Life Sciences, Vol. Printed in the USA
51, pp.
771-778
Pergamon
Press
DOWN REGULATION OF MASKED AND UNMASKED INSULIN RECEPTORS IN THE LIVER OF TRANSGENIC MICE EXPRESSING BOVINE GROWTH HORMONE GENE
A. Balbis, J.M. Dellacha, R.S. Calandra*, A. Bartke* and D. Turyn Instituto de Qufmica y Fisicoquimica Biol6gicas (UBA-CONICET) Facultad de Farmacia y Bioqufmica Junfn 956, (1113) Buenos Aires, Argentina *Department of Physiology, School of Medicine, Southern Illinois University Carbondale, IL 62901-6512, USA (Received
in final
form June 29,
1992)
Summary_ The interaction of insulin with its receptor was studied in microsomes from livers of transgenic mice expressing the bovine growth hormone gene with mouse metallothionein-1 promoter (MT/bGH) and in their normal (non-transgenic) littermates. Specific binding of xzsI-insulin was detected in hepatic microsomes from normal and transgenic mice with an apparent Kd of 8 and 200 nM, for high and low affinity sites, respectively. The transgenic MT/bGH mice had a marked hyperinsulinism without significant elevation of plasma glucose levels. Under identical conditions of preparation and incubation, microsomes from the transgenic male and female mice bound 39% and 34% less insulin than those from their litter mates. Scatchard's analysis indicates that this decrease in binding is due to a decrease in the number of receptor sites. In contrast to the marked decrease in insulin binding to unmasked receptors, the levels of masked (also called cryptic) insulin receptors were similar (or slightly increased) in transgenic mice microsomes as compared to those of their normal litter mates. Changes in insulin receptor levels occur in various states of altered insulin sensitivity and there is considerable amount of evidence that insulin itself is the principal regulator of the number of insulin receptors (1-3). Gaven et al. (4) were the first to show that high concentrations of insulin can reduce the number of insulin receptors in vitro. The process of receptor depletion, known as down regulation, was shown to be insulin dependent in direct proportion to their biological activity (5). It has been suggested that insulin receptors not accessible to the hormone in binding assays exist in liver cells and microsomal preparations (6-10). We have demonstrated the presence of masked insulin receptors in microsomes from the rat submaxillary gland, and the utility of high ionic strength as a tool for unmasking insulin receptors (11). The present study was undertaken to address some aspects of the mechanism of insulin receptor regulation which are not fully understood using transgenic animals with constitutive expression of a heterologous growth hormone (GH) gene. We felt that these animals offer a useful model for this type of study because they are characterized by a continuous excess of GH, coexisting with secondary effects which include high levels of insulin with no change in plasma glucose level (12-16). In order to examine the masked and unmasked insulin receptor regulation in hepatic microsomes from normal and transgenic mice, we measured the binding to microsomes in a low and a high ionic strength buffer. Materials and Methods Reagents. Insulin from porcine pancreas and bovine serum albumin (Cohn fraction V) were purchased from Sigma Chemical Co., St. Louis, MO, U.S.A. Sephadex G-50 was from Copyright
0024-3205/92 $5.00 + .00 © 1992 Pergamon Press Ltd All rights
reserved.
772
Insulin Receptors in Transgenic Mice
Vol. 51, No. i0, 1992
Pharmacia Fine Chem. AB, Uppsala, Sweden. Na lz5I (17.4 Ci/mg) was obtained from New England Nuclear, Boston, MA, U.S.A. Other chemicals were of analytical reagent quality or of the highest purity commercially available. Iodination and purification of porcine insulin. Porcine insulin was iodinated at room temperature with NalZSIby the chloramine-T method (19-20). Specific activity ranged from 150 to 200 ~CiAtg, equivalent to 0.6-0.8 iodine atoms per insulin molecule. Before use, labelled insulin was purified at 4°C by gel filtration on a Sephadex G-50 chromatographic column (1.3 x 38 cm) equilibrated and eluted with Krebs-Ringer bicarbonate buffer (120 mM NaC1, 4.9 mM KC1,2.5 mM CaCI2, 1.2 mM KH2PO4, 1.3 mM MgSO4, 25 mM NaHCO3), with a pH of 7.4, containing 1% (w/v) bovine serum albumin. Transgenic mice were derived from animals kindly provided by Dr. Thomas E. Wagner. These animals were originally produced by a micro-injection of the bovine GH (bGH) gene fused to the mouse metallothionein-I (MT) promoter/regulator (MT/bGH) into the pronucleus of fertilized mouse eggs (14,21-22). Transgenic animals had markedly accelerated post-weaning growth leading to a significant (approximately 50%) increase in adult body weight. Plasma levels of heterologous (b) GH were approximately 6-21 ng/ml in MT/bGH transgenic mice (16, and F. Buonomo & A. Bartke, unpublished observations). The animals were maintained under conditions of a controlled photoperiod (12 h light : 12 h darkness) and temperature (22_+2°C) with a constant access to food (Teklad 6% Rat/Mouse Diet 002; 24% protein, 6% fat) and tap water without zinc. Normal C57BL/6 x C3H F1 hybrid females purchased from the Jackson Laboratory (Bar Harbor, Maine, U.S.A.) were bred with transgenic male mice and the resulting transgenic and normal (non-transgenic) litter mates were used in the present study. Microsome Preparation. The microsomal fractions from mouse livers were obtained according to Koch et al. (23). The livers were homogenized in an Omnimixer (Ivan Sorval Inc., Norwalk, CT, U.S.A.; set at full speed) for 3 min at 0-2°C in sucrose (250 retool/l). The homogenate was centrifuged at 12000 X g for 3 min at 4°C and the supernatant adjusted to 100 mmol NaC1/1 and 200 mmol MgSOjI. This was then centrifuged at 105,000 X g for 60 rain at 4°C. The dissociation of bound insulin was carried out by dilution (1:100) with cold acid buffer (pH:5.0) for 1 hr, then this was again centrifuged at 105,000 X g for 60 min at 4°C. The resultant pellet, resuspended with a dounce-type homogenizer (Kontea Glass Co., Vineland, N J, U.S.A.) in Tris-HC1 buffer (50 mmol/l: pH 7.4), represents the microsomal fraction. This suspension was kept at -20°C until used. Aliquots of the microsome suspension were solubilized by heating tbr 30 min at 100°C in NaOH (1 mol/1) and the protein concentration was determined by the method of Lowry (24), using bovine serum albumin as a standard. Measurement of 1~ Insulin Binding. Binding assays were performed by incubating, for 50 min at 25°C, a tracer amount of lzsI-insulin (0.45 nM) with the microsomal fraction (60-150 ~tg protein) in a total volume of 240-360 I.tl. These conditions were appropriate in order to reach the equilibrium for specific insulin binding (25). Polystyrene tubes were used, and the incubation was made in two different media: (a) Krebs-Ringer bicarbonate (low ionic strength), or (b) KrebsRinger bicarbonate-2M (high ionic strength), the composition of which is similar to that detailed for Krebs-Ringer bicarbonate except that the NaC1 concentration is 2 M instead of 120 raM. When the incubation was completed, 12sI-insulin bound to the microsomes were separated from free labelled hormone by filtration through a glass microfiber filter (Whatman GF/B). The filter was rinsed with 10 ml of the respective ice-cold incubation buffer and the radioactivity retained was counted in a well gamma counter. The radioactivity absorbed by the glass microfiber filter in the absence of microsomes was less than 0.2% of the total counts. The nonspecific binding (1251insulin bound in the presence of 15 ~tM of unlabeled insulin) was subtracted from the total binding. Nonspecific binding was always less than 10% of the total binding. Unless otherwise specified, all determinations were carried out in triplicates and the results were expressed as mean values. The parameters of the hormone-receptors interaction (dissociation constant and binding capacity) were derived from the 'best fit' line according to the LIGAND computer program developed by P. Munson to provide a weighed, nonlinear least-square analysis
Vol. 51, NO. I0, 1992
Insulin Receptors in Transgenic Mice
773
of the two-sites on Scatchard plots (17). Statistical differences between groups were evaluated by student's t-tests. Results Comparison of the specific binding of 125I-insulin to liver microsomal preparations from normal and transgenic mice revealed a significant (approximately 35-40%) decrease of the specific insulin binding in preparations from MT/bGH transgenic mice in comparison with normal mice (Table I). Absolute values obtained in control mice were similar to values measured in matched experiments with Swiss strain mice (Data not shown). Equilibrium binding data obtained from competition curves yielded a non-linear Scatchard plot (Fig 1). Equilibrium dissociation constants, Kd, of the high and low affinity components did not differ significantly between normal and transgenic mice, while binding capacity was significantly higher in liver microsomes from normal mice than in preparations derived from transgenic animals (Table II). At the time of death, glucose levels in transgenic mice were 182 + 16 mg/dl in males and 182 + 25 mg/dl in females, while in normal (control) mice they were 164 + 40 and 167 + 28 mg/dl, respectively. Serum insulin levels were 157 + 14 and 108 + 3 ~tlU/ml for male and female transgenic mice and 16 + 5 I.tlU/ml for normal animals of either sex. The effect of ionic strength of the incubation buffer on lzsI-insulin binding to liver microsomes of normal and transgenic mice is shown in Table III and Fig 2. Since Kd of the high and low affinity components did not differ significantly in either of the experimental conditions (low and high ionic strength), a comparison of matched experiments gave total binding capacity for the samples. The amount of non-expressed (masked) insulin receptors (11) were calculated by subtracting the binding of the unmasked receptors from the total binding (Table IV). In order to minimize the differences in protein concentrations, we calculated the difference in the ratio of total binding and unmasked insulin receptors between microsomes from transgenic and control mice. There was a 44% and 49% difference for male and female mice, respectively, indicating that masked (cryptic) insulin receptors in transgenic animals were nearly normal or slightly increased. Discussion The ability of insulin to down-regulate its own receptor is well documented (1-6). Studies with insulin-resistant animal model systems, such as the ob/ob mouse (1) and others (2-3), as well as with several "in vitro" systems (4-6), have led some investigators to postulate that downregulation of the insulin receptor is the principal process by which cells modulate their sensitivity to the hormone. The liver of transgenic mice overexpressing GH genes provides an "in vitro" model system for the study of insulin receptor regulation. In the line of transgenic mMT/bGH mice used in the present study, plasma insulin concentration is elevated 7 to 10 times above the normal values, and this dramatic enhancement of insulin levels coexists with no change in plasma glucose (175 mg/100 ml), in agreement with previous studies in other lines of GH-expressing transgenic mice (26). We have demonstrated that the insulin receptors of these animals undergo ligandinduced down-regulation (Table I). It is evident that insulin-induced receptor down regulation cannot be explained by a change in the affinity of the receptor for its ligand (Fig. 1 and Table II) in agreement with the literature data (1-6). The fact that the number of binding sites at saturating levels of insulin (Fig. 1) was markedly reduced by chronic exposure of the cells to insulin, indicates that the number of expressed receptors decreases during down regulation (Table I and II). However, it should be mentioned that the livers of transgenic mice are approximately twice as large as the normal livers (3.05 + 0.20 vs 1.45 + 0.15 g) and therefore the total number of expressed (in low ionic strength) insulin receptors per liver was 34 + 3.9% higher in transgenic than in normal mice. The presence in membrane preparations of insulin receptors which are unable to bind to the hormone under the currently employed experimental conditions of binding assays have been
Insulin Receptors in Transgenic Mice
774
Vol. 51, No. 10, 1992
0.8 •
Control
O Transgenic 0.6
IL
0.4
0.2
0.0 0
I
I
I
I
2
4
6
8
B (mM) Fig. 1 Scatchard plots of the binding data obtained from competition of unlabeled insulin with 'zsI-insulin for binding to liver microsomes of normal (120 I.tg protein) and MT/bGH transgenic mice (144 ~g protein), in anionic strength of 0.12 M (low ionic strength). B & F represents bound and free hormones, respectively; data points are the means of triplicate determinations. TABLE I. Functional sp.ecific binding of insulin to liver microsomes from normal and MT/bGH transgenic mace. Sex Male Male Female Female
Mice Normal Transgenic Normal Transgenic
Insulin Bound* 98.5 + 7 60.1 + 6 93.4 + 8 61.6 + 6
%Bound 100 + 7 61 + 6 100 + 9 66 + 7
p
51, pp.
771-778
Pergamon
Press
DOWN REGULATION OF MASKED AND UNMASKED INSULIN RECEPTORS IN THE LIVER OF TRANSGENIC MICE EXPRESSING BOVINE GROWTH HORMONE GENE
A. Balbis, J.M. Dellacha, R.S. Calandra*, A. Bartke* and D. Turyn Instituto de Qufmica y Fisicoquimica Biol6gicas (UBA-CONICET) Facultad de Farmacia y Bioqufmica Junfn 956, (1113) Buenos Aires, Argentina *Department of Physiology, School of Medicine, Southern Illinois University Carbondale, IL 62901-6512, USA (Received
in final
form June 29,
1992)
Summary_ The interaction of insulin with its receptor was studied in microsomes from livers of transgenic mice expressing the bovine growth hormone gene with mouse metallothionein-1 promoter (MT/bGH) and in their normal (non-transgenic) littermates. Specific binding of xzsI-insulin was detected in hepatic microsomes from normal and transgenic mice with an apparent Kd of 8 and 200 nM, for high and low affinity sites, respectively. The transgenic MT/bGH mice had a marked hyperinsulinism without significant elevation of plasma glucose levels. Under identical conditions of preparation and incubation, microsomes from the transgenic male and female mice bound 39% and 34% less insulin than those from their litter mates. Scatchard's analysis indicates that this decrease in binding is due to a decrease in the number of receptor sites. In contrast to the marked decrease in insulin binding to unmasked receptors, the levels of masked (also called cryptic) insulin receptors were similar (or slightly increased) in transgenic mice microsomes as compared to those of their normal litter mates. Changes in insulin receptor levels occur in various states of altered insulin sensitivity and there is considerable amount of evidence that insulin itself is the principal regulator of the number of insulin receptors (1-3). Gaven et al. (4) were the first to show that high concentrations of insulin can reduce the number of insulin receptors in vitro. The process of receptor depletion, known as down regulation, was shown to be insulin dependent in direct proportion to their biological activity (5). It has been suggested that insulin receptors not accessible to the hormone in binding assays exist in liver cells and microsomal preparations (6-10). We have demonstrated the presence of masked insulin receptors in microsomes from the rat submaxillary gland, and the utility of high ionic strength as a tool for unmasking insulin receptors (11). The present study was undertaken to address some aspects of the mechanism of insulin receptor regulation which are not fully understood using transgenic animals with constitutive expression of a heterologous growth hormone (GH) gene. We felt that these animals offer a useful model for this type of study because they are characterized by a continuous excess of GH, coexisting with secondary effects which include high levels of insulin with no change in plasma glucose level (12-16). In order to examine the masked and unmasked insulin receptor regulation in hepatic microsomes from normal and transgenic mice, we measured the binding to microsomes in a low and a high ionic strength buffer. Materials and Methods Reagents. Insulin from porcine pancreas and bovine serum albumin (Cohn fraction V) were purchased from Sigma Chemical Co., St. Louis, MO, U.S.A. Sephadex G-50 was from Copyright
0024-3205/92 $5.00 + .00 © 1992 Pergamon Press Ltd All rights
reserved.
772
Insulin Receptors in Transgenic Mice
Vol. 51, No. i0, 1992
Pharmacia Fine Chem. AB, Uppsala, Sweden. Na lz5I (17.4 Ci/mg) was obtained from New England Nuclear, Boston, MA, U.S.A. Other chemicals were of analytical reagent quality or of the highest purity commercially available. Iodination and purification of porcine insulin. Porcine insulin was iodinated at room temperature with NalZSIby the chloramine-T method (19-20). Specific activity ranged from 150 to 200 ~CiAtg, equivalent to 0.6-0.8 iodine atoms per insulin molecule. Before use, labelled insulin was purified at 4°C by gel filtration on a Sephadex G-50 chromatographic column (1.3 x 38 cm) equilibrated and eluted with Krebs-Ringer bicarbonate buffer (120 mM NaC1, 4.9 mM KC1,2.5 mM CaCI2, 1.2 mM KH2PO4, 1.3 mM MgSO4, 25 mM NaHCO3), with a pH of 7.4, containing 1% (w/v) bovine serum albumin. Transgenic mice were derived from animals kindly provided by Dr. Thomas E. Wagner. These animals were originally produced by a micro-injection of the bovine GH (bGH) gene fused to the mouse metallothionein-I (MT) promoter/regulator (MT/bGH) into the pronucleus of fertilized mouse eggs (14,21-22). Transgenic animals had markedly accelerated post-weaning growth leading to a significant (approximately 50%) increase in adult body weight. Plasma levels of heterologous (b) GH were approximately 6-21 ng/ml in MT/bGH transgenic mice (16, and F. Buonomo & A. Bartke, unpublished observations). The animals were maintained under conditions of a controlled photoperiod (12 h light : 12 h darkness) and temperature (22_+2°C) with a constant access to food (Teklad 6% Rat/Mouse Diet 002; 24% protein, 6% fat) and tap water without zinc. Normal C57BL/6 x C3H F1 hybrid females purchased from the Jackson Laboratory (Bar Harbor, Maine, U.S.A.) were bred with transgenic male mice and the resulting transgenic and normal (non-transgenic) litter mates were used in the present study. Microsome Preparation. The microsomal fractions from mouse livers were obtained according to Koch et al. (23). The livers were homogenized in an Omnimixer (Ivan Sorval Inc., Norwalk, CT, U.S.A.; set at full speed) for 3 min at 0-2°C in sucrose (250 retool/l). The homogenate was centrifuged at 12000 X g for 3 min at 4°C and the supernatant adjusted to 100 mmol NaC1/1 and 200 mmol MgSOjI. This was then centrifuged at 105,000 X g for 60 rain at 4°C. The dissociation of bound insulin was carried out by dilution (1:100) with cold acid buffer (pH:5.0) for 1 hr, then this was again centrifuged at 105,000 X g for 60 min at 4°C. The resultant pellet, resuspended with a dounce-type homogenizer (Kontea Glass Co., Vineland, N J, U.S.A.) in Tris-HC1 buffer (50 mmol/l: pH 7.4), represents the microsomal fraction. This suspension was kept at -20°C until used. Aliquots of the microsome suspension were solubilized by heating tbr 30 min at 100°C in NaOH (1 mol/1) and the protein concentration was determined by the method of Lowry (24), using bovine serum albumin as a standard. Measurement of 1~ Insulin Binding. Binding assays were performed by incubating, for 50 min at 25°C, a tracer amount of lzsI-insulin (0.45 nM) with the microsomal fraction (60-150 ~tg protein) in a total volume of 240-360 I.tl. These conditions were appropriate in order to reach the equilibrium for specific insulin binding (25). Polystyrene tubes were used, and the incubation was made in two different media: (a) Krebs-Ringer bicarbonate (low ionic strength), or (b) KrebsRinger bicarbonate-2M (high ionic strength), the composition of which is similar to that detailed for Krebs-Ringer bicarbonate except that the NaC1 concentration is 2 M instead of 120 raM. When the incubation was completed, 12sI-insulin bound to the microsomes were separated from free labelled hormone by filtration through a glass microfiber filter (Whatman GF/B). The filter was rinsed with 10 ml of the respective ice-cold incubation buffer and the radioactivity retained was counted in a well gamma counter. The radioactivity absorbed by the glass microfiber filter in the absence of microsomes was less than 0.2% of the total counts. The nonspecific binding (1251insulin bound in the presence of 15 ~tM of unlabeled insulin) was subtracted from the total binding. Nonspecific binding was always less than 10% of the total binding. Unless otherwise specified, all determinations were carried out in triplicates and the results were expressed as mean values. The parameters of the hormone-receptors interaction (dissociation constant and binding capacity) were derived from the 'best fit' line according to the LIGAND computer program developed by P. Munson to provide a weighed, nonlinear least-square analysis
Vol. 51, NO. I0, 1992
Insulin Receptors in Transgenic Mice
773
of the two-sites on Scatchard plots (17). Statistical differences between groups were evaluated by student's t-tests. Results Comparison of the specific binding of 125I-insulin to liver microsomal preparations from normal and transgenic mice revealed a significant (approximately 35-40%) decrease of the specific insulin binding in preparations from MT/bGH transgenic mice in comparison with normal mice (Table I). Absolute values obtained in control mice were similar to values measured in matched experiments with Swiss strain mice (Data not shown). Equilibrium binding data obtained from competition curves yielded a non-linear Scatchard plot (Fig 1). Equilibrium dissociation constants, Kd, of the high and low affinity components did not differ significantly between normal and transgenic mice, while binding capacity was significantly higher in liver microsomes from normal mice than in preparations derived from transgenic animals (Table II). At the time of death, glucose levels in transgenic mice were 182 + 16 mg/dl in males and 182 + 25 mg/dl in females, while in normal (control) mice they were 164 + 40 and 167 + 28 mg/dl, respectively. Serum insulin levels were 157 + 14 and 108 + 3 ~tlU/ml for male and female transgenic mice and 16 + 5 I.tlU/ml for normal animals of either sex. The effect of ionic strength of the incubation buffer on lzsI-insulin binding to liver microsomes of normal and transgenic mice is shown in Table III and Fig 2. Since Kd of the high and low affinity components did not differ significantly in either of the experimental conditions (low and high ionic strength), a comparison of matched experiments gave total binding capacity for the samples. The amount of non-expressed (masked) insulin receptors (11) were calculated by subtracting the binding of the unmasked receptors from the total binding (Table IV). In order to minimize the differences in protein concentrations, we calculated the difference in the ratio of total binding and unmasked insulin receptors between microsomes from transgenic and control mice. There was a 44% and 49% difference for male and female mice, respectively, indicating that masked (cryptic) insulin receptors in transgenic animals were nearly normal or slightly increased. Discussion The ability of insulin to down-regulate its own receptor is well documented (1-6). Studies with insulin-resistant animal model systems, such as the ob/ob mouse (1) and others (2-3), as well as with several "in vitro" systems (4-6), have led some investigators to postulate that downregulation of the insulin receptor is the principal process by which cells modulate their sensitivity to the hormone. The liver of transgenic mice overexpressing GH genes provides an "in vitro" model system for the study of insulin receptor regulation. In the line of transgenic mMT/bGH mice used in the present study, plasma insulin concentration is elevated 7 to 10 times above the normal values, and this dramatic enhancement of insulin levels coexists with no change in plasma glucose (175 mg/100 ml), in agreement with previous studies in other lines of GH-expressing transgenic mice (26). We have demonstrated that the insulin receptors of these animals undergo ligandinduced down-regulation (Table I). It is evident that insulin-induced receptor down regulation cannot be explained by a change in the affinity of the receptor for its ligand (Fig. 1 and Table II) in agreement with the literature data (1-6). The fact that the number of binding sites at saturating levels of insulin (Fig. 1) was markedly reduced by chronic exposure of the cells to insulin, indicates that the number of expressed receptors decreases during down regulation (Table I and II). However, it should be mentioned that the livers of transgenic mice are approximately twice as large as the normal livers (3.05 + 0.20 vs 1.45 + 0.15 g) and therefore the total number of expressed (in low ionic strength) insulin receptors per liver was 34 + 3.9% higher in transgenic than in normal mice. The presence in membrane preparations of insulin receptors which are unable to bind to the hormone under the currently employed experimental conditions of binding assays have been
Insulin Receptors in Transgenic Mice
774
Vol. 51, No. 10, 1992
0.8 •
Control
O Transgenic 0.6
IL
0.4
0.2
0.0 0
I
I
I
I
2
4
6
8
B (mM) Fig. 1 Scatchard plots of the binding data obtained from competition of unlabeled insulin with 'zsI-insulin for binding to liver microsomes of normal (120 I.tg protein) and MT/bGH transgenic mice (144 ~g protein), in anionic strength of 0.12 M (low ionic strength). B & F represents bound and free hormones, respectively; data points are the means of triplicate determinations. TABLE I. Functional sp.ecific binding of insulin to liver microsomes from normal and MT/bGH transgenic mace. Sex Male Male Female Female
Mice Normal Transgenic Normal Transgenic
Insulin Bound* 98.5 + 7 60.1 + 6 93.4 + 8 61.6 + 6
%Bound 100 + 7 61 + 6 100 + 9 66 + 7
p
