:: Asian Journal of Andrology ::

Androgen receptor isoforms in human and rat prostate

^{Shu-Jie
XIA1, Gang-Yao HAO2, Xiao-Da TANG1}

^{1University
Department of Urology, Shanghai First People's Hospital, Fud}^{an
University, Shanghai,
200080, China}^{2Department of Urology, Second Affiliated}^{Hospital,
Shandong University, Jinan, 250033 , China}

Asian J Androl 2000 Dec; 2: 307-310

Keywords: androgen receptors; isoforms; prostate

Abstract

Aim: To investigate the androgen receptor (AR) isoforms and its variability of expression in human and rat prostatic tissues. Methods: Human benign prostatic hyperplasia (BPH) and prostatic cancer tissues were obtained from patients undergoing prostatectomy, and rat ventral prostate was incised 3 days after castration. Forty-one AR-positive BPH specimens, 3 prostatic cancer specimens, and 6 rat prostates were used. After processing at 4 , the tissues were examined by means of high resolution isoelectric focusing (IEF) technique to determine their AR isoforms. Results: From the prostatic specimens, 3 types of AR isoforms were detected with pI values at 6.5, 6.0, and 5.3. In human BPH tissues, 15/41 (36.6%) specimens showed all the three types of isoforms, while 19/41 (46.3%) showed 2 isoforms at various combinations and 7/41 (17.1%), 1 isoform. For the 3 prostatic cancer specimens, one showed 3 isoforms, one, 2 isoforms, and the other failed to show any isoform. All rat prostatic tissues showed 2 isoforms at different combinations. Binding of ³H-dihydrotestosterone (DHT) to the isoforms was inhibited by the addition of l00-fold excess of DHT or testosterone, but not progesterone, oestradiol or diethylstilboestrol. Conclusion: AR isoforms are different in different patients. Although their genesis is not clear, the therapeutic implication of the present observation appears to be interesting, that may help clarifying the individual differences in the response to hormonal therapy.

1 Introduction

Androgens (testosterone and 5 -dihydrotestosterone) play a crucial role in prostatic development and diseases^[1-3]. The androgen receptor (AR) mediates the actions of male sex steroids. Since the publication of an AR cDNA sequence from human prostate^[4]it has been acknowledged that androgen receptor is encoded by a single gene. However, there is evidence that the AR protein itself is heterogenous^[5,6] and a human prostate carcinoma cell line LNCaP can express an AR that binds testosterone (T) and R1881, resulting in ligands both capable of stimulating cells in culture.

Recently, two distinct clones of AR cDNA have been separated from rainbow trout and Japanese eel testis^[7,8] and two types of AR proteins from fibroblasts of human genital skin^[9].

Up to date, the hormonal therapy for both benign prostatic hyperplasia (BPH) and prostatic carcinoma is either by inhibiting androgen binding to its receptor or suppressing testosterone conversion to the more active androgen, DHT, within the prostate cells by 5-reductase inhibition. The latter form of treatment has been developed on the basis of the assumption that the androgen receptor is the only binding protein and that testosterone has little direct bioactivity on prostate cells. Clinicians have already disclosed the difference in the response to hormonal therapy in different patients suffering from BPH or prostatic cancer. Therefore, the relationship between the prostatic AR content in BPH and prostatic cancer and the response to endocrine therapy has been paid more attention to. Since only human beings and dogs are susceptible to BPH in mammals^[10], we have taken rat prostate as the control in the study.

This report describes studies performed on the AR-positive prostates of human beings and rats to investigate the existence of AR isoforms and to observe their different expressions.

2 Materials and methods

2.1 Tissue handling

Human benign prostatic hyperplasia (BPH) and prostatic cancer tissues were obtained from patients undergoing prostatectomy and were immediately stored in liquid nitrogen until processed. The ventral prostate of rats was used 3 days after castration. All tissue processing was performed at 4. The tissue was sliced and homogenized with!a polytron homogeniser in glycerol phosphate buffer (10% glycerol, 10 mmol/L phosphate, 1.5 mmol/L EDTA, 5 mmol/L monothioglycerol, pH 7.4) containing l ng/mL each of the protease inhibitors, aprotinin and soybean trypsin inhibitor (both from Sigma, USA). The homogenate was centrifuged for l0 min at 800g and the supernatant was further centrifuged for 60 min at 100,000g. The final supernatant was used for receptor analysis.

2.2 Dextran coated charcoal (DCC) assay of AR

ARs were measured by the single saturating dose (SSD) assay. Cytosol (3 mg-protein/mL cytosol) was incubated with 50 nmol/L tritiated DHT in the presence or absence of a l00-fold excess of unlabelled DHT, testosterone, oestradiol or diethylstilboestrol. These incubations were carried out at 4 for 24 h. Free hormone was separated from the bound by incubation with DCC buffer (0.5% w/v) charcoal and 0.05% (w/v) dextran T70 (Sigma Chemicals Ltd.) at 4 followed by centrifugation at 10,000g for 5 min. An aliquot of the supernatant was counted in a liquid scintillation counter. A further aliquot of these samples was taken for isoelectric focusing (IEF) analysis.

2.3 Isoelectric focusing

The IEF gels were cast in slabs of size 125260 mm and separation was conducted along the short side of the gel. Polyacrylamide gels (2 mm thick) contained 20% (v/v) glycerol. A pH gradient was achieved using 1.5% (w/v) LKB ampholine pH 3.5-10.7 (LKB, Bromma, Sweden) and l% (w/v) LKB ampholine pH 5-8. Gels were photopolymerized for at least 4 hours at room temperature by means of a TR 26 polymerization light, using riboflavin (0.004%, v/v). IEF was performed in a cold room and temperature of the cooling water was kept constant at 4. Electrode solution of 1mol/L NaOH (cathode) and 1 mol/L H₂SO₄ (anode) were used. Gels were pre-focused for 45 min at 20 mA/20 W/l200 V.

After DCC extraction, aliquots (150 L) of the radioactive (³H-DHT labelled) supernatants derived from SSD assay were loaded near the cathode. The runs were carried out for l.5 h, using a 3000xi CC power supply (LKB, system, Sweden) at 1200 V/20 mA/20 W constant power. A mixture of nine natural proteins (Bio-Rad) was used for pH calibration. After the run, the gels were cut into 2.5 mm slices and each slice was incubated with 5 mL scintillation cocktail (Ready-solv, Beckman, USA) for 24 hours at room temperature and the radioactivity was assayed.

The ligand specificity of these isoforms was confirmed by performing IEF on samples that had been incubated with labelled DHT and a 100-fold excess of unlabelled DHT, T, oestradiol or diethylstilboestrol. Unlabelled DHT and T competed with labelled DHT for binding all the three isoforms. Oestradiol and diethylstilboestrol had no effects on the isoform profiles obtained.

2.4 Protein determination

Proteins were determined by the method of Lowry, using BSA as the standard.

3 Results

Three types of AR isoforms were detected in prostatic specimens with three radioactive peaks on the gels focusing at pI 6.5, 6.0, and 5.3. In human BPH tissues, 15/41 (36.6%) specimens showed all the three types of isoforms, while 4/41(9.8%), 2 isoforms at 6.0 and 5.3, 10/41 (24.4%), 2 isoforms at 6.5 and 5.3, 5/41 (12.2), 2 isoforms at 6.5 and 6.0, 3/41 (7.3%), 1 isoform at 5.3, 2/41 (4.9%), 1 isoform at 6.0, and 2/41 (4.9%), 1 isoform at 6.5. For the 3 prostatic cancer specimens, one showed 3 isoforms, one, 2 isoforms at 6.5 and 6.0, and the other failed to show any isoform. Three (50%) rat prostatic tissues showed isoforms at 6.5, 6.0, and the other 3 (50%), at 6.0 and 5.3.

Binding of ³H-DHT to these three types of isoforms was inhibited by the addition of l00-fold excess of DHT and testosterone, but not progesterone, oestradiol and diethylstilboestrol. The data are summarized in Table l.

Table 1. AR isoforms of human and rat prostate.

Samples	pI			Occurrence (%)
Samples	6.5	6.0	5.3	Occurrence (%)
BPH	+	+	+	15/41 (36.5)
	+		+	10/41 (24.4)
		+	+	4/41 (9.8)
	+	+		5/41 (12.2)
	+			2/41 (4.9)
		+		2/41 (4.9)
			+	3/41 (7.3)
Prostatic cancer	+	+	+	1/3
	+	+		1/3
				1/3
Rat prostate	+	+		3/6 (50)
Rat prostate	+		+	3/6 (50)

In addition, two acidic non-specific binding proteins (i.e. non-displaceable steroid binding) focused at pI 4.0 and pI 5 .0 were found in all the specimens.

4 Discussion

The isoelectric point of the cytoplasmic androgen receptors obtained from human BPH and prostatic cancer specimens were found to be acidic (pH 6.5, 6.0 and 5.3). Previous data by Auf and Ghanadian^[11] using a synthetic ligand suggested that the pI of the androgen receptor isolated from BPH samples was 6.2. The reported value for AR of the rat ventral prostate is 5.8. The peak at pI 6.5 has not been reported before. It is we1l documented that BPH tissue homogenates contain high levels of sex hormone binding globulin (SHBG) that makes androgen receptor assay difficult when DHT is used as the ligand. Our data indicates that the peak at pI 6.5 is not due to steroid binding to SHBG, since neither oestradiol nor diethylstilboestrol could compete with labelled DHT for displacing binding from this peak. In fact, data presented here suggest that the non-displaceable steroid binding peak at pI 5.0 represents the SHBG. Puddefoot et al^[12] and Marsigliante et al^[13] also observed the same non-specific binding peak, when human breast tumor cytosol was labelled with ³H-oestradiol and subjected to separation by IEF. It was reported that androgen receptor is coded by single gene^[4,14,15]. It may be possible that the AR isoforms mentioned above arise as a consequence of protein modification or phosphorylation^[6]. However, more recently Takeo and Yamashita^[7] and Ikeuchi et al^[8] isolated two types of cDNA clones from rainbow trout and Japanese eel testis, and both of the cDNA could encode proteins. In fact, it has been confirmed that the translation of mRNA transcribed from AR cDNAs yielded 94- and 76-kDa proteins and smaller forms in a protein synthesis system of rabbit reticulocyte lysate that could bind to DNA and had high affinity towards androgens. The predicted molecular weight of the androgen receptor is approximately 90-95 kDa^[14], as confirmed by SDS-PAGE analysis of photoaffinity labelled receptor^[16]. Interestingly the AR expressed in LNCaP cells bind T and R-l88l, resulting in both ligands being capable of stimulating cell growth in culture. The androgen receptor heterogeneity observed in these cells is not due to mutation, because the same micro-heterogeneity (110-112 kDa doublet) is also found after expression of wild type androgen receptor protein in COS-l cell^[6]. Our data clearly showed that there were three isoforms in human prostatic tissues. The genetic basis for human prostate AR isoforms needs further studies.

Both human and dog prostates grow with aging^[1,2] and both men and dogs are susceptible to BPH^[10]. Therefore, we took rat prostate as the control. Expression of AR isoforms in rat prostates was similar to that of human prostate, i.e., pI at 6.5, 6.0, and 5.4. This phenomenon may be explained by the autoploidy of the cDNA structure between human being and rat^[14].

With reference to the significance of these three AR isoforms observed in this study, Sheridan^[17] has suggested that there may be two receptor proteins for T and DHT, since inhibition of 5-alpha reductase activity in the developing rat leads to the development of external female genitalia, yet the epididymis, vas deferentia and seminal vesicles are present, indicating T dependency. In humans, a rare group of genetic males with a defect in 5-alpha reductase activity have been noted to develop incomplete differentiation of the male external genitalia and prostate gland, whilst retaining normal development of Wolffian duct derivatives which undergo masculinization at puberty. In these subjects, however, the prostate remains rudimentary and BPH does not occur^[18]. Furthermore a study of castrated male rats implanted with the Dunning rat prostatic carcinoma R-3327 treated with testosterone and the 5-reductase inhibitor finasteride (MK906), showed that 5-reductase inhibition did not influence tumor growth. This suggests that the R-3327 tumor responds to testosterone directly^[19]. The androgen receptor binds to both T and DHT, but why T cannot similarly substitute for DHT to induce development of the prostate gland and male external genitalia? Indeed, Geller has demonstrated that intraprostatic testosterone is significantly elevated following treatment with potent 5 alpha reductase inhibitor, finasteride. It was definitely observed in our studies that both T and DHT are able to bind to all three isoforms thus it is possible that the differential tissue response to T and DHT may be a consequence of the ability of tissues to express these different isoforms of AR.

It has been reported that oestrogen receptor (ER) is encoded by a single gene. The ER in human breast cancer samples is heterogeneous, existing as four isoforms whose expression varies between tumors, of which expression or lack responds to endocrine treatment^[20,21]. The observations about ER indicate that there maybe a relationship between the lack and expression of AR isoforms and endocrine treatment. The data presented in this study suggests that several isoforms of AR can be expressed and the expression of isoforms can differ from patient to patient. One prostatic cancer specimen failed to express any isoform. The clinical significance of the different expression of AR isoforms remains to be resolved until more normal prostatic tissue AR isoforms are analyzed. But the therapeutic implications of the observations in this study are important, especially in the rapidly expanding field of pharmacotherapy for treatment of human benign and malignant prostatic diseases. The different expression of AR isoforms in prostatic tissues may explain the reason why the patients have different response to hormone therapy. The prostatic growth pattern is age and androgen dependent^[1-2] therefore different age group's prostate samples should be studied. Theoretically, synthetic steroids which can block binding to one or all types of AR isoformsmay be capable of completely inhibiting androgenic stimulation of benign and malignant prostatic cell growth without the induction of more generalized symptoms of hypogonadism.

References

[1] Xia SJ, Teng JB, Xu CX, Ma QZ, Wang Q. The human prostatic growth pattern with age -1305 cases report. Chin J Urol 1998; 19: 230.
[2] Xia SJ, Teng JB, Wang XL, Xu CX, Wang Q, Ma BL. The relationship of salivary testosterone level and human prostatic growth with aging. Chin J Androl 1999; 13: 214-7.
[3] Brinkmann AO, Jenster G, Kuipter GG. The human androgen receptor: structure/function relationship in normal and pathological situations. J Steroid Biochem Mol Biol 1992; 4l: 36l-8.
[4] Lubahn DB, Joseph DR, Sullivan PM, Willard HF, French FS, Wilson EW. Cloning of human androgen receptor complementary DNA and localization to the X chromosome. Science 1988; 240: 327-30.
[5] van Laar JH, Bolt-de Vries J, Zegers ND, Trapman J, Brinkmann AO. Androgen receptor heterogeneity and phosphorylation in human LNCaP cells. Biochem Biophys Res Commun 1990; 166: 193-200.
[6] Kuiper GG, de Ruiter PE, Brinkmann AO. Androgen receptor heterogeneity in LNCaP cells is caused by a hormone independent phosphorylation step. J Steroid Biochem Mol Biol 1992; 41: 697-700.
[7] Takeo J, Yamashita. Two distinct isoforms of cDNA encoding rainbow trout androgen receptors. J Biol Chem 1999; 274: 5674-80.
[8] Ikeuchi T, Todo T, Kobayshi T, Kobayashi T, Nagahama Y. cDNA cloning of a novel androgen receptor subtype. J Biol Chem 1999; 274: 25205-9.
[9] Wilson CM, McPhaul MJ. A and B forms of the androgen receptor are present in human genital skin fibroblasts. Proc Natl Acad Sci USA 1993; 91: 1234-8.
[10] Pierrepoint CG, Davies P, Lewis MH, Moffat DB. Examination of the hypothesis that a direct control system exists for the prostate and seminal vesicles. J Reprod Fertil 1975; 44: 395-409.
[11] Auf G, Ghanadian R. Analysis of androgen receptors in the human prostate by isoelectric- focusing in Polyacrylamide gel. J Steroid Biochem 1981; l4: 126l-7.
[12] Puddefoot JR, Anderson E, Vinson GP, Gilmore OJA. Heterogeneity of oestrogen receptors in human breast tumors. In: Peeters H, editor. Protides of the biological fluids. Oxford: Pergmon Press; 1987. p307-10.
[13] Marsigliante S, Baker VA, Puddefoot J, Barker S, Vinson GP. 4S Oestrogen receptor isoforms and their distribution in breast cancer samples. J Molec Endocrinol 1991; 7: 205-11.
[14] Chang C, Kokontis J, Liao S. Structural analysis of complementary DNA and amino acid sequences of human and rat androgen receptors. Proc Natl Acad Sci USA 1988; 85: 7211-5.
[15] Tilley WD, Marcela M, Wilson JD, McPhaul MJ. Characterization and expression of a cDNA encoding the human androgen receptor. Proc Natl Acad Sci USA 1989; 86: 327-31.
[16] Stamatiadis D, Dadoun F. Isoelectric focusing and 2D electrophoresis of the human androgen receptor. J Steroid Biochem Molec Bio 1992; 4l: 43.
[17] Sheridan PJ. Can a single androgen receptor fill the bill? Mol Cell Endocrinol 1991; 76: c39.
[18] Imperato-McGinley J, Guerrero L, Gautier T, Peterson RE. Steroid 5 alpha-reductase deficiency in man: an inherited form of male pseudohermaphroditism. Science 1974; 186: 1213-5.
[19] Brooks JR, Berman C, Nguyen H, Prahalada S, Primka RL, Rasmusson GH, Slater EE. Effect of castration, DES, flutamide, and the 5 alpha-reductase inhibitor, MK-906, on the growth of the Dunning rat prostatic carcinoma, R-3327. Prostate 1991; 18: 215-7.
[20] Baker VA, Puddefoot J, Marsigliante S, Barker S, Goode AW, Vinson GP. Oestrogen receptor isoforms, their distribution and relation to progesterone receptor levels in breast cancer samples. Br J Cancer 1992; 66: l083-7.
[21] Marsigliante S, Baker VA, Puddefoot J, Barker S, Vinson GP. 4S Oestrogen receptor isoforms and their distribution in breast cancer samples. J Mol Endocrinol 1991; 7: 205-11.

home

Correspondence to: Dr. Shu-Jie XIA, University Department of Urology, Shanghai First People's Hospital, 85 Wu Jin Rd. Shanghai 200080, China.
Tel: +86-21-6324 0090 Ext. 4708 Fax: +86-21-6324 0825
e-mail: XSJ@citiz.net
Received 2000-08-07 Accepted 2000-11-23