Imitation of contact inhibition by substrate-bound plasma membrane glycoproteins and lectins in serum-free hormone-supplemented cultures of GH3 cells

N-Acetylneuraminic acid-bearing glycoproteins and lectins isolated from the plasma membrane of GH₃ cells are coupled covalently to glass dishes and are used as substrates for cell culture. Under serum-free, hormone-supplemented culture conditions these coupled molecules inhibit the growth rate of GH₃ cells. This inhibition is non-toxic and is concentrationand time-dependent. For an inhibition of 50% 2 × 10⁶ coupled molecules contacting one cell are needed. The data suggest that specific plasma membrane-located glycoproteins and lectins are involved in the ‘density-dependent growth regulation’.     

Changes of the internal organization of the plasma membrane correlated to the regeneration potency of the cell

Internal organization of the plasma-membrane of rat thymocytes and pituitary tumor cells (GH₃) was experimentally altered by low temperature and either changing osmolarity or adding drugs destroying the cytoskeleton. These treatments induce reversible aggregation of intramembrane particles of the plasma membrane. Thin section electron microscopy of the hypotonically shocked cells show all the cell organelles to be extremely swollen and the cytoplasmic space to be rather empty. Returned to physiological conditions, the cells show normal morphological aspects, accompanied by ‘normal’ permeability properties of the plasma membrane; the aggregation of intramembrane particles is reversed. The proliferation behaviour of GH₃-cells is not affected by this treatment. This demonstrates a high regeneration potency of the mammalian cell and leads to the assumption that molecular components which are important for the survival of the cell must be structural (membrane) bound.     

Saccharin induces morphological changes and enhances prolactin production in GH4C1 cells

Summary The artificial sweetener saccharin inhibits binding of epidermal growth factor (EGF) to cultured rat pituitary tumor cells (GH₄C₁ cells). Saccharin also causes morphological alterations in these cells, resulting in pronounced elongation, stretching, and firmer attachment of cells to the culture dishes. These alterations in cell shape are similar to those observed after treatment of GH₄C₁ cells with EGF and with thyrotropin-releasing hormone (TRH), both of which enhance prolactin (PRL) production in these cells. After assaying for PRL in saccharin-treated cultures, it was observed that this sweetener is also capable of stimulating PRL production two-to sixfold in a dose-dependent manner. Enhancement of PRL production can be observed at 0.5 mM saccharin, yet this is 10 times less than the saccharin concentration required to alter cell shape. These effects of saccharin on cell morphology and on PRL production are reversible in GH₄C₁ cell cultures. When added to cultures along with maximal concentrations of EGF or TRH, the effects of saccharin on PRL production are additive, suggesting that the actions of saccharin are mediated by a somewhat different pathway from that of the peptide hormones. Pulse labeling studies indicate that the enhancement of PRL production is highly specific inasmuch as saccharin was found to decrease the overall rate of protein synthesis in these cells. Saccharin also causes a decrease in the rate of DNA synthesis under these treatment conditions. Mitomycin C, which similarly inhibited DNA synthesis, had no effect on cell morphology or PRL production. This investigation was supported by a Faculty Research Grant from Wheaton College     

Hormonal growth control of cells in culture

Summary Serum is the last undefined component in cell culture media. Our results indicate that the primary role of serum is to provide hormones and that serum can be replaced by a group of hormones. A rat pituitary cell line, GH₃, can grow in serum-free medium if the medium is supplemented with 3,3′,5-triiodothyronine, TSH-releasing hormone, transferrin, parathyroid hormone, insulin and three isoelectric focusing fractions of blood meal. The blood-meal components can be replaced by fibroblast growth factor and somatomedin C. The growth rate of GH₃ cells in hormone-supplemented serum-free medium is equal to that in serum-supplemented medium, and subculture in such medium is also possible. These results indicate that the replacement of the serum component is complete in the GH₃ system. The hormonal requirements of GH₃ cells and those of HeLa and mouse melanoma, M2R, were compared. Two generalizations could be made: (a) All three cell lines require insulin and transferrin. (b) There is a requirement for a hormone which localizes in the nucleus for each cell line. These generalizations seem to hold true for most of the other cell lines for which the hormonal requirements have been partially worked out. Since insulin is one of the universally required hormones, its effects on GH₃, HeLa and M2R were compared. Insulin stimulates glycogen synthesis in all three cell lines and facilitates fatty-acid synthesis in GH₃ and M2R. However, there is a difference in the effect of insulin on growth among the three cell lines. Insulin is an absolute requirement for GH₃ cells without which the cells cannot survive, whereas this is not the case for HeLa and M2R. The most stringent requirement for HeLa cells is for hydrocortisone, and for M2R, it is for transferrin. These results indicate that even though the necessity for some hormones is common, the degree of requirement may vary from one cell line to another. Whether this difference reflects the difference in the primary mode of action of the hormone on each cell type needs further investigation. Presented in the Opening Symposium on Nutritional Factors and Differentiation at the 28th Annual Meeting of the Tissue Culture Association, New Orleans, Louisiana, June 6–9, 1977. This work was supported by NIH Grant GM 17019. J. Larner was supported by Josiah Macy Foundation     

Neural precursors derived from human embryonic stem cells

Human embryonic stem (hES) cells provide a promising supply of specific cell types for transplantation therapy. We presented here the method to induce differentiation of purified neural precursors from hES cells. hES cells (Line PKU-1 and Line PKU-2) were cultured in suspension in bacteriological Petri dishes, which differentiated into cystic embryoid bodies (EBs). The EBs were then cultured in N₂ medium containing bFGF in poly-L-lysine-coated tissue culture dishes for two weeks. The central, small cells with 2–3 short processes of the spreading outgrowth were isolated mechanically and replated. The resulting neurospheres were cultured in suspension for 10 days, then dissociated into single cell suspension with a Pasteur pipette and plated. Cells grew vigorously in an attached way and were passed every 4–5 days. Almost all the cells were proved nestin positive by immunostaining. Following withdrawal of bFGF, they differentiated into neurons expressing β-tubulin isotype_III, GABA, serotonin and synaptophysin. Through induction of PDGF-AA, they differentiated into astrocytes expressing GFAP and oligodendrocytes expressing O₄. The results showed that hES cells can differentiate into typical neural precursors expressing the specific marker nestin and capable of generating all three cell types of the central nervous system (CNS)in vitro.     

Autocrine-paracrine inhibition of growth hormone and prolactin production by GH3 cell-conditioned medium

Summary In previous work we have shown that perifused GH₃ cells exhibit spontaneously accelerating growth hormone (GH) and prolactin (PRL) secretory rates. This behavior contrasts with GH and PRL secretion rates that are decreasing or stable over the same 3-d period in static cell culture. We now report that GH₃ cells maintained in serum-supplemented medium produce an autocrine-paracrine factor(s) which inhibits GH secretion in plate culture; PRL release is frequently reduced as well. The inhibitory effect of conditioned medium on GH secretion was concentration dependent, whereas PRL release was stimulated at low and inhibited at high concentrations over the same range. Extensive dialysis of conditioned medium using membranes with a molecular weight cut-off of 12 000–14 000 did not remove GH inhibition but produced a retentate that stimulated PRL secretion. Heat-inactivation of conditioned medium did not abolish inhibition of GH release but did remove the PRL-stimulatory effect. IGF-I added to fresh culture medium did not reproduce the GH-inhibitory effects of conditioned medium. We conclude that GH₃ cell secretory behavior in perifusion and plate culture systems may be partially explained by the production of an autocrine-paracrine factor: its accumulation in plate culture inhibits GH and PRL secretion whereas its removal, by perifusing medium, allows GH and PRL secretion to accelerate. Supported by grant DK33388 to M. E. S. from the National Institute of Health, Bethesda, MD, and in part by the Medical Research Service of the Veterans Administration, Washington, DC.     

Induction of distinct phenotypes in clonal and variant GH4 pituitary cells

Summary GH cells are a widely used cell strain for the investigation of mechanisms regulating hormone release and synthesis. This report identifies two inducible phenotypes of the GH₄ clone (epithelioid and motile) which may extend studies of this well-characterized cell line to different stages of pituitary cell development. GH₄C₁ cells treated in suspension with epidermal growth factor plus tetradecanoylphorbol acetate aggregate to form large epithelioid colonies with extensive cell-to-cell and cell-to-substratum adhesion. These cells cease replicating within 48 h, increase 50% in cell volume, and synthesize 40-fold more prolactin. A GH₄C₁ variant with enhanced substratum adhesion and little or no cell-to-cell adhesion (GH₄S₁), responds differently to this treatment. These cells cease replicating immediately, show increased cell separation, develop leading lamellae, and display locomotory activity. Each phenotype coexists in mixed cultures of GH₄C₁ and GH₄S₁ cells. This indicates that the different inducible response of the variant does not result from autocrine secretion. A molecular basis for cell-to-cell adhesion in GH₄ cells was investigated. GH₄C₁, but not the variant cells, express a 180 kDa immunoreactive protein indistinguishable from an isoform of the neural cell adhesion molecule. Therefore the absence of cell-to-cell adhesion and inability to develop extensive cell-to-cell adhesion characteristic of the epithelioid phenotype may result from altered expression of the neural cell adhesion molecule. These findings are important because they have defined an in vitro approach to investigate genetic and cellular changes associated with the development and progression of pituitary cell phenotype. This study was initiated in the laboratory of Dr. A. H. Tashjian, Jr., under the support of grant DK11011 from the National Institutes of Health, Bethesda, MD. The completion of this study was supported by the Medical University of South Carolina Biomedical Support Grant of 1987–1988 and the American Cancer Society grant 1N-175.     

Rat pituitary tumor cells in serum-free culture. I. Selection of thyroid hormone-responsive and autonomous cells

Summary The growth of GH₄C₁, GH₃, GH₁, and GH₃C₁₅ rat pituitary tumor cell lines was studied in a serum-free medium (designated TRM-1) formulated with 1∶1 (vol/vol) mixture of Ham's F12 nutrient mixture and Dulbecco's modified Eagle's medium (F12-DME) containing 15 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 50 μg/ml gentamicin supplemented with 10 μg/ml bovine insulin, 10 μg/ml human transferrin (Tf), 10 ng/ml selenous acid, 10 nM 3,5,3′-triiodothyronine (T₃), 50 μM ethanolamine (Etn), and 500 μg/ml bovine serum albumin. Of the lines evaluated, only the GH₁ failed to grow in TRM-1. Passage of the GH₄C₁ and GH₃ lines from serum-containing medium into TRM-1 caused an initial selection resulting in cells that grew progressively at higher rates and finally were maintained indefinitely in TRM-1. These populations showed a requirement for supraphysiologic concentrations of T₃ (1.0 to 10 nM). After adaptation of the GH₄C₁ line in TRM-1 for ≤20 generations, removal of components gave a less complex mixture containing 15 mM HEPES, 50 μ/ml gentamicin, 10 μg/ml Tf, 10 nM T₃, and 50 μM Etn (designated TRM-2) that supported serial passage of the cells. Under these conditions, thyroid hormone dependence was lost progressively. When T₃ was removed from TRM-2 adapted cells, a third population was selected that no longer required thyroid hormones and was only slightly stimulated by T₃. These studies demonstrated that the combination of serum-containing and serum-free conditions can be used to select pituitary cell populations that a) required both serum-factor(s) and T₃ for optimum growth, b) required supraphysiologic concentrations of T₃ without serum proteins other than Tf and albumin, and c) were completely autonomous in that they proliferated in medium supplemented only with Tf and nutrients without necessity of other serum factor(s) or T₃. This work was supported by grants CA-26617 and CA-38024 from the National Cancer Institute, Bethesda, MD, American Cancer Society grant BC-255, and grant 2225 from the Council for Tobacco Research, Inc., USA.     

Intrinsic tryptophan fluorescence of membranes of GH3 pituitary cells: quenching by thyrotropin-releasing hormone.

The intrinsic tryptophan fluorescence of membranes prepared from the GH₃ strain of hormone-producing pituitary cells was monitored by spectrofluorometry. Membranes of GH₃ cells have specific receptors which bind thyrotropin-releasing hormone (TRH). When TRH binds to GH₃ membranes there is quenching of tryptophan fluorescence. The kinetics of the change in fluorescence of GH₃ membranes and of TRH binding are similar. In addition, the concentration of TRH required to produce a half-maximum change in fluorescence is 10 nM, and that required for half-maximum binding of TRH to receptors is 11 nM. Inactive TRH analogs which do not bind to TRH receptors likewise do not alter GH₃ membrane fluorescence, and a pituitary cell strain which lacks TRH receptors does not change membrane fluorescence on incubation with TRH. We conclude that the TRH-receptor interaction in GH₃ membranes is associated with a change in membrane conformation that is readily measured by differential spectrofluorometry.     

Stereological and biochemical analysis of prolactin and growth hormone secreting rat pituitary cells in culture

Summary The secretion of prolactin is increased by treatment of prolactin producing rat pituitary cells with the hypothalamic tripeptide thyroliberin. To investigate the underlying mechanisms we used three closely related rat pituitary tumor cell strains (GH₁2C₁, GH₃ and GH₄C₁), which synthesize and spontaneously secrete prolactin and/or growth hormone. Growth hormone and prolactin released into the culture medium over a period of 24 h were measured by radioimmunoassay. Initial rates of synthesis were measured by immunoprecipitation of intracellular growth hormone and prolactin after incubation of cell cultures with ³H-leucine. The observed increase in prolactin synthesis and release was correlated with morphological effects of thyroliberin treatment. The volume density of Golgi complexes and the volume and surface densities of rough endoplasmic reticulum were compared in untreated cells and thyroliberin treated cells. As normal distribution could not be assumed, the non-parametric rank test of Wilcoxon was used whereby the densities calculated for each cell section were ranked. Alle three morphological parameters increased after thyroliberin treatment in cells secreting prolactin only (GH₄C₁), implying that the increase of prolactin secretion, at lest in part, is due to increased prolactin synthesis.     

Neural precursors derived from human embryonic stem cells

Before the successful isolation of human embryonic stem (hES) cells, many investigations had shown that mouse embryonic stem (mES) cells can be induced to differentiate into neural precursors which could be purified and differentiated to mature dopamine, motor, serotonin, GABA neurons, and oligodendrocytes and astrocytes in vitro[1―3]. mES cell-derived dopamine neurons have been shown capable of integrating into host brains after transplanting to the rodents of Park-inson’s disease model …     

Similarity Between Rat Brain Nicotinic α-Bungarotoxin Receptors and Stably Expressed α-Bungarotoxin Binding Sites

Abstract: The present results demonstrate stable expression of α-bungarotoxin (α-BGT) binding sites by cells of the GH₄C₁ rat pituitary clonal line. Wild-type GH₄C₁ cells do not express α-BGT binding sites, nor do they contain detectable mRNA for nicotinic receptor α2, α3, α4, α5, α7, β2, or β3 subunits. However, GH₄C₁ cells stably transfected with rat nicotinic receptor α7 cDNA (α7/GH₄C₁ cells) express the transgene abundantly as mRNA, and northern analysis showed that the message is of the predicted size. The α7/GH₄C₁ cells also express saturable, high-affinity binding sites for ¹²⁵I-labeled α-BGT, with a K_D of 0.4 nM and B_max of 3.2 fmol/10⁶ intact cells. ¹²⁵I-α-BGT binding affinities and pharmacological profiles are not significantly different for sites in membranes prepared either from rat brain or α7/GH₄C₁ cells. Furthermore, K_D and K_i values for ¹²⁵I-α-BGT binding sites on intact α7/GH₄C₁ cells are essentially similar to those for hippocampal neurons in culture. Sucrose density gradient analysis showed that the size of the α-BGT binding sites expressed in α7/GH₄C₁ cells was similar to that of the native brain α-BGT receptor. Chronic exposure of α7/GH₄C₁ cells in culture to nicotine or an elevated extracellular potassium concentration induces changes in the number of α-BGT binding sites comparable to those observed in cultured neurons. Collectively, the present results show that the properties of α-BGT binding sites in transfected α7/GH₄C₁ cells resemble those for brain nicotinic α-BGT receptors. If the heterologously expressed α-BGT binding sites in the present study are composed solely of α7 subunits, the results could suggest that the rat brain α-BGT receptor has a similar homooligomeric structure. Alternatively, if α-BGT binding sites exist as heterooligomers of α7 plus some other previously identified or novel subunit(s), the data would indicate that the α7 subunits play a major role in determining properties of the α-BGT receptor.     

Interaction of calcium and cyclic nucleotides in thyroliberin-stimulated prolactin release

The object of the present study was to determine the relative importance of Ca++ and cyclic nucleotides as “second messengers” in thyroliberin (TRH)-mediated prolactin (PRL) release in the GH₃ and GH₄ rat pituitary tumor cell lines. PRL, cyclic adenosine 3': 5'-monophosphate (cAMP), and cyclic guanosine 3': 5'-monophosphate (cGMP) were measured by radioimmunoassay (RIA) following TRH stimulation. TRH increased PRL release and cAMP levels in GH₃ and GH₄ cells, but cGMP increases were variable. Treatment with 1 mM theophylline increased PRL release and raised cAMP and cGMP. Addition of TRH to theophylline-pretreated cells produced further significant increases in PRL release without any additional increases in cAMP and cGMP. Co++, a Ca++ antagonist, abolished TRH-induced PRL release in a dose-dependent manner. The Co++ inhibition was partially reversed by Ca++ in GH₃ or GH₄ cells. Furthermore, the Ca++ ionophore A23187 stimulated PRL release. We conclude that Ca++ is the primary “second messenger” for TRH-mediated PRL release from GH₃ or GH₄ cells.     

The Lectins of Sophora japonica: II. Purification, Properties, and N-Terminal Amino Acid Sequences of Five Lectins from Bark

Five N-acetyl-galactosamine-specific lectins were isolated from the bark of the legume tree Sophora japonica. These lectins are immunologically and structurally very similar, but not identical, to the Sophora seed and leaf lectins. The carbohydrate specificities and hemagglutinin activities of these lectins are indistinguishable at pH 8.5 but their activities differ markedly at pH values below 8. All five lectins are tetrameric glycoproteins made up of different combinations of subunits of about 30,000, 30,100, 33,000 M_r containing 3% to 5% covalently attached sugar. These lectins are the overwhelmingly dominant proteins in bark, but they do not appear to be present in other tissues. Amino terminal sequence analysis indicates that at least two distinct lectin genes are expressed in bark.     

Medium flow rate modulates autocrine-paracrine feedback of GH and PRL release by perifused GH3 cells

Summary We previously documented both the spontaneous acceleration of growth hormone (GH) and prolactin (PRL) production by GH₃ cells during periffusion and the suppression of their production during plate culture. We here present the role played by medium flow itself in this differential behavior. Increasing rates of perifusion flow (pump rates of 1 to 5 ml/h, equivalent to chamber flow rates of 0.19 to 1.3 μl·min⁻¹·mm⁻² of cross-sectional area) were associated with enhanced GH and PRL secretion. Flow rate-dependent basal hromone secretion rates were established quickly and were stable for the first 10 to 14 h of perifusion. The previously documented independent, spontaneous, and continuously accelerating production of both hormones that followed during the subsequent 40 (PRL) to 60 (GH) h of perifusion was also shown to be flow-rate related. Any time the rate of medium flow was changed within an experiment, the rate of hormone secretion was modulated. However, that modulation did not interrupt ongoing flow-associated acceleration of hormone production once the latter had begun. In addition, GH₃ cell product(s) from one cell column reversibly inhibited secretion from cells in a downstream column. The inhibition did not occur when cells in the downstream column had been exposed to trypsin. Other work had suggested that neither GH, PRL, insulinlike growth factor-I, leucine, nor nutrient exhaustion were responsible for the effect. These data are consistent with autocrine-paracrine feedback regulation of GH₃ cells by a secretory product(s). Feedback would thus provide a mechanism to effect flow-rate-dependent modulation of GH and PRL release, and to explain accelerating hormone production during perifusion. This work was supported by a grant to M. E. S. from the National Institutes of Health (DK33388), Bethesda, MD, and in part, by the Medical Research Service of the Veterans Administration.     

Effects of TRH on cell proliferation of rat pituitary cells (GH3)

Summary Chronic treatment (more than 3 d) of GH₃ cells, cloned rat pituitary cells producing prolactin, with 100 nM TRH resulted in a 41% reduction in the rate of cell growth in a medium containing 0.5% fetal bovine serum. These effects of TRH appeared both in the medium containing a higher concentration of serum and in that containing six growth factors, i.e. insulin, transferrin, parathyroid hormone, fibroblast growth factor, triiodothyronine, and multiplication-stimulating activity (MSA) instead of serum. TRH stimulated prolactin production by GH₃ cells in a dose-dependent manner both in the serum-supplemented and serum-free media. On the other hand, TRH, at 1 nM, elicited a 130% stimulation in the cellular growth, whereas, at concentrations of more than 10 nM, it inhibited the growth significantly. In the defined culture system, it was demonstrated that TRH stimulated prolactin production in the presence or absence of six growth factors, whereas its inhibitory effects on cellular growth appeared only in the presence of MSA regardless of the presence or absence of the other five factors. Furthermore, it was shown that a dose-dependent stimulatory effect of MSA on the growth of GH₃ cells was suppressed by TRH. TRH exhibited only a stimulatory effect on cellular growth in the medium containing the five factors other than MSA. In conclusion, TRH could inhibit cell growth of GH₃ in the presence of MSA in the defined medium or MSA-like factor(s) in the serum-supplemented medium.     

Release of 6-keto-prostaglandin F1α and thromboxane B2 from mouse peritoneal macrophages during their adhesion and spreading on a glass surface

The release of 6-keto-prostaglandin F_1α (6KF_1α)_and of thromboxane B₂ (TXB₂) from cells were investigated using mouse peritoneal exudate cells (PECs) and non-cultured peritoneal macrophages. They were prepared by adhesion to glass dishes and incubated for 1 hr at 37°C in 5% Co₂ in air. Both the percentage of spreading macrophages and the release of 6KF_1α and TXb₂ increased in proportion to the incubation time. 6KF_1α and TXB₂ were released from the macrophages, not from the non-adherent cells. When PECs were incubated in silicon-coated glass dishes, the spreading of macrophages was hardly detected and lower amounts of 6KF_1α and TXB₂ were released from these cells compared with cells incubated in non-treated glass dishes. These findings suggest that adhesion with the correlated spreading of macrophages on glass dishes serve as a considerable physical factor for the release of 6KF_1α and TXB₂.     

Spreading of human fibroblasts in serum-free medium: Inhibition by dithiothreitol and the effect of cold insoluble globulin (plasma fibronectin)

We have tested the effect of dithiothreitol (DTT) treatment on the initial spreading of human fibroblasts in serum-free medium in tissue culture dishes. Cell spreading was inhibited following treatment of these cells with 10 mM DTT. Inhibition occurred when the cells were treated at 37°C but not at 4° and was reversible metabolically but not by the addition of sulfhydryl oxidizing reagents. The inhibition was overcome when DTT-treated human fibroblasts were plated on cold insoluble globulin (plasma fibronectin)—coated dishes. Under these conditions spreading appeared to be completely normal, including the formation of focal adhesions. Analysis of the fibronectin concentrations in the human fibroblasts following DTT treatment indicated that there was little decrease in the absolute level of activity as determined in a biological assay for BHK cell spreading on culture dishes. Analysis of the fibronectin distribution on the DTT-treated human fibroblasts by indirect immunofluorescence using a specific anti-CIG antiserum revealed that fibronectin was no longer deposited onto the culture dish surfaces. Even when the DTT-treated human fibroblasts spread in the presence of fetal calf serum, the cell fibronectin remained for the most part in a perinuclear location. These results indicate that DTT treatment of human fibroblasts prevents the normal translocation of fibronectin from a perinuclear location to the surface of the culture dish. This study further supports our hypothesis that the initial spreading in serum-free medium of fibroblasts from cell strains depends upon secretion of fibronectin onto the culture dish surface.     

The effects of epidermal growth factor on cell proliferation and prolactin production by GH3 rat pituitary cells

The effects of epidermal growth factor (EGF) were studied in rat pituitary tumor cells, GH₃, grown in serum-supplemented and serum-free chemically defined media. EGF (1 nM) increased the cell number to 132% of the control cultured in the defined medium during a 6-day incubation period, while it decreased the cell number to 60% of the control in the serum-supplemented medium. EGF altered the morphology of the cells grown in the defined medium more markedly to an elongated conformation than that of cells grown in the serum-supplemented medium. EGF also stimulated prolactin (PRL) production by culture in the presence or absence of serum. The effects of the cell density of GH₃ on the action of EGF were shown to appear in two ways. The mitogenic influence of EGF was more effective on, and more responsive to, high-density cells, whereas the stimulatory action on PRL production was less effective on high-density cells. However, the inhibitory effects on cellular growth appeared independently of cell densities. The results obtained with ¹²⁵I-EGF binding experiments indicated that the number of binding sites, affinity, and internalization of EGF receptors were similar in either serum-supplemented or serum-free culture. At low cell density, the number of available ¹²⁵I-EGF binding sites per cell was larger than at high cell density. These results suggested that there was no apparent correlation between EGF binding and its differing effects on the growth of GH₃ cultured in the serum-supplemented and the defined medium.