Biochemical Journal

Research article

CREBL2, interacting with CREB, induces adipogenesis in 3T3-L1 adipocytes

Xi Ma , Heyu Zhang , Lan Yuan , Hao Jing , Phil Thacker , Defa Li

Abstract

The factors that influence preadipocyte determination remain poorly understood. In the present paper, we report that CREBL2 [CREB (cAMP-response-element-binding protein)-like 2], a novel bZIP_1 protein, is up-regulated during MDI-induced preadipocyte differentiation. During both overexpression and under physiological conditions, CREBL2 interacted and was entirely co-localized with CREB. Overexpression of CREBL2 was sufficient to promote adipogenesis via up-regulating the expression of PPARγ (peroxisome-proliferator-activated receptor γ) and C/EBPα (CCAAT/enhancer-binding protein α) and accelerate lipogenesis accompanied with increased GLUT (glucose transporter) 1 and GLUT4. CREBL2 knockdown restrained adipogenic conversion and lipogenesis. Additionally, depletion of CREB could completely block the effects of overexpressed CREBL2, whereas an increase in CREB could not drive adipogenesis in the absence of CREBL2, indicating that the roles for CREBL2 on adipogenesis were CREB-dependent. Furthermore, siCREBL2 [siRNA (short interfering RNA) against CREBL2] could down-regulate CREB transcriptional activity and suppress CREB phosphorylation. CREB knockdown decreased the CREBL2 protein levels and vice versa. Collectively, the results of the present study indicate that CREBL2 plays a critical role in adipogenesis and lipogenesis via interaction with CREB.

  • adipogenesis
  • cAMP-response-element-binding protein (CREB)
  • cAMP-response-element-binding protein-like 2 (CREBL2)
  • lipogenesis
  • transcriptional activity

INTRODUCTION

One increasingly important effect of obesity is the generation of additional adipocytes in response to excess dietary energy intake and/or large increases in body fat composition [1]. It is well known that the generation of new adipocytes is controlled by a complex network of ‘adipocyte-specific’ transcription factors, cofactors and signalling molecules that regulate preadipocyte proliferation and adipogenesis [25]. However, the factors that influence preadipocyte determination remain poorly understood [6].

It is relatively well documented that during differentiation of 3T3-L1 preadipocytes, the cells first undergo cell growth and then they become quiescent and gradually acquire the morphological and biochemical characteristics of mature adipocytes [7,8]. The first step is characterized by marked changes in the pattern of C/EBP (CCAAT/enhancer-binding protein) β and C/EBPγ gene expression and the biomarkers of the second step are PPARγ (peroxisome-proliferator-activated receptor γ) and C/EBPα [9]. During lipogenesis of mature adipocytes, the α-oxidation pathway in adipocytes leads to a marked accumulation of lipids. There is a marked increase in the activity of peroxisomal fatty acid α-oxidation, which is reflected in the accumulation of odd chain length acyl moieties in major lipid species [10,11]. However, the underlying mechanisms for these changes and the functions of these pathways in adipocyte differentiation and lipid metabolism have not been fully explored.

CREBL2 [CREB (cAMP-response-element-binding protein)-like 2] was first identified in a search to find genes in a commonly deleted region on chromosome 12p13 flanked by ETV6 (Ets variant 6) and CDKN1B (cyclin-dependent kinase inhibitor 1B) genes [12]. The occurrence of CREBL2 deletion in malignancy suggests that CREBL2 may act as a tumour suppressor gene [13,14]. However, to our knowledge, no functional study has been performed on this hypothetical gene. The ORF (open reading frame) of CREBL2 encodes 123 amino acids with a predicted molecular mass of 14.0 kDa and a pI of 9.88. CREBL2 is highly conserved in human, mouse, rat, cow, chicken and zebrafish. Conserved domain analysis (http://www.ncbi.nlm.nih.gov/Structure/cdd/cddsrv.cgi?uid=154995) indicates that there is a predicted bZIP_1 (amino acid residues 27–75) domain in CREBL2, indicating that CREBL2 is a bZIP (basic leucine zipper) transcription factor of the CREB/ATF (activating transcription factor) family. In humans, more than 55 bZIP transcription factors have been cloned [15], and many of them are reported to be associated with preadipocyte differentiation and adipogenesis [25,1618]. CREBL2 shares the highest identity (41%) with CREB over a 48-base-long region, which encodes the CREB bZIP domain [19]. The CREB bZIP domain consists of approximately 30 amino acids rich in basic residues involved in DNA binding, followed by a leucine-zipper motif involved in protein dimerization [16,20], suggesting that CREBL2 encodes a protein with DNA-binding capabilities. Transcription factor dimerization can increase the selectivity of protein–DNA interactions and the diversity of DNA binding from a relatively small number of proteins [21]. Dimerization also results in establishment of complex regulatory networks [22], and dimers can interact with other necessary proteins for transcription [23]. In the present study, we indicate that CREBL2 is up-regulated during adipogenesis. Functional investigations confirm that CREBL2 can interact with CREB, the central adipocytic factor which was confirmed to be necessary and sufficient to promote adipogenic conversion and prevent apoptosis of mature adipocytes [2,16]. Using constitutively active and potent siRNA (short interfering RNA) against CREBL2, we investigated the effects of CREBL2 on preadipocyte differentiation and lipogenesis, as well as its potential roles in modulating the adipogenic regulators. The results presented suggest that CREBL2 may be a novel crucial adipogenesis-regulating protein.

MATERIALS AND METHODS

Reagents

RNA extraction reagent TRIzol®, the ThermoSCRIPT RT-PCR System and CNBr Sepharose 4B were obtained from Invitrogen Life Technologies. Rabbit antibodies against β-actin, histone H3.1, PPARγ, C/EBPα, CREB and P-CREB (Ser133-phosphorylated CREB), as well as insulin, 3-isobutyl-methyl-xanthine and dexamethasone were obtained from Sigma–Aldrich. Rabbit antibodies against GLUT (glucose transporter) 1 and GLUT4 were obtained from Santa Cruz Biotechnology. IRDye™ 800-conjugated secondary antibodies against rabbit IgG were purchased from LI-COR Biosciences.

cDNA cloning and vector construction

The full-length coding region of CREBL2 cDNA (GenBank® accession number NM_177687.3) (Supplementary Figure S1 at http://www.BiochemJ.org/bj/439/bj4390027add.htm) and CREB cDNA (GenBank® accession number NM_009952.2) were amplified from a mouse liver cDNA library (Clontech Laboratories) by PCR using specific primers (CREBL2: forward 5′-AGTCAGAGCCATATCCCCTTCTT-3′ and reverse 5′-ACACCATTGGCTTTGATCACGGACT-3′; CREB: forward 5′-AGCTTGTACCACCGGTAACTAAATG-3′ and reverse 5′-TCCATTTTCCACCTTAACAGGTGCC-3′). The purified PCR product was used to construct plasmids of pcDNA.3.1-CREBL2 and pDsred-CREBL2. After being confirmed by sequencing, all mammalian expression plasmids were extracted for transfection using the EndoFree Plasmid Maxi Kit (Qiagen).

siRNA synthesis

Double-stranded siRNAs against CREBL2 were designed, chemically synthesized and PAGE-purified, free of RNase contamination, by Sigma–Aldrich (Table 1). Potent siRNAs against CREB were designed and synthesized as described previously [16]. The non-silencing siRNA was confirmed to have no matches with the complete human and mouse genome by a BLAST search (http://www.ncbi.nlm.nih.gov). All siRNAs were dissolved to a concentration of 20 nM with the RNAi buffer according to the manufacturer's protocol.

View this table:
Table 1 Sense sequences of the indicated siRNAs

Polyclonal anti-CREBL2 antibody preparation

Antibodies against CREBL2 were generated by immunization of rabbits with KLH (keyhole-limpet haemocyanin)-coupled CREBL2 peptides (SRERAICALREELEMYKQWCMAMDQG), which were synthesized by solid-phase synthesis and purified by HPLC to 90% purity (Chinese Peptide Company). Rabbit polyclonal antibodies were purified using CNBr Sepharose 4B coupled with specific CREBL2 peptide. The antibodies were validated by immunofluorescence and Western blot analysis. The rabbits used in this experiment were maintained according to the guidelines of the China Agricultural University Animal Care and Use Ethics Committee (Beijing, China). All rabbits were individually housed in cages and had access to food and water ad libitum

Cell culture and adipocytic induction

3T3-L1 fibroblasts (American Type Culture Collection) were maintained in DMEM (Dulbecco's modified Eagle's medium) with 4.5 g/l glucose supplemented with 10% FBS (fetal bovine serum). Before adipocytic induction, confluent fibroblasts were cultured in DMEM supplemented with 10% FBS for 2 days. Differentiation of 3T3-L1 preadipocytes was induced by exposing the confluent cells to MDI reagent (2 μM insulin, 0.5 μM 3-isobutyl-methyl-xanthine and 1 μM dexamethasone) for 2 days, and then cells were exposed to insulin (2 μM) alone for an additional 2 days. Following this period, the medium containing 10% FBS was replenished every other day. After the differentiation process, at least 90% of the cells had accumulated lipid droplets, and were used as mature adipocytes.

Oil Red O staining

Differentiation of preadipocytes to mature adipocytes was confirmed by Oil Red O staining of lipid vesicles [2]. Oil Red O staining was conducted 8 days after electroporation. Cells were rinsed in PBS prior to fixing with 4% paraformaldehyde for 10 min. Then cells were incubated in propylene glycol after which Oil Red O stock solution was applied for 5 min. Cells were washed with 85% propylene glycol and then in distilled water. The nuclei were stained with haematoxylin. Photographs of small intestinal morphology were collected with an Axioskop-2 microscope (Olympus) and an Image Processing System (Visitron Systems).

Cell transfection

Transfection of 3T3-L1 preadipocytes or adipocytes was performed by electroporation. On the day of electroporation, semi-confluent preadipocytes or differentiated adipocytes were detached from cell culture plates. Approximately 2×106 cells were mixed with 10 μg of DNA (or 10 nmol of respective siRNA) in 350 μl of serum-free medium. The preadipocytes were electroporated with a 300 V, 5 ms pulse and the mature adipocytes (on day 6 after induction of adipogenesis) were electroporated with a 160 V, 10 ms pulse, using a BTX T820 square-wave electroporator in a 2 mm cuvette (BTX). Cells were immediately incubated with fresh medium before reseeding on to multiple-well plates for further analysis. Cells with more than 70% transfection efficiency were used for further experiments.

RT (reverse transcription)–PCR and Q-PCR (quantitative PCR) analysis

Cells were collected at the indicated time points. RNA was prepared using TRIzol® reagent and RT was performed with the ThermoSCRIPT RT-PCR System according to the manufacturer's protocol. CREBL2 and GAPDH (glyceraldehyde-3-phosphate dehydrogenase) were amplified by PCR using specific primers [GAPDH (internal control): forward 5′-GTGAAGGTCGGTGTGAACGGATTT-3′ and reverse 5′-CTCCTTGGAGGCCATGTAGGCCAT-3′]. PCR amplification was conducted with 30 cycles (for CREBL2) or 25 cycles (for GAPDH). PCR products were separated using 1.0% agarose gel electrophoresis and were visualized by ethidium bromide staining.

Q-PCR analysis was carried out using the TaqMan Sequence Detection System (Applied Biosystems) and the DNA double-strand-specific SYBR Green I dye (Roche Applied Science) according to the manufacturer's instructions. The relative level of mRNA expression was normalized against the amount of 18S rRNA in each sample.

Protein extraction and immunoblot analysis

Whole-cell lysate and nuclear protein extraction were prepared for different assays as described previously [24] with slight modifications. Cells were pelleted and lysed in ice-cold lysis buffer (10 mM Hepes, 0.15 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 0.05% SDS and 0.5% Nonidet P40). The total supernatant protein concentration was extracted. Cell nuclear proteins were extracted using the ProteoJET™ Cytoplasmic and Nuclear Protein Extraction Kit (Fermentas) according to the manufacturer's protocol. Total protein (30 μg) was separated by SDS/PAGE (12.5% gel), transferred on to nitrocellulose membranes, blocked with 5% non-fat dried skimmed milk, and incubated with the corresponding primary antibodies (overnight at 4 °C) followed by the IRDye™ 800-conjugated secondary antibodies. IR-fluorophores on the membrane were excited at 780 nm and emission (at 820 nm) was quantified using channel 800 of the LI-COR Infrared Imaging System (Odyssey Software) and analysed with Odyssey software.

Co-immunoprecipitation

The co-immunoprecipitation assays of CREBL2 and CREB were performed as described previously [24] with slight modifications. The pcDNA.3.1-CREBL2 vector was constructed with the c-Myc epitope tag (EQKLISEEDL) and the CREB cDNA was constructed with FLAG epitope tag (DYKDDDDK). The epitope-tagged constructs were transiently transfected into 3T3-L1 preadipocytes. Following transfection (48 h), the cells were harvested. Pelleted cells were homogenized in cell lysis buffer (1 ml, 4 °C; Sigma–Aldrich). Homogenates were centrifuged (20 min at 14000 g at 4 °C) and supernatants were combined with either anti-c-Myc or anti-FLAG M2 affinity agarose (Sigma–Aldrich) and then mixed overnight (4 °C). Immunoadsorbents were recovered and analysed using Western blotting.

For the endogenous co-immunoprecipitation assay, 1×107 3T3-L1 preadipocytes were collected and resuspended in 0.5 ml of cell lysis buffer supplemented with protease inhibitors and DNase. One-tenth of the lysate was saved for immunoblotting, and the rest was used for immunoprecipitation. Immunoprecipitation was performed using standard procedures using corresponding primary antibodies with Protein G beads [for anti-HA (haemagglutinin)]. Bound proteins were analysed using Western blotting.

Fluorescence microscopy

Cell transfection and fluorescence detection were performed as described previously [24]. Briefly, cells were transiently transfected with pEGFP-N1-CREB and pDsred-C3-CREBL2, then plated on to glass coverslips and cultured for 24 h. The cells were fixed, quenched and permeabilized. Fluorescence microscopy images were collected with an Axioskop-2 microscope (Olympus), standard filter sets (Leika MicroImaging) and Metavue software (Visitron Systems).

For endogenous immunofluorescence microscopy of CREB and CREBL2, the 3T3-L1 preadipocytes were plated on to glass cover slips and cultured for 24 h. Then, the cells were fixed, quenched and permeabilized. Cells were co-stained with FITC-conjugated anti-CREB antibodies and TRITC (tetramethylrhodamine β-isothiocyanate)-conjugated anti-CREBL2 antibodies. Afterwards the cells were detected.

Lipogenesis assay

Lipogenesis was assayed as described previously [25,26]. Briefly, after the indicated treatment, cells were treated with insulin for 15 min. Subsequently, cells were incubated with 5 mM D[U-14C]glucose (2 μCi per well) for 60 min at 37 °C. Cells were then washed on ice three times with ice-cold PBS, scraped into 1 ml of PBS and shaken vigorously with 5 ml of Betafluor scintillant (National Diagnostics). The samples were allowed to settle overnight and radioactivity in the organic phase was determined by liquid scintillation counting. Activities of lipogenesis were normalized to protein concentrations.

Glucose uptake assay

The glucose uptake assay was conducted as described previously with some modification [17]. Briefly, cells were serum-starved in DMEM containing 0.2% BSA for 4 h at 37 °C. Cells were then treated with 100 nM insulin for 15 min. Subsequently, glucose uptake was initiated by adding [3H]-2-deoxyglucose (50 μM, 1 μCi/ml) and measured as described previously [17]. Non-specific background uptake in the presence of the inhibitor cytochalasin B (20 μm) was subtracted from all values. Glucose uptake activities were normalized to protein concentrations.

Dual-luciferase reporter assay

The transcriptional activity of CREB was measured by luciferase activities using trans-reporting systems. Cells were co-transfected with 50 ng of pcDNA3.1-CREBL2 plasmid, 45 ng of pFR-Luc plasmid, 5 ng of pFA-CREB fusion trans-activator plasmid (PathDetect, Stratagene) and 4 ng of pRL-TK plasmids as the internal control by electroporation as described above. After 6 h, approx. 4×104 cells were seeded into one well of a 96-well culture plate. The negative control was performed with the pcDNA3.1 plasmid. At 24 and 48 h after transfection, the cells were stimulated with p-methoxyamphetamine (50 mM) and ionomycin (1 mM) for 6 h, and then lysed in standard lysis buffer. Using a Synergy4 Multifunction Microplate Reader (Bio-Tek Instruments) the cell lysates were assayed for both firefly and Renilla luciferase activities according to the manufacturer's protocol (Promega).

RESULTS

CREBL2 is up-regulated during 3T3-L1 preadipocyte differentiation

The presence of CREBL2 at the mRNA and protein levels was confirmed by expression profile analysis in various mouse tissues (Figures 1A and 1B). A ∼2.5-kb transcript specific for CREBL2 was expressed in the liver, skeletal muscle and adipose tissue, consistent with the bioinformatics analysis, and a band of ∼14 kDa consistent with the calculated molecular mass of CREBL2 protein was expressed in various tissues, indicating that CREBL2 could be widely expressed. In particular, CREBL2 could be highly expressed in adipose tissue, skeletal muscle and liver, suggesting that CREBL2 plays an important role in these tissues.

Figure 1 Expression profiles of CREBL2, as well as the expression levels of CREBL2 during MDI-induced adipocyte differentiation

(A) RT–PCR analysis of CREBL2 mRNA expression in mouse normal tissues. GAPDH was used as an internal control. (B) The total nuclear proteins were extracted and CREBL2 endogenous protein expression was indicated by Western blot analysis using a rabbit polyclonal anti-CREBL2 antibody in various tissues. Histone H3.1 expression was detected as an internal control. Differentiation of 3T3-L1 preadipocytes was induced as described in the Materials and methods section. mRNA samples from differentiating cells were prepared and the mRNA levels of CREBL2 were measured by agarose gel electrophoresis imaging (C) and Q-PCR analysis (D) as described in the Materials and methods section. Experiments were performed in triplicate, with similar results. Results from Q-PCR analysis were expressed as means±S.D. (n=3) for three independent experiments. The data at all other time points were normalized to the values of the control (day 0). *P<0.05. (E) Total nuclear proteins of differentiating 3T3-L1 preadipocytes during adipogenesis were extracted and the protein levels of CREBL2 were detected by Western blot. Histone H3.1 expression was detected as an internal control.

To study the expression pattern of CREBL2 during preadipocyte differentiation, endogenous CREBL2 expression was measured. The RT–PCR data indicated that, 48 h after adipocytic induction, CREBL2 mRNA levels were significantly increased in 3T3-L1 preadipocytes (P<0.05), and remained high during the 8-day differentiation period (Figure 1C), consistent with the results of Q-PCR analysis (Figure 1D). Additionally, endogenous CREBL2 protein expression could be significantly up-regulated, and continued to increase during the 8-day differentiation period (Figure 1E).

CREBL2 interacts and co-localized with CREB in 3T3-L1 preadipocytes

Accumulated research has indicated that bZIP transcription factors tend to form homo- or hetero-dimers through their leucine zipper region, in order to bind to specific target DNA sequences and exert transcriptional activity [20]. Previous studie have suggested that CREBL2 might interact with CREB [20,27,28]. Consequently, in the present study, the potential relationship was evaluated by co-immunoprecipitation analysis. As shown in Figure 2(A), all FLAG-tagged CREB and Myc-tagged CREBL2 were expressed in 3T3-L1 preadipocytes. Additionally, immunoprecipitation results demonstrated that Myc-tagged CREBL2 pulled down FLAG-tagged CREB, whereas the empty FLAG control plasmid did not, suggesting that CREBL2 was able to bind to CREB in preadipocytes. Moreover, the endogenous co-immunoprecipitation analysis further validated that CREBL2 could bind to CREB under physiological conditions, as shown in Figure 2(B).

Figure 2 CREBL2 interacted and co-localized with CREB in situ using both overexpression and the endogenous protein in 3T3-L1 preadipocytes

(A) Cell lysates transfected with the indicated constructs were immunoprecipitated with the anti-Myc antibody to precipitate Myc–CREBL2. An anti-FLAG antibody was used to detect CREB proteins in the immunoprecipitate. CREBL2 co-immunoprecipitated with CREB proteins in 3T3-L1 preadipocytes. (B) Cells lysates were immunoprecipitated with the rabbit-derived anti-CREB antibody to precipitate CREB proteins. The HRP (horseradish peroxidase)-tagged anti-CREBL2 antibody was used to detect corresponding proteins in the immunoprecipitates. CREBL2 co-immunoprecipitated with CREB proteins in 3T3-L1 preadipocytes, whereas the control rabbit-derived IgG did not. (C and D) Cells were co-transfected with pEGFP-CREB (green) and pDsred-CREBL2 (red) plasmids. Confocal microscopy images were taken (C, low-magnification images of the whole well; D, high-magnification images of a few representative cells). Co-expression is shown in yellow. Cell nucleus morphology was observed by DAPI (4′,6-diamidino-2-phenylindole) staining (blue). (E) 3T3-L1 preadipocytes co-stained with FITC-conjugated anti-CREB antibody (green) and TRITC-conjugated anti-CREBL2 antibody (red), for the endogenous immunofluorescence microscopy of CREBL2 and CREB. Cell nucleus morphology was observed by DAPI staining (blue), and yellow represents co-expression. Experiments were performed in triplicate with similar results. Ab, antibody; IP, immunoprecipitation; WB, Western blot.

Furthermore, we co-transfected fluorescent plasmids pEGFP-CREB (green) and pDsred-CREBL2 (red) into 3T3-L1 preadipocytes. Either pEGFP-CREB or pDsred-CREBL2 was localized in the nucleus but not in the nucleolus, and the distribution of overexpressed CREBL2 protein entirely overlapped with CREB (Figure 2C, low magnification; Figure 2D, high magnification). The subcellular distribution of CREBL2 was consistent with that of Saydam et al. [27] and those of CREB were in agreement with other studies [29]. Figure 2(C) also indicated that more than 80% of the cells were successfully transfected. Figure 2(E) shows that the endogenous subcellular location of CREBL2 and CREB, demonstrating that either CREBL2 or CREB localized in the nucleus as some highly concentrated spots and the distribution of CREBL2 entirely overlapped with CREB. These results confirm the interaction between CREBL2 and CREB, and suggest that they may form a complex which functions in the nucleus.

Overexpressed CREBL2 protein induced preadipocyte differentiation

Adipogenic conversion of CREBL2-transfected 3T3-L1 preadipocytes was measured by the appearance of Oil Red O staining of lipid droplets and by the appearance of adipocyte markers. As shown in Figure 3(A), similar to the MDI-induced 3T3-L1 adipocytes, numerous fat droplets were observed in the CREBL2-overexpressing cells, whereas few lipid droplets appeared in the cells transfected with the empty vector, which suggests that overexpressed CREBL2 protein induces preadipocyte differentiation. It is well known that ectopic overexpression of PPARγ and C/EBPα is sufficient to promote adipogenic conversion and increase or accelerate adipogenesis [16,30]. The results of Q-PCR/immunoblot analysis showed that PPARγ and C/EBPα mRNA levels in 3T3-L1 preadipocytes transfected with pcDNA3.1-CREBL2 were significantly increased over the control (P<0.05, Figures 3B and 3C), whereas PPARγ and C/EBPα protein levels also obviously increased in CREBL2-overexpressed cells (P<0.05), and approached the level of CREBL2 in MDI-inducted 3T3-L1 adipocytes (Figure 3D), supporting the idea that overexpressed CREBL2 protein induces 3T3-L1 preadipocyte differentiation via activation of PPARγ and C/EBPα expression.

Figure 3 Effects of overexpressed CREBL2 protein on preadipocyte differentiation and adipocyte lipogenesis

(A and D) 3T3-L1 preadipocytes were transiently transfected with pcDNA3.1 or pcDNA3.1-CREBL2. MDI-induced cells served as a positive control. On the indicated day, cells were stained or collected to perform Q-PCR and Western blot analysis. (A) On day 8, cells were stained with Oil Red O to visualize lipid vesicles and representative phase-contrast photomicrographs of transfected 3T3-L1 adipocytes were taken as described in the Materials and methods section. The white arrow indicates the cell with lipid droplets and lipid vesicles [upper panels, low-magnification (×40) images to show the whole well; lower panels, high-magnification (×400) images of a few representative cells]. On days 4 and 8, Q-PCR analysis of PPARγ (B) and C/EBPα (C) were performed. Experiments were performed in triplicate, with similar results. Data are expressed as means±S.D. (n=3). *P<0.05 for cells transfected with pcDNA3.1-CREBL2 compared with the control. (D) Transfected cells were collected and nuclear proteins were extracted. Protein levels of PPARγ and C/EBPα were determined by Western blot analysis on day 8. (EH) Mature 3T3-L1 adipocytes were transiently transfected with pcDNA3.1 or pcDNA3.1-CREBL2. At 24 or 48 h after transfection, cells were used to performed the lipogenesis assay and glucose uptake assay, as well as Western blot analysis. (E) Lipogenesis was assayed using 5 mM D-[U-14C]glucose (1 μCi per well) and normalized to total protein. Cells pre-treated with insulin for 15 min served as a positive control. Data are expressed as means±S.D. (n=3). *P<0.05. AU, arbitrary unit. (F) Glucose uptake was assayed using [3H]-2-deoxyglucose as described in the Materials and methods section. Cells pre-treated with insulin for 15 min were used as a positive control. Data are expressed as means±S.D. (n=3). *P<0.05. (G) Transfected cells were collected. Whole-cell lysates were prepared and analysed by immunoblot analysis using GLUT1 and GLUT4 antibodies, with β-actin expression as an internal control. (H) Relative quantification of P-CREB and total CREB levels. The relative level of protein expression was normalized against the amount of β-actin. *P<0.05.

Overexpressed CREBL2 protein induced adipocyte lipogenesis

We next examined the effects of overexpressed CREBL2 protein on adipocyte lipogenesis, which is important for adipocyte differentiation [17,31]. Overexpressed CREBL2 protein induced adipocyte lipogenesis, similar to the cells under insulin-stimulated conditions (P<0.05, Figure 3E). Glucose uptake is the major driving force for lipogenesis in adipocytes [17,32]. Similarly, overexpressed CREBL2 protein significantly accelerated glucose uptake in cells transfected with CREBL2 (P<0.05, Figure 3F). In addition, overexpression of CREBL2 led to significantly higher GLUT1 protein expression, to a greater extent than the increase in GLUT4 protein levels (Figures 3G and 3H). As GLUT4 resides in an intracellular location in the absence of insulin stimulation, these results suggested that GLUT1 accounts for the higher basal glucose uptake and lipogenesis in CREBL2-transfected cells (Figure 3F).

Knockdown of endogenous CREBL2 protein suppressed 3T3-L1 preadipocyte differentiation

To further explore the effect of CREBL2 on adipocyte differentiation, specific siRNA as a means of depleting CREBL2 was used. In the present study, non-silencing siRNA or the two siRNAs against CREBL2 (siCREBL2-1 and siCREBL2-2) were transfected into 3T3-L1 preadipocytes. The results show that either siCREBL2-1 or siCREBL2-2 significantly inhibited endogenous CREBL2 expression at both the mRNA (Figure 4A) and protein (Figure 4B) levels. In contrast, no effects were observed with non-silencing siRNA.

Figure 4 Effect of knockdown CREBL2 expression on preadipocyte differentiation and adipocyte lipogenesis

(A and B) Synthetic siCREBL2-1 and siCREBL2-2 were transfected into 3T3-L1 preadipocytes respectively, whereas non-siRNA served as a negative control, and endogenous CREBL2 expression was measured by RT–PCR (A) and immunoblot analysis (B). (CF) 3T3-L1 preadipocytes were transiently transfected with siCREBL2-1 or siCREBL2-2. On the indicated day, cells were stained, or collected to perform Q-PCR and Western blot analysis. (C) Cells were stained with Oil Red O and representative photographs were taken. A white arrow indicates the cell with lipid droplets and lipid vesicles [upper panels, low magnification (×40); lower panels, high magnification (×400). Q-PCR analysis of PPARγ (D) and C/EBPα (E) were performed. Experiments were performed in triplicate, with similar results. Data are expressed as means±S.D. (n=3). *P<0.05 for cells transfected with siCREBL2 compared with the control. (F) Transfected cells were collected and nuclear proteins were extracted. Protein levels of PPARγ and C/EBPα were determined by Western blot analysis. (GJ) Mature 3T3-L1 adipocytes were transiently transfected with siCREBL2-1 or siCREBL2-2. At 24 or 48 h after transfection, cells were used to performed a lipogenesis assay and glucose uptake assay, as well as Western blot analysis. (G) Cells were incubated in the presence or absence of insulin for 15 min. Lipogenesis was assayed using 5 mM D-[U-14C]glucose. Data are expressed as means±S.D. (n=3). *P<0.05. (H) Cells were incubated in the presence or absence of insulin for 15 min. Glucose uptake was assayed using [3H]-2-deoxyglucose. Data are expressed as means±S.D. (n=3). *P<0.05. (I) Transfected cells were collected. Whole-cell lysates were prepared and analysed by immunoblot assay using anti-GLUT1 and anti-GLUT4 antibodies, with β-actin expression as an internal control. (J) Relative quantification of P-CREB and total CREB levels. The relative level of P-CREB and total CREB expression was normalized against the amount of β-actin. *P<0.05.

The potent siRNA against CREBL2 was transfected into 3T3-L1 preadipocytes. After MDI induction, adipogenic conversion of transfected 3T3-L1 preadipocytes was measured. As shown in Figure 4(C), many fat droplets were observed in the cells transfected with non-silencing RNA. However, few lipid droplets appeared in the cells transfected with either siCREBL2-1 or siCREBL2-2, which suggests that CREBL2 might be crucial during preadipocyte adipogenesis. Accumulated research has indicated that during the MDI-induction of 3T3-L1 adipocyte differentiation, PPARγ and C/EBPα were the most important biomarkers in adipogenesis [8,30]. Accordingly, the results of Q-PCR and immunoblot analysis showed that during adipogenesis, both the mRNA (Figures 4D and 4E) and protein (Figure 4F) levels of PPARγ, as well as C/EBPα, in 3T3-L1 preadipocytes transfected with either siCREBL2-1 or siCREBL2-2 were significantly decreased compared with the control (P<0.05), suggesting that knockdown of CREBL2 protein suppress 3T3-L1 preadipocyte differentiation via inhibition of PPARγ and C/EBPα expression.

Down-regulating CREBL2 protein suppressed 3T3-L1 adipocytes lipogenesis

The effects of down-regulating CREBL2 protein on 3T3-L1 adipocytes lipogenesis were investigated by using a D-[U-14C]- glucose incorporation assay [25,26]. Interestingly, depletion of CREBL2 decreased lipogenesis under both basal and insulin-stimulated conditions (Figure 4G). Similar to its effect on lipogenesis, CREBL2 depletion significantly reduced glucose uptake under both basal and stimulated conditions (Figure 4H). These results indicate that CREBL2 is critical during preadipocyte lipogenesis.

In addition, following knockdown of CREBL2, the protein level of GLUT1 decreased significantly under basal conditions (Figures 4I and 4J), which may account for the lower basal glucose uptake and lipogenesis in CREBL2 siRNA-treated cells (Figures 4G and 4H). Conversely, GLUT4 levels hardly changed under basal conditions, but decreased more significantly than GLUT1 in response to insulin treatment (Figures 4I and 4J), which may explain the improved insulin-stimulated glucose uptake and lipogenesis in 3T3-L1 adipocytes with reduced CREBL2 expression (Figures 4G and 4H).

Depletion of CREB blocked CREBL2-induced adipogenesis and vice versa

To reveal the role for CREB on CREBL2-induced preadipocyte differentiation, a potent siRNA against CREB was designed as described previously [2], and co-transfected with the eukaryotic expression vector pcDNA.3.1-CREBL2. As shown in Figure 5(A), when co-transfected with siCREB, few lipid droplets appeared in the CREBL2-overexpressed cells compared with the control. In addition, depletion of CREB entirely inhibited the expression of PPARγ and C/EBPα at both mRNA (Figures 5B and 5C) and protein (Figure 5D) levels (P<0.05). These results illustrated that knockdown of endogenous CREB protein completely suppressed overexpressed CREBL2-induced preadipocyte differentiation, suggesting that CREB may be crucial for CREBL2-induced adipogenesis.

Figure 5 Effects of either CREBL2 or CREB knockdown on CREB-induced or CREBL2-induced preadipocyte differentiation respectively

(AD) 3T3-L1 preadipocytes were transiently transfected with siCREB and pcDNA3.1-CREBL2. On day 8, cells were stained or collected to perform Q-PCR and Western blot analysis. (A) Cells were stained with Oil Red O and representative photographs were taken. A white arrow indicates the cell with lipid droplets and lipid vesicles [upper panels, low magnification (×40); lower panels, high magnification (×400)]. Q-PCR analysis of PPARγ (B) and C/EBPα (C) were performed. Experiments were performed in triplicate, with similar results. Data are expressed as means±S.D. (n=3). *P<0.05 for cells co-transfected with siCREB and pcDNA3.1-CREBL2 compared with cells co-transfected with non-siRNA and pcDNA3.1-CREBL2. (D) Transfected cells were collected and nuclear proteins were extracted. Protein levels of PPARγ and C/EBPα were determined by Western blot analysis. (EH) 3T3-L1 preadipocytes were transiently transfected with siCREBL2-1 and pcDNA3.1-CREB. On day 8, cells were stained or collected to perform Q-PCR and Western blot analysis. (E) Cells were stained with Oil Red O and representative photographs were taken. A white arrow indicates the cell with lipid droplets and lipid vesicles [upper panels, low magnification (×40); lower panels, high magnification (×400)]. Q-PCR analysis of PPARγ (F) and C/EBPα (G) were performed. Data are expressed as means±S.D. (n=3). *P<0.05 for cells co-transfected with siCREBL2-1 and pcDNA3.1-CREB compared with cells co-transfected with non-siRNA and pcDNA3.1-CREB. (H) Transfected cells were collected and nuclear proteins were extracted. Protein levels of PPARγ and C/EBPα were determined by Western blot analysis.

Conversely, siCREBL2-1 was transfected with the eukaryotic expression vector pcDNA.3.1-CREB. As shown in Figure 5(E), when co-transfected with siCREBL2, few lipid droplets appeared in the CREB-overexpressed cells compared with the control. Similarly, depletion of CREBL2 inhibited the expression of PPARγ and C/EBPα at both the mRNA (Figures 5F and 5G) and protein (Figure 5H) levels in CREB-transfected 3T3-L1 preadipocytes (P<0.05). These results illustrate that increased CREB was unable to drive adipogenesis in the absence of CREBL2, suggesting that CREBL2 was crucial for CREB-induced adipogenesis and might regulate the CREB signal pathway.

CREBL2 modulated the transcriptional activity of CREB and phosphorylation of CREB

To reveal the role of CREBL2 on the transcriptional activity of CREB, siCREBL2-1 was transfected into preadipocytes with the CREB luciferase trans-reporting system. As shown in Figure 6(A), overexpressed CREBL2 significantly increased CREB reporter gene activity compared with the non-silencing control (P<0.05). In contrast, knockdown of endogenous CREBL2 protein significantly suppressed CREB reporter gene activity compared with the control (P<0.05, Figure 6B). Ser133 phosphorylation of CREB plays an important role in its transcriptional activity [2]. The results showed that endogenous CREBL2 depletion significantly reduced the phosphorylated level of CREB (Figure 6D). These data indicate that CREBL2 is important in maintaining the transcriptional activity and phosphorylation of CREB.

Figure 6 Effects of CREBL2 on the transcriptional activity of CREB, and effects of either CREBL2 depletion or CREB depletion on protein stability and phosphorylation

(A) 3T3-L1 preadipocytes were co-transfected with pcDNA3.1-CREBL2 and CREB trans-reporting plasmids stimulated and lysed as described in the Materials and methods section. The cell lysates were assayed for both firefly and Renilla luciferase activities. Data are expressed as means±S.D. (n=3). *P<0.05 for cells transfected with pcDNA3.1-CREBL2 compared with the control. (B) 3T3-L1 preadipocytes were co-transfected with siCREBL2, and CREB trans-reporting plasmids. The cell lysates were assayed for both firefly and Renilla luciferase activities. Data are expressed as means±S.D. (n=3). *P<0.05. (CE) 3T3-L1 preadipocytes were transiently transfected with siCREBL2-1. On the indicated day, cells were collected to perform Q-PCR analysis on the CREB mRNA level (C). Data are expressed as means±S.D. (n=3). (D) On day 8, nuclear proteins of transfected cells were extracted for Western blot analysis using antibodies specific for P-CREB and total CREB. (E) Relative quantification of P-CREB and total CREB levels. The relative level of P-CREB and total CREB expression was normalized against the amount of histone H3.1 protein. *P<0.05 for cells at the indicated time points compared with cells on day 0. (F and G) 3T3-L1 preadipocytes were transiently transfected with siCREB. On the indicated day, cells were collected to perform Q-PCR analysis on the CREBL2 mRNA level (F). Data are expressed as means±S.D. (n=3). (G) On day 8, nuclear proteins of transfected cells were extracted for Western blot analysis using antibodies specific for CREBL2. (H) Model of the interaction between CREBL2 and CREB, and its role in adipogenesis and lipogenesis. Taken together with findings from previous studies [2,16], the results of the present study indicate that either CREBL2 or CREB is necessary and sufficient to promote adipogenesis. Under physiological conditions, CREBL2 interacts with CREB in the nucleus and results in preadipocyte differentiation via up-regulation of the expression of PPARγ and C/EBPα, as well as adipocyte lipogenesis accompanied with increased GLUT1 and GLUT4. The interaction between CREB and CREBL2 can improve their protein stability and up-regulate the phosphorylation level and transcriptional activity of CREB.

CREBL2 knockdown resulted in decreased protein levels of CREB and vice versa

To reveal the effect of the interaction between CREBL2 and CREB on their expression and protein stability, siCREBL2-1 or siCREB was transfected into preadipocytes. Endogenous CREBL2 depletion had no effect on the mRNA level of CREB compared with the control (P>0.05, Figure 6C), whereas it reduced the protein level of CREB, especially the protein level of P-CREB (Figures 6D and 6E). On the other hand, endogenous CREB depletion had no effect on the mRNA level of CREBL2 (P>0.05, Figure 6F), whereas it significantly reduced the protein level of CREBL2 (Figure 6G). These results suggest that the interaction between CREBL2 and CREB might be important to maintain their protein stability.

DISCUSSION

To our knowledge, the present study is the first demonstration that CREBL2, a novel bZIP_1 domain-containing protein, interacted with CREB, the central adipogenic regulator. The interaction, which can be predicted from bioinformatics analysis [16,20] and some previous studies [20,27,28], was confirmed by our overexpression co-immunoprecipitation and co-localization assays. Furthermore, we validated that the interaction was endogenous and both endogenous proteins located in the nucleus but not the nucleolus, as well as the distribution of CREBL2 entirely overlapped with CREB. All of these suggested that the interaction is physiological and important.

Bioinformatics analysis indicates that CREBL2 is a bZIP transcription factor of the CREB/ATF family, most of them are reported to be associated with preadipocyte differentiation and adipogenesis [25,16,17,18]. In particular, CREB was necessary and sufficient to drive adipogenic conversion of 3T3-L1 preadipocytes [2,16,33]. Therefore we hypothesized that CREBL2 plays an important role in adipogenesis, which was identified by the results of the present study. First, levels of CREBL2 markedly increased during hormone-induced differentiation of 3T3-L1 adipocytes. Secondly, overexpression of CREBL2 was sufficient to promote adipogenesis in association with up-regulation of the expression of PPARγ and C/EBPα, the two master genes controlling the transcriptional regulation of adipogenesis [8]. Lastly, depletion of CREBL2 can completely block the process of preadipocyte differentiation accompanied with the decreased levels of PPARγ and C/EBPα expression. Collectively, these data document a critical role for CREBL2 in modulating adipocytic conversion which might, at least partly, be associated with the interaction with CREB (Figure 6H). Members of the CREM (cAMP-response element modulator) family of transcription factors, which are structurally related to CREB, are confirmed to be expressed in preadipocytes, although their participation in adipogenesis has not been studied [3436]. Previous research has indicated that the individual loss of any of these factors has little, if any, impact on adipocyte differentiation [20]. Thus we can deduce that, compared with the other members of the CREM family, CREBL2 not only shared similarities with CREB, such as regulation by MDI, ability to bind cAMP-response elements and potential structural similarities, but also had other characteristics including interaction with CREB and being essential for adipocyte development.

Differentiation of preadipocytes into adipocytes requires the co-ordination of a complex network of transcription factors, cofactors and signal molecules [25]. The insulin/IGF-1 (insulin-like growth factor 1)/IRS (insulin receptor substrate) signal pathway and activation of Akt are well characterized to play an important role in the regulation of genes involved in multiple early adipogenic events [33,37], and then the Akt signalling cascade appears to induce or activate PPARγ and C/EBPα during the induction of 3T3-L1 adipocyte differentiation [38]. In the present study, functional investigations revealed that overexpression of CREBL2 protein increased the levels of PPARγ and C/EBPα in adipocytes in the absence of insulin stimulation. Conversely, knockdown of CREBL2 protein markedly suppressed the expression of PPARγ and C/EBPα. These findings suggest that CREBL2 controls 3T3-L1 preadipocyte differentiation via modulation of PPARγ and C/EBPα expression, the classic pathway of adipogenesis. To our knowledge, such an insulin-independent role for CREBL2 has not previously been recognized. However, further research will need to be performed to elucidate the signal pathway.

Additionally, using constitutively active and potent siRNA of CREBL2, the results of the present study indicate that CREBL2 plays an important role in lipogenesis. It has been demonstrated that during lipogenesis of mature adipocytes [7,8] there is activation of a straight chain fatty acid α-oxidation pathway leading to a marked accumulation of lipids in the adipogenesis programme [16,20], which can be measured by the rate of glucose uptake in adipocytes, as well as the activity of GLUTs [31,39,40]. Studies have demonstrated that MDI-induced lipogenesis is mediated by GLUT4 and its membrane recruitment by the insulin receptor [35,41]. The results of the present study indicated that, without MDI induction, CREBL2 overexpression was sufficient to enhance the rate of glucose uptake and drive lipogenesis of 3T3-L1 adipocytes accompanied with the increased protein level of GLUT1, meanwhile, in CREBL2 siRNA-treated cells, the protein level of GLUT1 decreased significantly under basal conditions. Conversely, following knockdown of CREBL2, GLUT4 levels had hardly changed under basal conditions, but decreased more significantly than GLUT1 in response to insulin treatment. These results indicated that, in response to the presence or absence of insulin, modulation of CREBL2 on mature adipocyte lipogenesis might be mainly mediated by GLUT4 and GLUT1 respectively, which was different from the classical lipogenesis pathway [41,42]. More research is needed to fully elucidate this process.

It has been demonstrated that CREB was necessary and sufficient to drive adipogenic conversion of 3T3-L1 preadipocytes [2,16,33]. In the present study, we confirmed that CREBL2 was also sufficient to drive adipogenic conversion using constitutively active and potent siRNA of CREBL2. What is the relationship between them? The results of the present study showed that silence of CREBL2 could completely block the role of CREB in preadipocyte differentiation, which suggests that the modulation of CREB on adipogenesis may be CREBL2-dependent. As an important transcription factor, P-CREB binds to the cAMP-response element enhancer elements and initiates the transcription of target genes [42,43]. In the present study, we indicated that overexpressed CREBL2 can up-regulate the transcriptional activity of CREB, and knockdown of endogenous CREBL2 protein suppressed the transcriptional activity of CREB. Further research has indicated that potent siRNA against CREBL2 can decrease the protein level of CREB and P-CREB, but not the mRNA level, supporting the idea that CREBL2 could increase the protein stability of CREB. In particular, following knockdown of CREBL2, CREB phosphorylation is reduced to a greater extent than CREB protein levels, which suggested that CREBL2 may be more important to maintain the phosphorylated level of CREB. Collectively, the present study has suggested that, during preadipocyte adipogenesis, CREBL2 plays a critical role in the transcriptional activity of CREB by modulating the protein stability and P-CREB.

On the other hand, the results of the present study demonstrated that silencing of CREB could completely block the role of CREBL2 in preadipocyte differentiation, which suggests that the modulation of CREBL2 on adipogenesis may also be CREB-dependent. Potent siRNA against CREB can decrease the protein level of CREBL2, but not the mRNA level, which suggested that CREB can increase the protein stability of CREBL2. Hence, we can deduce that the interaction between CREB and CREBL2 increased the protein stability and phosphorylation, and then bound to the cAMP-response element enhancer element of the upstream regulatory regions to up-regulate the transcription activity of target genes (Figure 6H). Transcription factor dimerization can increase the selectivity of protein–DNA interactions [21]. Dimerization also results in establishing complex regulatory networks [22], and dimers can interact with other proteins necessary for transcription [23]. The bZIP domain in CREB and CREBL2 includes the basic region involved in DNA binding as well as the leucine zipper region involved in protein dimerization [16,20]. More research is needed to fully elucidate this process and its mechanism.

In summary, we have demonstrated that CREBL2, a novel bZIP_1 domain-containing protein, interacted and co-localized with CREB. Using constitutively active and potent siRNA against CREBL2, CREBL2 was identified to be sufficient to drive preadipocyte differentiation and adipocyte lipogenesis via modulation of the expression of adipogenic regulators. CREBL2 can modulate the transcriptional activity of CREB by regulating the protein stability and phosphorylated level of CREB, meanwhile CREB can stabilize the CREBL2 protein. These results indicated that CREBL2 might function as a positive regulator of CREB, and that it participates in adipogenesis at various stages of the differentiation programme. The findings of the present study highlight the importance of CREBL2 in the formation of new adipocytes, thus providing a more comprehensive picture of adipogenesis and lipogenesis.

AUTHOR CONTRIBUTION

Defa Li and Xi Ma conceived and designed the experiments and edited the paper. Xi Ma, Heyu Zhang and Lan Yuan performed the experiments, analysed the data and prepared the digital images. Xi Ma and Hao Jing drafted the paper. Phil Thacker provided excellent assistance in editing the paper.

FUNDING

This work was supported by the Chinese Universities Scientific Fund [grant number 2010JS004]; the National Natural Science Foundation of China [grant number 30800790]; and the Basic Research Program of China [grant number 2004CB117503].

Abbreviations: ATF, activating transcription factor; bZIP, basic leucine zipper; C/EBP, CCAAT/enhancer-binding protein; CREB, cAMP-response-element-binding protein; CREBL2, CREB (cAMP-response-element-binding protein)-like 2; CREM, cAMP-response element modulator; DMEM, Dulbecco's modified Eagle's medium; FBS, fetal bovine serum; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GLUT, glucose transporter; P-CREB, Ser133-phosphorylated CREB; PPARγ, peroxisome-proliferator-activated receptor γ; Q-PCR, quantitative PCR; RT, reverse transcription; siRNA, short interfering RNA; TRITC, tetramethylrhodamine β-isothiocyanate

References

View Abstract