Glucocorticoid receptor concentration and the ability to dimerize influence

nuclear translocation and distribution

Steven Robertson a, Janet P. Hapgood b, Ann Louw a,⇑

aDepartment of Biochemistry, University of Stellenbosch, Private Bag X1, Stellenbosch 7602, South AfricabDepartment of Molecular and Cell Biology, University of Cape Town, Private Bag X3, Cape Town 7701, South Africa

a r t i c l e i n f o

Article history:

Received 1 December 2011

Received in revised form 22 October 2012

Accepted 27 October 2012

Available online 20 November 2012

Keywords:

Glucocorticoid receptor levels

Dimerization deficient glucocorticoid

receptor

Observed nuclear import rate

Nuclear export rate

Nuclear foci

a b s t r a c t

Glucocorticoid receptor (GR) concentrations and the ability of the GR to dimerize are factors which influ-

ence sensitivity to glucocorticoids. Upon glucocorticoid binding, the GR is actively transported into the

nucleus, a crucial step in determining GR function. We examined the effects of GR concentration and

the ability to dimerize on GR nuclear import, export and nuclear distribution using both live cell micros-

copy of GFP-tagged GR and immunofluorescence of untagged GR, with both wild type GR (GRwt) and

dimerization deficient GR (GRdim). We found that the observed rate of GR nuclear import increases sig-

nificantly at higher GR concentrations, at saturating concentrations of dexamethasone (10�6 M) using

GFP-tagged GR, while with untagged GR it is only discernable at sub-saturating ligand concentrations

(10�10–10�9 M). Loss of dimerization results in a slower observed rate of nuclear import (2.5- to 3.3-fold

decrease for GFP-GRdim) as well as a decreased extent of GR nuclear localization (18–27% decrease for

untagged GRdim). These results were linked to an increased rate of GR export at low GR concentrations

(1.4- to 1.6-fold increase for untagged GR) and where GR dimerization is abrogated (1.5- to 1.7-fold

increase for GFP-GRdim). Furthermore, GR dimerization was shown to be required for the appearance

of discrete GC-dependent GR nuclear foci, the loss of which may explain the increased rate of GR export

for the GRdim. The reduction in the observed rate of nuclear import and increased rate of nuclear export

displayed at low GR concentrations and by the GRdim could explain the lowered glucocorticoid response

under these conditions.

� 2012 Elsevier Inc. All rights reserved.

1. Introduction

The glucocorticoid receptor (GR) mediates the effects of endog-

enous glucocorticoids (GCs) [37,49,73], as well as natural or syn-

thetic GCs used to treat inflammatory diseases [18,24,50]. The GR

is a ubiquitous ligand dependent transcription factor [50] and

essential for life [17,54]. In the absence of ligand the GR occurs

primarily in the cytoplasm in the form of a heteromeric complex

consisting of a heat shock protein (Hsp) 90 dimer, Hsp70, the small

acidic protein, p23, and one of the tertratricopeptide repeat (TPR)-

domain proteins [59]. Binding of a GC to the GR produces a confor-

mational change in the GR resulting in a change in the proteins

making up the heteromeric complex [4], GR dimerization [66]

and active import into the nucleus [26,72].

Nuclear import of the GR occurs quickly [39] and relies on the

association with Hsp90 [10,57], the TPR FK506-binding protein 52

(FKBP52) [23] and importin-a [77]. This complex is actively

shuttled into the nucleus by dynein [41] along the cytoskeleton

[35] through the nuclear pore complex [30,31,57]. Two nuclear

localization (NL) sequences have been identified in the human

GR, the NL1 sequence, which is situated within amino acids

479–506 [33,46,65], and the NL2 domain situated within amino

acids 526–777 [33,65]. The unliganded GR, although mostly

cytoplasmic, does exist in a dynamic equilibrium where a small

proportion of the population is actively shuttled into the nucleus

and allowed to diffuse back into the cytoplasm. Upon ligand activa-

tion this equilibrium shifts toward a predominantly import driven

state, which results in a primarily nuclear localization of the GR

[27,65]. Thus, the degree of nuclear localization reflects both the

rate of nuclear import as well as the rate of nuclear export [53].

There is clear evidence that the nuclear import rate is ligand

dependent [82] and that the degree of GR nuclear localization is a

critical factor in determining the level of GR function [40,42].

After ligand withdrawal the unliganded GR remains nuclear for

a considerable amount of time [85]. The retention of the GR in the

0039-128X/$ - see front matter � 2012 Elsevier Inc. All rights reserved.

http://dx.doi.org/10.1016/j.steroids.2012.10.016

Abbreviations: CpdA, compound A; DEX, dexamethasone; F, cortisol; FISH,

fluorescence in situ hybridization; GC, glucocorticoid; GR, glucocorticoid receptor;

GRE, glucocorticoid response element; MPA, medroxyprogesterone; NFjB, nuclearfactor-jB; Prog, progesterone; RU486, mifepristone.⇑ Corresponding author. Tel.: +27 21 8085873; fax: +27 21 8085863.

E-mail address: al@sun.ac.za (A. Louw).

Steroids 78 (2013) 182–194

Contents lists available at SciVerse ScienceDirect

Steroids

journal homepage: www.elsevier .com/locate /s teroids

nucleus is linked to GR association with Hsp90 in the nucleus [76]

and relies on the nuclear retention signal found within the hinge

region of the GR [11]. GR dissociation from DNA following ligand

withdrawal occurs rapidly [61] and is followed by the subsequent

localization of the GR to transcriptionally inactive areas of the nu-

cleus [85], prior to export of the GR from the nucleus or degrada-

tion of the GR by the proteasome [47]. It has been demonstrated

that nuclear export of the GR is independent of the exportin 1/

CRM1-directed nuclear export pathway [47], is an inactive process,

which occurs independently of ATP [85], and relies on the nuclear

export signal within amino acids 442–456 of the human GR [7].

Considering the slow rate of nuclear export and the fact that ATP

is not required for export, it is most likely that nuclear export of

the GR occurs through passive diffusion [67,76].

The nuclear import of GR has a half-time (t½) of 4 to 5 min fol-

lowing 10�6 M DEX stimulation [39,82] and has been shown to be

cell type [55], ligand [87] and ligand concentration dependent [45].

In addition, previous research demonstrated that once imported,

nuclear mobility [39] and the pattern of GR distribution [68] in

the nucleus are differentially affected by ligands and ligand con-

centration. Induction with the potent GR agonist, DEX, results in

discrete nuclear foci, while induction with the GR antagonist,

RU486, leads to diffuse nuclear localization of the activated GR

[45,67,82]. RNA fluorescence in situ hybridization (FISH) studies

have demonstrated active transcription close to receptor nuclear

foci [74,81,83]. GR export from the nucleus following the washout

of 10�6 M cortisol (F) shows a t½ of 8–9 h [39] and is ligand [39] as

well as ligand concentration dependent [11]. Thus, although ligand

type and ligand concentration have been investigated, the influ-

ence of GR concentration or the ability of GR to dimerize on the

rate of GR nuclear import and export or nuclear distribution has

not been previously examined.

Physiologically, the concentration of expressed GR varies con-

siderably between tissues, ranging from 4.1 fmol GR per mg pro-

tein in PBMCs [15] to as high as 893 fmol GR per mg protein in

the skin [38]. Considerable inter-individual variation, within the

same tissue type, has also been reported, primarily in cancerous

tissues [19,51]. Variations in GR concentration influence the re-

sponse to GC treatment within the same tissue between individu-

als [44], as well as between different tissue types [52]. Clinical

observations of patients broadly reveal hypersensitivity to GCs

brought about by increased GR levels [44] or GC resistance at re-

duced GR levels [13,69].

The ligand bound GR may exist in equilibrium as either a mono-

mer or dimer, although ligand binding shifts the equilibrium to-

wards more dimer [29,71]. Two regions of the GR have been

identified as influential in GR dimerization, the dimerization loop

(D-loop) of the DNA binding domain (DBD) (amino acids 458–

462 in the human GR) [22] and the ligand binding domain (LBD)

[8]. Dimerization of the GR has been demonstrated in the cyto-

plasm following ligand binding in live cells [63,66] and through

glycerol gradient centrifugation of purified GR [83]. Heck et al.

[43] created a dimerization deficient human GR mutant, through

the exchange of alanine to threonine at amino acid position 458,

termed the GRdim. This GR mutant is widely used to elucidate

the relevance of GR dimerization in GC signaling and has been

shown to display low affinity binding of the receptor to DNA

[22,43]. Furthermore, studies reveal that the GRdim generally has

a reduced capacity for transactivation relative to the GRwt

[34,43,60]. Although numerous other dimerization reduced GR

mutants exist [1], the GRdim (hGRA458T) is the most widely char-

acterized and as a result it is the one we will focus on. A recently

characterized natural mutation in the C-terminal zinc finger of

the DBD of the GR, referred to as GRR477H, has been linked to pri-

mary cortisol resistance in patients [64]. This mutation is thought

to affect GR homodimerization and results in a prolonged nuclear

import time [12], an inability to bind directly to DNA [12] and a re-

duced transactivation efficacy of the GR [62].

Unlike the majority of GR agonists that induce GR dimerization,

induction of GR by the selective GR agonist, CpdA [25,84,86], re-

sults in the abrogation of GR dimerization [28,63]. CpdA is a selec-

tive agonist in that it does not transactivate via the GR but retains

the ability to repress via the GR to the same extent as a full agonist

[25]. Recent findings by our group indicate that the action of DEX

through the mouse GRdim is similar to that of CpdA through the

mouse GRwt in immunofluorescent nuclear import and nuclear ex-

port assays [63] and piqued our interest in the influence of dimer-

ization on nuclear translocation.

GR nuclear translocation and distribution are crucial factors in

the behavior of GR and are known to be influenced by ligand type

and concentration; however, it is not known whether GR concen-

tration and the ability to dimerize affect these parameters. In order

to address this we determined the influence of GR concentration

and the ability of GR to dimerize on GR nuclear import, distribution

and export, utilizing physiologically relevant and statistically

different concentrations of GRwt and GRdim in parallel with induc-

tion by the dimerization-inducing agonist, DEX, and the dimeriza-

tion-abrogating, selective GR agonist, CpdA. We evaluated nuclear

import and export of untagged-GR in immunofluorescent studies

as well as GFP-tagged GR in live cell assays and in addition inves-

tigated nuclear distribution of ligand activated GR.

2. Materials and methods

2.1. Reagents

Dexamethasone (11b,16a)-9-fluoro-11,17,21-trihydroxy-16-methylpregna-1,4-diene-3,20-dione) (DEX), cortisol (11b,17a,21-trihydroxypregn-4-ene-3,20-dione or 17-hydroxycorticosterone)

(F), progesterone (4-pregnene-3,20-dione) (Prog), medroxyproges-

terone (6a-methyl-17ahydroxyprogesterone acetate) (MPA),

mifepristone (11b-(4-dimethyl amino)phenyl-17b-hydroxy-17-

(1-propynyl)estra-4,9-dien-3-one) (RU486), cycloheximide,

DEAE–Dextran and chloroquine diphosphate salt (chloroquine)

were purchased from Sigma–Aldrich. Compound A (2(4-acetoxy-

phenyl)-2-chloro-N-methyl-ethylammonium chloride) (CpdA)

was synthesized as described previously [48]. The [3H]-DEX (spe-

cific activity of 68–85 Ci/mmol) was obtained from AEC Amersham

Biosciences.

2.2. Plasmids

The pGL2-basic (empty vector) was obtained from Promega. The

pRS-hGRa (GRwt) was a gift from R. M. Evans [87] and pHis-

GRA458T (GRdim) from K. De Bosscher (University of Ghent, Bel-

gium) [6]. The pEGFP-C2-GR (GFP-GRwt) was provided by S.

Okret (Karolinska Institute, Sweden) [79]. The pEGFP-C2-GRA458T

(GFP-GRdim) was cloned by excising the wild type GR from pEGFP-

C2-GR with the restriction enzymes XmaI and SalI and replacing it

with the mutated GRdim sequence from pHisGRA458T. The pres-

ence of the mutation was confirmed through sequencing (primer,

forward 50-AGC TTC AGG ATG TCA TTA TGG AG-30 and reverse 50-

CCC CCC CCG GGG TTT TGA TGA AAC AGA-30). All plasmids were

verified by restriction enzyme digest.

2.3. Cell culture and DEAE–dextran transfection

Monkey kidney fibroblast cells (COS-1) purchased from Ameri-

can Type Culture Collection (ATCC) were maintained in high glu-

cose (4.5 g/ml) Dulbecco’s modified Eagle’s medium (DMEM)

(Sigma) with 2 mM glutamine (Merck), 44 mM sodium bicarbonate

S. Robertson et al. / Steroids 78 (2013) 182–194 183

(Invitrogen), and 1 mM sodium pyruvate (Invitrogen) (unsuppli-

mented DMEM) supplemented with 10% fetal calf serum (FCS)

(Highveld Biologicals, South Africa), 100 IU/ml of penicillin, and

100 lg/ml of streptomycin (Pen/Strep) (Invitrogen) (complete

DMEM). All transfections were done using the DEAE–Dextran

method [2]. Briefly, cultured cells where plated to achieve a density

of 70–80% confluence on the target day of transfection. The trans-

fection mix consisted of 11550 ng DNA/10 cm plate added to pre-

heated unsupplimented DMEM medium (7.5 ml/10 cm plate),

along with 0.1 mM chloroquin and 0.1 mg/ml DEAE–Dextran and

was incubated on the cells at 37 �C for 2 h followed by a 4 min

10% DMSO in PBS shock at 37 �C. The cells were then rinsed with

PBS and finally 15 ml complete DMEM (1% Pen/Strep; 10% FCS)

was added.

2.4. Western blots

COS-1 cells (2 � 106 cells/10-cm plate) were DEAE–Dextran

transfected with 38.5, 385, or 11550 ng GRwt, GRdim, GFP-GRwt

or GFP-GRdim DNA and filled to 11550 ng total plasmid DNA/

10 cm plate with empty vector. Twenty-four hours after transfec-

tion cells were replated (4 � 105 cells/well in 12-well plates) and

steroid starved in medium with 10% dextran-coated charcoal

stripped FCS (Highveld Biologicals, South Africa) and 1% Pen/Strep

(stripped DMEM). Twenty-four hours after replating cells were

washed twice with PBS before being lysed on ice in Buffer A

(10 mM Hepes pH7.5 (Invitrogen), 1.5 mM MgCl2, 10 mM KCl,

0.1% Nonidet P-40 (Roche Applied Science), and Complete Mini

protease inhibitor mixture (Roche Applied Science). Protein con-

centrations were determined using the Bradford method and

20 lg of protein/sample was separated on a 10% SDS–PAGE gel.

Following electrophoresis, proteins were electroblotted and trans-

ferred to Hybond-ECL nitrocellulose membrane (Amersham Biosci-

ences), which were probed for GR with the H-300 antibody from

Santa Cruz Biotechnology diluted 1:3000 and visualized using

ECL peroxidase-labeled anti-rabbit antibody AEC-Amersham Bio-

sciences diluted 1:10000 and ECL Western blotting detection re-

agents (GE Healthcare) on Hyperfilm (Amersham Biosciences).

Densitometric analysis of the immunoblots was carried out using

UN-SCAN-IT gel 6.1 software (Silk Scientific).

2.5. Whole cell saturation binding

COS-1 cells (2 � 106 cells/10-cm plate) were DEAE–Dextran

transfected with GRwt (38.5, 385, or 11550 ng DNA), GRdim (385

or 11550 ng DNA) or GFP-GRwt (38.5, 385, or 11550 ng DNA)

and filled to 11550 ng total plasmid DNA/10 cm plate with empty

vector. Twenty-four hours after transfection cells were replated

(1 � 105 cells/well in 24-well plates) in stripped DMEM. Twenty-

four hours after replating cells were incubated for 4 h at 37 �C with

increasing concentrations of [3H]-DEX (total binding) or [3H]-DEX

and a constant concentration of 60 lM unlabeled DEX (non-spe-

cific binding) in unsupplemented DMEM. Cells were then placed

on ice and washed three times, for 15 min each, with ice-cold

PBS containing 0.2% (w/v) BSA. Cells were lysed with 100 ll of pas-sive lysis buffer (0.2% (v/v) triton, 10% (v/v) glycerol, 2.8% (v/v) 1 M

tris–phosphate–EDTA and 0.29% (v/v) 0.5 M EDTA) and binding

was determined by scintillation counting in a 1900CA TRI-CARB li-

quid scintillation analyzer (Packard) using FLO-SCINT II (Perkin El-

mer). Total binding and non-specific binding were normalized to

protein concentration (Bradford assay [9]). Specific binding (total

binding – nonspecific binding) was determined and fmol GR per

mg protein was calculated using specific activity, Bmax and a

counting efficiency of 43%.

2.6. Live cell nuclear import

COS-1 cells (2 � 106 cells/10-cm plate) were DEAE–Dextran

transfected with 38.5, 385, or 11550 ng GFP-GRwt or GFP-GRdim

DNA and filled to 11550 ng total plasmid DNA/10 cm plate with

empty vector. Twenty-four hours after transfection cells were re-

plated (3 � 104 cells/well) into 8-well Lab-Tek chambered cover-

glass plates (Nunc, Denmark) and steroid starved in stripped

DMEM. Twenty-four hours after replating cells were analyzed at

the Stellenbosch University’s central analytical facility imaging

unit in the temperature-controlled chamber (37 �C) of an Olympus

Cell system attached to an IX-81 inverted fluorescence microscope

equipped with a F-view-II cooled CCD camera and a 150 W Xenon

lamp as light source, which is part of the MT20 excitation source.

An Olympus Plan Apo N 60X/1.4 oil objective and the Cell� imaging

software were used for image acquisition and analysis. In order to

ensure linear range of the emitted signal of various expression lev-

els, the full dynamic range of the CCD camera was utilized and rou-

tinely monitored by controlling for minimal pixel saturation. The

software allows indication of total pixel saturation, which is then

adjusted by xenon light intensity output and exposure time prior

to any acquisition. It was routinely confirmed that, prior to sample

acquisition of various conditions, the boundary conditions of light

intensity and exposure time were set appropriately, to enable the

utilization of the full dynamic range. The online intensity histo-

gram of the intensity distribution was assessed prior to image

acquisition. The GFP filter set (U-MGFP/XL, Olympus) excites at

470 nm (BP 460–490) and emission is collected at 506 nm

(BA510IF). The GFP signal was used to select cells for analysis,

using the entire cellular area as the ROI. Cells with a GFP emission

of 0–600 where selected from the medium GR population, GFP sig-

nals between 600 and 1200 from the high GR population and GFP

signals of >1200 from the very high GR population.

Cells were induced with 10�6 MDEX in unsupplemented DMEM

and GFP images were taken every minute over a 60 min period.

Nuclear import was quantified as the increase in GFP fluorescence

in the nucleus (region of interest) over the period of stimulation.

Fluorescence in the nucleus at the zero time point was subtracted

from all time points and a one phase exponential association curve

was fit to the data. The generated half time (t½) represents the time

it takes to achieve 50% of maximal GFP nuclear accumulation.

2.7. Live cell nuclear export

COS-1 cells were DEAE–Dextran transfected, replated and ste-

roid starved as for the live cell nuclear import assay. Twenty-four

hours after replating cells were induced with 10�9 M DEX for 1 h

after which they were rinsed 4 times with sterile PBS containing

5% BSA at 37 �C and stripped DMEM was added. Nuclear export

was analyzed at time points between 0 and 36 h after DEX washout

in the temperature-controlled chamber (37 �C) of an IX-81 Olym-

pus Cell system using the same hardware and software as for the

live nuclear import assay. Cells chosen for analysis were also se-

lected using the same criteria as for live nuclear import. Nuclear

export was quantified as the ratio of GFP fluorescence in the mid-

point of the nucleus over that in the midpoint between nuclear

membrane and cellular membrane. Data was fit to a one phase

exponential decay curve which generates a t½ to maximal cyto-

plasmic localization.

2.8. Immunofluorescent analysis of nuclear import

COS-1 cells (2 � 106 cells/10 cm plate) were DEAE–Dextran

transfected with GRwt (38.5 or 385 ng DNA) or GRdim (385 or

11550 ng DNA) and filled to 11550 ng total plasmid DNA/10 cm

plate with empty vector. Cells were replated 24 h later onto cover-

184 S. Robertson et al. / Steroids 78 (2013) 182–194

slips in 6-well plates (3 � 105 cells /well) and steroid starved in

stripped DMEM. Twenty-four hours after replating cells were in-

duced with 10�6 M DEX or 10�5 M CpdA for 0 to 60 min. After

induction cells were fixed and permeabilized by being placed on

ice, rinsed with 1 ml of �20 �C methanol, and incubated at

�20 �C for 15 min with another 1 ml of �20 �C methanol. Cells

were then washed three times with ice-cold PBS plus 0.2% BSA

and transferred to new 6-well plates containing 2 ml of blocking

buffer (PBS with 3% (v/v) FCS and 1% (w/v) BSA). Cells were incu-

bated for 1 h at room temperature and then washed twice with

ice-cold PBS plus 0.2% BSA. To visualize GR, cells were incubated

with the primary rabbit anti-GR antibody, H-300 (diluted 1:1000

in blocking buffer), overnight. Cells were then washed three times

with ice-cold PBS plus 0.2% BSA and incubated for 1 h at room tem-

perature with the secondary antibody (Alexa Fluor 488-tagged

anti-rabbit antibody (Molecular Probes)) diluted 1:500 in blocking

buffer. Nuclei were visualized using Hoechst 33258 stain (Sigma)

according to the manufacturer’s instructions. Cells were then

washed three times with ice-cold PBS and mounted on glass slides.

Cells were analyzed on an IX-81 Olympus Cell system using the

same hardware and software as for the live nuclear import assay

in a double-blind fashion, the DAPI filter set was used to stimulate

and visualize the Hoechst stain. Cells were allocated as either nu-

clear (where there was clear nuclear localization (>60% of signal

in nucleus) of the signal) or cytoplasmic and the percentage nucle-

ar of 50 total cells per slide counted was fit to a one phase expo-

nential association curve which generated t½ to maximal nuclear

localization as well as maximal nuclear localization values.

2.9. Immunofluorescent analysis of nuclear export

COS-1 cells were transfected, replated and steroid starved as for

the immunofluorescent nuclear import assay. Twenty-four hours

after replating cells were induced with 10�6 M DEX or 10�5 M

CpdA for 1 h, rinsed three times with sterile PBS containing 5%

BSA at 37 �C and incubated for time points ranging from 0 to

28 h in stripped DMEM. At the end of each time point cells were

fixed, permeabilized, fluorescently labeled, Hoechst stained and

mounted as for the immunofluorescent import assay. Cells were

analyzed on an IX-81 Olympus Cell system using the same hard-

ware and software as the live nuclear import assay in a double-

blind fashion. Cells were allocated as either nuclear (where there

was >60% nuclear localization of the signal) or cytoplasmic and

the percentage nuclear of 50 total cells counted per slide was fit

to a one phase exponential decay curve which generated t½ to

maximal cytoplasmic localization values.

2.10. Nuclear distribution

COS-1 cells were DEAE–Dextran transfected as for the live cell

nuclear import assay. Cells were replated 24 h later onto coverslips

in 6-well plates (3 � 105 cells /well) and steroid starved in stripped

DMEM. Twenty-four hours after replating cells were induced with

10�6 M DEX or 10�5 M CpdA for 1 h. After induction cells were

fixed and permeabilized as described for the immunofluorescent

assay. Cells were then washed three times with ice-cold PBS plus

0.2% BSA and mounted on glass slides. Cells were analyzed on an

IX-81 Olympus Cell system using the same hardware and software

as for the live cell nuclear import assay. The GFP signal was used to

select cells which displayed clear nuclear GFP-GR distribution for

analysis, using the entire cellular area as the ROI. Cells with a

GFP emission of 0–600 where selected from the medium GR

population, GFP signals between 600 and 1200 from the high GR

population and GFP signals of >1200 from the very high GR popu-

lation. Z-stack images of the nuclei were taken at various focal

planes and used to deconvolute a single nuclear image. As long a

line as possible was drawn through each nucleus avoiding nucleoli

and the Cell� imaging software was used to quantify the coefficient

of variation (CV) of GFP fluorescence intensity along this line. A

lower CV indicates a more random nuclear distribution [68].

2.11. Statistical analysis

Statistical analyses were carried out using GraphPad Prism ver-

sion 5.00 for Windows (GraphPad Software, San Diego California

USA), using one way analysis of variance (ANOVA) with either

Dunnett or Newman–Keuls post-tests or two tailed unpaired t

tests. Statistical significance of differences is indicated in figure

legends.

3. Results

3.1. Establishing a physiologically relevant model in which to compare

the effects of glucocorticoid receptor concentrations and ability of the

GR to dimerize

We selected COS-1 cells, as they contain little to no endogenous

GR [36] and could act as a ‘‘blank slate’’ for our studies to elucidate

the effects of GR concentration and dimerization on nuclear trans-

location and distribution. COS-1 cells were transiently transfected

with varying amounts (38.5, 385, or 11550 ng) of untagged GR wild

type (GRwt), untagged D-loop dimerization domain mutant GR

(GRdim), green fluorescent protein tagged GRwt (GFP-GRwt) or

Fig. 1. Saturation binding establishes three distinct and statistically different

populations of GR. COS-1 cells were transiently transfected with 38.5, 385 or

11550 ng GRwt, GRdim, GFP-GRwt, or GFP-GRdim and filled to 11550 ng total

plasmid DNA/10 cm tissue culture plate with the empty vector pGL2-basic. (A)

Representative Western blots of COS-1 cells transfected with the indicated

quantities of GFP-GRwt or GFP-GRdim. (B) Representative Western blots of COS-1

cells transfected with the indicated quantities of untagged GRwt or GRdim. (C)

Summary table of expressed GR from saturation binding results, fmol GR per mg

protein values were derived from the maximal binding (Bmax) value as described in

materials and methods. Statistical analysis on fmol GR per mg protein was carried

out using one-way ANOVA followed by Newman–Keuls multiple comparison post-

test, where conditions with different letters are statistically different from one

another (P < 0.05) and identifies three statistically different GR populations

designated medium, high and very high for GFP-tagged GR and low, medium and

high for untagged GR. Pooled results are shown from a minimum of two

independent experiments performed in triplicate (±SEM).

S. Robertson et al. / Steroids 78 (2013) 182–194 185

GFP-GRdim plasmid. Western blots indicated that expression of

GFP-GRwt and GFP-GRdim was equivalent (Fig. 1A), while that of

GRwt and GRdim (Fig. 1B) differed, with similar amounts of DNA

transfected. The fact that the untagged GRdim expressed at a lower

level than the GRwt may be due to the fact that these constructs

are in different vectors.

To quantify GR levels and to establish whether distinct and sta-

tistically different populations of GR, which are physiologically rel-

evant, were attained, whole cell saturation binding studies of the

transiently transfected GR’s were performed. Three statistically

different GR concentrations were identified within the GFP-tagged

GR group, designated as medium, high, and very high and within

the untagged GR group, designated as low, medium and high

(Fig. 1C). The GR levels range between 67 and 569 fmol GR per

mg protein, which reflects a physiologically relevant range of GR

concentrations expressed in human tissues [5,16,38].

Whole cell saturation binding was also performed on the un-

tagged GRdim as Western blots (Fig. 1B) suggested that this con-

struct expressed at a lower level that the untagged GRwt. Results

indicate that expression of GRdim after transfection of 385 and

11550 ng DNA is equivalent to expression of GRwt after transfec-

tion of 38.5 and 385 ng, respectively (Fig. 1C). Although saturation

binding was not performed on the GFP-GRdim, preliminary bind-

ing studies at a single radio-labeled ligand concentration revealed

that the expression of this construct was statistically the same as

that of GFP-GRwt (results not shown) as supported by Western

blots (Fig. 1A).

As we wanted to explore both the effect of GR concentration

and the ability to dimerize in tandem we decided to use all three

GFP-tagged GR levels, designated medium, high, and very high, in

subsequent live cell studies and only two untagged GR levels, des-

ignated low and medium, in further immunofluorescent analyses

as we could obtain equivalent levels of the GRdim construct.

Cells were pooled and replated after transfection to minimize

differences in transfection efficiency within experiments and GR

levels were monitored in all subsequent experiments. Transfection

efficiency varied between 7% and 15% for individual experiments

and between 10% and 12% for individual conditions, with an aver-

age transfection efficiency of 11% for all conditions. We found that

the receptor levels estimated from Western blots (intensity in Pix-

els) correlated well (R2 = 0.974) with the receptor levels measured

using whole cell binding (fmol GR/mg protein). Experiments that

displayed aberrant transfection rates of GR (GR levels which fell

outside of their respective population concentration) were

excluded.

3.2. The observed import rate of GFP-GR is influenced by receptor

concentration as well as the ability to dimerize in live cell nuclear

import analysis

In order to concurrently elucidate the effects of GR concentra-

tion and ability to dimerize on ligand induced nuclear import of

GR in single cells we conducted live cell nuclear import studies

on COS-1 cells transfected with 38.5, 385 and 11550 ng GFP-GRwt

or GFP-GRdim DNA (see Fig. 1C for concentrations of GR ex-

pressed). Following induction by test compound, live cell images

of nuclear import were taken every minute over a 60 min period

(Fig. 2A). Nuclear import was quantified as the increase in GFP

fluorescence in the nucleus over the period of stimulation, taking

the zero time point as 0% and the maximal fluorescence as 100%.

The entire nuclear area was selected as the region of interest

(ROI) and is indicated by the interior of the white border in

Fig. 2A. A one phase exponential association curve was fit to this

data which generated a half time (t½) to maximal nuclear localiza-

tion as illustrated for DEX (Fig. 2B).

Increasing GFP-GRwt concentration resulted in a significant de-

crease in t½, indicating an increase in the observed import rate

(t½ = 0.693/kobs), following DEX stimulation (Fig. 2C, black bars).

Specifically, a 1.7- and 2.5-fold increase in the observed import rate

relative to 38.5 ng was observed when 385 and 11550 ng GFP-

GRwt, respectively, was transfected. This significant increase in

the observed import rate (ranging from 1.3- to 1.9-fold for 385

and 11550 ng GFP-GRdim, respectively) was also seen for DEX

induction through the dimerization impaired D-loop mutant GFP-

GRdim (Fig. 2C, gray bars) suggesting that GR concentration also

influences its rate of nuclear import. Furthermore, we show that

the observed nuclear import rate increases significantly as GFP-

GRwt concentration increases, not only for DEX, but also for the

natural ligand cortisol (F), medroxyprogesterone acetate (MPA),

progesterone (Prog), and RU486, a GR antagonist (Fig. 2D). With

the exception of RU486, the rank order for ligand-induced ob-

served import rates at 38.5 ng GFP-GRwt concentration (DEX >

F > RU486 > MPA > Prog) is the same as the rank order for efficacy

for transactivation on the endogenous GRE-containing GILZ gene

[77], suggesting that for most ligands, import rate is a reliable indi-

cator of transactivation efficacy on endogenous genes.

The relevance of GR dimerization for nuclear import is clearly

demonstrated in that GFP-GRwt showed a significant increase in

the observed nuclear import rate when compared to GFP-GRdim,

at all GR concentrations (Fig. 2C). Specifically, at 38.5 ng GFP-GR

a 2.5-fold decrease in the observed import rate is seen with GFP-

GRdim relative to GFP-GRwt, while at 385 and 11550 ng the de-

crease observed is 3.2- and 3.3-fold, respectively.

We also tested nuclear import following stimulation with a

non-saturating DEX concentration (10�9 M) at 11550 ng GFP-

GRwt, which displayed a significant reduction in observed nuclear

import rate with a t1/2 of 11 min (results not shown), compared to

the 4 min following stimulation with 10�6 M DEX. In order to en-

sure a plateau in the level of nuclear import after 60 min and to

minimize the differences in receptor occupation due to differing li-

gand affinities we, however, decided to conduct most of our further

live cell nuclear import studies at saturating ligand concentrations.

3.3. The ability to dimerize influences maximal nuclear localization

while GR concentration affects the t½ of nuclear import in

immunofluorescent nuclear import assays

In order to compare our nuclear import results obtained with

GFP-GR in live cell nuclear import with that of untagged-GR,

COS-1 cells expressing low or medium concentrations (see

Fig. 1C for ng GR transfected) of GRwt or GRdim were induced with

the potent GR agonist, DEX (10�6 M), or the selective GR agonist,

CpdA (10�5 M), which has been shown to abrogate GR dimerization

[28,63]. Cells were classified as nuclear when they displayed pre-

dominantly nuclear localization of the GR, in other words where

GR concentration was clearly higher (>60% nuclear) in the nucleus

than that in the cytoplasm. The percentage nuclear cells was calcu-

lated as the ratio of cells expressing predominantly (>60%) nuclear

GR over total cells counted. Although, this method of classification

is commonly used [87], it is important to note that what is mea-

sured is in fact not absolute nuclear GR localization as for live cell

nuclear import, but predominantly (>60% nuclear) nuclear distri-

bution. A further difference between the two methods entails eval-

uation of maximal nuclear localization. As discussed above, for live

cell nuclear import maximal GFP fluorescence in the nucleus was

set as 100%, which implies that maximal import is always 100% un-

der all conditions. In contrast, for immunofluorescent nuclear im-

port assays maximal nuclear import (cells expressing >60%

nuclear GR/total cell counted) could theoretically differ.

A graphical representation of a full time course for the untagged

GRwt or GRdim at low concentration and induced with either DEX

186 S. Robertson et al. / Steroids 78 (2013) 182–194

(10�6 M) or CpdA (10�5 M) is presented in Fig. 3A. The percentage

of cells displaying maximal nuclear localization following DEX

induction is not significantly affected by GR concentration, how-

ever, it is significantly higher for GRwt (�95%) than for the GRdim

(�76.5%) (Fig. 3A and B). Furthermore, as induction with the

dimerization-abrogating selective GR agonist, CpdA, whether

through GRwt or GRdim, also resulted in a decrease in maximal im-

port (Fig. 3A and B), dimerization state clearly influences maximal

nuclear import. Both the study by Robertson et al. [63], with mouse

GR, as well as these results (Fig. 3A and B), with human GR, dem-

onstrate similar behavior for DEX through the dimerization im-

paired GRdim as for CpdA through the GRwt. We therefore

hypothesize that the dimerization-abrogating CpdA results in a

similar conformation of the GRwt as exists for the dimerization im-

paired GRdim mutant following DEX stimulation. Thus, the ability

of GR to dimerize, although not an absolute requirement for nucle-

ar import, does play a role in the extent of GR nuclear localization.

Although GR concentration did not significantly affect the t½ of

GR nuclear localization at saturating concentrations of DEX

(10�6 M) (Fig. 3C), at subsaturating DEX concentrations (10�9 or

10�10 M) GRwt concentration has a significant effect in increasing

the observed rate (Fig. 3C), but not the extent (results not shown),

of nuclear import. Specifically, at 10�9 M DEX a 2-fold decrease in

t½ is observed, while at 10�10M DEX the decrease is 2.9-fold. Fur-

thermore, whereas the observed nuclear import rate of the med-

ium concentration of GRwt following stimulation with 10�6, 10�9

or 10�10 M DEX, remains statistically similar, that of the low con-

centration of GRwt decreases significantly as stimulating DEX con-

centrations decreases (Fig. 3C).

3.4. The t½ of live cell nuclear export of GFP-GR is dimerization

dependent

The phenomena of an increase in observed nuclear import rate

at increased GR levels may be ascribed solely to the laws of mass

action namely, a faster reaction rate at increased concentration of

GR. However, as the observed import rate (kobs) comprises both

the true rate of nuclear import (kin) and the rate of nuclear export

(kout) [53], the observed rate of nuclear import may be affected by

the rate of export (kin = kobs � kout/[ligand]). Thus, a possible cause

of the decrease in the observed nuclear import rate through GRdim

(Figs. 2C and 3A) and the incomplete nuclear localization at both

low and medium GRdim concentrations (Fig. 3B) may thus be fas-

ter nuclear export. We therefore followed up our import studies

with an in depth analysis of the influence of GR concentration

and ability to dimerize on nuclear export.

In order to represent nuclear export of the GR in terms of both

the level of GFP-GR diffusion out of the nucleus and its concomi-

tant accumulation in the cytoplasm in live cell nuclear export as-

says we quantified nuclear export as the ratio of GFP in the

center of the nucleus divided by that in the mid-point of the cyto-

plasm (Fig. 4A). Initial studies on the GFP-GRwt following induc-

tion and washout of 10�6 M DEX revealed protracted nuclear

export rates with a t½ �20 h (results not shown). In order to visu-

0

10

20

30

GFP-GRdim

10

64

25§§

19

§§§

13

GFP-GRwt

38.5ng 385ng 11550ng

******

***

t½ t

o m

axim

al

nu

cle

ar

locali

zati

on

(m

in)0

510

60

Tim

e a

fter

DE

X a

dd

itio

n (

min

)A B C

D

0 10 20 30 40 50 60

0

50

100

38.5ng GFP-GRwt

385ng GFP-GRwt

11550ng GFP-GRwt

Time (min)

% N

uc

lea

r tr

an

slo

ca

tio

n

0

10

20

30

40

50

60

7038.5ng GFP-GRwt 385ng GFP-GRwt 11550ng GFP-GRwt

DEX F Prog RU486MPA

A**

a

***

A***

a,b

***

B***

b

***

c

*

a,b

**aa

b

b

a

D

C

t½ t

o m

axim

al

nu

cle

ar

locali

zati

on

(m

in)

Fig. 2. Higher GR concentrations and the ability to dimerize decrease live cell nuclear import t½. Live cell nuclear import studies were carried out as described in materials

and methods. (A) Represents a single cell expressing 38.5 ng of GFP-GRwt induced with 10�6 M DEX. GFP images were taken every minute over a 60 min period. The white

circle around the nucleus represents the region of interest (ROI). (B) Representative graph depicting the time course of nuclear import following 10�6 M DEX induction with

38.5, 385 and 11550 ng GFP-GRwt concentrations. (C) Half time (t½) to maximal nuclear localization of 10�6 M DEX stimulated cells expressing 38.5, 385, and 11550 ng GFP-

GRwt or GFP-GRdim. Statistical analysis comparing t½ to maximal nuclear localization of GFP-GRwt to GFP-GRdim was through two tailed unpaired t tests (⁄⁄⁄P < 0.001). One-

way ANOVA followed by Dunnett’s Multiple Comparison post test was used to compare within GFP-GRwt concentrations (��P < 0.01 and ���P < 0.001) and within GFP-GRdim

concentrations (§§P < 0.01 and §§§P < 0.001). (D) Half time (t½) to maximal nuclear localization of 10�6 M DEX, F, MPA, Prog or RU486 stimulated cells expressing 38.5, 385, or

11550 ng GFP-GRwt. Statistical analysis of t½ to maximal nuclear localization for each ligand comparing 38.5 to 385 and 11550 ng GFP-GRwt concentrations was carried out

using one-way ANOVA followed by Dunnett’s Multiple Comparison post test (⁄P < 0.05, ⁄⁄P < 0.01, ⁄⁄⁄P < 0.001). One-way ANOVA followed by Newman–Keuls post test was

used to compare ligands within the 38.5 ng GFP-GR concentration (lower case letters), the 385 ng GFP-GR concentration (capital letters), or the 11550 ng GFP-GR

concentration (lower case italics letters) populations. Conditions with different letters are statistically different from one another (P < 0.05). Pooled results from a minimum of

five cells, each from an independent experiment (±SEM), are shown in B–D.

S. Robertson et al. / Steroids 78 (2013) 182–194 187

alize complete nuclear export of the GR from the nucleus we thus

stimulated with 10�9 M DEX where the average t½ of nuclear

export was reduced to�5 h (Fig. 4B and C). These findings do, how-

ever, indicate that nuclear export is ligand concentration depen-

dent. As the live cell nuclear export was run over a 36 h period,

there was concern that production of newly synthesized GFP-GR

may influence the measured rate of nuclear export. We therefore

performed an experiment where cycloheximide, an inhibitor of

protein bio-synthesis [56], was incubated with the cells during

the assay. We found no significant difference between the export

rate of 38.5 ng concentrations of GFP-GRwt or GFP-GRdim with

or without cycloheximide (results not shown).

There was no statistical difference in nuclear export rate

(t½ = 0.693/kout), between the three concentrations of GFP-GRwt

or GFP-GRdim (Fig. 4C). Nuclear export is, however, significantly

slower through the GFP-GRwt than through the GFP-GRdim, at all

GR concentrations (Fig. 4C). Specifically, nuclear export is decreased

between 1.6- and 2-fold at 11550 and 38.5 ng, respectively.

3.5. GR concentration and the ability to dimerize influence GR export

rate following the washout of DEX in immunofluorescent assays

Immunofluorescent analysis of nuclear export was conducted

on COS-1 cells expressing low or medium concentrations of GRwt

or GRdim following the washout of either 10�6 M DEX or 10�5 M

CpdA. The export t½ value reflects the half time to less than 60% nu-

clear GR localization (Fig. 5A). The nuclear export of GRwt, as well

as GRdim, following DEX stimulation and washout was signifi-

cantly slower at the medium GR concentrations as compared to

the low GR concentration (Fig. 5B). Furthermore, nuclear export

of GR following DEX stimulation and washout is significantly

slower through GRwt than through GRdim at the medium GR con-

centration, displaying a similar, but not significant, trend at the low

GR concentration (Fig. 5B). These results suggest that GR is ex-

ported faster at low GR concentrations and that the ability to

dimerize enhances GR nuclear retention and thus results in a

slower export rate. CpdA, which abolishes GR dimerization [63], re-

sults in no significant differences in export rate between low and

medium GR concentrations or between GRwt and GRdim

(Fig. 5B), suggesting that the loss of dimerization effected by CpdA

attenuates the effect of GR concentration. The nuclear export rate

of the low and medium GRwt concentrations following CpdA

washout is similar and not significantly (P > 0.05) different from

the export rate of DEX washout at low GRwt concentration.

Nuclear export studies offer a possible explanation for the de-

crease in observed nuclear import rate (Figs. 2C and 3A) and lower

maximal localization levels (Fig. 3B) of GRdim and GFP-GRdim

when compared to that of GRwt and GFP-GRwt, as there is a signif-

icant trend towards faster nuclear export of the GRdim and GFP-

GRdim (Figs. 4C, 5B). This suggests that nuclear retention of GR

may be affected by the ability to dimerize. An increase in nuclear

export rate is associated with a reduced ability to bind to Hsp90

in the nucleus [67,76], which is reflected by a diffuse pattern of nu-

clear distribution [76]. We thus conclude our experimental work

with a study designed to determine whether GR concentration

and ability to dimerize influence nuclear distribution.

3.6. Nuclear distribution of the GR is dimerization dependent

We based our study of nuclear distribution of GFP-GRwt and

GFP-GRdim on those performed by Schaaf et al. [68] who quanti-

fied nuclear distribution in terms of the variation in fluorescent

intensity along a line drawn through the nucleus. The resulting

coefficient of variation (CV) along this nuclear line represents

0 10 20 30 40 50 60

0

50

100

hGRwt-DEX (t½ = 3.2 min)

hGRwt-CpdA (t½ = 4.0 min)

hGRdim-DEX (t½ = 4.9 min)

hGRdim-CpdA (t½ = 4.8 min)

Time

Nu

cle

ar

Imp

ort

(>60%

Nu

cle

ar)

0

10

20

30

40 Low [GRwt]

Medium [GRwt]

3.2 3.2

18.8

9.4

35.2

12.2

10-6M 10-9

M 10-10

M

DEX

a a

a

c

a,b

b

t ½ t

o m

axim

al n

ucle

ar

localizati

on

(m

in)

0

20

40

60

80

100

GRwt

Low [GR] Medium [GR] Medium [GR]Low [GR]

DEX CpdA

GRdim

******** ******

*94%

76%

96%

77%71% 70% 69% 71%

% N

uc

lea

r im

po

rt

A

C

B

Fig. 3. GR dimerization and concentration affects the extent and t½ of nuclear import in immunofluorescent study. Immunofluorescent analysis of nuclear import was

performed as described in materials and methods. (A) Representative graph of complete time course from cells expressing low GRwt or GRdim concentrations after

stimulation with 10�6 M DEX or 10�5 M CpdA. (B) Percentage maximal (i.e.>60%) nuclear localization from cells expressing low or medium concentrations of GRwt or GRdim

after stimulation with 10�6 M DEX or 10�5M CpdA. Statistical analysis was carried out using one-way ANOVA followed by Dunnett’s post test against DEX stimulated low

GRwt concentration (⁄P < 0.05, ⁄⁄P < 0.01, ⁄⁄⁄P < 0.001). (C) Half time (t½) to maximal nuclear localization of cells expressing low or medium GRwt stimulated by 10�6 M,

10�9M or 10�10M DEX. Statistical analysis was carried out using one-way ANOVA followed by Newman–Keuls post test. Conditions with different letters are statistically

different from one another (P < 0.01). (A–C) Show pooled results from three independent experiments (±SEM), where 50 cells were counted for each condition and time point.

188 S. Robertson et al. / Steroids 78 (2013) 182–194

the distribution of fluorescently labeled GR (Fig. 6A and B). A high

CV signifies a non-random nuclear distribution where discrete

foci or speckles of fluorescence are visible. A low CV indicates ran-

dom nuclear distribution typified by a diffuse distribution of

fluorescence.

The pattern of GFP-GRwt nuclear distribution following DEX

induction was not influenced by receptor concentration (Fig. 6C).

Our results demonstrate discrete nuclear foci and a CV value of

�18% following 10�6 M DEX stimulation (Fig. 6A and C), while

stimulation with CpdA results in random nuclear distribution of

the GR, which is reflected in a significantly lower percentage CV

than DEX (Fig. 6A and C). This suggests that the loss in dimerization

induced by CpdA stimulation at both 38.5 and 385 ng GFP-GRwt

concentrations results in random nuclear distribution of the GR.

In support of this theory the nuclear distribution of DEX stimulated

38.5 ng GFP-GRdim is significantly more diffuse and random than

that of the same concentration of GFP-GRwt, but similar to that ob-

tained with CpdA via the GRwt (Fig. 6C).

04

12

20

36

Tim

e a

fter

DE

X w

ash

ou

t (h

ou

rs)

A B

C GFP-GRwt GFP-GRdim

0 4 8 12 16 20 24 28 32 360

1

2

3

4

5

Time (hour)

nu

cle

ar/

cy

top

las

mic

flo

ure

cs

en

ce

0

2

4

6

8

4.9

2.5

4.8

2.7

4.6

2.8

38.5ng 385ng 11550ng

* ** *

t½ t

o m

ax

ima

l

cy

top

las

mic

lo

ca

liza

tio

n (

ho

urs

)

38.5ng GFP-GRwt (t½ = 4.9h)

38.5ng GFP-GRdim (t½ = 2.5h)

Fig. 4. GR dimerization decreases the rate of live cell nuclear export. Live cell analysis of nuclear export was performed as described in materials and methods. (A) A single

cell expressing a high GFP-GRwt concentration was induced with 10�9 M DEX for 1 h after which it was rinsed and nuclear export analyzed at 0, 2, 4, 12, 20 and 36 h after

washout. The small white circles represent the ROIs, situated at the mid-point of the nucleus and the mid-point between nuclear membrane and cellular membrane. (B) A

representative graph depicting the time course of nuclear export in cells expressing 38.5 ng GFP-GRwt or GFP-GRdim quantified as the ratio of GFP fluorescence in the nucleus

over that in the cytoplasm. A one phase exponential decay curve was fit to the data which generated a t½ to maximal cytoplasmic localization. (C) Half time (t½) to maximal

cytoplasmic localization of 38.5, 385, or 11550 ng GFP-GRwt or GFP-GRdim exposed to 10�9 M DEX. Statistical analysis was carried out using one tailed unpaired t test

(⁄P < 0.05, ⁄⁄P < 0.01). B and C show pooled data from a minimum of six individual cells from independent experiments (±SEM).

Medium [GR]

0

5

10

15

20

25**

Low [GR] Medium [GR]

10.9

13.3

21.4

15.8

ns

*

*

DEX

GRdimGRwt

12.511.5

8.9

12.6ns

ns

ns

ns

t ½ t

o m

ax

ima

l c

yto

pla

mic

loc

ali

za

tio

n (

ho

urs

)

Low [GR]

CpdA

A B

0 4 8 12 16 20 24 28

0

20

40

60

80

100

Low [GRwt], DEX

(t½ = 13.3h)

Medium [GRwt], DEX

(t½ = 21.4h)

Time (hours)

Nu

cle

ar

locali

zati

on

(>

60%

Nu

cle

ar)

Fig. 5. Increasing GR concentration and the ability to dimerize slows nuclear export as revealed by immunofluorescence. Immunofluorescence analysis of nuclear export was

performed as described in materials and methods. Cells expressing low and medium concentrations of GRwt or GRdim were stimulated with 10�6 M DEX or 10�5 M CpdA for

1 h, rinsed and incubated for 0, 4, 8, 12, 20, 24 and 28 h. The percentage cells displaying nuclear GR relative to the total number of GR containing cells counted was fit to a one

phase exponential decay curve. (A) Representative graph of the time course of nuclear export from cells with low and medium GRwt concentrations exposed to DEX, which

generates a t½ to maximal cytoplasmic localization. (B) Half time (t½) to maximal cytoplasmic localization of low or medium GRwt or GRdim exposed to 10�6 M DEX or

10�5 M CpdA. Statistical analysis was carried out on t½ to maximal cytoplasmic localization using two tailed unpaired t tests (⁄P < 0.05, ⁄⁄P < 0.01). Results show pooled data

from three independent experiments (±SEM), each performed on 50 cells per condition.

S. Robertson et al. / Steroids 78 (2013) 182–194 189

4. Discussion

In this study we have sought to determine the influence of GR

concentration and the ability to dimerize on nuclear translocation

and distribution of the GR. We used statistically different GR con-

centrations (Fig. 1C), which reflect physiological levels of the GR

[5,15,16,38] and a well characterized dimerization impaired GR

mutant, GRdim [6], which closely resembles the natural hGRR477H

mutant [12], in order to unravel the effects of these variables in tan-

dem. Furthermore, we employed two different techniques, live cell

assays of GFP-tagged GR and immunofluorescence of untagged-GR,

both of which are commonly used in the field [35,55,58,87,82].

The diverse techniques at times presented us with seemingly

contradictory findings and thus it is warranted to address the

methodological basis as well as strengths and weaknesses of each

technique as used by us before embarking on a discussion of the

results of our study. For live cell nuclear import assays the nucleus

of an individual cell was selected as the region of interest (ROI) and

fluorescence accumulation recorded automatically. Fluorescence

was set at 0% at the start of the assay while the maximal fluores-

cence was set at 100%, which implies that maximal localization

will always be 100% in this assay. In contrast, evaluation of nuclear

import using the immunofluorescence technique does not set max-

imal nuclear localization at 100% but rather relies on the manual

double blind counting of cells to quantify the percentage of cells

within a population of 50 cells displaying predominantly (>60%)

nuclear fluorescence. This technique is then methodologically most

suited to evaluate differences in maximal nuclear localization. For

both techniques the results are plotted against time and data fitted

to a one phase exponential association curve, which generates both

maximal localization and half time (t½) to maximal nuclear locali-

zation (Figs. 2B and 3A). The t½ in the case of the live cell assay is

the time taken for 50% of the maximal fluorescence to accumulate

in the nucleus of an individual cell, while for the immunofluores-

cence assay it is the time taken for 50% of the cells counted to

achieve >60% nuclear fluorescence. This suggests, as we indeed

found, that the t½ would be underestimated using the immunoflu-

orescence assay as in effect it potentially measures the time taken

for the cells taken to accumulate >30% (60/2) nuclear fluorescence.

This is best illustrated by comparing the t½-values of DEX import of

similar GR concentrations (Fig. 1C), specifically the 3.2 min of med-

ium GRwt concentration for untagged-GRwt (Fig. 3C) and the

10 min of the 38.5 ng GFP-GRwt (Fig. 2C). Inherently then the

immunofluorescent nuclear import assay is less sensitive as it

underestimates the observed import rate calculated using the live

cell import. Specifically, one would expect, as we found, that the

30% nuclear translocation of the live cell assay (3.7 min) would cor-

respond to the 50% nuclear translocation of the immunoflourescent

assay (3.2 min). Thus the live cell nuclear import assay may pres-

ent a more accurate t½-value. For the nuclear export assay the

same methodology was used for the immunoflourescent assay

while for the live cell assay both the nucleus and cytoplasm are se-

lected as ROI and data presented as the ratio of nuclear/cytoplas-

mic fluorescence. For both systems the data is fit to a one phase

exponential decay curve to generate t½ (Figs. 4B and 5A). Once

again the t½-values have slightly different meanings. In the case

of the live cell assay the t½ is the time taken for a 50% decrease

in the ratio of nuclear/cytoplasmic fluorescence in an individual

cell, while for the immunofluorescence assay it is the time taken

for 50% of the cells counted to return to <60% nuclear fluorescence.

When comparing the two curves (Figs. 4B and 5A) it may be clearly

seen that for the live cell assay the curve forms a nice bottom pla-

teau (Fig. 4B), not seen with the immunofluorescence assay

(Fig. 5A), making evaluation of t½ more accurate. Other factors that

may contribute further to differences in values are the difference

between manual and automatic recording of results as well as

the fact that the nuclear mobility of the GFP-tagged receptor may

be affected due to its large protein tag (32 kDA). However, despite

these caveats, comparison of parameters for conditions (such as GR

levels or dimerization state) within a specific assay remains valid.

Despite the fact that GR levels are generally not quantified in

nuclear localization studies, the t½-values we obtained for nuclear

import (4–10 min for GFP-GRwt (Fig. 2C), or 3.2 min (Fig. 3C) for

untagged-GRwt) are similar to those seen in the literature that re-

port t½-values of GFP-GRwt import following the addition of

10�6 M DEX of between 5 and 15 min [35,55,58,87]. Maximal nu-

0

5

10

15

20

25

38.5ng 385ng

CpdADEX18% 18%

14% 14%

[GFP-GRwt]

15% 15%

[GFP-GRdim]

38.5ng

a

b

a

bb b

CV

of

nu

cle

ar

dis

trib

uti

on

(%

)

0 2 4 6 8 10 12

20000

30000

40000

50000

Flu

ore

scen

ce

inte

nsit

y

Distance ( m)

GFP-GRwt Dex (26% CV)

GFP-GRwt CpdA (12% CV)A B

C

Dex

CpdA

38.5ng 385ng

[GFP-GRwt] [GFP-GRdim]

38.5ng

Fig. 6. CpdA results in diffuse nuclear distribution while DEX results in a non-random distribution of GRwt. The nuclear distribution assay was performed as described in

materials and methods. Cells expressing 38.5 or 385 ng GFP-GRwt or 38.5 ng GFP-GRdim were induced with 10�6 M DEX or 10�5 M CpdA for 1 h. Z-stack images of the nuclei

were taken at various focal planes and these were used to deconvolute a single nuclear image. As long a line as possible (white line) was selected in each nucleus, while

avoiding nucleoli and the Cell� imaging software was used to quantify the coefficient of variation (CV) of GFP fluorescence intensity along this line in 5 cells per condition

from 4 separate experiments (±SEM). (A) Representative deconvoluted nuclear images of 10�6 M DEX and 10�5 M CpdA stimulated cells. (B) Representative graph of

fluorescent intensity from two cells demonstrating fluorescent intensity along two nuclear lines with a CV of 12% and 26%, respectively. (C) Coefficient of variation (CV) values

of 38.5 or 385 ng GFP-GRwt and 38.5 ng GFP-GRdim exposed to 10�6 M DEX or 10�5 M CpdA. Statistical analysis of CV of nuclear distribution was carried out using one-way

ANOVA followed by Newman–Keuls post test. Conditions with different letters are statistically different from one another (P < 0.01).

190 S. Robertson et al. / Steroids 78 (2013) 182–194

clear localization of GFP-GRwt was also previously achieved

30 min after 10�7 M DEX stimulation [80,87], similar to results

from our study (Figs. 2B and 3A). Our live cell nuclear export stud-

ies revealed a dramatically faster rate of nuclear export (Fig. 4C) as

compared to immunofluorescent results (Fig. 5B). Although this

may be ascribed to the lower concentration of DEX used to induce

nuclear import (10�9 versus 10�6 M in live cell versus immunoflu-

orescent assay, respectively) as nuclear export has been shown to

be ligand concentration dependent [11], the fact that our export

rates also closely match values reported in the literature suggests

that these differences may also be due to the differences in meth-

odological features between the two assays. Specifically, export of

endogenous GR from isolated thymocytes assayed in the same

manner as for our immunofluorescent assay was shown to have

a t½ of �12 h following the washout of 10�6 M DEX [82], which

is nearly identical to the 13 h we observe for our low GRwt using

the same method (Fig. 5B), while export of GFP-GR after washout

of 10�6 M DEX was �3 h [87], equivalent to the �4 h we find

(Fig. 4C). Furthermore, our nuclear distribution assay displays sim-

ilar results (Fig. 6C) to those of Schaaf et al. [68] who reported a CV

of around 18% for 10�6 M DEX, despite the fact that we have not

used a confocal microscope but have relied on the deconvoluting

capability of the Cell� imaging software of an IX-81 inverted fluo-

rescence microscope.

Our results show for the first time that changes in GR concen-

trations within the physiological range of 67–569 fmol/mg protein

[15] modulate nuclear import of the receptor. Specifically, live cell

nuclear import studies revealed that the observed import rate of

GFP-GRwt at saturating DEX concentrations increased 2.5-fold

(Fig. 2C) as receptor concentration increased 3.4-fold (Fig. 1C).

While the less sensitive immunofluorescent analysis showed no

difference in the observed import rate of GRwt at saturating DEX

concentrations, at sub-saturating DEX concentrations it supports

the effect of GR concentration on observed import rate in that at

10�10 M DEX a 2.3-fold increase in GRwt concentration (Fig. 1C)

lead to a 2.9-fold increase in the observed import rate (Fig. 3C).

The extent of nuclear import, however, is not affected by GR con-

centration (Fig. 3B).

Although the increase in observed import rate at increased GR

concentrations may be ascribed solely to mass action, the fact that

the observed import rate reflects the equilibrium between import

and export rates (kin = kobs � kout/[ligand]) [53] suggests that a pos-

sible cause of the increase in observed nuclear import rate at high-

er GR levels may be due to a decrease in export rate. However, live

cell nuclear export of GFP-GRwt following the washout of sub-sat-

urating DEX concentrations revealed no significant GR concentra-

tion dependent differences (Fig. 4C). In contrast, though,

immunofluorescent nuclear export rate was significantly slower

as GRwt concentration increased (Fig. 5B), which may thus contrib-

ute to the increase in observed import rate seen at higher GRwt

concentrations. An increase in nuclear export rate is associated

with a reduced ability to bind to Hsp90 in the nucleus [67,76]

and may be linked to increased nuclear mobility as a result of a de-

crease in DNA affinity as revealed by agonist bound YFP-GR

[67,76,82]. However, our results suggest that GFP-GRwt concentra-

tion does not affect nuclear mobility (Fig. 6C), which is confirmed

by fluorescent recovery after photobleaching (FRAP) studies per-

formed by Schaaf et al. [68] who also demonstrated that GR con-

centration does not influence the rate of nuclear mobility.

To unravel the effect of GR dimerization on nuclear localization

we employed both the dimerization impaired GR mutant, GRdim

[6], and GRwt induced with CpdA, a GR ligand that has been shown

to abrogate GR dimerization [25,63]. Although the extent of nucle-

ar import at saturating DEX concentrations as examined by immu-

nofluorescent analysis was not affected by GR concentration

(Fig. 3B), GR dimerization is clearly required in order to achieve

maximal nuclear localization as GRdim results in a significantly

lower maximal nuclear localization as compared to GRwt (�77%

versus �95%) when induced with DEX. In support of this, GRwt

showed similarly reduced levels of nuclear localization following

stimulation with dimerization-abrogating CpdA (Fig. 3B). These re-

sults with human GR are similar to those presented by Robertson

et al. [63] for the mouse GR, where DEX stimulation of the mGRwt

resulted in 93% nuclear import while CpdA stimulation of mGRwt

resulted in only 66% nuclear import [63]. In addition, the observed

rate of nuclear import was influenced by the ability of the GR to

dimerize as the GFP-GRdim displayed significantly slower rates

of observed nuclear import as compared to the GFP-GRwt

(Fig. 2C), a result which is supported by the immunofluorescent as-

say finding that the low GRdim condition had a significant reduc-

tion in observed import rate compared to the low GRwt

concentration, which implies an effect of dimerization on observed

import rate (Fig. 3A).

The reduction in observed import rate seen with loss of dimer-

ization, whether through GRdim or GRwt induced with CpdA, may

be explained by the effect of GR dimerization on nuclear export

rates. Specifically, the live cell export rate through the GFP-GRdim,

at all receptor levels, is significantly increased when compared to

the same concentrations of GFP-GRwt (Fig. 4C). This is supported

by results from the immunofluorescent nuclear export assay where

GRdim, at both low and medium concentrations, after DEX wash-

out displays faster export rates (Fig. 5B), while GRwt after CpdA

washout also displays a significantly (P < 0.0001) increased rate

of nuclear export through the medium GRwt concentration when

compared to DEX washout (Fig. 5B). This suggests that export rate

is dimerization dependent and that loss of dimerization increases

the rate of GR export possibly through reduced nuclear retention.

Our nuclear distribution assay confirms a role for dimerization

in nuclear retention in that the nuclear distribution of both the

DEX induced GFP-GRdim as well as the CpdA induced GFP-GRwt

(Fig. 6C) is significantly more random than that observed with

DEX induced GFP-GRwt. Dimerization impaired GR mutants have

been shown to bind to DNA with a reduced capacity relative to

the wild type receptor [1,21,29,78], which would explain the ran-

dom distribution demonstrated by the GFP-GRdim and by GFP-

GRwt following CpdA stimulation (Fig. 6C). Distribution of acti-

vated GR in the nucleus has been shown to be influenced by ligand.

Specifically previous studies demonstrate that stimulation with

the agonist, DEX, results in a non-random particulate distribution

of the GR in the nucleus, while stimulation with the antagonist,

RU486, leads to a random distribution of the GR in the nucleus

[45,67,82]. Furthermore, these foci have been shown to correspond

with areas of the nucleus where active transcription takes place

[81,83]. Similarly to the antagonist RU486, CpdA induces random

nuclear distribution through both the GFP-GRwt and the GFP-

GRdim (Fig. 6C), which supports the evidence that CpdA has a re-

duced capacity to induce transactivation [25,63] via the GR.

The nuclear export and distribution studies offer a possible

explanation for the decrease in observed nuclear import rate

(Figs. 2C and 3A) and lower maximal localization levels (Fig. 3B)

of GRdim as well as the decrease in the observed nuclear import

rate at low GRwt levels (Figs. 2C and 3C). There is a significant

trend towards faster nuclear export of the GRdim as compared to

the GRwt (Figs. 4C and 5B) and for low concentrations of GRwt

as compared to medium concentrations (Fig. 5B). To reiterate, as

the observed nuclear import rate comprises both the rate of import

as well as that of export [53], we therefore propose that the faster

rate of nuclear export demonstrated by GRdim and at low GRwt

concentrations contributes to the reduction in the observed nucle-

ar import rate. Furthermore, the diffuse pattern of nuclear distribu-

tion elicited by stimulation with dimerization-abrogating CpdA

(Fig. 6C) and demonstrated by GRdim following DEX stimulation

S. Robertson et al. / Steroids 78 (2013) 182–194 191

(Fig. 6C), suggests that dimerization is necessary to retain the GR in

the nucleus, which contributes to slower export rates. This is sup-

ported by work done on the nucleoplasmic shuttling of SMAD that

suggests that an increase in import rate is mediated by increased

nuclear retention and the concomitant reduction of nuclear export

[70]. We thus propose that the reduced levels and rate of nuclear

localization shown by GRdim, as well as by CpdA stimulated GRwt,

may be linked to the increased nuclear export rate of undimerized

GR, which is caused by a decrease in nuclear retention as reflected

by the random pattern of nuclear distribution. Thus the influence

of increased GRwt concentration, namely faster nuclear import

and slower nuclear export is reduced by the introduction of the

D-loop GR mutation, GRdim, or by stimulation with CpdA, respec-

tively. As activated GR exists in equilibrium in either a monomeric

or dimeric state where increased GR concentrations shift the bal-

ance towards a higher concentration of dimerized receptor [14],

we theorize that an increased propensity for dimerization may

be responsible for the differential behavior of GR at increased

receptor concentrations.

These findings have physiological implications as GR concentra-

tion is known to vary between tissues [15,38], between individuals

[32] and in diseased states [51]. As the GR is a cytoplasmic tran-

scription factor which mediates the response to GCs via interaction

with GRE’s and other DNA bound transcription factors, its move-

ment into the nucleus and the duration of its nuclear localization

are crucial factors affecting signaling. Tissues which contain high

concentrations of GR would respond faster to GCs due to their in-

creased rate of nuclear import and maintain nuclear localization

for longer following GC withdrawal, because of their slower rate

of nuclear export, thus contributing to their enhanced sensitivity

to GC when compared to tissues with low GR concentrations. Fur-

thermore, the ability of the GR to dimerize has been linked in this

study to full nuclear localization, discreet focal nuclear distribution

and prolonged nuclear retention. These factors may help to explain

the reduced transactivative capacity of the naturally occurring

hGRR477H mutant [64] as well as the much characterized dimeriza-

tion deficient GRdim [34,43,60].

To ensure full occupancy (fractional occupancy >99%) of GR we

used mostly saturating concentrations of DEX and CpdA [63]. The

use of one concentration of DEX is relevant in terms of anti-inflam-

matory drug use as DEX is used pharmacologically at a constant

dose to treat inflammatory diseases [20]. However, it would be

interesting to expand the current study by investigating the effects

of GR concentration and dimerization on nuclear localization using

the natural ligand, cortisol, at more physiological concentrations

and within the context of pulsatile secretion, as recent literature

highlights differences in effects between synthetic ligands such

as DEX and endogenous ligands and pulsatile versus continuous

administration of ligand [3,75].

In summary, our results suggest that the ability of GR to dimer-

ize and increased GR concentration, promote an increase in the ob-

served nuclear import rate of GR and maximal nuclear localization,

a decrease in nuclear export rate and the appearance of discrete li-

gand-induced GR nuclear foci. As it has been shown, using a DNA

binding-deficient mutant of the GR, GFP-D4X [43], that DNA bind-

ing is required for complete nuclear localization of the GR [87], we

hypothesize that the ability to dimerize and the high affinity DNA

binding [71] which this facilitates, although not an absolute

requirement for nuclear retention, does decrease the rate of nucle-

ar export and thus increases the extent and rate of GR import.

Acknowledgements

We thank Carmen Langeveldt from the Department of Biochem-

istry, University of Stellenbosch for excellent technical assistance

in maintaining cells, Nicky Verhoog from the Department of Bio-

chemistry, University of Stellenbosch for assistance with export as-

says, Ben Loos from the Central Analytical Facility, University of

Stellenbosch for invaluable assistance with the fluorescence micro-

scope and Mauritz Venter, Department of Genetics, University of

Stellenbosch for creating the GFP-GRdim construct. The National

Research Foundation (NRF), South Africa is acknowledged for fund-

ing (Grant FA2005040500031 to A.L. and PhD bursary to S.E.R.).

Any opinion, findings and conclusions or recommendations ex-

pressed in this material are those of the author(s) and therefore

the NRF do not accept any liability in regard thereto.

References

[1] Adams M, Meijer OC, Wang J, Bhargava A, Pearce D. Homodimerization of theglucocorticoid receptor is not essential for response element binding:activation of the phenylethanolamine N-methyltransferase gene bydimerization-defective mutants. Mol Endocrinol 2003;17:2583–92. http://dx.doi.org/10.1210/me.2002-0305.

[2] Al-Moslih MI, Dubes GR. The kinetics of DEAE–dextran-induced cellsensitization to transfection. J Gen Virol 1973;18:189–93.

[3] Ashall L, Horton CA, Nelson DE, Paszek P, Harper CV, Sillitoe K, et al. Pulsatilestimulation determines timing and specificity of NF-kappaB-dependenttranscription. Science 2009;324:242–6. http://dx.doi.org/10.1126/science.1164860.

[4] Banerjee A, Periyasamy S, Wolf IM, Hinds Jr TD, Yong W, Shou W, et al. Controlof glucocorticoid and progesterone receptor subcellular localization by theligand-binding domain is mediated by distinct interactions withtetratricopeptide repeat proteins. Biochemistry 2008;47:10471–80. http://dx.doi.org/10.1021/bi8011862.

[5] Beattie CW, Hansen NW, Thomas PA. Steroid receptors in human lung cancer.Cancer Res 1985;45:4206–14.

[6] Beck IM, De Bosscher K, Haegeman G. Glucocorticoid receptor mutants: man-made tools for functional research. Trends Endocrinol Metab 2011. http://dx.doi.org/10.1016/j.tem.2011.03.009.

[7] Black BE, Holaska JM, Rastinejad F, Paschal BM. DNA binding domains indiverse nuclear receptors function as nuclear export signals. Curr Biol2001;11:1749–58.

[8] Bledsoe RK, Montana VG, Stanley TB, Delves CJ, Apolito CJ, McKee DD, et al.Crystal structure of the glucocorticoid receptor ligand binding domain revealsa novel mode of receptor dimerization and coactivator recognition. Cell2002;110:93–105.

[9] Bradford MM. A rapid and sensitive method for the quantitation of microgramquantities of protein utilizing the principle of protein–dye binding. AnalBiochem 1976;72:248–54.

[10] Caamano CA, Morano MI, Dalman FC, Pratt WB, Akil H. A conserved proline inthe hsp90 binding region of the glucocorticoid receptor is required for hsp90heterocomplex stabilization and receptor signaling. J Biol Chem1998;273:20473–80.

[11] Carrigan A, Walther RF, Salem HA, Wu D, Atlas E, Lefebvre YA, et al. An activenuclear retention signal in the glucocorticoid receptor functions as a stronginducer of transcriptional activation. J Biol Chem 2007;282:10963–71. http://dx.doi.org/10.1074/jbc.M602931200.

[12] Charmandari E, Kino T, Ichijo T, Zachman K, Alatsatianos A, Chrousos GP.Functional characterization of the natural human glucocorticoid receptor(hGR) mutants hGRalphaR477H and hGRalphaG679S associated withgeneralized glucocorticoid resistance. J Clin Endocrinol Metab2006;91:1535–43. http://dx.doi.org/10.1210/jc.2005-1893.

[13] Chikanza IC, Petrou P, Kingsley G, Chrousos G, Panayi GS. Defectivehypothalamic response to immune and inflammatory stimuli in patientswith rheumatoid arthritis. Arthritis Rheum 1992;35:1281–8.

[14] Cho S, Kagan BL, Blackford Jr JA, Szapary D, Simons Jr SS. Glucocorticoidreceptor ligand binding domain is sufficient for the modulation ofglucocorticoid induction properties by homologous receptors, coactivatortranscription intermediary factor 2, and Ubc9. Mol Endocrinol2005;19:290–311. http://dx.doi.org/10.1210/me.2004-0134.

[15] Chriguer RS, Elias LL, da Silva Jr IM, Vieira JG, Moreira AC, de Castro M.Glucocorticoid sensitivity in young healthy individuals: in vitro and in vivostudies. J Clin Endocrinol Metab 2005;90:5978–84. http://dx.doi.org/10.1210/jc.2005-0067.

[16] Chrousos GP, Vingerhoeds A, Brandon D, Eil C, Pugeat M, DeVroede M, et al.Primary cortisol resistance in man. A glucocorticoid receptor-mediateddisease. J Clin Invest 1982;69:1261–9.

[17] Cole TJ, Blendy JA, Monaghan AP, Krieglstein K, Schmid W, Aguzzi A, et al.Targeted disruption of the glucocorticoid receptor gene blocks adrenergicchromaffin cell development and severely retards lung maturation. Genes Dev1995;9:1608–21.

[18] Cosio BG, Torrego A, Adcock IM. Molecular mechanisms of glucocorticoids.Arch Bronconeumol 2005;41:34–41.

[19] Costlow ME, Pui CH, Dahl GV. Glucocorticoid receptors in childhood acutelymphocytic leukemia. Cancer Res 1982;42:4801–6.

192 S. Robertson et al. / Steroids 78 (2013) 182–194

[20] Czock D, Keller F, Franz MR, Häussler U. Pharmacokinetics andpharmacodynamics of systemically administered glucocorticoids. ClinPharmacokinet 2005;44:61–98.

[21] Dahlman-Wright K, Siltala-Roos H, Carlstedt-Duke J, Gustafsson JA. Protein–protein interactions facilitate DNA binding by the glucocorticoid receptorDNA-binding domain. J Biol Chem 1990;265:14030–5.

[22] Dahlman-Wright K, Wright A, Gustafsson JA, Carlstedt-Duke J. Interaction ofthe glucocorticoid receptor DNA-binding domain with DNA as a dimer ismediated by a short segment of five amino acids. J Biol Chem1991;266:3107–12.

[23] Davies TH, Ning YM, Sanchez ER. A new first step in activation of steroidreceptors: hormone-induced switching of FKBP51 and FKBP52 immunophilins.J Biol Chem 2002;277:4597–600. http://dx.doi.org/10.1074/jbc.C100531200.

[24] De Bosscher K, Van Craenenbroeck K, Meijer OC, Haegeman G. Selectivetransrepression versus transactivation mechanisms by glucocorticoid receptormodulators in stress and immune systems. Eur J Pharmacol2008;583:290–302. http://dx.doi.org/10.1016/j.ejphar.2007.11.076.

[25] De Bosscher K, Vanden Berghe W, Beck IM, Van Molle W, Hennuyer N,Hapgood J, et al. A fully dissociated compound of plant origin for inflammatorygene repression. Proc Natl Acad Sci USA 2005;102:15827–32. http://dx.doi.org/10.1073/pnas.0505554102.

[26] DeFranco DB. Navigating steroid hormone receptors through the nuclearcompartment. Mol Endocrinol 2002;16:1449–55.

[27] Defranco DB, Madan AP, Tang Y, Chandran UR, Xiao N, Yang J.Nucleocytoplasmic shuttling of steroid receptors. Vitam Horm1995;51:315–38.

[28] Dewint P, Gossye V, De Bosscher K, Vanden Berghe W, Van Beneden K, DeforceD, et al. A plant-derived ligand favoring monomeric glucocorticoid receptorconformation with impaired transactivation potential attenuates collagen-induced arthritis. J Immunol 2008;180:2608–15.

[29] Drouin J, Sun YL, Tremblay S, Lavender P, Schmidt TJ, de Lean A, et al.Homodimer formation is rate-limiting for high affinity DNA binding byglucocorticoid receptor. Mol Endocrinol 1992;6:1299–309.

[30] Echeverria PC, Mazaira G, Erlejman A, Gomez-Sanchez C, Piwien Pilipuk G,Galigniana MD. Nuclear import of the glucocorticoid receptor–hsp90 complexthrough the nuclear pore complex is mediated by its interaction with Nup62and importin beta. Mol Cell Biol 2009;29:4788–97. http://dx.doi.org/10.1128/MCB.00649-09.

[31] Echeverria PC, Picard D. Molecular chaperones, essential partners of steroidhormone receptors for activity and mobility. Biochim Biophys Acta2010;1803:641–9. http://dx.doi.org/10.1016/j.bbamcr.2009.11.012.

[32] Elakovic I, Perisic T, Cankovic-Kadijevic M, Matic G. Correlation betweenglucocorticoid receptor binding parameters, blood pressure, and body massindex in a healthy human population. Cell Biochem Funct 2007;25:427–31.http://dx.doi.org/10.1002/cbf.1323.

[33] Freedman ND, Yamamoto KR. Importin 7 and importin alpha/importin beta arenuclear import receptors for the glucocorticoid receptor. Mol Biol Cell2004;15:2276–86. http://dx.doi.org/10.1091/mbc.E03-11-0839.

[34] Frijters R, Fleuren W, Toonen EJ, Tuckermann JP, Reichardt HM, van derMaaden H, et al. Prednisolone-induced differential gene expression in mouseliver carrying wild type or a dimerization-defective glucocorticoid receptor.BMC Genomics 2010;11:359. http://dx.doi.org/10.1186/1471-2164-11-359.

[35] Galigniana MD, Scruggs JL, Herrington J, Welsh MJ, Carter-Su C, Housley PR,et al. Heat shock protein 90-dependent (geldanamycin-inhibited) movementof the glucocorticoid receptor through the cytoplasm to the nucleus requiresintact cytoskeleton. Mol Endocrinol 1998;12:1903–13.

[36] Gougat C, Jaffuel D, Gagliardo R, Henriquet C, Bousquet J, Demoly P, et al.Overexpression of the human glucocorticoid receptor alpha and beta isoformsinhibits AP-1 and NF-kappaB activities hormone independently. J Mol Med2002;80:309–18. http://dx.doi.org/10.1007/s00109-001-0302-6.

[37] Gross KL, Cidlowski JA. Tissue-specific glucocorticoid action: a family affair.Trends Endocrinol Metab 2008;19:331–9. http://dx.doi.org/10.1016/j.tem.2008.07.009.

[38] Guo WX, Antakly T, Cadotte M, Kachra Z, Kunkel L, Masood R, et al. Expressionand cytokine regulation of glucocorticoid receptors in Kaposi’s sarcoma. Am JPathol 1996;148:1999–2008.

[39] Hache RJ, Tse R, Reich T, Savory JG, Lefebvre YA. Nucleocytoplasmic traffickingof steroid-free glucocorticoid receptor. J Biol Chem 1999;274:1432–9.

[40] Hadley KE, Louw A, Hapgood JP. Differential nuclear localisation and promoteroccupancy play a role in glucocorticoid receptor ligand-specific transcriptionalresponses. Steroids 2011. http://dx.doi.org/10.1016/j.steroids.2011.05.007.

[41] Harrell JM, Murphy PJ, Morishima Y, Chen H, Mansfield JF, Galigniana MD, et al.Evidence for glucocorticoid receptor transport on microtubules by dynein. JBiol Chem 2004;279:54647–54. http://dx.doi.org/10.1074/jbc.M406863200.

[42] Hayashi R, Wada H, Ito K, Adcock IM. Effects of glucocorticoids on genetranscription. Eur J Pharmacol 2004;500:51–62. http://dx.doi.org/10.1016/j.ejphar.2004.07.011.

[43] Heck S, Kullmann M, Gast A, Ponta H, Rahmsdorf HJ, Herrlich P, et al. A distinctmodulating domain in glucocorticoid receptor monomers in the repression ofactivity of the transcription factor AP-1. EMBO J 1994;13:4087–95.

[44] Ho AD, Stojakowits S, Pralle H, Dorner M, Hunstein W. Glucocorticoid receptorlevel, terminal deoxynucleotidyl transferase activity and initial responsivenessto prednisone and vincristine in leukemia. Klin Wochenschr 1983;61:455–9.

[45] Htun H, Barsony J, Renyi I, Gould DL, Hager GL. Visualization of glucocorticoidreceptor translocation and intranuclear organization in living cells with agreen fluorescent protein chimera. Proc Natl Acad Sci USA 1996;93:4845–50.

[46] Jewell CM, Webster JC, Burnstein KL, Sar M, Bodwell JE, Cidlowski JA.Immunocytochemical analysis of hormone mediated nuclear translocation ofwild type and mutant glucocorticoid receptors. J Steroid Biochem Mol Biol1995;55:135–46.

[47] Liu J, DeFranco DB. Protracted nuclear export of glucocorticoid receptor limitsits turnover and does not require the exportin 1/CRM1-directed nuclear exportpathway. Mol Endocrinol 2000;14:40–51.

[48] Louw A, Swart P, de Kock SS, van der Merwe KJ. Mechanism for thestabilization in vivo of the aziridine precursor 2-(4-acetoxyphenyl)-2-chloro-N-methyl-ethylammonium chloride by serum proteins. Biochem Pharmacol1997;53:189–97.

[49] Lu NZ, Collins JB, Grissom SF, Cidlowski JA. Selective regulation of bone cellapoptosis by translational isoforms of the glucocorticoid receptor. Mol CellBiol 2007;27:7143–60. http://dx.doi.org/10.1128/MCB.00253-07.

[50] Lu NZ, Wardell SE, Burnstein KL, Defranco D, Fuller PJ, Giguere V, et al.International Union of Pharmacology. LXV. The pharmacology andclassification of the nuclear receptor superfamily: glucocorticoid,mineralocorticoid, progesterone, and androgen receptors. Pharmacol Rev2006;58:782–97. http://dx.doi.org/10.1124/pr.58.4.9.

[51] Lu YS, Lien HC, Yeh PY, Kuo SH, Chang WC, Kuo ML, et al. Glucocorticoidreceptor expression in advanced non-small cell lung cancer:clinicopathological correlation and in vitro effect of glucocorticoid on cellgrowth and chemosensitivity. Lung Cancer 2006;53:303–10. http://dx.doi.org/10.1016/j.lungcan.2006.05.005.

[52] Lu YS, Lien HC, Yeh PY, Yeh KH, Kuo ML, Kuo SH, et al. Effects of glucocorticoidson the growth and chemosensitivity of carcinoma cells are heterogeneous andrequire high concentration of functional glucocorticoid receptors. World JGastroenterol 2005;11:6373–80.

[53] Madan AP, DeFranco DB. Bidirectional transport of glucocorticoid receptorsacross the nuclear envelope. Proc Natl Acad Sci USA 1993;90:3588–92.

[54] Mittelstadt PR, Ashwell JD. Disruption of glucocorticoid receptor exon 2 yieldsa ligand-responsive C-terminal fragment that regulates gene expression. MolEndocrinol 2003;17:1534–42. http://dx.doi.org/10.1210/me.2002-0429.

[55] Nishi M, Ogawa H, Ito T, Matsuda KI, Kawata M. Dynamic changes insubcellular localization of mineralocorticoid receptor in living cells: incomparison with glucocorticoid receptor using dual-color labeling withgreen fluorescent protein spectral variants. Mol Endocrinol 2001;15:1077–92.

[56] Okret S, Poellinger L, Dong Y, Gustafsson JA. Down-regulation of glucocorticoidreceptor mRNA by glucocorticoid hormones and recognition by the receptor ofa specific binding sequence within a receptor cDNA clone. Proc Natl Acad SciUSA 1986;83:5899–903.

[57] Pemberton LF, Paschal BM. Mechanisms of receptor-mediated nuclear importand nuclear export. Traffic 2005;6:187–98. http://dx.doi.org/10.1111/j.1600-0854.2005.00270.x.

[58] Picard D, Yamamoto KR. Two signals mediate hormone-dependent nuclearlocalization of the glucocorticoid receptor. EMBO J 1987;6:3333–40.

[59] Pratt WB, Galigniana MD, Harrell JM, DeFranco DB. Role of hsp90 and thehsp90-binding immunophilins in signalling protein movement. Cell Signal2004;16:857–72. http://dx.doi.org/10.1016/j.cellsig.2004.02.004.

[60] Reichardt HM, Kaestner KH, Tuckermann J, Kretz O, Wessely O, Bock R, et al.DNA binding of the glucocorticoid receptor is not essential for survival. Cell1998;93:531–41.

[61] Reik A, Schutz G, Stewart AF. Glucocorticoids are required for establishmentand maintenance of an alteration in chromatin structure: induction leads to areversible disruption of nucleosomes over an enhancer. EMBO J1991;10:2569–76.

[62] Riml S, Schmidt S, Ausserlechner MJ, Geley S, Kofler R. Glucocorticoid receptorheterozygosity combined with lack of receptor auto-induction causesglucocorticoid resistance in Jurkat acute lymphoblastic leukemia cells. CellDeath Differ 2004;11(Suppl. 1):S65–72. http://dx.doi.org/10.1038/sj.cdd.4401413.

[63] Robertson S, Allie-Reid F, Berghe WV, Visser K, Binder A, Africander D, et al.Abrogation of glucocorticoid receptor dimerization correlates with dissociatedglucocorticoid behavior of compound a. J Biol Chem 2010;285:8061–75.http://dx.doi.org/10.1074/jbc.M109.087866.

[64] Ruiz M, Lind U, Gafvels M, Eggertsen G, Carlstedt-Duke J, Nilsson L, et al.Characterization of two novel mutations in the glucocorticoid receptor gene inpatients with primary cortisol resistance. Clin Endocrinol (Oxf)2001;55:363–71.

[65] Savory JG, Hsu B, Laquian IR, Giffin W, Reich T, Hache RJ, et al. Discriminationbetween NL1- and NL2-mediated nuclear localization of the glucocorticoidreceptor. Mol Cell Biol 1999;19:1025–37.

[66] Savory JG, Prefontaine GG, Lamprecht C, Liao M, Walther RF, Lefebvre YA, et al.Glucocorticoid receptor homodimers and glucocorticoid–mineralocorticoidreceptor heterodimers form in the cytoplasm through alternative dimerizationinterfaces. Mol Cell Biol 2001;21:781–93. http://dx.doi.org/10.1128/MCB.21.3.781-793.2001.

[67] Schaaf MJ, Cidlowski JA. Molecular determinants of glucocorticoid receptormobility in living cells: the importance of ligand affinity. Mol Cell Biol2003;23:1922–34.

[68] Schaaf MJ, Lewis-Tuffin LJ, Cidlowski JA. Ligand-selective targeting of theglucocorticoid receptor to nuclear subdomains is associated with decreasedreceptor mobility. Mol Endocrinol 2005;19:1501–15. http://dx.doi.org/10.1210/me.2005-0050.

[69] Schlaghecke R, Kornely E, Wollenhaupt J, Specker C. Glucocorticoid receptorsin rheumatoid arthritis. Arthritis Rheum 1992;35:740–4.

S. Robertson et al. / Steroids 78 (2013) 182–194 193

[70] Schmierer B, Hill CS. Kinetic analysis of Smadnucleocytoplasmic shuttlingreveals a mechanism for transforming growth factor beta-dependent nuclearaccumulation of Smads. Mol Cell Biol 2005;25:9845–58.

[71] Segard-Maurel I, Rajkowski K, Jibard N, Schweizer-Groyer G, Baulieu EE,Cadepond F. Glucocorticosteroid receptor dimerization investigated byanalysis of receptor binding to glucocorticosteroid responsive elementsusing a monomer-dimer equilibrium model. Biochemistry 1996;35:1634–42.http://dx.doi.org/10.1021/bi951369h.

[72] Shahin V. Route of glucocorticoid-induced macromolecules across the nuclearenvelope as viewed by atomic force microscopy. Pflugers Arch 2006;453:1–9.http://dx.doi.org/10.1007/s00424-006-0102-5.

[73] Simons Jr SS. What goes on behind closed doors: physiological versuspharmacological steroid hormone actions. Bioessays 2008;30:744–56. http://dx.doi.org/10.1002/bies.20792.

[74] Stavreva DA, Muller WG, Hager GL, Smith CL, McNally JG. Rapid glucocorticoidreceptor exchange at a promoter is coupled to transcription and regulated bychaperones and proteasomes. Mol Cell Biol 2004;24:2682–97.

[75] Stavreva DA, Wiench M, John S, Conway-Campbell BL, McKenna MA, Pooley JR,et al. Ultradian hormone stimulation induces glucocorticoid receptor-mediated pulses of gene transcription. Nat Cell Biol 2009;11:1093–102.http://dx.doi.org/10.1038/ncb1922.

[76] Tago K, Tsukahara F, Naruse M, Yoshioka T, Takano K. Regulation of nuclearretention of glucocorticoid receptor by nuclear Hsp90. Mol Cell Endocrinol2004;213:131–8. http://dx.doi.org/10.1016/j.mce.2003.10.057.

[77] Ronacher K, Hadley K, Avenant C, Stubsrud E, Simons Jr SS, Louw A, et al.Ligand-selective transactivation and transrepression via the glucocorticoidreceptor: role of cofactor interaction. Mol Cell Endocrinol 2009;299:219–31.

[78] Tanner TM, Denayer S, Geverts B, Van Tilborgh N, Kerkhofs S, Helsen C, et al. A629RKLKK633 motif in the hinge region controls the androgen receptor atmultiple levels. Cell Mol Life Sci 2010;67:1919–27. http://dx.doi.org/10.1007/s00018-010-0302-1.

[79] Tazawa H, OsmanW, Shoji Y, Treuter E, Gustafsson JA, Zilliacus J. Regulation ofsubnuclear localization is associated with a mechanism for nuclear receptorcorepression by RIP140. Mol Cell Biol 2003;23:4187–98.

[80] Usuku T, Nishi M, Morimoto M, Brewer JA, Muglia LJ, Sugimoto T, et al.Visualization of glucocorticoid receptor in the brain of green fluorescentprotein–glucocorticoid receptor knockin mice. Neuroscience2005;135:1119–28. http://dx.doi.org/10.1016/j.neuroscience.2005.06.071.

[81] van Royen ME, Cunha SM, Brink MC, Mattern KA, Nigg AL, Dubbink HJ, et al.Compartmentalization of androgen receptor protein–protein interactions inliving cells. J Cell Biol 2007;177:63–72. http://dx.doi.org/10.1083/jcb.200609178.

[82] Vicent GP, Pecci A, Ghini A, Piwien-Pilipuk G, Galigniana MD. Differences innuclear retention characteristics of agonist-activated glucocorticoid receptormay determine specific responses. Exp Cell Res 2002;276:142–54. http://dx.doi.org/10.1006/excr.2002.5532.

[83] Wrange O, Eriksson P, Perlmann T. The purified activated glucocorticoidreceptor is a homodimer. J Biol Chem 1989;264:5253–9.

[84] Wust S, Tischner D, John M, Tuckermann JP, Menzfeld C, Hanisch UK, et al.Therapeutic and adverse effects of a non-steroidal glucocorticoid receptorligand in a mouse model of multiple sclerosis. PLoS One 2009;4:e8202. http://dx.doi.org/10.1371/journal.pone.0008202.

[85] Yang J, Liu J, DeFranco DB. Subnuclear trafficking of glucocorticoid receptorsin vitro: chromatin recycling and nuclear export. J Cell Biol 1997;137:523–38.

[86] Yemelyanov A, Czwornog J, Gera L, Joshi S, Chatterton Jr RT, Budunova I. Novelsteroid receptor phyto-modulator compound a inhibits growth and survival ofprostate cancer cells. Cancer Res 2008;68:4763–73. http://dx.doi.org/10.1158/0008-5472.CAN-07-6104.

[87] Yoshikawa N, Makino Y, Okamoto K, Morimoto C, Makino I, Tanaka H. Distinctinteraction of cortivazol with the ligand binding domain confersglucocorticoid receptor specificity: cortivazol is a specific ligand for theglucocorticoid receptor. J Biol Chem 2002;277:5529–40. http://dx.doi.org/10.1074/jbc.M107946200.

194 S. Robertson et al. / Steroids 78 (2013) 182–194