Influence of cyclic nucleotides on receptor binding, immunological activity, and microheterogeneity of human choriogonadotropin synthesized in placental tissue culture

Molecular and Cellular Endocrinology, 39 (1985) 15 l-l 59 Elsevier Scientific Publishers Ireland, Ltd.

151

MCE 01263

Influence of cyclic nucleotides on receptor binding, immunological activity, and microheterogeneity of human choriogonadotropin synthesized in placental tissue culture Gerhard Hilf and Wolfgang E. Merz * Deportment of BiochemistT II, University of Heidelberg, Im Neuenheimer Feld 328, D 6900 Heidelberg (F. R.G.) (Received

10 August

1984; accepted

Kqywords: ratio of receptor binding activity/immunological biosynthesis; acidic hCG subpopulations.

activity;

5 November

activity

1984)

to stimulate

adenylate

cyclase

and

testosterone

Summary

The regulation of the biosynthesis of choriogonadotropin (hCG) in tissue culture by human first trimester placenta in the presence of the following cyclic nucleotides and 3-isobutyl-1-methylxanthine (IBMX) was studied (concentrations in parentheses): IBMX (0.1 mM), CAMP (1 mM) + IBMX (0.1 mM), cGMP (0.1 mM) + IBMX (0.1 mM), 8-bromo-CAMP (0.5 mM) and 8-bromo-cGMP (0.5 mM). The medium concentration of hCG follows an optimum curve at all conditions, showing highest values at day 3 of the culture. The efficacy of the substances to cause an increase in the hCG medium concentration was in the following order: IBMX < control < 8-bromo-cGMP = 8-bromo-CAMP c cGMP + IBMX < CAMP + IBMX. The synthesized hCG was examined with respect to its receptor binding activity (LH/hCG receptor of rat testes), the activities to stimulate adenylate cyclase as well as testosterone biosynthesis in purified mouse Leydig cells, the immunological activity, and the microheterogeneity in isoelectric focusing. Only in the presence of 8-bromo-CAMP, 8-bromo-cGMP, and cGMP + IBMX was hCG synthesized, which differs significantly in the investigated properties from hCG of the control cultures. Only in the presence of 8-bromo-cGMP is the ratio of receptor binding activity/immunological activity optimal (near 1). In the presence of both 8-bromo-CAMP and 8-bromo-cGMP, microheterogeneity of hCG in isoelectric focusing was diminished and the synthesis of more acidic hCG subpopulations was favoured.

Regulation of biosynthesis of human choriogonadotropin (hCG) by placental tissue as well as by tumours has been described by a number of authors (Hussa, 1977; Maruo et al., 1980; Hustin et al., 1980; Sasaki et al., 1983; Melmed et al., 1983). It has been demonstrated that the rate of hCG biosynthesis is influenced by the conditions for culture of placental tissue (Huot et al., 1979). Furthermore, hCG biosynthesis in placental tissue * To

whom correspondence

0303-7207/85/$03.30

should be addressed.

0 1985 Elsevier Scientific

Publishers

Ireland,

as well as in cultures of tumour tissue is stimulated by cyclic nucleotides (Chou et al., 1978; Huot et al., 1979; Chou, 1980; Cole et al., 1981; Kinoshita, 1981; Haning et al., 1982; Zeitler et al., 1983) and gonadoliberin (Haning et al., 1982; Siler-Khodr et al., 1983; Merz and Hilf, 1984), whereas the action of epidermal growth factor on hCG biosynthesis seems to be different in normal and malignant placental cultures (Beneviste et al., 1978; Huot et al., 1981). The relation of hCG/isolated subunits as well as molecular properties of the secreted Ltd.

152

hormone such as molecular weight or glycosylation (Hussa, 1977; Hustin et al., 1980; Maruo et al., 1980; Cox, 1981a, b) are changed during culture. The present paper describes the influence of cyclic nucleotides, especially of 8-bromo-CAMP and of 8-bromo-cGMP, on the hCG biosynthesis in tissue culture of first trimester placenta. The action of these substances on special aspects of hCG biosynthesis (ratio of hCG/isolated subunit in the medium) and on molecular properties of the secreted hormone (ratio of immunological activity/receptor binding activity, electrophoretic microheterogeneity) is studied. Materials and Methods Source of reagents Cyclic nucleotides, dispase (EC 3.4.24.4, from B. polymyxa) grade II, streptomycin sulphate and desoxyribonuclease (EC 3.1.21.1) grade II were purchased from Boehringer (Mannheim, F.R.G.). Insulin, dexamethasone, 3-isobutyl-l-methylxanthine (IBMX), ribonuclease A (EC 3.1.27.5) tissue culture medium 199, Norit A charcoal, and ampholyte for isoelectric focusing (Servalyte AG 2-11) were obtained from Serva Feinbiochemica GmbH (Heidelberg, F.R.G.). Penicillin G, sodium salt, was purchased from Hoechst AG (Frankfurt/Main, F.R.G.), amphotericin B and newborn calf serum from Biochrom (Berlin, F.R.G.), heparin from Braun (Melsungen, F.R.G.), Percoll, Dextran T-70 and CL-Sepharose 4B from Deutsche Pharmacia (Freiburg/Breisgau, F.R.G.). Buffer substances and the other chemical reagents were purchased from E. Merck AG (Darmstadt, F.R.G.). Na’251 for radioactive labelling was obtained from Amersham-Buchler (Braunschweig, F.R.G.). Placental tissue culture Human placental tissue was obtained from legal abortion (7th to 8th week of gestation), minced into pieces of 5-20 mg (wet weight), washed with medium (see below) and cultured for 7-8 days in Nunc multi-well tissue culture dishes in 1.5 ml tissue culture medium 199 containing per litre: 1 mg insulin, 393 pg dexamethasone, 116 mg penicillin G, sodium salt, 200 mg streptomycin sulphate

and 2.5 mg amphotericin B. Newborn calf serum was added to give a final concentration of 10% (v/v). The cultures were incubated at 37’C in a humidified atmosphere of 95% air, 5% CO, (CO, incubator Assab, Sundbyberg, Sweden). Medium was changed every 24 h. The medium was collected and stored in the frozen state (-20°C). Tissue was cultured in the presence of the following substances in order to examine their influence on hCG biosynthesis: IBMX (0.1 mM), CAMP (1 mM)+ IBMX (0.1 mM), cGMP (0.1 mM) + IBMX (0.1 mM), 8-bromo-CAMP (0.5 mM) and 8-bromo-cGMP (0.5 mM). Radioimmunoassay systems The medium concentrations of hCG and of its subunits were determined by means of radioimmunoassays using goat antisera against hCG and the isolated subunits prepared as described elsewhere (Merz et al., 1973). The radioimmunoassays were performed according to Herbert et al. (1965) using as trace ‘251-labelled hCG and ‘251labelled isolated subunits, respectively. Radioactive labelling was achieved by means of the chloramine-T method in accordance with Markkanen et al. (1980). Analysis of molecular properties of hCG synthesized in placental tissue culture Prior to analysis the media were thawed and desalted by chromatography on Sephadex G-25 (2.5 x 50 cm). The protein fraction was collected and lyophilized entirely. Receptor binding activity The ability of the synthesized hCG to bind to LH/hCG receptors of rat testes was determined by the radioligand-receptor test as described by Merz (1979). Dose response was composed of 5 doses separated by a constant factor of 1.6 and triplicate values per dose. The receptor binding activity is referred to the Second International Standard for Choriogonadotropin (Bioassay), kindly supplied by the WHO International Laboratory for Biological Standards (London, U.K.). Immunological activity Immunological activity dioimmunoassay (Herbert

was determined by raet al., 1965) using goat

153

anti-hCG serum (Merz et al., 1973). Dose-response curves were established from 5 doses (2, 5, 10, 20 and 50 ng hCG per ml) and triplicate values of each dose. The immunological activity is referred to the First International Standard for Choriogonadotropin (Immunoassay), kindly supplied by the WHO International Laboratory for Biological Standards (London, U.K.). Biological activity in vitro The ability of hCG to stimulate the synthesis of CAMP and testosterone in Leydig cells was used to determine the biological activity in vitro. Leydig cells of NMRI mice 96 months old) were prepared according to Lorenzen and Merz (unpublished) by enzymatic dispersion of decapsulated testes by means of dispase. The enzyme was dissolved in phosphate-buffered saline, pH 7.4 (2.5 g/l), containing per litre: 50 mg ribonuclease A, 50 mg desoxyribonuclease I and 100000 IU heparin. Each testis was incubated in 2 ml of the medium (30 min, 34”C, 95% air, 5% CO,). Thereafter the tissue was shaken, filtered through nylon gauze (100 pm) and applied to discontinuous osmolar Percoll density gradients (608, 40%, lo%, v/v) in tissue culture medium 199. After centrifugation (5 min at 60 x g followed by 15 min at 800 x g), the cells layered on top of the 60% Percoll solution were collected and washed twice with medium. lo5 Leydig cells (500 ~1) were incubated with 100 ~1 sample for 2 h at 34°C. Dose-response curves consisted of 5 doses (triplicate values) separated by a constant factor of 1.2. After incubation the cells were lysed (0°C; addition of 1.5 mM aqueous solution of IBMX), boiled (3 min) and stored at - 20°C. CAMP content was determined according to Gilman (1970) using binding protein from rabbit skeletal muscle (kindly supplied by F. Hofmann, Dept. of Pharmacology, University of Heidelberg). Testosterone content was determined by radioimmunoassay using dextran T-70-coated Norit A charcoal for separation of bound and free antigen. The anti-testosterone serum was kindly supplied by P. Vescei (Dept. of Pharmacology, University of Heidelberg). Statistical calculations Calculations of receptor binding, biological and

immunological activities and of 95% confidence limits were performed after variance analysis of spline-approximated dose-response curves and F-tests for parallelism of the dose-response curves of the hCG probes and the hCG standards (First International Standard for Choriogonadotropin (immunological activity) and Second International Standard for Choriogonadotropin (receptor binding and biological activities in vitro)). The calculations were performed by means of a HP 9825 A computer using our own software. Documentation of computations and printing of the dose-response curves were carried out with a HP 9871 A printer. The computer and the printer were purchased from Hewlett Packard (Frankfurt/ Main, F.R.G.). Isoelectric focusing hCG from placental tissue culture was examined for microheterogeneity in high resolution isoelectric focusing as previously described by Merkel and Merz (1984). Before being applied, hCG containing tissue culture medium was partially purified by immunoadsorption with rabbit antiserum against newborn calf serum immobilized on CL-Sepharose 4B in order to decrease the concentration of the serum proteins. When this procedure was omitted, thermal denaturation of proteins occurred in the course of isoelectric focusing, mainly in the albumin region, causing co-precipitation of hCG in part. After isoelectric focusing, the gels (on 5 x 5 cm slides) were cut into 40 slices of 1 mm width each and incubated with 250 ~1 RIA buffer for 6 h at room temperature in order to recover the proteins from the gel. Results

Medium concentration of hCG and isolated subunits A typical experiment out of a total number of similar cultures of 4 other placentas is presented in Fig. 1. At day 3 of culture, medium concentrations of hCG show optimal values at all conditions used (Fig. la). Thereafter a rapid decline is observed. The combination IBMX + CAMP stimulates the secretion of immunoreactive hCG best. The medium concentrations of the isolated subunits

154

never exceed one-third of the hCG concentrations on a molar basis (Fig. 1). The medium concentrations of the a-subunit declines during the whole culture period, whereas the medium concentration of the &subunit seems to show a smaller decrease (Fig. lb, c) and follows an optimum curve, respectively (in the presence of 8-bromo-CAMP and in other experiments; data not shown). 8-Bromo-CAMP shows only a moderate stimulating effect on hCG medium concentration. However, the release of the isolated P-subunit into the medium is si~ificantly increased whereas in case of the a-subunit the situation is reversed. In this

250

r" 200 3 2 5 150 E ';. aI b 3

100

50

0

231567

Days of Tissue Culture Fig. l(a). Influence of 1BMX and cyclic nucleotides on the medium concentration of immunoreactive hCG synthesized in cultures of first trimester placental tissue. Tissue was cultured without additions (+) and in the presence of 0.1 mM IBMX (x), 1 mM cAMP+O.l mM IBMX (0), 0.1 mM cGMP+O.l mM IBMX (A), 0.5 mM S-bromo-CAMP (0) and 0.5 mM S-bromo-cGMP (A). Range of medium concentrations (all conditions) of the isolated a-subunit (dotted area) and of the P-subunit (hatched area); depicted in detail in Fig. l(b. c).

60

.

1 2 Receptor Binding Act. Wlmgl

234567 Days

of Tissue

Culture

Fig. l(b,c). Release of the isolated a-subunit (b) and of the isolated @ubunit (c) into the tissue culture medium in control cultures (+ ) and in the presence of IBMX ( x), CAMP+ IBMX (0), cGMP+ IBMX (A), Sbromo-CAMP (0) and I-bromocGMP (A); same concentrations as in Fig. l(a).

Fig. 2. Ratios of receptor binding activity to immunological activity of hCG present in desalted lyophilized placental tissue culture medium collected at days 3, 4 and 5 of culture. Conditions for cultures: control (A), 0.1 mM IBMX (B), 1 mM cAMP+O.l mM IBh4X (C), 0.1 mM cGMP+O.l mM IBMX (D), 0.5 mM Sbromo-CAMP (E) and 0.5 mM 8-bromo-cGMP (F). The specific activities are given expressed in terms of IU/mg of lyophilized material. The horizontal and vertical bars represent the 95% confidence limits.

155

TABLE

1

RECEPTOR BINDING PLACENTAL TISSUE

AND IMMUNOLOGICAL ACTIVITY OF HUMAN CHORIOGONADOTROPIN CULTURE IN THE PRESENCE OF CYCLIC NUCLEOTIDES

Condition (medium concentration,

mM)

Day

Receptor

binding

Immunological

activity

IU/mg

95% confidence

limit

SYNTHESIZED

activity

IU/mg

95% confidence

Control IBMX (0.1) IBMX (O.l)+cAMP (1.0) IBMX (0.1) + cGMP (0.1) R-bromo-CAMP (0.5) 8-bromo-cGMP (0.5)

3 3 3 3 3 3

1510 7610 7590 11500 10570 11670

7130-8070 7330-7610 7320-7860 11020-12030 10250-10900 113Of-12040

9200 9360 8550 11160 10440 11520

8500-9960 8 640-10 140 6540-11130 10550-11810 9930-10990 10960-12120

Control 8-bromo-CAMP R-bromo-cGMP

5 5 5

6 640 6930 16067

6 290-7 040 6510-7350 15570-16580

7860 10650 18230

7470-8 250 10230-11100 17170-19390

(0.5) (0.5)

IN

limit

Receptor binding activity was determined by the radioligand-receptor test (rat testis homogenate) using the Second International Standard for Choriogonadotropin (Bioassay) as reference. Immunological activity was measured by radioimmunoassay; the First International Standard for Choriogonadotropin (Immunoassay) served as reference. In both cases the dose-response curves consisted of 5 dose points and triplicate values at each dose. The specific activity and the 95% confidence limits were calculated from the spline-approximated dose-response curves. The specific activity is based on the hCG content of the tissue culture medium, determined by radioimmunoassay prior to the binding tests.

respect tinct.

the action

of 8-bromo-cGMP

is less dis-

Receptor binding and immunological activity of the synthesized choriogonadotropin At day 3 of culture, when medium concentration of hCG is optimal, only the combination IBMX + cGMP and the 8-bromo-derivatives of CAMP and cGMP effect the biosynthesis of choriogonadotropin, which shows a significantly higher receptor binding activity and immunological activity than the hCG of the control cultures (Table 1). Thereafter, the receptor binding activity as well as the immunological activity of choriogonadotropin is decreased (data not shown), except in cultures which contained 8-bromo-cGMP. The optimum of receptor binding and immunological activity is reached in this case at day 5. The influence of the cyclic nucleotides on the ratio of the specific activities (receptor binding and immunological activity) in terms of IU/mg of desalted lyophilized culture medium is depicted in Fig. 2. It is evident that at day 3, only in the presence of 8-bromo-cGMP are choriogonadotropin molecules secreted into the medium which are equally potent to bind to the testicular LH/hCG receptor and to hCG antibodies. In the contra

cultures as well as in the presence of the other cyclic nucleotides, hCG is secreted with significantly higher immunoreactivity in comparison to the receptor binding activity. The discrepancy between these two activities is significantly increased in the control cultures at day 4 in favour of the immunological activity. In the presence of 8bromo-CAMP the ratio of immunological activity/receptor binding activity is further increased at day 5. The changes in the ratio of the receptor binding activity versus immunological activity during the whole culture period in the control cultures and in the presence of 8-bromo-CAMP and 8-bromocGMP are demonstrated in Fig. 3. The effect of 8-bromo-cGMP is clearly visible from Fig. 3C. In contrast to the other conditions, this nucleotide effects the synthesis and secretion of a hormone which shows an optimal ratio of receptor binding activity and immunological activity. Biological activity in vitro Preliminary results indicate that the abilities of hCG synthesized in placental tissue culture to stimulate CAMP and testosterone biosynthesis in purified mouse Leydig cells are also significantly changed in the course of the culture as effected by

156

1PH

I 6

,’

$06. FOL. 02. 0.

11

n 10

Ske

2 Receptor

L

6

8

Binding Act. lIU/mgl

Fig. 3. Changes in the ratio of receptor binding activity to immunological activity of hCG as found in placental tissue culture medium in the interval of days 2-8 of culture (arabic numerals) in control cultures (A), and in the presence of 0.5 mM 8-bromo-CAMP (B) and 0.5 mM 8-bromo-cGMP (C). The specific activities are expressed in terms of IU/mg of lyophilized material. The horizontal and vertical bars represent the 95% confidence limits.

the various culture conditions. At day 3 in the control cultures the activity to stimulate adenylate cyclase (9320 IU/mg hCG; 95% confidence limits 8740-9900) and the activity to stimulate testosterone biosynthesis (9720 IU/mg; 8030-12880) are not different. In the presence of 8-bromo-CAMP the activity to stimulate adenylate cyclase (11990 IU/mg; 11450-12 580) is significantly higher than in the control cultures. The stimulation of testosterone biosynthesis (11570 IU/mg; 9980-13950) corresponds well to the activity to stimulate adenylate cyclase. In the presence of 8-bromo-cGMP, slightly but not significantly lower values are reached at day 3 of culture

20 Number

30

LO

Fig. 4. Microheterogeneity of hCG synthesized in placental tissue cultures as demonstrated by isoelectric focusing in ultrathin layers (120 pm; polyacrylamide/IEF-agarose gel). The hCG content in slices of 1 mm width determined by radioimmunoassay is given. The placental tissue was cultured under the following conditions: at day 3, control cultures (panel a), cultures in the presence of 0.5 mM 8-bromo-CAMP (panel b), 0.5 mM 8-bromo-cGMP (panel c); in the presence of S-bromocGMP (panel d) at day 5.

for the activity to stimulate adenylate cyclase (10940 IU/mg; 10 320-11600) and for the stimulation of testosterone biosynthesis (10 610 IU/mg; 8800-12 620). Microheterogeneity in isoelectric focusing As urinary choriogonadotropin, hCG synthesized in the placental tissue culture shows microheterogeneity in isoelectric focusing (Fig. 4a). In the presence of 8-bromo-CAMP as well as of 8bromo-cGMP, the formation of acidic forms of the hormone is favoured (Fig. 4b-d). This effect on the microheterogeneity of the hormone coincides with the optima of the receptor binding activity and immunological activity of the molecules (day 3 in cultures in the presence of 8bromo-CAMP, day 5 in the presence of 8-bromocGMP).

157

Discussion The data presented in this paper demonstrate that cyclic nucleotides both stimulate and regulate hCG biosynthesis by tissue from first trimester placenta. The effectiveness of CAMP and the other cyclic nucleotides is not very high when referred to this parameter. However, when optimal conditions for the culture of the tissue are applied, hCG biosynthesis and secretion are not increased more than 2-3-fold even in the presence of gonadoliberin (Merz and Hilf, 1984). Also, the stimulation of the hCG biosynthesis in cultured choriocarcinoma cells by epidermal growth factor ranges in this order of magnitude (Beneviste et al., 1978). The regulating effect of cyclic nucleotides on the hCG biosynthesis is more distinct and is obvious from the changes in molecular properties of choriogonadotropin taking place under the influence of these substances. The 8-bromo-derivatives of the cyclic nucleowhich are more resistant to phostides, phodiesterase, effect a significant increase in receptor binding activity, activity to stimulate adenylate cyclase and testosterone biosynthesis as well as in immunological activity of secreted choriogonadotropin in relation to the control cultures. In the presence of CAMP in combination with the phosphodiesterase inhibitor 3-isobutyl-l-methylxanthine (IBMX), the highest amount of immunoreactive material identified by hCG antibodies is found in the medium; however, the receptor binding and immunological activities as well as the activity to stimulate adenylate cyclase and testosterone biosynthesis (data not shown) are not increased in relation to hCG found in the medium of the control cultures. In contrast, cGMP in combination with IBMX effects the secretion of a hCG molecule with significantly higher potency to bind to LH/hCG receptors as well as to hCG antibodies (Table 1). The activity of this hormone to stimulate adenylate cyclase of Leydig cells is also significantly increased (12030 IU/mg; 95% confidence limits 10320-14730). Its activity to stimulate testosterone biosynthesis (10 911 IU/mg) seems also to be elevated. This is remarkable since it was applied at a concentration (0.1 mM) which is lower by a factor of 10 than the CAMP medium

concentration used. In the presence of 8-bromoCAMP the receptor binding activity of secreted hCG is significantly lower than the activity to stimulate adenylate cyclase of Leydig cells. This seems to be reasonable from the fact that in the presence of this substance the pattern of microheterogeneous hCG molecules is shifted towards the anode in isoelectric focusing. As previously shown by Graesslin et al. (1973) and by us (Brossmer et al., 1971; Merz et al., 1974a), urinary human choriogonadotropin shows microheterogeneity in isoelectric focusing and other electrophoretic systems. As demonstrated by amino acid and carbohydrate analyses of the isolated hCG subpopulations, microheterogeneity is based on a different content of N-acetylneuraminic acid (Merz et al., 1974b). As shown in the present paper, hCG synthesized in the placental tissue culture is also microheterogeneous. The increase in the concentration of the more acidic hCG subpopulations in the presence of 8-bromo-CAMP in the medium of the tissue cultures is ascribed to a higher content of sialic acid; however, this has still to ,be proven experimentally. The lower receptor binding activity may be caused by the content of sialic acid being possibly higher in these molecules, since it has been shown that asialo- or partially desialylated hCG shows an increased receptor binding ability (Hilgenfeldt et al., 1976; Madnick et al., 1981; Amir et al., 1981). The activity of asialo-hCG to stimulate adenylate cyclase in target cells, however, seems to be decreased (Moyle et al., 1975). It is of special interest that choriogonadotropin, which shows a diminished microheterogeneity and a predominance of acidic hCG subpopulations, is also observed in tumour patients (Mann and Karl, 1983). The ability of 8-bromo-CAMP and of 8-bromocGMP to favour the biosynthesis of more acidic hCG molecules which show increased biological and immunological activity in vitro hints at a regulatory action on post-translational processes optimizing their efficiency, possibly by the improvement of the glycosylation of the hormone. This may cause an increase in the amount of hCG correctly combined from the isolated subunits, since it has been shown that the carbohydrate moiety influences the combination of the subunits in the course of biosynthesis of the pituitary

158

hormones TSH and LH, which are closely related to hCG (Weintraub et al., 1980; Strickland and Pierce, 1983). The processing of the carbohydrate chains of the a-subunit has been suggested to be rate-limiting for hCG biosynthesis in choriocarcinema cells (Ruddon et al., 1981). At least, the decrease in the medium concentration of the isolated a-subunit in the presence of 8-bromo-CAMP (Fig. lb) may hint at this direction. The study of the biosynthesis of hCG and the molecular properties is of special interest not only for gynaecology but also for tumour biology and diagnosis, since hCG is a good tumour marker (Vaitukaitis et al., 1976; Baylin and Mendelson, 1980). hCG secretion and subunit release is changed during chemotherapy and in the presence of tumour promoters (Melmed et al., 1983; Sasaki et al., 1983). The amount of isolated a-subunit is in some cases related to the malignancy of the tumour (Heitz et al., 1983). Furthermore, the ratio of hCG/isolated subunits is of diagnostic interest in ectopic pregnancy (L’Hermite-Baleriaux et al., 1982). Finally, the hCG-like substance found in normal non-pregnant human tissue shows varying ratios of immunological activity/biological activity in different organs (Braunstein et al., 1984). The biological significance of these changes in molecular properties of hCG and hCG-like substances is not at present clear. Acknowledgements These investigations were supported by the Deutsche Forschungsgemeinschaft, Bonn-Bad Godesberg (Me 545). The authors wish to thank Mr. Uwe Merkel for performing the isoelectric focusing experiments. References Amir, SM., Sullivan, R. and Ingbar, S.H. (1981) Endocrinology 109, 1203-1211. Baylin, S.B. and Mendelson, G. (1980) Endocr. Rev. 1, 45-77. Beneviste, R., Speeg Jr., K.V.. Carpenter. G., Cohen, S., Lindner, J. and Rabinowitz, D. (1978) J. Clin. Endocrinol. Metab. 46, 169-172. Braunstein, G.D., Rasor, J., McCready. J. and Wade, M.E. (1984) J. Clin. Endocrinol. Metab. 58, 170-175. Brossmer, R., Merz, W.E. and Hilgenfeldt, U. (1971) FEBS Lett. 18, 112-114.

Chou, J.Y. (1980) Cancer Res. 40, 4025-4030. Chou, J.Y., Wang, S.S. and Robinson, J.C. (1978) J. Clin. Endocrinol. Metab. 47, 46-50. Cole, L.A., Hussa, R.O. and Rao, C.V. (1981) Cancer Res. 41. 1615-1619. Cox, G.S. (1981a) Cancer Res. 41. 3087-3094. Cox, G.S. (1981b) Biochemistry 20, 4893-4900. Gilman, A.G. (1970) Proc. Natl. Acad. Sci. (U.S.A.) 67, 305-312. Graesslin, D., Weise, H.-C. and Braendle, W. (1973) FEBS Lett. 31, 214-216. Haning, R.V., Choi, L., Kiggens, A.J., Kuzma, D.L. and Summerville, J.W. (1982) J. Clin. Endocrinol. Metab. 55, 213-218. Heitz, P.U., Kasper, M., Kloppel, G., Polak, J.M., Oberholzer, M. and Vaitukaitis, J.L. (1983) Acta Endocrinol. (Kbh.) Suppl. 253. 75-76. Herbert, V., Lau, K.S., Gorrlieb, C.W. and Bleicher, S.J. (1965) J. Clin. Endocrinol. Metab. 25, 1375-1384. Hilgenfeldt, U., Merz, W.E. and Brossmer, R. (1976) Proc. Vth Int. Congr. Endocrinol., Hamburg, abstract 329, p. 135. Huot, R.I., Foidart, J.M. and Stromberg, K. (1979) In Vitro 15, 497-502. Huot, R.I., Foidart, J.M., Nardone, R.M. and Stromberg, K. (1981) J. Clin. Endocrinol. Metab. 53, 1059-1063. Hussa, R.O. (1977) J. Clin. Endocrinol. Metab. 44, 1154-1162. Hustin, J., Gaspard, U., Reuter, A., Hendrick, J.C. and Franchimont, P. (1980) In: The Human Placenta: Proteins and Hormones, Vol. 35, Eds.: A. Klopper, A. Genazzani and P.G. Crosignani (Academic Press, London-New York) pp. 259-266. Kinoshita, F., Nakai, Y., Nikaido, 0. and Imura, H. (1981) Cancer Res. Clin. Oncol. 100, 321-324. L’Hermite-Baleriaux, M., Van Exter, C., Deville, J.L., Gaspard, U. and Hechtermans, R. (1982) Acta Endocrinol. (Kbh.) 100,109-113. Madnick, H.M., Kalyan, N.K., Segal, H.L. and Bahl, O.P. (1981) Arch. Biochem. Biophys. 212, 432-442. Mann, K. and Karl, H.-J. (1983) Cancer 52, 654-660. Markkanen, S., Tollikkii, K., Jaaskelainen, K. and Rajaniemi, H. (1980) Hormone Res. 12, 32-45. Marne, T., Segal, S.J. and Koide, S.S. (1980) In: Chorionic Gonadotropin, Ed.: S.J. Segal (Plenum Press, New York) pp. 177-197. Melmed, S., Braunstein, G.D., Hershman, J.M. and Wade, M.E. (1983) Gynec. Oncol. 16, 49-55. Merkel, U. and Merz, W.E. (1984) Acta Endocrinol. (Kbh.) Suppl. 264, 120-121. Merz, W.E. (1979) Eur. J. B&hem. 101, 541-553. Merz, W.E. and Hilf, G. (1984) In: Recent Advances in Bioscience, Eds.: M. Schmidt-Gollwitzer and W. Elger (W. de Gruyter, Berlin) in press. Merz, W.E., Hilgenfeldt, U., Brockerhoff, P. and Brossmer, R. (1973) Eur. J. Biochem. 35. 297-306. Merz, W.E., Hilgenfeldt, U., Darner, M. and Brossmer, R. (1974a) Hoppe-Seyler’s Z. Physiol. Chem. 355, 1035-1045. Merz, W.E., Hilgenfeldt, U., Brossmer, R. and Rehberger, G. (1974b) Hoppe-Seyler’s Z. Physiol. Chem. 355, 1046-1050.

159

Moyle, W.R., Bahl, O.P. and M’ti, L. (1975) J. Biol. Chem. 250, 9163-9169. Ruddon, R.W., Bryan, A.H., Hanson, E.A., Perini, I., Ceccorulli, L.M. and Peters, R.P. (1981) J. Biol. Chem. 256, 5189-5196. Sasaki, N.. Matsuzaki, M., Okabe, T., Imai, Y., Kaneko, Y. and Matsuzaki, F. (1983) Experientia 39, 330-331. Siler-Khodr, T.M., Khodr, G.S., Vickery, B.N. and Nestor Jr., J.J. (1983) Life Sci. 32, 2741-2745.

Strickland, T.W. and Pierce, J.G. (1983) J. Biol. Chem. 258, 5927-5932. Vaitukaitis, J.L., Ross, G.T., Braunstein, G.D. and Rayford, P.L. (1976) Recent Progr. Hormone Res. 32,289-331. Weintraub, B.D., Staunard, B.S., Linnekin, D. and Marshall, M. (1980) J. Biol. Chem. 255, 5715-5723. Zeitler, P., Markboff, E. and Handwerger, S. (1983) J. Clin. Endoctinol. Metab. 57, 812-818.