Human Synovial Mesenchymal Stem Cells Expressed Immunoregulatory Factors IDO and TSG6 in a Context of Arthritis Mediated by Alphaviruses

Abstract

Infection by arthritogenic alphaviruses (aavs) can lead to reactive arthritis, which is characterized by inflammation and persistence of the virus; however, its mechanisms remain ill-characterized. Intriguingly, it has been shown that viral persistence still takes place in spite of robust innate and adaptive immune responses, characterized notably by the infiltration of macrophages (sources of TNF-alpha) as well as T/NK cells (sources of IFN-gamma) in the infected joint. Aavs are known to target mesenchymal stem cells (MSCs) in the synovium, and we herein tested the hypothesis that the infection of MSCs may promote the expression of immunoregulators to skew the anti-viral cellular immune responses. We compared the regulated expression via human synovial MSCs of pro-inflammatory mediators (e.g., IL-1β, IL6, CCL2, miR-221-3p) to that of immunoregulators (e.g., IDO, TSG6, GAS6, miR146a-5p). We used human synovial tissue-derived MSCs which were infected with O'Nyong-Nyong alphavirus (ONNV, class II aav) alone, or combined with recombinant human TNF-α or IFN-γ, to mimic the clinical settings. We confirmed via qPCR and immunofluorescence that ONNV infected human synovial tissue-derived MSCs. Interestingly, ONNV alone did not regulate the expression of pro-inflammatory mediators. In contrast, IDO, TSG6, and GAS6 mRNA expression were increased in response to ONNV infection alone, but particularly when combined with both recombinant cytokines. ONNV infection equally decreased miR-146a-5p and miR-221-3p in the untreated cells and abrogated the stimulatory activity of the recombinant TNF-α but not the IFN-gamma. Our study argues for a major immunoregulatory phenotype of MSCs infected with ONNV which may favor virus persistence in the inflamed joint.

Keywords: arthritis; arthritogenic alphaviruses; human synovial tissue-derived mesenchymal stem cells; immunoregulators.

Figure 1

Synovial tissue-derived MSCs were readily infected with ONNV. Human synovial tissue-derived MSCs were infected with ONNV at MOI 1 over 6 h or 24 h. (A) Total RNA was collected and the relative viral RNA expressions of ONNV E1, ONNV E2, and ONNV capsid were determined by qPCR (Sybr green technique). The expression was calculated by Delta Ct relative to the housekeeping gene (GAPDH). Reported values are means ± SEM of three independent experiments, and p-values were calculated using an unpaired t test: *** p < 0.001 as compared to control (uninfected cells). (B) Human synovial MSCs were probed with primary antibody rabbit anti-ONNV capsid (Dr A Merits, Estonia) and nuclei were stained with DAPI (blue). Next, an Alexa Fluor 594-conjugated donkey anti-rabbit (red) antibody was used (×60 oil objective).

Figure 2

No treatments had a significant effect on LDH release, mitochondrial activity, and cell proliferation. Human synovial tissue-derived MSCs were stimulated with TNF-α and IFN-γ at 20 ng/mL, or infected with ONNV at MOI 1 alone or in the presence of TNF-α or IFN-γ. (A) The amount of released LDH in the culture medium was determined using CytoTox 96^® non-radioactive cytotoxicity assay and expressed relative to the maximum release (Triton 1%). (B) Mitochondrial metabolic activity was determined via MTT assay. (C) Cells were probed with primary antibody rabbit anti-Ki67 and nuclei were stained with DAPI (blue). Next, an Alexa Fluor 594-conjugated donkey anti-rabbit (red) antibody was used (×60 oil objective). Reported values are means ± SEM of three independent experiments, and p-value was calculated using an ANOVA test with Bonferroni multiple comparison: ***: p < 0.001 as compared to control (untreated cells).

Figure 3

ONNV increased ISG15 mRNA expression in untreated cells or cells treated with TNF-α or IFN-γ. Human synovial tissue-derived MSCs were stimulated with TNF-α or IFN-γ at 20 ng/mL and/or infected with ONNV at MOI 1 over 6 h or 24 h. qPCR was performed to determine the expression of ISG15. Reported values are means ± SEM of three independent experiments, and p-value was calculated using an ANOVA test with Bonferroni multiple comparison: **: p < 0.01, ***: p < 0.001 as compared to control; ##: p < 0.01, ###: p < 0.001 as compared to TNF-α alone; $$$: p < 0.001 as compared to IFN-γ alone.

Figure 4

In the presence of TNF-α or IFN-γ, ONNV increased pro-inflammatory mediators. Human synovial tissue-derived MSCs were stimulated with TNF-α or IFN-γ at 20 ng/mL and/or infected with ONNV at MOI 1 over 6 h or 24 h. qPCR was performed to determine the expression of (A) IL-1β, (B) IL6, and (C) CCL2 under the different conditions. Reported values are means ± SEM of three independent experiments, and p-value was calculated using an ANOVA test with Bonferroni multiple comparison: *: p < 0.05, ***: p < 0.001 as compared to control (untreated cells); ##: p < 0.01, ###: p < 0.001 as compared to TNF-α alone; $: p < 0.05, $$: p < 0.01; and $$$: p < 0.001 as compared to IFN-γ alone.

Figure 5

ONNV increased IDO and TSG6 mRNAs expression in both untreated synovial tissue-derived MSCs and those treated with IFN-γ or TNF-α, respectively. (A) Human synovial tissue-derived MSCs were stimulated for 6 h or 24 h with IFN-γ at 20 ng/mL, or infected with ONNV at MOI 1, or infected with ONNV at MOI 1 in the presence of IFN-γ. (B) Cells were stimulated with TNF-α at 20 ng/mL, or infected with ONNV at MOI 1 alone or in the presence of TNF-α. qPCR was performed to determine the expression of IDO and TSG6 under the different conditions. Reported values are means ± SEM of three independent experiments, and p-value was calculated using an ANOVA test with Bonferroni multiple comparison: *: p < 0.05, **: p < 0.01, ***: p < 0.001 as compared to control; ###: p < 0.001 as compared to TNF-α alone; $$: p < 0.01 as compared to IFN-γ alone.

Figure 6

ONNV increased GAS6 mRNA expression but had no effect on PROS1 mRNA expression in synovial tissue-derived MSCs. Cells were infected with ONNV at MOI 1 over 6 h or 24 h. RNA was collected and the mRNA relative expression of GAS6 (A) and PROS1 (B) were determined by qPCR. Reported values are means ± SEM of three independent experiments, and p-values were calculated using an ANOVA test with Bonferroni multiple comparison: * p < 0.05 as compared to control (untreated cells).

Figure 7

No correlation between pro- and anti-inflammatory responses in cells infected with ONNV alone (A), stimulated with TNF-α (B), stimulated with TNF-α and infected with ONNV (C), stimulated with IFN-γ alone (D), and stimulated with IFN-γ and infected with ONNV (E). CCL2 has been used as pro-inflammatory gene, and IDO or TSG6 have been used as anti-inflammatory genes. Statistical analysis was performed using a non-parametric Spearman test.

Figure 8

ONNV can downregulate the expression of miR-146a-5p (alone or in response to TNF). Synovial tissue-derived MSCs were stimulated with TNF-α or IFN-γ at 20 ng/mL, or infected with ONNV at MOI 1 alone or in the presence of TNF-α or IFN-γ. qPCR was performed to determine the expression of miR-146a-5p (A) and miR-221-3p (B) under the different conditions. Reported values are means ± SEM of three independent experiments, and p-value was calculated using an ANOVA test with Bonferroni multiple comparison: *: p < 0.05, **: p < 0.01, ***: p < 0.001 as compared to control; #: p < 0.05, ###: p < 0.001 as compared to TNF-α alone.

Figure 9

An alphavirus (e.g., Chikungunya class III or ONNV class II, which are easier to study) might persist in human synovial tissue-derived MSCs through its capacity to modulate the expression of several immunoregulators, particularly in the context of innate and adaptive immune cell responses (i.e., macrophages producing TNF-α and IFN-γ produced by recruited T cells). In our study, we found that pro-inflammatory mi-RNA (e.g., miR-221) expression by MSCs is also downregulated by ONNV.