Fig 1.
Liver specific 15-PGDH expression protects mice against LPS/GalN-induced acute liver inflammation and tissue damage.
Wild type and 15-PGDH Tg mice were intraperitoneally injected with LPS (60ng/g body weight) plus GalN (800μg/g body weight). The mice were followed for survival or sacrificed at 5h after LPS/GalN injection to collect serum and liver tissue samples.(A) Survival curve of mice after LPS/GalN administration (n = 10). (B) Serum ALT and AST levels (n = 5). (C) Representative images of liver tissue sections (200×): H&E stain (upper panel), Caspase-3 IHC (mid panel), F4/80 IHC (lower panel). Quantified IHC results was shown below (n = 5). (D) Caspase-3, 8, and 9 activities in liver tissue homogenates (n = 3). The data were shown as fold changes compared to wild type mice. (E) The protein levels of apoptotic signaling molecules (Cleaved PARP, JNK, P-JNK) in liver tissue homogenates. GAPDH was used as loading control. (F) mRNA levels of pro-inflammatory cytokines (IL1β, IL6, TNF-α, MCP1 and CXCL2) in liver tissue homogenates (n = 3); the results are shown as fold changes compared to wild type mice. The quantitative data presented in this figure are expressed as means ± SE (*p<0.05, **p<0.01, ***p<0.001).
Fig 2.
15-PGDH expression in hepatocytes regulates Kupffer cell cytokine production.
Kupffer cells and hepatocytes were isolated and cultured separately. LPS (10ng/ml) was used to elicit Kupffer cell inflammatory response. TNF-α (25ng/ml) plus ActD (0.4μg/ml) was used to induce hepatocyte apoptosis. (A) mRNA level of pro-inflammatory cytokines (IL1β, IL6, TNF-α, MCP1 and CXCL2) and 15-PGDH in Kupffer cells after LPS treatment. The results are shown as fold changes compared to Kupffer cells isolated from wild type mice. (B) Accumulation of reactive oxygen species (ROS) in hepatocytes after TNF-α/ActD treatment. ROS was measured by dichlorofluorescin fluorescence assay and shown as fold change compared to untreated hepatocytes. (C) Hepatocyte apoptosis induced by TNF-α/ActD treatment. Apoptotic hepatocytes were stained by TUNEL assay. Representative images are showed in the left panel. Quantified results are showed in the right panel. (D) The protein levels of apoptotic signaling molecules (cleaved PARP, JNK, P-JNK) in hepatocytes after TNF-α/ActD treatment. GAPDH was used as loading control. (E) mRNA level of pro-inflammatory cytokines (IL1β, IL6, TNF-α, MCP1 and CXCL2) in WT Kupffer cells treated with hepatocyte CM followed by LPS stimulation. The results are presented as fold changes compared to Kupffer cells treated with CM of wild type hepatocytes. The quantitative data presented in this figure were obtained from three independent experiments and are expressed as means ± SE (*p<0.05, **p<0.01, ***p<0.001; N.S denotes no statistical significance).
Fig 3.
15-PGDH-derived 15-keto-PGE2 from hepatocytes inhibits Kupffer cell activation via PPAR-γ.
(A) PGE2 metabolites concentration in CM of hepatocyte as measured by the Prostaglandin E Metabolite assay. We observed that 15-PGDH overexpression enhanced PGE2 metabolite production and this effect was more apparent in the presence of arachidonic acid substrate (1μM); this effect was blocked by treatment with the 15-PGDH inhibitor (1μM). (B) mRNA levels of pro-inflammatory cytokines (IL1β, IL6, TNF-α, MCP1 and CXCL2) in WT Kupffer cells treated with DMSO/15-keto-PGE2 (10μM) followed by LPS. The results are shown as fold changes compared to Kupffer cells treated with DMSO. (C) mRNA levels of PPAR-γ downstream genes (CD36, ADRP, CPT1α and ABCG1) in WT Kupffer cells treated with CM of hepatocytes (upper panel) or DMSO/15-keto-PGE2 (10μM) (lower panel). The data are shown as fold changes compared to Kupffer cells treated CM of wild type hepatocytes (upper panel) or Kupffer cells treated with DMSO (lower panel). (D) DNA binding ability of PPAR-γ from mouse macrophages (RAW264.7) treated with CM of hepatocytes (upper panel) or DMSO/15-keto-PGE2 (10μM) (lower panel), as determined by EMSA assay using PPRE dsDNA probe. (E) mRNA levels of pro-inflammatory cytokines (IL1β, IL6, TNF-α, MCP1 and CXCL2) in WT Kupffer cells treated with CM of hepatocytes (upper panel) or DMSO/15-keto-PGE2 (10μM) (lower panel) (with or without GW9662 [10μM] or LPS treatment). For the upper panel, the results are shown as fold changes compared to Kupffer cells treated with CM of 15-PGDH Tg hepatocytes. For the lower panel, the results are shown as fold changes compared to Kupffer cells treated with GW9662. The quantitative data in this figure were obtained from three independent experiments and are expressed as means ± SE (*p<0.05, **p<0.01, ***p<0.001; N.S—no statistical significance).
Fig 4.
PPAR-γ antagonist, GW9662, restored the susceptibility of 15-PGDH Tg mice to LPS/GalN-induced acute liver injury.
Wild type and 15-PGDH Tg mice were intraperitoneally injected with GW9662 (1μg/g body weight) 2 h before LPS/GalN administration. Mice were sacrificed 5 h after LPS/GalN injection to collect serum and liver tissue samples. (A) Serum ALT and AST levels (n = 3). (B) Caspase-3, 8, and 9 activities in liver tissue homogenates (n = 3). The data are shown as fold changes compared to wild type mice treated with GW9662. (C) Representative images of liver tissue sections (200×): H&E stain (upper panel), caspase-3 IHC (mid panel), F4/80 IHC (lower panel). Quantified results are showed at the right panels (n = 3). (D) The protein levels of apoptotic signaling molecules (cleaved PARP, JNK, P-JNK) in liver tissue homogenates. GAPDH was used as loading control. (E) mRNA levels of pro-inflammatory cytokines (IL1β, IL6, TNF-α, MCP1 and CXCL2) in liver tissue homogenates (n = 3). The results are shown as fold changes compared to wild type mice treated with GW9662. The quantitative data in this figure are expressed as means ± SE (*p<0.05, **p<0.01, ***p<0.001; N.S—no statistical significance).