0013-7227/91/1286-2976$03.00/0 Endocrinology Copyright © 1991 by The Endocrine Society
Vol. 128, No. 6 Printed in U.S.A.
Prolactin Receptor Gene Expression in Rat Mammary Gland and Liver during Pregnancy and Lactation G. A. JAHN*, M. EDERY, L. BELAIR, P. A. KELLY, AND J. DJIANE Unite d'Endocrinologie Moleculaire (G.J., M.E., L.B., J.D.), Bdtiment des Biotechnologies, Institut National de la Recherche Aaronomique, 78350-Jouy-en-Josas, France; and Laboratory of Molecular Endocrinology (P.K.), McGill University, Royal Victoria Hospital, Montreal, Quebec, Canada H3A 1A1
ABSTRACT. The expression of two forms of PRL receptor messenger RNA was measured at different stages of pregnancy and lactation in mammary gland and liver from Sprague-Dawley rats, using 32P-labeled complementary DNA probes encoding the extracellular part of the receptor (E probe), common to the two forms and a probe encoding the intracellular part of the long form of the receptor (I probe), that only recognizes sequences specific to the long form of the receptor. Hybridizations were performed in Northern blots obtained from electrophoreses of poly (A+) enriched RNA preparations from mammary glands and livers of rats on days 0, 6, 12, 19, and 21 of pregnancy and 5, 10, 15, and 20 of lactation. The Northern blots were also hybridized with a chicken /3-actin probe, to correct for the amount of mRNA added and the different metabolic states of the tissues. Both tissues expressed the same forms of PRL receptor mRNAs, namely bands at 2.5, 3, and 5.5 kilobases encoding the long form of the receptor and a major band at 1.8 kilobases encoding the short form. The liver expressed all the receptor mRNA forms in much higher quantity than the mammary gland, independent of the reproductive state. In liver there was an increase of all the transcripts on day 19 of pregnancy, followed by an abrupt decline at the onset of lactation, to levels lower than those of virgin rats. In contrast, mammary gland PRL receptor mRNAs were low in virgin and pregnant animals,
P
ROLACTIN is involved in an impressive number of biological actions in different target organs, of which the most studied has been the initiation and maintenance of milk production by the mammary gland (1). The actions of PRL in its different target organs are initiated through interaction with specific, high affinity receptors at the cell surface. Different biocheminal studies have shown that the majority of PRL receptors have a molecular mass of approximately 40,000 although a form of mol wt 80,000 has also been characterized, particularly in the ovary (2-5). PRL receptor levels are differentially regulated according to the tissue examined. In rat liver, one of the tissues Received June 7, 1990. Address requests for reprints to: Dr. Jean Djiane, INRA, Batiment des Biotechnologies, 78350 Jouy-en-Josas, France. * Present address: Laboratorio de Reproduction y Lactancia, CRICYT-CONICET, C.C. 855, 5500 Mendoza, Argentina.
increased significantly at day 21 of pregnancy, and continued to increase throughout lactation. Treatment of day 19 pregnant rats with the antiprogesterone RU 486 induced, 24 h later, PRL receptor mRNAs in mammary gland but not in liver. There were no significant differences in the relative proportions of long to short forms of PRL receptor mRNAs at the different reproductive states, but the proportion of the long form was slightly greater in mammary gland than in liver. Membrane PRL receptor concentrations were also measured in the same tissues used for the mRNA study by binding to a 125I-labeled monoclonal antibody (U5), which specifically recognizes the PRL receptor at a site different from the hormone binding site. The quantity of receptor measured by U5 binding was approximately 3 times higher than that measured with 125I-labeled ovine PRL. The pattern of U5 binding to membranes observed during pregnancy and lactation in liver and mammary gland was parallel to that observed for PRL receptor mRNAs, with the exception that the increases observed in mammary gland on day 21 of pregnancy or after RU 486 treatment were not seen. The present results indicate that PRL receptors are differentially regulated in liver and mammary gland during pregnancy and lactation in the rat and that the decline of progesterone at the end of pregnancy induces PRL receptor mRNA in mammary gland but not in liver. (Endocrinology 128: 2976-2984, 1991)
with the highest PRL binding, there are sex differences in PRL receptor numbers (6), and estrogens (7, 8), PRL, and GH stimulate (9-12) binding. Receptor levels also vary during the estrous cycle (13, 14), pregnancy (6, 15, 16), and lactation (16,17). Although the mammary gland is one of the most studied target tissues for PRL, the concentration of receptors seems to be very low in rats (16, 18), being almost undetectable in tissue from virgin animals. In mammary tissue, PRL receptors increase during lactation (16, 20) and are induced by glucocorticoids (21) and PRL itself (22), while progesterone may exert a negative influence (22, 23). Recently, rat, rabbit, and human PRL receptor complementary DNAs have been cloned and sequenced (2426), and several forms of different size have been detected by hybridization in rat tissues (24, 27). Moreover, two different receptor cDNAs, differing in the length and sequence of the intracellular region, have been described
2976
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES (27) and encode receptors of different sizes. A cDNA encoding the short form of the receptor messenger RNA was obtained from a estrogen-treated liver cDNA library, and the long form was originally detected from a rat ovary library (24, 27). After binding of PRL to its receptor, the mechanism by which the PRL signal is transferred inside the cell remains unknown. The different forms of the PRL receptor could mediate different physiological actions of the hormone in the same or different tissues. In the present study, using a probe derived from the extracellular region of the receptor (E probe, common to both forms) and another obtained from the intracellular sequence of the long form (I probe), we have examined the expression of PRL receptor mRNAs in mammary gland and liver from rats at different stages of pregnancy and lactation. In order to compare concentrations of mRNA to concentrations of PRL receptors in membranes, we have also measured the binding of [ 125 I]PRL and [ 125 I]monoclonal antibody (U5, Ref. 28). This monoclonal antibody binds with high specificity to PRL receptor, at a site different from the PRL binding site, thus obviating the distortion of results produced by occupation of the receptors with endogenous hormone.
Materials and Methods Animals and experimental procedures Adult virgin female and male Sprague-Dawley rats were obtained from Iffa-Credo (Orleans, France) and kept in a light (lights on 0600-2000 h)- and temperature-controlled room; rat chow and tap water were available ad libitum. The rats were allowed a period of acclimation, after which daily vaginal smears were taken and on the night of proestrus females were caged with a fertile male. The presence of spermatozoa was ascertained in the vaginal smears the following morning, and this day was considered as day 0 of pregnancy. Groups of three rats each were killed by decapitation between 1030 and 1100 h on days 0, 6, 12, 19, and 21 of pregnancy and 5, 10, 15, and 20 of lactation. The day 0 rats were unmated animals on the day of estrus. The lactating rats nursed litters of eight pups. After decapitation, the whole inguinal and abdominal mammary glands and livers were quickly excised and frozen immediately in liquid nitrogen. The samples were kept at -80 C until portions were processed for RNA extraction and membrane preparation. In a group of pregnant rats, RU 486 [RU 38486; 17/3-hydroxy11 j8-(4-dimethyl-aminophenyl)-17-propinyl-estra-4,9-dien-3one, generously provided by Roussel-Uclaf, Romainville, France], was injected sc at a dose of 2 mg/kg in sesame oil (2 g/liter) at 1030 h on day 18 of pregnancy. These rats were killed 24 h later and mammary glands and livers obtained. RNA extraction and hybridization Total RNA was extracted using the guanidinium isothiocyanate-LiCl precipitation. Briefly, tissues were homogenized
2977
in 10 vol sterile 4 M guanidine isothiocyanate, 50 mM Tris HC1, pH 7.5, EDTA 10 mM, mercaptoethanol 8%, filtered through sterile gauze and centrifuged 10 min at 2000 X g. The supernatants were diluted with 5.5 vol sterile 4 M LiCl and allowed to precipitate for 48-72 h at 4 C. The precipitate was obtained by centrifugation at 15,000 X g during 90 min, washed twice with 3 M LiCl, and extracted twice with 5-15 ml TES buffer (Tris HC110 mM, pH 7.5 EDTA 1 mM, sodium dodecyl sulfate 0.2%). The pooled extraction supernatants were extracted with 1 vol of a mixture consisting of 0.5 vol water-saturated phenol -0.5 vol chloroform-isoamyl alcohol (24:1). The total RNAs in the aqueous phase were precipitated with 0.1 vol sodium acetate, 3 M pH 5.1 and 2.5 vol absolute ethanol. Aliquots of total RNAs were enriched in poly (A+) RNAs by two passages through an oligo dT cellulose column (29). The samples of poly (A+) RNAs were size separated by electrophoresis in agarose (1.3%, Seakem, FMC Bioproducts, Rockland, ME), formaldehyde vertical gels and transferred to nitrocellulose paper (Hybond C extra, Amersham, Arlington Heights, IL) (30). For size calibration of the electrophoreses, a 0.24-9.5 kilobase (kb) RNA ladder (BRL, Bethesda Research Laboratories Life Technologies, Inc., Rockville, MD) or a Pharmacia DRIgest III 0.07-23 kb/DNA ladder (Pharmacia, Bromma, Sweden) were run on each gel. The probes used for PRL receptor mRNA hybridization were gel-purified inserts from plasmids containing the cDNAs encoding the extracellular domain of the rat PRL receptor cDNA (probe E, nucleotides 328-619, Ref. 24), for the cytoplasmic domain of the long form of the receptor (probe I, nucleotides 1253 -1735, Ref. 27) and for chicken /3-actin. The filters were prehybridized for at least 4 h at 42 C in a solution containing 50% formamide, Denhardt's solution x 5, SSPE (1.15 M NaCl/0.01 M NaH2 PO4/1 mM EDTA) X 5, salmon DNA 100 n\, sodium dodecyl sulfate 0.1%, polyadenylic acid (5') type I, K+ salt (Sigma Chemical Co., St Louis, MO) 1 pg/ml. Hybridization solution contained, in addition to the above, 5% dextran and 500,000 cpm of the appropriate probe, previously labeled with 32P by random priming (31). Filters were hybridized overnight at 42 C and washed twice in 0.1 SSC (0.15 M NaCl/0.015 M sodium citrate, pH 7.0) with 0.1% sodium dodecyl sulfate and 1 mM EDTA at 65 C. The filters were hybridized first with the probe for the cytoplasmic part of the long form of the PRL receptor, stripped and rehybridized with the probe for the extracellular probe, and finally stripped once more and hybridized with the chicken /3-actin cDNA, as control for the differences in the amount of RNA loaded on the gel and also as control for the dilution of the specific mRNAs found in the lactating rat mammary gland samples in relation to the mammary glands of virgin or pregnant rats. Membrane preparation and measurement of membrane receptor Membranes were prepared from mammary gland and liver by homogenization of tissues in 5 vol sucrose 0.3 M and centrifugation for 15 min at 1500 X g. Supernatants were spun 90 min at 100,000 X g to obtain crude membrane pellets which were resuspended in 2 vol Tris HC1 25 mM, pH 7, MgCl2 10 mM (32). These preparations were frozen at -20 C until measurement of membrane binding to [125I]ovine (o) PRL or the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES
2978 I25
Endo«1991 Voll28«No6
I-L)5 binding to membranes
4 0 r Liver l25
30
I6r
I - U s binding to membranes
Liver
2 20
o 10
E o
w
0 4_ Mammary
6 Gland
12
19 21
m 0
6
12
19 21
20
Mammary Gland
E
§0.5
0 Day of
12 19 21 Pregnancy
10
15 20 Lactation
Oay of
12 19 21' Pregnancy
15 20 Lactation
FIG. 1. Concentration of membrane PRL receptor during pregnancy and lactation in rats. Relative receptor concentrations were measured by binding of [125I]U5 to liver or mammary gland membranes from rats on the indicated days of pregnancy and lactation as described in Materials and Methods. Results are expressed as means ± SEM of (A) counts per min of U5 specifically bound per mg protein and (B) counts per min of U5 specifically bound per g tissue of three replicate determinations. TABLE 1. Comparison of number of PRL receptor sites measured by Scatchard analysis performed with [125I]U5 or [125I]oPRL in liver and mammary gland membranes obtained from rats at different stages of pregnancy (P) and lactation (L) Binding (fmol/mg protein)
Liver P day 19 L day 5 L day 10 Mammary gland P day 19° L day 5 L day 10 L day 20 0
U5
oPRL
U5/oPRL
1488 90 189
459 40 49
3.2 2.3 3.7
counter (32). Scatchard analysis to compare the number of sites measured with U5 and oPRL binding was performed for some of the samples, by incubating 50,000 cpm of tracer and increasing concentrations (0.25 ng to 1 ng) of the corresponding unlabeled tracer. Tracers were labeled with 125I using the Chloramine T method (34), and labeled protein was separated from free iode using an acrylamide agarose column (ACA 54, LKB, Bromma, Sweden). Specific activities were 76-85 and 4-14 iiC\/iig for oPRL and U5, respectively.
Results 39 41 93 62
10 16 28 21
3.9 2.6 3.3 3.4
Treated with RU 486.
monoclonal antibody U5 (28). Aliquots were taken for protein measurement by the method of Lowry et al. (33). With the exception of mammary glands on day 0 and 6 of gestation, the tissues used for the U5 binding study were portions of the same tissues used for mRNA preparation. Binding assays were performed by incubating 25 jug (for liver samples) or 200 ng (for mammary gland samples) of membrane proteins suspended in 0.5 ml Tris HC1 25 mM, MgCl2 10 mM, BSA 0.1%, sodium azide 0.02% with 50,000 cpm of labeled tracer in the presence or absence of 1 i*g unlabeled tracer for measurement of specific or nonspecific binding. After incubation overnight at room temperature, tubes were diluted with 3 ml buffer and centrifuged, and the pellets were counted in a 7-
Evolution of membrane PRL receptor concentrations in mammary gland and liver during pregnancy and lactation PRL receptor concentrations in both tissues were measured by binding to a 125I-labeled monoclonal antibody, U5, which binds with great specificity to PRL receptor at a site different from the PRL binding site (28). Figure 1 shows the binding of [125I]U5 (counts per min/mg protein and per gram tissue) in liver and mammary gland. Scatchard analysis performed in some samples of pregnant and lactating liver and lactating mammary gland, and comparison with the specific binding values, showed that the number of receptors was always proportional to the specific binding of U5, using 50,000 cpm of [125I]U5. In order to correlate the number of sites detected by U5 with those detected with oPRL, we performed Scatchard analysis of some samples of pregnant
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES
FIG. 2. The different types of PRL receptor mRNAs in liver and mammary gland of rats. Northern blots performed with 25 ng poly (A+) RNA from pregnant rat liver or mammary glands were hybridized with a probe specific to the intracellular portion of the long form of the receptor (I probe), or the probe for the extracellular portion of the receptor (E probe), as indicated in Materials and Methods. Autoradiograms were exposed for the times indicated at the bottom of the lanes.
and lactating mammary gland and liver using both ligands. We were unable to use mammary glands from virgin animals for this study, since the binding of oPRL was too low to allow for accurate Scatchard analysis. The results (Table 1) showed that U5 consistently measured approximately 3 times the number of sites determined by PRL. Figure 1 shows that in liver membranes the number of PRL receptors is much greater than in mammary gland and that the number of sites varies inversely in both tissues. In liver, there is an increase during pregnancy and a decrease to the lowest values in lactation, while in mammary gland PRL receptor levels remain very low in virgin and pregnant rats and increase markedly after parturition. The increase seen in mammary gland receptor is more apparent when the results are expressed in terms of tissue wts due to the marked increase in protein content of the tissue as pregnancy and lactation progress (Fig. IB). PRL receptor mRNAs in mammary gland and liver of female rats Northern blot analysis of mRNAs obtained from liver with the I probe showed three bands, two at 2.5 and 3 kb of approximately the same intensity and a third, much weaker, at 5.5 kb (Fig. 2). When the same filters were hybrized with the E probe, a new signal, much stronger than all the others, was seen at 1.8 kb, with fainter bands at 2.5, 3, and 5.5 kb, which in fact were the same bands that hybridized with the I probe. When mammary gland mRNAs were analyzed with both probes, signals at 2.5, 3, and 5.5 were also obtained with the I probe. However, using the E probe, a strong band at 1.8 was seen, with faint signals observed at 2.5, 3, and 5.5 kb, which did not appear in all samples. Densitometric analysis of the autoradiographic films revealed that the proportion of the long form of the receptor in mammary gland was slightly higher than that obtained in liver. Thus, essentially the same forms of
2979
Mi. Ui .
PRL receptor mRNAs were found in mammary gland and liver, although the intensity of the bands obtained with both probes in mammary gland was much fainter than that observed in the liver (Fig. 2) Evolution of PRL receptor mRNAs during pregnancv and lactation Figure 3 shows the Northern blot analysis with both receptor probes and with an actin probe, of filters containing liver and mammary gland mRNAs obtained from the different stages of pregnancy and lactation. In liver tissue an increase was observed toward the end of pregnancy followed by an abrupt decrease, to very low values during lactation. In contrast, in mammary gland, mRNAs remained relatively low throughout pregnancy, started to increase at day 21 of pregnancy, and rose to the highest values during lactation. The actin signal in liver did not show any marked variation throughout pregnancy and lactation, whereas in mammary gland, there was a gradual decrease in the signal from the start of pregnancy, with only very faint signals being obtained in lactating tissues. This gradual diminution of the actin signal may be accounted for by dilution of the actinspecific mRNAs by the increased milk protein mRNAs in late pregnant or lactating tissue (37, 38), which also would dilute the PRL receptor-specific mRNAs. In order to quantify the amount of PRL receptor mRNAs present at each stage of pregnancy and lactation in liver and mammary gland, triplicate filters hybridized with both probes were analyzed by densitometry and corrected to uniform actin concentration. The values shown for the long form correspond to the sum of scan intensities of all the bands hybridizing with the I probe, and the values for the short form are the densitometric scan intensities of the 1.8 kb band hybridized with the E probe. Results, shown in Fig. 4, indicate that in both tissues the variations of long and short forms of the receptor were similar, and the relative proportions of both forms did not change significantly with the repro-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
2980
Day
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES
0
6
6ns
Endo«1991 Voll28«No6
fC 15 2( Lactatkn
FIG. 3. Evolution of PRL receptor mRNAs during pregnancy and lactation in liver (A) and mammary gland- (B). Northern blots of poly (A+) RNAs from liver and mammary glands from rats on different days of pregnancy and lactation were hybridized sequentially with I probe, E probe, and chicken /3-actin probe, as described in Materials and Methods. Autoradiographic films of Northern blots for liver were exposed 4 h for the E probe and 24 h for the other two probes, and those from mammary gland were exposed 6 days for the I and E probes and 24 h for the actin probe. Lanes from left to right represent days 0, 6, 12, 19, and 21 of pregnancy and 5, 10, 15, and 20 of lactation.
ductive stage. On the other hand, in liver, PRL receptor mRNAs increased at the end of pregnancy and fell to very low values during lactation, while in mammary gland mRNA levels were low in tissue from virgin or pregnant rats, started to increase on day 21 of pregnancy, and reached the highest values during lactation, with levels approximately 10 times higher than in virgin or pregnant rats. The proportion of long to short form was smaller in liver than in mammary gland, accounting for approximately 14% of the total while in the mammary gland long form accounted for approximately 33% of the total. The increase in late pregnancy in the liver was more marked for the short than for the long form.
Effect of RU 486 on PRL receptor mRNA expression in late pregnant rats The fall of progesterone that occurs at the end of pregnancy is known to induce lactogenesis and PRL secretion (35, 36). In order to determine whether the fall of progesterone is also able to increase expression of PRL receptors, rats were injected with the progesterone antagonist RU 486 at 1000 h on day 18 of pregnancy and killed 24 h later, at a time when the steroid has already induced lactogenesis. Figure 5 shows that RU 486 treatment induced a significant increase in PRL receptor mRNAs in mammary gland but not in liver. In contrast,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES PRL receptor mRNAS in liver O Long form 200
• Short form (1.8 kb)
100
OJ 0 Day of
12
19 21 pregnancy
15 20 lactation
PRL receptor mRNAS in mammary gland O Long form • Short form (1.8 kb)
0.5.
OJ
B
0 Day of
12
19 21 pregnancy
15 20 lactation
FIG. 4. Quantification of PRL receptor mRNAs from liver (A) and mammary gland (B) of rats during pregnancy and lactation. Triplicate Northern blots, hybridized sequentially with the I, E, and actin probes were analyzed by densitometry. The relative expression of the long form of the receptor (that hybridized with the I probe) was expressed as the ratio of intensities of the specific bands to the actin probe, all corrected for 24 h of exposure (means ± SEM of three replicate Northern blots).
membrane receptor levels measured with U5 did not increase after RU 486 in either tissue (Fig. 6). Discussion The present paper describes the measurement of PRL receptor and its mRNAs during pregnancy and lactation in the rat mammary gland and liver. Membrane PRL receptors were measured using an 125 -I labeled monoclonal antibody (28) instead of the classical PRL binding assay. This procedure has several advantages, since the monoclonal antibody gave greater sensitivity, allowing accurate measurement of the very low concentrations of receptor present in mammary glands from virgin rats and was independent of endogenous ligand (rat PRL or placental lactogens) bound to
2981
the receptor, since U5 binds to a site different from the PRL binding site (28). Scatchard analysis performed in parallel with labeled U5 or oPRL revealed that U5 always detected approximately 3 times more receptor than oPRL, but both measurements were fairly proportional; in other words, the different number of receptors observed in the different reproductive states and the different number of receptors measured in liver and mammary gland were maintained with both methods, thus indicating that U5 is able to accurately detect alterations in receptor numbers. The higher quantities of PRL receptor measured with U5 could be due to the fact that endogenous ligands are bound to receptors, which would be masked in the oPRL binding assay; and also that U5 may recognize receptor forms that are unable to bind PRL or finally that labeled PRL is degraded more than labeled U5 during the incubation, artificially lowering the binding values. Since we were unable to perform reliable Scatchard analysis in mammary gland membranes of virgin rats, we could not compare U5 with oPRL binding in states with low serum PRL levels. In summary, these results show that measurement of PRL receptors with U5 may be a more reliable and sensitive assay than the classical assay using PRL binding. The investigation of PRL receptor mRNAs showed that essentially the same mRNA transcripts were found in mammary gland and liver of female rats, namely, a predominant 1.8 kb form, that encodes the short form of the receptor and additional, minor bands at 2.5, 3, and 5.5 kb, that encode the long form of the receptor. The two noticeable differences between mammary gland and liver were that there was on the average between 30 (in lactating rats) to 1500 (in late gestation) times more PRL receptor mRNA in liver than in mammary gland, and that approximately 30% of total PRL receptor mRNA corresponds to the long form in the mammary gland compared to liver where only 14% corresponds to the long form of the receptor. The mammary gland appears to utilize PRL receptor mRNA more efficiently in the synthesis of the receptor protein than liver, or receptor mRNA half-life may be much greater. Although liver membranes had more receptors than mammary gland membranes, the liver to mammary gland ratios ranged from 10 in late gestation to 1 in lactation; these ratios are much smaller than the ratio observed for mRNAs (see above). The changes observed during pregnancy and lactation in PRL receptor mRNA were paralleled when membrane receptor was measured, but there were marked differences between mammary gland and liver, which showed opposite variations. Thus, in liver, there was a marked increase in receptor levels at the end of pregnancy,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
2982
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES
Endo • 1991 Voll28«No6
PRL rtctptor mRNAs
B
Mammary Gland
FIG. 5. Effect of RU 486 on the expression of PRL receptor mRNA in liver and mammary gland from rats on day 19 of pregnancy. Triplicate Northern blots of poly (A+) RNAs from liver and mammary glands from rats injected with RU 486 or oil (controls) at 1000 h on day 18 of pregnancy and killed 24 h later were hybridized sequentially with I, E, and actin probes. Panel A shows typical autoradiograms for the three probes; panel B shows the amount of each form of the receptor quantified by densitometry expressed as the ratio of densities of each specific form of the receptor (long or short as described in legend to Fig. 4) to the actin signal for 24 h of exposure (values are means ± SEM).
Liver
Liver
QI5
3OOr
O.IO
200
I I Controls EZ2 RU 466
M.G. E Prob*
1
I Prob«
I
.0.05
.. m
100
Chicken B - A c t i n
RU-486 Day 19 of gestation followed by a sharp fall after parturition and very low values throughout lactation. In contrast in mammary tissue, levels remained low through pregnancy, increased sharply on day 21, and continued to increase in lactation, confirming previous results (15, 16). After blockade of progesterone receptor with RU 486, PRL receptor mRNA increased in mammary gland, suggesting that the blockade of progesterone induces PRL receptor mRNA. It had been previously shown that PRL receptors (measured by binding to oPRL) were low in pregnant mammary tissue and increased sharply after ovariectomy-hysterectomy of the pregnant rat, to the levels found in lactating tissue (18). The authors concluded that the removal of the uterine-fetal lactogenic hormones, such as placental lactogen, unmasked preexistent PRL receptors and that in reality, PRL receptors were high in mammary gland from pregnant rats, but not measurable, since they were occupied by endogenous lactogenic hormones (18). Our results with U5 binding, as well as with PRL receptor mRNAs, demonstrate that PRL receptor concentrations in mammary gland during pregnancy are low, and that they are induced after the blockade of progesterone receptors, previous to parturition or when progesterone levels fall to basal levels, which also triggers PRL secretion and lactogenesis (35, 36). After the fall in progesterone at the end of pregnancy, PRL secretion increases, which in turn induces milk protein synthesis by acting on the transcription of milk protein genes and stabilization of their mRNA. Finally PRL is able to increase its own receptor and the sensitivity of the mammary gland to the hormone. This conclusion may be strengthened by the fact that progesterone inhibits PRL induction of
Lo«*
Start
Total
Long Short font) fonn (prebtOD (1.8 kb)
Total
its own receptor (22). It is interesting to note that on day 21 of pregnancy or on day 19 after RU 486 administration, membrane PRL receptors were not increased in mammary gland, when compared to values on day 19, although there was an increase in their mRNA; this phenomenon may correspond to the time lag between the transcription of specific mRNAs and the synthesis of the corresponding proteins or perhaps the increase in membrane receptors 24 h after RU 486 treatment was too small to be detected. Moreover, the action of RU 486 seemed to be restricted only to mammary tissue, since in liver there were neither modifications in PRL receptor mRNAs nor in membrane receptor, the levels of which were already very high. This indicates that progesterone probably inhibits PRL receptor gene expression specifically in the mammary gland during pregnancy. The intracellular mechanism of action of PRL after binding to its receptor that leads to activation of specific protein synthesis and other intracellular actions is still unknown. The different forms of the receptor, long and short, are probably responsible for different actions of the hormone within the same or in different tissues, and perhaps the different relation of long to short forms of the receptor in mammary gland, liver, and ovary (our present results and Refs. 24 and 27), could account for the different actions of PRL in these three tissues. We have recently shown that only the long form of PRL receptor is able to stimulate the transcription of milk protein genes (39). Another interesting question is, in spite of the hormonal environment to which the liver and mammary
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES •251-U5 binding to membranes
Liver 20 r Liver 40
Vol. 128, No. 6 Printed in U.S.A.
Prolactin Receptor Gene Expression in Rat Mammary Gland and Liver during Pregnancy and Lactation G. A. JAHN*, M. EDERY, L. BELAIR, P. A. KELLY, AND J. DJIANE Unite d'Endocrinologie Moleculaire (G.J., M.E., L.B., J.D.), Bdtiment des Biotechnologies, Institut National de la Recherche Aaronomique, 78350-Jouy-en-Josas, France; and Laboratory of Molecular Endocrinology (P.K.), McGill University, Royal Victoria Hospital, Montreal, Quebec, Canada H3A 1A1
ABSTRACT. The expression of two forms of PRL receptor messenger RNA was measured at different stages of pregnancy and lactation in mammary gland and liver from Sprague-Dawley rats, using 32P-labeled complementary DNA probes encoding the extracellular part of the receptor (E probe), common to the two forms and a probe encoding the intracellular part of the long form of the receptor (I probe), that only recognizes sequences specific to the long form of the receptor. Hybridizations were performed in Northern blots obtained from electrophoreses of poly (A+) enriched RNA preparations from mammary glands and livers of rats on days 0, 6, 12, 19, and 21 of pregnancy and 5, 10, 15, and 20 of lactation. The Northern blots were also hybridized with a chicken /3-actin probe, to correct for the amount of mRNA added and the different metabolic states of the tissues. Both tissues expressed the same forms of PRL receptor mRNAs, namely bands at 2.5, 3, and 5.5 kilobases encoding the long form of the receptor and a major band at 1.8 kilobases encoding the short form. The liver expressed all the receptor mRNA forms in much higher quantity than the mammary gland, independent of the reproductive state. In liver there was an increase of all the transcripts on day 19 of pregnancy, followed by an abrupt decline at the onset of lactation, to levels lower than those of virgin rats. In contrast, mammary gland PRL receptor mRNAs were low in virgin and pregnant animals,
P
ROLACTIN is involved in an impressive number of biological actions in different target organs, of which the most studied has been the initiation and maintenance of milk production by the mammary gland (1). The actions of PRL in its different target organs are initiated through interaction with specific, high affinity receptors at the cell surface. Different biocheminal studies have shown that the majority of PRL receptors have a molecular mass of approximately 40,000 although a form of mol wt 80,000 has also been characterized, particularly in the ovary (2-5). PRL receptor levels are differentially regulated according to the tissue examined. In rat liver, one of the tissues Received June 7, 1990. Address requests for reprints to: Dr. Jean Djiane, INRA, Batiment des Biotechnologies, 78350 Jouy-en-Josas, France. * Present address: Laboratorio de Reproduction y Lactancia, CRICYT-CONICET, C.C. 855, 5500 Mendoza, Argentina.
increased significantly at day 21 of pregnancy, and continued to increase throughout lactation. Treatment of day 19 pregnant rats with the antiprogesterone RU 486 induced, 24 h later, PRL receptor mRNAs in mammary gland but not in liver. There were no significant differences in the relative proportions of long to short forms of PRL receptor mRNAs at the different reproductive states, but the proportion of the long form was slightly greater in mammary gland than in liver. Membrane PRL receptor concentrations were also measured in the same tissues used for the mRNA study by binding to a 125I-labeled monoclonal antibody (U5), which specifically recognizes the PRL receptor at a site different from the hormone binding site. The quantity of receptor measured by U5 binding was approximately 3 times higher than that measured with 125I-labeled ovine PRL. The pattern of U5 binding to membranes observed during pregnancy and lactation in liver and mammary gland was parallel to that observed for PRL receptor mRNAs, with the exception that the increases observed in mammary gland on day 21 of pregnancy or after RU 486 treatment were not seen. The present results indicate that PRL receptors are differentially regulated in liver and mammary gland during pregnancy and lactation in the rat and that the decline of progesterone at the end of pregnancy induces PRL receptor mRNA in mammary gland but not in liver. (Endocrinology 128: 2976-2984, 1991)
with the highest PRL binding, there are sex differences in PRL receptor numbers (6), and estrogens (7, 8), PRL, and GH stimulate (9-12) binding. Receptor levels also vary during the estrous cycle (13, 14), pregnancy (6, 15, 16), and lactation (16,17). Although the mammary gland is one of the most studied target tissues for PRL, the concentration of receptors seems to be very low in rats (16, 18), being almost undetectable in tissue from virgin animals. In mammary tissue, PRL receptors increase during lactation (16, 20) and are induced by glucocorticoids (21) and PRL itself (22), while progesterone may exert a negative influence (22, 23). Recently, rat, rabbit, and human PRL receptor complementary DNAs have been cloned and sequenced (2426), and several forms of different size have been detected by hybridization in rat tissues (24, 27). Moreover, two different receptor cDNAs, differing in the length and sequence of the intracellular region, have been described
2976
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES (27) and encode receptors of different sizes. A cDNA encoding the short form of the receptor messenger RNA was obtained from a estrogen-treated liver cDNA library, and the long form was originally detected from a rat ovary library (24, 27). After binding of PRL to its receptor, the mechanism by which the PRL signal is transferred inside the cell remains unknown. The different forms of the PRL receptor could mediate different physiological actions of the hormone in the same or different tissues. In the present study, using a probe derived from the extracellular region of the receptor (E probe, common to both forms) and another obtained from the intracellular sequence of the long form (I probe), we have examined the expression of PRL receptor mRNAs in mammary gland and liver from rats at different stages of pregnancy and lactation. In order to compare concentrations of mRNA to concentrations of PRL receptors in membranes, we have also measured the binding of [ 125 I]PRL and [ 125 I]monoclonal antibody (U5, Ref. 28). This monoclonal antibody binds with high specificity to PRL receptor, at a site different from the PRL binding site, thus obviating the distortion of results produced by occupation of the receptors with endogenous hormone.
Materials and Methods Animals and experimental procedures Adult virgin female and male Sprague-Dawley rats were obtained from Iffa-Credo (Orleans, France) and kept in a light (lights on 0600-2000 h)- and temperature-controlled room; rat chow and tap water were available ad libitum. The rats were allowed a period of acclimation, after which daily vaginal smears were taken and on the night of proestrus females were caged with a fertile male. The presence of spermatozoa was ascertained in the vaginal smears the following morning, and this day was considered as day 0 of pregnancy. Groups of three rats each were killed by decapitation between 1030 and 1100 h on days 0, 6, 12, 19, and 21 of pregnancy and 5, 10, 15, and 20 of lactation. The day 0 rats were unmated animals on the day of estrus. The lactating rats nursed litters of eight pups. After decapitation, the whole inguinal and abdominal mammary glands and livers were quickly excised and frozen immediately in liquid nitrogen. The samples were kept at -80 C until portions were processed for RNA extraction and membrane preparation. In a group of pregnant rats, RU 486 [RU 38486; 17/3-hydroxy11 j8-(4-dimethyl-aminophenyl)-17-propinyl-estra-4,9-dien-3one, generously provided by Roussel-Uclaf, Romainville, France], was injected sc at a dose of 2 mg/kg in sesame oil (2 g/liter) at 1030 h on day 18 of pregnancy. These rats were killed 24 h later and mammary glands and livers obtained. RNA extraction and hybridization Total RNA was extracted using the guanidinium isothiocyanate-LiCl precipitation. Briefly, tissues were homogenized
2977
in 10 vol sterile 4 M guanidine isothiocyanate, 50 mM Tris HC1, pH 7.5, EDTA 10 mM, mercaptoethanol 8%, filtered through sterile gauze and centrifuged 10 min at 2000 X g. The supernatants were diluted with 5.5 vol sterile 4 M LiCl and allowed to precipitate for 48-72 h at 4 C. The precipitate was obtained by centrifugation at 15,000 X g during 90 min, washed twice with 3 M LiCl, and extracted twice with 5-15 ml TES buffer (Tris HC110 mM, pH 7.5 EDTA 1 mM, sodium dodecyl sulfate 0.2%). The pooled extraction supernatants were extracted with 1 vol of a mixture consisting of 0.5 vol water-saturated phenol -0.5 vol chloroform-isoamyl alcohol (24:1). The total RNAs in the aqueous phase were precipitated with 0.1 vol sodium acetate, 3 M pH 5.1 and 2.5 vol absolute ethanol. Aliquots of total RNAs were enriched in poly (A+) RNAs by two passages through an oligo dT cellulose column (29). The samples of poly (A+) RNAs were size separated by electrophoresis in agarose (1.3%, Seakem, FMC Bioproducts, Rockland, ME), formaldehyde vertical gels and transferred to nitrocellulose paper (Hybond C extra, Amersham, Arlington Heights, IL) (30). For size calibration of the electrophoreses, a 0.24-9.5 kilobase (kb) RNA ladder (BRL, Bethesda Research Laboratories Life Technologies, Inc., Rockville, MD) or a Pharmacia DRIgest III 0.07-23 kb/DNA ladder (Pharmacia, Bromma, Sweden) were run on each gel. The probes used for PRL receptor mRNA hybridization were gel-purified inserts from plasmids containing the cDNAs encoding the extracellular domain of the rat PRL receptor cDNA (probe E, nucleotides 328-619, Ref. 24), for the cytoplasmic domain of the long form of the receptor (probe I, nucleotides 1253 -1735, Ref. 27) and for chicken /3-actin. The filters were prehybridized for at least 4 h at 42 C in a solution containing 50% formamide, Denhardt's solution x 5, SSPE (1.15 M NaCl/0.01 M NaH2 PO4/1 mM EDTA) X 5, salmon DNA 100 n\, sodium dodecyl sulfate 0.1%, polyadenylic acid (5') type I, K+ salt (Sigma Chemical Co., St Louis, MO) 1 pg/ml. Hybridization solution contained, in addition to the above, 5% dextran and 500,000 cpm of the appropriate probe, previously labeled with 32P by random priming (31). Filters were hybridized overnight at 42 C and washed twice in 0.1 SSC (0.15 M NaCl/0.015 M sodium citrate, pH 7.0) with 0.1% sodium dodecyl sulfate and 1 mM EDTA at 65 C. The filters were hybridized first with the probe for the cytoplasmic part of the long form of the PRL receptor, stripped and rehybridized with the probe for the extracellular probe, and finally stripped once more and hybridized with the chicken /3-actin cDNA, as control for the differences in the amount of RNA loaded on the gel and also as control for the dilution of the specific mRNAs found in the lactating rat mammary gland samples in relation to the mammary glands of virgin or pregnant rats. Membrane preparation and measurement of membrane receptor Membranes were prepared from mammary gland and liver by homogenization of tissues in 5 vol sucrose 0.3 M and centrifugation for 15 min at 1500 X g. Supernatants were spun 90 min at 100,000 X g to obtain crude membrane pellets which were resuspended in 2 vol Tris HC1 25 mM, pH 7, MgCl2 10 mM (32). These preparations were frozen at -20 C until measurement of membrane binding to [125I]ovine (o) PRL or the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES
2978 I25
Endo«1991 Voll28«No6
I-L)5 binding to membranes
4 0 r Liver l25
30
I6r
I - U s binding to membranes
Liver
2 20
o 10
E o
w
0 4_ Mammary
6 Gland
12
19 21
m 0
6
12
19 21
20
Mammary Gland
E
§0.5
0 Day of
12 19 21 Pregnancy
10
15 20 Lactation
Oay of
12 19 21' Pregnancy
15 20 Lactation
FIG. 1. Concentration of membrane PRL receptor during pregnancy and lactation in rats. Relative receptor concentrations were measured by binding of [125I]U5 to liver or mammary gland membranes from rats on the indicated days of pregnancy and lactation as described in Materials and Methods. Results are expressed as means ± SEM of (A) counts per min of U5 specifically bound per mg protein and (B) counts per min of U5 specifically bound per g tissue of three replicate determinations. TABLE 1. Comparison of number of PRL receptor sites measured by Scatchard analysis performed with [125I]U5 or [125I]oPRL in liver and mammary gland membranes obtained from rats at different stages of pregnancy (P) and lactation (L) Binding (fmol/mg protein)
Liver P day 19 L day 5 L day 10 Mammary gland P day 19° L day 5 L day 10 L day 20 0
U5
oPRL
U5/oPRL
1488 90 189
459 40 49
3.2 2.3 3.7
counter (32). Scatchard analysis to compare the number of sites measured with U5 and oPRL binding was performed for some of the samples, by incubating 50,000 cpm of tracer and increasing concentrations (0.25 ng to 1 ng) of the corresponding unlabeled tracer. Tracers were labeled with 125I using the Chloramine T method (34), and labeled protein was separated from free iode using an acrylamide agarose column (ACA 54, LKB, Bromma, Sweden). Specific activities were 76-85 and 4-14 iiC\/iig for oPRL and U5, respectively.
Results 39 41 93 62
10 16 28 21
3.9 2.6 3.3 3.4
Treated with RU 486.
monoclonal antibody U5 (28). Aliquots were taken for protein measurement by the method of Lowry et al. (33). With the exception of mammary glands on day 0 and 6 of gestation, the tissues used for the U5 binding study were portions of the same tissues used for mRNA preparation. Binding assays were performed by incubating 25 jug (for liver samples) or 200 ng (for mammary gland samples) of membrane proteins suspended in 0.5 ml Tris HC1 25 mM, MgCl2 10 mM, BSA 0.1%, sodium azide 0.02% with 50,000 cpm of labeled tracer in the presence or absence of 1 i*g unlabeled tracer for measurement of specific or nonspecific binding. After incubation overnight at room temperature, tubes were diluted with 3 ml buffer and centrifuged, and the pellets were counted in a 7-
Evolution of membrane PRL receptor concentrations in mammary gland and liver during pregnancy and lactation PRL receptor concentrations in both tissues were measured by binding to a 125I-labeled monoclonal antibody, U5, which binds with great specificity to PRL receptor at a site different from the PRL binding site (28). Figure 1 shows the binding of [125I]U5 (counts per min/mg protein and per gram tissue) in liver and mammary gland. Scatchard analysis performed in some samples of pregnant and lactating liver and lactating mammary gland, and comparison with the specific binding values, showed that the number of receptors was always proportional to the specific binding of U5, using 50,000 cpm of [125I]U5. In order to correlate the number of sites detected by U5 with those detected with oPRL, we performed Scatchard analysis of some samples of pregnant
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES
FIG. 2. The different types of PRL receptor mRNAs in liver and mammary gland of rats. Northern blots performed with 25 ng poly (A+) RNA from pregnant rat liver or mammary glands were hybridized with a probe specific to the intracellular portion of the long form of the receptor (I probe), or the probe for the extracellular portion of the receptor (E probe), as indicated in Materials and Methods. Autoradiograms were exposed for the times indicated at the bottom of the lanes.
and lactating mammary gland and liver using both ligands. We were unable to use mammary glands from virgin animals for this study, since the binding of oPRL was too low to allow for accurate Scatchard analysis. The results (Table 1) showed that U5 consistently measured approximately 3 times the number of sites determined by PRL. Figure 1 shows that in liver membranes the number of PRL receptors is much greater than in mammary gland and that the number of sites varies inversely in both tissues. In liver, there is an increase during pregnancy and a decrease to the lowest values in lactation, while in mammary gland PRL receptor levels remain very low in virgin and pregnant rats and increase markedly after parturition. The increase seen in mammary gland receptor is more apparent when the results are expressed in terms of tissue wts due to the marked increase in protein content of the tissue as pregnancy and lactation progress (Fig. IB). PRL receptor mRNAs in mammary gland and liver of female rats Northern blot analysis of mRNAs obtained from liver with the I probe showed three bands, two at 2.5 and 3 kb of approximately the same intensity and a third, much weaker, at 5.5 kb (Fig. 2). When the same filters were hybrized with the E probe, a new signal, much stronger than all the others, was seen at 1.8 kb, with fainter bands at 2.5, 3, and 5.5 kb, which in fact were the same bands that hybridized with the I probe. When mammary gland mRNAs were analyzed with both probes, signals at 2.5, 3, and 5.5 were also obtained with the I probe. However, using the E probe, a strong band at 1.8 was seen, with faint signals observed at 2.5, 3, and 5.5 kb, which did not appear in all samples. Densitometric analysis of the autoradiographic films revealed that the proportion of the long form of the receptor in mammary gland was slightly higher than that obtained in liver. Thus, essentially the same forms of
2979
Mi. Ui .
PRL receptor mRNAs were found in mammary gland and liver, although the intensity of the bands obtained with both probes in mammary gland was much fainter than that observed in the liver (Fig. 2) Evolution of PRL receptor mRNAs during pregnancv and lactation Figure 3 shows the Northern blot analysis with both receptor probes and with an actin probe, of filters containing liver and mammary gland mRNAs obtained from the different stages of pregnancy and lactation. In liver tissue an increase was observed toward the end of pregnancy followed by an abrupt decrease, to very low values during lactation. In contrast, in mammary gland, mRNAs remained relatively low throughout pregnancy, started to increase at day 21 of pregnancy, and rose to the highest values during lactation. The actin signal in liver did not show any marked variation throughout pregnancy and lactation, whereas in mammary gland, there was a gradual decrease in the signal from the start of pregnancy, with only very faint signals being obtained in lactating tissues. This gradual diminution of the actin signal may be accounted for by dilution of the actinspecific mRNAs by the increased milk protein mRNAs in late pregnant or lactating tissue (37, 38), which also would dilute the PRL receptor-specific mRNAs. In order to quantify the amount of PRL receptor mRNAs present at each stage of pregnancy and lactation in liver and mammary gland, triplicate filters hybridized with both probes were analyzed by densitometry and corrected to uniform actin concentration. The values shown for the long form correspond to the sum of scan intensities of all the bands hybridizing with the I probe, and the values for the short form are the densitometric scan intensities of the 1.8 kb band hybridized with the E probe. Results, shown in Fig. 4, indicate that in both tissues the variations of long and short forms of the receptor were similar, and the relative proportions of both forms did not change significantly with the repro-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
2980
Day
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES
0
6
6ns
Endo«1991 Voll28«No6
fC 15 2( Lactatkn
FIG. 3. Evolution of PRL receptor mRNAs during pregnancy and lactation in liver (A) and mammary gland- (B). Northern blots of poly (A+) RNAs from liver and mammary glands from rats on different days of pregnancy and lactation were hybridized sequentially with I probe, E probe, and chicken /3-actin probe, as described in Materials and Methods. Autoradiographic films of Northern blots for liver were exposed 4 h for the E probe and 24 h for the other two probes, and those from mammary gland were exposed 6 days for the I and E probes and 24 h for the actin probe. Lanes from left to right represent days 0, 6, 12, 19, and 21 of pregnancy and 5, 10, 15, and 20 of lactation.
ductive stage. On the other hand, in liver, PRL receptor mRNAs increased at the end of pregnancy and fell to very low values during lactation, while in mammary gland mRNA levels were low in tissue from virgin or pregnant rats, started to increase on day 21 of pregnancy, and reached the highest values during lactation, with levels approximately 10 times higher than in virgin or pregnant rats. The proportion of long to short form was smaller in liver than in mammary gland, accounting for approximately 14% of the total while in the mammary gland long form accounted for approximately 33% of the total. The increase in late pregnancy in the liver was more marked for the short than for the long form.
Effect of RU 486 on PRL receptor mRNA expression in late pregnant rats The fall of progesterone that occurs at the end of pregnancy is known to induce lactogenesis and PRL secretion (35, 36). In order to determine whether the fall of progesterone is also able to increase expression of PRL receptors, rats were injected with the progesterone antagonist RU 486 at 1000 h on day 18 of pregnancy and killed 24 h later, at a time when the steroid has already induced lactogenesis. Figure 5 shows that RU 486 treatment induced a significant increase in PRL receptor mRNAs in mammary gland but not in liver. In contrast,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES PRL receptor mRNAS in liver O Long form 200
• Short form (1.8 kb)
100
OJ 0 Day of
12
19 21 pregnancy
15 20 lactation
PRL receptor mRNAS in mammary gland O Long form • Short form (1.8 kb)
0.5.
OJ
B
0 Day of
12
19 21 pregnancy
15 20 lactation
FIG. 4. Quantification of PRL receptor mRNAs from liver (A) and mammary gland (B) of rats during pregnancy and lactation. Triplicate Northern blots, hybridized sequentially with the I, E, and actin probes were analyzed by densitometry. The relative expression of the long form of the receptor (that hybridized with the I probe) was expressed as the ratio of intensities of the specific bands to the actin probe, all corrected for 24 h of exposure (means ± SEM of three replicate Northern blots).
membrane receptor levels measured with U5 did not increase after RU 486 in either tissue (Fig. 6). Discussion The present paper describes the measurement of PRL receptor and its mRNAs during pregnancy and lactation in the rat mammary gland and liver. Membrane PRL receptors were measured using an 125 -I labeled monoclonal antibody (28) instead of the classical PRL binding assay. This procedure has several advantages, since the monoclonal antibody gave greater sensitivity, allowing accurate measurement of the very low concentrations of receptor present in mammary glands from virgin rats and was independent of endogenous ligand (rat PRL or placental lactogens) bound to
2981
the receptor, since U5 binds to a site different from the PRL binding site (28). Scatchard analysis performed in parallel with labeled U5 or oPRL revealed that U5 always detected approximately 3 times more receptor than oPRL, but both measurements were fairly proportional; in other words, the different number of receptors observed in the different reproductive states and the different number of receptors measured in liver and mammary gland were maintained with both methods, thus indicating that U5 is able to accurately detect alterations in receptor numbers. The higher quantities of PRL receptor measured with U5 could be due to the fact that endogenous ligands are bound to receptors, which would be masked in the oPRL binding assay; and also that U5 may recognize receptor forms that are unable to bind PRL or finally that labeled PRL is degraded more than labeled U5 during the incubation, artificially lowering the binding values. Since we were unable to perform reliable Scatchard analysis in mammary gland membranes of virgin rats, we could not compare U5 with oPRL binding in states with low serum PRL levels. In summary, these results show that measurement of PRL receptors with U5 may be a more reliable and sensitive assay than the classical assay using PRL binding. The investigation of PRL receptor mRNAs showed that essentially the same mRNA transcripts were found in mammary gland and liver of female rats, namely, a predominant 1.8 kb form, that encodes the short form of the receptor and additional, minor bands at 2.5, 3, and 5.5 kb, that encode the long form of the receptor. The two noticeable differences between mammary gland and liver were that there was on the average between 30 (in lactating rats) to 1500 (in late gestation) times more PRL receptor mRNA in liver than in mammary gland, and that approximately 30% of total PRL receptor mRNA corresponds to the long form in the mammary gland compared to liver where only 14% corresponds to the long form of the receptor. The mammary gland appears to utilize PRL receptor mRNA more efficiently in the synthesis of the receptor protein than liver, or receptor mRNA half-life may be much greater. Although liver membranes had more receptors than mammary gland membranes, the liver to mammary gland ratios ranged from 10 in late gestation to 1 in lactation; these ratios are much smaller than the ratio observed for mRNAs (see above). The changes observed during pregnancy and lactation in PRL receptor mRNA were paralleled when membrane receptor was measured, but there were marked differences between mammary gland and liver, which showed opposite variations. Thus, in liver, there was a marked increase in receptor levels at the end of pregnancy,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
2982
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES
Endo • 1991 Voll28«No6
PRL rtctptor mRNAs
B
Mammary Gland
FIG. 5. Effect of RU 486 on the expression of PRL receptor mRNA in liver and mammary gland from rats on day 19 of pregnancy. Triplicate Northern blots of poly (A+) RNAs from liver and mammary glands from rats injected with RU 486 or oil (controls) at 1000 h on day 18 of pregnancy and killed 24 h later were hybridized sequentially with I, E, and actin probes. Panel A shows typical autoradiograms for the three probes; panel B shows the amount of each form of the receptor quantified by densitometry expressed as the ratio of densities of each specific form of the receptor (long or short as described in legend to Fig. 4) to the actin signal for 24 h of exposure (values are means ± SEM).
Liver
Liver
QI5
3OOr
O.IO
200
I I Controls EZ2 RU 466
M.G. E Prob*
1
I Prob«
I
.0.05
.. m
100
Chicken B - A c t i n
RU-486 Day 19 of gestation followed by a sharp fall after parturition and very low values throughout lactation. In contrast in mammary tissue, levels remained low through pregnancy, increased sharply on day 21, and continued to increase in lactation, confirming previous results (15, 16). After blockade of progesterone receptor with RU 486, PRL receptor mRNA increased in mammary gland, suggesting that the blockade of progesterone induces PRL receptor mRNA. It had been previously shown that PRL receptors (measured by binding to oPRL) were low in pregnant mammary tissue and increased sharply after ovariectomy-hysterectomy of the pregnant rat, to the levels found in lactating tissue (18). The authors concluded that the removal of the uterine-fetal lactogenic hormones, such as placental lactogen, unmasked preexistent PRL receptors and that in reality, PRL receptors were high in mammary gland from pregnant rats, but not measurable, since they were occupied by endogenous lactogenic hormones (18). Our results with U5 binding, as well as with PRL receptor mRNAs, demonstrate that PRL receptor concentrations in mammary gland during pregnancy are low, and that they are induced after the blockade of progesterone receptors, previous to parturition or when progesterone levels fall to basal levels, which also triggers PRL secretion and lactogenesis (35, 36). After the fall in progesterone at the end of pregnancy, PRL secretion increases, which in turn induces milk protein synthesis by acting on the transcription of milk protein genes and stabilization of their mRNA. Finally PRL is able to increase its own receptor and the sensitivity of the mammary gland to the hormone. This conclusion may be strengthened by the fact that progesterone inhibits PRL induction of
Lo«*
Start
Total
Long Short font) fonn (prebtOD (1.8 kb)
Total
its own receptor (22). It is interesting to note that on day 21 of pregnancy or on day 19 after RU 486 administration, membrane PRL receptors were not increased in mammary gland, when compared to values on day 19, although there was an increase in their mRNA; this phenomenon may correspond to the time lag between the transcription of specific mRNAs and the synthesis of the corresponding proteins or perhaps the increase in membrane receptors 24 h after RU 486 treatment was too small to be detected. Moreover, the action of RU 486 seemed to be restricted only to mammary tissue, since in liver there were neither modifications in PRL receptor mRNAs nor in membrane receptor, the levels of which were already very high. This indicates that progesterone probably inhibits PRL receptor gene expression specifically in the mammary gland during pregnancy. The intracellular mechanism of action of PRL after binding to its receptor that leads to activation of specific protein synthesis and other intracellular actions is still unknown. The different forms of the receptor, long and short, are probably responsible for different actions of the hormone within the same or in different tissues, and perhaps the different relation of long to short forms of the receptor in mammary gland, liver, and ovary (our present results and Refs. 24 and 27), could account for the different actions of PRL in these three tissues. We have recently shown that only the long form of PRL receptor is able to stimulate the transcription of milk protein genes (39). Another interesting question is, in spite of the hormonal environment to which the liver and mammary
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 July 2015. at 17:40 For personal use only. No other uses without permission. . All rights reserved.
PRL RECEPTOR GENE EXPRESSION IN RAT TISSUES •251-U5 binding to membranes
Liver 20 r Liver 40
