Growth hormone: species-specific stimulation of erythropoiesis in vitro.

The effects of purifed bovine and human growth hormone were tested in vitro with murine and human bone marrow by means of granulocyte-monocyte and erythroid progenitor cloning techniques. Nanogram concentrations of the growth hormones potentiated erythropoietin-stimulated erythropoiesis, but not granulopoiesis, in a species-specific manner.

Potentiation of erythropoiesis in vitro by dexamethasone.

The effect of dexamethasone on erythropoiesis was examined in vitro. Hematopoietic cells from 13-day mouse fetal livers were cultured for 48 h in the presence or absence of erythropoietin and erythroid colonies enumerated. Colony formation occurring in cultures containing no added erythropoietin was inhibited by the incorporation of antierythropoietin antibody, suggesting that these colonies formed in response to endogenous hepatic erythropoietin. Maximal colony formation was observed with 0.5 U/ml of sheep erythropoietin. Dexamethasone increased erythroid colony formation with peak stimulation at 10(-9) M. Dexamethasone potentiation was most marked in cultures containing less than maximally stimulating concentrations of erythropoietin. The cells required only a brief exposure to glucocorticosteroid to exhibit the augmented cloning capacity, and dexamethasone stimulation was inhibited by progesterone (10(-6) M). A comparable response to dexamethasone was observed in cultures of adult murine and human bone marrow erythroid progenitors, implying that the phenomenon is not peculiar to fetal cells and is not dependent on the presence of fetal hepatocytes. These data suggest that erythroid progenitor cells possess a glucocorticoid receptor mechanism that can modulate the response to erythropoietin in vitro.

Growth hormone induces resistance to the mitogenic action of insulin through local IGF-I. Studies in normal and Pygmy T-cell lines.

Growth hormone (GH) and insulin have both mitogenic and metabolic actions. The growth-promoting effects of GH in vivo are thought to be mediated by insulin-like growth factor-I (IGF-I), whereas the metabolic effects of GH are thought to be either direct or mediated by factors other than IGF-I. In previous studies using HTLV-II-transformed T-lymphoblast cell lines established from normal individuals, we have shown that GH preincubation induces resistance to the growth-promoting (mitogenic) action of insulin. In this study, using T-cell lines from 3 American control subjects, 1 African control subject, and 1 African Pygmy (the latter previously shown to be resistant to the growth-promoting actions of both IGF-I and GH), we examined the role of local IGF-I in the mediation of GH-induced resistance to the mitogenic action of insulin. In these studies, we quantified the stimulation of T-cell colony formation in response to insulin in the presence and absence of either GH or IGF-I.(ABSTRACT TRUNCATED AT 250 WORDS)

Use of in vitro clonogenic assays to differentiate acquired from genetic causes of insulin resistance.

Insulin resistance may be due directly to genetically programmed disorders of insulin action or acquired defects in which environmental factors influence insulin action. To address the issue of this distinction, we studied the ability of insulin to stimulate colony formation in primary cultures of erythroid progenitors (assumed to retain environmental influences) and immortalized T lymphocytes (presumed to reflect only genetic influences). Four patients with hyperinsulinemia and disturbed glucose metabolism were studied (2 patients with acanthosis nigricans, 1 of whom had circulating anti-insulin-receptor antibodies, 1 with partial lipodystrophy, and 1 with Cushing's syndrome). The mean colony-forming ability of their erythroid progenitor cells in response to insulin stimulation (less than or equal to 1.6 pM) was significantly blunted compared with control cells (P less than 0.05). The mean responsiveness of their immortalized T-lymphoblast cell lines to similar insulin concentrations was no different than that of control T-lymphocyte lines, consistent with an acquired cause for the observed insulin resistance in each case. A T-lymphocyte line from a patient with leprechaunism, however, showed no stimulation in response to physiological concentrations of insulin. With these same in vitro methodologies, there was normal T-lymphocyte line responsiveness to insulinlike growth factor I (IGF-I) or insulin concentrations greater than 8.6 pM; both of these responses could be completely blocked by preincubation with an antibody to the IGF-I receptor. These findings suggest that, despite resistance to physiological levels of insulin, the high circulating insulin concentrations present in the serum of these patients could mediate unwanted tissue-specific growth through an intact IGF-I receptor-effector mechanism.

Beta 2 receptor-mediated stimulation of Friend erythroleukemia cell growth by thyroid hormones.

Thyroid hormones are known to enhance normal erythroid colony growth (CFUE) and this enhancement depends on a functional beta 2-adrenergic receptor mechanism. we investigated the response of Friend cells to thyroid hormones, catecholamines, and other compounds influencing cellular cAMP activity. The thyroid hormones L-T3, L-T4, and "reverse T3" stimulated erythroleukemia colony growth in a serum-substituted methylcellulose culture system with peak activity at 10(-7) M. Various beta-adrenergic compounds enhanced Friend leukemia colony growth; however, the alpha-adrenergic agonist phenylephrine was inactive. Dibutyryl cyclic AMP and the phosphodiesterase inhibitor theophylline also enhanced Friend leukemia colony formation. Adrenergic antagonists with beta 2 specificity abrogated the stimulatory effect of L-T3, L-T4, and of "reverse T3" at equimolar concentrations. These experiments demonstrate that thyroid hormones, beta-adrenergic agonists, the phosphodiesterase inhibitor theophylline, and dbcAMP have a direct effect on the proliferation of Friend erythroleukemia cells. We conclude that thyroid hormones' action requires a functioning beta 2-adrenergic receptor mechanism. Thyroid hormones directly modulate the growth of neoplastic erythroid cells in a manner consistent with their effects on normal erythropoiesis.

Growth hormone induces insulin resistance in Laron dwarf cells via lactogenic receptors.

GH is the hormone primarily responsible for regulating body size within the genetic program. While GH has pleiotropic actions on cellular growth and metabolism, most of its effects are believed to be mediated by a single GH receptor. This receptor is not functional in tissues from patients with Laron dwarfism. We used human T-cell leukemia virus-immortalized T-lymphoblast cell lines from Laron dwarfs and normal individuals to examine the mechanism of GH-induced insulin resistance at the cellular level. GH (5-500 micrograms/L) caused a profound decrease in the sensitivity of normal T-lymphoblasts in response to all insulin concentrations (P < 0.0001 vs. insulin alone); pretreatment with GH and GH receptor antibody significantly improved sensitivity to all concentrations of insulin (P = NS vs. insulin alone). Preincubation with GH and PRL receptor antibody was associated with partial improvement in insulin sensitivity (P = 0.004 vs. insulin alone). Thus, in normal T-cell lines, the major pathway of GH-induced insulin resistance appears to be directed by the GH receptor, with a smaller effect mediated through the PRL receptor. While T-cell lines from Laron dwarfs do not respond to GH in clonal proliferation assays, GH (50 and 100 micrograms/L) caused profound insulin resistance in these cells (P = 0.008 and P < 0.0001, respectively, vs. insulin alone). GH receptor antibody did not abrogate this effect at any insulin concentration (P = NS vs. insulin alone), but there was partial restoration of insulin sensitivity when GH and PRL receptor antibody were coincubated (P = 0.0069 vs. insulin alone). Thus, in Laron T-cell lines, PRL and perhaps other lactogenic receptors appear to mediate GH-induced insulin resistance. The kinetics of GH-induced insulin resistance in Laron T-cells were also distinct from the pattern seen in normal T-cells, and unlike in normal cells, GH had no effect on insulin-like growth factor-I-induced clonal expansion of Laron T-cell lines (P = NS vs. insulin-like growth factor-I alone). These results provide evidence for an alternative pathway of GH action revealed in cells lacking classical growth responses to GH.

Natural and biosynthetic insulin stimulates the growth of human erythroid progenitors in vitro.

High concentrations of insulin are known to augment the growth of various cell types in vitro. We examined the effect of a purified porcine insulin and biosynthetic human insulin produced in E. coli on the growth of human erythroid progenitors in vitro. Both insulins stimulated peripheral blood erythroid colony formation within the physiological range. An approximately 2-fold augmentation in colony formation was seen at insulin concentrations of 8 ng/ml, and as little as 0.1 ng/ml (0.17 nM) caused detectable stimulation of colony formation. The effect of subnanomolar concentrations of insulin or erythropoiesis in vitro suggests that insulin could modulate erythropoiesis in vitro. Human responsiveness to insulin's growth-promoting activity can be directly assayed in vitro using peripheral blood.

Insulin-like growth factor I resistance in immortalized T cell lines from African Efe Pygmies.

Previous investigations suggested that resistance to GH was the cause of short stature of African Pygmies. Because many of the actions of GH are mediated by insulin-like growth factor I (IGF-I), we sought to determine whether Pygmy tissue was responsive to IGF-I. An initial effort to obtain HTLV-II-transformed T lymphoblast cell lines resulted in a single cell line that showed complete resistance to both IGF-I and GH in a clonal proliferation assay as well as decreased IGF-I binding. In the current study, we examined T cell lines from seven Efe Pygmy subjects, three neighboring Lese farmers, and six American controls and quantified clonal responses to IGF-I, GH, and insulin. The T cell lines from the Efe Pygmies were all completely resistant to the growth-promoting actions of IGF-I concentrations less than 250 micrograms/L and GH concentrations less than 500 micrograms/L. The Lese population, with whom there is admixture with the Efe population, showed heights and clonal responses to IGF-I and GH intermediate between those of Pygmies and American controls. The Pygmy T cell lines showed reduced clonal proliferation in response to high insulin concentrations known to act through the IGF-I receptor. These findings indicate that genetic IGF-I resistance is present in the T cell lines of Efe Pygmies and suggest that unresponsiveness to IGF-I may be responsible for their short stature.

T-lymphoblast cell lines from Laron dwarfs augment basal colony formation in response to extremely high concentrations of growth hormone.

The clinical entity of Laron dwarfism is characterized by resistance to both endogenous and exogenous GH and may be due to a deficiency or absence of functional GH receptors. We previously showed that two types of hematopoietic cells derived from these patients are resistant to the in vitro growth-promoting action of GH at concentrations below 500 micrograms/L. In the current study we found that Laron T-cell lines had a mean peak augmentation of basal colony formation of 22 +/- 3.4% above baseline in response to a GH concentration of 10,000 micrograms/L. Since cloned cDNAs for human and rabbit GH receptors and rat PRL receptors show a high degree of sequence homology, we undertook studies of PRL action in cells from patients with Laron dwarfism to determine if the Laron defect was also associated with PRL unresponsiveness. Quantitating the augmentation of colony formation by T-lymphoblast cell lines established from three Laron dwarfs, we found normal responsiveness to PRL at concentrations of 25-10,000 micrograms/L. It is, thus, possible that the responsiveness of Laron T-cell lines to very high concentrations of GH could be mediated through an intact PRL (or other lactogenic) receptor based on the known affinity of GH for these receptors in other systems. These data suggest that cells from patients with Laron dwarfism have normal in vitro responsiveness to PRL and that the defect in Laron dwarfism appears to be specific to the GH receptor-effector pathway. It remains to be determined whether intact alternative lactogenic receptor mechanisms subserve any clinical effects of GH in patients with Laron dwarfism.

A unique growth factor in patients with acromegaloidism.

Acromegaloidism is a syndrome characterized by features of acromegaly without biochemical evidence of excessive GH or somatomedin production. We searched for a growth factor in the serum of patients with this syndrome. Growth-promoting activity was measured by determining the stimulatory effect of whole and fractionated serum on colony formation by human erythroid progenitors in vitro. Sera from five subjects with acromegaloidism gave a mean (+/- SEM) stimulated colony growth of 211 +/- 4.0 colonies, in contrast to normal sera which yielded a mean colony growth of 100 +/- 11.0 (n = 9; P less than 0.001). When serum was chromatographed on a Sephadex G-200 column, the maximal stimulation of colony growth was found in the fractions coinciding with the descending slope of the second protein peak. Based on gel filtration chromatography, the estimated molecular weight was 70,000 daltons. Epidermal growth factor, nerve growth factor, fibroblast growth factor, and platelet-derived growth factor resulted in no substantial stimulation of colony growth under the conditions used. Although the erythroid progenitor cells of a Laron dwarf were unresponsive to 200 ng/ml human GH, they were clearly stimulated by serum from a patient with acromegaloidism. The present study describes the presence of a heretofore unidentified growth factor in the serum of subjects with acromegaloidism. This factor also stimulated the erythroid precursor cells of a Laron dwarf whose cells were unresponsive to GH. The physiological role of this growth factor in normal man as well as its pathogenic role in subjects with acromegaloidism remain to be established.

Growth hormone mediates the growth of T-lymphoblast cell lines via locally generated insulin-like growth factor-I.

It has become evident that locally produced insulin-like growth factors-I and -II (IGF-I and IGF-II) play an important role in the medication of GH action upon tissues. To explore this concept with respect to immunocompetent cells, we analyzed IGF production and clonogenic responsiveness of immortalized human T-cell lines established from seven normal controls and four Laron dwarfs. While the normal T-cell lines showed significant augmentation of basal colony formation in response to both IGF-I and GH, little increase in clonogenesis in response to GH was seen with the Laron T-cell lines. Assay of basal and GH-stimulated conditioned media demonstrated low, but measurable, levels of IGF-I and IGF-II from both normal and Laron T-cells. Under serum-free incubation conditions, GH stimulation of normal T-cell lines failed to generate significant increases in mean IGF-I or IGF-II concentration and no increase in the mean IGF-II concentration in conditioned medium were observed after GH stimulation of Laron T-cell lines. Nevertheless, the increased cloning efficiency of the normal T-cell lines in response to either GH or IGF-I was nearly completely abrogated by preincubation of cells with antibodies to either IGF-I or the type I IGF receptor. These studies, thus, support a role for locally generated IGF-I in the mediation of GH action on T-lymphocytes and indicate that this effect is mediated via the type I IGF receptor.

Decreased insulin-like growth factor I receptor expression and function in immortalized African Pygmy T cells.

Efe Pygmies of northeast Zaire have the shortest mean adult stature of any population on earth. Although various alterations in the GH/insulin-like growth factor I (IGF-I) axis have been suggested, the basis for short stature in the Pygmy is unknown. We previously described IGF-I unresponsiveness in a T lymphoblast cell line derived from an Efe Pygmy, and studies in five additional lines have confirmed severe IGF-I resistance in these cells. We have now performed experiments to determine the molecular basis for the IGF-I resistance in these cells. We found markedly decreased cell surface expression of IGF-I receptors with normal ligand binding affinity. The Pygmy IGF-I receptors were not autophosphorylated and did not transmit a signal in response to physiological concentrations of IGF-I. There was a substantially decreased level of IGF-I receptor messenger ribonucleic acid in the Pygmy cells with a normal messenger ribonucleic acid half-life. The nucleotide sequence of the full-length IGF receptor complementary DNA in Pygmy 1 showed no significant variation. These results indicate decreased IGF-I receptor gene transcription and IGF-I receptor signaling as the primary variation in the Pygmy cell lines. The findings point to the IGF-I receptor as the locus governing short stature in the African Pygmy and suggest that human stature may be genetically controlled by expression of the IGF-I receptor.

Diminished in vitro responsiveness of circulating erythroid progenitor cells to insulin as an indicator of insulin resistance.

While insulin resistance is considered characteristic of extreme obesity, it may be more difficult to demonstrate in less severe forms of obesity. We studied five moderately obese individuals [mean body mass index (MBMI), 34.1 +/- 1.85 (+/- SE) kg/m2], one massively obese patient (BMI, 50.2 kg/m2), and seven age-matched normal subjects (MBMI, 22.4 +/- 0.93 kg/m2). While two of the obese patients had normal glucose tolerance, all had fasting hyperinsulinemia (P less than 0.02 vs. normal subjects) and exaggerated insulin responses after oral glucose challenge, as defined by area under the 3-h insulin response curve (P less than 0.01 vs. normal subjects). That this hyperinsulinemia represented in vivo insulin resistance was supported by the glucose and insulin responses in four individuals to an iv glucose bolus analyzed by the minimal modeling technique. Study of monocyte insulin receptors revealed no reduction in total insulin binding in the four obese patients tested. Since physiological concentrations of insulin stimulate the in vitro growth of normal human erythroid progenitor cells (EPC), we reasoned that this response might be blunted in cells from individuals with endogenous insulin resistance. The mean peak EPC proliferative response (26.7 +/- 9.11% above baseline) in the obese hyperinsulinemic group was significantly less than the corresponding mean value in the control group (92.6 +/- 5.24% above baseline, P less than 0.001). These results suggest that the minimal modeling technique is a sensitive method for the in vivo demonstration of insulin resistance in moderately obese individuals and that EPC responsiveness to physiological concentrations of insulin reflects in vivo insulin sensitivity and may be used as an in vitro indicator of insulin resistance.

Tissues of the Laron dwarf are sensitive to insulin-like growth factor I but not to growth hormone.

Tissues from patients with Laron dwarfism are resistant to the actions of endogenous or exogenous GH. As a result, insulin-like growth factor I (IGF-I) levels are low, possibly contributing to the severe growth deficiency that occurs in patients with this syndrome. In this study, we found that erythroid progenitor cells and permanently transformed T-cell lines from two patients with Laron dwarfism responded in vitro to added IGF-I in concentrations ranging between 1-10 ng/mL despite no stimulatory response to added GH in concentrations of up to 500 ng/mL. Normal or near-normal responsiveness to insulin was also demonstrated. The persistence of GH resistance in the cultured T-cell lines confirms the primary genetic nature of the defect in Laron dwarfism. The preservation of in vitro growth responsiveness to IGF-I in hematopoietic tissue from the Laron dwarfs suggests that affected individuals are sensitive to this factor and may respond to it in vivo.

Tissue inhibitor of metalloproteinase-2 (TIMP-2) has erythroid-potentiating activity.

Tissue inhibitor of metalloproteinase (TIMP) was purified and molecularly cloned on the basis of its erythroid-potentiating activity (EPA). TIMP/EPA appears to be a bifunctional molecule with both growth factor and anti-enzymatic activity. Recently, a second TIMP-related molecule was identified and we have investigated its possible erythroid-potentiating activity. Native, purified human TIMP-2 was assayed for erythroid-potentiating activity using an in vitro erythroid burst formation assay and was compared with that of previously characterized recombinant EPA/TIMP-1. The results demonstrate that both members of the tissue inhibitor of metalloproteinase family, TIMP-1 and TIMP-2, possessed erythroid potentiating activity which was inhibited by antibodies developed to neutralize EPA. These results suggest that TIMP-2 shares a common structural domain with EPA/TIMP-1 that is responsible for the erythroid-potentiating activity of these inhibitors. Therefore, TIMP-1 and TIMP-2, with both anti-protease activity and growth factor activity, join a family of bifunctional molecules such as fibroblast growth factor and thrombin which have both enzymatic and growth factor activity.

Persistence of insulin resistance in polycystic ovarian disease after inhibition of ovarian steroid secretion.

Six nonobese women with polycystic ovarian disease (PCOD) showed significant hyperinsulinemia, compared with controls after oral glucose (P less than 0.05). As an indicator of insulin sensitivity, in vitro proliferation of erythrocyte progenitor cells of PCOD subjects exposed to physiologic concentrations of insulin was significantly blunted (P less than 0.001). Monocyte insulin receptor binding was not impaired in the PCOD subjects. Three of the PCOD patients were treated with a long-acting gonadotropin-releasing hormone agonist for 6 months, which resulted in marked suppression of ovarian androgen secretion but no demonstrable changes in in vivo or in vitro indicators of insulin resistance. Thus insulin resistance in PCOD subjects appears to be unrelated to ovarian hyperandrogenism (or acanthosis or obesity). Although certain tissues are insulin-resistant in PCOD patients, the ovary may remain sensitive and overproduce androgens in response to high circulating insulin levels.

Insulin and insulin-like growth factor-I responsiveness in polycystic ovarian syndrome.

OBJECTIVE: To assess insulin and insulin-like growth factor I (IGF-I) action in women with polycystic ovarian syndrome (PCOS). DESIGN: Hyperinsulinemia was determined by measuring the insulin responses during a 2-hour oral glucose tolerance test (OGTT). Quantification of in vivo insulin action was determined by a frequently sampled intravenous (IV) OGTT with minimal modeling analysis. In vitro sensitivity to insulin at physiological and supraphysiological concentrations and to IGF-I was assessed by examining colony formation of two hematopoietic cell populations, burst-forming units of the erythroid line (BFU-E) and human leukemia virus immortalized T-cell lines. (The proliferative responses of BFU-E, a primary tissue explant, are presumably conditioned by factors in the immediate blood-borne environment, whereas proliferative responses of T-cell lines are presumed to reflect intrinsic target-cell hormone sensitivity.) SETTING: Tertiary care research institution. PATIENTS: Eight patients (4 obese and 4 nonobese) with PCOS and three healthy women for reference controls. RESULTS: Nonobese (P less than 0.04) and obese patients with PCOS (P less than 0.01) both demonstrated significant hyperinsulinemia after OGTT. In vivo insulin resistance was observed in both nonobese (P less than 0.03) and obese PCOS subjects (P less than 0.01) using frequently sampled IV OGTT. Both nonobese (P less than 0.03) and obese patients with PCOS (P less than 0.01) had blunted in vitro clonal responses of BFU-E, with normal T-cell line clonal responsiveness to physiological levels of insulin and normal BFU-E and T-cell line clonal responses to IGF-I. CONCLUSIONS: These findings demonstrate the following in both nonobese and obese patients with PCOS: (1) there is in vivo hyperinsulinemia and resistance to insulin action on glucose disposal; (2) with BFU-E, there is in vitro resistance to the mitogenic action of insulin but normal responsiveness to IGF-I; and (3) there is normal in vitro mitogenic responsiveness of T-cell lines to both insulin and IGF-I. The intrinsically normal mitogenic responsiveness to insulin and, especially to IGF-I, whether or not under the influence of the bloodborne milieu, provides a mechanism whereby hyperinsulinemia could directly contribute to the ovarian abnormalities that characterize PCOS.

Insulin and IGF-I stimulate normal and virally transformed T-lymphocyte cell growth in vitro.

We used normal and HTLV-II-transformed T-lymphocytes as target cells to study clonal proliferative responses to physiologic and supraphysiologic concentrations of insulin and IGF-I. Responses of both growth factors were measured in the presence and absence of alpha IR-3, and IGF-I receptor-blocking antibody. A biphasic response to insulin was noted in all cell lines with the first peak [78 +/- 6.6% (mean +/- SE) above control] occurring at 1.4 or 1.6 nmol/liter and a second peak (84 +/- 4.9% above control) occurring at 18.0 nmol/liter. Following preincubation with alpha IR-3, the overall clonal profile in response to insulin was significantly reduced [F(7,56) = (10.4, p < .0001] as a result of blunting at high physiologic and supraphysiologic insulin concentrations, i.e., > or = 1.6 nmol/liter. As expected, the overall clonal profile in response to IGF-I was blocked by alpha IR-3 [F(4,32) = 11.6, p < .0001]. These data show that insulin at both physiologic and supraphysiologic concentrations, as well as IGF-I, stimulate virally transformed T-lymphoblast growth. The significant inhibition of growth responses to high concentrations of insulin and to IGF-I by alpha IR-3 suggests mediation of these effects through the IGF-I receptor. Similar studies were performed using freshly isolated, phytohemagglutinin (PHA)-stimulated T-lymphocytes. The overall response to insulin was significantly reduced compared to the profile of transformed T-lymphoblasts [F(7,70) = 4.9, p = .0002] as a result of blunting at physiologic insulin concentrations < 1.8 nmol/liter. In response to IGF-I, the clonal profile of PHA-stimulated T-lymphocytes was slightly reduced compared to that of virally transformed T-lymphoblasts [F(4,40) = 3.4, p = .0174]. Thus, both insulin and IGF-I receptor-effector mechanisms are involved in the growth of virally transformed T-lymphoblasts, whereas the IGF-I receptor-effector mechanism appears to play a more significant role in the growth of normal, mitogen-activated T-lymphocytes.

Characterization of purified human erythroid-potentiating activity.

Erythroid-potentiating activity (EPA), purified from serum-free medium conditioned by the Mo human T-lymphoblast cell line, is a 28,000 dalton glycoprotein. Removal of carbohydrate from EPA by digestion with endoglycosidase F generates a protein with an apparent molecular weight of 18,000. We have used purified EPA to develop rabbit antisera which neutralize the burst-promoting activity of purified EPA. These antisera immunoprecipitate a 28,000-dalton glycoprotein from metabolically labeled Mo cells. Purified EPA stimulates murine and human BFU-E and CFU-E but not myeloid or macrophage colonies.

Growth hormone modulation of murine erythroleukemia cell growth in vitro.

There are few studies showing a biological effect of growth hormone (somatotropin) on cell proliferation in vitro at physiological concentrations. We report here that Friend virus-infected erythroleukemia cells are responsive to growth hormone in vitro. Using a serum-free clonogenic assay we found as little as 0.1 ng of human growth hormone per ml caused a prominent stimulation of cell proliferation. Peak activity of human growth hormone occurred at 200 ng/ml, resulting in a 2-fold increase in cloning. Human chorionic somatomammotropin and the Cys(Cam)53-hGH(1-134) fragment of human growth hormone were also active, but a biologically inert oxidized human growth hormone had no growth-promoting effect in vitro. Cell proliferation was stimulated by insulin with peak potentiation occurring at 1 ng/ml, and prolactin had a demonstrable stimulatory effect between 50 and 100 ng/ml. These observations indicate that growth hormone and related polypeptides have a direct effect on the in vitro proliferation of erythroleukemia cells in the absence of serum. The results confirm a direct action of growth hormone on mammalian cells and suggest that pituitary hormones may affect leukemic cell growth.