- Email: [email protected]
Cholesteryl ester transfer protein gene expression during differentiation of human preadipocytes to adipocytes in primary culture
Atherosclerosis 142 (1999) 301 – 307
Cholesteryl ester transfer protein gene expression during differentiation of human preadipocytes to adipocytes in primary culture Benoit Gauthier, Malcolm Robb, Ruth McPherson * Lipoprotein and Atherosclerosis Group, Uni6ersity of Ottawa Heart Institute, Ottawa, K1Y 4E9 Canada Received 26 March 1998; received in revised form 7 July 1998; accepted 14 August 1998
Abstract The expression pattern of the CETP gene in relationship to that of LPL, adipsin, PPARg, C/EBPa, ADD1/SREBP1 and actin was examined by RT-PCR during differentiation of human fibroblastic preadipocytes to adipocytes in primary culture. Preadipocytes were isolated from subcutaneous fat obtained from healthy female subjects undergoing mammary reduction procedures, and induced to differentiate in culture. Morphologically, adipogenesis was confirmed by the accumulation of lipid droplets in cells. We show that the gene encoding CETP is expressed in preadipocytes and is present throughout differentiation as compared to LPL and adipsin which were detected in the majority of samples by day 2 or 3 of adipogenesis. The transcription factors, PPARg, ADD1/SREBP1 and C/EBPa were expressed by day 2, concomitant with the appearance of LPL and adipsin but subsequent to the appearance of CETP. CETP mRNA was not detectable in human skin fibroblasts. These studies demonstrate that CETP expression is induced at an early stage of commitment to the adipocyte lineage and may be activated by transcription factor(s), which are not members of the PPAR, ADD1/SREBP1 or C/EBP families. © 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Cholesterol; Pre-adipocyte; Human adipocyte differentiation; CETP; LPL; Adipsin; C/EBPa; PPARg; ADD1
1. Introduction Human adipose tissue stores a significant amount of total body cholesterol (greater than 25%) where, due to high cholesteryl esterase activity, it is stored as free cholesterol. Adipocytes are dependent on lipoproteins as a source of cholesterol since cholesterol synthesis is limited in these cells. Our laboratory has recently demonstrated that cholesteryl ester transfer protein (CETP), which has an established role in reverse cholesAbbre6iations: CETP, cholesteryl ester transfer protein; LPL, lipoprotein lipase; C/EBPa, CAATT enhancer binding protein alpha; PPARg, peroxisome proliferator-activated receptor gamma; ADD1, adipocyte determination and differentiation-dependent factor 1; HDL, high density lipoproteins; RT-PCR, reverse transcriptase polymerase chain reaction. * Corresponding author. Present address: Lab H453, 1053 Carling Ave, Ottawa, Canada K1Y 4E9. Tel. +1-613-761-5256; fax + 1-613761-5281; e-mail: [email protected].
terol transport, also promotes the selective uptake of HDL-derived cholesteryl esters by human adipocytes [1]. These findings indicate that CETP may play an important role in human adipocyte cholesterol homeostasis. Thus, molecular mechanisms governing CETP gene expression in adipocytes are of major interest. The human wildtype CETP (wtCETP) gene encompasses 16 exons [2]. A variant CETP cDNA has been described which lacks exon 9 (D9CETP) [3]. Significantly, D9CETP lacks neutral lipid transfer activity in the plasma compartment and appears to form an intracellular heterodimer with wtCETP, impairing its secretion [4]. In transgenic mice bearing the human wtCETP gene with its natural flanking sequences (NFR), hypercholesterolemia achieved by feeding a high cholesterol diet or crossing into an apoE knockout background resulted in a disproportionate increase in wtCETP mRNA [5].
0021-9150/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 2 1 - 9 1 5 0 ( 9 8 ) 0 0 2 4 5 - 7
302
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
Although CETP is highly expressed in human adipocytes [6] and is upregulated by cholesterol [7,8], few data are available on factors involved in its activation during adipogenesis [9]. We were therefore interested in characterizing the pattern of CETP gene expression during human adipogenesis with regard to the appearance of the major transcription factors known to be involved in the expression of a number of adipocyte specific genes. Our understanding of adipocyte differentiation is derived largely from studies with murine preadipose cell lines in culture, notably the C3H10T1/2 and NIH 3T3 fibroblastic cell lines and 3T3-L1 and 3T3-F442A preadipocytes [10,11]. The conversion of fibroblast-like preadipocytes into adipose cells is a process that continues throughout life and which can be replicated in cell culture. Much recent work has involved the use of preadipocyte cell lines that have already undergone commitment to the adipocyte lineage. The most extensively studied and characterized are mouse 3T3-L1 and 3T3-F442A cell lines. Although these immortalized preadipose cell lines have been very useful for the study of adipose development, they are aneuploid and often possess characteristics that differ from those of tissue preadipocytes. For example, differentiated 3T3 cells accumulate lipid poorly and production of leptin, adipsin and certain adipocyte-specific proteins is relatively impaired. For these reasons, we have chosen to prepare adipose tissue subfractions from fresh human tissue. We have successfully isolated and cultured primary human preadipocytes derived from subcutaneous fat obtained from healthy female subjects undergoing mammary reduction procedures. Using these cells as a model system for human adipogenesis we have studied the expression pattern during differentiation of genes encoding adipsin, lipoprotein lipase (LPL), C/EBPa, PPARg, ADD1/SREBP1, CETP and actin.
2. Methods
2.1. Subjects and sample collection Subcutaneous adipose tissue was obtained from four healthy female subjects (20 – 27 years old) undergoing reduction mammoplasty for cosmetic purposes. Subjects were all in good health, non-diabetic, not on lipid-active medication and demonstrated variable degrees of adiposity (body mass index 23 – 31). The study was approved by the Human Ethics Research Committee of the Ottawa Civic Hospital and written informed consent was obtained from all subjects. Tissue samples were dissected in the operating room after collection to obtain thin layers of subcutaneous fat tissue without any adhering breast tissue. Samples were collected in
DMEM/Ham F12 medium (Life Technologies Inc, Burlington, ON) and immediately transported to the laboratory for preadipocyte isolation.
2.2. Isolation of preadipocytes and cell culturing Fibroblastic preadipocytes were prepared by collagenase digestion of adipose tissue using a modified protocol of Radeau et al. [9,12]. The adipose tissue (80 g) was carefully dissected in phosphate buffer saline (PBS) to remove any adhering tissue and capillaries. Tissue (10 g) was then transferred to 25 ml of fresh DMEM/Ham F12 media containing 62 mg/l penicillin (Sigma, Oakville, ON), 50 mg/l streptomycin (Sigma), 2 mM L-glutamine (Life Technologies, Burlington, ON), 16 mM pantothenic acid (Sigma), 32 mM biotin (Sigma), 0.6 mg/ml collagenase A (Boehringer Mannheim, Montreal, PQ) and 6 mg/ml BSA (ICN, CA). Following incubation for 90 min at 37°C in a shaking incubator, 10% FCS was added to the digest and the mixture was filtered through a sterile 250 mm nylon mesh (T3; Tissage Tissus Techniques, Villeneuve-la-Garenne, France) to remove large fragments of undigested tissue before centrifugation at 1000 rpm for 10 min. The top layer, consisting of large mature adipocytes, was removed and the pellet containing the stromal vascular fraction (small mature fat cells and preadipocytes) was resuspended in 10 ml of fresh media. To further purify the stromal vascular fraction from cellular debris and contaminating red blood cells, the mixture was consecutively filtered through 100 and 40 mm nylon meshes. A sample was kept after the 40 mm mesh and labeled as the stromal vascular fraction (SVF). The remaining fraction was then filtered through a 25 mm mesh to separate the small mature adipocytes from preadipocytes. After centrifugation for 10 min at 1000 rpm, the pelleted preadipocytes were resuspended in 20 ml of media and divided equally among thirteen 60 mm dishes. The following day, four dishes of preadipocytes were set aside for protein and RNA analyses, while the nine remaining plates were replenished with differentiating medium (DMEM/Ham F12 supplemented with 5 mg/ml insulin (Sigma), 10 mg/ml transferrin (Sigma), 0.2 mM cPGI2 (Caymen Chemical), 1 mM hydrocortisone (Sigma) and 0.2 nM triiodothyronine (Sigma). Differentiating adipocytes were studied at 2, 3 and 7 days post preadipocyte isolation.
2.3. RNA isolation and re6erse transcriptase (RT) -PCR Total RNA was isolated from each sample using the acidic guanidinium isocyanate method as described by Chomczynski and Sacchi [13] or by using the Trireagent (Molecular Research Center, Inc, Cincinnati,
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
303
Table 1 Sequence of primers used for RT-PCR Gene
Primer sequences
Annealing temperature (°C)
Position of amplified fragment in cDNA
Adipsin
5%-ACC ACG ACG GCG CCA TCA CC-3% 5%-CCA CCT CCA TGC CCG CCA AT-3% 5%-TCA ATC ACA GCA GCA AAA CC-3% 5%-CCA CAT CTC CAA GTC CTC TC-3% 5%-GGA CAG ATC TGC AAA GAG ATC A-3% 5%-GAT TTC CTG GTT GGT GTT GAA G-3% 5%-GCC CCT CCA TCG TCC ACC GC-3% 5%-GGG CAC GAA GGC TCA TCA TT-3% 5%-CTG AGT AGG GGG AGC AAA TC-3% 5%-AAC CAA AAG CAA AGG GAG TC-3% 5%-ACC ACT CCC ACT CCT TTG ATA-3% 5%-GTC TGT CTC CGT CTT CTT GAT-3% 5-GGA GGG GTA GGG CCA ACG GCC T-3% 5%-CAT GTC TTC GAA AGT GCA ATC C-3%
66
530–955
425
59
380–936
556
59
722–1102
380
59
1139–1632
493
59
1989–2780
791
63
252–1541
1289
63
1–80
LPL CETP
b-Actin C/EBPa PPARg ADD1
OH) as described by the manufacturer. Reverse transcription of 2 mg total RNA was carried out in 20 ml reaction volume containing 0.1 pM oligodT, 1× first strand buffer (50 mM Tris – HCl (pH 8.3), 75 mM KCl and 3 mM MgCl2), 10 mM DTT, 0.5 mM dNTP and 10 U Moloney murine leukemia virus reverse transcriptase (Life Technologies, Burlington, ON). Reaction mixes were incubated for 2 h at 37°C, heated at 95°C for 3 min, and then chilled on ice. RNase A (35 mg/ml) was then added to each reaction. PCR was carried out in 25 ml reaction volumes containing 1×Vent buffer (NEB, Mississauga, ON), 5 ml of the reverse transcription mixture, 10 pmoles of primers as depicted in Table 1, 0.5 mM dNTPs and 1 U Vent DNA polymerase (NEB). The PCR conditions were 95°C for 5 min followed by 30 cycles at 95°C for 1 min, the annealing temperature for each individual gene as outline in Table 1 for 1 min, and 72°C for 1 min in a Stratagene Robocycler® 96 temperature cycler (LaJolla, CA). Due to relatively high levels of actin, this gene was amplified using only 20 cycles. The total number of cycles for each PCR reaction was chosen to remain within the exponential phase of the reaction and to avoid exceeding the limit of comparable intensities of the staining procedure. PCR reactions were performed in triplicate. PCR products (10 ml) were resolved by agarose (1.5%) gel electrophoresis and visualized by ethidium bromide staining. The expected size of each amplified fragment which corresponded to those obtained by gel electrophoresis are given in Table 1. The authenticity of each PCR fragment was confirmed by Southern blot analysis using the respective cDNA. One representative gel for each of four subjects is illustrated in Fig. 2.
PCR product (bp)
80
2.4. Protein extraction and western blot analysis Subsequent to RNA isolation using the Tri-reagent, the remaining organic phase of each time point was treated with ethanol and isopropanol to precipitate cellular proteins. Protein pellets were washed three times in a solution containing 0.3 M guanidine hydrochloride in 95% ethanol followed by a final wash in 100% ethanol. Protein extracts were dried and disolved in 1% SDS. Approximately 20 mg of protein was resolved on a 10% SDS-polyacrylamide gel before being transferred onto nitrocellulose for western blot analysis. CETP was detected using the monoclonal antibody TP2 (supplied by Dr R. Milne) and chemiluminescence as directed by the manufacturer (BMC, Germany).
3. Results Fibroblastic preadipocytes were harvested from subcutaneous fat adipose tissue of young female subjects undergoing mammary reduction procedures and induced to differentiate into mature adipocytes. Morphological changes accompanying differentiation were observed under light microscopy (Fig. 1). Overnight samples consisted of small preadipocytes, which were morphologically similar to fibroblasts, and scattered red blood cells (Fig. 1A and B). The addition of differentiating media to the cultured cells resulted, within 7 days, in the accumulation of lipids in these fibroblast-like cells, consistent with the overall appearance of adipocytes (Fig. 1C). The pattern of expression of LPL, adipsin, C/EBPa, PPARg, ADD1/SREBP1 and actin in relation to the appearance of CETP was then analyzed by RT-PCR (Fig. 2).
304
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
In general, LPL and adipsin, which are considered relatively early markers of differentiation, were not observed in preadipocytes (with the exception of adipsin in patient D and LPL in patient A) but were readily detectable by day 2 and fully present by day 7 as compared to mature adipocytes (compare lane 1 and
Fig. 1. Morphological appearance of (A) preadipocytes (B) cells 2 days after induction of differentiation and (C) 7 days after induction of differentiation as observed under light microscopy. Lipid droplets are indicated by the arrows.
6 of LPL and adipsin panels, Fig. 2). These were also detected in the SVF, which contains in addition to preadipocytes, small mature adipocytes (lane 2 LPL and adipsin panels, Fig. 2). The actin mRNA was also independently amplified to verify the integrity of RNA samples from the different patients. As shown in Fig. 2 (actin panel), this mRNA was detected in all samples with relatively little changes in levels throughout differentiation. In contrast to LPL and adipsin, high levels of CETP mRNA were observed in preadipocytes of all patients. These levels appeared to remain relatively constant throughout differentiation. Interestingly in patients B and C, levels of the D9 CETP mRNA, the alternatively spliced variant of the CETP mRNA in which exon 9 has been removed, were higher in the SVF as compared to mature adipocytes while levels of the wild type form remained constant (compare lanes 1 and 2 of CETP panel, Fig. 2). The D9 CETP could not be detected in any of the samples analyzed from patient A. Consistent with previous studies [9], we were unable to demonstrate the presence of CETP mRNA in human fibroblasts by RT-PCR (data not shown). Thus, CETP gene expression is evident in committed preadipocytes, but not in fibroblasts. To determine whether or not the protein encoded by the CETP gene was also present at an early stage of adipocyte commitment, RNA and protein extracts were simultaneously isolated from cells at each time point during differentiation and the presence of CETP was detected by RT-PCR and Western blot analysis. As in differentating adipocytes from patients studied previously, RT-PCR revealed the presence of CETP mRNA in human preadipocytes and throughout differentiation (Fig. 3A). Western blot analysis also demonstrated that CETP was present at the preadipocyte stage indicating that CETP is synthesized by human preadipocytes (Fig. 3B). It is therefore likely that CETP is induced early during development between the adipoblast (preadipocyte precursor) and preadipocyte stages. Interestingly CETP mRNA levels for this patient remained relatively constant during differentiation while protein levels decreased markedly indicating that CETP expression may be regulated post-transcriptionally during adipogenesis. Adipocyte differentiation from committed preadipocytes in culture has been shown to be orchestrated by two interdependently acting groups of transcription factors; PPARg and C/EBPa [10,11,14,15]. We were therefore interested in determining the expression pattern of these two families during human adipogenesis in relation to CETP expression. Transcripts for both PPARg and C/EBPa were absent in preadipocytes and gradually appeared during the course of differentiation with levels, by day 7, similar to those found in mature adipocytes (PPARg and C/EBPa panels, Fig. 2). With the exception of LPL in patient A, induction of
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
305
Fig. 2. Expression pattern of genes encoding adipsin, LPL, PPARg, C/EBPa, CETP and actin during human adipogenesis. Preadipocytes were isolated from four healthy subjects (A, B, C and D), cultured and induced to differentiate. RNA was isolated from adipocytes (lane 1), the stromal vascular fraction (lane 2), preadipocytes (lane 3), day 2 (lane 4), day 3 (lane 5) and day 7 (lane 6) of differentiation and analyzed by RT-PCR for the presence of specific mRNAs. One additional time point, day 5 (lane 5a) was included in patient D.
PPARg and C/EBPa coincided with the appearance of either LPL or adipsin supporting a possible regulatory function for these transcription factors in LPL and adipsin gene transcription during differentiation. However, these markers appeared later than CETP mRNA
indicating that neither PPARg nor C/EBPa is a prerequisite for the initial induction of CETP. Recently, ADD1 (adipocyte determination and differentation-dependent factor-1), a member of the SREBP family of transcription factors (equivalent to the SREBP-1c isoform of SREBP-1) was shown to be expressed early during adipogenesis and could therefore be a potential candidate involved in the initial induction of CETP expression [16]. However, RT-PCR revealed that ADD1/SREBP1 mRNA was not detectable until day 5 of differentiation excluding the possibility that this factor is involved in the initial induction of CETP gene expression during adipogenesis (Fig. 2, patient D).
4. Discussion
Fig. 3. Expression pattern of CETP during differentiation. (A) RTPCR of CETP and actin mRNA (B) Protein levels of CETP as detected by western blot analysis using TP2 antibodies. A, adipocytes; Pre, preadipocytes, 2, 3, 7 days after differentiation.
We have established a primary cell culture model using human preadipocytes isolated from subcutaneous fat, obtained at the time of reduction mammoplasty from healthy subjects, to study human adipogenesis. Differentiation of these cells, which are committed to adipocyte lineage, to adipocytes was confirmed by morphological and biochemical changes as determined by microscopy and RT-PCR of adipocyte specific mRNA for proteins involved in lipid storage. Adipsin and LPL, two markers of adipogenesis, were gradually induced over 7 days, consistent with the appearance of lipid droplets in cells. Small variation in the pattern of gene expression for LPL and adipsin amongst different patients indicates that differentiation is not induced with a
306
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
uniform time course. For example, in patients A and D, adipsin and LPL were present early during differentiation (preadipocyte samples) while in patients B and C, transcripts for these genes were detected only by day 3. Thus, even though differentiation was induced in culture, there was considerable heterogeneity in both the magnitude of mRNA induction and in the temporal appearance of specific markers during the differentiation process. Our results demonstrate that, in contrast to adipsin and LPL, CETP is expressed in preadipocytes with mRNA levels lower or very similar to those found in mature adipocytes (Fig. 2, CETP panel). With the exception of patient A, levels of CETP mRNA, as compared to actin mRNA, gradually decreased during the course of differentiation. This decrease in CETP mRNA abundance most likely results from the down regulation of the CETP gene due to the absence of lipoprotein cholesterol in the differentiating media [7,8]. The ratio of wild type to D9 CETP mRNA remained relatively constant during differentiation. However, we did observe higher levels of the D9 isoform in the stromal vascular fraction of patients B and C. This fraction is mainly composed of small lipid poor adipocytes, which we have previously demonstrated to express high levels of CETP mRNA as compared to mature adipocytes [9,12]. Yang and co-workers have shown using transgenic mice bearing the human CETP gene that the ratio of wild type to D9 CETP mRNA in the liver varied significantly dependent on the cholesterol content of the diet. A high cholesterol diet resulted in a 4.5-fold increase in mRNA levels of the wild type with no apparent changes in levels of the D9 CETP transcript [5]. In contrast, in human subjects, we reported that cholesterol feeding resulted in a significant increase in adipose tissue CETP mRNA abundance but no change in the relative proportion of wt and D9 CETP mRNA [17]. Similarly, cholesterol-induced variation in the ratio of the wild type to D9 CETP mRNA was not observed in the liposarcoma cell line SW872 that had been treated with either LDL or 25-OH cholesterol [18]. Thus, although CETP mRNA abundance increases in response to cholesterol, there does not appear to be a consistent effect of cellular cholesterol on the relative expression of the D9 transcript. We were also able to detect CETP throughout differentiation by Western blot analysis indicating that CETP mRNA is being translated efficiently. These findings are therefore consistent with our previous results in which we suggest that adipose cells express CETP at a very early stage of differentiation [9]. The expression patterns of PPARg and C/EBPa mRNA, two transcription factors involved in adipocyte differentiation, did not correlate with expression of CETP suggesting that these factors are not involved in the initial induction of CETP gene expression during
Fig. 4. Schematic representation of the expression pattern and relative levels of adipsin, LPL, PPARg, C/EBPa and CETP during preadipocyte differentiation as compared to mature adipocytes. This comparison is qualitative and not quantitative.
differentiation. However, the appearance of adipsin and LPL did coincide with that of PPARg and C/EBPa, consistent with previous reports indicating that adipsin and LPL are activated by these transcription factors [15,19]. Agellon and co-workers have shown that the transcription factor, C/EBPa, is involved in the constitutive hepatic expression of the CETP gene and that its low levels in proliferating or cultured cells may account for the reduced expression of CETP [20]. Our results suggest that, in human adipocytes, C/EBPa is not a prerequisite for CETP gene expression since CETP was found in cells not expressing this transcription factor. ADD1 is a third transcription factor involved in adipogenesis and the expression of ADD1 in differentiating mouse cell line precedes that of PPARg and C/EBPa [16]. However, the expression pattern of ADD1 mRNA during human preadipocyte differentiation did not correlate with that of CETP mRNA. These results suggest that CETP is induced at an early stage during commitment to the adipocyte by a yet uncharacterized transcription factor. The temporal sequence of expression of the major adipocyte genes during human adipogenesis is summarized in Fig. 4. Kim and Spiegelman have previously reported the induction pattern of adipsin, LPL, C/ EBPa, PPARg, ADD1/SREBP1 mRNAs during differentiation of the mouse preadipocyte cell lines 3T3-F442A and 3T3-L1 [16]. The 3T3-L1 cell line requires hormonal induction using dexamethasome, methylisobutylxanthine, insulin and serum to differentiate while the 3T3-F442A cell line differentiates spontaneously at confluency in the presence of insulin [21,22]. Comparison of the two mouse cell lines to the human
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
primary cell culture reveals differences in the expression pattern of several mRNAs. As opposed to the mouse model, in which ADD1 and PPARg are present in preadipocytes, these transcription factors appear later during human adipogenesis. Furthermore, in mouse adipocytic lines, LPL mRNA precedes the appearance of adipsin while in human, expression of both transcripts coincide. The expression pattern for C/EBPa appears to be identical in both human and mouse adipocytes. Since CETP is not expressed in murine cells, no correlation could be made with the human gene. These differences confirm the importance of a human primary cell model in the study of gene expression during adipogenesis Adipocytes do not synthesize cholesterol but maintain a relatively constant free cholesterol to triglyceride ratio. CETP plays an important role in human adipocyte cholesterol accumulation and expression of CETP very early during adipogenesis may reflect the need to accumulate cellular cholesterol [1] in preparation for triglyceride synthesis and lipid droplet formation.
Acknowledgements Supported by the Medical Research Council of Canada (group grant).
References [1] Benoist F, MacDonnell M, Lau P, Milne RW, McPherson R. Cholesteryl ester transfer protein (CETP) mediates the selective uptake of HDL-cholesteryl ester (CE) by human adipocytes. J Biol Chem 1997;272:23572–7. [2] Agellon LB, Quinet EM, Gillette TG, Drayna DT, Brown ML, Tall AR. Organization of the human cholesteryl ester transfer protein gene. Biochemistry 1990;29:1372–6. [3] Inazu A, Quinet EM, Wang S, Brown ML, Stevenson S, Barr ML, et al. Alternative splicing of the mRNA encoding the human cholesteryl ester transfer protein. Biochemistry 1992;31:2352 – 8. [4] Quinet E, Yang TP, Marinos C, Tall AR. Inhibition of the cellular secretion of cholesteryl ester transfer protein by a variant protein formed by alternative splicing of mRNA. J Biol Chem 1993;268:16891– 4. [5] Yang TP, Agellon LB, Walsh A, Breslow JL, Tall AR. Alternative splicing of the human cholesteryl ester transfer protein gene in transgenic mice-Exon exclusion modulates gene expression in response to dietary or developmental change. J Biol Chem 1996;271:12603– 9.
.
307
[6] Radeau T, Robb M, Lau P, McDonnell M, McPherson R. Age-related decline in adipose tissue cholesteryl ester transfer protein (CETP) mRNA abundance and plasma concentrations of CETP. Atherosclerosis 1998;139:369 – 76. [7] Oliveira HCF, Chouinard RA, Agellon LB, Bruce C, Ma LM, Walsh A, et al. Human cholesteryl ester transfer protein gene proximal promoter contains dietary cholesterol positive responsive elements and mediates expression in small intestine and periphery while predominant liver and spleen expression is controlled by 5%-distal sequences-cis-acting sequences mapped in transgenic mice. J Biol Chem 1996;271:31831– 8. [8] Gauthier BR, Robb M, McPherson R. Transcriptional up-regulation of CETP gene expression in human adipocytes by cholesterol is mediated by the interaction of SREBP-1a with a novel lipid response element (LRE) found in the CETP promoter. Circulation 1997;96:108. [9] Radeau T, Robb M, McDonnell M, McPherson R. Preferential expression of cholesteryl ester transfer protein mRNA by stromal-vascular cells of human adipose tissue. Biochim Biophys Acta 1998;1392:245 – 53. [10] Cornelius P, MacDougald OA, Lane MD. Regulation of adipocyte development. Ann Rev Nutr 1994;14:99 – 129. [11] MacDougald OA, Lane MD. Transcriptional regulation of gene expression during adipocyte differentiation. Ann Rev Biochem 1995;64:345 – 73. [12] Radeau T, Lau P, Robb M, McDonnell M, Ailhaud G, McPherson R. Cholesteryl ester transfer protein (CETP) mRNA abundance in human adipose tissue: Relationship to cell size and membrane cholesterol content. J Lipid Res 1995;36:2552–61. [13] Chomzynski P, Sacchi N. Single method of RNA isolation by guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987;162:156 – 9. [14] Vasseur-Cognet M, Lane MD. Trans-acting factors involved in adipogenic differentiation. Curr Opin Gen Dev 1993;3:238–45. [15] Schoonjans K, Peinado-Onsurbe J, Lefebvre AM, Heyman RA, Briggs MR, Deeb S, et al. Trans-acting factors involved in adipogenic differentiation. Biochim Biophys Acta 1996;1302:93– 109. [16] Kim JB, Spiegelman BM. ADD1/SREBP1 promotes adipocyte differentiation and gene expression linked to fatty acid metabolism. Genes Dev 1996;10:1096 – 107. [17] Martin L, Connelly P, Nancoo D, Wood N, Zhang Z, Quinet E, et al. Cholesteryl ester transfer protein and high density lipoprotein responses to cholesterol feeding in man: Relationship to apolipoprotein E genotype. J Lipid Res 1993;34:347 – 446. [18] Richardson MA, Berg DT, Johnston PA, McClure D, Grinnell BW. Human liposarcoma cell line, SW872, secretes cholesteryl ester transfer protein in response to cholesterol. J Lipid Res 1996;37:1162 – 6. [19] Wilkison WO, Min HY, Claffey KP, Satterberg BL, Spiegelman BM. Control of the adipsin gene in adipocyte differentiation. J Biol Chem 1990;265:477 – 82. [20] Agellon LB, Zhang P, Jiang XC, Mendelsohn L, Tall AR. The CCAAT/enhancer-binding protein trans-activates the human cholesteryl ester transfer protein gene promoter. J Biol Chem 1992;267:22336– 9. [21] Green H, Kehinde O. Sublines of mouse 3T3 cells that accumulate lipid. Cell 1974;1:113 – 6. [22] Green H, Kehinde O. Spontaneous heritable changes leading to increase adipose conversion in 3T3 cells. Cell 1976;7:105–15.
.
Cholesteryl ester transfer protein gene expression during differentiation of human preadipocytes to adipocytes in primary culture Benoit Gauthier, Malcolm Robb, Ruth McPherson * Lipoprotein and Atherosclerosis Group, Uni6ersity of Ottawa Heart Institute, Ottawa, K1Y 4E9 Canada Received 26 March 1998; received in revised form 7 July 1998; accepted 14 August 1998
Abstract The expression pattern of the CETP gene in relationship to that of LPL, adipsin, PPARg, C/EBPa, ADD1/SREBP1 and actin was examined by RT-PCR during differentiation of human fibroblastic preadipocytes to adipocytes in primary culture. Preadipocytes were isolated from subcutaneous fat obtained from healthy female subjects undergoing mammary reduction procedures, and induced to differentiate in culture. Morphologically, adipogenesis was confirmed by the accumulation of lipid droplets in cells. We show that the gene encoding CETP is expressed in preadipocytes and is present throughout differentiation as compared to LPL and adipsin which were detected in the majority of samples by day 2 or 3 of adipogenesis. The transcription factors, PPARg, ADD1/SREBP1 and C/EBPa were expressed by day 2, concomitant with the appearance of LPL and adipsin but subsequent to the appearance of CETP. CETP mRNA was not detectable in human skin fibroblasts. These studies demonstrate that CETP expression is induced at an early stage of commitment to the adipocyte lineage and may be activated by transcription factor(s), which are not members of the PPAR, ADD1/SREBP1 or C/EBP families. © 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Cholesterol; Pre-adipocyte; Human adipocyte differentiation; CETP; LPL; Adipsin; C/EBPa; PPARg; ADD1
1. Introduction Human adipose tissue stores a significant amount of total body cholesterol (greater than 25%) where, due to high cholesteryl esterase activity, it is stored as free cholesterol. Adipocytes are dependent on lipoproteins as a source of cholesterol since cholesterol synthesis is limited in these cells. Our laboratory has recently demonstrated that cholesteryl ester transfer protein (CETP), which has an established role in reverse cholesAbbre6iations: CETP, cholesteryl ester transfer protein; LPL, lipoprotein lipase; C/EBPa, CAATT enhancer binding protein alpha; PPARg, peroxisome proliferator-activated receptor gamma; ADD1, adipocyte determination and differentiation-dependent factor 1; HDL, high density lipoproteins; RT-PCR, reverse transcriptase polymerase chain reaction. * Corresponding author. Present address: Lab H453, 1053 Carling Ave, Ottawa, Canada K1Y 4E9. Tel. +1-613-761-5256; fax + 1-613761-5281; e-mail: [email protected].
terol transport, also promotes the selective uptake of HDL-derived cholesteryl esters by human adipocytes [1]. These findings indicate that CETP may play an important role in human adipocyte cholesterol homeostasis. Thus, molecular mechanisms governing CETP gene expression in adipocytes are of major interest. The human wildtype CETP (wtCETP) gene encompasses 16 exons [2]. A variant CETP cDNA has been described which lacks exon 9 (D9CETP) [3]. Significantly, D9CETP lacks neutral lipid transfer activity in the plasma compartment and appears to form an intracellular heterodimer with wtCETP, impairing its secretion [4]. In transgenic mice bearing the human wtCETP gene with its natural flanking sequences (NFR), hypercholesterolemia achieved by feeding a high cholesterol diet or crossing into an apoE knockout background resulted in a disproportionate increase in wtCETP mRNA [5].
0021-9150/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 2 1 - 9 1 5 0 ( 9 8 ) 0 0 2 4 5 - 7
302
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
Although CETP is highly expressed in human adipocytes [6] and is upregulated by cholesterol [7,8], few data are available on factors involved in its activation during adipogenesis [9]. We were therefore interested in characterizing the pattern of CETP gene expression during human adipogenesis with regard to the appearance of the major transcription factors known to be involved in the expression of a number of adipocyte specific genes. Our understanding of adipocyte differentiation is derived largely from studies with murine preadipose cell lines in culture, notably the C3H10T1/2 and NIH 3T3 fibroblastic cell lines and 3T3-L1 and 3T3-F442A preadipocytes [10,11]. The conversion of fibroblast-like preadipocytes into adipose cells is a process that continues throughout life and which can be replicated in cell culture. Much recent work has involved the use of preadipocyte cell lines that have already undergone commitment to the adipocyte lineage. The most extensively studied and characterized are mouse 3T3-L1 and 3T3-F442A cell lines. Although these immortalized preadipose cell lines have been very useful for the study of adipose development, they are aneuploid and often possess characteristics that differ from those of tissue preadipocytes. For example, differentiated 3T3 cells accumulate lipid poorly and production of leptin, adipsin and certain adipocyte-specific proteins is relatively impaired. For these reasons, we have chosen to prepare adipose tissue subfractions from fresh human tissue. We have successfully isolated and cultured primary human preadipocytes derived from subcutaneous fat obtained from healthy female subjects undergoing mammary reduction procedures. Using these cells as a model system for human adipogenesis we have studied the expression pattern during differentiation of genes encoding adipsin, lipoprotein lipase (LPL), C/EBPa, PPARg, ADD1/SREBP1, CETP and actin.
2. Methods
2.1. Subjects and sample collection Subcutaneous adipose tissue was obtained from four healthy female subjects (20 – 27 years old) undergoing reduction mammoplasty for cosmetic purposes. Subjects were all in good health, non-diabetic, not on lipid-active medication and demonstrated variable degrees of adiposity (body mass index 23 – 31). The study was approved by the Human Ethics Research Committee of the Ottawa Civic Hospital and written informed consent was obtained from all subjects. Tissue samples were dissected in the operating room after collection to obtain thin layers of subcutaneous fat tissue without any adhering breast tissue. Samples were collected in
DMEM/Ham F12 medium (Life Technologies Inc, Burlington, ON) and immediately transported to the laboratory for preadipocyte isolation.
2.2. Isolation of preadipocytes and cell culturing Fibroblastic preadipocytes were prepared by collagenase digestion of adipose tissue using a modified protocol of Radeau et al. [9,12]. The adipose tissue (80 g) was carefully dissected in phosphate buffer saline (PBS) to remove any adhering tissue and capillaries. Tissue (10 g) was then transferred to 25 ml of fresh DMEM/Ham F12 media containing 62 mg/l penicillin (Sigma, Oakville, ON), 50 mg/l streptomycin (Sigma), 2 mM L-glutamine (Life Technologies, Burlington, ON), 16 mM pantothenic acid (Sigma), 32 mM biotin (Sigma), 0.6 mg/ml collagenase A (Boehringer Mannheim, Montreal, PQ) and 6 mg/ml BSA (ICN, CA). Following incubation for 90 min at 37°C in a shaking incubator, 10% FCS was added to the digest and the mixture was filtered through a sterile 250 mm nylon mesh (T3; Tissage Tissus Techniques, Villeneuve-la-Garenne, France) to remove large fragments of undigested tissue before centrifugation at 1000 rpm for 10 min. The top layer, consisting of large mature adipocytes, was removed and the pellet containing the stromal vascular fraction (small mature fat cells and preadipocytes) was resuspended in 10 ml of fresh media. To further purify the stromal vascular fraction from cellular debris and contaminating red blood cells, the mixture was consecutively filtered through 100 and 40 mm nylon meshes. A sample was kept after the 40 mm mesh and labeled as the stromal vascular fraction (SVF). The remaining fraction was then filtered through a 25 mm mesh to separate the small mature adipocytes from preadipocytes. After centrifugation for 10 min at 1000 rpm, the pelleted preadipocytes were resuspended in 20 ml of media and divided equally among thirteen 60 mm dishes. The following day, four dishes of preadipocytes were set aside for protein and RNA analyses, while the nine remaining plates were replenished with differentiating medium (DMEM/Ham F12 supplemented with 5 mg/ml insulin (Sigma), 10 mg/ml transferrin (Sigma), 0.2 mM cPGI2 (Caymen Chemical), 1 mM hydrocortisone (Sigma) and 0.2 nM triiodothyronine (Sigma). Differentiating adipocytes were studied at 2, 3 and 7 days post preadipocyte isolation.
2.3. RNA isolation and re6erse transcriptase (RT) -PCR Total RNA was isolated from each sample using the acidic guanidinium isocyanate method as described by Chomczynski and Sacchi [13] or by using the Trireagent (Molecular Research Center, Inc, Cincinnati,
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
303
Table 1 Sequence of primers used for RT-PCR Gene
Primer sequences
Annealing temperature (°C)
Position of amplified fragment in cDNA
Adipsin
5%-ACC ACG ACG GCG CCA TCA CC-3% 5%-CCA CCT CCA TGC CCG CCA AT-3% 5%-TCA ATC ACA GCA GCA AAA CC-3% 5%-CCA CAT CTC CAA GTC CTC TC-3% 5%-GGA CAG ATC TGC AAA GAG ATC A-3% 5%-GAT TTC CTG GTT GGT GTT GAA G-3% 5%-GCC CCT CCA TCG TCC ACC GC-3% 5%-GGG CAC GAA GGC TCA TCA TT-3% 5%-CTG AGT AGG GGG AGC AAA TC-3% 5%-AAC CAA AAG CAA AGG GAG TC-3% 5%-ACC ACT CCC ACT CCT TTG ATA-3% 5%-GTC TGT CTC CGT CTT CTT GAT-3% 5-GGA GGG GTA GGG CCA ACG GCC T-3% 5%-CAT GTC TTC GAA AGT GCA ATC C-3%
66
530–955
425
59
380–936
556
59
722–1102
380
59
1139–1632
493
59
1989–2780
791
63
252–1541
1289
63
1–80
LPL CETP
b-Actin C/EBPa PPARg ADD1
OH) as described by the manufacturer. Reverse transcription of 2 mg total RNA was carried out in 20 ml reaction volume containing 0.1 pM oligodT, 1× first strand buffer (50 mM Tris – HCl (pH 8.3), 75 mM KCl and 3 mM MgCl2), 10 mM DTT, 0.5 mM dNTP and 10 U Moloney murine leukemia virus reverse transcriptase (Life Technologies, Burlington, ON). Reaction mixes were incubated for 2 h at 37°C, heated at 95°C for 3 min, and then chilled on ice. RNase A (35 mg/ml) was then added to each reaction. PCR was carried out in 25 ml reaction volumes containing 1×Vent buffer (NEB, Mississauga, ON), 5 ml of the reverse transcription mixture, 10 pmoles of primers as depicted in Table 1, 0.5 mM dNTPs and 1 U Vent DNA polymerase (NEB). The PCR conditions were 95°C for 5 min followed by 30 cycles at 95°C for 1 min, the annealing temperature for each individual gene as outline in Table 1 for 1 min, and 72°C for 1 min in a Stratagene Robocycler® 96 temperature cycler (LaJolla, CA). Due to relatively high levels of actin, this gene was amplified using only 20 cycles. The total number of cycles for each PCR reaction was chosen to remain within the exponential phase of the reaction and to avoid exceeding the limit of comparable intensities of the staining procedure. PCR reactions were performed in triplicate. PCR products (10 ml) were resolved by agarose (1.5%) gel electrophoresis and visualized by ethidium bromide staining. The expected size of each amplified fragment which corresponded to those obtained by gel electrophoresis are given in Table 1. The authenticity of each PCR fragment was confirmed by Southern blot analysis using the respective cDNA. One representative gel for each of four subjects is illustrated in Fig. 2.
PCR product (bp)
80
2.4. Protein extraction and western blot analysis Subsequent to RNA isolation using the Tri-reagent, the remaining organic phase of each time point was treated with ethanol and isopropanol to precipitate cellular proteins. Protein pellets were washed three times in a solution containing 0.3 M guanidine hydrochloride in 95% ethanol followed by a final wash in 100% ethanol. Protein extracts were dried and disolved in 1% SDS. Approximately 20 mg of protein was resolved on a 10% SDS-polyacrylamide gel before being transferred onto nitrocellulose for western blot analysis. CETP was detected using the monoclonal antibody TP2 (supplied by Dr R. Milne) and chemiluminescence as directed by the manufacturer (BMC, Germany).
3. Results Fibroblastic preadipocytes were harvested from subcutaneous fat adipose tissue of young female subjects undergoing mammary reduction procedures and induced to differentiate into mature adipocytes. Morphological changes accompanying differentiation were observed under light microscopy (Fig. 1). Overnight samples consisted of small preadipocytes, which were morphologically similar to fibroblasts, and scattered red blood cells (Fig. 1A and B). The addition of differentiating media to the cultured cells resulted, within 7 days, in the accumulation of lipids in these fibroblast-like cells, consistent with the overall appearance of adipocytes (Fig. 1C). The pattern of expression of LPL, adipsin, C/EBPa, PPARg, ADD1/SREBP1 and actin in relation to the appearance of CETP was then analyzed by RT-PCR (Fig. 2).
304
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
In general, LPL and adipsin, which are considered relatively early markers of differentiation, were not observed in preadipocytes (with the exception of adipsin in patient D and LPL in patient A) but were readily detectable by day 2 and fully present by day 7 as compared to mature adipocytes (compare lane 1 and
Fig. 1. Morphological appearance of (A) preadipocytes (B) cells 2 days after induction of differentiation and (C) 7 days after induction of differentiation as observed under light microscopy. Lipid droplets are indicated by the arrows.
6 of LPL and adipsin panels, Fig. 2). These were also detected in the SVF, which contains in addition to preadipocytes, small mature adipocytes (lane 2 LPL and adipsin panels, Fig. 2). The actin mRNA was also independently amplified to verify the integrity of RNA samples from the different patients. As shown in Fig. 2 (actin panel), this mRNA was detected in all samples with relatively little changes in levels throughout differentiation. In contrast to LPL and adipsin, high levels of CETP mRNA were observed in preadipocytes of all patients. These levels appeared to remain relatively constant throughout differentiation. Interestingly in patients B and C, levels of the D9 CETP mRNA, the alternatively spliced variant of the CETP mRNA in which exon 9 has been removed, were higher in the SVF as compared to mature adipocytes while levels of the wild type form remained constant (compare lanes 1 and 2 of CETP panel, Fig. 2). The D9 CETP could not be detected in any of the samples analyzed from patient A. Consistent with previous studies [9], we were unable to demonstrate the presence of CETP mRNA in human fibroblasts by RT-PCR (data not shown). Thus, CETP gene expression is evident in committed preadipocytes, but not in fibroblasts. To determine whether or not the protein encoded by the CETP gene was also present at an early stage of adipocyte commitment, RNA and protein extracts were simultaneously isolated from cells at each time point during differentiation and the presence of CETP was detected by RT-PCR and Western blot analysis. As in differentating adipocytes from patients studied previously, RT-PCR revealed the presence of CETP mRNA in human preadipocytes and throughout differentiation (Fig. 3A). Western blot analysis also demonstrated that CETP was present at the preadipocyte stage indicating that CETP is synthesized by human preadipocytes (Fig. 3B). It is therefore likely that CETP is induced early during development between the adipoblast (preadipocyte precursor) and preadipocyte stages. Interestingly CETP mRNA levels for this patient remained relatively constant during differentiation while protein levels decreased markedly indicating that CETP expression may be regulated post-transcriptionally during adipogenesis. Adipocyte differentiation from committed preadipocytes in culture has been shown to be orchestrated by two interdependently acting groups of transcription factors; PPARg and C/EBPa [10,11,14,15]. We were therefore interested in determining the expression pattern of these two families during human adipogenesis in relation to CETP expression. Transcripts for both PPARg and C/EBPa were absent in preadipocytes and gradually appeared during the course of differentiation with levels, by day 7, similar to those found in mature adipocytes (PPARg and C/EBPa panels, Fig. 2). With the exception of LPL in patient A, induction of
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
305
Fig. 2. Expression pattern of genes encoding adipsin, LPL, PPARg, C/EBPa, CETP and actin during human adipogenesis. Preadipocytes were isolated from four healthy subjects (A, B, C and D), cultured and induced to differentiate. RNA was isolated from adipocytes (lane 1), the stromal vascular fraction (lane 2), preadipocytes (lane 3), day 2 (lane 4), day 3 (lane 5) and day 7 (lane 6) of differentiation and analyzed by RT-PCR for the presence of specific mRNAs. One additional time point, day 5 (lane 5a) was included in patient D.
PPARg and C/EBPa coincided with the appearance of either LPL or adipsin supporting a possible regulatory function for these transcription factors in LPL and adipsin gene transcription during differentiation. However, these markers appeared later than CETP mRNA
indicating that neither PPARg nor C/EBPa is a prerequisite for the initial induction of CETP. Recently, ADD1 (adipocyte determination and differentation-dependent factor-1), a member of the SREBP family of transcription factors (equivalent to the SREBP-1c isoform of SREBP-1) was shown to be expressed early during adipogenesis and could therefore be a potential candidate involved in the initial induction of CETP expression [16]. However, RT-PCR revealed that ADD1/SREBP1 mRNA was not detectable until day 5 of differentiation excluding the possibility that this factor is involved in the initial induction of CETP gene expression during adipogenesis (Fig. 2, patient D).
4. Discussion
Fig. 3. Expression pattern of CETP during differentiation. (A) RTPCR of CETP and actin mRNA (B) Protein levels of CETP as detected by western blot analysis using TP2 antibodies. A, adipocytes; Pre, preadipocytes, 2, 3, 7 days after differentiation.
We have established a primary cell culture model using human preadipocytes isolated from subcutaneous fat, obtained at the time of reduction mammoplasty from healthy subjects, to study human adipogenesis. Differentiation of these cells, which are committed to adipocyte lineage, to adipocytes was confirmed by morphological and biochemical changes as determined by microscopy and RT-PCR of adipocyte specific mRNA for proteins involved in lipid storage. Adipsin and LPL, two markers of adipogenesis, were gradually induced over 7 days, consistent with the appearance of lipid droplets in cells. Small variation in the pattern of gene expression for LPL and adipsin amongst different patients indicates that differentiation is not induced with a
306
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
uniform time course. For example, in patients A and D, adipsin and LPL were present early during differentiation (preadipocyte samples) while in patients B and C, transcripts for these genes were detected only by day 3. Thus, even though differentiation was induced in culture, there was considerable heterogeneity in both the magnitude of mRNA induction and in the temporal appearance of specific markers during the differentiation process. Our results demonstrate that, in contrast to adipsin and LPL, CETP is expressed in preadipocytes with mRNA levels lower or very similar to those found in mature adipocytes (Fig. 2, CETP panel). With the exception of patient A, levels of CETP mRNA, as compared to actin mRNA, gradually decreased during the course of differentiation. This decrease in CETP mRNA abundance most likely results from the down regulation of the CETP gene due to the absence of lipoprotein cholesterol in the differentiating media [7,8]. The ratio of wild type to D9 CETP mRNA remained relatively constant during differentiation. However, we did observe higher levels of the D9 isoform in the stromal vascular fraction of patients B and C. This fraction is mainly composed of small lipid poor adipocytes, which we have previously demonstrated to express high levels of CETP mRNA as compared to mature adipocytes [9,12]. Yang and co-workers have shown using transgenic mice bearing the human CETP gene that the ratio of wild type to D9 CETP mRNA in the liver varied significantly dependent on the cholesterol content of the diet. A high cholesterol diet resulted in a 4.5-fold increase in mRNA levels of the wild type with no apparent changes in levels of the D9 CETP transcript [5]. In contrast, in human subjects, we reported that cholesterol feeding resulted in a significant increase in adipose tissue CETP mRNA abundance but no change in the relative proportion of wt and D9 CETP mRNA [17]. Similarly, cholesterol-induced variation in the ratio of the wild type to D9 CETP mRNA was not observed in the liposarcoma cell line SW872 that had been treated with either LDL or 25-OH cholesterol [18]. Thus, although CETP mRNA abundance increases in response to cholesterol, there does not appear to be a consistent effect of cellular cholesterol on the relative expression of the D9 transcript. We were also able to detect CETP throughout differentiation by Western blot analysis indicating that CETP mRNA is being translated efficiently. These findings are therefore consistent with our previous results in which we suggest that adipose cells express CETP at a very early stage of differentiation [9]. The expression patterns of PPARg and C/EBPa mRNA, two transcription factors involved in adipocyte differentiation, did not correlate with expression of CETP suggesting that these factors are not involved in the initial induction of CETP gene expression during
Fig. 4. Schematic representation of the expression pattern and relative levels of adipsin, LPL, PPARg, C/EBPa and CETP during preadipocyte differentiation as compared to mature adipocytes. This comparison is qualitative and not quantitative.
differentiation. However, the appearance of adipsin and LPL did coincide with that of PPARg and C/EBPa, consistent with previous reports indicating that adipsin and LPL are activated by these transcription factors [15,19]. Agellon and co-workers have shown that the transcription factor, C/EBPa, is involved in the constitutive hepatic expression of the CETP gene and that its low levels in proliferating or cultured cells may account for the reduced expression of CETP [20]. Our results suggest that, in human adipocytes, C/EBPa is not a prerequisite for CETP gene expression since CETP was found in cells not expressing this transcription factor. ADD1 is a third transcription factor involved in adipogenesis and the expression of ADD1 in differentiating mouse cell line precedes that of PPARg and C/EBPa [16]. However, the expression pattern of ADD1 mRNA during human preadipocyte differentiation did not correlate with that of CETP mRNA. These results suggest that CETP is induced at an early stage during commitment to the adipocyte by a yet uncharacterized transcription factor. The temporal sequence of expression of the major adipocyte genes during human adipogenesis is summarized in Fig. 4. Kim and Spiegelman have previously reported the induction pattern of adipsin, LPL, C/ EBPa, PPARg, ADD1/SREBP1 mRNAs during differentiation of the mouse preadipocyte cell lines 3T3-F442A and 3T3-L1 [16]. The 3T3-L1 cell line requires hormonal induction using dexamethasome, methylisobutylxanthine, insulin and serum to differentiate while the 3T3-F442A cell line differentiates spontaneously at confluency in the presence of insulin [21,22]. Comparison of the two mouse cell lines to the human
B. Gauthier et al. / Atherosclerosis 142 (1999) 301–307
primary cell culture reveals differences in the expression pattern of several mRNAs. As opposed to the mouse model, in which ADD1 and PPARg are present in preadipocytes, these transcription factors appear later during human adipogenesis. Furthermore, in mouse adipocytic lines, LPL mRNA precedes the appearance of adipsin while in human, expression of both transcripts coincide. The expression pattern for C/EBPa appears to be identical in both human and mouse adipocytes. Since CETP is not expressed in murine cells, no correlation could be made with the human gene. These differences confirm the importance of a human primary cell model in the study of gene expression during adipogenesis Adipocytes do not synthesize cholesterol but maintain a relatively constant free cholesterol to triglyceride ratio. CETP plays an important role in human adipocyte cholesterol accumulation and expression of CETP very early during adipogenesis may reflect the need to accumulate cellular cholesterol [1] in preparation for triglyceride synthesis and lipid droplet formation.
Acknowledgements Supported by the Medical Research Council of Canada (group grant).
References [1] Benoist F, MacDonnell M, Lau P, Milne RW, McPherson R. Cholesteryl ester transfer protein (CETP) mediates the selective uptake of HDL-cholesteryl ester (CE) by human adipocytes. J Biol Chem 1997;272:23572–7. [2] Agellon LB, Quinet EM, Gillette TG, Drayna DT, Brown ML, Tall AR. Organization of the human cholesteryl ester transfer protein gene. Biochemistry 1990;29:1372–6. [3] Inazu A, Quinet EM, Wang S, Brown ML, Stevenson S, Barr ML, et al. Alternative splicing of the mRNA encoding the human cholesteryl ester transfer protein. Biochemistry 1992;31:2352 – 8. [4] Quinet E, Yang TP, Marinos C, Tall AR. Inhibition of the cellular secretion of cholesteryl ester transfer protein by a variant protein formed by alternative splicing of mRNA. J Biol Chem 1993;268:16891– 4. [5] Yang TP, Agellon LB, Walsh A, Breslow JL, Tall AR. Alternative splicing of the human cholesteryl ester transfer protein gene in transgenic mice-Exon exclusion modulates gene expression in response to dietary or developmental change. J Biol Chem 1996;271:12603– 9.
.
307
[6] Radeau T, Robb M, Lau P, McDonnell M, McPherson R. Age-related decline in adipose tissue cholesteryl ester transfer protein (CETP) mRNA abundance and plasma concentrations of CETP. Atherosclerosis 1998;139:369 – 76. [7] Oliveira HCF, Chouinard RA, Agellon LB, Bruce C, Ma LM, Walsh A, et al. Human cholesteryl ester transfer protein gene proximal promoter contains dietary cholesterol positive responsive elements and mediates expression in small intestine and periphery while predominant liver and spleen expression is controlled by 5%-distal sequences-cis-acting sequences mapped in transgenic mice. J Biol Chem 1996;271:31831– 8. [8] Gauthier BR, Robb M, McPherson R. Transcriptional up-regulation of CETP gene expression in human adipocytes by cholesterol is mediated by the interaction of SREBP-1a with a novel lipid response element (LRE) found in the CETP promoter. Circulation 1997;96:108. [9] Radeau T, Robb M, McDonnell M, McPherson R. Preferential expression of cholesteryl ester transfer protein mRNA by stromal-vascular cells of human adipose tissue. Biochim Biophys Acta 1998;1392:245 – 53. [10] Cornelius P, MacDougald OA, Lane MD. Regulation of adipocyte development. Ann Rev Nutr 1994;14:99 – 129. [11] MacDougald OA, Lane MD. Transcriptional regulation of gene expression during adipocyte differentiation. Ann Rev Biochem 1995;64:345 – 73. [12] Radeau T, Lau P, Robb M, McDonnell M, Ailhaud G, McPherson R. Cholesteryl ester transfer protein (CETP) mRNA abundance in human adipose tissue: Relationship to cell size and membrane cholesterol content. J Lipid Res 1995;36:2552–61. [13] Chomzynski P, Sacchi N. Single method of RNA isolation by guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987;162:156 – 9. [14] Vasseur-Cognet M, Lane MD. Trans-acting factors involved in adipogenic differentiation. Curr Opin Gen Dev 1993;3:238–45. [15] Schoonjans K, Peinado-Onsurbe J, Lefebvre AM, Heyman RA, Briggs MR, Deeb S, et al. Trans-acting factors involved in adipogenic differentiation. Biochim Biophys Acta 1996;1302:93– 109. [16] Kim JB, Spiegelman BM. ADD1/SREBP1 promotes adipocyte differentiation and gene expression linked to fatty acid metabolism. Genes Dev 1996;10:1096 – 107. [17] Martin L, Connelly P, Nancoo D, Wood N, Zhang Z, Quinet E, et al. Cholesteryl ester transfer protein and high density lipoprotein responses to cholesterol feeding in man: Relationship to apolipoprotein E genotype. J Lipid Res 1993;34:347 – 446. [18] Richardson MA, Berg DT, Johnston PA, McClure D, Grinnell BW. Human liposarcoma cell line, SW872, secretes cholesteryl ester transfer protein in response to cholesterol. J Lipid Res 1996;37:1162 – 6. [19] Wilkison WO, Min HY, Claffey KP, Satterberg BL, Spiegelman BM. Control of the adipsin gene in adipocyte differentiation. J Biol Chem 1990;265:477 – 82. [20] Agellon LB, Zhang P, Jiang XC, Mendelsohn L, Tall AR. The CCAAT/enhancer-binding protein trans-activates the human cholesteryl ester transfer protein gene promoter. J Biol Chem 1992;267:22336– 9. [21] Green H, Kehinde O. Sublines of mouse 3T3 cells that accumulate lipid. Cell 1974;1:113 – 6. [22] Green H, Kehinde O. Spontaneous heritable changes leading to increase adipose conversion in 3T3 cells. Cell 1976;7:105–15.
.