In vitro aging of 3T3-L1 mouse adipocytes leads to altered metabolism and response to inflammation
- First Online:
- Cite this article as:
- Zoico, E., Di Francesco, V., Olioso, D. et al. Biogerontology (2010) 11: 111. doi:10.1007/s10522-009-9236-0
We used an in vitro model to evaluate the effects of cellular aging and inflammation on the gene expression and protein secretion profiles of adipocytes. 3T3-L1 mouse preadipocytes were cultured according to standard conditions and analyzed at different time points both at the basal state and after an acute stimulation with LPS. The mRNA levels of CCAAT/enhancer-binding protein (C/EBP)α, peroxisome proliferator-activated receptor (PPAR)γ and S100A1 were maximal during adipocyte differentiation and then significantly decreased. The expression of the GLUT4 and IRS-1 genes peaked during differentiation and then decreased in aged cells. The mRNA levels and secretion of adiponectin, quickly rose as adipocytes matured and then declined. The mRNA levels of IL6, as well as its secretion, increased as preadipocytes matured and became old cells; a similar trend was also found for MCP-1. LPS decreased the mRNA levels of C/EBPα and PPARγ at all time points, as well as those of GLUT4, IRS-1 and adiponectin. LPS significantly increased the mRNA levels of IL-6, as well as its secretion, with a similar trend also observed for MCP-1. These data suggest that aging adipocytes in vitro show a decline in pro-adipogenic signals, in genes involved in glucose metabolism and cytoskeleton maintenance and in adiponectin. These changes are paralleled by an increase in inflammatory cytokines; inflammation seems to mimic and amplify the effects of cellular aging on adipocytes.
Adipose tissue has been identified as an endocrine organ, at the interface of inflammation, insulin resistance and cardiovascular diseases (Kershaw and Flier 2004; Shoelson et al. 2006). In fact, obesity has been shown to be associated with a state of low grade inflammation, which results from the increased infiltration of immune cells into adipose tissue (Greenberg and Obin 2006; Weisberg et al. 2003; Xu et al. 2003). This infiltration may induce insulin resistance and the other manifestations of metabolic syndrome, whose prevalence is increasing, and is especially associated, with aging.
The effect of cellular aging on the intrinsic properties of adipocytes and therefore on the metabolism and function of whole adipose tissue is still unknown.
New adipocyte differentiation has also been described in the tissues of adult or old animals (Miller et al. 1984), however, the proliferation of preadipocytes and their capacity to differentiate declines between middle and old age in rats (Kirkland et al. 2002). In fact, it has recently been shown that the expression of the adipogenic transcription factors CCAAT/enhancer-binding protein (C/EBP)α and peroxisome proliferator-activated receptor (PPAR)γ, is significantly reduced in the cells of old animals (Karagiannides et al. 2006). It has been suggested that this age associated reduction in the expression of pro-adipogenic genes, may be associated with an increased expression of anti-adipogenic C/EBP family members, resulting in impaired adipogenesis (Cartwright et al. 2007). The reduced adipogenesis in old animals may also be due to the fact that their cells show an increased susceptibility to lipotoxicity and apoptosis (Guo et al. 2007).
While the early stages of adipogenesis have been the subject of several studies (Cartwright et al. 2007; Guo et al. 2007; Karagiannides et al. 2006; Kirkland et al. 2002; Miller et al. 1984), only one paper has focused on the function and metabolism of adipocytes at different stages of maturation, using an in vitro model of cellular aging (Yu and Zhu 2004). 3T3-L1 cells represent an established and validated experimental system in which to analyze the events that occur during the differentiation and development of adipose tissue under defined conditions (O’Shea Alvarez 1991). Yu and colleagues, using data from these type of adipocyte cultures, showed that an increase in insulin resistance, a decrease in glucose uptake and lipolysis rate and an impairment in cell metabolism, coupled with reduced leptin and adiponectin production, were all associated with cell aging (Yu and Zhu 2004).
While a few papers have described the effects of inflammation on the metabolism and function of adipocytes, using in vitro experimental systems, these data were limited to preadipocytes and young adipocytes (Chung et al. 2006; Poulain-Godefroy and Froguel 2007). The effects of inflammation on old adipocytes are still unknown.
The aim of this study was to evaluate the effects of in vitro aging on the metabolism of 3T3-L1 adipocytes and to examine in this system the effects of inflammation at different stages of cellular maturation. We used cultures of mouse 3T3-L1 adipocytes as a special model to study specific changes in the metabolism and function of adipocytes during maturation and aging under defined conditions.
3T3-L1 adipocyte culture
3T3-L1 pre-adipocytes (ECACC-Sigma–Aldrich) were seeded in 24-well plates at a density of 8,000 cells/cm2 and cultured according to standard conditions. Briefly, cells were grown in 5% CO2 in Dulbecco’s modified Eagle’s medium (D-MEM) containing 10% fetal bovine serum (FBS, Bio-Whittaker Europe, Cambrex, USA), 2 mM l-Glutamine and 1% penicillin-streptomycin mixture (Sigma–Aldrich), 50 μg/ml polymyxin B (Sigma–Aldrich) and 8 mg/l tylosin (Sigma–Aldrich). Two days post-confluence, cells were induced to differentiate in complete D-MEM with 20 μg/ml insulin, 0.5 mm IBMX and 1 μM dexamethasone for 3 days. Postinduction day (PID) 0 refers to preadipocytes just before induction. Fresh medium containing only insulin was added for a further 2 days according to ECACC-Sigma–Aldrich’ instructions. Thereafter, to avoid the influence of insulin on metabolic and functional profile of 3T3-L1 adipocytes, the cells were cultured in D-MEM only, up to PID 20 and the medium was changed every 2 days. All analyses were conducted at PID 5, 10, 15 and 20. Before assays, cells were washed and maintained overnight in D-MEM, without FBS. Assays were conducted both at the basal state and after stimulation with lipopolysaccharide (Sigma–Aldrich) (LPS 1 μg/ml for 16 h); concentrations and incubation times were as used in other similar experimental models (Ajuwon and Spurlock 2005) and as previously identified in our laboratory by time-concentration curve experiments.
The expression of the genes encoding C/EBPα, PPARγ, S100A1, GLUT4, IRS-1, adiponectin, IL-6, MCP-1, was evaluated. Cells were washed with PBS and homogenized with QIAshredder columns (Qiagen GMBH, Hilden, Germany) prior to RNA extraction. Total RNA was recovered using the RNeasy Mini kit (Qiagen) and DNase treated (RNase Free DNase set, Qiagen). The quality and quantity of RNA samples were determined using an Agilent 2100 bioanalyzer. For quantitative Real-time PCR, 35 ng of total RNA for each gene were reversed transcribed into cDNA in 20 μl reactions with Iscript cDNA Synthesis (Bio-Rad, Hercules, CA, USA). Aliquots of the reverse transcriptase reaction, or water only (negative control), were amplified using the QuantiTect SYBR Green PCR Kit (Qiagen) and 0.3 μM of pre-designed QuanTitect primer Assays for each gene (Qiagen, GMBH, Hilden, Germany). The Entrez gene IDs are 12606 for C/EBPα, 19016 for PPARγ, 20193 for S100A1, 20528 for GLUT4, 16367 for IRS-1, 11450 for adiponectin, 16193 for IL-6, 20296 for MCP-1, and 11461 for β-actin reference gene. Thermal cycling conditions for PCR reactions were 95°C for 15 min followed by 40 cycles of 94°C for 15 s, 55°C for 30 s and 72°C for 30 s. The abundance of each gene product was calculated by regression analysis against a standard curve generated by two-fold serial dilutions of positive PCR controls for each gene. The C/EBPα, PPARγ, S100A1, GLUT4, IRS-1, adiponectin, IL-6, MCP-1 values were normalized against β-actin for each sample. mRNA quantification was performed in triplicate for each well and in 3 wells for each experimental condition. Gene expression analysis was conduced by GeneX Software (Bio-Rad).
The concentration of cytokines (IL-6, MCP-1 and adiponectin) in conditioned media was measured using mouse-specific ELISA kits from Pierce Endogen (Rockford, IL) except for adiponectin, whose kit was from B-Bridge International (Otsuka Pharmaceuticals, Japan). The standard curve of each cytokine was, respectively, 20–0.51 μg/ml, 1,000–16 pg/ml and 8–0.25 ng/ml. Each experimental condition was tested in 3 different wells and measured in duplicate.
Trypan blue staining
Before staining, 3T3-L1 cells were washed with PBS. In order to quantify the percentage of permeable cells, cells were first released by trypsinization (0.5 mg/ml trypsin). 50 μl of resuspended cells were stained with an equal volume of trypan blue (0.04%) for 2 min and immediately examined under a microscope. Typically nonviable cells stained dark blue. The percentage of stained cells at each time point (PID 5, 10, 15 and 20) was determined by examining at least two independently stained cell populations. At least 400 individual cells were counted for each specimen.
Data are presented as means ± SD for each experimental condition. Data which did not show a normal distribution were log transformed before analyses. Differences between groups were evaluated by univariate analysis (ANOVA). A P value < 0.05 was used to determine statistical significance. All statistical analyses were performed using the SPSS statistical package (SPSS User’s Guide 1986).
The effects of cellular aging on the gene expression and protein secretion profiles of 3T3-L1 adipocytes
We used Real-Time PCR to determine the relative expression levels of genes encoding different cytokines, while their secretion into the culture media of 3T3-L1 adipocytes was detected using ELISA based techniques, at different stages of cell maturation. We used as a model of cellular aging cultures of 3T3-L1 adipocytes because in this system cell ages can be relatively synchronized; the ages of cellular cultures are usually indicated by the number of days following the induction of differentiation (PID).
The effects of inflammation on the gene expression and protein secretion profiles of 3T3-L1 adipocytes at different stages of cell maturation
In the second part of these experiments, LPS stimulation was performed for 16 h, on differently aged adipocyte cultures (PID 5, 10, 15 and 20), and the mRNA levels of different genes was measured as was the secretion into the culture medium of their respective proteins at each time point.
After LPS treatment, the mRNA levels of C/EBPα were significantly lower in mature and aged adipocytes (PID 10–20), when compared with non stimulated cells of the same age (Fig. 1). We also observed a trend towards a down-regulation of the expression of PPARγ mRNA (Fig. 1) and of S100A1 mRNA (Fig. 2) in LPS treated 3T3-L1 adipocytes compared to basal levels. Likewise, we found lower levels of mRNA of GLUT4 after stimulation with LPS, with significantly lower expression of this gene in mature adipocytes at PID 10, compared with unstimulated cells of the same age (Fig. 3). Moreover, IRS-1 mRNA levels tended to be lower in LPS treated adipocytes at all time points compared to cells of the same stage of maturation which had not been exposed to inflammatory stimuli (Fig. 3).
As expected, LPS treatment decreased the expression levels of adiponectin mRNA as well as its secretion into the culture medium; this was particularly significant in mature adipocytes (PID 10) (Fig. 4). Conversely, the mRNA levels and secretion of IL-6 increased significantly in LPS treated 3T3-L1 adipocytes, at all time points when compared to the basal state (Fig. 5). Similarly, we observed a trend towards an increase in MCP-1 mRNA expression and protein secretion in LPS treated 3T3-L1 adipocytes of different ages (Fig. 6).
This study characterized, in a particular in vitro model, chronological changes in adipocyte metabolism and function during prolonged cell culture and explore the effects of inflammation at different stages of cellular maturation. The cell culture data presented in these experiments shows that adipocyte aging is characterized by decreased levels of key differentiation genes, such as C/EBPα and PPARγ, as well as of S100A1 which is required for the maintenance of the fat cell phenotype and cytoskeleton. Finally in these experiments we observed a significant increase, in pro-inflammatory signals, such as IL-6 and MCP-1, as well as a decrease in adiponectin production and insulin sensitivity with cell maturation and aging.
The growth and aging of 3T3-L1 adipocytes was investigated using a culture system throughout a 20-day study period; approximately 3 days after induction, cells start to mature accumulating lipid droplets, and they reach full maturity between PID 5 and 10, before entering into a period characterized by cell aging, between PID 10 and 20 (Yu and Zhu 2004).
As expected, the expression of the C/EBPα and PPARγ encoding genes increased in the initial phases of adipogenesis. An initial increase in the C/EBP isoforms β and δ is known to promote the induction of PPARγ and C/EBPα (Rosen et al. 2002); furthermore it has been suggested that C/EBPα influences the C/EBPβ protein and consequently PPARγ expression through a positive feedback loop (Rosen et al. 2002). In our experiments the maximal increases in PPARγ mRNA expression observed at PID 10, may be, at least in part, mediated by an increase in C/EBPα mRNA levels during the earlier phases of adipogenesis.
During adipocyte differentiation and maturation, we also observed maximal levels of S100A1 expression, followed by a significant decrease with cell aging. The S100A1 protein is considered to be a marker of adipocyte differentiation, discriminating between the early stages of adipogenesis and fibroblasts (Cinti et al. 1989). Its expression quickly rose upon induction of differentiation in cultures of 3T3-L1, paralleling triacylglycerol accumulation inside the cell (Kato et al. 1988). Moreover, S100 proteins have been shown to have an important role in regulating cell morphology and cytoskeleton constituents (Donato 2003). Treatment of 3T3-L1 adipocytes with colchicine, which is known to destroy the microtubular cytoskeleton, inhibits the induction of S100 protein and suppresses the accumulation of triacylglycerol (Kato et al. 1988). Our data which show a significant decrease in S100A1 expression with cell aging may thus suggest a progressive impairment in the cytoskeleton of aged adipocytes.
Recent data suggest that C/EBPα and PPARγ may also play an important role in the maintenance of the fat cell phenotype and insulin sensitivity (Olofsson et al. 2008). In adipose tissue, several target genes involved in glucose and lipid metabolism, have been identified for C/EBPα, including adiponectin, 11-beta hydroxysteroid dehydrogenase, diacylglycerol acyltransferase 2, GLUT4 and leptin (Olofsson et al. 2008). Similarly, genes downstream of PPARγ, including aP2, carnitine palmytoil transferase- and PPARγ co-activator, have been shown to regulate the pathways of fatty acid handling as well as mitochondrial function (Lehrke and Lazar 2005).
Thus, it is feasible that the reduced expression of C/EBPα and PPARγ in mature and aging adipocytes as seen in our experimental model contributes, at least in part, to the metabolic profile of the senescent adipocyte characterized by impaired insulin sensitivity through reduced expression of GLUT4, IRS-1 and adiponectin. These observations confirm a previous study, where Yu and colleagues showed that older 3T3-L1 adipocytes are more insulin resistant, have decreased glucose uptake and fuel consumption as well as an impaired fatty acid re-esterification with reduced fatty acid release from adipocytes (Yu and Zhu 2004). These authors proposed a “cell-aging hypothesis”, as being responsible under various conditions for differences in the characteristics, function and metabolism of adipose tissue. Even a brief period of imbalance between adipocyte production and apoptosis (i.e. in adipocytes turnover rate) can lead to shifts in the average age of the adipocyte population, influencing whole tissue function and metabolism (Yu and Zhu 2004).
However, our findings and those of Yu (Yu and Zhu 2004) seem to suggest that, at least in in vitro experimental models, there is an age dependent deterioration in every aspect of adipocyte functionality. Interestingly in our experimental study, IL6 mRNA gene expression and secretion in culture medium, showed a significant increase from preadipocytes to mature and old cells. Similarly MCP-1 mRNA expression and secretion rose during the maturation and aging of adipocytes, while adiponectin mRNA expression and secretion declined, suggesting a shift towards a more pro-inflammatory profile with cell aging.
It has been suggested that adipocyte size is one of the factors which regulates cytokine release within adipose tissue, besides inflammatory stimuli and catecholamines (Coppack 2001). In fact, Skurk and colleagues, after separating different human adipocyte fractions according to mean cell volume, demonstrated a strong relationship between adipocyte size and the expression and secretion of adipocytokines, with increased production of proinflammatory adipokines in very large hypertrophic cells (Skurk et al. 2007). The constant increase in TG mass with cell aging, observed in 3T3-L1 by Yu and colleagues, may suggest an increase in adipocyte size with cell aging (Yu and Zhu 2004). Moreover, observations in normal rats of different ages, revealed a marked increase in the lipid content of older rats, accomplished exclusively by an increase in cell size (Hirsch and Han 1969). However, further studies are needed to confirm and support these data especially as some debate exists (Cartwright et al. 2007).
In the second part of these experiments we stimulated 3T3-L1 adipocytes with LPS, to see if there were differences between young and old adipocytes in their response to an acute inflammatory stimulus. LPS induced a significant decrease in the expression of the C/EBPα, PPARγ and S100A1 encoding genes, as well as in the mRNA levels of GLUT4 and IRS-1 and in adiponectin production. Furthermore, LPS treatment significantly increased the mRNA levels of IL-6, as well its secretion into the culture medium, with a similar trend also being observed for MCP-1 mRNA and protein levels.
Our findings suggest that older adipocytes maintain in culture their capacity to respond to acute inflammation. LPS initiates proinflammatory signaling through Toll-like-receptors (TLRs), expressed in both preadipocytes and adipocytes, which can activate the NF-κB pathway and consequently the production of proinflammatory cytokines and chemokines (Lin et al. 2000). A few papers (Chung et al. 2006; Poulain-Godefroy and Froguel 2007) have explored the potential induction of various pro-inflammatory cytokines by LPS in preadipocytes and mature adipocytes, showing that the production of pro-inflammatory products is more pronounced in preadipocytes than in differentiated adipocytes. However, the design of these studies make it difficult to draw conclusions about adipocytes at a stage older than mature cells.
From our experiments it seems likely that inflammation mimics and anticipates the effects of cellular aging in mature and young adipocytes, amplifying the effects of cellular aging itself at the later stages of adipocytes development.
Some limitations of our study should be recognized. First, these data are from a particular in vitro model of adipocyte aging, that, although recognized as an adequate experimental system in which to analyze the differentiation and development of adipose tissue, need to be confirmed in cultures of human adipocytes and then possibly in vivo. Second, the deterioration of adipocyte function could be due to an increasing proportion of nonviable adipocytes with cell aging, although we did not see differences in the percentage of trypan blue positive, nonviable cells throughout the experiment. The metabolic and functional changes of 3T3-L1 adipocytes at extreme ages (PID 20) when taken together point to a global decline of cell functionality that is difficult to interpret and needs further study. Finally metabolic and functional changes of aging adipocytes could perhaps have been better characterized by studying the changes in cell size and morphology with cellular aging in culture.
In conclusion our data suggest that in vitro cellular aging of adipocytes is characterized by declining pro-adipogenic signals, reductions in the expression of genes involved in glucose metabolism and cytoskeleton maintenance, as well as reductions in the production of adiponectin and increases in inflammatory cytokines. Indeed, inflammation seems to mimic and amplify the effects of cellular aging on adipocytes.
This study was supported by grants from MIUR project COFIN n 2005063885_005.