Activation of Toll-like receptor 2 promotes mesenchymal stem/stromal cell-mediated immunoregulation and angiostasis through AKR1C1

Abstract

Background: Mesenchymal stem/stromal cells (MSCs) maintain tissue homeostasis in response to microenvironmental perturbations. Toll-like receptors (TLRs) are key sensors for exogenous and endogenous signals produced during injury. In this study, we aimed to investigate whether TLRs affect the homeostatic functions of MSCs after injury. Methods: We examined the expression of TLR2, TLR3 and TLR4 in MSCs, and analyzed the functional significance of TLR2 activation using single-cell RNA sequencing. Additionally, we investigated the effects and mechanisms of TLR2 and its downstream activation in MSCs on the MSCs themselves, on monocytes/macrophages, and in a mouse model of sterile injury-induced inflammatory corneal angiogenesis. Results: MSCs expressed TLR2, which was upregulated by monocytes/macrophages. Activation of TLR2 in MSCs promoted their immunoregulatory and angiostatic functions in monocytes/macrophages and in mice with inflammatory corneal angiogenesis, whereas TLR2 inhibition attenuated these functions. Single-cell RNA sequencing revealed AKR1C1, a gene encoding aldo-keto reductase family 1 member C1, as the most significantly inducible gene in MSCs upon TLR2 stimulation, though its stimulation did not affect cell compositions. AKR1C1 protected MSCs against ferroptosis, increased secretion of anti-inflammatory cytokines, and enhanced their ability to drive monocytes/macrophages towards immunoregulatory phenotypes, leading to the amelioration of inflammatory corneal neovascularization in mice. Conclusion: Our findings suggest that activation of TLR2-AKR1C1 signaling in MSCs serves as an important pathway for the survival and homeostatic activities of MSCs during injury.

Keywords: Ado-keto reductase family 1 member C1; Cornea; Mesenchymal stem/stromal cell; Monocyte/macrophage; Toll-like receptor 2.

Figure 1

Expression of TLRs in MSCs and effect of TLR2 stimulation on PTGS2/COX2-PGE2 pathway. A. Experimental scheme. Human BM MSCs were treated with either Pam2CSK4 or anti-TLR2 Ab for 24 h. Alternatively, MSCs were cocultured in a Transwell system with PMA-differentiated THP-1 macrophages (THP-1 macro) for 18 h or with GM-CSF-stimulated murine BM monocytes (BM mono) for 5 d. Then, MSCs were assessed for expression of TLRs and PTGS2 and production of PGE2. B-D. Representative and quantitative flow cytometry results for TLR2, TLR3 and TLR4 in MSCs. FMO (fluorescence minus one) control for each Ab was used as gating control. MSCs were treated with Pam2CSK4 (100 ng/mL) or anti-TLR2 Ab (10 μg/mL). E, F. qRT-PCR quantification for TLR2, TLR3 and TLR4 in MSCs cocultured with THP-1 macrophages or BM monocytes. The mRNA levels are presented as the fold changes relative to MSCs cultured alone. G. qRT-PCR for PTGS2 and ELISA for PGE2 production in MSCs treated with Pam2CSK4 (0-1000 ng/mL). Shown are the mRNA levels relative to untreated MSCs. Data represent means ± SD from 2-3 independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: not significant, as analyzed by one-way ANOVA and Tukey's test (C, G) or by Student's t-test (D-F).

Figure 2

TLR2 signaling in MSC is crucial to induction of immunosuppressive monocytes/macrophages. A. Experimental scheme. THP-1 macrophages were cocultured with human MSCs in a Transwell system. Prior to coculturing, MSCs were subjected to pretreatment with Pam2CSK4 (100 ng/mL) or anti-TLR2 Ab (10 μg/mL) or transfected with TLR2 siRNA or SCR siRNA for 24 h. After 18 h of coculturing, THP-1 macrophages were analyzed for PTGS2 mRNA levels and the secreted levels of PGE2, IL-10 and AREG. B. qRT-PCR for PTGS2 in THP-1 macrophages. The mRNA levels are fold changes relative to THP-1 macrophages cultured alone. C. ELISA for PGE2, IL-10 and AREG in supernatants of THP-1 macrophages with or without MSC coculturing. D. Experimental scheme. Murine BM cells were cocultured with human MSCs under GM-CSF stimulation for 5 d. Prior to coculturing, MSCs were pretreated with Pam2CSK4 (100 ng/mL) or anti-TLR2 Ab (10 μg/mL) or transfected with TLR2 siRNA or SCR siRNA for 24 h. After 5 d of coculturing, BM cells were assayed. E. Representative and quantitative flow cytometry results for CD11b^hiLy6C^hiLy6G^lo and CD11b^midLy6C^midLy6G^lo cells in BM cells. F. qRT-PCR for Arg1, Nos2 and Ptgs2 in BM cells. Shown are the mRNA levels relative to BM cells cultured alone without GM-CSF. G. ELISA for PGE2, IL-10 and active TGF-β1 levels in cell-free supernatants of BM cells with or without MSC coculturing. Mouse-specific ELISA kits were used for measurement of IL-10 and active TGF-β1. The PGE2 assay kit recognizes both human and murine PGE2. H. Experimental scheme. MSCs were isolated from WT C57BL/6 or TLR2 KO mice. BM cells from WT C57BL/6 mice were cocultured with WT or TLR2 KO MSCs under GM-CSF for 5 d, and then subjected to assays. I, J. Representative flow cytometry results for CD11b, Ly6C, Ly6G and MHC class II in BM cells. K. Quantitative flow cytometric analysis for CD11b^hiLy6C^hiLy6G^lo cells, CD11b^midLy6C^midLy6G^lo cells and MHC class II^hiLy6C^hiLy6G^lo cells in BM cells. L. ELISA for murine TNF-α and IL-1β in supernatants of cell cultures. Data are means ± SD from 3 independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, as analyzed by one-way ANOVA and Tukey's test, Kruskal-Wallis test and Dunn's multiple-comparison test (CD11b^hiLy6C^hiLy6G^lo and MHC class II^hiLy6C^hiLy6G^lo cells in K) or Student's t-test (L).

Figure 3

TLR2 signaling in MSCs is essential for suppression of inflammatory angiogenesis in the cornea following sterile injury. A. Experimental protocol. Immediately after corneal suturing injuries, MSCs, Pam2CSK4-pretreated MSCs, TLR2 siRNA-transfected MSCs, SCR siRNA-transfected MSCs or HBSS (vehicle) were administered via tail vein injection in BALB/c mice. Seven days later, the corneas were subjected to assays. B, C. Representative corneal photographs and microphotographs of CD31/LYVE1-stained corneal whole-mounts. Scale bar: 500 μm for the first and second rows and 200 μm for the third row (magnified images of yellow-outlined insets from the first row). D, E. Clinical scoring of corneal NV and quantification of CD31-stained area in corneal whole-mounts. F, G. qRT-PCR for pro-angiogenic factors and pro-inflammatory cytokines in cornea. The mRNA levels are presented relative to those in BALB/c control corneas that had not received injury or treatment. H. Experimental protocol. Corneal sutures were applied to TLR2 KO mice, and either MSCs or HBSS were intravenously injected. After 7 d, the corneas were subjected to assays. I, J. Representative corneal photographs and microphotographs after CD31/LYVE1 immunostaining. Scale bar: 500 μm for the upper row and 200 μm for the lower row (magnified images of yellow-outlined insets from the upper row). K, L. Clinical scoring of corneal NV and measurement of CD31- and LYVE1-stained areas. Data represent means ± SD, where a circle indicates the data from an individual animal. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, as analyzed by one-way ANOVA and Tukey's test or by Student's t-test (Injury + MSC vs. Injury + MSC/Pam2CSK4 in F, G).

Figure 4

Single-cell transcriptional profiling of Pam2CSK4-treated and untreated MSCs. A. Experimental scheme. MSCs treated with Pam2CSK4 100 ng/mL for 24 h and untreated MSCs were subjected to scRNA-seq on the 10x Genomics Chromium platform. B. Combined UMAP and tSNE plots of Pam2CSK4-treated MSCs (blue) and untreated MSCs (orange) showing overlapping. C. tSNE plots of Pam2CSK4-treated MSCs and untreated MSCs depicting 7 clusters (C0-C6) in both cell libraries. D, E. Feature and dot plots displaying positive and negative markers for MSCs. F. Heatmap of top 10 DEGs in Pam2CSK4-treated MSCs vs. untreated MSCs as determined by RNA-Seq analysis. AKR1C1 was identified as the top DEG upregulated in Pam2CSK4-treated MSCs relative to untreated MSCs. G. ELISA for secreted levels of AKR1C1 in cell-free supernatants of MSC cultures treated with Pam2CSK4 (0-100 ng/mL). H. qRT-PCR and ELISA for mRNA and protein levels of AKR1C1 in MSCs transfected with TLR2 siRNA or control SCR siRNA. Means ± SD are presented. **p < 0.01 as analyzed by Kruskal-Wallis test and Dunn's multiple-comparison test (G) or Student's t-test (H).

Figure 5

AKR1C1 inhibition promotes MSC ferroptosis. A. Cell viability assay in MSCs transfected with AKR1C1 siRNA or control SCR siRNA or in MSCs treated with AKR1C1 inhibitor, 5-PBSA (10 or 100 ng/mL). B, C. Quantification of intracellular Fe²⁺ fluorescence intensity in MSCs and representative fluorescence images. Red indicates FerroOrange staining, while blue represents DAPI-stained nuclei. Scale bar: 50 μm for the upper row and 200 μm for the lower row. Data are means ± SD from 3-4 independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns: not significant, as analyzed by one-way ANOVA and Tukey's test.

Figure 6

AKR1C1 upregulates production of therapeutic factors in MSCs, whereas AKR1C1 inhibition abrogates it. A. Experimental scheme. MSCs were subjected to AKR1C1 inhibition or replenishment and analyzed for the expression of key therapeutic factors. B. qRT-PCR for key therapeutic factors in MSCs after transfection with AKR1C1 siRNA or SCR siRNA. C. qRT-PCR in MSCs with rhAKR1C1 replenishment (1 or 5 ng/mL). D, E. qRT-PCR and ELISA for mRNA and secreted protein levels of key therapeutic factors in MSCs with 5-PBSA treatment (10 or 100 ng/mL). Data are means ± SD from 2-3 independent experiments. *p < 0.05, ***p < 0.001, ****p < 0.0001, as analyzed by one-way ANOVA and Tukey's test, Kruskal-Wallis test and Dunn's multiple-comparison test (STC1, TGFB1, SDF1 in C and STC-1 in E) or Student's t-test (B).

Figure 7

AKR1C1 downregulates pro-inflammatory cytokines in monocytes and macrophages. A. Experimental scheme. BM cells were cultured under GM-CSF stimulation for 5 d in the presence or absence of rhAKR1C1 (1 or 5 ng/mL) or 5-PBSA (10 or 100 ng/mL). Then, the cells were assessed for MDSC marker Arg1 and pro-inflammatory cytokines. B. qRT-PCR for Arg1, Tnfa and Il1b in BM monocytes. The mRNA levels are presented as fold changes relative to GM-CSF-unstimulated BM cells. C. ELISA for TNF-α and IL-1β secretion in culture supernatants of BM monocytes. D. Experimental scheme. THP-1 macrophages were stimulated with LPS followed by ATP, and were treated with rhAKR1C1 (1 or 5 ng/mL) or 5-PBSA (10 or 100 ng/mL). 18 h later, the cells were analyzed for pro-inflammatory cytokine expression. E, F. qRT-PCR and ELISA for transcript and protein levels of TNF-α, IL-1β and IL-12B in LPS/ATP-stimulated THP-1 macrophages. Shown are the mRNA levels relative to LPS/ATP-unstimulated THP-1 macrophages. Means ± SD are presented. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, as analyzed by one-way ANOVA and Tukey's test.

Figure 8

AKR1C1 in MSCs is required for amelioration of inflammatory angiogenesis in the cornea following sterile injury. A. Experimental protocol. Immediately after corneal suturing injury, AKR1C1 siRNA-transfected MSCs, SCR siRNA-transfected MSCs or HBSS (vehicle) were intravenously administered into BALB/c mice. Seven days later, the corneas were assayed. B, C. Representative corneal photographs and microphotographs of corneal flat mounts with CD31/LYVE1 immunostaining. Scale bar: 500 μm or 200 μm (magnified views of yellow-outlined insets). D, E. Clinical scoring of corneal NV and quantification of LYVE1-stained area in corneal whole-mounts. F, G. Representative flow cytometry cytograms of CD11b^hiLy6C^hi cells in the spleen (F) and quantitation of CD11b^hiLy6C^hi cells in spleen and blood (G). H, I. qRT-PCR assays for pro-inflammatory cytokines in ocular draining cervical lymph nodes (DLNs) and blood cells. The mRNA levels are presented as fold changes relative to naïve BALB/c mice without injury or treatment. Data represent means ± SD, where a circle indicates the data from an individual animal. *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant, as analyzed by one-way ANOVA and Tukey's test.