Molecular and CeNular Endocrinology, 78 (19911229-236 0 1991 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/91/$03.50

229

MOLCEL 02533

Distinct placental lactogen and prolactin (lactogen) receptors in bovine endometrium Sybille S. Galosy ‘, Arieh Gertler 2, Gerard Elberg 2 and Don M. Laird

’

’ Department of Biological Sciences, Monsanto Company, Chesterfield, MO 63198, U.S.A., and 2 Department ofBiochemists and Human Nuir~tja~, Faculty o~Agricu~ture, The Hebrew Uni~ersi~ of Jerusalem, ~ehu~ot 76lOi3, Israel

(Received 21 January 1991; accepted 28 March 1991)

Key words: Placental lactogen receptor; Endometrium; (Bovine)

Summary Specific binding sites for bovine placental lactogen (bPL) and the lactogenic hormone, prolactin, have been detected in endometrial membranes isolated from uteri of mid-pregnant heifers. The specific binding of human growth hormone (hGH) (used to monitor the presence of lactogenic binding sites) and of bPL was increased appro~mately 4-fold following treatment of the membranes with 4 M MgCI,. Binding was found to be ligand specific, membrane protein concentration-, time- and temperature-dependent and reversible. Scatchard analysis of bPL and hGH competition binding data revealed curvilinear plots with dissociation constants for the high affinity sites of 4.1 X lo-" M and 6.4 x 10-i’ M, respectively. The maximum capacity of binding of bPL at the high affinity site was 21 fmol/mg membrane protein while approximately twice the level of binding was measured for hGH (39 fmol/mg). Both hGH and bGH, but not ovine prolactin, competed with [ 12sI]bPL for binding. The concentrations of hGH and bGH needed to effectively compete were however lOO-fold higher than those required for unlabeled bPL. No specific binding of radiolabeled bGH was detected in endometrial tissue suggesting the absence of bGH receptors. Preferential competition of [‘251]hGH binding was observed by prolactin and bPL. From these data it may be inferred that hGH binding is indicative of the presence of both lactogenic (prolactin) and bPL binding sites in endometrial tissue. The presence of distinct bPL receptors in the endometrium from mid-pregnant cows suggests a possible role for bPL in the maintenance of pregnancy.

Introduction Placental lactogens (PLs) are polypeptide hormanes produced by the placenta of a variety of mammals. Structural and functional similarities

Address for correspondence: D.M. Laird, Department of Biological Sciences, AA3C, Monsanto Company, 700 Chesterfield Village Parkway, Chesterfield, MO 63198, U.S.A.

to the pituitary hormones prolactin (PRL) and growth hormone (GH) suggest their membership in the GH/PRL polypeptide hormone gene family. PLs have been identified in a number of species including primates, rodents and ruminants (for review see Talamantes and Ogren, 1988). ~though the specific biological action(s) is uncertain, PL has been proposed to be involved in the regulation of fetal growth and metabolism (Freemark et al., 19891, development of the ma-

230

ternal mammary gland (Forsyth, 1986) and modulation of maternal intermediary metabolism (Handwerger et al., 1976). Initial investigations demonstrating PL’s ability to compete for binding with radiolabeled GH and PRL to postnatal tissues suggested that the biological effects of PL may be mediated through GH and/or PRL receptors (Shiu et al., 1973; Tsushima and Friesen, 1973; Lesniak et al., 1977). Recent findings (Freemark and Handwerger, 1983, 1984, 1986) indicating marked differences in the actions of PL and GH in fetal rat and ovine tissues as well as demonstration of the existence of specific PL receptors in tissues from fetal lambs and pregnant sheep (Freemark et al., 1987) suggest that the biological effects of PL are, at least in part, mediated through binding of the hormone to a distinct and unique receptor. Bovine PL (bPL1 has been found to be clearly different from PLs of other species including its close relatives, the sheep and goat. With an apparent molecular weight of 30-32 kDa on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), bPL exists as several isoforms having isoelectric points that range from 4.8 to over 5.5 (Byatt et al., 1986). Glycosylation of bPL may account for the comparatively high molecular weight and multiple isoforms (Shimomura and Bremel, 1988; Byatt et al., 19901. Like PLs from other non-primate species, the amino acid sequence of bPL has a higher percent homology to bovine prolactin (bPRL1 than to bovine growth hormone (bGH). Although biological activity of bPL has not been examined in the cow, bPL has been tested in explants of lactating bovine mammary gland and found to have comparable activity to oPRL and bPRL (Shamay, 1989). Heteroiogous systems have also shown that bPL possesses lactogenic activity in mouse mammary gland explants (Buttle and Forsyth, 1976), PRL-like and GH-like (Eakle et al., 1982; Murthy et al., 1982; Arima and Bremel, 1983) activities in radioreceptor assays and mitogenic activity in the Nb2 lymphoma cell assay (Schellenberg and Friesen, 1982). Both the presence of multiple receptor classes and inherent species differences in these systems often complicate and interfere with interpretation of the results. Therefore, identification of a target

tissue containing receptors specific for bPL may allow one to gain a clearer understanding of the biological role of bPLI as well as facilitate characterization of receptor/ ligand interaction. The present paper characterizes the binding of bPL to bovine endometrial membranes. Our findings suggest the presence of a distinct bPL receptor in endometrium. Materials and methods Hormones

Recombinant methionyl human growth hormone (hGH) was obtained from Genentech (San Francisco, CA, U.S.A.). Recombinant methionyl bGH and bPL were prepared at Monsanto. Ovine prolactin (oPRL) was purchased from Sigma (St. Louis, MO, U.S.A.). Preparation of radio~~beled hormones

[““I]hGH was prepared using a chloramine T procedure previously described (Greenwood et al., 3963). To 1 mCi carrier-free Na”“I (10 ~1) were added 5 pg of hormone in 0.4% (w/v) NaHCO, (20 ~0, 25 ~1 of 0.25 M Na,PO, (pH 7.4) and 15 ~1 0.01% (w/v> chloramine T. The reaction mixture was stirred for 5 min at room temperature (RT) prior to the addition of 10 ~1 0.1% (w/v) sodium metabisulfite (freshly prepared). Following an additional 2 min incubation, 150 ~1 of 0.25 M Na,PO, was added and the reaction mixture was applied to a Sephadex G-50 column to separate radiolabeled hGH from unincorporated [ ‘251]iodine. Bovine PL and bGH were iodinated by the Enzymobead/ lactoperoxidase reaction (Bio-Rad) according to the manufacturer’s instructions. Specific activity was determined by trichloroacetic acid precipitation prior to and following separation of radiolabeled hormone and free iodine. The specific activities of the radiolabeled hormones ranged from 100 to 200 yCi/,ug protein. Each radioligand migrated as a single radiographic band on SDS-PAGE. Membrane preparation

Uteri from 3- to 7-month pregnant, non-lactating heifers were collected at the time of sacrifice and placed immediately on ice. After all cotyledons and fetal-associated tissues were cut away

231

from the surface of the uterus, the endometrial membrane layer was carefully separated from the underlying uterine tissue using a scalpel. The tissue was finely minced and 100 g of tissue (wet weight) was suspended in 300 ml of buffer (25 mM Tris-HCl (pH 7.5), 10 mM M&l,, 300 mM sucrose, 500 KI units/ml aprotinin and 1 mM phenylmethylsulphonyl fluoride (PMSF)) and blended 30 s at high speed. The suspension was then filtered through two layers of gauze, homogenized for 30 s with a Polytron homogenizer, and centrifuged at 20,000 X g for 20 min at 4 * C. The resulting supernatants were then centrifuged at 100,000 X g for 60 min. The membrane pellet, consisting of the microsomal fraction, was resuspended in 25 mM Tris-HCI (pH 7.51, containing 10 mM MgCl,, aliquoted and stored at -80 o C until binding assays were performed. Where indicated, the microsomal fraction was thawed and MgCl, was added to a final concentration of 4 M, vortexed and incubated for 5 min at 25°C in an attempt to remove endogenously bound hormones. The suspension was diluted IO-fold with 25 mM Tris-HCl (pH 7.5) and the membranes were pelleted at 100,000 x g for 60 min at 4 o C. The MgCl,-treated microsomal fraction was resuspended in 25 mM Tris-HCl (pH 7.5) containing 25 mM MgCl,. The protein concentrations of non-treated and MgCl,-treated membranes were determined by the method of Bradford et al. (1976), using bovine serum albumin (BSA) as a standard, prior to performing receptor binding assays. Receptor

binding. In general, non-specific binding constituted 30-40% of the total amount of [‘*“I]hGH binding and 40-50% of the total amount of [‘*‘I]bPL binding. MgCl, treatment of the microsomal membranes resulted in slightly lower nonspecific binding. For competition assays, binding was assessed in the absence and presence of unlabeled bPL, hGH, bGH and oPRL at concentrations ranging from 0.008 to 2200 ng/ml. Scatchard analyses were performed with the program LIGAND (Munson and Rodbard, 1980). Human GH rather than bovine PRL was used as a tracer because of its higher bindability, characteristic to PRL receptor in several bovine tissues (Gertler et al., 1983; Cohen et al., 1987; Elberg et al., 1989). The reversibility of [ ‘2”I]bPL and [ ““I]hGH binding to endometrial membranes was assessed by incubating approximately 150,000 cpm with 0.2 mg membrane protein for 18 h at 25°C. At the end of the binding preincubation period excess umabeled hormone (2 pg) was added. The material was centrifuged at the indicated times and the amount of bound radiolabeled ligand was measured as described previousIy. Statistics

Figs. l-3 show the results of a representative experiment in which the mean f SE of triplicate determinations of each point were calculated. Each experiment was repeated at least 3 times. When no error bars are visible for particular data points, the error bar was smaller than the size of the symbol.

biff~~~g assays

Binding of radiolabeled hormones to endometrial membranes was performed using methods previously described (Haro et al., 1984) with minor modifications. Iodinated hormone ( - 180,000 cpm. 5-15 fmol) was incubated with various amounts of membrane protein in 25 mM Tris-HCl (pH 7.51, 10 mM MgC12 and 0.1% (w/v> BSA (final volume = 0.7 ml) for 18-20 h at 4 or 25 ’ C. The incubation was terminated by centrifugation at 12,000 X g for 15 min and aspiration of the supernatant. Specific binding was calculated by subtracting the amount of binding that occurred in the presence of excess (2 pg/ml) unlabeled hormone (non-specific binding) from the total

Results

initial experiments were performed to determine the existence of specific binding of [“2”I]bPL and [1251]hGH (an indicator of prolactin binding) to endometrial membranes isolated from the uteri of pregnant cows. The specific binding of both hormones to endometrial membranes is shown in Fig. 1. Binding of both hGH and bPL was dependent on the concentration of membrane protein in the assay. Specific binding of both [“‘I]hGH and [‘2sIlbPL was eIevated appro~mately 4-fold after preincubation of membranes with 4 M MgCl,. It is evident from this comparison that

232

6-

1

MEMBRANE

Fig. 1. The effect of membrane non-treated bovine endometrial

0

200

100

0

PROTEIN

200

400

MEMBRANE

(ug)

600

PROTEIN

800

(ug)

protein concentration on the binding of [‘*‘I]hGH (A) and [‘*‘I]bPL (B) to M&I,-treated and membranes. [“SI]hGH or [‘251]bPL ( * 180,000 cpm, 5-15 fmol) was incubated with MgC12-treated (0) or non-treated (0) endometrial membranes for 20 h at 25 o C.

endometrial membranes contain nearly 4 times more lactogen (hGH) binding than bPL binding. Fig. 2 shows the effect of incubation time and temperature on the specific binding of hGH and bPL to endometrial membranes. Specific binding of [ lZ51]hGH (Fig. 2A) at both 4°C and 25’ C reached a maximum at 24 h incubation and remained stable until 48 h. Binding of [‘251]bPL at

4” C was not maximal until 72 h of incubation, while maximal binding was achieved within 24 h at 25 “C. Interestingly, specific binding of [ ‘251]bPL at 25 OC was not stable beyond 24 h and decreased dramatically over extended incubation times. An incubation time of 24 h at 25 o C was routinely used for subsequent studies to assess the binding of both hGH and bPL because this

A

B

0

10

20 TIME

30 (h)

40

50

0

20

40 TIME

60

80

100

(h)

Fig. 2. The effect of incubation time and temperature on the binding of [ ‘251]hGH (A) and [‘Z51]bPL (B) to MgCl,-treated bovine endometrial membranes. 25 @g (A) and 200 @g (B) of microsomal protein were incubated with radiolabeled hormone ( _ 180,000 cpm) at either 4°C (0) or 25 “C (0) for the indicated times. Non-specific binding, determined at each time point, remained constant over the time course of the experiment.

233

protoco1 gave the highest level of binding within a convenient time period. The binding of [‘251]hGH and [12511bPL at 25 o C was partially reversible after the addition of a 2000-fold excess of unlabeled hormone (Fig. 3). Dissociation of both ligands appears to be biphasic with a rapid initial displacement of approximately 20% over the first 2.5 h and gradual increase to a maximum of 40-60% at 24 h. Competition by various hormones for hGH and bPL binding sites on the endometrium is shown in Fig. 4. Increasing concentrations of unlabeled hGH (8 ng/ml to 2 pg/ml) caused a dose-dependent inhibition of binding of [ “51]hGH to endometrial membranes (Fig. 4A). The halfmaximal competition of iodinated hormone was achieved with 2.4 ng/ml. Scatchard analysis of the competitive binding curves revealed a curvilinear plot (Fig. 4C), which was resolved into a higher affinity, lower capacity binding site with an apparent dissociation constant (K,) of 6.4 x 10-r’ M and a lower affinity, higher capacity site with an apparent I(, of 1.7 X 10m9 M. The maximum binding capacity for the high affinity site was estimated to be 39 fmol/mg membrane protein. The specific binding of [‘*‘I]hGH was also inhibited by the addition of bPL and oPRL (Fig. 4A), but the potencies of these hormones in compet-

ing for [‘251]hGH binding sites were 2.8% and 23.5% that of hGH, respectively. The dose-response curves for competition of binding [ 12”I]hGH by oPRL, but not bPL, was parallel to those with hGH. Concentrations of unlabeled bGH up to 2 pg/ml were ineffective in competing for [ ‘251]hGH binding and no specific binding of [‘““I]bGH could be detected (results not shown). Increasing concentrations of unlabeled bPL (8 ng/ml to 2 pg,/ml) competed for binding of [‘2”I]bPL in a doss-dependent manner with halfmaximal competition occurring at approximately 3 ng/ml (Fig. 4B). Although the specific binding of [ ‘2”I]bPL to endometrial membranes could be inhibited in a dose-dependent manner by unlabeled hGH and bGH, the concentrations of hGH and bGH required to inhibit 50% of the binding of [iZ”I]bPL to bovine endometrial membranes was shown to be lOO-fold higher than the necessary level of unlabeled bPL. Extremely high levels of oPRL were needed to compete for [12’I]bPL binding. The 40% inhibition observed at 2 pg/ml oPRL may be accounted for by the presence of low levels of contaminating oGH in the preparation. As seen for hGH binding, Scatchard analysis of the data from the competition assays revealed curvilinear plots. An apparent dissociation constant of 4.1 X lo-” M was calculated for the high affinity site. The number of high affinity receptors for bPL was estimated to be 21 fmol/mg membrane protein in endometrial tissue. Discussion

“P

30 TIME (h)

Fig. 3. [‘251]hGH and [ ‘251]bPL dissociation kinetics from bovine endometrium. The dissociation of specifically bound [‘2sflhGH (0) or [‘%lbPL (0) from membranes (200 /*g) preincubated with either hormone for 18 h at 25 “C was initiated by the addition of 2 p.g unlabeled hormone. Specific and non-specific binding were determined at each time point. Each point represents the mean of triplicate determinations.

Recent studies of Freemark et al. (1987) strongly suggest the existence of a distinct and unique PL receptor in fetal and maternal sheep liver. It has been postulated that the biological actions of PL in the tissues of postnatal animals are mediated through binding to somatotrophic (GH) and/or lactogenic (PRL) receptors. In a recent investigation by Grissom and Littleton (1988) the ontogeny of lactogenic binding activity in several tissues of pregnant rabbits was examined. Significant binding was measured in both uterine and ovarian tissue. Sub-uterine localization studies found prolactin binding predominantly to the endometrium.

60

0 10"

10°

HORMONE

10P

103

10'

10'

(ng/ml)

HORMONE

(ng/ml)

0.08-

0.06-

i 10

0

BOUND Competition was incubated

( A ). hPL

(

(fmoles)

binding of [“‘I]hGH

with 50 ~g of endometrial

A ),

oPRL

(01 or bGH

CO).

than IO’Z of the mean value. Scatchard radiolabeled

hormone

(A)

BOUND and [“‘l]bPL

membranes

(B)

to bovine endometrium.

the mean of triplicate

determinations.

analysis of the data from A and B far competition D. respectively.

ahxissa

Radiolabeled

The

ordinate

represents

the amount of hGH f C) or bPL fD)

In evaluating the presence of specific binding activities one must be concerned that the appropriate conditions have been chosen to carry out the binding analysis. A negative result may be due to inappropriate assay conditions rather than the actual absence of binding sites since both could lead to similar measurements. Therefore, several parameters were tested to optimize the conditions for studying both hGH- and bPL-binding to endometrial membranes. The conditions included pretreatment of the membranes with MgCl, and variations in temperature and time of

30

40

(fmoles) hormone

for 20 h at 25 OC in the presence of varying concentrations

Each point represents

is shown in C and

20

( - X5.000cpm)

of unlabeled

In all cases the SEM

binding with homologous

the ratio of bound/free

htiH

was less

ligand to the

hormone

and the

bound.

incubation. In contrast to recent findings (Hill et al., 19SS>, MgCI, treatment of endometrial membranes significantly increased the specific binding of both [“‘I]hGHand [ ‘2’I]bPL several fold. Although incubation of membranes with 4 M MgCl, has been reported to facilitate removal of endogenously bound hormones (Freemark and Handwerger, 19861, increases in specific binding could also result from the removal of peripheral or contaminating membrane-associated protein. MgCl, treatment of endometrial membranes resulted in a loss of 50-60% of the total protein

235

(data not shown). Therefore, the increase in binding could be accounted for by enrichment of membrane-bound bPL and bPRL receptors rather than removal of endogenously bound ligand. The presence of distinct bPL receptors in the endometrium is supported by several lines of evidence. First, specific, high affinity binding for bPL was measured, with an apparent dissociation constant of 4.1 x lo-” M and binding capacity of 21 fmol/mg protein. These measurements are in agreement with the affinity constants and binding capacities of sheep liver membranes previously reported for oPL (Chan et al., 1978; Freemark and Handwerger, 1986), K, = 1.1 X lo- ‘” M and 1.7 x lo-“’ M, B,,, = 10.5 fmol/mg and 31.6 fmol/mg, respectively. Secondly, specificity of binding for bPL was clearly demonstrated by competition binding analysis. Although hGH and bGH were seen to compete for binding of [ ‘2”I]bPL, levels lOO-fold the concentration of bPL were needed to give an equal level of competition. The ability of bPRL to compete with [‘*‘I]bPL was even lower (approx. 2000-fold). Thirdly, there was an absolute lack of detectable specific binding using [‘2sI]bGH as a tracer. Fourthly, when [“‘I]hGH was used as a tracer, bGH was unable to compete with its binding, even at a concentration of 2 Kg/ml. On the other hand bPRL could inhibit completely the binding of [ ‘2”I]hGH, although approximately 4-fold higher concentrations were required as compared to hGH. The relative ability of bPL to compete with [‘2”I]hGH was an additional 5-fold lower. These results suggest that: (a> there is no somatogenie-type receptor in bovine endometrium; (b) hGH binds mainly, if not solely to lactogen-type (PRL) receptors; (c) like in other bovine tissues (Gertler et al., 1983; Cohen et al., 1987; Elberg et al., 1989) the affinity of hGH for lactogen-type receptors is higher than that of bPRL; (d) [““I]hGH may also bind to bPL receptors (low affinity receptors) with approximately 30-fold lower affinity. Thus, in addition to the lactogen (PRL) receptors whose presence has previously been demonstrated in the endometrial tissue from rabbits (Grissom and Littleton, 1988) and sows (Young and Bazer, 19891, our results clearly demonstrate the existence of a distinct type of receptors, which exhibit high specificity for bPL,

in bovine endometrium. Furthe~ore, the difference observed in the binding kinetics of [‘*“I]hGH supports this conclusion. Although the physiological role of PL in fetal development is poorly understood, its growthpromoting activity has been demonstrated convincingly in hypophysectomized rats (Chan et al., 1976). Demonstration of distinct receptors for PL in endometrial tissue suggests that PL may also exert direct effects on maternal metabolism. Investigation of bPL-binding capacity of endometrium at various stages of gestation may reveal whether any relationship exists with plasma bPL concentrations. It is conceivable that bPL serves a role in the maintenance of pregnancy as well as sustaining normal fetal growth and development. Acknowledgements

We are grateful to Drs. Kevin Glenn and Gwen Krivi, each of whom reviewed the manuscript and made imp(~rtant suggestions for its improvement. In addition, we would like to thank Mary Cerveny for her help in preparing the manuscript. References Arima, Y. and Bremel, R.D. (1983) Endocrinology 113, 21862194. Bradford, M.M. (1976) Anal. Biochem. 72, 248-254. Buttle, H.L. and Forsyth, LA. (19761 J. Endocrinol. 68, 141146. Byatt, J.D., Shimomura, K., Duello, R.D. and Bremel, R.D. (19861 Endocrinology 119, 1343-1350. Byatt, J.D., Welply. J.K., Leimgruber, R.M. and Collier, R.J. (19901 ~ndocrinolo~ fin press). Chan. J.S.D., Robertson, H.A. and Friesen, H.G. (1976) Endocrinology 98, 65-76. Chan, J.S.D.. Robertson, H.A. and Friesen, H.G. (1978) Endocrinology 102, 632-640. Cohen, R., Ashkenazi, A., Elberg, G. and Gertler, A. (1987) J. Recept. Res. 7, 921-936. Eakle, K.A., Arima, Y., Swanson, P., Grimek. H. and Bremel, R.D. (19821 Endocrinology 110, 1758-1765. Elberg, G., Ashkenazi, A. and Gertler, A. (1989) Mol. Cell. Endocrinol. 61, 77-85. Forsyth, IA. (19861 J. Dairy Sci. 69, 886-903. Freemark, M. and Handwerger. S. (19831 Endocrinology 112, 402-404. Freemark. M. and Handwerger, S. (1984) Am. J. Physiol. 246, E2I-E24. Freemark, M. and Handwerger, S. (1986) Endocrinology 118, 613-618.

236 Freemark, M., Comer, M., Korner, G. and Handwerger, S. (1987) Endocrinology 120, 1865-1872. Freemark, M., Comer, M., Mularoni, T., D’Ercole, A.J., Grandis, A. and Kodack, L. (1989) Endocrinology 125, 1504-1512. Gertler. A., Ashkenazi, A. and Madar, Z. (1984) Mol. Cell. Endocrinol. 34. 5 l-57. Greenwood, F.C., Hunter, W.M. and Glover, J.S. (1963) Biochem. J. 89, 114-123. Grissom, F.E. and Littleton. G.K. (1988) Endocr. Res. 14, l-19. Handwerger, S., Fellows, R.E., Crenshaw, MC., Hurley, T., Barrett, J. and Maurer, W.F. (1976) J. Endocrinol. 69. 133-137. Haro. L.S., Collier, R.J. and Talamantes, F.J. (1984) Mol. Cell. Endocrinol. 38, 109-116. Hill, D.J., Freemark, M., Strain, A.J., Handwerger, S. and Milner, R.D.G. (1988) J. Clin. Endocrinol. Metab. 66, 1283-1290. Lesniak, M.A., Gorden, P. and Roth. J. (1977) J. Clin. Endocrinol. Metab. 44, 838-849.

Munson, P.J. and Rodbard, 0. (1980) Anal. Biochem. 107, 220-239. Murthy, G.S., Schellenberg, C. and Friesen, H.G. (1982) Endocrinology 11 I, 2117-2124. Schellenberg, C. and Friesen, H.G. (1982) Endocrinology 111, 2125-2128. Shamay, A. (1989) Control of Proliferation of Bovine Mammary Epithelial Cells in vitro by Hormones and Growth Factors. Ph.D. Thesis. The Hebrew University of Jerusalem, Rehovot, Israel. Shimomura. K. and Bremel, R.D. (1988) Mol. Endocrinol. 2. 845-853. Shiu, R.P.C., Kelly, P.A. and Friesen. H.G. (1973) Science 180, 96X-970. Talamantes, F. and Ogren, L. (1988) in The Physiology of Reproduction (Knobil, E. and Neill. J., eds.). pp. 20932144, Raven Press, New York. Tsushima, T. and Friesen, H.G. (1973) J. Clin. Endocrinol. Metab. 37. 334-337. Young, K.H. and Bazer, F.W. (1989) Mol. Cell. Endocrinol. 64, 145-154.