Published online before print April 28, 2009
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
activity and lead to enhanced leukotriene release in human monocytes
* Institute of Pharmaceutical Chemistry/Center for Drug Research, Development, and Safety (ZAFES), and
Pediatric Hematology and Oncology, Goethe University, Frankfurt/Main, Germany
1. Correspondence: Institute of Pharmaceutical Chemistry/Center for Drug Research, Development, and Safety (ZAFES), Goethe University, Max-von-Laue-Strasse 9, 60438 Frankfurt/Main, Germany. E-mail: b.sorg{at}pharmchem.uni-frankfurt.de
ABSTRACT
Toll-like receptors (TLRs) play an important role in innate immunity. They detect pathogen-associated receptor patterns (PAMPs) and initiate subsequent immune responses. Present studies investigate the influence of TLR2 ligands on leukotrienes (LT) formation in human monocytes. LTs are proinflammatory mediators derived from arachidonic acid (AA), which is released from membranes by phospholipase A2 (PLA2) enzymes. Pretreatment of MM6 cells with the TLR2 ligands LTA, FSL-1, or Pam3CSK4 resulted in an up to two- to threefold enhancement of ionophore-induced LT formation in a dose- and time-dependent manner and to an augmentation of ionophore-induced AA release with similar kinetics. Also in human peripheral blood mononuclear cells (hPBMC), TLR2 activators increased LT formation. Studies with PLA2 inhibitors indicated that the increase of AA release is a result of enhanced activity of group IV cPLA2 in MM6 cells. TLR2 ligands elicited the time-dependent activation of p38 MAPK and ERK1/2 pathways, which led to phosphorylation of cPLA2
at Ser505. Simultaneous inhibition of p38 MAPK and ERK1/2 pathways prevented the increase of cPLA2 phosphorylation and the augmentation of AA release. TLR2 ligand-induced increase of AA release was blocked by a neutralizing anti-hTLR2 antibody, indicating that TLR2 mediates augmented cPLA2 activation and subsequent LT biosynthesis.
Key Words: innate immunity • arachidonic acid • 5-lipoxygenase • Mono Mac 6
Introduction
TLRs are a family of pattern recognition receptors that play a crucial role in the detection of invading pathogens. Recognition of distinct PAMPs by individual TLR subtypes initiates the activation of NF-
B and of MAPKs, resulting in subsequent immune responses. TLR2 is activated by LTA, a cell wall component of Gram-positive bacteria. Besides, TLR2 can form heterodimers with other TLRs such as TLR1 and TLR6 and thus recognizes a wide spectrum of microbial components. The TLR2/6 heterodimer is activated by the synthetic diacylated lipopeptide FSL-1, whereas the synthetic triacylated lipopeptide Pam3CSK4 is recognized by TLR2/1. Recent studies have provided evidence that TLRs are not only critical in antimicrobial immunity but are also involved in autoimmune and inflammatory disorders [1
, 2
]. Particularly, TLR2 signaling has been implicated in the pathogenesis of atherosclerosis [3
].
As proinflammatory mediators, LTs likewise are associated with inflammatory diseases including asthma or atherosclerosis. They are synthesized from AA by 5-LO, which is expressed mainly in myeloid cells such as monocytes/macrophages [4
]. In the case of atherosclerotic tissue, LTs have been shown to be released from foam cells [4
, 5
]. AA is released from membrane phospholipids preferentially by two classes of the PLA2 enzyme family, namely, by group V sPLA2s and by group IV cPLA2s [6
]. The activity of cPLA2 is regulated by an increase in the intracellular calcium concentration and by phosphorylation events involving MAPKs. Phosphorylation of cPLA2 by ERK1/2 or p38 MAPK at Ser505 has been shown to increase its catalytic activity in vitro [7
, 8
]. However, phosphorylation alone was considered to be insufficient for direct cPLA2 activation without a concomitant increase of the intracellular calcium level [6
, 9
].
Knowledge about a possible interaction of TLR signaling and LT biosynthesis is limited. The TLR2 ligand Pam3CSK4 induced LTC4 formation directly in mouse mast cells, which was shown to be regulated by sPLA2 mediating extended ERK1/2 and cPLA2 phosphorylation [10
]. In contrast, TLR2 ligands failed to stimulate LT formation directly in human mast cells and human neutrophils [11
, 12
]. Other PAMPs such as PGN or mannan, however, were found to induce AA and LTB4 release directly in human neutrophils, but involvement of TLR2 herein remained uncertain [12
]. Others demonstrated that PGN and the TLR2 activator zymosan induce LTC4 release in human mast cells, but the mechanism was not described [11
]. Recently, the TLR7/8 ligand resiquimod was shown to prime human neutrophils for enhanced LTB4 formation upon stimulation with fMLP, PAF, and ionophore [13
]. Furthermore, also LPS-mediated priming of mouse RAW264.7 macrophages for enhanced AA release and LT formation was linked to TLR4 activation [14
].
In this study, we investigated the influence of TLR signaling on LT biosynthesis in human monocytes. Present results provide detailed insights into the underlying mechanisms of TLR2-mediated enhancement of LT release, which may play a role in monocyte/macrophage-mediated inflammatory disorders such as atherosclerosis.
MATERIALS AND METHODS
Materials
TLR ligands [type A CpG ODN 2216, type B CpG ODN 2006-G5, loxoribine, LPS from Escherichia coli 011:B4 strain, poly(I:C), lipoteichoic acid from Bacillus subtilis, FSL-1, Pam3CSK4], polymyxin B, and polyclonal anti-hTLR2 antibody were purchased from InvivoGen (Toulouse, France). Ca2+-ionophore A23187 and fMLP were from Sigma-Aldrich (Schnelldorf, Germany), and HPLC solvents were from Merck (Darmstadt, Germany). [3H]AA was purchased from Biotrend (Cologne, Germany). Bromoenollactone was from Sigma-Aldrich and DTT from AppliChem GmbH (Darmstadt, Germany). From Calbiochem (Merck, Darmstadt, Germany), we purchased the cPLA2 inhibitor PYR, Mnk1 inhibitor, and KN-62; from Axxora (Lörrach, Germany), we obtained SB203580 and PD98059. Antibodies against p-p38 MAPK, p44/42 MAPK, p-p44/42 MAPK, p-Mnk1, p-cPLA2 (Ser505) and IB were from New England Biolabs (Frankfurt a.M., Germany), and antibodies against p38 MAPK, Mnk1, and cPLA2 were from Santa Cruz (Heidelberg, Germany). Infrared dye-conjugated antibodies (IRDye®) were purchased from LI-COR® Biosciences (Bad Homburg, Germany).
Cells and cell culture
MM6 cells [15
] were maintained in RPMI-1640 medium (PAA Laboratories GmbH, Pasching, Austria) supplemented with 2 mM L-glutamine, 10% FBS (Biochrom AG, Berlin, Germany), 100 µg/mL streptomycin, 100 U/mL penicillin, 1 mM oxaloacetic acid, 1x MEM nonessential amino acids, and 10 µg/mL human insulin (kindly provided by Sanofi-Aventis, Frankfurt a.M., Germany) at 37°C, 5% CO2. MM6 cells were differentiated with 1 ng/mL natural hTGFβ1 (R&D Systems GmbH Wiesbaden, Germany) and 50 nM 1,25-dihydroxyvitamin D3 (Sigma-Aldrich) for 4 days, harvested by centrifugation (200 g, 10 min, room temperature), and washed once in PBS, pH 7.4. hPBMC were freshly isolated from leukocyte concentrates obtained from Municipal Hospital Frankfurt-Hoechst (Germany). In brief, cells were isolated by dextran sedimentation and density centrifugation (LSM 1077 lymphocyte separation medium, PAA Laboratories GmbH). hPBMC were washed twice with PBS, pH 7.4, and the monocyte content was elevated by adherence in RPMI-1640 medium supplemented with 25% human plasma for 3 h (37°C, 5% CO2). Cells were resuspended in PBS, pH 7.4, and harvested by centrifugation (200 g, 10 min, room temperature). Cell types were determined by FACS analysis: CD14+ monocytes, 30.2–52.9%; CD19+ B lymphocytes, 3.6–5.1%; CD3+CD56+ NK cells, 7.5–9.0%; CD3+ T lymphocytes, 25.2–43.8%.
Determination of metabolites of the 5-LO pathway
MM6 cells (3x106) or hPBMC (5x106) were resuspended in 1 mL PGC buffer (PBS containing 1 mg/mL glucose and 1 mM CaCl2). Cells were primed with TLR ligands as indicated. Then, LT formation was induced by addition of ionophore A23187 or fMLP at the indicated concentrations. After 10 min at 37°C, the reaction was stopped with 1 mL methanol, 30 µL 1 N HCl, 500 µL PBS, and 200 ng prostaglandin B1 as internal standard was added. After centrifugation (800 g, 10 min, room temperature), the samples were applied to C-18 solid-phase extraction columns (Clean-up® extraction columns from UCT, Bristol, PA, USA), preconditioned with 1 mL methanol and 1 mL water. The columns were washed with 1 mL water and 1 mL methanol 75% (v/v). Subsequently, metabolites of the 5-LO pathway were eluted with 300 µL methanol, diluted with 120 µL water, and analyzed by HPLC as described [16
] using a C-18 Radial-Pak column (Waters, GMbh, Eschborn, Germany). This analytical system detects LTB4, its all-trans isomers, and 5-HETE. 5-LO activity is expressed as ng LTs and 5-HETE/106 cells.
Determination of [3H]AA release
MM6 cells were resuspended at 2 x 106/mL in prewarmed RPMI-1640 medium containing 0.5 µCi/mL [3H]AA (specific activity, 7.4 MBq/µmol) and incubated for 3 h at 37°C, 5% CO2. Thereafter, cells were collected by centrifugation (1000 g, 10 min, room temperature), washed with PBS containing 2 mg/mL low endotoxin FAF-BSA (Sigma-Aldrich; 
0.1 ng/mg endotoxin according to the manufacturer’s specification) to remove unincorporated [3H]AA, and resuspended in PGC buffer/FAF-BSA at 3 x 106/0.5 mL or 2 x 105/0.2 mL. After labeling, MM6 cells were incubated with TLR ligands as indicated. [3H]AA release was induced by the addition of ionophore A23187, 2.5 µM for 10 min at 37°C unless otherwise noted. To stop the reaction, samples were placed on ice. After centrifugation (200 g, 5 min, 4°C), 100 µL of the supernatants was mixed with 3 mL of a liquid scintillation cocktail (Optiphase Hisafe 3, Perkin Elmer, Shelton, CT, USA) and analyzed with the Wallac 1409 liquid scintillation counter (Perkin Elmer). [3H]AA release is expressed as decays per minute/106 cells.
Preparation of cell lysates for analysis of protein phosphorylation
MM6 cells (3x106) were resuspended in 100 µL PGC buffer. For time course analyses, cells were stimulated with TLR ligands for different periods of time at 37°C as indicated. Alternatively, MM6 cells were stimulated with TLR ligands for 15 min at 37°C. Incubations were stopped by addition of 100 µL 2x SDS-PAGE sample loading buffer [SDS-b; 20 mM Tris/HCl, pH 8, 2 mM EDTA, 5% SDS (w/v), 10% β-ME] and placed on ice. Samples were heated at 95°C for 6 min and sonified subsequently (10 s).
SDS-PAGE and immunoblotting
Total cell lysates corresponding to 0.3 x 106 cells in 20 µL were mixed with 3 µL glycerol/0.1% bromphenol blue (1:1, v/v) and analyzed by SDS-PAGE using a Mini Protean® system (Bio-Rad, Munich, Germany) on a 10% gel. After electroblot on a nitrocellulose membrane (Hybond C, Amersham, Piscatway, NJ, USA) and blocking with Odyssey® blocking buffer (LI-COR® Biosciences) for 1 h/room temperature, membranes were incubated with primary antibodies overnight at 4°C. The respective antibody solutions contained one antibody specific for the phosphorylated form of the investigated protein, together with a second antibody, detecting both forms of the protein (derived from different species). Membranes were washed with PBS, pH 7.4/Tween 0.1% (v/v), and incubated with infrared dye-conjugated secondary antibodies (IRDye®, LI-COR® Biosciences) for 45 min/room temperature under protection from light. Two differentially labeled secondary antibodies (emission wavelength, 680 nm and 800 nm, respectively) were used for simultaneous detection of the phosphorylated form and the total amount of the investigated protein. Membranes were washed with PBS, pH 7.4/Tween 0.1% (v/v), and finally with PBS, pH 7.4. Visualization and quantitative analysis of protein bands were carried out with the Odyssey® infrared imaging system (LI-COR® Biosciences). For quantification, the intensities of bands representing the phosphorylated protein were corrected by the band intensities of the total protein. Enhancement of phosphorylation was calculated by relating the band intensities received after cell stimulation to band intensities derived from unstimulated cell samples.
Statistics
Results are given as mean + SD; n 
3. Statistical analysis was carried out by Student’s unpaired t-test (one-tailed). Differences were considered as significant for P < 0.05 (indicated as *, P<0.05, or **, P<0.01).
RESULTS
TLR2 ligands lead to enhanced LT and 5-HETE formation in human monocytes
To evaluate the effects of TLR ligands on LT biosynthesis in human monocytes, we first screened several activators of different TLR subtypes for their impact on LT and 5-HETE formation in MM6 cells. Preincubation with TLR4 ligand LPS, as well as with TLR2 ligands, but with none of the other TLR activators, led to an approximate twofold increase of ionophore-induced LT and 5-HETE formation (Fig. 1A
).
 View larger version (37K): [in a new window] |
Figure 1. LT and 5-HETE formation in human monocytes after priming with TLR ligands. (A) Differentiated MM6 (3x106 cells) were primed with TLR ligand or solvent as indicated 20 min/37°C [CpG 2216, 1 µM; CpG 2006-G5, 1 µM, TLR9; Loxoribine, 100 µM, TLR7; LPS, 10 µg/mL, TLR4; poly(I:C), 25 µg/mL, TLR3; LTA, 2 µg/mL, TLR2; FSL-1, 1 µg/mL, TLR2/6; Pam3CSK4, 1 µg/mL, TLR2/1] before stimulation of LT and 5-HETE formation with ionophore A23187, 2.5 µM for another 10 min/37°C. 5-LO metabolites were determined by HPLC as described in Materials and Methods. (B, i–iii) Differentiated MM6 cells were primed with different concentrations of TLR2 ligand or solvent as indicated 20 min/37°C before induction of LT and 5-HETE formation with ionophore A23187, 2.5 µM/another 10 min/37°C. (C, i–iii) Differentiated MM6 cells were primed with TLR2 ligands or solvent for different periods of time (LTA, 0.5 µg/mL; FSL-1, 0.5 µg/mL; Pam3CSK4, 5 µg/mL) before induction of LT and 5-HETE formation with ionophore A23187, 2.5 µM/another 10 min/37°C. (D) hPBMC (5x106 cells) were primed with TLR2 ligands (LTA, 0.5 µg/mL; FSL-1, 0.5 µg/mL; Pam3CSK4, 5 µg/mL) for 15 min/37°C. Then LT and 5-HETE formation was induced by addition of ionophore A23187 (iono), 0.025 µM, or fMLP, 1 µM, for another 10 min/37°C.
|
Furthermore, cells were preincubated with TLR2 ligands for different periods of time before ionophore stimulation (Fig. 1C , i–iii). All ligands showed comparable characteristics. The maximum response was obtained with a preincubation time of 15 min. However, the effect disappeared when preincubation was extended further up to 90 min. No enhanced LT formation was detectable when MM6 cells were treated with TLR2 ligands and ionophore simultaneously, indicating that TLR2 ligands seem to act as priming agents on human MM6 cells for an enhanced response to ionophore stimulation but do not act as direct costimuli.
To exclude cell line-specific effects, human primary cells were tested under analogous conditions (Fig. 1D) . We primed hPBMC with TLR2 ligands followed by stimulation with calcium ionophore. A significant increase of LTs and 5-HETE was detected when hPBMC were stimulated with 0.025 µM ionophore after priming. The use of high concentrations of calcium ionophore (2.5 µM) led to substantial LT and 5-HETE formation, which was not enhanced significantly by TLR2 ligands (data not shown). Additionally, fMLP at 1 µM was used as a physiological stimulus. In analogy, significant enhancement of LT and 5-HETE release was detected after fMLP stimulation. Taken together, TLR2 ligands showed a similar impact on hPBMC as on MM6 cells, particularly also upon fMLP activation.
TLR2 ligands enhance AA release in MM6 cells in a time-dependent manner
Next, we investigated whether enhanced TLR2 ligand-mediated LT formation is a result of an increased availability of AA in MM6 cells. Time course analyses were carried out as described above, now determining AA release (Fig. 2A
, i–iii). The results were similar to the time course of LT and 5-HETE formation. Simultaneous addition of TLR2 ligands with ionophore did not evoke an enhancement of AA release. Priming of the cells with TLR2 ligands, however, led to a time-dependent increase in free AA, peaking after preincubation of 10–15 min. For all ligands, the kinetics of enhanced AA generation was comparable, although the response upon Pam3CSK4 priming was slightly weaker.
 View larger version (16K): [in a new window] |
Figure 2. AA release from MM6 cells after priming with TLR2 ligands. TLR2 ligands: LTA, 0.5 µg/mL; FSL-1, 0.5 µg/mL; Pam3CSK4, 5 µg/mL; stimulus ionophore A23187, 2.5 µM; differentiated MM6 cells (2x106 cells/mL) were prelabeled with 0.5 µCi/mL [3H]AA for 3 h/37°C. Unincorporated [3H]AA was removed, and cells were resuspended in PGC buffer containing 2 mg/mL FAF-BSA (3x106/0.5 ml). (A, i–iii) [3H]AA-labeled MM6 cells were primed with TLR2 ligands or solvent for different periods of time before induction of AA release with ionophore A23187 for another 10 min/37°C. Free [3H]AA was determined as described in Materials and Methods. (B) [3H]AA-labeled MM6 cells were incubated with solvent, ionophore A23187, or TLR2 ligands for 10 or 20 min/37°C, and free [3H]AA was analyzed.
|
In summary, increased LT formation in monocytic cells seems to be a result of enhanced release of AA.
TLR2 ligand-mediated enhancement of AA release in MM6 cells depends on enhanced activity of cPLA2
AA release is catalyzed preferentially by group V sPLA2 and by group IV cPLA2 [6
]. To identify the PLA2 enzymes that play a role in TLR2 ligand-mediated priming, MM6 cells were treated with iPLA2 inhibitor bromoenollactone (BEL), the sPLA2 inhibitor dithiotreitol (DTT), or with the selective cPLA2 inhibitor pyrrolidine-1 (PYR) before priming and subsequent stimulation. As expected, BEL virtually did not repress AA release, in unprimed nor in primed MM6 cells, indicating an insignificant role of iPLA2 for AA release under these conditions (Fig. 3A
). Inhibition of sPLA2 diminished AA release considerably in primed as well as in unprimed cells. However, the magnitude of the TLR2 ligand-mediated priming effect remained constant despite elimination of sPLA2 activity. Inhibition of cPLA2 prevented AA release almost completely upon ionophore stimulation in MM6 cells, pointing to a crucial role of this enzyme for AA generation under these conditions. The residual PLA2 activity was very low, and the TLR2 ligand-mediated increase was not detectable anymore.
 View larger version (42K): [in a new window] |
Figure 3. Influence of PLA2 inhibitors on AA release from MM6 cells primed with TLR2 ligands. TLR2 ligands: LTA, 0.5 µg/mL; FSL-1, 0.5 µg/mL; Pam3CSK4, 5 µg/mL; stimulus, ionophore A23187, 2.5 µM. PLA2 inhibitors: BEL, 10 µM; DTT, 1 mM; PYR, 10 µM; differentiated MM6 (2x106 cells) were prelabeled with 0.5 µCi/mL [3H]AA for 3 h/37°C. Unincorporated [3H]AA was removed, and cells were resuspended in PGC buffer containing 2 mg/mL FAF-BSA. [3H]AA-labeled MM6 cells were pretreated with or without PLA2 inhibitors as indicated 15 min/37°C. Subsequently, cells were primed with TLR2 ligands or treated with solvent 15 min/37°C and thereupon stimulated for AA release with A23187 for another 10 min/37°C.
|
TLR2 ligands induce time-dependent activation of p38 MAPK and ERK1/2 in MM6 cells
Subsequently, we intended to elucidate the mechanisms of TLR2 ligand priming for enhanced cPLA2 activity. ERK1/2 and p38 MAPK have been shown to phosphorylate cPLA2 at Ser505, leading to enhanced cPLA2 activity in vitro [6
].
By immunoblotting, we analyzed activation of p38 MAPK in MM6 cell lysates upon incubation with TLR2 ligands. An increase of phosphorylated p38 MAPK was detectable for each of the three TLR2 ligands (Fig. 4A , i–iii). The ligand LTA significantly induced p-p38 MAPK two- to fourfold during 15 min of incubation in comparison with unstimulated cells, whereas FSL-1 and Pam3CSK4 led to a 1.5- to 3.5-fold and to a 1.5- to 2.5-fold increase, respectively. Activation of p38 MAPK was dependent on the incubation time. Time courses of p38 MAPK phosphorylation showed similar characteristics for each of the three TLR2 ligands with a maximal activation between 5 min and 15 min, followed by a decline to the initial level.
 View larger version (51K): [in a new window] |
Figure 4. Effect of TLR2 ligands on the activation of MAPK pathways in MM6 cells. TLR2 ligands: LTA, 0.5 µg/mL; FSL-1, 0.5 µg/mL; Pam3CSK4, 5 µg/mL. (A, i–iii, and B, i–iii) Differentiated MM6 (3x106 cells) were incubated with TLR2 ligands for different periods of time/37°C as indicated. As negative control, cells were treated with solvent for 15 min/37°C (solv 15). Incubations were stopped by boiling samples with SDS-PAGE loading buffer. Total cell lysates were prepared and then analyzed by SDS-PAGE and immunoblotting as described in Materials and Methods. To detect p38 MAPK or ERK1/2 activation, antibodies specific for the dually phosphorylated forms of the proteins were used, respectively. Visualization of protein bands was carried out by the use of infrared dye-conjugated antibodies (IRDye®), and analysis was performed with the Odyssey® infrared imaging system (LI-COR® Biosciences) as described in Materials and Methods. As loading control, the total amount of the proteins of interest (phosphorylated and unphosphorylated form) was determined. For quantification, the band intensities of the phosphorylated proteins were corrected by the band intensities of both forms. Phosphorylation levels are related to the negative control (dotted lines). (C) Differentiated MM6 cells (3x106cells) were treated with or without SB203580 (10 µM) for 15 min/37°C, before stimulation with or without TLR2 ligands for another 15 min/37°C. After termination of the incubation, total cell lysates were analyzed by SDS-PAGE and immunoblotting as described. (D) Differentiated MM6 cells (3x106 cells) were incubated with ionophore A23187, 5 µM (positive control), solvent (negative control), or TLR2 ligands for 15 min/37°C. After termination of the incubation, total cell lysates were analyzed by SDS-PAGE and immunoblotting as described. Mnk1 activation was determined by using an antibody specific for the dually phosphorylated form of the protein.
|
14-fold increase after 5 min and was abrogated almost completely after 10 min. We also investigated TLR2-induced ERK1/2 activation in the presence of SB203580, which selectively blocks p38 MAPK signaling (Fig. 4C) . ERK1/2 phosphorylation was more prominent when p38 MAPK signaling was blocked. Pam3CSK4 again led to a weaker response in comparison with LTA and FSL-1. In summary, both MAPK pathways were activated by TLR2 ligands.
As p38 MAPK-activated protein kinase Mnk1 is also known to influence cPLA2 activity by Ser727 phosphorylation [6
], we investigated Mnk1 activation. We observed that Mnk1, which is a downstream target of p38 MAPK and ERK1/2, was also activated by the TLR2 ligands (Fig. 4D)
. The ligands LTA and FSL-1 induced Mnk1 phosphorylation to a comparable extent, whereas Pam3CSK4 again displayed weaker activity.
TLR2 ligand-induced cPLA2 phosphorylation at Ser505 via p38 MAPK and ERK1/2 is the critical step for enhanced AA release in MM6 cells
Next, we addressed the question of whether TLR2 ligands induce cPLA2 phosphorylation at Ser505. We analyzed MM6 cell lysates for p-cPLA2 (Ser505) after treatment with the TLR2 activators. Simultaneously, we tested activation of the TLR2 signaling pathway by analysis of IB degradation. Each ligand evoked time-dependent phosphorylation of cPLA2 (Fig. 5A
). Quantitative analysis of p-cPLA2 signals after LTA treatment revealed an up to twofold increase of cPLA2 phosphorylation. Extended incubation led to a decrease of p-cPLA2 after 60–90 min. Conversely, LTA and FSL-1 initially led to IB degradation over time with a recovery of IB protein after 60–90 min. In contrast, no definite IB degradation could be determined after Pam3CSK4 treatment, indicating a poor activation of the TLR2 signaling pathway in comparison with the other ligands.
 View larger version (52K): [in a new window] |
Figure 5. Influence of kinase inhibitors on the effect of TLR2 ligands on cPLA2 phosphorylation and activity. TLR2 ligands: LTA, 0.5 µg/mL; FSL-1, 0.5 µg/mL; Pam3CSK4, 5 µg/mL. (A) Differentiated MM6 (3x106 cells) were incubated with TLR2 ligands for different periods of time/37°C as indicated, or cells were treated with solvent for 15 min/37°C. Incubations were stopped by boiling samples with SDS-PAGE loading buffer. Total cell lysates were analyzed by SDS-PAGE and immunoblotting. To detect cPLA2 phosphorylation by p38 MAPK and ERK1/2, an antibody specific for phosphorylated cPLA2 (Ser505) was used. As a control for the activation of TLR signaling pathways, IB degradation was analyzed over time. Visualization of protein bands was carried out by the use of infrared dye-conjugated antibodies as described. Simultaneous detection of total cPLA2 served as loading control. For quantification, the band intensities of phosphorylated cPLA2 were corrected by the band intensities of the total cPLA2 protein. Phosphorylation levels are related to the negative control. (B) Differentiated MM6 (3x106 cells) were treated with SB203580 (SB), 10 µM, and PD98059 (PD), 50 µM, with both inhibitors or without inhibitor as indicated for 15 min/37°C. Then, cells were stimulated with or without TLR2 ligands for another 15 min/37°C. After termination of the incubation, total cell lysates were analyzed by SDS-PAGE and immunoblotting. The increase of cPLA2 phosphorylation by TLR2 ligands is related to the cPLA2 phosphorylation level without TLR activation (dotted line), respectively. (C and D) Differentiated MM6 (2x106 cells) were prelabeled with 0.5 µCi/mL [3H]AA for 3 h/37°C. Unincorporated [3H]AA was removed, and cells were resuspended in PGC buffer containing 2 mg/mL FAF-BSA. Cells were pretreated with SB203580 (10 µM), PD98059 (50 µM), with the combination of both, with Mnk1 inhibitor (Inh), 20 µM, with CaMKII inhibitor KN-62, 10 µM, or without inhibitor as indicated 15 min/37°C. Thereupon, MM6 cells were primed with TLR2 ligands or treated with solvent 15 min/37°C and subsequently stimulated for AA release with A23187 for another 10 min/37°C. The increase of AA release by TLR2 ligands is related to the AA release level without TLR2 ligand priming (dotted line), respectively.
|
(Ser505) was linked to TLR2 ligand-triggered activation of p38 MAPK and ERK1/2. We compared the increase of p-cPLA2 in stimulated and unstimulated cells in the presence or absence of SB203580 or of PD98059, which inhibits ERK1/2 signaling selectively (Fig. 5B)
. In line with the results shown in Figure 5A
, we found an up to twofold increase of cPLA2 phosphorylation upon TLR2 ligand stimulation. In the presence of SB203580 or PD98059, however, LTA and also FSL-1-mediated increase of p-cPLA2 was reduced, but neither of the two inhibitors abolished the increase completely. Obviously, the two MAPK pathways were functionally redundant concerning cPLA2 phosphorylation. Only the presence of both inhibitors led to a total prevention of p-cPLA2 increase, which was significant for each of the three TLR2 ligands. Hence, both MAPK signaling pathways seem to play a role in TLR2 ligand-induced cPLA2 phosphorylation at Ser505. Control experiments analyzing IB protein confirm the activation of TLR signaling pathways by LTA, FSL-1, and essentially, to a weaker extent by Pam3CSK4.
By means of analogous inhibitor studies, we determined if MAPK-mediated phosphorylation of cPLA2 accounts for induced AA release after TLR2 ligand priming. The results (Fig. 5C)
are in accordance with the data about cPLA2 phosphorylation (Fig. 5B)
. Only a combined use of both inhibitors blocked TLR2 ligand effects completely. Thus, activation of two MAPK pathways and subsequent phosphorylation of cPLA2 seem to account for the TLR2 ligand-induced enhancement of ionophore-stimulated AA release.
Apart from p38 MAPK and ERK1/2, two more protein kinases have been suggested to regulate cPLA2 activity by phosphorylation [6
]. Mnk1 accounts for cPLA2 phosphorylation at Ser727, and CamKII was found to phosphorylate cPLA2 on Ser515. We used specific inhibitors for Mnk1 and CamKII, respectively, to investigate the role of these two cPLA2 kinases. However, neither of the two inhibitors showed a detectable impact on the effects of the TLR2 ligands (Fig. 5D)
.
Enhancement of AA release from MM6 cells by LTA, FSL-1, and Pam3CSK4 is dependent on TLR2
Finally, we checked whether the enhancing effect of LTA, FSL-1, and Pam3CSK4 on AA release is mediated by TLR2 activation. A neutralizing anti-hTLR2 antibody clearly inhibited the TLR2 ligand-induced priming effects and blocked LTA-, FSL-1-, and Pam3CSK4-induced enhancement of AA release almost completely (Fig. 6A
). In additional experiments, we excluded that the augmentation of AA and LT formation was triggered by LPS contaminations activating TLR4. The addition of LPS-binding reagent polymyxin B blocked the LPS-mediated increase of AA release and LT and 5-HETE formation significantly in MM6 cells. The TLR2 ligand-induced increase, however, was not affected in the presence of polymyxin B (Fig. 6 B and C)
. Hence, priming of MM6 cells by LTA, FSL-1, and Pam3CSK4 for enhanced AA release is mediated by TLR2.
 View larger version (26K): [in a new window] |
Figure 6. Influence of a TLR2-neutralizing antibody and of polymyxin B on TLR2 ligand priming of MM6 cells. TLR2 ligands: LTA, 0.5 µg/mL; FSL-1, 0.5 µg/mL; Pam3CSK4, 5 µg/mL; TLR4 ligand: LPS, 10 µg/mL; stimulus: ionophore A23187, 2.5 µM; the increase of AA release or LT and 5-HETE formation by TLR ligands is related to the response level without TLR ligand priming (dotted line), respectively. (A and B) Differentiated MM6 (2x106) were prelabeled with 0.5 µCi/mL [3H]AA for 3 h/37°C. Unincorporated [3H]AA was removed, and cells were resuspended in PGC buffer containing 1 mg/mL (A) or 2 mg/mL (B) FAF-BSA. [3H]AA-labeled MM6 cells were pretreated with anti-hTLR2 antibody (TLR2 Aby), 250 µg/mL (A), with polymyxin B, 100 µg/mL (B), or with solvent 15 min/37°C. Subsequently, cells were primed with TLR ligands or solvent as indicated 15 min/37°C and then stimulated for AA release with A23187 for another 10 min/37°C. (C) Differentiated MM6 cells (3x106) were preincubated with or without polymyxin B, 100 µg/mL, 15 min/37°C. After priming with TLR ligands or treatment with solvent 15 min/37°C, cells were stimulated with A23187 for another 10 min/37°C.
|
In this article, we report that the TLR2 ligands LTA, FSL-1, and Pam3CSK4 lead to an enhancement of ionophore- or fMLP-induced LT formation in human monocytes and provide detailed information about the underlying mechanisms.
In screening experiments, we show first that only ligands for TLR2 (LTA), for TLR2/6 (FSL-1), for TLR2/1 (Pam3CSK4), and also TLR4 activator LPS display enhancing effects on LT and 5-HETE formation in MM6 cells. Regarding TLR mRNA expression patterns in cellular subsets of hPBMC, it is known that high expression of TLR2 and also TLR4 is characteristic for monocytes and that TLR1 and TLR6 are expressed in several cell types of hPBMC including monocytes, B cells, pDCs, NK cells, and T cells. TLR7 and TLR3 are absent in monocytes, and monocytes did not respond to TLR9 activators in the absence of pDCs [17 , 18 ]. This expression pattern is compatible with our observations that no response to TLR7 ligand loxoribine, TLR3 ligand poly(I:C), and TLR9 activators CpG ODNs was detected in MM6 cells.
TLR2 ligand-mediated effects on LT biosynthesis were characterized in terms of their concentration dependence. For LTA and FSL-1, we could define an optimal concentration of 0.5 µg/mL. Slightly increasing responses at concentrations higher than 10 µg/mL may be caused by additional less-specific effects and/or effects on low-affinity receptors. For Pam3CSK4, it was reported in studies of lipopeptide-induced B cell activation that a concentration of 80 µg/mL induced an increase of intracellular calcium, and it was speculated that this may be ascribed to ionophore-like membrane interactions [19
]. Furthermore, Pam3CSK4, at concentrations greater than 25 µg/mL, was shown to activate superoxide formation in human neutrophils in a similar way as fMLP [20
]. Therefore, TLR2-dependent effects of Pam3CSK4 on LT formation may be enhanced by additional mechanisms at concentrations >10 µg/mL.
We also found a TLR2 ligand induced enhancement of LT and 5-HETE formation in human primary cells. Our preparations of hPBMC in large parts consisted of monocytes, which are known to express TLR2 highly [17 ]. hPBMC also included proportions of T lymphocytes and minor amounts of B lymphocytes and NK cells as determined by FACS analysis. Whereas T cells are considered to lack 5-LO [4 ], B cells were shown to express 5-LO. It was demonstrated, however, that in B lymphocytes, calcium ionophore is not sufficient to activate 5-LO [21 ]. Our experiments furthermore indicate that human primary NK cell preparations do not seem to synthesize LTs upon stimulation with calcium ionophore (unpublished data). Thus, enhanced LT and 5-HETE synthesis upon TLR2 ligand pretreatment in PBMC can be regarded as a monocytic response. TLR2 ligands enhanced LTs in PBMC, particularly upon activation by low ionophore concentrations and fMLP. This indicates that TLR2 ligands act on human primary monocytic cells in a similar manner as on MM6 cells and that MM6 cells seem to be a valid model to investigate TLR2-mediated LT formation in human monocytes.
Only a few studies that link TLR activation to LT formation and provide detailed information about the underlying mechanisms are available so far. Pam3CSK4-induced, TLR2-dependent LTC4 generation was described in mouse mast cells, which was shown to require cPLA2 and was found to be amplified by sPLA2 [10
]. In human mast cells and in hPMNL, Pam3CSK4 and LTA, LPS, muramyldipeptide, or Pam3CSK4 failed to stimulate LT formation directly, respectively [11
, 12
]. On the other hand, other PAMPs such as PGN or mannan were found to induce AA and LTB4 release directly in hPMNL, and a nonarchetypal involvement of TLR2 or participation of other receptors was suggested [12
]. PGN and TLR2 ligand zymosan induced LTC4 generation in human mast cells, but the mechanism was not described [11
].
Present investigations revealed that preincubation of TLR2 ligands was essential for enhancement of LT formation and that simultaneous addition of TLR2 ligands together with ionophore did not evoke augmented LT and 5-HETE synthesis. This suggests that TLR2 ligands do not stimulate LT biosynthesis directly in MM6 cells. Thus, TLR2 ligands rather seemed to prime human monocytes synergistically for an enhanced response upon stimulation.
Such priming effects for increased LT generation have been observed and investigated in numerous studies. Growth factors, cytokines, phorbol esters, LPS, and EBV have been identified and characterized as priming agents [22
]. GM-CSF and TNF- priming has been studied widely in neutrophils [23
24
25
26
27
28
], and it has been implicated in an enhanced endogenous generation of AA [23
, 24
, 26
, 27
]. Furthermore, GM-CSF has been shown to enhance LT synthesis, also when exogenous substrate was used to circumvent cPLA2 activity [22
, 29
]. As confirmed in our first screening experiment, LPS also exerts priming effects on LT formation in leukocytes, which was shown already in several studies reporting enhanced AA release, augmented LT formation [30
31
32
], and also phosphorylation of cPLA2 [30
]. Additionally, it was demonstrated that in neutrophils, LPS priming before fMLP stimulation evoked redistribution of cPLA2 and also 5-LO to the nuclear envelope leading to an increased capacity of LT biosynthesis [30
]. In another report, PMA, as a priming agent in MM6 cells and PMNL, was shown to up-regulate LT formation by increasing the capacity for 5-LO phosphorylation via activation of 5-LO kinases and for 5-LO translocation to the nuclear envelope [33
].
In recent investigations, priming for LT generation for the first time was linked to TLR activation [13
, 14
]. TLR7/8 ligand resiquimod was shown to act as a priming agent on human neutrophils via TLR8 activation, leading to enhanced LTB4 formation upon stimulation with fMLP, PAF, and ionophore. Resiquimod enhanced AA release following fMLP stimulation, induced cPLA2 phosphorylation, and promoted 5-LO translocation [13
]. Furthermore, Buczynski et al. [14
] suggested a PIP2-mediated increase of cPLA2 affinity to the membrane, accompanied by a PIP2-mediated increase of cPLA2 activity, together with a yet-undefined augmentation of 5-LO activity as an explanation for LPS-mediated priming of mouse RAW264.7 macrophages for enhanced AA release and LT formation.
In the present investigations, TLR2 ligands mediated an increase in ionophore-induced release of AA but not a direct stimulation of PLA2 activity. Furthermore, the effects on AA release and LT and 5-HETE synthesis were similar regarding time courses and intensity (two- to fivefold). The maximal effect of TLR2 ligands on AA release occurred somewhat earlier (10–15 min) than on LT and 5-HETE formation (15–20 min), probably reflecting the fact that the release of AA must occur prior to LT synthesis. Taken together, the results suggest that the ligand effect on AA release is the critical step for the enhancement of LT formation. Additionally, TLR2 ligands may have an enhancing influence on 5-LO activity itself or also on 5-LO translocation, but this has yet to be verified. Further studies revealed that the sPLA2 inhibitor DTT led to a considerable reduction of AA release in unprimed as well as in primed cells, indicating that sPLA2 is involved in AA release in ionophore-stimulated MM6 cells. The priming effect, though, was hardly affected by DTT, which indicates that essentially, cPLA2 was the PLA2 isoform involved under these conditions. Moreover, cPLA2 activity was found to be crucial, as its inhibition prevented AA release almost completely upon ionophore stimulation, and no appreciable sPLA2 activity remained detectable in the presence of the cPLA2 inhibitor. These data are consistent with reports demonstrating that activation of sPLA2 is dependent on previous AA release by cPLA2 in LPS-primed, PAF-stimulated P388D1 macrophages [34
, 35
].
Several studies have demonstrated that the activation of cPLA2 is not only regulated by Ca2+ but also by phosphorylation [6
, 36
]. It has been shown that phosphorylation at Ser505 of cPLA2 by p42 or p38 MAPK leads to a two- to threefold increase in catalytic activity in vitro [7
, 8
]. It was also suggested that cPLA2 activity is up-regulated in vivo via phosphorylation of Ser727 by Mnk1 [37
]. As a third serine residue playing a role for enzyme activity, Ser515 of cPLA2 was found to be phosphorylated by rat brain CaMKII in vitro and in norepinephrine-stimulated human vascular smooth muscle cells [38
]. In general, phosphorylation increases cPLA2 activation, but phosphorylation alone is considered to be insufficient for cPLA2-mediated cellular AA release, and a concomitant increase in the intracellular calcium level is required [6
, 9
].
TLR stimulation has been shown to result in activation of MAPKs (ERK, JNK, p38) [39
]. Here, we demonstrate that activation of ERKs and p38 MAPK is detectable also in MM6 cells challenged with TLR2 ligands, moreover, occurring within the time-frame of enhanced AA release and LT formation. Furthermore, time-dependent cPLA2 phosphorylation at Ser505 was evoked, which coincides with the intracellular release of AA. Maximal activation of both MAPKs appeared fast within 5 min and thus, seems to precede cPLA2 phosphorylation, AA generation, and finally, LT biosynthesis. In comparison with p38 MAPK activation (maximal increase approximately threefold), p42 phosphorylation was induced to a greater extent (12-fold). Phosphorylation of p42 was more transient and displayed slightly different characteristics depending on the respective TLR2 ligand. Combined use of a p38 MAPK and an ERK1/2 inhibitor led to complete prevention of a TLR2 ligand-induced increase of p-cPLA2 (Ser505) as well as of TLR2 ligand-mediated enhancement of AA release. This indicates that cPLA2 phosphorylation via both MAPK signaling pathways is crucial for the increase in cPLA2 activity by TLR2 ligand priming. Within this context, we note that in RAW264.7 macrophages, a time-dependent p38 MAPK and ERK1/2 activation and subsequent cPLA2 phosphorylation via TLR4 were shown to stimulate AA generation directly without additional calcium activation [40
]. Although Mnk1 activation (presumably via p38 and ERK1/2) was detected in the present studies upon TLR2 ligand treatment, a Mnk1 inhibitor and the CamKII inhibitor KN-62 did not affect enhanced AA release under priming conditions. Therefore, we suggest that phosphorylation of other serine residues of cPLA2 (Ser727 and Ser515) does not contribute substantially to enhanced cPLA2 activity after priming with TLR2 ligands.
Additionally, we demonstrate that neutralization of TLR2 by the use of an anti-hTLR2 antibody essentially abolished the priming effects of LTA and Pam3CSK4 on AA release, and also, the response to FSL-1 was reduced significantly. Possibly TLR6, as part of the heterodimeric TLR2/6 complex, is responsible for the remaining activity, as it still may be functional and capable of transducing partly an activation by FSL-1 under these conditions. The presence of polymyxin B did not alter the impact of TLR2 ligands on AA and LT release, assuring that TLR4 activation by LPS contaminations does not contribute to the observed effects. These data clearly confirm that the ligand effects are dependent on TLR2 activation.
In summary, our studies provide a detailed characterization of the TLR2-mediated regulation of AA release and subsequent LT formation in human monocytic cells. We have shown that TLR2 activation leads to enhanced LT formation, which links PAMP-mediated innate immune responses to the generation of lipid mediators that play a central role in the generation of inflammatory responses and in cardiovascular disease.
ACKNOWLEDGEMENTS
This work was supported by grants from the Deutsche Forschungsgemeinschaft (DFG/GRK 1172) and EC (LSHM-CT-2004–005033). This publication reflects only the authors’ views. The European Commission is not liable for any use that may be made of information herein. We thank Sven George for expert technical assistance.
FOOTNOTES
Abbreviations: 5-HETE=5(S)-hydroxy-6-trans-8,11,14-cis-eicosatetraenoic acid, 5-LO=5-lipoxygenase, AA=arachidonic acid, BEL=bromoenollactone, CamKII=Ca2+/calmodulin-dependent kinase II, cPLA2=cytosolic PLA2, FAF=fatty ODN=oligodeoxynucleotide acid free, FSL-1-2,3-Bis (palmitoyloxy)propyl-CGDPKHPKSF, h=human, iPLA2=independent PLA2, LT=leukotriene, LTA=lipoteichoic acid, MM6=Mono Mac 6, ODN=oligodeoxynucleotide, p=phospho, PAF=platelet-activating factor, Pam3CSK4=N-Palmitoyl-2,3-bis(palmitoyloxy)propyl-CSKKKK, PAMP=pathogen-associated molecular pattern, pDC=plasmacytoid dendritic cell, PGN=peptidoglycan, PIP2=phosphatidylinositol 4,5-bisphosphate, PLA2=phospholipase A2, PMNL=polymorphonuclear leukocyte, poly(I:C)=polyinosinic:polycytidylic acid, PYR=pyrrolidine-1, sPLA2=secretory PLA2
Received October 4, 2008; revised February 25, 2009; accepted March 23, 2009.
REFERENCES
activation Blood 110,561-567 priming of phospholipase A2 activation in human neutrophils. An alternative mechanism of priming J. Immunol. 146,1277-1285[Abstract] and low dose diacylglycerol Int. J. Biochem. Cell Biol. 28,771-776[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
