Antagonistic Interplay between MicroRNA-155 and IL-10 during Lyme Carditis and Arthritis

Abstract

MicroRNA-155 has been shown to play a role in immune activation and inflammation, and is suppressed by IL-10, an important anti-inflammatory cytokine. The established involvement of IL-10 in the murine model of Borrelia burgdorferi-induced Lyme arthritis and carditis allowed us to assess the interplay between IL-10 and miR-155 in vivo. As reported previously, Mir155 was highly upregulated in joints from infected severely arthritic B6 Il10-/- mice, but not in mildly arthritic B6 mice. In infected hearts, Mir155 was upregulated in both strains, suggesting a role of miR-155 in Lyme carditis. Using B. burgdorferi-infected B6, Mir155-/-, Il10-/-, and Mir155-/- Il10-/- double-knockout (DKO) mice, we found that anti-inflammatory IL-10 and pro-inflammatory miR-155 have opposite and somewhat compensatory effects on myeloid cell activity, cytokine production, and antibody response. Both IL-10 and miR-155 were required for suppression of Lyme carditis. Infected Mir155-/- mice developed moderate/severe carditis, had higher B. burgdorferi numbers, and had reduced Th1 cytokine expression in hearts. In contrast, while Il10-/- and DKO mice also developed severe carditis, hearts had reduced bacterial numbers and elevated Th1 and innate cytokine expression. Surprisingly, miR-155 had little effect on Lyme arthritis. These results show that antagonistic interplay between IL-10 and miR-155 is required to balance host defense and immune activation in vivo, and this balance is particularly important for suppression of Lyme carditis. These results also highlight tissue-specific differences in Lyme arthritis and carditis pathogenesis, and reveal the importance of IL-10-mediated regulation of miR-155 in maintaining healthy immunity.

Fig 1. Activation of macrophages is modulated by IL-10/miR-155 regulatory network.

Bone marrow-derived macrophages (BMDMs) and peritoneal macrophages were isolated from B6, Mir155^-/-, Il10^-/-, and DKO mice, and cells were stimulated with media alone or with B. burgdorferi. A, Expression of Mir155 in BMDMs stimulated with B. burgdorferi for 24 hours by quantitative real-time PCR (qRT-PCR). B, Peritoneal macrophages were incubated with GFP-expressing B. burgdorferi for 1 or 2 hours, and cells containing intracellular bacteria (GFP-positive) were determined by flow cytometric analysis. C, Cxcl9 and Cxcl10 expression in BMDMs stimulated with B. burgdorferi for 24 hours, by qRT-PCR. D, Cytokine levels in culture supernatant from BMDMs stimulated by B. burgdorferi for 24 hours, measured by ELISA. Statistically significant differences between strains were determined by ANOVA (Tukey’s post-hoc), and are indicated (* p<0.05), with adjusted p-values included for trends not achieving statistical significance (p<0.1). Results are from one experiment (n = 3 triplicates per mouse strain).

Fig 2. Description of Lyme carditis pathology.

A, subacute vasculitis/perivasculitis induced by B. burgdorferi occurs in 5 distinct locations (pancarditis) in the sub-gross image (left-side of figure) of the heart (Box 1: atrium/auricle; Box 2: muscular artery; Box 3: vascularized connective tissue at the base of the heart; Box 4: epicardium and myocardium of the right ventricle; Box 5: myocardium of the left ventricle) (rv = right ventricle; lv = left ventricle). The interaction with, and injury caused by, the spirochete elicits a subacute inflammatory response that consists predominately of an admixture of neutrophils, lymphocytes, and monocytes/macrophages (i.e., histiocytes), as well as much smaller numbers of other mononuclear inflammatory cells (potentially plasma cells and/or dendritic cells). A1, pectinate muscle (p) and reactive endothelium (arrows) of atria/auricles contain subacute inflammatory cells (from Mir155^-/-, Il10^-/-, and DKO mouse strains); A2, similar area from a control animal (endothelium [arrow]); B1, vasa vasorum (arrows) of muscular arteries (a) contains similar inflammatory cells (l = lumen of muscular artery) (from Mir155^-/-, Il10 ^-/-, and DKO mouse strains). B1 inset, subintimal areas (arrows) of muscular arteries contain similar inflammatory cells. B2, similar area of a muscular artery (a) from a control animal (vasa vasorum [arrows]) (l = lumen of muscular artery); C1, stromal tissues at the base of the heart where blood vessels exit and enter the ventricles and atria/auricles, respectively (from Mir155^-/-, Il10 ^-/-, and DKO mouse strains) contain inflammatory cells (arrows). C2, similar area from a control animal; D1, loose connective tissue of the epicardium contains similar inflammatory cells (arrows) (from Mir155^-/-, Il10 ^-/-, and DKO mouse strains). The cardiac myocytes (myocardium [m]) are reactive and enlarged. D2, similar area of the epicardium (arrows) from a control animal (myocardium [m]); and E1, myocardium (m) of the ventricles contains similar inflammatory cells (arrows) (from Mir155^-/-, Il10 ^-/-, and DKO mouse strains). The cardiac myocytes are reactive and enlarged. E2, similar area of the ventricular myocardium (m) from a control animal. H&E stains, images representative of two infection experiments, one at 2 and one at 3 weeks post-infection (n = 5 mice per strain for each infection).

Fig 3. Lyme carditis severity is modulated by both miR-155 and IL-10.

Comparison of the severity of subacute vasculitis/perivasculitis affecting stromal tissues at the base of the heart where blood vessels exit and enter the ventricles and atria/auricles, respectively between mouse strains at 2 weeks following infection with B. burgdorferi. The subacute inflammatory response consists predominately of an admixture of PMNs, lymphocytes, and monocytes/macrophages as well as much smaller numbers of other mononuclear inflammatory cells (potentially plasma cells and/or dendritic cells). H&E stains, images representative of two infection experiments, one at 2 and one at 3 weeks post-infection (n = 5 mice per strain for each infection).

Fig 4. Host defense and B cell responses are modulated by the IL-10/miR-155 regulatory network.

B6, Mir155^-/-, Il10^-/- and DKO mice were infected with B. burgdorferi and host defense was assessed. A, joints were collected from mice infected for 4 weeks, and ankle skin and hearts were collected from mice infected for 3 weeks, and bacterial numbers were determined by measuring expression of B. burgdorferi 16S rRNA, compared to β-actin. B, B. burgdorferi-specific IgG (total), IgG1, IgG2c, and IgG3 were measured by ELISA in serum from mice infected for 4 weeks. Statistically significant differences between strains were determined by ANOVA (Tukey’s post-hoc), and are indicated (* p<0.05), with adjusted p-values included for trends not achieving statistical significance (p<0.1). Representative of 2 independent experiments at 4 weeks post-infection, and once at 3 weeks post-infection (n≥ 5 mice per strain for each experiment).

Fig 5. IL-12 and IFNγ levels in serum of mice infected for 4 weeks.

Cytokines were measured by ELISA in serum collected from B6, Mir155^-/-, Il10^-/- and DKO mice infected with B. burgdorferi for 4 weeks. Statistically significant differences between cytokine levels before and after infection for each strain were determined by Student t-test, and between infected strains by ANOVA (Tukey’s post-hoc), and are indicated (* p<0.05, n≥ 8 mice per strain).

Fig 6. Proposed model: antagonistic interplay between IL-10 and miR-155 modulates immune response during infection with B. burgdorferi.

Host immune cells, such as leukocytes, secrete various innate and/or adaptive cytokines at the site of B. burgdorferi infection. IL-10 is also secreted, which acts to limit inflammation by suppressing pro-inflammatory cytokine activity. One mechanism of IL-10 immune modulation involves STAT3-dependent suppression of expression of miR-155, which is upregulated during infection through NF-κB activation, and which enhances cytokine signaling (such as JAK-STAT signaling and NF-κB signaling) and immune activation (IL-10/miR-155 regulatory circuit in bold).