Protection from lethal gram-negative bacterial sepsis by targeting Toll-like receptor 4

Abstract

Toll-like receptor 4 (TLR4), the signal-transducing molecule of the LPS receptor complex, plays a fundamental role in the sensing of LPS from gram-negative bacteria. Activation of TLR4 signaling pathways by LPS is a critical upstream event in the pathogenesis of gram-negative sepsis, making TLR4 an attractive target for novel antisepsis therapy. To validate the concept of TLR4-targeted treatment strategies in gram-negative sepsis, we first showed that TLR4(-/-) and myeloid differentiation primary response gene 88 (MyD88)(-/-) mice were fully resistant to Escherichia coli-induced septic shock, whereas TLR2(-/-) and wild-type mice rapidly died of fulminant sepsis. Neutralizing anti-TLR4 antibodies were then generated using a soluble chimeric fusion protein composed of the N-terminal domain of mouse TLR4 (amino acids 1-334) and the Fc portion of human IgG1. Anti-TLR4 antibodies inhibited intracellular signaling, markedly reduced cytokine production, and protected mice from lethal endotoxic shock and E. coli sepsis when administered in a prophylactic and therapeutic manner up to 13 h after the onset of bacterial sepsis. These experimental data provide strong support for the concept of TLR4-targeted therapy for gram-negative sepsis.

Fig. 1.

TLR4-deficient and MyD88-deficient mice are protected from lethal Gram-negative bacterial sepsis. (A–C) WT, TLR2^−/−, TLR4^−/−, and MyD88^−/− C57BL/6 mice were injected i.p. with 2 × 10⁹ cfu of E. coli O18 and treated with antibiotics as described in Materials and Methods. Plasma concentrations of TNF (A) and IL-6 (B) were measured 4 h after bacterial challenge. The horizontal line represents the median cytokine concentration (TNF: P < .005 for TLR4^−/− or MyD88^−/− vs. WT or TLR2^−/−, P = .15 for TLR4^−/− vs. MyD88^−/−, and P = .48 for WT vs. TLR2^−/−; IL-6: P < .05 and < .005 for TLR4^−/− and MyD88^−/− vs. WT or TLR2^−/−, P = .13 for TLR4^−/− vs. MyD88^−/−, and P = .31 for WT vs. TLR2^−/−). (C) Survival of TLR4^−/−, MyD88^−/−, TLR2^−/−, and WT mice (P < .001). Data points are from 1 experiment (n = 6 to 7 mice per treatment groups).

Fig. 2.

Anti-TLR4 antibodies bind to TLR4 and inhibit the activation of macrophages induced by LPS. (A) Anti-TLR4 antibodies binding to immobilized mTLR4-Fc but not to mGITR-Fc by ELISA. (B) Flow cytometry analysis of anti-TLR4 antibodies (gray area) binding to WT (i.e., TLR4^+/+) (Upper) but not to TLR4^−/− (Lower) thioglycollate-elicited mouse peritoneal macrophages. Background staining using control antibodies is shown in white. (C) NF-κB activity in RAW 264.7 macrophages transiently transfected with a trimeric κB site luciferase reporter vector and preincubated for 30 min with anti-TLR4 and control antibodies (100 μg/mL) before stimulation with LPS (10 ng/mL) or Pam₃CSK₄ (2 μg/mL) for 18 h. Data on relative luciferase activity are expressed as mean ± SD of 4 replicates from 1 representative experiment. *P = .001 for anti-TLR4 versus control antibodies. (D) Western blot analyses of phosphorylated-ERK1/2 (p-ERK1/2) and total ERK1/2 expression in bone marrow–derived macrophages preincubated for 20 min with 10 or 100 μg/mL of anti-TLR4 or control antibodies before stimulation with LPS (1 ng/mL) and Pam₃CSK₄ (1 μg/mL) for 20 min. (E–H) TNF and IL-6 production by RAW 264.7 macrophages (E) or mouse whole blood (F–H) preincubated for 30 min with anti-TLR4 or control antibodies (100 μg/mL) before stimulation with 100 ng/mL of LPS, 10 μg/mL of PGN, 0.1 μM CpG ODN (CpG), or 10⁶ cfu/mL of heat-killed E. coli O18 for 4 h. Data are expressed as mean ± SD of triplicates from 1 representative experiment. *.005 < P < .05 and **P < .005for anti-TLR4 versus control antibodies.

Fig. 3.

Anti-TLR4 antibodies inhibit cytokine production and protect mice from lethal endotoxemia when administered prophylactically and therapeutically. (A–C) Mice injected i.v. with 40 mg/kg of anti-TLR4 or control antibodies and TLR4^−/− mice were sensitized with D-galactosamine 15 min before an i.v. injection of 50 ng of E. coli O111:B4 LPS. Plasma concentrations of TNF (A) and of IL-6 (B) were measured 1 h after LPS injection. The horizontal line represents the median cytokine concentration. Control versus anti-TLR4 antibodies, P < .0001 for TNF and P = .005 for IL-6. (C) Prophylaxis. Survival of mice treated with anti-TLR4 versus control antibodies (n = 19 and 21 mice per treatment group; P = .0001) and TLR4^−/− mice (n = 8). Data points are from 4 independent experiments for antibody evaluation. (D) Therapy. Survival of BALB/c mice treated with anti-TLR4 or control antibodies (40 mg/kg i.p.) 4 h after i.p. injection of 1 mg of E. coli O111:B4 LPS (P = .025). Data points are from 1 experiment (n = 8 mice per treatment group).

Fig. 4.

Prophylactic and therapeutic administration of anti-TLR4 antibodies protect mice from lethal Gram-negative bacterial sepsis. (A–D) BALB/c mice were injected i.p. with anti-TLR4 or control antibodies (160 mg/kg for A–C and 200 mg/kg for D) given before (prophylactically; A and B) or after (therapeutically; C and D) an i.p. injection of a high (2 × 10⁹ cfu) inoculum (A–C) or low (2 × 10⁵ cfu) inoculum (D) of E. coli O18. (A) Plasma concentrations of TNF and IL-6 were measured 4 h after the bacterial challenge. The horizontal line represents the median cytokine concentration. P < .005 for TNF and IL-6. (B–D) Survival of mice treated prophylactically (B) (at −4, −0.5, and + 4 h) or therapeutically either early (+1 and + 4 h) (C) or late (+13 h) (D). P < .0001, .02, and .03, respectively. Data points are from 1 experiment (n = 10–12 mice per treatment group).