GPR183 antagonism reduces macrophage infiltration in influenza and SARS-CoV-2 infection

Abstract

Rationale: Severe viral respiratory infections are often characterised by extensive myeloid cell infiltration and activation and persistent lung tissue injury. However, the immunological mechanisms driving excessive inflammation in the lung remain poorly understood.

Objectives: To identify the mechanisms that drive immune cell recruitment in the lung during viral respiratory infections and identify novel drug targets to reduce inflammation and disease severity.

Methods: Preclinical murine models of influenza A virus and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection.

Results: Oxidised cholesterols and the oxysterol-sensing receptor GPR183 were identified as drivers of monocyte/macrophage infiltration to the lung during influenza A virus (IAV) and SARS-CoV-2 infection. Both IAV and SARS-CoV-2 infection upregulated the enzymes cholesterol 25-hydroxylase (CH25H) and cytochrome P450 family 7 subfamily member B1 (CYP7B1) in the lung, resulting in local production of the oxidised cholesterols 25-hydroxycholesterol (25-OHC) and 7α,25-dihydroxycholesterol (7α,25-OHC). Loss-of-function mutation of Gpr183 or treatment with a GPR183 antagonist reduced macrophage infiltration and inflammatory cytokine production in the lungs of IAV- or SARS-CoV-2-infected mice. The GPR183 antagonist significantly attenuated the severity of SARS-CoV-2 infection and viral loads. Analysis of single-cell RNA-sequencing data on bronchoalveolar lavage samples from healthy controls and COVID-19 patients with moderate and severe disease revealed that CH25H, CYP7B1 and GPR183 are significantly upregulated in macrophages during COVID-19.

Conclusion: This study demonstrates that oxysterols drive inflammation in the lung via GPR183 and provides the first preclinical evidence for the therapeutic benefit of targeting GPR183 during severe viral respiratory infections.

FIGURE 1

Influenza A virus (IAV) infection leads to upregulation of cholesterol 25-hydroxylase (CH25H) and cytochrome P450 family 7 subfamily member B1 (CYP7B1) expression in the lung and production of the oxysterols 25-hydroxycholesterol (25-OHC) and 7α,25-dihydroxycholesterol (7α,25-OHC). a) The biosynthetic pathway of 25-OHC and 7α,25-OHC. b) Experimental design: C57BL/6J mice were infected intranasally (i.n.) with 5500 plaque-forming units (PFU) of IAV. c) mRNA expression of Ch25h, Cyp7b1 and Hsd3b7 was measured by quantitative reverse transcription PCR at 3 days post infection (dpi) (D3) and 7 dpi (D7) normalised to Hprt. d) Quantitative analysis of CH25H, CYP7B1 and HSD3B7 protein labelling by immunohistochemistry (IHC). e) Representative IHC images of CH25H, CYP7B1 and HSD3B7 in lung sections of uninfected or IAV-infected mice. Scale bars: 100 µm (main) and 50 µm (inset). f) Concentrations of 25-OHC and 7α,25-OHC in the lungs (left) and bronchoalveolar lavage fluid (BALF) (right) at 3 dpi and 7 dpi. Data are presented as mean±sd of n=4 uninfected and n=6–10 infected mice per time point. ns: nonsignificant; U/I: mock infected. *: p<0.05; **: p<0.01.

FIGURE 2

Deletion of the Gpr183 gene or administration of a GPR183 antagonist reduces macrophage infiltration in influenza A virus (IAV)-infected lungs. C57BL/6J and Gpr183^−/− mice were infected intranasally with 5500 plaque-forming units (PFU) of IAV. a) Representative immunohistochemistry (IHC) images of ionised calcium-binding adapter molecule 1 (IBA1) in lung sections of IAV-infected C57BL/6J and Gpr183^−/− mice (left) and quantitative analysis (right). b) Experimental design: C57BL/6J mice and Gpr183^−/− mice were infected intranasally (i.n.) with 5500 PFU of IAV. Mice were subsequently treated orally with 7.6 mg·kg⁻¹ NIBR189 or vehicle control twice daily from 1 day post infection (dpi) until the end of the experiment. c) Representative IHC images of IBA1 in lung sections of C57BL/6J and Gpr183^−/− mice with the respective treatment groups at 3 dpi (D3) and 7 dpi (D7) (left) and quantitative analysis of IBA1 staining (right). Data are presented as mean±sd of n=6–12 infected mice per genotype and time point. Scale bars: 100 μm. a.u.: arbitrary units; U/I: mock infected; ns: nonsignificant. *: p<0.05; **: p<0.01; ***: p<0.001; ****: p<0.0001.

FIGURE 3

The GPR183 antagonist NIBR189 reduces macrophage infiltration and inflammatory cytokine production. C57BL/6J and Gpr183^−/− mice were infected intranasally with 5500 plaque-forming units (PFU) of influenza A virus (IAV). Mice were subsequently treated orally with 7.6 mg·kg⁻¹ NIBR189 or vehicle control twice daily from 1 day post infection (dpi) until the end of the experiment. a) Frequency of infiltrating macrophages (F480^high/CD11b⁺/Ly6G⁻/SigF⁻) and neutrophils (B220⁻CD3⁻Ly6G⁺) was determined by flow cytometry relative to total viable CD45⁺ immune cells at 3 dpi (left). Graphs show the frequency of macrophages and neutrophils (right). b) Cytokine measurements of interleukin 6 (IL-6), tumour necrosis factor (TNF), interferon β (IFNβ) and interferon λ (IFNλ) at 7 dpi measured by ELISA. Data are presented as mean±sd of n=5–12 infected mice per genotype and time point. U/I: mock infected; ns: nonsignificant. *: p<0.05; **: p<0.01.

FIGURE 4

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection leads to upregulation of cholesterol 25-hydroxylase (CH25H) and cytochrome P450 family 7 subfamily member B1 (CYP7B1) expression in the lung and production of the oxysterols 25-hydroxycholesterol (25-OHC) and 7α,25-dihydroxycholesterol (7α,25-OHC). C57BL/6J mice were infected intranasally with approximately 8×10⁴ plaque-forming units (PFU) of mouse49-adapted SARS-CoV-2. mRNA expression of a) Ch25h, Cyp7b1 and Hsd3b7 was measured by quantitative reverse transcription PCR at 2 days post infection (dpi) (D2) and 5 dpi (D5) normalised to Hprt. b) Quantitative analysis of CH25H, CYP7B1 and HSD3B7 protein by immunohistochemistry (IHC) labelling at D2 and D5 and c) representative IHC images of CH25H, CYP7B1 and HSD3B7 in lung sections in uninfected mice at D2. d) Concentrations of 25-OHC and 7α,25-OHC in the lungs (top) and bronchoalveolar lavage fluid (BALF) (bottom) at D2 and D5. Data are presented as mean±sd of n=3 uninfected mice and n=9–10 infected mice per time point. Scale bars: 50 μm. U/I: mock infected; a.u.: arbitrary units; ns: nonsignificant. *: p<0.05; **: p<0.01; ***: p<0.001; ****: p<0.0001.

FIGURE 5

GPR183 antagonism resulted in less severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection-induced weight loss and in reduced macrophage infiltration. C57BL/6J and Gpr183^−/− mice were infected intranasally with approximately 8×10⁴ plaque-forming units (PFU) of mouse-adapted SARS-63 CoV-2. Mice were subsequently treated orally with 7.6 mg·kg⁻¹ NIBR189 or vehicle control twice daily from 1 day post infection (dpi) until the end of the experiment. a) Experimental design. b) Weights of mice displayed as percentage of the weight at time of inoculation. c) Representative immunohistochemistry (IHC) images of ionised calcium-binding adapter molecule 1 (IBA1) in lung of C57BL/6J and Gpr183^−/− mice with the respective treatment groups at 2 dpi (D2) and 5 dpi (D5). Scale bars: 100 μm. d) Quantitative analysis of IBA1 at D2 and D5. Data are presented as mean±sd of n=9–12 infected mice per genotype and time point. a.u.: arbitrary units; U/I: uninfected; ns: nonsignificant. *: p<0.05; **: p<0.01; ***: p<0.001.

FIGURE 6

GPR183 antagonism led to a reduced inflammatory cytokine profile. C57BL/6J and Gpr183^−/− mice were infected intranasally with approximately 8×10⁴ plaque-forming units (PFU) of mouse-adapted severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Mice were subsequently treated orally with 7.6 mg·kg⁻¹ NIBR189 or vehicle control twice daily from 1 day post infection (dpi) until the end of the experiment. Relative expression of Tnf, Ifng, Ifnb and Ifnl at a) 2 dpi (D2) and b) 5 dpi (D5) in the lungs measured by reverse transcription quantitative PCR, normalised to Hprt. Data are presented as mean±sd of n=3 uninfected mice and n=9–12 infected mice per genotype and time point. U/I: mock infected; ns: nonsignificant. *: p<0.05; **: p<0.01; ***; p<0.001.

FIGURE 7

Mice treated with GPR183 antagonist had lower severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) loads. C57BL/6J and Gpr183^−/− mice were infected intranasally with approximately 8×10⁴ plaque-forming units (PFU) of mouse-adapted SARS-CoV-2. Mice were subsequently treated orally with 7.6 mg·kg⁻¹ NIBR189 or vehicle control twice daily from 1 day post infection (dpi) until the end of the experiment. a) Representative immunohistochemistry (IHC) images of viral nucleocapsid (N protein) expression at 2 dpi (D2) and 5 dpi (D5). b) Quantitative analysis of viral N protein expression of the treatment groups at D2. c) Viral load was assessed in the lung through the detection of Mpro RNA by reverse transcription quantitative PCR at D5, normalised to Hprt. Data are presented as mean±sd of n=9–12 infected mice per genotype and time point. Scale bars: 50 μm. a.u.: arbitrary units; U/I: mock infected; ns: nonsignificant. *: p<0.05; **: p<0.01.

FIGURE 8

Single-cell RNA-sequencing expression analysis of cells collected by bronchoalveolar lavage from healthy controls and COVID-19 patients. Summary heatmaps in the left panels show average normalised expression level of genes a) Ch25h, b) Cyp7b1, c) Hsd3b7 and d) Gpr183 per individual per cell type cluster. Summary heatmaps in the right panels show the average log fold change (logFC) in expression of each gene in 24 cell type clusters between moderate COVID-19 cases and healthy controls (M vs H), severe COVID-19 cases and healthy controls (S vs H), COVID-19 cases and healthy controls (M+S vs H), and severe and moderate COVID-19 cases (S vs M). The logFC values were estimated using negative binomial generalised linear models applied to raw UMI counts, adjusting for total UMI counts per cell, number of genes detected per cell and per cent mitochondrial counts per cell (NB regression); or non-parametric Wilcoxon rank sum test applied to normalised counts. Significant associations are highlighted with a single asterisk if they surpass Bonferroni significance (p<1.30×10⁻⁴) or a double asterisk if they were further expressed in at least 5% of cells in both groups with an absolute value of logFC >0.25. A logFC >0 suggests the expression level of the gene is higher among the focal group (e.g. moderate COVID-19 cases) compared to the other group (e.g. healthy controls), or vice versa. mDC: myeloid dendritic cell; pDC: plasmacytoid dendritic cell; NK: natural killer.

FIGURE 9

Graphical abstract of the role of GPR183 in the immune response to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and influenza A virus (IAV) infection. SARS-CoV-2 and IAV infection lead to the upregulation of cholesterol 25-hydroxylase (CH25H) and cytochrome P450 family 7 subfamily member B1 (CYP7B1), which results in the production of 7α,25-dihydroxycholesterol (7α,25-OHC). This oxysterol chemotactically attracts GPR183-expressing macrophages to the lungs where they produce pro-inflammatory cytokines. Pharmacological inhibition of GPR183 attenuates the infiltration of GPR183-expressing macrophages, leading to reduced production of inflammatory cytokines without negatively affecting antiviral responses.