FNDC4 acts as an anti-inflammatory factor on macrophages and improves colitis in mice

Abstract

FNDC4 is a secreted factor sharing high homology with the exercise-associated myokine irisin (FNDC5). Here we report that Fndc4 is robustly upregulated in several mouse models of inflammation as well as in human inflammatory conditions. Specifically, FNDC4 levels are increased locally at inflamed sites of the intestine of inflammatory bowel disease patients. Interestingly, administration of recombinant FNDC4 in the mouse model of induced colitis markedly reduces disease severity compared with mice injected with a control protein. Conversely, mice lacking Fndc4 develop more severe colitis. Analysis of binding of FNDC4 to different immune cell types reveals strong and specific binding to macrophages and monocytes. FNDC4 treatment of bone marrow-derived macrophages in vitro results in reduced phagocytosis, increased cell survival and reduced proinflammatory chemokine expression. Hence, treatment with FNDC4 results in a state of dampened macrophage activity, while enhancing their survival. Thus, we have characterized FNDC4 as a factor with direct therapeutic potential in inflammatory bowel disease and possibly other inflammatory diseases.

Figure 1. An N-terminal portion of FNDC4 is secreted.

(a) Graphical representation of the FNDC4 construct with N-terminal FLAG and C-terminal Myc tags as well as the peptide sequence. C, intracellular domain; FN III domain, Fibronectin type III domain; H, hydrophobic domain; SP, signal peptide. (b) Western blot against either FLAG or Myc, as indicated, in cell- and media samples from HEK293 cells transfected with the N-terminal FLAG and C-terminal Myc FNDC4 construct. The full-size blot image is presented in Supplementary Fig. 11a. Similar results were seen in at least three independent experiments. NT, non-transfected control.

Figure 2. Fndc4 gene expression is upregulated in inflammatory conditions.

(a) Fndc4 gene expression in the murine colon on induction of DSS-mediated colitis. DSS was administered via the drinking water, mice were sacrificed 7 days later, N=10 mice per group. NT, non-treated. Mean±s.e.m., *P<0.05 (Student's T-test). This experiment was performed three times with similar results. (b) Fndc4 gene expression in the kidney on induction of glomerulonephritis. Mice were injected with antiserum directed against the glomerular basement membrane (αGBM) and sacrificed 5 days post injection. NT, non-treated. N=5 mice per group. Mean±s.e.m., *P<0.05 (Student's T-test). This experiment was repeated once with similar results. (c–d) Human IBD cohort. FNDC4 gene expression in intestinal biopsies from inflamed versus non-inflamed regions of the same patients with ulcerative colitis (UC) (N=12) (c) or Crohn's disease (CD) (N=7) (d). *P<0.05, paired T-tests. (e) RNA in situ hybridization. Distal colon from DSS-treated mice. Fndc4 versus negative control (NC) probe. Arrows point at examples of positive signals (red).

Figure 3. TGFβ induces FNDC4 expression.

(a) FNDC4 expression in Caco-2 cells after 20 h treatment with 20 ng ml⁻¹ TGFβ1. N=4. Mean±s.e.m., *P<0.05, Student's T-test. (b) Luciferase assay. McA-RH7777 cells were transfected with an Fndc4 promoter reporter construct and treated with 20 ng ml⁻¹ TGFβ1 and/or TGFβ1R antagonist SB-431542 (SB) (20 μM) for 20 h. N=4–5. Mean±s.e.m., *P<0.05 (one-way analysis of variance with Tukey's post hoc tests).

Figure 4. FNDC4 improves colitis in mice.

DSS-induced colitis model. Mice were administered DSS from day 0 to 5 and were injected with hFc-FNDC4 or hFc control protein on days 0, 2 and 4. N=10 mice per group. This experiment was performed three times (of which the last experiment was blinded, blinding was performed by a colleague not involved in the study) with similar results. (a) DAI (disease activity index). **P<0.001, Mann–Whitney rank test. (b) Colon length. **P<0.001, Student's T-test. The non-treated (NT) control group consisted of 5 age- and gender-matched controls that did not receive DSS. (c) Body weight curve. *P<0.05, Student's T-test. (d–e) Gene expression, selected genes. Mean±s.e.m., *P<0.05 and q<0.10; two-sided Student's T-tests of hFc versus hFc-FNDC4, followed by the Benjamini-Hochberg false discovery rate (FDR) correction for multiple testing. (f) Histopathology. Colitis severity assessment and representative images. Scores are mean±s.e.m. *P<0.05, Mann–Whitney rank test.

Figure 5. FNDC4 acts on macrophages.

(a) Immune cell infiltration of DSS-mediated inflamed colon mostly consists of macrophages. Flow cytometry analysis of CD11b⁺-cells amongst CD45⁺ cells in the DSS colitis model. N=10 for the hFc control group, N=9 for the hFc-FNDC4 group. (b) FNDC4 binds to IgG Fc-gamma (IgG Fc-receptor negative) macrophages. The plots on the left show binding to peritoneal macrophages, the graph on the right shows binding to a panel of blood cells. Antibodies against CD11b and F4/80 were used to gate for macrophages. Binding of hFc-FNDC4 to cells was detected using a PE-conjugated antibody against human IgG. (c) ERK phosphorylation after 30 minutes of hFc-FNDC4 treatment. Total ERK served as loading control. This experiment was repeated twice with similar results. The full-size blot image is presented in Supplementary Fig. 11b. (d) Dose curve of hFc-FNDC4-mediated induction of Ass1 and Myo7a in bone marrow macrophages treated for 24 h with different concentrations of hFc-FNDC4 (100 nM hFc as control). Data are presented as mean±s.e.m., N=4. *P<0.05, One-way analysis of variance followed by Dunnett's post hoc tests.

Figure 6. FNDC4 treatment affects key macrophage processes and downregulates proinflammatory gene expression.

(a) Phagocytosis of fluorescently labelled zymosan A particles. Bone marrow macrophages were treated for 24 h with 100 nM hFc-FNDC4 or hFc control followed by incubation with zymosan A particles for 30 minutes. Mean±s.e.m., N=4. *P<0.05 (T-test). This experiment was performed twice with similar results. (b) Survival of bone marrow-derived macrophages after 48 h in starvation media, treated with 100 nM hFc-FNDC4 or hFc control. Mean±s.e.m., N=4. *P<0.05 (T-test). This effect was validated in two individual repeat experiments. (c) Gene expression of a selection of cytokines and chemokines on hFc-FNDC4 treatment (or hFc as carrier-control) (100 nM). N=6–7 (d) Gene expression of a selection of cytokines and chemokines on hFc-FNDC4 treatment (24 h, 100 nM) in the presence of TNFα. N=6–7. (e–f) Gene expression of M1 (e) and M2 (f) markers. N=6–7. (g) Gene expression of Tnf, Il1b and Nos2 after LPS-mediated polarization. Cells were treated with 10 ng ml⁻¹ LPS for 3 days before 24 h treatment with 100 nM hFc-FNDC4 or hFc control. N=4. (h) Gene expression of M2 markers Clec10a and Mrc1 after IL-4-mediated polarization. Cells were treated with 10 ng ml⁻¹ IL-4 for 3 days before 24 h treatment with 100 nM hFc-FNDC4 or hFc control. N=4. (i) Gene expression of the general leukocyte marker Ptprc (Cd45) and the general macrophage markers Emr1 (F4/80) and Itgam (Cd11b). N=3. Mean±s.e.m. NT, non-treated. *P<0.05 and q<0.10 for the comparison hFc-FNDC4 versus the corresponding hFc control group; one-way analysis of variance with Tukey's post hoc tests, followed by the Benjamini-Hochberg FDR correction. These genes were regulated to a similar extend in the gene expression array and similar effects on gene expression were seen in multiple (>3) independent validation experiments.

Figure 7. FNDC4 also acts on human macrophages.

Effects of hFc-FNDC4 treatment on proinflammatory gene expression in the basal (3–4 individual donors) (a) and LPS-stimulated state (2–3 individual donors) (b). Macrophages were cultured in the presence of macrophage colony-stimulating factor and treated with hFc-FNDC4 or hFc control for 24 h. Subsequently, RNA was isolated using trizol, reverse transcribed and subjected to RT-qPCR analysis. Mean±s.e.m., *P<0.05 (T-test).

Figure 8. FNDC4 knockout mice show increased inflammation and severity of DSS-induced colitis.

(a) Body weights of 12 week old mice, N=4. (b) Spleen weight, standardized to body weight at 16 weeks of age, N=4. (c) DAI (disease activity index). N=5 mice per group. *P<0.05, Mann–Whitney rank test. KO, Fndc4 knockout mice, WT, wild-type mice. (d) Colon length. **P<0.01, analysis of variance with Tukey's post hoc test. (e) Body weight curve. *P<0.05 and q<0.05, Student's T-test followed by the Benjamini-Hochberg false discovery rate correction. (f-g) Histopathology. Colitis severity assessment. N=11 mice per group, *P<0.05, Mann–Whitney rank test. (h) Gene expression, selected genes. *P<0.05; two-sided Student's T-tests, N=5 mice per group. Disease severity was assessed blinded and the experiment was repeated with similar results. Mean±s.e.m.

Figure 9. FNDC4 signals partly via STAT3.

(a) Socs3 gene expression after 4 h incubation with 100 nM hFc-FNDC4 or hFc control. Mean±s.e.m., N=3. *P<0.05 (T-test). (b) Phosphorylation of STAT3. Bone marrow-derived macrophages were starved overnight followed by 30 minutes treatment with 200 nM or 300 nM hFc-FNDC4, 200 nM hFc served as control. The full-size, uncropped blot image is presented in Supplementary Fig. 11c. Similar effects were observed in two repeated experiments. (c) STAT3 DNA binding to the Socs3 promoter region. Macrophages were treated with hFc-FNDC4 or hFc for 6 h. Data presented as percentage of input. Mean±s.e.m., N=4. *P<0.05 (T-test). (d) STAT3 inhibition reverses FNDC4-mediated gene regulation. Bone marrow-derived macrophages were pretreated with 50 μM S3I-201 STAT3 inhibitor or DMSO as control for 45 min and subsequently treated with 100 nM hFc-FNDC4 or hFc control for 6 h in the presence or absence of S3I-201. Mean±s.e.m., *P<0.05 and q<0.10, One-way analysis of variance followed by Tukey's post hoc tests and Benjamini-Hochberg false discovery rate-adjustment, N=4. (e) STAT3 inhibition reverses FNDC4-mediated improvements of macrophage survival. Bone marrow macrophages were pretreated with 50 μM S3I-201 STAT3 inhibitor or DMSO as control for 45 min and subsequently treated with 100 nM hFc-FNDC4 or hFc control for 48 h in starvation media in the presence or absence of S3I-201. NT, non-treated. Mean±s.e.m., *P<0.05, one-way analysis of variance followed by Tukey's post hoc tests, N=8 per condition. (f) Proteins showing a >1.3-fold increase or decrease in phosphorylation status in primary bone marrow macrophages on 30-min treatment with hFc-FNDC4 or hFc control. The fold change was calculated as the signal ratio of the paired antibodies for each pair of site-specific antibody and phospho site-specific antibody. Average ratios of two duplicates per condition. * Indicate that the fold induction in phosphorylation status is due to a reduction in binding of the total protein (conformational change affecting the accessibility of the epitopes; the array is performed using non-denatured protein).