Mesenchymal-stem-cell-induced immunoregulation involves FAS-ligand-/FAS-mediated T cell apoptosis

Abstract

Systemic infusion of bone marrow mesenchymal stem cells (BMMSCs) yields therapeutic benefit for a variety of autoimmune diseases, but the underlying mechanisms are poorly understood. Here we show that in mice systemic infusion of BMMSCs induced transient T cell apoptosis via the FAS ligand (FASL)-dependent FAS pathway and could ameliorate disease phenotypes in fibrillin-1 mutated systemic sclerosis (SS) and dextran-sulfate-sodium-induced experimental colitis. FASL(-/-) BMMSCs did not induce T cell apoptosis in recipients, and could not ameliorate SS and colitis. Mechanistic analysis revealed that FAS-regulated monocyte chemotactic protein 1 (MCP-1) secretion by BMMSCs recruited T cells for FASL-mediated apoptosis. The apoptotic T cells subsequently triggered macrophages to produce high levels of TGFβ, which in turn led to the upregulation of CD4(+)CD25(+)Foxp3(+) regulatory T cells and, ultimately, immune tolerance. These data therefore demonstrate a previously unrecognized mechanism underlying BMMSC-based immunotherapy involving coupling via FAS/FASL to induce T cell apoptosis.

Trial registration: ClinicalTrials.gov NCT00962923.

Copyright © 2012 Elsevier Inc. All rights reserved.

Figures

Figure 1. BMMSCs induce T cell apoptosis via Fas ligand (FasL)

(A) Schema of BMMSC transplantation procedure. 1×106 BMMSCs (n=5), FasL−/−gldBMMSCs (n=4) or FasL-transfected gldBMMSCs (FasL+gldBMMSCs, n=4) were infused into C57BL6 mice through the tail vein. All groups were sacrificed at indicated time points for sample collection. Zero hour represented that mice were immediately sacrificed after BMMSC injection. (B–E) BMMSC transplantation (BMMSC) induced transient reduction in the number of CD3+ T cells and increased AnnexinV+7AAD+ double positive apoptotic CD3+ T cells in peripheral blood mononuclear cells (PBMNCs; B, C) and bone marrow mononuclear cells (BMMNCs; D, E) at indicated time points, while FasL−/− BMMSCs from gld mice (gldBMMSCs) failed to reduce CD3+ T cells or elevate CD3+ T cell apoptosis in peripheral blood (B, C) and bone marrow (D, E). FasL-transfected gldBMMSC transplantation (FasL+gldBMMSC) partially rescued the capacity to reduce the number of CD3+ T cells and induce CD3+ T cell apoptosis in peripheral blood (B, C) and bone marrow (D, E). *P<0.05; **P<0.01; ***P<0.001 vs. gldBMMSC, #P<0.05; ###P<0.001 vs. FasL+gldBMMSC, $P<0.05; $$$P<0.001 vs. gldBMMSC. (F) When BMMSCs were infused into mice, TUNEL and immunohistochemistry staining showed that TUNEL-positive apoptotic cell (brown, white arrow) number in CD3-positive T cells (purple, yellow arrowhead) was higher in the BMMSC-injected group compared to the control group in bone marrow. (G) When BMMSCs were co-cultured with T cells, BMMSC-induced AnnexinV+7AAD+ double positive apoptotic T cells were significantly blocked by anti-FasL neutralizing antibody (1μg/mL) compared to IgG antibody control group. (H) TUNEL and immunohistochemistry staining showed that TUNEL-positive apoptotic T cells (brown, white arrow) were observed in CD3+ T cells (purple, yellow arrowhead) when co-cultured with BMMSCs in vitro. In the presence of anti-FasL neutralizing antibody (FasL Ab), TUNEL-positive cell percentage was significantly less than the untreated control group. (I) In addition, the number of BMMSC-induced AnnexinV+7AAD+ double positive apoptotic T cells was significantly blocked by caspase 3, 8, and 9 inhibitor treatments. The results were representative of three independent experiments. (J) Schematic diagram indicating that BMMSCs induce T cell apoptosis. (*P<0.05; **P<0.01; ***P<0.001. The bar graph represents mean±SD). See also Figures S1 and 2.

Figure 2. FasL is required for BMMSC-induced T cell apoptosis and upregulation of CD4+CD25+Foxp3+ regulatory T cells (Tregs)

(A, B) BMMSC transplantation (BMMSC, n=5) induced a transient reduction in the number of CD3+ T cells (A) and elevation of AnnexinV+7AAD+ double positive apoptotic CD3+ cells (B) in peripheral blood. Transplantation of FasL knockdown BMMSC (FasL siRNA BMMSC, n=3) failed to reduce CD3+ T cells (A) or increase the number of CD3+ apoptotic T cells (B) in peripheral blood. (C, D) BMMSC transplantation (BMMSC, n=5) showed a transient reduction of CD3+ T cells (C) and elevation of AnnexinV+7AAD+ double positive apoptotic CD3+ T cells (D) in bone marrow. Transplantation of FasL knockdown BMMSCs (FasL siRNA BMMSC, n=3) failed to reduce CD3+ T cells (C) or elevate CD3+ apoptotic T cells (D) in bone marrow. (E) BMMSC, but not FasL knockdown BMMSC, transplantation significantly upregulated levels of Tregs at 24 and 72 hours after transplantation in C57BL6 mice. (F) BMMSC transplantation resulted in a significant upregulation of Tregs when compared to the gldBMMSC transplantation group at 24 and 72 hours post-transplantation. FasL-transfected gldBMMSC transplantation (FasL+gldBMMSC) partially rescued BMMSC-induced upregulation of Tregs. (G) TGF-β level in peripheral blood was significantly increased in both BMMSC and FasL+gldBMMSC groups at 24 hours post-transplantation. FasL−/−gldBMMSC transplantation failed to upregulate TGF-β level. (H) Apoptotic pan T cells were engulfed by macrophages in vivo. Green indicates T cells, and red indicates CD11b+ macrophages. Bar=50μm. (I) BMMSC transplantation group increased the number of CD11b+ cells in peripheral blood when compared to the control group (C57BL6). Depletion of macrophages by clodronate liposome treatment showed the effectiveness in reducing CD11b+ cells in the BMMSC transplantation group (BMMSC+clodronate), as assessed by flow cytometric analysis. (J) TGF-β level was significantly increased in peripheral blood after BMMSC transplantation. Clodronate liposome treatment blocked BMMSC-induced upregulation of TGF-β (BMMSC+clodronate). (K) BMMSC transplantation upregulated the level of Tregs in peripheral blood compared to the control group (C57BL6). Clodronate liposome treatment inhibited BMMSC-induced Treg upregulation (BMMSC+clodronate). (L) Schematic diagram indicating that BMMSC-induced T cell apoptosis resulted in immune tolerance as evidenced by upregulation of Tregs. The results were representative of three independent experiments. (*P<0.05, **P<0.01, ***P<0.001. The bar graph represents mean±SD). See also Figure S3.

Figure 3. FasL is required for BMMSC-mediated amelioration of systemic sclerosis (SS) phenotypes

(A) Schema showing how BMMSC transplantation ameliorates SS phenotype. (B, C) BMMSC transplantation (n=6) showed a significantly reduced number of CD3+ T cells (B) and increased number of AnnexinV+7AAD+ double positive apoptotic CD3+ T cells (C) in SS mice as assessed by flow cytometric analysis. However, FasL−/−gldBMMSC (n=6) failed to reduce the number of CD3+ T cells (B) or elevate the number of apoptotic CD3+ T cells (C). (D–F) Tsk/+ SS mice showed elevated levels of antinuclear antibody (ANA, D) and anti-double strand DNA antibodies IgG (E) and IgM (F) when compared to control C57BL6 mice. BMMSC transplantation reduced the levels of ANA (D) and anti-double strand DNA antibodies IgG (E) and IgM (F). In contrast, FasL−/−gldBMMSC transplantation failed to reduce the levels of antinuclear antibody (ANA, D) or anti-double strand DNA IgG (E) and IgM (F) antibodies. (G) Creatinine level in serum was significantly increased in Tsk/+ mice. After BMMSC transplantation, creatinine level was significantly decreased to the level observed in C57BL6 mice. However, gldBMMSC transplantation failed to reduce the creatinine level. (H) The concentration of urine protein was significantly increased in Tsk/+ mice. BMMSC transplantation reduced urine protein to the control level. gldBMMSC transplantation failed to reduce urine protein levels in Tsk/+ mice. (I) Treg level was significantly decreased in Tsk/+ mice compared to C57BL6 mice. After BMMSC transplantation, Treg levels were significantly elevated, whereas gldBMMSC transplantation failed to increase Treg levels in Tsk/+ mice. (J) CD4+IL17+ Th17 cells were significantly increased in Tsk/+ mice compared to C57BL6 mice. Elevated Th17 level was significantly reduced in the BMMSC transplantation group, while gldBMMSC transplantation failed to reduce the Th17 level in Tsk/+ mice. (K) Hyperdermal thickness was significantly increased in Tsk/+ mice (Tsk/+, n=5) compared to control mice (C57BL6, n=5). BMMSC, but not FasL−/−gldBMMSC, transplantation reduced hyperdermal thickness. (*P<0.05, **P<0.01, ***P<0.001. The bar graph represents mean±SD). See also Figure S4.

Figure 4. FasL plays a critical role in BMMSC-mediated immune therapy for Dextran sulfate sodium (DSS)-induced experimental colitis

(A) Schema showing BMMSC transplantation in DSS-induced experimental colitis mice. (B, C) BMMSC transplantation (n=6) showed a significantly reduced number of CD3+ T cells at 24 hours post-transplantation (B) and increased number of AnnexinV+7AAD+ double positive apoptotic CD3+ T cells at 24–72 hours post-transplantation (C) in colitis mice as assessed by flow cytometric analysis. However, FasL−/−gldBMMSC (n=6) failed to reduce the number of CD3+ T cells (B) or elevate the number of apoptotic CD3+ T cells (C). (D) Colitis mice (colitis, n=5), BMMSC transplanted group, and gldBMMSC showed significantly reduced body weight from 5 to 10 days after DSS induction. The BMMSC transplantation group showed inhibition of body weight loss compared to the colitis and gldBMMSC transplantation groups at 10 days after DSS induction. (E) Disease activity index (DAI) was significantly increased in colitis mice compared to C57BL6 mice from 5 days to 10 days after DSS induction. BMMSC transplantation significantly reduced DAI score, but it was still higher than that observed in C57BL6 mice. FasL−/−gldBMMSC transplantation failed to reduce DAI score at all time points. (F) Treg level was significantly reduced in colitis mice compared to C57BL6 mice at 7days after DSS induction. BMMSC, but not FasL−/−gldBMMSC, transplantation upregulated the Treg levels in colitis mice. (G) Th17 cell level was significantly elevated in colitis mice compared to C57BL6 mice at 7 days after DSS induction. BMMSC, but not FasL−/−gldBMMSC, transplantation reduced the levels of Th17 cells in colitis mice from 7 to 10 days after DSS induction. (H) Hematoxylin and eosin staining showed the infiltration of inflammatory cells (blue arrows) in colon with destruction of epithelial layer (yellow triangles) in colitis mice. BMMSC, but not FasL−/−gldBMMSC, transplantation rescued disease phenotype in colon and reduced histological activity index. (I) Schematic diagram of BMMSC transplantation for immunotherapies. (Bar= 200μm; *P<0.05, **P<0.01, ***P<0.001. The bar graph represents mean±SD). See also Figure S5.

Figure 5. Fas plays an essential role in BMMSC-mediated CD3+ T cell apoptosis and up-regulation of Tregs via regulating monocyte chemotactic protein 1 (MCP-1) secretion

(AD) BMMSC transplantation (BMMSC) induced transient reduction in the number of CD3+ T cells and increase in the number of AnnexinV+7AAD+ double positive apoptotic CD3+ T cells in peripheral blood mononuclear cells (PBMNCs; A, B) and bone marrow mononuclear cells (BMMNCs, n=5; C, D) at indicated time points, while Fas−/− BMMSC from lpr mice (lprBMMSC, n=5) failed to reduce the number of CD3+ T cells or increase the number of CD3+ apoptotic T cells in peripheral blood (A, B) and bone marrow (C, D). (E, F) lprBMMSC transplantation failed to elevate Treg levels (E) and TGF-β (F) in C57BL6 mice compared to the BMMSC transplantation group at indicated time points. (G) lprBMMSC induced activated T cell apoptosis in a BMMSC/T cell in vitro co-cultured system, which was blocked by anti-FasL neutralizing antibody (1μg/mL). (H–K) Activated T cells (green) migrate to BMMSCs (red) in a transwell co-culture system (H). lprBMMSCs showed a significantly reduced capacity to induce activated T cell migration (I), which could be partially rescued by overexpression of MCP-1 (J) and totally rescued by overexpression of Fas (K) in lprBMMSCs. The results were representative of three independent experiments. (L) Quantitative RT-PCR analysis showed no significant difference between BMMSC and lprBMMSC in terms of MCP-1 expression level. However, overexpression of MCP-1 and Fas in lprBMMSC significantly elevated gene expression level of MCP-1. (M) Western blot showed that lprBMMSCs express a higher cytoplasm level of MCP-1 than BMMSC. Overexpression of Fas in lprBMMSC reduced the expression level of MCP-1 in cytoplasm. (N) ELISA analysis showed that MCP-1 secretion in culture supernatant was significantly reduced in lprBMMSCs compared to BMMSC. Overexpression of MCP-1 and Fas in lprBMMSCs significantly elevated MCP-1 secretion in culture supernatant. (O) ELISA data showed that knockdown Fas expression using siRNA resulted in reduction of MCP-1 level in culture medium compared to control siRNA group. (P–Q) Fas siRNA-treated BMMSCs (Q) showed reduced T cell migration in transwell co-culture system compared to control siRNA group (P). (*P<0.05, **P<0.01, ***P<0.001. The bar graph represents mean±SD). See also Figure S6.

Figure 6. MCP-1 plays an important role in T cell recruitment

(A) MCP-1−/−BMMSC transplantation showed a slightly reduced number of CD3+ T cells in peripheral blood, but the level of reduction was significantly less than that of the BMMSC transplantation group. (B) AnnexinV+7AAD+ double positive apoptotic CD3+T cell percentage was slightly increased in the MCP-1−/− BMMSC transplant group. (C) Treg level was slightly increased in the MCP-1−/− BMMSC-transplanted group at 72 hours post-transplantation, but significantly lower than the BMMSC transplantation group. (D) TGF-β level in serum was slightly increased in the MCP-1−/− BMMSC-transplanted group at 72 hours after transplantation compared to 0 hour, but the elevation level was lower than the BMMSC transplantation group. (E) When T cells were stimulated with CD3 and CD28 antibody and co-cultured with BMMSC or MCP-1−/− BMMSC in a transwell culture system, the number of migrated T cells was significantly higher in the BMMSC group (E) than the MCP-1−/− BMMSC group. (F) Schematic diagram showing the mechanism of BMMSC-induced immunotherapies. (**P<0.01, ***P<0.005; The bar graph represents mean±SD). See also Figure S7.

Figure 7. Allogenic MSC transplantation induces CD3+ T cell apoptosis and Treg upregulation in systemic sclerosis (SS) patients

(A) Schema of MSC transplantation in SS patients. (B) Flow cytometric analysis showed reduced number of CD3+ T cells from 2 to 72 hours post-transplantation. (C) AnnexinV+-positive apoptotic CD3+ T cells were significantly increased at 6 hours after MSC transplantation. (D) Flow cytometric analysis showed reduced number of CD4+ T cells from 2 to 72 hours post-transplantation. (E) Treg levels in peripheral blood were significantly increased at 72 hours after allogenic MSC transplantation. (F) Serum level of TGFβ was significantly increased in MSC transplantation group at 72 hours post-transplantation. (G, H) Modified Rodnan Skin Score (MRSS, G) and Health assessment Questionnaire disease activity index (HAQ-DI, H) were significantly reduced after allogenic MSC transplantation. (I) Representative images of skin ulcers prior to MSC transplantation (pre-MSC) and at 6 months post-transplantation (post-MSC). (J) The reduced ANA level was maintained at 12 months after MSC transplantation. (K) Real-time PCR analysis showed significantly decreased FasL expression in SS patient MSCs (SSMSC) compared to MSC from healthy donor (MSC). (L) SSMSC showed a significantly decreased capacity to induce T cell apoptosis compared to normal MSC in vitro. (M) SSMSC showed a reduced expression of Fas by real-time PCR analysis. (N) MCP-1 secretion level in SSMSC was significantly lower than that in MSC culture supernatant. (*P<0.05, **P<0.01, ***P<0.005; The bar graph represents mean±SD). See also Table S1.