Nestin+ Peyer's patch resident MSCs enhance healing of inflammatory bowel disease through IL-22-mediated intestinal epithelial repair

Abstract

Inflammatory bowel disease (IBD) is a chronic condition characterized by gastrointestinal tract inflammation and still lacks satisfactory treatments. Mesenchymal stromal cells (MSCs) show promising potential for treating IBD, but their therapeutic efficacy varies depending on the tissue of origin. We aim to investigate whether intestine Peyer's patch (PP)-derived MSCs have superior immunomodulatory effects on T cells and better therapeutic effects on IBD compared with bone marrow-derived MSCs. We isolated PPs-derived Nestin+ MSCs (MSCs^PP ) and bone marrow-derived Nestin+ MSCs (MSCs^BM ) from Nestin-GFP transgenic mice to explore their curative effects on murine IBD model. Moreover, we tested the effects of IL-22 knockdown and IL-22 overexpression on the therapeutic efficacy of MSCs^PP and MSCs^BM in murine IBD, respectively. We demonstrated that Nestin+ cells derived from murine PPs exhibit MSC-like biological characteristics. Compared with MSCs^BM , MSCs^PP possess enhanced immunoregulatory ability to suppress T cell proliferation and inflammatory cytokine production. Moreover, we observed that MSCs^PP exhibited greater therapeutic efficacy than MSCs^BM in murine IBD models. Interestingly, IL-22, which was highly expressed in MSCs^PP , could alleviate the severity of the intestinal inflammation, while knockdown IL-22 of MSCs^PP remarkably weakened the therapeutic effects. More importantly, IL-22 overexpressing MSCs^BM could significantly improve the symptoms of murine IBD models. This study systemically demonstrated that murine MSCs^PP have a prominent advantage in murine IBD treatment, partly through IL-22.

FIGURE 1

Isolation and characterization of Nestin+ cells derived from bone marrow or Peyer's patches in transgenic mice. (A) A graphical scheme of Nestin+ cell isolation. (B) Heat map showing the expression patterns of representative genes with relevant functions. (C) Expression of MSC‐related markers on bone marrow‐ and Peyer's patch‐derived Nestin+ cells. The proliferation (D) and secretion of IFN‐γ, TNF‐α, and IL‐17 (E) from splenic CD3+ T cells were analysed after T cells were cocultured with or without MSCs^BM/PP. Data were presented as the means ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. TNF, tumour necrosis factor; IFN, interferon; IL, interleukin; MSCs^BM, bone marrow‐derived Nes + MSCs; MSCs^PP, Peyer's patch‐derived Nes + MSCs.

FIGURE 2

MSCs^PP exhibit enhanced therapeutic potential compared with MSCs^BM in inflammatory bowel disease. Colitis severity assessments of Normal, IBD, MSCs^BM and MSCs^PP groups are shown, including measurement of (A) body weight, (B) disease activity index, (C) evaluation of gross morphology images, (D) colon length measurements, (E) H&E staining and corresponding histological scoring statistical histogram, (F) MPO, IFN‐γ, TNF‐α and IL‐17 expression levels as measured with corresponding murine ELISA kits, and (G) inflammatory cytokine level secreted by CD3+ T cells as assessed with flow cytometry. Scale bar, 100 μm. Data were presented as the means ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (n ≥ 3 mice in each group). Normal, control group treated with 0.9% NaCl; IBD, murine model group treated with TNBS; MSCs^BM, model group treated with MSCs^BM; MSCs^PP, model group treated with MSCs^PP.

FIGURE 3

Expression profile of MSCs^BM and MSCs^PP. (A) Venn diagram of the secretory proteins expressed in MSCs^BM and MSCs^PP. (B) The 20 most significant different secretory proteins in MSCs^PP compared with the expression of their counterparts in MSCs^BM. (C–E) The expression level of IL‐22 in MSCs^BM and MSCs^PP was confirmed by qPCR, flow cytometry and ELISA kit. (F) Immunofluorescence analysis of IL‐22 in Peyer's patches of Nestin‐GFP transgenic mice. IL‐22 + Nes + pp cells in each microscopic field were quantified and plotted in a histogram. Scale bars, 50 μm. (G) Flow cytometry analyses of CD45‐Ter119‐CD31‐IL‐22+ cells in Peyer's patches of Nestin‐GFP transgenic mice. Data were presented as the means ± SD (n = 3). ***p < 0.001, ****p < 0.0001. MSCs^BM, bone marrow‐derived Nes + MSCs; MSCs^PP, Peyer's patch‐derived Nes + MSCs.

FIGURE 4

Knockdown of IL‐22 in MSCs^PP by RNA interference. (A) The efficiency of siRNA‐mediated downregulation of IL‐22 was assessed at the RNA level and normalized with respect to the expression of gapdh. (B) Flow cytometry analysis of IL‐22 knockdown at the protein level. Colitis assessments including (C) body weight measurement, (D) disease activity index, (E) evaluation of gross morphology images, (F) colon length measurements, (G) H&E staining and corresponding histological scoring statistical histogram, and (H) inflammatory cytokine expression in CD3+ T cells detected by flow cytometry. Scale bar, 100 μm. Data were presented as the means ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (n ≥ 3 mice in each group). Normal, control group treated with 0.9% NaCl; IBD, murine model group treated with TBNS; MSCs^PPcon, model group treated with control MSCs^PP; MSCs^{PPsiIL‐22#2}, model group treated with IL‐22‐knockdown MSCs^PP.

FIGURE 5

MSCs^PP alleviate murine IBD partially through IL‐22. Colitis assessments included measurement of (A) body weight, (B) disease activity index, (C) evaluation of gross morphology images, (D) colon length measurement, (E) H&E staining and corresponding histological scoring statistical histogram, and (F) inflammatory cytokine expression in CD3+ T cells as assessed with flow cytometry. Scale bar, 100 μm. Data were presented as the means ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (n ≥ 3 mice in each group). Normal, control group treated with 0.9% NaCl; IBD, murine model group treated with TNBS; rIL‐22, model group treated with recombinant IL‐22 protein; IL‐22 B, model group treated with the IL‐22‐blocking antibody.

FIGURE 6

IL‐22 overexpressing MSCs^BM enhance the therapeutic potency. (A) The microscopical observation of MSCs^EGFP and MSCs^IL‐22 which transduced with control EGFP lentivirus and IL‐22 gene carrying‐GFP lentivirus, respectively, after 48 h. (B) The mRNA level of IL‐22 on MSCs^IL‐22 and MSCs^EGFP. (C) The protein level of IL‐22 on MSCs^IL‐22 and MSCs^EGFP detected by ELISA. The secretion of (D) IFN‐γ, (E) TNF‐α, and (F) IL‐17 from splenic CD3+ T cells were analysed after T cells were cocultured with or without MSCs^EGFP/IL‐22. Murine IBD severity assessments of Normal, IBD, MSCs^EGFP and MSCs^IL‐22 groups are shown, including measurement of (G) body weight loss, (H) disease activity index, (I) gross morphology images, (J) colon length measurements, (K) H&E staining analysis and corresponding histological scoring statistical histogram. Scale bars, 75 μm. Data were presented as the means ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (n ≥ 3 mice in each group).