- Email: [email protected]
Murine glucocorticoid receptors: New evidence for a discrete receptor influenced by H-2
ARCHIVES
OF BIOCHEMISTRY
Vol. 249, No. 2, September,
AND
BIOPHYSICS
pp. 316-325,1986
Murine Glucocorticoid Receptors: New Evidence for a Discrete Receptor Influenced by H-2 ALLEN Center
for
Craniofacial
S. GOLDMAN’
AND
MASUYUKI
KATSUMATA
Anmnalies and Department of Pediatrics, University of Illinois, Chicago, 808 South Wood St., Box 6998, Chicago, Illinois 60680 Received
March
l&1986,
and in revised
form
May
College
of Medicine
at
9, 1986
The glucocorticoid receptor contents in the lungs of females of two congenic strains of mice, BIO.A (H-2a) and BlO (H-Zb), differing only in the H-2 histocompatibility region of chromosome 17, have been measured by the dextran-charcoal method and by our previously described methods of molecular sieving and ion exchange chromatography [M. Katsumata, C. Gupta, and A. S. Goldman (1985) Arch. Biochem. Biophys. 243,3853951. As reported, two receptors, II and IB, are demonstrable by each column chromatographic method, and 5,5-diphenylhydantoin binds to receptor IB but not to receptor II. Receptor IB cannot be detected unless molybdate is added in cytosols prepared with hypotonic buffer [lo mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid and 10 mM dithiothreitol, pH 7.35) according to S. L. Liu, J. F. Grippo, R. P. Erickson, and W. B. Pratt (1984) J. Steroid Biochem. 21,633-6371, a method which has been reported to give maximal receptor levels. Using hypotonic buffer containing 10 mM molybdate we observed a small but significantly higher content of receptor IB in BIO.A mice than that in BlO mice, but no significant difference in receptor II or total receptor content. On the other hand, cytosols prepared with isotonic buffer (50 mM Tris-HCl, 120 mM NaCl, 1 mM EDTA, 10 mM dithiothreitol, and 10 mM molybdate, a modification of the buffer used in our previous report) contained significantly higher levels of receptor IB and of total binding in pulmonary cytosols of BIO.A as compared to those of BlO. There was no difference in receptor II content. Molybdate stabilizes receptor IB in both buffers. These results explain the apparent contradiction between our results and those of Liu et al. by showing that the hypotonic buffer used by them allows for determination of maximal levels of receptor II, but permits selective destruction of receptor IB. However, the use of isotonic buffer gives maximal values of both receptors II and IB. With isotonic buffer, it is demonstrated that only the level of receptor IB is influenced by H-Z-linked genes. 0 1986 Academic
Press, Inc.
lhydantoin (phenytoin, DPH)’ in rodents may utilize the same biochemical mechanism as that by which glucocorticoids pro-
Susceptibility of mice to cortisone(l-5) and phenytoin- (6, 7) induced cleft palate is influenced by genes linked to the H-2 histocompatibility locus on chromosome 17 of the mouse. Evidence from our laboratory has suggested that the clefting action of glucocorticoids and 5,5-dipheny’ To whom
0003-9861/86 Copyright All rights
correspondence
should
$3.00
0 1986 by Academic Press, Inc. of reproduction in any form reserved.
a Abbreviations used: DPH, 5,5-diphenylhydantoin; PLIP, phospholipase-inhibitory protein; DEX, dexamethasone, 9-cY-fluoro-16-a-methyl-ll~,l7a,21-trihydroxypregna-1,4-diene-3,2-dione; DTT, dithiothreitol; Hepes, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid.
be addressed. 316
MURINE
GLUCOCORTICOID
RECEPTOR
duce their anti-inflammatory effects (6-11). It is generally believed that the anti-inflammatory pathway of glucocorticoids involves a receptor-mediated production of phospholipase-inhibitory proteins (PLIPs) that inhibit the release of arachidonic acid from membrane phospholipids, with a subsequent diminished production of prostaglandins and thromboxanes by cyclooxygenase and of leukotrienes by lipoxygenase (12-14). Similar mechanisms appear to be involved in the induction of cleft palate, as exogenous arachidonic acid injected at the same time as glucocorticoids markedly reduces the teratogenic potency of the steroids, and indomethacin, an inhibitor of cyclooxygenase, blocks the corrective action of arachidonic acid (15,16). Indomethacin also blocks the corrective action of arachidonic acid on the inhibition of programmed cell death of embryonic palates produced by cortisol in culture, and inhibits this programmed cell death by itself (16). It has been also reported that phenylbutazone, another inhibitor of cyclooxygenase, blocks the fusion of embryonic palatal shelves in a culture model (1’7). We have suggested earlier that the way by which H-Z-linked genes affect susceptibility to glucocorticoid-induced cleft palate may be explained by their influence on glucocorticoid receptor levels in the embryonic palate, since the same H-Z-linked genes influence both susceptibility to glucocorticoid-induced cleft palate and glucocorticoid receptor levels (18-21). In this connection we have explained the H-2 influence on DPH-induced cleft palate by this same effect on susceptibility to DPH and glucocorticoid receptor levels (6, ‘7, 9, 22), since DPH competes with dexamethasone (DEX) for the glucocorticoid receptor affecting PLIP production and inhibition of prostaglandin production (10, 11, 22-24). If receptor level is a primary site of HZ influence on susceptibility and if this receptor level does affect the arachidonic acid cascade, one would expect to be able to show an H-2 influence at each step in the biochemical pathway. To date we have shown an H-2 influence directly in the embryonic palate on (i) specific [3H]DEX
INFLUENCED
BY
H-2
317
binding (19), (ii) induction of PLIPs by DEX (23), and (iii) inhibition of arachidonic acid release and prostaglandin synthesis (10). We have also shown that H-&linked genes influence the level of total specific binding of [3H]DPH in the same adult tissues, thymuses, and lungs in the same way as they influence glucocorticoid receptors (9,20) and that the DPH-induced inhibition of arachidonic acid release and prostaglandin synthesis in the embryonic palate is also influenced by H-2 (10, 23). Thus, these observations taken as a whole provide supportive evidence from a variety of different experimental approaches that DPH and glucocorticoids bind a common receptor affecting the teratogenic and antiinflammatory functions of glucocorticoids and that the level of this receptor is influenced by H-blinked genes. Recently, an effect of H-2 on binding capacity of glucocorticoids by pulmonary cytosols was reported by Liu et al. (25) to have been undetectable when glucocorticoid receptors were assayed in the presence of a sulfhydryl reagent for stabilizing the receptor which also maximizes binding capacity. This methodology included the use of an overnight incubation in the presence of 10 mM dithiothreitol (DTT) with or without 10 mM molybdate in 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (Hepes) buffer (hypotonic). This method gave values of receptor levels in the lung in the range of 250-300 fmol/mg cytosol protein for both A/J (H-2 a) and C5’7BL/6 (H-2 b) strains, and the levels were higher than we have previously reported (20). These levels required the presence of 10 mM DTT, but incubation with 10 mM molybdate did not affect the levels. We have just reported a study of pulmonary cytosols using isotonic buffer, 10 mM P-mercaptoethanol as sulfhydryl reagent, and [3H]DEX and [3H]DPH as ligands (22). In this study, receptors were measured in the same range by molecular sieving and ion exchange chromatography as those in the study by Liu et al. (25). This study has also provided evidence for the existence of the classical receptor (receptor II) which binds only DEX and a receptor
318
GOLDMAN
AND
which binds both DEX and DPH [similar to receptor IB of Litwack and Rosenfield (26)]. Receptor IB appears to mediate the anti-inflammatory and teratogenic effects of glucocorticoids and DPH, and may be controlled by gene(s) different from that of the classical receptor (22). Because of the apparent discrepancies between the conclusions of Liu et al. (25) and of us (20, 22), we have compared the use of an overnight incubation with 10 mM DTT and 10 mM molybdate either in hypotonic medium (method of Liu et al) or in isotonic medium (our modified method). We have also studied the use of both the dextran-charcoal and molecular sieving methods for determining the DEX-binding capacities of pulmonary cytosols of strain BIO.A and of its congenic partner strain, BlO. We report that determination of receptor level in cytosols prepared in hypotonic buffer preserves maximal levels of receptor II only, but preparation of cytosols in isotonic buffer allows the preservation of maximal levels of both receptors II and IB. Furthermore, the level of receptor IB is influenced by H-2, but that of receptor II is not. MATERIALS
AND
METHODS
Animals. BlO.A/SgSn (BIO.A) and C57BL/lOSn (BlO) mice were obtained from the Jackson Laboratory, Bar Harbor, Maine. Radioisotwpes. [aH]DEX, [1,2,4(N)-3H]DEX (41 Ci/ mmol), was purchased from Amersham Corporation. The purity of the isotope was confirmed to be more than 97% based on the radioactivity by TLC. Cytosol preparation Cytosols of adult female mouse lungs were prepared by homogenizing the tissue in 2 vol (w/v) of one of the following four kinds of buffer: isotonic buffer, consisting of 50 mM Tris-HCl, I20 mM NaCl, 1 mM EDTA, and 10 mM DTT with or without 10 mM sodium molybdate, pH 7.7; and hypotonic buffer consisting of 10 mM Hepes and 10 mM DTT with or without 10 mM sodium molybdate, pH 7.35. The homogenate was centrifuged at 100,OOOg for 60 min, and the supernatant was used as cytosol. [‘HJDEX-receptor complezea An aliquot of cytosol preparation was mixed with [aH]DEX (50 nM, final concentration) with or without a cold competitor (DEX or DPH, 5 PM). Cold DPH was dissolved in l/10 vol of 50 mM sodium carbonate, and added in the incubation mixture. The same volume of sodium carbonate solution was also added in the incubation mixtures of [aH]DEX with and without cold DEX as a
KATSUMATA control adjustment. The incubation was carried out at 0°C for 18 h. In some experiments, we incubated the mixtures for 60 min at 0°C followed by 25°C for 20 min to activate the receptors (27). Molecular sting chromatography. Sephadex G-200 and Sephadex G-25 medium gels were mixed to provide better separation and faster elution, equilibrated with 20 mM K phosphate buffer containing 50 mM KCl, pH 7.0, and packed in a column, 2 X 30 cm, as described previously (22). The sample volume was 1.0 ml, and the receptor complexes were eluted with the same buffer with a flow rate of 0.5 ml/min. Ferritin and cytochrome c were the inner markers. Ion exchange column chromatography with hydroxyapatite adsorption A DEAE-Sephadex A-50 column made from a 25-ml pipet (bed volume, 10 ml) was equilibrated with 20 mM K phosphate buffer containing 50 mM KCl, pH 7.0 (22). The salt concentration of the sample solution was adjusted by dilution with plain water. After the sample solution (2-5 ml) was loaded, the column was washed with 10 ml of the phosphate buffer mentioned above. The elution was carried out with a linear gradient system from the same buffer to 20 mM K phosphate buffer containing 1.0 M KCI, pH 7.0. The salt concentration in the eluate was determined by electrical conductivity. Each l.Oml fraction was mixed with 1.0 ml of a 25% suspension of hydroxyapatite, pH 6.9, and was incubated at 5°C for 10 min to adsorb the rH]DEX-carrying receptor (27,28). The receptor adsorbed by hydroxyapatite was pelleted by centrifugation at SOOg for 5 min, followed by two washings with 20 mM K phosphate buffer containing 50 mM KCl, pH 7.0. The hydroxyapatite pellet was counted directly by suspending it in Aquasol (New England Nuclear Corp.). Calculdim of the level of receptor II and IB. On the basis of our previous report (22), the contents of receptors IB and II were calculated from the results of molecular sieving chromatographic separation as follows. The total receptor content was determined by subtracting the rH]DEX bound to protein in the presence of cold DEX from the bound [aH]DEX in the absence of cold competitor. Receptor IB content was determined by subtracting the bound rH]DEX in the presence of cold DPH from the bound [3HpEX in the absence of cold competitor. Receptor II content was determined by subtracting the bound rH]DEX in the presence of cold DEX from the bound [aH]DEX in the presence of cold DPH. Determination of total receptor content by dextrancharcoal method. The total [aH]DEX binding capacity of pulmonary cytosol was determined by the dextrancharcoal method as previously described (19). Two kinds of cytosolic preparations of both BIO.A and BlO strains, hypotonic with molybdate and isotonic with molybdate, were measured and compared. Statistics. All results are expressed as means -C SD
MURINE
GLUCOCORTICOID
(number of observations). The for significance by the two-tailed
RECEPTOR
INFLUENCED
BY
H-2
319
results are analyzed Student’s t test.
RESULTS
Total Receptor Content Determined Dextran-Charcoal Method
by the
Table I shows that there is no significant difference between the total receptor content of BIO.A and BlO mice when hypotonic buffer is used for the preparation of cytosol, but that a highly significant large difference is observed when the isotonic buffer is used. The values we have shown in this table are quite close to the values of Liu et
al. (25). Comparison of the Chromatographic Patterns of Pulmonary Cytosolic Receptors of Bl0.A and BlO Female Mice Figure 1 shows in molecular sieving chromatography three DEX-labeled peaks: 7, 5, and 3.5 nm Stokes’ radii. The 7-nm peak contains DPH-blockable and DPHnonblockable complexes. The 5-nm peak is not affected by DPH, whereas the 3.5-nm peak is blocked by DPH. The ‘7-nm complex TABLE
I
COMPARISONOFDETERMINATIONOFTOTAL GLUCOCORTICOIDRECEPTORSINPLJLMONARY CYTOSOLS BYMETHODUSINGHYPOTONICBUFFERAND BYTHATUSING ISOTONICBUFFER Binding
Strain BIO.A B10 P value (t statistic)
Hypotonic (fmol/mg protein)
capacity Isotonic (fmol/mg protein)
243.5 + 25.5 (6) 236.5 f 6.2 (6)
332.0 + 25.5 (6) 261.2 -c 17.0 (6)
NS
Note. Numbers are Means + SD (number of determinations). The cytosol preparations, both hypotonic and isotonic, contained 10 mM sodium molybdate. The cytosolic preparations were incubated with 50 nM [3H]DEX in the presence or absence of 5 pM cold DEX at 0°C for 18 h; then each incubation mixture was treated with dextran-charcoal.
FRACTION
NUMBER
FIG. 1. Comparison of DEX-receptor complexes of pulmonary cytosols of BIO.A and BlO prepared with isotonic buffer and molybdate by molecular sieving. The pulmonary cytosol as prepared with isotonic buffer containing 10 mM sodium molybdate. The cytosol was incubated with 50 nM [3H]DEX in the presence and absence of cold competitors, DEX and DPH (5 PM), at 0°C for 18 h. Then a l-ml aliquot of each incubation mixture was gel-filtered through a Sephadex G-200-G-25 mixed gel column (2 X 30 cm) using 20 mM K phosphate buffer containing 50 mM KCl, pH 7.0. Left panel shows the molecular sieving pattern of the cytosol of BlO.A, right panel, that of BlO. 0, [3H]DEX in the absence of cold competitors; 0, [3H]DEX in the presence of 5 pM DPH, n , [3H]DEX in the presence of 5 pM DEX.
is a mixture of aggregates of activated receptor II and receptor IB, the 5-nm complex is activated receptor II, and the 3.5-nm complex is activated receptor IB, since each of these complexes is adsorbed by DNAcellulose (22). The content of receptor IB in pulmonary cytosol of the BIO.A strain is obviously higher than that of BlO, whereas the content of receptor II is not different (Fig. 1). A pronounced influence of H-2 can be seen. Figure 2 indicates that activated receptor IB elutes as both a wash and a 0.14 M KC1 peak in the DEAE-Sephadex A-50 column-hydroxyapatite method, whereas activated receptor II elutes as a 0.2 M KC1 peak, since each of these peaks is adsorbed by DNA-cellulose again as previously described (22). The DPH-blockable receptors, the wash, and 0.14 M complexes (receptor IB), are adsorbed by hydroxyapatite. Thus, the free ligand which also appears in the wash has been eliminated. This result with hydroxyapatite confirms our previous report determined by the dextran-charcoal.
320
GOLDMAN
AND
0.4 E 0.2 B
10
20
so FRACTION
40
10 20 NUtlEER
30
40
FIG. 2. Comparison of DEX-receptor complexes of pulmonary cytosols of BIO.A and BlO prepared with isotonic buffer and molybdate by DEAE-Sephadex A50 column chromatography. The preparation of cytosol and incubation with rH]DEX were carried out as described in Fig. 1. A l-ml aliquot of each incubation mixture was diluted with 1 ml of ice-cold water, and was loaded on a DEAE-Sephadex A-50 column (bed volume, 10 ml). rH]DEX-carrying proteins were eluted with 10 ml wash of 20 mM K phosphate buffer containing 50 mM KCl, pH 7.0, followed by KC1 linear gradient system from 50 mM to 1 M KCI. Each l-ml fraction was treated with hydroxyapatite. The [aH]DEX-carrying protein was adsorbed by hydroxyapatite, and directly determined by suspending the hydroxyapatite pellet in scintillation fluid after washing. Panels and symbols are the same as in Fig. 1.
The use of hydroxyapatite gives a better resolution between receptor IB and II than the previous method because there is more dissociation of ligand from the receptor IB complex with dextran-charcoal(22). Again BIO.A has obviously a higher content of receptor IB than BlO, whereas practically no strain difference is demonstrable in the content of receptor II (Fig. 2). The content determined by molecular sieving gives a value higher than that determined by the ion-exchange-hydroxyapatite method. From determinations using the same cytosolic preparation of BlO.A, molecular sieving gives 133 fmol/ mg protein of receptor IB and 165 fmol/ mg protein of II, whereas the ion-exchange-hydroxyapatite method gives 60 fmol/mg protein of receptor IB and 1’78 fmol/mg protein of receptor II, respectively. The total recovery of receptor proteins by molecular sieving chromatography is between 85 and 92% (assuming the dextran-charcoal method has 100% recovery).
KATSUMATA
Therefore, the content of receptor IB determined by the ion-exchange-hydroxyapatite method is approximately 40% of the total, and the lower value of recovery implies a dissociation of the bound ligand from the receptor IB in the process of separation. Figure 3 shows the molecular sieving pattern of the cytosols prepared with hypotonic buffer. A higher content of receptor IB is observed in BIO.A than in BlO, although much less than that in the cytosol prepared with isotonic buffer. Moreover, the amount of receptor IB in the ‘7- and 3.5nm peaks is much higher in the cytosol prepared with isotonic buffer (Fig. 1, left) than in the cytosol prepared with hypotonic buffer (Fig. 3, left), but the amount of receptor II is not affected by the osmolality of the buffer. The total receptor content and the contents of these two receptors in each buffer have been determined by molecular sieving chromatography and are shown in Table II. The total content is significantly higher in BIO.A than in BlO only when prepared with isotonic buffer. The content of receptor IB is significantly higher in BIO.A than in BlO by the use of either isotonic or hypotonic buffer. The content of receptor II is not significantly different in the two strains. The cytosol prepared with isotonic
-I
F
10
20
1
I
30 40 FRACTION
10
20 NUPlBER
30
40
1
FIG. 3. Comparison of DEX-receptor complexes oi pulmonary cytosols of BIO.A and BlO prepared with hypotonic buffer and molybdate by molecular sieving. The pulmonary cytosol was prepared using hypotonie buffer containing 10 mM sodium molybdate. Other procedures were the same as in Fig. 1. Panels and symbols are the same as in Fig. 1.
MURINE
GLUCOCORTICOID
RECEPTOR TABLE
INFLUENCED
BY
321
H-2
II
DEXAMETHASONE RECEPTORSIB AND II IN PULMONARY CYTOSOLICPREPARATIONS OF BIO.A AND BlO MICE AFTER MOLECULAR SIEVING Binding
capacity
(fmol/mg
Strain Hypotonic Receptor Receptor Total
protein) BIO.A
B10 buffer IB II
with
P value (t statistic) BlO vs BIO.A
molybdate. Incubation: 11 f 4.7 (7) 201 f 37.4 (7) 211 7t 34.8 (7)
overnight
at 24 f 203 f 227 f
0°C. 14.4 (8) 40.7 (8) 39.3 (8)
<0.05 NS NS
Isotonic buffer with molybdate. Incubation: overnight at 0°C. Receptor Receptor Total
15 f 5.1 (5) 207 + 27.8 (5) 222 f 32.6 (5)
IB II
P value (t statistic) Receptor IB Receptor II Total Note. The contents the following formula:
hypotonic
vs isotonic. NS NS NS
of the two receptors
Receptor
were
IB = ([3H]DEX
= [3H]DEX Receptor
II = [3H]DEX
without with without with
- ([3H]DEX
The cytosols were incubation mixture
111 f 46.5 (6) 199 k 40.3 (6) 309 + 15.8 (6)
from
competitor DPH
- [3H]DEX
of molecular
- [3H]DEX
- [3H]DEX
competitor DPH
the results
with
with
DEX)
- r3H]DEX
with
with
prepared as described in Table I. After the incubation was directly gel-filtered as described in Fig. 1.
sieving,
according
to
DEX)
DPH,
DEX. with
r3H]DEX
at 0°C for
18 h, the
buffer contains a significantly higher content of receptor IB than that prepared with hypotonic buffer, but no difference in the contents of receptor II is also observed. Molybdate Efect Figure 4 shows the effect of molybdate on the receptor content in the cytosol prepared with hypotonic buffer according to Liu et al. (25). With molybdate, receptor IB is detectable mainly as the 7-nm complex (Fig. 4, left), but without molybdate there is practically no receptor IB (Fig. 4, right), while the level of receptor II is not affected by the presence of molybdate. Eflect of Incubation Time Other than the osomolality of the buffer, the method of Liu et al. (25) differs from
FRACTION
NUllBER
FIG. 4. Effect of molybdate on the DEX receptors in the cytosol prepared with hypotonic buffer. Cytosol was prepared from lungs of Bl0.A mice using hypo-
tonic buffer with or without 10 mM sodium molybdate. The cytosol was incubated at 0°C for 18 h. Left shows the molecular sieving patterns of DEX tors in the cytosol containing molybdate; right that containing no molybdate. Symbols are the as in Fig. 1.
panel receppanel, same
322
GOLDMAN
AND
ours (22) in incubation time, overnight incubation at 0°C (25) and 60 min at 5°C followed by 20 min at 25°C incubation (22). When molybdate is absent, receptor IB disappears in the long-term incubation in the cytosol of isotonic buffer, but if the incubation time is short, we detect a high content of receptor IB (Fig. 6). Long-term incubation results in little change of the content of receptor II, but the size has changed to a smaller complex, from 5 to 3 nm, which is similar to the mero-receptor (29) (Fig. 6). In the presence of molybdate, the length of incubation causes little change (Fig. 5). In short-term incubation, we can detect a high con tent of receptor IB without molybdate, but still a little less than we obtained in the presence of molybdate (1’73 vs 123 fmol/mg protein) (Fig. 5, left and Fig. 6, left).
KATSUMATA
FRACTION
FIG. 6. Effect of incubation tors in the cytosol prepared cytosol was prepared using molybdate. Other conditions Fig. 5.
NUMBER
time on the DEX recepwithout molybdate. The isotonic buffer without are identical to those in
Efect of EDTA The presence or absence of 1 mM EDTA makes no appreciable difference in the receptor assay (data not shown).
FRACTION
NUMBER
FIG. 5. Effect of incubation time on the Dex receptors in the cytosol prepared with isotonic buffer with 10 mM molybdate. The cytosol was prepared from lungs of female BIO.A mice using isotonic buffer with 10 mM molybdate. The cytosol was incubated either at 0°C for 60 min followed by 20 min at 25’C (left) or 0°C for 18 h (right). Symbols are the same as in Fig. 1.
DISCUSSION
The present report has confirmed the report of Liu et al. (25) that the use of an overnight incubation with or without 10 mM molybdate in hypotonic buffer with 10 mM sulfhydryl reagent gives higher values of total glucocorticoid receptors in mouse lungs without an H-2 influence than those measured earlier by us with an H-2 influence by a method using a short incubation time of cytosols prepared without a sulfhydryl reagent (20). However, we recently have provided evidence for the existence of a discrete glucocorticoid receptor, receptor IB, which binds both DEX and DPH, which is adsorbed by DNA-cellulose, which mediates the anti-inflammatory and teratogenic actions of glucocorticoids, and which is different from the classical receptor II (22). The present results also indicate that this discrete receptor in the lungs is stabilized by molybdate. The main difference between the two methodologies rests on the fact that the method of Liu et al. (25) measures maximal levels of receptor II, while our present method measures maximal levels of both receptors II and IB. Thus, both methods show with column chromatography that the levels of receptor
MURINE
GLUCOCORTICOID
RECEPTOR
II are not affected by H-2. However, the measurement of the levels of receptor IB and its dependence on H-2-linked genes can be demonstrated only with isotonic buffer. The key factor in these differences is the use of hypotonic or isotonic buffer for the preparation of the cytosols. In general, those studies which have used isotonic (isoosmotic) buffers for the preparation of cytosols or in binding studies of cells in isotonic media show an influence of H-2 (9, 18-20, 30) or the existence of receptor IB (22, 26, 27, 31-36), while those which use hypotonic buffers fail to demonstrate an influence of H-2 (25,37,38). Because of the similarity of the properties of the receptor which binds both DEX and DPH to that described by Litwack and colleagues (26, 27, 31-34), we have considered it to be receptor IB as first designated by them (22). The activated receptor has a Stokes’ radius of 3.0-3.5 nm, and can aggregate to a much larger form. Unlike receptor II, IB has a wide specificity of steroid binding (22, 26), can bind DPH (22), and has the capacity to rebind ligand lost in the process of purification (22,32,39). The inhibition of the release of prelabeled [3H]arachidonic acid by DEX and DPH occurs only in the cells which have receptor IB, e.g., thymocytes (22). DPH binds only to receptor IB, i.e., 7- and 3.5-nm forms, and both forms of the DPH-receptor complex are adsorbed by DNA-cellulose (22). The DPH-receptor complex is incorporated into nuclei of thymocytes (11). DPH depresses the release of arachidonic acid from the cells such as thymocytes and lung cells, which have receptor IB (with II) (22, 23). Hepatoma G2 cells contain only receptor II, which does not bind DPH, and neither DPH nor DEX affects the release of arachidonic acid from these cells (22). Receptor IB is most likely, therefore, responsible for mediation of the production of phospholipase-inhibitory proteins induced by DEX (9,12-14,20) and DPH (11,22-24). It is thought that these proteins mediate the anti-inflammatory and teratogenic action of glucocorticoids (9, 12-16, 20) and DPH (11,22-24) by inhibiting the release of arachidonic acid from membrane phospho-
INFLUENCED
BY
H-2
323
lipids with a subsequent reduction in the synthesis of prostaglandins and leukotrienes. Receptor IB also appears to mediate the stimulation by glucocorticoids of cation transport in rat colonic epithelia, which contain only receptor IB (35). Recently, it has also been shown that rat embryonic yolk sac, the anlagen of the colonic epithelia, also contains only receptor IB (36, 50). In support of these observations is the finding that DPH also stimulates cation transport in rat jejunum (40), since we have shown that DPH binds to receptor IB as a glucocorticoid agonist (11, 22, 23). Receptor IB is not likely to be the “nicked” glucocorticoid receptor of Wrange and Gustafsson (29) or the proteolytic fragment of Sherman et al. (41) because of several observations. The present finding that the level of receptor II remains constant when the level of IB is markedly elevated by a change from hypotonic to isotonic buffer argues against the origin of receptor IB from II by proteolysis. This conclusion is supported by the facts that, unlike receptor II, the steroid-binding site of IB binds DPH and other steroids having a different molecular structure and, also unlike receptor II, activated EB can rebind ligand (22,32,39). The existence of receptor IB in rat liver has been confirmed by Hirota et al. (42), who have also provided evidence for the existence of a discrete third receptor which appears only under certain conditions and which mediates the induction of tryptophan dioxygenase, but not that of tyrosine aminotransferase by glucocorticoids. The induction of the activity of tyrosine aminotransferase by glucocorticoids (43-46) is mediated by receptor II in hepatoma G2 cells, which have receptor II exclusively (22). It has been shown that the major receptor involved in the killing or growth-inhibitory effect of glucocorticoids in glucocorticoid-sensitive cells, S49, is most probably receptor II, since these cells contain predominantly the receptor whose size is identical to that of the 5-nm receptor (47, 22). The killing or growth-inhibitory effect of glucocorticoids in S49 cells is coded by genes on mouse chromosome 18 (48). On
324
GOLDMAN
AND
the other hand, H-blinked genes on mouse chromosome 17 influence suceptibility to cortisone- (l-5) and DPH-induced (6, 7) cleft palate, DEX- and DPH-receptor levels (9, l&20), and the degree of production of PLIPs by DEX and DPH (23) and of inhibition of the arachidonic acid cascade (10). A separation of the growth-inhibitory function of glucocorticoids from that of inhibition of phospholipase by glucocorticoids has been demonstrated in embryonic palatal mesenchyme cells (49). Moreover, the phospholipase-inhibitory effect is evidently influenced by H-2, but the growthinhibitory effect is not (49). The growthinhibitory function of glucocorticoids is not different in degree in palate mesenchyme cells from strains different in H-2, but the degree of phospholipase-inhibitory action is (49). The separation of the function of the third receptor from that of receptor II suggests a third discrete receptor also under separate genetic control (42). Thus, this report taken with those studying genetic control, receptor location, receptor properties, and receptor function suggests the working hypothesis of receptor polymorphism with each receptor having a different function(s) under separate genetic control. This hypothesis suggests that receptor II mediates induction of tyrosine aminotransferase and growth-inhibitory functions of glucocorticoids and is coded by genes on chromosome 18, while genes in chromosome 17 influence the level of receptor IB and its effects on the arachidonic acid cascade and intestinal cation transport. The hypothesis favors the concept of receptor polymorphism in the controversy of whether receptor IB is an isomorphic form of receptor II found predominantly in kidney and colon (35) or yolk sac (36,50) or whether it is a proteolytic product of receptor II (29,41).
KATSUMATA 4. GASSER, MAN,
J. J., AND
SLAVKIN,
C. (1975)
ImmuGenetics
Proc.
A. S., BAKER, A.
S.,
M.
Immunc-
L. (1984)
M. K., AND
86. 8. GOLDMAN,
GASSER,
D. L.
18,17-22.
FISHMAN,
C.
Proc. Sot. Exp.
M. K. (1983) A.
S., BAKER,
AND HEROLD,
R. (1983)
L.,
AND
BAKER,
Biol.
Med.
173,82-
M. K.,
Proc
PIDDINGTON,
R.,
Sot. Exp. Bid
Med.
174,239-243. 9. GUPTA,
C., KATSUMATA,
(1984) 10.
GUPTA,
M., AND
Immunogenetics
GOLDMAN,
A. S.
20,667-676.
C., KATSUMATA,
(1985)
M., AND
Genet.
J. Craniofacial
GOLDMAN,
A. S.
Dew. Biol
5,277-
285. 11.
KATSUMATA, DORF,
M.,
GUPTA,
C. E., AND
C., BAKER,
GOLDMAN,
M. K., Suss-
A. S. (1982)
Science
218,1313-1315. 12. HIRATA,
F.,
MANIAN, 13.
14. 15.
SCHIFFMANN, K., SALOMON,
E., VENKATASUBRAD., AND AXELROD,
Nature
(1980) GHIARA,
(London)
P., MELI,
L., AND
PERSICO,
P. (1984) Biochem. Pharmacol. 33,1445-1450. TZORTZATOU, G. G., GOLDMAN, A. S., AND BouTWELL, W. C. (1981) Proc Sot. Exp. Biol Med.
MONTENEGRO,
(1982) 18.
287,147-149.
R., PARENTE,
166,321-324. 16. PIDDINGTON, R., HEROLD, (1983) Proc. Sot. Exp. 17.
J.
(1980) Proc. Natl. Acad. Sci. USA 77,2533-2536. BLACKWELL, G. J., CARNUCCIO, R., DI ROSA, M., FLOWER, R. J., PARENTE, L., AND PERSICO, P.
M.
Arch.
GOLDMAN,
A.,
R., AND
Biol.
AND
Oral Biol.
D. L. (1977)
GOLDMAN,
Med.
PAZ
DE
A. S.
174,336-342. LA
VEGA,
Y.
27,771-775.
A. S., KATSUMATA,
GASSER,
M., YAFFE,
Nature
S. J., AND
(Londm)
265,643-
645. 19.
KATSUMATA, M., BAKER, M. AND GASSER, D. L. (1981)
K., GOLDMAN,
A. S.,
Immunogenetics
13,
319-325. 20.
GUPTA, C., AND 994-996.
21.
GASSER,
23.
Actions X, pp.
KATSUMATA, GUPTA,
GOLDMAN,
D. L., AND
chemical ed.), Vol. York.
M.,
Arch. Biol.
GOLDMAN,
GUPTA,
C., AND
Acad 25.
LIU,
G., New A.
S.
243,385-395. A. S. (1985) Proc Sot.
178,29-35.
C., KATSUMATA, M., GOLDMAN, OLD, R., AND PIDDINGTON, R. (1984)
Sot.
in Bio-
(Litwack, Press, GOLDMAN,
216,
Biophys.
GOLDMAN,
Med.
Science
A. S, (1983)
of Hormones 357-381, Academic
B&hem
C., AND
A. S. (1982)
24. GUPTA,
35,289-302. 3. TYAN, M. L., AND MILLER, K. K. (1978) Exp. Biol. Med 158.618-621.
TYAN,
Immunogenetics
7. GOLDMAN,
Exp. F. C. (1977)
GOLD-
21,169-183.
6. GOLDMAN, (1983)
(1985) H.
J. J., AND
genetics
22.
nogenetics 2,213-218. 2. BIDDLE, F. G., AND FRASER,
D. D., AND
Acad. Sci. USA 78,
3147-3150. 5. BONNER,
REFERENCES 1. BONNER,
D. L., MELE, L., LEES, A. S. (1981) Proc. Natl.
A. S., HER-
Proc
Natl
Sci. USA 81,1140-1143.
S. L.,
GRIPPO,
J. F.,
ERICKSON,
R.
P., AND
MURINE
26. 27. 28. 29. 30. 31.
32.
33.
34.
35.
36.
GLUCOCORTICOID
RECEPTOR
PRATT, W. B. (1984) J. Steroid Biochem. 21,633637. LITWACK, G., AND ROSENFIELD, S. A. (1975) J. Biol. Chem. 250,6799-6805. WEBB, M. L., MILLER-DIENER, A. S., AND LITWACK, G. (1985) Biochemistry 24,1946-1952. ERDOS, T., BEST-BELPOMME, M., AND BESSADA, R. (1970) Anal. Biochem. 37,244-252. WRANGE, O., AND GUSTAFSSON, J.-A. (1978) J. Biol Chem. 253,856-865. SALOMON, D. S., AND PRATT, R. M. (1976) Nature (London) 264,174-177. MAYER, M., SCHMIDT, T. J., BARNEY, C. A., MILLER, A., AND LITWACK, G. (1983) J. Steroid Biochem. 18,111-120. LITWACK, G. (1976) in Glutathion: Metabolism and Function (Arias, I. A., and Jakoby, W. B., eds.), pp. 285-299, Raven Press, New York. MARKOVIC, R. D., EISEN, H. J., PARCHMAN, L. G., BARNETT, C. A., AND LITWACK, G. (1980) Biochemistry 19,4556-4564. OHL, V. S., MAYER, M., SEKULA, B. C., AND LITWACK, G. (1982) Arch. Biochem. Biophys. 217, 162-178. BASTL, C. P., BARNETT, C. A., SCHMIDT, T. J., AND LITWACK, G. (1984) J. Biol. Chem. 259, 11861195. CARBONE, J. P., MAGER, A. B., BALDRIDGE, R. C., KOSZALKA, T. R., AND BRENT, R. L. (1985) Teratology 31,36A.
INFLUENCED
BY
325
H-2
37. BUTLEY, M. S., ERICKSON, R. P., AND PRATT, W. B. (1978) Nature (London) 275,136-138. 38. HACKNEY, J. F. (1980) Teratology 21,31-51. 39. TSAWDAROGLOU, N. G., GOVINDAN, M. V., SCHMID, W., AND SEKERIS, C. E. (1981) Eur. J. B&hem. 114,305-313. 40. VAN REES, H., WOODBURY, D. M., AND NOACH, E. L. (1960) Arch. Int. Pharmuccdyn 182,437. 41. SHERMAN, M. R., STEVENS, Y.-W., AND TUAZON, F. B. (1984) Cancer Res. 44,3783-3796. 42. HIROTA, T., HIROTA, K., SANNO, Y., AND TANAKA, T. (1985) Endocrinology 117, 1788-1795. 43. KNOX, W. E., AND AUERBACH, V. H. (1955) J. Biol. Chem. 214.307-313. 44. HOSHINO, J., STUDINGER, G., AND KROGER, H. (1981) J. Steroid B&hem. 14,149-154. 45. MEYER, R. D., AND MCMORRIS, F. A. (1984) Somatic Cell Mol. Genet. 10, 153-159. 46. ROUSSEAU, 89.
G. G. (1975)
J. Steroid
Biochem.
6,75-
47. YAMAMOTO, K., GEHRING, U., STAMPFER, M. R.,AND SIBLEY, C. H. (1976) Recent Prog. Harm. Res. 32, 3-32. 48. FRANKE, 664.
U., AND GEHRING,
U. (1980)
Cell 22,657-
49. CHEPENIK, K. P., GEORGE, M., AND GREENE, R. M. (1985) Teratology 32,119-123. 50. ANDREWS, G. K. (1985) J. Stewid B&hem. 23,437443.
OF BIOCHEMISTRY
Vol. 249, No. 2, September,
AND
BIOPHYSICS
pp. 316-325,1986
Murine Glucocorticoid Receptors: New Evidence for a Discrete Receptor Influenced by H-2 ALLEN Center
for
Craniofacial
S. GOLDMAN’
AND
MASUYUKI
KATSUMATA
Anmnalies and Department of Pediatrics, University of Illinois, Chicago, 808 South Wood St., Box 6998, Chicago, Illinois 60680 Received
March
l&1986,
and in revised
form
May
College
of Medicine
at
9, 1986
The glucocorticoid receptor contents in the lungs of females of two congenic strains of mice, BIO.A (H-2a) and BlO (H-Zb), differing only in the H-2 histocompatibility region of chromosome 17, have been measured by the dextran-charcoal method and by our previously described methods of molecular sieving and ion exchange chromatography [M. Katsumata, C. Gupta, and A. S. Goldman (1985) Arch. Biochem. Biophys. 243,3853951. As reported, two receptors, II and IB, are demonstrable by each column chromatographic method, and 5,5-diphenylhydantoin binds to receptor IB but not to receptor II. Receptor IB cannot be detected unless molybdate is added in cytosols prepared with hypotonic buffer [lo mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid and 10 mM dithiothreitol, pH 7.35) according to S. L. Liu, J. F. Grippo, R. P. Erickson, and W. B. Pratt (1984) J. Steroid Biochem. 21,633-6371, a method which has been reported to give maximal receptor levels. Using hypotonic buffer containing 10 mM molybdate we observed a small but significantly higher content of receptor IB in BIO.A mice than that in BlO mice, but no significant difference in receptor II or total receptor content. On the other hand, cytosols prepared with isotonic buffer (50 mM Tris-HCl, 120 mM NaCl, 1 mM EDTA, 10 mM dithiothreitol, and 10 mM molybdate, a modification of the buffer used in our previous report) contained significantly higher levels of receptor IB and of total binding in pulmonary cytosols of BIO.A as compared to those of BlO. There was no difference in receptor II content. Molybdate stabilizes receptor IB in both buffers. These results explain the apparent contradiction between our results and those of Liu et al. by showing that the hypotonic buffer used by them allows for determination of maximal levels of receptor II, but permits selective destruction of receptor IB. However, the use of isotonic buffer gives maximal values of both receptors II and IB. With isotonic buffer, it is demonstrated that only the level of receptor IB is influenced by H-Z-linked genes. 0 1986 Academic
Press, Inc.
lhydantoin (phenytoin, DPH)’ in rodents may utilize the same biochemical mechanism as that by which glucocorticoids pro-
Susceptibility of mice to cortisone(l-5) and phenytoin- (6, 7) induced cleft palate is influenced by genes linked to the H-2 histocompatibility locus on chromosome 17 of the mouse. Evidence from our laboratory has suggested that the clefting action of glucocorticoids and 5,5-dipheny’ To whom
0003-9861/86 Copyright All rights
correspondence
should
$3.00
0 1986 by Academic Press, Inc. of reproduction in any form reserved.
a Abbreviations used: DPH, 5,5-diphenylhydantoin; PLIP, phospholipase-inhibitory protein; DEX, dexamethasone, 9-cY-fluoro-16-a-methyl-ll~,l7a,21-trihydroxypregna-1,4-diene-3,2-dione; DTT, dithiothreitol; Hepes, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid.
be addressed. 316
MURINE
GLUCOCORTICOID
RECEPTOR
duce their anti-inflammatory effects (6-11). It is generally believed that the anti-inflammatory pathway of glucocorticoids involves a receptor-mediated production of phospholipase-inhibitory proteins (PLIPs) that inhibit the release of arachidonic acid from membrane phospholipids, with a subsequent diminished production of prostaglandins and thromboxanes by cyclooxygenase and of leukotrienes by lipoxygenase (12-14). Similar mechanisms appear to be involved in the induction of cleft palate, as exogenous arachidonic acid injected at the same time as glucocorticoids markedly reduces the teratogenic potency of the steroids, and indomethacin, an inhibitor of cyclooxygenase, blocks the corrective action of arachidonic acid (15,16). Indomethacin also blocks the corrective action of arachidonic acid on the inhibition of programmed cell death of embryonic palates produced by cortisol in culture, and inhibits this programmed cell death by itself (16). It has been also reported that phenylbutazone, another inhibitor of cyclooxygenase, blocks the fusion of embryonic palatal shelves in a culture model (1’7). We have suggested earlier that the way by which H-Z-linked genes affect susceptibility to glucocorticoid-induced cleft palate may be explained by their influence on glucocorticoid receptor levels in the embryonic palate, since the same H-Z-linked genes influence both susceptibility to glucocorticoid-induced cleft palate and glucocorticoid receptor levels (18-21). In this connection we have explained the H-2 influence on DPH-induced cleft palate by this same effect on susceptibility to DPH and glucocorticoid receptor levels (6, ‘7, 9, 22), since DPH competes with dexamethasone (DEX) for the glucocorticoid receptor affecting PLIP production and inhibition of prostaglandin production (10, 11, 22-24). If receptor level is a primary site of HZ influence on susceptibility and if this receptor level does affect the arachidonic acid cascade, one would expect to be able to show an H-2 influence at each step in the biochemical pathway. To date we have shown an H-2 influence directly in the embryonic palate on (i) specific [3H]DEX
INFLUENCED
BY
H-2
317
binding (19), (ii) induction of PLIPs by DEX (23), and (iii) inhibition of arachidonic acid release and prostaglandin synthesis (10). We have also shown that H-&linked genes influence the level of total specific binding of [3H]DPH in the same adult tissues, thymuses, and lungs in the same way as they influence glucocorticoid receptors (9,20) and that the DPH-induced inhibition of arachidonic acid release and prostaglandin synthesis in the embryonic palate is also influenced by H-2 (10, 23). Thus, these observations taken as a whole provide supportive evidence from a variety of different experimental approaches that DPH and glucocorticoids bind a common receptor affecting the teratogenic and antiinflammatory functions of glucocorticoids and that the level of this receptor is influenced by H-blinked genes. Recently, an effect of H-2 on binding capacity of glucocorticoids by pulmonary cytosols was reported by Liu et al. (25) to have been undetectable when glucocorticoid receptors were assayed in the presence of a sulfhydryl reagent for stabilizing the receptor which also maximizes binding capacity. This methodology included the use of an overnight incubation in the presence of 10 mM dithiothreitol (DTT) with or without 10 mM molybdate in 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (Hepes) buffer (hypotonic). This method gave values of receptor levels in the lung in the range of 250-300 fmol/mg cytosol protein for both A/J (H-2 a) and C5’7BL/6 (H-2 b) strains, and the levels were higher than we have previously reported (20). These levels required the presence of 10 mM DTT, but incubation with 10 mM molybdate did not affect the levels. We have just reported a study of pulmonary cytosols using isotonic buffer, 10 mM P-mercaptoethanol as sulfhydryl reagent, and [3H]DEX and [3H]DPH as ligands (22). In this study, receptors were measured in the same range by molecular sieving and ion exchange chromatography as those in the study by Liu et al. (25). This study has also provided evidence for the existence of the classical receptor (receptor II) which binds only DEX and a receptor
318
GOLDMAN
AND
which binds both DEX and DPH [similar to receptor IB of Litwack and Rosenfield (26)]. Receptor IB appears to mediate the anti-inflammatory and teratogenic effects of glucocorticoids and DPH, and may be controlled by gene(s) different from that of the classical receptor (22). Because of the apparent discrepancies between the conclusions of Liu et al. (25) and of us (20, 22), we have compared the use of an overnight incubation with 10 mM DTT and 10 mM molybdate either in hypotonic medium (method of Liu et al) or in isotonic medium (our modified method). We have also studied the use of both the dextran-charcoal and molecular sieving methods for determining the DEX-binding capacities of pulmonary cytosols of strain BIO.A and of its congenic partner strain, BlO. We report that determination of receptor level in cytosols prepared in hypotonic buffer preserves maximal levels of receptor II only, but preparation of cytosols in isotonic buffer allows the preservation of maximal levels of both receptors II and IB. Furthermore, the level of receptor IB is influenced by H-2, but that of receptor II is not. MATERIALS
AND
METHODS
Animals. BlO.A/SgSn (BIO.A) and C57BL/lOSn (BlO) mice were obtained from the Jackson Laboratory, Bar Harbor, Maine. Radioisotwpes. [aH]DEX, [1,2,4(N)-3H]DEX (41 Ci/ mmol), was purchased from Amersham Corporation. The purity of the isotope was confirmed to be more than 97% based on the radioactivity by TLC. Cytosol preparation Cytosols of adult female mouse lungs were prepared by homogenizing the tissue in 2 vol (w/v) of one of the following four kinds of buffer: isotonic buffer, consisting of 50 mM Tris-HCl, I20 mM NaCl, 1 mM EDTA, and 10 mM DTT with or without 10 mM sodium molybdate, pH 7.7; and hypotonic buffer consisting of 10 mM Hepes and 10 mM DTT with or without 10 mM sodium molybdate, pH 7.35. The homogenate was centrifuged at 100,OOOg for 60 min, and the supernatant was used as cytosol. [‘HJDEX-receptor complezea An aliquot of cytosol preparation was mixed with [aH]DEX (50 nM, final concentration) with or without a cold competitor (DEX or DPH, 5 PM). Cold DPH was dissolved in l/10 vol of 50 mM sodium carbonate, and added in the incubation mixture. The same volume of sodium carbonate solution was also added in the incubation mixtures of [aH]DEX with and without cold DEX as a
KATSUMATA control adjustment. The incubation was carried out at 0°C for 18 h. In some experiments, we incubated the mixtures for 60 min at 0°C followed by 25°C for 20 min to activate the receptors (27). Molecular sting chromatography. Sephadex G-200 and Sephadex G-25 medium gels were mixed to provide better separation and faster elution, equilibrated with 20 mM K phosphate buffer containing 50 mM KCl, pH 7.0, and packed in a column, 2 X 30 cm, as described previously (22). The sample volume was 1.0 ml, and the receptor complexes were eluted with the same buffer with a flow rate of 0.5 ml/min. Ferritin and cytochrome c were the inner markers. Ion exchange column chromatography with hydroxyapatite adsorption A DEAE-Sephadex A-50 column made from a 25-ml pipet (bed volume, 10 ml) was equilibrated with 20 mM K phosphate buffer containing 50 mM KCl, pH 7.0 (22). The salt concentration of the sample solution was adjusted by dilution with plain water. After the sample solution (2-5 ml) was loaded, the column was washed with 10 ml of the phosphate buffer mentioned above. The elution was carried out with a linear gradient system from the same buffer to 20 mM K phosphate buffer containing 1.0 M KCI, pH 7.0. The salt concentration in the eluate was determined by electrical conductivity. Each l.Oml fraction was mixed with 1.0 ml of a 25% suspension of hydroxyapatite, pH 6.9, and was incubated at 5°C for 10 min to adsorb the rH]DEX-carrying receptor (27,28). The receptor adsorbed by hydroxyapatite was pelleted by centrifugation at SOOg for 5 min, followed by two washings with 20 mM K phosphate buffer containing 50 mM KCl, pH 7.0. The hydroxyapatite pellet was counted directly by suspending it in Aquasol (New England Nuclear Corp.). Calculdim of the level of receptor II and IB. On the basis of our previous report (22), the contents of receptors IB and II were calculated from the results of molecular sieving chromatographic separation as follows. The total receptor content was determined by subtracting the rH]DEX bound to protein in the presence of cold DEX from the bound [aH]DEX in the absence of cold competitor. Receptor IB content was determined by subtracting the bound rH]DEX in the presence of cold DPH from the bound [3HpEX in the absence of cold competitor. Receptor II content was determined by subtracting the bound rH]DEX in the presence of cold DEX from the bound [aH]DEX in the presence of cold DPH. Determination of total receptor content by dextrancharcoal method. The total [aH]DEX binding capacity of pulmonary cytosol was determined by the dextrancharcoal method as previously described (19). Two kinds of cytosolic preparations of both BIO.A and BlO strains, hypotonic with molybdate and isotonic with molybdate, were measured and compared. Statistics. All results are expressed as means -C SD
MURINE
GLUCOCORTICOID
(number of observations). The for significance by the two-tailed
RECEPTOR
INFLUENCED
BY
H-2
319
results are analyzed Student’s t test.
RESULTS
Total Receptor Content Determined Dextran-Charcoal Method
by the
Table I shows that there is no significant difference between the total receptor content of BIO.A and BlO mice when hypotonic buffer is used for the preparation of cytosol, but that a highly significant large difference is observed when the isotonic buffer is used. The values we have shown in this table are quite close to the values of Liu et
al. (25). Comparison of the Chromatographic Patterns of Pulmonary Cytosolic Receptors of Bl0.A and BlO Female Mice Figure 1 shows in molecular sieving chromatography three DEX-labeled peaks: 7, 5, and 3.5 nm Stokes’ radii. The 7-nm peak contains DPH-blockable and DPHnonblockable complexes. The 5-nm peak is not affected by DPH, whereas the 3.5-nm peak is blocked by DPH. The ‘7-nm complex TABLE
I
COMPARISONOFDETERMINATIONOFTOTAL GLUCOCORTICOIDRECEPTORSINPLJLMONARY CYTOSOLS BYMETHODUSINGHYPOTONICBUFFERAND BYTHATUSING ISOTONICBUFFER Binding
Strain BIO.A B10 P value (t statistic)
Hypotonic (fmol/mg protein)
capacity Isotonic (fmol/mg protein)
243.5 + 25.5 (6) 236.5 f 6.2 (6)
332.0 + 25.5 (6) 261.2 -c 17.0 (6)
NS
Note. Numbers are Means + SD (number of determinations). The cytosol preparations, both hypotonic and isotonic, contained 10 mM sodium molybdate. The cytosolic preparations were incubated with 50 nM [3H]DEX in the presence or absence of 5 pM cold DEX at 0°C for 18 h; then each incubation mixture was treated with dextran-charcoal.
FRACTION
NUMBER
FIG. 1. Comparison of DEX-receptor complexes of pulmonary cytosols of BIO.A and BlO prepared with isotonic buffer and molybdate by molecular sieving. The pulmonary cytosol as prepared with isotonic buffer containing 10 mM sodium molybdate. The cytosol was incubated with 50 nM [3H]DEX in the presence and absence of cold competitors, DEX and DPH (5 PM), at 0°C for 18 h. Then a l-ml aliquot of each incubation mixture was gel-filtered through a Sephadex G-200-G-25 mixed gel column (2 X 30 cm) using 20 mM K phosphate buffer containing 50 mM KCl, pH 7.0. Left panel shows the molecular sieving pattern of the cytosol of BlO.A, right panel, that of BlO. 0, [3H]DEX in the absence of cold competitors; 0, [3H]DEX in the presence of 5 pM DPH, n , [3H]DEX in the presence of 5 pM DEX.
is a mixture of aggregates of activated receptor II and receptor IB, the 5-nm complex is activated receptor II, and the 3.5-nm complex is activated receptor IB, since each of these complexes is adsorbed by DNAcellulose (22). The content of receptor IB in pulmonary cytosol of the BIO.A strain is obviously higher than that of BlO, whereas the content of receptor II is not different (Fig. 1). A pronounced influence of H-2 can be seen. Figure 2 indicates that activated receptor IB elutes as both a wash and a 0.14 M KC1 peak in the DEAE-Sephadex A-50 column-hydroxyapatite method, whereas activated receptor II elutes as a 0.2 M KC1 peak, since each of these peaks is adsorbed by DNA-cellulose again as previously described (22). The DPH-blockable receptors, the wash, and 0.14 M complexes (receptor IB), are adsorbed by hydroxyapatite. Thus, the free ligand which also appears in the wash has been eliminated. This result with hydroxyapatite confirms our previous report determined by the dextran-charcoal.
320
GOLDMAN
AND
0.4 E 0.2 B
10
20
so FRACTION
40
10 20 NUtlEER
30
40
FIG. 2. Comparison of DEX-receptor complexes of pulmonary cytosols of BIO.A and BlO prepared with isotonic buffer and molybdate by DEAE-Sephadex A50 column chromatography. The preparation of cytosol and incubation with rH]DEX were carried out as described in Fig. 1. A l-ml aliquot of each incubation mixture was diluted with 1 ml of ice-cold water, and was loaded on a DEAE-Sephadex A-50 column (bed volume, 10 ml). rH]DEX-carrying proteins were eluted with 10 ml wash of 20 mM K phosphate buffer containing 50 mM KCl, pH 7.0, followed by KC1 linear gradient system from 50 mM to 1 M KCI. Each l-ml fraction was treated with hydroxyapatite. The [aH]DEX-carrying protein was adsorbed by hydroxyapatite, and directly determined by suspending the hydroxyapatite pellet in scintillation fluid after washing. Panels and symbols are the same as in Fig. 1.
The use of hydroxyapatite gives a better resolution between receptor IB and II than the previous method because there is more dissociation of ligand from the receptor IB complex with dextran-charcoal(22). Again BIO.A has obviously a higher content of receptor IB than BlO, whereas practically no strain difference is demonstrable in the content of receptor II (Fig. 2). The content determined by molecular sieving gives a value higher than that determined by the ion-exchange-hydroxyapatite method. From determinations using the same cytosolic preparation of BlO.A, molecular sieving gives 133 fmol/ mg protein of receptor IB and 165 fmol/ mg protein of II, whereas the ion-exchange-hydroxyapatite method gives 60 fmol/mg protein of receptor IB and 1’78 fmol/mg protein of receptor II, respectively. The total recovery of receptor proteins by molecular sieving chromatography is between 85 and 92% (assuming the dextran-charcoal method has 100% recovery).
KATSUMATA
Therefore, the content of receptor IB determined by the ion-exchange-hydroxyapatite method is approximately 40% of the total, and the lower value of recovery implies a dissociation of the bound ligand from the receptor IB in the process of separation. Figure 3 shows the molecular sieving pattern of the cytosols prepared with hypotonic buffer. A higher content of receptor IB is observed in BIO.A than in BlO, although much less than that in the cytosol prepared with isotonic buffer. Moreover, the amount of receptor IB in the ‘7- and 3.5nm peaks is much higher in the cytosol prepared with isotonic buffer (Fig. 1, left) than in the cytosol prepared with hypotonic buffer (Fig. 3, left), but the amount of receptor II is not affected by the osmolality of the buffer. The total receptor content and the contents of these two receptors in each buffer have been determined by molecular sieving chromatography and are shown in Table II. The total content is significantly higher in BIO.A than in BlO only when prepared with isotonic buffer. The content of receptor IB is significantly higher in BIO.A than in BlO by the use of either isotonic or hypotonic buffer. The content of receptor II is not significantly different in the two strains. The cytosol prepared with isotonic
-I
F
10
20
1
I
30 40 FRACTION
10
20 NUPlBER
30
40
1
FIG. 3. Comparison of DEX-receptor complexes oi pulmonary cytosols of BIO.A and BlO prepared with hypotonic buffer and molybdate by molecular sieving. The pulmonary cytosol was prepared using hypotonie buffer containing 10 mM sodium molybdate. Other procedures were the same as in Fig. 1. Panels and symbols are the same as in Fig. 1.
MURINE
GLUCOCORTICOID
RECEPTOR TABLE
INFLUENCED
BY
321
H-2
II
DEXAMETHASONE RECEPTORSIB AND II IN PULMONARY CYTOSOLICPREPARATIONS OF BIO.A AND BlO MICE AFTER MOLECULAR SIEVING Binding
capacity
(fmol/mg
Strain Hypotonic Receptor Receptor Total
protein) BIO.A
B10 buffer IB II
with
P value (t statistic) BlO vs BIO.A
molybdate. Incubation: 11 f 4.7 (7) 201 f 37.4 (7) 211 7t 34.8 (7)
overnight
at 24 f 203 f 227 f
0°C. 14.4 (8) 40.7 (8) 39.3 (8)
<0.05 NS NS
Isotonic buffer with molybdate. Incubation: overnight at 0°C. Receptor Receptor Total
15 f 5.1 (5) 207 + 27.8 (5) 222 f 32.6 (5)
IB II
P value (t statistic) Receptor IB Receptor II Total Note. The contents the following formula:
hypotonic
vs isotonic. NS NS NS
of the two receptors
Receptor
were
IB = ([3H]DEX
= [3H]DEX Receptor
II = [3H]DEX
without with without with
- ([3H]DEX
The cytosols were incubation mixture
111 f 46.5 (6) 199 k 40.3 (6) 309 + 15.8 (6)
from
competitor DPH
- [3H]DEX
of molecular
- [3H]DEX
- [3H]DEX
competitor DPH
the results
with
with
DEX)
- r3H]DEX
with
with
prepared as described in Table I. After the incubation was directly gel-filtered as described in Fig. 1.
sieving,
according
to
DEX)
DPH,
DEX. with
r3H]DEX
at 0°C for
18 h, the
buffer contains a significantly higher content of receptor IB than that prepared with hypotonic buffer, but no difference in the contents of receptor II is also observed. Molybdate Efect Figure 4 shows the effect of molybdate on the receptor content in the cytosol prepared with hypotonic buffer according to Liu et al. (25). With molybdate, receptor IB is detectable mainly as the 7-nm complex (Fig. 4, left), but without molybdate there is practically no receptor IB (Fig. 4, right), while the level of receptor II is not affected by the presence of molybdate. Eflect of Incubation Time Other than the osomolality of the buffer, the method of Liu et al. (25) differs from
FRACTION
NUllBER
FIG. 4. Effect of molybdate on the DEX receptors in the cytosol prepared with hypotonic buffer. Cytosol was prepared from lungs of Bl0.A mice using hypo-
tonic buffer with or without 10 mM sodium molybdate. The cytosol was incubated at 0°C for 18 h. Left shows the molecular sieving patterns of DEX tors in the cytosol containing molybdate; right that containing no molybdate. Symbols are the as in Fig. 1.
panel receppanel, same
322
GOLDMAN
AND
ours (22) in incubation time, overnight incubation at 0°C (25) and 60 min at 5°C followed by 20 min at 25°C incubation (22). When molybdate is absent, receptor IB disappears in the long-term incubation in the cytosol of isotonic buffer, but if the incubation time is short, we detect a high content of receptor IB (Fig. 6). Long-term incubation results in little change of the content of receptor II, but the size has changed to a smaller complex, from 5 to 3 nm, which is similar to the mero-receptor (29) (Fig. 6). In the presence of molybdate, the length of incubation causes little change (Fig. 5). In short-term incubation, we can detect a high con tent of receptor IB without molybdate, but still a little less than we obtained in the presence of molybdate (1’73 vs 123 fmol/mg protein) (Fig. 5, left and Fig. 6, left).
KATSUMATA
FRACTION
FIG. 6. Effect of incubation tors in the cytosol prepared cytosol was prepared using molybdate. Other conditions Fig. 5.
NUMBER
time on the DEX recepwithout molybdate. The isotonic buffer without are identical to those in
Efect of EDTA The presence or absence of 1 mM EDTA makes no appreciable difference in the receptor assay (data not shown).
FRACTION
NUMBER
FIG. 5. Effect of incubation time on the Dex receptors in the cytosol prepared with isotonic buffer with 10 mM molybdate. The cytosol was prepared from lungs of female BIO.A mice using isotonic buffer with 10 mM molybdate. The cytosol was incubated either at 0°C for 60 min followed by 20 min at 25’C (left) or 0°C for 18 h (right). Symbols are the same as in Fig. 1.
DISCUSSION
The present report has confirmed the report of Liu et al. (25) that the use of an overnight incubation with or without 10 mM molybdate in hypotonic buffer with 10 mM sulfhydryl reagent gives higher values of total glucocorticoid receptors in mouse lungs without an H-2 influence than those measured earlier by us with an H-2 influence by a method using a short incubation time of cytosols prepared without a sulfhydryl reagent (20). However, we recently have provided evidence for the existence of a discrete glucocorticoid receptor, receptor IB, which binds both DEX and DPH, which is adsorbed by DNA-cellulose, which mediates the anti-inflammatory and teratogenic actions of glucocorticoids, and which is different from the classical receptor II (22). The present results also indicate that this discrete receptor in the lungs is stabilized by molybdate. The main difference between the two methodologies rests on the fact that the method of Liu et al. (25) measures maximal levels of receptor II, while our present method measures maximal levels of both receptors II and IB. Thus, both methods show with column chromatography that the levels of receptor
MURINE
GLUCOCORTICOID
RECEPTOR
II are not affected by H-2. However, the measurement of the levels of receptor IB and its dependence on H-2-linked genes can be demonstrated only with isotonic buffer. The key factor in these differences is the use of hypotonic or isotonic buffer for the preparation of the cytosols. In general, those studies which have used isotonic (isoosmotic) buffers for the preparation of cytosols or in binding studies of cells in isotonic media show an influence of H-2 (9, 18-20, 30) or the existence of receptor IB (22, 26, 27, 31-36), while those which use hypotonic buffers fail to demonstrate an influence of H-2 (25,37,38). Because of the similarity of the properties of the receptor which binds both DEX and DPH to that described by Litwack and colleagues (26, 27, 31-34), we have considered it to be receptor IB as first designated by them (22). The activated receptor has a Stokes’ radius of 3.0-3.5 nm, and can aggregate to a much larger form. Unlike receptor II, IB has a wide specificity of steroid binding (22, 26), can bind DPH (22), and has the capacity to rebind ligand lost in the process of purification (22,32,39). The inhibition of the release of prelabeled [3H]arachidonic acid by DEX and DPH occurs only in the cells which have receptor IB, e.g., thymocytes (22). DPH binds only to receptor IB, i.e., 7- and 3.5-nm forms, and both forms of the DPH-receptor complex are adsorbed by DNA-cellulose (22). The DPH-receptor complex is incorporated into nuclei of thymocytes (11). DPH depresses the release of arachidonic acid from the cells such as thymocytes and lung cells, which have receptor IB (with II) (22, 23). Hepatoma G2 cells contain only receptor II, which does not bind DPH, and neither DPH nor DEX affects the release of arachidonic acid from these cells (22). Receptor IB is most likely, therefore, responsible for mediation of the production of phospholipase-inhibitory proteins induced by DEX (9,12-14,20) and DPH (11,22-24). It is thought that these proteins mediate the anti-inflammatory and teratogenic action of glucocorticoids (9, 12-16, 20) and DPH (11,22-24) by inhibiting the release of arachidonic acid from membrane phospho-
INFLUENCED
BY
H-2
323
lipids with a subsequent reduction in the synthesis of prostaglandins and leukotrienes. Receptor IB also appears to mediate the stimulation by glucocorticoids of cation transport in rat colonic epithelia, which contain only receptor IB (35). Recently, it has also been shown that rat embryonic yolk sac, the anlagen of the colonic epithelia, also contains only receptor IB (36, 50). In support of these observations is the finding that DPH also stimulates cation transport in rat jejunum (40), since we have shown that DPH binds to receptor IB as a glucocorticoid agonist (11, 22, 23). Receptor IB is not likely to be the “nicked” glucocorticoid receptor of Wrange and Gustafsson (29) or the proteolytic fragment of Sherman et al. (41) because of several observations. The present finding that the level of receptor II remains constant when the level of IB is markedly elevated by a change from hypotonic to isotonic buffer argues against the origin of receptor IB from II by proteolysis. This conclusion is supported by the facts that, unlike receptor II, the steroid-binding site of IB binds DPH and other steroids having a different molecular structure and, also unlike receptor II, activated EB can rebind ligand (22,32,39). The existence of receptor IB in rat liver has been confirmed by Hirota et al. (42), who have also provided evidence for the existence of a discrete third receptor which appears only under certain conditions and which mediates the induction of tryptophan dioxygenase, but not that of tyrosine aminotransferase by glucocorticoids. The induction of the activity of tyrosine aminotransferase by glucocorticoids (43-46) is mediated by receptor II in hepatoma G2 cells, which have receptor II exclusively (22). It has been shown that the major receptor involved in the killing or growth-inhibitory effect of glucocorticoids in glucocorticoid-sensitive cells, S49, is most probably receptor II, since these cells contain predominantly the receptor whose size is identical to that of the 5-nm receptor (47, 22). The killing or growth-inhibitory effect of glucocorticoids in S49 cells is coded by genes on mouse chromosome 18 (48). On
324
GOLDMAN
AND
the other hand, H-blinked genes on mouse chromosome 17 influence suceptibility to cortisone- (l-5) and DPH-induced (6, 7) cleft palate, DEX- and DPH-receptor levels (9, l&20), and the degree of production of PLIPs by DEX and DPH (23) and of inhibition of the arachidonic acid cascade (10). A separation of the growth-inhibitory function of glucocorticoids from that of inhibition of phospholipase by glucocorticoids has been demonstrated in embryonic palatal mesenchyme cells (49). Moreover, the phospholipase-inhibitory effect is evidently influenced by H-2, but the growthinhibitory effect is not (49). The growthinhibitory function of glucocorticoids is not different in degree in palate mesenchyme cells from strains different in H-2, but the degree of phospholipase-inhibitory action is (49). The separation of the function of the third receptor from that of receptor II suggests a third discrete receptor also under separate genetic control (42). Thus, this report taken with those studying genetic control, receptor location, receptor properties, and receptor function suggests the working hypothesis of receptor polymorphism with each receptor having a different function(s) under separate genetic control. This hypothesis suggests that receptor II mediates induction of tyrosine aminotransferase and growth-inhibitory functions of glucocorticoids and is coded by genes on chromosome 18, while genes in chromosome 17 influence the level of receptor IB and its effects on the arachidonic acid cascade and intestinal cation transport. The hypothesis favors the concept of receptor polymorphism in the controversy of whether receptor IB is an isomorphic form of receptor II found predominantly in kidney and colon (35) or yolk sac (36,50) or whether it is a proteolytic product of receptor II (29,41).
KATSUMATA 4. GASSER, MAN,
J. J., AND
SLAVKIN,
C. (1975)
ImmuGenetics
Proc.
A. S., BAKER, A.
S.,
M.
Immunc-
L. (1984)
M. K., AND
86. 8. GOLDMAN,
GASSER,
D. L.
18,17-22.
FISHMAN,
C.
Proc. Sot. Exp.
M. K. (1983) A.
S., BAKER,
AND HEROLD,
R. (1983)
L.,
AND
BAKER,
Biol.
Med.
173,82-
M. K.,
Proc
PIDDINGTON,
R.,
Sot. Exp. Bid
Med.
174,239-243. 9. GUPTA,
C., KATSUMATA,
(1984) 10.
GUPTA,
M., AND
Immunogenetics
GOLDMAN,
A. S.
20,667-676.
C., KATSUMATA,
(1985)
M., AND
Genet.
J. Craniofacial
GOLDMAN,
A. S.
Dew. Biol
5,277-
285. 11.
KATSUMATA, DORF,
M.,
GUPTA,
C. E., AND
C., BAKER,
GOLDMAN,
M. K., Suss-
A. S. (1982)
Science
218,1313-1315. 12. HIRATA,
F.,
MANIAN, 13.
14. 15.
SCHIFFMANN, K., SALOMON,
E., VENKATASUBRAD., AND AXELROD,
Nature
(1980) GHIARA,
(London)
P., MELI,
L., AND
PERSICO,
P. (1984) Biochem. Pharmacol. 33,1445-1450. TZORTZATOU, G. G., GOLDMAN, A. S., AND BouTWELL, W. C. (1981) Proc Sot. Exp. Biol Med.
MONTENEGRO,
(1982) 18.
287,147-149.
R., PARENTE,
166,321-324. 16. PIDDINGTON, R., HEROLD, (1983) Proc. Sot. Exp. 17.
J.
(1980) Proc. Natl. Acad. Sci. USA 77,2533-2536. BLACKWELL, G. J., CARNUCCIO, R., DI ROSA, M., FLOWER, R. J., PARENTE, L., AND PERSICO, P.
M.
Arch.
GOLDMAN,
A.,
R., AND
Biol.
AND
Oral Biol.
D. L. (1977)
GOLDMAN,
Med.
PAZ
DE
A. S.
174,336-342. LA
VEGA,
Y.
27,771-775.
A. S., KATSUMATA,
GASSER,
M., YAFFE,
Nature
S. J., AND
(Londm)
265,643-
645. 19.
KATSUMATA, M., BAKER, M. AND GASSER, D. L. (1981)
K., GOLDMAN,
A. S.,
Immunogenetics
13,
319-325. 20.
GUPTA, C., AND 994-996.
21.
GASSER,
23.
Actions X, pp.
KATSUMATA, GUPTA,
GOLDMAN,
D. L., AND
chemical ed.), Vol. York.
M.,
Arch. Biol.
GOLDMAN,
GUPTA,
C., AND
Acad 25.
LIU,
G., New A.
S.
243,385-395. A. S. (1985) Proc Sot.
178,29-35.
C., KATSUMATA, M., GOLDMAN, OLD, R., AND PIDDINGTON, R. (1984)
Sot.
in Bio-
(Litwack, Press, GOLDMAN,
216,
Biophys.
GOLDMAN,
Med.
Science
A. S, (1983)
of Hormones 357-381, Academic
B&hem
C., AND
A. S. (1982)
24. GUPTA,
35,289-302. 3. TYAN, M. L., AND MILLER, K. K. (1978) Exp. Biol. Med 158.618-621.
TYAN,
Immunogenetics
7. GOLDMAN,
Exp. F. C. (1977)
GOLD-
21,169-183.
6. GOLDMAN, (1983)
(1985) H.
J. J., AND
genetics
22.
nogenetics 2,213-218. 2. BIDDLE, F. G., AND FRASER,
D. D., AND
Acad. Sci. USA 78,
3147-3150. 5. BONNER,
REFERENCES 1. BONNER,
D. L., MELE, L., LEES, A. S. (1981) Proc. Natl.
A. S., HER-
Proc
Natl
Sci. USA 81,1140-1143.
S. L.,
GRIPPO,
J. F.,
ERICKSON,
R.
P., AND
MURINE
26. 27. 28. 29. 30. 31.
32.
33.
34.
35.
36.
GLUCOCORTICOID
RECEPTOR
PRATT, W. B. (1984) J. Steroid Biochem. 21,633637. LITWACK, G., AND ROSENFIELD, S. A. (1975) J. Biol. Chem. 250,6799-6805. WEBB, M. L., MILLER-DIENER, A. S., AND LITWACK, G. (1985) Biochemistry 24,1946-1952. ERDOS, T., BEST-BELPOMME, M., AND BESSADA, R. (1970) Anal. Biochem. 37,244-252. WRANGE, O., AND GUSTAFSSON, J.-A. (1978) J. Biol Chem. 253,856-865. SALOMON, D. S., AND PRATT, R. M. (1976) Nature (London) 264,174-177. MAYER, M., SCHMIDT, T. J., BARNEY, C. A., MILLER, A., AND LITWACK, G. (1983) J. Steroid Biochem. 18,111-120. LITWACK, G. (1976) in Glutathion: Metabolism and Function (Arias, I. A., and Jakoby, W. B., eds.), pp. 285-299, Raven Press, New York. MARKOVIC, R. D., EISEN, H. J., PARCHMAN, L. G., BARNETT, C. A., AND LITWACK, G. (1980) Biochemistry 19,4556-4564. OHL, V. S., MAYER, M., SEKULA, B. C., AND LITWACK, G. (1982) Arch. Biochem. Biophys. 217, 162-178. BASTL, C. P., BARNETT, C. A., SCHMIDT, T. J., AND LITWACK, G. (1984) J. Biol. Chem. 259, 11861195. CARBONE, J. P., MAGER, A. B., BALDRIDGE, R. C., KOSZALKA, T. R., AND BRENT, R. L. (1985) Teratology 31,36A.
INFLUENCED
BY
325
H-2
37. BUTLEY, M. S., ERICKSON, R. P., AND PRATT, W. B. (1978) Nature (London) 275,136-138. 38. HACKNEY, J. F. (1980) Teratology 21,31-51. 39. TSAWDAROGLOU, N. G., GOVINDAN, M. V., SCHMID, W., AND SEKERIS, C. E. (1981) Eur. J. B&hem. 114,305-313. 40. VAN REES, H., WOODBURY, D. M., AND NOACH, E. L. (1960) Arch. Int. Pharmuccdyn 182,437. 41. SHERMAN, M. R., STEVENS, Y.-W., AND TUAZON, F. B. (1984) Cancer Res. 44,3783-3796. 42. HIROTA, T., HIROTA, K., SANNO, Y., AND TANAKA, T. (1985) Endocrinology 117, 1788-1795. 43. KNOX, W. E., AND AUERBACH, V. H. (1955) J. Biol. Chem. 214.307-313. 44. HOSHINO, J., STUDINGER, G., AND KROGER, H. (1981) J. Steroid B&hem. 14,149-154. 45. MEYER, R. D., AND MCMORRIS, F. A. (1984) Somatic Cell Mol. Genet. 10, 153-159. 46. ROUSSEAU, 89.
G. G. (1975)
J. Steroid
Biochem.
6,75-
47. YAMAMOTO, K., GEHRING, U., STAMPFER, M. R.,AND SIBLEY, C. H. (1976) Recent Prog. Harm. Res. 32, 3-32. 48. FRANKE, 664.
U., AND GEHRING,
U. (1980)
Cell 22,657-
49. CHEPENIK, K. P., GEORGE, M., AND GREENE, R. M. (1985) Teratology 32,119-123. 50. ANDREWS, G. K. (1985) J. Stewid B&hem. 23,437443.