High Level Expression of Wild Type and Variant Mouse Glucocorticoid Receptors in Chinese Hamster Ovary Cells
Margaret A. Hirst, Jeffrey P. Northrop, Mark Danielsen*, and Gordon M. Ringold
Institute of Cancer and Developmental Biology Syntex Research Palo Alto, California 94304 Howard Hughes Medical Institute (J.P.N.) Beckman Center Stanford University School of Medicine Stanford, California 94305
enous activity of the CHO GR. The overexpression of various forms of the GR in CHO cells serves as a paradigm for detailed biochemical analysis of this and other hormone-regulated transcription factors. (Molecular Endocrinology 4: 162-170, 1990)
We have isolated Chinese hamster ovary (CHO) cell lines expressing elevated levels of wild-type (W) and mutant forms of the glucocorticoid receptor (GR) using the technique of coamplification with a selectable dihydrofolate reductase (dhfr) cDNA. A prominent doublet at 90/92 kilodaltons was observed by Western blotting or labeling with [3H]-dexamethasone mesylate in extracts from cells transfected with W, the hormone binding mutant (NA), and the DNA binding mutant (NB). Quantification of receptor number by [3H]dexamethasone binding revealed the presence of approximately 106 receptors per cell in the W and NB-producing lines. This represents a 25to 50-fold increase in receptor density over control CHO cells which were not transfected with GR. Comparative quantitation by Western blotting of extracts from cells expressing GR showed that cells producing NA contain a level approximately 500-fold over control CHO cells. Function of the amptified receptors was examined by transient transfection with the glucocorticoid-responsive reporter plasmid pMMTV-chloramphenicol acetyl transferase (CAT). Our results indicate that inducible CAT activity increases with the abundance of W receptor and no evidence of saturability was observed even at the highest levels of receptor. This supports previous suggestions that the concentration of the hormone-regulated transcription factor is definitely limiting with regard to maximal transcription efficiency. Interestingly, cells expressing even highly amplified levels of NA-GR or NB-GR showed no inducible response above that seen with control CHO cells. Thus these mutations are exceedingly nonleaky and are not dominant over the low endog-
INTRODUCTION
Glucocorticoids, like other steroid hormones, exert their effects through a specific intracellular receptor protein that functions as a ligand-activated transcriptional regulator (1, 2). Upon hormone binding, the receptor apparently dissociates from a multimeric complex (3) containing at least one molecule of the heat-shock protein, hsp90 (4). The hormone-receptor complex subsequently binds tightly to a specific DNA sequence in target genes known as the glucocorticoid response element (GRE) whereupon either activation or repression of transcription may ensue. Biochemical and molecular dissection of wild-type and mutant forms of the glucocorticoid receptor has clearly revealed its modular composition. The approximately 86 kilodalton (kDa) receptor is composed of an approximately 10 kDa DNA binding domain containing two characteristic zinc fingers (5), flanked by a Cterminal hormone binding domain of approximately 25 kDa and an N-terminal transcription-modulating domain of approximately 50 kDa. The hormone binding domain serves to keep the receptor in a non-DNA binding conformation in the absence of ligand and contains one or more regions that interact with hsp90 (6). The Nterminal domain contains a strongly acidic region that functions to reduce nonspecific DNA interactions (7, 8). Regions in both the N-terminal domain (9) and C-terminal domain (10) can activate the transcriptional ca-
0888-8809/90/0162-0170$02.00/0 Molecular Endocrinology Copyright © 1990 by The Endocrine Society
162
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
GRs in CHO Cells
163
pacity of other DNA binding proteins deleted of their own activation functions. Substantial data has accumulated indicating that the GR undergoes covalent modification (11-13) and that its degree of phosphorylation may be altered by interaction with hormone (1215); however, the functional consequences of such modification remain to be explored. Generating a more detailed picture of the dynamics of GR function in intact cells and obtaining large quantities of purified receptor have been hampered by the relatively low abundance of receptor in normal tissues and cell lines. We have attempted to overcome this problem by generating stable Chinese hamster ovary (CHO) cell lines expressing elevated levels of W and mutant forms of the mouse GR. By cotransfection with a selectable dihydrofolate reductase (dhfr) cDNA we have succeeded in obtaining cells that express approximately 25-50 times more W or DNA-binding deficient receptor and as much as 400-500 times as much of a hormone-binding deficient receptor than do control CHO cells. Analysis of these cell lines indicates that the W-GR is fully functional as a transcriptional activator in these cells and that the inducibility of a GRE-containing mouse mammary tumor virus (MMTV) promoter increases in proportion to the level of receptor. Both forms of the mutant receptors even when expressed at high levels fail to activate transcription from the MMTV promoter and furthermore do not interfere with the function of the low levels of endogenous GR resident in the CHO cells. Last, as recently demonstrated, stable lines of CHO cells expressing copious levels of functional GR could potentially serve as a useful system in which to express foreign proteins of interest in a highly inducible fashion (16); this is of particular benefit if the protein to be expressed has deleterious effects on cell proliferation.
(data not shown). Faint bands considerably smaller than 90 kDa are also detected with the GR49 rat receptor monoclonal antibody which detects an epitope in the N-terminal domain of the GR (19). Thus the smaller forms of receptor detected in this analysis must arise by proteolytic degradation of receptor or premature termination of translation. In order to achieve higher levels of W-GR expression, several of these cell lines were exposed to higher concentrations of MTX in stepwise fashion. Pools of resistant cells at each concentration were analyzed by immunoblotting. The results shown in Fig. 1B indicate that, in general, resistance to increased levels of MTX yields cells that express more W-GR, corroborating previous studies demonstrating coamplification of unlinked cotransfected DNAs (20). The highest levels of W-GR expression were obtained at 1-3 x 10'6 M MTX. Surprisingly, selection of cells resistant to 10 4 M MTX did not yield higher levels of W-GR, perhaps due to antiproliferative effects of increased expression of a functional GR. A similar protocol was used to select cells overexpressing mutant forms NA and NB of the GR. As seen in Fig. 2, several clones were isolated expressing elevated levels of the DNA-binding mutant NB and extremely abundant levels of the hormone-binding mutant NA. These are compared to CHO cells which have been transfected with dhfr cDNA but not receptor and exhibit no detectable receptor. Clearly apparent in this immunoblot are the smaller forms of the receptor detected by the GR49 antibody. These are virtually identical to the fragments derived from W-GR (Fig. 1B) and thus are not a consequence of an altered structure due to the mutation at position 546.
RESULTS
To ascertain if the overexpressed 92 kDa receptor protein observed by immunoblot analysis is capable of binding hormone we incubated cell extracts with [3H] dex-mesylate and analyzed covalently labeled receptor by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis. As seen in Fig. 3, cells expressing elevated levels of immunoreactive W and NB-GR exhibit increased hormone binding activity at approximately 92 kDa that is specifically competed by a 200-fold excess of unlabeled dexamethasone. As expected, no labeling of the NA receptor could be detected (not shown). As in the immunoanalyses described above, a doublet of GR labeled with [3H]dex-mesylate is seen at 90/ 92 kDa. In the W cells, the upper band is labeled predominantly whereas the opposite is true in the NB cells. Conceivably the mutation in the DNA binding domain predisposes the receptor to increased proteolysis. However if this is so it must be limited to a minor clip since there is no excessive accumulation of smaller fragments observable in the NB receptor containing extracts labeled with [3H]dex-mesylate. Faint, specifically labeled smaller (-50 kDa) bands are seen in both
Immunoblot Analysis of CHO Cells Expressing W and Mutant GR After cotransfection of CHO d h f r cells (17) with dhfr and mouse GR cDNAs, dhfr+ cells were selected in a stepwise fashion for resistance to increasing concentrations of methotrexate (MTX). Three receptor cDNAs were used; W, a hormone binding mutant in which Glu546 has been changed to Gly (NA), and a DNA-binding mutant in which Arg484 has been changed to His (NB) (18). Immunoblot analysis of extracts from cells transfected with W receptor and grown in 10"7 M MTX is shown in Fig. 1 A. A prominent doublet at 90/92 kDa is observed in some of the transfectants; for comparison, immunoreactive receptor in an extract from HTC, rat hepatoma cells that contain 5-10 x 104-receptor per cell, is detectable but low in this immunoblot. Our evidence indicates that the upper band of this doublet corresponds to the size observed for mouse GR isolated from S49 cells and in vitro translated receptor
[3
H]Dexamethasone Mesylate (DEX-Mesylate) Binding to Overexpressed GR
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
— 200 — 116 — 92.5 — 66
— 45
— 31
— 21.5
W4
W12
W9
W5
B
i r
i r
VO
I
V©
I
— 200 — 116 — 92.5 — 66
— 45
— 31
— 21.5
164 The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
165
X V
• « *
13-4
o
12-5
GRs in CHO Cells
^j
z
z
z
z
Z
•4 Z
CQ
CQ
CQ
CQ
Z
Z
Z
Z
— 200 — 116 — 92.5 — 66
— 45
— 31 *****
— 21.5
Fig. 2. Immunological Analysis of CHO Cells Expressing NA and NB GR Cell extracts were prepared as described in Materials and Methods and analyzed as described in Fig. 1. NA lanes are CHO clones transfected with the GR hormone binding mutant cDNA. NA12-5 was selected in 10 s M MTX. NA1-4, NA2-4, NA5-4, NA13-4, NA4-4 were selected in 10/4 M MTX. NB lanes are CHO clones transfected with the GR DNA binding mutant cDNA. All were selected in 10"4 M MTX. HTC lane is a rat liver cell line. CHO lane is dhfr" CHO cells transfected with dhfr cDNA but not GR cDNA.
W and NB cells (Fig. 3, lanes 7 and 9). These bands become more apparent upon longer exposure and do not correspond to the smaller immunoreactive bands seen in Figs. 1 and 2. This is to be expected since the antibody and hormone affinity-label detect opposite ends of the receptor molecule. Quantification of Receptor Number In order to more carefully estimate receptor number in the overexpressing lines, we analyzed the amount of
[3H]dexamethasone binding in extracts from the various cells (Table 1). Receptors were quantified using a single concentration of 10~7 M [3H]dexamethasone. This is known to be a saturating dose in HTC cells which have a dissociation constant of approximately 10~9 M. It is possible that the overexpressed receptors in CHO cells may have a dissociation constant different than that of HTC cells. Therefore a significant decrease in receptor affinity for ligand may result in an underestimation of receptor levels in CHO cells. On a per cell basis, the highest levels of W and NB receptor obtained corre-
Fig. 1. Immunoanalysis of CHO Cells Expressing W GR Cell extracts were electrophoresed on a 10% SDS-polyacrylamide gel and transferred to nitrocellulose as described in Materials and Methods. A second antibody conjugated to alkaline phosphatase was used to detect the monoclonal antibody GR-49. W indicates CHO clones transfected with W GR. The number after the letter identifies the clone. The dashed number indicates the concentration of MTX in negative logs the clone was growing in at the time of assay. Multiple numbers specify the series of MTX concentrations used in the stepwise selection of cells. All lanes were loaded with 200 nQ protein. Molecular weight markers were visualized with Coomassie staining and are indicated at the right (in kilodaltons). A, Twelve CHO clones initially selected in 10'7 MTX and a rat liver cell line (HTC). B, Four CHO clones that were further selected in higher concentrations (10"7to 3 x I O ^ M ) MTX as indicated and analyzed as pools.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
Vol 4 No. 1
MOL ENDO-1990 166
HTC 1
2
W4-7 3
4
W9-7-6 5
6
W12-7-6 7
8
NB1-5 9
10 — 200 — 116 — 92.5
— 66
— 45
— 31
— 21.5
Fig. 3. Autoradiogram of [3H]Dex-Mesylate Labeled Cell Extracts Electrophoresed on a 10% SDS-Polyacrylamide Gel Cell extracts were prepared as described in Materials and Methods. Samples containing 200 ng protein were incubated with [3H] dex-mesylate at a final concentration of 1 x 10"7 M for 3 h at 4 C. Nonspecific binding was determined by addition of 200-fold molar excess of unlabeled dexamethasone. Cell extracts tested are labeled above the lane numbers. Odd numbered lanes, Extracts incubated with labeled hormone. Even numbered lanes, Extracts incubated with labeled and unlabeled hormone. Molecular weight markers were visualized with Coomasie stain and indicated at the right in kilodaltons.
spond to approximately 5-10 x 105 receptors. This compares with the 25- to 50-fold lower concentration (~2 x 104/cell) of GR in CHO cells and 5- to 10-fold lower level (~1 x 10s/cell) in HTC cells. The low level of hormone binding (equivalent to 2-3 x 104 receptors per cell) detected in NA-containing cells reflects the endogenous GR in CHO dhfr cells (21). Since the NA receptor cannot bind hormone we have estimated its relative abundance in the transfected cells by a semiquantitative immunological analysis. The results shown in Fig. 4 indicate that approximately 10 ^g cell protein from the cell line NA4-4 give rise to a band of equal intensity to that seen with 200 ^g cell protein from the NB2-5 cell line (containing ~5 x 105receptors per cell). Similar studies comparing NA4-4 with W4-76-4 yielded equivalent results (not shown). Based on
these experiments we conclude that NA-transfected cells contain as many as 10 x 106 receptors per cell. This number is only an approximation since the actual number has not been rigorously quantified. The reasons for our ability to select NA cells expressing higher levels of receptor than either NB or W-cells remain obscure. Functional Response to Glucocorticoids in CHO Cells Expressing Various Levels of GR Representative pools of cells containing various levels of W receptor were transiently transfected with the glucocorticoid responsive MMTV promoter-CAT fusion plasmid. The efficiency of transfection of each cell line was evaluated with pRSVCAT and all data have been normalized to the maximal expression seen in each
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
167
GRs in CHO Cells
Table 1. [3H]Dexamethasone [3H]Dex Binding in Cell Extracts Cell Line
W4-7-6-4 W5-7-6-3~6 W9-7-6-3-6 W-12-7-6 NB1-5 NB2-5 NB3-5 NA4-4 CHO HTC
Receptor Number/Cell (Minimum Estimation)
4.8 9.1 5.2 7.4 3.5 4.8 6.5 2.9 1.9 9.8
x x x x x x x x x x
10s 10s 10s 105 105 10s 10s 104 104 104
(Range)
(n)
(4.3-5.3) (8.5-9.7) (4.9-5.7) (6.2-8.5) (2.9-4.1)
(2) (2) (3) (2) (3) (1) (1) (1)
(6.8-11)
(3)
(D
Various CHO cell types (W, NB, NA, CHO) and rat liver cells (HTC) were removed from dishes by trypsin, counted with a hemacytometer, washed, and sonicated in SETBM buffer. Supematants from a 100,000 x g spin were incubated with 10~7 M [3H]dex for 3 h at 4 C. Total ([3H]dex) and nonspecific ([3H]dex plus 250-fold molar excess unlabeled dex) binding were analyzed in duplicate. Specific [3H]bound ligand (total minus nonspecific) was used to estimate minimum receptor number per cell. The range (Range) of values is shown followed by the number (n) of independent determinations. CHO cells were stably transfected with cDNA as follows: dhfr alone, CHO; dhfr and W-GR, W; dhfr and GR hormone binding mutant, NA; dhfr and GR DNA binding mutant, NB. The last number of each cell line designates the molar concentration in negative logs of MTX in the selection media.
experiment with this control vector. The results (Fig. 5
and Table 2) show that increased expression of W receptor gives rise to cells that support greater inducibility of the MMTV-CAT plasmid. Even at the highest levels of receptor expression we see no evidence of saturating the inducible response. Thus, receptor number appears to be a critical limiting factor in achieving increased hormone-dependent transcription from the MMTV promoter. Interestingly, there is a small increase in the basal level of CAT expression in cells expressing the highest concentration of receptor. This may be due to the presence of small amounts of corticoids in the medium (or serum) or alternatively may reflect spontaneous activation of receptor in a hormone-independent fashion. In cells overexpressing the mutant NA or NB receptors, we have been unable to detect any significant increase in glucocorticoid responsiveness using MMTVCAT. This documents that the mutants are indeed very tight and in addition are unable to interfere with the weak induction elicited by the endogenous CHO receptor. Moreover, the hormone binding mutant NA, even when expressed at approximately 10 x 106 receptor per cell, does not increase basal expression from the MMTV promoter. We surmise that it is maintained in a fully inactive form perhaps through its association with other proteins such as hsp 90 (3, 4). DISCUSSION The ability to study many aspects of GR structure, function, and metabolism are compromised by the rel-
atively low levels of GR present in most cells. To overcome this limitation, we have succeeded in overexpressing the mouse GR in CHO cells by the method of coamplification with a dhfr vector. Although we presume this method resulted in gene amplification of receptor, we did not directly examine this. This procedure is relatively easy to perform and can be extended to other receptors of interest. However the absolute level of expression cannot be predicted for any particular protein; we have, for example, obtained 10-fold higher expression of the hormone binding GR mutant NA than of either W or the DNA-binding mutant NB. It will be of interest to see whether this is fortuitously due to the sites of integration of the transfected DNAs or whether there is indeed selection against higher level of W-GR and NB-GR expression. Several revealing facts about transcription activation by the GR have emerged from analysis of the CHO cell transfectants. First, as has also been reported by Miesfeld et at. (22), increased levels of functional GR result in higher transcriptional activity from a glucocorticoidresponsive promoter. Moreover, our results indicate that the inducible expression of the MMTV-CAT plasmid is not saturated even with approximately 106 receptors per cell. Thus, not only is receptor a limiting factor, but the cell's capacity to use high levels of receptor is substantially greater than one might have predicted a priori. Second, the fact that we do not observe highly increased expression of MMTV-CAT in the absence of hormone indicates that there is excess capacity to maintain the receptor in the inactive state when in the unbound state. This presumably reflects the high cellular concentration of hsp 90 (and perhaps other proteins) with which the receptor is presumably associated in a high molecular weight complex (3). Whether or not the GR in the amplified CHO cells resides in such a complex (under hormone-free conditions) remains to be directly tested. If, as one would expect, this is so, then these cells may prove extremely valuable in exploring the structure of the inactive complex in finer detail. Third, it is of interest to note that the point mutations giving rise to the NA and NB receptors result in extremely tight phenotypes in which no inducibility of MMTV-CAT can be detected. It is perhaps also noteworthy that the NA mutant, even when present at approximately 107 molecules per cell remains recessive; that is, it does not interfere with the low level activity of the endogenous (104) receptors present in CHO cells. Since the inducibility of a GRE-containing promoter is markedly elevated in cells containing approximately 106 W receptors such a system would be useful for obtaining conditional high level expression of a protein that might otherwise be toxic to cells. The feasibility of this approach has been independently demonstrated in a recent publication (16). In addition to this practical use, the use of cell lines expressing high levels of GR for detailed study of receptor modification, receptorprotein interactions, and intracellular compartmentalization of W and mutant receptors will undoubtedly
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 168
Vol 4 No. 1
NA4-4 ug protein/lane
200
100
50
NB2-5 20
10
200
100
50
20 — 200
— 31
— 21.5 Fig. 4. Immunoblots of Various Cell Extracts at Increasing Protein Concentrations Electrophoresed on a SDS-Polyacrylamide Gel Cell extracts were prepared as described. Increasing protein was loaded from right to left for CHO clones NA4-4 and NB2-5. The numbers above each lane indicate the quantity of protein loaded.
AcCM
Dex Cell Line
-
HTC
W3-7
W9-7
W5-7
W9-7-6-3
CHO
NA4-4
NB1-5
Fig. 5. CAT activity from various cells transfected with pMMTVCAT As described in Materials and Methods, extracts were prepared from cells transiently transfected with the glucocorticoid responsive pMMTVCAT. After transfection, cells were treated with vehicle (-) or 10.-6 M dexamethasone (+) for 48 h. Equal quantities of cell extracts were assayed for CAT activity. Cells expressing W-GR in low (W3-7), medium (W9-7), high (W5-7), and very high (W9-7-6-3" 6 ) abundance are compared. NA4-4 and NB 1-5 are clones expressing high levels of mutant GR. Cells with normal receptor quantity is shown by HTC. CHO represents response to GR with no amplified receptor protein. CM, Unconverted [14C]chioramphenicol. AcCM, Acetylated [14C] chloramphenicol.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
169
GRs in CHO Cells
supplemented with 39.5 fig/m\ proline and 8% defined/supplemented bovine calf serum).
Table 2. Glucocorticoid-lnduced MMTVCAT Activity % Conversion Cell Line Vehicle ± SD
HTC NA4-4 CHO NB1-5 W3-7
W10-7 W9-7 W2-7 W5-7 W12-7 W5-3"6 W9-3"6
Dex ± SD
2.2 0.041 0.027 ± 0.004 0.072 ± 0.023 0.039 ± 0.009 0.06 ± 0.026 0.075 ±0.015 0.23 ±0.18 0.015 ±0.001 0.65 ± 0.1 0.077 0.8 2.5 ± 0.436 0.036 ± 0.008 8.1 ± 1.65 0.096 ± 0.007 12.1 0.14 0.16 ±0.11 17.5 ±0.89 0.11 ±0.035 28.0 ± 5.2 0.58 ± 0.058 29.0 ± 11.5
Fold Induction
(n)
53.6 2.7 1.5 3.1 43 10.4 69 84 86 109 254.5 50
Various CHO cell types (NA, CHO, NB, W) and rat liver cells (HTC) were transiently transfected with pMMTVCAT or pRSVCAT followed by treatment with ethanol (vehicle) or 10~6 M dex. Cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. The data is expressed as a percent conversion of the [14C]chloramphenicol to acetylated metabolites and was normalized to protein as well as CAT activity observed with the control plasmid RSVCAT. Percent conversion is the percent of total [14C] chloramphenicol added that is converted to acetylated products, SD is the standard deviation. Fold induction is the dex column divided by the vehicle column, (n) is the number of individual assays. CHO cells were stably transfected with cDNA as follows: dhfr alone, CHO; dhfr and W-GR, W; dhfr and GR hormone binding mutant, NA; dhfr and GR DNA binding mutant, NB. The last number of each cell type designates the molar concentration in negative logs of MTX in the selection media.
prove valuable to understanding the mechanism of receptor action. MATERIALS AND METHODS Cells and Media CHOdhfr" (17) were grown in nonselective media [Hams' F12 supplemented with 4% newborn and 4% fetal calf serum (Irving Scientific, Santa Clara, CA)]. The rat hepatoma cell line, JZ.1 (23) was grown in Dulbeccos Modified Eagles Medium (DME) supplemented with 8% defined/supplemented bovine calf serum (Hyclone Laboratories Inc., Logan, UT). Constructions The plasmids used here have been described previously. The plasmid pSV2dhfr (21) was cotransfected into CHOdhfr cells along with one of the following GR expression vectors described by Danielsen et al. (18). Plasmid pSV2Wrec contains a cDNA clone encoding the entire W mouse GR under the control of the SV40 early region promoter. Plasmids pSV2NArec and pSV2NBrec contain the mutant mouse GR cDNAs encoding the hormone binding minus and the nuclear transfer minus receptor proteins, respectively. Transfections Transfections were by the method of Graham and Van der Eb (24) using calcium phosphate precipitation as described (21). Briefly, 1 ng pSV2dhfr combined with 10 ng GR-containing plasmid were precipitated and added to 1 x 106 CHOdhfr cells. A glycerol shock was done 4 h after the addition of DNA with cells subsequently grown in nonselective media for 2 days. Cells were then replated into selective media (DME
MTX Selections Individual colonies which were able to grow in selective media (dhfr+) were further selected in increasing concentrations of MTX (Sigma, St. Louis, MO) as follows. Cells were plated at a density of 1 x 106 cells per 10-cm dish and 1 x 10"7 M MTX added to the selective media. Resistant cells were visible in 1-2 weeks. These pools of cells were cloned by the method of limiting dilution. Clones resistant to 1 x 10~7 M were further selected in higher concentrations to a maximum of 1 x 10""4 M MTX. Immunoblotting Western blot analysis of cellular extracts for GR content was performed as described (25) with the following changes. Cells were harvested by washing once with PBS and then scraping from plates using a rubber policeman. Cell pellets were solubilized in buffer (10 mM sodium phosphate, pH 7.2, 150 HIM NaCI, 1 % NP40, 1 % sodium deoxycholate, 0.1% SDS, 0.5 mM phenylmethylsulfonyl fluoride, and 5 ng/m\ Leupeptin). Insoluble debris were removed by centrifugation at 100,000 x g for 15 min. Typically, 100-200 ng protein were run per lane on 10% SDS-polyacrylamide gels. Proteins were determined by the method of Bradford (26). Blocking of filters containing transferred protein before antibody addition was done overnight at 4 C in 20% bovine calf serum diluted in PBS and then with 1 % BSA (Sigma) in T/S (10 mM Tris, pH 8.0, 150 mM NaCI) for 30 min at room temperature. The rest of the procedure was done at room temperature. Filters were washed with T/S/T (T/S + 0.05% tween 20) 2x 5 min and then incubated with the anti-GR monoclonal antibody, GR49 (gift from Dr. H. Westphal, Institut fur Molekularbiologie und Tumorforschung Marburg, West Germany)) diluted 1:1000 in T/S/T + 1 % BSA for 1.5 h. Filters were further washed and developed using the Proto-Blot system (Promega, Madison, Wl) containing an antimouse immunoglobulin G (H+L) alkaline phosphatase conjugated second antibody. Affinity Labeling Covalent affinity labeling of GR using [3H] dex- mesylate (New England Nuclear, Boston, MA; 9.9 Ci/mmol) was done as described (25) except that cells were broken by sonication on ice in buffer (20 mM Tris, pH 7.4, 20 mM sodium molybdate, 2 mM EDTA, 2 mM 2-mercaptoethanol, 2 mM 10% glycerol) and insoluble debris removed as above. Sodium salicylate (0.5 M) was used as a fluor. Transfections and Assays for CAT Activity Transient transfection of HTC and CHO cells was performed essentially as described (27) using Lipofectin at a final concentration of 20 Mg/ml in HBS (20 mM HEPES, pH 7.4, 150 mM NaCI). Typically, 10 fig plasmid pMMTVCAT or pRSVCAT were used per 10-cm dish. Growth media were added to cells after 2 h of exposure to the lipid/DNA mixture. The following day, the media changed and 1 x 10~6 M dexamethasone (Sigma) was added to the appropriate dishes. Cells were harvested 1 day later by scraping with a rubber policeman and assayed for CAT activity as described (28). [14C]chloramphenicol (NEN, 60 Ci/mmol) and its acetylated products were separated by TLC on silica-gel plates in a solvent of CHCI3:methanol (95:5). Radioactivity was assayed by cutting the silica-gel plates and counting in Aqua-Sol (Beckman). Hormone Binding Assays Cells were harvested using trypsin/EDTA. After removing an aliquot for cell quantification, cells were spun down and washed with magnesium and calcium free PBS. All following
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 170
Vol 4 No. 1
steps were carried out at 4 C. Cell pellets were resuspended in 0.25 M sucrose, 1.0 mwi EDTA, 10 ITIM Tris HCI, pH 7.8, 5 mM «ME, and 20 rrtM Na2MoO4 (SETBM) and soncicated. Cell extracts were then spun at 100,000 x g for 60 min. One hundred-microliter aliquots (0.1-0.3 mg protein) of the resulting supernatant were then incubated with 1 x 10~7 M [3H] dexamethasone (50 Ci/mmol, Amersham, Arlington Heights, IL) for 3 h at 4 C. Parallel samples containing 250-fold molar excess unlabeled dexamethasone were used to determine nonspecific binding which was never more than 20% of the total binding. Bound and free hormone were separated using hydroxylapatite (29).
13.
Acknowledgments
14.
We thank Barbara Dieckmann for her technical assistance and Dr. H. Westphal for providing the receptor antibody, GR49. We also thank Karen Benight for her expert preparation of the manuscript.
Received August 11, 1989. Revision received October 2, 1989. Accepted October 3,1989. Address requests for reprints to: Dr. Gordon M. Ringold, Syntex Research, Institute of Cancer and Developmental Biology, 3401 Hillview Avenue, Palo Alto, California 94304. This work was supported by NJEH Postdoctoral Fellowship ES05410 (to M.A.H.) and NIH Grant GM-25821 (to G.M.R.). Preliminary data regarding this work was previously reported at a satellite symposium of the International Endocrinology Meetings (1988) and published in Cancer Research Supplement. Danielsen, M., Hinck, L, and Ringold, G.M. (1989) Mutational analysis of the mouse glucocorticoid receptor. Cancer Res. [Suppl] 49:2286s-2291s * Present address: Department of Biochemistry and Molecular Biology, Georgetown University Medical School, 3900 Reservoir Road, Northwest, Washington, DC 20007.
REFERENCES 1. Evans RM1988 The steroid and thyroid hormone receptor superfamily. Science 240:889-895 2. Beato M 1989 Gene regulation by steroid hormones. Cell 56:335-344 3. Gehring U, Arndt H 1985 Heteromeric nature of glucocorticoid receptors. FEBS Lett 179:138-142 4. Sanchez ER, Toft DO, Schlesinger MJ, Pratt WB 1985 Evidence that the 90 kDa phosphoprotein associated with the untransformed L-cell glucocorticoid receptor is a murine heat shock protein. J Biol Chem 260:12398-12401 5. Miller J, McLachian AD, Klug A 1985 Repetitive zincbinding domains in the protein transcription factor 111A from xenopus oocytes. EMBO J 4:1609-1614 6. Pratt WB, Jolly DJ, Pratt DV, Hollenberg SM, Giguere V, Cadepon FM, Schweizer-Groyer G, Cartelli MG, Evans RM, Baulieu EE 1988 A region in the steroid-binding domain determines formation of the non-DNA binding, as glucocorticoid receptor complex. J Biol Chem 263:267273 7. Payvar F, Wrange O 1984 Relative selectivities and efficiencies of DNA binding by purified intact and proteasecleaved giucocorticoid receptor. In: Eriksson H, Gustafsson J-A, Hogberg B (eds) Steroid Hormone Receptors: Structure and Function. Elsevier/North-Holland Biomedical Press, New York, pp 267-282 8. Danielsen M, Northrop JP, Jonklaas J, Ringold GM 1987 Domains of the glucocorticoid receptor involved in specific and nonspecific deoxyribonucleic acid binding, hormone activation, and transcriptional enhancement. Mol Endocrinol 1:816-822 9. Godowski PJ, Picard D, Yamamoto KR 1988 Signal trans-
10. 11.
12.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
duction and transcriptional regulation by glucocorticoid receptor-Lex A fusion proteins. Science 241:812-816 Hollenberg SM, Evans RM 1988 Multiple and cooperative trans-activation domains of the human glucocorticoid receptor. Cell 55:899-906 Munck A, Brinck-Johnsen T 1968 Specific and non-specific physiochemicalinteractions of glucocorticoids and related steroids with rat thymus cells in vivo. J Biol Chem 243:5556-5565 Schmidt TJ, Litwack G 1982 Activation of the glucocorticoid-receptor complex. Physiol Rev 62:1131 -1192 Housley PR, Pratt WB 1983 Direct demonstration of glucocorticoid receptor phosphorylation by intact L-cells. J Biol Chem 258:4630-4635 Housley PR, Grippo JF, Dahmer MK, Pratt WB 1984 Inactivation, activation, and stabilization of glucocorticoid receptors. In: Litwack G (ed) Biochemical Actions of Hormones. Academic Press, Inc, Orlando, FL, 11:347-376 Orti E, Mendel DB, Smith, LI, Munck A 1989 Agonistdependent phosphorylation and nuclear dephosphorylation of glucocorticoid receptors in intact cells. J Biol Chem 264:9728-9731 Israel Dl, Kaufman RJ 1989 Highly inducible expression from vectors containing multiple GRE's in CHO cells overexpressing the glucocorticoid receptor. Nucleic Acids Res 17:4589-4604 Urlaub G, Chasin LA 1980 Isolation of Chinese hamster cell mutants deficient in dihydrofolate reductase activity. Proc Natl Acad Sci USA 77:4216-4220 Danielsen M, Northrop JP, Ringold GM 1986 The mouse glucocorticoid receptor: mapping of functional domains by cloning, sequencing and expression of wild-type and mutant receptor proteins. EMBO J 5:2513-2522 Westphal HM, Moldenhauer G, Beato M 1982 Monoclonal antibodies to the rat liver glucocorticoid receptor. Eur Mol Biol J 1:1467-1471 Wigler M, Perucho M, Kurtz D, Dana S, Pellicen A, Axel R, Silverstein S 1980 Transformation of mammalian cells with an amplifiable dominant-acting gene. Proc Natl Acad Sci USA 77:3567-3570 Lee F, Mulligan R, Berg P, Ringold GM 1981 Glucocorticoids regulate expression of dihydrofolate reductase cDNA in mouse mammary tumor virus chimaeric plasmids. Nature 294:228-232 Miesfeld R, Rusconi S, Godowski PJ, Maler BA, Okret S, Wilkstrom A-C,Gustafsson J-A, Yamamoto KR 1986 Genetic complementation of a glucocorticoidreceptor deficiency by expression of cloned receptor cDNA. Cell 46:389-399 Grove JR, Ringold GM 1981 Selection of rat hepatoma cells defective in hormone-regulated production of mouse mammary tumor virus RNA. Proc Natl Acad Sci USA 78:4349-4353 Graham R, Van der Eb A 1973 A new technique for the assay of infectivity of human adeonovirus S DNA. Virology 52:456-467 Northrop JP, Gametchu B, Harrison RW, Ringold GM 1985 Characterization of wild type and mutant glucocorticoid receptors from rat hepatoma and mouse lymphoma cells. J Biol Chem 260:6398-6403 Bradford NM 1976 A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principal of protein-dye binding. Anal Biochem 72:248-254 Feigner P, Gadek T, Holm M, Roman R, Chan H, Wenz M, Northrop J, Ringold GM, Danielsen M 1987 Lipofection: a highly efficient, lipid-mediated DNA-transfection procedure. Proc Natl Acad Sci USA 84:7413-7417 Gorman CM, Moffat LF, Howard BH 1982 Recombinant genomes which express chioramphenicol acetyltransferase in mammalian cells. Mol Cell Biol 2:1044-1051 Wagner RK, Jungblat PW 1976 Oestradiol and dihydrotestosterone receptors in normal and neoplastic human mammary cancer. Acta Endocrinol (Copenh) 82:105-120
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
Margaret A. Hirst, Jeffrey P. Northrop, Mark Danielsen*, and Gordon M. Ringold
Institute of Cancer and Developmental Biology Syntex Research Palo Alto, California 94304 Howard Hughes Medical Institute (J.P.N.) Beckman Center Stanford University School of Medicine Stanford, California 94305
enous activity of the CHO GR. The overexpression of various forms of the GR in CHO cells serves as a paradigm for detailed biochemical analysis of this and other hormone-regulated transcription factors. (Molecular Endocrinology 4: 162-170, 1990)
We have isolated Chinese hamster ovary (CHO) cell lines expressing elevated levels of wild-type (W) and mutant forms of the glucocorticoid receptor (GR) using the technique of coamplification with a selectable dihydrofolate reductase (dhfr) cDNA. A prominent doublet at 90/92 kilodaltons was observed by Western blotting or labeling with [3H]-dexamethasone mesylate in extracts from cells transfected with W, the hormone binding mutant (NA), and the DNA binding mutant (NB). Quantification of receptor number by [3H]dexamethasone binding revealed the presence of approximately 106 receptors per cell in the W and NB-producing lines. This represents a 25to 50-fold increase in receptor density over control CHO cells which were not transfected with GR. Comparative quantitation by Western blotting of extracts from cells expressing GR showed that cells producing NA contain a level approximately 500-fold over control CHO cells. Function of the amptified receptors was examined by transient transfection with the glucocorticoid-responsive reporter plasmid pMMTV-chloramphenicol acetyl transferase (CAT). Our results indicate that inducible CAT activity increases with the abundance of W receptor and no evidence of saturability was observed even at the highest levels of receptor. This supports previous suggestions that the concentration of the hormone-regulated transcription factor is definitely limiting with regard to maximal transcription efficiency. Interestingly, cells expressing even highly amplified levels of NA-GR or NB-GR showed no inducible response above that seen with control CHO cells. Thus these mutations are exceedingly nonleaky and are not dominant over the low endog-
INTRODUCTION
Glucocorticoids, like other steroid hormones, exert their effects through a specific intracellular receptor protein that functions as a ligand-activated transcriptional regulator (1, 2). Upon hormone binding, the receptor apparently dissociates from a multimeric complex (3) containing at least one molecule of the heat-shock protein, hsp90 (4). The hormone-receptor complex subsequently binds tightly to a specific DNA sequence in target genes known as the glucocorticoid response element (GRE) whereupon either activation or repression of transcription may ensue. Biochemical and molecular dissection of wild-type and mutant forms of the glucocorticoid receptor has clearly revealed its modular composition. The approximately 86 kilodalton (kDa) receptor is composed of an approximately 10 kDa DNA binding domain containing two characteristic zinc fingers (5), flanked by a Cterminal hormone binding domain of approximately 25 kDa and an N-terminal transcription-modulating domain of approximately 50 kDa. The hormone binding domain serves to keep the receptor in a non-DNA binding conformation in the absence of ligand and contains one or more regions that interact with hsp90 (6). The Nterminal domain contains a strongly acidic region that functions to reduce nonspecific DNA interactions (7, 8). Regions in both the N-terminal domain (9) and C-terminal domain (10) can activate the transcriptional ca-
0888-8809/90/0162-0170$02.00/0 Molecular Endocrinology Copyright © 1990 by The Endocrine Society
162
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
GRs in CHO Cells
163
pacity of other DNA binding proteins deleted of their own activation functions. Substantial data has accumulated indicating that the GR undergoes covalent modification (11-13) and that its degree of phosphorylation may be altered by interaction with hormone (1215); however, the functional consequences of such modification remain to be explored. Generating a more detailed picture of the dynamics of GR function in intact cells and obtaining large quantities of purified receptor have been hampered by the relatively low abundance of receptor in normal tissues and cell lines. We have attempted to overcome this problem by generating stable Chinese hamster ovary (CHO) cell lines expressing elevated levels of W and mutant forms of the mouse GR. By cotransfection with a selectable dihydrofolate reductase (dhfr) cDNA we have succeeded in obtaining cells that express approximately 25-50 times more W or DNA-binding deficient receptor and as much as 400-500 times as much of a hormone-binding deficient receptor than do control CHO cells. Analysis of these cell lines indicates that the W-GR is fully functional as a transcriptional activator in these cells and that the inducibility of a GRE-containing mouse mammary tumor virus (MMTV) promoter increases in proportion to the level of receptor. Both forms of the mutant receptors even when expressed at high levels fail to activate transcription from the MMTV promoter and furthermore do not interfere with the function of the low levels of endogenous GR resident in the CHO cells. Last, as recently demonstrated, stable lines of CHO cells expressing copious levels of functional GR could potentially serve as a useful system in which to express foreign proteins of interest in a highly inducible fashion (16); this is of particular benefit if the protein to be expressed has deleterious effects on cell proliferation.
(data not shown). Faint bands considerably smaller than 90 kDa are also detected with the GR49 rat receptor monoclonal antibody which detects an epitope in the N-terminal domain of the GR (19). Thus the smaller forms of receptor detected in this analysis must arise by proteolytic degradation of receptor or premature termination of translation. In order to achieve higher levels of W-GR expression, several of these cell lines were exposed to higher concentrations of MTX in stepwise fashion. Pools of resistant cells at each concentration were analyzed by immunoblotting. The results shown in Fig. 1B indicate that, in general, resistance to increased levels of MTX yields cells that express more W-GR, corroborating previous studies demonstrating coamplification of unlinked cotransfected DNAs (20). The highest levels of W-GR expression were obtained at 1-3 x 10'6 M MTX. Surprisingly, selection of cells resistant to 10 4 M MTX did not yield higher levels of W-GR, perhaps due to antiproliferative effects of increased expression of a functional GR. A similar protocol was used to select cells overexpressing mutant forms NA and NB of the GR. As seen in Fig. 2, several clones were isolated expressing elevated levels of the DNA-binding mutant NB and extremely abundant levels of the hormone-binding mutant NA. These are compared to CHO cells which have been transfected with dhfr cDNA but not receptor and exhibit no detectable receptor. Clearly apparent in this immunoblot are the smaller forms of the receptor detected by the GR49 antibody. These are virtually identical to the fragments derived from W-GR (Fig. 1B) and thus are not a consequence of an altered structure due to the mutation at position 546.
RESULTS
To ascertain if the overexpressed 92 kDa receptor protein observed by immunoblot analysis is capable of binding hormone we incubated cell extracts with [3H] dex-mesylate and analyzed covalently labeled receptor by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis. As seen in Fig. 3, cells expressing elevated levels of immunoreactive W and NB-GR exhibit increased hormone binding activity at approximately 92 kDa that is specifically competed by a 200-fold excess of unlabeled dexamethasone. As expected, no labeling of the NA receptor could be detected (not shown). As in the immunoanalyses described above, a doublet of GR labeled with [3H]dex-mesylate is seen at 90/ 92 kDa. In the W cells, the upper band is labeled predominantly whereas the opposite is true in the NB cells. Conceivably the mutation in the DNA binding domain predisposes the receptor to increased proteolysis. However if this is so it must be limited to a minor clip since there is no excessive accumulation of smaller fragments observable in the NB receptor containing extracts labeled with [3H]dex-mesylate. Faint, specifically labeled smaller (-50 kDa) bands are seen in both
Immunoblot Analysis of CHO Cells Expressing W and Mutant GR After cotransfection of CHO d h f r cells (17) with dhfr and mouse GR cDNAs, dhfr+ cells were selected in a stepwise fashion for resistance to increasing concentrations of methotrexate (MTX). Three receptor cDNAs were used; W, a hormone binding mutant in which Glu546 has been changed to Gly (NA), and a DNA-binding mutant in which Arg484 has been changed to His (NB) (18). Immunoblot analysis of extracts from cells transfected with W receptor and grown in 10"7 M MTX is shown in Fig. 1 A. A prominent doublet at 90/92 kDa is observed in some of the transfectants; for comparison, immunoreactive receptor in an extract from HTC, rat hepatoma cells that contain 5-10 x 104-receptor per cell, is detectable but low in this immunoblot. Our evidence indicates that the upper band of this doublet corresponds to the size observed for mouse GR isolated from S49 cells and in vitro translated receptor
[3
H]Dexamethasone Mesylate (DEX-Mesylate) Binding to Overexpressed GR
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
— 200 — 116 — 92.5 — 66
— 45
— 31
— 21.5
W4
W12
W9
W5
B
i r
i r
VO
I
V©
I
— 200 — 116 — 92.5 — 66
— 45
— 31
— 21.5
164 The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
165
X V
• « *
13-4
o
12-5
GRs in CHO Cells
^j
z
z
z
z
Z
•4 Z
CQ
CQ
CQ
CQ
Z
Z
Z
Z
— 200 — 116 — 92.5 — 66
— 45
— 31 *****
— 21.5
Fig. 2. Immunological Analysis of CHO Cells Expressing NA and NB GR Cell extracts were prepared as described in Materials and Methods and analyzed as described in Fig. 1. NA lanes are CHO clones transfected with the GR hormone binding mutant cDNA. NA12-5 was selected in 10 s M MTX. NA1-4, NA2-4, NA5-4, NA13-4, NA4-4 were selected in 10/4 M MTX. NB lanes are CHO clones transfected with the GR DNA binding mutant cDNA. All were selected in 10"4 M MTX. HTC lane is a rat liver cell line. CHO lane is dhfr" CHO cells transfected with dhfr cDNA but not GR cDNA.
W and NB cells (Fig. 3, lanes 7 and 9). These bands become more apparent upon longer exposure and do not correspond to the smaller immunoreactive bands seen in Figs. 1 and 2. This is to be expected since the antibody and hormone affinity-label detect opposite ends of the receptor molecule. Quantification of Receptor Number In order to more carefully estimate receptor number in the overexpressing lines, we analyzed the amount of
[3H]dexamethasone binding in extracts from the various cells (Table 1). Receptors were quantified using a single concentration of 10~7 M [3H]dexamethasone. This is known to be a saturating dose in HTC cells which have a dissociation constant of approximately 10~9 M. It is possible that the overexpressed receptors in CHO cells may have a dissociation constant different than that of HTC cells. Therefore a significant decrease in receptor affinity for ligand may result in an underestimation of receptor levels in CHO cells. On a per cell basis, the highest levels of W and NB receptor obtained corre-
Fig. 1. Immunoanalysis of CHO Cells Expressing W GR Cell extracts were electrophoresed on a 10% SDS-polyacrylamide gel and transferred to nitrocellulose as described in Materials and Methods. A second antibody conjugated to alkaline phosphatase was used to detect the monoclonal antibody GR-49. W indicates CHO clones transfected with W GR. The number after the letter identifies the clone. The dashed number indicates the concentration of MTX in negative logs the clone was growing in at the time of assay. Multiple numbers specify the series of MTX concentrations used in the stepwise selection of cells. All lanes were loaded with 200 nQ protein. Molecular weight markers were visualized with Coomassie staining and are indicated at the right (in kilodaltons). A, Twelve CHO clones initially selected in 10'7 MTX and a rat liver cell line (HTC). B, Four CHO clones that were further selected in higher concentrations (10"7to 3 x I O ^ M ) MTX as indicated and analyzed as pools.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
Vol 4 No. 1
MOL ENDO-1990 166
HTC 1
2
W4-7 3
4
W9-7-6 5
6
W12-7-6 7
8
NB1-5 9
10 — 200 — 116 — 92.5
— 66
— 45
— 31
— 21.5
Fig. 3. Autoradiogram of [3H]Dex-Mesylate Labeled Cell Extracts Electrophoresed on a 10% SDS-Polyacrylamide Gel Cell extracts were prepared as described in Materials and Methods. Samples containing 200 ng protein were incubated with [3H] dex-mesylate at a final concentration of 1 x 10"7 M for 3 h at 4 C. Nonspecific binding was determined by addition of 200-fold molar excess of unlabeled dexamethasone. Cell extracts tested are labeled above the lane numbers. Odd numbered lanes, Extracts incubated with labeled hormone. Even numbered lanes, Extracts incubated with labeled and unlabeled hormone. Molecular weight markers were visualized with Coomasie stain and indicated at the right in kilodaltons.
spond to approximately 5-10 x 105 receptors. This compares with the 25- to 50-fold lower concentration (~2 x 104/cell) of GR in CHO cells and 5- to 10-fold lower level (~1 x 10s/cell) in HTC cells. The low level of hormone binding (equivalent to 2-3 x 104 receptors per cell) detected in NA-containing cells reflects the endogenous GR in CHO dhfr cells (21). Since the NA receptor cannot bind hormone we have estimated its relative abundance in the transfected cells by a semiquantitative immunological analysis. The results shown in Fig. 4 indicate that approximately 10 ^g cell protein from the cell line NA4-4 give rise to a band of equal intensity to that seen with 200 ^g cell protein from the NB2-5 cell line (containing ~5 x 105receptors per cell). Similar studies comparing NA4-4 with W4-76-4 yielded equivalent results (not shown). Based on
these experiments we conclude that NA-transfected cells contain as many as 10 x 106 receptors per cell. This number is only an approximation since the actual number has not been rigorously quantified. The reasons for our ability to select NA cells expressing higher levels of receptor than either NB or W-cells remain obscure. Functional Response to Glucocorticoids in CHO Cells Expressing Various Levels of GR Representative pools of cells containing various levels of W receptor were transiently transfected with the glucocorticoid responsive MMTV promoter-CAT fusion plasmid. The efficiency of transfection of each cell line was evaluated with pRSVCAT and all data have been normalized to the maximal expression seen in each
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
167
GRs in CHO Cells
Table 1. [3H]Dexamethasone [3H]Dex Binding in Cell Extracts Cell Line
W4-7-6-4 W5-7-6-3~6 W9-7-6-3-6 W-12-7-6 NB1-5 NB2-5 NB3-5 NA4-4 CHO HTC
Receptor Number/Cell (Minimum Estimation)
4.8 9.1 5.2 7.4 3.5 4.8 6.5 2.9 1.9 9.8
x x x x x x x x x x
10s 10s 10s 105 105 10s 10s 104 104 104
(Range)
(n)
(4.3-5.3) (8.5-9.7) (4.9-5.7) (6.2-8.5) (2.9-4.1)
(2) (2) (3) (2) (3) (1) (1) (1)
(6.8-11)
(3)
(D
Various CHO cell types (W, NB, NA, CHO) and rat liver cells (HTC) were removed from dishes by trypsin, counted with a hemacytometer, washed, and sonicated in SETBM buffer. Supematants from a 100,000 x g spin were incubated with 10~7 M [3H]dex for 3 h at 4 C. Total ([3H]dex) and nonspecific ([3H]dex plus 250-fold molar excess unlabeled dex) binding were analyzed in duplicate. Specific [3H]bound ligand (total minus nonspecific) was used to estimate minimum receptor number per cell. The range (Range) of values is shown followed by the number (n) of independent determinations. CHO cells were stably transfected with cDNA as follows: dhfr alone, CHO; dhfr and W-GR, W; dhfr and GR hormone binding mutant, NA; dhfr and GR DNA binding mutant, NB. The last number of each cell line designates the molar concentration in negative logs of MTX in the selection media.
experiment with this control vector. The results (Fig. 5
and Table 2) show that increased expression of W receptor gives rise to cells that support greater inducibility of the MMTV-CAT plasmid. Even at the highest levels of receptor expression we see no evidence of saturating the inducible response. Thus, receptor number appears to be a critical limiting factor in achieving increased hormone-dependent transcription from the MMTV promoter. Interestingly, there is a small increase in the basal level of CAT expression in cells expressing the highest concentration of receptor. This may be due to the presence of small amounts of corticoids in the medium (or serum) or alternatively may reflect spontaneous activation of receptor in a hormone-independent fashion. In cells overexpressing the mutant NA or NB receptors, we have been unable to detect any significant increase in glucocorticoid responsiveness using MMTVCAT. This documents that the mutants are indeed very tight and in addition are unable to interfere with the weak induction elicited by the endogenous CHO receptor. Moreover, the hormone binding mutant NA, even when expressed at approximately 10 x 106 receptor per cell, does not increase basal expression from the MMTV promoter. We surmise that it is maintained in a fully inactive form perhaps through its association with other proteins such as hsp 90 (3, 4). DISCUSSION The ability to study many aspects of GR structure, function, and metabolism are compromised by the rel-
atively low levels of GR present in most cells. To overcome this limitation, we have succeeded in overexpressing the mouse GR in CHO cells by the method of coamplification with a dhfr vector. Although we presume this method resulted in gene amplification of receptor, we did not directly examine this. This procedure is relatively easy to perform and can be extended to other receptors of interest. However the absolute level of expression cannot be predicted for any particular protein; we have, for example, obtained 10-fold higher expression of the hormone binding GR mutant NA than of either W or the DNA-binding mutant NB. It will be of interest to see whether this is fortuitously due to the sites of integration of the transfected DNAs or whether there is indeed selection against higher level of W-GR and NB-GR expression. Several revealing facts about transcription activation by the GR have emerged from analysis of the CHO cell transfectants. First, as has also been reported by Miesfeld et at. (22), increased levels of functional GR result in higher transcriptional activity from a glucocorticoidresponsive promoter. Moreover, our results indicate that the inducible expression of the MMTV-CAT plasmid is not saturated even with approximately 106 receptors per cell. Thus, not only is receptor a limiting factor, but the cell's capacity to use high levels of receptor is substantially greater than one might have predicted a priori. Second, the fact that we do not observe highly increased expression of MMTV-CAT in the absence of hormone indicates that there is excess capacity to maintain the receptor in the inactive state when in the unbound state. This presumably reflects the high cellular concentration of hsp 90 (and perhaps other proteins) with which the receptor is presumably associated in a high molecular weight complex (3). Whether or not the GR in the amplified CHO cells resides in such a complex (under hormone-free conditions) remains to be directly tested. If, as one would expect, this is so, then these cells may prove extremely valuable in exploring the structure of the inactive complex in finer detail. Third, it is of interest to note that the point mutations giving rise to the NA and NB receptors result in extremely tight phenotypes in which no inducibility of MMTV-CAT can be detected. It is perhaps also noteworthy that the NA mutant, even when present at approximately 107 molecules per cell remains recessive; that is, it does not interfere with the low level activity of the endogenous (104) receptors present in CHO cells. Since the inducibility of a GRE-containing promoter is markedly elevated in cells containing approximately 106 W receptors such a system would be useful for obtaining conditional high level expression of a protein that might otherwise be toxic to cells. The feasibility of this approach has been independently demonstrated in a recent publication (16). In addition to this practical use, the use of cell lines expressing high levels of GR for detailed study of receptor modification, receptorprotein interactions, and intracellular compartmentalization of W and mutant receptors will undoubtedly
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 168
Vol 4 No. 1
NA4-4 ug protein/lane
200
100
50
NB2-5 20
10
200
100
50
20 — 200
— 31
— 21.5 Fig. 4. Immunoblots of Various Cell Extracts at Increasing Protein Concentrations Electrophoresed on a SDS-Polyacrylamide Gel Cell extracts were prepared as described. Increasing protein was loaded from right to left for CHO clones NA4-4 and NB2-5. The numbers above each lane indicate the quantity of protein loaded.
AcCM
Dex Cell Line
-
HTC
W3-7
W9-7
W5-7
W9-7-6-3
CHO
NA4-4
NB1-5
Fig. 5. CAT activity from various cells transfected with pMMTVCAT As described in Materials and Methods, extracts were prepared from cells transiently transfected with the glucocorticoid responsive pMMTVCAT. After transfection, cells were treated with vehicle (-) or 10.-6 M dexamethasone (+) for 48 h. Equal quantities of cell extracts were assayed for CAT activity. Cells expressing W-GR in low (W3-7), medium (W9-7), high (W5-7), and very high (W9-7-6-3" 6 ) abundance are compared. NA4-4 and NB 1-5 are clones expressing high levels of mutant GR. Cells with normal receptor quantity is shown by HTC. CHO represents response to GR with no amplified receptor protein. CM, Unconverted [14C]chioramphenicol. AcCM, Acetylated [14C] chloramphenicol.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
169
GRs in CHO Cells
supplemented with 39.5 fig/m\ proline and 8% defined/supplemented bovine calf serum).
Table 2. Glucocorticoid-lnduced MMTVCAT Activity % Conversion Cell Line Vehicle ± SD
HTC NA4-4 CHO NB1-5 W3-7
W10-7 W9-7 W2-7 W5-7 W12-7 W5-3"6 W9-3"6
Dex ± SD
2.2 0.041 0.027 ± 0.004 0.072 ± 0.023 0.039 ± 0.009 0.06 ± 0.026 0.075 ±0.015 0.23 ±0.18 0.015 ±0.001 0.65 ± 0.1 0.077 0.8 2.5 ± 0.436 0.036 ± 0.008 8.1 ± 1.65 0.096 ± 0.007 12.1 0.14 0.16 ±0.11 17.5 ±0.89 0.11 ±0.035 28.0 ± 5.2 0.58 ± 0.058 29.0 ± 11.5
Fold Induction
(n)
53.6 2.7 1.5 3.1 43 10.4 69 84 86 109 254.5 50
Various CHO cell types (NA, CHO, NB, W) and rat liver cells (HTC) were transiently transfected with pMMTVCAT or pRSVCAT followed by treatment with ethanol (vehicle) or 10~6 M dex. Cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. The data is expressed as a percent conversion of the [14C]chloramphenicol to acetylated metabolites and was normalized to protein as well as CAT activity observed with the control plasmid RSVCAT. Percent conversion is the percent of total [14C] chloramphenicol added that is converted to acetylated products, SD is the standard deviation. Fold induction is the dex column divided by the vehicle column, (n) is the number of individual assays. CHO cells were stably transfected with cDNA as follows: dhfr alone, CHO; dhfr and W-GR, W; dhfr and GR hormone binding mutant, NA; dhfr and GR DNA binding mutant, NB. The last number of each cell type designates the molar concentration in negative logs of MTX in the selection media.
prove valuable to understanding the mechanism of receptor action. MATERIALS AND METHODS Cells and Media CHOdhfr" (17) were grown in nonselective media [Hams' F12 supplemented with 4% newborn and 4% fetal calf serum (Irving Scientific, Santa Clara, CA)]. The rat hepatoma cell line, JZ.1 (23) was grown in Dulbeccos Modified Eagles Medium (DME) supplemented with 8% defined/supplemented bovine calf serum (Hyclone Laboratories Inc., Logan, UT). Constructions The plasmids used here have been described previously. The plasmid pSV2dhfr (21) was cotransfected into CHOdhfr cells along with one of the following GR expression vectors described by Danielsen et al. (18). Plasmid pSV2Wrec contains a cDNA clone encoding the entire W mouse GR under the control of the SV40 early region promoter. Plasmids pSV2NArec and pSV2NBrec contain the mutant mouse GR cDNAs encoding the hormone binding minus and the nuclear transfer minus receptor proteins, respectively. Transfections Transfections were by the method of Graham and Van der Eb (24) using calcium phosphate precipitation as described (21). Briefly, 1 ng pSV2dhfr combined with 10 ng GR-containing plasmid were precipitated and added to 1 x 106 CHOdhfr cells. A glycerol shock was done 4 h after the addition of DNA with cells subsequently grown in nonselective media for 2 days. Cells were then replated into selective media (DME
MTX Selections Individual colonies which were able to grow in selective media (dhfr+) were further selected in increasing concentrations of MTX (Sigma, St. Louis, MO) as follows. Cells were plated at a density of 1 x 106 cells per 10-cm dish and 1 x 10"7 M MTX added to the selective media. Resistant cells were visible in 1-2 weeks. These pools of cells were cloned by the method of limiting dilution. Clones resistant to 1 x 10~7 M were further selected in higher concentrations to a maximum of 1 x 10""4 M MTX. Immunoblotting Western blot analysis of cellular extracts for GR content was performed as described (25) with the following changes. Cells were harvested by washing once with PBS and then scraping from plates using a rubber policeman. Cell pellets were solubilized in buffer (10 mM sodium phosphate, pH 7.2, 150 HIM NaCI, 1 % NP40, 1 % sodium deoxycholate, 0.1% SDS, 0.5 mM phenylmethylsulfonyl fluoride, and 5 ng/m\ Leupeptin). Insoluble debris were removed by centrifugation at 100,000 x g for 15 min. Typically, 100-200 ng protein were run per lane on 10% SDS-polyacrylamide gels. Proteins were determined by the method of Bradford (26). Blocking of filters containing transferred protein before antibody addition was done overnight at 4 C in 20% bovine calf serum diluted in PBS and then with 1 % BSA (Sigma) in T/S (10 mM Tris, pH 8.0, 150 mM NaCI) for 30 min at room temperature. The rest of the procedure was done at room temperature. Filters were washed with T/S/T (T/S + 0.05% tween 20) 2x 5 min and then incubated with the anti-GR monoclonal antibody, GR49 (gift from Dr. H. Westphal, Institut fur Molekularbiologie und Tumorforschung Marburg, West Germany)) diluted 1:1000 in T/S/T + 1 % BSA for 1.5 h. Filters were further washed and developed using the Proto-Blot system (Promega, Madison, Wl) containing an antimouse immunoglobulin G (H+L) alkaline phosphatase conjugated second antibody. Affinity Labeling Covalent affinity labeling of GR using [3H] dex- mesylate (New England Nuclear, Boston, MA; 9.9 Ci/mmol) was done as described (25) except that cells were broken by sonication on ice in buffer (20 mM Tris, pH 7.4, 20 mM sodium molybdate, 2 mM EDTA, 2 mM 2-mercaptoethanol, 2 mM 10% glycerol) and insoluble debris removed as above. Sodium salicylate (0.5 M) was used as a fluor. Transfections and Assays for CAT Activity Transient transfection of HTC and CHO cells was performed essentially as described (27) using Lipofectin at a final concentration of 20 Mg/ml in HBS (20 mM HEPES, pH 7.4, 150 mM NaCI). Typically, 10 fig plasmid pMMTVCAT or pRSVCAT were used per 10-cm dish. Growth media were added to cells after 2 h of exposure to the lipid/DNA mixture. The following day, the media changed and 1 x 10~6 M dexamethasone (Sigma) was added to the appropriate dishes. Cells were harvested 1 day later by scraping with a rubber policeman and assayed for CAT activity as described (28). [14C]chloramphenicol (NEN, 60 Ci/mmol) and its acetylated products were separated by TLC on silica-gel plates in a solvent of CHCI3:methanol (95:5). Radioactivity was assayed by cutting the silica-gel plates and counting in Aqua-Sol (Beckman). Hormone Binding Assays Cells were harvested using trypsin/EDTA. After removing an aliquot for cell quantification, cells were spun down and washed with magnesium and calcium free PBS. All following
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
MOL ENDO-1990 170
Vol 4 No. 1
steps were carried out at 4 C. Cell pellets were resuspended in 0.25 M sucrose, 1.0 mwi EDTA, 10 ITIM Tris HCI, pH 7.8, 5 mM «ME, and 20 rrtM Na2MoO4 (SETBM) and soncicated. Cell extracts were then spun at 100,000 x g for 60 min. One hundred-microliter aliquots (0.1-0.3 mg protein) of the resulting supernatant were then incubated with 1 x 10~7 M [3H] dexamethasone (50 Ci/mmol, Amersham, Arlington Heights, IL) for 3 h at 4 C. Parallel samples containing 250-fold molar excess unlabeled dexamethasone were used to determine nonspecific binding which was never more than 20% of the total binding. Bound and free hormone were separated using hydroxylapatite (29).
13.
Acknowledgments
14.
We thank Barbara Dieckmann for her technical assistance and Dr. H. Westphal for providing the receptor antibody, GR49. We also thank Karen Benight for her expert preparation of the manuscript.
Received August 11, 1989. Revision received October 2, 1989. Accepted October 3,1989. Address requests for reprints to: Dr. Gordon M. Ringold, Syntex Research, Institute of Cancer and Developmental Biology, 3401 Hillview Avenue, Palo Alto, California 94304. This work was supported by NJEH Postdoctoral Fellowship ES05410 (to M.A.H.) and NIH Grant GM-25821 (to G.M.R.). Preliminary data regarding this work was previously reported at a satellite symposium of the International Endocrinology Meetings (1988) and published in Cancer Research Supplement. Danielsen, M., Hinck, L, and Ringold, G.M. (1989) Mutational analysis of the mouse glucocorticoid receptor. Cancer Res. [Suppl] 49:2286s-2291s * Present address: Department of Biochemistry and Molecular Biology, Georgetown University Medical School, 3900 Reservoir Road, Northwest, Washington, DC 20007.
REFERENCES 1. Evans RM1988 The steroid and thyroid hormone receptor superfamily. Science 240:889-895 2. Beato M 1989 Gene regulation by steroid hormones. Cell 56:335-344 3. Gehring U, Arndt H 1985 Heteromeric nature of glucocorticoid receptors. FEBS Lett 179:138-142 4. Sanchez ER, Toft DO, Schlesinger MJ, Pratt WB 1985 Evidence that the 90 kDa phosphoprotein associated with the untransformed L-cell glucocorticoid receptor is a murine heat shock protein. J Biol Chem 260:12398-12401 5. Miller J, McLachian AD, Klug A 1985 Repetitive zincbinding domains in the protein transcription factor 111A from xenopus oocytes. EMBO J 4:1609-1614 6. Pratt WB, Jolly DJ, Pratt DV, Hollenberg SM, Giguere V, Cadepon FM, Schweizer-Groyer G, Cartelli MG, Evans RM, Baulieu EE 1988 A region in the steroid-binding domain determines formation of the non-DNA binding, as glucocorticoid receptor complex. J Biol Chem 263:267273 7. Payvar F, Wrange O 1984 Relative selectivities and efficiencies of DNA binding by purified intact and proteasecleaved giucocorticoid receptor. In: Eriksson H, Gustafsson J-A, Hogberg B (eds) Steroid Hormone Receptors: Structure and Function. Elsevier/North-Holland Biomedical Press, New York, pp 267-282 8. Danielsen M, Northrop JP, Jonklaas J, Ringold GM 1987 Domains of the glucocorticoid receptor involved in specific and nonspecific deoxyribonucleic acid binding, hormone activation, and transcriptional enhancement. Mol Endocrinol 1:816-822 9. Godowski PJ, Picard D, Yamamoto KR 1988 Signal trans-
10. 11.
12.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
duction and transcriptional regulation by glucocorticoid receptor-Lex A fusion proteins. Science 241:812-816 Hollenberg SM, Evans RM 1988 Multiple and cooperative trans-activation domains of the human glucocorticoid receptor. Cell 55:899-906 Munck A, Brinck-Johnsen T 1968 Specific and non-specific physiochemicalinteractions of glucocorticoids and related steroids with rat thymus cells in vivo. J Biol Chem 243:5556-5565 Schmidt TJ, Litwack G 1982 Activation of the glucocorticoid-receptor complex. Physiol Rev 62:1131 -1192 Housley PR, Pratt WB 1983 Direct demonstration of glucocorticoid receptor phosphorylation by intact L-cells. J Biol Chem 258:4630-4635 Housley PR, Grippo JF, Dahmer MK, Pratt WB 1984 Inactivation, activation, and stabilization of glucocorticoid receptors. In: Litwack G (ed) Biochemical Actions of Hormones. Academic Press, Inc, Orlando, FL, 11:347-376 Orti E, Mendel DB, Smith, LI, Munck A 1989 Agonistdependent phosphorylation and nuclear dephosphorylation of glucocorticoid receptors in intact cells. J Biol Chem 264:9728-9731 Israel Dl, Kaufman RJ 1989 Highly inducible expression from vectors containing multiple GRE's in CHO cells overexpressing the glucocorticoid receptor. Nucleic Acids Res 17:4589-4604 Urlaub G, Chasin LA 1980 Isolation of Chinese hamster cell mutants deficient in dihydrofolate reductase activity. Proc Natl Acad Sci USA 77:4216-4220 Danielsen M, Northrop JP, Ringold GM 1986 The mouse glucocorticoid receptor: mapping of functional domains by cloning, sequencing and expression of wild-type and mutant receptor proteins. EMBO J 5:2513-2522 Westphal HM, Moldenhauer G, Beato M 1982 Monoclonal antibodies to the rat liver glucocorticoid receptor. Eur Mol Biol J 1:1467-1471 Wigler M, Perucho M, Kurtz D, Dana S, Pellicen A, Axel R, Silverstein S 1980 Transformation of mammalian cells with an amplifiable dominant-acting gene. Proc Natl Acad Sci USA 77:3567-3570 Lee F, Mulligan R, Berg P, Ringold GM 1981 Glucocorticoids regulate expression of dihydrofolate reductase cDNA in mouse mammary tumor virus chimaeric plasmids. Nature 294:228-232 Miesfeld R, Rusconi S, Godowski PJ, Maler BA, Okret S, Wilkstrom A-C,Gustafsson J-A, Yamamoto KR 1986 Genetic complementation of a glucocorticoidreceptor deficiency by expression of cloned receptor cDNA. Cell 46:389-399 Grove JR, Ringold GM 1981 Selection of rat hepatoma cells defective in hormone-regulated production of mouse mammary tumor virus RNA. Proc Natl Acad Sci USA 78:4349-4353 Graham R, Van der Eb A 1973 A new technique for the assay of infectivity of human adeonovirus S DNA. Virology 52:456-467 Northrop JP, Gametchu B, Harrison RW, Ringold GM 1985 Characterization of wild type and mutant glucocorticoid receptors from rat hepatoma and mouse lymphoma cells. J Biol Chem 260:6398-6403 Bradford NM 1976 A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principal of protein-dye binding. Anal Biochem 72:248-254 Feigner P, Gadek T, Holm M, Roman R, Chan H, Wenz M, Northrop J, Ringold GM, Danielsen M 1987 Lipofection: a highly efficient, lipid-mediated DNA-transfection procedure. Proc Natl Acad Sci USA 84:7413-7417 Gorman CM, Moffat LF, Howard BH 1982 Recombinant genomes which express chioramphenicol acetyltransferase in mammalian cells. Mol Cell Biol 2:1044-1051 Wagner RK, Jungblat PW 1976 Oestradiol and dihydrotestosterone receptors in normal and neoplastic human mammary cancer. Acta Endocrinol (Copenh) 82:105-120
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
