Structure and properties of members of the hGH family: a review.

This review will summarize the properties of five variant forms of human growth hormone: a disulfide dimer, a glycosylated form, 20 kD-hGH, and two peptides made up of portions of 22 kD-hGH. The two pituitary peptides (hGH(1-43) and hGH(44-191)) have, respectively, insulin-potentiating and anti-insulin properties. Both have been detected in serum. The shorter peptide may prove to be useful in decreasing the amount of exogenous insulin required by diabetics. The larger, strongly anti-insulin peptide, may be involved in diabetic retinopathy. We believe that this peptide is the long sought after diabetogenic substance of the pituitary gland.

Human growth hormone fragment (hGH44-91) produces insulin resistance and hyperinsulinemia but is less potent than 22 kDa hGH in the rat.

A 17 kDa fragment of human growth hormone (22 kDa hGH), identified as hGH44-191, has lower binding affinity for growth hormone receptors (GHRs), but has been reported to be more potent in producing glucose intolerance in yellow obese mice. Out aim was to investigate this anomaly by comparing acute development of hyperinsulinemia and insulin resistance ("diabetogenic activity") during hGH44-191 or 22 kDa hGH infusion in normal rats. Fasted awake make rats (350-370 g) were infused via a carotid cannula with saline (CON), 22 kDa hGH (at 0.125 micrograms/min), or hGH44-191 (at 0.64 or 0.32 micrograms/min) for 5.75 h. Over the last 2 h, a euglycemic hyperinsulinemic clamp (insulin infusion rate 0.25 U/kg/h) was performed. After 3.75 h infusion, 22 kDa hGH at 0.125 and hGH44-191 at 0.64 micrograms/min produced basal (preclamp) hyperinsulinemia compared to CON. During the clamp, insulin resistance was consistently produced by 22 kDa hGH at 0.125 and hGH44-191, at 0.64 micrograms/min compared to CON. Using specific radioimmunoassays for 22 kDa hGH and hGH44-191, we determined that under conditions of equivalent diabetogenic activity, molar circulating levels of hGH44-191 were 50-60-fold higher than 22 kDa hGH. It was concluded that whereas 22 kDA hGH and hGH44-191 are both capable of generating acute hyperinsulinemia and insulin resistance in the normal rat, the diabetogenic potency of hGH44-191 is not enhanced compared to 22 kDa hGH, and that diabetogenic potency is in accord with the reported lower binding affinity of hGH44-191 to the GHR.

Human growth hormone (hGH)-(44-191), a reportedly diabetogenic fragment of hGH, circulates in human blood: measurement by radioimmunoassay.

The purpose of these experiments was to determine whether a 17-kilodalton (K) fragment of human GH (hGH), hGH-(44-191), a peptide 10 times more potent than intact hGH in causing glucose intolerance in an animal model, circulates in human blood. Analysis of pituitary extracts and sera by Western blotting revealed the presence of a 17K hGH-immunoreactive band in both samples. Monoclonal antibodies to recombinant hGH-(44-191) cross-reacted with the 17K hGH-immunoreactive band of the pituitary. A RIA specific for hGH-(44-191) was developed using recombinant hGH-(44-191) as the tracer and standard, and a mouse anti-hGH-(44-191) serum as the source of antibody. The RIA detected hGH-(44-191) in human sera and pituitary extracts. The concentration of hGH-(44-191) in pituitary glands was 1/100th to 1/500th of the hGH concentration, but in serum its concentration averaged 1-2 times higher than that of hGH. Serum hGH-(44-191) concentrations were higher in pregnant than in nonpregnant women. The data document the existence of hGH-(44-191) in human pituitary gland and serum, and provide initial evidence that it may be physiologically produced. As its concentration is affected by the physiological state of the individual, it may play a role in the expression of the physiological and pathological actions of hGH.

Potential abnormalities in molecular forms of growth hormone in schizophrenia: a pilot study.

Growth hormone has been investigated in numerous studies involving patients with schizophrenia but has been measured only by radioimmunoassay (RIA). There have been no consistent abnormalities differentiating patients with schizophrenia from normal controls. In the current study, growth hormone (GH) variants were measured by Western blotting techniques, which resulted in the quantitation of 4 GH size variants: 27K (27,000 Daltons), 22K, 20K, and 17K. In the entire sample of 17 schizophrenic subjects, all GH variants were significantly higher than in the 14 normal controls. While there were no significant differences between the 2 groups in RIA GH values, the RIA values were generally higher in the schizophrenic group. In a subset of 12 schizophrenic patients whose RIA values were approximately equal to the controls, both the 27K and 22K GH variants remained significantly higher in the patient group. In the schizophrenic group, none of the GH variants or RIA GH changed significantly after 1 week of treatment with neuroleptic medication. These preliminary results suggest that certain GH forms may be elevated in schizophrenia, but further studies are needed.

Growth hormone and prolactin variants in normal subjects. Relative proportions in morning and afternoon samples.

There are multiple molecular forms of both growth hormone (GH) and prolactin (PRL). Traditionally the two hormones have been measured by radioimmunoassay (RIA) techniques. Recently, several molecular variants of these hormones have been discovered using Western blotting techniques: four GH size variants, 27K GH, 22K GH (the classical form), 20K GH (an alternatively-spliced form), and 17K GH, and two PRL structural variants, a glycosylated (G-PRL) and a nonglycosylated form. In this study, we measured these GH and PRL variants in 18 normal subjects in the morning in a fasting state and in the afternoon in a non-fasting state. Contrary to expectations, the predominant serum GH form in both morning and afternoon samples was found to be 17K, not 22K GH, accounting for 82-89% of the total circulating GH. The predominant serum PRL form was found to be the nonglycosylated variant, constituting 83-84% of the total circulating PRL. None of the GH or PRL variants were significantly different when comparing morning to afternoon samples. These results provide, for the first time, evidence for the existence of two new GH-immunoreactive components in human sera, the 17K and 27K GH, the former in proportions often higher than those of the classical 22K GH, and argue for the need to measure them individually.

Differential effects of glycosylated and nonglycosylated prolactin on islet cell division and insulin secretion.

A growing body of evidence suggests that prolactin (PRL) is a potent regulator of the structure and function of the islets of Langerhans, but PRL is a polymorphic hormone that exists in several molecular forms. Therefore, it is important to know whether glycosylated PRL, a major structural variant of the hormone in several species, has an effect different from that of the nonglycosylated PRL on islet function. This in vitro study examined the differential effects of glycosylated and nonglycosylated porcine PRL on cell division and insulin secretion from neonatal rat islets, and compared these results with those produced by homologous rat PRL. The nonglycosylated porcine PRL produced modest stimulation of cell division and insulin secretion from rat islets, but glycosylated porcine PRL had no significant effects. The stimulations produced by nonglycosylated porcine PRL were much weaker in comparison to those produced by the homologous rat PRL. The results show differential effects of the two structural variants of porcine PRL on rat islet function. Although these findings must be confirmed in a homologous system, the results present the possibility that the structural form of the PRL molecule available to the islet tissue may be crucial for its normal functioning.

Differences in GH secretion from individual somatotropes in rats genetically selected for fast and slow growth.

A reverse hemolytic plaque assay was used to examine effects of selection for fast (F) and slow (S) growth on growth hormone (GH) secretion by individual somatotropes. Anterior pituitaries (AP) from 32 male Charles River CD strain rats selected for F and S growth for 20 generations were used. Four rats per line were used at 4, 6, 8, or 10 wk of age. Body weight (P < 0.0001) of F rats was greater compared with S rats. AP (P < 0.05) were heavier at 8 and 10 wk of age in F compared with S line rats. Percentages of GH-secreting cells were unaffected by age (range = 32.7-35.5%) and line [F = 33.1 +/- 1.2% (SE) vs. S = 34.5 +/- 1.2%] or by human GH-releasing factor (hGRF). At 8 and 10 wk, mean plaque-forming area was greater (P < 0.0001) in F compared with S rats under both nonstimulated (2,704 +/- 202 vs. 1,648 +/- 202 microns2) and hGRF-stimulated secretion (4,503 +/- 202 vs. 2,682 +/- 202 microns2). Results indicate that differences in growth observed in the two lines may be due to a greater secretory capacity of individual somatotropes rather than to a greater percentage of somatotropes or sensitivity of somatotropes to secretagogue.

Prolactin variants.

Prolactin (PRL) is one of the most versatile hormones of the pituitary in terms of biologic actions, but how a single molecule is able to evoke so many different responses in the organism is not known. Research in recent years has uncovered a surprising degree of structural polymorphism for PRL, the different forms having variable biopotencies. Such findings lend credence to the hypothesis that the molecular heterogeneity of PRL is one of the mechanisms for creating diversity in the biologic actions of this hormone.

Dopa accumulates in the hypothalamic-hypophysial portal vessels and is taken into the anterior pituitary of NSD-1015-treated rodents.

DOPA was measured in the anterior pituitary and hypothalamic-hypophysial portal blood after treatment with NSD-1015, a DOPA decarboxylase inhibitor. NSD-1015 caused DOPA to accumulate in the anterior pituitary of mice and rats, and increased DOPA in the hypothalamic-hypophysial portal blood of rat. Serum prolactin was also increased. Interruption of the anterior pituitary blood supply from the hypothalamic-hypophysial system by cannulation of the entire pituitary stalk eliminated the NSD-1015-induced DOPA accumulation in the rat pituitary. We conclude that DOPA can be taken into the anterior pituitary from the portal blood of NSD-1015-treated rodents and that the anterior pituitary lacks tyrosine hydroxylase activity in both mice and rats.

Isolation and biochemical properties of four forms of glycosylated porcine prolactin.

Four isoforms of glycosylated prolactin (G-pPRL) were isolated from porcine pituitary glands by affinity chromatography and concanavalin A-Sepharose, based upon differences in their affinity for the lectin. Structural analysis indicated differences in the carbohydrate units of the four G-pPRLs. N-glycanase treatment cleaved the oligosaccharide from the G-pPRLs, establishing N-linked glycosylation. The binding of G-pPRLs to receptors from lactating rabbit mammary glands was only 3-8% that of nonglycosylated pPRL (NG-pPRL). The immunological crossreactivity of the G-pPRLs varied from 36 to 65% that of NG-pPRL. When tested in the pigeon crop sac bioassay, G-pPRLs were only 11-40% as active as NG-pPRL. The metabolic clearance rate of one of the G-pPRLs was slower and another faster than that of NG-pPRL. We conclude that there are several forms of G-PRL of variable immuno- and bio-potencies in the porcine pituitary, and that the current radioimmunoassay for the hormone does not measure the actual bioactivity.

Ontogeny of glycosylated and nonglycosylated forms of prolactin and growth hormone in porcine pituitary during fetal life.

Although prolactin (PRL) and growth hormone (GH) were long considered to be nonglycoprotein hormones of the pituitary, glycosylated forms of these hormones have nevertheless been discovered in recent years. We determined the ontogeny of glycosylated and nonglycosylated PRL and GH during the fetal life of the pig, an animal in whose pituitary the glycosylated variant of PRL has been found in high (40%) concentrations. Swine fetuses of both sexes from lean and obese animals of Duroc x Yorkshire crosses were examined at 60, 75, 90, and 105 days of age. No appreciable differences related to sex or phenotype were noted in any of the parameters measured; therefore, data for all animals within an age group were combined. Such averages revealed considerable amounts of GH in the fetal pituitary as early as 60 days of age, whereas PRL, although detectable by radioimmunoassay and immunoblotting at all ages tested, was not present in significant amounts until 105 days of age. From its first appearance, however, almost 70% of the PRL synthesized in the fetal pituitary was of the glycosylated type. In contrast to PRL, both the glycosylated and nonglycosylated forms of GH showed a steady rate of increase throughout the observation period. The immunoblotting analyses also revealed in the fetal pituitary several intensely staining 8- to 12-kDa PRL-immunoreactive peptides of unknown identity. The occurrence of significantly greater concentrations of glycosylated PRL than of non-glycosylated PRL in the fetal pituitary during late gestation offers new possibilities for the role of this hormone in the development of swine fetus.

Changes in the glycosylated and nonglycosylated forms of prolactin and growth hormone in lean and obese pigs during pregnancy.

The concentrations of glycosylated (G-PRL) and nonglycosylated (non-G-PRL) forms of PRL and GH were measured during pregnancy in pigs of lean and obese (high backfat thickness) lines. Pregnant sows of the two genetic lines were killed, in groups of five to eight, at 60, 75, 90, and 105 days of gestation, and their pituitary glands and plasma were analyzed for the two hormones by immunoblotting, lectin-binding, and RIA techniques. In both lean and obese pigs, pituitary concentrations of G-PRL and non-G-PRL increased with advance in pregnancy, but there were no significant changes in either form of pituitary GH. Plasma concentrations of radioimmunoassayable PRL also increased with advance in pregnancy, with no consistent changes in serum GH concentrations. The dominant PRL constituent in plasma during the second half of pregnancy was G-PRL, and its concentration either increased or remained constant with advance in pregnancy. In contrast, plasma non-G-PRL concentrations decreased with advance in pregnancy in both lines of pigs, resulting in a steady rise in the plasma G-PRL/PRL ratio toward term. Compared to lean pigs, obese pigs had less radioimmunoassayable PRL and GH in their plasma and less GH (glycosylated as well as nonglycosylated) in their pituitary glands, but obese pigs had more G-PRL in their pituitary glands than lean pigs, and their plasma G-PRL levels tended to be higher and non-G-PRL levels lower than those of lean pigs. Pituitary concentrations of non-G-PRL in the two lines of pigs were similar. Overall, the results show a preponderance of G-PRL over non-G-PRL in the plasma of pregnant sows, with a preferential secretion of the glycosylated form during the latter half of pregnancy. Furthermore, they indicate a prevalence of higher G-PRL/PRL ratios in the pituitary glands of obese than lean pigs. These findings raise the possibility of a functional role for the glycosylated variant of PRL in the initiation and/or maintenance of events associated with pregnancy and obesity in the pig.

Concentrations of glycosylated prolactin in the pituitary gland and plasma of lean and obese barrows.

We measured glycosylated PRL (G-PRL) in the pituitary gland and plasma of lean and obese barrows (castrated male pigs) by immunoblotting. We found that G-PRL constituted 40% or more of total monomeric PRL in the pituitary gland of these animals. Furthermore, pituitary G-PRL concentrations in obese barrows averaged 58% higher than in those of lean controls. The non-glycosylated PRL concentrations were also higher in the pituitaries of obese barrows in comparison to lean controls, but the magnitude of the difference was smaller (42%). Plasma concentrations of the two forms, or total immunoreactive PRL measurable by radioimmunoassay, did not differ significantly between the two groups. The pituitary data suggest that the state of PRL secretion is altered in obese barrows, and that the alteration is more pronounced for G-PRL than for the non-glycosylated counterpart.

Antibodies to newly recognized murine 13-18 KDa pituitary peptides crossreact with growth hormone and prolactin from several species, including man.

Recently we identified five novel peptides of Mr 13,000 to 18,000 (designated P13, P14, P16, P17, and P18 according to approximate Mr) in the anterior pituitary gland of rat and man that appeared related to GH and PRL in regulation and structure. We have now raised polyclonal antibodies in the rabbit to four of these peptides--P13, P14, P17 and P18--isolated from rat anterior pituitary; the rabbit injected with P16 did not produce antibodies. Besides reacting with their respective immunogens, antisera to all four peptides crossreacted, quite unexpectedly, with human GH and with human, porcine, and ovine PRL. Antiserum to P17, in addition, crossreacted very strongly with rat PRL, while P18 antiserum crossreacted not only with human GH but also with its 20K and cleaved variants. These results provide strong evidence for the structural relatedness of these peptides to GH and PRL, and raise the possibility that they may be related functionally as well.

Three growth hormone- and two prolactin-related novel peptides of Mr 13,000-18,000 identified in the anterior pituitary.

Electrophoretic analysis of murine anterior pituitary extract revealed five major proteins of Mr 13,000-18,000 (designated P13, P14, P16, P17, and P18 according to Mr), three of which, P16, P17, and P18, were markedly influenced by estradiol benzoate and perphenazine in a manner similar to that of growth hormone, and two, P13 and P14, to that of prolactin. Tyrosine peptide mapping showed partial resemblance of fingerprints for P16 and P17 (and possibly P18) to those for growth hormone, and of P13 and P14 to those for prolactin. Both P14 and P18 bound to Concanavalin A. None of the peptides crossreacted with antibodies to growth hormone or prolactin. The concentrations of P13 and P14 in pituitaries from lactating rats and in a prolactinoma were distinctly higher than normal. All five peptides were secreted into the medium during the in vitro incubation. These results suggest that P16, P17 and P18 are growth hormone- and P13 and P14 prolactin-related secretory proteins of the pituitary.

Glycosylated prolactin in porcine plasma: immunoblotic measurement from birth to one year of age.

As much as 40% of PRL in the pituitary gland of the pig is glycosylated. To help determine the physiological significance of this structural variant of PRL, we have measured glycosylated PRL (G-PRL) in the plasma of growing pigs from birth to 1 yr of age. An immunoblotting method developed originally for human plasma was used. With some modifications, the method could detect G-PRL in as little as 0.2 ml porcine plasma. Concanavalin-A affinity chromatography confirmed the glycoprotein nature of the plasma G-PRL band. Quantitative estimates of the immunoblotting results revealed marked differences with age in the secretion of the two monomeric forms of PRL. G-PRL concentrations averaged 138% higher than those of non-G-PRL between birth and 2 months of age, but lower thereafter. Chronologically, both forms displayed similar patterns between birth and 2 months, the concentration remaining unchanged or decreasing slightly. After 2 months, however, concentrations of non-G-PRL increased markedly; the increase was characterized by great fluctuations. G-PRL concentrations, on the other hand, increased only moderately, but the increase was consistent until the end of the study period. The results demonstrate the circulating nature of G-PRL in the pig and suggest that the variant may play a physiological role in the growth and development of the animal.

Effects of gonadal steroids on somatotroph function in the rat: analysis by the reverse hemolytic plaque assay.

The mechanism by which gonadal steroids modulate GH secretion is not known. We have used the reverse hemolytic plaque assay to examine whether gonadal steroid-induced modulation of GH secretion is effected by changes in the population of somatotrophs and/or alterations in their secretory properties. Two groups of Sprague-Dawley rats were studied: group 1 (n = 6) comprised male (M), castrate (Cx), and testosterone-replaced castrate male (Cx + T) rats and group 2 (n = 5) consisted of male (M), female (F), and 17 beta-estradiol-replaced castrate male (Cx + E) rats. The number of plaque-forming cells (expressed as both absolute number and a percentage of all cells) was determined, and secretory status was assessed by measuring plaque areas in response to 0, 0.01, 0.1, 1, 10, and 100 nM GHRH. While mean basal GH plaque areas were similar among the treatment groups of group 1, the maximal GH plaque area was significantly decreased in Cx [16.8 +/- 2.4 vs. 26.4 +/- 3.9 X 10(6) microns2 (mean +/- SEM); P less than 0.05], but not in Cx + T (27.5 +/- 4.1 microns2) rats. The GHRH EC50 was unaffected by castration or T replacement. The percentage and absolute population of somatotrophs were reduced in Cx, but not in Cx + T, rats, while the numbers of lactotrophs remained unchanged in these treatment groups. For group 2, the mean peak GH plaque area was reduced in Cx + E (16.5 +/- 2.9 microns2; P less than 0.001) compared to that in M rats (36.2 +/- 2.3 microns2), but was not significantly different from that in F (13.0 +/- 1.5 microns2) rats. The EC50 was significantly (P less than 0.025) greater in Cx + E (10.9 +/- 2.3 nM) and F (7.9 +/- 1.6 nM) compared to M rats (2.8 +/- 0.7 nM). The absolute somatotroph and lactotroph populations were increased in Cx + E compared to M and F rats, as were the populations of other pituitary cell types. Testosterone enhances GH secretion by increasing the secretory capacity, but not the sensitivity, of somatotrophs to GHRH and by recruiting the function of a subpopulation of somatotrophs. Estradiol reduces the secretory capacity and sensitivity of somatotrophs to GHRH, but increases the population of somatotrophs, lactotrophs, and non-GH- and non-PRL-secreting cells.(ABSTRACT TRUNCATED AT 400 WORDS)

Structural and immunologic evidence for a small molecular weight ("21K") variant of prolactin.

The mol wt (Mr) of intact murine PRL is approximately 23,000. Immunoblotting analysis shows a 21,000 Mr band in fresh rat and mouse pituitary extracts that cross-reacts strongly with PRL antibodies. The band becomes markedly altered by stimuli known to influence PRL secretion, such as nursing, estradiol benzoate, and perphenazine. By two-dimensional electrophoresis, it splits into at least three components, two of which have more acidic pI than PRL. The tyrosine peptide maps of the three proteins resemble that of PRL. These results indicate that the 21,000 Mr band consists of a cluster of proteins that are structural variants of PRL, raising the possibility that these variants result from alternative splicing of the PRL gene transcript.