TLR2 ligands induce NF-κB activation from endosomal compartments of human monocytes

Abstract

Localization of Toll-like receptors (TLR) in subcellular organelles is a major strategy to regulate innate immune responses. While TLR4, a cell-surface receptor, signals from both the plasma membrane and endosomal compartments, less is known about the functional role of endosomal trafficking upon TLR2 signaling. Here we show that the bacterial TLR2 ligands Pam3CSK4 and LTA activate NF-κB-dependent signaling from endosomal compartments in human monocytes and in a NF-κB sensitive reporter cell line, despite the expression of TLR2 at the cell surface. Further analyses indicate that TLR2-induced NF-κB activation is controlled by a clathrin/dynamin-dependent endocytosis mechanism, in which CD14 serves as an important upstream regulator. These findings establish that internalization of cell-surface TLR2 into endosomal compartments is required for NF-κB activation. These observations further demonstrate the need of endocytosis in the activation and regulation of TLR2-dependent signaling pathways.

Figure 1. Effect of endocytosis inhibitors on TLR2 mediated induction of TNF in primary human monocytes.

Monocytes were treated with pharmacological endocytosis inhibitors during 45min prior to treatment for 4h or 24h with LTA (1μg/ml) and Pam₃CSK₄ (100ng/ml). (A) Dose response of the effect of chlorpromazine (CPZ), chloroquine (CHQ) and Dynasore (Dyn) on TNF secretion in TLR2 ligand-activated monocytes. TNF response to LTA and Pam₃CSK₄ after 24h is strongly reduced by all three endosomal pathway inhibitors. (B) Effect of CPZ, CHQ and Dyn on TNF mRNA expression in LTA- and Pam₃CSK₄-activated monocytes at 4h. TNF mRNA response to LTA and Pam₃CSK₄ is reduced by all three endosomal pathway inhibitors. (C) Effect of CPZ, CHQ and Dyn on TNF mRNA and TNF production after 4h and 24h, respectively, in IFNγ-activated monocytes. Data were normalized to the TNF production observed in the absence of inhibitors. For all panels data are represented as mean +/- SD of at least 3 independent experiments. *:p ≤ 0.05; ** p ≤ 0.005; *** p ≤ 0.0005.

Figure 2. Effect of endocytosis inhibitors on TLR2 mediated induction of NF-κB in HEK-Blue2™ cells.

(A-C) Dose response of the effect of CPZ, CHQ and Dyn on NF-κB activity in LTA- and Pam3CSK4-activated HEK-Blue2™ cells, which express soluble alkaline phosphatase (SEAP) under control of a promoter containing five NF-κB binding sites. LTA- and Pam3CSK4-induced NF-κB activity was reduced by all endocytic pathway inhibitors. NF-κB activity was monitored in cells supernatant by SEAP enzyme activity measured with HEK-Blue™ Detection Medium, which contains a specific SEAP substrate. Data are represented as mean +/- SD of at least 3 independent experiments. *:p ≤ 0.05; **:p ≤ 0.005; ***:p ≤ 0.0005

Figure 3. Effects of clathrin knockdown on NF-κB responses to LTA and Pam₃CSK₄.

(A) The right panel shows quantification by western blot of the heavy chain of clathrin in HEK-Blue2™ and representative experience obtains with HEK-Blue2™ and HEK-Blue4™cells treated with stealth siRNA for the heavy chain of clathrin or Stealth RNAi™ negative control duplex. The left panel shows NF-κB activity of HEK-Blue2™ and HEK-Blue4™cells siRNA-treated for 72h and then activated with Pam3CSK4, LTA or LPS (100ng/ml, 1μg/ml, 100ng/ml, respectively) for 24h. Cells were tested for NF-κB activity by measuring SEAP activity in cell supernatants. NF-κB activity is presented as percentage of production in mock-siRNA transfected HEK-Blue2™or HEK-Blue4™ cells. Data are represented as mean +/- SD of at least 3 independent experiments.

Figure 4. LTA and Pam₃CSK₄ internalization allows NF-κB activation.

(A) HEK-Blue2™ or HEK-Blue4™ cells were treated with LTA-biotin, Pam₃CSK₄-biotin, LPS-biotin beads-bound ligands as well as Streptavidin-beads. After beads isolation, the amount of beads-associated HEK-Blue2™ (BA-TLR2) and HEK-Blue4™ (BA-TLR4) cells was measured by flow cytometry using anti-TLR2 and anti-TLR4 antibodies. Results are expressed as the ratio of the geometric mean of TLR2 or TLR4 fluorescence (GMEAN) ± SD of three experiments. Final GMEAN values are the result of GMEAN subtraction from isotype control. (B) HEK-Blue2™ or HEK-Blue4™ cells for LPS, were treated with soluble LTA-biotin (1μg/ml), Pam₃CSK₄-biotin (100ng/ml), LPS-biotin (100ng/ml) or LTA-biotin, Pam₃CSK₄-biotin, LPS-biotin beads-bound ligands (containing the same concentrations of TLR ligands as the soluble forms) and ligands depleted supernatants for 24h. NF-κB activity is monitored by SEAP enzyme activity measured with HEK-Blue™ Detection Medium, which contains a specific SEAP substrate. Data are represented as mean +/- SD of at least 3 independent experiments.

Figure 5. LTA and Pam₃CSK₄ do not induce IRF-3 phosphorylation or IFN-β production.

(A) Human monocytes were treated with LTA, Pam₃CSK₄, or LPS (1μg/ml, 100ng/ml and 100ng/ml respectively) for the indicated time periods. IFN-β expression was assayed by quantitative PCR. Data are represented as mean +/- SD of results obtained with monocytes from at least 3 different blood donors. (B) Human monocytes were treated with LTA, Pam₃CSK₄ or LPS (1μg/ml, 100ng/ml and 100ng/ml respectively) for indicated time and the degradation of IκB (NF-κB) and phosphorylation of IRF3 analyzed by Western blot. Data are representative of 3 independent experiments. (C) Human monocytes were treated with 50μM of blocking peptides for MyD88 and TRIF or control peptide (CP) during 60min prior to be activated for 24h with LTA (1μg/ml), Pam₃CSK₄ (100ng/ml) or LPS (100ng/ml). MyD88 blocking peptide reduces TNF response to LTA and Pam₃CSK₄ after 24h. Data are represented as mean +/- SD of results obtained with monocytes from at least 3 different blood donors.

Figure 6. CD14 controls LTA and Pam₃CSK₄ internalization and NF-κB activation.

HEK-Blue2™ cells are TLR2⁺ and CD14⁺ (CD14) while HEK-TLR2 cells are TLR2⁺ and CD14^- cells (no CD14). (A and C) HEK-Blue2™ cells ((red dot) and HEK-TLR2 cells (■ were treated with LTA-biotin 1μg/ml or Pam₃CSK₄-biotin 100ng/ml and TLR2 ligands endocytosis was measured by flow cytometry at the times indicated. Displayed are the mean +/- SD of the percentage of corrected fluorescence index (MFI) of specific extracellular TLR2 ligands staining at each time point. (B and D) HEK-Blue2™ cells ((red dot) and HEK-TLR2 cells (■ were treated with LTA-biotin 1μg/ml or Pam₃CSK₄-biotin 100ng/ml and intracellular accumulation of TLR2 ligands was measured by flow cytometry at the times indicated. Displayed are the mean +/- SD of the percentage of corrected fluorescence index (MFI) of specific intracellular TLR2 ligands staining at each time point. (E) HEK-Blue2™ cells and HEK-TLR2 cells were activated by indicated TLR2 ligands (LTA 1μg/ml or Pam₃CSK₄ 100ng/ml) and IL-8 production was assayed by ELISA. Data are represented as mean +/- SD of at least 5 independent experiments. (F) Monocytes were treated with blocking antibody against CD14 (10μg/ml) during 45min prior to be activated with Pam₃CSK₄ 100ng/ml or LTA 1μg/ml. After 24h, TNF secretion was assayed by ELISA. CD14 blocking antibody decreases significantly TNF production. Data are represented as mean +/- SD of at least 4 independent experiments. (G) HEK-Blue2™ cells (CD14⁺ cells) and HEK-TLR2 cells (no CD14 cells) were activated with LTA 1μg/ml or Pam₃CSK₄ 100ng/ml. The presence of CD14 and activation of IκB (NF-κB) were analyzed by Western blot. CD14 controls NF-κB activation in TLR2 ligands-activated cells. Data are representative of 3 independent experiments.