Ifih1 gene dose effect reveals MDA5-mediated chronic type I IFN gene signature, viral resistance, and accelerated autoimmunity

Abstract

Type I IFNs (IFN-I) are normally produced during antiviral responses, yet high levels of chronic IFN-I expression correlate with autoimmune disease. A variety of viral sensors generate IFN-I in their response, but other than TLRs, it is not fully known which pathways are directly involved in the development of spontaneous immune pathologies. To further explore the link between IFN-I induced by viral pathways and autoimmunity, we generated a new transgenic mouse line containing multiple copies of Ifih1, a gene encoding the cytoplasmic dsRNA sensor MDA5 with proven linkage to diabetes and lupus. We show that MDA5 overexpression led to a chronic IFN-I state characterized by resistance to a lethal viral infection through rapid clearance of virus in the absence of a CD8(+) or Ab response. Spontaneous MDA5 activation was not sufficient to initiate autoimmune or inflammatory pathology by itself, even though every immune cell population had signs of IFN activation. When combined with the lupus-susceptible background of the FcγR2B deficiency, MDA5 overexpression did accelerate the production of switched autoantibodies, the incidence of glomerulonephritis, and early lethality. Thus, MDA5 transgenic mice provide evidence that chronic elevated levels of IFN-I are not sufficient to initiate autoimmunity or inflammation although they might exacerbate an ongoing autoimmune pathology.

Figure 1. MDA5 Transgenic mice display a chronic IFN-I signature with minor pathological consequences

(A) Ifih1 genomic and mRNA levels were measured by real-time PCR on ear DNA or splenic RNA from mice of the indicated genotype. β-actin was used as standard. (B) IFN-I-target gene expression was measured by real-time PCR on RNA from spleens (CD4-depleted or CD43-negative) from mice of the indicated genotype. TLR7 Tg mice contain between 8 and 16 copies of TLR7 genomic DNA (16). Each dot represents an individual mouse. Data are pooled from multiple experiments. For B, a statistical T-test was performed. *P<0.05, **P<0.01.

Figure 2. Evidence of cellular IFN-I activation, yet normal lymphocyte homeostasis in MDA5 Tgs

(A) Ly6C and Ly6A/E surface expression was evaluated by flow cytometry on CD8⁺, B220⁺ CD4⁺ and CD11c⁺ cells from WT and MDA5 Tg spleens or bone marrow. (B) PDCA-1 surface expression levels were evaluated on B cells and cDCs in the spleen as well as cDCs and monocytes in the bone marrow of WT and MDA5 Tg mice. (C–E) Splenocytes from MDA5 Tg and WT littermate controls were gated by flow cytometry using the indicated markers. Each dot represents absolute numbers of cells from an individual mouse. Mice were between 2 and 6 months of age with little age-dependent differences observed. A statistical T-test was performed comparing WT to MDA5 Tg. **P<0.01. (F) Representative dot plots of Ly6C staining on monocytes with a gate showing the frequency of CD11b⁺Ly6C^Hi cells. In A, B and F, representative cytometric plots from at least three mice in each group are shown.

Figure 3. Cellular phenotype in MDA5 Tgs is dependent on type I IFN and is initiated in radio-resistant cells

(A) mRNA levels of the gene indicated were measured by real time PCR. Values are plotted as fold average WT. n=3–5 mice in each group. (B) Surface expression of Ly6C and Ly6A/E measured by flow cytometry on splenic cells gated as indicated below each graph. N=3–5 mice in each group. (C) T cell-depleted congenically marked bone marrow (WT CD45.1⁺) was transferred into lethally irradiated WT or MDA5 Tg CD45.2⁺ recipients (see schematic). Spleen cells from reconstituted mice were analyzed three months post-transfer, gated CD45.1 positive (donor-derived cells) and stained with the indicated markers for flow cytometry. Absolute numbers of cells per gate for individual mice are displayed. A statistical T-test was performed. Bone marrow transfer experiment was conducted twice with similar results. One experiment is depicted. *P<0.05, **P<0.01, ***P<0.001.

Figure 4. MDA5 Tg mice are resistant to lethal VSV infection

(A) WT or MDA5 Tgs mice were infected with 2×10⁶ PFU VSV and at 32 and 72 hrs post-infection, spleen and liver RNA was isolated. VSV L-gene mRNA was detected using real-time PCR. Results are expressed as values from a standard curve generated from in vitro infected vero cells (see methods). Each dot represents a single mouse. Data are pooled from 2 independent experiments. (B) WT or MDA5 Tgs were challenged with 200×10⁶ PFU VSV i.v. (the LD50) and monitored for survival. (C) WT or MDA5 Tg mice were infected i.v. with 2×10⁶ PFU VSV and splenic antigen-specific CD8 T cells were measured 8 days later using fluorochrome-labeled H-2K^b/peptide pentamers. Each dot represents absolute number of cells from the indicated gate. Three independent experiments were performed with similar results. A representative experiment is shown. (D) WT or MDA5 Tg mice were infected with VSV, sera were collected 16 and 69 days later and tested for neutralizing activity to VSV in vitro. Neutralizing antibody titer is determined by the dilution required to inhibit 50% of plaque colonies. One of two similar experiments shown. n=3 mice in each group. For A and D, a statistical T-test was performed. For B, Log-rank (Mantel-Cox) and Gehan-Breslow-Wilcoxon tests were performed with similar P values. *P<0.05, **P<0.01.

Figure 5. The MDA5 Transgene enhances B cell responses in FcγR2B^−/− mice

(A) 3–6 month old WT, MDA5 Tg, FcγR2B^−/− and MDA5 Tg FcγR2B^−/− mice were analyzed for the presence of class-switched (B220⁺IgG2a⁺), GC (B220⁺FAS⁺GL7⁺) and plasma blast (B220^intCD138⁺) B cells using flow cytometry. Additionally, data points for WT and MDA5 Tg animals from Figure 2 are plotted here for comparison. Absolute numbers for individual mice are displayed. (B) Mean fluorescence intensity of ICOS and numbers for CD4⁺CD44^HiCXCR5⁺ T follicular helper cells were measured on splenocytes purified from mice of the indicated genotype. Each dot represents an individual mouse. (C) Total IgG, IgG1 and IgG2a was measured in the sera of 2–4 month old WT, MDA5 Tg, FcγR2B^−/− and MDA5 Tg FcγR2B^−/− mice by ELISA. Each dot represents a single mouse. *P<0.05, **P<0.01, ***P<0.001.

Figure 6. The MDA5 Transgene accelerates autoimmunity in FcγR2B^−/− mice

(A) Cumulative survival was monitored for FcγR2B^+/−, MDA5 Tg FcγR2B^+/−, FcγR2B^−/− and MDA5 Tg FcγR2B^−/− mice for 10 months (left panel). The level of proteinuria was measured in the urine using a dip stick (Right Panel). Data is represented as the percentage of mice that are proteinuria free throughout the 10-month period. (B) Spleen weights from 2–4 month FcγR2B^−/− and MDA5 Tg FcγR2B^−/− mice were quantified. Spleen weights for WT and MDA5 Tgs from Figure 1D are plotted for comparison. (C) Total IgG + IgM anti-nuclear antibody levels were quantified using an ANA ELISA. (D) Kidneys from 2.5–4 month old mice of the indicated genotype were stained with H&E and kidney sections were analyzed for glomerulonephritis. Glomeruli score was determined by measuring individual glomeruli from multiple fields of view. WT n=3 mice, FcγR2B^−/− n=6 mice and MDA5 Tg FcγR2B^−/− n=8 mice. For A–D each dot represents a single mouse. Data are pooled from multiple experiments. For A, Log-rank (Mantel-Cox) and Gehan-Breslow-Wilcoxon tests were performed with similar P values. *P<0.05, **P<0.01, ***P<0.001.