Main

Twenty-one years ago Bonnet et al.(1) studied adipose tissue cellularity in untreated hypopituitary secretion in children and its changes after substitution with hGH. Initially, the patients' fat cells were larger in size but their number was subnormal when compared with healthy children of the same chronologic age or the same bone age. After initiation of hGH treatment, size and number of fat cells were shifted toward normal. From their findings Bonnet et al. concluded that GH contributes to the regulation of fat cell volume as well as to the formation of new fat cells(1). Later studies showed that, in adipose tissue samples and in isolated adipocytes from various animal sources, GH is involved in the control of lipogenesis and lipolysis and thereby influences fat cell volume(2). However, in preadipocyte cell lines of mouse origin, GH was found to behave as an adipogenic factor stimulating the differentiation of these cells into mature adipocytes. The results of these and many related studies have been recently summarized(3, 4). Apart from these findings in animal models and cell lines, the effects of GH on the differentiation of human adipocyte precursor cells has not been investigated so far. Despite the increasing number of children and adults treated with hGH, the mechanisms of hGH action in man are also still unknown.

It has been recently shown that most fibroblast-like cells present in the stromal fraction of human adipose tissue are able to differentiate into mature adipocytes(5). The applied culture system may also represent a useful tool for getting an insight into the role of hGH in the development of human adipocytes. Using cultured human adipocyte precursor cells obtained from the adipose tissue of children and adults, we investigated the effect of hGH on fat cell formation. Because factors involved in adipose conversion may exert their action by different mechanisms(6), we also studied the effect of hGH on the proliferation of human adipocyte precursor cells as well as on glucose and lipid metabolism of newly developed fat cells.

METHODS

Subjects. Adipose tissue samples (100-200 mg) were obtained from the s.c. adipose tissue of 10 children (2 mo to 12 y old) undergoing herniotomies. In addition, s.c. adipose tissue samples (200-700 g) were obtained from three women (20-45 y old) undergoing surgical reduction of the abdominal s.c. fat depots and from seven women (18-42 y old) undergoing surgical mammary reduction. All subjects were healthy; the women were of normal weight or moderately obese, with a body mass index ranging between 24.0 and 30.9 kg/m2. Informed consent was obtained from the patients and from the parents of the children. The study was in accordance with the ethical standards of the responsible local committee on human experimentation.

Materials. Culture media (DMEM and Ham's F12 medium), FCS, HBSS, penicillin-streptomycin, and trypsin-EDTA were obtained from Life Technologies, Inc. (Gaithersburg, MD). Tissue culture plastic ware was from both Flow Laboratories (Irvine, Scotland) and Life Technologies (Berlin, Germany). Type I collagenase was purchased from Worthington (Freehold, NJ) and human transferrin, biotin, pantothenate, and BSA from Sigma Chemical Co. (St. Louis, MO). 3-D-[3H]Glucose (sp act, 15 Ci/mmol), 2-deoxy-D-[3H]glucose (sp act, 14.4 Ci/mmol), and [3H]thymidine(sp act, 110 Ci/mmol) were purchased from Amersham (Buckinghamshire, England). A mouse MAb raised against human IGF-I was obtained from Serotec (Kidlington, England). All of the other reagents were of the highest chemical grade and purchased from Sigma Chemical Co.. Recombinant hGH, insulin, and IGF-I were kindly provided by Novo Nordisk (Gentofte, Denmark). hGH was labeled with[125I]iodine to a sp act of 60-80 mCi/mg and purified as previously reported(7).

Cell preparation and cell culture. Stromal-vascular cells from human adipose tissue were isolated and cultured as described earlier, with minor modifications(5). The isolated cells were inoculated into 12-well plates (4.5 cm2/well) at a density of 105 cells/cm2 and maintained in the serum-enriched medium for 12-18 h in a humified atmosphere of 95% air and 5% CO2 at 37 °C. Thereafter they were washed three times with prewarmed DMEM and finally cultured in a serum-free medium consisting of DMEM/Ham's F12 medium (1:1, vol/vol) supplemented with 15 mM NaHCO3, 15 mM HEPES, 33 mM biotin, 17 mM pantothenate, penicillin, and streptomycin (basal medium). Culture media were changed every 5 d.

Cell proliferation experiments. To determine changes of cell number during the whole culture time in the presence and absence of hGH, cells were cultured in 10 μg/mL transferrin, 1 mM insulin, 200 pM triiodothyronine, and 10 nM cortisol (TITC) medium, detached with HBSS containing 0.05% trypsin and 0.02% EDTA on d 1 and 15, and counted in a hemocytometer. Alternatively, cell number was determined on the culture dishes using a net micrometer (Zeiss, Oberkochen, FRG) at a 100-fold magnification. In the latter case, five different areas each representing 10 mm2 were counted, and the mean cell number per dish was calculated. The two methods yielded comparable results (data not shown). In addition to the direct cell counts, [3H]thymidine incorporation was used as a measure of DNA synthesis as described earlier(8) with minor modifications. The inoculation density was 2 × 104 cells/cm2, and the incubation time with [3H]thymidine (0.2 mCi/well) was 48 h. In some experiments, a mouse monoclonal anti-human IGF-I antibody was present at a 1:200 dilution.

Induction of adipose differentiation. To induce adipose differentiation, cells were cultured in basal medium supplemented with 10μg/mL transferrin, 1 mM insulin, 200 pM triiodothyronine, and 10 nM cortisol (TITC medium).

Morphology. The differentiation of adipocyte precursor cells into mature adipocytes was assessed by conventional microscopy at a 100-fold magnification. Cells were considered as differentiated when their cytoplasm was completely filled with lipid droplets. Adipose conversion was additionally assessed by Oil Red O staining. In some experiments, the proportion of differentiated cells in the monolayers was estimated by direct counting using a net micrometer (Zeiss, Oberkochen, Germany).

Determination of triglyceride content and GPDH activity. The cellular triglyceride content and GPDH (EC 1.1.1.8) activity were determined as described earlier(8). The activity of GPDH was used as a specific marker for differentiation. Its increase during adipose differentiation has been shown to be directly related to the increase in cellular GPDH mRNA(9). Enzyme activity was expressed in milliunits per mg of total cellular protein, 1 mU being equal to the oxidation of 1 nmol of NADH/min.

Determination of glucose uptake. Glucose uptake was determined as described earlier(8) with minor modifications. Adipocyte precursor cells were maintained from d 1 in ITC medium, and hGH was added or not at concentrations as indicated to adipocyte precursor after differentiation into adipocytes on d 13. On day d 14, cells were washed twice with glucose-free KRH buffer, pH 7.4, containing 1% BSA, and maintained thereafter in the same buffer for 3 h at 37 °C before the glucose uptake experiments. One hour before measuring glucose uptake, 1 μM insulin was added. 2-Deoxy-D-glucose, both 3H-labeled (0.3 mCi/well) and unlabeled(final concentration 0.1 mM), was added for 5 min. The incubation was terminated by washing the cells twice with ice-cold HBSS containing 200 mM phloretin. Noncarrier-mediated uptake was assessed in the presence of 40 mM cytochalasin B and used for correction of the raw data.

Incorporation of [3H]glucose into cellular lipids. Cultures were prepared as described above. On d 12, cells were incubated with 0.1 μCi of 3-D-[3H]glucose in ITC medium for 48 h at 37 °C. Then, the medium was removed, and cells were washed, dissolved in 1 mL 0.5 N NaOH, and transferred to 25-mL plastic vials. Cellular lipids were extracted by adding 9 mL of a scintillation mixture (1 L of toluene containing 0.30 g of 2,2′-p-phenylen-bis-5-phenyloxazole and 5.00 g of 2,5-diphenyloxazole) and counted. All values were corrected for the blanks.

Determination of free glycerol. Adipocyte precursor cells were maintained in ITC medium. The medium was changed on d 12, and hGH was added for 48 h to the cultures of newly differentiated adipocytes at the concentrations indicated. The free glycerol concentration of the culture medium was determined on d 14 using the bioluminescence method described by Kather and Wieland(10).

GH binding studies. 125I-Labeled hGH binding to the cells was measured on d 14. Initially, cell monolayers were washed twice with 2 mL of HBSS. Then, 150 pM 125I-labeled hGH was added in the absence or presence of rising concentrations of unlabeled hGH, in a total volume of 0.5 mL in KRH buffer, pH 7.4, containing 0.1 mM glucose and 1% BSA. The incubation was performed at room temperature (22 °C) for 4.5 h and stopped by aspirating the supernatant and rapidly washing the monolayers with 2 mL of ice-cold HBSS. After solubilization of the cells in 0.5 N NaOH, the radioactivity of the samples was counted in a γ-counter.

Other assays. The cellular protein content was measured according to Bradford et al.(11). The IGF-I content in the culture medium was kindly measured by Dr. W. Blum, Giessen, Germany, with a specific RIA(12).

Studies in adipose tissue samples from children. Because only a small number of stromal cells could be obtained from the small samples of adipose tissue from the children, we were able to investigate only the proliferation and the differentiation of these cells by microscopical counting of the cell number and determination of the proportion of differentiated cells.

Statistical analysis. Results were expressed as means ± SD of at least three experiments performed in duplicate. Comparison between cell preparations maintained under various culture conditions was performed using a t test.

RESULTS

Effects of hGH on the proliferation of adipocyte precursor cells. As shown in Figure 1A, addition of hGH to cultures of human adipocyte precursor cells resulted in a dose-dependent increase in cell number on d 15. There was no statistically significant difference in the increase in cell number between the cultures obtained from adults and from children. The mitogenic effect of hGH was also investigated by measuring the incubation of [3H]thymidine into the cellular DNA of adipocyte precursor cells over 48 h (Fig. 1B). hGH stimulated [3H]thymidine incorporation dose dependently. This stimulation was completely inhibited by addition of a mouse anti-IGF-I MAb. The IGF-I concentration in the culture medium was significantly higher in cells exposed to 5 nM hGH for 72 h than in unexposed cells (Fig. 2). Similar to hGH, addition of IGF-I also dose dependently stimulated [3H]thymidine incorporation (data not shown). Thereby, these results indicate that the mitogenic activity of hGH was mediated by a stimulation of IGF-I synthesis.

Figure 1
figure 1

Stimulation of proliferation and DNA synthesis of human adipocyte precursor cells from children and adults by hGH. (A) Effect of hGH on total cell number on d 15 in cultures of human adipocyte precursor cells obtained from adults and children. The increase in cell number is related to cell number on d 1. Data represent means ± SD of five experiments in duplicate. The asterisk (*) indicates a statistically significant difference compared with control cultures without hGH (p< 0.05). B, effect of hGH on the incorporation of[3H]thymidine into cellular DNA. Cells were cultured in a serum-free medium supplemented with 100 nM insulin, 200 pM triiodothyronine, 100 nM cortisol, and 10 μg/mL transferrin, and in the absence or presence of hGH at the indicated concentrations. In addition, in some cultures a MAb against human IGF-I was added. The incorporation of [3H]thymidine into cellular DNA for 48 h was measured as described in “Methods.” Data are given as percentage of control and represent means ± SD of three experiments in triplicate. The asterisk (*) indicates a statistically significant difference compared with control cultures without hGH (p< 0.05).

Full size image
Figure 2
figure 2

Effect of hGH on IGF-I secretion by human adipocyte precursor cells in primary culture. Cells were incubated on d 1 for 72 h with or without 5 nM hGH. IGF-I concentration was measured in the culture medium using a specific RIA (15). Data represent means ± SD of three experiments in duplicate. The asterisk (*) indicates a statistically significant difference compared with control cultures without hGH (p< 0.01).

Full size image

Differentiation of human adipocyte precursor cells into mature adipocytes. Initially, the cultured stromal cells from human adipose tissue samples had a fibroblast-like appearance, and no intracellular lipids were visible by conventional microscopy (Fig. 3A). During the following days, cells acquired a round shape and accumulated lipid droplets. On d 14, most of the cells were completely filled with multiple lipid droplets (Fig. 3B). In the serum-free medium supplemented with ITC medium, up to 75% of the stromal cells from the human adipose tissue differentiated into mature adipocytes. There was no statistically significant difference in the percentage of differentiated cells between the cultures obtained from children and those obtained from adults (Fig. 4A). During differentiation the triglyceride content of the cells rose from undetectable levels on d 1 to 1.5 ± 0.5 mg/mg of cellular protein in the newly differentiated adipocytes on d 14. Similarly, GPDH activity increased from undetectable values on d 1 to more than 1200 mU/mg of protein on d 14 (data not shown).

Figure 3
figure 3

Photomicrographs of primary cultures of human adipocyte precursor cells and newly differentiated adipocytes (magnification 100-fold).(A) Undifferentiated cells on d 2. (B) Differentiated adipocytes on d 14. Cells were cultured in a serum-free medium supplemented with 100 nM insulin, 200 pM triiodothyronine, 100 nM cortisol, and 10 μg/mL transferrin. (C) Cell culture on d 14 when cells were cultured in the same medium as in B in the presence of 5 nM hGH.

Full size image
Figure 4
figure 4

Inhibition of differentiation of human adipocyte precursor cells from children and adults by hGH. (A) Effects of hGH on the percentage of newly formed adipocytes in cultures prepared from adipose tissue samples of adults and children. Data represent means ± SD of five experiments in duplicate. The asterisk (*) indicates a statistically significant difference compared with control cultures without hGH (p < 0.01). (B) Activity of GPDH in cultures of human adipocyte precursor cells on d 14. Cells were cultured in a serum-free medium supplemented with 100 nM insulin, 200 pM triiodothyronine, 100 nM cortisol, 10 μg/mL transferrin, and hGH at concentrations as indicated. Data represent means ± SD of three experiments in duplicate.*p < 0.05; **p < 0.01.

Full size image

Effects of hGH on the differentiation of human adipocyte precursor cells. The presence of hGH markedly diminished the differentiation of adipocyte precursor cells. This hGH-induced decrease in the proportion of differentiated cells on d 14 (Fig. 3C) was similar in cultures from the children and from the adults (Fig. 4A). The inhibitory effect of hGH on GPDH activity was dose-dependent and was statistically significant at concentrations of 0.5 nM hGH and higher(p < 0.05) (Fig. 4B). The increase in triglyceride content was also reduced by hGH (data not shown).

Effects of hGH on glucose uptake and lipogenesis. In the culture system used, glucose uptake is the rate-limiting step for lipid synthesis and accumulation, because the serum-free medium was devoid of exogenous lipids. Insulin-stimulated uptake of 2-deoxy-D-[3H]glucose markedly increased during adipose differentiation from 0.04 pmol/106 cells on d 4 to 0.32 pmol/106 cells on d 14. We performed experiments on glucose transport and incorporation into extractable lipids using cultures of human adipocytes differentiated in TITC medium. Addition of hGH on d 13 to newly differentiated adipocytes for 24 h reduced cellular glucose uptake and incorporation of glucose into lipids in a dose-dependent manner(Table 1).

Table 1 Effects of hGH on 2-deoxy-D-glucose uptake and lipogenesis in in vitro differentiated human adipocytes
Full size table

Effects of hGH on lipolysis. To study the effect of hGH on lipolysis the glycerol content of the culture medium of newly differentiated adipocytes was measured. As shown in Figure 5, cells cultured in TITC medium without addition of hGH released only small amounts of glycerol into the medium (5.2 ± 1.9 μmol/L/24 h). Addition of hGH on d 13 for 24 h significantly stimulated lipolysis at concentrations of 0.5 and 5 nM (25 ± 8 and 82 ± 13 μmol/L/24 h, each p < 0.01 compared with TITC medium) (Fig. 5). The above mentioned effects of hGH on adipose differentiation and metabolism could not be inhibited by a mouse anti-IGF-I MAb.

Figure 5
figure 5

Glycerol release in in vitro differentiated human adipocytes in the absence or presence of hGH. Cells were incubated or not with the indicated concentrations of hGH for 48 h. The glycerol concentration of the culture medium was measured using a bioluminescence assay. Data represent means ± SD of three experiments in duplicate. The asterisk (*) indicates a statistically significant difference compared with control cultures without hGH (p < 0.01).

Full size image

Binding of125I-labeled hGH to newly differentiated human adipocytes. A direct action of hGH on human fat cells requires the presence of specific GH receptors. To demonstrate their existence, binding experiments were performed using 125I-labeled hGH. The binding of 125I-labeled hGH to cultured adipocytes on d 14 was both temperature- and time-dependent. Maximum binding was obtained after an incubation time of 4.5 h at 22 °C. Competition experiments were performed by incubating cells with 150 pM 125I-labeled hGH and increasing concentrations (0-500 nM) of unlabeled hGH (Fig. 6). Scatchard analysis of the data yielded a linear plot (Fig. 6, inset) with an apparent dissociation constant(Kdapp) of 1.08 nM. The estimated number of GH binding sites was 7000/cell.

Figure 6
figure 6

125I-Labeled hGH binding in differentiated human adipocyte precursor cells on d 14. 125I-Labeled hGH binding was determined in KRH buffer for 3.5 h at room temperature (22 °C) at increasing concentrations of unlabeled hGH. 125I-Labeled hGH binding is expressed as ratio bound/total hormone concentration and represented as a function of unlabeled hGH concentration. (Inset) The same data are represented in Scatchard coordinates (bound/free vs bound).

Full size image

DISCUSSION

In the present study we found a stimulatory effect of hGH on the proliferation of human adipocyte precursor cells which was mediated by an increase in cellular IGF-I production. In addition, we were able to show that hGH decreases the differentiation of human adipocyte precursor into mature adipocytes to a similar extent as recently demonstrated in rat adipocyte precursor cells(8). In newly developed adipocytes, hGH was able to decrease glucose uptake and lipogenesis and to stimulate lipolysis. These results may help to explain the changes in adipose tissue cellularity of children with hGH deficiency before and after treatment with hGH(1, 13). They suggest that hGH is able to enlarge the pool of adipocyte precursor cells in human adipose tissue and, in addition, to behave as an antiadipogenic hormone which not only interferes with adipose differentiation but also reduces the lipid content of mature fat cells due to various metabolic effects.

The observation that hGH inhibits adipose differentiation is in agreement with earlier reports on the effect of GH on rat and pig preadipocytes(8, 14), but is in contrast to results in clonal cell lines where hGH appears to be an essential factor for the induction of the differentiation process(4, 6). Based on results obtained in 3T3-F442A cells, Green et al.(15) had originally proposed a “dual effector model” of GH action. In a first step, GH triggers precursor cells to enter the differentiation program, and in a second step IGF-I acts as a mitogen under the control of GH to stimulate the clonal growth of these committed cells. This concept seems to hold true for clonal cell lines but not for the differentiation of human adipocyte precursor cells. However, the discrepancy between our findings and the results of studies in clonal cell lines may be explained by differences of the developmental stage of the cells. Human adipocyte precursor cells may have already undergone critical cell divisions in vivo and may be in a later stage of the differentiation process(16). These cells appear to be already committed to become adipocytes in contrast to established cell lines(4, 6).

The mechanisms by which hGH reduces terminal differentiation of adipocyte precursor cells are not fully understood. However, there is growing evidence from in vitro studies(14, 17, 18) that hGH may inhibit differentiation due to a direct antiadipogenic action resulting in an alteration of critical metabolic pathways. Recently it has been shown that hGH inhibits cortisol-stimulated LPL activity in samples of human adipose tissue in vitro(17). This inhibition, however, occurred without affecting LPL mRNA levels. A direct lipolytic action of hGH in human adipose tissue, as seen in this study, has been recently demonstrated in freshly isolated human fat cells(18).

Interestingly, there was no significant difference in the differentiation rate and the response to hGH between the cultures obtained from the adipose tissue of children and those obtained from adults. There is convincing evidence that the formation of new fat cells can occur throughout life with decreasing capacity during aging(5, 19). The results presented in this study are to some degree in contrast to an earlier report, where a difference in the capacity of fat cell formation in children and adults was observed and no effect of hGH was detectable(20). However, these results were obtained in a serum-containing culture system, and the use of serum may have concealed possible GH effects. In addition, the adipogenic factors used in our system may cause a maximum stimulation of adipose differentiation which is not identical with the complex physiologic regulation of fat cell formation in vivo.

The finding that human adipocyte precursor cells produce IGF-I and that this production is stimulated by hGH confirms the results of recent studies, indicating that the adipose tissue organ produces large amounts of IGF-I and IGF-binding proteins. In rat white adipose tissue, Peter et al.(21) found considerably higher levels of IGF-I mRNA and IGF-I protein than in most other tissues. The production rate in adipose tissue was in the same range as that in liver(22). An increased IGF-I synthesis was also found in cultured stromal-vascular cells from rabbit and pig adipose tissue(22–24), which was under the control of GH(23, 24) and glucocorticoids(22). The results of these studies and our results suggest that locally produced IGF-I plays an important role in the regulation of adipose tissue growth.

The clinical observation that adipose tissue is a target tissue for the action of hGH as well as the biologic effects of hGH on human adipocyte precursor cells and adipocytes observed in our study depend on the presence of specific GH receptors. From our GH-binding data we estimated a number of 7000 GH-receptors per human adipocyte and a Kd of 1.08 nM. Both results are agreement with earlier observations obtained in freshly isolated human adipocytes(25, 26) and indicate that fat cells in their physiologic environment are able to respond to hGH.

The cell culture model used in our study represents a valuable tool to further investigate the diffentiation process of human adipocytes. The serum-free culture system has the advantage that effects of specific hormones can be studied directly without interference of unknown serum factors. Furthermore, the metabolic effects on the newly developed adipocytes can be studied over longer periods of time than has been possible with freshly isolated adipocytes due to their high fragility. Although it is evident that an in vitro system cannot fully represent the physiologic situation, the results of our study suggest that hGH is able to enlarge the pool of precursor cells via stimulation of IGF-I synthesis and, at the same time, to reduce the lipid content of developing and mature fat cells.