| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pathology and Program in Molecular Biology (Y.W.,
V.G.T., S.K.N.) University of Colorado Health Sciences Center
Denver, Colorado 80262
Department of Pharmacology and
Physiology (K.K.C., P.R.H.) University of South Carolina School of
Medicine Columbia, South Carolina 29208
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The receptors for progesterone (PR) and for glucocorticoids (GR) are two closely related members of the nuclear receptor family of transcription factors. They share very similar domain structures and functional characteristics. Each N-terminal domain of the two receptors, while nonhomologous, has ligand-independent transcription activation function; the DNA-binding domain (DBD) exhibits 86% sequence identity; the short hinge region has little identity between receptors but is the location of a major nuclear localization signal (NLS); and finally, the C-terminal ligand-binding-domain (LBD) exhibits 54% identity between human (h) GR and hPR, overlapping ligand-binding specificity, and a ligand-dependent transcription activation function (3). The DBD and the LBD have been characterized extensively both structurally and functionally. For both receptors an optimal recognition site is an inverted hexameric palindrome separated by 3 bp, PuGNACANNNTGTNCPy (4). In the absence of hormone, receptor monomers form an inactive complex with molecular chaperones, most notably heat shock protein 90 (hsp90) (5). Upon binding of an agonist ligand to the receptor, it dissociates from the chaperones and undergoes dimerization. The dimerized receptor binds to recognition sites in target promoters and recruits coactivators, resulting in increased transcription initiation at those promoters. In a number of cases, both receptors can mediate induction of the same genes (6, 7), although there are poorly understood influences of chromatin that can differentially modulate induction by GR and PR (8). How can two receptors with such remarkable similarities mediate such dissimilar biological effects? The mechanism underlying the distinct biological effects of glucocorticoids and progestins, even in cells where both receptors are present, is a question of significant interest.
Despite the remarkable similarities between these two receptors, previous biochemical and immunochemical studies have shown that the subcellular localization of GR vis a vis PR in the absence of ligands is quite distinct. Although both receptors appear to continuously shuttle, at any given moment GR is found largely in the cytoplasm (9–12), while PR localizes to the nucleus (13). Hormone binding results in tight association of the receptors with the nucleus. The multipartite NLSs of GR and PR have been mapped in some detail (14–18). The major NLS is located just C-terminal to the DBD in the hinge domain and is comprised of a stretch of clustered basic residues. Two additional basic clusters in the second zinc finger of the DBD contribute to the overall NLS activity. Interestingly, it has been shown that the NLS in GR, which is cytoplasmic in the absence of hormone, is just as potent as that of PR and ER, which are both nuclear. A mutant PR, whose NLS has been replaced by the NLS of GR, also localized to the nucleus in the absence of ligand (14). Thus, current hypothesis posits that the differential localization of GR and PR is not determined by the NLS but rather that the GR LBD can mask the NLS. This view is based on the following observations: 1) LBD-truncated GR is constitutively nuclear and has constitutive transcriptional activity (19–21); 2) when the LBD is moved to the N terminus of GR, leaving the NLS at the C-terminal end of the receptor, the receptor is constitutively nuclear (22); 3) an antibody (AP64) against the major NLS can react with liganded GR but not unliganded receptor (23). It is unclear whether the LBD itself or the hsp90 bound to the LBD masks the NLS, although some evidence favors hsp90 (24). However, the hypothesis that the LBD or associated proteins masks the NLS has been challenged (25, 26).
In this study, we seek to define the receptor domain that determines the differential localization between GR and PR. We use receptors tagged with green fluorescent protein (GFP) to monitor the subcellular localization of steroid receptors by confocal microscopy. GFP from the jellyfish, Aequorea victoria, has been developed into an extremely useful tool to monitor protein localization and trafficking. Many proteins, when fused to GFP, maintain their normal function and localization (27). The utility of GFP has been extended by the selection of enhanced variants that exhibit increased fluorescence; in addition, codon usage has been humanized to improve translation. Fluorescence can be monitored in living cells, thereby avoiding artifacts caused by biochemical fractionation or fixation. GFP is proving to be an extremely useful means of dissecting the localization of nuclear receptors and the dynamics of receptor trafficking (26, 28–31). Here, we use GFP-tagged receptors to monitor the differential subcellular localization between GR and PR. By assessing a series of GFP-tagged receptor chimeras, we show that it is the LBD that determines the differential localization of the two highly related receptors. Chimeras between the two LBDs implicate multiple features within the N-terminal half of the LBD in the specification of the nuclear or cytoplasmic localization of the receptor in the absence of hormone. However, the context of this region within the receptor is critical since it cannot by itself promote nuclear import or export. Mutagenesis studies of steroid receptors along with ligand-LBD crystal structures from different members of the nuclear receptor family, including PR (32, 33), implicate residues in the N-terminal portion of the LBD in ligand binding specificity. Nonetheless, we show that there is not a strict correlation between the ligand binding specificity of the receptor, its subcellular localization, or capacity for transactivation, implying that distinct structural features determine these properties.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Receptor localization was assessed by confocal microscopy in living
cells. Similar results were seen in all three lines, and confocal data
for COS-1 and E82.A3 cells are shown in Fig. 1
. GFP itself is distributed throughout
the cell; this distribution is unaffected by hormones. GFP-GR is found
predominantly in the cytoplasm in the absence of hormone. Treatment of
GFP-GR-expressing cells with dexamethasone is accompanied by the rapid
movement of receptor to the nucleus (t1/2 
5
min). Nuclear receptor remains excluded from nucleoli. In contrast,
GFP-PR localizes predominantly to the nucleus in the absence of
hormone, although cytoplasmic GFP-PR is sometimes seen, especially when
the receptor is overexpressed. GFP-PR remains nuclear after hormone
addition as expected.
|
). These results demonstrate that the
GFP-GR and GFP-PR chimeras recapitulate the cellular distribution
expected of GR and PR as well as the wild-type transcriptional
activity.
|
). The presence of a unique BclI site
adjacent to the hinge-LBD border of hPR (35) facilitates the
construction of GFP-tagged receptors in which the LBDs of PR and GR are
exchanged. Thus, G/P contains the entire N terminus, DBD, and hinge
from hGR and the entire LBD from hPR with GFP at the N terminus of the
chimeric receptor. Conversely, P/G is the complement where only the LBD
is derived from GR. Unliganded G/P is nuclear in both COS-1 cells and
E82.A3 cells, whereas unliganded P/G is cytoplasmic, redistributing to
the nucleus on dexamethasone addition (Fig. 2B). Thus, while the
multipartite NLS is necessary for nuclear import of both GR and PR
(14, 15, 16, 17, 18), it is the LBD that determines the differential localization
of the unliganded receptor. The chimeric receptors retained
transcriptional activity. P/G was fully activated by the synthetic
glucocorticoid dexamethasone and G/P by the synthetic progestin R5020
(Fig. 2C).
Nuclear Localization Specificity Maps to the N-Terminal Half of the
LBD but Is Separable from Ligand Binding Specificity
To localize further the region within the LBD that is responsible
for the differential localization of GR and PR, additional chimeras
within the LBD were created. Portions of the PR LBD were replaced by
the homologous portions of GR, taking advantage of three natural
restriction sites in PR that divide the LBD into four segments (see
Fig. 3A and Materials and
Methods). Thus, in the expression vectors
P/P1-9G1
and P/P1-7G, C-terminal PR sequences have been
replaced with 82 or 123 amino acids of GR sequence, respectively. When
expressed in cells, both receptors were nuclear without ligand (Fig. 3B), like PR itself, indicating that the C-terminal half of the
LBD has little role in differential receptor localization.
Furthermore, since P/P1-7G
differs from the P/G only in the origin of the N-terminal half of the
LBD yet exhibits nuclear rather than cytoplasmic localization, it
suggests that sequences encompassed by the 123 amino acids of the
helices 1–7 segment determine differential localization of GR and PR.
The construction and analysis of additional vectors confirmed this
suggestion. G/P1-9G and
G/P1-7G were nuclear in
the absence of ligand just as
P/P1-9G and
P/P1-7G (Fig. 3),
confirming that sequences N-terminal to the LBD are not involved in the
differential localization of GR and PR.
|
), indicating that sequences within the helices 4–7
segment play a role in specifying localization. However, helices 4–7
are insufficient to specify localization, because the chimera
P/GP4-12, the converse of
G/P1-3G, is, like
G/P1-3G, also cytoplasmic
in the absence of hormone (Fig. 3). If an LBD chimera and its converse
both exhibit a similar localization pattern, one of two possibilities
exists: either both chimeras exhibit the default localization phenotype
because the localization signal has been disrupted or, conversely, both
chimeras exhibit a directed localization because the signal has
multiple components that can each act independently. Thus, the finding
that the converse pair,
P/GP4-12 and
G/P1-3G, are both
cytoplasmic could signify that this is the default localization,
suggesting that in the chimeras an NLS present in PR has been
disrupted. This could be either a single, contiguous determinant or a
multipartite signal. Alternatively, nuclear localization is the default
and determinants from both the GR helices 4–7 and GR helices 1–3
regions can independently specify cytoplasmic localization,
e.g. by serving as nuclear export signals. We believe this
latter alternative to be less likely, in part because placing an NLS at
the N terminus of GR creates a constitutively nuclear receptor (Ref. 22
and L. Taraseviciene and S. K. Nordeen, unpublished results).
Additional results below address these possibilities and further
localize the determinants. Chimera G/P1-3G redistributes to the nucleus upon addition of the glucocorticoid dexamethasone. In contrast, P/GP4-12 does not. However, P/GP4-12 can bind the progestin R5020, albeit poorly, since it only partially redistributes to the nucleus after exposure to ligand. Since P/GP4-12 binds progestins in preference to glucocorticoids yet is cytoplasmic in the absence of ligand, it suggests that localization and hormone binding specificity are separable properties of receptors. Additional chimeras confirm this suggestion as detailed below.
Multiple Subdomains of the LBD Contribute to Localization
The next series of chimeras were constructed to determine which
domains within the N-terminal half of the LBD are involved in receptor
localization (Fig. 4A).
P/P1-speG is a chimera in which the fusion was
made at an introduced SpeI site within the third helix. The
converse chimera is
G/GPspe-12 in which the
only LBD sequences from GR are the first helix, spacer, and the first 4
amino acids of the third helix. When expressed in cells, both of these
chimeric receptors are cytoplasmic like GR (Fig. 4B). The fusion of PR
and GR sequences at the introduced SpeI site within helix 3
may have disrupted an NLS. To test this, the four GR residues of the
third helix of G/GPspe-12
LBD (WRIM) were mutated to PR sequence (SSLL) to create
G/GP3-12 where the entire
helix 3 to C-terminal sequence is from PR. In the absence of ligand,
this chimera is still largely cytoplasmic. This suggests that PR LBD
sequence N-terminal to helix 3, along with the helices 4–7 domain
implicated above, contribute to the specification of a PR-like nuclear
localization.
|
Results presented above (Fig. 3) demonstrated that
G/P1-3G is cytoplasmic and
that G/P1-7G is nuclear.
The next series of constructs attempted to refine the requirement for
PR sequences from helices 4 through 7 in nuclear localization. Some of
the greatest sequence disparity between the LBD of GR and PR occurs in
helix 7. Furthermore, sequences in helix 7, helix 6, and the ß-turn
structure preceding helix 6 have been implicated in differential ligand
recognition (36). We therefore constructed chimeras based on
G/P1-3G and
G/P1-speG replacing GR sequence with PR helix 7
or a larger PR block encompassing the ß-turn, helix 6, and helix 7
(Fig. 5, A and B). All of the
resulting double chimeras,
G/P1-3GP7G,
G/P1-3GP6-7G,
G/P1-speGP7G, and
G/P1-speGP6-7G,
retained a cytoplasmic phenotype in the absence of ligand (Fig. 5C),
indicating a role for PR sequences in helix 4 and/or helix 5 in the
acquisition of a nuclear phenotype. The double LBD chimeras all had
altered ligand recognition specificity compared with the progenitors,
G/P1-3G and
G/P1-speG. All double chimeras could translocate
to the nucleus in response to R5020, and all exhibited a reduced
ability of dexamethasone to induce translocation (Fig. 5C).
|
, A and B). It exhibits largely nuclear localization in the absence of
ligand (Fig. 5C). This chimera represented the receptor with the
smallest region of PR that retained a predominantly nuclear phenotype
in the absence of ligand. This chimera binds glucocorticoids, as
evidenced by the fact that the remaining cytoplasmic receptor could be
driven to the nucleus upon addition of dexamethasone and is
transcriptionally active in response to dexamethasone (see below).
Thus, several different chimeras demonstrated that a receptor
binding a glucocorticoid ligand could be nuclear in the absence of
ligand and, conversely, that a progestin binding receptor could be
cytoplasmic in the ligand-free state. Therefore, the structural
features that determine localization in the ligand-free state and those
that determine ligand specificity are distinct and separable.
The Sequences That Specify Nuclear Localization Do Not Act as an
Independent NLS
To test whether a region of PR that can specify nuclear
localization contains an independent NLS, the 123-amino acid domain
encompassing helices 1–7 was fused to a dimer of GFP (Fig. 6A). A GFP dimer has a molecular
mass of about 54 kDa and is distributed mostly in cytoplasm
(C>N). When expressed in COS cells, the fusion protein
(GFP)2-PR1-7
is predominantly cytoplasmic (Fig. 6B). This result indicates that the
helices 1–7 domain does not itself serve as an independent NLS.
However, this conclusion must be tempered by the uncertainty that
helices 1–7 can fold properly in the absence of the rest of the
LBD.
|
). In
parallel, direct ligand binding assays were performed to examine the
suppositions made with regard to ligand specificity based on
ligand-mediated receptor redistribution. Ligand binding was assessed by
whole-cell binding assays after transient expression of receptors in
COS-1 cells (Fig. 7B). These studies confirm that the structural
features governing ligand binding and activation capacity of a receptor
are not coincident with those that determine its distribution in the
absence of hormone. Also, as detailed below, a number of
inferences can be made concerning the role of different domains in
ligand specificity and transactivation.
|
20% of
wild type), it is only minimally activated even at very high levels of
dexamethasone. Replacement with additional GR sequence,
G/P1-5G, still results in
a receptor with a largely nuclear distribution. However, this chimeric
receptor now exhibits a strong preference for binding to and activation
by glucocorticoids. These data concur with the implications of the
ligand-mediated redistribution data that the ß-turn helices 6–7
region possesses key determinants for steroid binding specificity. Additional chimeras confirm the key role of this region. Chimera G/P1-3G binds glucocorticoids similarly to GR and is transcriptionally activated by dexamethasone. When the ß-turn-helix 6-helix 7 region is changed to PR sequence (G/P1-3GP6-7G), the resulting double chimera does not bind to glucocorticoids nor respond to them. Instead, this chimera displays significant binding to R5020 (albeit reduced from wild-type PR) and a proportionate transactivation response. When only helix 7 instead of the ß-turn-helix 6-helix 7 region is changed to PR sequence in another double chimera, G/P1-3GP7G, the receptors can still relocalize in response to high levels of R5020 although neither R5020 nor dexamethasone binding or transactivation is observed. The observation that this double chimeric receptor can redistribute to the nucleus in response to ligand but fail to transactivate will be addressed in the Discussion.
Along with the ß-turn-helices 6–7 region, helix 3, which forms part of the ligand binding pocket (32, 33), contributes to ligand binding and specificity. G/GP3-12 binds well to progestins and undetectably to glucocorticoids. Replacing helix 3 with the homologous GR sequence (P/GP4-12) abrogates R5020 binding. Even changing only four amino acids at the N terminus of helix 3 to make G/GPspe-12 reduces R5020 binding 4-fold compared with G/GP3-12. Thus, together these chimeras indicate that the best R5020 binding is seen when the entire helix 3 through helix 7 region is derived from PR. Interestingly, the origin of helix 3 is of lesser consequence to glucocorticoid binding. For example, G/P1-3G binds dexamethasone well. However, now changing the helices 6–7 domain to PR sequence (G/P1-3GP6-7G) abrogates glucocorticoid binding. This binding can be partially recovered by changing part of helix 3 to GR sequence (G/P1-speGP6-7G) as can a proportionate degree of transactivation. Thus, while the helix 3 sequences may be less important in glucocorticoid specificity than progestin specificity, they do contribute in the context of additional domains of the LBD. Together, these data indicate that sequences governing ligand specificity overlap, but are not coincident with, the functional domains that govern differential localization of the two receptors.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Among the steroid receptors, there is general agreement that hormone-free PR and the estrogen receptor are nuclear (13, 40). Recent studies employed GFP-tagged receptors to compare the localization of the two isoforms of PR (31). In this work, the distribution of the two isoforms differed somewhat, with GFP-PR-B on average showing slightly more nuclear than cytoplasmic fluorescence and GFP-PR-A predominantly nuclear staining. We also find some GFP-PR-B transfected cells with detectable cytoplasmic fluorescence, although most of the cells exhibit predominantly or entirely nuclear fluorescence. In contrast to PR, GR and the mineralocorticoid receptor are cytoplasmic (9, 10, 11, 12, 41, 42). Both nuclear and cytoplasmic localization of the androgen receptor have been reported (43, 44, 45, 46, 47). In this work we have mapped the sequence domains of GR and PR that control the differential distribution of the two receptors in a ligand-free state. Although the sequences of the GR and PR that render the receptor capable of nuclear localization upon hormone addition have been mapped to two clusters of basic amino acids in the DBD and another in the hinge (14, 15, 16, 17, 18), previous work in which the LBD of the PR was replaced with that of GR implicated the LBD in determining the differential distribution of the two receptors in the absence of hormone (14). We have extended this work using receptor chimeras to delineate the sequences that control the differential distribution. Our approach has employed an enhanced GFP tag to follow the receptor in contrast to most previous studies on receptor localization and trafficking that have used indirect immunofluorescence. A green fluorescent protein tag permits the visualization of receptor proteins in living, unfixed cells and therefore represents a valuable alternative approach to studies on receptor trafficking. In addition, due to the similarity of the LBDs of GR and PR, receptor chimeras used in our study maintain the receptor integrity and overall conformation compared with receptor deletion mutants, as evidenced by the ability to bind hormone and, in most cases, to transactivate in response. This is not unexpected. The crystal structure of the LBD of hPR bound to progesterone has been solved to a 1.8 Å resolution (32, 33). Despite amino acid identity as low as 15%, PR exhibits an overall structure very similar to other members of the nuclear receptor family for whom the crystal structure is available.
Notwithstanding the overall conformational similarity and extensive sequence identity of the LBD, GR and PR exhibit distinct ligand binding properties and distinct subcellular distribution patterns in the ligand-free state. The examination of many chimeric LBDs has shown that determinants of ligand binding specificity and differential localization are separable. The region of helices 1–5 determines subcellular localization specificity, while helix 7 and the larger ß-turn-helix 6-helix 7 domain along with helix 3 make important contributions to ligand recognition. Therefore, a glucocorticoid binding chimeric receptor could be nuclear, and conversely a progestin binding chimeric receptor could be cytoplasmic in the ligand-free state.
Interestingly, a subset of chimeras (the double chimeras, Figs. 5 and 7) could redistribute from cytoplasm to nucleus upon addition of
ligand, yet they exhibited greatly reduced or ablated transactivation
capability. GR is driven to the nucleus by ligand relatively quickly
(t1/2 5 min) yet GR that has been withdrawn from
hormone redistributes into the cytoplasm only slowly
(t1/2 4 h) even though nucleocytoplasmic
shuttling continues and reestablishment of an 8S complex has occurred
(25, 48). In light of these kinetics, it stands to reason that the
addition of hormone to a chimeric receptor whose structure reduces
hormone affinity because of a high off rate may cause some nuclear
accumulation even though the receptors are only occupied a small
fraction of the time. The chimeric receptor may not be able to assume a
proper conformation to bind coactivators and thus fail to promote
transactivation. Alternatively, a high ligand off rate may not
allow sufficient time for the completion of the full set of
receptor-promoted steps necessary to initiate a round of transcription
initiation before the entire process is aborted by ligand
dissociation.
The examination of the GFP-tagged receptor chimeras has narrowed the domain that determines the differential distribution of unoccupied GR and PR to the N-terminal 87 amino acids of the LBD (helices 1–5) and suggest that multiple determinants may be involved in specifying nuclear localization. This region of GR and/or PR has been implicated in binding of chaperone proteins and in ligand binding. The present work implies that this region may have yet another role in interacting with nuclear trafficking proteins to determine kinetics of the import and export steps and therefore the equilibrium distribution of the protein. The difference in nucleocytoplasmic trafficking between ligand-free GR and PR suggests that the helices 1–5 region of the GR LBD may have a distinct conformation compared with PR. As mentioned earlier, GR that has been withdrawn from hormone redistributes into the cytoplasm only slowly even though nucleocytoplasmic shuttling continues and reestablishment of an 8S complex has occurred (25, 48). It may be that this hormone-withdrawn state represents an LBD conformation that is more PR-like and that reverts slowly to a conformation that is more typical of the naive receptor. Binding of the antiprogestin RU486 to GR appears to entrain a conformation that favors nuclear localization as GR withdrawn from RU486 fails to redistribute to the cytoplasm (Refs. 25, 48 and P. R. Housley, unpublished). Data with RU486 and with the pure antiestrogen ICI182 780 (49) suggest that disruption of the normal nucleocytoplasmic trafficking of receptors may play a role in the mechanism of action of steroid antagonists. The widespread clinical use of steroid antagonists makes this an important question to understand more fully.
The functional consequences of the differential localization of PR and GR are not clear. However, there are indications that the localization of PR may be regulated in a developmental or tissue-specific fashion (50). This regulation suggests that there are indeed biological consequences of receptor localization. Preliminary findings that suggest a molecular target have come from yeast two-hybrid studies. These studies indicate that PR has a strong SH3 binding domain and can physically interact in vitro with cell signaling molecules, such as the tyrosine kinase src, and regulate activity (51). Since these cell-signaling molecules are present predominantly in compartments other than the nucleus, such findings highlight the importance of a thorough understanding of steroid receptor trafficking.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
was constructed by cloning a
BamHI-XhoI fragment from pI9 (52) into the GFP-C2
vector (CLONTECH Laboratories, Inc., Palo Alto, CA).
GFP-hPR was constructed by cloning an EcoRI-ScaI
fragment from hPR1 (35) into the GFP-C1 vector (CLONTECH Laboratories, Inc.). Receptor chimeras G/P, P/G,
P/P1-9G,
P/P1-7G,
G/P1-3G, and
P/GP4-12 were constructed
by cloning fragments of hGR amplified by PCR using Pfu polymerase
(Stratagene, La Jolla, CA). Appropriate restriction sites
were incorporated into the primers to facilitate the in-frame junction
with naturally occurring restriction sites within hPR (see Fig. 3A).
The BstBI site in the PR LBD immediately follows the coding
region encompassing LBD helix 3, HindIII follows helix 7,
and Ecl136 II cuts near the C terminus of helix 9. All PCR-generated
sequences were confirmed by sequencing. Receptor chimeras
G/P1-9G and
G/P1-7G were constructed
by replacing the LBD in G/P with the chimeric LBD from
P/P1-9G and
P/P1-7G, respectively,
utilizing the BclI site at the junction between the hinge
and the LBD. SpeI sites were introduced into the LBD of
GFP-GR and GFP-PR-B by Quick-Change site-directed mutagenesis
(Stratagene), and the products were confirmed by DNA
sequencing. Receptor chimeras P/P1-speG and
G/GPspe-12 were
constructed by using these SpeI sites. Chimera
G/P1-speG was constructed by replacing the LBD in
G/P with the LBD fragment from P/P1-speG at
the BclI site. Chimera
G/GP3-12 was constructed
from G/GPspe-12 by
Quick-Change site-directed mutagenesis (Stratagene),
changing the sequences encoding the four amino acids immediately
N-terminal of the introduced SpeI site from GR-like (WRIM)
to PR-like (SSLL). The changes were confirmed by DNA sequencing. All
the double chimeric receptors were constructed using the gene Splicing
by Overlapping Extension (SOE) procedure (53, 54).
G/P1-3GP6-7G
and
G/P1-3GP7G
were constructed using
G/P1-3G and
G/P1-7G as progenitors.
G/P1-speGP6-7G
and G/P1-speGP7G were
constructed using G/P1-speG and
G/P1-7G as progenitors.
G/P1-5G was constructed
using G/P1-7G and GFP-GR
as progenitors. Additional details on plasmid constructions,
including the oligonucleotides used for PCR and for site-directed
mutagenesis, are available on request.
Cell Culture and Transfection
Mouse fibroblast E82.A3 and Ltk- cells
were maintained in MEM (Life Technologies, Inc.,
Gaithersburg, MD) supplemented with 5% FBS (HyClone Laboratories, Inc.), 10 mM HEPES, and nonessential
amino acids. Transient transfections of both lines were performed using
a diethylaminoethyl (DEAE)/Dextran method as previously described (55).
For fluorescence experiments, cells were plated on coverslips in
culture dishes and transfected with 5 µg/ml of receptor expression
vector. Cells were maintained in medium containing charcoal-stripped
serum before fluorescence imaging. Fluorescence was assessed 24–30 h
after transfection and 1 h after vehicle or hormone addition. For
quantitation of reporter gene expression, cells were transfected with 1
µg/ml of receptor expression vector. Cells were cotransfected with 2
µg/ml of pAHluc and 0.1 µg/ml of pCMVß-gal. The response of the
former was used to assess hormone response, and the latter served as an
internal transfection control. The promoter of pAHluc is an
AvaI-HpaII fragment spanning nearly the entire
MMTV long terminal repeat. Cells were treated with hormone for 20–24 h
beginning the second day after transfection. Extracts were prepared by
first washing the cells, harvesting them in 0.5 ml of cell lysis
buffer, and then pelleting debris (55). For luciferase assays, 25 µl
of soluble lysate were used and for ß-galactosidase assays, 2.5 µl
were used. Luciferase and ß-galactosidase assays were assessed using
a Monolight 3010 luminometer (Analytical Luminescence Laboratory, San Diego, CA) as previously described (55). Data
are reported as luciferase activity normalized to
ß-galactosidase activity in the same transfection.
COS-1 cells were maintained in DMEM (Life Technologies, Inc.) supplemented with 10% FBS (Life Technologies, Inc.). Transient transfection of COS cells was performed using a modified DEAE/Dextran method with 200 µg/ml DEAE/Dextran and 30 µM chloroquine. Cells were incubated in the DNA-DEAE/Dextran-chloroquine transfection mixture for 2 h at 37 C. The transfection mixture was aspirated and the cells subjected to a 1-min shock (55). Cells were then washed twice with PBS and refed with DMEM with charcoal-stripped serum. Hormone or vehicle were added 18–20 h after transfection and fluorescence was monitored 1 h thereafter.
Fluorescence Microscopy
Transfected cells on coverslips were analyzed by confocal
scanning laser microscopy with a MRC instrument (Bio-Rad Laboratories, Inc., Hercules, CA). Living cells in medium were
scanned at low laser power to avoid photobleaching and at a sufficient
depth to correctly assess the presence of intranuclear signal. The
figures show representative cells from each transfected DNA; at least
50–100 cells from each transfection were inspected and scored as
described in the legend to Fig. 3.
Whole-Cell Hormone Binding Assays
For hormone binding assays, 5 µg/ml of each construct was
transfected as described into COS-1 cells in six-well plates (R5020
assays) or a T150 flask (dexamethasone assay). For the latter, cells
were replated into six-well dishes 24 h after transfection so that
they would be 60–70% confluent at the time of harvest. Forty eight
hours after transfection, three wells were treated with
3H-labeled hormone for total hormone binding,
three wells were treated with 3H-labeled hormone
plus excess unlabeled hormone for nonspecific binding, and three wells
were left untreated for protein determination. For R5020 binding assay,
cells were treated with 1 nM
3H-R5020 ± 100 nM unlabeled
R5020. For dexamethasone binding assay, cells were treated with 20
nM 3H-dexamethasone ± 20
µM unlabeled dexamethasone. After incubation for 4 h
at 37 C, the cells were washed five times with cold PBS. Hormone was
then extracted with ethanol at room temperature for 30 min. Ethanol was
transferred into scintillation vials for quantitation of bound hormone.
To obtain specific binding (picomoles/mg protein) the following formula
was used: total binding - nonspecific binding - specific
binding from empty vector transfected cells. The latter term is
included to subtract the small amount of specific dexamethasone binding
seen in COS-1 cells (a few percent or less of that seen in transfected
cells).
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
This work has been supported by NIH Grants DK-37061 and DK-47951 to S.K.N. and to P.R.H., respectively.
1 In the naming system that will be followed for
all remaining GFP-tagged receptor chimeras, the first letter designates
the origin of the receptor sequences comprising the Nterminal
domain through the hinge. The slash denotes the hinge-LBD border. The
source of each segment of the LBD is indicated by P or G. For receptors
with chimeric LBDs, subscripts denote the helices or other sequence
features encompassed by the PR segments in the chimeric LBD. The
GR-derived segments can be inferred. In all cases GFP is present at the
N terminus of the protein. See schematics of each receptor chimera in
the figures for exact boundaries of the junctions and other
details. 
Received for publication July 7, 2000. Revision received October 3, 2000. Accepted for publication October 9, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  H. Becker, A. Sturm, J. E. Bron, K. Schirmer, and N. R. Bury The A/B Domain of the Teleost Glucocorticoid Receptors Influences Partial Nuclear Localization in the Absence of Hormone Endocrinology, September 1, 2008; 149(9): 4567 - 4576. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-H. Man, H.-Y. Li, P.-J. Zhang, T. Zhou, K. He, X. Pan, B. Liang, A.-L. Li, J. Zhao, W.-L. Gong, et al. PIAS3 induction of PRB sumoylation represses PRB transactivation by destabilizing its retention in the nucleus Nucleic Acids Res., November 14, 2006; 34(19): 5552 - 5566. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. M. Price, E. L. Hansen, and T. N. Oliver Immunofluorescent Localization of a Novel Progesterone Receptor(s) in a T47D-Y Breast Cancer Cell Line Lacking Genomic Progesterone Receptor Expression Reproductive Sciences, December 1, 2005; 12(8): 610 - 616. [Abstract] [PDF]
|
|
|
|
 |
|
 K. De Bosscher, W. Vanden Berghe, and G. Haegeman The Interplay between the Glucocorticoid Receptor and Nuclear Factor-{kappa}B or Activator Protein-1: Molecular Mechanisms for Gene Repression Endocr. Rev., August 1, 2003; 24(4): 488 - 522. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kawajiri, T. Ikuta, T. Suzuki, M. Kusaka, M. Muramatsu, K. Fujieda, M. Tachibana, and K.-i. Morohashi Role of the LXXLL-Motif and Activation Function 2 Domain in Subcellular Localization of Dax-1 (Dosage-Sensitive Sex Reversal-Adrenal Hypoplasia Congenita Critical Region on the X Chromosome, Gene 1) Mol. Endocrinol., June 1, 2003; 17(6): 994 - 1004. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Miura, R. Ouchida, N. Yoshikawa, K. Okamoto, Y. Makino, T. Nakamura, C. Morimoto, I. Makino, and H. Tanaka Functional Modulation of the Glucocorticoid Receptor and Suppression of NF-kappa B-dependent Transcription by Ursodeoxycholic Acid J. Biol. Chem., December 7, 2001; 276(50): 47371 - 47378. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Maccarrone, H. Valensise, M. Bari, N. Lazzarin, C. Romanini, and A. Finazzi-Agro Progesterone Up-Regulates Anandamide Hydrolase in Human Lymphocytes: Role of Cytokines and Implications for Fertility J. Immunol., June 15, 2001; 166(12): 7183 - 7189. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Minamino, V. Gaussin, F. J. DeMayo, and M. D. Schneider Inducible Gene Targeting in Postnatal Myocardium by Cardiac-Specific Expression of a Hormone-Activated Cre Fusion Protein Circ. Res., March 30, 2001; 88(6): 587 - 592. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
