Phenotypical and Functional Characteristics of in Vitro-Expanded Adipose-Derived Mesenchymal Stromal Cells from Patients with Systematic Sclerosis

Abstract

Mesenchymal stromal cells (MSCs) have received attention as an ideal source of regenerative cells because of their multipotent differentiation potential. Adipose tissue is an attractive source of MSCs. Recent studies have shown that autologous fat grafting may be effective in the treatment of systemic sclerosis (SSc), but no specific study exists that aimed at investigating whether adipose tissue-derived stromal cells (ADSCs) from SSc patients maintain normal phenotypic and functional characteristics. The purpose of the current study was to investigate whether ADSCs from patients with SSc (SSc-ADSCs) are phenotypically and functionally identical to those from healthy controls (HC-ADSCs). Adipose tissue samples were obtained from 10 patients with SSc and from 8 HCs. Both MSC populations were evaluated for their capacity to (a) express specific MSC surface antigens by flow cytometry analysis, (b) proliferate, (c) differentiate along the adipogenic and osteogenic lineages, (d) suppress in vitro lymphocyte proliferation induced by a mitogenic stimulus, and (e) support endothelial cell (EC) tube formation. ADSCs from SSc patients and HCs showed similar surface phenotype and multilineage differentiation capabilities. In PBMC proliferation inhibition assays, no significant differences were observed between SSc- and HC-ADSCs. Using ADSC/EC cocultures, both SSc- and HC-ADSCs improved tube formation by both HC- and SSc-ECs. This effect was enhanced under hypoxic conditions in all of the cocultures. SSc-ADSCs exhibited the same phenotypic pattern, proliferation and differentiation potentials, and immunosuppressive properties as those from HCs. The proangiogenic activity shown by SSc-ADSCs, namely, under hypoxic conditions, suggests that autologous ADSC grafting may represent a possible therapeutic option for SSc.

Introduction

Adult stem/stromal cells, namely those of mesenchymal origin, hold great promise for use in tissue repair and regeneration, and there is increasing interest in both their biology and future therapeutic applications. Bone marrow-derived mesenchymal stromal cells (BM-MSCs) have been most extensively characterized. These cells can differentiate into multiple cell phenotypes, including bone, fat, and cartilage¹. In addition, BM-MSCs exhibit immunomodulatory effects on various activated lymphoid cells, including T cells, B cells, natural killer cells (NKCs), and dendritic cells^2-5. The ability of BM-MSCs to modulate immune responses and promote tissue repair in preclinical studies has prompted the exploration of MSCs in a wide range of experimental and clinical autoimmune disorders and degenerative diseases^6-10. In systemic lupus, both autologous and allogeneic MSCs were able to suppress inflammation and reduce damage to kidneys^11-15. As with systemic sclerosis (SSc), both autologous and allogeneic MSCs were able to induce some improvement in skin fibrosis and skin ulcer healing, with features of neovascularization^16-18. However, some data suggest that BM-MSCs from patients with SSc have impaired proliferation potential, partially defective immunosuppressive capacity, and show a more senescent phenotype^19-24.

Apart from BM-MSCs, interest has rapidly grown on the plasticity and therapeutic potential of MSCs isolated from adipose tissue²⁵. Adipose tissue is an attractive source of adult stem cells because of its abundance and surgical accessibility. Adipose tissue-derived stromal cells (ADSCs) possess many characteristics, such as differentiation potential, that are usually recognized to be similar to those of BM-MSCs. It has been shown that these cells have the capacity to differentiate into cells of different lineages and are able to at least partially restore damaged tissue by inducing angiogenesis, which is likely as a result of their anti-inflammatory and immune modulating properties^26-28. Compared with BM-MSCs, ADSCs offer several advantages, including ease of isolation, less donor morbidity, relative abundance, and rapidity of expansion^28,29. Furthermore, ADSCs have a higher proliferative capability^30,31, display a lower senescence ratio³¹, exhibit higher genetic stability, and retain differentiation potential in long-term culture^32-34. Although it has been initially shown that BM-MSCs and ADSCs exhibit very similar immunosuppressive properties^28,35, some additional in vitro results suggest that ADSCs can be more effective suppressors of an immune response than BM-MSCs^36,37. Moreover, human ADSCs exhibit greater proangiogenic activity than human BM-MSCs, an effect thought to be mostly mediated by matrix metalloproteinases (MMP-3 and MMP-9)³⁸. Moreover, in response to hypoxia, ADSCs stimulate angiogenesis more strongly than BM-MSCs^38-40, as shown by the higher upregulation of genes for proangiogenic factors and downregulation of antiangiogenic genes in ADSCs. In addition, hypoxia stimulates the expression of a C-X-C chemokine receptor type 4 (CXCR4) receptor on the surface of ADSCs, which promotes the directed migration and accumulation of these cells in damaged areas⁴⁰.

Based on these outstanding findings, autologous fat grafting has been applied in the treatment of different skin diseases including localized scleroderma and SSc^41-46. This procedure has been shown to be effective in improving mouth-opening capacity and in inducing neoangiogenesis of the perioral region of SSc patients who have perioral skin involvement⁴³. Furthermore, autologous fat grafting also appeared to be effective in inducing rapid healing of SSc-related digital ulcers unresponsive to traditional therapy⁴⁴. Finally, it was demonstrated that autologous grafting of ADSCs improves distal ischemic crisis caused by Raynaud's phenomenon in patients with SSc^45,46. Up to now, however, it is unknown whether ADSCs from patients with SSc are phenotypically and functionally similar to those of healthy controls (HCs), or are defective in proliferative potential and immunosuppressive capacity, as demonstrated for SSc-BM-MSCs^19-24.

In this study, the phenotypical and functional characteristics of ADSCs from patients with SSc and HCs have been compared. In particular, the angiogenic potential of both cell populations has been assessed by evaluating their ability to support blood vessel formation.

Materials and Methods

Adipose Tissue Samples

Human adipose tissue (AT) samples were obtained by elective liposuction procedures from 10 patients with the diffuse form of SSc undergoing autologous fat grafting for the treatment of fibrotic perioral changes. All the patients fulfilled the American College of Rheumatology–European League Against Rheumatism (ACR-EULAR) new criteria for SSc⁴⁷ and were classified as having diffuse cutaneous SSc (dcSSc) on the basis of the criteria of LeRoy et al.⁴⁸. The presence of severe extracutaneous involvement, treatment with immunosuppressive therapies (including prednisone equivalent >10 mg) during the 3 months before enrolment time, current pregnancy, and lactation were considered as exclusion criteria. As controls, AT samples were obtained by the same liposuction procedure from eight age- and sex-matched HCs undergoing aesthetic surgery. AT was harvested by the Coleman technique, as previously described⁴³.

The study was approved by the G. Pini Hospital ethics committee and performed according to the criteria of the Declaration of Helsinki. Written informed consent was obtained from all patients who took part in the study.

Isolation and Culture of ADSCs

Lipoaspirate (∼7 g) was washed three to four times to remove excess blood by mixing with an equal volume of phosphate-buffered saline (PBS; Euroclone, Pero, Italy) and allowed to settle for 5 min to separate the aqueous phase from the fat fraction. The fat was then transferred to a 150-mm tissue culture Petri dish (Corning Inc., Corning, NY, USA), where it was minced into fragments of about 5 mm³. The tissue fragments were then distributed over the surface of the dish. The explants were cultured in α-minimum essential medium (α-MEM; Life Technologies, Paisley, UK) supplemented with 5% allogeneic human platelet lysate (hPL; Transfusion Centre, ASST Papa Giovanni XXIII, Bergamo, Italy). A minimal quantity (10 ml) of prewarmed culture medium was added over the tissue explants, such that the explants still remained in contact with the surface of the culture dish. The dishes were maintained at 37°C with 5% humidified CO₂. The explant tissue was removed on days 5-7 after plating, and the outgrown cells were cultured in fresh medium for another 7-10 days until a confluent monolayer was obtained. These initial cells, referred to as passage 1 (P1), were further subcultured at a seeding density of 100-500 cells/cm² and serially passaged.

Growth Kinetics

For population doubling (PD) evaluation in long-term cultures, ADSCs were plated at 500 cells/cm² in T25 flasks (Nalge-Nunc International, Rochester, NY, USA) in α-MEM supplemented with 5% hPL. The cells were counted and passaged at a confluence of 70%-90% and replated at the same cell density. PD was calculated according to the following equation: PD=(logN_f-logN_i) × 3.32, where N_f is the number of viable cells at the end of each passage, and N_i is the cell number seeded. Cumulative PD was calculated by adding the PD value obtained for each passage to the sum of PDs obtained from the previous passages. Since the cell number of plastic-adherent cells could be counted only starting from P1, the cumulative doubling rates were calculated starting at P2. The PD time (PD_t; in hours) was calculated using the time interval between cell seeding and harvesting divided by the PD for that passage.

Immunophenotypic Characterization of ADSCs

Immunophenotyping of the expanded ADSCs was performed by flow cytometry when all the cell cultures were at two in vitro passages. Cells were labeled with the following antibodies: CD10-fluorescein isothiocyanate (FITC), CD14-FITC, CD26-phycoerythrin (PE), CD31-FITC, CD34-PE, CD45-FITC, CD73-PE, CD90-FITC, CD106-FITC, human leukocyte antigen–antigen D related (HLA-DR)-FITC, HLA-ABC-PE, CD146-PE (BD Biosciences, San Jose, CA, USA), CD105-PE (Caltag Laboratories, Burlingame, CA, USA), isotype-matched immunoglobulin G (IgG)-FITC, and IgG-PE control antibodies (BD Biosciences). Analysis was performed on a fluorescent-activated cell sorting (FACS) Canto II (BD Biosciences) on at least 5,000 events and using the software FACSDiva (BD Biosciences).

ADSC Multilineage Differentiation

The adipogenic and osteogenic differentiation capacity of ADSCs was determined between P2 and P5. For osteogenic differentiation, cells were plated in six-well culture plates (BD Falcon, Milan, Italy) at a density of 3,000 cells/cm² in MSC expansion medium and incubated in a humidified atmosphere with 5% CO₂ at 37°C. At 60%-70% confluence, the medium was replaced with osteogenic induction medium (Differentiation Media BulletKit^™ Osteogenic; Lonza, Basel, Switzerland). After 3 weeks of culture with fresh medium replacement every 3 days, the cells were fixed with precooled 70% ethanol for 1 h at room temperature. Osteogenesis was assessed using Alizarin red S (Sigma-Aldrich, St. Louis, MO, USA) staining to detect the deposition of intracellular calcium.

For adipogenic differentiation, cells were plated in six-well culture plates at a density of 10,000 cells/cm² in MSC expansion medium and incubated in a humidified atmosphere with 5% CO₂ at 37°C. At 90% confluence, the growth medium was changed alternatively every 3-4 days from adipogenic induction to adipogenic maintenance medium (Differentiation Media BulletKit^™ Adipogenic; Lonza) following the manufacturer's instructions. After 3 weeks of culture, the cells were fixed with 10% formalin (Sigma-Aldrich) for 30 min at room temperature. Adipogenesis was assessed using Oil red O solution (ORO; Sigma-Aldrich) staining to detect intracellular lipids.

The images were analyzed under an inverted microscope (Axiovert 25; Zeiss, Oberkochen, Germany).

Inhibition of Peripheral Blood Mononuclear Cell (PBMC) Proliferation by ADSCs

To assess the ability of ADSCs to suppress PBMC proliferation, ADSCs were thawed and seeded in 48-well flat bottom plates at 20 × 10³ cells/well in α-MEM + 10% heat-inactivated AB serum and allowed to adhere to the plastic surface overnight. PBMCs, isolated from two healthy donors by Ficoll-Hypaque density centrifugation technique (Lympholyte^®-H; Cedarlane Laboratories, Hornby, Canada), were labeled with 1 mM carboxyfluorescein diacetate succinimidyl ester (CFDA-SE; Molecular Probes, Eugene, OR, USA). CFDA-SE-labeled PBMCs (10⁵/well) were then plated in the presence of either SSc-ADSCs or HC-ADSCs (ratio: PBMC/MSC, 5:1). In a second experiment, 10⁵ CFDA-SE-labeled PBMCs isolated from two different SSc patients were plated in the presence of autologous SSc-ADSCs at the same PBMC/MSC ratio. Phytohemoagglutinin (PHA; 2 μg/ml; Sigma-Aldrich) was added as mitogenic stimulus.

After 5 days of coculture, PBMCs were collected by gentle pipetting, stained with CD3-allophycocyanin (BD Biosciences, Milano, Italy), and approximately 10,000 events were acquired on a FACSCanto II (BD Biosciences) and analyzed using ModFitLT^™ (Verity Software House, Topsham, ME, USA) to obtain the proliferation index (PI), which is related to the number of cell divisions the PBMCs underwent.

Endothelial Cell Isolation and Culture

After approval of the San Salvatore University Hospital ethics committee and written informed consent from patients, microvascular endothelial cells (ECs) were obtained from skin biopsy of four patients with dcSSc of recent onset (disease duration less than 3 years calculated since the first non-Raynaud's symptom). Patients discontinued corticosteroids, oral vasodilators, intravenous prostanoids, or other potentially disease-modifying drugs, at least 1 month before biopsies. None assumed immunosuppressants. Four frozen EC samples obtained from skin samples of age- and sex-matched healthy volunteers were used as control (HC-ECs).

Biopsy samples (1 × 0.5 cm) were placed into a 50-ml tube containing 15 ml of trypsin (Sigma-Aldrich) and then digested for 45 min at 37°C. The cells were cultured in EC growth medium, EGM^™-2-MV (Lonza), at 37°C in a humidified atmosphere of 5% CO₂.

Before the cells reached confluence, after approximately 1 week, the heterogeneous pool of cells was exposed to a CD31⁺ selection, performed with the Dynabeads^™ magnetic CD31 microbead cell sorting system (Life Technologies). The beads rapidly target and partially coat the ECs expressing the CD31 receptor.

After incubation, the cells were placed in a magnet [Dynal^® Magnetic Particle Concentrator (MPC^®)-S; Life Technologies] for 2 min, following the manufacturer's recommended protocol for washings and final extraction. The CD31⁻ cells were removed during the successive washings. The positive selected cells were 99% ECs with a specific phenotype (CD31, CD34, and CD144). The cells were used at third passages (P3).

In Vitro Angiogenesis Assay and Hypoxia

Tube formation ability was evaluated using a Matrigel assay. Matrigel (8.6 mg/ml; BD Biosciences) was used at 1:1 dilution with EGM2-MV, without supplement. ECs and ADSCs were labeled, before coculture in Matrigel, using the green fluorescent dye PKH67 and red fluorescent dye PKH26 (Sigma-Aldrich), respectively, according to the manufacturer's instructions.

ECs and ADSCs were seeded alone and in coculture ECs/ADSCs in a 2:1 ratio, and incubated in hypoxia (1% O₂) (according to Hayashi et al.⁴⁹) and in normoxia (5% CO₂/95% air). After 4 h, the images were acquired using an Olympus BX53 fluorescence microscope (Olympus Italia Srl, Segrate, Italy).

The total tube length of each well was measured as branching index=(master junction/area) × 1,000 and photographed. Results are expressed as median (range) of triplicate experiments (p=0.02).

Statistical Analysis

Statistical analysis was performed by standard procedures using SPSS^® Statistic21 (IBM^® Software Group, Chicago, IL, USA). All results are expressed as mean±standard deviation (SD). To estimate the probability of differences, we have adopted the Mann–Whitney U-test.

Results

Patient Characteristics

ADSCs were isolated by explant culture technique performed with AT obtained from 10 SSc patients (SSc-ADSC) and 8 HCs (HC-ADSC) undergoing liposuction for clinical or aesthetic reasons, respectively. The median age of SSc patients was 39 years (mean: 38±13), while the median age of HCs was 40 years (mean: 41±16). All the patients had the diffuse form of the disease. Their median disease duration was 58 months (ranging from 30 to 74), while the median modified Rodnan skin score (mRss) was 5 (ranging from 2 to 10). Eight patients showed pulmonary involvement with evidence of interstitial lung disease at high-resolution computed tomography (HRCT), with a median lung diffusion capacity (DLCO) equal to 55% of the predicted value (ranging from 41 to 100).

Proliferative Capacity of ADSCs From SSc Patients and HCs

The cell isolation rate was 100% for both types of tissue donor. Within 5-7 days of culture explant, spindle-shape cells with typical mesenchymal morphology could be observed growing out the adipose tissue onto the culture dish. After tissue removal, the adhered cells continued to proliferate until reaching 70%-90% confluence after a further 7-10 days. The mean yield obtained at P1 for SSc-ADSCs was 1.7±2.0 × 10⁵cells/g of initial tissue plated, while for HC-ADSCs it was 2.2±1.6 × 10⁵cells/g (not statistically significant; p = 0.7).

To assess the proliferative capacity of both SSc- and HC-ADSCs, the cells were serially passaged, and cumulative PDs were calculated for all passages. As shown in Figure 1, both ADSC types showed very similar growth kinetics in terms of rates of expansion and cumulative PD (Fig. 1A and B). In a mean of 50±6 days, SSc-ADSCs expanded over a mean of 25.5±2.9 PDs, while HC-ADSCs reached a mean of 26.4±3.0 PDs in a mean of 53±3 days (P6). MSC growth was also evaluated in terms of PD_t; the mean PD_t between P1 and P4 for SSc-ADSCs was 32.8±6.5 h, whereas it was 31.8±5.4 h for HC-ADSCs (Fig. 1C). There was no statistically significant difference in terms of proliferative capacity between HC- and SSc-ADSCs.

Figure 1.

Proliferative capacity of both SSc-ADSCs and HC-ADSCs. Long-term expansion capacity for 10 different ADSCs from SSc patients and 8 different ADSCs from HCs. Cell numbers were determined at the end of every passage, and cumulative PDs were calculated in relation to the cell numbers plated (A). Mean cumulative PDs for SSc-ADSCs and HC-ADSCs from passage 2 (P2) to P6 (B). Mean PD time for SSc-ADSCs and HC-ADSCs between P2 and P4 (C). SSc, systemic sclerosis; ADSCs, adipose tissue-derived stromal cells; HC, healthy control; PD, population doubling.

Immunophenotypic Characterization of ADSCs From SSc Patients and HCs

The phenotype of both SSc- and HC-ADSCs was analyzed by flow cytometry, and the results of different cell surface marker expression are shown in Table 1 and Figure 2A, where the percentage of positive cells for each marker is indicated. Both MSC preparations showed a high expression of typical MSC markers CD73, CD90, and CD105 as well as of CD10 and CD26, with a higher mean fluorescence intensity (MFI) of CD90 in comparison with the other positive markers (Fig. 2B); a lower expression was found for CD146 [melanoma cell adhesion molecule (MCAM)]. Furthermore, both cell populations demonstrated negative results for expression of the typical hematopoietic and endothelial markers such as CD14, CD31, CD34, and CD45 as well as for CD106, confirming data already reported^50,51. In addition, the presence of HLA-ABC proteins and the absence of HLA-DR were observed. Based on these results, ADSCs from patients with SSc and HCs were indistinguishable from a phenotypic point of view.

Table 1.

Phenotypic Characterization of SSc-ADSCs and HC-ADSCs

Marker	SSc-ADSCs % Positive Cells (±SD)	HC-ADSCs % Positive Cells (±SD)
CD45	0.8±0.4	0.7±0.1
CD34	1.6±2.0	0.0±0.1
CD14	1.1±0.8	0.0±0.0
CD31	0.3±0.6	0.0±0.0
CD73	98.9±0.6	98.2±1.7
CD90	99.8±0.2	99.8±0.1
CD105	98.8±0.7	98.9±0.1
HLA-ABC	94.6±4.0	92.6±3.9
HLA-DR	1.0±0.6	1.3±1.8
CD10	91.9±15.2	88.0±2.8
CD26	97.7±1.4	96.8±3.0
CD106	1.3±1.8	0.0±0.0
CD146	22.2±12.8	30.1±16.5

FACS analysis was performed when all cultures were at two in vitro passages. Data represent the percentage of positive cells for each marker analyzed (means±SD).

Figure 2.

Immunophenotype of SSc-ADSCs and HC-ADSCs. Mean of percentage of positive cells from 10 different samples of systemic sclerosis patient-derived adipose tissue-derived stromal cells (SSc-ADSCs) and 8 healthy control-derived adipose tissue-derived stromal cells (HC-ADSCs) is represented (A). Mean of mean fluorescence intensity (MFI)-positive cells from 10 different samples of SSc-ADSCs and 8 HC-ADSCs is represented (B).

Morphology and Multilineage Differentiation Potential of ADSCs From SSc and HCs

ADSCs isolated from SSc patients displayed a predominantly fibroblast-like morphology similar to ADSCs isolated from normal donors and expanded in the same culture conditions (Fig. 3A).

Figure 3.

Morphology and differentiation potential of SSc-ADSCs and HC-ADSCs. (A) No differences in cell morphology were observed between undifferentiated healthy control-derived adipose tissue-derived stromal cells (HC-ADSCs) and systemic sclerosis patient-derived adipose tissue-derived stromal cells (SSc-ADSCs). (B) Adipogenic differentiation revealed by formation of lipid droplets stained with Oil red O after 18 days of induction for HC-ADSCs and SSc-ADSCs. (C) Osteogenic differentiation revealed by Alizarin red (AR) staining after 21 days of induction for HC-ADSCs and SSc-ADSCs. Both SSc-ADSCs and HC-ADSCs are equally competent to differentiate toward adipocytes or bone cells. Magnification: 400x.

A typical feature of MSCs is their multipotent differentiation potential. Both ADSC populations were induced to differentiate along the adipogenic and osteogenic lineages using specific culture media. SSc- and HC-ADSCs exhibited an equal potential to differentiate into adipocytes as examined by ORO staining (Fig. 3B) with a marked change in the cellular morphology together with the accumulation of lipid vacuoles. When tested for differentiation capabilities along osteogenic lineage, comparable levels of calcium deposition were detected by Alizarin red staining in both the ADSC preparations after 21 days of induction in the osteogenic medium (Fig. 3C). We conclude that both SSc- and HC-ADSCs are equally competent to differentiate toward bone or adipocytes.

Immunosuppressive Properties of ADSCs From SSc and HCs

The immunosuppressive activity of ADSCs, derived from both HC donors and SSc patients, was assessed in two different experiments by coculture with PHA-stimulated PBMCs. In the first set of experiments, PHA-treated PBMCs isolated from two normal donors were cocultured for 5 days in the presence of either SSc-ADSCs (three donors) or HC-ADSCs (three donors). As shown in Figure 4A, both SSc- and HC-derived ADSCs inhibited a mitogen-stimulated lymphoproliferation with no difference between SSc patients and HCs.

Figure 4.

Immunomodulatory effect of ADSCs on mitogen-stimulated PBMCs. (A) ADSCs isolated from three SSc (SSc-ADSCs) patients and three healthy controls (HC-ADSCs) were cocultured at a peripheral blood mononuclear cell (PBMC)/mesenchymal stromal cell (MSC) ratio of 5:1 with carboxyfluorescein diacetate succinimidyl ester (CFDA-SE)-labeled PBMCs stimulated with 2 μg/ml phytohemoagglutinin (PHA). CD3⁺ cell proliferation was evaluated at day 5, and data were expressed relative to CD3⁺ cells alone (assigned to 100%). Results are expressed as mean±SD of the two experiments. Each bar represents a different donor of ADSCs (SSc donors: #1, #2, and #3; HC donors: #4, #5, and #6). (B) Fluorescence profile of PHA-stimulated PBMCs isolated from two SSc patients labeled with CFDA-SE and co-cultured with autologous SSc-ADSCs (ratio: PBMC/MSC, 5:1). Each peak corresponds to a T-cell division. The PI (a statistic generated by ModFit) correlated with the number of cell divisions the PBMC underwent. The data were normalized to the PI of PHA-stimulated PBMCs alone (assigned to 100%). PBMC inhibition by ADSCs is shown as a reduction of the PI.

To further confirm the SSc-ADSC capacity to inhibit T-cell proliferation, we set up a second experiment by coculturing ADSCs isolated from two different SSc patients (#1 and #2) with PHA-stimulated PBMCs isolated from the same patients (autologous setting) (Fig. 4B). Although PBMCs from patient #1 proliferated much more than PBMCs from patient #2 (PI = 26.8 vs. 6.6, respectively), coculture with autologous SSc-ADSCs significantly reduced the PI to 2.1 and 2.9, respectively.

ADSCs Assist EC Tube Formation

In the normoxic condition, HC-ECs, cultured alone into Matrigel, formed small numbers of organized tube-like structures (Fig. 5A). When HC-ECs were cocultured with HC-ADSCs, we observed a significant improvement in tube formation ability (Fig. 5B). SSc-ADSCs cocultured with HC-ECs mirrored the results obtained with HC-ADSCs (Fig. 5C).

Figure 5.

Tubular-like structure formation by HC-ECs. (A–B1) Phase-contrast pictures of tubular-like structure formation in normoxia: endothelial cells from healthy controls (HC-ECs) cultured alone form a small number of short tube-like structures (A); HC-ECs/adipose tissue-derived stromal cells from healthy controls (HC-ADSCs) coculture: the picture shows that mesenchymal stromal cells (MSCs) support the HC-ECs' ability to perform tube-like structures (B); HC-EC/adipose tissue-derived stromal cells from systemic sclerosis patients (SSc-ADSC) coculture: SSc-ADSCs mirror HC-ADSCs' behavior (B1). (C–D1) Phase-contrast pictures of tubular-like structure formation in hypoxia: HC-ECs cultured alone—the hypoxia improves the HC-EC ability to form a well-organized tube network (C); HC-EC/HC-ADSC coculture: the hypoxia increases the HC-ADSCs' ability to form a well-organized tube-like network (D); HC-EC/SSc-ADSC coculture: SSc-ADSCs mirror the HC-ADSCs' behavior (D1). The fluorescent images show the EC (green) and MSC (red) distribution. Pictures are representative of all experiments.

Of note, the hypoxic condition improved the HC-EC ability to form a well-organized tube network when compared to the HC-EC ability in normoxic condition (Fig. 5D). When HC-ECs were cocultured in hypoxia with HC-ADSCs, MSCs supported the tube formation. It is noteworthy that the organization of the tube was improved when compared with HC-EC/HC-ADSC cocultures in normoxia (Fig. 5E). HC-EC/SSc-ADSC cocultures mirrored the results obtained with HC-ECs/HC-ADSCs (Fig. 5F).

When cultured in the normoxic condition, the SSc-ECs formed a significantly lower number of vessels when compared with HC-ECs (Fig. 6A). The coculture with both HC- and SSc-ADSCs improved the organization of the tubes (Fig. 6B and C). The hypoxic stimulus induced a significant increase in SSc-EC ability to form an organized tube (Fig. 6D), and the coculture with both HC- and SSc-ADSCs improved the organization of the tube network (Fig. 6E and F), but the branching index value for SSc-ECs was significantly lower when compared to HC-ECs cultured in the same culture condition (Fig. 7).

Figure 6.

Tubular-like structure formation by SSc-ECs. (A–B1) Phase-contrast pictures of tubular-like structure formation in normoxia: endothelial cells from systemic sclerosis patients (SSc-ECs) cultured alone form a small number of short tube-like structures (A); SSc-EC/adipose tissue-derived stromal cells from healthy controls (HC-ADSC) coculture: the organization of the tube was improved by the presence of HC-ADSCs in the SSc-EC culture (B); ADSCs from systemic sclerosis patients (SSc-ADSCs), mirroring HC-ADSCs, improve SSc-ECs' ability to form tube-like structures (B1). (C–D1) Phase-contrast pictures of tubular-like structure formation in hypoxia: SSc-ECs cultured alone: the hypoxic condition improves the SSc-ECs' ability to form tube-like structures (C); SSc-EC/HC-ADSC cocultures: HC-ADSCs support the tube formation (D); SSc-EC/SSc-ADSC cocultures: SSc-ADSCs mirror the HC-ADSCs' skill to improve tube-like structures (D1). The fluorescent images showed the EC (green) and mesenchymal stromal cell (MSC) (red) distribution. Pictures are representative of all experiments.

Figure 7.

Comparison in tubular-like structure formation by HC- and SSc-ECs in the presence of HC- and SSc-ADSCs in the normoxia and hypoxia conditions. The picture shows the capacity of adipose tissue-derived stromal cells from healthy controls (HC-ADSCs) and SSc-ADSCs to enhance the tubular formation by HC- and SSc-ECs. As shown in the graphic, the coculture increases the ability to form tube-like structures by HC-ECs in the normoxia (A) and more in the hypoxia (B) conditions, both by HC-ADSCs and SSc-ADSCs. Furthermore, no difference was observed using ADSCs of patients and controls. The formation of tube-like structures by SSc-ECs was low in comparison to HC-ECs and was slightly but significantly improved by the presence of HC-ADSCs and SSc-ADSCs in the normoxic condition. The hypoxic stimulus increased the tubular formation by HC-ECs and SSc-ECs in comparison to the normoxic condition. This effect was higher in the presence of HC-ADSCs and SSc-ADSCs. The total tube length of each well was measured as branching index = (master junction/area) × 1,000 and photographed. Results are expressed as median (range) of triplicate experiments. *p = 0.002; **p = 0.0001.

Discussion

This study demonstrates that ADSCs isolated and expanded from patients with SSc exhibit in vitro the same phenotypic and proliferative characteristics, and similar capacity to differentiate themselves toward osteogenic and adipogenic lineages in comparison with MSCs from HCs. Furthermore, when compared to the same cell type from HCs, SSc-ADSCs maintain an intact capability to suppress lymphocyte proliferation induced by a mitogenic stimulus. Finally, both SSc- and HC-ADSCs are able to support either HC- or SSc-ECs to perform tube formation in normoxic condition. The latter capability appears to be enhanced under the hypoxic condition and, in a comparable way, in all types of MSC-EC cocultures including those where SSc-ECs were tested. This is a noticeable result considering that in a previous study cultured SSc-ECs had shown a decreased spontaneous ability to develop a tube network, which is not improved by cocultures with different BM-MSCs⁵².

Overall, the results of the present study show that ADSCs from patients with SSc can be considered phenotypically and functionally comparable to the same cell type isolated and cultured from normal individuals. Conversely, most of the studies on BM-MSCs from patients with SSc have shown that these cells are somehow defective in their proliferative potential and may express a partially different phenotype^19,20. SSc-BM-MSCs have been found to be preferentially differentiated toward myofibroblast-like cells and express increased TGF-β-related transcripts⁵³, probably as a consequence of SSc-related TGF-β overproduction. Furthermore, SSc-BM-MSCs have also been reported to be more oriented to endothelial differentiation, but numerically and functionally ineffective for this purpose^19,20. Although SSc-derived BM-MSCs have been shown to constitutively overexpress proangiogenic factors⁵⁴, they are less effective than HC-BM-MSCs in inducing tube formation in cocultures with both HC- and SSc-ECs, which probably contributes to the insufficient vascular repair in SSc^20,52.

As to the immunosuppressive activity of SSc-BM-MSCs, some studies have provided evidence that these cells may exert a normal capacity to modulate proliferation of mononuclear cells in vitro. However, their senescent aspect and the need for some enrichment of culture medium to maintain this ability suggest that the SSc-BM-MSCs immunosuppressive power can be impaired in some way^21,23.

It has been postulated that some of the defective features of SSc-BM-MSCs may be ascribed to the changes in their microenvironment that may be related to the disease process^20,53,54. Why ADSCs from patients with SSc may preserve their normal phenotypic expression and functional behavior remains purely speculative at present. As locally activated biological pathways condition the MSCs resident in different sites, similarly the specifically released mediators can differently influence these cells during the disease evolution. Adult MSCs reside in specific local niche providing structural support and molecular signals to regulate MSC quiescence, self-renewal, and activation for a biologically specific regulation of their own site. In the BM niche, the hematopoietic stem cells, osteoblasts, osteoclasts, ECs, perivascular cells, and abundant reticular cells surround the MSCs. A variety of autocrine and paracrine stimuli, coming from the different cellular components of the niche, are effective in regulating self-maintenance and activating the stem cells of both hematopoietic and mesenchymal origin⁵⁵. Conversely, mature adipocytes and ECs are the most important cellular components in the AT niche, and these cells influence MSCs through direct cell-to-cell interactions and indirect paracrine effects. Inflammatory and metabolic mediators locally released in conditions such as diabetes and obesity induce local hypoxia in adipose tissue and increase the secretion of hormones, growth factors, and adipokines to initiate the stimulation of ADSCs with consequent proliferation, adipocyte differentiation, and microvessel formation⁵⁶. Therefore, it is not so surprising that MSCs isolated from different sites and coming from both HC and patients with different diseases may show dissimilar biological potentials.

The fact that SSc-ADSCs maintain the same proangiogenic and immunosuppressive characteristics as those of HC-ADSCs may really open new scenarios for their autologous therapeutic use in SSc, where immune activation and endothelial damage, with consequent microvascular bed loss, represent two fundamental pathological processes.

The finding that MSC-induced neoangiogenesis in vitro may be enhanced in the hypoxic condition is an additional favorable aspect in this perspective. The in vitro results can be the biological counterpart of the in vivo observation that autologous fat grafting is able to support neoangiogenic changes in perioral skin and fingers of patients with SSc^43,44 and consequently induce improvement of lip skin elasticity and prompt healing of indolent digital ulcers. Even the observation by Granel et al.⁴⁵ that autologous grafting of AT stromal vascular fraction in the hands of patients with SSc predominantly improves vascular peripheral manifestations, such as Raynaud's phenomenon, represents an indirect confirmation of the neoangiogenic action of ADSCs. This hypothesis is further substantiated by the significant reduction of avascular areas and dystrophic capillaries recorded in nailfold video-capillaroscopy of the grafted patients in this study⁴⁵.

The different immunosuppressive activities exhibited by MSCs have also been postulated to be potentially useful in the perspective of a cellular-based therapy in SSc, as well as in other autoimmune disorders. MSCs have been found to be capable of inhibiting the function of different immune cells of both innate and adaptive immunity by downregulating the replication of activated T cells, differentiation and maturation of dendritic cells, and, conversely, by upregulating the regulatory T cells^57-61. Autoimmune diseases are commonly characterized by expansion of dendritic cells, which work as the antigen-presenting cells to activated T-cell-specific clones whose proliferation is not adequately controlled by defective regulatory T cells. Thus, MSCs appear to possess the right profile to modulate the immunological abnormalities characterizing these pathologic conditions. Preliminary pilot trials have been conducted in systemic lupus erythematosus and SSc using either autologous or allogeneic BM-MSCs and have shown promising results⁹.

The demonstration provided by the present study that SSc-derived ADSCs possess unaltered proliferative, angiogenetic, as well as immune-regulatory potential, suggests that the use of this cell type could be the preferential choice in future experimental therapeutic applications of cell-based therapy in SSc. This hypothesis has been further supported by the results of a recent study in a mouse model of SSc where human ADSCs showed higher anti-inflammatory and antifibrotic properties than BM-MSCs⁶².

In conclusion, the present study shows that ADSCs isolated and expanded from patients with SSc maintain a normal phenotype and an unchanged proliferation capacity. Their unaltered immunosuppressive and neoangiogenetic potentials strongly support the use of these cells, which are abundant and easily accessible in the fat tissue, in future cell-based therapeutic trials in SSc.