibitory factor from the receptor and that this step is necessary for receptor activation. To test this hypothesis, we determined whether dilution of cytosol before the DNA-cellulose assay enhanced Rp binding to DNA. For this purpose, a 10-min exposure of cytosol to DNA-cellulose was used to measure only phase I (pre-activated) receptor binding to DNA (Fig. 3 ) . In the absence of dithiothreitol, dilution with buffer caused a small, but reproducible, increase in the amount of activated Rp in cytosol. Addition of 50 m dithiothreitol to buffer-diluted cytosol caused a marked increase in activated Rp. Clearly, dilution of cytosol and sulfhydryl reduction are requisites for optimal receptor activation; neither manipulation alone produces the maximal response. Dilution of cytosol with heattreated, receptor-free cytosol did not cause an increase in activated receptor, either in the presence or absence of dithiothreitol. This finding suggests that cytosol contains a factor(s) which block(s) activation of Rp. Previous work has shown that steroid receptors may be activated by dilution of the cytosol (5,23,25), by passage through gel fiitration columns (5,24-26) or by dialysis (25), suggesting that some agent in cytosol acts as an inhibitor of activation. The action of such an inhibitor may explain the rate of DNA binding of Rp during phase I1 of the DNA binding assay. An alternative possibility is that the receptor may dissociate into subunits upon dilution and thereby become activated. Other proteins in cytosol can interfere with steroid receptor binding to DNA (27), but this is unlikely for the case of uterine Rp, since the dilution-dependent change that stimulates Rp activation seems to occur in pretreatment of the cytosol rather than in the DNA-cellulose binding assay. Finally, the possibility exists that the receptor must undergo a change in conformation upon transition from the activated to unactivated state (28,29).
Preliminary efforts to separate the putative inhibitor of activation from the receptor have not been successful. The factor is not removed from uterine cytosol by passage through Sephadex G-50 in the presence of 50 m~ dithiothreitol (data not shown) or by dialysis (Table 11). This would indicate that the inhibitor has an M, > 30,000 and probably is dissimilar to the inhibitor(s) of glucocorticoid receptor activation present in rat liver cytosol (5,25). Although activation of Rp appears to be regulated by dissociation of an inhibitor, the possibility that a conformational change occurs upon dissociation of the inhibitor cannot be excluded. Furthermore, it remains to be determined whether the sulfhydryl groups important for activation of Rp are located on the receptor or the inhibitor, or both (perhaps a disulfide bridge links the molecules covalently).
Recently, Hughes et al. (30) showed that purified chick oviduct progesterone receptor subunit A binds in a saturable manner to DNA containing limited nicks, but that binding became nonsaturable as the single-stranded character of the DNA increased. They concluded that the steroid receptor may act as a helix-destabilizing protein. In light of these results, the interaction of steroid receptors with DNA takes on renewed importance in the mechanism of steroid hormone action. Our studies demonstrate that sulfhydryl groups on the DNA-binding domain of the receptor play an essential role in this mechanism.