| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Dame Roma Mitchell Cancer Research Laboratories (L.M.B., P.J.N., G.B., C.S.Y.C., C.R., K.S., M.L., A.O., M.Y., M.P.B., W.D.T.), Department of Medicine, The University of Adelaide and Hanson Institute, and School of Molecular and Biomedical Science (M.M.C.), the University of Adelaide, Adelaide, South Australia 5000, Australia
Address all correspondence and requests for reprints to: Lisa M. Butler, Dame Roma Mitchell Cancer Research Laboratories, The University of Adelaide, Hanson Institute, P.O. Box 14, Rundle Mall, Adelaide, South Australia 5000, Australia. E-mail: lisa.butler{at}imvs.sa.gov.au.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The transcriptional activity of the AR is mediated by three activation functions, AF-1 and AF-5 located in the amino terminal transactivation domain (AR-NTD), and AF-2 located in the ligand binding domain (LBD). AF-1 consists of at least two overlapping transactivation functions, AF-1a and AF-1b (10). Variants of the human AR that retain inhibitory subdomains and are unable to transactivate target genes have the potential to act in a dominant-negative manner to inhibit androgen signaling through the endogenous AR in prostate cancer cells. We have constructed human AR variants with deletions or inactivating mutations in AF-1, AF-5, and/or AF-2, to generate transcriptionally inactive ARi. Here, we demonstrate that AR inhibitor (ARi)-410 is a dominant-negative inhibitor of the activity of both wtAR and mutant ARs identified in clinical prostate cancer, and can suppress the growth of AR-positive LNCaP prostate cancer cells in culture.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). The criteria for dominant-negative activity was set as an ARi variant that has no intrinsic transactivation activity in the presence of the native ligand 5
-dihydrotestosterone (DHT) and can inhibit agonist-induced wtAR activity by at least 50% when cotransfected into prostate cancer cells at a 1:1 molar ratio. The intrinsic activity of each of the ARi constructs, depicted in Fig. 1, was compared with the wild-type AR (wtAR) using a transient transfection assay in AR-negative PC-3 human prostate carcinoma cells. The wtAR activated the androgen-responsive probasin promoter (ARR3-tk-luciferase) 50- to 60-fold in the presence of DHT (1 nM; Fig. 2A). Compared with wtAR, four of the ARi constructs (ARi-410 and ARi-410-Q; and ARi-532 and ARi-532-Q) did not show any appreciable transactivation activity in the presence of DHT (Fig. 2, A and B). In contrast, ARi-297 and ARi-297-Q had approximately 25% of wtAR activity (Fig. 2, A and B), indicating that AF-1a and AF-1b are not the sole contributors to the transactivation capacity of the receptor. All receptor variants truncated (t) at amino acid 707 showed constitutive, ligand-independent activity (Fig. 2C), although for constructs ARi-297-t and ARi-532-t this activity was reduced by approximately 80% compared with AR-t707. Immunoblot analysis with the AR-C19 antibody, raised against the C-terminal 19 amino acids of the wtAR protein, demonstrated that the ARi-297, ARi-410, and ARi-532 constructs were expressed at comparable protein levels after transient transfection in COS-1 cells and were of the predicted molecular masses of 92, 60, and 49 kDa, respectively (Fig. 2D).
|
|
), with ARi-410 achieving 50% inhibition at a 1:1 molar ratio to wtAR and a maximal inhibition of approximately 95% at a 4:1 molar ratio. The amount of ARi transfected did not appreciably alter the level of wtAR (Fig. 3B), suggesting that the ARi-mediated decrease in wtAR activity was not a consequence of decreased wtAR expression, or effects on wtAR steady-state protein levels.
|
). Although the DHT-induced transactivation activity of ARi-410 is approximately 10-fold lower than wtAR, we observed a qualitatively similar effect of these two coregulators on ARi-410 activity. The DHT-induced activity of ARi-532, which had weak dominant-negative ability, was enhanced by GRIP1, but SMRT had no effect, indicating that a region between amino acids 410 and 532 is necessary for binding of SMRT, which may explain the reduced capacity of ARi-532 to repress AR function compared with ARi-410.
Specificity of ARi Action
Further characterization of ARi-410 indicated that it was able to inhibit activity of wtAR on three different AR-regulated reporters, containing the probasin (ARR3-tk), prostate-specific antigen (PSA), and mouse mammary tumor virus (MMTV) promoters respectively (Fig. 4A). We also demonstrated that ARi-410 preferentially inhibits AR activity rather than that of other nuclear receptors. We tested the ability of ARi-410 to suppress the activity of the vitamin D receptor (VDR), which is expressed in prostate cancer cells and has a predominantly antiproliferative effect on prostate cancer cell growth (11), the estrogen receptor (ER) and progesterone receptor (PR). In transiently transfected PC-3 cells, ARi-410 did not inhibit VDR activity induced by 1,25-dihydroxyvitamin D3 (Fig. 4B). Similarly, we detected no inhibition of ER activity by ARi-410 (data not shown). ARi-410 did cause some inhibition of the PR, a more closely related steroid receptor, with an approximately 35% reduction in PR activity with equimolar amounts of ARi-410. However the maximal inhibition of PR activity by ARi-410 was 50% (compared with 95% inhibition of AR activity), when ARi-410 was expressed at a 4-fold molar excess of the PR (Fig. 4C), which does not fit our criteria of a dominant-negative inhibitor.
|
), known to be promiscuous with respect to ligand activation in prostate cancer cell lines and clinical tumors (12, 13, 14). Similarly, ARi-410 was able to effectively inhibit the function of the endogenous AR-T875A variant in LNCaP cells by up to 90% (Fig. 5B). The effect of transfected ARi-410 on prostate cancer cell proliferation was assessed in androgen-responsive, AR-positive LNCaP cells and androgen-independent, AR-negative PC-3 cells. A green fluorescent protein (GFP)-ARi-410 fusion protein, which inhibits wtAR activity in a DHT-dependent, dominant-negative manner comparable to ARi-410 (data not shown), was used to select cells expressing ARi-410 after transfection. The proliferation of LNCaP cells expressing GFP-ARi-410 was significantly reduced (P < 0.05) at d 3 and 5 of culture compared with control cells expressing the GFP empty vector control alone (Fig. 5C). In contrast, there was no significant difference in the growth of PC-3 cells expressing either GFP-ARi-410 or GFP (Fig. 5C). Immunoblot analysis confirmed that the expression of GFP-ARi-410 was maintained in transfected cells throughout the experimental period (Fig. 5C).
|
). The wtAR was detected using the U407 AR antibody, which was raised against amino acids 200–220 of the wtAR and consequently selectively detects the full-length wtAR and not ARi-410. Whereas both wtAR and GFP-ARi-410 were predominantly located in the cytoplasm in the absence of DHT (
90% of cells with cytoplasmic localization), both receptors were localized in the nucleus in the presence of DHT (90% of cells) and exhibited the same punctate pattern of nuclear fluorescence as observed when the receptors were expressed individually (Fig. 6, A and B). The LNCaP AR-A875 receptor variant also translocated into the nucleus irrespective of the presence of GFP-ARi-410, in response to either DHT or the AR antagonist hydroxyflutamide, the latter being a known activator of the AR-A875 variant (data not shown). These results indicate that ARi-410 blocks transcriptional function of the wtAR in the nucleus, and not the capacity to traffic from the cytoplasm.
|
, upper panel), indicative of an interaction between AR and ARi-410. Immunoblot analysis with U407 detected only full-length wtAR as expected (Fig. 7, lower panel).
|
). Similarly, ARi-410 showed a 20-fold induction of an N/C interaction in the presence of DHT (10 nM; Fig. 8A), and this interaction was abolished by the incorporation of the E895Q substitution into the LBD (Fig. 8A). To assess whether disruption of the N/C interaction in wtAR would affect the dominant-negative activity of ARi-410, we generated variants of ARi-410 with mutation of the 23FQNLF27 or 432WHTLF436 peptides that mediate the N/C interaction (20). These variants (ARi-410-AQNAA and ARi-410-AHTAA) were equally effective as the original ARi-410 in inhibition of wtAR function (Fig. 8B). In addition, we tested the ability of ARi-410 to inhibit the function of the AR-E895Q variant. Although AR-E895Q cannot undergo an N/C interaction, it retains up to 80% of wtAR transactivation activity (19). ARi-410 inhibited the function of both wtAR and AR-E895Q activity in a dose-dependent manner (Fig. 8C). Taken together with our data demonstrating that ARi-410-Q is an effective inhibitor of wtAR activity (Fig. 3A), it appears that N/C interaction is not required for dominant-negative activity and that dimerization of AR and ARi-410 occurs primarily through alternative receptor domains.
|
DBD) was considerably less effective than ARi-410 in inhibiting wtAR activity (Fig. 9A). Immunoblotting of the lysates used to measure transactivation activity confirmed that the difference in dominant-negative activity between ARi-410 and ARi-410-DBD was not due to differences in steady-state levels of the two proteins (Fig. 9A).
|
), indicating that DNA binding per se is not required for dominant-negative activity. The A594T amino acid substitution has been shown to prevent dimerization of the AR through the DBD as well as disrupting DNA binding (23, 24), and when incorporated into ARi-410 (i.e. ARi-410-A594T) did not affect dominant-negative activity. Taken together, the results suggest that, although the DBD is required for optimal dominant-negative activity of ARi-410, this activity is not dependent on DNA binding per se, nor on dimerization through the DBD.
We next addressed whether dimerization through the LBD was involved in dominant-negative activity. Although dimerization through the LBD of the AR has been demonstrated (25), a specific sequence mediating dimerization has not yet been defined. We mutated a proline residue in the AR (P764), corresponding to a conserved amino acid in the LBD of the GR that is required for GR dimerization (26), to generate ARi-410-P764A. When compared with ARi-410, this variant showed substantially reduced dominant-negative activity on wtAR (Fig. 10). Again, immunoblot analysis confirmed that ARi-410 and ARi-410-P764A were expressed at comparable levels in the transfected cells used in the assay (Fig. 10). These results suggest that dimerization through the LBD may contribute to optimal inhibitory activity of ARi-410 on wtAR.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
A desirable attribute of an ARi is that its effect is specific for the AR. Accordingly, we have demonstrated that the most potent ARi, ARi-410, does not inhibit the activity of the VDR, which is commonly expressed in prostate cancer cells, exhibits predominantly antiproliferative effects on prostate cancer cell growth, and is known to impact on other steroid hormone signaling pathways, including the AR (11, 29). Similarly, no inhibitory effect of ARi-410 was observed for the ER, and a substantially reduced inhibitory effect was observed on the PR. Taken together with our observation that ARi-410 can suppress the growth of AR-positive (LNCaP) but not AR-negative (PC-3) prostate cancer cells, our data strongly indicate that the efficacy of ARi-410 is dependent on AR-mediated cell growth pathways. Squelching by ARi-410 of cofactors that interact with the AR could explain the partial inhibition of PR activity with an excess of ARi-410.
Dominant-negative human ARs have also been generated by Bramlett et al. (30) by fusing histone deacetylase 1 or the Kruppel-associated box transcriptional repressor domain to an inactive AR variant containing a deletion of the first 122 amino acids of the protein (30). Although those constructs were potent inhibitors of AR function, it is possible that incorporation of exogenous repressor domains may reduce specificity of action by impacting on other signaling pathways. ARi-410, the most effective inhibitor of AR function tested in the current study, is structurally similar to a dominant-negative rat AR (residues 46–408 deleted) generated by Palvimo et al. (31). Because ARi-410 retains part of the ligand-independent transactivation function, AF-5, ARi-532 was designed to remove both AF-1 and AF-5, to determine whether complete loss of both activation functions would generate a more potent ARi. Like ARi-410, ARi-532 had no intrinsic transactivation activity, but a comparison of the relative efficacies of ARi-410 and ARi-532 to inhibit wtAR activity indicates that retention of the 123-amino acid portion of AF-5 (i.e. amino acids 410–532) is required for optimal dominant-negative activity. This region has been implicated in ligand-independent activation of AR (32, 33), stabilization of AR N/C interaction (15, 20) and, consistent with our findings, interaction with AR coregulators (34, 35, 36, 37). In addition, AR gene mutations identified in clinical prostate cancer colocate to this region of the receptor (7, 38, 39). Our data indicate that, although both ARi-410 and ARi-532 can interact with GRIP1, indicating that GRIP1 acts primarily through the LBD as expected, only ARi-410 can interact with SMRT. This suggests that a region between 410 and 532 is sufficient for AR repression by SMRT and for optimal inhibition of wtAR function by an ARi. Further definition of the specific elements required for ARi activity may provide further insight into both AR function and the requirements for dominant-negative activity.
The mechanism by which transcriptionally inactive AR variants suppress androgen signaling may involve the formation of inactive receptor heterodimers (40), such that the inhibitory effects of an ARi may be a function of the relative levels of ARi homodimers, AR-ARi heterodimers and AR homodimers present in the prostate cancer cell. ARi-410 translocates to the nucleus with wtAR in the presence of DHT, with a characteristic punctate pattern previously observed for wtAR (41). A specific interaction was demonstrated between wtAR and ARi-410 by coimmunoprecipitation, supporting the hypothesis that heterodimers can be formed between the endogenous AR and ARi-410. This association was also observed in the absence of ligand, indicating that heterodimers may form even when the receptors are localized in the cytoplasm. In experiments designed to delineate the regions of the AR required for dominant-negative activity, we determined that ARi-410 does not appear to require the so-called N/C interaction associated with agonist-induced activation of the wtAR (15, 16, 17, 18). A variant of ARi-410 that lacked the DBD (i.e. ARi-410-DBD), another key dimerization interface for nuclear receptors (21), had substantially reduced ability to inhibit wtAR activity. Introduction of mutations into the DBD that prevent binding to response elements of target genes (i.e. ARi-410-C617Y) or dimerization through the DBD (i.e. ARi-410-A594T) did not substantially interfere with the dominant-negative activity of ARi-410 on AR function, indicating that the loss of dominant-negative activity associated with ARi-410-DBD involved other sequences or motifs within this region. In contrast, incorporating a mutation to disrupt dimerization of AR through the LBD (i.e. ARi-410-P764A) substantially reduced the ability of ARi-410 to suppress AR function. Taken together, these findings indicate that the DBD and dimerization through the LBD are essential for optimal dominant-negative receptor activity and may provide a means of generating a smaller ARi that is more easily delivered to target cells.
In summary, we have constructed an AR variant that can effectively inhibit wtAR activity in vitro by greater than 95%. This level of AR inhibition is comparable to the maximal level achieved using a thousand-fold excess of specific AR antagonists, such as hydroxyflutamide or bicalutamide. We propose that specifically targeting the AR by ARi has the potential to be equally effective at inhibiting the growth of prostate cancer cells compared with current treatment modalities, which typically target the ligand rather than the receptor. Moreover, there is the potential to use this approach in combination with other AR-targeting agents such as specific AR blocking antisera, AR antisense oligonucleotides, AR-based peptides, or AR mRNA hammerhead ribozymes to achieve more complete blockade of AR signaling (27, 28, 42). Irrespective of their long-term clinical utility, ARi will be invaluable reagents to study AR function, undertake preclinical studies to assess the role of AR in resistance to androgen ablation therapy (AAT), and to determine whether complete abrogation of AR signaling—using an ARi either alone or in combination with other AR-targeting agents—will prevent regrowth of prostate cancer after failure of AAT.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Plasmids
The luciferase-linked reporter plasmids probasin-ARR3-tk-luciferase, PSA-luciferase (pGL3PSA540), and MMTV-luciferase were provided by Dr. Robert Matusik (Vanderbilt University Medical Center, Nashville, TN), Bristol-Myers Squibb, and Dr. Ron M. Evans (Salk Institute, La Jolla, CA), respectively. The GRIP1 plasmid was provided by Drs. Gerry Coetzee and Mike Stallcup (University of Southern California, Los Angeles, CA). The pCYP24-luciferase reporter and VDR expression plasmids were provided by Drs. Prem Dwivedi and Brian May (University of Adelaide, South Australia). The PR and PRE-luciferase plasmids were provided by Prof. Christine Clarke (Westmead Millennium Institute, Sydney, New South Wales). The base substitutions required to generate the AR-L699H, -T875A, -E895Q, -C617Y, -A594T, -P764A, AQNAA. and AHTAA AR variants were introduced into an expression vector carrying the complete human AR gene coding sequence (pCMV3.1AR) (44), by PCR-based megaprimer in vitro mutagenesis, as described previously (45), and subcloned into the ARi-410 construct. The AR variant truncated at amino acid 707 (pCMV-AR-t707) was generated by introducing an XbaI site at 707, and cloning the XbaI and HindIII fragment into the pCMV3.1AR plasmid.
Structure of AR Variants Designed to Inhibit wtAR Function
The ARi-532, ARi-410, and ARi-297 receptor variants (Series 1) were constructed from the basic AR structure as shown in Fig. 1, A and B, with specific deletions in the AR-NTD. All of these AR variants were able to bind radiolabeled DHT with similar affinity to wtAR (data not shown). An E to Q substitution at amino acid 895 in AF-2, which has been shown to disrupt the AR N/C interaction and reduce receptor activity, while retaining ligand and DNA binding at normal affinity (46, 47), was introduced into each of the above ARi constructs to potentially improve efficacy by preventing intramolecular interaction (Series 2, Fig. 1C). The third modification was to truncate each of the three basic ARi structures in the C-terminal at amino acid 707 (Series 3, Fig. 1D). In the background of an intact N-terminal AR-NTD, this truncation results in a constitutively active AR.
Generation of an AR Variant with Amino Acids 39–410 Deleted (ARi-410).
This construct was based on the rat dominant-negative AR generated previously (31). An insert containing this deletion was generated by Splicing Overlap Extension PCR (SOE-PCR) (48) using two sets of primer pairs:
5'-GCAGAGCTCGTTTAGTGAACC (pCMV-sense)
5'-GCTGCACCCGCGCCATGCAGGCCCGGGTTCTGGATCACTTC (AR256extv2-antisense),
5'-CTGCATGGCGCGGGTGCAGC (AR1393-sense),
5'-GGGCACTCTGCTCACCATGC (AR1673-antisense)
PCR was performed on the pCMV3.1AR template using the Expand PCR kit (Roche, Indianapolis, IN). After treatment with Klenow, a third PCR was performed on templates generated in the previous two PCRs using the pCMV-sense and X4-antisense primers. The NheI, BstEII digestion product of the third PCR was ligated to the large fragment of pCMV3.1AR, pCMV-AR-t707 or pCMV-AR895 digested with the same restriction enzymes. Clones containing the deletion were identified by digestion with EcoRI, and SmaI digestion in the truncated receptor pCMV-AR-t707.
Generation of an AR Variant with Amino Acids 39–532 Deleted (ARi-532).
These constructs were generated as described above, using primer pairs:
5'-GCAGAGCTCGTTTAGTGAACC (pCMV-sense)
5'-CGTAAGGTCCGGACTAGCTACCCGGGTTCTGGATCACTTC (AR256ext-antisense)
5'-TAGCTACTCCGGACCTTACG (AR1759-sense)
5'-ACACACTACACCTGGCTCAAT (X4-antisense)
A third PCR was performed with primers pCMVsense and X4-antisense on Klenow-treated templates generated in the first two PCRs. This PCR product was digested with EagI and AspII and ligated to the large fragment of pCMV3.1AR, pCMV-AR-t707, or pCMV-AR-E895Q digested with the same restriction enzymes.
Generation of an AR Variant Containing Mutations Inactivating AF1a and Deletion of AF1b (ARi-297).
These constructs, designed to inactivate AF1a and AF1b (10), were generated as described above using the following primer pairs:
5'-CCCGGCTTAAGCAGCTGCTCCGCTGACCTTAAAGACAACAACAGCGAGGCCAGCACCATGC (ARN184–185),
5'-GCCAGTGGAAAGTTGTAGTAGCTGTCGTCTAGCAGAGAAC (AR128extv1as)
5'-TACTACAACTTTCCACTGGC (AR1243-sense),
5'-GGGCACTCTGCTCACCATGC (AR1673-antisense)
Underlined bases in primer ARN184–185 introduce inactivating mutations into AF1a (184N and 185N) as described previously (10). The AflII, BstEII digestion product of the third PCR was ligated to the large fragment of pCMV3.1AR, pCMV-AR-t707 or pCMV-AR-E895Q digested with the same restriction enzymes.
Generation of ARi-410 Variant with the DBD Deleted (Amino Acids 553–622).
An insert containing this deletion was generated as described above using the following primer pairs:
5'-TAGCCCCCTACGGCTACACT (N166)
5'-ACACACTACACCTGGCTCAAT (X4AS)
5'-CTATTACTTTCCACCCACTCTGGGAGCCCGGAAGCTG (AR-DBDdS)
5'-CCGGGCTCCCAGAGTGGGTGGAAAGTAATAGTCATTG (AR-DBDdAS)
The Bsu36I, TthIII 1 digestion product of the third PCR was ligated to the large fragment of pCMV3.1ARi-410 digested with the same restriction enzymes.
GFP-ARi Construct.
The chimeric GFP-AR plasmid was generously provided by Dr. Marco Marcelli (Baylor College of Medicine, Houston, TX). The GFP-ARi construct was generated from the GFP-AR construct using the same cloning strategy described above for the pCMV3.1AR plasmid. The pEGFP-C1 (CLONTECH, Mountain View, CA) encodes a red-shifted variant of wild-type GFP, due to a double amino acid substitution of Phe-64 to Leu and Ser-65 to Thr.
Cell Culture
Transactivation assays.
PC-3AR– (43) and COS-1 cells were maintained in RPMI 1640 medium containing glutamine and 5% fetal calf serum (FCS). LNCaP cells were maintained in RPMI 1640 medium containing glutamine and 10% FCS. For transactivation studies, cells (1.5 x 104 cells/well) were seeded in 96-well plates and allowed to attach overnight. Cells were then transfected with 100 ng of reporter-luciferase plasmid (ARR3-tk-luciferase, PSA-luciferase or MMTV-luciferase) and 2.5 ng of wtR expression vector (pCMV3.1AR), using LipofectAMINE 2000 transfection reagent (Invitrogen, Carlsbad, CA), according to the manufacturer’s instructions. For the inhibitory studies, the ARi constructs were cotransfected with wtAR at 0.5:1 to 4:1 molar ratios based on relative construct lengths. An equivalent molar amount of ectopic promoter was maintained in each experiment by addition of the appropriate amount of parental pCMV3.1 expression vector. Equal amounts of plasmid DNA were similarly maintained by addition of noneukaryotic vector DNA pBS(sk–). Three hours after transfection, the medium was removed from the cells and replaced with phenol red-free RPMI 1640 containing 5% dextran-coated charcoal-stripped FCS supplemented with the appropriate amount of ligand or an equivalent volume of carrier (ethanol). Cells were lysed 36 h later and assayed for luciferase activity using a Luciferase assay kit (Promega, Madison, WI) and detected using a plate-reading luminometer (Top Count, Packard).
Coactivator and corepressor experiments were performed similarly by transfection of PC-3 cells (1 x 104 cells/well) with an equal molar ratio of AR, ARi-410, or ARi-532 expression vectors, a 5-fold molar ratio of control (pVP16:SV40-T), pSG5:GRIP1 or pCMX:SMRT expression vectors, and 100 ng of ARR3-tk-luciferase reporter construct. The total amount of DNA in each transfection mix was kept constant by the addition of empty noneukaryotic vector DNA pBS(sk–). Cells were cultured for 24 h in medium containing 10 nM DHT, lysed, and assayed for luciferase activity as described above.
N/C Interaction Assay
N/C interaction assays were performed in COS-1 cells (1.5 x 104 cells/well) in 96-well plates. Each well was cotransfected with an equal molar amount of a pM-AR-LBD (encoding AR amino acids 644–917) and a pVP16-AR-NTD (encoding AR amino acids 1–538) vector (maximum 5 ng of each vector), and 25 ng of the pGK1 reporter. DNA was balanced by adding an appropriate amount of pBS-sk(–). After transfection, cells were cultured for 24 h in phenol red-free medium in the presence or absence of DHT (10 nM). Luciferase activity was measured in cell lysates as detailed above. To demonstrate specificity of the N/C interaction, each experiment also included cotransfection of each AR vector with an equal molar ratio of either pM-53 and pVP16-T (expressing p53 and simian virus 40 large T antigen, respectively) as appropriate, and culturing with or without DHT. No significant activity over background was observed for any control reaction in either the presence or absence of ligand during the course of this study.
Proliferation Assays
LNCaP or PC-3 cells were seeded 48 h before transfection in 10-cm tissue culture plates at a density of 3 x 106 cells/plate. LipofectAMINE 2000 reagent was used to transfect 24 µg of either pEGFP-C1 empty plasmid or GFP-ARi-410 plasmid per plate. After 12 h, cells were harvested and sorted by gating for GFP fluorescence using a Becton Dickinson FACStar Plus cytometer with Cell Quest software. Enhanced GFP (EGFP) or GFP-ARi-410 expressing cells were seeded in triplicate at 5 x 104 cells/well in 48-well plates. After 24 h (d 0), and at the indicated time points, triplicate samples were harvested by trypsinization. Cell viability was assessed by Trypan blue exclusion, and viable cells were counted using a hemocytometer. The remaining cells from triplicate EGFP and GFP-ARi-410 samples were pooled. Cells were washed in PBS then lysed in RIPA buffer. Samples were analyzed by immunoblot for AR and GFP-ARi-410 expression, using the AR C-19 antibody.
Immunoblotting and Immunoprecipitation
Three AR antibodies were used in the current studies. The N-20 and C-19 rabbit polyclonal antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) were used to detect both wtAR and the ARi variants. The U407 rabbit polyclonal antibody was raised against a peptide corresponding to amino acids 200–220 of the AR (49) and was used to selectively detect wtAR in the confocal studies because this sequence is not present in the ARi-410 or ARi-532 variants. Total proteins (30 µg) extracted from transfected cells were electrophoresed on 7.5% sodium dodecyl sulfate-polyacrylamide gels, transferred to Hybond-C membrane (Amersham, Little Chalfont, Buckinghamshire, UK) and immunostained using the relevant AR antibody. Immunoreactivity was detected using horseradish peroxidase-conjugated secondary antibodies (Silenus; Chemicon International, Temecula, CA) and visualized using ECL reagents (Amersham).
For immunoprecipitation of AR and ARi-410, COS-1 cells (2 x 106) were plated in 10-cm culture dishes and allowed to attach for 16 h. Cells were cotransfected with AR (5 µg) and ARi-410 (5 µg) using LipofectAMINE 2000 according to the manufacturer’s protocol. Five hours after transfection, the media were replaced with phenol red-free RPMI 1640 medium containing 5% dextran-coated charcoal-stripped FCS with 1 nM DHT or equal amount of vehicle (ethanol). Cells were lysed in IGEPAL buffer [100 mM NaCl, 50 mM Tris (pH 8.0), 20 mM HEPES (pH 7.5), 1% Igepal CA-630 and 0.5% sodium deoxycholate]. Lysates were incubated with 5 µg/ml AR-U407 antibody or 5 µg/ml of rabbit IgG at 4 C for 60 min with rotation. Rabbit IgG Dynal beads (Dynal Biotech, Invitrogen, Mount Waverley, Victoria, Australia) were added to the lysates and incubated at 4 C for 60 min with rotation. The Dynal beads were washed with IGEPAL buffer, before solubilizing the protein in IGEPAL buffer and sample buffer, boiling, and centrifugation. Proteins were resolved via SDS-PAGE gel, as described above, and AR species were detected by immunoblotting with the AR-U407 or C-19 antibodies.
Transfections with GFP-AR or GFP-ARi-410 in PC-3 Cells
PC-3 cells (7 x 104 cells/well) were cultured into eight-well chamber slides (Nuclon Lab-Tek II Chamber slide; Naperville, IL) in 600 µl of 5% FBS RPMI and allowed to adhere for 48 h. To investigate the localization of GFP-ARi-410 or GFP-AR individually, cells were transfected with 1 µg/well of plasmid, using LipofectAMINE 2000 reagent (Invitrogen) as per the manufacturers’ instructions with minor modifications. Twenty hours after transfection, media were removed and replaced with 600 µl phenol red-free RPMI containing 5% charcoal-stripped FCS and DHT, vehicle (ethanol) or the indicated ligand. After 24 h incubation with ligand, cells were fixed for 10 min in 4% paraformaldehyde followed by fixation in methanol (–20 C) for 3 min and 1 min in acetone (–20 C) to permeabilize cells. Acetylated histone H3, which indicated the location of the nucleus, was detected using a rabbit polyclonal antibody (Upstate Biotechnology, Lake Placid, NY) and an Alexa 594-conjugated secondary antibody. To examine the colocalization of wtAR and GFP-ARi-410 when the receptors are coexpressed, cells were transfected with equimolar amounts of GFP-ARi-410 and pCMV3.1AR, to a total of 2 µg/well of plasmid, The wtAR was detected using the rabbit polyclonal antibody U407 (which does not detect the ARi), followed by an Alexa 594 conjugated secondary antibody. The slides were mounted with fluorescent mounting medium (DAKO, Carpinteria, CA) and air-dried overnight before confocal microscopy.
Confocal Microscopy and Imaging Analysis
The images of GFP-ARi and GFP-AR or wtAR were produced using a Bio-Rad Radiance 2100 confocal microscope (Bio-Rad Microscience Ltd., Hemel Hempstead, UK) and Olympus IX70 inverted microscope. The GFP was excited with Ar 488-nm laser line and the emission was viewed through a HQ515/30-nm narrow band barrier filter in photomultiplier tube 1 (PMT1). The red fluorescence (Alexa 594) was excited with Green HeNe 543-nm laser line and the emission was viewed through a long pass barrier filter (E600LP) to allow only red light wavelengths longer than 600 nm to pass through PMT2. Automatically, all signals from PMTs 1 and 2 were merged. The image data were analyzed using a Confocal Assistant software program (Todd Clark Brelje, University of Minnesota). The expression and subcellular localization of GFP-ARi, GFP-AR or wtAR after the different treatments were examined in at least 60 cells from four independent experiments. To quantify the subcellular distribution, fluorescent cells were classified into the following categories: completely nuclear (N), nuclear > cytoplasmic (N >C), cytoplasmic > nuclear (C>N), and completely cytoplasmic (C), as described previously (50).
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Current address for P.J.N.: Department of Immunology, Allergy and Arthritis, Flinders Medical Centre, Bedford Park, South Australia 5042, Australia.
Current address for C.S.Y.C.: Department of Pediatric Endocrinology, Princess Margaret Hospital for Children, Western Australia 6008, Australia.
Current address for M.Y.: Department of Microbiology, Institute of Medical and Veterinary Science, Adelaide, South Australia 5000, Australia.
All authors have nothing to declare.
First Published Online January 19, 2006
1 L.M.B., M.M.C., and P.J.N. contributed equally to this work. 
Abbreviations: AAT, Androgen ablation therapy; AF, activation function; AR, androgen receptor; ARi, androgen receptor inhibitor; ARR3-tk-luciferase, the androgen-responsive probasin promoter; DBD, DNA binding domain; DHT, 5-dihydrotestosterone; EGFP, enhanced GFP; FCS, fetal calf serum; GFP, green fluorescent protein; GRIP1, glucocorticoid receptor-interacting protein 1; LBD, ligand binding domain; MMTV, mouse mammary tumor virus; N/C, amino-carboxy terminal; NTD, amino-terminal transactivation domain; PMT, photomultiplier tube; PSA, prostate-specific antigen; SMRT, silencing mediator for retinoid and thyroid receptors; SOE, splicing overlap extension; VDR, vitamin D receptor; wtAR, wild-type AR.
Received for publication October 8, 2004. Accepted for publication January 11, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
