Fig 1.
IFI16 binds to LPS of different bacterial origin and inflammatory activity.
Coomassie brilliant blue staining of pull-down assays performed with 3 μg of recombinant IFI16 (A) or GST (B) in the presence or absence of biotin-labeled lipopolysaccharide (LPS) from E. coli O111:B4 (biotin-LPS-EB). (C) Saturation binding experiments performed with 2 μg/ml of IFI16 (red circles) and increasing amount of biotin-LPS-EB. Binding was detected by ELISA using HRP-conjugated streptavidin. Optical density (OD) of samples was measured at 450 nm. An excess of recombinant GST (blue circles) or BSA (green circles) and pre-treatment of biotin-LPS-EB with polymyxin B (PMB, empty circles) were used as negative controls. Data are expressed as mean values ± SD of three independent experiments. (D) Surface plasmon resonance (SPR) analysis of LPS-EB binding to immobilized IFI16. After immobilization of IFI16 on the CM5 sensor chip surface, increasing concentration of LPS-EB (3.125–100 nM) diluted in running buffer were injected over immobilized IFI16. Data are representative of three independent experiments. (E) Ex-vivo interaction analysis between increasing amount of recombinant IFI16 and formalin-fixed gram-negative (E. coli and K. pneumonia; pink circles and pink squares, respectively) or gram-positive (S. aureus, S. epidermidis, S. pyogenes; blue circles, squares and triangles, respectively) bacteria. Data are expressed as mean values ± SD of three independent experiments. (F) Lipid A structures of LPS derived from E. coli O111:B4 or F583 LPS (LPS-EB and LPS-F583, respectively; strong TLR4 agonists), P. gingivalis (LPS-PG; weak TLR4 agonist) and R. sphaeroides (LPS-RS, TLR4 antagonist). For LPS-PG, which harbors a mixture of di-, mono- and de-phosphorylated penta- or tetra-acylated lipid A moieties, a single isoform is represented for simplicity. PS-outer chain = polysaccharide outer chain. (G) Saturation binding experiments with increasing amount of recombinant IFI16 (8 to 12,288 ng/ml) and 10 μg/ml of LPS-EB (red line), LPS-F583 (green line), LPS-PG (blue line) or LPS-RS (purple line). Anti-IFI16 antibodies against the N-terminus of the protein and HRP-labelled anti-rabbit IgG were added as primary and secondary antibody, respectively, and binding was detected by ELISA at 450 nm. Data are expressed as mean values ± SD of three independent experiments.
Table 1.
Full-length IFI16 and IFI16 domains binding affinities to LPS.
Fig 2.
FI16 binds to the lipid A moiety of LPS through its HINB domain.
(A) Structures of di- or mono-phosphorylated lipid A from E. coli F583 (DPLA and MPLA, respectively) and detoxified LPS (detox-LPS) derived from E. coli O111:B4. PS-outer chain = polysaccharide outer chain. (B) Saturation binding experiments with increasing amount of recombinant IFI16 (from 8 to 12,288 ng/ml) and 10 μg/ml of MPLA (green line), DPLA (red line) or detox-LPS (purple line). Binding was detected by ELISA as described in the legend to Fig 1G. Data are expressed as mean values ± SD of three different experiments. (C) Competition ELISA assay for LPS-EB binding to IFI16 with increasing amount of LPS-EB, MPLA or detox-LPS as competitors. Briefly, microtiter plates were coated with 1 μg/ml LPS-EB, then 2 μg/ml of IFI16 were added to the wells in the presence of increasing concentration (5 to 25 μg/ml) of competitor. Binding was detected by ELISA as described in the legend to Fig 1F. Data are expressed as mean values ± SD of three independent experiments (***P<0.001, **P<0.01, Student’s t test). (D) Domain organization of the IFI16 protein. The numbers represent the amino acid positions based on NCBI Reference Sequence NP_005522. From the N- to the C-terminal (left to right), IFI16 comprises a pyrin domain involved in protein-protein interaction, and two hematopoietic interferon-inducible nuclear protein with 200-amino-acid repeats (HINA and HINB) domains, which are a hallmark of the absent in melanoma 2-like receptors (ALRs). S/T/P = serine/threonine/proline-rich repeats, which are regulated by alternative mRNA splicing. (E) Coomassie brilliant blue staining of pull-down assays performed with 3 μg of recombinant PYRIN, HINA, and HINB domains, in presence or absence of biotin-labeled LPS from E. coli O111:B4 (biotin-LPS-EB). (F) Saturation binding experiments performed by using increasing amount (8 to 4,096 ng/ml) of recombinant PYRIN, HINA or HINB domains (orange, purple and blue lines, respectively) and 10 μg/ml of LPS-EB, LPS-PG, MPLA or DPLA. Anti-IFI16 antibodies against the N- or C-terminus of the protein and HRP-labeled anti-rabbit IgG were added as primary and secondary antibody, respectively, and binding detected in ELISA at 450 nm. Data are expressed as mean values ± SD of three independent experiments.
Fig 3.
IFI16 proinflammatory activity is potentiated by the strong TLR4 activator LPS-EB.
(A) qRT-PCR analysis of IL-6, IL-8, TNF-α and IL-1β mRNA expression levels in 786-O or THP-1 cells stimulated for 24 h with IFI16 (25 μg/ml), IFI16ΔHINB (25 μg/ml), LPS from E. coli O111:B4 (LPS-EB, 10 ng/ml or 1 μg/ml), IFI16/LPS-EB complex (preincubated O/N at 4°C), IFI16ΔHINB/LPS-EB (preincubated O/N at 4°C), or left untreated (mock). Values are normalized to GAPDH mRNA and plotted as fold induction over mock-treated cells. qRT-PCR data are presented as mean values of biological triplicates. Error bars indicate SD (*P < 0.05, **P < 0.01, ***P < 0.001; two-way ANOVA followed by Dunnett’s test). (B) Protein concentration of IL-6, IL-8 and TNF-α evaluated by ELISA in supernatants derived from 786-O or THP-1 cells stimulated for 24 h as described in A. Data are expressed as mean values ± SD of three independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001; two-way ANOVA followed by Dunnett’s test).
Fig 4.
Weak TLR4-activating LPS variants and the TLR4 antagonist LPS-RS do not potentiate IFI16 proinflammatory activity.
(A) qRT-PCR analysis of IL-6, IL-8 and TNF-α mRNA expression levels in 786-O or THP-1 cells stimulated for 24 h with or without IFI16 (25 μg/ml), LPS from E. coli F583 (LPS-F583, 10 ng/ml) or LPS from R. sphaeroides (LPS-RS, 10 ng/ml), MPLA (10 ng/ml), DPLA (10 ng/ml) or in the presence of one of the following complexes: IFI16/LPS-F583, IFI16/LPS-RS, IFI16/MPLA or IFI16/DPLA. Values are normalized to GAPDH mRNA and plotted as fold induction over mock-treated cells. qRT-PCR data are presented as mean values of biological triplicates. Error bars indicate SD, and the P values refer to comparisons between IFI16 vs. IFI16/LPS or IFI16/lipid A complex-treated cells (*P < 0.05, **P < 0.01, ***P < 0.001; two-way ANOVA followed by Dunnett’s test). (B) Protein concentration of IL-6, IL-8 and TNF-α evaluated by ELISA in supernatants derived from 786-O or THP-1 cells stimulated for 24 h as described in the legend to panel A. Data are expressed as mean values ± SD of three independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant; two-way ANOVA followed by Dunnett’s test). The P values are relative to comparisons between IFI16- and IFI16/LPS- or IFI16/lipid A-treated cells.
Fig 5.
IFI16 exerts its proinflammatory activity in a TLR4/MyD88-dependent fashion.
(A-C) Protein concentration of IL-6, IL-8 and TNF-α measured by ELISA in supernatants derived from 786-O or THP-1 cells transfected for 48 h with scramble control (siCTRL), or siRNAs against TLR4 (siTLR4, A), MD2 (siMD2, B) or MyD88 (siMyD88, C). Cells were then stimulated for 24 h with IFI16 (25 μg/ml), LPS from E. coli O111:B4 (LPS-EB, 10 ng/ml) or IFI16/LPS-EB complex (preincubated O/N at 4°C), or left untreated (mock). Data are expressed as mean values ± SD of three independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant; unpaired Student’s t-test for comparison of silenced cells vs. their relative control counterpart). (D) Schematic representation of the probe containing the NF-κB binding site (highlighted in grey). (E) 786-O cells or THP-1 cells were stimulated with IFI16 (25 μg/ml), LPS from E. coli O111:B4 (LPS-EB, 10 ng/ml) or IFI16/LPS-EB complex (preincubated O/N at 4°C), or left untreated (mock). After 2 h, the cells were lysed and the nuclear fraction was analyzed for NF-κB binding activity using the Universal EZ-TFA transcription factor assay colorimetric kit and the probe described in D. Data are expressed as mean values ± SD of three independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001; two-way ANOVA followed by Dunnett’s test).
Fig 6.
Endogenous IFI16 is released by UVB-exposed U2OS cells and triggers proinflammatory cytokines production in a TLR4-dependent fashion.
(A) Western blot analysis of IFI16 in culture supernatants and total cell extracts of UVB-exposed (0 or 800 Jm-2) U2OS or U2OS-IFI16-/- cells at 16 h after treatment. β-actin cellular expression was used for protein loading control. Data are representative of three independent experiments with similar results. (B) Protein concentration of IL-6, IL-8 and TNF-α evaluated by ELISA in supernatants derived from THP-1 cells stimulated for 24 h in the presence or absence of anti-TLR4 neutralizing antibodies (10 μg/ml) using conditioned medium from UVB-exposed (0 or 800 Jm-2) U2OS and U2OS-IFI16-/- cells, or complete medium (mock), preincubated (O/N at 4°C), or not, with LPS from E. coli O111:B4 (LPS-EB, 10 ng/ml), or LPS from R. sphaeroides (LPS-RS, 10 ng/ml). Values were normalized to the initial protein concentration of the analyzed cytokines in the supernatants used for the treatment. Data are expressed as mean values ± SD of three independent experiments. The P values refer to comparison in each group with cells treated only with the medium without any addiction (white bar and black border; *P < 0.05, **P < 0.01, ***P < 0.001; two-way ANOVA followed by Dunnett’s test).
Fig 7.
IFI16 binds to TLR4 in vitro and in vivo.
(A) 786-O cells were stimulated for 1 h in the presence or absence of the indicated concentrations of IFI16, LPS from E. coli O111:B4 (LPS-EB), or IFI16/LPS-EB complex. Total cell extracts, untreated or DNase I-treated, were subjected to immunoprecipitation using a TLR4 monoclonal antibody. Immunoprecipitates and whole-cell lysates were analyzed by immunoblotting with anti-IFI16, anti-TLR4 or anti-MD2 antibodies. β-actin protein expression was used for protein loading control. Data are representative of three independent experiments with similar results. (B) Surface plasmon resonance (SPR) analyses of IFI16 binding to immobilized TLR4. After immobilization of TLR4 on the CM5 sensor chip surface, increasing concentration of IFI16 (31.25–1,000 nM) diluted in running buffer were injected over immobilized TLR4. IFI16 binds to TLR4 with an equilibrium dissociation constant (KD) of 0.13 μM. Data are representative of three independent experiments.
Fig 8.
The IFI16/LPS complex proinflammatory activity is not affected by the presence of free LPS.
(A) qRT-PCR analysis of IL-6, IL-8, TNF-α and mRNA expression levels in 786-O or THP-1 cells stimulated for 24 h with IFI16 (25 μg/ml), LPS from E. coli O111:B4 (LPS-EB, 10 ng/ml), IFI16/LPS-EB complex, IFI16 + LPS-EB (not complexed), IFI16/LPS-EB complex + LPS-EB (10 ng/ml), LPS from Rhodobacter sphaeroides (LPS-RS, 10 ng/ml), IFI16/LPS-RS complex, IFI16 + LPS-RS (not complexed), IFI16/LPS-RS complex + LPS-RS (10 ng/ml), or left untreated (mock). Values were normalized to GAPDH mRNA and plotted as fold induction over mock-treated cells. The P values refer to comparisons with IFI16-treated cells (*P < 0.05, **P < 0.01, ***P < 0.001; one-way ANOVA followed by Dunnett’s test). (B) Protein concentration of IL-6, IL-8 and TNF-α was measured by ELISA in supernatants derived from 786-O or THP-1 cells stimulated for 24 h as described in A. Data are expressed as mean values ± SD of three independent experiments. The P values refer to comparisons with IFI16-treated cells (*P < 0.05, **P < 0.01, ***P < 0.001; one-way ANOVA followed by Dunnett’s test). (C) Surface plasmon resonance (SPR) analyses of IFI16, LPS-EB and IFI16/LPS-EB complex binding to immobilized TLR4/MD2 receptor. After immobilization of recombinant TLR4/MD2 on the CM5 sensor chip surface, increasing concentration of the different analytes—31.25–1000 nM for IFI16, 3.125–100 μM for LPS-EB, 31.25–1000 nM for IFI16/LPS-EB complex—diluted in running buffer were injected over immobilized TLR4/MD2. IFI16, LPS-EB, and IFI16/LPS-EB bind to TLR4/MD2 with an equilibrium dissociation constant (KD) of 0.68 μM, 0.15 nM, and 4.01 μM, respectively. Data are representative of three independent experiments.
Table 2.
Binding kinetics of IFI16, LPS-EB and IFI16/LPS-EB to TLR4/MD2 receptor.
Fig 9.
Proposed model depicting the inflammatory activities and binding kinetics to TLR4 of LPS and IFI16, alone or in combination.
The relative inflammatory activities, from low to high, are reported in the upper part of the scheme (orange arrow). The relative binding kinetics to TLR4 are reported in the lower part of the scheme (green arrow). The thickness of the black arrows is directly proportional to the ability of the pathway to induce NF-κB activation.