| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Internal Medicine, The University of Texas Southwestern Medical Center, Dallas, Texas 75235
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-dihydrotestosterone. The cDNA nucleotide sequence
predicts a protein approximately 917 amino acids long comprised of a
highly conserved DNA-binding domain, a carboxy-terminal hormone-binding
domain, and a large amino terminus (3, 4, 5, 6). The form of the AR predicted
by the translation of the complete open reading frame (termed AR-B)
migrates with an apparent molecular mass of 110 kDa on
SDS-polyacrylamide gels. During our studies of affected members of a family with complete testicular feminization, we detected a novel shortened form of the human AR (7). Subsequent experiments revealed that this AR isoform [termed AR-A because of its similarity in structure to the A form of the progesterone receptor (PR)] differed from the structure predicted for the B form of the AR in that it lacked an intact amino terminus, yet contained intact DNA- and hormone- binding domains. Further analysis revealed that this form of the AR is expressed in a variety of normal tissues, albeit at low levels, and is derived from initiation of translation at methionine-188 instead of methionine-1 of the AR open reading frame (8, 9).
The detection of two distinct forms of the AR raised a number of issues. First, since the original family in which the AR-A form of the receptor was detected exhibited a phenotype of complete testicular feminization, it was not clear whether this defect of androgen action was caused by the reduced levels of receptor expressed or to the impaired function of this form of the AR (7). Second, as noted above, the structures of the A and B forms of the AR bear a striking resemblance to the more fully characterized A and B forms of the PR. A number of studies have suggested that the A and B forms of the PR exhibit distinct differences of function, with respect to the target genes that they regulate, the ligands to which they respond, and the way in which they interact with target DNA sequences (10, 11, 12, 13, 14, 15). The similarities evident between the isoforms of the AR and PR suggested that specific roles might exist for the AR-A and AR-B isoforms in the regulation of androgen-responsive genes as well.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). The effect of this change was
assessed by measuring the level of expression in transfected cells
using antibodies that recognize both receptor isoforms (amino acid
residues 200–220 in AR-B). As shown in Fig. 2, transfection of the resulting
expression plasmid (PS367) into COS cells resulted in the synthesis of
levels of immunoreactive AR equivalent to those observed after
transfection with a cDNA encoding the full-length AR (CMV 3.1).
|
|
) revealed substantial differences
in the dose-response curves of the two isoforms. After transfection
with low concentrations of cDNAs encoding either isoform, increasing
activation of the reporter gene was observed. Although in both
instances the maximum stimulation of the reporter gene was achieved
between 0.1 µg and 0.4 µg of AR cDNA, at the region of peak
activity AR-A exhibits approximately half the level of reporter gene
activity measured for cells transfected with the AR-B cDNA (Fig. 3A).
This result suggests that the amino-terminal segment (amino acid
residues 1–188) that is not present in the AR-A isoform is important
for the formation of receptor complexes capable of stimulating maximal
activation of the MMTV promoter in CV1 cells.
|
). This
result suggests that the two AR isoforms interact with components of
the transcription apparatus in a fashion that is promoter-specific.
Experiments conducted transfecting cDNAs encoding both isoforms
demonstrated no evidence that the AR-A isoform acted in a dominant
negative fashion (data not shown).
Cell- and Promoter-Specific Behavior of AR Isoforms in Stimulating
the Activity of Androgen-Responsive Reporter Genes
To determine how the A and B isoforms would function in different
cell types to activate these model androgen-responsive reporter genes,
we transfected 200 ng of AR-A and AR-B cDNA in parallel with either the
MMTV or PRE2-tk reporter plasmids into the CV1, DU145, and
PPC1 cell lines and assayed reporter gene activity after incubation
with 2 nM DHT in each experiment. This protocol was chosen
as this amount of AR cDNA induced the activity of the
androgen-responsive genes by both isoforms to maximal levels (Fig. 3).
When CV1 cells were transfected with cDNAs encoding the AR
isoforms and the MMTV promoter reporter plasmids, AR-A exhibited
approximately half of the level of activity observed in transfections
using the AR-B cDNA. These results correspond to those obtained in the
dose response curves presented in Fig. 3A. Although the measurements of
AR function in the different cell lines cannot be compared directly, it
is clear that similar patterns in the relative activities of the A and
B isoforms are seen when AR function is measured in the DU145 and PPC1
cell lines using the MMTV reporter, as when performed in the CV1 cell
line. In each instance, the B form of the AR is approximately twice as
active in stimulating reporter gene activity as the A form assayed in
parallel (Fig. 4A).
|
, results consistent with those depicted in Fig. 3B).
When assayed in the DU145 cell line, the two isoforms exhibited similar
capacities to activate the reporter gene. In the PPC1 cell line, the
level of reporter gene activation achieved by AR-A was lower than that
of AR-B. These differences suggest that the AR-A and AR-B
hormone-receptor complexes interact with the PRE2-tk
promoter in a fashion that varies among cell types. Despite these
differences, each reporter gene used displayed similar levels of
activation in response to saturating concentrations of DHT and
testosterone (data not shown).
AR-A and AR-B Cannot Be Distinguished by Antiandrogens in
Cotransfection Assay
The A and B forms of the PR have functions that can be
distinguished pharmacologically. To examine whether this was true for
the A and B AR isoforms, we examined the response of these isoforms to
different AR agonists and antagonists. These assays were performed by
transfecting the AR-A and AR-B cDNAs in parallel into CV1 cells in
combination with the MMTV-luciferase reporter gene and treating the
transfected cells with varying concentrations of antiandrogens alone or
in combination with saturating concentrations of 5-DHT. The results
that we obtained are shown in Fig. 5 and
are intriguing in several respects.
|
). In these experiments,
the inclusion of 0.5 µM hydroxyflutamide inhibited the
reporter gene activity induced by saturating concentrations of 5-DHT
by more than 90%. Additional inhibition of AR function was not evident
when the concentration of hydroxyflutamide was increased further.
Instead, as the concentration of hydroxyflutamide was increased,
increasing levels of reporter gene activation were observed. When these
experiments were performed using hydroxyflutamide alone, it was
possible to demonstrate that the agonism observed at the higher
hydroxyflutamide concentrations was a function of the properties of
hydroxyflutamide itself.
When this same type of experiment was repeated using flutamide, a
different result was observed. Flutamide showed less potent capacity to
antagonize DHT action than hydroxyflutamide, requiring concentrations
of 2 µM to achieve a 90% inhibition of DHT-stimulated AR
function. Of interest, as the concentrations of flutamide were
increased (from 0.5 to 2 µM), only a progressive
inhibition of AR function was observed. In keeping with this
observation, flutamide was devoid of any intrinsic agonism in
functional assays when added alone to the transfected cells (Fig. 5B).
Experiments in which side-by-side comparisons were performed confirmed
this finding (Fig. 5C).
The findings presented above are not confined to hydroxyflutamide, and
other high-affinity AR antagonists—even those that have been suggested
to be pure antiandrogens—exhibit a mixture of agonist and antagonist
properties. A representative experiment is shown in Fig. 5D for Casodex
(bicalutamide). At low concentrations, Casodex exhibits a potent
capacity to antagonize the activation of the AR-B by 2 nM
5-DHT. As was observed with hydroxyflutamide, however, as the
concentration of Casodex is increased, increasing agonism becomes
evident. As shown in Table 1, the AR
antagonists demonstrate similar behaviors when assayed in the PPC-1 and
DU145 cell lines.
|
, A, B, and D, the activity of
AR-A isoform was measured in parallel with those of the AR-B isoform.
In each instance, the patterns of response observed for the AR-B and
AR-A isoforms were similar.
Activation of the AR Isoforms Is Not Always Accompanied by
Increased Levels of Immunoreactive AR
Androgen causes the levels of AR to increase when assayed by
ligand binding (17) or by immunoblot assays (18), and this increase
appears to result from a change in the half-life of the receptor
protein. To determine whether the agonism observed for the AR
antagonists (such as hydroxyflutamide and Casodex) also caused changes
in the levels of AR, we performed immunoblot assays of cells
transfected with cDNAs encoding the different isoforms after treatment
with either no hormone or with the different antagonists. The results
of these experiments are shown in Fig. 6
and are interesting in two respects. The first is that the level of AR
does not change in cells transfected with AR-B and treated with AR
antagonists, even under circumstances in which substantial agonism is
observed (e.g. at 5 µM hydroxyflutamide).
These experiments also show, unexpectedly, that the increase in the
level of AR-B that is observed after treatment with androgen agonists,
such as mibolerone, is not observed in cells transfected with the AR-A
isoform. This finding suggests that the amino- terminal segment of AR-B
is critical for the increase in receptor levels after treatment with
ligand.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
In our prior studies, a patient with complete testicular feminization expressed only a shortened form of AR (AR-A), which appeared to be derived from initiation of translation at methionine-188 (7). The low and relatively constant level of AR-A that was detected in the fibroblasts of this patient was also found to be present in the fibroblasts of normal subjects and in most target tissues (8). The unfavorable context surrounding the AUG triplet encoding methionine-188 appeared to account for the low level of expression in vivo and in transfection studies and complicated studies to analyze the functional capacity of the two isoforms. For this reason, we constructed an AR-A expression vector (PS367) in which the sequence upstream of methionine-1 was introduced 5' to methionine-188. As the result of these changes, the expression levels between AR-A and AR-B were similar and permitted a more thorough analysis of their functional activities.
When assayed in CV1 cells, AR-A and AR-B showed different dose-response curves in transfection assays using an MMTV-luciferase reporter gene fusion to measure receptor function. The activities of both isoforms were maximal at low concentrations of transfected cDNA, and decreasing levels of reporter gene activity were observed as increasing amount of cDNA are added. Of note, at the level where maximal activation was achieved, the AR-B isoform was approximately twice as active as the AR-A isoform. When such assays are repeated using a reporter gene controlled by the PRE2-tk promoter in CV1 cells, the level of induction by AR-B was half that stimulated by AR-A. This finding suggests that the amino-terminal segment of the AR (amino acid residues 1–187) interacts differently with the proteins that regulate transcription from these two promoters. These inferences are further supported by the results of assays using these same reporter genes in different cell lines. In these experiments, while the activity of the AR-B was higher using the MMTV promoter in all of the cell lines, the PRE2-tk promoter gene fusion was activated to varying levels by the AR-A in the DU145, PPC1, and CV1 cell lines. The demonstration that the amino terminus confers upon the AR-B isoform activities not exhibited by the AR-A isoform is consistent with studies that have localized a distinct transactivation domain to the corresponding segment of the human PR (22).
In some instances, isoforms of nuclear receptors have demonstrated differential responses to pharmacological manipulations (10, 11, 12). For this reason, we explored the effects of a variety of AR antagonists on AR function. In these experiments, using a reporter gene controlled by the MMTV promoter in CV1 cells, a variety of antiandrogens failed to display any major qualitative differences in the activities of AR isoforms. In each instance the results of assays using AR-A were similar to those obtained using AR-B.
While the results of these functional studies did not reveal unique activities of the AR isoforms, the properties exhibited by the different antagonists in these experiments were intriguing. First, it is evident that of the compounds tested, only flutamide failed to exhibit agonism in our assays. While low concentrations of each of the higher affinity antagonists (Casodex, hydroxyflutamide) effectively blocked the stimulation of the AR by saturating concentrations of agonist (DHT), increasing agonism was evident as the concentrations of these AR antagonists were further increased. Of note, our observations differ somewhat from the findings of other groups who have examined the activities of hydroxyflutamide and Casodex in reporter gene assays. In two studies, these compounds exhibited little agonistic activity and appeared to function as pure antagonists (23, 24). By contrast, Kemppainen and Wilson (25) noted the agonism inherent in the behaviors of hydroxyflutamide and Casodex. The results reported in the present study are in best agreement with these latter investigations. The reasons for the discordances are not clear, but may be explained in part by the reporter genes used, by the concentrations of antagonist assayed, or by the specific cell types employed.
The mechanisms by which ligands for nuclear receptors exhibit both agonism and antagonism are not clear. Molecules exhibiting such a mixture of activities have been identified that act to antagonize the actions of ligands for many members of the nuclear receptor family, including androgens, estrogens, and progestins (12, 24, 25, 26, 27, 28). Detailed studies of the activities of such compounds have suggested models in which the receptor assumes distinct conformations that permit the liganded receptor to interact with the transcription apparatus in a productive or nonproductive fashion (29, 30). The degree of agonism observed has been found to vary substantially, depending on cell type and promoter context. In the case of the ER, the agonistic properties of some antiestrogens have been correlated with the activity of a specific transactivation domain (TAF-1) located within the amino terminus of the receptor (27, 28, 31). Recent work suggests that the levels and capacity of the liganded receptor to recruit specific coactivators and repressors to the transcription complex are important determinants of the degree of agonism or antagonism that is observed (32, 33).
While such models provide a framework with which to explain the tissue- and cell type-specific behaviors of some antagonists, they do not provide a ready explanation for one aspect of the behaviors of the AR antagonists studied in the current work. Namely, these models do not offer a rationale for the emergence of agonistic behavior of some AR antagonists at concentrations well above that required to inhibit the activation of the AR by saturating concentrations of DHT. Such observations could be reconciled by a contrasting view of the actions of steroid hormone receptor antagonists, such as that afforded by Jensen and co-workers (34, 35). In this view, a two-site model has been proposed to assist in accounting for the mixed agonist-antagonist properties displayed by some receptor antagonists. In such a model, in addition to competing for the cognate hormone-binding site, antihormones react with a second domain in the receptor, which is not recognized by cognate ligand and which plays a role in antagonist action (34, 35). Such a model would consider the difference between antihormones of different classes to result from their different relative affinities for the two binding regions. While such models have not been experimentally tested extensively, differences in the number of binding sites have been detected using [3H]tamoxifen and estradiol as ligands in binding assays. In a similar vein, the demonstration that a mutant PR deleted for the receptor carboxy terminus is able to bind and respond to RU486 but not to progesterone is also consistent with the existence of separable sites for the binding of progesterone antagonists and agonists within the PR hormone-binding domain (36). Finally, with respect to the mechanisms by which AR antagonists function, the existence of additional binding sites would offer a ready explanation for the appearance of agonism at concentrations of antagonist well above that needed to interrupt the action of AR agonists.
Flutamide is rapidly and extensively metabolized in vivo to hydroxyflutamide by hydroxylation. While both molecules have been shown to block the effects of androgens in target tissues (37, 38, 39, 40), flutamide has been estimated to do so with a 25-fold lower affinity relative to hydroxyflutamide for AR in vitro (41). For this reason, subsequent attempts to develop higher affinity AR antagonists have used hydroxyflutamide as a starting point. Our results suggest that antagonists with such structures (particularly with larger side chain substituents) may possess, by their very nature, a certain degree of inherent agonism that is unmasked at higher concentrations. It is conceivable that increased formation or accumulation of metabolites (such as hydroxyflutamide) within cells might contribute to the apparent resistance of prostate cancers that occurs in patients treated with such compounds. Such a mechanism would offer an alternative explanation for the flutamide withdrawal syndrome, particularly in those cases in which an AR mutation is not identified.
Finally, the addition of ligand to cultured cells has been shown to result in an increase in the level of AR, measured either using ligand-binding assays (17) or Western blotting (18). Previous studies have demonstrated that this effect is posttranslational and requires both an intact ligand-binding domain and the amino-terminal segment of the receptor molecule. The studies of Zhou et al. concluded that an interaction of the amino and carboxy termini of the protein were required to observe a ligand-induced stabilization of the receptor (42). These inferences were further supported by the results of immunoblot assays that measured the levels of expression of the intact AR and a number of mutant ARs containing deletions of varying sizes within the amino terminus (43). The present work suggests that amino acids within the most amino-terminal segments of the receptor participate directly in these interactions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-Dihydrotestosterone was purchased from Steraloids,
Inc. (Wilton, NH).
Site-Directed Mutagenesis
The segment 5' to methionine-188 in the original plasmid
encoding the A-form of the AR [CMV776 (Ref.7)] was modified by
replacing it with the sequence that is 5' to methionine-1 in the AR
expression plasmid CMV3.1. This mutagenesis was accomplished by using
two oligonucleotides (AR-As:
5'-ACACAGATCTAGGTGGAAGATTCAGCCAAGCTCAAGGATGCAACTCCTTCAGCAACAGCAGCAGGAA-3'
and AR-Aas: 5'-GGCTGAGGGTGACCCAGAACCGGGT-3') as
primers and the plasmid CMV3.1 (6) as template in a PCR reaction. The
resulting PCR fragment, encoding the methionine-188 as the initiator
methionine [i.e. deleting nucleotides 163–723 of the AR
cDNA sequence (6)], was digested with the restriction enzymes
NcoI and BglII and ligated into the 0.6-kb
fragment purified after digestion of a sample of the CMV 3.1 hAR
expression plasmid digested with the same restriction endonucleases.
Vector fragments derived from the expression vector (CMV3.1) cleaved
with the same enzymes. As shown in Fig. 1, the resulting plasmid
(designated PS367) encodes a protein that is identical to hAR-B, except
it lacked the segment encoding amino acids 1–187 of the hAR open
reading frame. In this environment, the initiator methionine of the
hAR-B (the codon encoding methionine-188 in CMV 3.1) is predicted to be
a much better fit to the Kozak consensus sequence for translation
initiation (16).
Cell Culture and Transient Transfection Assays
Stock cultures of the CV1 (monkey kidney fibroblast-like cell
line) and DU145 cell lines (a metastatic human prostate carcinoma
cell line) were obtained from the American Type Culture Collection and
were maintained in MEM (GIBCO/BRL, Gaithersburg, MD) containing 10%
(vol/vol) FCS and 1% penicillin and streptomycin. PPC1 cell line
(considered to be a variant of PC-3 cell line) was obtained from Arthur
Brothman (Salt Lake City, UT) and maintained in RPMI 1640
containing 10% (vol/vol) FCS and 1% penicillin and
streptomycin.
Transfection assays were performed as described (43). The day before transfection, cells were trypsinized and plated at a density of 2 x 105 cells per well in six-well plates (each well 35 mM diameter) for reporter gene assays of AR function and a density of 1 x 106 cells per 10-cm dish for immunoblot analyses. Each six-well plate or 10-cm dish was transiently transfected by the addition of calcium phosphate precipitate containing the indicated concentrations of hAR isoform expression plasmid, the androgen-responsive reporter plasmid (10 µg), and 1 µg of a control plasmid (CMV-ß-galactosidase) in 12 ml (for six-well plates) or 10 ml (for each 10-cm dish) of culture medium for 24 h. After these incubations, the medium was removed and replaced with fresh MEM containing 5% charcoal-stripped serum alone or containing the various ligands. Forty-eight hours later, the cell cultures were harvested and assayed for luciferase activity and ß-galactosidase activity or harvested for immunoblot assays. In each experiment, measurements of the function of individual ARs were assessed in at least three separate transfections (wells) for each in the absence or presence of hormone. The results of these individual measurements were averaged and compared with the results obtained using the AR-B included in each experiment. In the text and legends, the functional assay results are presented either as stimulated luciferase values or as fold induction. This latter value is calculated by dividing the stimulated luciferase values by the basal luciferase values.
Immunoblots
Immunoblots were prepared as previously described. After
transfer, the filters were incubated with affinity-purified antibody
(anti-Internal A antibody) from rabbit U407 (8) that recognizes amino
acids 200–220 of the hAR protein, which is preserved in both the AR-A
and AR-B isoforms.
|
FOOTNOTES |
|---|
This work was supported by NIH Grants DK-03892 and 47657, a grant from the Robert A. Welch foundation (I-1090), and a grant from the Perot Family Foundation.
Received for publication July 14, 1997. Revision received January 2, 1998. Accepted for publication January 29, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Aquila, E. Middea, S. Catalano, S. Marsico, M. Lanzino, I. Casaburi, I. Barone, R. Bruno, S. Zupo, and S. Ando Human sperm express a functional androgen receptor: effects on PI3K/AKT pathway Hum. Reprod., October 1, 2007; 22(10): 2594 - 2605. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Z. Lu, S. E. Wardell, K. L. Burnstein, D. Defranco, P. J. Fuller, V. Giguere, R. B. Hochberg, L. McKay, J.-M. Renoir, N. L. Weigel, et al. International Union of Pharmacology. LXV. The Pharmacology and Classification of the Nuclear Receptor Superfamily: Glucocorticoid, Mineralocorticoid, Progesterone, and Androgen Receptors Pharmacol. Rev., December 1, 2006; 58(4): 782 - 797. [Full Text] [PDF]
|
|
|
|
 |
|
 J. W. Gatson, P. Kaur, and M. Singh Dihydrotestosterone Differentially Modulates the Mitogen-Activated Protein Kinase and the Phosphoinositide 3-Kinase/Akt Pathways through the Nuclear and Novel Membrane Androgen Receptor in C6 Cells Endocrinology, April 1, 2006; 147(4): 2028 - 2034. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Servert, J. Garcia-Castro, V. Diaz, D. Lucas, M. A. Gonzalez, C. Martinez-A, and A. Bernad Inducible model for {beta}-six-mediated site-specific recombination in mammalian cells Nucleic Acids Res., January 3, 2006; 34(1): e1 - e1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Buchanan, S. N. Birrell, A. A. Peters, T. Bianco-Miotto, K. Ramsay, E. J. Cops, M. Yang, J. M. Harris, H. A. Simila, N. L. Moore, et al. Decreased Androgen Receptor Levels and Receptor Function in Breast Cancer Contribute to the Failure of Response to Medroxyprogesterone Acetate Cancer Res., September 15, 2005; 65(18): 8487 - 8496. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Modugno Ovarian Cancer and Polymorphisms in the Androgen and Progesterone Receptor Genes: A HuGE Review Am. J. Epidemiol., February 15, 2004; 159(4): 319 - 335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Brenowitz and K. Lent Act locally and think globally: Intracerebral testosterone implants induce seasonal-like growth of adult avian song control circuits PNAS, September 17, 2002; 99(19): 12421 - 12426. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Qi, C. Fillion, Y. Labrie, J. Grenier, A. Fournier, L. Berger, M. El-Alfy, and C. Labrie AIbZIP, a Novel bZIP Gene Located on Chromosome 1q21.3 That Is Highly Expressed in Prostate Tumors and of Which the Expression Is Up-Regulated by Androgens in LNCaP Human Prostate Cancer Cells Cancer Res., February 1, 2002; 62(3): 721 - 733. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. McPhaul Molecular Defects of the Androgen Receptor Recent Prog. Horm. Res., January 1, 2002; 57(1): 181 - 194. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Vermeulen Androgen Replacement Therapy in the Aging Male--A Critical Evaluation J. Clin. Endocrinol. Metab., June 1, 2001; 86(6): 2380 - 2390. [Full Text] [PDF]
|
|
|
|
|
|
M. J. McPhaul and J. E. Griffin Male Pseudohermaphroditism Caused by Mutations of the Human Androgen Receptor J. Clin. Endocrinol. Metab., October 1, 1999; 84(10): 3435 - 3441. [Full Text] [PDF]
|
|
|
|
|
|
Y.-S. Zhu, L.-Q. Cai, J. J. Cordero, W. J. Canovatchel, M. D. Katz, and J. Imperato-McGinley A Novel Mutation in the CAG Triplet Region of Exon 1 of Androgen Receptor Gene Causes Complete Androgen Insensitivity Syndrome in a Large Kindred J. Clin. Endocrinol. Metab., May 1, 1999; 84(5): 1590 - 1594. [Abstract] [Full Text]
|
|
|
|
|
|
T. S. Sperry and P. Thomas Characterization of Two Nuclear Androgen Receptors in Atlantic Croaker: Comparison of Their Biochemical Properties and Binding Specificities Endocrinology, April 1, 1999; 140(4): 1602 - 1611. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE |
100 µg).