Article Text

Extended report
Bone marrow-derived mesenchymal stem cells from early diffuse systemic sclerosis exhibit a paracrine machinery and stimulate angiogenesis in vitro
  1. Serena Guiducci1,
  2. Mirko Manetti1,2,
  3. Eloisa Romano1,
  4. Benedetta Mazzanti3,
  5. Claudia Ceccarelli1,
  6. Simone Dal Pozzo3,
  7. Anna Franca Milia1,
  8. Silvia Bellando-Randone1,
  9. Ginevra Fiori1,
  10. Maria Letizia Conforti1,
  11. Riccardo Saccardi3,
  12. Lidia Ibba-Manneschi2,
  13. Marco Matucci-Cerinic1
  1. 1Department of Biomedicine, Division of Rheumatology, AOUC, Excellence Centre for Research, Transfer and High Education DENOthe, University of Florence, Florence, Italy
  2. 2Department of Anatomy, Histology and Forensic Medicine, University of Florence, Florence, Italy
  3. 3Department of Haematology, University of Florence, Florence, Italy
  1. Correspondence to Professor Marco Matucci-Cerinic, Department of Biomedicine, Division of Rheumatology, DENOthe Centre, University of Florence, Viale Pieraccini 18, 50139 Florence, Italy; cerinic{at}unifi.it

Abstract

Objective To characterise bone marrow-derived mesenchymal stem cells (MSCs) from patients with systemic sclerosis (SSc) for the expression of factors implicated in MSC recruitment at sites of injury, angiogenesis and fibrosis. The study also analysed whether the production/release of bioactive mediators by MSCs were affected by stimulation with cytokines found upregulated in SSc serum and tissues, and whether MSCs could modulate dermal microvascular endothelial cell (MVEC) angiogenesis.

Methods MSCs obtained from five patients with early severe diffuse SSc (SSc-MSCs) and five healthy donors (H-MSCs) were stimulated with vascular endothelial growth factor (VEGF), transforming growth factor β (TGFβ) or stromal cell-derived factor-1 (SDF-1). Transcript and protein levels of SDF-1 and its receptor CXCR4, VEGF, TGFβ1 and receptors TβRI and TβRII were evaluated by quantitative real-time PCR, western blotting and confocal microscopy. VEGF, SDF-1 and TGFβ1 secretion in culture supernatant was measured by ELISA. MVEC capillary morphogenesis was performed on Matrigel with the addition of MSC-conditioned medium.

Results In SSc-MSCs the basal expression of proangiogenic SDF-1/CXCR4 and VEGF was significantly increased compared with H-MSCs. SSc-MSCs constitutively released higher levels of SDF-1 and VEGF. SDF-1/CXCR4 were upregulated after VEGF stimulation and CXCR4 redistributed from the cytoplasm to the cell surface. VEGF was increased by SDF-1 challenge. VEGF, TGFβ and SDF-1 stimulation upregulated TGFβ1, TβRI and TβRII in SSc-MSCs. TβRII redistributed from the cytoplasm to focal adhesion contacts. SSc-MSC-conditioned medium showed a greater proangiogenic effect on MVECs than H-MSCs. Experiments with blocking antibodies showed that MSC-derived cytokines were responsible for this potent proangiogenic effect.

Conclusion SSc-MSCs constitutively overexpress and release bioactive mediators/proangiogenic factors and potentiate dermal MVEC angiogenesis.

Statistics from Altmetric.com

Introduction

Systemic sclerosis (SSc, scleroderma) is characterised by widespread vasculopathy, lack of angiogenesis and progressive fibrosis of the skin and internal organs.1 In early SSc, endothelial cell activation and damage and altered vessel permeability with formation of perivascular infiltrates are commonly observed.1 2 In advanced SSc, intimal proliferation, vessel narrowing or obliteration, loss of angiogenesis and perivascular and stromal fibrosis predominate.2 3 The progressive vascular involvement alters vascular tone control and reduces blood flow, leading to tissue ischaemia and clinical manifestations such as Raynaud's phenomenon, digital ulcers or gangrene.3 4 In SSc, despite impaired blood flow and chronic tissue hypoxia, there is no evidence of significant angiogenesis and vascular repair.3 Moreover, fibroblasts become constitutively activated, transdifferentiate into contracting myofibroblasts and produce high amounts of collagen and other extracellular matrix components that accumulate in tissues provoking organ dysfunction.1

In SSc, treatments including disease-modifying therapies are available for specific organ-based complications, and some therapies have led to improvements in skin thickening and pulmonary function.5 6 Nevertheless, there is a need for further therapies able to stop or reverse fibrosis and disease progression. Recent evidence suggests that, in SSc, autologous stem cell-based therapies by topical application or systemic delivery may be effective for the treatment of peripheral vascular complications, including healing of digital ulcers and gangrene of the extremities.7,,9

Adult bone marrow-derived mesenchymal stem cells (MSCs) can be defined as multipotent cells of non-haematopoietic origin which are able to differentiate into various types of end-stage specialised mesenchymal cell phenotypes such as osteoblasts, chondrocytes, adipocytes, tenocytes and other connective tissue cells.10 11 The ability to differentiate into multiple cell lineages makes these cells attractive for therapeutic purposes. Besides their potential to differentiate, the secretion of a broad range of bioactive molecules by MSCs—such as growth factors, cytokines and chemokines that exert beneficial effects on other cells—may determine their most biologically significant role under injury conditions.11 12 The paracrine effects of MSCs include trophic, antiapoptotic, supportive (eg, stimulation of mitosis, proliferation and differentiation of organ-intrinsic precursor or stem cells), chemoattractant and proangiogenic actions, as well as immunomodulatory properties.12

Recent studies have shown that bone marrow-derived MSCs from patients with SSc (SSc-MSCs) exhibit the same phenotypic, proliferative and differentiative potential and immunosuppressive properties as MSCs from healthy controls (H-MSCs) and could therefore be considered in an autologous setting.13 However, other reports have described failure of bone marrow-derived SSc-MSCs to differentiate into osteogenic and adipogenic lineages,14 and their commitment towards the endothelial phenotype was also defective.15 To the best of our knowledge, no data are currently available in patients with SSc on the production of bioactive molecules by MSCs.

In order to investigate the possible involvement of bone marrow-derived MSCs in the pathogenesis of SSc and their potential therapeutic properties, we characterised SSc-MSCs for the mRNA and protein expression of bioactive factors participating in MSC recruitment at sites of injury, angiogenesis and fibrosis. In particular, we focused on stromal cell-derived factor-1 (SDF-1/CXCL12) and its receptor CXCR4, vascular endothelial growth factor (VEGF) and transforming growth factor β1 (TGFβ1) and its receptors TβRI and TβRII which are key molecules involved in the paracrine effects of MSCs12 and have been implicated in the pathophysiology of SSc.1 3 We also analysed whether the production of these bioactive mediators by cultured MSCs could be affected by stimulation with cytokines that are commonly found upregulated in SSc serum and lesional tissues, and whether MSCs were able to modulate dermal microvascular endothelial cell (MVEC) angiogenesis in vitro.

Methods

Patients and controls

Five patients with early severe diffuse cutaneous SSc with rapidly progressive disease16 underwent autologous haematopoietic stem cell transplantation (HSCT). All patients were women, positive for antinuclear and antitopoisomerase I antibodies, and had a disease duration <3 years calculated since the first non-Raynaud's symptom of SSc. The mean age was 41 years (range 18–44) and the mean disease duration was 21 months (range 16–32). All patients presented with ground-glass opacity on high-resolution CT scan, had a modified Rodnan skin thickness score ≥15 (of a maximum of 51) and had no cardiac and renal involvement. None of the patients had end-stage organ failure and no patient was extensively pretreated with cyclophosphamide. Previous treatments were calcium channel blockers and intravenous prostanoids. Bone marrow aspiration is routinely performed at our institution before each HSCT to exclude subclinical haematopoietic disorders. Before HSCT, patients underwent physical examination, laboratory testing and instrumental examination to evaluate internal organ involvement. All patients were assessed according to international guidelines.17 Bone marrow aspirations were performed to collect MSCs for assay just before prostanoid infusion and at least 1 month after the previous infusion. Calcium channel blockers were discontinued at least 3 weeks before bone marrow aspiration. Control MSCs were obtained from five healthy donors (five women; mean age 39 years (range 20–44)). Exclusion criteria for control subjects were familial histories of autoimmune diseases and/or chronic inflammatory diseases.

Isolation and culture of MSCs

Bone marrow cells were harvested from patients with SSc before HSCT and from healthy donors by aspiration from the posterior superior iliac crest. In order to enrich the total nucleated cell fraction, whole bone marrow aspirate was centrifuged for 10 min at 700g and the interface between plasma and the red cell pellet (buffy coat) was recovered, diluted 1:10 in Hanks' balanced salt solution (EuroClone, Milan, Italy) and counted. Cells were plated in 75 cm2 flasks (1.6×105 total nucleated cells/cm2) in low glucose Iscove's modified Dulbecco's medium (IMDM; with L-glutamine and HEPES 25 mM; EuroClone) with 50 µg/ml gentamicin (Schering-Plough, Milan, Italy) and 10% fetal bovine serum (FBS; Hyclone, South Logan, Utah, USA) and incubated at 37°C in a 5% CO2 atmosphere. When about 80% of the flask surface was covered, the adherent cells were incubated for 5–10 min at 37°C with 0.05% trypsin and 0.02% EDTA (Eurobio, Courtaboeuf, France), harvested, washed and resuspended in complete medium. Cell expansion was obtained with successive cycles of trypsinisation and reseeding. Cells were used until the third passage in culture.

Determination of colony-forming unit–fibroblastoid frequency and analysis of the osteogenic and adipogenic differentiation of MSCs

The number of colony-forming unit–fibroblastoid (CFU-F) colonies was used as a surrogate marker for MSC progenitor frequency. Two dishes (100 mm diameter) were seeded with 5×105 total nucleated cells from the bone marrow buffy coat. After incubation for 14 days, visible colonies formed by 50 or more cells were counted and reported as the number of CFU-F colonies/106 total nucleated cells seeded. The osteogenic and adipogenic differentiation of MSCs was assessed as previously described.15

Immunophenotyping of MSCs by flow cytometric analysis

First passage MSCs were analysed for the expression of surface antigens by flow cytometry. Aliquots were incubated with the following conjugated mouse monoclonal antihuman antibodies: phycoerythrin (PE)-conjugated CD34, fluorescein isothiocyanate (FITC)-conjugated CD45, PE-conjugated CD14, PE-conjugated CD29, FITC-conjugated CD44, PE-conjugated CD166, PE-conjugated CD90, PE-conjugated CD73, FITC-conjugated HLA-DP, HLA-DQ and HLA-DR, FITC-conjugated HLA-A, HLA-B and HLA-C (BD PharMingen, San Diego, California, USA) and PE-conjugated CD105 (Ancell, Bayport, Minnesota, USA). Non-specific fluorescence and cell morphological features were determined by incubation with isotype-matched mouse monoclonal antibodies (BD PharMingen). Gating acquisition was performed according to previously described methods.18

Cell stimulation

MSCs were seeded in 75 cm2 flasks (1.6×105 cells/cm2) and cultured in IMDM with 10% FBS. Cells were washed three times with serum-free medium and starved in IMDM with 2% FBS overnight. MSCs were then stimulated with IMDM supplemented with recombinant human VEGF (50 ng/ml), TGFβ (1 ng/ml) or SDF-1 (250 ng/ml) (all from PeproTech, Rocky Hill, New Jersey, USA) for 6 and 24 h. mRNA and protein expression of SDF-1 (SDF-1α), CXCR4, VEGF (VEGF-A), TGFβ1, TβRI and TβRII were evaluated by quantitative real-time PCR, western blotting and immunocytochemistry.

RNA purification, cDNA synthesis and quantitative real-time PCR

Total RNA was isolated from MSCs using the RNeasy Micro Kit (Qiagen, Milan, Italy). First strand cDNA was synthesised using the QuantiTect Reverse Transcription kit (Qiagen). For mRNA quantification, SYBR Green real-time PCR was performed using the ABI Prism 7000 Sequence Detection System (Applied Biosystems, Milan, Italy) with melting curve analysis. Predesigned oligonucleotide primer pairs were obtained from Qiagen (QuantiTect Primer Assay). The assay IDs were QT00087591 Hs_CXCL12, QT00223188 Hs_CXCR4, QT01682072 Hs_VEGFA, QT00000728 Hs_TGFB1, QT00083412 Hs_TGFBR1, QT00014350 Hs_TGFBR2 and QT00199367 Hs_RRN18S. The PCR mixture contained 1 μl cDNA, 0.5 μM sense and antisense primers, 10 μl 2× QuantiTect SYBR Green PCR Master Mix containing SYBR Green I dye, ROX passive reference dye, HotStarTaq DNA Polymerase, dNTP mix and MgCl2 (Qiagen). Amplification was performed according to a standard protocol recommended by the manufacturer. Non-specific signals caused by primer dimers or genomic DNA were excluded by dissociation curve analysis, non-template controls and samples without enzyme in the reverse transcription step. 18S ribosomal RNA was measured as an endogenous control to normalise for the amounts of loaded cDNA. Differences were calculated with the threshold cycle (Ct) and comparative Ct method for relative quantification. All measurements were performed in triplicate.

Western blotting

MSCs were scraped in ice-cold lysis buffer (50 mM Tris HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.25% sodium dodecyl sulfate (SDS)) supplemented with protease inhibitor cocktail (Roche, Mannheim, Germany). The solution was cleared by centrifugation for 30 min at 4°C at 15 000 rpm and assayed for protein content using Bradford's method. Fifty µg of proteins was electrophoresed in SDS 12–20% polyacrylamide gel under reducing conditions and blotted to a nitrocellulose membrane (Amersham Biosciences, Piscataway, New Jersey, USA). Membranes were blocked in 5% non-fat dry milk with 0.05% Tween-20 in phosphate buffered saline (PBS) for 1 h at room temperature and incubated overnight at 4°C with the following primary antihuman antibodies: rabbit polyclonal anti-SDF-1α (1:1000 dilution; Cell Signaling Technology, Danvers, Massachusetts, USA), anti-CXCR4 (1:2000; Abcam, Cambridge, UK), anti-VEGF-A (1:1000; Abcam), rabbit monoclonal anti-TGFβ1 (1:1000; Cell Signaling Technology), rabbit polyclonal anti-TβRI and anti-TβRII (both 1:1000; Abcam). After incubation with horseradish peroxidase-conjugated anti-rabbit IgG (Cell Signaling Technology), immune complexes were detected with the enhanced chemiluminescence detection system (Amersham Biosciences). Membranes were exposed to autoradiographic films (Amersham Biosciences) for 1–30 min. Blots were stripped and reprobed with rabbit monoclonal antihuman α-tubulin antibodies (1:1000; Cell Signaling Technology) to confirm similar loading of the gels and efficiency in electrophoretic transfer. Immunoreactive bands were quantified with densitometry using ImageJ software (NIH, Bethesda, Maryland, USA).

Immunocytochemistry and confocal laser scanning microscopy

MSCs were grown to semiconfluence on glass coverslips and then starved in IMDM with 2% FBS overnight. They were stimulated as described above and fixed in 3.7% buffered paraformaldehyde. Cells were permeabilised with 0.1% Triton X-100 in PBS. Non-specific antibody binding was blocked in PBS with 1% bovine serum albumin. The cells were incubated overnight with the following primary antihuman antibodies: mouse monoclonal anti-SDF-1α (1:100; R&D Systems, Minneapolis, Minnesota, USA), rabbit polyclonal anti-CXCR4 (1:50; Abcam), anti-VEGF-A (1:100; Abcam), anti-TGFβ1 (1:50; Abcam), anti-TβRI (1:50; Abcam) and anti-TβRII (1:100; Abcam). Alexa 488-conjugated goat anti-mouse or anti-rabbit IgG (1:200; Molecular Probes, Eugene, Oregon, USA) were used as secondary antibodies. Negative controls were obtained by incubation with isotype-matched normal IgG. MSCs were examined with a Leica TCS SP5 confocal laser scanning microscope (Leica Microsystems, Mannheim, Germany) equipped with a Leica PlanApo ×63 oil immersion objective and a HeNe/Argon laser source for fluorescence measurements. Series of optical sections (1024×1024 pixels each) at intervals of 0.8 µm were obtained and superimposed to create a single composite image.

ELISA

The levels of VEGF, SDF-1 (SDF-1α) and TGFβ1 in culture supernatants were measured with a Quantikine ELISA kit (R&D Systems) according to the manufacturer's instructions.

Preparation of MSC-conditioned medium

Confluent cultures of MSCs at basal condition or prestimulated with growth factors were washed twice with PBS and incubated overnight in IMDM with 2% FBS. The culture supernatant was centrifuged at 1500 rpm for 10 min and either used immediately or stored at −20°C.

In vitro capillary morphogenesis assay

Human dermal MVECs (Lonza, Milan, Italy) were cultured in complete endothelial cell growth medium (EGM) supplemented with EGM-MV BulletKit (Lonza). In vitro capillary morphogenesis assay was performed in 96-well plates covered with Matrigel (BD Biosciences, Bedford, Massachusetts, USA). Matrigel (50 µl; 10–12 mg/ml) was pipetted into culture wells and polymerised for 30–60 min at 37°C. MVECs (30×103 cells/well) were incubated in EGM containing 2% FBS, VEGF (50 ng/ml), TGFβ (1 ng/ml), SDF-1 (250 ng/ml) or in MSC-conditioned medium. In some experiments the MSC-conditioned medium was preincubated with goat polyclonal anti-VEGF-A, anti-SDF-1 or anti-TGFβ1 blocking antibodies at 3 µg/ml (all from R&D Systems). Irrelevant isotype-matched IgG were used to verify the specificity of the effect. The plates were photographed at 6 and 24 h. The results were quantified at 24 h by measuring the percentage field occupancy of capillary projections, as determined by image analysis. Six to nine photographic fields from three plates were scanned for each point.

Statistical analysis

Data are represented as mean±SD. The Student t test for independent data was used to test the probability of significant differences between groups; p<0.05 was considered statistically significant.

Results

Expansion of MSCs from patients with SSc and controls in culture

Bone marrow-derived SSc-MSCs and H-MSCs were expanded in culture. Culture-expanded confluent MSCs displayed both spindle-shaped cells and large flat cells (see figure 1A,B in online supplement). Morphological features were typical of MSCs. MSCs were used until the third-passage, and showed no significant differences in morphology or in proliferation rate.

Findings of the CFU-F assay

Total nucleated cells from the bone marrow buffy coat were used in the CFU-F assay. The mean±SD number of CFU-F colonies in cells from patients with SSc (47±24/106 total nucleated cells) was not different from that in control cells (51±9/106 total nucleated cells).

Immunophenotype of MSCs

SSc-MSCs and H-MSCs were uniformly positive for CD29, CD44, CD166, CD90, CD73, HLA-A, HLA-B, HLA-C and CD105. HLA-DP, HLA-DQ and HLA-DR were expressed in <4% of the population. There was no contamination by haematopoietic cells, as indicated by negative findings on flow cytometry for markers of haematopoietic lineage including CD14, CD34, and CD45. There was no statistically significant difference in the immunophenotype of SSc-MSCs and H-MSCs.

In vitro differentiation of SSc-MSCs and H-MSCs

Osteogenic differentiation of MSCs was determined after 21 days of stimulation. Alizarin red S staining showed aggregates or nodules of hydroxyapatite-mineralised matrices that were intensely red-stained in both SSc-MSCs and H-MSCs (data not shown). MSCs treated with adipogenic medium were successfully differentiated towards adipogenic lineages in both SSc patients and controls. Lipid vacuoles stained orange-red after 21 days (see figure 1C,D in online supplement).

Expression of bioactive molecules and receptors in SSc-MSCs and H-MSCs

We first evaluated the expression of the SDF-1/CXCR4 axis which is implicated in the recruitment of MSCs at sites of injury and in angiogenesis. Next, we analysed the expression of proangiogenic VEGF, as well as TGFβ1/TβRs that participate in vascular remodelling and fibrosis, in SSc-MSCs and H-MSCs at basal conditions and after different stimuli for 6 and 24 h. No significant differences were observed between 6 and 24 h stimulation experimental points in SSc-MSCs and H-MSCs (data not shown). Therefore, only the results of 6 h stimulation are shown in the figures.

SDF-1/CXCR4 expression

The findings of SDF-1 and CXCR4 mRNA and protein expression in SSc-MSCs and H-MSCs are shown in figures 1 and 2, respectively. At basal condition a statistically significant increase was seen in SDF-1 transcript levels in SSc-MSCs compared with H-MSCs (1.34±0.14 vs 1.0±0.19, p<0.005). SDF-1 mRNA expression was significantly upregulated after VEGF stimulation in both SSc-MSCs (p<0.005) and H-MSCs (p<0.001) (figure 1A). These results were confirmed at the protein level by western blotting and confocal laser scanning microscopy (CLSM) analyses (figure 1B,C). At basal condition, SSc-MSCs constitutively released higher levels of SDF-1 in the culture supernatant than H-MSCs (p<0.001) (figure 1D). Basal transcript levels of CXCR4 were significantly higher in SSc-MSCs than in H-MSCs (1.42±0.14 vs 1.0±0.05, p<0.001; figure 2A). Instead, western blotting and CLSM analyses showed that CXCR4 protein expression was similar in basal SSc-MSCs and H-MSCs (figure 2B,C). CXCR4 mRNA and protein were both significantly upregulated after VEGF, TGFβ and SDF-1 stimulation in both SSc-MSCs and H-MSCs (for mRNA, all p<0.001), showing the maximum effect after challenge with VEGF (figure 2). The induction of CXCR4 expression was more impressive in SSc-MSCs than in H-MSCs after all stimuli (figure 2). In particular, CLSM analysis revealed that, after VEGF and TGFβ stimulation, in SSc-MSCs CXCR4 was redistributed from the cytoplasm to the cell surface where it was localised at filopodia (figure 2C). Flow cytometric analysis confirmed the increased cell surface expression of CXCR4 (data not shown).

Figure 1

Expression of stromal cell-derived factor-1α (SDF-1α) in bone marrow-derived mesenchymal stem cells (MSCs) from patients with SSc (SSc-MSC) and healthy controls (H-MSC) at basal conditions and after challenge with vascular endothelial growth factor (VEGF), transforming growth factor β (TGFβ) and SDF-1. (A) Quantification of mRNA expression by real-time reverse transcription PCR. The basal level of SDF-1 mRNA in H-MSC was set to 1; the other results are normalised to this value. Bars represent the mean±SD values of triplicate determinations of five H-MSC and five SSc-MSC lines. *p<0.005 basal SSc-MSC vs basal H-MSC; **p<0.005 VEGF-stimulated SSc-MSC vs basal SSc-MSC; §p<0.001 VEGF-stimulated H-MSC vs basal H-MSC. (B) SDF-1 protein expression analysed by western blotting. The results of a representative experiment are shown. The expression levels of basal H-MSC were set to 100%; the other results are normalised to this value. α-tubulin was measured as a loading control for normalisation. Numbers on the right indicate molecular weight (kDa). (C) VEGF expression analysed by confocal microscopy. Representative microphotographs are shown. Original magnification ×63. (D) Determination of SDF-1 levels in the culture supernatant of H-MSC and SSc-MSC with ELISA. Bars represent the mean±SD values of triplicate determinations of 5 H-MSC and 5 SSc-MSC lines. *p<0.001 basal SSc-MSC vs basal H-MSC. SSc, systemic sclerosis.

Figure 2

Expression of CXCR4 in bone marrow-derived mesenchymal stem cells (MSCs) from patients with SSc (SSc-MSC) and healthy controls (H-MSC) at basal condition and after challenge with vascular endothelial growth factor (VEGF), transforming growth factor β (TGFβ) and stromal cell-derived factor-1 (SDF-1). (A) Quantification of mRNA expression by real-time reverse transcription PCR. The basal level of CXCR4 mRNA in H-MSC was set to 1; the other results are normalised to this value. Bars represent the mean±SD values of triplicate determinations of five H-MSC and five SSc-MSC lines. *p<0.001 basal SSc-MSC vs basal H-MSC; **p<0.001 vs basal SSc-MSC; §p<0.001 vs basal H-MSC. (B) CXCR4 protein expression analysed by western blotting. Results of a representative experiment are shown. The expression levels of basal H-MSC were set to 100%; the other results are normalised to this value. α-tubulin was measured as a loading control for normalisation. Numbers on the right indicate molecular weight (kDa). (C) CXCR4 expression analysed by confocal microscopy. Arrowheads indicate localisation of CXCR4 at filopodia. Representative microphotographs are shown. Original magnification ×63. SSc, systemic sclerosis.

VEGF expression

At basal condition, a statistically significant increase was detected in VEGF transcript levels in SSc-MSCs compared with H-MSCs (1.34±0.2 vs 1.0±0.1, p<0.05; figure 3A). These results were confirmed at the protein level (figure 3B,C). VEGF mRNA and protein expression levels were significantly increased after SDF-1 stimulation in both SSc-MSCs (for mRNA, p<0.001) and H-MSCs (for mRNA, p<0.005), with a more prominent induction in SSc-MSCs (figure 3A–C). SSc-MSCs consistently released higher levels of VEGF in the culture supernatant than H-MSCs (p<0.001), and VEGF secretion was significantly increased after SDF-1 stimulation in both SSc-MSCs and H-MSCs (both p<0.001 compared with unstimulated cells; figure 3D).

Figure 3

Expression of vascular endothelial growth factor (VEGF) in bone marrow-derived mesenchymal stem cells (MSCs) from patients with SSc (SSc-MSC) and healthy controls (H-MSC) at basal condition and after challenge with VEGF, transforming growth factor β (TGFβ) and stromal cell-derived factor-1 (SDF-1). (A) Quantification of mRNA expression by real-time reverse transcription PCR. The basal level of VEGF mRNA in H-MSC was set to 1; the other results are normalised to this value. Bars represent the mean±SD values of triplicate determinations of five H-MSC and five SSc-MSC lines. *p<0.05 basal SSc-MSC vs basal H-MSC; **p<0.001 SDF-1 stimulated SSc-MSC vs basal SSc-MSC; §p<0.005 vs basal H-MSC. (B) VEGF protein expression analysed by western blotting. Results of a representative experiment are shown. Bands are consistent with VEGF165 (22 kDa) and VEGF189 (28 kDa) isoforms. The expression levels of basal H-MSC were set to 100%; the other results are normalised to this value. α-tubulin was measured as a loading control for normalisation. Numbers on the right indicate molecular weight (kDa). (C) VEGF expression analysed by confocal microscopy. Representative microphotographs are shown. Original magnification ×63. (D) Determination of VEGF levels in the culture supernatant of H-MSC and SSc-MSC with ELISA. Bars represent the mean±SD values of triplicate determinations of five H-MSC and five SSc-MSC lines. *p<0.001 basal SSc-MSC vs basal H-MSC; **p<0.001 SDF-1 stimulated SSc-MSC vs basal SSc-MSC; §p<0.001 SDF-1 stimulated H-MSC vs basal H-MSC. SSc, systemic sclerosis.

TGFβ1/TβR expression

At basal condition, no difference was observed in TGFβ1 mRNA and protein expression between SSc-MSCs and H-MSCs (figure 4AC). In SSc-MSCs only, stimulation with VEGF, TGFβ and SDF-1 strongly upregulated the expression of TGFβ1 mRNA and protein (for mRNA, all p<0.001 compared with unstimulated cells), reaching the maximum effect after TGFβ challenge (figure 4A–C). The basal levels of TGFβ1 secreted by SSc-MSCs were not different from those in H-MSCs (figure 4D). After stimulation with VEGF and SDF-1, SSc-MSCs released significantly higher levels of TGFβ1 in the culture supernatant (both p<0.001 compared with unstimulated cells), but this effect was not observed in H-MSCs (figure 4D).

Figure 4

Expression of transforming growth factor β1 (TGFβ1) in bone marrow-derived mesenchymal stem cells (MSCs) from patients with SSc (SSc-MSC) and healthy controls (H-MSC) at basal condition and after challenge with vascular endothelial growth factor (VEGF), TGFβ and stromal cell-derived factor-1 (SDF-1). (A) Quantification of mRNA expression by real-time reverse transcription PCR. The basal level of TGFβ1 mRNA in H-MSC was set to 1; the other results are normalised to this value. Bars represent the mean±SD values of triplicate determinations of five H-MSC and five SSc-MSC lines. *p<0.001 vs basal SSc-MSC. (B) TGFβ1 protein expression analysed by western blotting. Results of a representative experiment are shown. The expression levels of basal H-MSC were set to 100%; the other results are normalised to this value. α-tubulin was measured as a loading control for normalisation. Numbers on the right indicate molecular weight (kDa). (C) TGFβ1 expression analysed by confocal microscopy. Representative microphotographs are shown. Original magnification ×63. (D) Determination of TGFβ1 levels in the culture supernatant of H-MSC and SSc-MSC with ELISA. Bars represent the mean±SD values of triplicate determinations of five H-MSC and five SSc-MSC lines. *p<0.001 vs basal SSc-MSC. SSc, systemic sclerosis.

At basal condition, a statistically significant decrease was detected in TβRI mRNA expression in SSc-MSCs compared with H-MSCs (0.64±0.02 vs 1.0±0.08, p<0.001; figure 5A). TβRI protein was also downregulated in basal SSc-MSCs compared with H-MSCs (figure 5B). VEGF stimulation strongly upregulated TβRI expression in SSc-MSCs (for mRNA, p<0.001), while the opposite effect was observed in H-MSCs. Moreover, TβRI mRNA and protein were significantly upregulated by TGFβ and SDF-1 both in SSc-MSCs and H-MSCs (figure 5A,B). For TβRII, the basal mRNA and protein expression levels were significantly higher in SSc-MSCs than in H-MSCs (for mRNA, 2.10±0.11 vs 1.0±0.02, p<0.001; figure 5C,D). In SSc-MSCs only, stimulation with VEGF, TGFβ and SDF-1 significantly increased TβRII expression at both the mRNA and protein levels (for mRNA, all p<0.001 compared with unstimulated cells; figure 5C,D). In particular, TβRII was clearly redistributed from the cytoplasm to focal adhesion contacts on the cell surface of SSc-MSCs after all stimuli (figure 5D).

Figure 5

Expression of TβRI and TβRII in bone marrow-derived mesenchymal stem cells (MSCs) from patients with SSc (SSc-MSC) and healthy controls (H-MSC) at basal condition and after challenge with vascular endothelial growth factor (VEGF), transforming growth factor β (TGFβ) and stromal cell-derived factor-1 (SDF-1). (A and C): Quantification of (A) TβRI and (C) TβRII mRNA expression by real-time reverse transcription PCR. The basal mRNA level in H-MSC was set to 1; the other results are normalised to this value. Bars represent the mean±SD values of triplicate determinations of five H-MSC and five SSc-MSC lines. *p<0.001 basal SSc-MSC vs basal H-MSC; **p<0.001 vs basal SSc-MSC; §p<0.005 vs basal H-MSC; §§p<0.05 vs basal H-MSC. (B and D): (B) TβRI and (D) TβRII protein expression analysed by western blotting and confocal microscopy. The expression levels of basal H-MSC were set to 100%; the other results are normalised to this value. α-tubulin was measured as a loading control for normalisation. Numbers on the right indicate molecular weight (kDa). Representative confocal microphotographs are shown. Arrowheads in (D) indicate localisation of TβRII at focal adhesion contacts. Original magnification ×63. SSc, systemic sclerosis.

Modulation of endothelial cell angiogenesis by SSc-MSC- and H-MSC-conditioned medium

To study whether MSCs were able to modulate endothelial cell angiogenesis in vitro we performed studies of capillary morphogenesis on dermal MVECs plated on Matrigel in the presence of MSC-conditioned medium. After 6 h of plating on Matrigel, MVECs showed abundant networks of branching cords of cells. By 24 h MVECs formed an interconnected network of anastomosing cells that had a honeycomb appearance (figure 6). The exogenous addition of proangiogenic growth factors (VEGF, SDF-1, TGFβ) or MSC-conditioned medium was able to augment in vitro capillary morphogenesis by MVECs (figure 6A). In particular, the proangiogenic effect of SSc-MSC-conditioned medium was comparable to that of recombinant human VEGF or SDF-1, and significantly greater than that of H-MSC-conditioned medium (p<0.001; figure 6A). The maximum proangiogenic effect was observed in the presence of conditioned medium of SDF-1-prestimulated SSc-MSCs (figure 6A). Preincubation with anti-VEGF-A, anti-SDF-1 or anti-TGFβ1 blocking antibodies strongly and significantly reduced the proangiogenic effect of SSc-MSC-conditioned medium (all p<0.001; figure 6B). The maximum inhibitory effect on SSc-MSC-conditioned medium-induced MVEC angiogenesis was observed with a combination of the three cytokine blocking antibodies (figure 6B).

Figure 6

Modulation of dermal microvascular endothelial cell (MVEC) angiogenesis by conditioned medium (c.m.) of SSc-MSC and H-MSC. Capillary morphogenesis was performed on Matrigel. (A) Representative light microscopy images of MVECs incubated for 24h with basal endothelial medium, recombinant human VEGF (50 ng/ml), and H-MSC or SSc-MSC c.m are shown. Capillary morphogenesis was quantified by measuring the percent field occupancy of capillary projections, as determined by image analysis. Six to nine photographic fields from 3 plates were scanned for each point. Bars represent the mean±SD values of triplicate determinations. Conditioned media of 5 H-MSC and 5 SSc-MSC lines were tested. *p<0.001 vs basal medium. **p<0.001 SSc-MSC c.m. vs H-MSC c.m. (B) Representative light microscopy images of MVECs incubated for 24h with SSc-MSC c.m, and SSc-MSC c.m preincubated with anti-VEGF, anti-TGFb1 and anti-SDF-1 blocking antibodies are shown. Capillary morphogenesis was quantified as described above. Bars represent the mean±SD values of triplicate determinations. The proangiogenic effect of SSc-MSC c.m was set to 100%; the other results are normalized to this value. Irrelevant goat IgG were used to verify the specificity of the effect of blocking antibodies. *p<0.001 vs SSc-MSC c.m.

Discussion

Our data show that bone marrow-derived MSCs from patients with early severe and rapidly progressive diffuse SSc overexpress bioactive mediators and proangiogenic growth factors that may allow them to participate in different pathways of the disease. The paracrine effects of SSc-MSCs appear to be further influenced by the local microenvironment, as these cells upregulate and release bioactive mediators in response to different exogenous stimuli. Moreover, we demonstrate that SSc-MSCs are able to promote dermal endothelial cell sprouting angiogenesis in vitro. Early endothelial injury is thought to be the primary event in the development of SSc.2 3 Therefore, in early-stage SSc, the proangiogenic behaviour of culture-expanded autologous MSCs might be of primary importance for therapeutic strategies to counteract vascular damage.

Recent experimental data have provided evidence that bone marrow progenitor cells may play a crucial role in the modulation of the angiogenic and fibrogenic processes.12 19 20 It has been reported that MSCs can produce a variety of cytokines and chemokines involved in the regulation of cell migratory properties, differentiation, proliferation and immunomodulation.11 12 MSCs can contribute to tissue repair by paracrine effects (eg, secretion of trophic mediators) and by entering injured areas after tissue damage as a consequence of local chemoattractants or by systemic mobilisation.12 MSCs expanded in culture can be introduced by direct injection into the damaged tissue or systemically by intravenous injection.9 11 12 Once MSCs arrive at the damaged site they can produce a wide array of bioactive factors potentially contributing to tissue repair.12

SSc is a devastating disease characterised by lack of angiogenesis and progressive fibrosis of the skin and internal organs.1 For this reason, potential stem cell-based therapies have received increasing interest.5 7,,9 The aim of the current study was to investigate the possible involvement of bone marrow-derived MSCs in the pathogenesis of SSc and their potential therapeutic properties. In particular, we evaluated the expression of molecules involved in the paracrine effects of MSCs and in the recruitment of MSCs at sites of injury, as well as key mediators of the angiogenic and fibrotic processes. Furthermore, we stimulated cultured MSCs with different cytokines (VEGF, TGFβ, SDF-1) that are implicated in the pathogenesis of SSc and increased in the serum and lesional tissues of patients with SSc.1 3 21,,24

MSCs display a restricted pattern of chemokine receptors, including CXCR4, favouring their migration into tissues upon specific chemotactic triggers such as SDF-1.25 These receptors represent the basis for MSC homing to multiple organs where they may undergo a programme of tissue-specific differentiation.25 In our study we found that, following stimulation with VEGF, TGFβ and SDF-1, CXCR4 was overexpressed in SSc-MSCs, suggesting an upregulation of this chemokine receptor in an attempt by the patient's bone marrow to respond to the injury by MSC mobilisation. Moreover, under stimuli, CXCR4 was redistributed from the cytoplasm to the cell surface indicating a possible activation of the SDF-1/CXCR4 axis. SSc-MSCs constitutively released higher levels of SDF-1 than H-MSCs. The maximum effect on SDF-1/CXCR4 upregulation was observed after VEGF stimulation. In turn, in SSc-MSCs challenged with SDF-1, VEGF expression and release were significantly increased. Previous studies have shown that SDF-1 and CXCR4 are upregulated by VEGF and that the interaction of SDF-1 with CXCR4 further amplifies neovascularisation by increasing VEGF release by endothelial cells.26 27 Therefore, this SDF-1/CXCR4/VEGF positive loop seems highly active in SSc-MSCs and could contribute to MSC recruitment after tissue injury and their participation in promoting angiogenesis. However, previous studies have shown that SSc-MSCs failed to differentiate into endothelial cells and responded to a less extent to VEGF and SDF-1 in chemoinvasion assays compared with control MSCs.15 On the other hand, in this study we show that SSc-MSCs are highly active in promoting dermal MVEC sprouting angiogenesis in vitro, and that MSC-derived cytokines are responsible for these proangiogenic effects. Therefore, SSc-MSCs might not be able to incorporate into damaged or newly formed vessels but may participate in vascular repair through paracrine effects on MVECs by releasing a wide array of proangiogenic and chemotactic factors such as VEGF and SDF-1. In agreement with this hypothesis, recent studies have shown that engineered MSCs overexpressing VEGF and SDF-1 promote angiogenesis and improve cardiac function after acute experimental myocardial infarction in rats.28 VEGF released by MSCs may stimulate the activation, proliferation and migration of vascular endothelial cells, significantly enhancing angiogenesis in the ischaemic tissues, while MSC-derived SDF-1 could play a major role in dictating the homing of CXCR4-expressing circulating endothelial progenitor cells to injured tissues.29,,32

The proangiogenic effect of MSCs has been also demonstrated in a murine model of hindlimb ischaemia.33 VEGF and other proangiogenic mediators (eg, basic fibroblast growth factor (bFGF), placental growth factor and monocyte chemoattractant protein-1 (MCP-1)) were detected in MSC-conditioned medium and VEGF and bFGF were also around the cells injected into tissues.12 It was also demonstrated that MSC-conditioned medium contains high amounts of angiogenic and anti-apoptotic factors such as VEGF, TGFβ and MCP-1 which inhibit apoptosis of endothelial cells cultured under hypoxic conditions and promote the formation of capillary-like structures in vitro.34 Recently, bone marrow-derived MSCs have been shown to support the formation of vessel-like structures by endothelial cells in vitro in a medium devoid of VEGF and bFGF. In addition to soluble angiogenic factors, MSCs provided extracellular matrix components that serve as a substrate for sprouting endothelial cells.12 In this context, our findings demonstrate that SSc-MSCs exhibit an intact, or even potentiated, proangiogenic paracrine machinery that could be considered in an autologous therapeutic approach. Our data also support the hypothesis that the local SSc microenvironment may influence this peculiar MSC phenotype. The ability to preferentially dock at sites of injured tissues adds to the regenerative properties of MSCs as it increases the likelihood of systemically delivered cells finding the areas where their paracrine effects are most needed in a therapeutic setting.12 In agreement with these findings, we recently reported that intravenous infusion of culture-expanded autologous bone marrow-derived MSCs could regenerate the peripheral vascular network in a case of SSc complicated by acute gangrene of the extremities.9 However, further studies using in vivo preclinical models will be required to fully evaluate the molecular basis of MSC paracrine effects.

We also demonstrated that stimulation with VEGF, TGFβ and SDF-1 strongly upregulated the expression of the multifunctional cytokine TGFβ1 in SSc-MSCs, reaching a maximum effect under TGFβ challenge in an autocrine way. SSc-MSCs could therefore participate in vascular repair by producing proangiogenic molecules and also vessel maturating factors such as TGFβ1.35 This may allow the formation of functionally mature vessels composed of endothelial cells and mural cells (ie, pericytes and vascular smooth muscle cells). In addition, the potential transition of MSCs to pericytes or smooth muscle cells on newly formed vessels might serve to stabilise the forming vasculature both in vitro and in vivo.11 36 Further studies will be required to verify whether SSc-MSCs might be able to differentiate into vascular mural cells.

However, in the complex scenario of SSc we cannot exclude the possibility that MSCs might also participate in the fibrotic process. In fact, we found that SSc-MSCs highly expressed the profibrotic cytokine TGFβ1 and responded in an autocrine manner to TGFβ stimulation, a feature that has been consistently shown in dermal and pulmonary fibroblasts of patients with SSc.1 Moreover, SSc-MSCs highly expressed TβRs, particularly TβRII which was redistributed to the cell surface after different stimuli. These data suggest that SSc-MSCs may not only release TGFβ1 in a paracrine manner but also might be highly responsive to autocrine TGFβ1. Furthermore, we cannot exclude the possibility that MSCs could differentiate into fibroblasts/myofibroblasts or act as supporting cells with trophic effects on circulating fibroblast precursors such as CXCR4-expressing fibrocytes via release of SDF-1.37 Indeed, an understanding of the source and mechanisms of the recruitment of fibroblasts/myofibroblasts is critical in the pathogenesis of fibrotic disorders. While these cells were classically thought to derive exclusively from resident fibroblasts, recent studies indicate that they may originate from epithelial/endothelial cells or even precursor cells.38 39 In this context, a profibrotic role of bone marrow-derived circulating fibroblast-like cells has been reported.20 39 In other studies, however, a protective rather than a profibrotic effect of MSCs was demonstrated.40 These contradictory data suggest that the role of bone marrow-derived precursor cells in the pathogenesis of fibrosis deserves further in-depth investigation.

Some limitations should be considered when interpreting these results. We studied only MSCs from a relatively small number of patients with early severe and rapidly progressive diffuse SSc that are usually selected for HSCT. As SSc is a complex heterogeneous disorder, further studies should address whether MSCs from patients with limited SSc or longstanding disease may exhibit the same genuine proangiogenic phenotype.

In conclusion, our results show that early diffuse SSc-MSCs constitutively overexpress and release bioactive mediators and proangiogenic growth factors and potentiate endothelial cell angiogenesis in vitro. Regarding the possible use of autologous MSCs for the treatment of SSc, further functional studies are required to investigate more fully the paracrine effects of these cells in different disease mechanisms such as the angiogenic and fibrotic processes.

References

Supplementary materials

  • Web Only Data

    Files in this Data Supplement:

Footnotes

  • SG and MM contributed equally to this work.

  • Funding This study was supported by Fondazione Cassa di Risparmio di Pistoia e Pescia.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval The study was approved by the Institutional Review Board. All subjects gave written informed consent.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.