Vol. 254, No. li, Issue of September 10, pp. 8534-8539. 19i9 … · 2002-12-05 · Vol. 254, No. li, Issue of September 10, pp. 8534-8539. 19i9 I’rmfed in I! S.A Thyroid Hormone

Vol. 254, No. li, Issue of September 10, pp. 8534-8539. 19i9 I’rmfed in I ! S.A

Thyroid Hormone Receptors ALTERATION OF HORMONE-BINDING SPECIFICITY*

(Received for publication, October 2, 1978, and in revised form, May 29, 1979)

Norman L. Eberbardt, Janet C. Ring, Keith R. Latham,+ and John D. Baxter5

From the Howard Hughes Medical Institute, Endocrine Research Division, Department of Medicine, Department of Biochemistry and Biophysics, and the Metabolic Research Unit, University of California, San Francisco, California 94143

In previous studies of nuclear extracts containing solubilized receptors for thyroid hormones, triiodothyronine (Ts) but not thyroxine (T4) binding activity was lost, even though in the starting material TI and T4 appeared to bind to the same proteins (Latham, K. R., Ring, J. C., and Baxter, J. D. (1976) J. Biol. Chem. 251, 7388-7397). In the current studies this paradoxical loss of T3 but not T4 binding activity was studied in greater detail. Competition, iodoacetamide inhibition and binding site concentration studies with [‘251]T3 and [lZ51]T4 were consistent with the idea that T3 and T4 are bound by the same protein(s) in the initial extracts. When such preparations were heated at 50°C [““IIT but not [1251]T4 binding activity was rapidly lost. There was no change in the T4 binding site concentration and no change or a modest reduction in affinity. By contrast, heating markedly reduced the concentration of high affinity T3 binding sites; the affinity of the remaining sites for T3 was 0.1% that of the starting material. When the pH of the extract was lowered from 7.6 to 6.0 the high affinity T3 binding site concentration was markedly reduced but the T4 binding site concentration and affinity were minimally affected. T3 was a more avid competitor than T, for [‘251]T4 binding at pH 7.6 but had weak competitor activity at pH 6.0. These heat- and pH-induced influences could occur if the receptors are destroyed and a T4-binding species is generated or exposed. Alternatively, the data can be explained by a model in which the thyroid hormone receptor can be converted to a form which retains binding activity for T4, but which has reduced affinity for TB; in this case a fundamental unit of the receptor could be more similar than previously apparent to certain cytosol or plasma proteins that bind T4 more avidly than T,.

Binding proteins for triiodothyronine (T:J and thyroxine (T4) have been well characterized in the plasma (l-5), cytosol (6, 7), and in the nucleus of target cells (8-19). A class of intranuclear proteins have been designated as receptors on the basis of several lines of correlative evidence. These receptors: (a) are acidic chromosomal proteins which after solubi- lization have an apparent molecular weight of approximately 50,000 to 70,000 (8, 9); (b) are DNA binding proteins or bind to DNA through association with other proteins in the extract

* This work was supported by National Institutes of Health Grant l-ROl-AM-18878. The costs of publication of this article were de- frayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

+ Present address, Dept. of Medicine, Uniformed Services Univer- sity of the Health Sciences, Bethesda, Md. 20014.

5 Investigator of the Howard Hughes Medical Institute.

(9, 10); (c) bind T:s,’ Tq, and other thyroid hormone analogues in proportion to their biological potency (12-15); (d) bind biologically active hormones with apparent equilibrium dissociation constants that are similar to physiologically active concentrations of the hormones (13,16); and (e) are detectable in all known thyroid hormone-target tissues examined (17).

Although the precise molecular mechanism(s) of thyroid hormone action is unknown, it has been shown that thyroid hormones can increase the synthesis of total RNA in certain target cells (20). More recently, it has been demonstrated that thyroid hormones can also increase levels of specific messen- ger RNAs, even in tissues where the hormone does not affect total RNA levels (21, 23). Accordingly, a current hypothesis of thyroid hormone action holds that the nuclear receptor mediates an effect of the hormone, resulting in changes in chromatin activity which increase the levels of specific mRNAs.

In contrast to the nuclear receptor, which binds Tcs more avidly than Td (8, 12-15), certain cytosol proteins (6) and plasma binding proteins (1, 2), thyroxine binding globulin (TBG) and thyroid hormone-binding prealbumin (TBPA), bind T,, more avidly than T:3. Since these binding proteins do not bind TLs, Tq, or other biologically active analogues in relation to their biological potency, they are not considered to be receptors.

The x-ray crystallographic structure of TBPA has been determined (3). This is the only hormone binding protein for which such detailed structural information is available; it may serve as a model for thyroid hormone receptors. The crystallographic data suggest that TBPA has a DNA binding site. In earlier investigations of plasma thyroid hormone-binding proteins, we failed to detect high affinity DNA binding (10). However, the finding of a structure in TBPA that resembles a DNA binding site and the data that suggests that the receptor may be a DNA binding protein suggests that these two proteins have similarities other than their common thyroid hormone-binding capability. In addition, the molecular weight of the receptor and TBPA are quite similar (4, 5,8,9).

In our previous efforts to preliminarily characterize and purify the solubilized receptors (8), we found that T3 binding activity was lost in fractions where there was little, if any, loss of Tq binding activity. This result was unexpected since the data suggested that T8 and T4 were binding to the same protein in the starting material. In extending these studies, we now present additional evidence that high affinity Ts binding activity can be selectively decreased by heating or changes in pH. The results suggest that either a new Tq binding species is being generated or that the receptor can be converted to a form which binds T4 more avidly than T3. The

’ The abbreviations used are: T.3, 3,3’,5-triiodo-L-thyronine; Tq, 3,3’,5,5’-tetraiodo-L-thyronine; TBPA, thyroid hormone-binding prealbumin; and TBG, thyroid hormone-binding globulin.

8534

by guest on September 7, 2020

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

latter model implies that there may be an even closer similar- ity of the receptor to either TBPA or certain other cytosol binding proteins than is obvious from previous studies and that the intranuclear receptor may be derived from a protein structurally similar to a plasma or cytosol binding protein.

MATERIALS AND METHODS

Reagents-3,5-[3’-1’51]Triiodo-L-thyronine (approximately 900 pCi/pg, 1350 dpm/fmol) and 3,5-[3’,5’-““Iltetraiodo-L-thryonine (approximately 1200 $!i/pg, 2100 dpm/fmol) were purchased from 40 80 120 l&l New England Nuclear and unlabeled 3,3’,5-triiodo-I,-thyronine and 3,3’,5,5’-tetraiodo-r-thyronine were obtained from Sigma Chemical

BOUND IpM! BOUND (pM1

Co. These hormones were routinely monitored for purity as previously FIG. 1. Scatchard analysis of [‘““IIT, ( A) and [“‘I]T3 (B) bind-

described (8). Iodoacetamide was from Sigma Chemical Co. MEM ing to the initial chromatin extract from rat liver before (0)

Joklik medium and fetal calf serum were from the Cell Culture and after (0) heating for 10 min at 50°C. Control and heated

Facility, University of California, San “rancisco. Coomassie brilliant samples were treated and assayed as described under “Materials and

blue G-250 was obtained from Eastman Kodak Co. Methods” except that the concentrations of [‘251]T3 and [‘?]T, were

Buffers-Buffers had the following compositions: Buffer A, 10 mM varied from 35 to 6200 and 400 to 11,000 PM, respectively. The protein

Tris.HCI (pH 7.5), 0.32 M sucrose, 2 mM MgC11, 0.24 mM spermine; concentrations in the incubations of the control and heated extracts

Buffer B, 20 IIIM (N-[Tris(hydroxymethyl)methyl]glycine (Tricine, were 330 and 241 pg/ml, respectively. The equilibrium dissociation

pH 7.5), 2 mM CaC12, 1 mM MgC12, 0.5% Triton X-106; Buffer C, 20 constants for [““I]T:r and [‘““IIT, binding by the control extracts are

mM Tris.HCl (pH 8.0), 0.25 M sucrose, 1 IIIM EDTA, 0.2 mM dithio- 0.50 and 10.1 nM, respectively.

threitol, 5% glycerol; NaCl/P,, 25 InM sodium phosphate (pH 7.4), 0.1 M NaCl; Buffer G, 50 tnM sodium phosphate (pH 7.6), 200 nlM experiments we have determined that this method of calculating the (NH,)SO.,, 1 mM EDTA, 0.2 mM dithiothreitol, 5% glycerol. Mea- free hormone concentration agrees with the actual measurement of surements of pH were made at 22”C, the temperature at which the free hormone which was accomplished by eluting [““I]TB or [‘*“I]T., assays were performed. from individual G-25 Sephadex columns with 0.25 N NaOH, 50 mM

Preparation of Chromatin Entracts-The solubilized receptor was NaCl. In all cases, the line drawn through the individual Scatchard obtained from purified rat liver nuclei as described previously (8). data points represents the least squares linear regression line.’ These extracts, which were routinely stored in liquid nitrogen with Proteins were determined by the method of Lowry et al. (25) or by no loss of hormone binding activity for periods up to 6 months, were the Coomassie blue dye binding assay (26). The dye binding assay centrifuged after thawing (37°C) at 50,006 x g for 20 min at 4°C in agreed with Lowry protein determinations for the chromatin extracts order to remove a cryoprecipitate. The centrifugation at this point and partially purified receptor preparations. DNA was measured by removed a small amount of protein (25%) as well as 80% of the DNA the method of Burton (27). which was present in the crude extract without loss of hormone binding activity. In all of the experiments described herein the extract RESULTS was treated in this manner prior to subsequent manipulation,

Cultured rat pituitary cells (subline GHB) were grown in suspension Binding of T3 and T4 by Nuclear Extracts-Previous stud-

culture at 37°C in MEM Joklik medium supplemented with 10% fetal ies have shown that there are generally equivalent concentra- calf serum. The cells were grown to a density of approximately 10” tions of Ta and Tq binding sites in solubilized extracts from cells/ml. All of the procedures below were performed at 4°C unless otherwise specified. One to 1.5 liters of cells were centrifuged at 500

both rat liver (8) and GH cell nuclei (28) as measured by

X g for 10 min and the supernatant was discarded. The cell pellet was Scatchard analyses of the binding data. These findings are

washed twice with NaCl/P, by resuspending it in 400 ml of NaCl/P, confirmed by the Scatchard analyses of [issI]Ts and [‘251]T4

and centrifugation at 500 x g for 10 min. The cell pellet was subse- binding by rat liver nuclear extracts, shown in Fig. 1 (compare

quently suspended in 40 ml of Buffer A and homogenized with a A and B, control data). These data suggest that Ts and Tq Potter-Elvehjem homogenizer (six strokes). The homogenate was may be bound by the same proteins in the nuclear extract. centrifuged at 700 x g for 10 min and the supernatant was saved for subsequent isolation of cytoplasmic components (liquid nitrogen).

Additionally, data from a number of laboratories (8, 15, 28-

The nuclear pellet was washed twice by resuspension in 40 ml of 30) have demonstrated that [is’I]Ts binding by nuclear ex-

Buffer B (Triton X-109) followed by centrifugation at 700 x g for 10 tracts is inhibited by unlabeled Tq. Moreover, the binding of

min. The nuclear pellet was suspended in 40 ml of Buffer C containing [“51]T4 by nulcear extracts from rat liver is completely in-

200 mM ammonium sulfate and sonicated (Heat Systems Ultrasonics, hibited by unlabeled T:r and Tq (8). This finding is confirmed

Inc., model 185) three times for 20 s at 80 watts with intermittent in Fig. 2 (A and B, control data) which shows the inhibition cooling. The gelatinous suspension was centrifuged at 50,000 x g for of [i2’I]T4 binding to rat liver and GH cell nuclear extracts by 1 h and the supernatant was stored in liquid nitrogen. Prior to assay, the solubilized GH.3 cell receptor was treated in an identical manner

unlabeled T:j and Tq. In both cases, [‘251]T4 binding is com-

as the rat liver receptor (described above). pletely inhibited by unlabeled Ts, and T:s is the more avid

Heat Treatment of the Chromatin Extracts-Five to 10 ml of the comuetitor. These frndines demonstrate that both hormones

chromatin extracts were incubated in a water bath with intermittent are bound by the same protein(s) that has a higher affinity for

stirring at 50°C for various times; aliquots were removed and chilled Ts than Tq. on ice and centrifuged at 15,000 x g for 10 min at 4°C prior to assay. Effect of Heat on Receptor Binding of T:s and T4-In our

Binding Assay-All binding reactions were carried out at 22’C for 2 h in Buffer G at a final volume of 0.5 ml containing 1 nM radiolabeled

previous studies (8) with the solubilized receptor from rat

hormone unless otherwise specified. In all cases the data have been liver nuclei, we found that after quaternary aminoethyl

corrected for the decay of “’ I. The standard binding assay has been ‘Two problems arise when it is desired to measure low affinity described in detail (8) and was used throughout this study. In all of binding by the methodology employed in these studies. First, the gel- the binding reported in these experiments, the nonspecific hormone filtration assay perturbs the equilibrium. Therefore, there can be binding (binding not competed by 1 pM radioinert TJ or T,) has been dissociation of weakly bound hormone and consequently an under- subtracted from the total binding. estimate of the binding. Secondly, and more importantly, the assay

In analyses of data according to Scatchard (24), the free hormone depends on the amount of nonspecific hormone binding being low has been determined by subtracting the total amount of [““I]T.I or relative to the amount of specific binding. This is the case at lower [‘““I]T? bound (specific binding plus nonspecific binding) from the hormone concentrations (0.1 to 10 nM), but nonspecific binding is total [““I]T.1 or [‘?]T., present in each incubation (determined by linearly related to the hormone concentration and would greatly counting 50 ~1 of each assay mixture). This ensures that the nonspe- exceed specific binding at the 0.1 to 10 PM ranges required for cific binding is not included in the free hormone pool. In separate substantial binding by a lower affinity component.

Thyroid Hormone Receptor: Alteration of Binding Specificity


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

8536 Thyroid Hormone Receptor: Alteration of Binding Specificity

coMmlToR CloG,~ Ml coMFmtm wG,,M)

FIG. 2. Competition for [‘““IJT4 binding by unlabeled Ts (0, 0) and T, (Cl, M) in the control (0, B) and heated (50°C for 10 min. 0, Cl) chromatin extracts from rat liver nuclei (A) and GHs cell nuclei (B). Binding of radiolabeled hormone is shown as per cent of control binding. The 100% values for [‘*“IIT., binding by the rat liver extracts (A) and the GHs cell extracts (B) were 21.0 and 46.8 PM,

respectively. Samples were treated and assayed as described under “Materials and Methods” except that the concentration of unlabeled hormone was varied as indicated.

TIME (minutes)

FIG. 3. Effect of heating on binding of [rz61jT3 (0) and [‘*“I]T4 (0) by solubilized rat liver (A) and GH3 cell (B) nuclear thyroid hormone receptors. Aliquots of the heated extracts were removed at the specified times, chilled on ice, and centrifuged at 12,006 x g for 10 min at 4°C prior to assay to remove precipitated protein. &says were performed as described under “Materials and Methods” and the data represent specific binding of the respective hormones. The binding is plotted as the per cent of binding in the unheated control. With rat liver extracts (A) the 100% values for [‘251]T3 and [““I]T4 binding were 95.6 and 26.6 PM, respectively. The 100% values for the binding of [‘251]T3 and [‘*“I]T4 by the GHa cell extracts (B) were 96.0 and 24.2 PM, respectively.

(QAE)-Sephadex chromatography of the crude nuclear extract a component, not detected in the original extract, was observed which bound Tq more avidly than Ts. In addition, after storage of such QAE-Sephadex purified fractions, there was a further specific loss of T3 binding activity without a concomitant effect on Tq binding activity. In view of the evidence previously discussed that Ts and Tq are bound by the same protein(s), this seemingly paradoxical result suggested that some property of the receptor, important for Ts binding activity, was lost after storage at 4°C or purification.

To study the selective loss of TB binding activity under more controlled conditions, we examined the effect of heating the nuclear extracts on [‘251]TZ and [1Z51]T4 binding activity. When a crude extract from rat liver nuclei is heated at 50°C (see “Materials and Methods”), there is a rapid loss of TS binding activity; by 10 mm 90% of it is lost (Fig. 3A). By contrast, at 10 min T4 binding activity remains at the control level; it is reduced only by 40% after 20 min, at which time TZ binding activity is nearly abolished (Fig. 3A). Similar results were obtained with the solubilized receptor from GH3 cell nuclei as shown in Fig. 3B.

To further analyze the nature of the heat-induced change in receptor properties, we compared the binding of TS and Td by Scatchard analyses (Fig. 1, A and B) before and after heating. The binding of [‘?]T., by the control and heated (50°C for 10 min) extracts was indistinguishable (Fig. IA); there was no significant change after heating in the T4 affinity or binding site concentration. In four such experiments, the number of Tq binding sites as estimated by the Scatchard

technique was within 10% of that in the control extract and the mean equilibrium dissociation constant was 6.3 + 3.7 nM (S.D.) prior to heating and 9.1 -+ 4.4 ILM (S.D.) after heating at 50°C for 10 min. By contrast, the binding of TB by the receptor after heating has been drastically reduced (Fig. 1B). In other studies, to assess the nature of the loss, the kinetics of the effect were examined by Scatchard analyses of heated samples taken at various time points (data not shown). The Tq binding site concentration remained constant through 10 min, whereas there was a progressive decrease in the high affinity TB binding site concentration. The residual T3 binding activity of Fig. 1B measured in the heated sample represents the sum of a small amount of high affinity TS binding activity and the weaker T3 binding activity demonstrated more clearly by the competition studies discussed below.

Several possibilities could explain the nature of the loss of T3 binding activity. TB might not bind to the heat-treated extract at all. Alternatively, the affinity of the residual binding activity for TB could be too low to be detected by the direct assay procedures employed. To understand the nature of the loss of T3 binding activity in more detail, we examined the capability of T3 and Tq for competing with [ 1251]T4 for binding by control and heat-treated nuclear extracts from rat liver and GH3 cell nuclei (Fig. 2, A and B). The capability of T4 for inhibiting [1251]T4 binding in the heated extracts is similar to the case with the initial extracts. TB can inhibit [lz51]Tq binding, but its ability to do so has been reduced by about lOOO- fold. These data suggest that Ts is still capable of binding to the proteins in the heated extract, albeit with an affinity that is about 0.001 times that of the original affinity.

Inhibition of T3 and T4 Binding by Iodoacetamide-Two possibilities could be envisioned to account for the differential influences of heat on Ts and Tq binding activity discussed above. First, a single protein which binds TS more avidly than Tq might be altered by heat treatment such that it retained its Tq, but not Ts, binding capability. Alternatively, a Tq binding protein which was activated by heat treatment might be generated which could compensate for the loss of a protein which preferentially bound TS.

In order to obtain additional information about the similar- ity of the binding properties of T1 and Tq in control and heat- treated preparations, studies of the inhibition of [‘251]T3 and [1251]T4 binding by iodoacetamide were performed; it was

IODOACETAMIDE (mM)

FIG. 4. Dixon analysis. A, Dixon analysis of the inhibition of the binding of [‘*“IIT, (0) and [‘*“IIT (0) to the initial cbromatin extract from rat liver nuclei by iodoacetamide. B, Dixon analysis of the inhibition of the binding of control (0) and heated (e 50°C for 10 min, “Materials and Methods”) cbromatin extracts from rat liver nuclei by iodoacetamide. Assays were performed as described under “Materials and Methods.” Iodoacetamide was included at the indicated concentrations, the concentrations of [“‘I]Ts and [‘251]T4 were kept constant at 1 nM, and the receptor concentration was varied by adding various aliquots (25, 50, and 100 ~1) of the initial or heated chromatin extract (see “Materials and Methods”). Linear regression analysis of the data for the initial extract indicates that the inhibition constant was 3.87 + 0.25 mM (SD., N = 3) when measured with TS and 3.53 + 0.15 mM (SD., N = 5) when measured with T4; the inhibition constant was 3.45 f 0.25 (SD., N = 3) when measured with Tq in the heated extract.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Thyroid Hormone Receptor: Alteration of Binding Specificity

known that Ts-receptor binding was sensitive to agents which reacted with sulfhydryl groups (31).

Fig. 4A shows a Dixon analysis (32) of the inhibition of 1 nM [‘““IIT or [lz51]T4 binding by the crude extract at varying concentrations of iodoacetamide and receptor. As indicated, the data for both hormones result in a series of nonparallel lines intersecting at the same point on the abscissa. The inhibition constant for iodoacetamide (judged by the intercept on the abscissa) is identical for [1251]T3 or [lz51]T4 binding. These results are consistent with the idea that in the starting material TS and Tq are bound by the same protein(s). A comparison of the inhibition of [1251]T4 binding by iodoacetamide in the control and heated extracts is shown in Fig. 4B. It is evident from the data that heating does not alter the characteristics of receptor inhibition by iodoacetamide. These data, therefore, do not provide any support for the possibility that the T4 binding species in the heated extracts are different from that in the initial extract.

Effect of pH on Receptor Binding of T3 and Td-Previous studies indicated that the properties of the receptor for binding T3 and Tq were differentially affected by changes in pH (8). Optimal binding of T3 by the receptor occurred at pH 7.6, whereas optimal binding of T4 was observed at pH 6.0. These differences in binding might be related to a specific influence on the binding site which allows for the discrimination be- tween TS and Tq. Since previous studies could be interpreted to indicate that protonation of the phenolic hydroxyl influ- enced binding of the hormones (8), we examined binding of TS and T4 at pH 6.0 where the phenolic hydroxyl for both TS and T4 is protonated. This would minimize the effect of the state of ionization of the hormones on the binding.

Fig. 5 (A and B) shows Scatchard analyses of the binding of [1251]T3 and [1251]T4 by the receptor at pH 7.6 and 6.0. At pH 7.6 (Fig. 5A) the binding site concentrations for TS and Tq are similar. By contrast, at pH 6.0 (Fig. 5B) the high affinity Tq binding site concentration greatly exceeds that for Tz. Al- though the T4 site concentration in the experiment of Fig. 5B (pH 6.0) exceeds by about 30% that at pH 7.6 (Fig. 5A), it cannot be concluded that the lower pH increased the Tq site concentration since different samples were used in these experiments. In addition, the lower pH increased the dissociation constant for Tq from 4.6 to 9.3 no and for TI from 0.37 to 2.1 no (Fig. 5, A and B). Thus, the concentration of high affinity

BOUND (pM) BOUND (PM)

FIG. 5. Scatchard analysis of the binding of [“‘I]T3 (0) and [12?]T4 (0) by the initial chromatin extract from rat liver nuclei at pH 7.6 (A) and pH 6.0 (B). Samples of the chromatin extracts were dialyzed against Buffer G (see “Materials and Methods”), pH 7.6 or pH 6.0, for 12 h at 4°C. Following dialysis, the samples were centrifuged at 50,006 x g for 20 min at 4°C and aliquots were incubated in Buffer G, pH 7.6 or 6.0, as described (see “Materials and Methods”). The [“51]T, concentration was varied from 200 to 13,200 PM, and the [‘251]T3 concentration was varied from 60 to 12,000 PM.

The protein concentrations of the incubations at pH 7.6 and 6.0 were 230 and 222 pg/ml, respectively. The equilibrium dissociation constants for [1251]T3 and [?]T4 binding by the nuclear extracts at pH 7.6 are 0.37 and 4.6 I-N, respectively. The corresponding values at pH 6.0 are 2.1 and 9.3 no, respectively.

IO 9 8 7 6

COMPETITOR I-LOG,, M)

FIG. 6. Competition for [12’qT, binding by unlabeled Ts (0, 0) and Td (a W) in the initial chromatin extract from rat liver nuclei at pH 7.6 (0,U) and pH 6.0 (0, W). The chromatin extracts were treated as described in the legend to Fig. 5 and the assays were performed as described (see “Materials and Methods”) except that the [““IIT concentration was 7.2 mu. The 100% values for [‘251]T4 binding by the receptor at pH 7.6 and 6.0 were 72.8 and 56.4 PM,

respectively.

IONK STRENGTH (mM)

FIG. 7. Effect of increasing ionic strength on the binding of [12”I]T3 (0,O) and [1251JT4 (0, W) by the initial chromatin extract from rat liver nuclei. The initial buffer composition was: 50 mM

sodium phosphate (pH 7.6), 20 mu ammonium sulfate, 1 mM EDTA, 0.2 mM dithiothreitol and the ionic strength was increased by adding aliquots of 1 M KC1 (0, 0) or 1 M ammonium sulfate (0, W). Assays were performed as described under “Materials and Methods” except that the [““IIT, concentration was 7.2 no. The 100% values for [““IIT and [‘*51]T, binding were 148 and 62.2 PM, respectively.

T3 binding sites is markedly decreased by lowering the pH, whereas only minimal changes in Tq binding occur.

To determine if the reduced TZ binding activity at pH 6.0 resulted from a loss of binding sites or was the result of an alteration of some of the receptors such that their affinity is too low to be measured by the direct approach used in Fig. 5, competition analyses of the binding of [1251]T4 by the receptor were performed as indicated in Fig. 6. Unlabeled T4 competes with [1251]T4 for binding equally well at pH 7.6 or pH 6.0. By contrast, at pH 6.0 unlabeled TS competes much less avidly than at pH 7.6 for the sites which bind [1251]T4. At pH 6.0 TI is still capable of binding to the sites which bind Tq, albeit with an affinity that is lower than the affinity of Tq for the sites. These data indicate that pH differentially affects the binding of TS and T.,.

Ionic Strength Dependence of TZ and T4 Binding by the Receptor-The influence of salt on T3 and Tq binding to the receptor is shown in Fig. 7. Binding of [1251]T~ and [‘251]T4 are both diminished in a parallel fashion with increasing KC1 concentrations above an ionic strength of 206 mM; identical results were also obtained with NaCl. Ammonium sulfate also decreased the binding but higher ionic strengths were required (Fig. 7). In other studies (not shown), it appears that the receptor may be more susceptible to chloride in the presence


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

8538 Thyroid Hormone Receptor: Alteration of Binding Specificity

BDiJND (pM) BOUND (PM)

FIG. 8. Scatchard analysis of the binding of [“‘IIT (A) and [““I]T4 (B) by the initial chromatin extract from rat liver nuclei at an ionic strength of 206 mu (0,Q and 706 mra (0, B). Buffer conditions were as described in the legend to Fig. 7 and KC1 was used to increase ionic strength. Assays were carried out as described (see “Materials and Methods”); the [‘*“I]T3 and [““IIT concentrations were varied from 30 to 1700 PM and 600 to 18,000 PM, respectively. The protein concentration in the incubations was 230 pg/ml.

of phosphate than in the presence of Tris. Scatchard analyses of the effect of ionic strength on the binding of [1251]T3 and [lz51]T4 by the receptor are shown in Fig. 8 (A and B). These data indicate that as the ionic strength is increased from 206 to 706 mM there is a reduction in the affinity of either TS (Fig. 8A) or T4 (Fig. 8B) for the receptor. There may also be a small reduction in the high affinity Ta binding capacity (Fig. 8A). In addition, there may be a further selective loss of high affinity Ts binding activity when the ionic strength is reduced below 100 mM (data not shown). Thus, the dominant effect of higher ionic strength is on the affinity of the binding of both TB and Tq. This influence differs from that due to heating or pH.

DISCUSSION

In the current studies of the solubilized nuclear thyroid hormone receptor, several types of manipulations resulted in a decrease in TB binding activity without a similar change in T4 binding activity. As demonstrated in Figs. 2 to 6, heat treatment and alterations in pH of the nuclear extract are each capable of selectively reducing the concentration of high affinity TS binding sites with no decrease in the concentration of Tq binding sites (Figs. 1 and 5). As judged by competition studies for [‘251]T4 binding (Figs. 2 and 6), TZ binds to the material in the heat- or pH 6.0-treated preparations with about 0.001 the affinity it has for receptors in the starting material at pH 7.6. These influences were similar to those we previously observed with QAE-Sephadex-purified material (8). Further, in more recent studies, we have observed the same phenomenon in Sephadex G-IOO-purified (approximately go-fold) receptors.” We found that, if these preparations are kept concentrated, the concentration of high affinity Ts and Tq binding sites are identical and TS is a more avid competitor for [12”I]T4 binding. When these preparations are diluted, there is a loss of high affinity TS but not Tb binding activity. All factors that influence binding do not promote these changes; the affinity of TS and Tq for binding to the receptor are both decreased with higher ionic strength (Figs. 7and8,AandB).

Ordinarily, such selective losses of TS binding activity in a crude system would suggest that a TS binding species which does not bind Tq was destroyed, leaving a Tq binding species. However, several lines of evidence indicate that in the initial extracts, TS and Tq are bound by the same protein(s). Thus, there are similar concentrations of Ta and Tq binding sites in the initial nuclear extracts (8, 28; Figs. 1, A and B, and 5A).

3 Eberhardt, N. L., Ring, J. C., Latham, K. R., and Baxter, J. D. (1979) Proc. Natl. Acad. Sci. U. S. A., in press.

More importantly, the binding of radiolabeled Ts and Tq is completely inhibited by unlabeled TS and Tq, indicating that both hormones are bound to the same protein(s). Further, all of the Tq binding can be more avidly inhibited by T3 than Tq (Figs. 2 and 6), suggesting that all of the Tb binding is by sites that actually have a higher affinity for TS. Thus, any destruction of a TB binding species should have resulted in a major loss of Ts binding activity; this was not the case (Figs. lA, 3, and 5).

The question of whether TS and Tq are binding to the same protein was further explored by taking advantage of the fact that receptor binding activity was known to be sensitive to reagents that react with sulfhydryl groups (31). The binding of Ts and Tq was inhibited by iodoacetamide in a parallel fashion (Fig. 4A); the inhibition constant was similar for either ligand. These findings are also consistent with the idea that the hormones are bound by the same protein(s). Under the conditions of these experiments (pH 7.6) akylation of histidine residues of proteins by the inhibitor is markedly depressed (33). Thus, the inhibition of binding by iodoacetamide could also indicate that a cysteine or lysine residue located at or near the binding site or at an adjacent site is important for maintenance of the integrity of the binding site.

Although the loss in T3 binding activity cannot be explained only by destruction of a species that binds T3 but not Tq, the results might be explained if the receptors were denatured and the heating or pH influences (or as in the earlier studies (8) the storage of partially purified receptors) resulted in a simultaneous generation or exposure of a new binding species that had an affinity for Tq identical to that of the receptors. Since the affinity and number of binding sites for Tq is unchanged through 10 min of heating at 5O”C, whereas the number of high affinity TS binding sites decreases markedly, the generation of any new T4 binding species would have to be kinetically and quantitatively coupled to the loss of the Ts binding species. In addition, it would have to occur equally in extracts of nuclei from pituitary cells or liver and, therefore, may be general to tissues that contain thyroid hormone receptors. Finally, any newly formed or exposed Tq binding species would have to be inhibited by iodoacetamide identically to the receptors (Fig. 4B). Heating, pH changes, and possibly dilution and purification are known to denature proteins and any hydrophobic regions exposed could contain new binding sites. For instance, we have noted an increase in nonspecific binding for glucocorticoids, TS, and Td after heating but do not detect this with dilution of partially purified material and or changes in pH. However, the probability that this type of influence would generate a class of specific binding sites with a site concentration and affinity for Tq equal to and kinetically coupled to the loss of receptor binding activity seems to us less likely. Nevertheless, these data do not exclude the possibility that there is a simultaneous loss of receptors and exposure or generation of new sites. Alternatively, the hypothesis must be considered that heating, pH, and purification and dilution influence the properties of the receptors such that they retain their capability for high affinity Tq binding but are able to bind Ts with only 0.001 the affinity of the original receptors for this ligand. In any event, the finding that there can occur selective influences on TS binding without parallel changes in Tq binding should also alert investigators purifying these proteins or studying them in other ways (34-39) to these possibilities.

The hypothesis that there might exist a form of the receptor with a different hormone-binding specificity would imply that specific influences on the binding site are required for the receptor to bind TS, the major hormone that occupies the receptor in vivo with high affinity (40). Such influences on the


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Thyroid Hormone Receptor: Alteration of Binding Specificity 8539

binding could be allosteric or by direct modification of the binding sites. The altered form of the receptor could be generated in a number of ways, including proteolysis, covalent modifications, interactions with small ligands, or interactions with other macromolecules. In other studies, we have obtained additional evidence that supports the hypothesis that a basic unit of the receptor with altered hormone-binding properties can exist and that also suggests that it is the dissociation of a receptor-associated protein (possibly a histone or histone-like species) that is responsible for the generation of an altered form of the receptor. A basic protein present in the extracts of chromatin or purified histones (but not other basic proteins such as lysozyme, cytochrome c, and poly-L-lysine) will recon- stitute high affinity T3 binding activity in preparations that have lost activity.” In preliminary studies, we have sought to obtain some indication of the physical characteristics of the receptor before and after heating, acidification, or purification. Whereas the major portion of the high affinity TB or Tq receptor binding activity is included by Sephadex G-100 corresponding to a molecular weight of about 50,000 (as previously reported, Ref. 8), the Tq binding activity after heating is excluded by the column, suggesting that these species aggre- gate with themselves or other macromolecules in the preparation. Thus, an elucidation of the molecular characteristics of the altered receptors must await the development of better methodology.

The thyroid hormone receptors are found in chromatin and have not been detected in the cytosol of hormone-responsive cells. Based on this information and the current studies, the hypothesis might also be considered that the intranuclear receptors are derived from a “core” subunit of the receptor that is synthesized in the cytosol. Such a subunit might also have properties similar to the binding activity in preparations that have been heated in terms of binding T, more avidly than TB. Such a core subunit could gain its high affinity TX binding properties upon its localization in chromatin. In preparations that have lost Ts binding activity, the estimated equilibrium dissociation constants4 of the binding activity for T4 and Ts are 10 to 50 nM and 0.7 pM, respectively (Fig. 3), similar to TBPA (1). These findings and the similarities of TBPA to the receptor in molecular weight (4, 5, 8, 9) and a putative DNA binding site (3) could imply that TBPA (or perhaps other proteins that bind Tq more avidly than T3) and the receptor are more closely related than has been previously apparent. In fact, it is possible that such “core” receptors contain subunits (or modified subunits) in common with one or more of the other proteins. Such a situation might explain the evolutionary stability of proteins, such as the plasma binding proteins which have no known obligatory functions; a loss of these might also mean a loss of the receptor. Finally, if the thyroid hormone receptor can be converted to a form with a binding specificity that resembles another thyroid hormone-binding protein, then receptors and other proteins need not differ in some fundamental binding structure even if they do differ in the native state in the way they recognize various hormones.

REFERENCES 1. Ferguson, R. N., Edelhoch, H., Saroff, H. A., and Robbins, J.

’ The apparent equilibrium dissociation constant was calculated from the equation (8): KX = Csn.K~/(K~ + T), where KH is the apparent equilibrium dissociation constant of the competitor (unlabeled T.]), KT is the apparent equilibrium dissociation constant of [‘““I]T4, C&O is the concentration of competitor that produces 50% inhibition, and 2’ is the concentration of [“‘I]T+

2.

3. 4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20. 21.

22.

23

24 25

26. 27. 28.

29.

30.

31.

32.

33.

34.

35.

36. 37.

38.

39.

40.

(1975) Biochemistry 14, 282-289 Snyder, S. M., Cavalieri, R. R., Goldfine, I. D., Ingbar, S. H., and

Jorgensen, E. C. (1976) J. Biol. Chem. 251,6489-6494 Blake, C. C. F., and Oatley, S. J. (1977) Nature 268, 115-120 Blake, C. C. F., Swan, I. D. A., Rerat, C., Berthou, J., Laurent, A.,

and Rerat, B. (1971) J. Mol. Biol. 61, 217-224 Kanda, Y., Goodman, D. S., Canfield, R. E., and Morgan, F. J.

(1974) J. Biol. Chem. 249,6796-6805 Davis, P. J., Handwerger, B. S., and Glaser, F. (1974) J. Biol.

Chem. 249,6208-6217 DiIIman, W., Surks, M. I., and Oppenheimer, J. H. (1974) Endo-

crinology 96,492-498 Latham, K. R., Ring, J. C., and Baxter, J. D. (1976) J. Biol. Chem.

251,7388-7397 Surks, M. I., Koerner, D., Dillman, W., and Oppenheimer, J. H.

(1973) J. Biol. Chem. 248,7066-7072 MacLeod, K. M., and Baxter, J. D. (1975) Biochem. Biophys. Res.

Commun. 62,577-583 MacLeod, K. M., and Baxter, J. D. (1976) J. Biol. Chem. 251.

7380-7387 Oppenheimer, J. H., Schwartz, H. L., Dillman, W., and Surks, M.

I. (1973) Biochem. Biophys. Res. Commun. 55. 544-550 Samuels, H. H., Tsai, J.-S.; and Cintron, R. (1973) Science 181,

1253-1256 De Groot, L. H., and Torresani, J. (1976) Endocrinology 96,357-

369 Koerner, D., Schwartz, H. L., Surks, M. I., Oppenheimer, J. H.,

and Jorgensen, E. C. (1975) J. Biol. Chem. 250, 6417-6423 Tsai, J. S., and Samuels, H. H. (1974) Biochem. Biophys. Res.

Commun. 59, 420-428 Oppenheimer, J. H., Schwartz, H. L., and Surks, M. I. (1974)

Endocrinology 95,897-903 Spindler, B. J., MacLeod, K. M., Ring, J., and Baxter, J. D. (1975)

J. Biol. Chem. 250,4113-4119 Oppenheimer, J. H., Schwartz, H. L., Surks, M. I., Koerner, D.,

and Dillman, W. H. (1976) Recent Prog. Hormone Res. 32, 529-565

Tata, J. R., and Widnell, C. C. (1966) Biochem. J. g&604-620 Martial, J. A., Baxter, J. D., Goodman, H. M., and Seeburg, P. H.

(1977) Proc. Natl. Acad. Sci. U. S. A. 74, 1816-1820 Seo, H., Vassart, G., Brocas, H., and Refetoff, S. (1977) Proc.

Natl. Acad. Sci. U. S. A. 74, 2054-2058 Kurtz, D. T., Sippel, A. E., and Feigelson, P. (1976) Biochemistry

15, 1031-1036 Scatchard, G. (1949) Ann. N. Y. Acad. Sci. 51, 73-77 Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

(1951) J. Biol. Chem. 193, 265-275 Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 Burton, K. (1956) Biochem. J. 62, 315-323 Samuels. H. H.. and Tsai. J. S. (1973) Proc. Natl. Acad. Sci. U.

S. A. iO,3488-3492 ’ DeGroot, L. J., and Strausser, J. L. (1974) Endocrinology 95,74-

83 Oppenheimer, J. H., Schwartz, H. L., Koerner, D., and Surks, M.

-1: (1974) J. Clin. Invest. 53, 768-777 DeGroot, L. J., Refetoff, S., Strausser, J., and Barsano, C. (1974)

Proc. Natl. Acad. Sci. U. S. A. 71, 4042-4046 Dixon, M., and Webb, E. (1964) in Enzymes (Boyer, P. D., ed)

2nd Ed, nn. 328-330. Academic Press. New York Fruchter, I? G., and Crestfled, A. M. (i967) J. Biol. Chem. 242,

5807-5812 Samuels, H. H., Stanley, F., and Shapiro, L. E. (1977) J. Biol.

Chem. 252,6052-6060 Holm, A. C., Lemarchand-Beraud, T., Scazziga, B. R., and Cuttle-

lod, S. (1975) Acta Endocrinol. 80, 642-656 Schussler, G. C., and Orlando, J. (1978) Science 199, 686-687 Burman, K. D., Lubes, Y., Wright, F. D., and Wartofsky, L. (1977)

Endocrinology 101, 1331-1334 DeGroot, L. J., Coleoni, A. H., Rue, P. A., Seo, H., Martino, E.,

and Refetoff, S. (1977) Biochem. Biophys. Res. Commun. 79, 173-178

Samuels, H. H., Stanley, F., and Shapiro, L. E. (1976) Proc. Natl. Acad. Sci. U. S. A. 73, 3877-3881

Surks, M. I., and Oppenheimer, J. H. (1977) J. Clin. Inuest. 60, 555-562


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

N L Eberhardt, J C Ring, K R Latham and J D BaxterThyroid hormone receptors. Alteration of hormone-binding specificity.

1979, 254:8534-8539.J. Biol. Chem.

http://www.jbc.org/content/254/17/8534.citation

Access the most updated version of this article at

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/254/17/8534.citation.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/content/254/17/8534.citation

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;254/17/8534&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/254/17/8534.citation

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=254/17/8534&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/254/17/8534.citation

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/254/17/8534.citation.full.html#ref-list-1

http://www.jbc.org/

Vol. 254, No. li, Issue of September 10, pp. 8534-8539. 19i9 … · 2002-12-05 · Vol. 254, No. li, Issue of September 10, pp. 8534-8539. 19i9 I’rmfed in I! S.A Thyroid Hormone

Documents