Evolution of Prolactin Receptors in Rabbit Mammary Gland During Pregnancy and Lactation JEAN DJIANE,* PHILIPPE DURAND,* AND PAUL A. KELLYf *Laboratoire de Physiologie de la Lactation, Institut National de la Recherche Agronomique C.N.R.Z., 78350 Jouy-en-Josas, France, and \MRC Group in Molecular Endocrinology Centre Hospitalier de I'Universite Laval 2705 Boulevard Laurier, Quebec GIV 4G2, Canada linear increase in the weight of the mammary gland from Day 14 of pregnancy until Day 6 of lactation and this increase is unaffected (except at Day 6 of lactation) by CB 154 treatment. It was observed that prolactin receptors remain at a relatively low and constant level while mammary development (mammogenesis) takes place. The onset of milk secretion parallels a striking increase (>500%) in the number of prolactin receptors (expressed per total mammary gland or per cell) after parturition has occurred. These results are discussed with emphasis on the mechanisms through which hormonal balances during pregnancy and lactation may modulate the amount of receptors per cell, hence its sensitivity to lactogenic hormones. (Endocrinology 100: 1348, 1977)
ABSTRACT. The numbers and affinity of prolactin receptors in the rabbit mammary gland were determined during pregnancy and early lactation under conditions in which the endogenous lactogenic hormone was depleted by means of the compound CB 154. In untreated rabbits the number of prolactin binding sites per mg protein increased from 25 ± 3 (SE) fmol at Day 14 of gestation to 54.8 ± 5.8 fmol/mg at Day 22, after which binding declined to 14.2 ± 8.5 fmol/mg, then increased in late pregnancy and during lactation to 110.5 ±11.5 fmol/mg at Day 28. In animals treated with CB 154, binding was always higher than in non-treated animals, with a peak during pregnancy of 149 ± 24 fmol/mg at Day 22. After declining in late pregnancy, the number of receptors was highest at Day 6 of lactation (257.4 ± 34.6 fmol/mg). There is an almost
T
HE OCCURRENCE of a specific receptor for prolactin in mammary tissue wasfirstestablished by Turkington (1) using prolactin covalently bound to Sepharose beads, and was confirmed by Falconer and Birkenshaw (2,3) by means of autoradiographic methods using radioiodinated prolactin. Shiu et al. (4), working with rabbit mammary glands, found this receptor to be located in the particulars membrane fraction obtained at 100,000 x g and thoroughly investigated the kinetic aspects of hormonereceptor interactions on this preparation (5). The receptor was purified by the same authors (6) and recently a specific antibody against this protein was prepared (7), which led them to demonstrate unequivocally that the action of prolactin on mammary explants stringently required the presence
Received September 7, 1976. Supported in part by the Centre National de la Recherche Scientifique (grant no. 655-2152) and the Delegation Generale a la Recherche Scientifique et Technique (grant no. 75-7-1651).
and availability of the receptor on the membranes. The induction of the receptor molecule under various endocrine conditions has been extensively studied in rat liver, since this organ seems to contain this protein in sufficient amounts and its membranes are easier to fractionate than those of the mammary gland. Kelly et al. (8) first detected higher prolactin binding in female than male rat liver, and noted that in females there was a marked increase in binding following puberty and during pregnancy. Posner et al. (9) then described an estrogenmediated increase of binding sites in liver, but later (10) demonstrated that this phenomenon could be amplified or even mediated by prolactin itself. Costlow et al. (11) reached similar conclusions, and recently the same group (12) found that prolactin binding estimated in slices of rat mammary gland during pregnancy and lactation increases to a considerable extent at the time of parturition, whereas it is almost undetectable at earlier periods. This latter
1348
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
PROLACTIN RECEPTORS IN THE MAMMARY GLAND
finding could have been due to endogenous saturation of the receptors, since in rats (as well as primates and ruminants) the placenta is known to secrete large amounts of a lactogenic hormone (13). In the rabbit, however, there does not appear to be a placental lactogen (13,14). In the present work the rabbit has been used to assess and titrate the amount of prolactin receptors during pregnancy and lactation, as well as the state of occupancy of these receptors by endogenous prolactin at various times during these periods. This was achieved by using the pharmacological compound Bromocryptine (CB 154, Sandoz), which is known to depress selectively prolactin secretion from the pituitary gland (15,16), and shows no effect on in vitro prolactin binding by rabbit mammary gland membranes (5).
1349
Ultraturrax homogenizer) was not filtered through cheese cloth in order to avoid any loss of material; it was rehomogenized in a 100 ml Potter homogenizer, centrifuged first at 15,000 x g (20 min), the pellet was discarded and the supernatant centrifuged at 100,000 x g (90 min) as described in the original method (4). The pellet obtained accounted for about 70% of the total binding sites of the tissue, as previously shown by Shiu et al. (4), whereas 30% of these sites remained with the first 15,000 x g pellet. With lactating mammary glands, slight contaminations of the pellets with precipitated casein micelles frequently occurred and were not eliminated to avoid any loss of material. The 100,000 X g crude membrane pellets were suspended in 25 mM Tris buffer, pH 7.6, containing 10 mM MgCl2. Protein contents were assayed according to the method of Lowry et al. (17) and the suspension was frozen (—20 C) until the binding assay was performed. Prolactin iodination
Materials and Methods Animals New-Zealand rabbits (6 months old) were used during their first pregnancy and kept in air and temperature conditioned cots. Mating day was considered as Day 0 and the animals were sacrificed at 4 day intervals, starting on Day 14. Some experiments were performed at Day 6, but very low values for receptor content were obtained and, since the mammary region at this time consists mainly of muscle and connective tissue, these data were not included. The in vivo desaturation procedure consisted of treatment with 2 mg of CB 154 injected sc at 36, 24 and 12 h before the animals were sacrificed. The 36 h pretreatment (3 injections) has been shown to produce maximal desaturation of prolactin, since no additional sites are unmasked when the treatment is extended to 72 h (unpublished observations). Mammary membranes Mammary glands were removed, cleaned and frozen at —80 C and then stored at —20 C. Aliquots of 5 g of this tissue were used for the membrane preparation according to Shiu et al. (4), with the following modifications: the first homogenate (0.3M sucrose, 0 C, prepared with an
Ovine PRL (NIH-P-S7, 24 IU/mg) was iodinated according to the method of Greenwood et al, (18), as modified by Kann (19). The specific activity of the prolactin was approximately 80 fiCi/fig. The specificity of its binding, as assayed with varying hormonal dilutions, appears to be satisfactory (20) and non-specific binding averaged from 15-20% of total radioactivity bound, when to the labelled hormone (1 ng) a 1000-fold excess of unlabelled prolactin (1 ju,g) was added. Binding assay The assay was performed according to Shiu and Friesen (4) at 20 C in the above mentioned Tris/MgCl2 buffer (pH 7.6) containing 0.1% bovine serum albumin. Four hundred fig of membrane protein were incubated in a final volume of 0.5 ml with approximately 105 cpm labelled prolactin and containing increasing amounts of unlabelled prolactin (50 to 1000 fmol for the assays, and 1 fig for assessing nonspecific binding). Up to 70% of the iodinated prolactin molecules could be specifically bound to prolactin receptors in receptor rich membranes (e.g., lactating animals 5-6 days, desaturated with CB 154) at protein amounts up to 600 fig. The specific binding plateaued at higher protein concentra-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
1350
DJIANE, DURAND AND KELLY
tions, but the variability of the assays at these high concentrations of protein prevented any clear-cut conclusions as to whether this phenomenon was due merely to the kinetic parameters of the saturation reaction as recently discussed by Chang et al. (21), to a "protein effect" at high concentrations, or to the inability of a fraction of the iodinated molecules to interact with the receptors. In all cases, when used under the conditions of the "specific binding" assay (i.e., difference between radioactivity bound in the presence of an excess of unlabelled hormone and radioactivity bound in its absence), or at the onset of saturation (see below, Fig. 2), the iodinated prolactin yielded entirely consistent results, which appear to ensure at least the relative accuracy of the amounts of available sites titrated. Chemicals CB 154 (2 bromo-a-ergocryptine-methanesulfonate) was kindly supplied by Sandoz, Basel, and bovine serum albumin (fraction V) from I.C.N. Pharmaceutical, Cleveland, Ohio. All reagents were of analytical grade.
Results Figure 1 shows the time course for the incubation of mammary gland membranes with ovine [125I]iodo-PRL. An incubation period of 6 h was necessary for equilibrium,
Endo • 1977 Vol 100 • No 5
which agrees with the kinetic data of Shiu and Friesen (5). However, a small, but reproducible intermediary plateau has been observed, at approximately half-saturation (2-3 h of incubation), suggesting that two distinct sets of binding sites might be present. This phenomenon seems also to be observed in Shiu and Friesen's data (ref. 5, Fig. 2, 23 C). Further experiments will be necessary to decide if this could be due to the fact that half of the receptor sites are buried in inverted vesicles, as has been demonstrated by Bennett and Cuatrecases for insulin receptors (22). In any case, it seems clear that at least 80% of the high affinity sites have been titrated under our experimental conditions, provided that no loss of receptor molecules has taken place when the cell lipids were removed after the first centrifugation: this seems difficult to assert, since no equilibration of this lipid layer with [125I]iodo-prolactin has been attempted, owing to the fact that it could not be mixed with or suspended in an aqueous phase. The problem of possible contributions of various cell populations to prolactin binding is further considered below (see Discussion). Saturation curves of prolactin specifically bound to mammary gland membranes from rabbits at various stages of pregnancy or
FIG. 1. Binding kinetics of ovine [125I]iodo-PRL to membranes from animals at 28 days of lactation O O and from animals at 14 days of lactation treated with CB 154 to desaturate the binding sites • • (see Materials and Methods). Each point represents the mean (± SEM) for 3 animals. Incubation was performed with 50,000 cpm of ovine [125I]iodo-PRL during the periods indicated, at 20 C. Results are expressed as fmoles specifically bound/400 fig protein.
2.
as i
23
hours.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
PROLACTIN RECEPTORS IN THE MAMMARY GLAND
1351 Uxto D 28 (4)
LOCtO D 14(4)
D.26{4)
200
400
600
800
1000 Total Prolactin . ( f moles)
FiG. 2. Prolactin binding as a function of increasing amounts of prolactin at the indicated stages of pregnancy and lactation (Lacta.). The competition data with increasing amounts of unlabelled prolactin are expressed as saturation curves. Each point represents the mean (±SEM) for the number of animals indicated in parentheses. These data are obtained with membrane preparations from animals which were not desaturated in vivo by CB 154 treatment.
lactation (not treated with CB 154) are shown in Fig. 2. Representative Scatchard plots calculated from saturation kinetics (23) are depicted in Fig. 3. Panel A shows data from untreated animals, panel B data from CB 154-treated animals (original saturation curves not shown). The binding affinity in all cases remained constant (Ka = 2.5 to 3.2 x 109M-1), whereas the number of binding sites varied depending on the stage of pregnancy or lactation. The development of prolactin binding sites expressed per mg of membrane protein in rabbit mammary glands during pregnancy and lactation in untreated and CB 154-treated animals is shown in Fig. 4. In untreated animals, the number of prolactin binding sites was 25 ± 3 fmol/mg protein at day 14 of pregnancy then the binding increased to reach 54.8 ± 5.8 fmol/mg at day 22
of pregnancy. This value is significantly higher than the values observed at day 14 (P < 0.01) and at day 26 (P < 0.01), but not significantly different from the values observed at day 18 and day 29. Binding increased during lactation to a maximum of 110.5 ± 11.5 fmol/mg at day 28 (P < 0.01 when comparing day 28 of lactation with day 22 of pregnancy). In animals treated with CB 154, binding was always higher than in non-treated animals, but the variability of the values obtained was greater, hence the peak observed at day 22 (140 ± 24 fmol/mg) did not differ significantly from the other values of gestation. In early lactation, binding increased strikingly. The highest number of receptors was observed at day 6 (257.4 ± 34.6 fmol/ mg) and significantly declined at day 14 to 172 ± 10 fmol/mg (P < 0.01). These results
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
DJIANE, DURAND AND KELLY
1352 A = untreated animals a — D [>28 lactation • — a D 14 H o—o
D 22 pregnancy
A—^
D 18
»
.--..
D 14
M
Endo • 1977 Vol 100 • No 5
B= CB154 treated animals u. CD
0.8.
. \
A—A D 6 lactation • — • DI4 '1 V \
A — A D 18 pregnancy • _ . DI4 '• \
0.6.
FIG. 3. Representative scatchard analyses of ovine [l25I]iodoPRL binding in membranes from untreated rabbits (A) and CB 154-treated rabbits (B).
0.4. 0.4. 0.3.
0.2. 0.2. O.I.
25
50 75 B f moles oPRL bound
«
5O
22 26 OftEGNANCV
FlG. 4. Evolution of the number of prolactin receptors (expressed as fmoles prolactin bound per mg of membrane protein) throughout pregnancy and lactation. The values were obtained by analysis of the Scatchard plots (Fig. 3) of the saturation data (Fig. 2). (n) = number of animals, • • = animals desaturated in vivo by CB 154 (36 h), O 0 = controls, a n d l = ±SEM.
100
I5OB
f moles oPRL bound
clearly demonstrate that a) in vivo desaturation with CB 154 for 36 h results in unmasking a considerable amount of endogenously saturated binding sites, which depending upon the period considered may vary from 50% to more than 70% of the number of total sites (titrated under our desaturation conditions); b) this total amount fluctuates to a considerable degree, with a steady increase during early pregnancy, a decrease in late, a sharp rise at the onset of lactation and a second decrease following the establishment of lactation. Since the amount of crude membrane protein in the mammary gland itself varies to a considerable degree, with a rise at late pregnancy, due to the development of an endoplasmic reticulum characteristic of increased protein synthesis (24), it was considered relevant to express the number of receptors per total mammary gland or per mg DNA, the latter accounting for the number of cells. Figure 5 shows the weight increase of the mammary glands of untreated and CB 154treated animals during the experiment and demonstrates that 36 h depression of prolactin secretion does not modify this increase, except at the early lactational stage,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
PROLACTIN RECEPTORS IN THE MAMMARY GLAND
1353
FIG. 5. Weights of rabbit total mammary glands during the course of pregnancy and lactation. For other details see legend to Fig. 4.
18
when milk secretion is known to be highly sensitive to prolactin levels (25) which may account for the decrease observed. It has been verified repeatedly (results not shown) that the 5 g samples taken from the mammary glands to perform the various assays are representative of entire mammary tissue. Figure 6 portrays the variations of receptor levels in whole mammary gland (panel A) in untreated and CB 154-treated animals and reveals that the sudden increase in prolactin binding sites at parturition is greatly amplified when compared to Fig. 4, since mammary growth is still linear at this period. Panel B shows the amount of receptors per mg P DNA (i.e., per cell): this figure does not point out a real decrease of receptors during late pregnancy (except for CB 154-treated animals at 29 days of pregnancy), as might have been postulated, if only protein concentrations were considered (Fig. 4), and shows that the striking increase of binding sites at parturition is even much higher than may have been expected from the receptor protein ratios. In addition this figure suggests that at parturition almost
22 26 PREGNANCY
28 Doys
all sites are desaturated, a paradox which is masked in Fig. 4 due to the high milk protein contents of the controls (untreated) at the onset of lactation. This apparent desaturation may be artefactual, as discussed below. Discussion Mammary development and the onset of lactation are usually considered to be due to sequential phases of hormonal stimulation (26) clearly separated by parturition, which causes both prolactin secretion at high levels (27,28) and the complete withdrawal of progesterone. It has recently been demonstrated (29) that in the pseudopregnant rabbit prolactin stimulation acts most probably at the transcriptional level and that large quantities of mRNAs for milk proteins are detectable only in the presence of prolactin and in the absence of progesterone. The present results demonstrate that this biphasic aspect of mammary growth and secretion also applies to prolactin receptors: comparatively low levels of receptors per total mammary tissue (Fig. 6 A) or per cell (Fig. 6 B) are detected during pregnancy
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
Endo i L977 Vol 100 ( No 5
DJIANE, DURAND AND KELLY
1354
o 500.
E 400.
300.
200.
100.
5.
10.
14 18 22 26 29| Pregnancy
p
6
14 Lactation
28 days
14
18 22 26 291
6
Pregnoncy
Lactation
p
14
28 days
FIG. 6. Change in number of prolactin receptors per total mammary gland (panel A) or per mg of DNA phosphorus (P DNA) (panel B) in the course of pregnancy and lactation. The data for DNA contents of the mammary gland at the indicated periods, are those reported by Denamur (30). For other details see legend to Fig. 4.
and these levels suddenly increase by al- truly reflects a regulatory step, with no most 500% at parturition. In vivo desatura- major contribution by a desaturation phetion by CB 154 constantly increases the nomenon, as could be the case for species available binding sites by 100-200%, but secreting high levels of a placental lactothis inevitably raises the question as to genic hormone. whether the sudden burst of free receptors When comparing receptor levels of unat parturition could not be due to a desatura- treated or CB 154 desaturated animals durtion of previously occupied sites by the ing pregnancy and lactation (Fig. 6 A and B), withdrawal of some endogenous placental the question arises as to whether 36 h mammotropin as postulated by Holcomb et suppression of prolactin secretion by the al. (12) for the rat. The rabbit is considered drug could not be itself a parameter modulatnot to possess such a hormone (13,14) ing these levels: prolactin has been shown to and this is the main reason why this increase its own receptors in the liver species has been chosen to perform the pres- (10,11), hence CB 154 should decrease these ent experiments. Moreover, a series of levels. Some evidence that this might be the experiments not included in the present case at 29 days of pregnancy is provided by manuscript, demonstrate that when CB 154 the apparent desaturation of all receptors treatment is applied for two days prior to at this time, a paradox which has already parturition (which suppresses the prolactin been mentioned and may be explained, if surge at birth), such an increase of receptor one assumes that in the CB 154-treated level in the mammary gland is not observed. animals the number of receptors has Hence it may be assumed that this increase dropped approximately to control values.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
PROLACTIN RECEPTORS IN THE MAMMARY GLAND The experiments mentioned above, where suppression of the prolactin by CB 154 at the time of parturition prevents the increase of receptors, lend some support to this hypothesis. For other periods of pregnancy, where no sudden induction of receptors is observed, it seems unlikely that a short treatment with CB 154 could extensively modify receptor levels. During lactation, it appears that the number of receptors in CB 154treated animals is lower at Day 14 of lactation than at Day 6, whereas free receptors (untreated animals) remain fairly constant; this might be correlated with a decrease of prolactin levels during the second half of lactation, but more experiments are needed to establish prolactin levels and receptor variations at these periods in the rabbit. A remarkable aspect of the experiments reported consists in the demonstration of a roughly reciprocal correlation between phases of mammary growth (as estimated by the increase of total DNA) and receptor amplification (receptors per unit DNA or per cell). Thus, during the first period of pregnancy (up to 20 days), the levels of DNA do not increase to a great extent (30), whereas the number of receptors per cell is increased by a factor of 3 to 4. From this time until the day of parturition about 80% of the total mammary growth takes place and the level of receptors per cell remains constant or even slightly decreases. The fact that the main increase of receptors takes place at parturition, while 20% of the growth phenomenon is still to take place (up to the fifth day of lactation (30)) does not seem to contradict this reciprocal relationship, since the mammary cells are probably not completely synchronized. Nevertheless, this phenomenon is reminiscent of the behavior of synchronized cell cultures, where mitosis limits protein synthesis (31,32). The sudden burst of specific binding sites which occurs when mammary tissue switches from the developmental to the secretory phase raises the puzzling problem as to whether this qualitative switch could
1355
be mediated merely by the amount of information communicated to the cell through a parallel increase of receptor and hormonal levels. No definite answer can be presently suggested to this question, but it must be kept in mind that additional hormonal shifts are taking place at parturition, mainly concerning progesterone and corticosteroids, which may modulate the overall phenomenon, either independently of prolactin effects or even by participating in the regulation of prolactin receptor levels, hence of the cells sensitivity to the latter hormone. An additional aspect concerns the onset of differentiation, which is presently believed to be triggered by estrogens, in addition to prolactin (9). This may be difficult to analyze with intact mammary tissue, considering the small numbers of epithelial cells present when pregnancy starts: at 14 days pregnancy or even at 6 days (results not shown), receptors are detected, but the initial cell differentiation must have already taken place. It must be finally emphasized that no completely satisfactory parameter to express the concentration of receptors in such an heterogeneous cell population as the mammary gland is presently available. It has already been mentioned that protein concentrations are inadequate, since they include changes in a parameter such as casein synthesis, which should not be accounted for, when expressing membrane receptor concentrations. The amount of DNA reflects the total number of cells, irrespective of their nature; hence adipocytes, fibroblasts, myoepithelial cells, and others contribute to DNA, possibly without contributing prolactin receptors, although the latter point remains to be established. No specific marker enzyme or protein for the epithelial mammary cell is presently known; moreover, any such marker would be likely to fluctuate itself, according to the phases of mammary growth and secretion. Despite these shortcomings, prolactin receptor concentrations, as described in the present work, should adequately reflect the physio-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
DJIANE, DURAND AND KELLY
1356
logical regulations, since their main variations remain identical, irrespective of the parameters with which they are correlated. Acknowledgments We wish to thank Professor H. Clauser for much helpful advice during the course of this work and during redaction of the manuscript. We are greatly indebted to Dr. G. Kann for the supply of the iodinated prolactin. References 1. Turkington, R. W., Biochem Biophys Res Commun 41: 1362, 1970. 2. Falconer, I. R., Biochem J 126: 8p., 1972. 3. Birkenshaw, N., and I. R. Falconer, J Endocrinol 55: 323, 1972. 4. Shiu, R. P. C , P. A. Kelly, and H. G. Friesen, Science 180: 968, 1973. 5. Shiu, R. P. C., and H. G. Friesen, Biochem J 140: 301, 1974. 6. Shiu, R. P. C., and H. G. Friesen, J Biol Chem 249: 7902, 1974. 7. Shiu, R. P. C , and H. G. Friesen, Science 192: 259, 1976. 8. Kelly, P. A., B. I. Posner, T. Tsushima, and H. G. Friesen, Endocrinology 96: 532, 1974. 9. Posner, B. I., P. A. Kelly, and H. G. Friesen, Proc Natl Acad Sci USA 71: 2407, 1974. 10. Posner, B. I., P. A. Kelly, and H. G. Friesen, Science 188: 57, 1975. 11. Costlow, M. E., R. A. Buschow, and W. L. McGuire, Life Sci 17: 1457, 1976. 12. Holcomb, H. H., M. E. Costlow, R. A. Buschow, and W. L. McGuire, Biochim Biophys Ada 428: 104, 1976.
Endo • 1977 Vol 100 • No 5
13. Kelly, P. A., T. Tsushima, R. P. C. Shiu, and H. G. Friesen, Endocrinology 99: 765, 1976. 14. Talamantes, Jr, F., Am Zool 15: 279, 1975. 15. del Pozo, E., and E. Fliickiger, In Pasteels, J. L., and C. Robyn (eds.), Proceedings of the International Symposium on Human Prolactin, Brussels, Excerpta Medica, Amsterdam, 1973, p. 291. 16. Floss, H. G., J. M. Cassady, and J. E. Roberts, J Pharm Sci 62: 699, 1973. 17. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. RandallJ Biol Chem 193: 265, 1951. 18. Greenwood, F. C., W. M. Hunter, and J. S. Glover, Biochem J 89: 114, 1963. 19. Kann, G.,CRAcadSci [D](Paris) 272: 2808,1971. 20. Djiane, J., and G. Kann, CR Acad Sci [D] (Paris) 280: 2785, 1975. 21. Chang, K. J., S. Jacobs, and P. Cuatrecasas, Biochim Biophys Ada 406: 294, 1975. 22. Bennett, V., and P. Cuatrecasas, Biochim Biophys Ada 311: 362, 1973. 23. Scatchard, G., Ann NY Acad Sci 51: 660, 1949. 24. Bousquet, M., J. E. Flechon, and R. Denamur, Z Zellforsch 96: 418, 1969. 25. Taylor, J. C , and M. PeakerJ Endocrinol 67: 313, 1975. 26. Denamur, R . J Dairy Res 38: 237, 1971. 27. Kann, G., and R. DenamurJ Reprod Fertil 39: 473, 1974. 28. Morishige, W. K., G. J. Pepe, and I. Rothchild, Endocrinology 92: 1527, 1973. 29. Houdebine, L. M., and P. Gaye, Mol Cell Endocrinol 3: 37, 1975. 30. Denamur, R., CR Acad Sci [D] (Paris) 256: 4748, 1963. 31. Martin, D., Jr., G. M. Tomkins, and D. Granner, Proc Natl Acad Sci USA 63: 842, 1970. 32. Johnson, T. C , and J. J. Holland, J Cell Biol 27: 565, 1965.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
ABSTRACT. The numbers and affinity of prolactin receptors in the rabbit mammary gland were determined during pregnancy and early lactation under conditions in which the endogenous lactogenic hormone was depleted by means of the compound CB 154. In untreated rabbits the number of prolactin binding sites per mg protein increased from 25 ± 3 (SE) fmol at Day 14 of gestation to 54.8 ± 5.8 fmol/mg at Day 22, after which binding declined to 14.2 ± 8.5 fmol/mg, then increased in late pregnancy and during lactation to 110.5 ±11.5 fmol/mg at Day 28. In animals treated with CB 154, binding was always higher than in non-treated animals, with a peak during pregnancy of 149 ± 24 fmol/mg at Day 22. After declining in late pregnancy, the number of receptors was highest at Day 6 of lactation (257.4 ± 34.6 fmol/mg). There is an almost
T
HE OCCURRENCE of a specific receptor for prolactin in mammary tissue wasfirstestablished by Turkington (1) using prolactin covalently bound to Sepharose beads, and was confirmed by Falconer and Birkenshaw (2,3) by means of autoradiographic methods using radioiodinated prolactin. Shiu et al. (4), working with rabbit mammary glands, found this receptor to be located in the particulars membrane fraction obtained at 100,000 x g and thoroughly investigated the kinetic aspects of hormonereceptor interactions on this preparation (5). The receptor was purified by the same authors (6) and recently a specific antibody against this protein was prepared (7), which led them to demonstrate unequivocally that the action of prolactin on mammary explants stringently required the presence
Received September 7, 1976. Supported in part by the Centre National de la Recherche Scientifique (grant no. 655-2152) and the Delegation Generale a la Recherche Scientifique et Technique (grant no. 75-7-1651).
and availability of the receptor on the membranes. The induction of the receptor molecule under various endocrine conditions has been extensively studied in rat liver, since this organ seems to contain this protein in sufficient amounts and its membranes are easier to fractionate than those of the mammary gland. Kelly et al. (8) first detected higher prolactin binding in female than male rat liver, and noted that in females there was a marked increase in binding following puberty and during pregnancy. Posner et al. (9) then described an estrogenmediated increase of binding sites in liver, but later (10) demonstrated that this phenomenon could be amplified or even mediated by prolactin itself. Costlow et al. (11) reached similar conclusions, and recently the same group (12) found that prolactin binding estimated in slices of rat mammary gland during pregnancy and lactation increases to a considerable extent at the time of parturition, whereas it is almost undetectable at earlier periods. This latter
1348
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
PROLACTIN RECEPTORS IN THE MAMMARY GLAND
finding could have been due to endogenous saturation of the receptors, since in rats (as well as primates and ruminants) the placenta is known to secrete large amounts of a lactogenic hormone (13). In the rabbit, however, there does not appear to be a placental lactogen (13,14). In the present work the rabbit has been used to assess and titrate the amount of prolactin receptors during pregnancy and lactation, as well as the state of occupancy of these receptors by endogenous prolactin at various times during these periods. This was achieved by using the pharmacological compound Bromocryptine (CB 154, Sandoz), which is known to depress selectively prolactin secretion from the pituitary gland (15,16), and shows no effect on in vitro prolactin binding by rabbit mammary gland membranes (5).
1349
Ultraturrax homogenizer) was not filtered through cheese cloth in order to avoid any loss of material; it was rehomogenized in a 100 ml Potter homogenizer, centrifuged first at 15,000 x g (20 min), the pellet was discarded and the supernatant centrifuged at 100,000 x g (90 min) as described in the original method (4). The pellet obtained accounted for about 70% of the total binding sites of the tissue, as previously shown by Shiu et al. (4), whereas 30% of these sites remained with the first 15,000 x g pellet. With lactating mammary glands, slight contaminations of the pellets with precipitated casein micelles frequently occurred and were not eliminated to avoid any loss of material. The 100,000 X g crude membrane pellets were suspended in 25 mM Tris buffer, pH 7.6, containing 10 mM MgCl2. Protein contents were assayed according to the method of Lowry et al. (17) and the suspension was frozen (—20 C) until the binding assay was performed. Prolactin iodination
Materials and Methods Animals New-Zealand rabbits (6 months old) were used during their first pregnancy and kept in air and temperature conditioned cots. Mating day was considered as Day 0 and the animals were sacrificed at 4 day intervals, starting on Day 14. Some experiments were performed at Day 6, but very low values for receptor content were obtained and, since the mammary region at this time consists mainly of muscle and connective tissue, these data were not included. The in vivo desaturation procedure consisted of treatment with 2 mg of CB 154 injected sc at 36, 24 and 12 h before the animals were sacrificed. The 36 h pretreatment (3 injections) has been shown to produce maximal desaturation of prolactin, since no additional sites are unmasked when the treatment is extended to 72 h (unpublished observations). Mammary membranes Mammary glands were removed, cleaned and frozen at —80 C and then stored at —20 C. Aliquots of 5 g of this tissue were used for the membrane preparation according to Shiu et al. (4), with the following modifications: the first homogenate (0.3M sucrose, 0 C, prepared with an
Ovine PRL (NIH-P-S7, 24 IU/mg) was iodinated according to the method of Greenwood et al, (18), as modified by Kann (19). The specific activity of the prolactin was approximately 80 fiCi/fig. The specificity of its binding, as assayed with varying hormonal dilutions, appears to be satisfactory (20) and non-specific binding averaged from 15-20% of total radioactivity bound, when to the labelled hormone (1 ng) a 1000-fold excess of unlabelled prolactin (1 ju,g) was added. Binding assay The assay was performed according to Shiu and Friesen (4) at 20 C in the above mentioned Tris/MgCl2 buffer (pH 7.6) containing 0.1% bovine serum albumin. Four hundred fig of membrane protein were incubated in a final volume of 0.5 ml with approximately 105 cpm labelled prolactin and containing increasing amounts of unlabelled prolactin (50 to 1000 fmol for the assays, and 1 fig for assessing nonspecific binding). Up to 70% of the iodinated prolactin molecules could be specifically bound to prolactin receptors in receptor rich membranes (e.g., lactating animals 5-6 days, desaturated with CB 154) at protein amounts up to 600 fig. The specific binding plateaued at higher protein concentra-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
1350
DJIANE, DURAND AND KELLY
tions, but the variability of the assays at these high concentrations of protein prevented any clear-cut conclusions as to whether this phenomenon was due merely to the kinetic parameters of the saturation reaction as recently discussed by Chang et al. (21), to a "protein effect" at high concentrations, or to the inability of a fraction of the iodinated molecules to interact with the receptors. In all cases, when used under the conditions of the "specific binding" assay (i.e., difference between radioactivity bound in the presence of an excess of unlabelled hormone and radioactivity bound in its absence), or at the onset of saturation (see below, Fig. 2), the iodinated prolactin yielded entirely consistent results, which appear to ensure at least the relative accuracy of the amounts of available sites titrated. Chemicals CB 154 (2 bromo-a-ergocryptine-methanesulfonate) was kindly supplied by Sandoz, Basel, and bovine serum albumin (fraction V) from I.C.N. Pharmaceutical, Cleveland, Ohio. All reagents were of analytical grade.
Results Figure 1 shows the time course for the incubation of mammary gland membranes with ovine [125I]iodo-PRL. An incubation period of 6 h was necessary for equilibrium,
Endo • 1977 Vol 100 • No 5
which agrees with the kinetic data of Shiu and Friesen (5). However, a small, but reproducible intermediary plateau has been observed, at approximately half-saturation (2-3 h of incubation), suggesting that two distinct sets of binding sites might be present. This phenomenon seems also to be observed in Shiu and Friesen's data (ref. 5, Fig. 2, 23 C). Further experiments will be necessary to decide if this could be due to the fact that half of the receptor sites are buried in inverted vesicles, as has been demonstrated by Bennett and Cuatrecases for insulin receptors (22). In any case, it seems clear that at least 80% of the high affinity sites have been titrated under our experimental conditions, provided that no loss of receptor molecules has taken place when the cell lipids were removed after the first centrifugation: this seems difficult to assert, since no equilibration of this lipid layer with [125I]iodo-prolactin has been attempted, owing to the fact that it could not be mixed with or suspended in an aqueous phase. The problem of possible contributions of various cell populations to prolactin binding is further considered below (see Discussion). Saturation curves of prolactin specifically bound to mammary gland membranes from rabbits at various stages of pregnancy or
FIG. 1. Binding kinetics of ovine [125I]iodo-PRL to membranes from animals at 28 days of lactation O O and from animals at 14 days of lactation treated with CB 154 to desaturate the binding sites • • (see Materials and Methods). Each point represents the mean (± SEM) for 3 animals. Incubation was performed with 50,000 cpm of ovine [125I]iodo-PRL during the periods indicated, at 20 C. Results are expressed as fmoles specifically bound/400 fig protein.
2.
as i
23
hours.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
PROLACTIN RECEPTORS IN THE MAMMARY GLAND
1351 Uxto D 28 (4)
LOCtO D 14(4)
D.26{4)
200
400
600
800
1000 Total Prolactin . ( f moles)
FiG. 2. Prolactin binding as a function of increasing amounts of prolactin at the indicated stages of pregnancy and lactation (Lacta.). The competition data with increasing amounts of unlabelled prolactin are expressed as saturation curves. Each point represents the mean (±SEM) for the number of animals indicated in parentheses. These data are obtained with membrane preparations from animals which were not desaturated in vivo by CB 154 treatment.
lactation (not treated with CB 154) are shown in Fig. 2. Representative Scatchard plots calculated from saturation kinetics (23) are depicted in Fig. 3. Panel A shows data from untreated animals, panel B data from CB 154-treated animals (original saturation curves not shown). The binding affinity in all cases remained constant (Ka = 2.5 to 3.2 x 109M-1), whereas the number of binding sites varied depending on the stage of pregnancy or lactation. The development of prolactin binding sites expressed per mg of membrane protein in rabbit mammary glands during pregnancy and lactation in untreated and CB 154-treated animals is shown in Fig. 4. In untreated animals, the number of prolactin binding sites was 25 ± 3 fmol/mg protein at day 14 of pregnancy then the binding increased to reach 54.8 ± 5.8 fmol/mg at day 22
of pregnancy. This value is significantly higher than the values observed at day 14 (P < 0.01) and at day 26 (P < 0.01), but not significantly different from the values observed at day 18 and day 29. Binding increased during lactation to a maximum of 110.5 ± 11.5 fmol/mg at day 28 (P < 0.01 when comparing day 28 of lactation with day 22 of pregnancy). In animals treated with CB 154, binding was always higher than in non-treated animals, but the variability of the values obtained was greater, hence the peak observed at day 22 (140 ± 24 fmol/mg) did not differ significantly from the other values of gestation. In early lactation, binding increased strikingly. The highest number of receptors was observed at day 6 (257.4 ± 34.6 fmol/ mg) and significantly declined at day 14 to 172 ± 10 fmol/mg (P < 0.01). These results
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
DJIANE, DURAND AND KELLY
1352 A = untreated animals a — D [>28 lactation • — a D 14 H o—o
D 22 pregnancy
A—^
D 18
»
.--..
D 14
M
Endo • 1977 Vol 100 • No 5
B= CB154 treated animals u. CD
0.8.
. \
A—A D 6 lactation • — • DI4 '1 V \
A — A D 18 pregnancy • _ . DI4 '• \
0.6.
FIG. 3. Representative scatchard analyses of ovine [l25I]iodoPRL binding in membranes from untreated rabbits (A) and CB 154-treated rabbits (B).
0.4. 0.4. 0.3.
0.2. 0.2. O.I.
25
50 75 B f moles oPRL bound
«
5O
22 26 OftEGNANCV
FlG. 4. Evolution of the number of prolactin receptors (expressed as fmoles prolactin bound per mg of membrane protein) throughout pregnancy and lactation. The values were obtained by analysis of the Scatchard plots (Fig. 3) of the saturation data (Fig. 2). (n) = number of animals, • • = animals desaturated in vivo by CB 154 (36 h), O 0 = controls, a n d l = ±SEM.
100
I5OB
f moles oPRL bound
clearly demonstrate that a) in vivo desaturation with CB 154 for 36 h results in unmasking a considerable amount of endogenously saturated binding sites, which depending upon the period considered may vary from 50% to more than 70% of the number of total sites (titrated under our desaturation conditions); b) this total amount fluctuates to a considerable degree, with a steady increase during early pregnancy, a decrease in late, a sharp rise at the onset of lactation and a second decrease following the establishment of lactation. Since the amount of crude membrane protein in the mammary gland itself varies to a considerable degree, with a rise at late pregnancy, due to the development of an endoplasmic reticulum characteristic of increased protein synthesis (24), it was considered relevant to express the number of receptors per total mammary gland or per mg DNA, the latter accounting for the number of cells. Figure 5 shows the weight increase of the mammary glands of untreated and CB 154treated animals during the experiment and demonstrates that 36 h depression of prolactin secretion does not modify this increase, except at the early lactational stage,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
PROLACTIN RECEPTORS IN THE MAMMARY GLAND
1353
FIG. 5. Weights of rabbit total mammary glands during the course of pregnancy and lactation. For other details see legend to Fig. 4.
18
when milk secretion is known to be highly sensitive to prolactin levels (25) which may account for the decrease observed. It has been verified repeatedly (results not shown) that the 5 g samples taken from the mammary glands to perform the various assays are representative of entire mammary tissue. Figure 6 portrays the variations of receptor levels in whole mammary gland (panel A) in untreated and CB 154-treated animals and reveals that the sudden increase in prolactin binding sites at parturition is greatly amplified when compared to Fig. 4, since mammary growth is still linear at this period. Panel B shows the amount of receptors per mg P DNA (i.e., per cell): this figure does not point out a real decrease of receptors during late pregnancy (except for CB 154-treated animals at 29 days of pregnancy), as might have been postulated, if only protein concentrations were considered (Fig. 4), and shows that the striking increase of binding sites at parturition is even much higher than may have been expected from the receptor protein ratios. In addition this figure suggests that at parturition almost
22 26 PREGNANCY
28 Doys
all sites are desaturated, a paradox which is masked in Fig. 4 due to the high milk protein contents of the controls (untreated) at the onset of lactation. This apparent desaturation may be artefactual, as discussed below. Discussion Mammary development and the onset of lactation are usually considered to be due to sequential phases of hormonal stimulation (26) clearly separated by parturition, which causes both prolactin secretion at high levels (27,28) and the complete withdrawal of progesterone. It has recently been demonstrated (29) that in the pseudopregnant rabbit prolactin stimulation acts most probably at the transcriptional level and that large quantities of mRNAs for milk proteins are detectable only in the presence of prolactin and in the absence of progesterone. The present results demonstrate that this biphasic aspect of mammary growth and secretion also applies to prolactin receptors: comparatively low levels of receptors per total mammary tissue (Fig. 6 A) or per cell (Fig. 6 B) are detected during pregnancy
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
Endo i L977 Vol 100 ( No 5
DJIANE, DURAND AND KELLY
1354
o 500.
E 400.
300.
200.
100.
5.
10.
14 18 22 26 29| Pregnancy
p
6
14 Lactation
28 days
14
18 22 26 291
6
Pregnoncy
Lactation
p
14
28 days
FIG. 6. Change in number of prolactin receptors per total mammary gland (panel A) or per mg of DNA phosphorus (P DNA) (panel B) in the course of pregnancy and lactation. The data for DNA contents of the mammary gland at the indicated periods, are those reported by Denamur (30). For other details see legend to Fig. 4.
and these levels suddenly increase by al- truly reflects a regulatory step, with no most 500% at parturition. In vivo desatura- major contribution by a desaturation phetion by CB 154 constantly increases the nomenon, as could be the case for species available binding sites by 100-200%, but secreting high levels of a placental lactothis inevitably raises the question as to genic hormone. whether the sudden burst of free receptors When comparing receptor levels of unat parturition could not be due to a desatura- treated or CB 154 desaturated animals durtion of previously occupied sites by the ing pregnancy and lactation (Fig. 6 A and B), withdrawal of some endogenous placental the question arises as to whether 36 h mammotropin as postulated by Holcomb et suppression of prolactin secretion by the al. (12) for the rat. The rabbit is considered drug could not be itself a parameter modulatnot to possess such a hormone (13,14) ing these levels: prolactin has been shown to and this is the main reason why this increase its own receptors in the liver species has been chosen to perform the pres- (10,11), hence CB 154 should decrease these ent experiments. Moreover, a series of levels. Some evidence that this might be the experiments not included in the present case at 29 days of pregnancy is provided by manuscript, demonstrate that when CB 154 the apparent desaturation of all receptors treatment is applied for two days prior to at this time, a paradox which has already parturition (which suppresses the prolactin been mentioned and may be explained, if surge at birth), such an increase of receptor one assumes that in the CB 154-treated level in the mammary gland is not observed. animals the number of receptors has Hence it may be assumed that this increase dropped approximately to control values.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
PROLACTIN RECEPTORS IN THE MAMMARY GLAND The experiments mentioned above, where suppression of the prolactin by CB 154 at the time of parturition prevents the increase of receptors, lend some support to this hypothesis. For other periods of pregnancy, where no sudden induction of receptors is observed, it seems unlikely that a short treatment with CB 154 could extensively modify receptor levels. During lactation, it appears that the number of receptors in CB 154treated animals is lower at Day 14 of lactation than at Day 6, whereas free receptors (untreated animals) remain fairly constant; this might be correlated with a decrease of prolactin levels during the second half of lactation, but more experiments are needed to establish prolactin levels and receptor variations at these periods in the rabbit. A remarkable aspect of the experiments reported consists in the demonstration of a roughly reciprocal correlation between phases of mammary growth (as estimated by the increase of total DNA) and receptor amplification (receptors per unit DNA or per cell). Thus, during the first period of pregnancy (up to 20 days), the levels of DNA do not increase to a great extent (30), whereas the number of receptors per cell is increased by a factor of 3 to 4. From this time until the day of parturition about 80% of the total mammary growth takes place and the level of receptors per cell remains constant or even slightly decreases. The fact that the main increase of receptors takes place at parturition, while 20% of the growth phenomenon is still to take place (up to the fifth day of lactation (30)) does not seem to contradict this reciprocal relationship, since the mammary cells are probably not completely synchronized. Nevertheless, this phenomenon is reminiscent of the behavior of synchronized cell cultures, where mitosis limits protein synthesis (31,32). The sudden burst of specific binding sites which occurs when mammary tissue switches from the developmental to the secretory phase raises the puzzling problem as to whether this qualitative switch could
1355
be mediated merely by the amount of information communicated to the cell through a parallel increase of receptor and hormonal levels. No definite answer can be presently suggested to this question, but it must be kept in mind that additional hormonal shifts are taking place at parturition, mainly concerning progesterone and corticosteroids, which may modulate the overall phenomenon, either independently of prolactin effects or even by participating in the regulation of prolactin receptor levels, hence of the cells sensitivity to the latter hormone. An additional aspect concerns the onset of differentiation, which is presently believed to be triggered by estrogens, in addition to prolactin (9). This may be difficult to analyze with intact mammary tissue, considering the small numbers of epithelial cells present when pregnancy starts: at 14 days pregnancy or even at 6 days (results not shown), receptors are detected, but the initial cell differentiation must have already taken place. It must be finally emphasized that no completely satisfactory parameter to express the concentration of receptors in such an heterogeneous cell population as the mammary gland is presently available. It has already been mentioned that protein concentrations are inadequate, since they include changes in a parameter such as casein synthesis, which should not be accounted for, when expressing membrane receptor concentrations. The amount of DNA reflects the total number of cells, irrespective of their nature; hence adipocytes, fibroblasts, myoepithelial cells, and others contribute to DNA, possibly without contributing prolactin receptors, although the latter point remains to be established. No specific marker enzyme or protein for the epithelial mammary cell is presently known; moreover, any such marker would be likely to fluctuate itself, according to the phases of mammary growth and secretion. Despite these shortcomings, prolactin receptor concentrations, as described in the present work, should adequately reflect the physio-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
DJIANE, DURAND AND KELLY
1356
logical regulations, since their main variations remain identical, irrespective of the parameters with which they are correlated. Acknowledgments We wish to thank Professor H. Clauser for much helpful advice during the course of this work and during redaction of the manuscript. We are greatly indebted to Dr. G. Kann for the supply of the iodinated prolactin. References 1. Turkington, R. W., Biochem Biophys Res Commun 41: 1362, 1970. 2. Falconer, I. R., Biochem J 126: 8p., 1972. 3. Birkenshaw, N., and I. R. Falconer, J Endocrinol 55: 323, 1972. 4. Shiu, R. P. C , P. A. Kelly, and H. G. Friesen, Science 180: 968, 1973. 5. Shiu, R. P. C., and H. G. Friesen, Biochem J 140: 301, 1974. 6. Shiu, R. P. C., and H. G. Friesen, J Biol Chem 249: 7902, 1974. 7. Shiu, R. P. C , and H. G. Friesen, Science 192: 259, 1976. 8. Kelly, P. A., B. I. Posner, T. Tsushima, and H. G. Friesen, Endocrinology 96: 532, 1974. 9. Posner, B. I., P. A. Kelly, and H. G. Friesen, Proc Natl Acad Sci USA 71: 2407, 1974. 10. Posner, B. I., P. A. Kelly, and H. G. Friesen, Science 188: 57, 1975. 11. Costlow, M. E., R. A. Buschow, and W. L. McGuire, Life Sci 17: 1457, 1976. 12. Holcomb, H. H., M. E. Costlow, R. A. Buschow, and W. L. McGuire, Biochim Biophys Ada 428: 104, 1976.
Endo • 1977 Vol 100 • No 5
13. Kelly, P. A., T. Tsushima, R. P. C. Shiu, and H. G. Friesen, Endocrinology 99: 765, 1976. 14. Talamantes, Jr, F., Am Zool 15: 279, 1975. 15. del Pozo, E., and E. Fliickiger, In Pasteels, J. L., and C. Robyn (eds.), Proceedings of the International Symposium on Human Prolactin, Brussels, Excerpta Medica, Amsterdam, 1973, p. 291. 16. Floss, H. G., J. M. Cassady, and J. E. Roberts, J Pharm Sci 62: 699, 1973. 17. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. RandallJ Biol Chem 193: 265, 1951. 18. Greenwood, F. C., W. M. Hunter, and J. S. Glover, Biochem J 89: 114, 1963. 19. Kann, G.,CRAcadSci [D](Paris) 272: 2808,1971. 20. Djiane, J., and G. Kann, CR Acad Sci [D] (Paris) 280: 2785, 1975. 21. Chang, K. J., S. Jacobs, and P. Cuatrecasas, Biochim Biophys Ada 406: 294, 1975. 22. Bennett, V., and P. Cuatrecasas, Biochim Biophys Ada 311: 362, 1973. 23. Scatchard, G., Ann NY Acad Sci 51: 660, 1949. 24. Bousquet, M., J. E. Flechon, and R. Denamur, Z Zellforsch 96: 418, 1969. 25. Taylor, J. C , and M. PeakerJ Endocrinol 67: 313, 1975. 26. Denamur, R . J Dairy Res 38: 237, 1971. 27. Kann, G., and R. DenamurJ Reprod Fertil 39: 473, 1974. 28. Morishige, W. K., G. J. Pepe, and I. Rothchild, Endocrinology 92: 1527, 1973. 29. Houdebine, L. M., and P. Gaye, Mol Cell Endocrinol 3: 37, 1975. 30. Denamur, R., CR Acad Sci [D] (Paris) 256: 4748, 1963. 31. Martin, D., Jr., G. M. Tomkins, and D. Granner, Proc Natl Acad Sci USA 63: 842, 1970. 32. Johnson, T. C , and J. J. Holland, J Cell Biol 27: 565, 1965.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 August 2016. at 14:32 For personal use only. No other uses without permission. . All rights reserved.
