Cells isolated from adipose tissue fraction of LAs in primary culture and their descendants obtained during the first ten passages of subcultivation were subjected to phenotypic characterization in order to establish the crucial passage when a pure hADSC population was achieved. Zuk et al.[5] demonstrated the presence of multiple CD marker antigens, similar to those observed on MSCs, on the surface of processed lipoaspirate (PLA) cells.

Considering this, the surface phenotype of hADSCs was examined via quantitative q-PCR reaction. All PLA descendants were shown to be positive for stromal stem cell surface markers CD44, CD73 and CD90, as revealed by agarose electrophoresis of gradient RT-PCR products (Figure 2a). In addition, no expression of CD34 was observed starting with the third passage cells.

CD44, CD73 and CD90 antigens protein expression on the surface of hADSCs was confirmed by Western Blot analysis (Figure 2b). Moreover, hADSCs are negative for the hematopoietic marker CD34 starting with the third passage and positive for the stromal cell markers CD44, CD73 and CD90 as highlighted by immunofluorescence microscopy (Figure 2c).

All these data suggest that the third passage hADSCs express the markers associated with BM-MSCs. Thus, the pattern of surface antigens displayed by hADSCs in our study is in accordance with the phenotype of stem cells isolated from the bone marrow, umbilical-cord blood [30] and resembles with the results reported by Zuk et al.[20] and Gronthos et al.[4]. Consequently, only cells obtained starting with the third passage can be considered suitable PLA descendants for further adipogenic studies. Based on senescence studies we performed (data not shown), hADSCs between the third and the seventh passages proved to be appropriate for prospective investigations.

The differentiation potential of hADSC was extensively studied in the last decade [2,6,9,22,31]. It is already known that these cultured cells are able to undergo adipogenesis, chondrogenesis and osteogenesis in the presence of certain inducers. The adipogenic differentiation process differs by species, donor’s age, sex and tissue location [31] and thereby requires the adjustment of experimental parameters. However, for tissue engineering purposes, the differentiation process must be controlled in terms of clonal expansion rate and adipogenesis kinetics.

In our study, the adjustment of adipogenic differentiation was performed by means of adipogenic cocktail variation. Thereby, a set of main inducers consisting of 3-iso-butyl-1-methylxanthine (IBMX), dexamethasone (DEX) and troglitazone was established. IBMX is known to upregulate C/EBP, a transcriptional factor responsible for Krüppel-like transcription factor 5 (KLF5) activation. Subsequently, KLF5 triggers PPARγ2 promoting further transcription of the target genes [32]. The synthetic glucocorticoid DEX is a powerful adipogenic inducer for precursor cells isolated from different species, which probably acts by the inhibition of pref-1 expression [31]. DEX increases the C/EBP and PPARγ expression in 3T3-L1 preadipocytes. Tiazolidindione-induced adipogenesis was observed in multiple cell culture models since troglitazone, a PPARγ ligand, is a crucial adipogenic factor [31]. Insulin was a constant compound in the adipogenic cocktail. Although the amount of insulin receptors on the surface of progenitor cells is low, the insulin effect increases during the adipogenic process. Insulin promotes hADSCs differentiations by activating PI3-kinase and Akt-signaling pathways, which subsequently stimulate the expression of adipocyte specific genes, such as FAS, leptin and adiponectin [2].

Considering ATE as our final goal, four different protocols (P₀, P₁, P₂ and P₃) were proposed for the identification of an adequate adipogenic strategy. These protocols were created by mixing five different adipogenic media (MD₀–MD₄), as described in Tables 1 and 2 Taking into account the effect of each pro-adipogenic molecule, P₀–P₃ were designed for the administration of inducers separately from the differentiation supporting factors (biotin, indomethacine, hydrocortisone, T3 and transferrin). Moreover, insulin was added in all cocktails (MD₀–MD₄) and IBMX, DEX and troglitazone were considered as the main adipogenic inducers. Thus, P₁ was based on the incubation of the hADSC monolayers with the most complex adipogenic cocktail (MD₁). P₂ consisted in the same strategy for the first three days of induction, followed by the exclusion of the main adipogenic inducers from the differentiation medium (MD₂). To provide a slow cell differentiation, P₃ was designed using the minimal set of pro-adipogenic molecules required for the activation and maintenance of the process (MD₃, MD₄). All these strategies were related to the use of a reference adipogenic medium (MD₀), administered constantly for up to 21 days (P₀).

Oil Red O staining of monolayers exposed to MD₀-MD₄ (P₀–P₃) over an incubation period of 15 days was used to reveal that all protocols successfully induced and sustained adipogenesis. However, significant differences between the stained samples (Figure 3a) were identified by phase contrast microscopy.

Thereby, the cell exposure to the standard medium (MD₀) led to a fast lipid accumulation. The same results were obtained by incubating hADSCs with MD₁ (P₁), probably due to its chemical complexity. In contrast, P₃ treatment resulted in a low number of differentiated cells with decreased triacylglycerol (TAG) deposits. The most balanced rate of differentiation was observed after P₂ administration, when cells accumulated numerous small lipid droplets located around the nucleus. Taking into account the ATE possible applications, we decided to select P₂ for further studies.

Consequently, hADSCs’ incubation with MD₁ and MD₂ (P₂) for up to 21 days led to TAG accumulation starting with the tenth day after adipogenic induction. At 21 days post-adipogenic induction, cell dimensions increased several times and the morphology changed completely, from spindle to spherical shape, due to the presence of large lipid vesicles. Mature adipocytes gradually lost adherence to culture surface and floated into the medium (Figure 3b).

Rahman et al.[33], performed a proteomic analysis of 3T3-L1 preadipocytes subjected to adipogenesis two days post confluence by supplementing the growth culture medium with 0.5 mM IBMX, 0.25 μM DEX and 10 μg/mL insulin for two days and, afterwards, only with 10 μg/mL insulin until the end of the experiment. Eight days after this treatment, Oil Red O staining revealed that 90% of the cells expressed intracellular lipid accumulation. In contrast to these results, our study demonstrated that hADSCs, exposed to more complex adipogenic media (MD₁ and MD₂), started the TAG-accumulation at 10 days post induction, probably as a consequence of their more undifferentiated state compared to 3T3-L1 preadipocytes.

Aditionally, cell counting was performed at three days post adipogenic induction with P₂, at the moment when MD₁ was exchanged with MD₂, and revealed an increase in cell number as compared to the initial moment of induction (T₀) (data not shown). These results suggest that the composition of the media allowed cell proliferation simultaneously with hADSCs conversion to preadipocytes. These data are in accordance with phase contrast micrographs in Figure 3b, which display higher cellular density in samples exposed to P₂ at three days post induction, as compared to T₀. Moreover, in the view of prospective regenerative applications, the modulation of adipogenic conditions in order to control the equilibrium between cell proliferation and differentiation could potentially result in a greater amount of tissue formed from a set number of original cells.

Double fluorescence staining for perilipin and actin filaments allowed us to relate the changes in actin cytoskeleton to the adipogenic status during differentiation. Notably, perilipin fluorescence detection is correlated with the presence of intracellular lipid droplets highlighted by Oil Red O staining at 10 days post induction (Figure 3). Perilipin was not detected in undifferentiated cells and its levels of expression increased remarkably during the exposure to the adipogenic conditions ensured by P₂ (MD₁ and MD₂), suggesting its high potential as an adipogenic inducer (Figure 4).

Due to the progressive TAG accumulation during adipogenesis, the cell morphology changed from spindle to spherical shape (Figure 4). Our study reveals that the increasing cell volume during adipogenesis determines the reorganization of the actin filaments. These observations are in accordance with Verstraeten et al.[34], who demonstrated that actin stress fibers, vimentin and microtubules reorganize into a cortical network lining membrane caves, cage-like structures surrounding lipid droplets and into a network under the plasma membrane, respectively. What is more, Kanzaki et al.[35] showed that insulin stimulation increased cortical actin levels, which was subsequently followed by increased actin polymerization in the perinuclear region.

The expression of the typical mesenchymal stem cells markers CD44, CD73 and CD90 was investigated at the transcriptional level by q-PCR, whereas their presence on cell surface was assessed via Western blotting and immunofluorescence microscopy. In addition, the same techniques were used to examine the expression of CD34 as a hematopoietic stem cell marker during successive passaging.

Third passage hADSCs were exposed to four original adipogenic media (MD₁, MD₂, MD₃ and MD₄) over a 21 days period of incubation and the process was evaluated by comparison with a standard adipogenic medium (MD₀, ECACC 811D-250, Cell Applications, San Diego, CA, USA), which served as a positive control. Cells were exposed to this standardized formula of adipogenic differentiation medium, according to the manufacturer’s instructions, since the composition of the medium was not fully provided. MD₁-MD₄ adipogenic media were prepared by supplementing the CM with the following adipogenic inducers: IBMX, DEX, troglitazone, insulin, biotin, indomethacine, hydrocortisone, triiodothyronine and transferrin, (all purchased from Sigma- Aldrich Co., Steinheim, Germany), as described in Table 1.

The response of hADSCs to the adipogenic conditions was characterized in terms of intracellular lipid droplets accumulation, actin filaments distribution associated with the protein levels of perilipin and gene expression profiles of PPARγ2, perilipin, aP2 and FAS. These changes were monitored up to 21 days of adipogenesis in a time dependent manner. Moreover, a sample of undifferentiated hADSCs nonexposed to the adipogenic conditions (T₀) was used in our study to serve as starting point.