SEARCH

SEARCH BY CITATION

Keywords:

  • Adult stem cells;
  • Self-renewal;
  • Fibroblast growth factor;
  • Long-term expansion;
  • Adipogenesis

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Adipose tissue-derived stem cells offer tremendous potential for regenerative medicine. However, characterization of their self-renewal ability has not been performed yet, although it is a crucial feature for in vitro expansion of undifferentiated cells and in vivo maintenance of stem cell pools. We have undertaken the identification of molecular events that are involved in in vitro self-renewal of human multipotent adipose-derived stem (hMADS) cells from young donors, by assessing their proliferation rate, their ability to grow at the single-cell level (clonogenicity), and their differentiation potential. As hMADS cells are propagated in culture, cell morphology changes dramatically, concomitantly to a progressive decrease in proliferation, clonogenicity, and differentiation potential. This decrease is associated with a decrease in fibroblast growth factor 2 (FGF2) expression and can be circumvented by chronic treatment with exogenous FGF2. Moreover, analysis of FGF2 secretion revealed that it is exported to hMADS cell surface without being released into the culture medium, suggesting a strictly autocrine loop. Indeed, treatment of FGF2-expressing hMADS cells with PD173074, a specific FGF receptor inhibitor, decreases dramatically their clonogenicity and differentiation potential. Thus, hMADS cells express a functional autocrine FGF loop that allows maintenance of their self-renewal ability in vitro. Finally, inhibition of mitogen-activated protein kinase kinase 1 reduces the clonogenic potential of hMADS cells but does not affect their differentiation potential, indicating that the extracellular signal-related kinases 1/2 signaling pathway is partly involved in FGF2-mediated self-renewal. Together, our data clearly identify the key function of FGF2 in the maintenance of self-renewal of adipose tissue-derived stem cells.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

The prospective use of stem cells for the restoration of injured or diseased tissues has opened new fields of research. Plasticity is the first requirement for this therapeutic potential. Several studies have demonstrated that this feature is not restricted to embryonic stem cells. A number of stem cells isolated from adult tissues have proven to be multipotent in vitro and have been successfully used for tissue repair in vivo [1, 2]. Much attention has been paid to mesenchymal stem cells from human bone marrow because of their extended plasticity [3, 4]. Within this category, multipotent adult progenitor cells are able, at the single-cell level, to differentiate in vitro into multiple mesodermal, endodermal, and ectodermal lineages [5]. Moreover, when injected into the mouse blastocyst, they give rise to viable chimeric offspring, thus demonstrating their ability to incorporate into virtually all tissues in vivo [6]. Recently, adipose tissue has been identified as another source of multipotent adult stem cells [7, [8]9]. From a cell therapy perspective, adipose tissue presents several advantages compared with bone marrow as it is very large and can be easily removed by surgery with little trouble. Several research teams have successfully isolated, from adipose tissue, cell populations that are able to differentiate into mesenchymal cell types. They have been termed adipose-derived adult stem [9], processed lipoaspirate [7], and adipose tissue-derived stromal cells [10]. Recently, Rodriguez et al. characterized human multipotent adipose-derived stem (hMADS) cells from the stroma-vascular fraction isolated from infant adipose tissue [8, 11, 12]. After being cultured for more than 100 population doublings, these cells display a normal diploid karyotype. They retain multilineage differentiation potential as they undergo differentiation into adipocytes, osteoblasts, and myocytes in vitro [8, 11, 12]. Moreover, after transplantation into skeletal muscle of dystrophin-deficient (mdx) mice, they are able to restore long-term expression of dystrophin [13]. Thus, hMADS cells provide a powerful system for studying commitment and differentiation toward various lineages, as well as for cell therapy.

In addition to plasticity, self-renewal is a crucial feature of stem cells. It is defined by the ability to proliferate while maintaining an undifferentiated phenotype. The importance of self-renewal is best illustrated in vivo for hematopoietic stem cells (HSCs), which are able to give rise to both HSC and differentiated cells. This enables HSCs to perpetuate themselves while retaining the ability to regenerate mature blood cells throughout the entire lifespan of an organism [14]. In vitro, self-renewal has been defined by the ability of a cell to be maintained for many passages with no alterations of its developmental potential. It is still uncertain, however, how self-renewal is maintained in vivo and in vitro. The first clues have come from mouse embryonic stem (ES) cells, which appear to require extracellular stimuli such as leukemia inhibitory factor and bone morphogenic protein 4 to self-renew [15, [16]17]. A few intrinsic mediators of pluripotency have been identified in these cells, including Oct4, Sox2, and Nanog [16, [17]18]. Whether the mechanisms involved in self-renewal in embryonic versus adult stem cells are conserved is not known. A few signaling pathways have been shown to be involved in the regulation of adult stem cell self-renewal. Whereas the Notch and Wnt pathways participate in the self-renewal of hematopoietic, intestinal epithelial, and neural stem cells [19, 20], Hedgehog may play a role in the maintenance of neural stem cells in several regions of the mammalian central nervous system [21, 22]. Previous reports have identified the Wnt pathway as a promoter of proliferation and inhibitor of osteogenesis in mesenchymal stem cells [23, [24], [25]26]. However, no clear data demonstrate its role in self-renewal of these cells. Thus, the molecular mechanisms responsible for self-renewal of mesenchymal stem cells remain unclear at present. As in vitro expansion of mesenchymal stem cells is necessary for subsequent engraftment, it is crucial to identify intrinsic and extrinsic factors that contribute to their propagation.

Among the large number of intrinsic and extrinsic factors that could affect self-renewal of mesenchymal stem cells, we focused our attention on the fibroblast growth factor (FGF) pathway. This pathway, and especially FGF2, has been identified as a major candidate for self-renewal regulation in human ES cells [27, [28]29]. FGF2 may also be important for maintaining the neural stem cells pool in the mouse brain subventricular zone [30]. Moreover, FGF2 increases lifespan of bone marrow stromal cell primary cultures when cultivated at low cell density [31] and has been reported to support proliferation as well as the osteogenic and chondrogenic differentiation potential of these cells [32, [33], [34]35]. Finally, a transcriptome screening of undifferentiated versus differentiated hMADS cells revealed that this factor is expressed by undifferentiated cells but not by their differentiated derivatives, which suggests that it might be involved in self-renewal (unpublished data). In addition to these proliferative properties, FGF2 was shown to mediate osteoblast, chondrocyte, and neuron differentiation [36, [37]38], as well as endothelial cell migration [39]. Thus, its role seems to be context-dependent, and additional investigations are required to clearly identify the role of FGF2 in mesenchymal stem cells.

FGF2 belongs to the 22-member family of fibroblast growth factors. Five isoforms of FGF2 have been characterized, representing alternative translation products from a single mRNA. The 22-, 22.5-, 24-, and 34-kDa isoforms have been found to localize in the nucleus and trigger an active intracrine signaling pathway, whereas the AUG-initiated isoform of 18 kDa is mostly cytosolic [40, 41]. FGF2 binds to two classes of receptors: high-affinity transmembrane receptor tyrosine kinases (fibroblast growth factor receptors [FGFRs]) and low-affinity receptors, which are heparan sulfate proteoglycans (HSPGs). The FGFRs belong to a family of five genes (FGFR1–FGFR5), from which alternative splicings generate several isoforms. HSPGs not only are involved in FGF2 storage and protection from proteolysis but also provide a higher affinity of FGF2 for FGFRs [42, 43].

In this study, we investigated the role of FGF2 in hMADS cell self-renewal capacity in vitro. We assessed self-renewal by measuring the proliferation rate, the ability to grow at the single-cell level, and the differentiation potential of hMADS cells. The data presented here clearly identify FGF2 as a key factor in the maintenance of hMADS cell self-renewal and highlight some of the transduction pathways involved in this process.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Cell Culture

hMADS cells were obtained from the stroma of human adipose tissue as described previously [13]. The cell populations that have been extensively studied were isolated from the pubic region fat pad of a 5-year-old male donor (hMADS2) and a 4-month-old male donor (hMADS3) [13]. Proliferation medium was composed of Dulbecco's modified Eagle's medium (low glucose) containing 10% fetal calf serum (FCS), and 100 U/ml penicillin and streptomycin. After reaching 90% confluence, adherent cells were dissociated in 0.25% trypsin EDTA and seeded at 4,500 cells per cm2. PD173074 was used to inhibit FGFR signaling. Final concentration (75 nM) was determined as being the lowest concentration that inhibits the proliferative effect of 2 ng/ml of FGF2. At that concentration, no toxicity and no effect on epidermal growth factor (EGF) and platelet-derived growth factor (PDGF) pathways were observed (data not shown). Cultures were maintained at 37°C in a humidified gassed incubator, 5% CO2 in air. Media were changed every other day.

hMADS Cell Differentiation

Adipocyte differentiation were performed as described previously [13]. Cells were plated at high density (15,000 cells per cm2), and differentiation was induced 24 hours after plating. Glycerol-3-phosphate dehydrogenase (GPDH) activity was performed in triplicate wells, using the method described previously [44]. GPDH is an enzyme that is required for the formation of triglycerides. Oil red O staining was performed as described previously [45].

Cell Proliferation Assays

hMADS cells were plated into 100-mm diameter dishes (2.5 × 105 cells per dish). After the appropriate time, cells were trypsinized as described above and counted with a Coulter counter.

Clonal Assays

Cells were plated at a density of 10 cells per cm2 in 100-mm2 dishes. Fifteen days after plating, cells were fixed with 0.25% glutaraldehyde and stained with 0.1% crystal violet. Colonies containing at least 40 cells were counted under a light microscope. Medium was changed three times a week.

Reverse Transcription-Polymerase Chain Reaction Analysis

Total RNA was extracted using TRI-Reagent kit (Euromedex, Souffelweyersheim, France, http://www.euromedex.com) according to the manufacturer's instructions, and reverse transcription-polymerase chain reaction (RT-PCR) analysis was conducted as described previously [11]. All primers sequences are detailed in supplemental online Table 1. An aliquot of PCR products was analyzed on 2% ethidium bromide-stained agarose. For quantitative PCR, final reaction volume was 25 μl, including specific primers (0.4 μM), 5 ng of reverse-transcribed RNA, and 12.5 μl of SYBR green master mix (Eurogentec, Angers, France, http://www.eurogentec.com). Quantitative PCR conditions were as follows: 2 minutes at 50°C; 10 minutes at 95°C; and 35 cycles of 15 seconds at 95°C, 1 minute at 60°C. Real-time PCR assays were run on an ABI Prism 7700 real-time PCR machine (PerkinElmer Life and Analytical Sciences, Boston, http://www.perkinelmer.com).

Enzyme-Linked Solid Phase Immunosorbent Assay

Quantikine enzyme-linked solid phase immunosorbent assay (ELISA) kit (R&D Systems Inc., Minneapolis, http://www.rndsystems.com) was used to detect FGF2 in cell culture supernatants, fetal calf serum, and cell surface washes. Cells were washed with 2 M NaCl (20 mM HEPES, pH 7.4) for 5 seconds to remove HSPG-bound FGF2.

Preparation of Cell Extracts and Western Blot Analysis

Whole cell extracts, SDS-polyacrylamide gel electrophoresis, blotting, and enhanced chemiluminescence were performed as described previously [45]. Primary antibodies were goat anti-FGF2 (R&D Systems), mouse anti-phospho-ERK1/2, and rabbit anti-ERK1/2 (Cell Signaling Technology, Beverly, MA, http://www.cellsignal.com). Secondary horseradish peroxidase-conjugated antibody was purchased from Promega (Madison, WI, http://www.promega.com).

Materials and Chemicals

Cell culture media and reagents were purchased from Gibco-BRL (Gaithersburg, MD, http://www.gibcobrl.com), and FCS was purchased from Dutscher S.A. (Brumath, France, http://www.dutscher.com). FGF2 was from Peprotech (Rocky Hill, NJ, http://www.peprotech.com). Rosiglitazone (BRL4953) was a gift from Dr J.F. Dole (GlaxoSmithKline, King of Prussia, PA, http://www.gsk.com). PD173074 was a gift from GlaxoSmithKline Research and Development Ltd. (Hertfordshire, U.K.). U0126 was purchased from Sigma-Aldrich (St. Louis, http://www.sigmaaldrich.com).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

hMADS Cells Are Dependent on FGF2 Supplementation for Long-Term Propagation in Culture

Four populations of hMADS cells were isolated from different donors (hMADS1, hMADS2, hMADS3, and hMADS6). Two of them (hMADS2 and hMADS3) were extensively studied and yielded similar results. They were shown to be able to proliferate in culture for more than 30 passages, which represents approximately 150 population doublings, without reaching senescence [13]. However, during propagation in vitro, we noticed that the cell morphology changed dramatically, from spindle-shaped cells to large and flat cells (Fig. 1A). This morphological change was accompanied by a change in cell proliferation ability. Two major stages of cell proliferation could be identified: a “fast-cycling” stage, which extended until population doubling 70–80, followed by a “slow-cycling” stage. Fast-cycling cells exhibited an average population doubling time of 36 hours, whereas slow-cycling cells only divided every 72–96 hours (Fig. 1B), despite the presence of 10% FCS in culture medium. Thus, although FCS contains a broad range of growth factors, it was not sufficient to maintain hMADS cells in a fast-cycling state.

thumbnail image

Figure Figure 1.. Overview of the change in proliferation status and morphology of hMADS cells as the number of population doublings increases. (A): Light microscopy photographs of fast- and slow-cycling human multipotent adipose-derived stem cells stained with crystal violet. Scale bars = 75 μm. (B): Average population doubling time of fast- and slow-cycling cells. ∗, p < .05.

Download figure to PowerPoint

In an effort to find culture conditions to maintain hMADS cells in a fast-cycling state, growth factors such as FGF2, EGF, and PDGF were added separately to serum-containing medium. FGF2 only was able to sustain proliferation over several passages (data not shown). It is noteworthy that FCS did not contain FGF2, as checked by ELISA (data not shown). As shown in Figure 2A, FGF2 was able to maintain hMADS cells in a fast-cycling state until at least passage 25. During this time course, hMADS cells retained their initial spindle-shaped morphology (Fig. 2B).

thumbnail image

Figure Figure 2.. Effect of FGF2 on proliferation and morphology of human multipotent adipose-derived stem (hMADS) cells during in vitro propagation. (A): Comparative evolution of doubling times of hMADS cells cultured with or without 2 ng/ml FGF2. (B): Light microscopy photographs of hMADS cell maintained with or without 2 ng/ml FGF2. Scale bars = 350 μm. Abbreviation: FGF2, fibroblast growth factor 2.

Download figure to PowerPoint

FGF2 Allows hMADS Cells to Retain Their Differentiation Potential During In Vitro Expansion

The maintenance of a differentiation potential during propagation is a critical component of self-renewal. To assess hMADS cell differentiation potential during in vitro propagation, we focused on their ability to differentiate into their three major derivatives (e.g., adipocytes, osteoblasts, and chondrocytes).

As illustrated in Figure 3A, the ability of hMADS cells from two different donors to differentiate into adipocytes was decreased in slow-cycling cells compared with fast-cycling cells. However, treatment of slow-cycling cells with FGF2, exclusively during propagation, completely restored their adipogenic potential, as monitored by GPDH activity (Fig. 3B; supplemental online Fig. 1).

thumbnail image

Figure Figure 3.. Effect of FGF2 on differentiation potential of fast- and slow-cycling human multipotent adipose-derived stem (hMADS) cells. Fast- and slow-cycling cells were maintained 5 days with or without (control) 2 ng/ml FGF2 during propagation. Then, FGF2 was removed, and 36 hours later, cells were induced to undergo differentiation into adipocytes. (A): Light microscopy photographs of fast- (passage 15) and slow-cycling (passage 25) hMADS cells that were maintained in the absence of FGF2 and then subjected to adipogenesis for 10 days. Lipid droplets accumulation was revealed by oil red O staining. Scale bars = 350 μm. (B): Adipogenesis, quantified by GPDH activity measurement, of fast- and slow-cycling cells pretreated with FGF2 or not pretreated. Columns indicate mean ± SE (n = 3). ∗, p < .05; ∗∗, p < .01. Abbreviations: FGF2, fibroblast growth factor 2; GPDH, glycerol-3-phosphate dehydrogenase.

Download figure to PowerPoint

Similar results were obtained when hMADS cells were induced to differentiate into osteoblasts and chondrocytes (supplemental online Fig. 2): differentiation potential decreased in slow-cycling cells but could be maintained when these cells were cultivated with FGF2. Together, these data indicate that FGF2 prevents the loss of differentiation capacity observed during hMADS cell long-term propagation.

FGF2 Allows hMADS Cells to Maintain Their Ability to Propagate at the Single-Cell Level

The ability to expand at the single-cell level in vitro is a major feature of self-renewing cells. The clonogenic potential of hMADS cells was assessed at different propagation time points. Fast-cycling cells were able to form clones with an efficiency of (11.6% ± 0.08%), whereas slow-cycling cells exhibited lower clonogenic efficiency (5.7% ± 0.04%). Addition of FGF2 to slow-cycling cells almost completely restored their clonogenic potential, raising the clonogenic efficiency to 10.1% ± 0.27 (Fig. 4). Therefore, FGF2 improves the ability of long-term cultured cells to proliferate at the single-cell level.

thumbnail image

Figure Figure 4.. Effect of FGF2 on human multipotent adipose-derived stem (hMADS) cell clonogenicity. Fast- (passage 15) and slow-cycling (passage 25) cells were seeded at low density (10 cells per cm2) and maintained in proliferation medium containing or not containing FGF2 (2 ng/ml). After 15 days, colonies containing more than 40 cells were scored. Results are expressed as the ratio of the number of scored colonies over the initial number of seeded cells. Similar clonogenicity measurements were performed on hMADS cells from three other donors and yielded similar results. Columns indicate mean ± SE (n = 3). ∗∗, p < .01. Abbreviation: FGF2, fibroblast growth factor 2.

Download figure to PowerPoint

Together, these data indicate that fast-cycling hMADS cells display the main characteristics of self-renewing multipotent stem cells. They lose this property and turn into slow-cycling cells during propagation in culture. Nevertheless, FGF2 supplementation prevents this process, allowing hMADS cells to be propagated at a multipotent state.

Acquisition of Slow-Cycling State Involves Irreversible Decrease in FGF2 Expression

To explain why slow-cycling cells required exogenous FGF2 to self-renew whereas fast-cycling cells did not, we hypothesized that the latter may initially express and secrete FGF2, whereas slow-cycling cells may have lost this ability. We therefore investigated the expression of FGF2 protein in both types. As illustrated in Figure 5A, slow-cycling cells displayed a marked decrease in the expression of FGF2 (isoforms of 18, 22/22.5, and 24 kDa) compared with fast-cycling cells. It is worth mentioning that FGF2 treatment of hMADS cells allowed the maintenance of their self-renewal abilities but did not sustain their FGF2 expression (Fig. 5A). This suggests that unlike the decrease in proliferation and differentiation, the decrease in FGF2 expression became irreversible in slow-cycling cells.

thumbnail image

Figure Figure 5.. Expression of FGF2 and FGFR1 by human multipotent adipose-derived stem (hMADS) cells. (A): Western blot for FGF2. Fifty micrograms of hMADS cells protein extracts of fast- (passage 15) and slow-cycling (passage 25) cells were resolved on a 15% polyacrylamide gel, transferred, and blotted with anti-FGF2 antibody. The right lane corresponds to cells treated with FGF2 from passages 15–25. (B): Screening for FGFR expression was performed by conventional reverse transcription-polymerase chain reaction. (C): Secretion of FGF2 by fast- and slow-cycling hMADS cells was assessed by enzyme-linked immunosorbent assay. One hundred microliters of 48-hour conditioned media and 100 μl of 2 M NaCl 20 mM HEPES (pH 7.4) washes were assayed according to the manufacturer's instructions. Columns indicate mean ± SE (n = 3). Abbreviations: FGF2, fibroblast growth factor 2; FGFR, fibroblast growth factor receptor; kD, kilodaltons.

Download figure to PowerPoint

We next checked whether fast-cycling cells express FGFRs by RT-PCR. Only one receptor, FGFR1, was detected (Fig. 5B). Quantitative PCR analysis showed that this receptor was equally expressed in fast- and slow-cycling cells (supplemental online Fig. 2).

The expression of FGF2 and FGFR1 by hMADS cells suggests the existence of an active autocrine/paracrine loop. Therefore, we checked for the secretion of FGF2 by hMADS cells by performing an ELISA on conditioned media from fast- and slow-cycling cells. As shown in Figure 5C, no detectable levels of FGF2 were found in any conditioned medium. This could reflect an association of FGF2 to cell surface, presumably to HSPG, rather than a failure to export FGF2 through the plasma membrane [46]. To remove putative interactions between HSPGs and FGF2, cell layers were washed with 2 M NaCl at neutral pH. Significant amounts of FGF2 were detected in washes from fast-cycling cells, whereas only very low amounts of FGF2 were found in those from slow-cycling cells (Fig. 5C). Together, these results show that hMADS cells express a FGF2 autocrine signaling loop that irreversibly declines as the cells are expanding. This decline results, at least in part, from a decrease in FGF2 production, which may explain why slow-cycling cells become dependent on FGF2 supplementation to maintain their self-renewal ability.

FGF2 Autocrine Signaling Is Necessary for hMADS Cell Self-Renewal During In Vitro Propagation

To demonstrate that the autocrine FGF2 signaling was necessary for their self-renewal, hMADS cells were cultured in the presence of PD173074, a specific inhibitor of FGF receptor phosphorylation [47], and then were assessed for their clonogenic and differentiation potential.

When seeded at clonal density in the presence of PD173074, fast-cycling cells exhibited a potent decrease in their clonogenic efficiency compared with nontreated cells (Fig. 6A). Thus, signaling through FGFR1 promotes hMADS cell expansion at the single-cell level. When fast-cycling cells were treated for one passage with PD173074 and then induced to differentiate, adipogenesis was strongly impaired compared with untreated cells (Fig. 6B). In conclusion, the inhibition of FGF2 signaling in fast-cycling cells impairs both their clonogenic potential and their differentiation abilities, indicating its crucial role in the regulation of hMADS cell self-renewal.

thumbnail image

Figure Figure 6.. Effect of PD173074 on clonogenicity and differentiation potential of fast-cycling (passage 15) human multipotent adipose-derived stem (hMADS) cells. (A): Effect of PD173074 on the clonogenicity of fast-cycling cells. Cells were treated with PD173074 (75 nM) or not treated during the clonal assay and then analyzed as described in Figure 4. (B): Effect of PD173074 treatment during proliferation on adipogenesis of fast-cycling hMADS cells. Cells were seeded and treated with PD173074 (75 nM) during 5 days and then processed as described in Figure 3. Columns indicate mean ± SE (n = 3). ∗∗, p < .01. Abbreviations: DMSO, dimethyl sulfoxide; GPDH, glycerol-3-phosphate dehydrogenase.

Download figure to PowerPoint

The ERK1/2 Pathway Is Partly Responsible for FGF2-Induced Self-Renewal

We next tried to identify signaling pathways that could mediate FGF2 effects on hMADS cell self-renewal. Because in many cell types the ERK1/2 pathway is known to be activated by FGF2, we investigated its role in hMADS cell self-renewal.

Addition of FGF2 to slow-cycling hMADS cells induced ERK1 and ERK2 phosphorylation. This activation was first observed after a 10-minute exposure, and the levels of phosphorylated ERK1/2 started to decrease after 24 hours (Fig. 7A).

thumbnail image

Figure Figure 7.. Involvement of the ERK1/2 pathway in FGF2-mediated self-renewal. (A): Western blot against phospho-ERK1/2 and ERK1/2 on FGF2 stimulated slow-cycling human multipotent adipose-derived stem (hMADS) cell (passage 25) extracts. Thirty micrograms of total extracts were resolved on a 12% polyacrylamide gel and then blotted with anti-phospho-ERK1/2 and anti-ERK1/2 antibodies. (B): Cumulative number of slow-cycling hMADS cells (passage 25) propagated for three passages under the following conditions: control (open circles), 2 ng/ml FGF2 (filled circles), 10 μM U0126 (open squares), or 2 ng/ml FGF2 and 10 μM U0126 (filled squares). (C): Clonogenicity efficiency of slow-cycling hMADS cells (passage 25) treated either with 2 ng/ml FGF2 or simultaneously with 10 μM U0126 and 2 ng/ml FGF2. Columns indicate mean ± SE (n = 3). ∗∗, p < .01. (D): Light microscopy photographs of slow-cycling hMADS cells (passage 25) that were treated either with 2 ng/ml FGF2 or simultaneously with 10 μM U0126 and FGF2 during propagation (5 days) and then processed as described in Figure 3. Lipid droplet accumulation was revealed by oil red O staining. Scale bars = 350 μm. Abbreviations: ERK, extracellular signal-related kinase; FGF2, fibroblast growth factor 2.

Download figure to PowerPoint

To assess whether the activation of the ERK1/2 pathway was necessary for the effect of FGF2 on hMADS cell self-renewal, we evaluated the effect of U0126 on FGF2-mediated proliferation, clonogenicity, and differentiation potential. U0126 specifically inhibits mitogen-activated protein kinase kinase 1, which is the mitogen-activated protein kinase kinase that activates ERK1/2.

As illustrated in Figure 7B, treatment of slow-cycling hMADS cells with U0126 alone did not significantly alter their proliferation rate in the absence of FGF2. In contrast, FGF2-mediated increase in proliferation was counteracted in the presence of U0126. Furthermore, the clonogenic potential of slow-cycling hMADS cells was also impaired by simultaneous treatment with FGF2 and U0126 (Fig. 7C). Thus, these results suggest that FGF2 promotes proliferation and clonogenicity of hMADS cells, at least in part, by activating the ERK1/2 pathway.

Finally, we evaluated the effect of U0126 on the maintenance of adipocyte differentiation potential mediated by FGF2. Slow-cycling hMADS cells were treated with FGF2 in the absence or presence of U0126 for two passages and then were induced to differentiate. The results shown in Figure 7D indicate that U0126 treatment during hMADS cell proliferation did not affect FGF2-mediated maintenance of differentiation potential. These results were confirmed by GPDH activity assay (data not shown). In conclusion, these data suggest that the ERK1/2 pathway is partly responsible for FGF2-mediated self-renewal of hMADS cells, as it is involved in proliferation and clonogenic efficiency but not in the maintenance of differentiation potential.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Given their convenient isolation and extensive proliferative capacities in vitro, adipose tissue-derived stem cells are a promising source of human stem cells for regenerative medicine. Nevertheless, little characterization of their self-renewal properties has been achieved so far. Indeed, to generate sufficient amounts of stem cells for transplantation, one should be able to expand them in vitro without altering their developmental potential. This is a major difficulty with human bone marrow stem cell expansion. These cells are plastic at early passages but lose their multipotency as they are propagated in culture. For instance, they seem to lose their adipogenic potential upon reaching passage 12 in culture [48]. We also observe a progressive loss of differentiation potential during expansion of adipose tissue-derived stem cells in serum-containing medium. We show herein that a decrease in FGF2 signaling is responsible not only for this decline in differentiation capacity but also for a reduced ability of the cells to grow at the single-cell level. Exogenous FGF2 restored both of these properties, indicating that FGF2 plays a crucial role in self-renewal of adipose tissue-derived stem cells during propagation in vitro.

To our knowledge, there has been only one recent report dealing with the expansion of bone marrow mesenchymal stem cells in fully chemically defined conditions [49], which would be best suited for subsequent therapeutic use of the cells. Still, expansion of adipose tissue-derived mesenchymal stem cells under these kinds of conditions has never been described. The media that are currently used contain serum at high concentrations, which includes a broad range of growth factors. Our data presented here indicate that the widely used 10% FCS-containing media may not be optimal for long-term propagation of mesenchymal stem cells. Indeed, adipose tissue-derived stem cells cultured under these conditions progressively lose FGF2 expression, resulting in an impairment of their self-renewal ability. Whether long-term exposure to high serum concentration itself is responsible for the decrease in FGF2 expression is unclear at present. Nevertheless, exposure to high serum concentrations has previously been reported to impair propagation of primary rodent cells, including Schwann cells and oligodendrocyte precursor cells [50, 51]. In these cell types, impairment of propagation was attributed to telomere-independent premature senescence. In contrast, the progressive loss of self-renewal ability of adipose tissue-derived stem cells during propagation does not seem to be caused by senescence, since the cells display no senescence-associated β-galactosidase activity [13] and since this process is fully reversible upon addition of FGF2. However, even though hMADS cells are able to retain their stem cells features during extensive propagation in medium containing high serum concentrations in the presence of FGF2, it would be worth optimizing fully chemically defined conditions allowing efficient propagation of undifferentiated mesenchymal stem cells. Defined conditions would allow convenient identification of factors involved in self-renewal and would probably cause fewer damages to the cells than high serum exposure during long-term propagation. Preliminary data suggest that hMADS cells can be propagated in low-serum conditions and that FGF2 is necessary for maintenance of their self-renewal abilities under these conditions (data not shown).

Self-renewal implies the coordination of two features: proliferation and maintenance of differentiation capacities. These two processes are not necessarily regulated by a common pathway. In the hematopoietic system, for instance, HOXB4 appears to regulate proliferation of HSCs, whereas the Notch pathway regulates their differentiation potential [52]. Our data concerning the ERK1/2 pathway suggest that FGF2 controls proliferation and differentiation of adipose tissue-derived stem cells through distinct downstream effectors. Indeed, we showed that the ERK1/2 pathway is involved in FGF2-mediated clonogenicity of hMADS cells but not in the maintenance of their differentiation potential. Moreover, the observation that exogenous FGF2 does not improve clonogenicity of fast-cycling cells, whereas it increases their differentiation potential (Figs. 3B, 4), corroborates the hypothesis that distinct transduction pathways act to promote clonogenicity versus maintenance of the differentiation potential. The respective pathways and targets genes responsible for these two features remain to be elucidated.

In conclusion, our study identifies for the first time FGF2 as a key factor for long-term propagation and self-renewal of human adipose tissue-derived stem cells. If the understanding of the self-renewal of adipose tissue-derived stem cells in vitro is crucial for therapeutic use of these cells, one should also consider the importance of this process in vivo. As these cells are isolated from adipose tissue and are able to differentiate into adipocytes, they might constitute a reservoir for adipocyte precursors. This pool of precursor cells may be responsible for fat mass enlargement according to their ability to proliferate and then to differentiate after stimulation by the appropriate signals [53]. It has been reported that FGF2 is expressed in human adipose tissue [54]. Since this factor is able to enhance proliferation and to maintain adipogenic potential of hMADS cells, it could play an important role in physiological and pathological modifications of adipose tissue, such as in aging and obesity.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

We thank the members of the Stem Cells and Differentiation and Biology of Adipose Tissue groups for helpful discussions. Special thanks are due to N. Billon for helpful suggestions and critical reading of the manuscript. This work was supported in part by Centre National de la Recherche Scientifique and Association pour la Recherche sur le Cancer funds (Grant 3721). Z.L.E. is supported by a fellowship from Yves-Saint Laurent Beauté. D.C. is an Institut National de la Santé et de la Recherche Médicale established investigator.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
daniR1_Supplemental_Figure_1.pdf94KSupplemental Figure 1
daniR1_Supplemental_Figure_2.pdf297KSupplemental Figure 2
daniR1_Supplemental_Figure_3.pdf116KSupplemental Figure 3
Supplemental_Legends.pdf15KSupplemental Legend

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.