CD24-Siglec axis is an innate immune checkpoint against metaflammation and metabolic disorder

Abstract

The molecular interactions that regulate chronic inflammation underlying metabolic disease remain largely unknown. Since the CD24-Siglec interaction regulates inflammatory response to danger-associated molecular patterns (DAMPs), we have generated multiple mouse strains with single or combined mutations of Cd24 or Siglec genes to explore the role of the CD24-Siglec interaction in metaflammation and metabolic disorder. Here, we report that the CD24-Siglec-E axis, but not other Siglecs, is a key suppressor of obesity-related metabolic dysfunction. Inactivation of the CD24-Siglec-E pathway exacerbates, while CD24Fc treatment alleviates, diet-induced metabolic disorders, including obesity, dyslipidemia, insulin resistance, and nonalcoholic steatohepatitis (NASH). Mechanistically, sialylation-dependent recognition of CD24 by Siglec-E induces SHP-1 recruitment and represses metaflammation to protect against metabolic syndrome. A first-in-human study of CD24Fc (NCT02650895) supports the significance of this pathway in human lipid metabolism and inflammation. These findings identify the CD24-Siglec-E axis as an innate immune checkpoint against metaflammation and metabolic disorder and suggest a promising therapeutic target for metabolic disease.

Keywords: CD24; NASH; Siglec-E; Siglecs; insulin resistance; metabolic syndrome; metaflammation; obesity; sialic acid-binding immunoglobulin-like lectins; sialylation.

Figure 1.. CD24 deficiency aggravates diet-induced metabolic disorder in mice.

(A-J) Cd24^−/− mice and WT littermates were fed a HFD for 12 weeks. (A) Body weight of mice in the indicated groups. n = 8 per group. (B) Representative photographs, inguinal and epididymal fat pads of mice in the indicated groups. (C) Body composition of mice was detected by DEXA. n = 8 per group. (D) TC, TG, FFA, LDL-C and HDL-C levels of mice in the indicated groups. n = 8 per group. (E) Fasting blood glucose levels of mice in the indicated groups. n = 8 per group. (F) GTT and ITT results of mice in the indicated groups. n = 8 per group. (G) Representative photographs of livers from mice in the indicated groups. (H) Liver weight and liver/body weight ratio of mice in the indicated groups. n = 8 per group. (I) Representative images of H&E staining of liver sections. Scale bar, 100 μm. Graph shows the quantitation of lipid droplet area. n = 8 per group. (J) Relative mRNA levels of key metabolic genes in the livers of mice in the indicated groups. n = 6 per group. (K-N) Cd24^−/− mice and littermate controls were fed a CD-HFD for 12 weeks. (K) Body weight, liver weight and liver/body weight ratio of mice in the indicated groups. n = 9 per group. (L) Serum levels of ALT and AST of mice in the indicated groups. n = 9 per group. (M) H&E and Masson’s trichrome staining of liver sections. Scale bar, 100 μm. Graph shows the quantitation of liver fibrosis area. n = 9 per group. (N) Relative mRNA levels of profibrotic genes in the livers of mice in the indicated groups. n = 6 per group. Data are mean ± SEM and representative of two or three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, unpaired Student’s t-test (C-E, H-N), two-way analysis of variance (ANOVA) (A, F). See also Figure S1.

Figure 2.. CD24Fc alleviates obesity-related metabolic dysfunction in mice.

(A-I) WT mice were fed a HFD and concurrently treated with CD24Fc or IgGFc control twice a week for 8 weeks. (A) Body weight of mice in the indicated groups. n = 8 per group. (B) Body composition of mice was detected by DEXA. n = 8 per group. (C) TC, TG, FFA, LDL-C and HDL-C levels of mice in the indicated groups. n = 6 per group. (D) Fasting blood glucose levels of mice in the indicated groups. n = 6 per group. (E) GTT and ITT results of mice in the indicated groups. n = 8 per group. (F) Insulin-stimulated phosphorylation of Akt in the livers of mice. Graph shows the quantitation of p-Akt relative to total Akt. n = 3 per group. (G) Liver weight and liver/body weight ratio of mice in the indicated groups. n = 6 per group. (H) Representative images of H&E staining of liver sections. Scale bar, 100 μm. Graph shows the quantitation of lipid droplet area. n = 6 per group. (I) Relative mRNA levels of key metabolic genes in the livers of mice in the indicated groups. n = 6 per group. (J-M) WT mice were fed a CD-HFD and simultaneously treated with CD24Fc or IgGFc control twice a week for 8 weeks.

(J) Body weight, liver weight and liver/body weight ratio of mice in the indicated groups. n = 8 per group. (K) Serum levels of ALT and AST of mice in the indicated groups. n = 8 per group. (L) H&E and Masson’s trichrome staining of liver sections. Scale bar, 100 μm. Graph shows the quantitation of liver fibrosis area. n = 8 per group. (M) Relative mRNA levels of profibrotic genes in the livers from mice in the indicated groups. n = 6 per group. Data are mean ± SEM and representative of two or three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, unpaired Student’s t-test (B-D, G-M), two-way analysis of variance (ANOVA) (A, E, F). See also Figure S2.

Figure 3.. Siglec-E is the CD24 receptor that protects against metabolic syndrome.

(A) Interactions between endogenous CD24 and recombinant Siglecs. n = 3 per group. (B) Direct interactions between CD24Fc and recombinant Siglecs. n = 3 per group.

(C and D) Siglec-deficient mice were maintained on a normal diet. Serum levels of total cholesterol (C) and fasting blood glucose (D) were detected at 8 months of age. n = 5–12 per group.

(E and F) WT and Siglec-deficient mice were fed a HFD for 12 weeks. Body weight (E) and body composition (F) were detected. n = 6–8 per group.

(G-M) Siglece^−/− mice and WT littermates were fed a HFD for 12 weeks. (G) TC, TG, FFA, LDL-C and HDL-C levels of mice in the indicated groups. n = 8 per group. (H) Fasting blood glucose levels of mice in the indicated groups. n = 8 per group. (I) GTT and ITT results of mice in the indicated groups. n = 6 per group. (J) Photographs of representative livers from mice in the indicated groups. (K) Liver weight and liver/body weight ratio of mice in the indicated groups. n = 8 per group. (L) Representative images of H&E staining of liver sections. Scale bar, 100 μm. Graph shows the quantitation of lipid droplet area. n = 8 per group. (M) Relative mRNA levels of key metabolic genes in the livers from mice in the indicated groups. n = 8 per group. Data are mean ± SEM and representative of two or three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, unpaired Student’s t-test (G, H, K-M), one-way analysis of variance (ANOVA) (A-D, F) or two-way analysis of variance (ANOVA) (E, I). See also Figures S3 and S4.

Figure 4.. Siglec-E signaling is required for CD24-mediated protection against metabolic disorder.

(A-F) Siglece^−/− mice and WT littermates were fed a HFD for 8 weeks, followed by injection of CD24Fc or IgGFc control twice a week for 4 weeks while continuing HFD. (A) TC, TG, LDL-C and HDL-C levels of mice in the indicated groups. n = 7–8 per group. (B) Fasting blood glucose levels of mice in the indicated groups. n = 7–8 per group. (C) GTT and ITT results of mice in the indicated groups. n = 6 per group. (D) Body weight of mice in the indicated groups. n = 7–8 per group. (E) Liver weight and liver/body weight ratio of mice in the indicated groups. n = 7–8 per group. (F) Representative images of H&E staining of liver sections. Scale bar, 100 μm. Graph shows the quantitation of lipid droplet area. n = 7–8 per group. Data are mean ± SEM and representative of two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, two-way analysis of variance (ANOVA).

Figure 5.. CD24-Siglec-E axis represses metaflammation to ameliorate metabolic disorder.

(A-D) WT, Cd24^−/− and Siglece^−/− mice were fed a HFD for 12 weeks. (A) Serum concentrations of inflammatory cytokines of mice in the indicated groups. n = 8 per group. (B) Representative images of H&E staining of liver and eWAT sections. Arrows indicate immune cells infiltrates. Scale bar, 50 μm. (C) Relative mRNA levels of inflammatory genes in liver and eWAT from mice in the indicated groups. n = 6 per group. (D) Immunoblotting analysis of NF-κB signaling pathway in the livers. β-actin was used as internal loading control. n = 3 per group. (E and F) Siglece^−/− mice and WT littermates were fed a HFD for 8 weeks, followed by injection of CD24Fc or IgGFc control twice a week for 4 weeks. (E) Serum levels of inflammatory cytokines of mice in the indicated groups of mice. n = 7–8 per group. (F) Relative mRNA levels of inflammatory genes in liver (left) and eWAT (right) from mice in the indicated groups. n = 6 per group. (G-L) Cd24^−/− mice and WT littermates were fed a HFD for 8 weeks, followed by injection of neutralizing antibodies (anti-TNFα, anti-IL-6 and anti-IL-1β monoclonal antibodies) twice a week for 4 weeks while continuing HFD. (G) TC, TG, LDL-C and HDL-C levels of mice in the indicated groups. n = 6–7 per group. (H) Fasting blood glucose levels of mice in the indicated groups. n = 6–7 per group. (I) GTT and ITT results of mice in the indicated groups. n = 5–6 per group. (J) Body weight of mice in the indicated groups. n = 6–7 per group. (K) Liver weight and liver/body weight ratio of mice in the indicated groups. n = 6–7 per group. (L) Representative images of H&E staining of liver sections. Scale bar, 100 μm. Graph shows the quantitation of lipid droplet area. n = 6–7 per group. Data are mean ± SEM and representative of two or three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, one-way analysis of variance (ANOVA) (A, C), two-way analysis of variance (ANOVA) (E-L). See also Figure S5.

Figure 6.. Sialoside-based recognition of CD24 by Siglec-E modulates metabolic inflammation.

(A-B) Peritoneal macrophages from WT, Cd24^−/− and Siglece^−/− mice were stimulated with palmitate or BSA control for 16 hours before collection. (A) TNF-α and IL-6 cytokines production in macrophages treated with palmitate or BSA control. n = 3 per group. (B) Immunoblotting analysis of NF-κB signaling pathway in macrophages treated with palmitate or BSA control. n = 3 per group. (C and D) Peritoneal macrophages from WT and Siglece^−/− mice were stimulated with palmitate or BSA control, concurrently treated with CD24Fc or control IgGFc for 16 hours before collection. (C) TNF-α and IL-6 cytokines production in macrophages with or without palmitate or CD24Fc treatment. n = 3 per group. (D) Immunoblotting analysis of NF-κB signaling in macrophages with or without palmitate or CD24Fc treatment. n = 3 per group. (E) Co-immunoprecipitation of CD24 and Siglec-E in WT and Cd24^−/− spleen cells. (F) Co-immunoprecipitation of CD24 and Siglec-E in spleen cells with or without NanA treatment. (G) The associations of CD24Fc or desialylated CD24Fc with Siglec-E in spleen cells. Protein A/G beads were used to pull down Fc. (H) Co-immunoprecipitation of Siglec-E and SHP-1 in WT and Cd24^−/− spleen cells. (I) Co-immunoprecipitation of Siglec-E and SHP-1 in WT and Cd24^−/− spleen cells treated with CD24Fc or IgGFc control. (J) Co-immunoprecipitation of Siglec-E and SHP-1 in WT and Cd24^−/− peritoneal macrophages treated with CD24Fc or IgGFc control. (K) Co-immunoprecipitation of CD24 and Siglec-E in spleen cells from WT mice fed a NCD or HFD. Sialylation level of CD24 was detected by SNA and MAL II lectin blotting. (L) Serum levels of free sialic acid in NCD or HFD-fed mice. n = 8 per group. (M) Sialylation levels on PBMCs from NCD or HFD-fed mice were detected by flow cytometry with MAL II lectins. n = 8 per group. Data are mean ± SEM and representative of two or three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, unpaired Student’s t-test (L, M), two-way analysis of variance (ANOVA) (A, C). See also Figures S6 and S7.

Figure 7.. CD24Fc regulates lipid metabolism and inflammation in humans.

(A) The human subjects were treated with a single dose of placebo or CD24Fc at different doses (10, 30, 60, 120 and 240 mg per injection). Plasma LDL-C levels were detected at pre-dosing baseline, and at 7, 14 and 42 days after dosing. Relative LDL-C levels were normalized to the baseline. LDL-C of Day 14 samples in 10 mg group were not measured. n = 5–10 per group. (B and C) RNA-sequencing was performed on PBMC samples obtained on Day −1 (pre-treatment) and Day 3 (post-treatment) from subjects receiving 240 mg of CD24Fc. All samples that pass quality control were used.

(B) Volcano plot analysis. Significantly differentially expressed genes (P < 0.05 and fold change > 1.5) were highlighted in red (up-regulated) and blue (down-regulated). Selected genes related to inflammation were indicated. (C) Heat map of alterations in expression of inflammatory genes. (D) Expression of inflammatory genes in PBMC samples was validated by real-time PCR. Gene expression levels were calculated after normalization to the housekeeping gene GAPDH. Data are mean ± SEM. *p < 0.05, **p < 0.01, one-way analysis of variance (ANOVA) for multiple comparisons (A), linear regression analysis for dose-dependency (A) or paired t-test (D). See also Figure S7.