Hepatocytes are a dominant source of HMGB1 in the innate immune response during sepsis
Author(s):
Meihong Deng; Melanie Scott, University of Pittsburgh; Timothy Billiar, University of Pittsburgh
Background: High-mobility group protein B1 (HMGB1) is a nucleus-DNA binding protein, which is secreted into extracellular milieu as an inflammatory cytokine during sepsis. However, the dominant source of circulating HMGB1 in sepsis is unknown.
Hypothesis: We hypothesized that hepatocytes (HCs) are one important source of circulating HMGB1 during sepsis.
Methods: To test the hypothesis, we generated HC-specific HMGB1-/- mice. C57BL/6 (WT), cell-specific knockout control (flox), and Hepatocyte-specific HMGB1-/- (HC-HMGB1-/-) mice underwent cecal ligation and puncture (CLP) for 18h.
Results: In WT mice, HMGB1 translocated from nucleus to cytoplasm in HC visualized by immunofluorescence staining with a dramatic increase of circulating HMGB1 level assessed by ELISA after CLP. Surprisingly, specific deletion of HMGB1 in HC resulted in more than 75% reduction in the plasma HMGB1 level (HC-HMGB1-/-: 11.8±3.3 ng/mL) compared to the control mice (WT: 59±7.3 ng/mL, Flox: 53±15.04 ng/mL) after CLP, suggesting that HC-HMGB1 was the major source of circulating HMGB1 during sepsis. Importantly, specific deletion of HMGB1 in HC significantly lowered bacterial load in peritoneum compared to controls (HC-HMGB1-/-: 7.0x105±6.0x105 vs WT: 1x107±1.5x106 CFU/mL and Flox: 7.33x107±3.06x106CFU/mL, p<0.05) at 18h after CLP, with a decrease in systemic inflammation indicated by significantly lower circulating IL-6 and IL-1β. Furthermore, fewer peritoneal apoptotic PMNs (Ly6G+ve, and TMR+ve cell) were detected by flow cytometry in HC-HMGB1-/- (3.12±0.21%) mice than in control mice (WT: 8.43±0.67% and Flox: 6.98±0.53%). This correlated with the HMGB1 levels in peritoneal lavage fluid and inversely correlated with the bacterial load. These results suggested that the release of HMGB1 from HC might contribute to cell death in PMNs during sepsis. To test whether HC release HMGB1 actively or passively from necrotic process, liver histology and liver function test (ALT) was performed. No obvious necrotic areas were seen in liver section in any mice strain after CLP. ALT levels elevated after CLP; however no significantly difference was detected between the strains. It suggested that HMGB1 was not released passively from necrotic HCs.
Conclusions: Together, these results suggested that HMGB1 released by HCs during sepsis maybe an active process that regulates bacterial clearance via induction of cell death in peritoneal PMNs.