Biochemistry and pharmacology of rabbit cardiac growth hormone (GH) receptors.

In this report we present the first in-depth description of the biochemical and pharmacological properties of rabbit cardiac GH receptors. The apparent M(r)'s of the [125I]human (h) GH-receptor complexes were 380, 205, 90, 62, 52 and 38 kDa as demonstrated by an autoradiograph of affinity-labelled cardiac GH receptors separated under non-reducing conditions by SDS PAGE. The [125I]hGH-cardiac GH receptor complexes were disulfide-linked since the M(r)s of the complexes diminished to 170, 116, 97, 71, 45 and 38 kDa under reducing conditions, indicating the presence of multiple receptors, receptor-associated macromolecules or receptor and ligand in various ratios. The pharmacology of cardiac GH receptors is not typical of GH receptors present in other tissues. In radio receptor assays, both bovine GH and ovine prolactin were 50-fold and 100-fold less potent, respectively, than unlabelled hGH, in blocking the binding of [125I]hGH to cardiac binding sites and were, therefore, extremely weak antagonists. Similarly, neither bovine GH nor ovine prolactin blocked the [125I]hGH affinity-labelling of cardiac GH receptors compared to equivalent doses of unlabelled hGH. Parameters which characterize the kinetics for the association, dissociation and equilibrium binding of [125I]hGH to cardiac GH receptors were ascertained. Association kinetics for the binding of [125I]hGH to heart GH receptors exhibited a maximum specific binding at 17 h and 25 degrees C. The association of [125I]hGH to heart GH receptors was reversible with approximately 15 h required for half of the specifically bound [125I]hGH to dissociate. The coupling of [125I]hGH to heart GH receptors was optimum at pH 6 and the strength of the equilibrium binding, as measured by the ED50, was approximately 2 ng/ml. These data indicate that the cardiac GH receptors are pharmacologically distinct and that there is a M(r) heterogeneity in the [125I]hGH receptor complexes.

Precipitation of dilute chromatographic samples (ng/ml) containing interfering substances for SDS-PAGE.

SDS-PAGE of chromatographic fractions requires prior removal of salts, detergents, denaturants, or organic solvents which may perturb the electrophoretic separation. Likewise, to successfully visualize minute amounts of protein present in chromatographic fractions, they must often be concentrated before analysis by SDS-PAGE. In this study, we used a dye precipitation procedure for simultaneous removal of interfering substances and concentration of dilute samples (ng/ml) before analysis by SDS-PAGE. Nanogram amounts of protein (143 ng) were effectively precipitated with a pyrogallol red-molybdate reagent from commonly used chromatographic buffers containing various interfering solutes or solvents. Proteins were successfully precipitated from solution in the presence of organic solvents (acetonitrile, methanol, 2-propanol), chaotropic agents (6 M urea, 6 M guanidine-HCl), a protein stabilizer (40% sucrose), metal chelators (30 mM EDTA and 30 mM EGTA), or high salt (1.0 M NaCl). Detergents, at concentrations up to twice their critical micelle concentrations, from the nonionic class (Triton X-100, Tween 20) or from the zwitterionic class (3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate) did not inhibit protein precipitation. Some interference was observed when proteins were precipitated in the presence of ammonium sulfate (0. 5-2.0 M). Proteins did not precipitate in the presence of ionic detergents (SDS and cetyltrimethylammonium bromide). The sensitivity of the combined pyrogallol red-molybdate precipitation/SDS-PAGE procedure is approximately 7 ng. Two other methods of precipitating proteins (trichloroacetic acid and phenol-ether) both exhibited varying degrees of effectiveness, ranging from 714 to 7 ng/ml, in the precipitation of individual proteins. In summary, the pyrogallol red-molybdate protein precipitation procedure facilitates the SDS-PAGE analysis of dilute protein samples (ng/ml) from chromatographic fractions of various compositions. The method is useful for rapid pilot-scale protein fractionation and facilitates the ongoing propensity of researchers to work with minuscule amounts of protein.

Generation, characterization and utilization of anti-human growth hormone 1-43, (hGH1-43), monoclonal antibodies in an ELISA.

The major isoform of hGH is a polypeptide of 191 amino acids. Human GH1-43 is an amino terminal segment of hGH1-191 which comprises the first 43 amino acids. This peptide is a potent regulator of glucose homeostasis. To facilitate our understanding of the physiological regulation of hGH1-43 an assay to measure its levels in biological fluids and extracts is needed. This communication describes the development of anti-hGH1-43 monoclonal antibodies and their use in the development of an indirect competitive ELISA for the quantification of hGH1-43. Hybridomas were produced by the fusion of FOX-NY myeloma cells with spleen cells taken from a mouse immunized with hGH1-43. The hybridomas were screened for production of antibodies to hGH1-43 by antibody capture ELISA. Hybridomas which produced antibodies reactive to hGH1-43 were cloned by limiting dilution. Three monoclonal hybridomas, CCL-1, CCL-2, and CCL-3 were subsequently obtained. These hybridomas secreted antibodies that were highly reactive towards hGH1-43 but minimally reactive towards hGH1-191. The isotypes of the mAbs secreted by CCL-1, CCL-2 and CCL-3 were all IgG1 kappa as shown by isotype specific antibody capture analysis. An indirect competitive ELISA with a detection limit that ranged from 1 to 10 ng/ml was developed using mAbs from monoclonal hybridoma CCL-3. Dose-response curves for competing hGH1-191, hPRL, and hPL indicated minimal cross-reactivity of mAbs with these hormones and conversely, a high degree of specificity for hGH1-43. Dose-response curves for dilutions of human serum and pituitary extract were parallel to the standard. The availability of a sensitive assay for the measurement of hGH1-43 will help us answer questions regarding the biosynthesis, regulation of secretion, and role of hGH1-43 in the control of glucose homeostasis.

Divalent metal cation chelators enhance chromatographic separation of structurally similar macromolecules: separation of human growth hormone isoforms.

Human GH isoforms were separated by anion-exchange chromatography using a linear NaCl gradient in the presence and absence of EDTA and EGTA. SDS-PAGE showed that glycosylated 24-kDa hGH did not appreciably separate from other hGH variants in the absence of metal chelators. However, in the presence of metal chelators, glycosylated 24-kDa hGH separated from the bulk of the hGH isoforms. Human GH isoforms were also separated by size-exclusion chromatography in the presence and absence of metal chelators. Glycosylated 24-kDa hGH eluted with the bulk of the hGH isoforms in both separations. The inclusion of metal chelators in chromatographic buffers to alter the charge and/or size of proteins by stripping their metals may be a generally useful strategy in their fractionation.

High-resolution polyacrylamide gel electrophoresis of carbohydrates derivatized with a visible dye.

A technique for carbohydrate analysis that is both inexpensive and easily performed is currently unavailable. In this communication we address the problem and have outlined a method for labeling saccharides with a visible dye, 4-amino-1,1'-azobenzene-3, 4'-disulfonic acid, which has an absorption maximum of 489 nm and an extinction coefficient of 37,615, to facilitate visible detection at low levels. The visible dye was coupled by a reductive amination to different sugars. The sugar-dye adducts were then separated by electrophoresis on alkaline polyacrylamide gels. The gels were scanned with a densitometer, and visible sugar-dye adducts were qualitatively analyzed by identifying them according to their mobilities. The sugar-dye adducts were quantified by determining their densitometric volume. The kinetics of the reductive amination reactions, performed at 37 degrees C, were different for each of three saccharides tested. The rate constants for glucose and fucose were 1.31 times greater and 1.8 times greater, respectively, than that of maltotriose. The reductive amination reactions were essentially complete after approximately 16 h under the given experimental conditions. A linear dose-response relationship was observed between the amount of sugar (monosaccharide, trisaccharide, or heptasaccharide) in the reductive amination reaction. The quantity of saccharide-dye adduct that could be visually detected for glucose, maltotriose, and maltoheptaose, was 25, 25, and 50 nmol, respectively. Sugar-dye adducts were separated from one another by varying the acrylamide concentration in the polyacrylamide gels. Sugar-dye adducts of monosaccharides, disaccharides, trisaccharides, and heptasaccharides were separated on alkaline 30% polyacrylamide gels with mobilities of 0.778, 0.667, 0.639, and 0.375. Adducts of glucose, fucose, galactose, and mannose were separated with mobilities of 0.844, 0.833, 0.820, and 0.810, respectively, on a 30 to 40% gradient polyacrylamide gel. Adducts of glucose and glucose derivatives were separated on a 35% polyacrylamide gel. This technique provides an inexpensive and easily performed method of carbohydrate analysis to laboratories that do not have the highly trained personnel nor the expensive equipment needed for other methods of carbohydrate analysis. The method is most applicable to research problems where sensitivity (20 pmol) is not a problem. The simplicity of the method also makes it easily incorporated into teaching laboratories.

Glycosylated human growth hormone (hGH): a novel 24 kDa hGH-N variant.

We have identified a human pituitary protein as a novel glycosylated variant of hGH. Isolation of the denatured protein included separation of human pituitary extract by Sephadex G-100 chromatography in ammonium bicarbonate, followed by Sephadex G-100 chromatography in 10% acetic acid, with subsequent DEAE Sephacryl chromatography in ammonium bicarbonate, and finally by preparative SDS PAGE. The pituitary protein has a molecular weight of 24 kDa as determined by SDS PAGE analysis and is thus larger than the normal 22 kDa hGH. N-Terminal amino acid sequence analysis of the first twenty-six residues reveals that this protein is not derived from the hGH-V gene but is rather a hGH-N gene product. Assays for the detection of glycoconjugates (periodate oxidation, sialidase treatment, trifluoromethanesulfonic acid treatment) indicate that the hGH variant has carbohydrate moieties. The discovery of new hGH raises questions about the role of glycosylation in the structure/function relationships of this hormone.

Human growth hormone deletion mutant (hGH44-191) binds with high affinity to lactogenic receptors but not to somatogenic receptors.

The carboxyl-terminal hGH fragment, hGH44-191, displays diabetogenic activity. To understand whether this biological activity is mediated through somatogenic or lactogenic receptors, we investigated the ability of hGH44-191 to bind both receptor classes. We found that hGH44-191 could not compete with [125I]hGH or [125I]bGH for bovine liver somatogenic binding sites. Additionally, hGH44-191 could not displace [125I]hGH from the somatogenic receptor sites when human liver microsomes were used as the receptor source. In contrast, hGH44-191 effectively competed with [125I]hGH for lactogenic receptor sites of lactating mammary gland microsomes and of bovine liver microsomes. In summary, hGH44-191 does not bind to somatogenic receptors but does not bind to lactogenic receptors. These data suggest that the biological actions of hGH44-191 could be mediated through lactogenic receptors.

A 12-kilodalton N-glycosylated growth hormone-related peptide is present in human pituitary extracts.

This study was designed to investigate whether N-linked glycosylation could account for the presence of glycosylated GH forms (G-GH) in human pituitary extracts. The study was carried out in commercially available pituitary GH preparations (pitGH). Recombinant GHs obtained from eu- or prokaryotic cells were used as controls. Radioiodinated GHs were incubated in tubes containing Concanavalin-A (Con-A) attached to a Sepharose 4B matrix. Pituitary G-GH forms were selectively displaced from Con-A by adding N-acetyl-D-glucosamine or methyl-alpha-D-mannopyranoside and electrophoresed. Autoradiographies of these gels identified a 12-kilodalton (12K) band as the glycosylated form. Due to the fact that this peptide was partially immunoprecipitated with an anti-GH serum and the absence of detectable PRL in the pitGH extracts, it would indicate that such a glycopeptide was GH related. Endoglycosidase-F treatment of pitGH extracts induced a decrease in the mol wt of that 12K peptide, as indicated by the changes observed in its mobility on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. This demonstrated that the sugar moieties were N-linked. When recombinant GHs were assayed by a similar method, no specific binding to Con-A was detected. NH2-terminal amino acid sequence analysis from the 12K band demonstrated that this band was composed by three peptides. Peptide 1 corresponds to the GH-N 102-119 sequence. Interestingly, peptide 2 exhibits GH-V 1-18 sequence, while peptide 3 seems to be a novel GH-related peptide. Taken together, these data suggest that the pituitary G-GH form found in human pituitary extracts is derived not from the "normal" GH-N gene, but, rather, from the GH-V gene or another unidentified gene. The potentially important pathophysiological implications of this finding need to be investigated.

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.

Glycosylated human prolactin: alterations in glycosylation pattern modify affinity for lactogen receptor and values in prolactin radioimmunoassay.

The receptor-binding properties of monomeric nonglycosylated human PRL (hPRL), glycosylated hPRL that does not bind to Concanavalin-A-Sepharose (G1-hPRL) and glycosylated hPRL that binds to Concanavalin-A-Sepharose (G2-hPRL) were tested in the lactating rabbit mammary gland RRA for lactogenic hormones. Variations in the glycosylation pattern of G-hPRL altered its receptor-binding properties, suggesting that the site of glycosylation may be proximal to the receptor-binding region. Relative potencies for the displacement of [125I]hPRL by hPRL, G1-hPRL, and G2-hPRL were 100%, 40%, and 26%, respectively. Relative potencies for displacement of [125I]G1-hPRL by hPRL, G1-hPRL, and G2-hPRL were 100%, 44%, and 69%, respectively; however, the displacement curve for G2-hPRL was not parallel to the others. When G2-hPRL was radiolabeled, there was no specific binding to lactogenic receptors. The presence of PRL receptor subtypes and/or kinetic cooperativity was suggested by the complexity of the binding isotherms. The immunoreactivities of the PRLs were tested in a homologous RIA, using polyclonal antiserum. The modification of the glycosylation pattern of hPRL significantly altered the RIA values for PRL. When hPRL was used as the radiotracer, the percent cross-reactivities of hPRL, G1-hPRL, and G2-hPRL were approximately 100%, 23%, and 17%, respectively. When G1-hPRL was used as the radiotracer, the percent cross-reactivities were approximately 100%, 135%, and 54% for hPRL, G1-hPRL, and G2-hPRL; however, the displacement curve for hPRL was not parallel to those of the glycosylated hPRLs. When G2-hPRL was used as the radiotracer, the percent cross-reactivities were approximately 100%, 32%, and 37% for hPRL, G1-hPRL, and G2-hPRL. These data point out that the glycosylation heterogeneity of hPRL is a factor that affects the diagnostic accuracy of hPRL determinations. Specific RIAs for each PRL are needed so that we can have valid and reliable measurements of each PRL isoform and consequently gain a better understanding of PRL's complex biological role.

Effects of hypophysectomy and the insulin-like and anti-insulin pituitary peptides on carbohydrate metabolism in yellow Avy/A (BALB/c x VY)F1 hybrid mice.

The amino-terminal portion of human growth hormone, residues 1-43 (hGH1-43), has insulin-potentiating action, while a hyperglycemic pituitary peptide (HP), which co-purifies with human growth hormone (hGH), is antagonistic to the action of insulin. The effects of hGH, hGH1-43, and HP on glucose metabolism were assessed in young (4-5 weeks) and adult (6-8 months) hypophysectomized yellow Avy/A mice which lacked any interfering endogenous pituitary hormones, and compared with age-matched intact obese yellow Avy/A and lean agouti A/a mice. Treatment with hGH1-43 or HP did not promote body growth in hypophysectomized yellow mice; but after 2 weeks of treatment with hGH, there was a significant increase in body weight (P less than 0.05). Treatment with HP raised blood glucose and lowered insulin concentrations in obese yellow mice, but not in agouti or hypophysectomized yellow mice. The severely impaired glucose tolerance of the hypophysectomized yellow mice was improved by acute (60 min) and chronic (3 days) treatment with hGH1-43 as well as by 2 weeks of treatment with hGH; in contrast, HP had no effect. Glucose oxidation in adipose tissue from obese yellow mice was low and showed essentially no response to stimulation by insulin at doses lower than 1000 microunits/ml. Basal glucose oxidation rates in adipose tissue taken from agouti and hypophysectomized yellow mice were significantly higher (P less than 0.001) than those in tissue from obese yellow mice, and the rates responded significantly (P less than 0.05) to 100 microunits/ml insulin. The insulin binding affinities in liver membranes from agouti mice were higher than those from either obese or hypophysectomized yellow mice. The insulin receptor densities were similar in both agouti and obese yellow mice, but higher in hypophysectomized yellow mice (P less than 0.05). Treatment with hGH1-43 slightly increased, although not significantly, the insulin receptor density in yellow obese mice while hGH showed essentially no change. Therefore, hypophysectomy appeared to increase tissue response and decrease insulin resistance by increasing receptor numbers and lowering the circulating insulin levels. Furthermore, the insulin-like action of hGH was elicited directly in vivo by hGH1-43 in hypophysectomized yellow mice.

Studies on dissociation of mouse prolactin from mouse hepatic receptors.

The influence of pH, temperature, ethylene glycol, urea, chaotropic anions and excess unlabelled secreted mouse prolactin (smPRL) on the dissociation kinetics of 125I-iodosmPRL from mouse hepatic receptors was investigated. The destabilization of smPRL-receptor complexes by chaotropic anions followed the typical trend of the Hofmeister series: I- greater than Br- greater than Cl- greater than F-. Increasing the temperature of the dissociation reaction from 8 degrees C to 23 degrees C and 30 degrees C caused partial dissociation of 125I-iodosmPRL-receptor complexes. Dissociation of 125I-iodosmPRL from mouse hepatic receptors was pH dependent, with the slowest rate of dissociation occurring at pH 8 and the fastest rate of dissociation occurring at pH 5 and 6. Both ethylene glycol and urea accelerated the rate of dissociation of 125I-iodosmPRL from mouse hepatic receptors in a concentration-dependent manner. Dissociation of 125I-iodosmPRL from mouse hepatic receptors was 6-fold faster in the presence of excess unlabelled smPRL than in its absence. The results of these investigations suggest that both protonation/de-protonation reactions and hydrophobic interactions play important roles in stabilizing the smPRL-receptor complex. In addition, they suggest that cooperative interactions may be involved in the binding of smPRL to mouse hepatic receptors.

Thermodynamics and kinetics of mouse prolactin-hepatic receptor interaction.

Kinetic and thermodynamic parameters associated with the binding of secreted mouse prolactin (smPRL) to mouse hepatic receptors were investigated. When the reaction temperature was increased from 8 degrees C to 37 degrees C, the association rate constant k+1, increased approximately 5-fold, from 2.3 X 10(4) M-1 . S-1 to 12.6 X 10(4) M-1 . S-1. An Arrhenius plot indicated that there was a linear relationship between ln (k+1) and 1/T. When the reaction temperature was increased from 8 degrees C to 37 degrees C, the equilibrium binding constant, Ka, decreased approximately 1.5-fold, from 2.8 X 10(8) M-1 to 1.9 X 10(8) M-1. When the pH of the binding reaction was lowered from 9.0 to 6.2, Ka decreased approximately 3-fold, from 2.6 X 10(8) M-1 to 0.9 X 10(8) M-1. Transition state thermodynamic parameters for the formation of the smPRL-receptor complex, represented by delta G+', delta H+' and delta S+', were +45.7 kJ/mol, +41.2 kJ/mol and -15.1 J/(mol . K), respectively. Parameters for the equilibrium reaction, described by delta G0', delta H0' and delta S0', were -47.6 kJ/mol, -10.6 kJ/mol and +124 J/(mol . K), respectively. Over the temperature range studied, a Van't Hoff plot of the binding constants demonstrated a linear relationship between ln (ka) and 1/T, indicating that changes in enthalpy for the binding reaction were temperature independent. The binding reaction was largely entropically driven (delta S0' greater than 0), suggesting that hydrophobic interactions are involved in forming the smPRL-receptor complex.

Interaction of mouse prolactin with mouse hepatic receptors.

Lactogenic receptors are usually studied in heterologous systems where prolactin is derived from one species and receptors prepared from another. In such systems the foreign prolactin could be seen as a growth hormone by the host tissue. We have therefore developed a homologous radioreceptor assay using secreted mouse prolactin (smPRL) and mouse hepatic receptors. In this system, monovalent anions augment the smPRL-receptor interaction in the order F- greater than Cl- greater than Br- greater than I-. Divalent cations (Mg2+, Ca2+, Sr2+), phosphate and acetate also increase smPRL binding. Temperature and pH optima are at 8 degrees C and pH 8.3, respectively. Under optimum conditions, the percent total, specific and nonspecific binding are 55%, 45% and 10%, respectively. At infinite receptor concentration the maximum specific bindability of labeled smPRL is 50%. The effects of ions on binding of smPRL to the receptor show that hydrophobic forces participate in smPRL-receptor coupling. The biphasic dissociation kinetics show initial and final rate constants of 1.56 X 10(-4)/s and 7.62 X 10(-6)/s, respectively. The lactogenic receptor does not bind mouse growth hormone; however, it binds both mouse placental lactogen (mPL) and smPRL with equilibrium association constants of 3.90 X 10(8) M-1 and 2.25 X 10(8) M-1, respectively, suggesting that smPRL and mPL share biological roles by acting through the same receptor.

Secreted mouse prolactin (PRL) and stored ovine PRL. II. Role of amides in receptor binding and immunoreactivity.

Native PRL and des-amido forms 1, 2, and 3 were tested for their individual immunochemical and receptor-binding abilities. The results show that deamidation of either secreted mouse PRL or stored ovine PRL alters their binding in a radioreceptor assay. For each accumulated deamidation there was a statistically significant (P less than 0.05) decrease in the binding potency of either stored ovine PRL or secreted mouse PRL to a cell membrane receptor preparation. RIAs indicated that there was a statistically significant (P less than 0.05) decrease in PRL's immunopotency toward polyclonal antisera only when select residues deamidated. This study suggests that all the asn/gln residues which deamidate in order to make des-amido forms 1, 2, and 3 constitute part of the receptor-binding domains of both secreted and stored PRLs, while only a fraction of those same residues form portions of PRL's antigenic sites. Thus PRL's receptor-binding surface is separated from its antigenic sites, with only partial overlap being indicated. Our data also indicate that there are no major structural differences in the receptor-binding domains of secreted and stored PRLs. We also see that the receptor-binding domain of PRL has been highly conserved throughout evolution. Since the binding affinity of PRL to a membrane-bound receptor can be altered through deamidation, we view the process as a plausible regulatory mechanism for controlling the quantitative action of PRL at a given target organ.

Secreted mouse prolactin (PRL) and stored ovine PRL. I. Biochemical characterization, isolation, and purification of their electrophoretic isoforms.

The biochemical nature of the electroisomers of secreted mouse PRL and stored ovine PRL were compared. When examined on alkaline polyacrylamide gels, they exhibited electrophoretic heterogeneity. The electrophoretic isomers had the same molecular size by examination of Ferguson plots and therefore differed only in net negative charge at alkaline pH. There was no apparent charge heterogeneity in the PRL preparations when they were electrophoresed at a pH below the pKa of the side-chain carboxyl groups of aspartic acid and glutamic acid; however, they exhibited size heterogeneity owing to aggregation. The slowest migrating electroisomers (from alkaline gels) converted spontaneously into faster migrating forms at 37 C in either acid (pH 4.0) or alkaline (pH 8.0) environments. The rate constant (determined at 37 C) for the transformation of the native PRL into faster migrating isoforms was 10.76 X 10(6) sec-1 at pH 8.0 and 0.50 X 10(6) sec-1 at pH 4.0. When secreted mouse PRL was incubated in alkaline (pH 10.0) conditions at 25 C, ammonia was released as the conversion reaction occurred. These results indicate that the electroisomers separated at alkaline pH are charge isoforms differing only in the number of free carboxyl groups of glutamic and/or aspartic acid residues. The des-amido isoforms were fractionated and purified. Polyacrylamide gel electrophoresis at alkaline pH was used to separate the charge isomers from each other. Each isoform was then excised from the separating matrix and recovered via electrophoretic elution. The homogeneity of each isoform was monitored by polyacrylamide gel electrophoresis.

Protein hormones: detection and extraction of functional forms from alkaline polyacrylamide gels.

A method is presented for rapidly staining zones (30 min) in alkaline polyacrylamide gels in the absence of fixatives. Native functional proteins are recovered in homogeneous form after excision of the visible zones from the polyacrylamide matrix. Removal of dye from excised zones is facilitated because only the surface of the gel is stained. Biological activity is then recovered from the gel slices by simple diffusion. The technique makes use of the sensitivity of Coomassie Brilliant Blue G-250 for the detection of proteins (less than 1 microgram). Structural variants of prolactin are isolated and recovered by this method. The method is applicable to studies requiring analytical and semi-preparative electrophoresis of proteins, especially pituitary hormones.

Homologous somatotropin radioreceptor assay utilizing recombinant bovine growth hormone.

A homologous radioreceptor assay using recombinant bovine growth hormone and bovine liver membranes is described. The total specific binding of 125I-labeled recombinant bovine growth hormone to the 100 000 X g pellet was 48% in 24 h at 25 degrees C. Hormone binding was partially reversible, with 40% of the radiolabeled hormone being irreversibly bound. The amount of specific binding varied with assay pH, with the optimum occurring at pH 7.8. Specific binding was temperature-dependent, with greater specific binding occurring at 25 degrees C than at 5 degrees C or 37 degrees C during a 24 h period. Recombinant bovine growth hormone, human growth hormone, ovine growth hormone and recombinant porcine growth hormone competed effectively with 125I-labeled recombinant bovine growth hormone for binding sites, while bovine prolactin and ovine prolactin were needed in amounts 10(6)-fold the concentration of recombinant bovine growth hormone to displace the radiolabeled hormone. Surprisingly, human placental lactogen did not displace the radiolabeled hormone.