Uncoupling of 3T3-L1 gene expression from lipid accumulation during adipogenesis

Abstract

Adipocyte differentiation comprises altered gene expression and increased triglyceride storage. To investigate the interdependency of these two events, 3T3-L1 cells were differentiated in the presence of glucose or pyruvate. All adipocytic proteins examined were similarly increased between the two conditions. In contrast, 3T3-L1 adipocytes differentiated with glucose exhibited significant lipid accumulation, which was largely suppressed in the presence of pyruvate. Subsequent addition of glucose to the latter cells restored lipid accumulation and acute rates of insulin-stimulated lipogenesis. These data indicate that extracellular energy is required for induction of adipocytic proteins, while only glucose sustained the parallel increase in triglyceride storage.

1. Introduction

The ability to store and release energy to maintain energy balance throughout life is critical for human survival. The adipocyte is a key component in this process due to its immense capacity for energy storage and mobilization. During the past 20 years, there has been an explosion in the incidence of obesity and secondary development of other diseases such as diabetes, atherosclerosis, and cancer. Accordingly, significant research efforts have focused on understanding the molecular mechanisms involved in the etiology of obesity. As part of this research, it is important to understand the impact of nutrient availability in regulating adipose formation and function.

Differentiation of preadipocytes to adipocytes entails the increased expression of numerous proteins, including the transcription factors peroxisome proliferator-activated receptor γ (PPARγ) isoforms and CCAAT/enhancer-binding protein α (C/EBPα), proteins involved in glucose and lipid metabolism, insulin signaling components and adipokines such as adiponectin and leptin [1–3]. In parallel, during adipogenesis there is a dramatic increase in triglyceride storage and formation of lipid droplets. Triglyceride synthesis involves the esterification of three free fatty acid (FFA) molecules to a glycerol-3-phosphate carbon backbone; conversely, triglyceride hydrolysis releases three FFAs and glycerol for use by other tissues during increased energy demand. The extremely low level of glycerol kinase expression in adipocytes [4,5] prevents the glycerol released during lipolysis from being recycled back to glycerol-3-phosphate for reuse during lipogenesis. Thus, insulin-stimulated glucose uptake and metabolism is required to provide glycerol-3-phosphate for triglyceride synthesis. Additional research exists, however, which suggests that pyruvate and other non-glucose precursors can generate 3-glycerol phosphate via the glyceroneogenic pathway [6–8]. Yet the effect of nutrient availability during adipogenesis on the interdependence of increased gene expression and lipid accumulation has not been extensively studied. We examined the effects of varying extracellular energy sources during differentiation of 3T3-L1 adipocytes, and found that these two markers of adipogenesis can be regulated separately.

2. Materials and Methods

2.1 3T3-L1 cell culture and differentiation

3T3-L1 fibroblasts were cultured and differentiated as previously described [9], with the modification that DMEM with 0 glucose/pyruvate, 25 mM glucose or 25 mM pyruvate was used as indicated. Following completion of the differentiation protocol, cells were then maintained in DMEM plus 10% FBS media and the indicated nutrient for an additional 1–4 days. Oil Red O (Sigma) staining of 3T3-L1 adipocytes was performed as described [10].

2.2 Preparation and analysis of cellular lysates

Cells were washed three times with phosphate-buffered saline (PBS) on ice. For immunoblotting experiments, cells were scraped into homogenization buffer (50 mM Hepes (pH 7.4), 150 mM NaCl, 10 mM NaF, 10 mM EDTA, 10% glycerol, 0.5% Triton X-100, and protease inhibitors added just before use). Lysates were centrifuged for 10 min, 10,000 X g, 4°C, and supernatants were transferred to new tubes. Protein determinations were performed using the Bradford method. For immunoblots, Laemli buffer was added to lysates prior to boiling 2 min, with the exception of samples analyzed with GLUT-antibodies which were diluted with Laemli buffer lacking 2-mercaptoethanol and were heated for 10 min at 37°C. All samples (50–100 μg of protein) were resolved on 10% sodium dodecyl sulfate polyacrylamide gels and transferred to nitrocellulose (Schleicher and Schuell, Keene, NH). Western blots were probed as described [11] with antibodies against glycogen synthase and adiponectin (Chemicon, Temecula, CA), protein phosphatase-1 (PP1; sc7482), C/EBPα, C/EBPβ, PPARγ and β subunit of the insulin receptor (IRβ) (Santa Cruz Biotechnology, Santa Cruz, CA), phospho-Akt (Thr 308) and phospho-MAP kinase (Cell Signaling Solutions), perilipin (Research Diagnostics Inc, Flanders, NJ) and GLUT4 and GLUT1 (Alpha Diagnostics Inc, San Antonio, TX). Blots were then incubated with horseradish peroxidase-conjugated goat anti-rabbit or goat anti-mouse IgG (Bio-Rad) and developed using ECL reagent (GE Healthcare).

2.3 Metabolic assays

Partitioning of glucose into lipid versus glycogen was determined as previously described [12], with the following modifications: 3T3-L1 cells were differentiated in 12-well dishes and serum-starved for 1 – 2.5 hours in DMEM containing 5 mM glucose and 0.5% FBS. After 15 min treatment −/+ 100 nM insulin, 1 μCi of [¹⁴C]-glucose was added to all wells. Cells were incubated at 37°C for 30 min, washed three times with cold PBS, and collected in 1 ml of PBS. 500 μl of the cell suspension was added to 500 μl of PBS for overnight extraction with 4 ml of Betafluor (National Diagnostics) and measurement of glucose incorporation into lipid by scintillation counting. 500 μl of the cell suspension was added to 500 μl of 60% KOH for glycogen synthesis determination as previously described [11].

3. Results

3.1 Lipid accumulation but not gene expression varies with available energy sources during 3T3-L1 differentiation

To examine the impact of varying extracellular energy source on adipogenesis, 3T3-L1 cells were differentiated by a standard protocol [9], in the absence or presence of varying concentrations of glucose or pyruvate. Oil Red-O staining indicated that cells do not accumulate lipid in the absence of both glucose and pyruvate. In addition, cells accumulated more lipid with glucose as the primary extracellular energy source than in the presence of pyruvate (Fig. 1a). At higher magnification, glucose appeared to produce a 3–4-fold greater accumulation of lipid than pyruvate (Fig. 1a), indicating that although both glucose and pyruvate were sufficient energy sources for adipogenesis, glucose more effectively promoted lipid accumulation.

Fig. 1.

Modulation of 3T3-L1 adipocyte differentiation by varying extracellular energy source. 12-well plates of 3T3-L1 fibroblasts were differentiated (Diff) by addition of DMEM containing 10% FBS, 167 nM insulin, 0.5 mM MIX, 0.25 μM dexamethasone, and the indicated amount of glucose or pyruvate (0–25 mM). After three days, the cells were placed in DMEM containing 10% FBS, 167 nM insulin, and the indicated amount of glucose or pyruvate for two more days. Cells were then cultured (Cul) in DMEM plus 10% FBS and the indicated amount of glucose or pyruvate for an additional four days. (a), Oil Red-O staining of the cells. Higher magnification (10X) of the 25 mM condition is also shown. In parallel, cell extracts were prepared and probed with anti-glycogen synthase (GS), anti-perilipin, anti-adiponectin (Adn), and anti-protein phosphatase-1 (PP1) antibodies (b). Fib: fibroblasts maintained in calf serum media throughout the differentiation protocol. Results are representative of 3–5 independent experiments.

The reduction in lipid accumulation in the presence of pyruvate could be due to either an inhibition of adipogenic gene expression during differentiation and/or an inability of the cells to synthesize triglyceride. To examine the effect of different energy sources on induction of adipocyte-specific genes, 3T3-L1 cells were differentiated in the presence of varying concentrations of glucose or pyruvate. Cell lysates were prepared and the expression of markers of terminal 3T3-L1 differentiation was analyzed by immunoblotting (Fig. 1b; also Fig. 2b). Protein levels of the lipid droplet-associated protein perilipin, the adipokine adiponectin, and glycogen synthase were similarly up-regulated in the presence of either glucose or pyruvate as compared to undifferentiated fibroblasts, although maximal expression of all three proteins occurred at lower glucose vs. pyruvate concentrations (Fig. 1b). Interestingly, in cells differentiated in the absence of both glucose and pyruvate, perilipin and adiponectin expression was not detectable, although glycogen synthase levels were fully induced. In contrast, levels of PP1, which do not change during 3T3-L1 differentiation [13], were constant across all of the conditions (Fig. 1b), indicating that alterations in protein expression were not due to differences in cell viability. Taken together, these results indicate that both extracellular pyruvate and glucose can induce expression of genes indicative of terminal adipocyte differentiation.

Fig. 2.

Effects of addition of glucose to culture media on adipocytic profile. 12-well plates of 3T3-L1 fibroblasts were differentiated (Diff) with no addition (0), 25 mM glucose (Glu) or 25 mM pyruvate (Pyr) as in Figure 1. After removal of the insulin-containing media, cells were cultured for four days in DMEM plus 10% FBS with no addition (0), 25 mM glucose (Glu) or 25 mM pyruvate (Pyr) (Cul). (a), Oil Red-O staining of cells. (b), Cell lysates from replicate plates were analyzed by anti-perilipin, anti-adiponectin (Adn), anti-PPARγ, anti-C/EBPα, anti-C/EBPβ, and anti-protein phosphatase-1 (PP1) immunoblotting. Fib: fibroblasts maintained in calf serum media throughout the differentiation protocol. Results are representative of 2–4 independent experiments.

3.2 Restoration of triglyceride synthesis by addition of glucose to the culture medium

Since gene expression was not markedly affected by differentiation with different energy sources, the decrease in lipid accumulation in the presence of pyruvate could be due to a metabolic block on triglyceride synthesis. Therefore, we next induced cells to differentiate in the presence or absence of glucose or pyruvate and then cultured for 4 more days in media containing 25 mM glucose. Oil Red-O staining revealed that addition of glucose to cells differentiated in pyruvate was sufficient to increase lipid accumulation to levels comparable to that of cells differentiated in the presence of glucose throughout (Fig. 2a). Glucose addition, however, was not sufficient to promote lipid accumulation in cells differentiated in the absence of both glucose and pyruvate.

To determine if the addition of glucose post-differentiation had any effect on protein expression, cell lysates were analyzed by immunoblotting. As in Fig. 1b, cell differentiation in the presence of either 25 mM glucose or pyruvate induced similar expression of perilipin, adiponectin (Fig. 2b) and glycogen synthase (Fig. 3a) vs. cells differentiated in the absence of both glucose and pyruvate. Additionally, the increased expression of the adipocytic transcription factors PPARγ and C/EBPα showed an identical pattern. Expression levels of C/EBPβ and PP1 were similar under all conditions tested and were again used as protein loading controls. Inclusion of glucose in the culture media of cells differentiated in the absence or presence of pyruvate had no discernable effect on the expression of any proteins that were examined.

Fig. 3.

Effects of glucose and pyruvate on insulin signaling in 3T3-L1 adipocytes. 12-well plates of 3T3-L1 fibroblasts were differentiated (Diff) and cultured (Cul) as in Figure 2 (Glu or G: glucose) (Pyr or P: pyruvate). (a), Cell extracts were prepared and probed with anti-glycogen synthase (GS), anti-insulin receptor β (IRβ), anti-GLUT4, anti-GLUT1 and anti-protein phosphatase-1 (PP1) antibodies. (b), Replicate wells were stimulated −/+ 0-100 nM insulin for 5 min and cell extracts were prepared. Insulin signaling was determined by analyzing the lysates using anti-phospho-MAPK (pMAPK) and anti-phospho-Akt (pAkt) antibodies. Results are representative of 2–4 independent experiments.

3.3 Induction of insulin action in 3T3-L1 adipocytes differentiated in glucose or pyruvate

Since addition of glucose to 3T3-L1 cells differentiated in the presence of pyruvate restored lipid accumulation without any detectable change in adipocytic gene expression, insulin signaling and metabolic action were next examined. Cells were differentiated and cultured for 4 days as above, and levels of IRβ and facilitative glucose transporters were determined. Immunoblotting revealed that IRβ and GLUT4 expression were strongly up-regulated by inclusion of either glucose or pyruvate during differentiation. In contrast, GLUT1 levels were dramatically increased in cells differentiated in the absence of both glucose and pyruvate, while subsequent addition of glucose to the culture media decreased GLUT1 expression (Fig. 3a). Both GLUT4 and IRβ expression were lower in cells differentiated in pyruvate vs. glucose, and their levels were not increased upon inclusion of glucose into the culture media (Fig. 3a). Interestingly, glycogen synthase exhibited a higher electrophoretic mobility under the 0 glucose/pyruvate or 25 mM pyruvate differentiation conditions, indicative of increased protein dephosphorylation, and addition of glucose to the culture media reversed this effect.

Next, insulin sensitivity was determined by measurement of insulin-stimulated MAPK and Akt phosphorylation. 3T3-L1 cells were differentiated and cultured as above and then stimulated with 0–100 nM insulin for 5 min (Fig. 3b). Cell lysates were prepared and analyzed by phospho-specific immunoblotting. A stepwise increase in the phosphorylation of MAPK was evident in cells differentiated in glucose, but was not seen in cells differentiated in pyruvate or the absence of both glucose and pyruvate (Fig. 3b, top panel). The addition of 25 mM glucose to the culture media specifically restored the ability of insulin to phosphorylate MAPK in cells differentiated in the presence of pyruvate (Fig. 3b). Cells differentiated under all conditions exhibited a similar robust phosphorylation of Akt with increasing concentrations of insulin. Total levels of MAPK and Akt were identical in all culture conditions (data not shown). These results indicate that modulation of nutrient availability during 3T3-L1 adipogenesis had differential effects on insulin signaling pathways.

To directly measure insulin-stimulated energy storage, 3T3-L1 fibroblasts were differentiated and cultured as above, and then rates of glucose synthesis into glycogen or triglyceride were assayed simultaneously. Cells differentiated under standard 25 mM glucose conditions exhibited maximal rates of insulin-stimulated glycogen synthesis (Fig. 4a) and lipogenesis (Fig. 4b). In contrast, cells differentiated in the presence of pyruvate, or absence of both pyruvate and glucose, had significantly reduced rates of insulin-stimulated glucose storage (Figs. 4a, 4b) that were comparable to levels obtained in undifferentiated fibroblasts. However, basal rates of glycogen synthesis were elevated in the cells differentiated in the absence of both glucose and pyruvate, which may be a result of the enhanced GLUT1 expression and dephosphorylation of glycogen synthase detected in Fig. 3a. Interestingly, addition of 25 mM glucose to the culture medium of cells differentiated in the presence of pyruvate fully restored insulin-stimulated rates of glycogen synthesis and lipogenesis (Figs. 4a, 4b). Moreover, basal lipogenic rates were also significantly increased under these conditions (Fig. 4b), potentially underlying the increase in lipid accumulation seen in Fig. 2a. Cumulatively, these results indicate that both increased adipogenic gene expression and presence of extracellular glucose are required for maximal induction of insulin-stimulated signaling and glucose metabolism.

Fig. 4.

Measurement of glucose incorporation into glycogen and lipid in 3T3-L1 adipocytes. 12-well plates of 3T3-L1 fibroblasts (Fib) were differentiated (Diff) and cultured (Cul) as in Figure 2 (G: glucose) (P: pyruvate). After a 1–2.5 hour serum starvation, cells were stimulated in the absence (Basal) or presence (Insulin) of 100 nM insulin for 15 min. 1 μCi of [¹⁴C]glucose was added to all wells, and cells were incubated for an additional 30 min. Cells were collected in 1 ml of PBS, and 500 μl of each sample was used to determine glucose incorporation into glycogen (a) or lipid (b) from the same well. Results are representative of 3 independent experiments, each performed in triplicate.

4. Discussion

The differentiation of fibroblasts or preadipocytes to adipocytes comprises an up-regulation of adipocyte-specific genes involved in regulation of gene transcription, lipid metabolism and insulin signaling, which underlies the increased lipid accumulation and acquisition of insulin sensitivity [1,2]. We have utilized the 3T3-L1 cell line to investigate the effect of varying energy sources on adipocyte differentiation. 3T3-L1 preadipocytes will differentiate in the presence of fetal bovine serum, dexamethasone, isobutylmethylxanthine, and insulin. Together, this cocktail activates an adipogenic program, which induces the expression of specific transcription factors such as C/EBPα and PPARγ [14]. The cells then undergo terminal differentiation, and begin to produce lipid droplets and express multiple metabolic regulators characteristic of mature fat cells [1,3].

However, the interconnections between changes in metabolic pathways and the coordinated increase in adipocytic gene expression and triglyceride accumulation during adipogenesis had not been extensively studied. The present results demonstrated that varying extracellular energy sources during adipocyte differentiation resulted in the uncoupling of adipocyte gene expression from lipid accumulation. Although inclusion of either glucose or pyruvate in the differentiation media induced a comparable increase in the levels of all genes examined that normally increased during 3T3-L1 differentiation, robust cellular lipid accumulation only occurred in the presence of extracellular glucose. In contrast, 3T3-L1 cells differentiated in the absence of both glucose and pyruvate did not exhibit any characteristics of mature adipocytes. There was no induction of any genes studied that are indicative of terminal adipocyte differentiation nor was there any visible lipid accumulation. However, protein levels of GLUT1 were dramatically increased under these conditions. Further, glycogen synthase appeared markedly dephosphorylated (indicative of activation), and basal levels of glycogen synthesis were also enhanced. These data are in agreement with previous work in 3T3-L1 cells that demonstrated a role for glucose availability and glycogen levels in regulating GLUT1 expression and localization [15–19], and indicate that 3T3-L1 cells respond to a deficit of intracellular energy storage by several mechanisms. These compensatory responses to energy deprivation were completely reversed by inclusion of glucose in the culture medium following differentiation.

3T3-L1 cells differentiated in the presence of pyruvate expressed all adipocytic genes examined but accumulated only minimal amounts of lipid. Although the levels of IRβ and GLUT4 were slightly depressed as compared to adipocytes differentiated in glucose-containing medium, phosphorylation of Akt by insulin was unaffected indicating preservation of insulin signaling via the PI3-kinase pathway. The limited amount of lipid accumulation under these conditions suggested that although these cells were synthesizing some 3-glycerol phosphate from pyruvate via glyceroneogenesis [6–8,20], this pathway was unable to fully support maximal triglyceride synthesis. When glucose was added back to cells for four days after completion of the differentiation protocol, there was no detectable change in the expression of any proteins analyzed. However, triglyceride accumulation was fully restored, indicating that adipocytic proteins underlying lipid metabolism were fully functional in cells differentiated in pyruvate. Further, inclusion of glucose in the medium completely restored insulin-stimulated MAPK phosphorylation and rates of glycogen synthesis and lipogenesis. In fact, basal lipogenic rates were also elevated under these culture conditions, suggesting that the cells differentiated in pyruvate medium might be attempting to “catch-up” when glucose was provided, enabling generation of glycerol-3-phosphate and subsequently triglyceride.

In summary, results indicated that terminal differentiation to a mature adipocyte required an energy source such as glucose or pyruvate. 3T3-L1 cells cultured in the absence of an energy source were unresponsive to the components of the adipogenic differentiation cocktail. In contrast, cells differentiated in the presence of either glucose or pyruvate exhibited similar patterns of expression of adipocyte-marker genes. However, pyruvate was not able to sustain the normal increase in lipid accumulation during adipogenesis, but triglyceride synthesis was fully recovered upon restoration of glucose to the extracellular medium. Thus, during adipogenesis the changes in gene expression can be uncoupled from changes in energy storage.