Brought to you by Portland Press Ltd.
Published on behalf of the International Federation for Cell Biology
Cancer Cell death Cell cycle Cytoskeleton Exo/endocytosis Differentiation Division Organelles Signalling Stem cells Trafficking
Cell Biology International (2010) 34, 737–746 (Printed in Great Britain)
A new protocol for functional analysis of adipogenesis using reverse transfection technology and time-lapse video microscopy
Elke Grönniger1, Sonja Wessel1, Sonja Christin Kühn, Jörn Söhle, Horst Wenck, Franz Stäb and Marc Winnefeld2
Research and Development, Research Special Skincare, Beiersdorf AG, Unnastrasse 48, 20245 Hamburg, Germany

Since the worldwide increase in obesity represents a growing challenge for healthcare systems, research focusing on fat cell metabolism has become a focal point of interest. Here, we describe a small interfering RNA (siRNA)-technology-based screening method to study fat cell differentiation in human primary preadipocytes that could be further developed towards an automated middle-throughput screening procedure. First, we established optimal conditions for the reverse transfection of human primary preadipocytes demonstrating that an efficient reverse transfection of preadipocytes is technically feasible. Aligning the processes of reverse transfection and fat cell differentiation utilizing peroxisome proliferator-activated receptor γ (PPARγ)-siRNA, we showed that preadipocyte differentiation was suppressed by knock-down of PPARγ, the key regulator of fat cell differentiation. The use of fluorescently labelled fatty acids in combination with fluorescence time-lapse microscopy over a longer period of time enabled us to quantify the PPARγ phenotype. Additionally, our data demonstrate that reverse transfection of human cultured preadipocytes with TIP60 (HIV-1 Tat-interacting protein 60)–siRNA lead to a TIP60 knock-down and subsequently inhibits fat cell differentiation, suggesting a role of this protein in human adipogenesis. In conclusion, we established a protocol that allows for an efficient functional and time-dependent analysis by quantitative time-lapse microscopy to identify novel adipogenesis-associated genes.

Key words: adipogenesis-associated gene, fluorescently labelled fatty acid, human adipogenesis, human preadipocyte, quantitative time-lapse video microscopy, reverse siRNA transfection

Abbreviations: DPBS, Dulbecco’s phosphate-buffered saline, EtOH, ethanol, EthD-1, ethidium homodimer, GAPDH, glyceraldehyde-3-phosphate dehydrogenase, TIP60, HIV-1 Tat-interacting protein 60, PPARγ, peroxisome proliferator-activated receptor γ, RNAi, RNA interference, siRNA, small interfering RNA

1These authors contributed equally to this work.

2To whom correspondence should be addressed (email

1. Introduction

The discovery of leptin secretion by adipocytes (Zhang et al., 1994; Campfield et al., 1995; Flier, 1995) was the starting point to study the cellular characteristics underlying fat metabolism (Rosen et al., 2000; Gregoire, 2001).

Adipocytes form during the period of childhood/adolescence, and their number remains constant in adulthood in both lean and obese individuals (Spalding et al., 2008). Since approximately 10% of adipocytes are renewed annually at all adult ages, the differentiation of fat cells represents an interesting therapeutic target for anti-obesity treatments. In this context, the identification of new adipogenesis-related genes might lead to the development of new clinical applications.

The use of small interfering RNA (siRNA) is currently the method of choice to study loss of function phenotypes. To characterize so far unknown adipogenesis-related genes, we established a protocol for the reverse transfection of cultured human preadipocytes. This method is based on a technique developed by Erfle et al. (2007) for the production of cell arrays for reverse transfection of tissue culture cells. This procedure has been successfully put into practice for a genome-wide RNA interference screening in human cells, and it offers multiple advantages. First, only a small amount of siRNA transfection solution is needed per spot. Due to this small volume, a high sample density can be reached, which accelerates data collection. Secondly, costs are reduced due to the minimal amount of siRNA needed per spot (Wheeler et al., 2005). Finally, an internal control is automatically provided, since reverse transfected (on the spots) and non-transfected (outside the spots) cells are cultivated without physical separation. This fact is especially important for the analysis of fat cell differentiation, since the conversion of preadipocytes into adipocytes can be monitored in control and reverse transfected cells cultivated under identical experimental conditions. Contrary to normal transfection, reverse transfection opens the possibility of manufacturing reproducible siRNA arrays created from the same sample source. After spotting, arrays can be stored for several months without loss of transfection efficiency (Silva et al., 2004; Erfle et al., 2007). Also, sample preparation for immunostaining does not require further automation.

In recent years, time-lapse video microscopy has become a powerful readout tool for siRNA screenings (Neumann et al., 2006). Contrary to assays based on end-point analysis, the time-resolved measurement provided by time-lapse video microscopy makes it possible to follow the kinetics of biological processes, such as the onset of differentiation or differences in fat accumulation rates. Consequently, discrimination between early and late events in the course of experiments becomes feasible.

Here, we present an experimental setup that allows for the identification of novel adipogenesis-associated genes utilizing reverse siRNA transfection of primary human preadipocytes. To benefit from the experimental advantages described by Erfle et al. (2004), we first developed a method to analyse human adipogenesis by aligning the processes of siRNA reverse transfection and differentiation in living cultured human preadipocytes. By additionally integrating the use of fluorescently labelled fatty acids, we further refined our protocol, aiming to qualitatively and quantitatively determine fat accumulation during (pre)adipocyte differentiation.

2. Materials and methods

2.1. Primary human (pre)adipocytes

Primary human preadipocytes isolated from thighs or waists of two different healthy female subjects were obtained from Zenbio Inc. Cells were cultured as previously described (Söhle et al., 2009).

For subsequent differentiation experiments, cells were (dependent on the cell density) incubated in basal growth medium (Cambrex) containing 10% fetal calf serum, 2 mM l-glutamine, 100 units/ml penicilline and 100 μg/ml streptomycin (Cambrex) for 3 or 4 days following transfection. Differentiation into adipocytes was initiated by addition of 10 μg/ml insulin, 1 μM dexamethasone, 200 μM indomethacin and 500 μM isobutylmethylxanthine (Cambrex) to the medium. For our studies, we used cells up to the third passage.

2.2. siRNA oligonucleotides

Oligonucleotides (Qiagen) used for siRNA experiments are based on human target sequences and have the following ordering numbers: GAPDH (glyceraldehyde-3-phosphate dehydrogenase) [Hs_GAPDH_3_HP_Validated siRNA]; PPARγ (peroxisome proliferator-activated receptor γ) [Hs_PPARG_1_HP siRNA]; TIP60 (HIV-1 Tat-interacting protein 60) [Hs_HTATIP_8_HP Validated siRNA]; control: scrambled siRNA [AllStars Neg. control siRNA].

2.3. Reverse transfection of human preadipocytes

To utilize siRNA technology for the genome-wide analysis of adipogenesis, the following protocol for reverse transfection was established as described by Erfle et al. (2007).

The transfection solution (0.2–2 μl) was manually spotted on to Lab-Tek chamber slides, coverslips or Permanox (Nunc) or 96-well plates (Greiner Bio-One) using a pipette. After spotting, slides or plates were immediately placed overnight into a gel drying box. After drying, slides can be stored at −20°C for several weeks without apparent loss of transfection efficiencies.

For reverse transfection experiments, cells were carefully dispensed to the centre of the slide or plate (16130 or 24190 cells/cm2 depending on the point of initiation of differentiation). Preadipocytes were then incubated for 3 or 4 days before initiation of adipogenesis or the performance of immunofluorescence microscopic analysis.

2.4. Determination of cell viability

A two-colour fluorescence cell viability assay based on the determination of live (esterase activity, calcein AM) and dead (ethidium homodimer) cells was used to evaluate possible cytotoxic effects of the reverse siRNA transfection (2 μl spot volume).

Briefly, preadipocytes were seeded in 96-well plates on spotted GAPDH–siRNA or scrambled–siRNA. After 4 days of incubation, cells were washed with 1×DPBS (Dulbecco’s phosphate-buffered saline; 200 μl/cavity; Cambrex). As a positive control for dead cell staining, control cells were treated with 70% EtOH (ethanol) for 30 min. After incubation for 30 min in 2 μM calcein AM (live cells) and 4 mM EthD-1 (ethidium homodimer, dead cells) (Molecular Probes) solution (LIVE/Dead® Viability/Cytotoxicity Kit) fluorescence was determined at 530 nm (calcein AM) and 645 nm (EthD-1) using the 96-well plate reader Safire 1 (Tecan). For restricting the detection zone to the spot area, a detection mask was used (Oris-Assay, Platypus Technologies).

2.5. Immunofluorescence microscopic analysis

Preadipocyte populations were reverse transfected (0.2 μl spot volume) on Lab-Tek chamber slides, coverslips and incubated for 4 days. Immunofluorescence analysis was performed as described previously (Söhle et al., 2009). For GAPDH staining, a primary antibody directed against GAPDH (1:500 in DPBS; sc-47724) was used. The secondary antibody was labelled with Cy3 (1:200 in DPBS; 715-165-150; donkey–anti-mouse IgG, Cy3 conjugated; Jackson), and cell nuclei were stained using Hoechst 33342 (1 μg/ml; Invitrogen). Results were determined using the fluorescence microscope Axio Observer (Zeiss).

2.6. Determination of triglyceride accumulation (end-point analysis)

For determination of triglyceride accumulation, preadipocytes were reverse transfected in 96-well plates using scrambled–siRNA or PPARγ–siRNA as described above (0.2–2 μl spot volume). After 3 or 4 days of incubation, differentiation was initiated, and cells were incubated for another 7–14 days (depending on the readout mode). For triglyceride and parallel cell nuclei staining, cells were incubated in the dark for 10 min with 4.5 μl AdipoRedTM (Cambrex) and 5 μl of Hoechst 33342 (1 μg/ml) solution in a total volume of 200 μl of DPBS. Spots containing siRNA were observed in bright-field illumination, which allows clear identification of each spot. Also, results were determined utilizing the fluorescence microscope Axio Observer (Zeiss) employing filters adapted to each dye. For quantification of data, images of AdipoRedTM/Hoechst 33342-stained cells were processed utilizing the ImageJ software (Rasband, 2008). The number of cell nuclei according to Hoechst 33342 staining was determined. In parallel, the area filled with yellow fluorescing lipid vesicles was calculated and normalized to the number of cell nuclei. Thus, the percentage of triglycerides per cell grown in a PPARγ–siRNA area in comparison with cells grown in the vicinity of a siRNA spot was measured.

2.7. Time-lapse video microscopy

Human preadipocytes were seeded on Lab-Tek chamber slides, cover Permanox, which were spotted with PPARγ–siRNA or scrambled–siRNA (2 μl spot volume) as described above. After incubation for 4 days, differentiation was initiated. After 3 additional days, cells were transferred to an incubation chamber (humidified atmosphere of 5% CO2 at 37°C) associated with a ScanR microscope (Olympus), and time-lapse video microscopy was started. Bright-field images were acquired for 4 days with a lapse time of 15 min.

2.8. Documentation of human adipogenesis using fluorescently labelled fatty acids

The Quencher-Based Technology Fatty Acid Uptake Assay Kit (QBTTM Fatty Acid Uptake Kit; Molecular Devices) consists of a proprietary formulation of the quenching agent Q-Red.1 (Molecular Devices) and 4,4-difluoro-5-methyl-4-bora-3a,4a-diaza-S-indacene-3-dodecanoic acid (BODIPY-FA; Molecular Probes, Inc.) (Liao et al., 2005). QBT Fatty Acid Uptake Assay stock solutions were dissolved completely by adding 10 ml of 1× HBSS buffer (1×Hank’s balanced salt solution with 20 mM Hepes and 0.2% fatty-acid-free bovine serum albumin). Following subsequent vortex mixing, this solution was mixed with differentiation medium in a ratio of 1:1.

To document the course of human adipogenesis, 7.5×103 human preadipocytes were seeded in a cavity of a 96-well plate and incubated in differentiation medium containing fluorescently labelled fatty acids and the non-toxic cell-impermeable quenching agent as described above. Incorporation of fluorescently labelled fatty acids was determined daily (days 1–16) utilizing the 96-well plate reader Infinite M 200 (Tecan) and visually documented with an inverted microscope Axio Observer (Zeiss). Each well was scanned on days 4, 8, 12 and 16 by generating 2×2 mosaic images.

Additionally, cell viability was determined by measurement of plasma membrane integrity (see above) after 16 days of cell cultivation in the presence of fluorescently labelled fatty acids.

2.9. Quantitative assessment of time-lapse video microscopy

For knock-down experiments, 7.5×103 human preadipocytes were reverse transfected in a 96-well plate with PPARγ–siRNA, TIP60–siRNA and scrambled–siRNA (2 μl spot volume). Differentiation was initiated 3–4 days after reverse transfection. To identify lipid vesicles during adipogenesis, fluorescently labelled fatty acids (QBTTM Fatty Acid Uptake Kit) were added to the differentiation medium as described above. Incorporation of fatty acids into cells was measured in real time using fluorescence microscopy (ScanR microscope). Fluorescence intensity directly correlates with the amount of lipids in the cell. Images were acquired on days 7–10 after induction of differentiation with a lapse time of 60 min. Quantification of fluorescence intensity was performed using the ScanR software (Olympus) by determining the total intensity of the fluorescence per image and point in time.

2.10. Software used for Figure preparation

For Figure 1, AxioVision Rel.4.7 (Zeiss), cellF v.2.4 (Olympus), PowerPoint and Excel. For Figure 2, AxioVision Rel.4.7 (Zeiss), ImageJ 1.40g, PowerPoint and Excel. For Figure 3, ScanR (Olympus), ImageJ 1.40g and PowerPoint. For Figure 4, cellF v.2.4 (Olympus), PowerPoint and Excel. For Figure 5, ScanR (Olympus), ImageJ 1.40g, PowerPoint and Excel.

3. Results

3.1. Reverse transfection of human preadipocytes

To determine whether a transfection of human preadipocytes is technically feasible, we investigated the knock-down efficiency of GAPDH–siRNA in cultured human preadipocytes. As confirmed by immunofluorescence analysis, significant siRNA-induced depletion of GAPDH was achieved 4 days after seeding the cells on the 0.2 μl of GAPDH–siRNA spot (Figure 1Ab). In contrast, cells cultivated on a scrambled–siRNA spot (that served as control) showed strong fluorescence signals caused by GAPDH staining (Figure 1Aa). In order to investigate cell viability of reverse transfected cell populations, esterase activity (Figure 1Ba) and plasma membrane integrity (Figure 1Bb) were analysed. Cells reverse transfected with scrambled siRNA or GADPH–siRNA displayed comparable esterase activities. Furthermore, cells reverse transfected with scrambled siRNA or GAPDH–siRNA showed just a minor EthD-1 staining compared with control cells incubated in the presence of 70% EtOH (Figures 1Bb and 1Bc). Taken together, these results indicate that the viability of cultured cells was not affected by the reverse transfection process.

3.2. Combination of reverse transfection and fat cell differentiation

To achieve the goal of identifying novel adipogenesis-related genes, the processes of reverse transfection and fat cell differentiation needed to be aligned experimentally. This was performed by establishing a protocol for the reverse transfection of human preadipocytes using PPARγ–siRNA. PPARγ was chosen because this nuclear receptor is the master regulator of adipogenesis and therefore an ideal gene to coordinate the reverse transfection with the process of fat cell differentiation. The most critical point here is the cell density. While conditions for fat cell differentiation are optimal at confluence, for reverse transfection, a cell density of approximately 30–50% is needed to induce a high knock-down efficiency. Therefore, reverse transfection experiments were carried out at a low preadipocyte density followed by an intermediate step of cultivation, allowing transfected preadipocytes to proliferate. According to the protocol described by Xu et al. (2006), we induced fat cell differentiation on either day 3 or 4. At these points in time, cell density had reached an optimal value for inducing adipogenesis.

To verify that the specific PPARγ–siRNA used for our experiments induced a significant PPARγ knock-down, cells were transfected conventionally, and the PPARγ–mRNA level was determined using qRT-PCR. Compared with controls, PPARγ–mRNA levels clearly decreased after transfection with PPARγ-specific siRNA (Supplementary Figure S1 at Also, even 6 days after siRNA transfection, reduced protein levels of PPARγ could be detected (data not shown).

In order to investigate if the experimental setup allows fat cell differentiation – indicated by fat accumulation – and to analyse whether a PPARγ knock-down influences adipogenesis, we examined triglyceride accumulation in cell populations growing on cell-culture dishes with a scrambled–siRNA or a PPARγ–siRNA spot located in the centre (indicated by the dashed lines, Figures 2Aa and 2Ab). On day 14, after induction of differentiation, reverse transfection with PPARγ–siRNA substantially decreased fat accumulation, compared with controls as observed by brightfield microscopy. Moreover, cells stained with AdipoRedTM (triglyceride staining), showed a lack of the AdipoRedTM signal within the area of the PPARγ–siRNA spot compared with cells growing in the vicinity of this spot. Notably, the number of nuclei visualized with Hoechst 33342 was comparable in all investigated areas (Figure 2Ac).

To quantify the observed phenotype induced by PPARγ–siRNA, we used the image analysis software ImageJ (Rasband, 2008) and compared the amount of triglycerides per cell grown on a PPARγ–siRNA spot, a scrambled–siRNA spot or in a non-treated area. With this method, the number of nuclei was determined from the Hoechst 33342-stained images, and the amount of triglycerides was determined from the total intensity of yellow fluorescence caused by the AdipoRedTM staining. The total amount of triglycerides per cell was calculated, and the value of the untreated cells was set at 100% (Dragunow et al., 2007). Figure 2(B) shows the substantial inhibition of adipogenesis in preadipocytes after reverse transfection with PPARγ–siRNA (25±10%) in comparison with non-transfected cells (100%). Transfection with scrambled–siRNA did not have any effect (101±10%). Taken together, these results demonstrate a successful adaptation of the conditions for reverse transfection as well as adipogenesis.

3.3. Time-lapse video microscopy

Based on the established protocol for reverse transfection combined with adipogenesis, we integrated the application of live-cell imaging using time-lapse video microscopy to follow the kinetics of fat accumulation after siRNA treatment. Since triglyceride incorporation is a rather slow process, time-lapse video microscopy needed to be carried out over several days. For this reason, (pre)adipocytes growing on Lab-Tek chamber slides were time-lapse recorded every 15 min for 91.5 h, starting 3 days after induction of differentiation (Supplementary Movies S1 and S2 at As illustrated in Figure 3(A), control transfected cells had already entered the process of adipogenesis on day 3 after induction of fat cell differentiation, as evidenced by the appearance of a partially spherical phenotype. After completion of time-lapse video microscopy, many of these cells were filled with multiple lipid droplets (Figures 3B–3D). In contrast, on day 3, after induction of adipogenesis, cells growing on a PPARγ–siRNA spot (Figure 3E) still displayed the typical preadipocyte spindle-shaped morphology, and no signs of fat cell differentiation could be observed. Moreover, Figures 3(F)–3(H) demonstrated that, even after cultivation for 4 additional days, only few lipid droplets had been formed in PPARγ-knock-down cells, indicating that the knock-down was sufficient to inhibit adipogenesis. Notably, the repeated illumination and prolonged incubation on the microscope dish did not induce toxic side effects in the reverse-transfected cell populations.

3.4. Analysis of human adipogenesis using fluorescently labelled fatty acids

To quantify the dynamic process of adipogenesis over time, incorporation of fluorescently labelled fatty acids was determined daily (days 1–16 after induction of differentiation), using a fluorescence reader in combination with a microscope. Figure 4(A) shows that, during adipogenesis, cells accumulated the fluorescently labelled fatty acids, as evidenced by an increasing fluorescence signal. Figure 4(B) depicts the increase in fluorescence intensity over a time period of 16 days. Based on these results, all subsequent experiments were carried out between days 7 and 10 after induction of differentiation. Notably, cell viability was not affected by the use of the fluorescently labelled fatty acids, as indicated by the absence of cell death in cell populations on day 16 after induction of differentiation (Figure 4C).

3.5. Time-lapse video microscopy of siRNA-transfected preadipocytes

To quantitatively evaluate the role of different genes in adipogenesis, preadipocytes were reverse transfected with either PPARγ–siRNA, TIP60–siRNA or scrambled–siRNA. Incorporation of fluorescently labelled fatty acids was determined between days 7 and 10 after induction of differentiation in real time, using fluorescence microscopy (Figure 5A, Supplementary Movies S3, S4 and S5 at

Quantification of triglyceride accumulation is shown in Figure 5(B). Cells cultivated for 10 days on a scrambled–siRNA spot exhibited a similar amount of triglycerides (103±13%) compared with cells growing on spots without siRNA (set as 100%). In contrast, cells growing on PPARγ–siRNA and TIP60–siRNA spots showed a decreased triglyceride accumulation of (29±4%) and (8±1%), respectively. Overall, these results indicate that PPARγ as well as TIP60–siRNA reduced preadipocyte differentiation and, therefore, interfere with adipogenesis.

4. Discussion

siRNAs are a powerful tool for the analysis of gene function (Elbashir et al., 2001). Applying this technique in the field of obesity and fat metabolism research, Ashfari et al. (2003) performed a genome-wide RNAi (RNA interference) analysis of genes associated with Caenorhabditis elegans’ fat storage regulation. Although C. elegans is an excellent model organism for RNAi experiments, it has been widely accepted that significant differences in tissue biology exists between species.

Therefore, our aim was to establish a protocol for a middle-throughput screening using human primary preadipocytes to identify new adipogenesis-associated genes. Specifically, we describe the adaptation of siRNA transfection technology for the functional analysis of human fat cell differentiation over a longer period of time. Overall, the method is based on the reverse transfection of primary human preadipocytes with siRNA.

To tackle this matter experimentally, we first developed the optimal conditions for the reverse transfection of human preadipocytes. Secondly, we aligned the processes of reverse transfection and fat cell differentiation to reduce gene expression during adipogenesis and analyse the resulting phenotype. Thirdly, we cultured reverse transfected preadipocytes in the presence of fluorescently labelled fatty acids and quantitatively assessed fat cell differentiation using time-lapse video microscopy.

Since the transfection of primary cells is generally challenging (Moffat et al., 2006), our initial goal was to determine ideal transfection conditions. We adapted protocols developed by Erfle et al. (2007) and by Xu et al. (2006) for the delivery of siRNA into living cultured cells and for siRNA transfection of human preadipocytes, respectively.

Based on the work of Erfle et al. (2007), we spotted the transfection reagent LipofectamineTM 2000 and the respective siRNA in a gelatin/fibronectin matrix directly on the coverslide or cell culture dish/96 well plate (Erfle and Pepperkok, 2007). The reverse transfection efficiency in our experiments was approximately 95%, 4 h after seeding cells on the siRNA spots. In comparison, ‘conventional’ transfection of human preadipocytes only yielded a transfection efficiency of 83% (data not shown). These results are in line with observations by Neumann et al. (2006) showing a high transfection efficacy of 99% when reverse transfected HeLa cells were analysed with a fluorescently labelled siRNA.

A crucial parameter for reverse transfection is the siRNA transfection complex spot volume. In their protocol, Erfle et al. (2007) utilized solid pins creating spots with a diameter of 400 μm each. Using our experimental setup, the results showed that manual spotting of 0.2 μl creates reproducible spots (approximately 1.2 mm in diameter), which allow for the investigation of a representative phenotype (Figures 1A and 2A). However, for quantitative analysis, we increased the spotting volume of the transfection solution (Figures 2B and 5). This adaptation increased the reproducibility of experiments by increasing the number of cells investigated.

Since PPARγ is described as the master regulator of adipogenesis, its gene expression is not substantially increased until induction of adipogenesis (Chawla et al., 1994; Tontonoz et al., 1994). Accordingly, a significant knock-down of this protein by siRNA transfection can only be observed after induction of (pre)adipocyte differentiation. To test and optimize our experimental approach, we reverse transfected preadipocytes with PPARγ–siRNA, induced differentiation on day 4 and carried out a triglyceride staining (AdipoRedTM), as well as a bright-field microscopic analysis 14 days after induction of differentiation. Importantly, our results clearly showed a reduced fat cell differentiation compared with the control situation, indicating that adipogenesis was successfully inhibited after reverse transfection with PPARγ–siRNA.

Although staining of triglycerides with AdipoRedTM is a well-accepted tool to identify the accumulation of intracellular lipids, a drawback of this method is that an AdipoRedTM staining constitutes an end-point assay, since even low concentrations of this reagent cause toxic effects in cell populations. In general, end-point assays only detect phenotypes of genes when the protein level is adequately reduced at the point in time the assay is performed. Moreover, it is difficult to discriminate the primary phenotype caused by gene knock-down from secondary cellular responses (Neumann et al., 2006).

To overcome these limitations, we combined the methods of reverse siRNA transfection and time-lapse microscopy. Time-lapse imaging allows for the detection of early as well as late phenotypes and minimizes loss of information. Three days after induction of differentiation, PPARγ–siRNA transfected preadipocytes were transferred to an incubation chamber and analysed for 4 days with a time lapse of 15 min, using bright-field microscopy. Our results show that, after reverse transfection of preadipocytes with PPARγ–siRNA and subsequent induction of differentiation, cells displayed a substantially decreased amount of lipid droplets compared with control cells (Figure 3 and Supplementary Movies S1 and S2).

However, a disadvantage of phase-contrast microscopy is the fact that quantification is difficult to perform. To address this shortcoming and to quantify the process of adipogenesis in a continuous manner, we adapted a fatty acid uptake assay that allows fatty acid transport to be measured in real time, using standard fluorescence microscopy. This assay has already been successfully put into practice for the quantification of cellular fatty acid uptake in 3T3-L1 cells over a period of 1 h (Liao et al., 2005). To test if this method can be adapted to the quantitative long-term observation of (pre)adipocyte differentiation, we analysed the uptake of fluorescently labelled fatty acids into (pre)adipocytes over a period of 16 days. After optimization of cell density and concentration of fluorescently labelled fatty acids, our results showed that, on day 4 after induction of differentiation, fatty acids were hardly detectable. On day 8, however, single preadipocytes clearly showed a fluorescence signal, whereas after 12 days, the majority of cells incorporated the fluorescently labelled fatty acids into lipid droplets. Saturation was almost reached after 16 days. Importantly, cell characteristics were not affected, and toxic side effects were not observed.

In a next step, we investigated if fluorescently labelled fatty acids can be successfully employed using the technique of video microscopy in combination with reverse transfected cells. Since previous experiments revealed a first weak fluorescence signal 8 days after initiation of differentiation, video microscopy was started one day earlier on day 7 to analyse the first steps of adipogenesis.

The fluorescent signal increased continuously over time, confirming that the fluorescence intensity of the fluorescently labelled fatty acids stayed constant during an observation period of 4 days (time lapse of 60 min). This augmentation also verified differentiation of cell populations, as indicated by the intracellular formation of lipid droplets. Moreover, no differences could be observed between the control cell populations either transfected with or without scrambled siRNA. In contrast, PPARγ–siRNA-transfected cells did not significantly accumulate fatty acids. Taken together, the fatty acid uptake assay can be used for quantification of the differentiation process of live (pre)adipocytes.

As described above, our protocol was optimized on the reverse transfection using PPARγ–siRNA. As a first test to decrease expression of other genes, we next investigated the influence of a TIP60 gene knock-down on the differentiation of cultured human preadipocytes.

TIP60 has been shown to regulate PPARγ transcriptional activity in mice cells (van Beekum et al., 2008). In that study, the authors used a siRNA-mediated down-regulation of TIP60 expression resulting in inhibition of the differentiation of 3T3-L1 cells into adipocytes. However, to our knowledge, in human cells, the role of TIP60 in adipogenesis has not been determined so far. Our data demonstrate an inhibition of fat cell differentiation after reverse transfection of human cultured preadipocytes with TIP60–siRNA (Supplementary Movie S5).

In conclusion, we established a protocol that allows for the identification of novel adipogenesis-associated genes. This assay is based on effective reverse transfection of human primary preadipocytes with siRNA. Further, by using fluorescently labelled fatty acids in combination with fluorescent time-lapse microscopy over a longer period of time, we were able to quantitatively determine fat accumulation during adipogenesis. Our protocol provides a basis for functional analysis of fat cell differentiation in human primary cells and is suitable for the further expansion to and development of an automated middle-throughput screening method.

Author contribution

Elke Grönniger combined the reverse transfection process with induction of adipogenesis. She performed lipid (AdipoRed) and protein staining (GAPDH) and the fluorescence microscopic analysis. Moreover, she assisted with interpretation of the results and helped draft the manuscript. Sonja Wessel performed time-lapse video microscopy, analysed microscopic images and quantified the results concerning the different phenotypes. In addition, she assisted with interpretation of the results and helped draft the manuscript. Elke Grönniger and Sonja Wessel contributed equally to this work. Sonja Christin Kühn performed experiments using fluorescently labelled fatty acids, and Jörn Söhle investigated the cytotoxic side effects of reverse transfected cells. Horst Wenck and Franz Stäb assisted with the interpretation of the results. Marc Winnefeld supervised the analyses and helped draft the manuscript. All authors read and approved the final manuscript.


We kindly thank Dr Erfle (MitoCheck Project Group, EMBL, Heidelberg, Germany) for a helpful discussion and advice concerning reverse transfection arrays. We also thank Mr B. Larisch for the critical review of the manuscript prior to submission.


This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.


Ashrafi, K, Chang, FY, Watts, JL, Fraser, AG, Kamath, RS and Ahringer, J (2003) Genome-wide RNAi analysis of Caenorhabditis elegans fat regulatory genes. Nature 421, 268-72
Crossref   Medline   1st Citation  

Campfield, LA, Smith, FJ, Guisez, Y, Devos, R and Burn, P (1995) Recombinant mouse OB protein: evidence for a peripheral signal linking adiposity and central neural networks. Science 269, 546-49
Crossref   Medline   1st Citation  

Chawla, A, Schwarz, EJ, Dimaculangan, DD and Lazar, MA (1994) Peroxisome proliferator-activated receptor (PPAR) gamma: adipose-predominant expression and induction early in adipocyte differentiation. Endocrinology 135, 798-800
Crossref   Medline   1st Citation  

Dragunow, M, Cameron, R, Narayan, P and O’Carroll, S (2007) Image-based high-throughput quantification of cellular fat accumulation. J Biomol Screen 12, 999-1005
Crossref   Medline   1st Citation  

Elbashir, SM, Harborth, J, Lendeckel, W, Yalcin, A, Weber, K and Tuschl, T (2001) Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494-98
Crossref   Medline   1st Citation  

Erfle, H, Simpson, J, Bastiaens, PI and Pepperkok, R (2004) siRNA cell arrays for high-content screening microscopy. BioTechniques 37, 454-62
Medline   1st Citation  

Erfle, H, Neumann, B, Liebel, U, Rogers, P, Held, M and Walter, T (2007) Reverse transfection on cell arrays for high content screening microscopy. Nat Protoc 2, 392-9
Crossref   Medline   1st Citation   2nd   3rd   4th   5th   6th  

Erfle, H and Pepperkok, R (2007) Production of siRNA- and cDNA-transfected cell arrays on noncoated chambered coverglass for high-content screening microscopy in living cells. Methods Mol Biol 360, 155-61
Medline   1st Citation  

Flier, JS (1995) The adipocyte: storage depot or node on the energy information superhighway? Cell 80, 15-8
Crossref   Medline   1st Citation  

Gregoire, FM (2001) Adipocyte differentiation: from fibroblast to endocrine cell. Exp Biol Med 226, 997-1002
1st Citation  

Liao, J, Sportsman, R, Harris, J and Stahl, A (2005) A real-time quantification of fatty acid uptake using a novel fluorescence assay. J Lipid Res 46, 597-602
Crossref   Medline   1st Citation   2nd  

Moffat, J, Grueneberg, DA, Yang, X, Kim, SY, Kloepfer, AM and Hinkle, G (2006) A lentiviral RNAi library for human and mouse genes applied to an arrayed viral high-content screen. Cell 124, (6), 1283-98
Medline   1st Citation  

Neumann, B, Held, M, Liebel, U, Erfle, H, Rogers, P, Pepperkok, R and Ellenberg, J (2006) High-throughput RNAi screening by time-lapse imaging of live human cells. Nat Methods 3, 385-90
Crossref   Medline   1st Citation   2nd   3rd  

Rasband, WS (1997–2008) ImageJ., U.S. National Institutes of Health, Bethesda, Maryland, USA
1st Citation  

Rosen, ED, Walkey, CJ, Puigserver, P and Spiegelman, BM (2000) Transcriptional regulation of adipogenesis. Genes Dev 14, 1293-307
Medline   1st Citation  

Silva, JM, Mizuno, H, Brady, A, Lucito, R and Hannon, GJ (2004) RNA interference microarrays: high-throughput loss-of-function genetics in mammalian cells. Proc Natl Acad Sci USA 101, 6548-52
Crossref   Medline   1st Citation  

Söhle, J, Knott, A, Holtzmann, U, Siegner, R, Grönniger, E and Schepky, A (2009) White tea extract induces lipolytic activity and inhibits adipogenesis in human subcutaneous (pre)-adipocytes. Nutr Metab (London) 6, 20
Crossref   1st Citation   2nd  

Spalding, KL, Arner, E, Westermark, PO, Bernard, S, Buchholz, BA and Bergmann, O (2008) Dynamics of fat cell turnover in humans. Nature 453, 783-7
Crossref   Medline   1st Citation  

Tontonoz, P, Hu, E, Graves, RA, Budavari, AI and Spiegelman, BM (1994) mPPAR gamma 2: tissue-specific regulator of an adipocyte enhancer. Genes Dev 8, 1224-34
Crossref   Medline   1st Citation  

van Beekum, O, Brenkman, AB, Grøntved, L, Hamers, N, van den Broek, NJ and Berger, R (2008) The adipogenic acetyltransferase Tip60 targets activation function 1 of peroxisome proliferator-activated receptor gamma. Endocrinology 149, 1840-9
Crossref   Medline   1st Citation  

Wheeler, DB, Carpenter, AE and Sabatini, DM (2005) Cell microarrays and RNA interference chip away at gene function. Nat Genet 37(Suppl), S25-30
Medline   1st Citation  

Xu, Y, Mirmalek-Sani, S-H, Yang, X, Zhang, J and Oreffo, RO (2006) The use of small interfering RNAs to inhibit adipocyte differentiation in human preadipocytes and fetal-femur-derived mesenchymal cells. Exp Cell Res 312, 1856-64
Crossref   Medline   1st Citation   2nd  

Zhang, Y, Proenca, R, Maffei, M, Barone, M, Leopold, L and Friedman, JM (1994) Positional cloning of the mouse obese gene and its human homologue. Nature 372, 425-32
Crossref   Medline   1st Citation  

Received 9 October 2009/11 February 2010; accepted 1 April 2010

Published as Cell Biology International Immediate Publication 1 April 2010, doi:10.1042/CBI20090299

© The Author(s) Journal compilation © 2010 Portland Press Limited

ISSN Print: 1065-6995
ISSN Electronic: 1095-8355
Published by Portland Press Limited on behalf of the International Federation for Cell Biology (IFCB)