Growth hormone (GH) excess is associated with secondary hyperinsulinemia, but the molecular mechanism and consequences of this alteration are poorly understood. To address this problem we have examined the levels and phosphorylation state of the insulin receptor (IR) and the insulin receptor substrate-1 (IRS-1), the association between IRS-1 and the p85 subunit of phosphatidylinositol 3-kinase (PI 3-kinase) as well as the PI 3-kinase activity in the livers of GH-transgenic mice. As expected, IR levels were reduced in the liver of GH-transgenic mice (55% of normal values) as determined by immunoblotting with an anti-IR beta-subunit antibody. IR and IRS-1 phosphorylation as determined by immunoblotting with antiphosphotyrosine antibody were increased in basal conditions by 315% and 560% respectively. After a bolus administration of insulin in vivo, IR phosphorylation increased by 40% while IRS-1 phosphorylation did not change. Insulin administration to control (normal) mice produced 670% and 300% increases in the IR and IRS-1 phosphorylation respectively. In the GH-transgenic animals, basal association of PI 3-kinase with IRS-1 as well as PI 3-kinase activity in liver was increased by 200% and 280% respectively, and did not increase further after administration of insulin in vivo, indicating a complete insensitivity to insulin at these levels. In conclusion, GH excess and the resulting secondary hyperinsulinemia were associated with alterations at the early steps of insulin action in liver. IR concentration was reduced, while IR and IRS-1 phosphorylation, IRS-1/PI 3-kinase association, and PI 3-kinase activity appeared to be maximally activated under basal conditions, thus making this tissue insensitive to further stimulation by exogenous insulin in vivo.

To study the effects of homologous mouse GH (mGH) on the presence and characteristics of serum GH-binding protein (GHBP) we have used transgenic mice expressing GH-releasing hormone (GHRH) as a model. Chromatographic techniques allowed the characterization of GHBP bioactivity, and immunological techniques were used to determine its concentration and molecular components. Chromatographic separation of labeled human GH or mGH cross-linked to serum GHBPs showed two GH-binding serum fractions in normal as well as in transgenic mice serum. SDS-PAGE of this material revealed a specific band of 66 kDa and another higher molecular weight broad band, which, in the presence of 2-mercapto-ethanol, is converted into the 66 kDa fraction.Since normal mice serum has an mGH concentration of 0. 40+/-0.06 nM and a GHBP concentration of 5.7+/-1.1 nM, while the high-affinity site for mGH has a K(d)</+/-27 nM, only a small percentage (2.9%) of total serum mGH is bound to GHBP in the sera of these mice. In transgenic mice serum, where the mGH concentration is 60 times higher (23+/-2.7 nM), 22.5% of total serum mGH is bound to serum GHBP. These values agree with the experimental data (4+/-2% and 17+/-4% for normal and transgenic mice serum respectively).The concentration of GHBP in GHRH transgenic mice was found to be increased four- to tenfold, depending on the technique used. This increment closely resembles the increased concentration of GHBP in the serum of transgenic bovine GH (bGH) mice, in which peripheral bGH levels are grossly elevated. Our results support the idea that the circulating levels of mGH in normal mouse serum are capable of influencing the levels of GHBP in peripheral circulation in a way similar to that of bGH, in spite of the different affinities of these two hormones. The fact that the up-regulation of GHBP occurs, even though a small percentage of mGH is bound in these animals, strongly suggests the existence of a physiological function for GHBP. These results also question some of the assigned or attributed physiological roles of GHBP, at least in the mouse, since only a negligible percentage of total mGH would be prevented from degradation and/or renal filtration by binding to GHBP. This small percentage of bound mGH also invalidates its role as a reservoir or a buffer of mGH concentration during pulses of GH release or rapid changes of mGH levels. Our results also demonstrate the presence of high molecular weight forms of GH-GHBP complexes that could be dissociated by dilution or in the presence of 2-mercapto-ethanol.

Overexpression of bovine growth hormone (bGH) in transgenic (PEPCK-bGH) mice induces resistance to insulin, which is compensated by a major increase in insulin levels. In these animals, hepatic insulin receptors (InsRs) are downregulated while tyrosine kinase activity of wheat germ agglutinin (WGA)-purified InsRs towards exogenous substrates is unexpectedly increased. By normalizing insulinemia, we attempted to determine whether the alterations detected in the early steps of insulin signal transduction are due to exposure to chronically high GH levels or are secondary to hyperinsulinemia. Transgenic PEPCK-bGH animals were treated with a single intraperitoneal administration of streptozotocin (STZ) or were deprived of food for 48 h, to normalize insulin levels. Both fasting and STZ treatment were effective in reducing insulin blood levels to control values or below, while GH levels remained unchanged (STZ treatment) or increased (fasted animals). In the liver of untreated transgenic mice, the number of InsRs as determined by 125I-insulin binding was significantly diminished (65+/-5% and 60+/-6% of normal values in microsomes and solubilized membranes respectively;P<0.01 vs control mice). In treated transgenic mice, the number of InsRs increased to values similar to or slightly higher than those found in normal control mice (STZ-treated: 139+/-26% and 126+/-8%; fasted: 128+/-5% (P<0.05) and 102+/-1.5%, for microsomes and solubilized membranes respectively). Neither treatment altered InsR affinity. InsR concentration in liver as determined by immunoblotting using an antibody against the beta-subunit of the insulin receptor was found to be reduced in transgenic mice (69+/-3% of normal values,P<0.001) and was normalized after both STZ treatment (105+/-4%) and fasting (109+/-4%). Insulin-stimulated autophosphorylation activity of InsRs in transgenic mice was increased (154+/-13%,P<0.01 compared with the control group), essentially normalized by STZ treatment (96+/-14%), and reduced by fasting, to below the values measured in normal control mice (56+/-15%,P<0.05). The potential influence of basal serine/threonine (Ser/Thr) phosphorylation of the InsR beta-subunit on the regulation of the InsRs from transgenic mice was also investigated. The autophosphorylation activity of WGA-purified InsRs from all groups of mice studied was essentially unchanged after dephosphorylation with alkaline phosphatase or mild trypsinization. Consequently, our results suggest that the observed changes in InsR number and autophosphorylation activity in the liver of bGH transgenic mice are directly related to changes in insulin blood levels, and that Ser/Thr phosphorylation is apparently not involved in the regulation of the InsR autophosphorylation activity in this model of insulin resistance.

Abstract

Ames dwarf mice that do not express growth hormone (GH) or prolactin (PRL) genes were used to study the effects of GH deficiency on the presence and the characteristics of GH-binding protein (GHBP) in serum. Chromatographic techniques were used to allow characterization of biological rather than immunological activity of GHBP. Two GH-binding fractions were found in dwarf mice serum, one with low affinity and high capacity (GHBP_I) and one with high affinity, low capacity and lower molecular mass (GHBP_II). Serum concentration of the high-affinity GHBP was 0·73 ± 0·03 nm with a K _d of 6·3 ± 1·7 nm. Since Ames dwarf mice have no GH in the circulation, all the GHBP is free. Interestingly, the concentration of GHBP in dwarf mice was similar to the levels of free GHBP measured in normal mice from the same line. Moreover, this value (0·7 nm) closely resembles the concentration of free GHBP in the serum of transgenic mice overexpressing GH, in which peripheral GH levels are grossly elevated. These observations can be interpreted as evidence that the levels of free GHBP in mouse serum are independent of GH concentration, and that GH influences only the levels of bound GHBP in peripheral circulation.

Journal of Endocrinology (1997) 153, 319–325

We have examined the regulation of hepatic growth hormone receptors (GH-R) and serum GH binding proteins (GHBP) in transgenic mice expressing an antagonist of bovine growth hormone (bGH), G119K-bGH, and consequently exhibiting a growth suppressed dwarf phenotype. Specific GHBP could be measured in transgenic dwarf mouse serum only by immunological methods (RIA), because these mice have a very high concentration of mutated bGH in circulation (> 1 microgram/ml) and, therefore, almost all GHBP is bound to G119K-bGH and cannot be quantitated in binding assays. The concentrations of GHBP were 0.6 +/- 0.4 nM and 1.7 +/- 0.4 nM for normal and dwarf mice respectively. The concentrations of free GHBP in normal mice and in transgenic mice expressing wild-type GH can be calculated using chromatographic techniques as the dissociation constant (Kd) and the ratio of bound 125I-GH to free 125I-GH in the serum ([GHBP]free = B/F.Kd). In agreement with the assumption that GHBP reflects GH-R status, liver uptake of injected labeled bGH was greatly reduced in transgenic dwarfs in comparison with normal mice or with transgenic mice expressing wild-type bGH (liver/blood ratio of 0.48 +/- 0.21, 2.7 +/- 0.2, and 1.3 +/- 0.3 respectively) indicating that the high concentration of the mutated bGH (G119K-bGH) prevents labeled bGH uptake, as was expected from the dwarf phenotype. 125I-bGH taken up by the liver of transgenic dwarf mice was found in a smaller molecular species than in normal mice, compatible with the presence of 1:1 [(GH-R):GH] complexes instead of the 2:1 [(GH-R)2:GH] or 2:2 [(GHBP)2:(GH)2] complexes found in normal mice. The concentration of IGF-I, the principal mediator of GH activity, in the G119K-bGH transgenic mice was correlated with the concentration of free GHBP. This allowed us to use free GHBP concentration as a marker of the effects of the active endogenous hormone (mGH) on liver receptors in the presence of different concentrations of the antagonist of GH. The levels of GHBP in serum, as well as the concentration of GH-R in liver microsomes from mice expressing the bGH antagonist, are up-regulated by the high concentration of G119K-bGH (85%), but significantly less so than that which could be expected for the same concentration of native GH (220-275%). This up-regulation suggests that the G119K-bGH antagonist is internalized and induces synthesis of the receptor and of the binding protein.

In the present study we have used hypopituitary Ames dwarf mice, which lack GH, prolactin and TSH, to investigate the consequences of the deficiency of these hormones on glucose homeostasis and on the initial components of the insulin signal transduction pathway in the liver. Ames dwarf mice displayed hypersensitivity to insulin since they maintained lower fasting glucose concentrations (73% of control values), had significantly reduced amounts of insulin (58% of control values), and exhibited an increased hypoglycemic response to exogenous insulin. Probably as a result of reduced insulin production, Ames dwarf mice displayed intolerance to glucose. The insulin-stimulated phosphorylation of the insulin receptor (IR) tended to be increased in the liver of Ames dwarf mice, while IR receptor protein content was increased by 38%. Insulin-stimulated phosphorylation of insulin receptor substrate (IRS)-1 and IRS-2 was increased by 61 and 72% respectively, while IRS-1 and IRS-2 protein levels were increased by 76 and 95%. The insulin-stimulated association of the p85 regulatory subunit of phosphatidylinositol (PI) 3-kinase with IRS-1 was increased by 28%, but unaltered with IRS-2. Interestingly, while the insulin-stimulated phosphotyrosine-derived PI 3-kinase activity was not changed, insulin-stimulated protein kinase B activation was increased by 41% in this tissue. These alterations may account for the insulin hypersensitivity exhibited by these animals. The present findings in long-lived Ames dwarf mice add to the evidence that insulin signaling is importantly related to the regulation of aging and life span.

Growth hormone (GH) deficiency is associated with increased sensitivity to insulin, but the molecular mechanisms involved in this association are poorly understood. In the current work, we have examined the consequences of the absence of the biological effects of GH on the first steps of the insulin signaling system in vivo in liver of mice with targeted disruption of the GH receptor/GH binding protein gene (GHR-KO mice). In these animals, circulating insulin concentrations are less than 4 microIU/ml, and glucose concentrations are low, concordant with a state of insulin hypersensitivity. The abundance and tyrosine phosphorylation state of the insulin receptor (IR), the IR substrate-1 (IRS-1), and Shc, the association between IRS-1 and the p85 subunit of phosphatidylinositol (PI) 3-kinase, the IRS-1- and the phosphotyrosine-associated PI 3-kinase in liver were examined. We found that, in liver of GHR-KO mice, the lack of GHR and GH eff! ects is associated with: (1) increased IR abundance, (2) increased insulin-stimulated IR tyrosine phosphorylation, (3) normal efficiency of IRS-1 and Shc tyrosine phosphorylation and (4) normal activation of PI 3-kinase by insulin. These alterations could represent an adaptation to the low insulin concentrations displayed by these animals, and may account for their increased insulin sensitivity.

Transgenic mice overexpressing GH present a marked GH signaling desensitization, reflected by low basal phosphorylation levels of the tyrosine kinase JAK2, and signal transducer and activator of transcription-5 (STAT5) and a lack of response of these proteins to a high GH dose. To evaluate the mechanisms involved in the regulation of JAK2 activity by high GH levels in vivo, the content and subcellular distribution of SH2-Bβ were studied in GH-overexpressing transgenic mice. SH2-B is a member of a conserved family of adapter proteins characterized by the presence of a C-terminal SH2 domain, a central pleckstrin homology (PH) domain, and an N-terminal proline rich region. The isoform SH2-Bβ modulates JAK2 activity by binding to the phosphorylated enzyme, further increasing its activity. However, it may also interact with non-phosphorylated inactive JAK2 via lower affinity binding sites, preventing abnormal activation of the kinase. SH2-Bβ may also function as an adapter protein, acting as a GH signaling mediator.

We now report that, in an animal model of GH excess in which JAK2 is not phosphorylated, although it is increased in the membrane-fraction, both the level of SH2-Bβ, and especially its association to membranes, are augmented (67% and 13-fold vs normal mice values respectively), suggesting SH2-Bβ could modulate JAK2 activity in vivo.

ABSTRACT

The characteristics of the specific binding of labelled insulin to turtle thyroid microsomes were investigated. Binding experiments were performed in Krebs–Ringer bicarbonate buffer (pH 7·4) at 25 or 4 °C for different periods of time. Dissociation of the labelled insulin from the binding sites was also evaluated. It was found that the binding is dependent on time, temperature and microsomal protein concentration, with an optimum pH of 8·0. Unlabelled insulin and pro-insulin competed with the labelled insulin, binding in direct proportion to their biological activities, while glucagon and growth hormones did not compete for the binding sites. Scatchard plot analysis established the presence of binding sites of high and low affinities, and the rate of dissociation of bound insulin was considerably increased by the addition of unlabelled insulin. Both results are compatible with a negative co-operativity site–site interaction model. Trypsin abolished the insulin binding. These findings indicate that the microsomes from the turtle thyroid gland contain specific binding sites for insulin. However, preincubation of microsomes with phospholipase C or S-adenosyl-l-methionine (SAM), or incubation in the presence of 2 mol NaCl/l did not increase the specific insulin binding. Therefore, the binding properties are similar to those observed in mammalian insulin-responsive tissues except for the absence of the effects of 2 mol NaCl/l, phospholipase C or SAM, which suggests the absence of masked insulin-binding sites.

J. Endocr. (1986) 108, 157–162

Serine phosphorylation of the insulin receptor (IR) has been proposed to exert an inhibitory influence on its tyrosine kinase activity. Previous works using site-directed mutagenesis suggested that serine 994 of the IR (IR Ser 994) might be part of an inhibitory domain of the receptor. In this study we examined whether this residue is subjected to phosphorylation in vivo. We used a site-phosphospecific antibody to determine the extent of phosphorylation of IR Ser 994 in insulin target tissues from two animal models of insulin resistance with different IR kinase (IRK) activity: obese (fa/fa) Zucker rats and transgenic mice overexpressing bovine growth hormone (PEPCK-bGH mice).Phosphorylation at IR Ser 994 was markedly increased in liver of obese rats. This alteration appeared to be tissue-selective since no phosphorylation on Ser 994 was detected in IRs isolated from skeletal muscle of these animals. On the other hand, the phosphorylation level of IR Ser 994 was very low in liver of PEPCK-bGH mice and did not differ from that of the control group. We have also demonstrated that protein kinase (PK) C isoforms alpha, betaI and zeta are able to promote the in vitro phosphorylation of the IR on Ser 994. Differential findings in these two models of insulin resistance might thus reflect increased PKC activity resulting from increased lipid availability in obese Zucker rats. Our results suggest that Ser 994 is a novel in vivo IR phosphorylation site that might be involved in the regulation of the IRK in some states of insulin resistance.