Porphyromonas gingivalis lipopolysaccharide antagonizes Escherichia coli lipopolysaccharide at toll-like receptor 4 in human endothelial cells

Abstract

E. coli lipopolysaccharide (LPS) induces cytokine and adhesion molecule expression via the toll-like receptor 4 (TLR4) signaling complex in human endothelial cells. In the present study, we investigated the mechanism by which Porphyromonas gingivalis LPS antagonizes E. coli LPS-dependent activation of human endothelial cells. P. gingivalis LPS at 1 micro g/ml inhibited both E. coli LPS (10 ng/ml) and Mycobacterium tuberculosis heat shock protein (HSP) 60.1 (10 micro g/ml) stimulation of E-selectin mRNA expression in human umbilical vein endothelial cells (HUVEC) without inhibiting interleukin-1 beta (IL-1beta) stimulation. P. gingivalis LPS (1 micro g/ml) also blocked both E. coli LPS-dependent and M. tuberculosis HSP60.1-dependent but not IL-1beta-dependent activation of NF-kappaB in human microvascular endothelial (HMEC-1) cells, consistent with antagonism occurring upstream from the TLR/IL-1 receptor adaptor protein, MyD88. Surprisingly, P. gingivalis LPS weakly but significantly activated NF-kappaB in HMEC-1 cells in the absence of E. coli LPS, and the P. gingivalis LPS-dependent agonism was blocked by transient expression of a dominant negative murine TLR4. Pretreatment of HUVECs with P. gingivalis LPS did not influence the ability of E. coli LPS to stimulate E-selectin mRNA expression. Taken together, these data provide the first evidence that P. gingivalis LPS-dependent antagonism of E. coli LPS in human endothelial cells likely involves the ability of P. gingivalis LPS to directly compete with E. coli LPS at the TLR4 signaling complex.

FIG. 1.

P. gingivalis LPS antagonizes E. coli LPS-activation of E-selectin expression in HUVEC as measured by enzyme-linked immunosorbent assay. Cells were treated for 4 h with the indicated doses of LPSs, and E-selectin expression was determined as described in Materials and Methods. Data are represented as the mean absorbance from duplicate samples. (A) Ability of E. coli LPS (EcLPS), P. gingivalis LPS (PgLPS), and P. gingivalis LPS_RE to activate E-selectin expression in HUVEC. (B) Ability of P. gingivalis LPS to antagonize E. coli LPS-dependent activation of E-selectin expression in HUVEC. (C) Ability of P. gingivalis LPS_RE to antagonize E. coli LPS-dependent activation of E-selectin expression in HUVEC. Results shown are representative of three independent experiments.

FIG. 2.

P. gingivalis LPS (PgLPS) antagonizes E. coli LPS (EcLPS) induction but not Il-1β induction of E-selectin mRNA expression in HUVEC. Cells were plated in six-well culture dishes and treated with the indicated activators for 4 h. Total RNA was harvested, and RT-PCR analysis was performed to detect E-selectin and β-actin mRNA expression as described in Materials and Methods. The resulting RT-PCR products (lower panels) were imaged and subjected to densitometric analysis. E-selectin mRNA expression was normalized to β-actin mRNA expression, and the resulting values were expressed as E-selectin mRNA induction relative to the unstimulated control (upper panel). Results shown are representative of three independent experiments.

FIG. 3.

P. gingivalis LPS (PgLPS) and msbBLPS antagonize both E. coli LPS (EcLPS) induction and HSP60.1 induction of E-selectin mRNA expression in HUVEC. Cells were plated in six-well culture dishes and treated with the indicated stimulants and doses for 4 h. Total RNA was harvested, and RT-PCR analysis was performed to detect E-selectin and β-actin mRNA expression as described in Materials and Methods. The resulting RT-PCR products (lower panels) were imaged and subjected to densitometric analysis. E-selectin mRNA expression was normalized to β-actin mRNA expression, and the resulting values were expressed as E-selectin mRNA induction relative to the unstimulated control (upper panel). Results shown are representative of three independent experiments.

FIG. 4.

P. gingivalis LPS (PgLPS) antagonizes E. coli LPS (EcLPS)-dependent activation of NF-κB in HMEC-1. Cells were plated in 24-well culture dishes and cotransfected with pNF-κB-TA-Luc and pβ-actin Renilla luciferase as described in Materials and Methods. The following day, transfected cells were exposed to the indicated stimulants and doses for 5 h and washed with PBS, and the resulting cell lysates were analyzed for firefly luciferase activity and Renilla luciferase activity. Firefly luciferase values were normalized to Renilla luciferase values, and the resulting values are represented as NF-κB activation. Results are presented as means ± standard deviations of triplicate determinations and are representative of three independent experiments. Two independent preparations of P. gingivalis LPS that were used for these experiments yielded similar results.

FIG. 5.

P. gingivalis LPS (PgLPS) and msbB LPS antagonize both E. coli LPS (EcLPS)-dependent and M. tuberculosis HSP60.1-dependent activation of NF-κB in HMEC-1 cells. Cells were plated in 24-well culture dishes and cotransfected with pNF-κB-TA-Luc and pβ-actin Renilla luciferase as described in Materials and Methods. The following day, transfected cells were exposed to the indicated stimulants and doses for 5 h and washed with PBS, and the resulting cell lysates were analyzed for firefly luciferase activity and Renilla luciferase activity. Firefly luciferase values were normalized to Renilla luciferase values, and the resulting values are represented as NF-κB activation. Results are presented as means ± standard deviations of triplicate determinations and are representative of two independent experiments. Two independent preparations of P. gingivalis LPS that were used for these experiments yielded similar results.

FIG. 6.

P. gingivalis LPS -dependent activation of NF-κB is mediated by TLR4 in HMEC-1 cells. Cells were plated in 24-well culture dishes and cotransfected with pNF-κB-TA-Luc, pβ-actin Renilla luciferase, and either (A) the irrelevant plasmid control, pDisplay, (B) phTLR2, or (C and D) pmuTLR4P712H. The following day, transfected cells were exposed to the indicated doses of P. gingivalis LPS (PgLPS), E. coli LPS (EcLPS), or IL-1β for 5 h and washed with PBS, and the resulting cell lysates were analyzed for firefly luciferase activity and Renilla luciferase activity. Firefly luciferase values were normalized to Renilla luciferase values, and the resulting values are represented as NF-κB activation. Results are presented as means ± standard deviations of triplicate determinations. Asterisks indicate statistically significant (P < 0.05) differences between P. gingivalis LPS- and peptidoglycan-treated cells and the unstimulated control. Results shown are representative of three independent experiments.

FIG. 7.

Ectopic expression of mCD14 does not reduce the ability of P. gingivalis LPS (PgLPS) to antagonize E. coli LPS (EcLPS) in HMEC-1 cells. Cells were plated in 24-well culture dishes and cotransfected with pNF-κB-TA-Luc, pβ-actin Renilla luciferase, and either pDisplay or phmCD14 as indicated in panels A and B. The following day, transfected cells were exposed to the indicated doses for 5 h and washed with PBS, and the resulting cell lysates were analyzed for firefly luciferase activity and Renilla luciferase activity. Firefly luciferase values were normalized to Renilla luciferase values, and the resulting values are represented as NF-κB activation. Each treatment was performed in triplicate and expressed as the standard deviation of the mean. (A) P. gingivalis LPS-dependent agonism in HMEC-1 cells transiently expressing human mCD14. (B) P. gingivalis LPS-dependent antagonism of E. coli LPS in HMEC-1 cells transiently expressing human mCD14. Results shown are representative of two independent experiments.

FIG. 8.

Pretreatment of HUVEC with P. gingivalis LPS (PgLPS) does not influence the ability of E. coli LPS (EcLPS) to induce E-selectin mRNA expression. HUVEC were plated in six-well culture dishes and treated for 2 h with either growth medium alone or growth medium containing P. gingivalis LPS (1 μg/ml). Subsequently, the cells were treated with the indicated LPSs for 4 h. Total RNA was harvested, and RT-PCR analysis was performed to detect E-selectin and β-actin mRNA expression as described in Materials and Methods. The resulting RT-PCR products (lower panels) were imaged and subjected to densitometric analysis. E-selectin mRNA expression was normalized to β-actin mRNA expression, and the resulting values are expressed as E-selectin mRNA induction relative to the unstimulated control (upper panel). Results shown are representative of three independent experiments.