Nutritional regulation of the insulin-like growth factors.

Nutrition is one of the main regulators of circulating IGF-I. In humans, serum IGF-I concentrations are markedly lowered by energy and/or protein deprivation. Both energy and proteins are critical in the regulation of serum IGF-I concentrations. Indeed, after fasting, optimal intake of both energy and protein is necessary for the rapid restoration of circulating IGF-I. We believe, however, that in adult humans energy may be somewhat more important than protein in this regard. While the lowest protein intake is able to increase IGF-I in the presence of adequate energy, there is a threshold energy requirement below which optimal protein intake fails to raise IGF-I after fasting. When energy intake is severely reduced, the carbohydrate content of the diet is a major determinant of responsiveness of IGF-I to GH. The essential amino acid content of the diet is also critical for the optimal restoration of IGF-I after fasting, when protein intake is reduced. The exquisite sensitivity of circulating IGF-I to nutrients, the nycthemeral stability of its concentrations and its relative short half-life constitute the basis for its use as a marker of both nutritional status and adequacy of nutritional rehabilitation. For these indications, IGF-I measurement is more sensitive and more specific than measurement of the other nutrient-related serum proteins (albumin, prealbumin, transferrin, retinol-binding protein). Animal models have been developed to investigate the mechanisms responsible for the nutritional regulation of IGF-I. There is no doubt that many mechanisms are involved (Fig. 12). Decline of serum IGF-I in dietary restriction is independent of the diet-induced alterations in pituitary GH secretion. The role of the liver GH receptors is dependent on the severity of the nutritional insult. In severe dietary restriction (fasting), a marked decrease of the number of somatogenic receptors supports the role of a receptor defect in the decline of circulating IGF-I. In contrast, in less severe forms of dietary restriction (protein restriction), the decline of IGF-I results from a postreceptor defect in the GH action at the hepatic level. Nutritional deprivation decreases hepatic IGF-I production by diminishing IGF-I gene expression. Decline in IGF-I gene expression is mainly caused by nutrient deficiency and less importantly by the nutritionally induced hormonal changes (insulin and T3). Diet restriction also increases the clearance and degradation of serum IGF-I through changes in the levels of circulating IGFBPs.(ABSTRACT TRUNCATED AT 400 WORDS)

Dietary components that regulate serum somatomedin-C concentrations in humans.

Dietary components responsible for the regulation of somatomedin-C in humans were assessed in five adult volunteers of normal weight who were fasted for 5 d on three occasions, then refed three diets of differing composition. The serum somatomedin-C decreased from a mean prefasting value of 1.85 +/- 0.39 U/ml (+/- 1 SD) to 0.67 +/- 0.16 U/ml at the end of fasting (P less than 0.005). After refeeding for 5 d with a normal diet, the mean serum somatomedin-C increased to 1.26 +/- 0.20 U/ml. A protein-deficient (32% of control), isocaloric diet resulted in a significantly smaller increase, to a mean value of 0.90 +/- 0.24 U/ml (P less than 0.05). A diet deficient in both protein and energy led to a further fall 0.31 +/- 0.06 U/ml. The changes in somatomedin-C during fasting and refeeding correlated significantly with mean daily nitrogen balance (r = 0.90). We conclude that both protein and energy intake are regulators of serum somatomedin-C concentrations in adult humans, and energy intake may be of greater importance. The correlation between changes in somatomedin-C and nitrogen balance suggests that the former are directly related to changes in protein synthesis and may be helpful in assessing the response to nutritional therapy.

Enhancement of the anabolic effects of growth hormone and insulin-like growth factor I by use of both agents simultaneously.

The use of growth hormone (GH) as an anabolic agent is limited by its tendency to cause hyperglycemia and by its inability to reverse nitrogen wasting in some catabolic conditions. In a previous study comparing the anabolic actions of GH and IGF-I (insulin-like growth factor I), we observed that intravenous infusions of IGF-I (12 micrograms/kg ideal body wt [IBW]/h) attenuated nitrogen wasting to a degree comparable to GH given subcutaneously at a standard dose of 0.05 mg/kg IBW per d. IGF-I, however, had a tendency to cause hypoglycemia. In the present study, we treated seven calorically restricted (20 kcal/kg IBW per d) normal volunteers with a combination of GH and IGF-I (using the same doses as in the previous study) and compared its effects on anabolism and carbohydrate metabolism to treatment with IGF-I alone. The GH/IGF-I combination caused significantly greater nitrogen retention (262 +/- 43 mmol/d, mean +/- SD) compared to IGF-I alone (108 +/- 29 mmol/d; P < 0.001). GH/IGF-I treatment resulted in substantial urinary potassium conservation (34 +/- 3 mmol/d, mean +/- SE; P < 0.001), suggesting that most protein accretion occurred in muscle and connective tissue. GH attenuated the hypoglycemia induced by IGF-I as indicated by fewer hypoglycemic episodes and higher capillary blood glucose concentrations on GH/IGF-I (4.3 +/- 1.0 mmol/liter, mean +/- SD) compared to IGF-I alone (3.8 +/- 0.8 mmol/liter; P < 0.001). IGF-I caused a marked decline in C-peptide (1,165 +/- 341 pmol/liter; mean +/- SD) compared to the GH/IGF-I combination (2,280 +/- 612 pmol/liter; P < 0.001), suggesting maintenance of normal carbohydrate metabolism with the latter regimen. GH/IGF-I produced higher serum IGF-I concentrations (1,854 +/- 708 micrograms/liter; mean +/- SD) compared to IGF-I only treatment (1,092 +/- 503 micrograms/liter; P < 0.001). This observation was associated with increased concentrations of IGF binding protein 3 and acid-labile subunit on GH/IGF-I treatment and decreased concentrations on IGF-I alone. These results suggest that the combination of GH and IGF-I treatment is substantially more anabolic than either IGF-I or GH alone. GH/IGF-I treatment also attenuates the hypoglycemia caused by IGF-I alone. GH/IGF-I treatment could have important applications in diseases associated with catabolism.

Hormonal control of immunoreactive somatomedin production by cultured human fibroblasts.

Human growth hormone (hGH) is known to be a potent stimulator of somatomedin secretion in vivo. The induction of somatomedin by growth hormone has been difficult to study in vitro, however, because no organ containing a high concentration of somatomedin has been identified. Because fetal mouse explants have been shown to produce somatomedin in vitro, we have undertaken studies to determine whether postnatal human fibroblast monolayers also produce somatomedin, and if so, whether its production is regulated by other hormones. Quiescent human fibroblasts were exposed to serum-free minimum essential medium, and the medium was assayed for somatomedin concentration using a specific radioimmunoassay for somatomedin-C. A progressive rise in immunoreactive somatomedin to 0.08 U/ml per 10(5) cells per 24 h was observed over 72 h of incubation. This was an underestimation of the actual concentration of immunoreactive somatomedin since the amount measured following acid treatment was at least fourfold higher than in the untreated medium. Growth hormone stimulated immunoreactive somatomedin production in a dose-dependent manner: 5 ng hGH/ml = 0.1 U/ml per 10(5) cells; 50 ng hGH/ml = 0.25 U/ml per 10(5) cells. Platelet-derived growth factor and fibroblast growth factor were also stimulatory, but epidermal growth factor, thyroxine, or cortisol had no effect. Media that had been exposed to human fibroblasts stimulated DNA synthesis in BALB/c 3T3 fibroblasts (a cell type that does not produce somatomedin). Medium-derived immuno-reactive somatomedin eluted from Sephacryl S-200 in two major peaks (150,000 and 8,000 mol wt). The higher molecular weight peak is similar to the one observed when whole serum was used. These studies provide a model system for studying the humoral and nonhumoral factors that control the biosynthesis of somatomedin by human tissues. Since immunoreactive somatomedin production may be a rate-limiting factor for fibroblast growth, the delineation of the hormonal control of somatomedin production should lead to a better understanding of the mechanisms controlling human fibroblast growth.

Estimation of somatomedin-C levels in normals and patients with pituitary disease by radioimmunoassay.

The development of a radioimmunoassay for somatomedin-C has for the first time made it possible to discriminate between serum concentrations of a single peptide or closely related group of peptides and the net somatomedin activity measured by less specific bioassay and radioreceptor techniques. Antibodies to human somatomedin-C were raised in rabbits using a somatomedin-C ovalbumin complex as the antigen. A variety of peptide hormones at concentrations up to 1 muM are not recognized by the antibody. Insulin at concentrations >0.1 muM cross reacts in a non-parallel fashion; purified somatomedin-A is only 3% as active as somatomedin-C; and radiolabeled cloned rat liver multiplication stimulating activity does not bind to the antibody. Immunoreactive somatomedin-C can also be quantitated in the sera of a variety of subhuman species. Unusual assay kinetics, which are manifest when reactants are incubated under classic "equilibrium" assay conditions, appear to result from the failure of (125)I-somatomedin-C to readily equilibrate with the somatomedin-C serum binding protein complex. It is, therefore, necessary to use nonequilibrium assay conditions to quantitate somatomedin-C in serum. With this assay it is possible to detect somatomedin-C in normal subjects using as little as 0.25 mul of unextracted serum. Serum somatomedin-C concentrations in normal subjects were lowest in cord blood and rose rapidly during the first 4 yr of life to near adult levels. In 23 normal adult volunteers, the mean serum somatomedin-C concentration was 1.50+/-0.10 U/ml (SEM) when compared to a pooled adult serum standard. 19 children with hypopituitary dwarfism had concentrations below 0.20 U/ml. 17 of these were below 0.1 U/ml, the lower limit of sensitivity of the assay. The mean concentration in 14 adults with active acromegaly was 6.28+/-0.37 U/ml (SEM), five times greater than the normal volunteers. Significant increases in serum somatomedin-C concentrations were observed in 8 of 10 hypopituitary children within 72 h after the parenteral administration of human growth hormone. Three patients with Cushing's disease had elevated serum somatomedin-C concentrations (2.61+/-0.14 U/ml [SEM]). Three patients with hyperprolactinemia had normal concentrations (1.74+/-0.11 U/ml [SEM]).The important new discovery brought to light by quantitation of immunoassayable somatomedin in patient sera is that all previously used assays detect, in addition to somatomedin-C, serum substances that are not under as stringent growth hormone control.

Molecular cloning and characterization of a new fasting-inducible short-chain dehydrogenase/reductase from rat liver(1).

We have isolated a 668 bp cDNA from fasted rat liver, designated RLF98, by suppression subtractive hybridization (SSH). The full-length RLF98 cDNA, cloned by rapid amplification of cDNA ends (RACE), is 1113 bp long with an open reading frame of 912 bp. This cDNA encodes a protein of 303 amino acid residues with a calculated molecular weight of 32433 Da. In vitro transcription and translation of the full-length RLF98 cDNA produced a protein of about 33 kDa. The RLF98 protein shares strong amino acid sequence homology with members of the short-chain dehydrogenase/reductase (SDR) family. Northern analysis of RNA from rat liver revealed a transcript of 1.1 kb. Fasting increased this mRNA 2.7-fold. While the RLF98 gene is widely expressed in rat tissues, its level of expression is highly variable. Expression in liver and kidney is abundant and is more than 10 times that observed in other tissues. Our data indicate that the RLF98 is a new member of the SDR family that is upregulated by fasting. Additional experiments including purification of recombinant RLF98 protein are in progress to define the specific function of this protein and the role it plays during fasting-induced catabolism.

Identification of a new gene (rat TM6P1) encoding a fasting-inducible, integral membrane protein with six transmembrane domains.

We describe the isolation of a new gene that encodes a membrane-integrated protein with six transmembrane domains, termed TM6P1. A 403-bp expressed sequence tag was isolated from fasted rat liver subtracted cDNA library, and its full-length cDNA is 1482 bp long. It contains an open reading frame of 816 bp and is predicted to encode a 271-amino acid protein with a deduced mass of 29520 Da. A sequence homology search failed to show significant correspondence to any known protein in the databank. TM6P1 has six highly hydrophobic domains that are predicted to be transmembrane helices. Consistent with this prediction, the TM6P1-EGFP fusion protein was shown to localize to the plasma membrane. TM6P1 mRNA is widely expressed in rat tissues, with placenta and liver being the most abundant sites. Fasting increased TM6P1 mRNA nearly two-fold in liver. Taken together, our data suggest that TM6P1 is a unique new membrane integral protein that might have a function important during fasting-induced catabolism.

Low plasma somatomedin-C in streptozotocin-induced diabetes mellitus. Correlation with changes in somatogenic and lactogenic liver binding sites.

To determine the mechanism for the decrease of somatomedin levels in insulin-dependent diabetes, the relationships among plasma immunoreactive somatomedin-C (Sm-C), plasma growth hormone (GH) and prolactin (PRL), and the somatogenic and lactogenic binding sites in liver were assessed in rats with nonketotic diabetes mellitus of different duration (1 wk or 1 mo) and severity (50 or 60 mg streptozotocin/kg BW). One week after administration of 60 mg streptozotocin (STZ)/kg, plasma Sm-C concentrations were significantly decreased (0.23 +/- 0.03 versus 0.43 +/- 0.03 U/ml in controls; mean +/- SEM, P less than 0.01). In contrast, plasma GH concentrations, bovine GH (bGH) binding, and human GH (hGH) binding were not significantly changed. After 1 mo of diabetes, no further decrease in plasma Sm-C content was observed despite a reduction in plasma GH and PRL concentrations and reduced hepatic bGH binding capacity (5 +/- 2 versus 38 +/- 4 fmol/mg protein; P less than 0.01). In the group of rats injected with 50 mg STZ/kg, the Sm-C was reduced at 1 mo but hepatic GH binding was not. In a second study, diabetic rats (75 mg STZ/kg) were treated after 3 wk with insulin (10 U lente per day for 7 days). This treatment normalized Sm-C levels and partially restored the GH binding capacity (treated: 49 +/- 4 fmol/mg protein versus untreated diabetics: 28 +/- 6 fmol/mg protein; P less than 0.01 and versus controls: 68 +/- 4 fmol/mg protein; P less than 0.05).2+ suggest that in an early phase of nonketotic diabetes, the low plasma Sm-C is not due primarily to reduced GH receptor number.(ABSTRACT TRUNCATED AT 250 WORDS)

Serum somatomedin-C concentrations in a rabbit model of diabetic pregnancy.

We have developed and validated a method for measuring immunoreactive somatomedin-C (Sm-C) in serum of rabbits, and have shown that during midgestation (11-26 days; gestation = 31 days) serum Sm-C concentrations are higher in normal pregnant animals than in pregnant diabetic animals. Sm-C concentrations in the serum of 28-day gestation fetuses of diabetic rabbits (3.14 +/- 0.25 U/ml) were significantly higher than in the fetuses of nondiabetic rabbits (2.31 +/- 0.23 U/ml; P less than 0.05). Fetuses from litters of the most severely hyperglycemic diabetic mothers (glucose greater than 250 mg/dl) had higher serum Sm-C (3.66 +/- 0.41 U/ml) than those of mothers who were mildly hyperglycemic (2.71 +/- 0.2 U/ml). Although these differences were not statistically significant, fetuses from the former litters accounted in great part for the difference between the fetuses of diabetic and normal pregnancy. The diabetes-related increment in Sm-C does not appear to be due to insulin, since fetal insulin concentrations were not different between the normal and diabetic litters (normal, 50.0 +/- 3.6 microU/ml versus diabetic, 49.6 +/- 7.6 microU/ml). Despite their elevation in serum Sm-C, fetuses from litters of diabetic rabbits were growth retarded in weight (26.8 +/- 6.9 g and 33.8 +/- 6.9 g, diabetic versus normal pregnancy; P less than 0.05) and in length (7.9 +/- 0.7 cm and 8.6 +/- 0.7 cm, diabetic versus normal pregnancy; P less than 0.025). We speculate that these discrepancies between growth and Sm-C might be secondary to the toxic effects of glucose on embryonic growth and that later in gestation, the excessive energy provided to the fetus might stimulate Sm-C synthesis.

Hemopoietic growth factors show promise as therapeutic agents.

Hemopoietic growth factors such as erythropoietin and colony-stimulating factors may have important therapeutic applications.

Contributions of growth hormone receptor and postreceptor defects to growth hormone resistance in malnutrition.

Malnutrition results in poor growth and is associated with resistance to growth hormone (GH) action. The mechanisms involved in the GH resistance depend on the severity and the timing of the nutritional insult. Stringent dietary restrictions such as fasting may produce GH resistance by reducing the number of GH receptors. Less severe nutritional deprivation such a short-term protein restriction may cause GH insensitivity mainly through postreceptor mechanisms.

Regulation of components of the ubiquitin system by insulin-like growth factor I and growth hormone in skeletal muscle of rats made catabolic with dexamethasone.

To investigate whether the anabolic effects of insulin-like growth factor I (IGF-I) and GH are mediated through regulation of the ubiquitin (Ub) pathway, we examined the effect of IGF-I (0.35 microg/100 g) and/or GH (0.3 mg/100 g BW) on the expression of Ub and Ub-conjugating (E2) enzyme messenger RNAs (mRNAs) in skeletal muscle of rats made catabolic by treatment with dexamethasone (Dex; 0.5 mg/100 g BW) for 3 days. Dex caused a significant loss of body and gastrocnemius weight (14% and 25%, respectively) concurrent with an increase in the 2.8- and 1.2-kb transcripts of Ub (14.3- and 12-fold increases, respectively), the 1.8- and 1.2-kb transcripts of 14-kDa Ub-conjugating enzyme (E2-14 kDa; 5.6- and 7.7-fold increases, respectively), the 4.4- and 2.4-kb transcripts of Ub-E2G (6.5- and 8.2-fold increases, respectively), and the 2E isoform of the 17-kDa E2 mRNA (3.5-fold increase). Injections of IGF-I in Dex-treated animals ameliorated the body weight loss, and the gastrocnemius tended to be heavier. This improvement was also accompanied by a significant suppression of transcripts for Ub (58% and 66% decreases), E2-14 kDa (58% and 68% decreases), and Ub-E2G (78% decrease), whereas the 2E isoform of the 17-kDa E2 was only modestly affected (20% decrease). GH restored serum IGF-I levels to normal in Dex-treated rats, but had no effect on the body weight loss or on any of the studied components of the Ub pathway. Administration of IGF-I to the Dex/GH-treated animals decreased the activated mRNAs of the Ub pathway in a manner and degree similar to those observed in the Dex/ IGF-I group. In summary, these results provide evidence that IGF-I regulates the expression of mRNAs encoding components of the Ub pathway during catabolism and suggest a possible mechanism for the antiproteolytic actions of IGF-I. On the other hand, GH, which is believed not to affect proteolysis but only protein synthesis, had no effect on any of the mRNAs studied.

Low serum somatomedin-C in insulin-dependent diabetes: evidence for a postreceptor mechanism.

In insulin-dependent diabetic rats, plasma somatomedin (Sm) levels are low and are not corrected by GH treatment, suggesting GH resistance. To define the mechanism of this GH-resistant state, the number and affinity constant of bovine liver GH-binding sites and the serum Sm-C responses to injections of bovine GH were determined in control (diluent-injected) and diabetic (streptozotocin-injected; 40 mg/kg BW) hypophysectomized rats. The affinity constants (Ka) of the GH-binding sites of control (0.92 +/- 0.07 X 10(9) M-1) and diabetic animals (0.68 +/- 0.04 X 10(9) M-1) were not significantly different (P less than 0.1). Likewise, there were no significant differences in the liver GH-binding capacities between control and diabetic hypophysectomized rats, whether these capacities were expressed as picomoles per liver (26.99 +/- 3.43 vs. 22.27 +/- 2.55, controls vs. diabetics), picomoles per mg DNA (1.26 +/- 0.15 vs. 1.10 +/- 0.12), or femtomoles per mg protein (30.95 +/- 4.08 vs. 29.98 +/- 2.70). Despite the absence of alterations in liver GH-binding sites, the Sm-C responses 24 h after sc injections of graded doses of bovine GH were severely blunted in the diabetic animals. The maximal serum Sm-C response in the controls was 0.81 +/- 0.12 U/ml, but was only 0.09 +/- 0.01 U/ml in the diabetics (P less than 0.01). The dose of GH required to achieve the half-maximal Sm-C response (ED50) was similar in diabetic and nondiabetic rats (700-900 micrograms). The absence of significant alterations in liver GH binding and the decreased maximal serum Sm-C response without changes in the ED50 suggest that the GH-resistant state in insulin-dependent diabetes is due to a postreceptor defect.

Effect of fasting on insulin-like growth factor (IGF)-IA and IGF-IB messenger ribonucleic acids and prehormones in rat liver.

The insulin-like growth factor I (IGF-I) gene generates by alternative splicing two IGF-I messenger RNAs (mRNAs) coding for IGF-I prehormones with different E domain sequences. In rats, these two mRNAs differ by the presence (IGF-IB) or absence (IGF-IA) of a 52-bp insert in the E domain coding region. The purpose of this study was to investigate the effect of nutritional perturbation on IGF-IA and -IB expression in rat liver. Northern blot analysis of liver mRNA revealed that the 1.5-1.9 kb and 0.9-1.2 kb IGF-I mRNA species were decreased in rats fasted for 48 h compared with either fasted-refed (48 h of each) or control-fed rats (each, P < 0.01), whereas the 7.5 kb IGF-I mRNA was decreased only when compared with the fasted-refed animals. Using semiquantitative RT-PCR, the IGF-IA transcript (114 bp amplicon) was not altered, whereas the IGF-IB transcript (166 bp amplicon) was decreased in fasted rats compared with the other two groups (both P < 0.01). We confirmed the RT-PCR results by RNase protection assay (RPA), observing that the IGF-IA (224 and 100 bases protected) was not decreased and that the IGF-IB transcript (376 bases protected), accounting for only 23% of the total IGF-I transcripts of control fed rats, was decreased by fasting. Because the results from RT-PCR and RPA do not necessarily predict full-length translatable mRNA, we subjected hepatic IGF-I transcripts to in vitro translation, and we immunoprecipitated IGF-IA and -IB prehormones. Both prehormones were translated principally from exon 1-containing mRNAs, with molecular weights of about 17K and 18K, representing 80% and 20% of the total IGF-I prehormones observed in control fed rats, respectively. Both peptides were reduced in fasted rats compared with controls (P < 0.01), and refeeding restored both. By immunoblotting of the protein extract from liver of fasted rats, IGF-IA was decreased by 77% compared with control-fed animals. Refeeding returned IGF-IA to normal. The lack of reduction of IGF-IA transcript at the alternative splice site suggests that posttranscriptional mechanisms are responsible for the reduction in steady-state IGF-I mRNAs that occurs during fasting. Additionally, we present evidence that biosynthesis of IGF-IA and -IB prehormones by liver is impaired at a posttranscriptional level.

Reduction of insulin-like growth factor-I (IGF-I) in protein-restricted rats is associated with differential regulation of IGF-binding protein messenger ribonucleic acids in liver and kidney, and peptides in liver and serum.

To investigate the influence of the insulin-like growth factor binding-proteins (IGFBPs) on the nutritional regulation of IGF-I's actions, we compared the gene expression of IGF-I and the six IGFBPs in liver and kidney of protein-restricted (P5) and normally fed (P15) young rats. Using poly(A)+ Northern blot analysis, we observed a decrease in IGF-I messenger RNA (mRNA) at steady state in liver (-50%) and kidney (-60%). The increases in IGFBP-1 mRNA were parallel in these two tissues (liver, 5.7-fold; kidney, 4-fold). In contrast, the expression of the other IGFBP genes exhibited organ-specific regulation during protein restriction; although IGFBP-2 mRNA increased in liver in the P5 group (3-fold), it decreased slightly in kidney (-15%). IGFBP-3 mRNA declined by 30% in liver and was unchanged in kidney. IGFBP-4 mRNA increased by 50-88% in liver and was not modified in kidney. IGFBP-5 mRNA was not detected in liver and was identical in kidney of P15 and P5 rats. IGFBP-6 mRNA was not changed in either liver or kidney during protein restriction. To determine whether the changes in IGFBP mRNAs induced by protein restriction were associated with changes in the respective peptides, IGFBPs in supernatants of liver homogenates and in serum of the same rats were measured by ligand blot analyses. IGFBP-1 and IGFBP-2 Western immunoblot analyses were also performed in serum. By ligand blot, a 45,000 mol wt (M(r)) band (IGFBP-3) decreased in liver and serum of P5 rats, paralleling the changes in liver IGFBP-3 mRNA. A 30,000 M(r) band, consistent with IGFBP-1 and/or IGFBP-2, increased in liver. By immunoblot in serum, IGFBP-1 was only detectable in P5 rats, whereas IGFBP-2 decreased in the P5 group. By ligand blot, a 24,000 M(r) band (IGFBP-4) declined slightly in serum (not detected in liver). Our study shows that protein restriction regulates the expression of four of six IGFBPs in rats, and this regulation is organ specific. The nutritional regulation of IGFBP peptides in biological fluids, in particular serum, seems to involve additional mechanisms.

Reduction of hepatic insulin-like growth factor I (IGF-I) messenger ribonucleic acid (mRNA) during fasting is associated with diminished splicing of IGF-I pre-mRNA and decreased stability of cytoplasmic IGF-I mRNA.

The mechanisms by which fasting decreases liver insulin-like growth factor I (IGF-I) messenger RNA (mRNA) abundance have not been defined completely. In the present study, we have examined the effects of fasting in rats on hepatic IGF-I gene transcription, IGF-I pre-mRNA splicing, and cytoplasmic IGF-I mRNA stability. Using the in vitro nuclear run-on transcription technique, we observed that fasting did not change IGF-I gene transcription activity [76 +/- 32 densitometric units (DU) for fasted vs. 58 +/- 23 DU for control-fed rats; P = 0.1], whereas IGF-binding protein-1 (IGFBP-1) gene transcription, a positive control, was increased more than 2-fold (729 +/- 157 DU for fasted vs. 261 +/- 56 DU for control-fed rats; P < 0.05). This implies that fasting-induced reduction of liver IGF-I mRNA is due to events other than a decreased rate of IGF-I gene transcription. By measuring nonspliced (pre-mRNA) and spliced IGF-I transcripts in liver nuclear RNA using ribonuclease protection assays, we found that IGF-I pre-mRNA was increased in fasted rats (measured as the percentage of beta-actin: 34.0 +/- 5.5% for fasted vs. 8.1 +/- 3.8% for control-fed rats; P < 0.01), whereas spliced IGF-I transcript remained unchanged (measured as the percentage of beta-actin: 60.9 +/- 9.2% for fasted vs. 79.0 +/- 6.2% for control-fed rats; P = 0.75). We then compared this pattern of splicing to IGF-I pre-mRNA splicing in hypophysectomized rats subjected to GH stimulation and to IGFBP-1 pre-mRNA splicing in the same fasting experiment. One hour after GH injection, we observed a coordinate increase in both nonspliced and spliced IGF-I transcripts in liver nuclei of hypophysectomized rats. Fasting increased both IGFBP-1 pre-mRNA and spliced transcript. Taken together, these results indicate that the increase in IGF-I pre-mRNA in liver nuclei during fasting is caused by delayed pre-mRNA splicing, rather than increased IGF-I gene transcription. To examine the possible effect of fasting on hepatic IGF-I mRNA stability, we used an in vitro model of nutrient deprivation (fewer amino acids in culture medium) of rat hepatocyte primary culture. Each of the three major IGF-I mRNA species exhibited a shortened half-life in the amino acid-deprived media. The 7.5-kb IGF-I mRNA, however, was degraded faster than the two smaller IGF-I mRNA species. This may indicate that fasting decreases the stability of liver IGF-I mRNA in vivo. In summary, these results suggest that fasting regulates hepatic IGF-I gene expression mainly at the posttranscriptional level by delaying IGF-I pre-mRNA splicing, which attenuates mature IGF-I mRNA generation, and by accelerating the rate of degradation of IGF-I mRNA in cytoplasm.

Neonatal treatment with monosodium glutamate: effects of prolonged growth hormone (GH)-releasing hormone deficiency on pulsatile GH secretion and growth in female rats.

Administration of monosodium glutamate (MSG) to neonatal rodents produces permanent lesions of hypothalamic arcuate neurons that secrete GH-releasing hormone (GHRH). The present study was intended to determine the consequences of GHRH deficiency on the pulsatile GH secretory pattern and growth in MSG-treated female rats and to compare these effects with those observed in male littermates. Male and female rats were injected with MSG [4 mg/g body wt (BW), sc] or saline (controls) on days 2, 4, 6, 8, and 10 after birth. Immunoreactive GHRH concentrations were decreased in the hypothalamus (by 60%) and in the median eminence (by 95%) of adult male and female MSG-treated rats. In contrast, somatostatin concentrations were unaffected. BW and linear growth were severely impaired in male MSG-treated rats, but in MSG-lesioned females BW was not different from controls, and the attenuation of longitudinal growth was less severe and the obesity more pronounced than in males. These sex differences occurred despite similar reductions (by 55%) in serum insulin-like growth factor I concentrations in male and female MSG-treated rats. MSG treatment also produced decreases in pituitary wt and GH content (by 60%), independent of sex. Pulsatile GH secretion was studied by serial blood sampling of chronically cannulated, freely moving rats. Plasma GH patterns were analyzed by the PULSAR program. Compared to controls, treatment with MSG led to a marked inhibition (by 90%) of GH secretion in both sexes. Significant reductions in GH pulse amplitude (-95%) and pulse duration (-62%) were observed in males, whereas pulse amplitude (-85%), pulse frequency (-67%), and baseline GH concentrations (-80%) were markedly reduced in females. The GH responses to an iv bolus injection of rat GHRH (1 microgram/rat) was severely blunted in both male and female MSG-treated rats. This study demonstrates that GHRH deficiency in female rats results in a marked inhibition of GH pulses, as in males, but also causes severe and sex-specific reductions in GH basal secretion and pulse frequency. These observations suggest that hypothalamic GHRH secretion in female rats is more continuous than in males and is a determinant of the elevated interpulse secretion of GH. Moreover, body wt and linear growth are less severely affected by arcuate lesions in female animals, compared to males. These sex-related differences in growth rates may result in part from the tendency of female MSG-lesioned rats to become more obese than males, and the development of obesity, in turn, may antagonize the factors that tend to slow linear growth.(ABSTRACT TRUNCATED AT 400 WORDS)

Acute down-regulation of the somatogenic receptors in rat liver by a single injection of growth hormone.

Prolonged continuous administration of GH induces somatogenic receptors in rat liver. However, because GH secretion is pulsatile and the effect of acute changes in serum GH concentrations on liver GH receptors is unknown, we measured total (MgCl2-treated homogenates) and free (water-treated homogenates) GH-binding sites in the livers of hypophysectomized (hypox) rats killed between 1 and 24 h after a single sc injection of rat GH (100 micrograms/100 g BW; n = 29). Control hypox rats (n = 10) were studied immediately or 3 h after injection of vehicle. GH injection caused profound decreases in both total and free liver GH receptors, but these changes followed different kinetic patterns. Free receptors declined rapidly (to 17% of control), reaching a nadir at the same time (1 h) as the maximal GH concentration in serum. These free receptors then increased, returning to normal 12 h after GH injection. In contrast, total GH receptors were slightly increased at 1 h, decreased to their minimal value at 6 h (53% of control), and returned to normal at 12 h. Serum immunoreactive somatomedin-C/insulin-like growth factor I concentrations peaked 12 h after GH injection. Total and free liver GH receptors were quantitated in hypox rats that had been injected 3 h previously with doses of rat GH from 2.5-500 micrograms/100 g BW or with vehicle. Both total and free binding sites decreased in a dose-dependent manner; the maximal responses were 40% and 90% below control values, respectively. Half-maximal reductions in GH binding were achieved when 10 micrograms GH/100 g BW were given. These data suggest that a surge of GH in serum leads to a time- and dose-dependent down-regulation of the liver somatogenic binding sites and are consistent with ligand-induced internalization and degradation of the receptor.

Insulin-like growth factor I stimulates the synthesis and release of prolactin from human decidual cells.

Recent studies suggest a role for insulin-like growth factor I (IGF-I) in the regulation of hormone release from placental, gonadal, and pituitary tissues. To examine whether IGF-I may also regulate the release of PRL from human decidual tissue, we have investigated the effect of recombinant human IGF-I on PRL release from monolayer cultures of human decidual cells exposed to IGF-I for up to 4 days. IGF-I (10-1000 ng/ml) stimulated a sustained dose-dependent increase in PRL release (half-maximal concentration, 25 ng/ml) beginning 48 h after initial exposure, but had no effect on the intracellular PRL content. The amounts of PRL released from maximally stimulated cultures on days 3 and 4 were 168 +/- 3% (mean +/- SEM) and 258 +/- 8% of control values, respectively. IGF-I-mediated effects were inhibited by cycloheximide (3.6 microM), suggesting that the increase in PRL was the result of newly synthesized hormone. The increase in PRL release was not due to a generalized effect on protein release, since IGF-I had no effect on the release of trichloroacetic acid-precipitable [35S]methionyl proteins. Radioligand competition studies indicate that the biological actions of IGF-I are mediated through interaction with the IGF-I receptor. Binding of radiolabeled IGF-I to decidual cells in suspension was specific, saturable, and displacable by unlabeled IGF-I, with a potency nearly 10 times greater than that of insulin. Furthermore, exposure of decidual cells to a monoclonal antibody to the IGF-I receptor (alpha-IR3) completely inhibited both IGF-I-mediated PRL release and specific binding of [125I]IGF-I to decidual cells. Since the actions of IGF-I occurred at physiological concentrations, these findings strongly support a role for IGF-I in the regulation of PRL secretion by human decidua.

Differential regulation of insulin-like growth factor I (IGF-I) and IGF binding protein-1 messenger ribonucleic acids by amino acid availability and growth hormone in rat hepatocyte primary culture.

To determine the mechanisms involved in the nutritional regulation of insulin-like growth factor I (IGF-I) production and action, we studied the regulation of IGF-I and IGF-Binding Protein 1 (IGFBP-1) gene expression by GH and amino acid availability in rat hepatocyte primary culture. Hepatocytes were isolated by in situ collagenase perfusion and cultured on Matrigel in serum-free medium containing insulin and hydrocortisone. Rat GH (500 ng/ml) increased IGF-I messenger RNA (mRNA) abundance 6.9-fold at 24 h, as measured by Northern Blot using an IGF-I-specific riboprobe. In contrast, IGFBP-1 mRNA levels were decreased by 41% after 24 h of rat GH treatment. Hepatocytes were incubated for 24 h in three media differing in their amino acid concentrations: 0.2X, 1X, and 5X the normal rat plasma concentration. Amino acid deprivation (0.2X) decreased the abundance of IGF-I mRNAs (-56% after 24 h), whereas amino acid excess (5X) increased it (+70%) in comparison to the 1X medium. In contrast, amino acid deprivation increased IGFBP-1 mRNA abundance (+69%), whereas excess decreased it (-75%). Studies of the interaction between GH and amino acids, accomplished by the simultaneous manipulation of the two, suggest that each factor modulates the IGF-I mRNA and the IGFBP-1 mRNA and protein response to the other. We conclude that the IGF-I and IGFBP-1 genes are regulated in opposite ways by GH and amino acid availability. Our observations suggest that amino acids and GH regulate the production of IGF-I directly and exert indirect effects on IGF-I action by regulating the production of IGFBP-1.