Expression profiling during adipocyte differentiation of 3T3-L1 fibroblasts

Expression profiling during adipocyte differentiation of 3T3-L1 fibroblasts

Gene 299 (2002) 95–100 www.elsevier.com/locate/gene Expression profiling during adipocyte differentiation of 3T3-L1 fibroblasts Bart A. Jessen*, Greg...

178KB Sizes 0 Downloads 76 Views

Gene 299 (2002) 95–100 www.elsevier.com/locate/gene

Expression profiling during adipocyte differentiation of 3T3-L1 fibroblasts Bart A. Jessen*, Greg J. Stevens Drug Safety Evaluation, Pfizer Global Research and Development/Agouron Pharmaceuticals, Inc., 10724 Science Center Drive, San Diego, CA 92121, USA Received 13 May 2002; received in revised form 28 August 2002; accepted 17 September 2002 Received by R. Di Lauro

Abstract The 3T3-L1 cell line is a well-established and commonly used in vitro model to assess adipocyte differentiation. Over the course of several days confluent 3T3-L1 cells can be converted to adipocytes in the presence of an adipogenic cocktail. Changes in gene expression were measured by DNA microarrays at three time points (24 h, 4 days, and 1 week) during the course of differentiation from preadipocytes to mature adipocytes. Several functional categories of genes were affected by adipocyte conversion. In addition, seven genes were found to be commonly altered by 5-fold or more by adipocyte conversion at all three time points. Lipocalin 2, haptoglobin, serum amyloid A3, stearoylCoA desaturase, and 11b-hydroxysteroid dehydrogenase 1 were induced while actin a2 and procollagen VIII a1 were suppressed by adipocyte differentiation. Further study of the regulation of these genes and pathways will lead to an increased understanding of the biochemical pathways involved in adipocyte differentiation and possibly to the identification of new therapeutic targets for treatment of obesity and other metabolic diseases. q 2002 Elsevier Science B.V. All rights reserved. Keywords: DNA microarray; 11b-Hydroxysteroid dehydrogenase 1; Lipocalin 2; Haptoglobin

1. Introduction 3T3-L1 cells, originally derived from mouse embryos (Green and Kehinde, 1974), have served as a useful in vitro model for adipocyte differentiation and function. When stimulated to differentiate with a cocktail containing dexamethasone, methylisobutylxanthine, fetal bovine serum, and insulin, these cultures take on adipocyte characteristics (Rubin et al., 1978). The most apparent of these changes is the accumulation of lipid by converted 3T3-L1 adipocytes, which can be detected microscopically. Several adipocyte differentiation markers were identified or studied using the 3T3 L1 culture system. These markers Abbreviations: 11bHSD1, 11b-hydroxysteroid dehydrogenase 1; aP2, adipocyte protein 2; PPARg, peroxisome proliferator activated receptor g; C/EBPa, CAAT enhancer-binding protein a; col 8, collagen VIII a1; EST, expressed sequence tag; GOS2, G0/G1 switch protein 2; HP, haptoglobin; LC2, lipocalin 2; MAP, mitogen-activated protein; pref-1, preadipocyte factor 1; rhoB, ras-related homolog B; SCD1, sterol CoA desaturase 1; SREBP, sterol response element binding protein. * Corresponding author. Tel.: þ 1-858-622-6039; fax: þ 1-858-6225999.. E-mail address: [email protected] (B.A. Jessen).

include the fatty acid binding protein aP2, the triglyceride metabolizing enzyme glycerol-3-phosphate dehydrogenase, and the secreted product adipsin (Spiegelman et al., 1983). In addition, 3T3 L1 cultures were employed to study regulatory factors such as the adipogenic transcription factors the peroxisome proliferator activated receptor gamma (PPARg) (Tontonoz et al., 1994), the CAAT enhancer-binding protein a, b, and d (C/EBPa, -b, -d) (Darlington et al., 1998), and the sterol response element binding proteins (SREBP) (Kim and Spiegelman, 1996), as well as the anti-adipogenic factor, preadipocyte factor 1 (pref-1) (Smas et al., 1997). Many of the genes associated with the differentiation and maintenance of the adipocyte phenotype could be involved in metabolic disorders such as type II diabetes and obesity. Several genes associated with adipocyte differentiation are often mutated or aberrantly expressed in naturally occurring or transgenic mouse models of obesity and metabolic disorders (Robinson et al., 2000). In this study we present genes differentially expressed during adipocyte conversion as measured by DNA microarray technology. While many of the genes found were well-established adipogenic

0141-933/02/$ - see front matter q 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 1 1 9 ( 0 2 ) 0 1 0 1 7 - X

96

B.A. Jessen, G.J. Stevens / Gene 299 (2002) 95–100

markers, several others represent potential new therapeutic targets for obesity and metabolic diseases.

3. Results 3.1. Expression time course of selected adipocyte differentiation markers

2. Materials and methods

2.1. Cell culture 3T3-L1 cells (ATCC CL-173) were grown in DMEM containing 10% calf serum and 10 mg/ml gentamycin. Upon confluence, medium was changed and cultures were treated with adipocyte conversion cocktail containing 10% fetal bovine serum, 0.5 mM methylisobutylxanthine, 1 mM dexamethasone, and 1 mg/ml insulin. Subsequent medium changes contained 1 mg/ml insulin and occurred every three days until the cells were harvested.

2.2. DNA Microarray

A subset of genes selected by their sensitivity to regulation by inhibitors of adipocyte differentiation (data not shown) were measured for their expression level at several time points after initiation of adipocyte conversion. Northern analysis (Fig. 1) shows that while annexin VIII was barely detectable one day after conversion, aP2, 11bHSD1, spot14, and GOS2 were not detected. At day 3, high expression levels of all but the spot14 gene were measured. The maximum difference between expression levels in preadipocyte and converted adipocytes occurred on days 4 and 5. By days 7 and 9 the expression levels of these five genes had greatly decreased in converted cultures while the expression of 11bHSD1 and spot14 increased in preadipocytes. The fold change between mature and preadipocytes was several times greater on days 4 and 5 than on days 7– 9 for the genes compared in the time course. 3.2. Analysis of genes affected by adipocyte differentiation

RNA was isolated from 3T3-L1 cultures for microarray analysis using single step extraction with Trizol reagent (Gibco BRL). cDNA was synthesized from 10– 20 mg total RNA using the Superscript Choice System (Gibco BRL, protocol modified by Afftmetrix) and T7-(dT)24 primer. Biotin-labeled cRNA was prepared using the Bioarray High Yield RNA Transcript Labeling Kit (Enzo). cRNA was fragmented to an average size of 50 – 200 bp before hybridizing to Mu 11k sub B or MG 74A V2 mouse genome arrays (Affymetrix) containing 6,595 or 12,422 probe sets, respectively. Scanned Genechips were analysed by selecting genes that were determined to be present in at least two samples per time point and increased or decreased in at least two of the replicates at each time point by Affymetrix analysis software. Only average fold changes from three replicate experiments of greater than 5 or less than 2 5 are shown. This fold change level was chosen to select the most specific and significant genes involved in the dramatic morphological change that occurs during adipocyte conversion of 3T3-L1 cells.

2.3. Northern blotting Total RNA was fractionated on a 1% denaturing agarose gel and transferred to nylon membranes. Northern blots were hybridized with 32P-labeled cDNAs for 11bHSD-1, annexin VIII, spot 14, and GOS2-like genes and stripped with 0:1£ SSC/0.1% SDS at 95 8C between hybridizations. The density of the blots was determined by phosphorimaging (Typhoon, Molecular Dynamics) and the values expressed as mRNA band density normalized to 18S rRNA.

RNA from preadipocytes and adipocytes harvested 24 h, 4 days and 1 week after initiation of adipocyte conversion was probed with mouse genome arrays. The 24-h, 4-day, and 1-week conversion protocols resulted in alterations in expression levels of 24, 186, and 70 genes, respectively. Fig. 2 shows the percentage of genes affected, categorized by function at the three time points. The largest percentage (12 –22%) of genes at all time points belonged to the unknown category. These included unidentified ESTs and genes for which functions have not been determined. Gene products associated with cell cycle, cytoskeleton, and extracellular matrix and cell adhesion were represented at similar levels across the time points. The remaining functional categories displayed time point-dependent differences in expression levels. Genes involved in the acute phase inflammatory response were affected at the 24-h time point at a rate four times greater than that of the later time points. Genes involved with apoptotic pathways showed a similar temporal pattern with the 24-h time point having the most changes. Transport related genes had the fewer changes at the 4-day time point than at 24 h and were unaffected at 1 week. The pathways involved in amino acid and carbohydrate metabolism were not affected at 24 h, but were equally altered at the later time points. Fatty acid and lipid metabolizing gene products were affected at rates that increased slightly with time. Genes involved with fatty acid and lipid binding and those secreted as serum components changed with higher frequency at 1 week than at 4 days and were not affected at 24 h. DNA metabolism and signal transduction related genes were similarly altered at the earlier two time points but were far less affected at the 1-week time point. The remaining

B.A. Jessen, G.J. Stevens / Gene 299 (2002) 95–100

97

Fig. 1. Differentiation dependence of gene expression. Confluent 3T3-L1 cultures were subjected to the adipocyte conversion protocol (þ ) or continued to be fed with growth media (2) for up to 9 days. Cells were harvested at the days indicated (1, 3, 4, 5, 7, 8 and 9) and 10 mg of total RNA was used for northern blot analysis. 18s rRNA was measured by fluoro-imaging of ethidium bromide staining and mRNAs were detected from a single northern blot by individual hybridizations with 32P labeled cDNA probes. The probes were stripped between hybridizations.

functional categories showed a bi-phasic temporal pattern, with transcriptionally related genes being most altered and steroid metabolizing and proteolytic processing genes being least affected at 4 days. Several genes were altered 5-fold or more after adipocyte conversion at all three time points (Fig. 3). Lipocalin 2 (LC2), haptoglobin (HP), serum amyloid A3 (SAA3), stearoyl-CoA desaturase (SCD1), and 11bHSD1 were induced by adipocyte differentiation, while actin a2 and collagen VIII a1 (col 8) were suppressed. There was minimal difference in the induction level of LC2 between the time points. Both HP and SCD1 increased with time while SAA3 was only slightly more elevated at 1 week than the earlier time points. 11bHSD1, actin a2, and col 8 displayed a biphasic response with the 4-day time point having the greatest fold change due to adipocyte conversion.

4. Discussion Interpretation of changes in the expression of large numbers of genes and alterations in functional pathways is difficult without knowing the relationships between the

individual genes. However, some rational conclusions can be made in the context of what is known about adipocyte differentiation. Progression of the conversion from committed preadipocytes to mature adipocytes is charted with the appearance of early, intermediate and late expression markers and triglyceride accumulation (Gregoire et al., 1998). In addition, genes that are inhibitory to adipogenesis are suppressed during differentiation. Most of the regulation of these adipogenic and inhibitory genes is reported to be at the transcription level (Gregoire et al., 1998). Our purpose in this study was to identify genes whose mRNA expression changes during adipocyte differentiation. A recent publication details a similar study profiling the expression differences between differentiated 3T3-L1 adipocytes and preadipocytes (Guo and Liao, 2000). The experimental design and technology of the present study differed significantly from the previous report. The previous report evaluated a single replicate at a single time point during differentiation (6 days), as opposed to the three time points and triplicate experiments shown in the present study. In addition the previous report used cDNA EST probes spotted on membranes while the current study used the more specific oligonucleotide probe format. Interestingly, of the

98

B.A. Jessen, G.J. Stevens / Gene 299 (2002) 95–100

Fig. 2. Cellular pathways affected by adipocyte differentiation. Confluent 3T3-L1 cultures (n ¼ 3) were treated with adipocyte conversion media for 24 h, 4 days, or 1 week before harvesting and RNA isolation. Total RNA (10–20 mg) was used to generate biotin labeled cRNA samples to be probed by Mu11k subB (24 h and 1 week) or MG U74A (4 days) DNA arrays containing 6,595 and 12,422 probe sets, respectively. Those genes whose average expression differed by . 5-fold when compared to time matched non-converted cultures were grouped by functional category. The percent of the total number of genes affected that each category represented at the three time points is shown.

differentially expressed genes ð. 10 fold) found in the previous report, none matched the gene bank accession numbers of those reported in this study. In fact, only five probes sets in that report are found on the arrays used in this study. These include W50665 (cadherin), W96985 (tyrosine phosphatases d), AA050453 (ATP receptor), threonyltRNA synthase (W97102), Tubulin a-4 (W11746) and W11665 (leucyl-tRNA synthase) from the Mu 11k sub B array and AW047630 (EST) from the MG 74A V2 array. These differences could reflect the different probes present on the two array formats and the differences in study design. Although growth arrest is required for the initiation of adipocyte differentiation in many adipogenic cell lines (Amri et al., 1986), clonal expansion is one of the first steps in response to adipogenic factors and is required in 3T3-L1 cells (Kuri-Harcuch and Marsch-Moreno, 1983). The dependence on clonal expansion could explain the expression of the anti-apoptotic and cell cycle genes. The DNA replication, which takes place during clonal expansion, could also explain the greater fraction of DNA and nucleotide metabolism and nuclease related transcripts in the earlier versus later time points. Induction of PPARg and C/EBPa are among the early changes in gene expression reported in adipocyte differentiation (Brun et al., 1996). The fact that these transcription factors were induced 14- and 7-fold, respectively, on day 4 is consistent with the timing reported previously. The frequency of altered expression of transcriptional regulators was higher on day 4 than at the other time points. The

decrease in transcription factor mRNA levels after 1 week could reflect that adipogenic transcripts have reached equilibrium and the need for increased transcription has diminished. Signal transduction pathways comprised of receptors, kinases, phosphatases, and GTPases are responsible for translating hormone, growth factor, and other environmental signals into synthesis or modification of new gene products. These pathways are represented to a greater extent in the earlier time points (24 h and 4 days) than at 1 week. This could be consistent with the completion of the

Fig. 3. Genes altered throughout adipocyte differentiation. The seven genes that were found to be altered 5-fold or more by adipocyte conversion at all time points tested (24 h, 4 days, and 1 week) and their respective average fold changes are shown.

B.A. Jessen, G.J. Stevens / Gene 299 (2002) 95–100

differentiation process and a lessening of the need for direction of new transcript synthesis. Morphological changes are also noted in the adipocyte differentiation process. Adipocytes adopt a spherical shape, changing from their fibroblastic precursor. This process occurs independent of lipid accumulation (Kuri-Harcuch et al., 1978) and therefore is likely the result of changes in structural protein content. The presence of structural gene products involved in forming the cytoskeleton and extracellular matrix among those differentially expressed is therefore not surprising. Also, the finding that expression of genes responsible for the metabolism of fatty acids and lipid gradually increases from 24 h to 1 week after adipocyte conversion was to be expected. Terminal differentiation of adipocytes is associated with a dramatic increase in lipid production and the production of secreted gene products (Gregoire et al., 1998). The increase in lipid production produces the need for proper lipid storage in the form of lipid and fatty acid binding proteins. The increase in lipid binding proteins and secreted serum constituents is reflected in the array results as these functional categories are represented to a much greater extent at the 1-week time point than at earlier time points. Several transcripts remain elevated at all three time points, suggesting their importance throughout the adipocyte differentiation process. LC2, also referred to as 24p3, is a dexamethasone-inducible, secreted, hydrophobic ligand transporting protein (Garay-Rojas et al., 1996). The protein has been shown to have micromolar affinity towards fatty acids and retinoids, shedding some light on its possible role as a carrier of hydrophobic molecules (Chu et al., 1998). Haptoglobin is an acute-phase protein whose serum levels rise during inflammation and whose physiological role is thought to be that of an antioxidant and angiogenic factor (Friedrichs et al., 1995). Another acute phase serum constituent, SAA3, has recently been shown to be upregulated in both adipose tissue and 3T3-L1 cultures by hyperglycemic conditions and has been proposed to play a role in the pathogenesis of type II diabetes (Lin et al., 2001). SCD1 is an enzyme that catalyses the initial desaturation of long-chain fatty acids and represents the first regulatory step in the formation of long-chain unsaturated fatty acids (Enoch et al., 1976). 11bHSD1 is expressed in white adipose tissue and has been implicated in the modulation of glucocorticoid action by enzymatically activating inert intracellular glucocorticoids (Napolitano et al., 1998). Actin a2 is a cytoskeletal protein normally associated with vascular smooth muscle (Min et al., 1988), while col VIII is an extracellular matrix product of endothelial cells, keratinocytes and masts cells. Col VIII may also be involved in vascular smooth muscle migration and angiogenesis (Shuttleworth, 1997). These structural proteins may be involved in maintenance of the morphology of preadipocyte fibroblast. In addition to the genes previously reported to be associated with adipocyte differentiation, the expression of

99

several genes not typically associated with this process was detected in the array. Of particular interest are those genes involved in signal transduction. The ras-related homolog B (rhoB) is up regulated 14-fold by adipocyte conversion at 24 h. RhoB is a GTPase found to be an immediate-early gene inducible by growth factors (Jahner and Hunter, 1991). In addition, the zinc-finger transcription factor, Kruppel-like factor 5, has been described as a delayed early response gene that regulates cellular proliferation (Sun et al., 2001). This gene was also elevated at the 24-h time point by , 9fold. In the context of adipocyte differentiation, rhoB and Krupple-like factor 5 may be involved in the clonal expansion phase. The differential expression at day 4 was marked by the presence of several kinases. In addition to the two genes associated with the mitogen-activated protein (MAP) kinase pathway, which has previously been linked to adipocyte differentiation (Aubert et al., 1999), several other kinases were present. Among these were the homeodomaininteracting protein kinase 2, a putative pim 1-like kinase, sphingosine kinase 1, cGMP-dependent kinase type II, and kinase interacting protein 2. With the exception of sphingosine kinase, which may be involved in cell growth and survival, the precise role of these other kinases in unclear. These genes as well as the unidentified ESTs warrant further investigation into their potential role in adipocyte differentiation and as possible targets for obesityrelated diseases.

References Amri, E.Z., Dani, C., Doglio, A., Grimaldi, P., Ailhaud, G., 1986. Coupling of growth arrest and expression of early markers during adipose conversion of preadipocyte cell lines. Biochem. Biophys. Res. Commun. 137, 903–910. Aubert, J., Belmonte, N., Dani, C., 1999. Role of pathways for signal transducers and activators of transcription, mitogen-activated protein kinase in adipocyte differentiation. Cell. Mol. Life Sci. 56, 538–542. Brun, R.P., Kim, J.B., Hu, E., Altiok, S., Spiegelman, B.M., 1996. Adipocyte differentiation: a transcriptional regulatory cascade. Curr. Opin. Cell Biol. 8, 826–832. Chu, S.T., Lin, H.J., Huang, H.L., Chen, Y.H., 1998. The hydrophobic pocket of 24p3 protein from mouse uterine luminal fluid: fatty acid and retinol binding activity and predicted structural similarity to lipocalins. J. Pept. Res. 52, 390 –397. Darlington, G.J., Ross, S.E., MacDougald, O.A., 1998. The role of C/EBP genes in adipocyte differentiation. J. Biol. Chem. 273, 30057– 30060. Enoch, H.G., Catala, A., Strittmatter, P., 1976. Mechanism of rat liver microsomal stearyl-CoA desaturase. Studies of the substrate specificity, enzyme–substrate interactions, the function of lipid. J. Biol. Chem. 251, 5095–5103. Friedrichs, W.E., Navarijo-Ashbaugh, A.L., Bowman, B.H., Yang, F., 1995. Expression and inflammatory regulation of haptoglobin gene in adipocytes. Biochem. Biophys. Res. Commun. 209, 250–256. Garay-Rojas, E., Harper, M., Hraba-Renevey, S., Kress, M., 1996. An apparent autocrine mechanism amplifies the dexamethasone- and retinoic acid-induced expression of mouse lipocalin-encoding gene 24p3. Gene 170, 173–180.

100

B.A. Jessen, G.J. Stevens / Gene 299 (2002) 95–100

Green, H., Kehinde, O., 1974. An established pre-adipose cell line and its differentiation in culture. Cell 3, 127–133. Gregoire, F.M., Smas, C.M., Sul, H.S., 1998. Understanding adipocyte differentiation. Physiol. Rev. 78, 783–809. Guo, X., Liao, K., 2000. Analysis of expression profiling during 3T3-L1 preadipocyte differentiation. Gene 251, 45 –53. Jahner, D., Hunter, T., 1991. The ras-related gene rhoB is an immediateearly gene inducible by v-Fps, epidermal growth factor, platelet-derived growth factor in rat fibroblast. Mol. Cell. Biol. 11, 3682–3690. Kim, J.B., Spiegelman, B.M., 1996. ADD1/SREBP1 promotes adipocyte differentiation and gene expression linked to fatty acid metabolism. Genes Dev. 10, 1096– 1107. Kuri-Harcuch, W., Marsch-Moreno, M., 1983. DNA synthesis and cell division related to adipose differentiation of 3T3 cells. J Cell Physiol. 114, 39–44. Kuri-Harcuch, W., Wise, L.S., Green, H., 1978. Interruption of the adipose conversion of 3T3 cells by biotin deficiency: differentiation without triglyceride accumulation. Cell 14, 53– 59. Lin, Y., Rajala, M.W., Berger, J.P., Moller, D.E., Barzilai, N., Scherer, P.E., 2001. Hyperglycemia-induced production of acute phase reactants in adipose tissue. J. Biol. Chem. 276, 42077–42083. Min, B.H., Strauch, A.R., Foster, D.N., 1988. Nucleotide sequence of a mouse vascular smooth muscle alpha-actin cDNA. Nucleic Acids Res. 16, 10374. Napolitano, A., Voice, M.W., Edwards, C.R., Seckl, J.R., Chapman, K.E., 1998. 11b-Hydroxysteroid dehydrogenase 1 in adipocytes: expression

is differentiation dependent and hormonally regulated. J. Steroid Biochem. Mol. Biol. 64, 251–260. Robinson, S.W., Dinulescu, D.M., Cone, R.D., 2000. Genetic models of obesity and energy balance in the mouse. Annu. Rev. Genet. 3, 687 –745. Rubin, C.S., Hirsch, A., Fung, C., Rosen, O.M., 1978. Development of hormone receptors and hormone responsiveness in vitro. Insulin receptors and insulin sensitivity in the preadipocyte and adipocyte forms of 3T3-L1 cells. J Biol. Chem. 253, 7570–7578. Shuttleworth, C.A., 1997. Type VIII collagen. Int. J. Biochem. Cell Biol. 29, 1145–1148. Smas, C.M., Chen, L., Sul, H.S., 1997. Cleavage of membrane-associated pref-1 generates a soluble inhibitor of adipocyte differentiation. Mol. Cell. Biol. 17, 977–988. Spiegelman, B.M., Frank, M., Green, H., 1983. Molecular cloning of mRNA from 3T3 adipocytes. Regulation of mRNA content for glycerophosphte dehydrogenase and other differentiation-dependent proteins during adipocyte development. J. Biol. Chem. 258, 10083–10089. Sun, R., Chen, X., Yang, V.W., 2001. Intestinal-enriched Kruppel-like factor (Kruppel-like factor 5) is a positive regulator of cellular proliferation. J. Biol. Chem. 276, 6897–6900. Tontonoz, P.E., Hu, E., Graves, R.A., Budavari, A.I., Spiegelman, B.M., 1994. mPPAR gamma 2: tissue-specific regulator of an adipocyte enhancer. Genes Dev. 8, 1224–1234.