Rapid and Efficient Production of Human Functional Mast Cells through a Three-Dimensional Culture of Adipose Tissue–Derived Stromal Vascular Cells

Mast cells (MC) are immune cells located in virtually all organs at strategic locations close to host–environment interfaces, such as skin, vascular and mucosal barriers, or nerve endings (1). MC are easily distinguishable from other cells by their high content of basophilic and electronic dense secretory granules (2) containing a huge array of molecules such as histamine, serotonin, cytokines, or serine proteases. Besides these preformed mediators, MC also release de novo synthetized cytokines, chemokines, and eicosanoids. Via these mediators, MC are involved in incalculable biological processes such as allergy, vascular permeability control, pain, or tumor progression (2–6).

However, studies of MC biology are limited because of lack of reproducible in vitro models, especially in humans. Indeed, MC cell lines such as HMC-1, LUVA or LAD-2 do not behave like in vivo MC (7–9). HMC-1 cells do not possess the IgE receptor (FCεRI) (9), and immortalized LUVA cells do not respond to stem cell factor (SCF), which is required for the normal development of MC in vitro (8). To date, most in vitro studies involving MC are based on CD34⁺ cells isolated from the bone marrow, cord blood, or peripheral blood and differentiated into functional MC (10). However, this method is a long lasting process that usually requires ∼10 wk to obtain pure mature MC. Primary human MC can also be obtained via their direct isolation from tissues such as skin (11). However, this method is limited by the scarcity of mature MC and the difficulties in obtaining fresh human tissues. The development of in vitro models that produce easily and readily high numbers of mature MC is therefore required for a better understanding of MC functions in health and diseases.

White adipose tissue (WAT) is an easy access tissue which is present throughout the human body and represents around 20% of the body weight in healthy humans. We and others have previously shown that in mice and in humans, WAT contains progenitors and/or mature MC (12, 13). MC accumulation has been described in the WAT in obesity and food allergy, but their roles are still debated (13–16).

In this study, we present an efficient method to produce mature and functional human connective tissue–type MC (MC_TC) within 3 wk from the stromal vascular fraction (SVF) of WAT. This method produces a homogeneous MC population expressing typical MC peptidases (tryptase, chymase, and carboxypeptidase A3) with good yield and high purity. These WAT-MC are able to degranulate upon FcεRI aggregation or following stimulation with C5a, substance P (SP), or compound 48/80. WAT-MC exhibit the same phenotype than MC derived from peripheral blood CD34⁺ cells, and both release, upon activation by anti-IgE, identical eicosanoids and proinflammatory cytokines (TNF-α, IL-6, GM-CSF).

Abs used are listed in Supplemental Table I.

Twenty-four patients were used in this study. After informed consent obtained from the patients, s.c. WAT was obtained from the Department of Plastic Surgery at Rangueil University Toulouse Hospital and processed rapidly after surgery. This study was approved by the Institutional Review Board, Rangueil University Toulouse Hospital, and was performed in accordance with the principles of the Declaration of Helsinki (1964) and in agreement with French bioethics laws of July 7th, 2011. SVF was obtained as previously described (17). Briefly, 10 to 50 g of WAT was digested with collagenase A (Serva Electrophoresis, Heidelberg, Germany) under agitation for 45 min at 37°C. The cell suspension was filtered through a 100-μm cell strainer (Becton Dickinson, Franklin Lakes, NJ) and centrifuged 10 min at 600 × g. The pellet was incubated in hemolysis buffer (155 mM NH₄Cl and 20 mM Tris, pH 7.6) to eliminate RBCs. Cells of the SVF were then centrifuged 10 min at 600 × g and counted, and their viability was assessed by trypan blue exclusion.

SVF cells were plated in StemSpan SFEM II (STEMCELL Technologies, Vancouver, Canada) supplemented with penicillin 100 μg/ml, streptomycin 100 μg/ml, amphotericin 0.25 μg/ml (Life Technologies BRL, Grand Island, NY), 100 ng/ml human SCF (hSCF) (PeproTech, Rocky Hill, NJ), and 2% methylcellulose (Sigma-Aldrich, Saint-Louis, MO) (18), at 1.7 × 10⁵ cells per well in ultra-low attachment 96-well plates (Corning Costar, Corning, NY). One week later, spheroids were discarded and nonaggregated cells were kept in culture. Half of the medium was changed once a week.

To test the combination of SCF, IL-3, and IL-6 on MC growth and purity, human SVF cells were cultured after isolation in medium containing 100 ng/ml hSCF and 30 ng/ml human IL-3 (PeproTech) for 1 wk and then in medium supplemented with 100 ng/ml hSCF and 100 ng/ml human IL-6 (PeproTech). To test the effect of IL-4 on FcεR1 cell surface expression, 100 ng/ml IL-4 (PeproTech) was added to the medium for 1 wk after 3 wk of culture.

MC derived from peripheral blood CD34⁺ cells were obtained as previously described (19). Briefly, CD34⁺ cells were isolated from buffy coats (Etablissement français du sang) and grown in StemSpan medium (STEMCELL Technologies) supplemented with recombinant human IL-6 (50 ng/ml; PeproTech), human IL-3 (10 ng/ml; PeproTech), and 3% supernatant of Chinese hamster ovary transfectants secreting murine SCF (a gift from Dr. P. Dubreuil, Marseille, France; 3% corresponds to ∼50 ng/ml SCF) for 1 wk. Cells were then grown in IMDM Glutamax I, sodium pyruvate, 2-ME, 0.5% BSA, Insulin-Transferrin-Selenium (all from Invitrogen), ciprofloxacin (10 μg/ml; Sigma-Aldrich), IL-6 (50 ng/ml), and 3% supernatant of Chinese hamster ovary transfectants secreting murine SCF.

A total of 5 × 10⁴ cells were cytocentrifuged onto glass slides and air dried. Slides were incubated in sulfuric ether solution for 30 min, rinsed with tap water, and stained with toluidine blue (VWR, Allison Park, PA). Slides were then rinsed twice with propanol 60% for 15 s. After dehydration and air drying, slides were examined under a light microscope (DM3000; Leica).

A total of 5 × 10⁴ cells were cytocentrifuged onto Superfrost microscope slides and air dried. Slides were fixed with formalin 10% (Sigma), washed with PBS, and then incubated with PBS containing normal horse serum 2%, Triton X-100 0.2% for 1 h at room temperature. Slides were incubated overnight in a humidified chamber at 4°C with mAb against tryptase or with biotinylated mAb against chymase and then washed with PBS–Triton X-100 0.2%. For chymase detection, cytospins were incubated with streptavidin–Alexa 488 conjugate (Invitrogen, CA) and counterstained with DAPI. For tryptase immunostaining, OptiView DAB IHC Detection Kit (Roche, Switzerland) was used and nuclei were stained with hematoxylin. Observations were performed under a light microscope (DM3000; Leica) or Nikon Eclipse TE2000-5 inverted fluorescent microscope.

Cultured cells were stained in PBS containing FcR-blocking reagent (Becton Dickinson). For intracellular detection of tryptase and chymase, cells were first permeabilized with the Cytofix/Cytoperm kit (BD Biosciences, Heidelberg, Germany) according to the manufacturer’s instructions. Staining was performed by direct immunofluorescence with conjugated mouse anti–human mAbs, as compared with their matched isotype control mAb. Cell viability was assessed by DAPI labeling. Cells were then washed in PBS and analyzed on a fluorescence-activated cell sorter LSR Fortessa (BD Biosciences). Data were acquired with FACSDiva software (BD Biosciences) and analyzed with Kaluza 1.5a software (Beckman Coulter, Roissy Charles de Gaulle, France).

MC degranulation was determined by measuring the release of β-hexosaminidase as previously described (20). MC were sensitized with human IgE for 16 h; the cells were then washed and distributed in 96-well flat-bottom plates at a density of 1 × 10⁵ cells in 50 ml of Tyrode’s buffer. The cells were adapted to 37°C for 20 min and then treated with prewarmed of the indicated stimuli: PMA plus ionomycin (respectively, 0.1 and 1 μg/ml) (Sigma-Aldrich), SP (50 μm), compound 48/80 (50 μg/ml) (Sigma), C5a (50 ng/ml) (R&D Systems, Minneapolis, MN), or anti-human IgE (Santa Cruz Biotechnology, Dallas, TX) diluted in Tyrode’s buffer for 45 min at 37°C. The release of β-hexosaminidase in the supernatants was measured by means of enzymatic colorimetric analysis with p-nitrophenyl N-acetyl-b-d-glucosamine (Sigma). To measure lipid mediators and cytokines production, cells were stimulated overnight at 37°C in medium as indicated above and cell supernatants were kept for measurements.

Supernatant lipid extraction was performed by adding 5 μl of internal standard mixture (LxA₄-d5 and LTB₄-d4 at 400 ng ml⁻¹ in MeOH) and 300 μl of cold methanol to 500 μl of cell supernatants. Samples were centrifuged at 3000 rpm for 15 min at 4°C. Supernatants were collected and submitted to solid-phase extraction using HRX-50 mg 96-well (Macherey Nagel, Hoerd, France) as previously described (21). Solvent was evaporated under nitrogen and samples were dissolved with MeOH.

Liquid chromatography/tandem mass spectrometry measurement was used to quantify arachidonic acid metabolites: 6kPGF_1α, TXB₂, PGE₂, PGD₂, 8-isoPGA₂, 11β-PGF2α, PGF2α, LxA₄, LxB₄, LTB₄, 5,6-DiHETE, 15-HETE, 12-HETE, 8-HETE, 5-HETE, 14,15-EET, 11,12-EET, 8,9-EET, 5,6-EET, and 5-oxo-ETE (21). To simultaneously separate 19 lipids of interest and three deuterated internal standards, liquid chromatography/tandem mass spectrometry analysis was performed on ultra-high performance liquid chromatography system (Agilent LC1290 Infinity) coupled to Agilent 6460 Triple Quadrupole MS (Agilent Technologies) equipped with electrospray ionization operating in negative mode. Reverse-phase ultra-high performance liquid chromatography system was performed using ZorBAX SB-C18 column (Agilent Technologies) with a gradient elution. The mobile phases consisted of water, acetonitrile, and formic acid (75:25:0.1; v/v/v) (A) and acetonitrile, formic acid (100:0.1, v/v) (B). The linear gradient was as follows: 0% B at 0 min, 85% B at 8.5 min, 100% B at 9.5 min, 100% B at 10.5 min and 0% B at 12 min. The flow rate was 0.35 ml/min. The autosampler was set at 5°C and the injection volume was 5 μl. Data were acquired in multiple reaction monitoring mode with optimized conditions. Peak detection, integration, and quantitative analyses were done using MassHunter Quantitative analysis software (Agilent Technologies). For each standard, calibration curves were built using 10 solutions at concentrations ranging from 0.95 to 500 ng/ml. A linear regression with a weight factor of 1/X was applied for each compound. The limit of detection and the limit of quantification (LOQ) were determined for the 27 compounds using signal to noise ratio. The limit of detection corresponded to the lowest concentration leading to a signal to noise over three, and LOQ corresponded to the lowest concentration leading to a signal to noise over 10. No values under the LOQ were considered. Importantly, blank samples were evaluated, and their injection showed no interference (no peak detected) during the analysis.

Cytokines (IL-6, TNF-α, GM-CSF) were measured by a multiplex immunoassay (BioLegend, San Diego, CA) according to the manufacturer’s instructions. Data were acquired on a BD FACSCanto, and analysis was performed using LEGENDplex software (BioLegend).

To quantify intracellular tryptase and chymase amounts, 1 × 10⁶ MC were disrupted with 100 μl 1% Triton X-100 in PBS, and the cell lysates were immediately processed. Enzymatically active tryptase was measured spectrophotometrically by adding 50 μl of 0.5 mmol/l Tosyl-Gly-Pro-Lys-p-nitroanilide (T6140; Sigma-Aldrich) 100 mmol/l Tris-HCl (pH 8), and 100 mmol/l NaCI to 50 μl of cell lysate. Enzymatically active chymase was quantified by adding 50 μl of 1 mmol/l Succinyl-Ala-Ala-Pro-Phe-p-nitroanilide (S7388; Sigma-Aldrich), 300 mmol/l Tris (pH 8), and 1.5 mol/l NaCl to 50 μl of cell lysate. Changes in absorbances at 405 nm were immediately registered for 5 min at 30 s intervals at room temperature.

All results are expressed as mean ± SEM. Statistical analyses were performed using GraphPad Prism 5 software, and Student t test was applied.

Culturing SVF of human s.c. WAT in ultra-low adherent plates and in serum-free medium supplemented in SCF led to the formation of spheroids between 5 and 7 d after seeding (Fig. 1A). As soon as spheroids were formed, dispersed cells surrounding the spheroids were observed (Fig. 1A). These cells formed a homogenous population after 3 wk in culture. Analyses by flow cytometry showed that this homogenous cell population expressed the MC cell surface markers CD45, CD117, and FcεR1 (Fig. 1B, 1C). In addition, the presence of metachromatic cytoplasmic granules and round unilobular nucleus identified by toluidine blue staining on cytocentrifuged cells confirmed the MC nature of the cultured cell type (Fig. 1D). MC purity reached 91.8 ± 2.1% after 3 wk in culture and remained as high for the following 3 wk (Fig. 1E).

Altogether, these results show that three-dimensional culture of human SVF cells led to the production of highly pure MC in only 3 wk without any prior purification.

With this three-dimensional culture, 5.3 ± 3.2 × 10⁴ MC/g of WAT were obtained after 3 wk, and this yield doubled every week, reaching 4.8 ± 2.7 × 10⁵ MC/g of WAT after 6 wk (Fig. 2A). WAT-MC stopped expanding shortly after 6 wk and could be maintained in culture for 2 mo. Interestingly, the yield increased as long as the spheroids remained compact. Indeed, if spheroids were seeded in fresh medium after 1 wk, WAT-MC were observed de novo around spheroids. These MC are called G2 for second generation (Fig. 2B). As shown in Fig. 2C, G2 WAT-MC underwent a similar expansion compared toMC produced initially (G1 WAT-MC). We were able to generate up to four successive generations of MC, leading to a total production of 10⁶ MC/g of WAT after 6 wk.

To test whether cytokines classically added to expand CD34⁺ cell–derived MC were able to increase WAT-MC yield, cultures were performed with or without SCF (100 ng/ml), IL-6 (100 ng/ml), and/or IL-3 (30 ng/ml) for 5 wk. As expected, in the absence of SCF, no MC were obtained in vitro (data not shown), demonstrating that this cytokine is required for MC development, as previously described (22). The effects of IL-3 and IL-6 on WAT-MC yield were tested by using the protocol classically described for CD34⁺ cell–derived MC (10). SVF of human s.c. WAT was seeded in medium with SCF and IL-3 for the first week and then with SCF and IL-6 for the following weeks. Addition of IL-3 and IL-6 in the medium modified neither MC purity after 3 wk (Fig. 2D) nor WAT-MC yield evaluated after 5 wk (Fig. 2E).

These results demonstrate that SCF alone is required and sufficient for WAT-MC production in culture. In contrast, this cell culture system does not require IL-3 or IL-6, suggesting that WAT-MC produced in vitro differ from MC obtained from bone marrow CD34⁺ purified cells.

MC can be divided into two distinct populations based on tryptase and chymase expression. MC express either tryptase alone (mucosal-type MC [MC_T]) or tryptase and chymase MC_TC (23). As demonstrated by flow cytometry and immunofluorescence experiments, WAT-MC expressed typical MC_TC peptidases (i.e., tryptase, chymase, and carboxypeptidase A3) (Fig. 3A–C), and the mean fluorescence intensities relative to tryptase, chymase, and carboxypeptidase A3 were similar between WAT-MC and CD34⁺ cell–derived MC (Fig. 3A). Moreover, tryptase and chymase activity were similar between both MC types (Fig. 3D, 3E). WAT-MC also expressed MrgX2, a G protein coupled receptor specific for MC_TC and that recognizes several basic secretagogues (Fig. 3F) (24). In addition, treatment of WAT-MC with IL-4 (100 ng/ml) for 5 d led to an increase in cell surface expression of FcεRI (Fig. 3G), demonstrating that these cells behave as mature MC and not as progenitors (25, 26).

Taken together, these results indicate that WAT-MC produced in vitro are mature MC_TC.

WAT-MC functionality was characterized by the standard degranulation assay using Ag (DNP-HSA) after sensitization with IgE or SP, compound 48/80, or other polycationic molecules. WAT-MC, previously sensitized with IgE, showed a dose-dependent degranulation when stimulated with increasing concentrations of anti-IgE as detected by β-hexosaminidase release (Fig. 4A). MC_TC, in contrast to MC_T, can also be activated by basic compounds such as C3a, C5a, compound 48/80, or SP (25). Consistent with our previous results regarding the phenotype of cultured cells, stimulation of WAT-MC by C5a (50 ng/ml), SP (50 μm), or compound 48/80 (50 μg/ml) induced their degranulation (Fig. 4B).

Upon activation, MC release preformed and newly synthesized mediators. We thus quantified eicosanoids produced and released by WAT-MC after sensitization with IgE and stimulation with anti-IgE. Stimulated WAT-MC released PGs (PGE₂, PGD₂, 8-isoPGA₂, 15d-PGJ₂) (Fig. 4C). In contrast, stimulation of WAT-MC has no effect on HETE concentrations (data not shown) or other arachidonic metabolites, including LTB4 (Fig. 4C). Interestingly, the same pattern of arachidonic acid metabolites released after IgE sensitization and stimulation with anti-IgE was obtained with CD34⁺ cell–derived MC (Fig. 4D). Moreover, as shown in Fig. 4E and 4F, WAT-MC as well as CD34⁺ cell–derived MC produced TNF-α, IL-6, and GM-CSF following PMA/ionomycin activation or FcεRI aggregation.

Altogether, these results demonstrate that three-dimensional culture of SVF cells leads to the production of mature and functional human MC_TC within 3 wk.

In this study, we report a novel time-saving and reproducible method to produce highly pure human MC in 3 wk from WAT. WAT-MC produced with this method are mature and functional MC_TC. WAT-MC behave like MC_TC obtained from peripheral blood–derived CD34⁺ pluripotent progenitor cells, which is, to date, the gold standard strategy to generate fully mature and functional primary MC lines. Indeed, WAT-MC and CD34⁺ derived MC exhibit similar phenotypes and release similar cytokines and lipid mediators following FcεRI aggregation. The amount of lipid mediators and cytokines produced by both MC types were in the same range, although they cannot be compared because they were not obtained from the same donor.

The method described here present two major advantages as compared with methods using CD34⁺ progenitors. First, it is a time saving method. Indeed, we are able to generate an almost pure population of mature functional MC_TC in only 3 wk, whereas in most of studies, MC are generated after 9–14 wk in culture (10). In some protocols however, the culture is shorter, but in these cases MC purity is not high or MC obtained are a mix population between MC_T and MC_TC. For example, a protocol has been developed to generate MC in 3 wk using CD34⁺ cells from discarded autologous stem cell concentrates, but their purity greatly varied among donors (from 55 to 90%), and only 4–20% of MC contained both tryptase and chymase (26). The shortest protocol with same range of MC_TC purity requires 9 wk of culture of CD34-positive cells from buffy coat (27). Second, we are not limited by the amount of starting materials because adipose tissue is an abundant and easily accessible tissue.

Our protocol, however, does not improve MC yield compared with protocols generating MC from CD34⁺ cells. A total of 10⁶ to 2 × 10⁷ mature MC can be generated from 50 to 100 ml of blood or buffy coat in 9 to 14 wk (27–29). With our protocol, 5 ± 3 × 10⁴ MC/g of WAT were obtained after 3 wk, reaching 5 ± 3 × 10⁵ MC/g of WAT after 6 wk. This yield can be increased by at least 2-fold if spheroids are maintained in culture.

In our model, we observe that spheroids are able to continuously produce MC for four generations, suggesting that WAT-MC seem to derive from MC progenitors present in the spheroids. This observation is supported by the fact that we previously demonstrated that, in mice, MC in adipose tissue come from resident MC progenitors (12). Moreover, we have quantified the proliferation of mature MC in the spheroids by using BrdU incorporation assay, and our results show that ∼10% of the mature MC population proliferate (data not shown). The lack of robust markers of MC progenitors allowing their identification in vivo did not allow us to conclude about the origin of WAT-MC.

Our culture conditions are based on what is currently used to obtain MC from CD34⁺ cells. First, serum-free culture conditions have been used to prevent batch to batch variability in serum composition and thus increase reproducibility. Indeed, variations in proliferation and maturation of MC have been reported depending on the presence of serum in culture media. Serum almost suppresses MC development of CD34⁺ cord blood cells and inhibits proliferation of skin derived MC when added at the beginning of the culture (30, 31). In contrast, it increases MC maturation when added at later stages (31). Second, we used IL-3 and IL-6, both known to increase the yield of MC derived from CD34⁺ cells mainly by acting on early stage of differentiation (4, 10, 32, 33). The fact that IL-3 and IL-6 were not required for WAT-MC production suggests that MC progenitors present in the WAT are less immature than bone marrow CD34⁺ cells. Moreover, this hypothesis is consistent with the faster production rate of WAT-MC compared with MC derived from CD34⁺ cells.

In conclusion, we have developed a simple, robust, and time saving protocol to obtain mature and functional human MC_TC which behave like MC derived from peripheral blood CD34⁺ cells for all tested parameters. This short term culture protocol will be useful for the evaluation of MC biology in human as well as in mice and is a suitable tool to screen drug effect on human mature MC. In addition, because of the generalized localization and the easy access of WAT, this newly developed MC culture provides the possibility to study individual MC populations from patients in physiological or pathological conditions such as mastocytosis.

We thank Camille Lejon and Cyprien Pelcavel for their technical assistance. We thank Audrey Varin for helpful discussions.

The authors have no financial conflicts of interest.