Treatment of Collagen-Induced Arthritis Using Immune Modulatory Properties of Human Mesenchymal Stem Cells

Abstract

Mesenchymal stem cells (MSCs) have immune modulatory properties. We investigated the potential therapeutic effects of human bone marrow (BM)-, adipose tissue (AD)-, and cord blood (CB)-derived MSCs in an experimental animal model of rheumatoid arthritis (RA) and explored the mechanism underlying immune modulation by MSCs. We evaluated the therapeutic effect of clinically available human BM-, AD-, and CB-derived MSCs in DBA/1 mice with collagen-induced arthritis (CIA). CIA mice were injected intraperitoneally with three types of MSCs. Treatment control animals were injected with 35 mg/kg methotrexate (MTX) twice weekly. Clinical activity in CIA mice, degree of inflammation, cytokine expression in the joint, serum cytokine levels, and regulatory T cells (Tregs) were evaluated. Mice treated with human BM-, AD-, and CB-MSCs showed significant improvement in clinical joint score, comparable to MTX-treated mice. Histologic examination showed greatly reduced joint inflammation and damage in MSC-treated mice compared with untreated mice. Microcomputed tomography also showed little joint damage in the MSC-treated group. MSCs significantly decreased serum interleukin (IL)-1β, tumor necrosis factor (TNF)-α, IL-6, and interferon-γ and increased IL-10 and transforming growth factor-β levels. Tregs were increased in mice treated with MSCs compared to untreated or MTX-treated mice. Human BM-, AD-, and CB-MSCs significantly suppressed joint inflammation in CIA mice. The cells decreased proinflammatory cytokines and upregulated anti-inflammatory cytokines and induced Tregs. Therefore, our study suggests that the use of human BM-, AD-, and CB-MSCs could be an effective therapeutic approach for RA.

Keywords

Rheumatoid arthritis (RA)Mesenchymal stem cells (MSCs)Immune modulation Regulatory T cells (Tregs)

Introduction

Rheumatoid arthritis (RA) is a chronic inflammatory disease that primarily involves synovial joints. Uncontrolled active RA causes joint damage, disability, decreased quality of life, and cardiovascular and other comorbidities (35). During the past decade, disease-modifying antirheumatic drugs and biologic agents have been introduced for the treatment of RA (31,32,39). However, they have limitations, including incomplete treatment response, adverse effects requiring drug withdrawal, fall off in efficacy over time, high cost of biologic agents, and refractory cases (11,19,36). Consequently, there is a need to establish safe and effective advanced therapeutic modalities for RA to overcome the shortcomings of current treatments.

Mesenchymal stem cells (MSCs) are multipotent progenitor cells that can differentiate into tissues of mesenchymal lineage, including bone, cartilage, and adipose tissue (13,29). Most studies of MSCs have focused on the regenerative properties of the cells; however, MSCs also have unique immunoregulatory properties (6,22). MSCs have potent immunoregulatory effects that might be mediated through direct cell-to-cell contact or secretion of soluble factors. The relative ease of harvesting MSCs and their stable phenotype in culture make the cells an attractive tool for cellular therapy in alloimmunity, autoimmunity, and inflammation (22). This phenomenon has led to an increasing number of clinical trials, such as in graft-versus-host disease (21), Crohn's disease (8,14), and multiple sclerosis (9).

With regard to RA, several studies have reported on the effects of allogenic or xenogenic MSC treatment in collagen-induced arthritis (CIA) mice, a representative animal model of RA (2,17,27). However, the therapeutic effects of MSCs, which derive from three sources, bone marrow (BM), adipose tissue (AD), and cord blood (CB), and the specific mechanisms underlying their immune modulatory effects in models of CIA are not fully understood.

In this study, we investigated and compared the therapeutic efficacy of human BM-, AD-, and CB-derived MSCs in CIA. We determined the optimal cell dosages and injection frequencies for therapeutic efficacy and assessed time- and dose-dependent changes in the therapeutic response. Furthermore, to explore the mechanism underlying immune modulation by MSCs in CIA, we examined time- and dose-dependent changes in the levels of proinflammatory and anti-inflammatory cytokines and regulatory T cells (Tregs) after the MSC administration.

Materials and Methods

Treatment Protocol for CIA

Treatment was begun after the onset of disease, when arthritis had become well established, approximately 3 weeks after the primary immunization, and clinical assessment was continued for the following 4 weeks.

The first part of the experiment was designed for the purpose of investigating and comparing the therapeutic efficacy of the three types of MSCs and exploring the mechanism of immune modulation by MSCs. Mice were injected with MSCs according to the two administration schedules. In the first schedule, CIA mice were injected intraperitoneally daily with 100 μl phosphate-buffered saline (PBS control, n = 5; Invitrogen, Carlsbad, CA, USA) or 100 μl PBS containing 1 × 10⁶ BM-MSCs, AD-MSCs, or CB-MSCs for 5 consecutive days (total 5 × 10⁶ cells, n = 5 per each group). In the second schedule, mice were injected intraperitoneally with 100 μl PBS containing 2.5 × 10⁶ BM-MSCs, AD-MSCs, or CB-MSCs twice over a 3-day interval (total 5 × 10⁶ cells, n = 5 per each group). To compare therapeutic efficacy, treatment control mice (n = 5) were injected intraperitoneally with 35 mg/kg methotrexate (MTX) twice weekly for 4 weeks.

The second part of the experiment was for the purpose of exploring the mechanism of immunomodulation by MSCs. We determined the optimal cell dosages for therapeutic efficacy and to assess time-dependent changes in therapeutic response. To investigate dose-dependent changes in therapeutic efficacy in CIA mice, human BM-MSCs (2.5 × 10⁵, 2.5 × 10⁶, or 2.5 × 10⁷) were injected intraperitoneally twice over a 3-day interval (total 5 × 10⁵, 5 × 10⁶, or 5 × 10⁷ cells, n = 5 per each group). To assess time-dependent changes in therapeutic responses, human BM-MSCs (2.5 × 10⁶) were injected twice over a 3-day interval (total 5 × 10⁶, n = 20). Twenty mice were enrolled, and five mice were sacrificed according to the experimental schedule every week in order to elucidate the changes in levels of serum cytokines and Treg levels at each time point. Finally, all mice were sacrificed 28 days after final MSC injection. The serum, lymph nodes, spleen, and limbs of all animals were collected for analysis.

Induction of CIA

All procedures involving animals were carried out in accordance with the Laboratory Animals Welfare Act, the Guide for the Care and Use of Laboratory Animals, and the Guidelines and Policies for Rodent Experiments provided by the Institutional Animal Care and Use Committee of Yonsei University Health System, Seoul, Korea. Approval by our institutional review board was obtained (2010-0372-1, Yonsei University, College of Medicine, Seoul, Korea).

Male DBA/1J mice (8 weeks old; Central Lab Animal, Inc., Seoul, Korea) were injected intradermally at the base of the tail with 200 μg bovine type II collagen (CII; Chondrex, Redmond, WA, USA) and emulsified in Freund's complete adjuvant (1:1, v/v; Chondrex) containing 200 μg Mycobacterium tuberculosis H37Ra (Chondrex). Two weeks later, the mice were given intradermal booster injections of 100 μg CII in incomplete Freund's adjuvant (1:1, v/v; Chondrex).

Mice were monitored twice weekly for signs of arthritis based on paw swelling and arthritis scores. Clinical arthritis was scored on a scale of 0–4 (3). Each paw was graded, and the grades were summed to yield the arthritis score for each animal (maximum possible score, 16). Arthritis scoring was performed by two independent observers. Mice were randomly distributed to each treatment group (n = 5) at 1 week after the second booster injection of CII when arthritis scores were between 4 and 6. Mice with an arthritis score below 4 or one paw with an arthritis score of 4 at that time point were excluded from the experiment to minimize the variance of arthritis scores between groups.

Culture of BM-, AD-, and CB-MSCs

All frozen stocks of MSCs were provided at passage 1 by the Cell Therapy Center, Severance Hospital (Yonsei University, College of Medicine, Seoul, Korea) after being obtained with donor consent.

Adipose-derived (female, 32 years old), bone marrow-derived (male, 41 years old), and cord blood-derived (female, 42 weeks full-term baby) MSCs were characterized by fluorescence-activated cell sorting (FACS) analysis for typical surface antigens: surface expression of CD29, CD44, CD73, CD90, and CD105 and absence of surface expression of CD45, CD34, CD14, and human leukocyte antigen (HLA)-DR. For FACS analysis, cells were incubated with fluorescent primary antibodies for 30 min. All antibodies were diluted 1:200. The labeled cells were analyzed using a BD LSRΠ (BD Biosciences, San Jose, CA, USA). To test the multipotency of the cells, MSCs were treated in adipogenic, chondrogenic, and osteogenic media (R&D Systems, Minneapolis, MN, USA) for 14 days following the manufacturer's protocol. Standard Oil red O (Abcam, Cambridge, UK), Alcian blue (Abcam), and von Kossa (Abcam) stains were used according to manufacturer's protocol to identify adipocyte, chondrocyte, and osteoblast cells, respectively (20,40).

Cells were plated at a density of 1 × 10⁶ cells per 10-cm cell culture dish (BD Biosciences) with Dulbecco's modified Eagle medium (low glucose) supplemented with 10% fetal bovine serum, 1% glutamine, 1% nonessential amino acids, 0.01% 2-mercaptoethanol, and 1% penicillin–streptomycin (all from Invitrogen). MSCs were cultured at 37°C in an atmosphere containing 5% CO₂, and the medium was changed every 3–4 days. All MSCs were cultured until passage 6, harvested, and stored in liquid nitrogen until use.

Histological and Immunohistochemical Assessments of CIA

At day 28 from final MSC injection, mice were anesthetized with a standard dose of Zoletil (Virbac, Carros, France) and Rompun (Bayer, Barmen, Germany) and euthanized for analysis. Formalin (10%; Merck & Co. Inc., Kenilworth, NY, USA)-fixed limbs were decalcified in EDTA (30%; Merck) for 4 weeks and embedded in paraffin (Merck) using standard histologic techniques. Serial 4-μm sections were cut and stained with hematoxylin and eosin (Sigma-Aldrich, St. Louis, MO, USA) to examine morphologic features and determine the histological arthritis score. Sections were evaluated histopathologically and scored for inflammatory cell infiltration, synovial hyperplasia, and bone erosion, according to previously published criteria (4).

For immunohistochemical analysis, paw and knee joint sections were prepared and incubated with specific antibodies directed against murine tumor necrosis factor-α (TNF-α; 1:500; Hycult Biotechnology, Uden, Netherlands), interleukin (IL)-1β (1:500; Santa Cruz Biotechnology, Santa Cruz, CA, USA), IL-6 (1:250; Santa Cruz Biotechnology), and IL-10 (1:1,000; Abcam) followed by horseradish peroxidase-conjugated secondary antibodies and staining with diaminobenzidine (Dako, Glostrup, Denmark). The antigen retrieval process was done by microwave oven heating and enzyme digestion using protease (Sigma-Aldrich) before the application of the primary antibodies (38).

Expression of markers in the synovial tissues of paw and knee joints was scored semiquantitatively on a 4-point scale (0–3) independently by two blinded observers, and the average scores were calculated (23).

Microcomputed Tomography (MicroCT)

All groups of mice were observed for 28 days after final MSC injection, sacrificed, and the legs were excised and fixed in 4% formalin for 2 days. The paws (from the tip of the toes to the end of the ankle) were scanned by NanoFocus Ray microCT scanner (Nano Focus Ray Inc., Chun-ju, Korea). Images were acquired at 80 kVp, 150 mA, and 5 s/frame with 360 views. The estimated radiation dose was approximately 6.9 mGy using image acquisition protocol and reconstructed into three-dimensional structures and evaluated using NFR Polarys software (Exxim Computing Corporation, Pleasanton, CA, USA).

Measurement of Inflammatory Cytokines

Serum levels of the inflammatory cytokines murine IL-1β, IL-6, interferon (IFN)-γ, TNF-α, and IL-10 were determined using the Luminex multiplex cytokine assay (Luminex 200 System; Millipore, Billerica, MA, USA) according to the manufacturer's recommendations. The level of the cytokine transforming growth factor (TGF)-β was measured using a commercially available mouse TGF-β1 enzyme-linked immunosorbent assay (ELISA) kit (R&D Systems).

Separation of T Cells and Tregs by Flow Cytometry

Lymph nodes and spleens were collected from all experimental animals. Lymphocytes and splenocytes were obtained by mechanical shredding, collected in complete RPMI medium (Invitrogen), and pooled within each experimental group. T lymphocytes were isolated using a Pan T Cell Isolation Kit (Miltenyi Biotec, Bergisch Gladbach, Germany) by depletion of non-T cells. Tregs were stained with anti-CD4-phycoerythrin, anti-CD25-peridinin chlorophyll protein (BD Pharmingen, San Diego, CA, USA), and anti-Helios-allophycocyanin (BioLegend, San Diego, CA, USA). Cells were fixed and permeabilized with Fix/Perm buffer (eBioscience, San Diego, CA, USA) and stained with anti-forkhead box P3 (FoxP3)-fluorescein isothiocyanate (eBioscience). Cells were analyzed on a FACS Calibur flow cytometer using Flowjo software (Tree Star, Inc., Ashland, OR, USA).

Treg Suppression Assay

To measure the suppressive capacity of Tregs generated in the presence of MSCs, CD4⁺CD25⁺ Tregs, and CD4⁺ T cells were isolated from spleens and stained with 5 mM carboxyfluorescein diacetate succinimidyl ester (CFSE; eBioscience). Anti-CD3/CD28 antibodies (5 μg/ml; BD Biosciences) were suspended in RPMI-1640 (Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen) as stimulator. CD4⁺ T cells were cocultured with CD4⁺CD25⁺ Tregs at a ratio of 10:1 (1 × 10⁵ CD4⁺:1 × 10⁴ CD4⁺CD25⁺). Suppression of T-cell proliferation was measured by the reduction in the CFSE concentration using flow cytometry (FACS Caliber; BD).

Measurement of Immunoglobulin G Antibodies to Collagen Type II (CII)

Serum was collected from animals in each group of mice 28 days after final MSC injection and stored at −70°C until assayed. Serum anti-CII immunoglobulin G (IgG) levels were determined using a commercially available ELISA kit (Chondrex) according to the manufacturer's instructions.

Statistical Analysis

All values represent the mean ± standard deviation (SD). Statistical analysis of group differences was carried out using a one-way analysis of variance (ANOVA) test (SPSS 12.0; IBM, Armonk, NY, USA). A value of p < 0.05 was considered statistically significant after Bonferroni correction.

Results

Therapeutic Effects of BM-MSCs, AD-MSCs, and CB-MSCs in CIA Mice

Treatment with each type of MSC attenuated the severity of arthritis significantly during the progression of CIA (Fig. 1A,B). In particular, the clinical arthritis score of mice treated with 5 × 10⁶ BM-MSCs decreased to approximately 50% of that of untreated mice. AD-MSC- and CB-MSC-treated mice also showed a decrease in clinical arthritis scores, but to a lesser extent than seen in BM-MSC-treated mice. No statistically significant difference was observed in arthritis severity of mice treated with 1 × 10⁶ MSCs daily for 5 days and those treated with 2.5 × 10⁶ MSCs twice over a 3-day interval, suggesting that the injection schedule did not influence the outcome.

Figure 1.

Therapeutic effects of BM-, AD-, and CB-MSCs in CIA. (A and B) The severity of arthritis, assessed by clinical arthritis scoring, was significantly lessened in mice treated with MSCs compared to untreated CIA mice. Mice were intraperitoneally injected daily with 1 × 10⁶ BM-, AD-, or CB-MSCs for 5 days (A) or with 2.5 × 10⁶ BM-, AD-, or CB-MSCs twice over a 3-day interval (B). (C) Histopathological evaluation and hematoxylin and eosin staining (middle rows) showed severe inflammation of hind legs and synovial proliferation with cartilage destruction in untreated CIA mice. B, bone; C, cartilage; JC, joint cavity; ST, synovial tissue. Fields shown in upper row of stained sections are shown at higher magnification in panels directly below. MicroCT analysis (bottom panel) showed severe bone erosion in untreated mice, with markedly less joint destruction in MSC-treated mice. Conditions indicated by numbers 1–9 are described at the bottom of the figure (n = 5 per each group). (D) MSC treatment significantly reduced inflammatory cell infiltration, synovial hyperplasia, and bone erosion 28 days after MSC final injection. (E) Immunohistochemical analysis showed that inflammatory cytokines IL-1β, IL-6, and TNF-α were highly expressed in the joints of untreated CIA mice. IL-1β, IL-6, and TNF-α staining were decreased in MSC-treated mice compared to untreated mice. Expression of these cytokines was less in BM-MSC- and AD-MSC-treated mice than in CB-MSC-treated mice. IL-10 expression was significantly induced in MSC-treated mice. All values are the mean ± SD. Scale bars: 200 um. *p < 0.05 versus untreated CIA mice.

The inflammatory signs of arthritis, such as erythema or swelling, were marked in untreated CIA mice. Histological analysis of the knee joints revealed severe inflammatory cell infiltration, synovial hyperplasia, and bone and cartilage erosion, whereas these parameters were significantly reduced in all BM-MSC-, AD-MSC-, and CB-MSC-treated mice. The BM-MSC-treated mice showed the best therapeutic result of the three types of MSCs (Fig. 1C,D). These results suggest that the three types of MSCs have therapeutic efficacy in CIA mice, comparable to that of MTX.

Three-dimensional microCT analysis showed severe cartilage destruction and bone erosion in untreated control mice, whereas MSC- and MTX-treated groups showed markedly less joint destruction (Fig. 1C).

MSCs Reduced Expression of IL-1β, IL-6, and TNF-α and Increased Expression of IL-10 in Inflamed Joints of CIA Mice

Immunohistochemical analysis of paw and knee joint tissue in untreated CIA mice showed significant immunopositive staining for IL-1β, IL-6, and TNF-α that was localized primarily in synovium. IL-1β, IL-6, and TNF-α staining were decreased in MSC-treated mice compared to untreated mice. Expression of these cytokines was less in BM-MSC- and AD-MSC-treated mice than in CB-MSC-treated mice. IL-10 expression was significantly induced in MSC-treated mice (Fig. 1E).

MSCs Decreased Serum Levels of IL-1β, IL-6, and TNF-α and Increased Levels of IL-10

We measured serum cytokines of all mice 4 weeks after MSC final injection. The inflammatory cytokines IFN-γ, IL-1β, IL-6, and TNF-α were decreased in MTX-treated and all MSC-treated mice compared to untreated mice (Fig. 2A–D). In contrast, IL-10 levels were significantly increased in all MSC-treated mice, but not in MTX-treated mice (Fig. 2E).

Figure 2.

Serum proinflammatory and anti-inflammatory cytokine and Treg levels. (A–E) IFN-γ, IL-6, IL-1β, and TNF-α were significantly reduced in MTX-treated mice and all MSC-treated mice. IL-10 levels were significantly increased in all MSC-treated, but not in MTX-treated, mice. (F) Flow cytometric analysis showed that expression of CD4⁺CD25⁺FoxP3⁺ Tregs was significantly increased in lymph nodes of all MSC-treated, but not in MTX-treated, mice. (G and H) Quantitative analysis of CD4⁺CD25⁺FoxP3⁺ Tregs in lymph nodes (G) and spleen (H). The number of Tregs was slightly higher in the BM-MSC-treated group than in the AD-MSC- and CB-MSC-treated groups. Conditions indicated by numbers 1–9 are described below (n = 5 per each group). *p < 0.05 versus untreated CIA mice. Values are the mean ± SD.

Induction of Tregs Following Administration of MSCs in CIA Mice

Mice treated with BM-MSCs, AD-MSCs, and CB-MSCs had significantly higher numbers of CD4⁺CD25⁺ FoxP3⁺ Tregs in both lymph node and spleen than control mice (Fig. 2F). The number of Tregs was slightly higher in the BM-MSC-treated group than in the AD-MSC- and CB-MSC-treated groups (Fig. 2G,H). These results show that MSC treatment induced the generation of CD4⁺CD25⁺FoxP3⁺ Tregs in CIA mice.

BM-MSCs Suppressed the Severity of CIA in a Time- and Dose-Dependent Manner

According to our data, BM-MSCs had better therapeutic efficacy than the other types of MSCs. Thus, further experiments were performed to explore the mechanism of immune modulation by MSCs, treating the mice with two injections of 2.5 × 10⁵, 2.5 × 10⁶, or 2.5 × 10⁷ BM-MSCs. To investigate dose- and time-dependent changes of arthritis severity in CIA mice after treatment, we assessed the clinical arthritis score and histopathological changes in the hindpaw joints of these mice.

BM-MSC treatment resulted in a significant reduction in clinical arthritis scores, indicating a strong therapeutic effect, in a dose- and time-dependent manner (Fig. 3A). However, there was no significant difference in therapeutic outcome between treatment with 5 × 10⁶ and 5 × 10⁷ BM-MSCs. The decreases in inflammatory cell infiltration, synovial hyperplasia, and cartilage and bone erosion were less marked in mice treated with 5 × 10⁵ BM-MSCs than in those treated with the two higher doses (Fig. 3B,C). Treatment with 5 × 10⁶ BM-MSCs ameliorated the inflammatory cell infiltration and synovial hyperplasia in a time-dependent manner (Fig. 3D,E).

Figure 3.

Dose- and time-dependent outcomes of arthritis in CIA mice after BM-MSC treatment. (A) Clinical arthritis scores were decreased in all BM-MSC- and MTX-treated mice compared with untreated mice (n = 5 per each group). (B and C) Histologic examination and hematoxylin and eosin staining showed robust infiltration of inflammatory cells and destruction of bone structure in untreated mice. Mild infiltration of inflammatory cells and synovial hyperplasia were observed in mice treated with 5 × 10⁶ and 5 × 10⁷ BM-MSCs. (D and E) Treatment with 5 × 10⁶ BM-MSCs ameliorated the inflammatory cell infiltration and synovial hyperplasia in a time-dependent manner. Square outlined areas indicate fields shown at higher magnification in panels directly below. Values are the mean ± SD. Scale bars: 200 μm. *p < 0.05 versus untreated CIA mice.

Dose- and Time-Dependent Changes in Serum Cytokine Levels

To elucidate the mechanism underlying the reduction in severity of CIA in mice treated with BM-MSCs, we examined levels of proinflammatory and anti-inflammatory cytokines in mice 4 weeks after BM-MSC final injection.

The proinflammatory cytokines IFN-γ, IL-6, IL-1β, and TNF-α were significantly decreased in mice treated with 5 × 10⁶ and 5 × 10⁷ BM-MSCs and MTX, but not in mice treated with 5 × 10⁵ cells (Fig. 4A–D). In contrast, the levels of the anti-inflammatory cytokines IL-10 and TGF-β were significantly increased in 5 × 10⁶ and 5 × 10⁷ cells of BM-MSC-treated CIA mice, but not in MTX-treated mice. The increases in the levels of the anti-inflammatory cytokines were much smaller in mice treated with 5 × 10⁵ BM-MSCs than in those treated with the two higher doses, and there was no significant difference in IL-10 and TGF-β levels between mice treated with 5 × 10⁶ and 5 × 10⁷ BM-MSCs (Fig. 4E,F).

Figure 4.

Dose- and time-dependent changes in levels of proinflammatory and anti-inflammatory cytokines. (A–F) Levels of IFN-γ, IL-6, TNF-α, and IL-1β were significantly decreased in mice treated with 5 × 10⁶ and 5 × 10⁷ BM-MSCs and MTX, but not in mice treated with 5 × 10⁵ cells. Levels of IL-10 and TGF-β were significantly increased in 5 × 10⁶ and 5 × 10⁷ cells of BM-MSCs treated CIA mice, but not in MTX-treated mice (1 = wild type, 2 = CIA untreated, 3 = MTX treatment, 4 = BM-MSC 5 × 10⁵, 5 = BM-MSC 5 × 10⁶, 6 = BM-MSC 5 × 10⁷). (G–L) Levels of IFN-γ, IL-6, TNF-α, and IL-1β were decreased, while levels of IL-10 and TGF-β were increased, 2 weeks after treatment with 5 × 10⁶ BM-MSCs. Interestingly, increased levels of TGF-β were observed 1 week after treatment. Values are the mean ± SD. *p < 0.05 versus untreated CIA mice (n = 5).

We also investigated the changes in cytokine levels at different time points after administration of 5 × 10⁶ BM-MSCs. One week after treatment, no significant changes were seen in levels of any of the cytokines, with the exception of TGF-β, which, interestingly, was increased. However, 2 weeks after treatment, IFN-γ, IL-6, IL-1β, and TNF-α were significantly decreased (Fig. 4G–J), and IL-10 and TGF-β levels were increased (Fig. 4K,L). Clinical arthritis scores also began to improve at the 2-week time point.

Induction of Tregs by BM-MSCs in Mice with CIA

We investigated the proportion of Tregs in lymph nodes and spleen of CIA mice at different time points after administration of different doses of BM-MSCs. The induction of Tregs was significantly increased in lymph nodes (Fig. 5A) and spleen (Fig. 5B) of mice treated with 5 × 10⁶ and 5 × 10⁷ BM-MSCs, compared to MTX-treated and untreated CIA mice (Fig. 5C). Tregs were not elevated in mice treated with 5 × 10⁵ BM-MSCs, suggesting that this dose is insufficient for induction of Tregs, whereas a dose of 5 × 10⁶ cells is sufficient; no difference was seen between doses of 5 × 10⁶ and 5 × 10⁷ BM-MSCs. In addition, the induction of Tregs was significantly increased 2 weeks after treatment (Fig. 5D–F), when the clinical arthritis scores began to improve and changes were seen in levels of the anti-inflammatory cytokines IL-10 and TGF-β.

Figure 5.

Induction of Tregs by BM-MSCs in mice with CIA. (A–C) Expression of CD4⁺CD25⁺FoxP3⁺ Tregs was significantly increased in lymph nodes (A) and spleen (B) of mice treated with 5 × 10⁶ and 5 × 10⁷ BM-MSCs, compared with untreated and MTX-treated mice (p < 0.05) (n = 5 per each group) (1 = wild type, 2 = CIA untreated, 3 = MTX treatment, 4 = BM-MSC 5 × 10⁵, 5 = BM-MSC 5 × 10⁶, 6 = BM-MSC 5 × 10⁷). Results of flow cytometric analysis are shown in (C). (D–F) Expression of Tregs was significantly increased in lymph nodes (D) and spleen (E) of CIA mice 2 weeks after treatment with 5 × 10⁶ BM-MSCs, when the clinical scores and levels of anti-inflammatory cytokine IL-10 began to improve. Results of flow cytometric analysis are shown in (F). Values are the mean ± SD. *p < 0.05 versus untreated CIA mice.

Changes in Helios Expression and Treg Suppressive Activity by MSC Treatments

To identify the activity of Tregs induced by BM-, AD-, and CB-MSCs treatment, we investigated the expression of Helios, a transcription factor that is preferentially expressed by Tregs and binds the FoxP3 promoter region of Tregs in the mammalian immune system. Helios expression was significantly increased, consistent with the increased expression of FoxP3 4 weeks after BM-, AD-, and CB-MSC treatment (Fig. 6A). We also assessed the functionality of Tregs using an in vitro Treg suppression assay. Tregs showed potent suppressive activity, through inhibition of T-cell proliferation in each MSC-treated group compared to untreated CIA mice (Fig. 6B); these increases were slightly higher in Tregs induced by BM-MSCs than those by AD-MSCs and CB-MSCs. Helios expression and inhibition of T-cell proliferation were observed slightly higher in the BM-MSC group.

Figure 6.

Helios expression and Treg suppressive activity. (A) Levels of FoxP3⁺Helios⁺ Tregs were increased 28 days after treatment with 5 × 10⁶ BM-, AD-, and CB-MSCs, compared with untreated CIA mice (n = 5 per each group). (B) Treg suppression assay showed Treg suppressive activity in BM-, AD-, and CB-MSC-treated mice compared to untreated CIA mice. The red line represents the Tregonly group, and the blue line represents the proliferation of CD4⁺ T cells (n = 5 per each group). CFSE, carboxyfluorescein diacetate succinimidyl ester.

To identify the activity of Tregs induced by AD-, BM-, and CB-MSC treatment, we investigated the expression of Helios, a transcription factor that is preferentially expressed by Tregs and binds the FoxP3 promoter region in Tregs in the mammalian immune system. Helios expression was significantly increased, consistent with the increased expression of FoxP3 4 weeks after AD-, BM-, and CB-MSC treatment (Fig. 6A). We also assessed the functionality of Tregs using an in vitro Treg suppression assay. Tregs showed potent suppressive activity through inhibition of T-cell proliferation in the AD-, BM-, and CB-MSC-treated groups compared to untreated CIA mice (Fig. 6B); however, these increases were slightly smaller than those induced by BM-MSCs.

Dose- and Time-Dependent Change in Anti-Collagen II IgG (IgG-CII) Titer in BM-MSC-Treated CIA Mice

Serum IgG-CII levels in mice were measured 4 weeks after treatment with BM-MSCs to determine whether the treatment was associated with a change in the humoral immune response to type II collagen. Serum IgG-CII concentration was reduced in mice treated with 5 × 10⁶ and 5 × 10⁷ BM-MSCs and in MTX-treated mice (Fig. 7A), and it was significantly decreased 2 weeks after treatment with 5 × 10⁶ BM-MSCs (Fig. 7B).

Figure 7.

Dose- and time-dependent changes in anti-collagen II IgG titers in BM-MSC-treated CIA mice. (A) Serum concentration of anti-CII IgG was reduced in mice treated with 5 × 10⁶ and 5 × 10⁷ BM-MSCs or MTX. (B) Serum concentration of anti-CII IgG was significantly decreased 2 weeks after treatment with 5 × 10⁶ BM-MSCs (n = 5 per each group). Values are the mean ± SD. *p < 0.05 versus untreated mice.

Discussion

Several studies have reported on the therapeutic effects of allogenic or xenogenic MSC treatment in collagen-induced arthritis (CIA) mice, a representative animal model of RA (2,17,27). However, others have failed to demonstrate such effects (5,37). Potential reasons for the discrepancies among these results include the following: source of MSCs (murine, syngeneic vs. human, allogeneic), tissue of origin (BM, AD, CB), timing of treatment, number of stem cells injected, route of injection (IV, IP, IA), and treatment regimen (a single injection of MSCs vs. daily injections for 5 consecutive days) (28).

Therefore, we tried to find the best condition for clinical application of human MSCs for RA treatment. We compared the three types of human MSCs (BM-, AD-, CB-MSCs) and determined the optimal number of MSCs injected and the treatment regimen for the CIA model, which shares clinical, histological, and immunologic features with RA. Also, to explore the mechanism underlying immune modulation by MSCs in CIA, we examined the levels of proinflammatory and anti-inflammatory cytokines and Tregs after MSC administration. To the best of our knowledge, there is no study directly comparing the difference in the anti-inflammatory effect among the three types of MSCs in vivo.

BM-, AD-, and CB-MSCs share a common phenotype (CD73⁺, CD90⁺, CD105⁺, CD45^–, CD14^–, CD11b^–, and CD34^–) and can differentiate into adipocytes, osteoblasts, and chondrocytes. These three types of MSCs have an immunosuppressive effect mediated by soluble mediators such as indoleamine 2, 3-dioxygenase (IDO), nitric oxide (NO), prostaglandin E2 (PGE2), IL-10, TGF-β, hepatocyte growth factor (HGF), heme oxygenase-1, and human leukocyte antigen (HLA)-G5 (13). However, there is a difference in the amount of soluble cytokine receptors and the gene expression of paracrine factors between each of the MSCs (15,18). Soluble TNF receptor 1 and soluble vascular endothelial growth factor receptor 1 (VEGFR1) are highly expressed in BM-MSCs (15). The mRNA expression of vascular endothelial growth factor-D (VEGF-D), IL-8, and insulin-like growth factor-1 (IGF-1) are significantly higher in AD-MSCs than in other MSCs (18).

We demonstrated that the administration of BM-, AD-, and CB-MSCs reduced clinical arthritis scores, synovial inflammation and proliferation, and cartilage and bone erosion in CIA mice. The therapeutic effect of MSCs successfully lasted for 4 weeks after a one-time MSC administration, and the therapeutic efficacy of MSCs was comparable to that of MTX. Our results are consistent with those of previous studies demonstrating that human AD-MSCs (17) and human CB-MSCs (27) reduce the severity of experimental arthritis. We observed that BM-MSCs had a greater therapeutic efficacy than AD-MSCs or CB-MSCs in CIA in view of the severity of arthritis and histopathological evaluation. IL-1β, IL-6, and TNF-α were expressed less in BM-MSC- and AD-MSC-treated mice than in CB-MSC-treated mice for immunohistochemical analysis. Among the three types of MSCs, there were similarities in the amount of serum proinflammatory cytokines and anti-inflammatory cytokines. The number of Tregs was higher in the BM-MSC-treated mice than in the AD-MSC- and CB-MSC-treated mice, and Treg suppression assay showed that Tregs induced by BM-MSCs had stronger immunosuppressive activity than those induced by AD-MSCs or CB-MSCs.

As for clinical application, determining the optimal dosage of the MSCs is very important from the point of view of effort and cost. There has not been any report regarding the cell dose-dependent therapeutic effects of MSCs in CIA. Therefore, we investigated the dose-dependent therapeutic effect of MSCs. Although minimal therapeutic effects were seen with 5 × 10⁵ MSCs, treatment with 5 × 10⁶ cells was sufficient for the suppression of arthritis and the induction of Tregs. However, there was no significant difference between the cell number of 5 × 10⁶ and 5 × 10⁷ in their therapeutic effect and Treg induction on CIA.

Our study showed there is no difference in effectiveness between the two administration schedules (daily injection of 1×10⁶ MSCs for 5 consecutive days vs. two injections of 2.5×10⁶ cells over a 3-day interval). The effects of a one-time MSC treatment lasted for at least 4 weeks. As the number of MSCs delivered through the tail vein of mice is limited, we used an intraperitoneal injection method in our experiment. MSCs showed similar cell migration and therapeutic effect given by intravenous or intraperitoneal injection (28,30). Gonzalez et al. reported that the intra-articular injection of MSCs was less effective than the intraperitoneal route in CIA mice (17), showing that the therapeutic effects of MSCs are not simply due to a direct effect in the joints. They confirmed the therapeutic effect was attributed by viable human MSCs, not by dead human MSCs, and also showed that the injection of human myoblasts and dermal fibroblasts had no therapeutic effects (17).

MSC treatment decreased expression of the proinflammatory cytokines IFN-γ, IL-6, IL-1β, and TNF-α to the same extent as seen with MTX treatment. Specifically, MSCs significantly increased levels of the anti-inflammatory cytokines IL-10 and TGF-β and, unlike MTX, induced Tregs, suggesting different therapeutic mechanisms for MSCs and MTX. When the clinical arthritis scores began to improve, the levels of proinflammatory cytokines decreased, whereas the levels of anti-inflammatory cytokines and Tregs increased. MSC treatment was associated with increased levels of the anti-inflammatory cytokines IL-10 and TGF-β. These two are major anti-inflammatory cytokines from Treg cells, which play critical roles in the homeostatic regulation of the autoreactive T-cell repertoire and the induction of peripheral tolerance in vivo (24,41). The increase in IL-10 and TGF-β production suggests human MSCs may induce IL-10 and TGF-β producing mouse Tregs (33) that suppress inflammation in CIA mice. Interestingly, the increase in TGF-β levels was identified 1 week prior to that of IL-10 in BM-MSC-treated mice. Further study is needed to determine why TGF-β and IL-10 are expressed sequentially at different points of time. The MSC treatment of CIA mice strongly induced CD4⁺CD25⁺FoxP3⁺ Tregs within 2 weeks after treatment, suggesting that the immunomodulatory activity of the MSCs is sustained by the action of Tregs. Augello et al. reported that MSC viability is not required for the cells' long-term immunosuppressive action in CIA (2); MSCs were, in fact, detectable in the recipient for no more than 10 days after treatment. Cell-tracking experiments in that study showed that MSCs did not localize to the joints, suggesting that the clinical improvement observed was not due to the direct MSC repair of the joints (2). During this period, the MSCs were able to exert an immunosuppressive effect on other cells to inhibit the pathogenic immune reaction.

The therapeutic effect of MSCs in CIA might be related to Treg induction by MSCs. We demonstrated the induction of Tregs after MSC administration in a dose-dependent manner and their immunosuppressive properties. For the induction of Tregs, 5 × 10⁵ BM-MSCs were insufficient, whereas a dose of 5 × 10⁶ cells was sufficient, suggesting that the number of injected MSCs is important for Treg induction. To confirm whether the induced Tregs after MSC treatment have immunosuppressive properties, we conducted a Treg suppression assay in a coculture system of CD4⁺ T cells from CIA mice and Tregs from MSC-treated mice. The inhibition of CD4⁺ T-cell proliferation by Tregs showed that MSCs induced immunocompetent Tregs in CIA. From the result of our study, Tregs induced by MSC have more immune suppression activity than those of untreated CIA mice. MSC therapy can reset the immune system by reducing the deleterious Th1/Th17 response and enhancing the protective Treg response. The ability of MSCs to generate de novo Tregs may be advantageous therapeutically when compared to neutralizing antibodies against single-cytokine signaling, in terms of both safety and efficacy (28).

Recent studies of MSCs and their role in immune modulation show their potential for use in cell therapy (43,45). In this study, we used human MSCs in completely MHC-mismatched mice for evaluating the therapeutic effects of MSCs. Data show that MSCs are not rejected by the immune system, even after allogenic or xenogenic transplantation, because T cells do not recognize them. They express few or no class II MHC molecules or costimulation molecules, such as CD80 and CD86 (12,34). Data also show that allogenic MSCs might be functional in clinical settings (16,26). In animal models, human MSCs have been successfully used to ameliorate experimental autoimmune encephalomyelitis (44) and CIA mice (17,27), and therapy with human AD-MSCs prevented the recurrence of autoimmunity in lupus-prone mice (7). MSCs engraft and function across the barrier between species, which was confirmed in several different cross-species models (25). Considering the published data, our study using human MSCs for the treatment of CIA mice is justifiable upon these grounds.

Helios, a member of the Ikaros family of zinc finger transcription factors, is found at high levels in a subpopulation of Tregs (42) and plays an important role in Treg homeostasis. Helios is coexpressed in CD4⁺FoxP3⁺ Tregs, which are pivotal to maintaining self-tolerance and preventing autoimmunity (1). Our data show that induction of FoxP3⁺Helios⁺ Tregs was significantly increased in MSC-treated mice compared to untreated CIA mice. It is still controversial whether Helios expression in human and murine Tregs can be used to discriminate between naturally occurring thymus-derived Tregs (nTregs) and peripherally induced Tregs (iTregs) that develop from naive T cells under different conditions (10). The question remains as to whether Tregs induced by MSC treatment are derived from nTregs or iTregs or activated by other soluble factors in CIA.

In this study, human BM-, AD-, and CB-MSCs significantly suppressed joint inflammation in CIA mice. The cells decreased proinflammatory cytokines and upregulated anti-inflammatory cytokines and induced Tregs. Our results show that the immune modulatory effect of MSCs was associated with Treg induction, consistent with the increase in anti-inflammatory cytokine expression in CIA. Therefore, our study suggests that the use of human BM-, AD- and CB-MSCs could be an effective therapeutic approach for RA. According to our data, BM-MSCs had a greater therapeutic efficacy than the other types of MSCs in CIA. Further studies investigating the immune modulatory mechanism of MSCs from a perspective of the interactions with other immune cells and secretomes are required for the development of RA treatment using MSCs.

Footnotes

Acknowledgments

This research was supported by a grant from the Korean Health Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (Grant No. HI13C1270). The authors declare no conflicts of interest.

References

Akimova

; Beier

U. H.

; Wang

; Levine

M. H.

; Hancock

W. W.

Helios expression is a marker of T cell activation and proliferation. PLoS One 6(8): e24226; 2011.

Augello

; Tasso

; Negrini

S. M.

; Cancedda

; Pennesi

Cell therapy using allogeneic bone marrow mesenchymal stem cells prevents tissue damage in collagen-induced arthritis. Arthritis Rheum. 56(4): 1175–1186; 2007.

Brand

D. D.

; Latham

K. A.

; Rosloniec

E. F.

Collagen-induced arthritis. Nat. Protoc. 2(5): 1269–1275; 2007.

Camps

; Ruckle

; Ji

; Ardissone

; Rintelen

; Shaw

; Ferrandi

; Chabert

; Gillieron

; Francon

; Martin

; Gretener

; Perrin

; Leroy

; Vitte

P. A.

; Hirsch

; Wymann

M. P.

; Cirillo

; Schwarz

M. K.

; Rommel

Blockade of PI3Kgamma suppresses joint inflammation and damage in mouse models of rheumatoid arthritis. Nat. Med. 11(9): 936–943; 2005.

Chen

; Hu

; Liao

; Sun

; Han

; Song

; Zhao

RC.

Flk-1mesenchymal stem cells aggravate collagen-induced arthritis by up-regulating interleukin-6. Clin. Exp. Immunol. 159(3): 292–302; 2010.

Chen

F. C.

; Tuan

R. S.

Mesenchymal stem cells in arthritic diseases. Arthritis Res. Ther. 10(5): 223; 2008.

Choi

E. W.

; Shin

I. S.

; Park

S. Y.

; Park

J. H.

; Kim

J. S.

; Yoon

E. J.

; Kang

S. K.

; Ra

J. C.

; Hong

S. H.

Reversal of serologic, immunologic, and histologic dysfunction in mice with systemic lupus erythematosus by long-term serial adipose tissue-derived mesenchymal stem cell transplantation. Arthritis Rheum. 64(1): 243–253; 2012.

Ciccocioppo

; Bernardo

M. E.

; Sgarella

; Maccario

; Avanzini

M. A.

; Ubezio

; Minelli

; Alvisi

; Vanoli

; Calliada

; Dionigi

; Perotti

; Locatelli

; Corazza

G. R.

Autologous bone marrow-derived mesenchymal stromal cells in the treatment of fistulising Crohn's disease. Gut 60(6): 788–798; 2011.

Connick

; Kolappan

; Crawley

; Webber

D. J.

; Patani

; Michell

A. W.

; Du

M. Q.

; Luan

S. L.

; Altmann

D. R.

; Thompson

A. J.

; Compston

; Scott

M. A.

; Miller

D. H.

; Chandran

Autologous mesenchymal stem cells for the treatment of secondary progressive multiple sclerosis: An open-label phase 2a proof-of-concept study. Lancet Neurol. 11(2): 150–156; 2012.

10.

Curotto de Lafaille

M. A.

; Lafaille

J. J.

Natural and adaptive foxp3+ regulatory T cells: More of the same or a division of labor?

Immunity 30(5): 626–635; 2009.

11.

Curtis

J. R.

; Xi

; Patkar

; Xie

; Saag

K. G.

; Martin

Drug-specific and time-dependent risks of bacterial infection among patients with rheumatoid arthritis who were exposed to tumor necrosis factor alpha antagonists. Arthritis Rheum. 56(12): 4226–4227; 2007.

12.

Deans

R. J.

; Moseley

A. B.

Mesenchymal stem cells: Biology and potential clinical uses. Exp. Hematol. 28(8): 875–884; 2000.

13.

Djouad

; Bouffi

; Ghannam

; Noel

; Jorgensen

Mesenchymal stem cells: Innovative therapeutic tools for rheumatic diseases. Nat. Rev. Rheumatol. 5(7): 392–399; 2009.

14.

Duijvestein

; Vos

A. C.

; Roelofs

; Wildenberg

M. E.

; Wendrich

B. B.

; Verspaget

H. W.

; Kooy-Winkelaar

E. M.

; Koning

; Zwaginga

J. J.

; Fidder

H. H.

; Verhaar

A. P.

; Fibbe

W. E.

; van den Brink

G. R.

; Hommes

D. W.

Autologous bone marrow-derived mesenchymal stromal cell treatment for refractory luminal Crohn's disease: Results of a phase I study. Gut 59(12): 1662–1669; 2010.

15.

Elman

J. S.

; Li

; Wang

; Gimble

J. M.

; Parekkadan

A comparison of adipose and bone marrow-derived mesenchymal stromal cell secreted factors in the treatment of systemic inflammation. J. Inflamm. (Lond.) 11(1): 1; 2014.

16.

English

; Mahon

B. P.

Allogeneic mesenchymal stem cells: Agents of immune modulation. J. Cell. Biochem. 112(8): 1963–1968; 2011.

17.

Gonzalez

M. A.

; Gonzalez-Rey

; Rico

; Buscher

; Delgado

Treatment of experimental arthritis by inducing immune tolerance with human adipose-derived mesenchymal stem cells. Arthritis Rheum. 60(4): 1006–1019; 2009.

18.

Hsiao

S. T.

; Asgari

; Lokmic

; Sinclair

; Dusting

G. J.

; Lim

S. Y.

; Dilley

R. J.

Comparative analysis of paracrine factor expression in human adult mesenchymal stem cells derived from bone marrow, adipose, and dermal tissue. Stem Cells Dev. 21(12): 2189–2203; 2012.

19.

Kavanaugh

A. F.

B cell targeted therapies: Safety considerations. J. Rheumatol. Suppl. 77: 18–23; 2006.

20.

Kim

; Heo

J. S.

; Kim

H. S.

; Kim

H. O.

Isolation of mesenchymal stem cells from the mononuclear cells remaining in the bone marrow processing kit. Korean J. Blood Transfus. 21(3): 280–288; 2010.

21.

Le Blanc

; Frassoni

; Ball

; Locatelli

; Roelofs

; Lewis

; Lanino

; Sundberg

; Bernardo

M. E.

; Remberger

; Dini

; Egeler

R. M.

; Bacigalupo

; Fibbe

; Ringden

Developmental Committee of the European Group for Blood and Marrow Transplantation. Mesenchymal stem cells for treatment of steroid-resistant, severe, acute graft-versus-host disease: A phase II study. Lancet 371(9624): 1579–1586; 2008.

22.

Le Blanc

; Mougiakakos

Multipotent mesenchymal stromal cells and the innate immune system. Nat. Rev. Immunol. 12(5): 383–396; 2012.

23.

Lee

S. W.

; Kim

J. H.

; Park

Y. B.

; Lee

S. K.

Bortezomib attenuates murine collagen-induced arthritis. Ann. Rheum. Dis. 68(11): 1761–1767; 2009.

24.

Levings

M. K.

; Bacchetta

; Schulz

; Roncarolo

M. G.

The role of IL-10 and TGF-beta in the differentiation and effector function of T regulatory cells. Int. Arch. Allergy Immunol. 129(4): 263–276; 2002.

25.

; Ezzelarab

M. B.

; Cooper

D. K.

Do mesenchymal stem cells function across species barriers? Relevance for xenotransplantation. Xenotransplantation 19(5): 273–285; 2012.

26.

Liang

; Zhang

; Hua

; Wang

; Lu

; Shi

; Hou

; Zeng

; Gilkeson

G. S.

; Sun

Allogenic mesenchymal stem cells transplantation in refractory systemic lupus erythematosus: A pilot clinical study. Ann. Rheum. Dis. 69(8): 1423–1429; 2010.

27.

Liu

; Mu

; Wang

; Long

; Liu

; Li

; Sun

; Guo

; Zhang

; Guo

; Yu

; Li

; Liu

; Huang

; Wang

; Li

; Gu

; Liu

; Li

Therapeutic potential of human umbilical cord mesenchymal stem cells in the treatment of rheumatoid arthritis. Arthritis Res. Ther. 12(6): R210; 2010.

28.

MacDonald

G. I.

; Augello

; De Bari

Role of mesenchymal stem cells in reestablishing immunologic tolerance in autoimmune rheumatic diseases. Arthritis Rheum. 63(9): 2547–2557; 2011.

29.

Malgieri

; Kantzari

; Patrizi

M. P.

; Gambardella

Bone marrow and umbilical cord blood human mesenchymal stem cells: State of the art. Int. J. Clin. Exp. Med. 3(4): 248–269; 2011.

30.

Meyerrose

T. E.

; De Ugarte

D. A.

; Hofling

A. A.

; Herrbrich

P. E.

; Cordonnier

T. D.

; Shultz

L. D.

; Eagon

J. C.

; Wirthlin

; Sands

M. S.

; Hedrick

M. A.

; Nolta

J. A.

In vivo distribution of human adipose-derived mesenchymal stem cells in novel xenotransplantation models. Stem Cells 25(1): 220–227; 2007.

31.

Paulus

H. E.

The use of combinations of disease-modifying antirheumatic agents in rheumatoid arthritis. Arthritis Rheum. 33(1): 113–120; 1990.

32.

Paulus

H. E.

Protocol development for combination therapy with disease-modifying antirheumatic drugs. Semin. Arthritis Rheum. 23(2 Suppl 1): 19–25; 1993.

33.

Prevosto

; Zancolli

; Canevali

; Zocchi

M. R.

; Poggi

Generation of CD4+ or CD8+ regulatory T cells upon mesenchymal stem cell-lymphocyte interaction. Haematologica 92(7): 881–888; 2007.

34.

Ryan

J. M.

; Barry

F. P.

; Murphy

J. M.

; Mahon

B. P.

Mesenchymal stem cells avoid allogeneic rejection. J. Inflamm. (Lond.) 2: 8; 2005.

35.

Scott

; Wolfe

; Huizinga

T. W.

Rheumatoid arthritis. Lancet 376(9746): 1094–1108; 2010.

36.

Schnabel

; Gross

W. L.

Low-dose methotrexate in rheumatic diseases–Efficacy, side effects, and risk factors for side effects. Semin. Arthritis Rheum. 23(5): 310–327; 1994.

37.

Schurgers

; Kelchtermans

; Mitera

; Geboes

; Matthys

Discrepancy between the in vitro and in vivo effects of murine mesenchymal stem cells on T-cell proliferation and collagen induced arthritis. Arthritis Res. Ther. 12(1): R31; 2010.

38.

Shi

S. R.

; Liu

; Taylor

C. R.

Standardization of immunohistochemistry for formalin-fixed, paraffin-embedded tissue sections based on the antigen-retrieval technique: From experiments to hypothesis. J. Histochem. Cytochem. 55(2): 105–109; 2007.

39.

Singh

J. A.

; Furst

D. E.

; Bharat

; Curtis

J. R.

; Kavanaugh

A. F.

; Kremer

J. M.

; Moreland

L. W.

; O'Dell

; Winthrop

K. L.

; Beukelman

; Bridges

S. L.

Jr. ; Chatham

W. W.

; Paulus

H. E.

; Suarez-Almazor

; Bombardier

; Dougados

; Khanna

; King

C. M.

; Leong

A. L.

; Matteson

E. L.

; Schousboe

J. T.

; Moynihan

; Kolba

K. S.

; Jain

; Volkmann

E. R.

; Agrawal

; Bae

; Mudano

A. S.

; Patkar

N. M.

; Saag

K. G.

2012 update of the 2008 American College of Rheumatology recommendations for the use of disease-modifying antirheumatic drugs and biologic agents in the treatment of rheumatoid arthritis. Arthritis Care Res. (Hoboken) 64(5): 625–639; 2012.

40.

Sohn

H. S.

; Heo

J. S.

; Kim

H. S.

; Choi

; Kim

H. O.

Duration of in vitro storage affects the key stem cell features of human bone marrow-derived mesenchymal stromal cells for clinical transplantation. Cytotherapy 15(4): 460–466; 2013.

41.

Taylor

; Verhagen

; Blaser

; Akdis

C. A.

Mechanisms of immune suppression by interleukin-10 and transforming growth factor-beta: The role of T regulatory cells. Immunology 117(4): 433–442; 2006.

42.

Thornton

A. M.

; Korty

P. E.

; Tran

D. Q.

; Wohlfert

E. A.

; Murray

P. E.

; Belkaid

; Shevach

E. M.

Expression of Helios, an Ikaros transcription factor family member, differentiates thymic-derived from peripherally induced Foxp3+ T regulatory cells. J. Immunol. 184(7): 3433–3441; 2010.

43.

Wang

; Liao

; Tan

Mesenchymal-stem-cell-based experimental and clinical trials: current status and open questions. Expert Opin. Biol. Ther. 11(7): 893–909; 2011.

44.

Zappia

; Casazza

; Pedemonte

; Benvenuto

; Bonanni

; Gerdoni

; Giunti

; Ceravolo

; Cazzanti

; Frassoni

; Mancardi

; Uccelli

Mesenchymal stem cells ameliorate experimental autoimmune encephalomyelitis inducing T-cell anergy. Blood 106(5): 1755–1761; 2005.

45.

Zhao

; Wehner

; Bornhauser

; Wassmuth

; Bachmann

; Schmitz

Immunomodulatory properties of mesenchymal stromal cells and their therapeutic consequences for immune-mediated disorders. Stem Cells Dev. 19(5): 607–614; 2010.