Molecular and CellularEndoc~noIo~, 13 (19’79) 11- 23 0 Elsevier/North-Holland Scientific Publishers, Ltd.
11
TITRATION OF TOTAL BINDING SITES FOR GROWTH HORMONE IN RABBIT LIVER Q~~TITAT~E
MODIFICATIONS
OF THESE SITES DURING PREGNANCY
Patrick GERASIMO ‘, Jean DJIANE ’ ** and Paul A. KELLY * r Laboratoire de Physiotogiede la Lactation, lnstitut National de f4 Recherche Agronomique, C.N.R.Z., 78350 Jouy-en-Josas, France, and 2 Medical Research Council Group in Molecular Endocrinology, Centre Hospitalierde WniversitP Lava/, 2 705 Boulevard Laurier, QuCbec Gl V 4C2, Canada Received 10 July 1978; Accepted 26 September 1978
The present work outlines various kinetic parameters of the interaction between bCH (bovine growth hormone) and a receptor located on the membranes of rabbit liver. A dissociation procedure, which strips the hormone off its receptor has been worked out, by varying MgCI2 concentrations and times of contact; exposure to 4 M MgCls for 15-30 min was found optimal for dissociation, without denaturing the receptor, as shown by the possibility of rebinding the hormone to the desaturated membranes. This method has been applied to titrate growth hormone receptors in rabbit liver, during pregnancy and lactation. The first half of pregnancy is characterized by a gradual increase of receptor levels and low saturation by endogenous hormone; conversely at parturition, there occurs a striking increase in receptors, which, however, may be demonstrated only after desaturation. During the lactational period, the receptors remain in part saturated. The results suggest that growth hormone levels increase at parturition and during lactation in the rabbit and hence may play an important role during lactation. The factors which modulate receptor levels at the same period remain unknown. Keywords: receptors; growth hormone; liver membranes; pregnancy.
It is now generally accepted that the first step in the action of peptide hormones on target organs consists in their specific binding to receptor sites located on the cell periphery. Binding sites specific for mammalian growth hormones are most abundant in membrane preparations from rabbit liver (Tsushima and Friesen, 1973; Posner et al., 1974; Kelly et al., 1974), whereas rat-liver receptors interact mainly with lactogenic hormones, such as prolactin, placental lactogen and primate growth hormone (Posner et al., 1974). However, no titration of total binding sites for bovine growth hormone (bGH) has been performed under various physiological conditions * To whom all correspondence
should be addressed.
P. Gerasimo et al.
12
and no systematic investigations on the in vivo or in vitro desaturation of the receptors (a prerequisite for their titration) has to our knowledge been undertaken Since we have recently shown (Djiane et al., 1977; Djiane and Durand, 1977) that prolactin receptor levels in the rabbit mammary gland fluctuate considerably during pregnancy and that these levels could be considered to reflect the organ’s sensitivity to the hormone, it seemed appropriate to test the response of a specific target tissue, the rabbit liver, to growth hormone, which shares many aspects of its mode of action with lactogenic hormones. The present study reports on the in vitro desaturation of growth-ho~one receptors prior to their titration. This step has been found to be technically feasible and reveals variations in the receptor levels, which, though less marked than with prolactin receptors in the mammary gland, are qualitatively quite similar. MATERIAL AND METHODS Animals 6-month-old female New Zealand rabbits (average weight 2.5 kg) have been used during their first pregnancy and maintained in air conditioned buildings. The day of coitus was considered as day 0. For all methodolo~cal work rabbits at 17 days pregnancy have been used.
The membranes from rabbit livers were prepared according to Tsushima and Friesen (1973) and the pellet at 100 OOOg was used. The livers were frozen and kept at -20°C. 5 g of tissue were cut into small fragments, homogenized in 25 ml of 0.3 M sucrose (using an Ultraturrax homogenizer) and centrifuged at 15 000 g (20 min). The pellet was kept for DNA assays, and the supernatant centrifuged for 90 min at 100 OOOg. The resulting pellet was suspended in 25 mM T&buffer pH 7.6, containing 10 mM MgCls (standard buffer). Protein concentrations were determined according to Lowry et al. (1951). The suspension was frozen at -20°C until subsequent binding studies were performed. It has pre~ously been shown that rabbit-liver membranes remain stable (in terms of GH binding) for at least 3 months. Iodination of growth hormone
Bovine growth hormone (bGH, N.I.H. 0743B, biological activity 1 IU/mg) which was generously supplied by the NIAMDD, N.I.H., was iodinated according to Greenwood et al. (1963) with the modifications introduced by Martal (1972) to a specific activity of approximately 80 &i/pg. The preparation was stable for periods not exceeding 2 weeks. Binding experiments
Binding tests were perfo~ed
using methods adapted from Tsushima and Friesen
Growth hormone receptors in rabbit liver
13
(1973) and Posner (1976): 200 &gof membranes (estimated as proteins) were incubated at 37°C for 1.5 h in the presence of iodinated growth hormone (total counts 10’ cpm). Some experiments were also performed at 4°C for periods up to 50 h. The volumes were adjusted to 0.5 ml with standard buffer containing 0.1% serum albumin and merthiolate 0.01%. All assays were performed in duplicate. Non-specific binding was assayed simultaneously, after adding a lOOO-fold excess of unlabelled hormone. Since it has been shown previously (Shiu et al., 1973) that membranes stored frozen in the presence of CaCla or MgClatend to sediment at low speeds, which has been confirmed with the material used in the present work (results not shown), the tubes were centrifuged at 3000g and the bound radioactivity in the pellet was determined in a Packard scintilIation counter. Specific binding was estimated by subtracting the non-specific from total binding values. Affinity constant (iu,) and binding capacities (N) were determined by Scatchard analysis (1949) of competition curves using a constant amount of [‘251]bGH (1 X IO5 cpm) and increasing concentrations of unlabelled bGH (1 to 20 ng/tube). The kinetic association and dissociation constants were calculated from the data of Figs. 1 and 5 according to Catt et al. (1976). In vitro dissociation of the receptorfhormone complex 200 1.18of membrane protein corresponding to 100 pl of membrane suspension
previously rehomogenized and suitably diluted with standard buffer were suspended in 600 4 of 4 M MgCla using a Vortex mixer and the mixture kept at room temperature for 30 min. Thereafter the solution was diluted adding 3 ml of standard buffer (0.1% BSA) and centrifuged at 3000g. The pellet was rinsed and centrifuged another time. When previously incubated with [12SI]bGH the finai pellet was counted directly. In all cases it was resuspended in 2OOfi of standard buffer (0.1% BSA) using the Vortex mixer and used to perform further binding assays.
RESULTS (A) Binding characte~stics of growth hormone to its receptor
Fig. 1 shows the time and temperature dependence of specific binding of the hormone to its receptor. Equ~ib~um is obtained rather slowly, but the kinetics match almost exactly the binding of prolactin to its receptors in the mammary gland, as observed in a previous work (Djiane et al., 1977). The inset of Fig. 1 demonstrates however that there is no advantage in extending the incubation unduly, since non-specific binding increases linearly up to 6 h at 37’C and eventually reaches a percentage (vs. total binding) high enough to interfere with the precision of the assay at this temperature. The dependence of the amount of membrane proteins on binding is shown in Fig. 2 which demonstrates that up to 40% of the labelled hormone may be bound to the receptors, and that the binding
P. Gerasimo et al.
I
z
0
2
4
6 Incubation
/
/
i I
I
20
time
50
’
’
’
’
1
hours
I
100
( hours)
Fig. 1. Effect of time and temperature on specific binding of ’ z SI-labelled bovine growth hormone to liver membranes. Membranes were incubated with 9 X lo4 cpm of [’ * 51]bGH in absence or in presence of an excess of unlabelled bGH. Results are expressed as percentage of the total [t2SI]bGH specifically bound per 200 gg of membrane protein. A, 4°C; o, 20°C; n, 37°C. The inset shows data at 37’C where total binding (m) and non-specific binding (0) are represented.
is roughly linear up to 400 pg protein. We have not attempted to assess if the reduction in binding at high protein concentrations is due to alterations of the unbound iodinated hormone or to a “protein effect”, which shifts the equilibrium of the binding reaction.
40, 0’
i .c 0
0’
.Y
q
0 .I? 20.c, ‘c x 0 z
/ / d” J loo
200
I
I
1000
500 Amount
pf
membrane
I,
I
2000 protein
per
tube
( Irg)
Fig. 2. Effect of the amount of membrane protein on binding of (12sI]bGH membranes.
to rabbit-liver
Growth hormone receptors in rabbit liver
200
400
600
800 total
15
1000 GH
t moles
26 f tnola~
GH
bound
JO /2OOup
protein
Fig. 3. (a) Binding equilibrium using unlabelied bCH to dilute the specific activity of the “% labelled bGH ( 10s cpm). The results are expressed as bGH bound (fmol) calculated from the specific binding of the tracer. (b) Scatchard ptot of the data obtained in (a).
Saturation kinetics of [ lzs I] bGH binding are outlined in Fig. 3 in the presence of suitable dilutions with cold hormone (saturation curve, Fig. 3a) or as a Scatchard plot (Fig. 3b). Some uncertainties remain concerning this latter plot, since previous papers have observed for rat liver Scatchard plots with single (Kelly et al., 1974; Posner et al., 1974) or multiple slopes (Herington et al., 1976). We failed to observe any obvious deviations from linearity and the class of low affinity binding sites described by Herington et al. (1976) were not clearly distinguishable from nonspecific binding sites. The specificity of [‘251]bGH binding to rabbit-liver membranes is shown in Fig. 4: growth hormones (either from bovine or human origin) readily displaced binding of the tracer, whereas human placental lactogen, at the concentrations used, failed to reduce binding. A less expected result was observed with ovine pituitary prolactin, the binding of which greatly exceeded its reported (1%) contamination with growth hormone. No explanation can presently be suggested for this phenomenon, which is under investigation. Binding could be reversed in the presence of an excess of unlabelled bovine growth hormone (Fig. 5). A kinetic dissociation constant (K-i = 4 X lo-’ M-i set-‘) could be deduced from this plot, which, when combined with the kinetic association constant (K+r = 1.25 X lo6 M-’ set-i) calculated at the same temperature from Fig. 1, led to an estimate for the dissociation constant KD = 3 X IO-” M about 17 times lower than the K, deduced from the Scatchard plot of Fig. 3 (ir, = 5 X lo-” M). This paradoxical behavlour has been recently reported (Garnier et al., 1975; Ross et al., 1977) for the interactions of other hormones with their membrane receptors and will be discussed below, since it may bear on some essential steps of the mechanism of hormone action. It was obvious, considering the relatively slow rate of dissociation of the
P. Gerasimo
et al.
hpL
aPri
I
1 2
I
4
I
I
I
8
12
20 Hormone
I 40
III 60
loo
concentration
(ng)
Fig. 4. Hormonal specificity of binding of bGW to rabbit-liver membranes. Competition for binding of 1251-labelled bGH (100% = binding in the absence of unlabelled hormone) in the presence of increasing amounts of unlabelled hormones.
hormone/receptor complex that in vitro desaturation of growth-hormone receptors, a prerequisite for their titration, would not be feasible under the conditions for membrane incubations. Therefore, it was necessary to use methods, which in preliminary experiments (Leblanc et al., 1977) and parallel studies on the disso-
Fig. 5. Dissociation of [ 1251]bGH from rabbit-liver membrane. Specific binding was determined as indicated in the Material and Methods section. To certain sets of tubes, an excess of unlabelled bovine growth hormone (A, 100 ng/tube; v, 1000 ng/tube) was added after an incubation period of 1.5 h (shown by arrow). The amount of 12sI-labelked bovine growth hormone remaining bound was followed up to 8.5 h.
Growth hormone receptors in rabbit liver
17
ciation of prolactin from rat-liver membranes (Kelly, Leblanc and Djiane, manuscript in preparation) had been shown to greatly accelerate the dissociation of the hormone/receptor complex, via an increase in the ionic strength using high molar concentrations of MgC12. (B) Desaturation of growth-hormone receptors Fig. 6a shows the in vitro desaturation kinetics at 20°C of receptor-bound tracer amounts of labelled growth hormone in the presence of increasing molarities of MgClz (0 to 5 M). Complete dissociation was observed only at concentrations of MgC12 equal or superior to 4 M, at least for the incubation periods used. Fig. 6b shows that reincubation at 37°C of the desaturated membranes under identical conditions and with the same amount of Iabelled growth hormone as used in the first binding study (Fig. 6a) results in essentially the same amounts of hormone bound at equilibrium, indicating that the desaturated specific binding sites remain accessible to the ligand. Fig. 7 (a and b) illustrates the same kinetics, but this time the membranes have been first fully saturated with unlabelled growth hormone either in vitro (Fig. 7a) or in vivo (Fig. 7b) (for experimental details, see the legend of the figure). On Fig. 7a, preincubation with unlabelled growth hormone leaves no specific sites available to the tracer (point 0 on the abscissa axis) and these sites gradually reappear to varying extents between 0 and 120 min, depending on the dissociation period and on MgCl? concentrations. An adequate recovery of all the sites pre-
a
15 30
60
b
120
time
(mn1
Fig. 6. (a) In vitro dissociation of tracer amounts of [ 12sI]bGH from membranes using buffer only (control, C) or increasing molar concentrations of MgCla for the indicated times. Initial binding was performed at 37’C for 2 h. (b) Dissociation (D) of [rZSI]bGH using the same conditions and symbols as described in (a) for 30 min, and subsequent binding (B) of fresh tracer to the depleted membranes for 2 h at 37°C.
18
P. Gerasimo et al. Q
b
Fig. 7. Dissociationof bGH bound to its receptor sites fo~owing in vitro (a) or in vivo (b) saturation with unlabeliedbGH. In part (a), control membraneswere incubatedwith 10s cpm Ii2 sI]bGA for 2 h at 37°C (0); samples were incubated with 2 *g/ml bGH for the same period and fohowittg centrifugation were exposed to increasing concentrations of MgCla for the periods indicated. In part (b), in vivo saturation was performed by an iV injection of 2.5 mg of bovine growth hormone to two rabbits. A controt rabbit was kiiled at 0 time, a bGH injected rabbit after a period of 30 min and another after 60 min. Specific binding at 0 time (0); specific binding at 0 time after MgCIa treatment (+); specific binding at 30 min (n); specific binding at 30 min after &$lz treatment (8); specific biding at 60 min fA); specific binding at 60 min after i%gC& treatment (A));D represents specific binding after desaturation with 4 M MgCla for 30 min.
viousiy occupied is observed with 4 M MgC12solely, whereas 5 M MgC& seems progressively to damage the membranes or else to sdubilize the receptors. High molar concentrations of MgClz (4 M-S M) also result in a maximal 30% loss of membrane protein. Fig. 7b illustrates similar experiments with receptors saturated in viva by i.v, injections of large amounts of unabetted growth hormone, Dissociation was performed in vitro, under the optimal conditions determined previously (MgCfl 4 M, 30 min). It appears that full saturation of the binding sites cannot be obtained in vivo which raises some questions on their accessibility within the cell; on the ather hand the binding seems fully reversible even if the delay between injection of the hormone and removal of the liver is extended up to 60 min.
Specific binding of growth hormone has been estimated either before or after desaturation of the membranes during pregnancy and pseudopregnancy, as shown in Fig. 8 (a and b). The values have been expressed both with respect to the amounts of proteins initiahy present in the membrane pellets (Fig. 8a) and to the amounts of proteins found before and after the desaturat~on procedure (for preg-
Growth hormone receptors in rabbit Ever
1 b
I 6
I 17
I 24
19
I 28
I I
I I4
days
Pr~gnWcy Lactation Fig, 8. Evolution of specific binding throughout pregnancy and lactation and throughout pseudopregnancy before and after in vitro depletion by 4 M MgCla for 30 min. (a) Values are expressed in percent of total amount of labelled growth hormone, used in each tube. Specific binding throughout pregnancy and lactation (0). Specific binding throughout pregnancy and lactation after MgCla treatment (0). SpecifTc binding throughout pseudopregnancy (*). Specific binding throughout pseudopregnancy after MgCla treatment (0). The numbers in parentheses refer to the number of animals in each group. (b) Curves are similar to that described previously, with the desaturation curve adjusted to 200 pg protein to account for the constant loss of 30 percent of the proteins during the dissociation process.
nant rabbits only, Fig. 8b). Obviously this should not interfere with the assessment of relative increases or decreases during the periods considered, and in fact the curves in both figures are almost identical. It appears however that a constant loss
P. Gerasimo
Zd
0d
17
11
TITRATION OF TOTAL BINDING SITES FOR GROWTH HORMONE IN RABBIT LIVER Q~~TITAT~E
MODIFICATIONS
OF THESE SITES DURING PREGNANCY
Patrick GERASIMO ‘, Jean DJIANE ’ ** and Paul A. KELLY * r Laboratoire de Physiotogiede la Lactation, lnstitut National de f4 Recherche Agronomique, C.N.R.Z., 78350 Jouy-en-Josas, France, and 2 Medical Research Council Group in Molecular Endocrinology, Centre Hospitalierde WniversitP Lava/, 2 705 Boulevard Laurier, QuCbec Gl V 4C2, Canada Received 10 July 1978; Accepted 26 September 1978
The present work outlines various kinetic parameters of the interaction between bCH (bovine growth hormone) and a receptor located on the membranes of rabbit liver. A dissociation procedure, which strips the hormone off its receptor has been worked out, by varying MgCI2 concentrations and times of contact; exposure to 4 M MgCls for 15-30 min was found optimal for dissociation, without denaturing the receptor, as shown by the possibility of rebinding the hormone to the desaturated membranes. This method has been applied to titrate growth hormone receptors in rabbit liver, during pregnancy and lactation. The first half of pregnancy is characterized by a gradual increase of receptor levels and low saturation by endogenous hormone; conversely at parturition, there occurs a striking increase in receptors, which, however, may be demonstrated only after desaturation. During the lactational period, the receptors remain in part saturated. The results suggest that growth hormone levels increase at parturition and during lactation in the rabbit and hence may play an important role during lactation. The factors which modulate receptor levels at the same period remain unknown. Keywords: receptors; growth hormone; liver membranes; pregnancy.
It is now generally accepted that the first step in the action of peptide hormones on target organs consists in their specific binding to receptor sites located on the cell periphery. Binding sites specific for mammalian growth hormones are most abundant in membrane preparations from rabbit liver (Tsushima and Friesen, 1973; Posner et al., 1974; Kelly et al., 1974), whereas rat-liver receptors interact mainly with lactogenic hormones, such as prolactin, placental lactogen and primate growth hormone (Posner et al., 1974). However, no titration of total binding sites for bovine growth hormone (bGH) has been performed under various physiological conditions * To whom all correspondence
should be addressed.
P. Gerasimo et al.
12
and no systematic investigations on the in vivo or in vitro desaturation of the receptors (a prerequisite for their titration) has to our knowledge been undertaken Since we have recently shown (Djiane et al., 1977; Djiane and Durand, 1977) that prolactin receptor levels in the rabbit mammary gland fluctuate considerably during pregnancy and that these levels could be considered to reflect the organ’s sensitivity to the hormone, it seemed appropriate to test the response of a specific target tissue, the rabbit liver, to growth hormone, which shares many aspects of its mode of action with lactogenic hormones. The present study reports on the in vitro desaturation of growth-ho~one receptors prior to their titration. This step has been found to be technically feasible and reveals variations in the receptor levels, which, though less marked than with prolactin receptors in the mammary gland, are qualitatively quite similar. MATERIAL AND METHODS Animals 6-month-old female New Zealand rabbits (average weight 2.5 kg) have been used during their first pregnancy and maintained in air conditioned buildings. The day of coitus was considered as day 0. For all methodolo~cal work rabbits at 17 days pregnancy have been used.
The membranes from rabbit livers were prepared according to Tsushima and Friesen (1973) and the pellet at 100 OOOg was used. The livers were frozen and kept at -20°C. 5 g of tissue were cut into small fragments, homogenized in 25 ml of 0.3 M sucrose (using an Ultraturrax homogenizer) and centrifuged at 15 000 g (20 min). The pellet was kept for DNA assays, and the supernatant centrifuged for 90 min at 100 OOOg. The resulting pellet was suspended in 25 mM T&buffer pH 7.6, containing 10 mM MgCls (standard buffer). Protein concentrations were determined according to Lowry et al. (1951). The suspension was frozen at -20°C until subsequent binding studies were performed. It has pre~ously been shown that rabbit-liver membranes remain stable (in terms of GH binding) for at least 3 months. Iodination of growth hormone
Bovine growth hormone (bGH, N.I.H. 0743B, biological activity 1 IU/mg) which was generously supplied by the NIAMDD, N.I.H., was iodinated according to Greenwood et al. (1963) with the modifications introduced by Martal (1972) to a specific activity of approximately 80 &i/pg. The preparation was stable for periods not exceeding 2 weeks. Binding experiments
Binding tests were perfo~ed
using methods adapted from Tsushima and Friesen
Growth hormone receptors in rabbit liver
13
(1973) and Posner (1976): 200 &gof membranes (estimated as proteins) were incubated at 37°C for 1.5 h in the presence of iodinated growth hormone (total counts 10’ cpm). Some experiments were also performed at 4°C for periods up to 50 h. The volumes were adjusted to 0.5 ml with standard buffer containing 0.1% serum albumin and merthiolate 0.01%. All assays were performed in duplicate. Non-specific binding was assayed simultaneously, after adding a lOOO-fold excess of unlabelled hormone. Since it has been shown previously (Shiu et al., 1973) that membranes stored frozen in the presence of CaCla or MgClatend to sediment at low speeds, which has been confirmed with the material used in the present work (results not shown), the tubes were centrifuged at 3000g and the bound radioactivity in the pellet was determined in a Packard scintilIation counter. Specific binding was estimated by subtracting the non-specific from total binding values. Affinity constant (iu,) and binding capacities (N) were determined by Scatchard analysis (1949) of competition curves using a constant amount of [‘251]bGH (1 X IO5 cpm) and increasing concentrations of unlabelled bGH (1 to 20 ng/tube). The kinetic association and dissociation constants were calculated from the data of Figs. 1 and 5 according to Catt et al. (1976). In vitro dissociation of the receptorfhormone complex 200 1.18of membrane protein corresponding to 100 pl of membrane suspension
previously rehomogenized and suitably diluted with standard buffer were suspended in 600 4 of 4 M MgCla using a Vortex mixer and the mixture kept at room temperature for 30 min. Thereafter the solution was diluted adding 3 ml of standard buffer (0.1% BSA) and centrifuged at 3000g. The pellet was rinsed and centrifuged another time. When previously incubated with [12SI]bGH the finai pellet was counted directly. In all cases it was resuspended in 2OOfi of standard buffer (0.1% BSA) using the Vortex mixer and used to perform further binding assays.
RESULTS (A) Binding characte~stics of growth hormone to its receptor
Fig. 1 shows the time and temperature dependence of specific binding of the hormone to its receptor. Equ~ib~um is obtained rather slowly, but the kinetics match almost exactly the binding of prolactin to its receptors in the mammary gland, as observed in a previous work (Djiane et al., 1977). The inset of Fig. 1 demonstrates however that there is no advantage in extending the incubation unduly, since non-specific binding increases linearly up to 6 h at 37’C and eventually reaches a percentage (vs. total binding) high enough to interfere with the precision of the assay at this temperature. The dependence of the amount of membrane proteins on binding is shown in Fig. 2 which demonstrates that up to 40% of the labelled hormone may be bound to the receptors, and that the binding
P. Gerasimo et al.
I
z
0
2
4
6 Incubation
/
/
i I
I
20
time
50
’
’
’
’
1
hours
I
100
( hours)
Fig. 1. Effect of time and temperature on specific binding of ’ z SI-labelled bovine growth hormone to liver membranes. Membranes were incubated with 9 X lo4 cpm of [’ * 51]bGH in absence or in presence of an excess of unlabelled bGH. Results are expressed as percentage of the total [t2SI]bGH specifically bound per 200 gg of membrane protein. A, 4°C; o, 20°C; n, 37°C. The inset shows data at 37’C where total binding (m) and non-specific binding (0) are represented.
is roughly linear up to 400 pg protein. We have not attempted to assess if the reduction in binding at high protein concentrations is due to alterations of the unbound iodinated hormone or to a “protein effect”, which shifts the equilibrium of the binding reaction.
40, 0’
i .c 0
0’
.Y
q
0 .I? 20.c, ‘c x 0 z
/ / d” J loo
200
I
I
1000
500 Amount
pf
membrane
I,
I
2000 protein
per
tube
( Irg)
Fig. 2. Effect of the amount of membrane protein on binding of (12sI]bGH membranes.
to rabbit-liver
Growth hormone receptors in rabbit liver
200
400
600
800 total
15
1000 GH
t moles
26 f tnola~
GH
bound
JO /2OOup
protein
Fig. 3. (a) Binding equilibrium using unlabelied bCH to dilute the specific activity of the “% labelled bGH ( 10s cpm). The results are expressed as bGH bound (fmol) calculated from the specific binding of the tracer. (b) Scatchard ptot of the data obtained in (a).
Saturation kinetics of [ lzs I] bGH binding are outlined in Fig. 3 in the presence of suitable dilutions with cold hormone (saturation curve, Fig. 3a) or as a Scatchard plot (Fig. 3b). Some uncertainties remain concerning this latter plot, since previous papers have observed for rat liver Scatchard plots with single (Kelly et al., 1974; Posner et al., 1974) or multiple slopes (Herington et al., 1976). We failed to observe any obvious deviations from linearity and the class of low affinity binding sites described by Herington et al. (1976) were not clearly distinguishable from nonspecific binding sites. The specificity of [‘251]bGH binding to rabbit-liver membranes is shown in Fig. 4: growth hormones (either from bovine or human origin) readily displaced binding of the tracer, whereas human placental lactogen, at the concentrations used, failed to reduce binding. A less expected result was observed with ovine pituitary prolactin, the binding of which greatly exceeded its reported (1%) contamination with growth hormone. No explanation can presently be suggested for this phenomenon, which is under investigation. Binding could be reversed in the presence of an excess of unlabelled bovine growth hormone (Fig. 5). A kinetic dissociation constant (K-i = 4 X lo-’ M-i set-‘) could be deduced from this plot, which, when combined with the kinetic association constant (K+r = 1.25 X lo6 M-’ set-i) calculated at the same temperature from Fig. 1, led to an estimate for the dissociation constant KD = 3 X IO-” M about 17 times lower than the K, deduced from the Scatchard plot of Fig. 3 (ir, = 5 X lo-” M). This paradoxical behavlour has been recently reported (Garnier et al., 1975; Ross et al., 1977) for the interactions of other hormones with their membrane receptors and will be discussed below, since it may bear on some essential steps of the mechanism of hormone action. It was obvious, considering the relatively slow rate of dissociation of the
P. Gerasimo
et al.
hpL
aPri
I
1 2
I
4
I
I
I
8
12
20 Hormone
I 40
III 60
loo
concentration
(ng)
Fig. 4. Hormonal specificity of binding of bGW to rabbit-liver membranes. Competition for binding of 1251-labelled bGH (100% = binding in the absence of unlabelled hormone) in the presence of increasing amounts of unlabelled hormones.
hormone/receptor complex that in vitro desaturation of growth-hormone receptors, a prerequisite for their titration, would not be feasible under the conditions for membrane incubations. Therefore, it was necessary to use methods, which in preliminary experiments (Leblanc et al., 1977) and parallel studies on the disso-
Fig. 5. Dissociation of [ 1251]bGH from rabbit-liver membrane. Specific binding was determined as indicated in the Material and Methods section. To certain sets of tubes, an excess of unlabelled bovine growth hormone (A, 100 ng/tube; v, 1000 ng/tube) was added after an incubation period of 1.5 h (shown by arrow). The amount of 12sI-labelked bovine growth hormone remaining bound was followed up to 8.5 h.
Growth hormone receptors in rabbit liver
17
ciation of prolactin from rat-liver membranes (Kelly, Leblanc and Djiane, manuscript in preparation) had been shown to greatly accelerate the dissociation of the hormone/receptor complex, via an increase in the ionic strength using high molar concentrations of MgC12. (B) Desaturation of growth-hormone receptors Fig. 6a shows the in vitro desaturation kinetics at 20°C of receptor-bound tracer amounts of labelled growth hormone in the presence of increasing molarities of MgClz (0 to 5 M). Complete dissociation was observed only at concentrations of MgC12 equal or superior to 4 M, at least for the incubation periods used. Fig. 6b shows that reincubation at 37°C of the desaturated membranes under identical conditions and with the same amount of Iabelled growth hormone as used in the first binding study (Fig. 6a) results in essentially the same amounts of hormone bound at equilibrium, indicating that the desaturated specific binding sites remain accessible to the ligand. Fig. 7 (a and b) illustrates the same kinetics, but this time the membranes have been first fully saturated with unlabelled growth hormone either in vitro (Fig. 7a) or in vivo (Fig. 7b) (for experimental details, see the legend of the figure). On Fig. 7a, preincubation with unlabelled growth hormone leaves no specific sites available to the tracer (point 0 on the abscissa axis) and these sites gradually reappear to varying extents between 0 and 120 min, depending on the dissociation period and on MgCl? concentrations. An adequate recovery of all the sites pre-
a
15 30
60
b
120
time
(mn1
Fig. 6. (a) In vitro dissociation of tracer amounts of [ 12sI]bGH from membranes using buffer only (control, C) or increasing molar concentrations of MgCla for the indicated times. Initial binding was performed at 37’C for 2 h. (b) Dissociation (D) of [rZSI]bGH using the same conditions and symbols as described in (a) for 30 min, and subsequent binding (B) of fresh tracer to the depleted membranes for 2 h at 37°C.
18
P. Gerasimo et al. Q
b
Fig. 7. Dissociationof bGH bound to its receptor sites fo~owing in vitro (a) or in vivo (b) saturation with unlabeliedbGH. In part (a), control membraneswere incubatedwith 10s cpm Ii2 sI]bGA for 2 h at 37°C (0); samples were incubated with 2 *g/ml bGH for the same period and fohowittg centrifugation were exposed to increasing concentrations of MgCla for the periods indicated. In part (b), in vivo saturation was performed by an iV injection of 2.5 mg of bovine growth hormone to two rabbits. A controt rabbit was kiiled at 0 time, a bGH injected rabbit after a period of 30 min and another after 60 min. Specific binding at 0 time (0); specific binding at 0 time after MgCIa treatment (+); specific binding at 30 min (n); specific binding at 30 min after &$lz treatment (8); specific biding at 60 min fA); specific binding at 60 min after i%gC& treatment (A));D represents specific binding after desaturation with 4 M MgCla for 30 min.
viousiy occupied is observed with 4 M MgC12solely, whereas 5 M MgC& seems progressively to damage the membranes or else to sdubilize the receptors. High molar concentrations of MgClz (4 M-S M) also result in a maximal 30% loss of membrane protein. Fig. 7b illustrates similar experiments with receptors saturated in viva by i.v, injections of large amounts of unabetted growth hormone, Dissociation was performed in vitro, under the optimal conditions determined previously (MgCfl 4 M, 30 min). It appears that full saturation of the binding sites cannot be obtained in vivo which raises some questions on their accessibility within the cell; on the ather hand the binding seems fully reversible even if the delay between injection of the hormone and removal of the liver is extended up to 60 min.
Specific binding of growth hormone has been estimated either before or after desaturation of the membranes during pregnancy and pseudopregnancy, as shown in Fig. 8 (a and b). The values have been expressed both with respect to the amounts of proteins initiahy present in the membrane pellets (Fig. 8a) and to the amounts of proteins found before and after the desaturat~on procedure (for preg-
Growth hormone receptors in rabbit Ever
1 b
I 6
I 17
I 24
19
I 28
I I
I I4
days
Pr~gnWcy Lactation Fig, 8. Evolution of specific binding throughout pregnancy and lactation and throughout pseudopregnancy before and after in vitro depletion by 4 M MgCla for 30 min. (a) Values are expressed in percent of total amount of labelled growth hormone, used in each tube. Specific binding throughout pregnancy and lactation (0). Specific binding throughout pregnancy and lactation after MgCla treatment (0). SpecifTc binding throughout pseudopregnancy (*). Specific binding throughout pseudopregnancy after MgCla treatment (0). The numbers in parentheses refer to the number of animals in each group. (b) Curves are similar to that described previously, with the desaturation curve adjusted to 200 pg protein to account for the constant loss of 30 percent of the proteins during the dissociation process.
nant rabbits only, Fig. 8b). Obviously this should not interfere with the assessment of relative increases or decreases during the periods considered, and in fact the curves in both figures are almost identical. It appears however that a constant loss
P. Gerasimo
Zd
0d
17
