Pro-inflammatory cytokines from Kupffer cells downregulate hepatocyte expression of adrenomedullin binding protein-1

Abstract

Polymicrobial sepsis is characterized by an early, hyperdynamic phase followed by a late hypodynamic phase. Adrenomedullin (AM), a vasodilatory peptide, inhibits this transition from the early phase to the late phase. Adrenomedullin binding protein-1 (AMBP-1) enhances AM-mediated activities. The decrease of AMBP-1 levels in late sepsis reduces the vascular response to AM and produces the hypodynamic phase. Studies have indicated that the administration of LPS downregulates AMBP-1 production in the liver. Since hepatocytes are the primary source of AMBP-1 biosynthesis in the liver, we employed a co-culture strategy using hepatocyte and Kupffer cells to determine whether LPS directly or by increasing pro-inflammatory cytokines from Kupffer cells downregulates AMBP-1 production. Hepatocytes and Kupffer cells isolated from rats were co-cultured and treated with LPS for 24 h. LPS significantly attenuated AMBP-1 protein expression in a dose-dependent manner. Since AMBP-1 is basically a secretory protein, cell supernatants from co-culture cells treated with LPS were examined for AMBP-1 protein levels. LPS treatment caused a dose related decrease in AMBP-1 protein secretion. Similarly, LPS treatment produced a significant decrease in AMBP-1 protein expression in hepatocytes and Kupffer cells cultured using transwell inserts. LPS had no direct effect on AMBP-1 levels in cultured hepatocytes or Kupffer cells alone. To confirm that the observed effects in co-culture were due to the cytokines released from Kupffer cells, hepatocytes were treated with IL-1beta or TNF-alpha for 24 h and AMBP-1 expression was examined. The results indicated that both cytokines significantly inhibited AMBP-1 protein levels. Thus, pro-inflammatory cytokines released from Kupffer cells are responsible for downregulation of AMBP-1.

Figure 1

LPS treatment downregulates AMBP-1 in a co-culture of hepatocytes and Kupffer cells. A. Cells in co-culture for 4 hrs were treated with LPS (1.0, 10 and 100 ng/ml). 24 hr post treatment, cells were lysed in HEPES buffer containing protease inhibitors and protein levels were measured. Equal amounts of protein (25 μg) were electrophoresed on 3–8% NuPAGE Tris-Acetate gels under non-reducing conditions and Western blotted with anti-human complement factor H antibody (AMBP-1). Duplicate aliquots of protein (25 μg) were electrophoresed on 4–12% NuPAGE Bis-Tris gels under reducing conditions and Western blotted with anti-β-actin antibody (β-actin). B. The ratios between AMBP-1 and β-actin protein expressions (calculated as arbitrary densitometric values obtained from AMBP-1 protein expression divided by β-actin values in corresponding lanes) are shown. Data are presented as mean ± SE (n=6) and compared by one-way ANOVA and Student-Newman-Keuls method. *P<0.05 versus control.

Figure 2

LPS treatment produced no increase in AMBP-1 secretion in a co-culture of hepatocytes and Kupffer cells. A. Cell supernatants from co-culture experiments treated with LPS were concentrated 10 fold with Centricon-50 and a 2 μl aliquot was analyzed by Western blot under non-reducing conditions using anti-AMBP-1 antibody. B. Arbitrary densitometric values obtained from the blot are shown. Data are presented as mean ± SE (n=3).

Figure 3

LPS treatment attenuates AMBP-1 expression in culture of hepatocytes and Kupffer separated by the use of 0.4 μm transwell inserts. A. Equal amount of proteins (25μg) from cell lysates were run on 3–8% NuPAGE gels under non-reducing conditions and Western blotted using anti-AMBP-1 antibody. Duplicate aliquots of protein (25 μg) were electrophoresed on 4–12% NuPAGE Bis-Tris gels under reducing conditions and Western blotted with anti-β-actin antibody (β-actin). B. The ratios between AMBP-1 and β-actin protein expressions are shown. Data are presented as mean ± SE (n=5) and compared by one-way ANOVA and Student-Newman-Keuls method. *P<0.05 versus control.

Figure 4

LPS treatment has no direct effect on AMBP-1 expression in hepatocytes. A. Hepatocytes in culture for 4 hrs were treated with LPS (1.0, 10 and 100 ng/ml) and incubated for 24 hrs at 37°C. Cells were lysed and equal amount of proteins (25 μg) were electrophoresed and Western blotted with either AMBP-1 or anti-β-actin antibody as described in the legend to Figure 1. B. The ratios between AMBP-1 and β-actin protein expression are shown. Data are presented as mean ± SE (n=6) and compared by ANOVA and Student-Newman-Keuls method.

Figure 5

AMBP-1 protein expression is extremely low in Kupffer cells and LPS had no further effect on its expression. A. Kupffer cells in culture for 24 hrs were treated with LPS (1.0, 10 and 100 ng/ml) and incubated for 24 hrs at 37°C. Cells were lysed and equal amount of proteins (15 μg) were electrophoresed and Western blotted with either AMBP-1 or anti-β-actin antibody. B. The ratios between AMBP-1 and β-actin are shown. Data are presented as mean ± SE (n=6) and compared by ANOVA and Student-Newman-Keuls method.

Figure 6

AMBP-1 protein expression in hepatocytes is significantly attenuated by IL-1β or TNF-α treatment. A. Hepatocytes in culture for 4 hrs were treated with IL-1β or TNF-α at 1.0 and 10 ng/ml and incubated for 24 hrs at 37°C. Cells were lysed and equal amount of proteins (25 μg) were electrophoresed and Western blotted with either AMBP-1 or anti-β-actin antibody. B. The ratios between AMBP-1 and β-actin protein expression are shown. Data are presented as mean ± SE (n=6) and compared by ANOVA and Student-Newman-Keuls method. *P<0.05 versus control.