Disruption of hepatocyte Sialylation drives a T cell-dependent pro-inflammatory immune tone

Abstract

Through the catalysis of α2,6-linked sialylation, the enzyme ST6Gal1 is thought to play key roles in immune cell communication and homeostasis. Of particular importance, glycans with terminal α2,6-sialic acids are known to negatively regulate B cell receptor signaling and are associated with an immunosuppressive tumor microenvironment that promotes T cell anergy, suggesting that α2,6-sialic acids are a key immune inhibitory signal. Consistent with this model, mice harboring a hepatocyte-specific ablation of ST6Gal1 (H-cKO) develop a progressive and severe non-alcoholic fatty liver disease characterized by steatohepatitis. Using this H-cKO mouse, we have further discovered that loss of hepatocyte α2,6-sialylation not only increases the inflammatory state of the local tissue microenvironment, but also systemic T cell-dependent immune responses. H-cKO mice responded normally to innate and passively induced inflammation, but showed significantly increased morbidity in T cell-dependent house dust mite-antigen (HDM)-induced asthma and myelin oligodendrocyte glycoprotein (MOG) peptide-induced experimental autoimmune encephalomyelitis (EAE). We further discovered that H-cKO mice have a profound shift toward effector/memory T cells even among unchallenged mice, and that macrophages from both the liver and spleen expressed the inhibitory and α2,6-sialic acid-specific glycan binding molecule CD22. These findings align with previously reported pro-inflammatory changes in liver macrophages, and support a model in which the liver microenvironment sets a systemic immune tone that is regulated by tissue α2,6-sialylation and mediated by liver macrophages and systemic T cells.

Keywords: Asthma; EAE; Glycobiology; IgG; Inflammation; Liver; Macrophage; ST6Gal1; Sialic acid; Sialylation; T cell.

Figure 1.. H-cKO mice show no difference upon intranasal LPS challenge

A) Mouse lungs were challenged with LPS via i.n. delivery and harvested 24 hours (n=4-8 per group). B) Cellular differentials on a Hematrue system revealed no difference between wild type ST6^f/f and H-cKO mice. C) Lungs were fixed and stained with H&E, again revealing no difference in the extent of inflammation.

Figure 2.. H-cKO mice show no difference in thioglycollate-induced peritonitis

A) Thioglycollate was injected i.p. in ST6^f/f and H-cKO mice, and peritoneal lavage was performed at 24 hours (n=4-7 per group). B) Lavage harvested cell differentials were collected, showing no difference between wild type and H-cKO mice. C) Cell differentials from blood were also collected (n=2-5), and again no differences were seen between strains.

Figure 3.. H-cKO mice show no difference in DSS-induced colitis

A) Mice were provided 2.4 % DSS in their drinking water for 7 days, then moved to normal water thereafter. Weight was measured daily until harvest at day 19 (n=3-5 per group). B-C) No difference was seen in overall weight loss between wild type ST6^f/f and H-cKO mice, as measured by change in starting weight (B) or as normalized to the strain-specific negative controls (C).

Figure 4.. H-cKO mice show no difference in disease burden in CAI arthritis

A) Mice were injected i.p. with collagen-specific monoclonal cocktail on day 0 and LPS on day 3 to generate arthritis. Disease score and weight were monitored daily until harvest at day 20 (n=6 per group). B) No difference was seen in disease score, which reflects digit, palm, and ankle swelling, between wild type ST6^f/f and H-cKO mice. C-D) Likewise, no difference in average weight (C) or percent weight loss (D) was observed.

Figure 5.. H-cKO mice exhibit increased HDM-induced asthma tissue pathology

A) Mice were i.n. dosed with HDM a total of 10 times over 14 days as indicated, with harvest including bronchoalveolar lavage and tissue pathology (n=5-8 per group). B) Lavage harvested cell differentials failed to show a significant difference between wild type ST6^f/f and H-cKO mice, as measured by total while blood cells (WBC), neutrophils, macrophages, lymphocytes, and eosinophils. Basophils were reduced in H-cKO, but the total number of cells was considered biologically insignificant. C) H&E-stained tissue sections were scored in a blind fashion based on the degree of tissue leukocyte infiltration, epithelial hyperplasia, and dissemination of inflammatory foci throughout the tissue. H-cKO were significantly increased compared to wild type ST6^f/f mice (p<0.05). D) Representative H&E tissue sections at 4x and 10x.

Figure 6.. H-cKO mice exhibit increased EAE disease burden

A) Mice were injected with MOG emulsified in CFA on day 0. PTX was administered i.p. 1 hour and 24 hours after MOG. Disease score and weight were monitored daily until harvest on day 22 (n=4 per group). B) H-cKO mice developed more severe disease, as measured by evaluating tail and limb paralysis (*p<0.05). C) H-cKO mice also displayed greater weight loss as a result of disease (*p<0.05). D) Luxol Fast Blue staining of myelin in fixed spinal cords revealed significantly more myelin loss (arrows) in H-cKO mice compared to wild type ST6^f/f controls. E) H&E staining of fixed spinal cords further revealed increased leukocyte infiltration.

Figure 7.. Unchallenged H-cKO mice show profoundly increased antigen-experienced T cells

A) Spleen cells were isolated from unchallenged wild type ST6^f/f and H-cKO mice and analyzed by flow cytometry (n=9-12 per strain; *p<0.05). B-C) H-cKO mice showed a 4-fold decrease in the number of CD62L⁺CD44⁻ naïve CD4⁺ (B) and CD8⁺ (C) T cells. D-E) H-cKO mice also showed a 6- and 15-fold increase in the number of CD62L⁻CD44⁺ effector/memory antigen experienced CD4⁺ (D) and CD8⁺ (E) T cells. F) The number of activated CD25⁺CD4⁺ T cells was significantly increased in H-cKO mice. G) The number of F4/80⁺ macrophages among all single cells was unchanged.

Figure 8.. Non-B cell leukocytes express CD22

Spleen and liver cells were isolated from unchallenged wild type ST6^f/f and H-cKO mice and analyzed by flow cytometry (n=3 per strain). A-B) Approximately 50 % of all CD45⁺CD19⁻ leukocytes in both the liver (A) and spleen (B) expressed CD22. C) As a control, CD19⁺ B cells were found to be nearly 100% CD22⁺. D-E) Among liver (D) and splenic (E) CD45⁺CD11b⁺Ly6G⁻ macrophages, approximately 5 % and 20 %, respectively, expressed CD22. In all cases, no difference was observed between wild type ST6^f/f and H-cKO mice.