Hydrogen sulfide promotes immunomodulation of gingiva-derived mesenchymal stem cells via the Fas/FasL coupling pathway

Background Mesenchymal stem cells derived from gingiva (GMSCs) display profound immunomodulation properties in addition to self-renewal and multilineage differentiation capacities. Hydrogen sulfide (H2S) is not only an environmental pollutant, but also is an important biological gas transmitter in health and disease. Methods We used an in-vitro coculture system and a mouse colitis model to compare the immunomodulatory effects between control and H2S-deficient GMSCs. The flow cytometry analysis was used for T-cell apoptosis and T-helper 17 (Th17) and regulatory T (Treg) cell differentiation. Results We revealed that GMSCs exerted their immunomodulatory effect by inducing T-cell apoptosis, promoting Treg cell polarization, and inhibiting Th17 cell polarization in vitro. The levels of H2S regulated the immunomodulatory effect of GMSCs. Mechanically, H2S deficiency downregulated the expression of Fas in GMSCs, resulting in reduced secretion of monocyte chemotactic protein 1 (MCP-1), which in turn led to decreased T-cell migration to GMSCs mediated by MCP-1. Moreover, H2S deficiency downregulated the expression of Fas ligand (FasL) in GMSCs. The Fas/FasL coupling-induced T-cell apoptosis by GMSCs was attenuated in H2S-deficient GMSCs. Consistent with this, H2S-deficient GMSCs showed attenuated therapeutic effects on colitis in vivo, which could be restored by treatment with the H2S donor, NaHS. Conclusions These findings showed that H2S was required to maintain immunomodulation of GMSCs, which was mediated by Fas/FasL coupling-induced T-cell apoptosis. Electronic supplementary material The online version of this article (10.1186/s13287-018-0804-6) contains supplementary material, which is available to authorized users.

Metabolic alterations by H 2 S play important roles in health and the development of disease [14]. Several studies have reported that H 2 S regulates the proliferation and differentiation of mesenchymal stem cells (MSCs) [15][16][17]. However, the role of H 2 S in the immunomodulatory capacity of GMSCs remains to be elucidated.
In this study, we showed that H 2 S was required to maintain the immunomodulatory effect of GMSCs, which is mediated by promoting Fas/FasL couplinginduced T-cell apoptosis.

Reagents
NaHS was purchased from Sigma-Aldrich. CBS, CSE, and MCP-1 siRNA were purchased from Santa Cruz Biotechnology.

Isolation of mouse bone marrow mesenchymal stem cells
Bone marrow cells were flushed out from the bone cavities of femurs and tibias with 2% heat-inactivated FBS (Equitech-Bio, Kerrville, TN, USA) in PBS. Single cell suspensions of all nuclear cells were obtained by passing through a 70-μm cell strainer (BD Biosciences). All nuclear cells were seeded into 100-μm culture dishes (Corning, Corning, NY, USA) and initially incubated for 48 h at 37°C in 5% CO 2 . To eliminate the nonadherent cells, the cultures were washed twice with PBS. The attached cells were cultured for 16 days. The BMMSCs were cultured with α-MEM supplemented with 20% FBS, 2 mM L-glutamine (Invitrogen), 55 mM 2-mercaptoethanol (Invitrogen), 100 U/ml penicillin, and 100 mg/ml streptomycin (Invitrogen).

T-lymphocyte isolation
Mouse T cells and CD4 + CD25 − T lymphocytes were isolated from mouse total spleen cells using a magnetic sorting Pan T and CD4 + CD25 + regulatory T-cell isolation kit (Miltenyi Biotec, Auburn, CA, USA), according to the manufacturer's instructions.
For the T-cell migration assay, a transwell coculture system was used. PKH26-stained GMSCs (1 × 10 4 cells) were seeded on the lower chamber of a 12-well culture plate (Corning) with a transwell and incubated for 12 h. The CFSE (10 μM)-labeled active T cells were loaded onto the upper chamber of the transwell and cocultured for 24 h. The culture transwells were then observed acquired using a laser-scanning microscope (LSM510; Zeiss). The green-labeled T-cell number was counted and normalized by the red-labeled number of GMSCs in five representative images as reported previously [18].

Immunofluorescent staining
GMSCs were seeded onto a four-well chamber slide and incubated for 12 h at 37°C in 5% CO 2 . The slides were fixed in 4% paraformaldehyde, followed by 0.01% Triton X-100 treatment for 10 min. The slides were then blocked with normal serum matched to secondary antibodies for 1 h, followed by incubation with the specific or isotype-matched antibodies (1:100) overnight at 4°C. The slides were then treated with Rhodamin/FITC-conjugated secondary antibodies (1:200; Jackson Immuno Research, West Grove, PA, USA) for 1 h under 20-25°C in the dark and mounted using VECTASHIELD® Mounting Medium containing 4′6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, Burlingame, CA, USA). The slides were then observed and images acquired using a laserscanning microscope (LSM510; Zeiss) and the images were processed with LSM 5 Release 4.2 software after acquisition.

Quantitative PCR
For quantitative PCR (qPCR) analysis, total RNA was extracted with an RNeasy Mini kit (Qiagen, Valencia, CA, USA), and cDNA was prepared using a SuperScript® III Reverse Transcriptase (RT) kit (Invitrogen). qPCR was performed using SYBR® Green Supermix (Bio-Rad, Hercules, CA, USA) on a Bio-Rad CFX96 Real Time System, according to the manufacturer's instructions.

Enzyme-linked immunosorbent assay
Culture supernatants from GMSCs were collected. The samples were centrifuged and used for MCP-1 enzymelinked immunosorbent assay (ELISA) analysis (R&D Systems), according to the manufacturer's instructions.

Flow cytometry analysis
The following antibodies were used for surface staining: anti-mouse CD146-PE, CD90-PE, CD105-PE, CD73-PE, CD34-PE, CD45-PE, CD4-Percp, CD3-APC, and CD25-APC. Staining of intracellular cytokines or transcription factors was performed using Foxp3 staining buffer (eBioscience) according to the manufacturer's instructions. Anti-mouse Foxp3-PE and IL-17-PE were used for intracellular staining. The Annexin V-7AAD apoptosis kit (BD Biosciences) was used for analysis of T-cell apoptosis. The data were acquired using the FACS calibur platform (BD Biosciences).

Overexpression of FasL in GMSCs
For overexpression of FasL, 293 T cells for lentivirus production were plated in a 1-cm dish and cultured until 80% confluent. The plasmids were mixed in proper proportions (5:3:2) with gene expression vectors (Addgene17620), psPAX (Addgene12260), and pCMV-VSV-G (Addgene8454), and mixed in opti-MEM with Lipofectamine LTX (Invitrogen) according to the manufacturer's instructions. The EGFP expression plasmid (Addgene17618) was used as a control. The supernatant was collected 24 and 48 h after transfection and filtered through a 0.45-μm filter to remove cell debris. For infection, the supernatant containing the lentivirus was added into the GMSC culture in the presence of 4 μg/ml polybrene and the transgenic FasL expression was validated by observation of green fluorescent protein [19].

Knockdown assay in GMSCs
For knockdown of MCP-1 expression in GMSCs, siRNA transfection was performed according to the manufacturer's instructions. Fluorescein-conjugated control siRNA was used as a control and as a method of evaluating transfection efficiency. All siRNA products were purchased from Santa Cruz Biotechnology (Dallas, TX, USA).

GMSC infusion in acute colitis mice
Acute colitis was induced by administering 3% (w/v) dextran sulfate sodium (DSS) (molecular weight 36,000-50,000 Da; MP Biochemicals, France) through drinking water, and mice were fed ad libitum for 10 days. GMSCs or NaHSpretreated GMSCs were infused (0.2 × 10 6 cells) into colitis mice (n = 6) intravenously at day 3 after feeding DSS water. In the control group, mice were given PBS (n = 6). Samples from all mice were harvested at day 10 after the DSS water was provided. Induced colitis was evaluated as described previously [20].

Statistical analysis
Comparisons between two groups were analyzed using the two-tailed Student's t test using SPSS 18.0 statistical software. Comparisons between more than two groups were analyzed using one-way analysis of variance (ANOVA). A Bonferroni method was used to consider multiple comparisons. P < 0.05 was considered significant.

Mouse GMSCs produced H 2 S
We found that GMSCs isolated from mouse gingiva expressed the main catalytic enzymes of H 2 S synthesis, cystathionine β-synthase (CBS), and cystathionine γ-lyase (CSE), as did BMMSCs (Fig. 1a and Additional file 1: Figure S1A). a c d b Fig. 1 GMSCs produced H 2 S. a GMSCs expressed CBS and CSE, as assessed by western blot analysis. b Flow cytometry analysis showed 47.1% and 54.6% of GMSCs positive for CBS and CSE, respectively. c GMSCs positive for both CD146 and CBS or CSE, as assessed by immunofluorescence staining. d H 2 S donor NaHS treatment elevated H 2 S production in culture supernatant, while CBS siRNA or CSE siRNA treatment downregulated H 2 S concentrations (n = 5). *P < 0.05, **P < 0.01, ***P < 0.001. Scale bar: 20 μm. Error bars: mean ± SD. All experimental data verified in at least three independent experiments. CBS cystathionine β-synthase, CSE cystathionine γ-lyase, GMSC gingiva-derived mesenchymal stem cell, BMMSC bone marrow mesenchymal stem cell, FSC forward scatter, DAPI 4′6-diamidino-2-phenylindole, H 2 S hydrogen sulfide, si small interfering Flow cytometry revealed that 47.1% and 54.6% of GMSCs were positive for CBS and CSE, respectively (Fig. 1b). GMSCs were positive for both CD146 and CBS or CSE (Fig. 1c). GMSCs produced H 2 S in the culture supernatant, which could be upregulated by H 2 S donor NaHS treatment and downregulated by CBS or CSE siRNA treatment ( Fig. 1d and Additional file 1: Figure S1B). These data all supported the production of H 2 S by GMSCs.

H 2 S was essential for GMSCs to induce T-cell apoptosis
To analyze the immunomodulation of GMSCs, we employed GMSC and T-cell coculture systems and found that GMSCs induced T-cell apoptosis, as reported previously [1]. Both the early (AnnexinV + /7AAD − cells) and later (AnnexinV + /7AAD + cells) T-cell apoptosis induced by GMSCs decreased after CBS or CSE siRNA treatment ( Fig. 2a and Additional file 1: Figure S2A). Moreover, a b c d *P < 0.05, **P < 0.01, ***P < 0.001. Error bars: mean ± SD. All experimental data verified in at least three independent experiments. CBS cystathionine β-synthase, CSE cystathionine γ-lyase, GMSC gingiva-derived mesenchymal stem cell, si small interfering GMSCs derived from Cbs −/− mice induced less T-cell apoptosis than control ones. This defect of Cbs −/− GMSCs could be partially restored by NaHS treatment (Fig. 2b and Additional file 1: Figure S2B). MSCs have been reported to play a role in immunomodulation by promoting Treg cell differentiation and inhibiting Th17 cell differentiation [21]. Here we revealed that GMSCs promoted Treg cell polarization and inhibited Th17 polarization in a coculture system, and that these effects were attenuated by CBS or CSE siRNA treatment (Additional file 1: Figure  S2C, D). Furthermore, Cbs −/− GMSCs showed a decreased capacity to promote Treg cell differentiation and inhibit Th17 cell differentiation compared with control GMSCs, and these abilities were restored by NaHS treatment (Fig. 2c, d). These results suggested that treatment of GMSCs by H 2 S induced T-cell apoptosis and regulated T-helper cell polarization.

Immunomodulation of GMSCs via Fas/FasL-mediated T-cell apoptosis
To explore the underlying mechanism of how H 2 S regulates the immunomodulatory effects of GMSCs, we analyzed the expression of FasL, a key member of the apoptosis pathway, in GMSCs and found that FasL expression was decreased in Cbs −/− GMSCs compared with controls (Fig. 3a). When cocultured with T cells, GMSCs induced T-cell apoptosis, which was blocked by treatment with FasL neutralizing antibody (Fig. 3b). Furthermore, when we cocultured T cells and GMSCs in cell-cell contact, the compromised capacity of Cbs −/− GMSCs to induce T-cell apoptosis was restored by overexpression of FasL ( Fig. 3c and Additional file 1: Figure  S3). The results suggested that FasL was essential in GMSC-induced T-cell apoptosis. However, when GMSCs were cocultured with T cells in a transwell system, FasL overexpression failed to enhance the decreased immunomodulation of Cbs −/− GMSCs (Fig. 3d), which indicated that soluble factors may be involved in H 2 S-mediated immunomodulation by GMSCs. We then compared the cytokine profiles between control and Cbs −/− GMSCs using a cytokine array and found that the chemokines CXCL-10, CXCL-1, and MCP-1 were significantly decreased in Cbs −/− GMSCs (Fig. 4a, b). We next found that fewer activated T cells migrated to Cbs −/− GMSCs compared with control GMSCs in a transwell coculture system (Fig. 4c). Monocyte chemotactic protein 1 (MCP-1), a T-cell chemoattractant cytokine, has been reported to play a crucial role in MSCmediated T-cell apoptosis [18,22]. We knocked down MCP-1 using siRNA and found that fewer activated T cells migrated to GMSCs after this treatment (Fig. 4d). The death receptor, Fas, regulates the secretion of cytokines and exosomes including MCP-1 and C-X-C motif chemokine 10 (CXCL-10) [18]. We found that Fas a b c d  Figure S4A). Fas and FasL expression in Cbs −/− GMSCs were upregulated by treatment with NaHS (Fig. 4e). MCP-1 expression was elevated in Cbs −/− GMSCs compared with controls, and this elevated expression was attenuated by NaHS treatment (Additional file 1: Figure  S4B-D). MCP-1 secretion of Cbs −/− GMSCs in the culture supernatant was reduced relative to control GMSCs, and could be upregulated by NaHS treatment (Additional file 1: Figure S4E). Moreover, NaHS treatment upregulated the expression of Fas and FasL in Cbs −/− GMSCs ( Fig. 4f and Additional file 1: Figure S4F) and increased the number of activated T cells that migrated to Cbs −/− GMSCs (Fig. 4g). Taken together, these results indicated

H 2 S promoted the therapeutic effects of GMSCs in the treatment of colitis
To further study the role of H 2 S in immunomodulation by GMSCs, we compared the immunotherapeutic effects between control and Cbs −/− GMSCs in mice with dextran sulfate sodium (DSS)-induced colitis. Control GMSCs, Cbs −/− GMSCs, or NaHS-treated GMSCs (0.2 × 10 6 cells) were systemically infused into mice with colitis at day 3 after 3% DSS treatment (Fig. 5a). The disease activity index (DAI), determined by the presence of sustained weight loss and bloody diarrhea/loose feces, was elevated in the mice with colitis compared with the control group. After infusion of control GMSC or NaHS-pretreated Cbs −/− GMSCs, but not Cbs −/− GMSCs, the DAI score decreased (Fig. 5b). Infusion of control GMSCs, but not Cbs −/− GMSCs, partially restored the reduced body weight of mice with DSSinduced colitis. NaHS pretreatment enhanced the ability of Cbs −/− GMSCs to restore the reduced body weight of mice with colitis (Fig. 5c). Decreased numbers of Treg cells and elevated numbers of Th17 cells were observed in the mice with colitis at day 10 post DSS administration (Fig. 5d, e).  Fig. 5d, e). Infiltration of inflammatory cells, impairment of colon structures, and the absence of an epithelial layer were observed in histological analysis of mice with colitis. Infusion of control GMSCs or NaHS-pretreated Cbs −/− GMSC, but not Cbs −/− GMSC, partially restored the histological appearance of the colon in DSS-induced colitis mice, eliminating the inflammatory cells and showing evidence for recovery of the epithelial structure (Fig. 5f). Together, these results confirmed that H 2 S is important for maintaining the immunomodulatory effects of GMSCs (Fig. 6).

Discussion
H 2 S has been reported to play an important role in MSCs by enhancing the proliferation and differentiation [16]. Here we showed, for the first time, that a physiological level of H 2 S was required to maintain the immunomodulatory capacity of GMSCs. H 2 S deficiency in GMSCs led to reduced T-cell apoptosis when cocultured with GMSCs. Furthermore, H 2 S-deficient GMSCs showed compromised therapeutic effects on experimental colitis in mice. Fas/FasL interaction induces oligomerization and aggregation of the Fas receptor, leading to apoptosis after protein-protein interactions. A previous study indicated that MSC-induced immune tolerance was associated with FasL-triggered T-cell apoptosis via the Fas/FasL death pathway [18] In the present study, we showed that H 2 S was required to maintain the immunoregulatory function of GMSCs by enhancing the expression of FasL.
Conversely, H 2 S-deficient GMSCs with decreased expression of FasL showed a significantly decreased immunomodulation. Cell-cell contact was essential for Fas/ FasL-mediated T-cell apoptosis. In our coculture system, FasL overexpression rescued the decrease in T-cell apoptosis induced by H 2 S-deficient GMSCs in a cell-cell contact manner, but not in a transwell chamber. These results indicated that there may be some soluble factors in addition to FasL participating in the regulation of immunomodulation of GMSCs. Furthermore, we showed that the expression of chemokines associated with cell trafficking, including CXCL-10, CXCL-1, and MCP-1, were significantly decreased in H 2 S-deficient GMSCs compared with control GMSCs. The MCP-1/CCR2 signaling axis has been implicated in the trafficking of T cells, macrophages, and myeloid-derived suppressor cells in tumors and immune disorders [23][24][25]. It has been reported that MCP-1 produced by MSCs mobilized macrophages to modulate the Th1/Th17 immune response, which contributed to the therapeutic effects of MSCs on experimental autoimmune encephalitis (EAE) or experimental autoimmune uveitis (EAU) [24,26,27]. MCP-1 secreted by MSCs was also reported to recruit activated T cells to MSCs to cause immune tolerance [18]. Here we showed that MCP-1 produced by GMSCs recruited activated T cells to GMSCs, which in turn resulted in GMSC-induced T-cell apoptosis through the Fas/FasL death pathway.
Recent studies have shown that Fas controls exosome release by MSCs via epigenetic regulation [28]. Here we found that Fas expression, which was regulated by the Fig. 6 Schematic showing that H 2 S metabolism regulates the immunomodulatory capacity of GMSCs via Fas/FasL coupling-induced T-cell apoptosis. MCP-1 monocyte chemotactic protein 1, FasL Fas ligand, H 2 S hydrogen sulfide, CBS cystathionine β-synthase, CSE cystathionine γ-lyase levels of H 2 S in GMSCs, controlled MCP-1 secretion. We also found that Fas-mediated MCP-1 secretion in GMSCs stimulated the migration of activated T cells in a transwell chamber, which reflected what happens in vivo when T cells migrate to transplanted GMSCs. FasL expressed in GMSCs then induced T-cell apoptosis via the Fas/FasL death pathway in a cell-cell dependent manner. Furthermore, the level of H 2 S played a critical role in promoting both Fas and FasL expression to cause immune tolerance. However, the underlying mechanism of Fas-dependent control of MCP-1 secretion needs to be studied further.
The large number of antigens present in commensal bacteria of the oral microflora makes the induction of immune tolerance in oral mucosa unique. Disturbances of the immune network contribute to the development of gingivitis and periodontitis [14]. H 2 S plays crucial roles in regulating the immunomodulation of GMSCs and H 2 S donor treatment could elevate the immunomodulation of GMSCs. Manipulation of H 2 S levels may therefore be an alternative choice for preventing or treating inflammation and infections, including colitis and periodontal disease.

Conclusions
The level of H 2 S controlled the immunomodulatory effect of GMSCs, which was mediated by Fas/FasL coupling-induced T-cell apoptosis (Fig. 6).

Additional file
Additional file 1: is Figure S1 showing mouse GMSCs produced H 2 S, Figure S2 showing H 2 S is required in GMSCs to induce T-cell apoptosis, Figure S3 showing efficacy of FasL overexpression, as assessed by western blot analysis, and Figure S4