Perturbation of the Actin Cytoskeleton in Human Hepatoma Cells Influences Interleukin-6 (IL-6) Signaling, but Not Soluble IL-6 Receptor Generation or NF-κB Activation

Abstract

The transcription factor nuclear factor-kappa B (NF-κB) is critically involved in inflammation and cancer development. Activation of NF-κB induces the expression and release of several pro-inflammatory proteins, which include the cytokine interleukin-6 (IL-6). Perturbation of the actin cytoskeleton has been previously shown to activate NF-κB signaling. In this study, we analyze the influence of different compounds that modulate the actin cytoskeleton on NF-κB activation, IL-6 signaling and the proteolytic generation of the soluble IL-6 receptor (sIL-6R) in human hepatoma cells. We show that perturbation of the actin cytoskeleton is not sufficient to induce NF-κB activation and IL-6 secretion. However, perturbation of the actin cytoskeleton reduces IL-6-induced activation of the transcription factor STAT3 in Hep3B cells. In contrast, IL-6R proteolysis by the metalloprotease ADAM10 did not depend upon the integrity of the actin cytoskeleton. In summary, we uncover a previously unknown function of the actin cytoskeleton in IL-6-mediated signal transduction in Hep3B cells.

Keywords: NF-κB; STAT3; actin; interleukin-6; interleukin-6 receptor.

Figure 1

Short-term perturbation of the actin cytoskeleton does not activate NF-κB. (A) Stimulation of Hep3B cells with TNFα (100 U/mL) for 30 min 24 h, 48 h and 72 h after seeding. Representative immunoblot of phospho (p)p65 expression. GAPDH was visualized to verify equal loading. Data present the quantification of pp65 expression normalized to GAPDH. Data are shown as means ± SD of three independent experiments. Statistical significance was determined using two-tailed unpaired t tests (**: p < 0.01; ***: p < 0.001). (B) Stimulation of Hep3B with TNFα and cytochalasin B at increasing concentrations for 15 min. Representative immunoblots for pIκBα, IκBα pp65, p65 and GAPDH. (C,D) Quantification of expression of (C) pp65/p65 and (D) IκBα/GAPDH. Data are shown as means ± SD of three independent experiments. (E–G) The experiments were performed as described for panels B-D, but cells were pre-incubated with different amounts of jasplakinolide instead of CB. Statistical significance was determined using one-way ANOVA with Dunnett’s multiple comparisons test (all treated samples were compared to the DMSO-treated cells; *: p < 0.05; ****: p < 0.001; all other samples were not significantly different). CB: cytochalasin B; JP: jasplakinolide.

Figure 2

Long-term perturbation of the actin cytoskeleton does not activate NF-κB or induce IL-6 secretion. (A–F) The experiments were conducted as described in the legend to Figure 1, but the cells were stimulated for 120 min with the compounds indicated. Data are shown as means ± SD of two or three independent experiments. Statistical significance was determined using one-way ANOVA with Dunnett’s multiple comparisons test (all treated samples were compared to the DMSO-treated cells; *: p < 0.05; **: p < 0.01; ***: p < 0.001; ****: p < 0.001; all other samples were not significantly different). (G) ELISA quantification of IL-6 in the culture supernatant after pre-treatment of Hep3B cells with cytochalasin B (15 µM), cytochalasin D (1 µg) and jasplakinolide (0.25 µM) for 120 min. Data are shown as means ± SD of three independent experiments. Statistical significance was determined using one-way ANOVA with Dunnett’s multiple comparisons test and revealed no differences. CB: cytochalasin B; CD: cytochalasin D; JP: jasplakinolide.

Figure 3

Perturbation of the actin cytoskeleton interferes with STAT3 signaling in Hep3B cells. (A,B) Hep3B cells were pretreated with cytochalasin B (15 µM), cytochalasin D (1 µg), jasplakinolide (250 nM) or DMSO (0.1%) as a control for 120 min. IL-6 (10 ng/mL) was added for 15min in order to activate STAT3 signaling. Representative western blot images from one experiment are shown, the quantitative data are shown as means ± SD of three independent experiments. (C,D) The experiments were performed as described for panels A and B, but HepG2 cells were used. Representative western blot images from one experiment are shown, the quantitative data are shown as means ± SD of three independent experiments. Statistical significance was determined using one-way ANOVA with Sidaks’s multiple comparisons test (cells pre-treated with actin-modulating compounds and stimulated with IL-6 were compared to IL-6-stimulated cells only; ****: p < 0.001; ns = no significant difference CB: cytochalasin B; CD: cytochalasin D; JP: jasplakinolide.

Figure 4

Perturbation of the actin cytoskeleton does not influence IL-6R proteolysis by ADAM10. (A,B) ELISA quantification of sIL-6R in culture supernatant after pretreatment of Hep3B cells with (A) cytochalasin B (15 µM) and D (1 µg) or (B) cytochalasin D (1 µg) and jasplakinolide (0.25 µM) for 120 min. Proteolysis by ADAM10 was activated via ionomycin (1 µM) for 60 min. Data are shown as means ± SD of three independent experiments. (C,D) The experiments were performed as described for panels A and B, but HepG2 cells were used. Data are shown as means ± SD of three independent experiments. Statistical significance was determined using one-way ANOVA with Sidaks’s multiple comparisons test (cells pre-treated with actin-modulating compounds and stimulated with ionomycin were compared to ionomycin-stimulated cells only). ns = no significant difference; CB: cytochalasin B; CD: cytochalasin D; JP: jasplakinolide; IM: ionomycin.