Search

We have examined the effects of infusing recombinant human growth hormone (hGH), insulin-like growth factor-I (IGF-I), the truncated IGF-I analogue, des(1-3)IGF-I, and insulin over a 7-day period in streptozotocin-induced diabetic rats. IGF-I at a dose of 1.05 or 1.08 mg/kg per day in two experiments increased body weight and nitrogen retention above those of vehicle-infused controls to about 30% of the improvement achieved with 25 or 30 units of insulin/kg per day, but only in the second experiment were the differences statistically significant (P less than 0.05). A 2.5-fold higher IGF-I dose, or des(1-3)IGF-I at 1.08 mg/kg per day, gave effects that were approx. 70% of those obtained with insulin. hGH at 1.38 mg/kg per day was not effective. The IGF peptides, unlike insulin, did not ameliorate the diabetic glucosuria. The improvements in nitrogen balance could be accounted for in part by increases in muscle protein synthesis. Muscle protein breakdown, as assessed by 3-methylhistidine excretion, was inhibited by insulin, but not by the IGF peptides. Carcass fat increased substantially following insulin administration. This did not occur with the IGF peptides, suggesting that IGF predominantly stimulates the growth of lean tissue. IGF-I concentrations and IGF-I-binding proteins in plasma were increased by IGF-I, especially at the higher dose, whereas hGH produced only a transient increase in IGF-I. Des(1-3)IGF-I induced binding proteins, but had only a slight effect on measured IGF-I concentrations. We conclude that IGF peptides stimulate muscle protein synthesis and improve nitrogen balance in diabetes without obviously influencing the abnormal carbohydrate metabolism. Moreover, des(1-3)IGF-I is at least as potent as the full-length IGF-I.

The aim of this study was to assess the regulation of insulin-like growth factor-binding proteins (IGFBPs) by IGFs in primary cultures of rat articular chondrocytes (RAC). Employing Western ligand blotting, immunoprecipitation and Northern blot analysis, RAC were found to secrete IGFBP-5 (29 kDa) and IGFBP-4 (24 kDa) as the predominant IGFBPs, as well as IGFBP-2 (32–30 kDa) and IGFBP-3 (43–39 kDa) as the minor species. Treatment of cells with IGF-I and IGF-II resulted in a dose-dependent increase of IGFBP-5 and a small increase in IGFBP-4 in conditioned media (CM). Des(1–3) IGF-I and [Gln6, Ala7,Tyr18, Leu19] IGF-II ([QAYL] IGF-II), which bind to the type 1 IGF receptor but not to IGFBPs, also induced IGFBP-5 peptide, although the increase was less than with IGF-I or IGF-II treatment of RAC. [Leu27] IGF-II, which does not bind to the type 1 IGF receptor but binds to IGFBPs, resulted in little induction of IGFBP-5, while [QAYL-Leu27] IGF-II, which has reduced affinity for both the type 1 IGF receptor and IGFBPs, did not increase IGFBP-5. These data suggest that the increase in IGFBP-5 in CM is modulated by both the type 1 IGF receptor and the interaction between IGFs and IGFBPs. Northern blotting analysis showed that IGF-I, IGF-II and des(1–3) IGF-I treatment of RAC increased steady state levels of IGFBP-5 mRNA, suggesting that the IGF-mediated increase in IGFBP-5 is transcriptionally modulated. Interestingly, the increase in IGFBP-5 peptide levels and mRNA were not parallel, suggesting the possibility of post-translational modifications of IGFBP-5, such as those seen with IGFBP-5 protease. IGFBP-5 protease activity was detectable in untreated CM, whereas treatment with IGF-I and IGF-II partially protected IGFBP-5 from proteolysis. In summary, treatment of RAC with IGF-I and IGF-II results in dose-dependent increases in both IGFBP-5 peptide in the CM and mRNA levels. These changes are mediated by interactions via the type 1 IGF receptor as well as IGFBPs, both transcriptionally and post-translationally.Journal of Endocrinology(1996) 148,355–369

1. Synthetic peptides corresponding to the five, seven, nine and eleven C-terminal amino acids of the tetradecapeptide bombesin as well as bombesin itself and gastrin-releasing peptide have been evaluated in Swiss 3T3 cells in order to define the minimal peptide length needed for biological responsiveness. 2. Gastrin-releasing peptide, bombesin, the undecapeptide and nonapeptide had nearly equipotent abilities to compete for binding of labelled gastrin-releasing peptide to the cell receptors and showed half-maximal competition at 5-10 nM. The heptapeptide and pentapeptide were ineffective. 3. Cross-linking experiments demonstrated specific binding of gastrin-releasing peptide to a 100 kDa receptor subunit. 4. Total cell protein synthesis was stimulated equally by the nonapeptide and longer peptides with a half-maximal effect at 0.5 nM, while a more than 1000-fold higher concentration of the heptapeptide was required to produce a similar response. Comparable results were found when insulin was also present. 5. Neither an inhibition of protein breakdown nor a stimulation of DNA labelling could be demonstrated by bombesin or gastrin-releasing peptide. 6. We conclude that a C-terminal peptide ligand comprising more than seven but no more than nine amino acids is required to achieve high-affinity binding and receptor-mediated responses via the bombesin receptor.

During pregnancy, changes in the IGF axis are associated with changes in maternal metabolism and nutrient repartitioning which are necessary to meet the demands of a growing conceptus. The aim of this study was to assess the IGF axis, maternal weight changes and food intake in female New Zealand White rabbits (n=7) prior to breeding (day 0) and serially throughout pregnancy until term (day 30-31).The total weight of the pregnant does progressively increased from 4·03±0·06 kg (mean ± s.e.m.) on day 0 to 4·47 ±0·07 kg on day 30 (P<0·001). Maternal tissue mass (total weight minus estimated conceptus weight) increased until day 18, plateaued to day 22/23, and then significantly declined. On day 30, the maternal tissue mass was not significantly different from the non-pregnant value, such that the final increase in total weight was due to conceptus growth. Although the does were fed ad libitum, food intake did not change until day 29 when it decreased to approximately 50% of previous intake (P<0·01).Maternal serum IGF-I was 499 ± 32 ng/ml on day 0, reached a peak of 832 ± 160 ng/ml on day 21 (P<0·02), and then declined to 341 ± 49 ng/ml on day 30. In contrast, serum IGF-II increased dramatically from a non-pregnant level of 85 ±14 ng/ml to 16 295 ± 2488 ng/ml on day 23 (P<0·001), and then rapidly declined (3335 ± 954 ng/ml, day 30). Changes in serum IGF-binding proteins (IGFBPs) followed a pattern similar to IGF-II, as assessed by Western-ligand blotting. All IGFBPs, especially the 45–40 kDa IGFBP-3 doublet, increased dramatically between days 12 and 24 of pregnancy, and then declined towards term.In conclusion, we observed unique and dramatic changes in the maternal serum IGF axis that corresponded to periods of maternal weight gain and loss. The tissue source of IGFs and IGFBPs remains undetermined, although it is of note that the time when major changes in the IGF axis were first observed coincided with the time of functional change from yolk sac to placenta in the rabbit.Journal of Endocrinology(1996) 148,121–130

Pregnancy in rodents is associated with important maternal metabolic changes. Early pregnancy is considered to be an anabolic phase of nutrient storing, while in late pregnancy, a catabolic phase develops to help meet the metabolic demands of the rapidly growing conceptus. Similarly, major changes also occur in the IGFs, IGF-binding proteins (IGFBPs) and GH axis. In the rat, maternal serum IGF-I levels increase from early to mid-pregnancy, after which IGF-I levels decline. Conversely, as IGF-I levels decline, pituitary-derived rat GH increases twofold. This coincides with a decrease in IGFBP-3 and the appearance of an IGFBP protease. However, the physiological role of these changes is unclear.The aim of our study was to examine the roles of GH, IGFs and IGFBPs during pregnancy in a unique isolated GH-deficient rat model, spontaneous dwarf rats.Pregnancy in GH-deficient dams resulted in a significantly reduced litter number, and maternal weight gain (day 21–day 1) was reduced by 28% when compared with dams with normal GH levels (GH-normal dams). In the sera of GH-normal dams, IGF-I levels increased 2·6-fold by day 4 of pregnancy and then progressively declined to below non-pregnant levels. Serum IGF-II levels were low and remained unchanged. Western-ligand blot analysis showed that IGFBP-3 was present in non-pregnancy and early pregnancy sera, but declined dramatically after day 12. This decline in IGFBP-3 coincided with the detection of an IGFBP protease. In contrast, in non-pregnancy sera from GH-deficient rats, serum IGF-I concentrations were 10% of the levels seen in GH-normal females, and neither IGF-I nor IGF-II concentrations changed with pregnancy. Furthermore, in GH-deficient dams, serum IGFBP-3 (as assessed by Western-ligand blot analysis) was low in non-pregnancy and early pregnancy sera, and became undetectable by day 12 of pregnancy. The low concentrations of IGFBP-3 in early pregnancy serum from GH-deficient rats coincided with IGFBP-proteolytic activity, and the decline in serum IGFBP-3 after day 16 was the result of increased protease activity.In conclusion, isolated GH deficiency results in: (1) reduced maternal weight gain and correspondingly smaller litter sizes; (2) low and unchanged maternal serum IGF-I levels; (3) low, but declining, serum IGFBP-3 levels; and (4) activity of IGFBP protease(s) detectable in early and late pregnancy, which may modulate the bioavailability and bioactivity of IGFs by regulating IGFBP-3.Journal of Endocrinology(1993) 136,479–489

The ratio of the serum concentrations of insulin-like growth factors (IGF) to IGF-binding protein (IGFBP)-3 is highly correlated (Baxter & Martin 1986). During pregnancy in the baboon, this ratio is perturbed; serum IGFBP-3 concentrations increase 10-fold, yet IGF-I levels are unaltered and IGF-II is increased only 2-fold (Giudice et al. 1993). The aims of this study were to determine the molecular distribution of IGFBP-3 and to identify the tissue source and form(s) of IGFBP-3 during pregnancy in the baboon. Serum of non-pregnant and pregnant baboons, and conditioned media of decidua and placental explant cultures were characterized using neutral size-exclusion chromatography in combination with Western ligand blot, Western immunoblot, an IGFBP-3 radioimmunoassay (RIA) and an IGFBP-3 protease assay. Localization of immunoreactive IGFBP-3 was determined by immunocytochemistry, and expression of IGFBP-3 mRNA in the placental and decidual explants was examined by Northern blot analysis.RIA confirmed that immunoreactive IGFBP-3 is increased 10-fold in pregnancy serum compared with non-pregnancy serum. Size-exclusion chromatography combined with an IGFBP-3 RIA revealed that, unlike non-pregnancy serum where 70% of the immunoreactive IGFBP-3 elutes in the 150 kDa ternary complex, equal amounts of immunoreactive IGFBP-3 were measured in pregnancy serum in the ≤150 and 60 kDa IGFBP regions. Western analysis revealed that non-pregnancy serum contained predominantly a 45–40 kDa IGFBP-3 doublet and a 28 kDa immunoreactive form of IGFBP-3, while in pregnancy serum IGFBP-3 existed as a 45–40 kDa doublet, as well as 26–28 kDa and 18 kDa immunoreactive forms. These alternative forms of IGFBP-3 were not attributable to detectable IGFBP-3 protease activity.To identify the source of the increased serum levels of IGFBP-3 during pregnancy, we examined explant culture media of baboon decidua and placenta. Size-exclusion chromatography combined with RIA and Western analysis revealed that: (1) more immunoreactive IGFBP-3 was produced by decidual cultures than by placental explants, but less 45–40 kDa IGFBP-3 was present in decidua; (2) the immunoreactive forms of IGFBP-3 detected in decidua were similar to those found in maternal serum; (3) placental explants secreted only 45–40 kDa IGFBP-3 in culture. IGFBP-3 was immunohistochemically localized to the cells of placental villi, and to the perinuclear region of the decidual cells and staining for IGFBP-3 was more intense in the decidua than in the placenta.Northern analysis of the explant cultures revealed two IGFBP-3 mRNA transcripts of 2·4 and 1·7 kb in both decidua and placenta which may account for the different immunoreactive forms of IGFBP-3 detected in the baboon. However analysis of non-pregnancy liver also revealed two IGFBP-3 transcripts of 2·4 and 1·7 kb. These data suggest that the two transcripts are not solely pregnancy-associated and levels of protein may be the reason for detection of multiple immunoreactive IGFBP-3 fragments in pregnancy.Journal of Endocrinology(1995) 147,449–461

The ability of insulin-like growth factor-I (IGF-I) to protect against losses of body protein during periods of dietary nitrogen restriction has been evaluated in young rats. Recombinant human IGF-I was administered by osmotic pumps at dose rates of 0, 1·2 or 2·9 mg/kg per day over a 7-day period beginning with the transfer of animals from an 18% to a 4% protein diet. A fourth group received the potent truncated IGF-I analogue, des(1–3)IGF-I, at a dose of 1·2 mg/kg per day over a comparable 7-day period. Plasma IGF-I levels were reduced by 60% following nitrogen restriction, a reduction that was partly prevented by IGF-I administration, especially at the higher dose, but not measurably by des(1–3)IGF-I. The major IGF-binding protein circulating in blood, IGFBP-3, demonstrated a similar pattern of change.A significant (P<0·05) protection of body weight was achieved in the low dose IGF-I and des(1–3)IGF-I groups, but only after differences in food intake had been eliminated by analysis of covariance. Nitrogen balances were not significantly different unless analysis of covariance was used to adjust for the nitrogen intakes, whereupon all treatment groups showed improved balance, especially the animals treated with the low IGF-I dose and des(1–3)IGF-I (both P<0·01). The rate of muscle protein breakdown calculated from the urinary excretion of 3-methylhistidine was not significantly altered by the treatments, but fell progressively throughout the 7 days. The fractional rate of muscle protein synthesis measured on the final day was increased by 31, 26 and 21% respectively by the low and high doses of IGF-I and by des(1–3)IGF-I. Organ weights (g/kg body weight) showed no effects of IGF-I treatment except for 16% increases in the weight of kidneys in the high dose IGF-I and the des(1–3)IGF-I groups. Carcass analyses demonstrated higher water and lower fat contents (all P< 0·01) in the same groups. These results suggest that exogenous IGF-I and especially des(1–3)IGF-I can partly protect body protein reserves during nitrogen restriction.Journal of Endocrinology(1991) 128,97–105

Insulin-like growth factor-binding proteins (IGFBPs) in rat serum, lymph, amniotic fluid and cerebrospinal fluid (CSF), and in rat cell-conditioned media were characterized using a combination of gel-permeation chromatography, Western immunoblots and Westernligand analysis. Adult serum and abdominal lymph contained a 200 kDa IGFBP (the putative type-II IGF receptor) and a 150 kDa IGFBP that contained subunits of 40–50 kDa aligning with porcine IGFBP-3 on Western-ligand blots. In addition, both fluids contained the smaller IGFBPs: a 30 kDa IGFBP which was immunoreactive with IGFBP-2 antiserum, a 28 kDa IGFBP which electrophoresed with human IGFBP-1, and a 24 kDa IGFBP. In contrast, fetal serum and amniotic fluid lacked the 150 kDa and the 28 kDa IGFBPs. CSF contained only a 30 kDa IGFBP, but this was not IGFBP-2. Several IGFBPs were detected in media conditioned by liver, bone and muscle cells. Liver-derived cells and some hepatoma cell lines produced similar patterns upon ligand blot analysis, i.e. IGFBPs of 30 kDa (which reacted with IGFBP-2 antiserum), 28 kDa and 24 kDa. A hepatoma cell line, HTC, and a smooth muscle cell line contained only an IGFBP of 26 kDa. Skeletal muscle-derived cells (L6 myoblasts) produced a 28 kDa, a 26 kDa and a 24 kDa IGFBP. Both calvarial osteoblasts and osteogenic sarcoma cells produced an IGFBP of 30 kDa that cross-reacted with IGFBP-2 antisera. In addition, osteogenic sarcoma cells produced a 28 kDa and a 24 kDa IGFBP. These results allow us partially to classify and to compare the IGFBPs in rat fluids and those produced by cultured cells.Journal of Endocrinology(1990) 127,391–400

Insulin-like growth factor-I (IGF-I), IGF-II and IGF-binding proteins (IGFBP) were examined in rat serum during pregnancy and lactation. IGF-I concentrations determined after acid column chromatography of serum were low during the last third of pregnancy. IGF-II was undetectable in pregnant and non-pregnant rats. IGF-binding protein (IGFBP) concentrations, measured as high molecular mass activity in the IGF-I RIA and the IGF-II RRA of acid column fractions, paralleled the changes observed with IGF-I. Western ligand blot analysis of serum from non-pregnant rats revealed a 40–50 kDa IGFBP aligning with IGFBP-3, a smaller 28–30 kDa doublet and 24 kDa IGFBP. Serum from rats in late pregnancy lacked IGFBP-3, whereas the smaller IGFBP persisted during late pregnancy. IGFBP-3 reappeared in postpartum animals. The fall in serum IGF-I is consistent with a maternal catabolic state during late pregnancy which may maximize substrate availability for the developing fetus.Journal of Endocrinology(1990) 127,383–390

The aim of this study was to assess the molecular size distribution of insulin-like growth factors (IGFs) complexed to IGF-binding proteins (IGFBPs) in serum of non-pregnant and pregnant women. Sera were fractionated on a size-exclusion column at pH 7·4 to resolve different IGF–IGFBP complexes from unbound IGFs. Each fraction was further chromatographed on a size-exclusion column under acid conditions to dissociate IGFs from binding proteins prior to measurement of the IGF content of the complexes. The IGF-containing fractions from the acid column were assayed specifically for IGF-I and IGF-II. Serum pooled from pregnant women contained more IGF-I and IGF-II than serum from non-pregnant women. In both groups most of the IGF-I and IGF-II was found in a large (150 kDa) complex in serum and the remainder was present in complexes eluting in the 40–50 kDa region at pH 7·4. Some free IGF-I was also detected. Serum fractions collected by size-exclusion chromatography at pH 7·4 were also analysed by Western-ligand blotting to characterize the IGFBPs in the two main IGFBP size classes. In serum pooled from non-pregnant women, the 150 kDa IGF–IGFBP complexes contained a 40–50 kDa IGFBP doublet following Western-ligand blot analysis. This IGFBP co-migrated with a pure IGFBP-3 standard. IGFBPs of 34, 28 and 24 kDa all contributed to the 40–50 kDa IGF–IGFBP complexes of the pH 7·4 chromatograph. In serum pooled from pregnant women, the 40–50 kDa IGFBP-3 doublet and 34 kDa IGFBP were not evident in any fractions from the pH 7·4 column. Thus, although the amounts of IGF-I and IGF-II in 150 kDa IGF–IGFBP complexes were increased during late pregnancy, IGFBP-3 measured by Western-ligand blot analysis of this complex was greatly diminished. The large amount of IGF-I and IGF-II in 150 kDa complexes is strong evidence for the presence of IGFBP-3 in serum during pregnancy, because IGFs and IGFBP-3 are normally present in equimolar amounts in this complex. We suggest that during pregnancy in women, the IGFBP-3 in the 150 kDa complex becomes unstable and this may explain the failure of the Western-ligand blot to detect IGFBP-3.Journal of Endocrinology(1991) 131,491–497

During late pregnancy in the rat, circulating levels of insulin-like growth factor-I (IGF-I) and some IGF-binding proteins (IGFBP) decline. The aim of the present study was to determine the relationship of GH to circulating IGF and IGFBP in the late-pregnant rat and to examine the effects on maternal, fetal and placental growth of preventing the decline in serum IGF and IGFBP concentrations. During the first 9 days of pregnancy, IGF-I concentrations increased from 340 to 500 μg/l. Recombinant human (rh) GH at 2·4 mg/kg per day and rhIGF-I at 1·4 mg/kg per day were infused into pregnant rats via osmotic mini pumps during the second half of pregnancy. After pump implantation on day 11 of pregnancy, only IGF-I infusion significantly increased circulating IGF-I. A maximum IGF-I concentration of 907 μg/l was measured on day 14 during treatment with IGF-I, after which the serum concentration decreased to 510 μg/l by day 20 of pregnancy. The serum IGFBPs were examined using a Western ligand blot technique. Infusion of neither GH nor IGF-I returned the IGFBPs to non-pregnant levels. Administration of IGF-I slightly increased IGFBP-3 and a smaller 32 kDa IGFBP at days 17 and 20 of pregnancy.Neither fetal nor placental weight was significantly different between treatment groups. However, administration of IGF-I significantly increased maternal weight gain during the 10-day treatment period. Thus, pregnant rats infused with IGF-I gained 99±4 g (mean ± s.e.m., n= 10) compared with rats treated with GH or vehicle which gained 72±4 g (n= 9) and 77±4 g (n= 10) respectively. The increase in maternal weight after administration of IGF-I was not due to increased litter size, fetal or placental weight. The increased maternal weight gain after IGF administration, without affecting fetal and placental weights, suggests a modification in the mode of maternal nutrient repartitioning during late pregnancy.Journal of Endocrinology(1991) 130,395–400

During pregnancy there are dramatic changes in the endocrine and metabolic status of the mother: growth hormone (GH) is an important regulator of growth and development. A proportion of GH is bound by specific GH binding proteins (GHBP) that closely resemble the GH receptor (GHR). In the rodent both GHBP and the GHR are considered to be GH dependent, and consequently during pregnancy the increase in serum GH is associated with an increase in GHBP. To examine whether an increase in maternal GH is obligatory for elevation of maternal GHBP or GHR during pregnancy, we used a unique GH-deficient (GHD) strain of rats, to avoid the methodological complications of hypophysectomy and assessed serum GH, GHBP and hepatic GHR binding during the course of pregnancy. In GH normal rats, serum GH concentrations increased twofold and GHBP levels increased threefold; there was no change in hepatic GHR binding. In GHD rats, serum GH concentrations were low and did not increase during pregnancy. Nonetheless, levels of both serum GHBP and hepatic GHR binding increased to that measured in normal rats. Thus, an increase in maternal GH concentration is not required for the gestational upregulation of maternal GHBP or hepatic GHR binding, suggesting that other hormones may be essential in modulating the GH axis during pregnancy.

The major IGFBPs in CSF are IGFBP-2 and IGFBP-4. whereas IGFBP-3 is a minor component in CSF of healthy subjects. We investigated IGFBP-3 levels in CSF from patients with brain tumors, leukemia or meningitis. IGFBP-3 was measured by radioimmuno-assay with aIGFBP-3g I, a rabbit polyclonal antibody. Further, as proteolysis of IGFBP-3 is purt of the modulation of IGF activity IGFBP-3 fragmentation was quantified by densitometric analysis of [125I]IGFBP-3 protease assays. We examined CSFs of 24 children with malignant brain tumors. 18 children with leukemia and 13 children with meningitis. CSFs of 38 children, who received a lumbal puncture in order to exclude meningitis, were used to define a normalative range for IGFBP-3 concentration and IGFBP-3 protease activity in normal CSF. Elevated IGFBP-3 concentrations were found in CSF of 17 of all 24 (71%) brain tumor patients and 7 of 8 (87%) brain tumor patients who had microscopically detectable malignant cells in CSF. 13 of 14 (93%) patients with medulloblastoma or ependymoma and all 7 medulloblasioma/ependymoma patients with malignant cells in CSF had elevated IGFBP-3 concentrations in CSF. IGFBP-3 protease activity in CSF was elevated in 15 of 17 (88%) patients with histological high grade (WIIO III°/IV°) brain tumors. 5 of 6 (83%) patients with acute leukemia and microscopically detectable malignant cells in CSF at the time of diagnosis showed elevated IGFBP-3 concentrations in CSF that normalized under chemotherapy. leukemia patients without any detectable malignant cells in the CSF had normal IGFBP-3 concentrations in CSF. We conclude that in CSF of children with high grade malignant brain tumors or CNS leukemia. IGFBP-3 is elevated. We hypothesize that this phenomenon could be caused by local production of IGFBP-3 by the brain tumor tissue and secretion into CSF or by local secretion of IGFBP-3 by malignant cells that spread into CSF. This hypothesis is supported by the observation that only 2 of 13 (15%) meningitis patients had slightly elevated IGFBP-3 concentrations in CSF while high numbers of non malignant inflammatory cells were present in all cases. Further studies will analyse the origin of elevated IGFBP-3 levels in CSF of tumor patients and will show whether these findings are of diagnostic or prognostic value.

The insulin-like growth factors (IGF) arc potent polypeptide mitogens that are carried in the blood, primarily by a GH-dependent protein, IGF binding protein-3 (IGFBP-3). This IGFBP can be fragmented by IGFBP-proteases, subsequently altering the affinity for IGFs. We have characterised the serum and urine IGFBP profiles of children both pre- and post GH treatment. Techniques used were an IGFBP-3 RIA, Western ligand blot (WLB) and an IGFBP-3 protease assay. In the sera of pretreatment GHD children, IGFBP-3 levels were low by both RIA and WLB analysis, and there was no protease activity. Serum concentrations of IGFBP-3 were corrected following GH administration. In the urine, we have established a normal range of IGFBP-3 and IGFBP-3 protease activity in subjects between 5-44 years. Urinary IGFBP-3 (uIGFBP-3) is age-dependent, increasing from 40 ng/mg Cr to 60 ng/mg Cr at age 11-15, after which levels decline to 18 ng/mg Cr by 25 years. Furthermore, little uIGFBP-3 protease activity is delected. In pretreatment GHD subjects, immunoreactive uIGFBP-3 as determined by RIA were generally higher than age-matched controls, although there was some overlap with the normal range. Following GH therapy, uIGFBP-3 levels declined to within the normal range. This was in striking contrast to serum, where serum IGFBP-3 levels increased with GH administration. Although immunoreactive uIGFBP-3 were high, IGFBP-3 was undetectable by WLB. This difference was due to significant protease activity in these subjects, especially in the younger age groups. Following GH therapy, uIGFBP-3 became detectable by WLB and the protease activity diminished. Thus, in the urine of GHD children, an IGFBP-3 protease activity was delected that was apparently GH dependent. Furthermore, the urinary IGFBP-3 protease activity is a serine-type protease which is not metal ion dependent. The source of the IGFBP-3 protease is uncertain, although its appearance solely in the urine of GHD patients is a unique finding. Further assessment of this proteolytic activity may provide insight as to the renal function and clearance in GHD patients.