Poly(I:c)-priming Improved the Therapeutic Efficacy of Mesenchymal Stromal Cells on Experimental Colitis by Promoting the Expression of Indoleamine 2,3-dioxygenase


 Background:Inflammatory bowel disease is a chronic and excessive inflammation of the colon and small intestine. We previously reported that priming of mesenchymal stromal cells (MSC) with poly(I:C) induced them to express indoleamine 2,3-dioxygenase (IDO). We tried to find out whether the poly(I:C)-primed MSCs have better therapeutic efficacy on the experimental colitis in the IDO1-dependent manner. Methods:To compare the therapeutic effects between the naïve MSCs and primed MSCs on murine colitis, mice (C57BL6) were administered with 2.5% dextran sodium sulfatein drinking water for 5 days and injected with MSCs intraperitoneally on days 1 and 3 following DSS ingestion. The disease activity index score and body weight loss were significantly improved after injection of the primed MSCs. Results:Mice receiving the poly(I:C)-primed MSCs showed a reduced disease activity index and less weight loss. Colon tissue from the same mice presented attenuated pathological damage, increased Paneth cells, increased IDO1-expressing cells, and better proliferation of enterocytes. The primed MSC treatment upregulated the mRNA expression of intestinal stem cell markers (Lgr5, Olfm4, and Bmi1), enterocyte differentiation markers (Muc2, Alpi, Chga, and occludin), and regulatory T (Treg) cells (Foxp3). The same treatment decreased inflammatory cell infiltration to lymphoid organs and the level of pro-inflammatory cytokines (IL-1β, TNF-α, IL-6, and MCP-1) in colon tissue. Notably, in vivo pharmacologic inhibition of the IDO1 activity blocked the Foxp3 up-regulation in colon tissue and diminished the protective effects of the primed MSC. Conclusions: The priming of MSCs with poly(I:C) is a promising new strategy to improve the therapeutic efficacy of MSC and is worth further research.

Recently, the allogeneic adipose-derived MSC product (Cx601) showed favorable outcomes in the phase III clinical trial and obtained the landmark approval in Europe for the treatment of complex perianal stulas in Crohn's disease (16). Other studies also found the positive effects of locally injected MSCs, but they indicated that the e cacy decreased over time and that there was perhaps a need for repeated treatments (17). In contrast to the success of local injection, systemic MSC therapy given intravenously for the luminal IBD has shown limited e cacy and inconsistent results (18). More mechanistic studies in representative murine models of IBD are needed to bring the immune modulation by MSC to the clinic.
In our previous work, we primed murine BM-derived MSCs with various toll-like receptor (TLR) ligands and found that only the TLR3 ligand polyinosinic-polycytidylic acid [poly(I:C)] signi cantly increased the expression of IDO. Furthermore, poly(I:C)-treated MSCs could improve the pathologic scores of Dextran sulfate sodium (DSS)-induced colitis more effectively than could unstimulated MSCs (19). In the present study, we demonstrated that the poly(I:C)-primed MSCs increased the Treg frequency, decreased in ammation, stimulated epithelial regeneration, and had better therapeutic e cacy on the murine IBD model than did unstimulated MSCs.

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

Priming of murine BM-derived MSCs
The BM-derived MSCs were isolated from C57BL/6 mice and expanded, as described in our previous work (19). In brief, BM cells were ushed out from femurs and tibias, plated in 75 cm 2 tissue culture asks at a concentration of 1 × 10 6 cells/mL in the complete culture medium, and incubated at 37℃ and 5% CO 2 .
Non-adherent cells were removed after three days, and the remaining cells were passaged into a new ask when the cells reached 70 ~ 80% con uency. To do priming, we harvested cells at the 90% con uency and plated them in 12-well plates (5 × 10 4 cells/ well) in the complete culture medium supplemented with recombinant mouse IFN-γ (100 ng/mL, R&D Systems, Minneapolis, MN, USA).
Poly(I:C) (TLR3 ligand, 10 µg/mL, Sigma-Aldrich, St Louis, MO, USA) was added to the culture medium for stimulation. The primed MSCs were collected after 24 h and used for in vitro and in vivo experiments.

DSS-induced murine colitis and MSC treatments
Experimental colitis was induced by administration of 2.5% DSS (molecular weight 36,000 ~ 50,000; MP Biomedicals, Santa Ana, CA, USA) in drinking water ad libitum for six days (day 0 ~ 5). Nine-week-old mice were randomly allocated into one of three groups: DSS only (DSS control), DSS with unstimulated MSCs treatment (DSS + MSCs), and DSS with the treatment of MSCs stimulated with IFN-γ and poly(I:C) (DSS + primed-MSCs). Unstimulated MSCs or the primed (stimulated) MSCs were injected intraperitoneally (i.p.) on days 1 and 3. We assessed the severity of colitis using body weight and the disease activity index (DAI), which evaluates stool consistency and the presence of fecal blood as previously described (20). All mice were sacri ced on day 9 to harvest the spleen, mesenteric lymph node (mLN), and colon tissue.

L-1MT preparations
To prepare L-1-methyl tryptophan (L-1MT, Sigma-Aldrich, St. Louis, MO) for oral gavage, 1 g of L-1MT was added to a 15-ml conical tube with 7.8 ml Methocel/Tween [0.5% Tween 80/ 0.5% Methylcellulose (v/v in water; both from Sigma-Aldrich)]. The following day, the L-1MT concentration was adjusted to 85 mg/ml by adding an additional 4 ml Methocel/Tween and mixing again. For in vitro use, 1-MT was prepared as a 100 mmol/L stock in 0.1 N NaOH, adjusted to pH 7.4 and stored at -20 °C protected from light.
Hematoxylin-eosin (H&E) staining and Immunohistochemistry (IHC) We subjected Formalin-xed, para n-embedded tissue sections to H&E staining for microscopic examination. Slides were scored by a pathologist (blinded to experimental group). Pathologic severity of IBD was assessed by three parameters (epithelial loss, crypt destruction, and in ammatory cell in ltration), as described before (20).
Tissue sections (4 ) were mounted on super frost glass sliders and depara nized in xylene and a graded series of ethanol, followed by antigen retrieval. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide. Nonspeci c binding sites were saturated by exposure to 10% normal goat serum diluted in phosphate buffered saline (PBS) for 60 min. We incubated slides overnight at 4℃ with primary antibodies against mouse Ki-67 (1:100 dilution, Abcam, CB, UK), Lysozyme (1:250 dilution, Abcam), IDO-1 (1: 400 dilution, Biolegend, San Diego, CA), then washed with PBS for 10 min. Biotinylated goat anti-rabbit IgG and rabbit anti-rat IgG (Vector Laboratories, Burlingame, CA) secondary antibodies were applied to tissue sections, and the slides were incubated at room temperature for 30 min. After the sections were washed and incubated for 30 minutes with peroxidase-conjugated streptavidin (Dako, Glostrup, Denmark) at room temperature, 3,3'-diaminobenzidine was added to visualize antigens. Sections were counterstained with Mayer's hematoxylin, dehydrated, cleared, and mounted. We prepared negative control tissue samples in the same manner as described above, except that the primary antibody was omitted or replaced with an isotype control antibody (R&D Systems, Minneapolis, MN).

Quantitative RT-PCR
We isolated total RNA from colon homogenates with Trizol® (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. One microgram of total RNA was reverse transcribed into cDNA. We did quantitative assessment of target mRNA levels by quantitative RT-PCR using a CFX96TM real-time PCR detection system (Bio-Rad, Hercules, CA, USA). The quantity of mRNA was calculated using the 2 − ΔΔCt method, and the level of β-actin was used to normalize total RNA quantities. The sequences of forward and reverse primers are shown in Table 1.

Treg generation
We did the following experiments to analyze the effect of MSCs on T-cell proliferation. We isolated untouched T cells from splenocytes using the Pan T Cell Isolation Kit (Miltenyi Biotec, Bergisch Gladbach, Germany). We cocultured 2 × 10 5 splenocytes with or without MSCs (primed or unstimulated) in the presence or absence of 2 /ml anti-CD3/CD28 (ebioscience, San Diego, CA, USA) plus 10 ng/ml recombinant murine TGF-β (R&D Systems) and 50 ng/ml recombinant murine IL-2 (R&D Systems) for 72 hours. Then, T cells were harvested and surface stained for CD4, CD25, and Foxp3. A competitive IDO1 inhibitor, L-1-MT (Sigma-Aldrich, St. Louis, MO, USA) was added to some wells.

Statistical analysis
All values are expressed as mean ± standard error of the mean. We did statistical comparisons between groups using a parametric independent sample t-test if there were ≥ 5 animals per group, or used the Mann-Whitney test if there were < 5 animals per group.

Results
The poly(I:C)-primed MSCs were more effective than unstimulated MSCs to ameliorate the DSS-induced colitis in mice To identify the immunomodulatory effects of the poly(I:C)-primed MSC, we used the murine DSS-induced colitis model (Fig. 1A). MSCs were primed with poly(I:C) as described before (26) and injected i.p. to mice on days 1 and 3 (DSS + primed MSCs group). Other mice were treated with the same volume of saline (DSS group) or unstimulated MSC (DSS + MSCs group) on days 1 and 3. Compared to the DSS group and DSS + unstimulated MSC group, the DSS + primed MSCs group had reduced DAI scores and less weight loss ( Fig. 1B-C). Additionally, the primed MSC treatment shortened the colon length less, which is the anatomic marker of colonic in ammation (Fig. 1D). A pathologic examination con rmed the symptomatic improvement of the DSS-induced colitis, which was caused by the primed MSC. Figure 1E shows that the typical pathologic ndings of the DSS-induced colitis (epithelial loss, crypt destruction, and in ammatory cell in ltration) were signi cantly ameliorated by the unstimulated MSC and further by the primed MSC. These results demonstrated the improved anti-in ammatory effects of the primed MSC compared to the unstimulated MSC.
The poly(I:C)-primed MSCs signi cantly stimulated the intestinal stem cell (ISC) proliferation, enterocyte differentiation, and epithelial regeneration We did the IHC staining of a cell proliferation marker, Ki-67, using colon tissues, which were harvested on day 9. The Ki-67 expression was prominently increased in the DSS + primed MSC group more than in other groups, suggesting the intestinal epithelial layer was recovered signi cantly faster ( Fig. 2A). Next, we investigated the ISC proliferation and Wnt/β-catenin signaling pathway, since they play a pivotal role in maintaining intestinal homeostasis (21). Total RNA was isolated from colon tissues, and the mRNA of target genes was quanti ed by a real-time RT-PCR method. Markers of active ISC (Lgr5 and OLFM4) and quiescent '4+' ISC (Bmi1) were signi cantly increased in the primed MSC group more than in the other groups. The expression of Axin2, a target gene of the Wnt/β-catenin signaling pathway, was increased in the primed MSC group (Fig. 2B). The mRNA levels of Muc2 (goblet cells), Alpi (enterocytes), Chga (enteroendocrine cells), and occludin (epithelial tight junction) were prominently upregulated in the primed MSC group more than in the other groups. (Fig. 2C). The IHC staining of lysozyme revealed that the primed MSC treatment signi cantly increased Paneth cells and preserved crypt structures better than the unstimulated MSC did (Fig. 2D). These results con rmed that the poly(I:C) priming, i.e., TLR3 stimulation, increased the MSC's e cacy in promoting intestinal epithelial regeneration and maintain homeostasis.
The poly(I:C)-primed MSC decreased in ammatory response in the spleen, mLN, and colon tissue IBD is a chronic in ammatory state of the gastrointestinal tract, which involves effector T cells and in ammatory cytokines (1). Therefore, we studied the anti-in ammatory and immunomodulatory effects of the primed MSC. Spleen, mLN, and colon tissues were harvested on day 9, as described in Fig. 1A. Like other ndings, the primed MSC treatment reduced the in ammatory cell in ltration into lymphoid organs more effectively than the unstimulated MSC did. The numbers of monocytes (CD11b + ), dendritic cells (CD11c + ), and CD4 + T cells signi cantly decreased in the spleen and mLN from mice that had received the primed MSC treatment (Fig. 3A-C). The mRNA levels of in ammatory cytokines (IL-1β, TNF-α, and IL-6) and a chemotactic factor (MCP-1) in colon tissue decreased more in the primed MSC group than in other groups (Fig. 3D). The mRNA expression of immune-suppressive cytokine (IL-10) in colon tissue was elevated in the primed MSC group (Fig. 3F). These ndings can account for the increased e cacy of the poly(I:C)-primed MSC on the IBD.

The poly(I:C)-primed MSC expanded Treg cells in an IDO1dependent manner
The Foxp3 + Treg cells were functionally defective or its frequency was signi cantly lower in active IBD patients (22). Interestingly, the anti TNFα therapy, such as in iximab, signi cantly increased the frequency of functional Foxp3 + Treg cells in patients with active IBD (23). In this study, we investigated the change of Foxp3 + Treg cell frequencies in lymphoid organs and colon tissue after treatments of the poly(I:C)primed or unstimulated MSCs for DSS-induced colitis. The proportion of Treg (CD25 + Foxp3 + CD4 + ) to CD4 + T cells in spleen was signi cantly higher in the primed MSC group than in the unstimulated MSC group and the DSS control group. Difference of Treg proportion between the primed and unstimulated MSC treatments was not prominent in mLN (Fig. 4A). The mRNA of Foxp3 in colon tissue increased considerably more in the primed MSC group than in the other groups. The unstimulated MSC could not increase the Foxp3 expression compared to the DSS control group, although they showed some antiin ammatory effects on the IBD. We suppose that the ability of MSCs to induce the Treg cell expansion is the most important improvements induced by the poly(I:C) priming. To further con rm the capacity of the primed MSCs for expanding Treg, we did in vitro experiments. In brief, T cells were isolated from splenocytes of B6 mice and co-cultured with the primed or unstimulated MSC. The primed MSC caused a signi cantly higher increase of Treg (CD25 + Foxp3 + ) proportion to CD4 + T cells than the unstimulated MSCs did (Fig. 4C). Since we demonstrated that the poly(I:C) priming of MSC induced the considerable IDO1 up-regulation (19), we tried to nd out whether the primed MSC promoted Treg cells in an IDO1dependent manner. Figure 4D shows that a competitive IDO1 inhibitor, L-1-MT, diminished the increment of Treg (CD25 + Foxp3 + ) proportion induced by the primed MSCs, con rming our hypothesis.

The poly(I:C)-primed MSCs increased the expression of colonic IDO1 and COX2
We explored how the IDO1 and PGE2 pathways were affected by the MSC treatment, since they promote intestinal homeostasis by limiting in ammatory responses and protecting the epithelium (24). We did the IHC staining of IDO1 using colon tissues, which were harvested on day 9, as described in Fig. 1A. The primed MSC treatment induced a signi cant increment of IDO1-expressing cells in intestinal crypts than the unstimulated MSCs did, suggesting the role of IDO1 in maintaining the intestinal epithelial layer (Fig. 5A). The real-time RT PCR con rmed the same nding for the IDO1 expression (Fig. 5B). We found that the poly(I:C)-primed MSCs induced a signi cantly higher expression of Cyclooxygenase 2 (COX2) and Prostaglandin E synthase 3 (PTGES3) genes in colon tissue than did the unstimulated MSC, suggesting that the local PGE2 level may be increased in the in amed site (Fig. 5B).
Pharmacologic IDO1 inhibition decreased the therapeutic e cacy of the poly(I:C)-primed MSCs on the DSS-induced colitis.
Next, we tried to nd out whether IDO1 is essential for the protective effects of the primed MSCs. We induced experimental colitis and injected MSCs, as described in Fig. 1A. In addition, a competitive IDO inhibitor, L-1-MT (200 mg/kg), was administered by oral gavage from day 1 to day 5. Figure 6 suggests clearly that the pharmacologic IDO1 inhibition abolished the bene cial effects of the primed-MSC treatment, such as less weight loss, reduced DAI scores, and preserved colon length (Fig. 5B-D). Importantly, the mRNA expressions of Foxp3 and IDO1, which were increased by the primed-MSC treatment, were down-regulated by the IDO1 inhibition. In contrast, the expression of IL-10 was not affected by the IDO1 inhibition, implying that IL-10 was possibly secreted from monocytes rather than from Treg cells in the colon tissue of the experimental IBD, or that other mediators besides IDO1 might be involved in the IL-10 production (Fig. 5E). These results altogether suggest that IDO1 is an important mediator of the poly(I:C)-primed MSCs in promoting intestinal epithelial recovery and suppressing in ammatory responses.

Discussion
Several clinical trials using MSCs have reported that they have been safe and effective treatments for IBD until now. However, each trial used different procedures for tissue harvest, MSC isolation, and cell culture.
The routes and timing of MSC administration were also diverse. This aspect made the obtained results frequently inconsistent and unimpressive (25). The cell priming (also referred to as pre-conditioning) is one of the most studied approaches to overcome this problem and has been known to improve the immune-modulatory attributes of MSCs. Previous studies demonstrated that MSC priming with proin ammatory cytokines, such as IFN-γ and TNF-α, increased the secretion of anti-in ammatory cytokines (IDO1, PGE2, TGF-β, and hepatocyte growth factor), the expression of chemokine ligands (CXCL9, CXCL10, and CXCL11), adhesion proteins (VCAM-1 and ICAM-1), and immune checkpoint molecules, such as PD-L1 (26,27). In experimental colitis, human IFN-γ-primed MSCs showed higher migration rates to in ammatory sites and a signi cant reduction of mucosal damage and in ammatory responses than did non-primed MSC (28). There is still some room for improvement in the IFN-γ priming of MSCs, because its effects can be inconsistent (29) and transient (30). A few studies have been published regarding the priming of MSCs with TLR ligands, since TLRs are involved with the immune-modulating functions of MSCs (31). Waterman and colleagues reported that TLR4-primed MSCs mostly elaborated pro-in ammatory cytokines, whereas TLR3-primed MSCs expressed mostly immunosuppressive mediators (32). We previously showed that the priming of MSCs with a TLR3 ligand, poly(I:C), in the presence of IFN-γ increased IDO1 production and increased the immunomodulatory effects of MSCs. Other TLR ligands, Pam3CSK4 (TLR1/2), peptidoglycan (TLR2), LPS (TLR4), agellin (TLR5), FSL-1 (TLR2/6), R848 (TLR7/8), and CpG (TLR9) did not increase IDO1 expression compared to IFN-γ alone (19). Therefore, we hypothesized that the poly(I:C) priming of MSCs could increase the therapeutic e cacy in an IDO1-dependent manner.
The up-regulation of IDO1 is observed not only in the experimental murine colitis but also from the clinical samples of human IBD. CD103 + gut dendritic cells can express IDO1 and support Treg conversion while suppressing Th1/Th17 differentiation to limit gut in ammation (33). In the homeostatic state, gut expression of IDO1 is low and limited to the cells of the lamina propria. However, the in ammatory cytokines, including IFN-γ, TNF-α, and IL-1β, stimulate the IDO1 expression in epithelial cells, which become a signi cant source of IDO1 activity in IBD. The IDO1 expression is more apparent in epithelial cells near the sites of ulceration (34). In the experimental colitis model, the administration of the TLR-9 agonist improved clinical and histological parameters via the induction of IDO1, and the inhibition of IDO1 activity abrogated the protective effects (35). Gurtner et al. also demonstrated that the administration of IDO1 inhibitor 1-DL-MT worsened the IBD activity in mice, suggesting that IDO1 downregulated Th1 responses within the intestinal tract (36). The protective effects of IDO1 were con rmed in a mouse model of graft versus host disease. Jasperson et al. showed that IDO1 −/− mice exhibited greater colitis severity, T-cell in ltration, and mortality (37). They also demonstrated that induction of IDO1 primarily in professional APCs by a TLR-7/TLR-8 agonist reduced colon injury and decreased lethality (38). IDO1 secreted from MSCs also polarized the differentiation of monocytes into IL-10-producing CD206 + M2 macrophages, which in turn promote T-cell suppression (39). In addition, IDO1 blocked the intestinal bacterial growth and mediated epithelial barrier protection induced by IL-27 (40). It is worthy of note that IDO1 expression supported epithelial proliferation independently of the effects on adaptive immunity through the activation of the Wnt/β-catenin signaling pathway in the colitis-associated cancer model (azoxymethane/DSS) (41). Taken together, IDO1 expression by APCs may be critical to suppressing in ammatory T-cell responses, whereas epithelial IDO1 activity functions to limit microbial invasion and promote epithelial repair. Further studies are required to nd out which one predominantly contributes to ameliorating IBD.

Conclusion
In this study, we demonstrated that the poly(I:    The poly(I:C)-primed MSCs decreased in ammatory response in the spleen, mLN, and colon tissue. Mice with the DSS-induced colitis were sacri ced on day 9 to harvest the spleen, mLN, and colon tissues.