Recent developments regarding steroid receptors and antihormones

Muturitas, 8 (1986) 133-140

133

Elsevier MAT 00354

Recent developments regarding steroid receptors and antihormones Etienne INSERM

Baulieu

U33, Lab Hornones, 94270 Bihtre, France

(Received 19 June 1985; revision received 28 January 1986; accepted 19 March 1986) (Key words: Steroid receptors, RU 486, Progesterone, Tamoxifen, Endometrium)

Introduction There are five distinct classes of steroid hormone, namely oestrogen, androgen, progestogen, glucocorticosteroid and mineralocorticosteroid. The corresponding physiological hormones are oestradiol, testosterone, progesterone, cortisol (or corticosterone) and aldosterone, respectively. Depending on the type of target cell involved, each hormone controls different activities which may be classified into very diverse biological categories, for instance cell growth, cell differentiation, specific protein synthesis, etc. However, for each steroid hormone, whatever the type of effect it exerts, there is apparently a single and specific receptor, proper to the hormone itself and not to the response. All five types of receptor (ER: oestrogen receptor; PR: progesterone receptor; AR: androgen receptor; GR: glucocorticosteroid receptor and MR: mineralocorticosteroid receptor) are intracellular. (Only in Xenopus lueuis oocytes, in which progesterone causes meiosis to .resume, has a membrane receptor for steroid hormones been identified [19,20]). The receptors are all soluble proteins which are extractable without the use of detergent.

Research findings Autoradiography studies using radioactive steroids have demonstrated an accumulation of hormone apparently exclusively in the nucleus, where it is presumably bound to its receptor, as would be expected for the initiation of a hormonal response at the gene transcription level. In the absence of hormone, most if not all of the receptor is indeed found to be present in the nucleus, as has recently been visualized in chick oviduct, pituitary and hypothalamus as well as other chick organs, using anti-progesterone receptor antibodies [1,2] (Fig. 1). That the receptor Correspondence to: Professor E.E. Baulieu, Universite de Paris-Sud, Facultd de MCdecine de Bidtre, Dept. de Chimie Biologique, 78, rue du General Leclerc, 94270 Bic&tre, France

0378-5122/86/$03.50

0 1986 Elsevier Science Publishers B.V. (Biomedical Division)

134

135

is initially synthetised in the cytoplasm, as are all proteins, is certain, as is the fact that (active) hormone-receptor complexes accumulate in the nucleus. It would seem to be of little importance whether or not, in the absence of hormone, there is a preferential distribution of the receptor within the cytoplasm or the nucleus since hormone-receptor complexes may not act in the cytoplasm at all. In practice, that is ‘biochemically’, receptor in the absence of hormone can be extracted with a low-salt solution containing buffer, and will appear in the ‘cytosol’, a high-speed supernatant of target cell homogenates. Hormone-receptor complexes, however, have a greater affinity for the nuclear fraction and appear to be biochemically nuclear, while a high-salt solution is necessary for their extraction. Work is currently aimed at isolating the ‘acceptor’ to which hormone-receptor complexes are ‘naturally’ attached in the nucleus [3]. Many experiments have substantiated the notion of a two-state system for receptors, which may be either ‘non-transformed-non-activated’ or ‘transformedactivated’, the latter state being induced by hormone binding in the physiological context of the cell. Untransformed receptor is observed in the absence of hormone; this has a relatively low affinity for the nucleus and is therefore the ‘cytosol receptor’. However, this does not mean that all or part of the receptor does not originally accumulate in the nucleus, from which it is easily extracted. In low-salt buffer, all cytosol steroid receptors have a sedimentation coefficient of 8S, which can be stabilised by molybdate ions (Fig. 2). Such stabilisationGmparts a well-defined ‘8s’ sedimentation coefficient and resistance to the ‘activation’ effect of heating or increased salt concentration. It also indicates that the hormone binding property is more stable under denaturing conditions which would otherwise alter it (heating, proteolysis). In the activated state the receptor has a greater affinity for cell nuclei and many polyanions (e.g., DNA), while the rate of dissociation of the hormone from the receptor is decreased. In order to learn more about the function-

Fig. 1. Autoradiography and histoimmunocytology of chicken oviduct and pituitary cells [1,2]. (1) Autoradiogram (frozen sections) of chicken,oviduct explant incubated in vitro with [3H]ORG 2058. Cells of the luminal epithelium, glands and stroma concentrate the tritiated synthetic progestogen in their nuclei. Some stromal cells (arrowed) display stronger radioactive labelhng than nkghbouring glandular or epithelial cells. Note the decreasing gradient of labelling intensity from the periphery (top) to the centre (bottom) of the explant. (2) Frozen section of the same oviduct explant as in (1) stained by means of an immunoperoxidase technique using IgG-RB as primary antibody. The progesterone receptor is revealed as [ 3H]ORG 2058 in the same cell types (luminal epithelium, glands and stroma). In some stromal cells, the intensity of reaction with the antibody is greater (arrows) than in the neighbouring cells of the glands or luminal epithelium. There is no difference in intensity between the periphery (top) and centre (bottom) of the explant. It is thus demonstrated that the receptor is nuclear (detected by antibody) even in the absence of hormone (detected by radioactive labelling). (3) Frozen section of oviduct processed successively for autoradiography and immunohistochemistry. The nuclear accumulation of [ 3H]ORG 2058 is observed in the same cells in which the progesterone receptor is detected by IgG-RB. (4) Same technique as in (3), using a frozen section of chicken pituitary injected with [ ‘H]ORG 2058. The weak, non-specific background staining makes it easier to see the complete congruence between the hormonal and immunological techniques in revealing the progesterone receptor. Besides being important as regards the cell biology of the receptor, these results show for the first time the double detection of a receptor via the (radioactive) hormone and (an antibody to) the receptor protein.

136 6-

bottom t

GO

P

t

tt

top

I 30 Fraction number I

wi20

Fig. 2. Sucrose and glycerol gradient centrifugation analysis of purified 8S-PR in chick oviduct cytosol. (0), Labelled with [3H]progesterone; (O), idem after an additional incubation with 5 pM of non-radioactive progesterone for 30 min at 25°C.

of hormone-receptor complexes by comparing the so-called activated and non-activated forms of the receptor, we first purified the molybdate-stabilised 8s progesterone receptor (8S-PR) from the chick oviduct cytosol [4]. Subsequent sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of purified 8s revealed the presence of an intense 90 kDa peptide (detected by staining with Coomassie blue, a dye for protein determination) and two other proteins with molecular weights of 110 kDa and 79 kDa respectively (which show up better with silver nitrate staining). These last two proteins have already been described by Sherman, Schrader and O’Malley and their associates, under the designations B and A subunits, respectively. Purification of activated PR from salt-containing cytosol allowed us to obtain the llO-kDa B and the 79-kDa A proteins directly. Following affinity chromatography, these were separated using DEAE (diethylarninoethyl) cellulose chromatography. We obtained polyclonal antibodies (IgG-G3) [5] against the purified 8s receptor and against the B subunit (IgG-RB) [6], and a monoclonal antibody after injection of 8s receptor (BF4) [7]. Immunoblotting indicated that the BF4 antibody reacted only with the 90 kDa protein and IgG-RB only with the B and A subunits, while IgG-G3 antibodies recognised all three proteins (Fig. 3). The 90-kDa protein has recently been purified by immunoadsorption. Recent studies have indicated that the 90 kDa protein does not bind hormones. In the chick PR, the non-activated 8S-PR is composed of one molecule of a progesterone-binding protein (either llO-kDa B or 79-kDa A), and 2 molecules of 90 kDa. We have recently presented evidence for this oligomeric structure [8]. The 90-kDa protein is also remarkable because it seems to be present in the 8s forms of other steroid receptors [9]. Indeed, in the chick cytosol, we observed the interaction of the BF4 monoclonal antibody with ER, AR and GR, as well as with PR. Displacement on the gradient centrifugation pattern was observed only for the 8s ing

_I 36

10 ?, ACRlLUllDl

fraction

No

Fig. 3. (Left) 8S-PR I and II studied by ion exchange chromatography; composition analysed by SDS-PAGE (both contain 90 kDA; I contains more 79 kDa A and II more 110 kDa. B). (Right) Immunoblotting. 8%PR is seen in lanes 4 and 8 of 10% acrylamide SDS-PAGE. It is revealed by either IgG-G3 or BF, [9]. Purified 79 kDA A and 110 kDa B subunits obtained from a KCI-4s preparation of cytosol are seen in lanes 2 and 3 revealed by IgG-G3, but not in lanes 6 and 7, treated with BF,.

form, while the respective steroid-binding units did not react with this antibody. In the chick oviduct, at least 1% of the protein is accounted for by the 90-kDa protein in the cytosol. This preponderantly cytoplasmic protein may, besides interacting with components of the steroid receptors, also play other roles. It is in fact very similar to one of the recently described ‘heat-shock’ proteins of 90 kDa (M.G. Catelli, in preparation). In mammals, it has also been found that the oestrogen receptor in the calf uterus and the progesterone receptor in the rabbit uterus include, in their 8s forms, a 90-kDa non-hormone binding component (G. Redeuilh and M. Renoir, in preparation). The discovery of protein kinase activity in purified receptor fractions was both remarkable and unexpected [lo]. There is a llO-kDa B-associated protein kinase, which is not only active in the presence of Mg2+ (although not in that of Ca2’), but also phosphorylates histones and the j%subunit. Further extensive study is necessary before these enzymatic activities of receptor constituents can be understood in terms of hormone action. The same is true as regards the significance of receptor structure and activation. Pathophysiology

Steroid receptors are not found at the same cellular concentrations in different cell types, even within the same organ. Indeed, the hormonal control of concentra-

138 TABLE

I

CHANGES IN PR AND ER CONCENTRATIONS OESTRADIOL (DATA OBTAINED FROM DIFFERENT Changes

in number

Heterologous

of receptor

+ Oestrogen Oestradiol + Progesterone

a T,

Change

up; J , change

BY PROGESTERONE SYSTEMS)

AND

sites per cell a

hormones

Progesterone

CAUSED MODEL

receptor 1 receptor t

Homologous

hormones

Progesterone

+ Progesterone receptor receptor t

1

Oestradiol + Oestradiol

down.

tion may vary in the different cells [ll]. In a given cell, the concentration of steroid receptors will vary according to whether conditions are physiological or pathological [12]. For instance, variations in PR activity in the guinea pig uterus during the oestrus cycle are rather well explained by the induction of synthesis of PR by oestradiol at the pro-oestrus stage and the inactivation (‘down-regulation’) of binding by progesterone during the luteal phase [13]. This regulation and these changes in concentration can be put to clinical use despite the multiplicity and complexity of the hormonal controls (Table I). In the human endometrium there are changes in ER and PR during the cycle and in early pregnancy, all of which can be ‘logically’ explained by the effects of oestradiol and progesterone themselves. In post-menopausal women with endometrial cancer, the disease apparently produces an endometrium that is at risk for oestrogen, with too much ER and not enough PR [14,15] (Table II). During the normal cycle and early pregnancy, there are enough progesterone-PR complexes to be a target for antiprogesterone action, as has been observed with the antiprogesterone RU 486 [16]. As was shown in our studies on the chick oviduct using antioestrogens [17], it would seem that antihormone has only to ‘travel’ by itself to the nucleus in order to exchange with the endogenous hormone as a function of the respective concentrations and affinities. The theory that antihormone should act via binding to a cytoplasmic receptor is not therefore tenable. The antioestrogen 4-hydroxytamoxifen has a strong affinity for the ER. At one time there was a ‘theory’ suggesting that anti-steroid hormone should show rapid dissociation from the receptor: this is not

TABLE

II

OESTRADIOL AND PROGESTERONE RECEPTOR IN HUMAN ENDOMETRIUM MAL AND CANCEROUS WOMEN (NUMBER OF SITES PER CELL)

Norman women - proliferative phase Normal women - luteal phase Post-menopausal women - cancer

ER

PR

8800 a 3600 9ooo

12400 4400 4ooo

OF NOR-

a The figures are values from several studies, including [15]. Note that in endometrial cancer the ER value is relatively high and that of PR relatively low, indicating endometrium that is at risk for oestrogen.

139

// 2F

v\/N - I’

GT tt 1000 OH 30

“C3C_CH3

0

’

RU

486

x -AdA

5

10

i

al

days

Fig. 4. RU 486 and pregnancy interruption. Administration of 4 X 200 mg of RU 486 ( fi) The horizontal line indicates the duration of the bleeding and the two short vertical lines signal the abortion itself. ?? , progesterone; t, E, (oestradiol); A, cortisol; and 0, j3hCG in the plasma.

the case and pharmacological screening for antihormones should not be based on the affinity-to-receptor values for the tested compound if the object is to detect antihormones. RU 486 [16] possesses the features of a 1Pnortestosterone progestogen derivative with an extra ring attached at the Cl1 position. This substituent seems to be responsible for the non-agonist characteristics of the molecule, while the compound also retains a strong affinity for effective antagonism. It is remarkable that this antiprogesterone (and other anti steroid hormones synthesised by Roussel-Uclaf) (Fig. 4) should share, with antioestrogens of the triphenylethylene series (such as tamoxifen), an additional cyclic structure which is just outside the portion of the steroid moiety in front of the Cl-ClO-C9-Cll-Cl2 region. This may indicate that there are analogies between neighbouring portions of the binding sites of the receptors, including room for a substitution on the /3 side of the molecule. At the same time, any interaction would entail a conformational change abolishing hormone action. Our work also embraces another area of interest in the antihormonal field. This relates to a paradoxical effect that has been described with the antioestrogen tamoxifen, which is a very good ‘pure’, non-agonist, antagonist compound in the chick oviduct system [17]. We have observed that, in the presence of progesterone or glucocorticosteroid, tamoxifen acquires some oestrogen-like properties [18]. The mechanism of such a ‘potentiation’ is unknown, but we look upon it as an invitation to exercise caution, even after careful testing in experimental model systems. We make progress by using molecular and cellular approaches, and we are certainly keen to apply antihormones in the field of human pharmacology. However, we must always exercise the greatest care and continually question our own research findings so that our predictions will be as comprehensive as they can be and patients will benefit from the best possible safeguards. A systematic review of the field has not been presented here. References to other laboratory contributions will be found in the publications cited.

140

References 1 Gasc JM, Ennis BW, Baulieu, EE, Stumpf WE Recepteur de la progesterone darts l’oviducte de poulet: double revelation par immunohistochemie avec des anticorps antirecepteur et par autoradiograpbie a l’aide dun progestagene tritit. C R Acad Sci Paris 1983; 297: 477-482. 2 Gasc JM, Renoir JM, Radanyi C, Joab I, Tuohimaa P, Baulieu EE. Progesterone receptor in the chick oviduct: an immunohistochemical study with antibodies to distinct receptor components. J Cell Biol 1984; 99: 1193-1201. 3 Mass01 N, Lebeau MC, Baulieu EE. Interaction of oestrogen and progesterone receptors with specific subfractions of laying-hen oviduct chromatin. Biochem J 1984; 217: 309-316. 4 Renoir JM, Yang CR, Formstecher P, Lustenberger P, Wolfson A, Redeuilh G, Mester J, Richard-Foy H, Baulieu EE. Chick oviduct progesterone receptor: purification of a molybdate-stabilized form and preliminary characterization. Eur J Biochem 1982; 127: 71-79. 5 Renoir JM, Radanyi C, Yang CR, Baulieu EE. Antibodies against chick oviduct progesterone receptor. Cross-reactivity with mammalian progesterone receptors. Em J Biochem 1982; 127: 81-86. 6 Toucbimaa P, Renoir JM, Radanyi C, Mester J, Joab I, Buchou T, Baulieu EE. Antibodies against highly purified B-subunit of the chick oviduct progesterone receptor. Biochem Biophys Res Commun 1984; 119: 433-439. I Radanyi C, Joab I, Renoir JM, Richard-Foy H, Baulieu EE. Monoclonal antibody to chicken oviduct progesterone receptor. Proc Nat1 Acad Sci USA, 1983 80: 2854-2858. 8 Renoir JM, Buchou T, Mester J, Radanyi C, Baulieu EE. Oligomeric structure of the molybdatestabilized, non-transformed ‘8s’ progesterone receptor from chicken oviduct cytosol. Biochemistry 1985. 9 Joab I, Radanyi C, Renoir JM, Buchou T, Catelli MG. Immunological evidence for a common non-hormone-binding component in ‘non-transformed’ chick oviduct receptors of four steroid hormones. Nature 1984; 308: 850-853. 10 Garcia T, Tuohimaa P, Mester J, Buchou T, Renoir JM, Baulieu EE. Protein kinase activity of purified components of the chicken oviduct progesterone receptor. Biochem Biophys Res Commun 1983; 113: 960-966. 11 Mester J, Martel D, Psychoyos A, Baulieu EE. Hormonal control of oestrogen receptor in uterus and receptivity for ovoimplantation in the rat. Nature 1974; 250: 776-778. 12 Baulieu EE. Steroid receptors and hormone receptivity: New approaches in pharmacology and therapeutics. J Am Med Assoc 1975; 234: 404-409. 13 Milgrom E, Luu Thi M, Atger M, Baulieu EE. Mechanisms regulating the concentration and the conformation of progesterone receptor(s) in the uterus. J Biol Chem 1973; 248: 6366-6374. 14 Levy C, Robe1 P, Gautray JP, De Brux J, Verma U, Descomps B, Baulieu EE, Eychenne B., Estradiol and progesterone receptors in human endometrium: normal and abnormal menstrual cycles and early pregnancy. Am J Obstet Gynecol 1980; 136: 646-651. 15 Mortel R, Levy C, Wolff JP, Nicolas JC, Robe1 P. Baulieu, EE. Female sex steroid receptors in postmenopausal endometrial carcinoma and biochemical response to an antiestrogen. Cancer Res 1981; 41: 1140-1147. 16 Herrmann W, Wyss R, Riondel A, Philibert D, Teutsch G, Sakiz E, Baulieu EE. Effet d’un steroide anti-progesterone chez la femme: interruption du cycle menstruel et de la grossesse au debut CR Acad Sci Paris, 1984: 294: 933-938. 17 Sutherland R, Mester J, Baulieu EE. Tamoxifen is a potent ‘pure’ anti-oestrogen in chick oviduct. Nature 1977; 267: 434-435. 18 Binart N, Mester J, Baulieu EE, Catelli MG. Combined effects of progesterone and tamoxifen in the chick oviduct. Endocrinology 1982; 111: 7-16. 19 Baulieu EE, Schorderet-Slatkine S. In: Porter R, Whelan J, eds. Molecular biology of egg maturation. Bath: Pitman Press, 1983: 137-158. 20 Blondeau JP, Baulieu EE. Progesterone receptor characterized by photoaffinity labelling in the plasma membrane of Xenopus la&s oocytes. Biochem J 1984; 219: 785-792.