Mitogenic and Receptor Activities of Human Growth Hormone 108-129 "

We have selectively synthesized a number of peptides encompassing the region of helix 3 of growth hormone (GH). These peptides and native human (h) GH have been evaluated for mitogenic and receptor activities in 3T3-F442A preadipocytes. In this system, wild type hGH is anti-mitogenic. In contrast, hGH 108-129 stimulated DNA synthesis while other GH-derived peptides were ineffective. hGH (L) 108-129 had an ECao of about 0.2 nM and was maximally effective at about 0.5 nM in stimulating ['Hlthymidine incorporation in 3T3F442A cells. hGH (L) 108-129 was mitogenically as active as insulin-like growth factor-I and more active than insulin. It was less effective than transforming growth factor-@. By cell cycle analysis, hGH (L) 108129 increased the proportion of cells in S/Gz/M phases to 28%. hGH, when coincubated with hGH (L) 108129, blocked the mitogenic response of the peptide. A monoclonal antibody to the GH receptor significantly reduced binding of "9-hGH to its receptor but had no effect on binding of '261-hGH (L) 108-129. Affinity cross-linking of lZ6I-hGH to its receptor was not duplicated with 12"I-hGH (L) 108-129. No other GHpeptides or insulin competed for binding of '261-hGH 108-129. Scatchard analysis indicated a Kd of 6.2 nM with 6.6 x lo6 binding siteslcell for hGH (L) 108-129. These studies indicate that hGH (L) 108-129, a sequence encompassing helix 3 of hGH, acts by binding to a site other than the GH receptor and evokes high mitogenic responses.

has been reported (2-4) to have growth promoting activity in the hypophysectomized rat.bGH 96-133 has been well characterized chemically (2, 5) and physically (6,7), including NMR spectroscopy (8,9).The importance of residues within helix 3 of GH for growth promotion has been further emphasized by studies employing site directed mutagenesis (10-18).On the basis of binding of two molecules of growth hormonebinding protein to a molecule of GH (10-12), it has been suggested that there are two different binding sites on GH.The residues in helix 3 contribute to binding site 2.
We have developed a serum-free hormonally defined medium (SFM) (19) in which the differentiation, mitogenic, and metabolic actions of GH can be assayed in 3T3-F442A fibroblasts (19,ZO).With the above noted indications that helix 3related peptides may have growth promoting activity, we have synthesized a series of peptides including bGH 7-34 (encompassing helix l), hGH (L) 108-129, hGH (D) 108-129 (prepared with all D-amino acids), and hGH 113-130.These have been assayed for mitogenic activity.In addition, binding to target cells with biologically active peptides has been studied.We report here that hGH (L) 108-129 has mitogenic activity and binds to a receptor other than the GH receptor.The other GH-derived peptides were biologically inactive.

EXPERIMENTAL PROCEDURES
Materials-General biochemicals and chemicals of required high grade (tissue culture, ultra pure and sequencing grade) were from Sigma, Boehringer Mannheim, J. T. Baker Chemical Co.,and Pierce Chemical Co. [3H]Thymidine (20 Ci/mmol) was from Du-Pont New England Nuclear.Na[1Z51] (2176 Ci/mmol) was from Amersham Corp.
Solvents for HPLC of chromatography grade were from J. T. Baker Chemical Co. and Pierce Chemical Co.
Stocks of Swiss 3T3-F442A fibroblasts were generously supplied by Dr.Howard Green (Boston, MA) and passed in this laboratory as previously described (19).Tissue culture plasticware was purchased from Corning Glass Works (Corning, NY).Calf serum, fetal calf serum, trypsin, glutamine, and 6-well plates were obtained from Life Technologies, Inc. Recombinant hGH (met-hGH) was a gift from Genentech (South San Francisco, CA) and Eli Lilly (Indianapolis, IN).Bovine transferrin, bovine insulin, bovine fetuin, and T1 were obtained from Sigma.Bovine serum albumin (BSA Fraction V) was purchased from ICN Biomedicals, Inc. (Lisle, IL).Murine epidermal growth factor, human insulin-like growth factor I (IGF-I), and transforming growth factor-& (TGF-P) were obtained from Collaborative Research (Waltham, MA).Anti-GHR antibody (mAb 263) was obtained from Agen, Inc. (Parsippany, NJ).Disuccinimidyl suberate was obtained from Pierce Chemical CO.
Chemical Methods-Peptides were prepared by the solid-phase synthesis method developed by Merrifield (21) using an Applied Biosystem 431A automatic peptide synthesizer.Peptides were purified to homogeneity by HPLC with a Waters 600 multi-solvent delivery system.Purity was established by amino acid analysis and NH2-terminal sequencing with an Applied Biosystems 477A automatic protein sequencer with a 120 phenylthiohydantoin detector based on the published method (22).All the peptides were homogeneous by HPLC analysis (data not shown).The amino acid compo-sition and sequences were the same as the published sequences (23-26, data not shown).Synthesis, amino acid analysis, and sequencing were performed in the Microchemistry Core Facility of the Sloan-Kettering Institute.Laemmli SDS-polyacrylamide gel electrophoresis was run as described (27).
PHIThymidine Incorporation Assay-Cells were inoculated in 6well dishes (9.6-cm2 area) in a volume of 3 ml at a density of 2 X lo4 cells/well.The following day, medium was removed and cells were washed three times with HBSS (Hanks' buffered salt solution), and SFM containing various concentrations of peptides was added to cells.Incubations with peptides were carried out for 1-2 days at 37 "C.There was little difference in stimulation of [3H]thymidine incorporation between 1 and 2 day incubations with peptides.Cells were labeled with [3H]thymidine (10-20 pCi/ml) 3 h prior to cell harvest.
Following labeling, cells were solubilized with 0.5% SDS at 37 'C for 10 min.Portions of cell lysates were precipitated with 10% trichloroacetic acid and counted by scintillation counting.Radioactivity of each sample was normalized by protein concentration determined by the micro-bicinchinonic acid protein assay.
Flow Cytornetry-Cells (lo6) were detached by trypsin and were harvested by centrifugation at 1000 X g for 10 min.Cell pellets were suspended in 0.5 ml of PBS buffer and an equal volume of 95% ethanol was added.After fixing overnight at 4 "C, cells were harvested by centrifugation at 1000 X g for 10 min and cell pellets were resuspended in PBS containing 50 pg/ml propidium iodide and 50 pg/ml RNase A. The staining reaction was carried out for at least 2 h at room temperature, and the DNA content of each sample was analyzed by an Epics profile I1 flow cytometer.The propidium iodide staining does not distinguish Go cells from GI cells.
B i d i n g Assay-Monolayer cultures of 3T3-F442A cells (2 X lo5 cells/well) were washed with HBSS, and then binding buffer (HBSS, 10 mM HEPES, pH 7.2, 0.1% BSA) containing 0.4 nM of T -h G H (L) 108-129, and indicated amounts of hGH, hGH (L) insulin, were added.The binding reaction was carried out for 3 h at 4 "C on a platform shaker.Unlabeled ligand was removed by washing monolayers with ice-cold binding buffer.Cell lysates were prepared using extraction buffer (150 mM NaCI, 20 mM Tris-HC1, pH 8.0, 1 mM MgC12, 0.1 mM ZnClz, 0.5% Nonidet P-40, 2 mM phenylmethylsulfonyl fluoride).Aliquots were saved for protein determination.Bound peptide was measured by counting samples in a gamma counter.Specific binding was defined as the difference between total binding and binding observed at a 1000-2000-fold excess of hGH (L) 108-129.Binding affinity and number of binding sites were determined by Scatchard analysis (29).
Affinity Cross-linking-Monolayer cultures of 3T3-F442A cells (lO'/well) were washed with HBSS.Binding buffer (2 ml) containing "'I-hGH (1 nM) or lZ5I-hGH (L) 108-129 (1 nM) was added.The binding reaction was carried out for 3 h at 4 "C.Binding was terminated by removing unbound ligand by washing three times with HBSS.Disuccinimidyl suberate (0.5 mM) in PBS was then added to the cells and the cross-linking reaction was carried out at 4 "C for 30 min.Cells were lysed in extraction buffer (20 mM HEPES, pH 7.4, 1% Triton X 100,10% glycerol, 50 mM sodium fluoride, 1 mM sodium orthovanadate, 10 pg/ml leupeptin, 0.5 mM phenylmethylsulfonyl fluoride) and lysates were clarified by centrifugation at 30,000 x g for 30 min.Supernatants were boiled for 5 min and were subjected to SDS-polyacrylamide electrophoresis gels (7.5%).Gels were fixed, dried, and exposed on Kodak XAR-5 film at -70 "C.

RESULTS
Mitogenic Activity-The mitogenicity of peptides and hormones was evaluated by [3H]thymidine incorporation and by FACS analysis of DNA content in 3T3-F442A cells.Native hGH was anti-mitogenic with an ECso about 2.0 nM in unstimulated cells (Table I and Ref. 30).hGH (L) 108-129 stimulated [3H]thymidine incorporation about 5-fold with an ECso between 0.1 and 0.5 nM and a maximal effect noted at 0.5 nM (Fig. 1).No other peptide derived from the hGH sequence, hGH 113-130, hGH (D) 108-129 or hGH itself was mitogenic based on [3H]thymidine incorporation in doses up to 1 pM (Fig. l), The mitogenic effect of hGH (L) 108-129 was not as effective as TGF-/3 which had an ECso of about 1 PM and was maximally effective at 50 p~ (Fig. 2).hGH (L) 108-129 had potency similar to IGF-I (Fig. 2).Insulin was less potent in stimulating [3H]thymidine incorporation, although IGF-I was more potent than insulin (Fig. 2).
Since hGH and hGH (L) 108-129 show opposite effects on [3H]thymidine incorporation, it was important to determine whether hGH-treated 3T3-F442A cells are resistant to the mitogenic effects of hGH (L) 108-129.Table I shows  with hGH and/or hGH (L) 108-129.As expected, hGH and hGH (L) 108-129 show opposite effects when incubated separately with cells.When incubated together in molar ratios (hGH/hGH (L) 108-129) of 2:l and 201, the mitogenic response to 0.5 nM hGH (L) 108-129 was completely offset and the cells manifested the prevailing anti-mitogenic response of hGH (Table I).
To examine whether the mitogenic effects of hGH (L) 108-129 are accompanied by cell cycle transition in 3T3-F442A cells, cell cycle analysis was carried out by analysis of DNA content using a FACS analyzer.As expected, cells maintained in SFM are primarily in the G1/G, phase (Fig. 3a).Cells grown with 10% FCS showed 45% of the cells in S/G2/M phases (Fig. 3b).hGH (L) 108-129 (0.5 nM) increased the number of cells in S/G2/M phases to 28% (Fig. 3c).hGH 113-130 did not cause any shift in the cell cycle phases of 3T3-F442A cells as is seen in Fig. 3d.
To investigate the possibility that the mitogenic response to hGH (L) 108-129 was mediated by secretion of IGF-I, 3T3-F442A cells were incubated with IGF-I antibody in the presence of hGH (L) 108-129, as well as IGF-I.Although both peptides individually gave a strong mitogenic response, the presence of an antibody to IGF-I blocked only the IGF-I response, without effect on the mitogenesis induced by hGH (L) 108-129 (data not shown).
Binding Characteristics of hGH (L) 108-129-A series of experiments was carried out to determine the binding characteristics of the synthetic peptide hGH (L) 108-129 to 3T3-F442A cells.The ICs0 (concentration of 50% inhibition of binding) of "'I-hGH (L) 108-129 was between 10-20 nM (Fig. 4).Binding was effectively competed for by an excess of hGH and hGH (L) 108-129, but not by bGH 7-34, hGH 113-130 or insulin.hGH (L) 108-129 binding also showed saturation kinetics (data not shown).The optimal concentrations of hGH and hGH (L) 108-129 for their effect on cellular growth were lower than those of the K d values, suggesting that the respective receptors do not have to be fully occupied, i.e. there are spare receptors.
By Scatchard analysis, hGH had a binding affinity of approximately 0.2 nM and approximately 28,000 binding sites on 3T3-F442A cells (data not shown).Under similar conditions hGH 108-129 bound with an affinity of approximately 5 nM and had 560,000 binding sites/cell (Fig. 5).Affinity cross-linking of lZ5I-hGH to its receptor was not duplicated with "'I-hGH 108-129.
To determine whether hGH and hGH 108-129 bound to the GHR, competitive binding studies were performed in the absence and presence of the anti-GHR antibody, mAb 263.It was noted (Fig. 6) that mAb 263, in a dose-related fashion from 10"' to pg/ml inhibited binding of T -h G H to 3T3-F442A cells with complete inhibition of binding at concentrations of mAb 263 of pg/ml and greater (Fig. 6).In contradistinction there was no dose-related inhibition of binding of lZ5I-hGH 108-129 in any concentration range of mAb 263 between and pg/ml and the change in binding was generally less than 50% (Fig. 7).

DISCUSSION
The results of experiments in the present study indicate that hGH (L) 108-129 is a biologically active peptide with properties dissimilar from hGH. hGH is anti-mitogenic (30, 31), whereas hGH (L) 108-129 is mitogenic (Fig. 1).The pg/ml Mab 263 FIG. 6. Binding of "'I-hGH in the presence of GHR antibody.Binding studies were performed as described in legend to Fig. 4. 3T3-F442A cells were incubated in binding buffer with increasing concentrations of mAb 263 for 2 h at 4 "C.Cells were washed three times with the same buffer.lZ51-hGH was incubated with increasing concentrations of GHR antibody, mAb 263, and incubation was continued at 4 "C for 2 h.activity of hGH (L) 108-129 is not as great as TGF-0, similar to IGF-I, and greater than insulin.As with TGF-P, the mitogenic response was biphasic, i.e. increasing in the subnanomolar range and decreasing at higher concentrations.By cell cycle analysis, the increase (28%) of cells in S/Gz/M phases was similar to the increase (45%) induced by FCS (Fig. 3).The mitogenic response did not appear to be mediated by IGF-I as indicated by a lack of inhibition by an IGF-I antibody.
The mitogenic response to hGH (L) 108-129 was blocked by hGH.This is not likely due to competition between hGH and hGH (L) 108-129 for binding to the same receptor since they have different Scatchard analyses and binding specificities.Other mechanisms for reducing the mitogenic response are possible.In this assay it seems clear that hGH (L) 108-129 is not an antagonist because it does not affect the antimitogenic activity of hGH in 3T3-F442A calls.This also confirms the anti-mitogenic effect of hGH since it prevents the mitogenic effect of hGH (L) 108-129.This is consistent with the ability of hGH to block the mitogenic effect of insulin and platelet-derived growth factor in 3T3-F442A fibroblasts in SFM (31).
Studies of site-directed mutagenesis of hGH and x-ray crystallography of the complex of growth hormone-binding protein (hGH)' (10)(11)(12) suggest that helix 3 forms one of two binding sites on hGH.Notwithstanding the identity of the 22 amino acid sequence of hGH (L) 108-129 and a corresponding sequence in hGH and the putative determinants in this sequence for hGH site 2 binding to its receptor, there are few data from this investigation to indicate binding of the peptide hGH (L) 108-129 to the GHR.There was some competition of hGH for T -h G H (L) 108-129 binding (Fig. 4).This might reflect competition of a portion of GHR for the hGH-binding site 2 or cross-binding to another receptor as occurs in other systems, e.g.insulin and IGF-1.Scatchard analysis indicated 28 X lo3 binding sites for hGH with a Kd of 2 X 10"' M while the corresponding values for hGH (L) 108-129 were 5.6 X lo5 M binding sites/cell and a Kd of 5.2 X lo-' M. hGH offset the mitogenic response to hGH (L) 108-129 but not in stoichiometric ratios to suggest competitive antagonism.The binding of lZ5I-hGH (L) 108-129 could not be blocked by the antibody to GHR as could the binding of hGH nor could the GHR be affinity labeled with lZ5I-hGH (L) 108-129 as it could with '"I-hGH.
It is of note to compare the binding to 3T3-F442A cells with different biological responses.hGH binds to 3T3-F442A cells with a Kd of 0.77-1.8nM (32), while the ECso for the adipogenic response is about 0.1 nM (33).The anti-mitogenic effect occurs with an ECso of approximately 0.05 nM (30).This indicates that the half-maximal anti-mitogenic response occurs when approximately 5% of the sites are occupied.In the present study, hGH (L) 108-129 binding is 5.2 nM, and the ECW for mitogenesis is approximately 0.2 nM, when only 1% of the receptors are occupied, i.e. suggesting that there are spare receptors.It is of note that hGH (L) 108-129 has a Kd of binding of 5.2 nM compared to a value of about 0.2 to 1 nM for hGH (32).This lower affinity for the smaller peptide would be consistent with binding only at the site 2 residues of hGH present in hGH (L) 108-129.The lower affinity is not solely due to the size of the fragment since we have found that the 38-amino-acid sequence bGH 96-133 binds with a Kd of approximately 100 nM with corresponding mitogenic activity.' The numbers of binding sites ( P a x ) determined by a receptor other than GHR.This would be consistent with the fact that hGH with a K d of about 0.2-1 nM is less effective than hGH (L) 108-129 with a K d of 5.2 nM in competing for binding of '"I-hGH (L) 108-129 (Fig. 4).
Based on competition assays, it would appear that nonmitogenic peptides do not bind to the receptor that binds hGH (L) 108-129 because these non-mitogenic peptides were not able to block binding of hGH (L) 108-129.The mechanism of mitogenicity by this synthetic peptide is not yet clear since the immediate consequence(s) of ligand binding to GHR in 3T3-F442A cells is not known.
There are no reports from this study or elsewhere to indicate that hGH 108-129 is derived from natural sources either by de nouo biosynthesis or biodegradation of hGH.There is, however, reason to study the biological and biochemical properties of helix 3 of GH.Sequences related to this helicogenic peptide are of interest as isolated peptides or within the larger polypeptide of GH.In addition to chemically synthesized hGH 108-129, hGH 95-133 and bGH 96-133, prepared by chemical synthesis or enzymatic digestion, have proliferative activity, albeit at lower doses.'In addition, modification of sequences within helix 3 of GH by site-directed mutagenesis is associated with changes in growth promoting (14,17,34,35) and binding activity (16).Our system affords the unique opportunity for the study of GH structure and function since native hormone and helix 3 of GH elicit opposite effects on cellular proliferation.

FIG. 3 .
FIG. 1. Effects of growth hormone peptides on ['Hlthymi-dine incorporation in 3T3-F442A cells.3T3-F442A cells were maintained in serum-free medium with the indicated concentrations of peptides for 48 h.Levels of DNA synthesis were determined by measurement of [3H]thymidine (10 pCi/ml) incorporation in a 3 h labeling 21 h after the addition of peptides.Data were normalized on the basis of protein concentration, compared to incorporation in serum-free medium, and are presented as the mean f S.E. ( n = 3).Non-stimulated incorporation represents 34,400 counts/min/mg protein.

FIG. 4 .FIG. 5 .
FIG.4.Binding characteristics of hGH (L) 108-129.3T3-F442A cells were incubated in binding medium containing 0.4 nM of radiolabeled hGH (L) 108-129 in the presence of indicated concentrations of competitors.Nonspecific binding activity was defined as the amount of ligand bound in the presence of a 1000-fold excess of competitor.,035 1

FIG. 7 .
FIG. 7. Binding of hGH (L) 108-129 in the presence of GHR antibody.Radiolabeled hGH (L) 108-129 was used as the binding ligand with the same protocol as Fig. 6.

*
M. Sonenberg, S. Guller, K-Y.Wu, R. E. Corin, and D. L. Allen, unpublished results.Scatchard analysis of lZ5I-hGH and T -h G H (L) 108-129 binding are 28 x lo3 and 560 x lo3.There may also be several species of GHRs to which hGH and hGH (L) 108-129 bind differently.It is also possible that hGH (L) 108-129 binds to