Lipopolysaccharide- TLR-4 Axis regulates Osteoclastogenesis independent of RANKL/RANK signaling

Abstract

Background: Lipopolysaccharide (LPS) is an endotoxin and a vital component of gram-negative bacteria's outer membrane. During gram-negative bacterial sepsis, LPS regulates osteoclast differentiation and activity, in addition to increasing inflammation. This study aimed to investigate how LPS regulates osteoclast differentiation of RAW 264.7 cells in vitro.

Results: Herein, we revealed that RAW cells failed to differentiate into mature osteoclasts in vitro in the presence of LPS. However, differentiation occurred in cells primed with receptor activator of nuclear factor-kappa-Β ligand (RANKL) for 24 h and then treated with LPS for 48 h (henceforth, denoted as LPS-treated cells). In cells treated with either RANKL or LPS, an increase in membrane levels of toll-like receptor 4 (TLR4) receptor was observed. Mechanistically, an inhibitor of TLR4 (TAK-242) reduced the number of osteoclasts as well as the secretion of tumor necrosis factor (TNF)-α in LPS-treated cells. RANKL-induced RAW cells secreted a very basal level TNF-α. TAK-242 did not affect RANKL-induced osteoclastogenesis. Increased osteoclast differentiation in LPS-treated osteoclasts was not associated with the RANKL/RANK/OPG axis but connected with the LPS/TLR4/TNF-α tumor necrosis factor receptor (TNFR)-2 axis. We postulate that this is because TAK-242 and a TNF-α antibody suppress osteoclast differentiation. Furthermore, an antibody against TNF-α reduced membrane levels of TNFR-2. Secreted TNF-α appears to function as an autocrine/ paracrine factor in the induction of osteoclastogenesis independent of RANKL.

Conclusion: TNF-α secreted via LPS/TLR4 signaling regulates osteoclastogenesis in macrophages primed with RANKL and then treated with LPS. Our findings suggest that TLR4/TNF-α might be a potential target to suppress bone loss associated with inflammatory bone diseases, including periodontitis, rheumatoid arthritis, and osteoporosis.

Keywords: Bone Resorption; Lipopolysaccharides; Osteoclasts; RANK ligand.

Fig. 1

LPS induces osteoclast differentiation in RANKL-primed cells. a The diagrammatic sketch illustrates the treatment strategy related to RANKL and RANKL-primed LPS-treated groups. b and c Representative images of TRAP stained osteoclasts in response to the treatment strategy presented in panel A. d The diagrammatic sketch explains the treatment strategy performed to evaluate the osteoclastogenic ability of LPS. e, f, and g Representative images of TRAP stained osteoclasts in response to the treatment strategy presented in panel D. h The number of TRAP-positive multinucleated osteoclasts were counted in all treatment groups (n = 3). Statistical analysis was performed to compare the number of osteoclasts in the treatment groups with the control group (RANKL). One-way ANOVA was applied, and values are expressed as mean ± standard deviation (SD). **P < 0.01. Magnification is X100 in panels B, C, and E-G

Fig. 2

Analysis of RANK expression in LPS- and RANKL-mediated osteoclastogenesis. a Equal amounts of membrane lysate proteins were used for immunoblotting analyses with antibodies against RANK (~ 90 kDa) and GAPDH (loading control; ~ 37 kDa). Protein levels were quantified by densitometry, corrected for the sample load based on the GAPDH level, and expressed as a fold increase relative to the control lane (−). The results represent one of three experiments performed. b and c Immunostaining with an antibody against RANK was performed in non-permeabilized RANKL (b) and LPS (c) -stimulated osteoclasts. White arrows indicate mature osteoclast. Red arrows indicate mononuclear cells. The results represent one of three experiments performed. d Identification of the time-dependent effect of OPG on LPS-induced osteoclast differentiation. The diagrammatic sketch demonstrates the treatment strategy of RAW cells with LPS (5 μg/mL) and OPG (120 ng/mL). e Representative images of TRAP stained osteoclasts in response to the treatment strategy is shown in panel D. TRAP stained osteoclasts in panels A and C were imaged with a 4× objective (magnification: 40X), and panels B and D were imaged with a 10× objective (magnification: 100X). f The number of TRAP-positive multinucleated osteoclasts were counted in both groups from three different experiments. Statistical analysis was performed to compare the number of osteoclasts in the LPS + OPG group with the control group (LPS). T-test was applied, and the difference between groups is not statistically significant. Scanned uncropped autoradiograms are presented in Additional file 5, Figure S5. Corresponding immunoblots are shown in panel A

Fig. 3

Evaluation of the involvement of TLR4 signaling in RANKL-induced osteoclastogenesis. a Immunoblotting analyses with antibodies against TLR4 (~ 95 kDa) and GAPDH (loading control; ~ 37 kDa) are shown. Protein levels were quantified by densitometry, corrected for the sample load based on GAPDH expression, and expressed as a fold increase relative to the control lane (−). The results represent one of three experiments performed. b Representative images of TRAP stained osteoclasts in response to treatment with RANKL and TAK-242 (5 μM/mL). TRAP stained osteoclasts in panels A and B were obtained with a 4× objective (magnification: 40X), while those in panels C and D were obtained with a 10× objective (magnification: 100X). c The number of TRAP-positive multinucleated osteoclasts were counted in both groups (n = 3). Statistical analysis was performed to compare the number of osteoclasts in the RANKL+TAK-242 group with the control group (RANKL). The t-test was applied; the difference between groups is not statistically significant. d TRAP stained osteoclasts in response to treatment with LPS (5 μg/mL) and LPS/TAK-242 (5 μM/mL) are shown. TRAP stained osteoclasts in panels A and C were obtained with a 4× objective (magnification: 40X), while those in panels B and D were obtained with a 10× objective (magnification: 100X). e The number of TRAP-positive multinucleated osteoclasts were counted in both groups (n = 3). Statistical analysis was performed to compare the number of osteoclasts in the LPS + TAK-242 group with the control group (LPS) using the t-test. **P < 0.01 vs LPS group. f Effect of TAK-242 on LPS-induced TNF-α production from RAW cells-derived osteoclast. ELISA determined the concentrations of TNF-α in the culture medium. One-way ANOVA was applied, and values are expressed as mean ± standard deviation (SD). **P < 0.01 vs. LPS group. Scanned uncropped autoradiograms are presented in Additional file 5, Figure S5. Corresponding immunoblots are shown in panel A

Fig. 4

Analysis of the regulatory role of TNF-α in LPS- induced osteoclastogenesis. a Measurement of TNF-α production in response to RANKL and LPS stimulation of RAW cells. ELISA determined the TNF-α concentrations in the culture medium. One-way ANOVA was applied, and values are expressed as mean ± standard deviation (SD). ***P < 0.001 vs. (−) control and RANKL- treated cells. b The diagrammatic sketch demonstrates the treatment strategy of RAW cells with LPS (5 μg/mL) and anti-TNF-α (2 μg/mL). c In response to the treatment strategy, representative images of TRAP stained osteoclasts are shown in panel B. TRAP stained osteoclasts in panels A and C were obtained with a 4X objective (magnification: 40X); panels in B and D were obtained with a 10× objective (magnification: 100X). d The number of TRAP-positive multinucleated osteoclasts were counted in both groups (n = 3). Statistical analysis was performed to compare the number of osteoclasts in the LPS+ anti- TNF-α group with the control group (LPS). The t-test was applied, and values are expressed as mean ± standard deviation (SD). **P < 0.01 vs. LPS group. LPS, lipopolysaccharide; RANKL, receptor activator of nuclear factor kappa-B ligand; TNF-α, tumor necrosis factor α

Fig. 5

Effect of RANKL, LPS, and anti-TNF-α / LPS on the membrane levels of TNFR-1 and TNFR-2. Panels a, c, and d Immunoblotting analysis for TNFR1 (a; top panel, ~ 55 kDa) and TNFR-2 (middle panel in a, ~ 68 kDa) in osteoclasts differentiated with RANKL and LPS are shown. Representative immunoblotting showing the membrane levels of TNFR-1 and -2 (Panel a). Protein levels in untreated RAW cells (−) are shown in lane 1. Membrane levels of TNFR-1 (n = 4) and TNFR-2 (n = 3) were quantified in an Un-Scan IT software, corrected for the GAPDH level, and provided as percentage surface level of receptors (Panels c and d). **p < 0.001 vs. LPS-treated cells (c and d). One-way ANOVA was applied, and values are expressed as mean ± SD of four and three independent experiments for TNFR-1 and TNFR-2, respectively. TNFR-2 blot in A was stripped and blotted with a GAPDH antibody (bottom panel in a). Panels b and e-h Immunoblotting analysis demonstrates the effect of LPS (lane 1) and LPS+ neutralizing antibody to TNF-α (Lane 2) on the membrane levels of TNFR1 (b; top panel) and TNFR-2 (b; middle panel). Protein levels were quantified by densitometry, corrected for the sample load based on GAPDH level, and provided as a fold-change relative to LPS- treated control cells (Panels e and g). The Table (f and h) provides the average pixel value of the protein bands (TNFR-1 in f and TNFR-2 in h) from two experiments and fold changes in the surface levels of interest proteins. The experiment was performed twice and demonstrated remarkably related results. Raw data are provided in the (Additional file 7, Figure S7; A-C)

Fig. 6

A schematic model is summarizing the possible role of LPS in osteoclastogenesis via TLR-4. Bacterial LPS induces osteoclast formation in RANKL-primed cells (pre-osteoclasts) via TLR4 signaling. LPS/TLR4 activation results in the secretion of TNF-α, which directly stimulates osteoclastogenic signals in pre-osteoclasts in a paracrine/autocrine manner via TNFR-2. LPS, lipopolysaccharide; RANKL, receptor activator of nuclear factor kappa-B ligand; TNF-α, tumor necrosis factor α; TLR4, Toll-like receptor 4; TNFR-2, tumor necrosis factor receptor-2