- Email: [email protected]
B6 rat prolactin cells
Molecular and Cellular Endocrinology, 18 (1980) 123 -136 0 Elsevier/North-Holland Scientific Publishers, Ltd.
EFFECT OF 17fl-ESTRADIOL GH3/B6 RAT PROLACTIN
ON THYROLIBERIN
123
RESPONSIVENESS IN
CELLS
N. BRUNET, D. GOURDJI and A. TIXIER-VIDAL Groupe de Neuroendocrinologie 75231 Paris Cedex 05 (France)
Cellulaire, Coll;ge de France, 1 I, Place
Marcelin'Berthelot,
Received 15 October 1979; accepted 7 February 1980
GH3/B6 rat prolactin cells were used to analyse at the cellular level the mechanisms by which 17p-estradiol (Ea) regulates TRH responsiveness of prolactin cells. Before experiments, cells were grown for up to 7 days in 3 different media: normal medium (N) containing 15% horse serum and 2.5% fetal calf serum, CD medium prepared with charcoal-dextran extracted serum and CDE medium supplemented with 4 X 10 -8 M Ea. The binding of 3H-TRH (30 min at 37°C) and the TRH-induced percent increase of prolactin release as a function of TRH doses were compared in the 3 conditions. Preculture in Ea enriched medium increased by 50% the number of TRH high-affinity binding sites without modifying their affinity, increased by up to 3 times the percent of the TRH-induced stimulation of prolactin release and improved by one order of magnitude the EDse of the TRH effect on prolactin release. The presence of HEPES (10 mM) during TRH challenge masked the effect of Ea on the increase in number of binding sites but respected its potentiating effect on prolactin release Keywords:
17pestradiol; thyroliberin; prolactin release; binding; cell culture.
Estrogens and thyroliberin (TRH) are, separately, well known potent stimulators of prolactin (PRL) secretion in vivo as well as in vitro. In addition, an interaction between estrogens and TRH has been observed in vivo. A facilitating effect of 17P-estradiol (Es) pretreatment on the TRH-induced PRL release has been described in the human female (Rakoff et al., 1973) and more recently in female rhesus monkeys (Quadri et al., 1979). In female rats, E, pretreatment has been shown to enhance both the number of TRH-binding sites in pituitary homogenates and the stimulation of PRL release in response to a TRH i.v. injection (De Lean et al., 1977 a,b). These experiments suggested that estrogens exert their action by modulating the number of TRH receptors in pituitary target cells. Further analysis at tlfe cellular level was nevertheless limited by the fact that in these experiments at least two pituitary cell types - thyrotropic cells and PRL cells - were modified in both their secretory activity and mitotic activity. Experiments with primary cultures of dispersed anterior pituitary cells although allowing to localize at the pituitary level a site of this interaction (Raymond et al., 1978; Padmanabhan and
124
N, Brunet, D. Gourdji, A. Tixier- Vidal
Convey, 1979) did not overcome this drawback because of their cellular heterogeneity. Clonal rat PRL cell lines which offer a homogeneous population of target cells might represent an appropriate model to study such an interaction. Indeed, GH3 rat pituitary cell lines (Tashjian et al., 1968) which secrete PRL and, to a lesser extent growth hormone (Tashjian et al., 1970) have been shown to respond to TRH (Tashjian et al., 1971; Gourdji et al., 1972) and to E2 (Tashjian and Hoyt, 1972). TRH exerts a biphasic stimulating action on PRL secretion: an early effect on the release of a stored PRL pool, and a delayed effect on the synthesis of PRL (Morin et al., 1975; Dannies et al., 1976). 3H-THR binds to intact GH3 cells. Equilibrium binding is reached after 15-30 min at 37” (Gourdji et al., 1973; Hinkle and Tashjian, 1973). This binding is also dose-dependent and 2 apparent dissociation constants have been found: K - 4 X lo-’ M, in the range of the physiological doses (0.05-50 nM TRH) > andok- D2= 6 X lo-* M for higher doses, up to 850 nM (Tixier-Vidal et al., 1975; Gourdji et al., 1976; Gourdji, in press). GH3 cells are also responsive to Ea. After several days in Ez supplemented medium, PRL production by GH3 cells was enhanced (Tashjian and Hoyt, 1972). Although the same group reported later that these cells have lost this property (Dannies et al., 1977), other authors (Haug and Gautvik, 1976) using the same cell line have found that Ez (IO-’ M) increases prolactin production following treatment for 4 days up to 10 days and this effect was additive to that of TRH (3 X lo-’ M) (Haug and Gautvik, 1976). The study of the effect of E, in cells which are routinely grown in serum-supplemented medium is hampered by the presence of non-negligible amounts of this steroid in the serum. Using GH3/B6 cells, a subclone of the GH3 line, which were precultured for one week in medium supplemented with charcoal dextran extracted serum (CD medium), we have previously shown that Ea (4 X 10m8 M) is able to reestablish the normal cell morphology, to maintain the number of TRH-binding sites and to potentiate the TRH-induced stimulation of PRL release (Brunet et al., 1977). In addition PRL production was increased over normal level after one-week culture in CD medium supplemented with Ez (Tixier-Vidal et al., 1978). We report here the results of a study which has been performed using the same experimental model in order to analyse the mechanism by which E2 interacts on TRH binding and TRH responsiveness by GH3/B6 cells. It is shown that Ea increases the number of TRH-binding sites without modifying their affinity for TRH, increases by up to 3 times the percent of the TRH-induced stimulation of PRL release and improves by one order of magnitude the ED,c of the TRH effect .on PRL release. After completing this manuscript we had access to the paper of Gershengorn et al. (1979) who also found that in GH3 cells estrogens increase the number of TRH receptors without modifying their affinity. These authors nevertheless did not correlate the binding with the dose effect of TRH on rapid PRL release.
125
TRH binding and PRL release after estrogen treatment
MATERIALS
AND METHODS
Cell culture
Cells were grown in control conditions (N medium) as previously described et al., 1973) in Ham’s FIO medium enriched with 15% heat inactivated horse serum (Eurobio, Paris), 2.5% fetal calf serum (Eurobio, Paris) and with antibiotics (penicillin 50 mu/ml plus streptomycin 50 pg/ml, or gentiamicin 40 ,ug/ml). A single batch of horse serum as well as of fetal calf serum was used for all of the experiments. CD medium was prepared using horse serum from the same batch previously treated with charcoal dextran according to Kirkland et al. (1976). CDE medium was freshly prepared by adding Es after serial dilutions in FIO medium starting from an ethanol solution. The final Es concentration in the 3 media (N, CD, CDE) is shown in Table 1. (Gourdji
Experimental schedules GH3/B6 cells previously grown in N medium where harvested by mild trypsinisation (0.25%) and plated in tissue-culture petri dishes (Falcon, 3.5 cm diameter, 2 or 2.5 X lo5 cells/2 ml medium) in the 3 media, N, CD, and CDE resp. They were incubated at 37°C in a humidified atmosphere of 5% CO, in air. The cells were grown in those conditions for 3, 5 or 7 days, the medium being renewed every 2 or 3 days. At the end of each of these different periods, the cells were assayed for TRH responsiveness. Binding experiments 24 h before experiments the culture medium was renewed and in some experiments supplemented with 10 mM HEPES (N-2-hydroxyethylpiperazine-N’-2 ethanesulfonic acid). On the day of experiment, the last culture medium was discarded and the cells were washed with 1 ml of warmed Ham’s FlO. They were then covered with 1 ml of appropriate medium (N, CD or CDE), with or without 10 mM HEPES, supplemented with various doses of 3H-TRH (60 Ci/mmole) (prepared by
Table 1 17p-Estradiol (Ez) and triiodothyronine Hormones
E2 T3
(T3) concentrations
in the 3 culture media
Media N
CD
CDE
0.04 nM 0.60 nM
0.01 nM 0.25 nM
40 nM 0.25 nM
Hormones were measured by specific radioimmunoassays in duplicate, in heat inactivated horse serum before and after treatment with charcoal-dextran, and in fetal calf serum.
126
N. Brunet, D. Gourdji, A. Tixier- Vidal
DURATION
OF
PRECULTURE (days)
Fig. 1. Cell proliferation as a function of time in culture in each of the 3 media. GH3/B6 cells were seeded at 200 000 cells/dish. At increasing time intervals, for each medium, 3 dishes were used for cell counting. Statistical comparisons were done using the Fisher F test. N medium (e), CD medium (A), CDE medium (a). Each point represents mean of triplicate k SEM. * P < 0.05 CDE vs. CD; ** P < 0.01 CD or CDE vs. N.
Table 2 Binding of 3H-TRH (27 nM, 30 min, 37°C) by GH3/B6 cells grown in N medium or CD medium or CDE medium for 3, 5 or 7 days Media
3H-TRH (fmoles/pg cell DNA) Preculture duration
N
CD CD!!~
~_
3 days
5 days
7 days
24.2 f 0.4 15.8 f 0.7 a 18.75 + 0.5 b,c
22.0 f I .l 16.1 f 0.8 a 22.4 f 0.4 ’ ~~___~
18.3 * 0.7 11.5 *0.4a 18.2 * 0.9 c ____
~~_
Results are expressed as fmoles of bound 3H-TRH per pg cell DNA. Each value represents mean of quadruplicate dishes + SEM. Statistical comparisons were done according to the Fisher F test. a P < 0.01 CD vs. N; b P < 0.01 CDE vs. N; c P < 0.01 CDE vs. CD.
TRH binding and PRL release after estrogen treatment
127
Levine-Pinto, CEA, Saclay, according to Pradelles et al., 1972) and incubated 30 min at 37’C in a controlled atmosphere. After cooling the dishes at 0-4°C the media were collected for prolactin determination. The cells were washed 4 times with ice-cold Ham’s FIO. They were then scraped in 1 ml of PBS. One aliquot of the cell sonicate (0.5 ml) was used for radioactivity determination and another aliquot for proteins and DNA determinations. 3H-TRH uptake was determined by the measure of the radioactivity bound to the cells as previously described (FaivreBauman et al., 1975). The non-compatible binding was estimated on parallel dishes and never exceeded 10%. Triplicate or quadruplicate dishes were used for each type of medium and each dose of 3H-TRH. Results were expressed as fmoles of TRH bound per gg DNA. Protein measurement Cell proteins were measured by Folin’s reagent according to Lowry et al. (19.51) using bovine serum albumin as standard. DNA measurement DNA was measured by the method of Karsten and Wollenberger modified by Bournaud et al. (1977).
(1972) slightly
Prolactin determination PRL medium content was measured by radioimmunoassay, using the rat PRL kit kindly provided by the NIAMDD rat hormone distribution program. Calculations Statistical significances were measured using the Fisher test. Binding data were analysed according to Scatchard (1949).
RESULTS (A) Celi proliferation Cell proliferation was estimated by cell counts, using a hemocytometer, in the 3 culture media as a function of time in culture (3-5-7 days). As shown in Fig. 1, culture in CD medium, supplemented or not with EZ, significantly stimulated cell proliferation. As compared to the initial number of cells (200 000), after 7 days of culture, the number of cells/dish was increased by S times in medium Ninstead of 7-8 times in medium CD or CDE. In contrast, there was no significant modification of the DNA content/cell (23-29 pg DNA/lo6 cells). The protein content per cell was not modified (37-41 pg/lO’ cells) except in cells grown in CD medium for 7 days, where it was significantly decreased (34 pg/105 cells) (F test: P< 0.01 CD vs. CDE). In the following experiments, the binding of 3H-TRH was therefore expressed per pg DNA, which represents a valuable index of the cell number.
33.1 f 3.5 a
17.1 f 3 29.9 f 1.7 e
19.1 a 4 41.4 f 4 e 75
115
35.5 + 9.7 a
15.9 f: 2.9 32.5 + 6.3 e
16.9 * 2 38.5 + 3.7 e
PRL ng/ng cell DNA
PRL ng/ng cell DNA % increase
5 days
3 days
100
128
% increase
12.1 f 0.9
7.4 f 0.8 b*c,d 16.6 * 1.7 e
15.3 f 2.4 28.1 f 2.7 e
PRL ng/Mg cell DNA
7 days
124
83
% increase
aH-TRH 65.1 f 4 e 197 89.8 i 7.8 e 253 37.6 i: 3.3 e 211 Results are expressed as ng PRL per ng cell DNA. Each value represents the mean of quadruplicate dishes +_S.E.M. Statistical comparisons were done according to the Fisher F test. Comparisons between the 3 media without TRH: ap =G0.05 CDE vs. N on days 3 or 5; b P < 0.01 CD vs. d P Q 0.1CD on day 7 vs. CD on day 5. Comparisons between TRH and control for each of the 3 CDEonday 7;cPG0.01 CDvs.Nonday7; media, e P < 0.01 TRH vs. control on days 3,s or 7.
0
CDE
H-;RH
0 aH-TRH
3
Treatment
CD
N
Media
Table 3 Release of PRL within 30 min at 37’C in the absence or presence of 3H-TRH (27 nM) by GH3/B6 cells grown in N medium, or CD medium, or CDE medium for 3,s or 7 days
“- :
$ $ a.
r “% P
2
TRH binding and PRL release after estrogen treatment
129
(B) Effect of duration of preculture in the 3 different media on responsiveness to a single dose of TRH (1) 3H-TRH binding (Table 2). When the binding of 27 nM 3H-TRH was tested as a function of the duration of preculture in the 3 media, it appeared that: (a) CD medium preculture for 3-7 days significantly reduced the number of TRH binding sites and (b) CDE medium preculture maintained the normal level of TRH binding sites after 5 days as well as 7 days of preculture. (2) PRL release (Table 3). The basal PRL release within 30 min remained approximately at the same level after 3-7 days of preculture in N medium. It was also maintained in CD medium for the first 5 days, but thereafter steeply decreased. Addition of E, into CD medium resulted in a significant augmentation of the basal PRL release already visible after 3 days. After 7 days in CDE medium, the absolute value of basal PRL release decreased but remained significantly higher than in CD medium. The % of TRH-induced increase of PRL release by cells grown in N medium presented a maximum after 5 days of preculture (Table 3). For cells grown in CD medium the % TRH effect first decreased after 3 days, then increased and reached after 7 days a value higher than that observed in N medium. Cells grown in CDE medium displayed already after 3 days of preculture an increase of the % TRH effect as compared to that observed in CD medium as well as N medium. This potentiating effect of CDE was maintained and even increased after 5 and 7 days. It appeared therefore that 7 days of preculture offered the optimal condition to study the effect of Ez on the number of TRH-binding sites (Table 2) and remained suitable to study the concomitant effect of TRH on PRL release (Table 3). (C) Dose effect of TRH on cells precultured for 7 days in the 3 different media (I) 3H-TRH binding. Fig. 2 shows that preculture for 7 days in CDE medium significantly increased the binding of 3H-TRH as compared to that observed for cells grown in CD medium as well as in N medium. The increase was statistically significant (F test, PS 0.01) within a large dose range, from 10W9M up to 10m6 M. Scatchard plot of the data (Fig. 3) yielded in any medium conditions curvilinear graphs, which is consistent with previous findings on GH3 cells grown in regular culture medium and is representative of either negative cooperativity or heterogeneity of binding sites (Gourdji, in press). This characteristic of TRH binding does not depend therefore on the presence or absence of E,, pretreatment. Assuming the existence of 2 independent classes of binding sites the slope of the asymptote drawn to the curve at its extremity near the y axes approximates affinity constant for highaffinity binding sites in each type of medium respectively: 1.3 X lOa M-r for N medium, 4.0 X lOa M-r for CD medium, 1.9 X lo8 M-’ for CDE medium. The comparison of the KA, in the 3 different media indicates that Ez does not alter the
N. Brunet, D. Gourdji, A. Tixier-Vidal
130
60 _
3H_TRH
[M)
Fig. 2. Binding of 3H-TRH as a function of 3H-TRH dose by GH3/B6 cells grown for 7 days in the 3 different media. Each point represents mean of triplicate +_SEM. Statistical comparisons between CDE and CD medium were done for each dose using the Fisher F test. ** P 4 0.01 CDE vs. N or CD. N medium (*), CD medium (A), CDE medium (m).
affinity of 3H-TFW-binding sites. The discrete increase in CD medium as compared to those observed in N and CDE media would require statistical analysis of several independent experiments to be taken into account. Estimations of the affinity constant for 3H-TRH of GH3 cells grown in control medium from 5 experiments, each performed in triplicate over a 7-year period in our laboratory, range from 1.3 to 5.6 X 10s M-’ (see Gourdji, 1979 and Fig. 3). The same variation most probably would occur in repeated experiments with E, enriched medium. The theoretical number of high-affinity binding sites per cell for each type of medium as indicated by the intercept of the asymptote on the x axis were 450 000, 225 000 and 530 000 for medium N, CD and CDE resp. This allowed to conclude that estradiol increases by about 50-60% the number of high-affinity binding sites without modifying their affinity.
TRH binding and PRL release after estrogen treatment
131
a _
&
c -
’ _L,
6
5 -
Kh, i 1.9xlO'td '\F
KAZ;1.5x10',d \
4-
\.
3-
' '\, 7. “+. -‘-;----____~~ >. -.:.\y ---.\. . ~--.A__
2l-
~,/ 10
20 BOUND
30 (fmoles
n
‘--._
40
50
60
Irs DNA)
Fig. 3. Scatchard plot of the 3H-TRH binding data of Fig. 2. Cells precultured in N medium (a), CD medium (b) and CDE medium (c).
As concerns the low-affinity binding sites the slope of the second asymptote drawn to the curve at its extremity near the y axis approximates the affinity constant for each type of medium resp. They were not clearly different: 3.8 X lo7 M-i for N medium, 2.6 X lo7 M-’ for CD medium and 1.5 X lo7 M-i for CDE medium resp. The theoretical number of total binding sites per cell, as indicated by the intercept of the second asymptote on the x axis was slightly increased in CDE medium. It appears therefore that estradiol acts mostly on the number of high-affinity binding sites.
N. Brunet, D. Gourd@, A. Tixier- Vidal
132
When the binding experiments were performed in presence of 10 mM HEPES there was again no significant modification of the values of KA, whatever the medium condition. Moreoever these values were similar to those observed in absence of HEPES. In contrast, the increase in number of high-affinity binding sites after estradiol treatment was totally masked in presence of HEPES (data not shown). (2) PRL release. Fig. 4 represents the % increase over basal of PRL release under TRH exposure at 37°C for 30 min as a function of TRH dose in the 3 culture media. The % increase was clearly enhanced in CDE medium precultured cells as compared to control cells as well as to CD medium precultured cells. Curiously enough for high TRH doses (1O-8-1O-7 M) the % increase was also enhanced in CD medium as compared to N medium, although less than in CDE medium. The halfmaximum of the TRH-induced increase was observed at 3 X I 0m9M TRH in control cells, 1 X lo-’ M TRH in CD medium and 1 X 10V9 M TRH in CDE medium. It appears therefore that Es improved by 1 order of magnitude the ED5,, or apparent
_1 10JU
10-p
10-T
3H_TRH
[M]
IO“
10-IU
IO--
1O-y
3H_TRH
10-7
[M]
Fig. 4. % increase of PRL release following exposure for 30 min at 37°C to increasing doses of 3H-TRH by GH3/B6 cells grown for 7 days in the 3 different media (same experiments as in Fig. 2). Each point represents mean of triplicate. N medium (a), CD medium (A), CDE medium (m). Fig. 5. % increase of PRL release following exposure for 30 min at 37°C in the presence of 10 mM HEPES to increasing doses of 3H-TRH by GH3/B6 cells grown for 7 days in the 3 different media. Each point represents the mean of quadruplicate. N medium (0). CD medium (A), CDE medium (m).
TRH binding and PRL release after estrogen treatment
133
K, of the short-term biological effect of TRH. This is consistent with an increase in number of binding sites. When the experiments were performed in presence of 10 mM HEPES (Fig. 5) there was again a clear enhancement of the TRH-induced % increase of PRL release in CDE medium precultured cells as compared to both control cells and CD medium precultured cells. The ED5,, of TRH effect was improved in CDE medium (1.5 X low9 M) as compared to N medium (6 X 10M9M). In CD medium, nevertheless, the presence of HEPES highly reduced the % increase of the TRH-induced stimulation of PRL release which increased in a very narrow dose range from 1.2 X 10m9 M to 5 X 1O-9 M. This does not permit estimation of EDsa. Thus the presence of HEPES although masking the effects of EZ on the increase in number of binding sites respected the potentiating effect of Ez pretreatment on the TRH-induced increase of PRL release.
DISCUSSION In the present experiments we used a homogeneous population of target cells, GH3/B6 rat PRL cells and culture medium supplemented with charcoal-dextran extracted serum (CD medium) in order to analyse at the cellular level the mechanisms by which E, regulates TRH responsiveness in PRL cells. The results permit us to draw the following conclusions. Preculture for 5-7 days in E2 (4 X lo-’ M) supplemented CD medium increases by 50%. the number of high affinity TRHbinding sites per cell without modifying their affinity and improved by one order of magnitude (1 X 10e9 M instead of 1 X lo-’ M) the ED,0 of the TRH-induced stimulation of PRL release. These effects of Ea on TRH responsiveness by a tumor-derived clonal rat prolactin cell line are in complete agreement with previous findings of De Lean et al. (1977a) obtained in vivo, in male and female rats and showing that Ez pretreatment increases the number of TRH-binding sites in pituitary homogenates without modifying their dissociation constant and enhances PRL release in response to a single dose of TRH. In spite of their tumoral origin, GH3/B6 cells therefore behave as normal pituitary cells. They offer over pituitary homogenates and in vivo experiments the advantage to localize at a single cell level the modifications induced by E, pretreatment. Indeed, we can correlate both 3H-TRH binding and PRL release to the cell number by measuring DNA in the same culture dishes. As in vivo, a several day exposure of the cells to Ea is necessary in order to reveal its effect on TRH binding and cell responsiveness. This indicates that the mechanisms by which Ez exerts these effects involve complex intracellular regulations. It is known that GH3 cells possess receptors for estrogens (Haug et al., 1978; Gershengorn et al., 1979). The mechanisms by which E, increases the number of binding sites are still entirely unknown. The fact that this effect is not accompanied by a modification of the apparent association constant indicates that E2 does not modify their proper-
134
N. Brunet, D. Gourdji, A. Tixier-Vidal
ties but only their number by altering their synthesis or degradation or by unmasking receptors incorporated into the plasma membrane but not exposed to ligand in normal conditions. The characteristics of the modulation of TRH-binding sites by E2 that we found in GH3/B6 cells are in agreement with those recently reported by Gershengorn et al. (1979) using parent cell lines, GH3 and GC, and a slightly different experimental schedule. In our conditions, cells were exposed to E2 without preculture in CD medium because previous observations showed us that the cell morphology was not restored in such conditions. Our cells were precultured in presence or absence of Es for longer periods than in Gershengorn’s experiments, in order to permit comparison with regular (N) culture medium. This shows that Ez alone is not able to increase the basal PRL release above normal levels after 7 days of treatment (Table 3) and suggests that other factors are lacking in CD medium (see discussion below). The potentiation of the TRH-induced stimulation of PRL release that we observed is consistent with an increased number of binding sites although the link between these two effects of estradiol is most probably not simple. Two aspects may be distinguished in this potentiating action: (1) an increase of the % stimulation for a given dose (260% instead of 50% for 3 X lo-’ M TRH) and (2) a shift toward low doses (one order of magnitude) of the EDa,, in absence of HEPES (Fig. 4). As compared to the 50% increase in number of TRH-binding sites, this strongly suggests the existence, beyond the TRH-binding step of a mechanism of amplification. The increase in intracellular PRL content following exposure to E, (Haug et al., 1976) represents only one aspect of this mechanism which for the most part escapes to the present investigations. The effects of the presence in the medium of 10 mM HEPES during TRH exposure were rather surprising. This completely masked the effect of E, pretreatment on the increase in number of high-affinity binding sites, but respected its potentiating action on the % of the TRH-induced stimulation of PRL release (Fig. 5). In fact, we used HEPES in order to improve the stability of the pH during the binding experiments. This unexpected effect of HEPES does not seem to be nevertheless related to the pH which was not strongly altered when experiments were performed in absence of HEPES, due most probably to the presence of serum in the incubation medium. A possible interpretation of the effect of HEPES in cells which were not routinely grown in presence of that buffer might derive from the fact that HEPES is an amine compound. It has been recently shown that such compounds inhibit the receptor-mediated uptake of lysosomal enzymes by fibroblasts (Sand0 et al., 1979) as well as the clustering of receptors following binding of ocz-macroglobulin or epidermal growth factor at the surface of fibroblasts (Maxfield et al., 1979). This effect has been related to an alteration of the lateral mobility of membrane components. Such an interpretation might explain in our experiments the inhibition by HEPES of the effect of Ez pretreatment on TRH-binding sites. This would reinforce the hypothesis of a membrane modification following
TRH binding and PRL release after estrogen treatment
135
pretreatment with an unmasking of TRH-binding sites. The persistence in those conditions of the potentiating effect of Es on the % of the TRH-induced stimulation of PRL release might be related only to the increase of PRL synthesis following a 7-day exposure to E,. The charcoal-dextran treatment of the horse serum which participates in the composition of culture medium was imposed by the necessity to eliminate estrogens which they contain at a non-negligible level (Table 1). Nevertheless such treatment obviously extracts other serum components such as other sex steroids, glucocorticoids, thyroid hormones and small unidentified peptides. The presence or absence of these various compounds may interact in a complex manner on the proper effects of EZ. Indeed it has been shown that TRH-binding sites in GH3 cells are regulated positively by glucocorticoids (Tashjian et al., 1977) and negatively by Ta (Perrone and Hinkle, 1978). This may explain the fact that the differences in number of binding sites and EDso which were observed between N medium and CDE medium do not appear directly related to their respective Ez concentrations. In contrast the comparison between CD medium and CDE medium undoubtedly reveals the role of E2 in increasing the number of TRH-binding sites and improving the EDsc of the TRH stimulation of PRL release. In order to further analyse the mechanism of these effects an ideal way to eliminate such unknown serum factors consists in the use of serum-free chemically defined medium (Hayashi and Sato, 1976; Wu et al., 1978). Such experiments are now in progress in our laboratory (Brunet et al., 1979). E,
ACKNOWLEDGEMENTS We are indebted to Dr. D. Garnier (Biologie de la Reproduction, Rennes) for performing the estradiol radioimmunoassays and to Professor Ph. Leloup (Museum d’Histoire Naturelle, Paris) for performing the Ts radioimmunoassays. We are grateful to Mrs. M.F. Moreau and D. Grouselle for their excellent technical assistance. This work was supported by grants from INSERM (ATP Nos. 24-75-47 and 76-4-496).
REFERENCES Bournaud, F., Gourdji, D., Mongongu, S., and Tixier-Vidal, A. (1977) Neuroendocrinology 24, 183-194. Brunet, N., Gourdji, D., Moreau, M.F., Grouselle, D., Bournaud, F., and Tixier-Vidal, A. (1977) Ann. Biol. Anim. Biochim. Biophys. 17,413-424. Brunet, N., Rizzino, A., Tixier-Vidal, A., and Gourdji, D. (1979) In: First International Congress on Hormones and Cancer (Rome, Italy) abstr., in press. Dannies, P.S., Gautvik, K.M., and Tashjian Jr., A.H. (1976) Endocrinology 98, 1147-1158. Danmes, P.S., Yen, P.M., and Tashjian Jr., A.H. (1977) Endocrinology 101, 1151-1156 De Lean, A., Ferland, L., Drouin, J., Kelly, P.A., and Labrie, F. (1977a) Endocrinology 100, 1496-1504.
136
N. Brunet, D. Gourdji, A. Tixier- Vidal
De Leon, A., Garon, M., Kelly, P.A., and Labrie, F. (1977b) Endocrinology 100, 1505-1510. Faivre-Bauman, A., Gourdji, D., Grouselle, D., and Tixier-Vidal, A. (1975) Biochem. Biophys. Res. Commun. 67,50-57. Gershengorn, M.C., Marcus-Samuels, B.E., and Geras, E. (1979) Endocrinology 105, 171-176. Gourdji, D. (1979) In: Biochemical Endocrinology: Synthesis and Release of Adenohypophyseal Hormones, Cellular and Molecular Mechanisms, Eds.: McKerns and M. Jutiszz (Plenum, New York) pp. 463-493. Gourdji, D., Kerdelhue, B., and Tixier-Vidal, A. (1972) C.R. Acad. Sci. Ser. D 274, 437-440. Gourdji, D., Tixier-Vidal, A., Morin, A., Pradelles, P., Morgat, J.L., Fromageot, P., and Kerdelhue, B. (1973) Exp. Cell. Res. 82, 39-46. Gourdji, D., Bourneau, F., Faivre-Bauman, A., and Tixier-Vidal, A. (1976) Colloque de Synthese (1976) Serie Action Thematique, Rapport No. 7: Actions Thematiques 22, 35, Neuroendocrinologie (INSERM, Paris) pp. 351-359. Haug, E., and Gautvik, K.M. (1976) Endocrinology 99, 1482-1489. Haug, E., Tjernshaugen, H., and Gautvik, K.M. (1976) J. Cell. Physiol. 91, 15-29. Haug, E., Naess, O., and Gautvik, K.M. (1978) Mol. Cell. Endocrinol. 12, 81-95. Hayashi, I., and Sato, G.H. (1976) Nature (London) 259, 1322134. Hinkle, P.M., and Tashjian Jr., A.H. (1973) J. Biol. Chem. 248, 6180-6186. Karsten, U., and Wollenberger, A. (1972) Anal. Biochem. 46, 135-148. Kirkland, W.L., Sorrentino, J.M., and Sirbasku, D.A. (1976) J. Nat]. Cancer Inst. 56, 11591164. Lowry, O.H., Rosebrough, N.F., Farr, A.L., and Randall, R.J. (1951) J. Biol.Chem. 193, 265275. Maxfield, F.R., Willmgham, M.C., Davies, P.J.A., and Pastan, I. (1979) Nature (London) 277, 661-663. Morin, A., Tixier-Vidal, A., Gourdji, D., Kerdelhue, B., and Grouselle, D. (1975) Mol. Cell Endocrinol. 3, 351-373. Padmanabhan, V., and Convey, E.M. (1979) Mol. Cell. Endocrinol. 14, 103-112. Perrone, M.H., and Hinkle, P.M. (1978) J. Biol. Chem. 253, 5168-5173. PradelIes, P., Morgat, J.L., Fromageot, P., Oliver, C., Jacquet, P., Gourdji, D., and Tixier-Vidal, A. (1972) FEBS Lett. 22, 19-22. Quadri, SK., Oyama, T., and Spies, H.G. (1979) Endocrinology 104, 1649-1655. Rakoff, J.S., Siler, T.M., Sinha, Y.N., and Yen, S.S.C. (1973) J. Clin. Endocrinol. Metab. 37, 641-644. Raymond, V., Beaulieu, M., Labrie, F., and Boissier, J. (1978) Science 200, 1173-1175. Sando, G.N., Titus-Dillon, P., Hall, C.W., and Neufeld, E.F. (1979) Exp. Cell. Res. 119, 359364. Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 53,660-672. Tashjian Jr., A.H. and Hoyt Jr., R.F. (1972) In: Molecular Genetics and Developmental Biology, Ed.: M. Sussman (Prentice-Hall, Englewood cliffs, NJ) pp. 353 -387. Tashjian Jr., A.H., Yasumura, Y., Levine, L., Sato, G.H., and Parker, M.L. (1968) Endocrinoology 82, 342-352. Tashjian Jr., A.H., Bancroft, F.C., and Levine, L. (1970) J. Cell Biol. 47, 61-70. Tashjian Jr., A.H., Barowsky, N.J., and Jensen, D.K. (1971) Biochem. Biophys. Res. Commun. 43,516-523. Tashjian Jr., A.H., Osborne, R., Maina, D., and Knaian, A. (1977) Biochem. Biophys. Res. Commun. 79,333-340. Tixier-Vidal, A., Gourdji, D., PradeIIes, P., Morgat, J.L., Fromageot, P., and Kerdelhue, B. (1975) In: Hypothalamic Hormones, Vol. 6, Eds.: M. Motta, P.G. Grosignani and L. Martini (Academic Press, New York) pp. 89-107. Tixier-Vidal, A., Brunet, N., and Gourdji, D. (1978) In: Progress in Prolactin Physiology and Pathology, Eds.: C. Robyn and M. Harter (Elsevier/North-Holland, Amsterdam) pp. 29-43. Wu, R., and Sato, G.H. (1978) J. Toxicol. Environ. Health 4,427-448.
EFFECT OF 17fl-ESTRADIOL GH3/B6 RAT PROLACTIN
ON THYROLIBERIN
123
RESPONSIVENESS IN
CELLS
N. BRUNET, D. GOURDJI and A. TIXIER-VIDAL Groupe de Neuroendocrinologie 75231 Paris Cedex 05 (France)
Cellulaire, Coll;ge de France, 1 I, Place
Marcelin'Berthelot,
Received 15 October 1979; accepted 7 February 1980
GH3/B6 rat prolactin cells were used to analyse at the cellular level the mechanisms by which 17p-estradiol (Ea) regulates TRH responsiveness of prolactin cells. Before experiments, cells were grown for up to 7 days in 3 different media: normal medium (N) containing 15% horse serum and 2.5% fetal calf serum, CD medium prepared with charcoal-dextran extracted serum and CDE medium supplemented with 4 X 10 -8 M Ea. The binding of 3H-TRH (30 min at 37°C) and the TRH-induced percent increase of prolactin release as a function of TRH doses were compared in the 3 conditions. Preculture in Ea enriched medium increased by 50% the number of TRH high-affinity binding sites without modifying their affinity, increased by up to 3 times the percent of the TRH-induced stimulation of prolactin release and improved by one order of magnitude the EDse of the TRH effect on prolactin release. The presence of HEPES (10 mM) during TRH challenge masked the effect of Ea on the increase in number of binding sites but respected its potentiating effect on prolactin release Keywords:
17pestradiol; thyroliberin; prolactin release; binding; cell culture.
Estrogens and thyroliberin (TRH) are, separately, well known potent stimulators of prolactin (PRL) secretion in vivo as well as in vitro. In addition, an interaction between estrogens and TRH has been observed in vivo. A facilitating effect of 17P-estradiol (Es) pretreatment on the TRH-induced PRL release has been described in the human female (Rakoff et al., 1973) and more recently in female rhesus monkeys (Quadri et al., 1979). In female rats, E, pretreatment has been shown to enhance both the number of TRH-binding sites in pituitary homogenates and the stimulation of PRL release in response to a TRH i.v. injection (De Lean et al., 1977 a,b). These experiments suggested that estrogens exert their action by modulating the number of TRH receptors in pituitary target cells. Further analysis at tlfe cellular level was nevertheless limited by the fact that in these experiments at least two pituitary cell types - thyrotropic cells and PRL cells - were modified in both their secretory activity and mitotic activity. Experiments with primary cultures of dispersed anterior pituitary cells although allowing to localize at the pituitary level a site of this interaction (Raymond et al., 1978; Padmanabhan and
124
N, Brunet, D. Gourdji, A. Tixier- Vidal
Convey, 1979) did not overcome this drawback because of their cellular heterogeneity. Clonal rat PRL cell lines which offer a homogeneous population of target cells might represent an appropriate model to study such an interaction. Indeed, GH3 rat pituitary cell lines (Tashjian et al., 1968) which secrete PRL and, to a lesser extent growth hormone (Tashjian et al., 1970) have been shown to respond to TRH (Tashjian et al., 1971; Gourdji et al., 1972) and to E2 (Tashjian and Hoyt, 1972). TRH exerts a biphasic stimulating action on PRL secretion: an early effect on the release of a stored PRL pool, and a delayed effect on the synthesis of PRL (Morin et al., 1975; Dannies et al., 1976). 3H-THR binds to intact GH3 cells. Equilibrium binding is reached after 15-30 min at 37” (Gourdji et al., 1973; Hinkle and Tashjian, 1973). This binding is also dose-dependent and 2 apparent dissociation constants have been found: K - 4 X lo-’ M, in the range of the physiological doses (0.05-50 nM TRH) > andok- D2= 6 X lo-* M for higher doses, up to 850 nM (Tixier-Vidal et al., 1975; Gourdji et al., 1976; Gourdji, in press). GH3 cells are also responsive to Ea. After several days in Ez supplemented medium, PRL production by GH3 cells was enhanced (Tashjian and Hoyt, 1972). Although the same group reported later that these cells have lost this property (Dannies et al., 1977), other authors (Haug and Gautvik, 1976) using the same cell line have found that Ez (IO-’ M) increases prolactin production following treatment for 4 days up to 10 days and this effect was additive to that of TRH (3 X lo-’ M) (Haug and Gautvik, 1976). The study of the effect of E, in cells which are routinely grown in serum-supplemented medium is hampered by the presence of non-negligible amounts of this steroid in the serum. Using GH3/B6 cells, a subclone of the GH3 line, which were precultured for one week in medium supplemented with charcoal dextran extracted serum (CD medium), we have previously shown that Ea (4 X 10m8 M) is able to reestablish the normal cell morphology, to maintain the number of TRH-binding sites and to potentiate the TRH-induced stimulation of PRL release (Brunet et al., 1977). In addition PRL production was increased over normal level after one-week culture in CD medium supplemented with Ez (Tixier-Vidal et al., 1978). We report here the results of a study which has been performed using the same experimental model in order to analyse the mechanism by which E2 interacts on TRH binding and TRH responsiveness by GH3/B6 cells. It is shown that Ea increases the number of TRH-binding sites without modifying their affinity for TRH, increases by up to 3 times the percent of the TRH-induced stimulation of PRL release and improves by one order of magnitude the ED,c of the TRH effect .on PRL release. After completing this manuscript we had access to the paper of Gershengorn et al. (1979) who also found that in GH3 cells estrogens increase the number of TRH receptors without modifying their affinity. These authors nevertheless did not correlate the binding with the dose effect of TRH on rapid PRL release.
125
TRH binding and PRL release after estrogen treatment
MATERIALS
AND METHODS
Cell culture
Cells were grown in control conditions (N medium) as previously described et al., 1973) in Ham’s FIO medium enriched with 15% heat inactivated horse serum (Eurobio, Paris), 2.5% fetal calf serum (Eurobio, Paris) and with antibiotics (penicillin 50 mu/ml plus streptomycin 50 pg/ml, or gentiamicin 40 ,ug/ml). A single batch of horse serum as well as of fetal calf serum was used for all of the experiments. CD medium was prepared using horse serum from the same batch previously treated with charcoal dextran according to Kirkland et al. (1976). CDE medium was freshly prepared by adding Es after serial dilutions in FIO medium starting from an ethanol solution. The final Es concentration in the 3 media (N, CD, CDE) is shown in Table 1. (Gourdji
Experimental schedules GH3/B6 cells previously grown in N medium where harvested by mild trypsinisation (0.25%) and plated in tissue-culture petri dishes (Falcon, 3.5 cm diameter, 2 or 2.5 X lo5 cells/2 ml medium) in the 3 media, N, CD, and CDE resp. They were incubated at 37°C in a humidified atmosphere of 5% CO, in air. The cells were grown in those conditions for 3, 5 or 7 days, the medium being renewed every 2 or 3 days. At the end of each of these different periods, the cells were assayed for TRH responsiveness. Binding experiments 24 h before experiments the culture medium was renewed and in some experiments supplemented with 10 mM HEPES (N-2-hydroxyethylpiperazine-N’-2 ethanesulfonic acid). On the day of experiment, the last culture medium was discarded and the cells were washed with 1 ml of warmed Ham’s FlO. They were then covered with 1 ml of appropriate medium (N, CD or CDE), with or without 10 mM HEPES, supplemented with various doses of 3H-TRH (60 Ci/mmole) (prepared by
Table 1 17p-Estradiol (Ez) and triiodothyronine Hormones
E2 T3
(T3) concentrations
in the 3 culture media
Media N
CD
CDE
0.04 nM 0.60 nM
0.01 nM 0.25 nM
40 nM 0.25 nM
Hormones were measured by specific radioimmunoassays in duplicate, in heat inactivated horse serum before and after treatment with charcoal-dextran, and in fetal calf serum.
126
N. Brunet, D. Gourdji, A. Tixier- Vidal
DURATION
OF
PRECULTURE (days)
Fig. 1. Cell proliferation as a function of time in culture in each of the 3 media. GH3/B6 cells were seeded at 200 000 cells/dish. At increasing time intervals, for each medium, 3 dishes were used for cell counting. Statistical comparisons were done using the Fisher F test. N medium (e), CD medium (A), CDE medium (a). Each point represents mean of triplicate k SEM. * P < 0.05 CDE vs. CD; ** P < 0.01 CD or CDE vs. N.
Table 2 Binding of 3H-TRH (27 nM, 30 min, 37°C) by GH3/B6 cells grown in N medium or CD medium or CDE medium for 3, 5 or 7 days Media
3H-TRH (fmoles/pg cell DNA) Preculture duration
N
CD CD!!~
~_
3 days
5 days
7 days
24.2 f 0.4 15.8 f 0.7 a 18.75 + 0.5 b,c
22.0 f I .l 16.1 f 0.8 a 22.4 f 0.4 ’ ~~___~
18.3 * 0.7 11.5 *0.4a 18.2 * 0.9 c ____
~~_
Results are expressed as fmoles of bound 3H-TRH per pg cell DNA. Each value represents mean of quadruplicate dishes + SEM. Statistical comparisons were done according to the Fisher F test. a P < 0.01 CD vs. N; b P < 0.01 CDE vs. N; c P < 0.01 CDE vs. CD.
TRH binding and PRL release after estrogen treatment
127
Levine-Pinto, CEA, Saclay, according to Pradelles et al., 1972) and incubated 30 min at 37’C in a controlled atmosphere. After cooling the dishes at 0-4°C the media were collected for prolactin determination. The cells were washed 4 times with ice-cold Ham’s FIO. They were then scraped in 1 ml of PBS. One aliquot of the cell sonicate (0.5 ml) was used for radioactivity determination and another aliquot for proteins and DNA determinations. 3H-TRH uptake was determined by the measure of the radioactivity bound to the cells as previously described (FaivreBauman et al., 1975). The non-compatible binding was estimated on parallel dishes and never exceeded 10%. Triplicate or quadruplicate dishes were used for each type of medium and each dose of 3H-TRH. Results were expressed as fmoles of TRH bound per gg DNA. Protein measurement Cell proteins were measured by Folin’s reagent according to Lowry et al. (19.51) using bovine serum albumin as standard. DNA measurement DNA was measured by the method of Karsten and Wollenberger modified by Bournaud et al. (1977).
(1972) slightly
Prolactin determination PRL medium content was measured by radioimmunoassay, using the rat PRL kit kindly provided by the NIAMDD rat hormone distribution program. Calculations Statistical significances were measured using the Fisher test. Binding data were analysed according to Scatchard (1949).
RESULTS (A) Celi proliferation Cell proliferation was estimated by cell counts, using a hemocytometer, in the 3 culture media as a function of time in culture (3-5-7 days). As shown in Fig. 1, culture in CD medium, supplemented or not with EZ, significantly stimulated cell proliferation. As compared to the initial number of cells (200 000), after 7 days of culture, the number of cells/dish was increased by S times in medium Ninstead of 7-8 times in medium CD or CDE. In contrast, there was no significant modification of the DNA content/cell (23-29 pg DNA/lo6 cells). The protein content per cell was not modified (37-41 pg/lO’ cells) except in cells grown in CD medium for 7 days, where it was significantly decreased (34 pg/105 cells) (F test: P< 0.01 CD vs. CDE). In the following experiments, the binding of 3H-TRH was therefore expressed per pg DNA, which represents a valuable index of the cell number.
33.1 f 3.5 a
17.1 f 3 29.9 f 1.7 e
19.1 a 4 41.4 f 4 e 75
115
35.5 + 9.7 a
15.9 f: 2.9 32.5 + 6.3 e
16.9 * 2 38.5 + 3.7 e
PRL ng/ng cell DNA
PRL ng/ng cell DNA % increase
5 days
3 days
100
128
% increase
12.1 f 0.9
7.4 f 0.8 b*c,d 16.6 * 1.7 e
15.3 f 2.4 28.1 f 2.7 e
PRL ng/Mg cell DNA
7 days
124
83
% increase
aH-TRH 65.1 f 4 e 197 89.8 i 7.8 e 253 37.6 i: 3.3 e 211 Results are expressed as ng PRL per ng cell DNA. Each value represents the mean of quadruplicate dishes +_S.E.M. Statistical comparisons were done according to the Fisher F test. Comparisons between the 3 media without TRH: ap =G0.05 CDE vs. N on days 3 or 5; b P < 0.01 CD vs. d P Q 0.1CD on day 7 vs. CD on day 5. Comparisons between TRH and control for each of the 3 CDEonday 7;cPG0.01 CDvs.Nonday7; media, e P < 0.01 TRH vs. control on days 3,s or 7.
0
CDE
H-;RH
0 aH-TRH
3
Treatment
CD
N
Media
Table 3 Release of PRL within 30 min at 37’C in the absence or presence of 3H-TRH (27 nM) by GH3/B6 cells grown in N medium, or CD medium, or CDE medium for 3,s or 7 days
“- :
$ $ a.
r “% P
2
TRH binding and PRL release after estrogen treatment
129
(B) Effect of duration of preculture in the 3 different media on responsiveness to a single dose of TRH (1) 3H-TRH binding (Table 2). When the binding of 27 nM 3H-TRH was tested as a function of the duration of preculture in the 3 media, it appeared that: (a) CD medium preculture for 3-7 days significantly reduced the number of TRH binding sites and (b) CDE medium preculture maintained the normal level of TRH binding sites after 5 days as well as 7 days of preculture. (2) PRL release (Table 3). The basal PRL release within 30 min remained approximately at the same level after 3-7 days of preculture in N medium. It was also maintained in CD medium for the first 5 days, but thereafter steeply decreased. Addition of E, into CD medium resulted in a significant augmentation of the basal PRL release already visible after 3 days. After 7 days in CDE medium, the absolute value of basal PRL release decreased but remained significantly higher than in CD medium. The % of TRH-induced increase of PRL release by cells grown in N medium presented a maximum after 5 days of preculture (Table 3). For cells grown in CD medium the % TRH effect first decreased after 3 days, then increased and reached after 7 days a value higher than that observed in N medium. Cells grown in CDE medium displayed already after 3 days of preculture an increase of the % TRH effect as compared to that observed in CD medium as well as N medium. This potentiating effect of CDE was maintained and even increased after 5 and 7 days. It appeared therefore that 7 days of preculture offered the optimal condition to study the effect of Ez on the number of TRH-binding sites (Table 2) and remained suitable to study the concomitant effect of TRH on PRL release (Table 3). (C) Dose effect of TRH on cells precultured for 7 days in the 3 different media (I) 3H-TRH binding. Fig. 2 shows that preculture for 7 days in CDE medium significantly increased the binding of 3H-TRH as compared to that observed for cells grown in CD medium as well as in N medium. The increase was statistically significant (F test, PS 0.01) within a large dose range, from 10W9M up to 10m6 M. Scatchard plot of the data (Fig. 3) yielded in any medium conditions curvilinear graphs, which is consistent with previous findings on GH3 cells grown in regular culture medium and is representative of either negative cooperativity or heterogeneity of binding sites (Gourdji, in press). This characteristic of TRH binding does not depend therefore on the presence or absence of E,, pretreatment. Assuming the existence of 2 independent classes of binding sites the slope of the asymptote drawn to the curve at its extremity near the y axes approximates affinity constant for highaffinity binding sites in each type of medium respectively: 1.3 X lOa M-r for N medium, 4.0 X lOa M-r for CD medium, 1.9 X lo8 M-’ for CDE medium. The comparison of the KA, in the 3 different media indicates that Ez does not alter the
N. Brunet, D. Gourdji, A. Tixier-Vidal
130
60 _
3H_TRH
[M)
Fig. 2. Binding of 3H-TRH as a function of 3H-TRH dose by GH3/B6 cells grown for 7 days in the 3 different media. Each point represents mean of triplicate +_SEM. Statistical comparisons between CDE and CD medium were done for each dose using the Fisher F test. ** P 4 0.01 CDE vs. N or CD. N medium (*), CD medium (A), CDE medium (m).
affinity of 3H-TFW-binding sites. The discrete increase in CD medium as compared to those observed in N and CDE media would require statistical analysis of several independent experiments to be taken into account. Estimations of the affinity constant for 3H-TRH of GH3 cells grown in control medium from 5 experiments, each performed in triplicate over a 7-year period in our laboratory, range from 1.3 to 5.6 X 10s M-’ (see Gourdji, 1979 and Fig. 3). The same variation most probably would occur in repeated experiments with E, enriched medium. The theoretical number of high-affinity binding sites per cell for each type of medium as indicated by the intercept of the asymptote on the x axis were 450 000, 225 000 and 530 000 for medium N, CD and CDE resp. This allowed to conclude that estradiol increases by about 50-60% the number of high-affinity binding sites without modifying their affinity.
TRH binding and PRL release after estrogen treatment
131
a _
&
c -
’ _L,
6
5 -
Kh, i 1.9xlO'td '\F
KAZ;1.5x10',d \
4-
\.
3-
' '\, 7. “+. -‘-;----____~~ >. -.:.\y ---.\. . ~--.A__
2l-
~,/ 10
20 BOUND
30 (fmoles
n
‘--._
40
50
60
Irs DNA)
Fig. 3. Scatchard plot of the 3H-TRH binding data of Fig. 2. Cells precultured in N medium (a), CD medium (b) and CDE medium (c).
As concerns the low-affinity binding sites the slope of the second asymptote drawn to the curve at its extremity near the y axis approximates the affinity constant for each type of medium resp. They were not clearly different: 3.8 X lo7 M-i for N medium, 2.6 X lo7 M-’ for CD medium and 1.5 X lo7 M-i for CDE medium resp. The theoretical number of total binding sites per cell, as indicated by the intercept of the second asymptote on the x axis was slightly increased in CDE medium. It appears therefore that estradiol acts mostly on the number of high-affinity binding sites.
N. Brunet, D. Gourd@, A. Tixier- Vidal
132
When the binding experiments were performed in presence of 10 mM HEPES there was again no significant modification of the values of KA, whatever the medium condition. Moreoever these values were similar to those observed in absence of HEPES. In contrast, the increase in number of high-affinity binding sites after estradiol treatment was totally masked in presence of HEPES (data not shown). (2) PRL release. Fig. 4 represents the % increase over basal of PRL release under TRH exposure at 37°C for 30 min as a function of TRH dose in the 3 culture media. The % increase was clearly enhanced in CDE medium precultured cells as compared to control cells as well as to CD medium precultured cells. Curiously enough for high TRH doses (1O-8-1O-7 M) the % increase was also enhanced in CD medium as compared to N medium, although less than in CDE medium. The halfmaximum of the TRH-induced increase was observed at 3 X I 0m9M TRH in control cells, 1 X lo-’ M TRH in CD medium and 1 X 10V9 M TRH in CDE medium. It appears therefore that Es improved by 1 order of magnitude the ED5,, or apparent
_1 10JU
10-p
10-T
3H_TRH
[M]
IO“
10-IU
IO--
1O-y
3H_TRH
10-7
[M]
Fig. 4. % increase of PRL release following exposure for 30 min at 37°C to increasing doses of 3H-TRH by GH3/B6 cells grown for 7 days in the 3 different media (same experiments as in Fig. 2). Each point represents mean of triplicate. N medium (a), CD medium (A), CDE medium (m). Fig. 5. % increase of PRL release following exposure for 30 min at 37°C in the presence of 10 mM HEPES to increasing doses of 3H-TRH by GH3/B6 cells grown for 7 days in the 3 different media. Each point represents the mean of quadruplicate. N medium (0). CD medium (A), CDE medium (m).
TRH binding and PRL release after estrogen treatment
133
K, of the short-term biological effect of TRH. This is consistent with an increase in number of binding sites. When the experiments were performed in presence of 10 mM HEPES (Fig. 5) there was again a clear enhancement of the TRH-induced % increase of PRL release in CDE medium precultured cells as compared to both control cells and CD medium precultured cells. The ED5,, of TRH effect was improved in CDE medium (1.5 X low9 M) as compared to N medium (6 X 10M9M). In CD medium, nevertheless, the presence of HEPES highly reduced the % increase of the TRH-induced stimulation of PRL release which increased in a very narrow dose range from 1.2 X 10m9 M to 5 X 1O-9 M. This does not permit estimation of EDsa. Thus the presence of HEPES although masking the effects of EZ on the increase in number of binding sites respected the potentiating effect of Ez pretreatment on the TRH-induced increase of PRL release.
DISCUSSION In the present experiments we used a homogeneous population of target cells, GH3/B6 rat PRL cells and culture medium supplemented with charcoal-dextran extracted serum (CD medium) in order to analyse at the cellular level the mechanisms by which E, regulates TRH responsiveness in PRL cells. The results permit us to draw the following conclusions. Preculture for 5-7 days in E2 (4 X lo-’ M) supplemented CD medium increases by 50%. the number of high affinity TRHbinding sites per cell without modifying their affinity and improved by one order of magnitude (1 X 10e9 M instead of 1 X lo-’ M) the ED,0 of the TRH-induced stimulation of PRL release. These effects of Ea on TRH responsiveness by a tumor-derived clonal rat prolactin cell line are in complete agreement with previous findings of De Lean et al. (1977a) obtained in vivo, in male and female rats and showing that Ez pretreatment increases the number of TRH-binding sites in pituitary homogenates without modifying their dissociation constant and enhances PRL release in response to a single dose of TRH. In spite of their tumoral origin, GH3/B6 cells therefore behave as normal pituitary cells. They offer over pituitary homogenates and in vivo experiments the advantage to localize at a single cell level the modifications induced by E, pretreatment. Indeed, we can correlate both 3H-TRH binding and PRL release to the cell number by measuring DNA in the same culture dishes. As in vivo, a several day exposure of the cells to Ea is necessary in order to reveal its effect on TRH binding and cell responsiveness. This indicates that the mechanisms by which Ez exerts these effects involve complex intracellular regulations. It is known that GH3 cells possess receptors for estrogens (Haug et al., 1978; Gershengorn et al., 1979). The mechanisms by which E, increases the number of binding sites are still entirely unknown. The fact that this effect is not accompanied by a modification of the apparent association constant indicates that E2 does not modify their proper-
134
N. Brunet, D. Gourdji, A. Tixier-Vidal
ties but only their number by altering their synthesis or degradation or by unmasking receptors incorporated into the plasma membrane but not exposed to ligand in normal conditions. The characteristics of the modulation of TRH-binding sites by E2 that we found in GH3/B6 cells are in agreement with those recently reported by Gershengorn et al. (1979) using parent cell lines, GH3 and GC, and a slightly different experimental schedule. In our conditions, cells were exposed to E2 without preculture in CD medium because previous observations showed us that the cell morphology was not restored in such conditions. Our cells were precultured in presence or absence of Es for longer periods than in Gershengorn’s experiments, in order to permit comparison with regular (N) culture medium. This shows that Ez alone is not able to increase the basal PRL release above normal levels after 7 days of treatment (Table 3) and suggests that other factors are lacking in CD medium (see discussion below). The potentiation of the TRH-induced stimulation of PRL release that we observed is consistent with an increased number of binding sites although the link between these two effects of estradiol is most probably not simple. Two aspects may be distinguished in this potentiating action: (1) an increase of the % stimulation for a given dose (260% instead of 50% for 3 X lo-’ M TRH) and (2) a shift toward low doses (one order of magnitude) of the EDa,, in absence of HEPES (Fig. 4). As compared to the 50% increase in number of TRH-binding sites, this strongly suggests the existence, beyond the TRH-binding step of a mechanism of amplification. The increase in intracellular PRL content following exposure to E, (Haug et al., 1976) represents only one aspect of this mechanism which for the most part escapes to the present investigations. The effects of the presence in the medium of 10 mM HEPES during TRH exposure were rather surprising. This completely masked the effect of E, pretreatment on the increase in number of high-affinity binding sites, but respected its potentiating action on the % of the TRH-induced stimulation of PRL release (Fig. 5). In fact, we used HEPES in order to improve the stability of the pH during the binding experiments. This unexpected effect of HEPES does not seem to be nevertheless related to the pH which was not strongly altered when experiments were performed in absence of HEPES, due most probably to the presence of serum in the incubation medium. A possible interpretation of the effect of HEPES in cells which were not routinely grown in presence of that buffer might derive from the fact that HEPES is an amine compound. It has been recently shown that such compounds inhibit the receptor-mediated uptake of lysosomal enzymes by fibroblasts (Sand0 et al., 1979) as well as the clustering of receptors following binding of ocz-macroglobulin or epidermal growth factor at the surface of fibroblasts (Maxfield et al., 1979). This effect has been related to an alteration of the lateral mobility of membrane components. Such an interpretation might explain in our experiments the inhibition by HEPES of the effect of Ez pretreatment on TRH-binding sites. This would reinforce the hypothesis of a membrane modification following
TRH binding and PRL release after estrogen treatment
135
pretreatment with an unmasking of TRH-binding sites. The persistence in those conditions of the potentiating effect of Es on the % of the TRH-induced stimulation of PRL release might be related only to the increase of PRL synthesis following a 7-day exposure to E,. The charcoal-dextran treatment of the horse serum which participates in the composition of culture medium was imposed by the necessity to eliminate estrogens which they contain at a non-negligible level (Table 1). Nevertheless such treatment obviously extracts other serum components such as other sex steroids, glucocorticoids, thyroid hormones and small unidentified peptides. The presence or absence of these various compounds may interact in a complex manner on the proper effects of EZ. Indeed it has been shown that TRH-binding sites in GH3 cells are regulated positively by glucocorticoids (Tashjian et al., 1977) and negatively by Ta (Perrone and Hinkle, 1978). This may explain the fact that the differences in number of binding sites and EDso which were observed between N medium and CDE medium do not appear directly related to their respective Ez concentrations. In contrast the comparison between CD medium and CDE medium undoubtedly reveals the role of E2 in increasing the number of TRH-binding sites and improving the EDsc of the TRH stimulation of PRL release. In order to further analyse the mechanism of these effects an ideal way to eliminate such unknown serum factors consists in the use of serum-free chemically defined medium (Hayashi and Sato, 1976; Wu et al., 1978). Such experiments are now in progress in our laboratory (Brunet et al., 1979). E,
ACKNOWLEDGEMENTS We are indebted to Dr. D. Garnier (Biologie de la Reproduction, Rennes) for performing the estradiol radioimmunoassays and to Professor Ph. Leloup (Museum d’Histoire Naturelle, Paris) for performing the Ts radioimmunoassays. We are grateful to Mrs. M.F. Moreau and D. Grouselle for their excellent technical assistance. This work was supported by grants from INSERM (ATP Nos. 24-75-47 and 76-4-496).
REFERENCES Bournaud, F., Gourdji, D., Mongongu, S., and Tixier-Vidal, A. (1977) Neuroendocrinology 24, 183-194. Brunet, N., Gourdji, D., Moreau, M.F., Grouselle, D., Bournaud, F., and Tixier-Vidal, A. (1977) Ann. Biol. Anim. Biochim. Biophys. 17,413-424. Brunet, N., Rizzino, A., Tixier-Vidal, A., and Gourdji, D. (1979) In: First International Congress on Hormones and Cancer (Rome, Italy) abstr., in press. Dannies, P.S., Gautvik, K.M., and Tashjian Jr., A.H. (1976) Endocrinology 98, 1147-1158. Danmes, P.S., Yen, P.M., and Tashjian Jr., A.H. (1977) Endocrinology 101, 1151-1156 De Lean, A., Ferland, L., Drouin, J., Kelly, P.A., and Labrie, F. (1977a) Endocrinology 100, 1496-1504.
136
N. Brunet, D. Gourdji, A. Tixier- Vidal
De Leon, A., Garon, M., Kelly, P.A., and Labrie, F. (1977b) Endocrinology 100, 1505-1510. Faivre-Bauman, A., Gourdji, D., Grouselle, D., and Tixier-Vidal, A. (1975) Biochem. Biophys. Res. Commun. 67,50-57. Gershengorn, M.C., Marcus-Samuels, B.E., and Geras, E. (1979) Endocrinology 105, 171-176. Gourdji, D. (1979) In: Biochemical Endocrinology: Synthesis and Release of Adenohypophyseal Hormones, Cellular and Molecular Mechanisms, Eds.: McKerns and M. Jutiszz (Plenum, New York) pp. 463-493. Gourdji, D., Kerdelhue, B., and Tixier-Vidal, A. (1972) C.R. Acad. Sci. Ser. D 274, 437-440. Gourdji, D., Tixier-Vidal, A., Morin, A., Pradelles, P., Morgat, J.L., Fromageot, P., and Kerdelhue, B. (1973) Exp. Cell. Res. 82, 39-46. Gourdji, D., Bourneau, F., Faivre-Bauman, A., and Tixier-Vidal, A. (1976) Colloque de Synthese (1976) Serie Action Thematique, Rapport No. 7: Actions Thematiques 22, 35, Neuroendocrinologie (INSERM, Paris) pp. 351-359. Haug, E., and Gautvik, K.M. (1976) Endocrinology 99, 1482-1489. Haug, E., Tjernshaugen, H., and Gautvik, K.M. (1976) J. Cell. Physiol. 91, 15-29. Haug, E., Naess, O., and Gautvik, K.M. (1978) Mol. Cell. Endocrinol. 12, 81-95. Hayashi, I., and Sato, G.H. (1976) Nature (London) 259, 1322134. Hinkle, P.M., and Tashjian Jr., A.H. (1973) J. Biol. Chem. 248, 6180-6186. Karsten, U., and Wollenberger, A. (1972) Anal. Biochem. 46, 135-148. Kirkland, W.L., Sorrentino, J.M., and Sirbasku, D.A. (1976) J. Nat]. Cancer Inst. 56, 11591164. Lowry, O.H., Rosebrough, N.F., Farr, A.L., and Randall, R.J. (1951) J. Biol.Chem. 193, 265275. Maxfield, F.R., Willmgham, M.C., Davies, P.J.A., and Pastan, I. (1979) Nature (London) 277, 661-663. Morin, A., Tixier-Vidal, A., Gourdji, D., Kerdelhue, B., and Grouselle, D. (1975) Mol. Cell Endocrinol. 3, 351-373. Padmanabhan, V., and Convey, E.M. (1979) Mol. Cell. Endocrinol. 14, 103-112. Perrone, M.H., and Hinkle, P.M. (1978) J. Biol. Chem. 253, 5168-5173. PradelIes, P., Morgat, J.L., Fromageot, P., Oliver, C., Jacquet, P., Gourdji, D., and Tixier-Vidal, A. (1972) FEBS Lett. 22, 19-22. Quadri, SK., Oyama, T., and Spies, H.G. (1979) Endocrinology 104, 1649-1655. Rakoff, J.S., Siler, T.M., Sinha, Y.N., and Yen, S.S.C. (1973) J. Clin. Endocrinol. Metab. 37, 641-644. Raymond, V., Beaulieu, M., Labrie, F., and Boissier, J. (1978) Science 200, 1173-1175. Sando, G.N., Titus-Dillon, P., Hall, C.W., and Neufeld, E.F. (1979) Exp. Cell. Res. 119, 359364. Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 53,660-672. Tashjian Jr., A.H. and Hoyt Jr., R.F. (1972) In: Molecular Genetics and Developmental Biology, Ed.: M. Sussman (Prentice-Hall, Englewood cliffs, NJ) pp. 353 -387. Tashjian Jr., A.H., Yasumura, Y., Levine, L., Sato, G.H., and Parker, M.L. (1968) Endocrinoology 82, 342-352. Tashjian Jr., A.H., Bancroft, F.C., and Levine, L. (1970) J. Cell Biol. 47, 61-70. Tashjian Jr., A.H., Barowsky, N.J., and Jensen, D.K. (1971) Biochem. Biophys. Res. Commun. 43,516-523. Tashjian Jr., A.H., Osborne, R., Maina, D., and Knaian, A. (1977) Biochem. Biophys. Res. Commun. 79,333-340. Tixier-Vidal, A., Gourdji, D., PradeIIes, P., Morgat, J.L., Fromageot, P., and Kerdelhue, B. (1975) In: Hypothalamic Hormones, Vol. 6, Eds.: M. Motta, P.G. Grosignani and L. Martini (Academic Press, New York) pp. 89-107. Tixier-Vidal, A., Brunet, N., and Gourdji, D. (1978) In: Progress in Prolactin Physiology and Pathology, Eds.: C. Robyn and M. Harter (Elsevier/North-Holland, Amsterdam) pp. 29-43. Wu, R., and Sato, G.H. (1978) J. Toxicol. Environ. Health 4,427-448.