Lipopolysaccharide-Mimetic Activities of a Toll-Like Receptor 2-Stimulatory Substance(s) in Enterobacterial Lipopolysaccharide Preparations

Cell culture and reagents.The human embryonic 293 cell line (obtained from the Human Science Research Resources Bank, Tokyo, Japan) was grown in Dulbecco's modified Eagle medium (Gibco BRL, Rockville, Md.) supplemented with 10% (vol/vol) heat-inactivated fetal calf serum (FCS; Gibco BRL), penicillin (100 U/ml), and streptomycin (100 μg/ml). E. coli O111 LPS purified by phenol extraction followed by gel filtration (Ec O111a), E. coli O111 LPS purified by phenol extraction (Ec O111b), and E. coli F583 Rd LPS (Ec F583 Rd) purified by phenol-chloroform-petroleum ether extraction were obtained from Sigma (St. Louis, Mo.). E. coli J5 Rc LPS (Ec J5 Rc) purified by the method of Galanos et al. (8) was obtained from List Biological (Campbell, Calif.). E. coli R3 F653 R-type LPS (Ec R3 F653) and Salmonella enterica serovar Minnesota Re595 LPS (Sm Re595) were purified by the method of Galanos et al. (8), and S. enterica serovar Minnesota S-type LPS (Sm S-type) and S. enterica serovar Abortusequi LPS (Sab) were purified by phenol extraction in our laboratory. LPS preparations were subjected to sodium dodecyl sulfate-15% polyacrylamide gel electrophoresis, and LPS was visualized with a silver staining kit (Bio-Rad, Hercules, Calif.). In addition, electrophoretic transfer to polyvinylidene difluoride membranes was performed, and protein was detected by using a colloidal gold reagent with a gold enhancement kit (Bio-Rad), which allows detection of as little as 10 to 100 pg of protein. Peptidoglycan from Staphylococcus aureus and polymyxin B were obtained from Sigma. The E. coli murein lipoprotein-derived synthetic lipopeptide tripalmitoyl-Cys-Ser-Ser-Asn-Ala (Pam₃CSSNA) was obtained from Bachem (Bubendorf, Switzerland) and suspended in 25 mM octyl glucoside. Synthetic E. coli-type lipid A (compound 506), Salmonella-type lipid A (compound 516), and a lipid A precursor (compound 406, also known as compound IVa) were kindly provided by Daiichi Kagaku (Tokyo, Japan). An antibody (no. 1060) against the equine infectious anemia virus tag epitope (amino acid sequence: ADRRIPGTAEE) was a generous gift from Nancy Rice (NCI-Frederick Cancer Research and Development Center). The anti-mouse CD14 antibody M-20 was purchased from Santa Cruz Biotechnology (Santa Cruz, Calif.). Recombinant human LPS-binding protein (LBP) was prepared as previously described (26).

Phenol fractionation.LPS preparations were fractionated according to the modified phenol extraction method described by Hirschfeld et al. (13) with a slight modification, as follows. LPS preparations were dissolved in purified water containing 0.2% triethylamine and 0.5% deoxycholate and further extracted twice with purified water containing 0.2% triethylamine and 0.5% deoxycholate. The final phenol phase was dialyzed against methanol and air dried. The residue was dissolved in purified water by adding a trace amount of triethylamine and then used as the phenol extract (PEX; yield: 15 to 20%). The water phase obtained at the first phenol extraction was further extracted with phenol, and the final water phase was extensively dialyzed against purified water. To the dialysate, a 1/10 volume of 3 M sodium acetate (pH 4) and 3 volumes of cold ethanol were added. The mixture was incubated overnight at −20°C and then centrifuged at 4°C and at 10,000 × g for 15 min. The precipitate was washed with cold 70% ethanol and then air dried. The dried precipitate was dissolved in purified water by adding a trace amount of triethylamine and was used as the water fraction (WEX; yield: 40 to 60%).

Expression plasmids.Mammalian expression plasmids for human CD14, TLR4, and MD-2 were described previously (22-24). The region encoding human TLR2 was amplified by reverse transcription-PCR from total RNA prepared from THP-1 cells and cloned into mammalian expression vector pcDNA3 (Invitrogen, Carlsbad, Calif.). The region encoding TLR2 minus its signal peptide sequence of the construct described above was subcloned into the downstream portion of a modified pcDNA3 vector; in the product of this vector the coding sequence for the preprotrypsin signal peptide sequence precedes that for the NH₂-terminal equine infectious anemia virus tag epitope. The NF-κB-dependent luciferase reporter plasmid pELAM-L was described previously (26).

NF-κB reporter assay.The NF-κB-dependent luciferase reporter assay was performed as described elsewhere (23, 26). Briefly, human embryonic kidney 293 cells (3 × 10⁵ to 5 × 10⁵/well) were plated in six-well dishes and on the following day were transfected by the calcium phosphate precipitation method with the indicated amounts of the respective expression plasmids along with 0.2 μg of pELAM-L described above and 0.05 μg of reporter plasmid pRL-TK (Promega, Madison, Wis.) for normalization. At 24 h after transfection, cells were stimulated for 6 h, and the reporter gene activity was measured according to the manufacturer's (Promega) instructions.

Limulus amoebocyte gelation activity.Activation of the proclotting enzyme of the horseshoe crab was tested by using a quantitative Limulus assay reagent (Endospecy; Seikagaku Kogyo, Tokyo, Japan) as described previously (37). Briefly, each sample (50 μl) was incubated with the same volume of the assay reagent for 30 min at 37°C and optical density at 405 nm was measured.

TLR2-stimulatory activities of various LPS preparations.We first tested the TLR2-stimulatory activities of commercial and our own LPS preparations obtained from E. coli and Salmonella in 293 cells expressing human CD14 and TLR2 (Fig. 1). Among eight LPS preparations tested, six induced NF-κB-dependent reporter activity in a concentration-dependent manner. Sab and Sm Re595, up to 10 μg/ml, did not induce this activity. Only J5 Rc showed a detectable amount of protein contamination with silver-enhanced colloidal gold staining (data not shown). Expression of human MD-2 in addition to CD14 and TLR2 did not affect these activities (data not shown).

• Open in new tab
• Download powerpoint

FIG. 1.

Activation of NF-κB in response to LPS preparations in 293 cells expressing CD14 and TLR2. 293 cells were transiently transfected with CD14 (0.1 μg) and TLR2 (0.1 μg) plasmids, together with an NF-κB-dependent luciferase reporter plasmid. After 24 h, cells were either left unstimulated (open bars) or stimulated for 6 h with LPS preparations in the presence of 10% (vol/vol) FCS, and luciferase activity was then measured. Values are means ± standard errors from two to four independent experiments. LPS preparations used were Ec O111a, Ec O111b, Ec R3 F653, Ec J5 Rc, Ec F583 Rd, Sab, Sm S-type, and Sm Re595.

To characterize the TLR2-stimulatory substances contained in LPS preparations, the E. coli O111a LPS preparation, which showed no detectable protein contamination, was subjected to modified phenol fractionation. The preparation was extracted with phenol in the presence of deoxycholate, and the resulting phenol fraction (PEX) and water fraction (WEX) were analyzed for TLR2-stimulatory activity (Fig. 2). PEX induced NF-κB-dependent reporter activity in 293 cells expressing CD14 and TLR2, and this effect was TLR2 dependent because no activation was observed in 293 cells transfected with a control vector or cells expressing only CD14 (data not shown). PEX induced stronger activity than the original LPS preparation, whereas WEX showed no TLR2-stimulatory activity (Fig. 2A). On the other hand, WEX activated NF-κB-dependent reporter activity in 293 cells expressing CD14, TLR4, and MD-2, and this activity was equivalent to that of the original LPS preparation (Fig. 2B). PEX also possessed TLR4-stimulatory activity, although the activity was approximately 100-fold weaker than that of the original LPS preparation (Fig. 2B). The activities of these preparations in 293 cells expressing CD14, TLR4, and MD-2 paralleled their Limulus activities (Fig. 2C).

• Open in new tab
• Download powerpoint

FIG. 2.

TLR2- and TLR4-mediated activation of NF-κB in response to phenol and water extracts prepared from an E. coli O111 LPS preparation and their Limulus amoebocyte gelation activities. 293 cells were transiently transfected with plasmids encoding either CD14 (0.1 μg) and TLR2 (0.1 μg) (A) or CD14 (0.1 μg), TLR4 (2 ng), and MD-2 (2 ng) (B), together with an NF-κB-dependent luciferase reporter plasmid. After 24 h, cells were either left unstimulated (○) or stimulated for 6 h with E. coli O111 LPS, WEX, or PEX obtained from E. coli O111 LPS in the presence of 10% (vol/vol) FCS, and luciferase activity was then measured. Values are means ± standard errors from one to four independent experiments. (C) Limulus amoebocyte gelation activities of E. coli O111 LPS, WEX, and PEX were measured by using a quantitative Limulus assay reagent, and the chromogen released was measured at an optical density (O.D.) of 405 nm.

CD14 and LBP are required for the TLR2-stimulatory activity of PEX.We next asked whether the TLR2-stimulatory activity of PEX was attributable to bacterial lipoprotein or peptidoglycan contaminations. To answer this question, we compared their modes of action in TLR2-mediated activation of NF-κB. In 293 cells expressing human CD14 and TLR2, PEX induced NF-κB-dependent reporter activity. This activity was CD14 dependent because expression of only TLR2 significantly reduced the activity (Fig. 3A), although the expression levels of TLR2 in cells transfected with TLR2 and in cells transfected with CD14 plus TLR2 were equivalent (Fig. 3A, inset). The phenol extract prepared from Sm S-type also activated NF-κB in a CD14- and TLR2-dependent manner (data not shown). On the other hand, the activation of NF-κB in response to the synthetic bacterial lipopeptide Pam₃CSSNA was not CD14 dependent because comparable activation was seen in 293 cells expressing TLR2 alone and CD14 plus TLR2 (Fig. 3B). The TLR2 ligand peptidoglycan also activated NF-κB in a CD14-dependent manner (Fig. 3C). In 293 cells transfected with a control vector or cells expressing only human CD14, no stimulation of reporter activity was observed with PEX, Pam₃CSSNA, or peptidoglycan at concentrations up to 10 μg/ml (data not shown).

• Open in new tab
• Download powerpoint

FIG. 3.

CD14-dependent activation of NF-κB by PEX. 293 cells were transiently transfected with either an empty vector (0.1 μg) plus a TLR2 plasmid (0.1 μg; ○) or plasmids encoding CD14 (0.1 μg) and TLR2 (0.1 μg) (•), together with an NF-κB-dependent luciferase reporter plasmid. After 24 h, cells were stimulated for 6 h with PEX (A), a synthetic murein lipopeptide (Pam₃CSSNA; B), or peptidoglycan (PG; C) in the presence of 10% (vol/vol) FCS, and luciferase activity was then measured. Values are means ± standard errors from at least three independent experiments. A part of the cellular extract prepared for the measurement of the luciferase activities was analyzed for the expression of TLR2 and CD14 by Western blotting (A, inset).

The activity of PEX was also enhanced by the presence of serum (Fig. 4). In the presence of 10% serum, 0.1 μg of PEX/ml induced NF-κB-dependent reporter activity comparable to that observed with 1 μg of PEX/ml in the absence of serum. When human LBP was used instead of serum, the reporter activity in response to PEX was also increased, and only 1/10 the amount of PEX was sufficient to produce activity levels similar to those observed in the absence of LBP or serum (Fig. 4A). On the other hand, reporter activities in response to peptidoglycan and Pam₃CSSNA were not affected by the presence of serum (Fig. 4B).

• Open in new tab
• Download powerpoint

FIG. 4.

Serum-dependent activation of NF-κB by PEX. 293 cells were transiently transfected with plasmids encoding CD14 (0.1 μg) and TLR2 (0.1 μg), together with an NF-κB-dependent luciferase reporter plasmid. After 24 h, cells were washed with phosphate-buffered saline twice and stimulated for 6 h with PEX (A), a synthetic murein lipopeptide (Pam₃CSSNA; B), or peptidoglycan (PG; B) in Dulbecco's modified Eagle medium in the absence (open bars) or presence (hatched bars) of LBP (100 ng/ml) or FCS (10% [vol/vol]) (solid bars), and luciferase activity was then measured. Values are means ± standard errors from at least four independent experiments. Responses were compared with responses in the absence of LBP or serum by a two-tailed Student t test (∗, P < 0.05; ∗∗, P < 0.01).

LPS-mimetic activity of PEX.The TLR2-stimulatory activity of PEX was CD14 and LBP dependent, thus closely resembling the effect of LPS on TLR4-mediated activity. Therefore, we next asked whether the TLR2-stimulatory activity in response to PEX is inhibited by polymyxin B. Polymyxin B at 10 to 50 μg/ml inhibited LPS-induced activation of NF-κB in 293 cells expressing CD14, TLR4, and MD-2. A similar extent of inhibition was observed in PEX-induced activation of NF-κB in 293 cells expressing CD14 and TLR2, and 50 μg of polymyxin B/ml inhibited activation almost completely (Fig. 5A). On the other hand, peptidoglycan-induced activation of NF-κB was not affected by polymyxin B, and there was only a slight inhibition in Pam₃CSSNA-induced activation of NF-κB in 293 cells expressing CD14 and TLR2 (Fig. 5B).

• Open in new tab
• Download powerpoint

FIG. 5.

Polymyxin B and lipid A precursor inhibit PEX-induced activation of NF-κB. (A and B) 293 cells were transiently transfected with plasmids encoding either CD14 (0.1 μg) and TLR2 (0.1 μg) (A, left five bars, and B) or CD14 (0.1 μg), TLR4 (2 ng), and MD-2 (2 ng) (A, right five bars), together with an NF-κB-dependent luciferase reporter plasmid. After 24 h, cells were either left untreated (open bars) or treated with the indicated concentrations of polymyxin B followed by either PEX (0.1 μg/ml; A), LPS (10 ng/ml; A), peptidoglycan (PG; 10 μg/ml; B), or Pam₃CSSNA (0.1 μg/ml; B) for 6 h in the presence of 100 ng of LBP/ml, and luciferase activity was then measured. Luciferase activities were expressed relative to the activity in response to PEX, LPS, PG, or Pam₃CSSNA in the absence of polymyxin B. Values are means ± standard errors (SE) from at least three independent experiments. (C and D) 293 cells were transiently transfected with plasmids encoding either CD14 (0.1 μg) and TLR2 (0.1 μg) (C and D, left four bars) or TLR2 (0.1 μg; D, right four bars), together with an NF-κB-dependent luciferase reporter plasmid. After 24 h, cells were either left untreated (open bars) or treated with the indicated concentrations of 506, 516, or 406 followed by either PEX (0.1 μg/ml; C), PG (10 μg/ml; D), or Pam₃CSSNA (0.1 μg/ml; D) for 6 h in the presence of 100 ng of LBP/ml, and luciferase activity was then measured. Luciferase activities were expressed relative to the activity in response to PEX, PG, or Pam₃CSSNA in the absence of lipid A compounds. Values are means ± SE from at least three independent experiments. Responses were compared with responses in the absence of any lipid A compounds by a two-tailed Student t test (∗, P < 0.05; ∗∗, P < 0.01.

The observation that polymyxin B inhibited PEX-induced activation of NF-κB strongly suggested the involvement of an LPS-like substance(s) in this activation. We thus examined the effect of synthetic lipid A compounds on PEX-induced activation in 293 cells expressing CD14 and TLR2. The activity was not significantly affected by either synthetic E. coli-type lipid A (compound 506) or Salmonella-type lipid A (compound 516), whereas a lipid A precursor (compound 406) significantly inhibited activation, and complete inhibition was observed at 10 μg of 406/ml (Fig. 5C). On the other hand, 406 inhibited neither peptidoglycan- nor Pam₃CSSNA-induced activation of NF-κB (Fig. 5D). These synthetic lipid A compounds by themselves did not activate NF-κB (data not shown) in 293 cells expressing CD14 and TLR2.