miR-221-3p Drives the Shift of M2-Macrophages to a Pro-Inflammatory Function by Suppressing JAK3/STAT3 Activation

Abstract

Objectives: Macrophages are conventionally classified as pro-inflammatory (M1) and anti-inflammatory (M2) functional types. There is evidence for a predominance of macrophages with an inflammatory phenotype (M1) in the rheumatoid arthritis (RA) synovium. MicroRNAs (miRs) play a pivotal role in regulating the inflammatory response in innate immune cells and are found at dysregulated levels in RA patients. Here we explored miRs that tune the inflammatory function of M2-macrophages. Methods: Expression profiles of miR-221-3p and miR-155-5p were analyzed in clinical samples from RA, other inflammatory arthritis (OIA), osteoarthritis (OA), and healthy donors (HD) by qPCR. In vitro generated macrophages were transfected with miR-mimics and inhibitors. Transcriptome profiling through RNA-sequencing was performed on M2-macrophages overexpressing miR-221-3p mimic with or without LPS treatment. Secretion of IL-6, IL-10, IL-12, IL-8, and CXCL13 was measured in M1- and M2-macrophages upon TLR2/TLR3/TLR4-stimulation using ELISA. Inflammatory pathways including NF-κB, IRF3, MAPKs, and JAK3/STAT3 were evaluated by immunoblotting. Direct target interaction of miR-221-3p and predicted target sites in 3'UTR of JAK3 were examined by luciferase reporter gene assay. Results: miR-221-3p in synovial tissue and fluid was increased in RA vs. OA or OIA. Endogenous expression levels of miR-221-3p and miR-155-5p were higher in M1- than M2-macrophages derived from RA patients or HD. TLR4-stimulation of M1- and M2-macrophages resulted in downregulation of miR-221-3p, but upregulation of miR-155-5p. M2-macrophages transfected with miR-221-3p mimics secreted less IL-10 and CXCL13 but more IL-6 and IL-8, exhibited downregulation of JAK3 protein and decreased pSTAT3 activation. JAK3 was identified as new direct target of miR-221-3p in macrophages. Co-transfection of miR-221-3p/miR-155-5p mimics in M2-macrophages increased M1-specific IL-12 secretion. Conclusions: miR-221-3p acts as a regulator of TLR4-induced inflammatory M2-macrophage function by directly targeting JAK3. Dysregulated miR-221-3p expression, as seen in synovium of RA patients, leads to a diminished anti-inflammatory response and drives M2-macrophages to exhibit a M1-cytokine profile.

Keywords: JAK3/STAT3; RNA-seq; TLRs; innate immunity; macrophages (M1/M2); microRNAs; rheumatoid arthritis.

Figure 1

Comparison of miR-221-3p and miR-155-5p expression levels in clinical samples from healthy donors (HD), patients with rheumatoid arthritis (RA), other inflammatory arthritis (OIA) and osteoarthritis (OA). (A) miR expression levels in the synovial fluid of RA patients were compared to patients with OIA (psoriatic- and spondyloarthritis) and OA, and miR expression in synovial tissue was compared between biopsy samples from RA and OA patients. (B) miR expression levels were measured in plasma, PBMCs and isolated CD14⁺ cells from the peripheral blood of RA patients and HD. miR expression was measured by qRT-PCR and normalized to control miRs (miR-16-5p, miR-103-5p, miR-15b-5p, or spike-in cel-miR-39) and expressed as 2^−ΔCT by boxplot with min/max whiskers. N = 4–6, *p < 0.05. (C) −ΔCT level of miR-221-3p was plotted against miR-155-5p level from synovial fluid, tissue and CD14⁺ cells from blood and the Pearson or Spearman correlation and p-values were computed. Data are plotted with regression coefficients (solid line) and 95% confidence intervals (broken lines). (D) Expression levels of tested miRs were compared in M1- and M2-macrophages generated from CD14⁺ cells of PBMCs from HD or RA patients and presented as 2^−ΔCT by boxplot with min/max whiskers. N = 3–6, *p < 0.05.

Figure 2

Regulation of miR 221-3p and miR-155-5p in M1- and M2-macrophages by TLR ligands. Macrophages differentiated from CD14⁺ cells from (A) HD or (B) RA blood were stimulated with either 300 ng/ml Pam3 (TLR2), 10 μg/ml PolyIC (TLR3) or 100 ng/ml LPS (TLR4) for 24 h. Changes in miR expression levels were measured by qRT-PCR. Values were normalized to RNU48, miR-103a-3p, or miR-15b-5p and presented as 2^−ΔCT by boxplot with min/max whiskers. N = 3–6, *p < 0.05 compared to untreated cells (ctrl).

Figure 3

Gene expression profile of M2-macrophages transfected with miR-221-3p mimic plus/minus LPS stimulation. (A) RNA-seq was performed of in vitro generated M2-macrophages transfected with control miR (ctrl) or miR-221-3p mimic (221) and stimulated with LPS for 8.5 h or left untreated (NS). The heatmap displays the centered and scaled gene expression levels of genes differentially expressed between miR-221-3p transfected and control cells in LPS-stimulated conditions. To better visualize the differences across transfected and control cells, the expression levels were corrected for donor-specific effects (see RNA-seq methods in the Supplementary File 1). (B) Volcano plot showing the differentially expressed genes between cells transfected with miR-221-3p mimic and control cells in LPS-stimulated cells. Significant genes are colored in red (up-regulated) and blue (down-regulated). For clarity, only genes with an absolute log2-fold change larger than 1 are annotated.

Figure 4

Effects of miR-221-3p mimics on cytokine and chemokine release in M1- and M2-macrophages. (A) M1- and M2-macrophages differentiated from CD14⁺ cells from HD blood were transfected with mimic of miR-221-3p (221 mimic) or a respective control miR (ctrl mimic) and stimulated with 100 ng/ml LPS or left untreated. Secreted cytokines and chemokines were measured after 8.5 and 24 h by ELISA and values are expressed as mean ± S.D. N = 6–10, *p < 0.05. (B) M2-macrophages differentiated from CD14⁺ cells from blood of RA patients were transfected with mimic of miR-221-3p (221 mimic) or a respective control miR (ctrl mimic) and stimulated with 100 ng/ml LPS or left untreated. Secreted cytokines and chemokines were measured after 8.5 and 24 h by ELISA and values are expressed as mean ± S.D. N = 6–11, *p < 0.05. (C) Anti-inflammatory activity of generated M2-macrophages from HD or RA patients was calculated after 8.5 and 24 h LPS-stimulation by the ratio of secreted cytokines IL-10 to IL-6 or chemokines CXCL13 to IL-8. Values are expressed as mean ± S.D. N = 6–11, *p < 0.05.

Figure 5

JAK3 as top downregulated gene of unstimulated M2-macrophages transfected with miR-221-3p. (A) Volcano plot showing the differentially expressed genes between cells transfected with miR-221-3p mimic and control cells in untreated conditions. Significant genes are colored in red (up-regulated) and blue (down-regulated). (B) Significantly enriched KEGG pathways among differentially expressed genes upon transfection in untreated conditions after gene set enrichment analysis. The absolute log2 fold-change (x-axis) is shown relative to the significance (y-axis). Significant pathways are colored in yellow (up-regulated pathway) and blue (down-regulated pathway). The dot size indicates the number of genes annotated in each pathway.

Figure 6

miR-221-3p is directly interfering with JAK3/STAT3 signaling in TLR4-stimulated M2-macrophages. (A) Location of two predicted miR-221-3p targets, Site-A and Site-B, within the 3'UTR region of the human JAK3 mRNA. Sequence of Site-A contains seed match region identical to already established miR-221-3p target on CDKN1B gene (gray shading). (B) Luciferase reporter assay of constructs carrying JAK3 Site-A and JAK3 Site-B target sites, co-transfected with increasing concentration of miR-221-3p (221 mimic) or RNA duplex carrying randomized base pairs (negative ctrl). Luciferase activity represents Renilla/firefly values normalized 0 nM and is expressed as mean ± S.D. N = 3–8, *p < 0.05. Differentiated M2-macrophages derived from CD14⁺ cells of HD blood were transfected with miR-221-3p mimic (221 mimic) or a respective control miR (ctrl mimic) and stimulated with 100 ng/ml LPS for (C) 16 h and (D) 1 h or left untreated. Protein levels of (C) JAK3 and phosphorylated STAT3 and nuclear shuttling of (D) p65 (NF-κB) and IRF3 were analyzed by SDS-PAGE and visualized by immunoblotting. TBP and β-Tubulin were used as loading controls for nuclear protein extract or whole cell lysates, respectively. Shown are representative images and quantification of N = 3–5. Quantification of protein expression was measured using ImageJ software and represented as a relative density ratio (protein of interest to loading control). Values are expressed as mean ± S.D. N = 3–5, *p < 0.05.

Figure 7

JAK3 activity was inhibited in generated M2-macrophages from blood of HD or RA patients using two selective JAK inhibitors FM-381 and Tofacitinib. Cells were treated 1 h before LPS stimulation with 1 μM JAK inhibitor or DMSO control. Cytokine and chemokine secretion were measured by ELISA after stimulation with LPS for 24 h. Values are expressed as mean ± S.D. N = 3–8, *p < 0.05.

Figure 8

Effects of combined miR-155-5p/miR-221-3p mimics on pro- and anti-inflammatory cytokine secretion in TLR4-stimulated M2-macrophages. Cells were differentiated from CD14⁺ cells derived from HD blood. (A) M2-macrophages were transfected with mimics of miR-155-5p and miR-221-3p, either alone (155 or 221 mimic) or combined (155/221 mimic) and a control miR (ctrl mimic). M2-macrophages were then stimulated with 100 ng/ml LPS or left untreated. Secreted cytokines were measured after 24 h by ELISA and values are expressed as mean ± S.D. N = 15–18, *p < 0.05. (B) Anti-inflammatory activity of generated M2-macrophages transfected with the mentioned mimic combination setup was calculated after 24 h LPS-stimulation by the ratio of secreted cytokines IL-10 to IL-12 or IL-10 to IL-6. Values are expressed as mean ± S.D. N = 15–18, *p < 0.05.