Female C57BL/6 mice (9~10 weeks old) were purchased from Japan SLC Inc. (Shizuoka, Japan). All animal experiments were approved by the institutional Animal Care and Use Committees of the Catholic University of Korea (Seoul, Korea).

The bone marrow-derived cells were isolated from C57BL/6 mice by flushing the femurs and tibias with complete culture medium, which comprised DMEM (WelGENE, Daegu, Korea) supplemented with 10% heat-inactivated FBS (WelGENE), 2 mM glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin (Gibco BRL, Gaithersburg, MA, USA). Briefly, the isolated cells were plated in 75 cm² tissue culture flasks at a concentration of 1×10⁶ cells/mL in the complete culture medium and incubated at 37℃ and 5% CO₂. After 3 days, the non-adherent cells were removed. When the cells had reached 70~80% confluence, the cells were trypsinized and passaged into a new flask. A homogenous cell population was obtained after culturing for 3~5 weeks. MSCs from passages 8~14 were used for all experiments.

When the MSCs were 90% confluent, the cells were harvested and plated in 12-well-plates (5×10⁴ cells/well) in the complete culture medium with or without recombinant mouse IFN-γ (100 ng/mL, R&D Systems, Minneapolis, MN, USA). For TLR ligands treatment, peptidoglycans (PGN, TLR2 ligand, 5 µg/mL), poly(I:C) (TLR3 ligand, 10 µg/mL), LPS (TLR4 ligand, 10 µg/mL) (all from Sigma-Aldrich, St Louis, MO, USA), Pam3CSK4 (TLR1/2 ligand, 1 µg/mL), RecFLA-ST (flagellin, TLR5 ligand, 100 ng/mL), FSL-1 (TLR2/6 ligand, 100 µg/mL), R848 (TLR7/8 ligand, 1 µg/mL), and ODN 1826 (CpG, TLR9 ligand, 1 µg/mL) (all from InvivoGen, San Diego, CA, USA) were used for stimulation. After stimulation for 24 h, the MSCs were collected for immunophenotypic analysis or the measurement of IDO expression.

Normal MSCs or stimulated MSCs were stained for FITC-conjugated anti-CD45, CD117, allophycocyanin-conjugated anti-Flk-1, PE-conjugated anti-CD34, Alexa Fluor® 700-conjugated anti-Sca-1 (all from eBioScience, San Diego, CA, USA), allophycocyanin-conjugated anti-CD44, and Brilliant Violet 605™-conjugated anti-CD11b (all from BD Biosciences, San Diego, CA, USA). Data were analyzed using an LSRII flow cytometer (BD Biosciences).

Total RNA was isolated from cultured MSCs and colon homogenates with Trizol® (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. One microgram of total RNA was reverse transcribed into cDNA. IDO and β-actin (ACTB) transcripts were amplified using the following primers: IDO forward 5′-ATTGGTGGAAATCGCAGCTTC-3′, reverse 5′-ACAAAGTCACGCATCCTCTTAAA-3′); ACTB forward 5′-AGCTGCGTTTTACACCCTTT-3′, reverse 5′-AAGCCATGCCAATGTTGTCT-3′). Quantitative assessment of target mRNA levels was performed by quantitative RT-PCR using a CFX96™ real-time PCR detection system (Bio-Rad, Hercules, CA, USA). The quantity of mRNA was calculated using the 2 ^−ΔΔCt method, and ACTB was used to normalize total RNA quantities.

Forty micrograms of each protein sample was separated by 10% SDS-PAGE and transferred to a nitrocellulose membrane (Amersham Pharmacia Biotech, Buckinghamshire, UK). Membranes were blocked with 5% skim milk (Difco/Becton Dickinson, Atlanta, GA, USA) at 4℃ overnight and incubated for 2 h with anti-IDO antibody (1:1000 dilution, BioLegend, San Diego, CA, USA) at room temperature. The membrane was washed three times with TBST and incubated with secondary antibody for 2 h at room temperature. HRP-conjugated rat IgG antibody (1:1000 dilution, Santa Cruz Biotechnology, Dallas, TX, USA) in 5% skim milk was used as the secondary antibody. The target protein was detected using the SuperSignal™ West Dura extended duration enhanced chemiluminescence substrate (Thermo Fisher Scientific, Waltham, MA, USA).

Colitis was induced by administration of 4% DSS (molecular weight 36,000~50,000; MP Biomedicals, Santa Ana, CA, USA) in drinking water ad libitum for 7 days. 9-week-old mice were divided into three groups: DSS treatment group (DSS control), DSS with unstimulated MSCs treatment group (DSS+MSCs), and DSS with MSCs stimulated with IFN-γ and poly(I:C) group (DSS+sti-MSCs). Unstimulated MSCs or stimulated MSCs were administered by i.v. or i.p. injection on days 1 and 3 after the administration of DSS. The severity of disease was assessed using the disease activity index (DAI), which evaluates stool consistency and the presence of fecal blood, as previously described (9).

All values are expressed as mean±SEM. Statistical comparisons between groups were performed using the Mann–Whitney U test.

The MSC cultures were assayed routinely for the presence of MSC-related cell surface antigens by flow cytometric analysis. Murine bone marrow-derived MSCs were negative for CD34, CD45, CD11b, CD117, and Flk-1 but most of them expressed Sca-1 and CD44 (10). To investigate the effect of TLR ligand stimulation on the immunophenotype of MSCs, we cultured MSCs in the presence of IFN-γ with or without Pam3CSK4, PGN, poly(I:C), LPS, flagellin, FSL-1, R848, and CpG. MSCs were harvested after 24 h and their phenotype was analyzed. All MSC groups were negative for CD34, CD45, CD11b, CD117, and Flk-1 and positive for Sca-1 and CD44. There were no significant differences in MSC immunophenotype among the treatment groups (Fig. 1).

To determine the effects of activating each TLR on the expression of the tryptophan-degrading enzyme IDO, which is responsible for TLR-mediated kynurenine production in MSCs, MSC/TLR ligand coculture experiments were performed. In the presence of IFN-γ, we cultured MSCs with or without stimulation by Pam3CSK4, PGN, poly(I:C), LPS, flagellin, FSL-1, R848, and CpG, because IDO is expressed by MSCs in response to inflammation. The mRNA expression of IDO was significantly increased in TLR ligand-treated MSCs compared with that in unstimulated MSCs (Fig. 2). We also demonstrated that the protein expression of IDO was increased in stimulated MSCs compared with that in unstimulated MSCs (Fig. 2B). Poly(I:C) preconditioning of MSCs significantly increased their expression of IDO, compared with that of MSCs without TLR ligand treatment. In contrast, the other tested TLR ligands had no significant effect on IDO expression.

Our in vitro experiments demonstrated that poly(I:C) can induce increased IDO expression in MSCs. We next investigated whether there is a beneficial effect of using MSCs preconditioned with poly(I:C) for cell-based therapy to reduce the disease activity of DSS-induced colitis. Mice that were orally administered with 4% DSS for 7 days showed a significant increase in DAI, bloody diarrhea, and sustained weight loss. To examine the effect of using MSCs with an elevated expression of IDO for cell-based treatment in the DSS-induced colitis model, the mice were treated with unstimulated and poly(I:C)-stimulated MSCs by i.v. or i.p. injection on days 1 and 3 after the administration of DSS (Figs. 3 and 4). Body weight changes and the DAI score were monitored daily. The i.v.-injected groups showed increased body weight loss after MSC injection compared with DSS controls (Fig. 3B). There were no significant differences in DAI score among the three groups (Fig. 3C). The i.p.-injected groups showed no significant differences in body weight loss and DAI score (Figs. 4B and C). DSS-induced colitis is associated with a reduced colon length in diseased mice. Therefore, we measured the colon length at day 9 after the administration of DSS (11). Colon lengths did not significantly differ among the i.v.-injected (Fig. 3D), whereas the i.p. injection of poly(I:C)-stimulated MSCs ameliorated the shortening of colon length compared to that in the DSS control group without MSCs injection (Fig. 4D).

Since the colon was a major site for injury by DSS, the organ was excised and examined by quantitative RT-PCR assays for the expression of IDO. In agreement with the pathologic findings, poly(I:C)-treated MSCs showed distinct changes in their IDO expression profiles after they were administered inside the peritoneal cavity (Fig. 5). Quantitative RT-PCR assays showed an upregulation of IDO expression in colon when poly(I:C)-stimulated MSCs were i.p. injected but not when they were i.v. injected in the DSS-colitis model. The injection of unstimulated MSCs did not induce an increase of IDO expression regardless of the injection route.