The Mitogenic and Antiapoptotic Actions of Ghrelin in 3T3-L1 Adipocytes

Abstract

Ghrelin, a stomach-derived hormone, induces adiposity when administered to rodents. Because ghrelin receptor is abundantly expressed in adipose tissue, we investigated the role of ghrelin in adipocyte biology. We observed ghrelin receptor expression in 3T3-L1 preadipocytes and adipocytes. Treatment of preadipocytes with ghrelin induced cellular proliferation and differentiation to mature adipocytes, as well as basal and insulin-stimulated glucose transport, but it inhibited adipocyte apoptosis induced by serum deprivation. Exposure of 3T3-L1 cells to ghrelin caused a rapid activation of MAPKs, especially ERK1/2. Chemical inhibition of MAPK blocked the mitogenic and antiapoptotic effects of ghrelin. Ghrelin also stimulated the insulin receptor substrate-associated phosphatidylinositol 3-kinase/Akt pathway in 3T3-L1 preadipocytes and adipocytes, whereas inhibition of this pathway blocked the effects of ghrelin on cell proliferation, antiapoptosis and glucose uptake. These findings suggest that the direct effects of ghrelin on proliferation, differentiation, and apoptosis in adipocytes may play a role in regulating fat cell number. These effects may be mediated via activation of the MAPK and phosphatidylinositol 3-kinase/Akt pathways.

GHRELIN IS A 28-amino acid peptide produced primarily in the stomach and secreted into the systemic circulation (1). It was originally identified as an endogenous ligand of GH secretagog receptor (GHS-R) and shown to have a potent stimulatory effect on GH release in humans and rodents (1, 2). In addition to being an endogenous GH secretagog, ghrelin appears to play an important role in the regulation of body weight. Administration of ghrelin stimulates food intake and adiposity (3, 4). The ghrelin receptor, GHS-R, is highly expressed in the central nervous system (5), and central administration of a small amount of ghrelin has been shown to increase food intake (4), suggesting that ghrelin increases adipose tissue primarily through a central acting mechanism.

Ghrelin has also been shown to have multiple regulatory functions in peripheral tissues (6–8). Expression of GHS-R mRNA, as well as binding sites for the synthetic GHS-R ligand [¹²⁵I]-Try-Ala-Hexarelin, have been observed in peripheral organs, including adipose tissue (9, 10), suggesting that ghrelin may act directly on adipocytes to increase adipose tissue mass. We therefore investigated whether ghrelin plays a direct role in adipocyte biology. We found that ghrelin stimulates the proliferation and differentiation of 3T3-L1 adipocytes, as well as suppressing apoptosis in these cells. In addition, we found that ghrelin activates the MAPK and phosphatidylinositol 3 kinase (PI3K)-Akt pathways, both of which mediate the mitogenic and antiapoptotic effects of ghrelin.

RESULTS

GHS-R Expression in 3T3-L1 Adipocytes

When we assayed expression of the ghrelin receptor, GHS-R1a, in 3T3-L1 preadipocytes and differentiated adipocytes, we found that GHS-R1a mRNA and protein were present in 3T3-L1 cells, as well as in the rat hypothalamus and pituitary gland (Fig. 1, A and B). GHS-R1a abundance in 3T3-L1 cells was dependent on the stage of adipocyte differentiation, being higher in terminally differentiated adipocytes than in preadipocytes or early differentiating adipocytes (Fig. 1A). In contrast, preproghrelin mRNA expression was found in rat stomach but not in 3T3-L1 peadipocytes and differentiated adipocytes (Fig. 1C).

Fig. 1.

Expression of GHSR-1a mRNA (A) and Protein (B) in 3T3-L1 Cells Total RNA and protein were extracted from rat pituitary gland and hypothalamus and from 3T3-L1 cells at different stages of differentiation. The GHS-R1a mRNA expression was analyzed by RT-PCR and compared with that of GAPDH. GHSR-1a protein expression was determined by immunoblot assay after immunoprecipitation using GHS-R1a antibody. C, Expression of preproghrelin mRNA in rat stomach and 3T3-L1 cells.

Ghrelin Stimulates Adipocyte Proliferation

Adipose tissue mass is regulated by the process of adipocyte acquisition (proliferation, differentiation) and loss (apoptosis, dedifferentiation) (11). To determine the effect of ghrelin on adipocyte proliferation, we cultured 3T3-L1 preadipocytes for 24 h in serum-deprived medium containing IGF-I or ghrelin and determined the number of cells using a cell proliferation assay. IGF-I, a well-known adipocyte mitogen (12), increased adipocyte cell number (Fig. 2A). Ghrelin at concentrations of 10⁻⁷ to 10⁻¹³m also significantly increased adipocyte cell numbers, with the most pronounced effect observed at 10⁻¹¹m (Fig. 2A). In cell cycle analysis, ghrelin (10⁻⁹ to 10⁻¹⁴m) also stimulated cell cycle progression from G1 to S phase (Table 1).

Fig. 2.

Ghrelin Stimulates Adipocyte Mitogenesis A, 3T3-L1 preadipocytes were starved overnight and treated with IGF-I (10 nm) or ghrelin (10⁻⁷m to 10⁻¹⁵m) for 24 h at 37 C. Cell proliferation was determined by MTS assay. Assays were performed in triplicate for each sample and repeated three times. *, P < 0.05; **, P < 0.01 vs. control. B, Immunoblot analysis of c-Myc and phospho-Rb protein (Ser780). 3T3-L1 preadipocytes were maintained for 12 h in a serum-free medium containing 50 nm insulin or 10⁻⁷m to 10⁻¹⁷ M ghrelin before immunoblot analysis. C and D, The mRNA expression levels of c-myc and thymidine kinase. Cells were maintained in serum free medium containing 50 nm insulin or 10⁻¹¹m ghrelin for 18 h (c-myc) or 12 h (thymidine kinase), and levels of specific mRNA were assayed by RT-PCR and compared with levels of GAPDH mRNA. *, P < 0.05 vs. control.

Table 1.

Effect of Ghrelin on Cell Cycle Progression in 3T3-L1 Preadipocytes

	G0–G1 (%)	S (%)	G2-M (%)
Control	86.6 ± 0.3	8.3 ± 0.3	5.0 ± 0.3
Insulin 5× 10−8m	82.6 ± 0.3a	10.3 ± 0.5a	7.1 ± 0.5a
Ghrelin 10−7m	86.6 ± 0.2	7.9 ± 0.5	5.5 ± 0.4
Ghrelin 10−9m	85.8 ± 0.2b	9.8 ± 0.4b	4.4 ± 0.3
Ghrelin 10−11m	85.3 ± 0.3b	8.9 ± 0.3	5.8 ± 0.2b
Ghrelin 10−13m	85.6 ± 0.2b	9.2 ± 0.2b	5.2 ± 0.3
Ghrelin 10−14m	85.5 ± 0.1b	9.5 ± 0.3b	5.0 ± 0.1
Ghrelin 10−15m	87.3 ± 0.3	7.6 ± 0.4	5.1 ± 0.4

3T3-L1 cells were grown in serum-free medium with or without insulin or various concentrations of ghrelin for 24 h before cell cycle analysis.

^a

P < 0.01 vs. control.

^b

P < 0.05 vs. control.

The c-Myc protein is a potent cell cycle activator that induces DNA synthesis (13). Cell cycle activation by c-Myc protein in serum-deprived 3T3-L1 cells involves the inactivation of retinoblastoma (Rb) protein through phosphorylation and subsequent activation of E2F-mediated gene expression, such as that encoding thymidine kinase (14). Treatment of 3T3-L1 preadipocytes with ghrelin increased the levels of c-myc mRNA and c-Myc protein (Fig. 2, B and C), as well as stimulating the phosphorylation of Rb protein and the expression of thymidine kinase mRNA (Fig. 2, B and D).

Ghrelin Stimulates Adipocyte Proliferation

Insulin and IGF-I have been shown to stimulate both the proliferation and differentiation of adipocytes (15, 16). To determine whether ghrelin affects adipocyte differentiation, we maintained 3T3-L1 cells in DMEM containing 10% fetal bovine serum (FBS) with or without insulin and/or ghrelin for 6–8 d after induction of differentiation. Oil red O staining revealed that the formation of lipid droplets, a marker for adipocyte differentiation, was significantly increased by insulin (Fig. 3A). Treatment of cells with ghrelin (10⁻¹¹m or 10⁻¹³m) induced the accumulation of more lipid droplets than observed in FBS-treated cells, but fewer than that observed in full dose insulin-treated cells (Fig. 3A). Cotreatment of ghrelin (10⁻¹¹m or 10⁻¹³m) and half-dose (0.5 μg/ml) insulin further stimulated adipocyte differentiation compared with treatment of ghrelin alone (Fig. 3A). We also found that insulin and ghrelin increased the level of mRNA expression of the adipogenic genes, peroxisome proliferator-activated receptor (PPAR)-γ, adipocyte determination and differentiation-dependent factor (ADD)1, and adipose protein 2/fatty acid-binding protein (aP2) (Fig. 3B).

Fig. 3.

Effect of Ghrelin on Adipocyte Differentiation A, Lipid accumulation during adipocyte differentiation. Following induction of differentiation, cells were subsequently maintained in DMEM containing 10% FBS alone (I), 10% FBS plus insulin (1 μg/ml) (II) 10% FBS plus 10⁻¹¹m ghrelin (III), 10% FBS plus 10⁻¹¹m ghrelin and 0.5 μg/ml insulin (IV), 10% FBS plus 10⁻¹³m ghrelin (V), or 10% FBS plus 10⁻¹³m ghrelin and 0.5 μg/ml insulin (VI) for 6 d. Intracellular fat accumulation was determined by oil red O staining. B, Expression of PPAR-γ, ADD1, and aP2 mRNA during adipocyte proliferation. 3T3-L1 cells were differentiated in DMEM containing 10% FBS alone (control), 10% FBS plus 1 μg/ml insulin (insulin), or 10% FBS plus 10⁻¹³m ghrelin (ghrelin). Cells were harvested on the postdifferentiation second and sixth days for determination of the mRNA levels by Northern blot analysis.

Ghrelin Prevents Adipocyte Apoptosis Induced by Serum Starvation

After growth factor deprivation, human and murine adipocytes have been observed to undergo apoptosis, which may contribute to a decrease in adipocyte number (11). We therefore assayed the effect of ghrelin on adipocyte apoptosis. Apoptosis was determined by measurement of fragmented DNA fraction and terminal deoxynucleotidyl transferase-mediated dUTP biotin nick-end labeling (TUNEL) analysis. Serum withdrawal for 48 h increased the fraction of apoptotic DNA in 3T3-L1 preadipocytes (Fig. 4A). Consistent with previous studies (17, 18), 10 nm IGF-I reduced apoptosis induced by serum starvation in these cells, as did ghrelin, at concentrations of 10⁻⁷ to 10⁻¹³m (Fig. 4A). Similarly, serum starvation increased TUNEL-positive cells, which was attenuated by treatment of ghrelin (TUNEL-positive cells: control, 4 ± 1%; serum deprivation, 56 ± 5%; 10⁻¹¹m ghrelin, 18 ± 3%; 10⁻¹³m ghrelin, 15 ± 2% of PI stained cells) (Fig. 4B).

Fig. 4.

Effect of Ghrelin on Serum Starvation-Induced Apoptosis in 3T3-L1 Adipocytes A, 3T3-L1 preadipocytes were starved for 48 h in the presence or absence of IGF-I (10 nm) or ghrelin (10⁻⁷m to 10⁻¹⁵m). DNA fragmentation, a marker of apoptosis, was analyzed by flow cytometer. *, P < 0.05; **, P < 0.01 vs. control. B, TUNEL staining of 3T3-L1 preadipocytes. Cells were starved for 48 h in the presence or absence of 10⁻¹³m ghrelin. Fragmented DNA was labeled with FITC-conjugated dUTP and DNA labeled with PI. FITC and PI labels were observed using a confocal microscope.

Ghrelin Activates the ERK1/2 Pathway

MAPKs, especially ERK1/2, are implicated in cell cycle control and differentiation (19, 20). We investigated the ability of ghrelin to activate the ERK1/2 pathway, measured as serine residue-specific phosphorylation, in 3T3-L1 preadipocytes and adipocytes. Treatment of 3T3-L1 preadipocytes for 5 min with insulin or IGF-I increased ERK1/2 phosphorylation. Similarly, treatment with ghrelin or the peptidyl GHS-R1a ligand, GH-releasing peptide 6 (GHRP-6), activated ERK1/2 in 3T3-L1 preadipocytes (Fig. 5A). We further determined the dose response and time course of ghrelin-induced ERK1/2 activation at different stages of adipocyte differentiation. We found that ghrelin concentrations ranging from 10⁻⁷ to 10⁻¹⁵m caused a rapid and strong activation of ERK1/2 in preadipocytes and terminally differentiated adipocytes, but ghrelin-induced ERK1/2 activation was significantly reduced in early differentiating adipocytes (Fig. 5B). Time course experiments showed that, in preadipocytes and terminally differentiated adipocytes, ERK1/2 activation occurred as early as 2 min after ghrelin stimulation, peaked between 5 and 10 min, and lasted for at least 30 min (Fig. 5C). In early stages of adipocyte differentiation, however, ghrelin-induced activation of ERK1/2 persisted no longer than 5 min.

Fig. 5.

Ghrelin Activates ERK1/2 in Adipocytes A, ERK1/2 activation in 3T3-L1 preadipocytes. Cells were treated with 3 × 10^{− 8}m insulin, 10⁻⁸m IGF-I, 10⁻⁷m ghrelin, or 10⁻⁵m or 10⁻⁷m GHRP-6 for 5 min. ERK1/2 activation was analyzed by immunoblotting for phosphorylated ERK1/2 (Thr202/Tyr204) and total ERK. B, Dose responsiveness of ghrelin-induced phosphorylation of ERK1/2. Cells were treated with 10⁻⁷ to 10⁻¹⁵m ghrelin for 5 min and assayed by immunoblotting as above. C, Time course of ghrelin-induced phosphorylation of ERK1/2. Cells were treated with 10⁻⁹m ghrelin for 2, 5, 10, 20, and 30 min and assayed by immunoblotting as above. D–G, Upstream signaling pathways mediating ghrelin-induced ERK1/2 activation. Cells were preincubated with 100 ng/ml pertussis toxin (Px) for 24 h, 200 μm PD98059 for 2 h, 200 nm wortmannin for 30 min, 5 μm GF109203X for 30 min, 200 nm staurosporine (Stauro) for 10 min, or 10 μm phorbol 12-myristate 13-acetate (PMA) for 15 min and then treated with 10⁻⁹m ghrelin for 5 min. Each experiment was repeated three times. The phospho-ERK1/2 band intensity was normalized to β-actin band intensity and reported relative to the increase in control cells. *, P < 0.005 vs. untreated control; †, P < 0.05 vs. ghrelin-treated cells.

The ghrelin receptor GHS-R is a G protein-coupled receptor (GPCR) (5), and several signaling pathways have been associated with GPCR-mediated ERK1/2 activation (21–23). We found that pretreatment of 3T3-L1 cells with the Gi inhibitor, pertussis toxin, the MAPK kinase inhibitor, PD98059, the PI3K inhibitor, wortmannin, or the protein kinase C (PKC) inhibitors, staurosporine and GF109203X, significantly attenuated ERK1/2 phosphorylation in response to ghrelin, suggesting that all these pathways contribute to ghrelin-induced MAPK activation (Fig. 5, D–G).

Ghrelin Activates the PI3K/Akt Pathway

It was recently reported that, in cardiomyocytes, ghrelin and the GHS-R ligand, hexarelin, can activate the PI3K/Akt pathway, an important signaling pathway for cell survival, adipocyte differentiation, and glucose transport (7). We therefore assayed the effect of ghrelin on the PI3K/Akt pathway in 3T3-L1 cells. In preadipocytes, ghrelin induced an increase in insulin receptor substrate 1 (IRS-1)-associated PI3K activity, which occurred within 5 min and persisted for at least 30 min (Fig. 6A). Akt activation, as measured by phosphorylation, was detected beginning 30 min after ghrelin treatment and gradually increased over the next 24 h (Fig. 6B). In differentiated adipocytes, ghrelin also induced Akt phosphorylation (Fig. 6, C and D). Wortmannin, an inhibitor of PI3K, prevented ghrelin-stimulated Akt phosphorylation (Fig. 6E), suggesting that the activation of PI3K is an important upstream signal leading to Akt activation.

Fig. 6.

Ghrelin Activates PI3K/Akt Pathways A, IRS-1-associated PI3K activity in response to ghrelin. 3T3-L1 preadipocytes were grown to confluence, incubated for 2 h in serum-free DMEM, and then treated with 50 nm insulin for 30 min or 10⁻¹¹m ghrelin for the indicated time. B–D, Activation of Akt induced by treatment of ghrelin. Cells were treated with 10⁻¹¹m ghrelin for 5 min, 30 min, 1 h, and 24 h, and Akt activation was determined by Western blot analysis of phospho-Akt (Ser473). Phospho-Akt band intensity was normalized to β-actin band intensity and reported relative to the increase in control cells. Each experiment was repeated three times. *, P < 0.05; **, P < 0.01 vs. control. E, Effect of PI3K inhibitor on ghrelin-induced Akt activation. 3T3-L1 preadipocytes were treated with 200 nm wortmannin for 30 min and then with 10⁻¹¹m and 10⁻¹³m ghrelin for 1 h.

MAPK and PI3K/Akt Pathways Mediate Ghrelin-Induced Mitogenesis and Antiapoptosis

When we tested the effects of the MAPK kinase inhibitor, PD98059, and the PI3K inhibitor, wortmannin, we found that both inhibited ghrelin-induced cell proliferation, as well as the antiapoptotic effects of ghrelin (Fig. 7, A and B).

Fig. 7.

ERK1/2 and PI3K Pathways Mediate Ghrelin-Induced Mitogenesis and Antiapoptosis A, Effect of ERK1/2 and PI3K inhibition on ghrelin-stimulated cell proliferation. 3T3-L1 preadipocytes were starved overnight and treated with 10⁻¹¹m ghrelin in the presence or absence of 30 μm PD98059 or 200 nm wortmannin for 12 h at 37 C, and cell proliferation was determined by MTS assay. Assays were performed in triplicate for each sample and repeated three times. *, P < 0.05 vs. control; †, P < 0.05 vs. ghrelin-treated cells. B, Effect of ERK1/2 and PI3K inhibition on the antiapoptotic activity of ghrelin. 3T3-L1 preadipocytes were serum starved for 48 h to induce apoptosis and simultaneously treated with 10⁻¹¹m ghrelin in the presence or absence of 30 μm PD98059 or 200 nm wortmannin. *, P < 0.005 vs. 10% FBS-treated group; †, P < 0.05 vs. serum-starved group; and ‡, P < 0.05 vs. serum-starved, ghrelin-treated group.

Ghrelin Increases Glucose Transport through PI3K/Akt Activation

The PI3K/Akt pathway has been implicated in insulin-stimulated glucose transport (24, 25). Because ghrelin activated the PI3K/Akt pathway in 3T3-L1 cells, we wished to determine whether ghrelin can also stimulate glucose transport via this pathway. We therefore incubated terminally differentiated 3T3-L1 adipocytes with insulin or/and ghrelin overnight and assayed glucose transport. We found that insulin (50 nm) and ghrelin (10⁻¹¹m or 10⁻¹³m) increased glucose transport (Fig. 8A). Moreover, cotreatment of insulin (25 nm) and ghrelin (10⁻¹¹m to 10⁻¹⁵m) induced a further increase in glucose transport (Fig. 8B). The PI3K inhibitor, wortmannin, completely blocked this ghrelin-induced increase in glucose transport (Fig. 8A), suggesting that PI3K may mediate the effect of ghrelin on glucose transport in these adipocytes.

Fig. 8.

Effect of Ghrelin on Glucose Transport in 3T3-L1 Cells Differentiated 3T3-L1 cells were incubated overnight in serum-free medium containing 0.1% BSA in the presence or absence of 50 nm insulin, 10⁻⁷m to 10⁻¹⁵m ghrelin, 10⁻¹³m ghrelin plus 200 nm wortmannin (A) or 25 nm insulin plus 10⁻⁷m to 10⁻¹⁵m ghrelin (B). Uptake of [¹⁴C]deoxyglucose was assayed. Assays were performed in triplicate for each sample and repeated three times. *, P < 0.05; **, P < 0.01 vs. control; †, P < 0.05 vs. insulin-treated group.

DISCUSSION

In agreement with previous studies demonstrating the mitogenic actions of ghrelin in hepatoma and prostate tumor cells (26, 27), we have shown that ghrelin has a direct mitogenic effect on 3T3-L1 preadipocytes. The minimum effective concentration of ghrelin observed in our experiments was comparable to the circulating concentration in humans (∼100 fm) (28, 29), suggesting that, at physiological concentrations, ghrelin can act as an adipocyte mitogen. We also found that ghrelin stimulated fat accumulation as well as the mRNA expressions of adipogenic markers (PPAR-γ, ADD1, and aP2) during adipocyte differentiation. Our finding is in line with that of Choi et al. (30) that ghrelin stimulated adipocyte differentiation in primary cultured rat adipocytes. However, Zhang et al. (31) have recently reported that ectopic overexpression of ghrelin gene in 3T3-L1 cells promoted adipocyte proliferation but inhibited adipocyte differentiation.

The reason for the discrepancy between our findings and those of Zhang et al. is unclear at present. Prolonged exposure to higher dose of ghrelin may inhibit adipocyte differentiation. We found that only lower dose of ghrelin (10⁻¹¹ and 10⁻¹³m) can stimulate adipocyte differentiation. Furthermore, MAPK activation (mitogenic signaling pathway) in response to ghrelin was significantly reduced in early differentiating adipocytes. Reduction in MAPK activation induced by IGF-I has been suggested to be important for induction of adipocyte differentiation (32). In contrast, epidermal growth factor 1-induced MAPK activation is sustained in early differentiating adipocytes, and epidermal growth factor 1 has an inhibitory effect on adipocyte differentiation (32). Thus, reduced MAPK signaling in response to exogenous ghrelin may be of help for induction of adipocyte differentiation. Similarly, the effects of persistent exposure to endogenous ghrelin on adipocyte biology may differ from that of exogenous ghrelin. Further study will be warranted to clarify the argument.

We also found that ghrelin significantly reduced adipocyte apoptosis induced by serum deprivation, a finding in agreement with a recent report showing that ghrelin reduces apoptosis in cardiomyocytes and endothelial cells (7). Adipocyte apoptosis, which occurs in rodents as well as in humans with cancer-related cachexia, has been hypothesized to decrease adipocyte cell number (11). Thus, the reduction in adipocyte apoptosis due to treatment with ghrelin may contribute to an increase in adipocyte cell number coupled with an increase in adipocyte proliferation.

Several other factors have been implicated in the regulation of adipocyte number. Insulin, IGF-I, PPAR-γ ligands, retinoids, and corticosteroids have been found to increase adipocyte number, whereas TNF-α and leptin have been shown to decrease the number of these cells (11). Our findings that ghrelin receptor is expressed in adipocytes, and that physiological concentrations of ghrelin can directly stimulate adipocyte proliferation/differentiation and prevent adipocyte apoptosis, identify ghrelin as a new regulator of adipocyte number.

Previous studies have demonstrated that in cardiomyocytes, endothelial cells, and hepatocytes, ghrelin induces ERK1/2 activation (7, 26), an important mechanism mediating the mitogenic effects of IGF-I and insulin in adipocytes (33, 34). In our study, ghrelin strongly activated ERK1/2 in 3T3-L1 adipocytes, and the ERK inhibitor, PD98059, inhibited the mitogenic and antiapoptotic activities of ghrelin, thus indicating that ghrelin acts on adipocytes through ERK1/2 activation.

GHS-R1a, which is abundantly expressed in the pituitary and hypothalamus, mediates the endocrine effects of ghrelin (5, 35). We and others (30) have shown that cultures of murine 3T3-L1 adipocytes and primary rat adipocytes express functional GHS-R1a. In addition, ERK1/2 was similarly activated by the GHS-R1a agonist, GHRP-6, in the present study. Thus the ghrelin-induced activation of ERK1/2 and mitogenesis may be mediated though GHS-R1a, in contrast to cardiomyocytes, in which the activation of this pathway is mediated via a novel, yet-to-be-identified receptor, distinct from GHS-R1a (7).

Several signaling pathways have been associated with the GPCR-mediated activation of ERK1/2. Activation of the Gi/o subtype of GPCR can activate ERK1/2 through the direct interaction of the Gβγ subunits with P21^ras and the raf-1 kinase complex, which act upstream of ERK pathway (21). A synthetic GHS-R ligand has been shown to activate PKC pathways (36), which may induce MAPK activation via P21^ras-dependent and -independent mechanisms (22, 37). On the other hand, a PI3K inhibitor has been shown to inhibit ERK1/2 activation induced by insulin and platelet derived growth factor, suggesting a role for PI3K in ERK activation (23, 38). In our study, pretreatment with a Gi/o inhibitor (pertussis toxin), PKC inhibitors (staurosporin and GF109203X), or a PI3K inhibitor (wortmannin) significantly attenuated ghrelin-induced ERK1/2 phosphorylation. These findings suggest that multiple signaling pathways are involved in ghrelin signaling leading to MAPK activation.

Finally, we have shown that exposure of 3T3-L1 cells to ghrelin induced increases in IRS-1-associated PI3K activity and Akt phosphorylation. Inhibition of PI3K blocked the effects of ghrelin on adipocyte proliferation and apoptosis. Furthermore, ghrelin increased basal and insulin-stimulated glucose transport, whereas coadministration of a PI3K inhibitor blocked this effect of ghrelin on glucose transport. These findings indicate that the PI3K/Akt pathway acts to mediate the effects of ghrelin in 3T3-L1 cells.

In summary, we have demonstrated that ghrelin has direct regulatory effects on proliferation, differentiation, apoptosis, and glucose transport in 3T3-L1 adipocyte cell lines. We have also shown that treatment with ghrelin strongly activated the ERK1/2 and PI3K/Akt pathways in these cells. These findings indicate that ghrelin may exert its adipogenic action not only through central mechanisms but also through direct effects in adipocytes.

MATERIALS AND METHODS

Materials

Biologically active ghrelin was purchased from Phoenix Pharmaceuticals (Mountain View, CA). GHRP-6 was obtained from the American Peptide Company (Sunnyvale, CA). PD98059 was purchased from Calbiochem (La Jolla, CA), and insulin, IGF-I, pertussis toxin, staurosporine, GF109203X, and wortmannin were from Sigma Chemical Co. (St. Louis, MO). Antiphospho-antibodies to ERK1/2 (Thr202/Tyr204), Akt (Ser473), and Rb (Ser780) were from Cell Signaling Technology (Beverly, MA), antibodies to c-myc and β-actin were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and antibody to GHSR-1a was from Phoenix Pharmaceuticals. All tissue culture reagents were purchased from Life Technologies (Gaithersburg, MD), and all other materials were from Sigma, unless otherwise indicated.

Cell Culture

3T3-L1 preadipocytes were maintained in DMEM containing 10% calf serum, 100 U/ml penicillin, and 100 U/ml streptomycin. For routine induction of differentiation, confluent cells were treated with 25 μm dexamethasone, 0.5 mm isobutylmethylxanthine, and 5 μg/ml insulin for 60 h and maintained in DMEM containing 10% FBS and 1 μg/ml insulin for 6–8 d unless otherwise stated.

Cell Proliferation Assay

Cell proliferation was monitored with the Cell titer 96 aqueous one-solution cell proliferative assay (Promega Corp., Madison, WI). Cells (5 × 10³/well) were seeded in a 96-well culture plate, maintained for 24 h in DMEM containing 10% calf serum, incubated overnight in serum-free DMEM with 0.1% BSA to synchronize the cell cycle, and treated with various compounds at 37 C for 24 h. To each well was added 20 μl of 2-(4′,5′-dimethyl-2′-thiazolyl)-3-(4″-sulfophenyl) (MTS) solution; the MTS tetrazolium compound is reduced by the reduced nicotinamide adenine dinucleotide phosphate or reduced nicotinamide adenine dinucleotide produced by dehydrogenase enzymes in metabolically active cells into a colored, soluble formazan product. The plate was kept for 3 h in a CO₂ incubator, and the absorbance at 490 nm was recorded with a 96-well plate ELISA reader (PerkinElmer Corp., Norwalk, CT). All samples were assayed in triplicate, and each experiment was repeated at least three times.

Cell Cycle Analysis

Cells were seeded in 12-well culture plates, maintained for 24 h in DMEM containing 10% calf serum, incubated overnight in serum-free DMEM with 0.1% BSA to synchronize the cell cycle, and treated with various compounds at 37 C for 15 h. The cells were harvested, fixed in 70% ethanol for at least 2 h, washed once with PBS, and incubated with 5 μg/ml propidium iodide (PI) and 100 μg/ml RNase for 10 min at room temperature in the dark room. Cell cycle was analyzed with a FACSCalibur argon laser cytometer (Becton Dickson, Franklin Lakes, NJ). Percentage of cells in each phase of cell cycle was calculated and expressed as mean ± sem.

Apoptosis Assay

Cells were maintained in serum-free DMEM with 0.1% BSA, with or without ghrelin or IGF-I, for the indicated times. The cells were prepared with the same method of cell cycle analysis. Analysis of stained cells was performed on a flow cytometer. The percentage of apoptotic cells was determined from DNA histogram as a ratio of cells with hypodiploid DNA content (the sub-G1 peak) to the total number of cells.

TUNEL Staining

Cells were plated on polylysine-coated slides, fixed with 4% paraformaldehyde in 0.1 m PBS for 1 h at room temperature, rinsed with 0.1 m PBS, pH 7.4, and permeabilized with 1% Triton X-100 in 0.01 m citrate buffer, pH 6.0. DNA fragmentation was detected using TUNEL detection kit (Roche Clinical Laboratories, Indianapolis, IN), which specifically labeled 3′-hydroxyl termini of DNA strand breaks using fluorescein isothiocyanate (FITC)-conjugated dUTP (39). DNA was also labeled with 25 μg/ml PI DNA-binding dye for 5 min. FITC and PI labels were observed with a confocal microscope using an excitation wavelength of 488 nm, and detection was in the range of 500–530 nm for FITC and of 600–630 nm for PI. The percentage of apoptotic cells was calculated as the number of apoptotic cells per number of total cells × 100%.

Oil Red O Staining

Differentiated adipocytes were fixed with 10% formalin and stained with 0.35% oil red O, and the resultant liquid droplets were visualized by light microscopy.

Semiquantitative RT-PCR

Total RNA from rat tissue and 3T3-L1 cells was extracted using Trizol reagent (Invitrogen, San Diego, CA) and quantified by spectrophotometry (Beckman, Fullerton, CA). An aliquot of cDNA synthesized from 2 mg of total RNA was amplified using primer sets for GHS-R1a (5′-TTC TGC CTC ACT GTG CTC TAC AGT-3′ and 5′-GGA CAC CAG GTT GCA GTA CTG GCT-3′), preproghrelin (5′-ATG CTC TGG ATG GAC ATG GC-3′ and 5′-TAC TTG TCA GCT GGC GCC TC-3′), c-myc (5′-ACG ATG GAT CCT ATC ACC AGC AAC AGC AGA GCC AG-3′ and 5′-ATC GAG AAT TCG AAT CGG ACG AGG TAC AGG ATT TG-3′), or thymidine kinase (5′-CCA TGC GGA TCC AAC GAG GGC AAG ACA GTA ATT GTC-3′ and 5′-CCA TGC GAA TTC TCT CTG AGA GTC CAA CCT GGG TAG-3′) The amplification protocol consisted of 35 (GHS-R1a and ghrelin), 27 (c-myc), 40 (thymidine kinase), or 30 [glyceraldehyde-3-phosphate dehydrogenase (GAPDH)] cycles of denaturation at 94 C for 30 sec, annealing at 56 C for 30 sec, and extension at 72 C for 2 min. The resulting products, of 134 bp (GHS-R1a), 310 bp (ghrelin), 450 bp (c-myc), and 550 bp (thymidine kinase), were visualized by ethidium bromide staining of 1.2% agarose gel electrophoresis and normalized relative to amplification of the same samples with primers for GAPDH.

Northern Blot Analysis

Total RNA was prepared using Trizol reagent, and 20 μg/lane was electrophoresed on 1.2% agarose-formaldehyde gels and transferred to nylon membranes in 10× standard sodium citrate. Membranes were hybridized for 20 h at 68 C with PPAR-γ, ADD1, and aP2 cDNA (from Dr. Jae-Bum Kim, Seoul National University) probes labeled with the Rediprime labeling kit (Amersham Pharmacia, Piscataway, NJ) and [³²P]dCTP. The blots were hybridized in the presence of Quickhyb Solution (Stratagene, La Jolla, CA) at 68 C, washed twice with 1× standard sodium citrate, 0.1% sodium dodecyl sulfate at 68 C, and exposed to x-ray film.

Western Blot Analysis

Cells were lysed in a buffer containing 20 mm Tris-HCl (pH 7.4), 1 mm EDTA, 140 mm NaCl, 1% (wt/vol) Nonidet P-40, 1 mm Na₃VO₄, 1 mm phenylmethylsulfonyl fluoride, 50 mm NaF, and 10 μg/ml aprotinin. Cell lysates were centrifuged at 13,000 rpm for 15 min to remove insoluble materials, separated by 10% SDS-PAGE, and electrotransferred to a nitrocellulose membrane for 1 h. The membrane was soaked in blocking buffer (1× Tris-buffered saline, 0.1% Tween-20, 5% nonfat dry milk) for 2 h and incubated overnight at 4 C with the primary antibody. Blots were developed using a horseradish peroxidase-linked antirabbit secondary antibody and a chemiluminescent detection system (New England Biolabs, Beverly, MA). For determination of GHS-R1a protein expression, cell lysates were immunoprecipited using anti-GHS-R1a antibody and 20 μl protein G sephalose bead overnight, and the immunoprecipitates were subjected to Western blot analysis.

Measurement of PI3K Activity

PI3K activity was assayed using a protocol adapted from Krook et al. (40). 3T3-L1 preadipocytes and differentiated adipocytes were grown to confluency in 10-cm dishes, incubated for 2 h in serum-free DMEM, and treated with insulin (50 nm) for 30 min or ghrelin (10⁻¹¹m) for the indicated time. Incubations were terminated by aspirating the medium and rinsing briefly with ice-cold PBS. To each culture were added 500 μl of ice-cold freshly prepared PI3K lysis buffer A [20 mm Tris (pH 8.0), 137 mm NaCl, 2.7 mm KCl, 1 mm MgCl₂, 500 μm NaVO₃, 1% Nonidet P-40, 10% (wt/vol) glycerol, 10 μg/ml leupeptin, and 200 μm phenylmethylsulfonyl fluoride], and the lysates were immunoprecipitated with IRS-1 antibody coupled to protein A-sepharose (Amersham Pharmacia) overnight at 4 C. Reaction products were resolved by thin layer chromatography and quantified using a phosphor imager (Bio-Rad Laboratories, Hercules, CA).

Glucose Transport Assay

3T3-L1 adipocytes in 12-well culture plates were cultured overnight in serum-free medium containing 0.1% BSA in the presence or absence of insulin (25 or 50 nm) and/or ghrelin (10⁻⁷ to 10⁻¹⁵m) at the indicated concentrations before assay. To each well were added 10 μl [U-¹⁴C]2-deoxyglucose (0.5 μCi/sample) and glucose to a final concentration of 0.01 mm, and glucose uptake was determined in triplicate (41).

Data Analysis

All data are presented as mean ± sem. Comparisons between groups were by ANOVA followed by the post hoc least significance difference test. Significance was defined as P < 0.05.

Acknowledgments

We thank Dr. Jae-Bum Kim (Seoul National University) for providing 3T3-L1 cells.

This work was supported by a grant from the Korean Ministry of Health & Welfare (02-PJ1-PG3-20504-0007, 03-PJ1-PG1-CH05-0005). M.S.K. was a postdoctoral fellow supported by the 2001 BK21 project for Medicine, Dentistry and Pharmacy, Seoul National University.

M.S.K. and C.Y.Y. contributed equally to this work and should both be considered as first authors.

Abbreviations

 
• ADD

  Adipocyte determination and differentiation-dependent factor;

 
• aP2

  adipose protein 2/fatty acid-binding protein;

 
• FBS

  fetal bovine serum;

 
• FITC

  fluorescein isothiocyanate;

 
• GAPDH

  glyceraldehyde-3-phosphate dehydrogenase;

 
• GHRP

  GH-releasing peptide;

 
• GHS-R

  GH secretagog receptor;

 
• GPCR

  G protein-coupled receptor;

 
• IRS-1

  insulin receptor substrate 1;

 
• MTS

  2-(4′,5′-dimethyl-2′-thiazolyl)-3-(4″-sulfophenyl);

 
• PI

  propidium iodide;

 
• PI3K

  phosphatidylinositol 3 kinase;

 
• PKC

  protein kinase C;

 
• PPAR

  peroxisome proliferator-activated receptor;

 
• Rb

  retinoblastoma;

 
• TUNEL

  terminal deoxynucleotidyl transferase-mediated dUTP biotin nick-end labeling.