Lipopolysaccharide-induced c-Jun NH2-terminal kinase activation in human neutrophils: role of phosphatidylinositol 3-Kinase and Syk-mediated pathways.

Polymorphonuclear leukocytes (neutrophils) respond to lipopolysaccharide (LPS) through the up-regulation of several pro-inflammatory mediators. We have recently shown that LPS-stimulated neutrophils express monocyte chemoattractant protein 1 (MCP-1), an AP-1-dependent gene, suggesting that LPS activates the c-Jun N-terminal kinase (JNK) pathway in neutrophils. Previously, we have shown the activation of p38 MAPK, but not JNK, in suspended neutrophils stimulated with LPS but have recently shown activation of JNK by TNF-alpha in an adherent neutrophil system. We show here that exposure to LPS activates JNK in non-suspended neutrophils and that LPS-induced MCP-1 expression, but not tumor necrosis factor-alpha (TNF-alpha) or interleukin-8 (IL-8), is dependent on JNK activation. In addition, LPS stimulation of non-suspended neutrophils activates Syk and phosphatidylinositol 3-kinase (PI3K). Inhibition of Syk with piceatannol or PI3K with wortmannin inhibited LPS-induced JNK activation and decreased MCP-1 expression after exposure to LPS, suggesting that both Syk and PI3K reside in a signaling pathway leading to LPS-induced JNK activation in neutrophils. This Syk- and PI3K-dependent pathway leading to JNK activation after LPS exposure in non-suspended neutrophils is specific for JNK, because inhibition of neither Syk nor PI3K decreased p38 activation after LPS stimulation. Furthermore we show that PI3K inhibition decreased LPS-induced Syk activation suggesting that PI3K resides upstream of Syk in this pathway. Finally, we show that Syk associates with Toll-like receptor 4 (TLR4) upon LPS stimulation further implicating Syk in the LPS-induced signaling pathway in neutrophils. Overall our data suggests that LPS induces JNK activation only in non-suspended neutrophils, which proceeds through Syk- and PI3K-dependent pathways, and that JNK activation is important for LPS-induced MCP-1 expression but not for TNF-alpha or IL-8 expression.

nition of bacterial products, such as lipopolysaccharide (LPS), 1 and release of inflammatory mediators and antibacterial products are hallmarks of the involvement of these cells in the pathophysiology of inflammatory conditions, such as the sepsis syndrome and acute lung injury (1)(2)(3)(4)(5)(6)(7)(8). An improved understanding by which neutrophils respond to LPS, and more specifically the signaling pathways utilized in induction of an inflammatory response, may facilitate novel therapies in the treatment of LPS-induced inflammatory diseases.
A variety of signaling pathways have been elucidated in the response to LPS. The best described of which is the pathway leading to NFB activation through Toll-like receptor 4 (TLR4). Binding of LPS, in the context of LPS-binding protein and CD14, leads to the association of MyD88 with TLR4, the subsequent recruitment and activation of the interleukin-1 receptor-associated kinase to the TLR4-MyD88 complex, and the association of interleukin-1 receptor-associated kinase with TNF receptor activating factor 6 (9 -11). Although the NFB pathway is of undoubted importance, the role of MAPK pathways in the response to LPS is less certain. The MAPK family consists of three members: p42/44 ERK, which is activated by chemoattractants and growth factors, and p38 and c-Jun NH 2terminal kinase (JNK), which are stress-and cytokine-activated (12). MAPKs are involved in perpetuation of the inflammatory response and are activated through well described signaling cascades (13). Whether similar upstream components as described above, or alternatively some other combination of intermediates, lead to LPS-induced activation of MAPKs remains a matter of some dispute.
Activation of the MAPKs occur after LPS stimulation in MyD88 knockout mice, although with delayed kinetics compared with wild type mice, suggesting that MyD88 is involved, but not essential, in LPS-induced MAPK activation and that other MyD88-independent pathways exist (14). Recently an adaptor molecule homologous to MyD88, Mal/Tirap, has been shown to be involved in the activation of the MAPKs by TLR4 (15,16). In a fashion similar to that of MyD88 knockout mice, Tirap knockout mice demonstrate delayed activation of the MAPKs after LPS exposure (17,18). These findings suggest the presence of pathways independent of MyD88 and Mal/Tirap, which lead to the activation of the MAPKs after LPS stimulation. Recently, PI3K and the small G-proteins Cdc42 and Rac1, which are involved in the activation of MAPKs (19 -31), particularly JNK (26 -30, 32-35), have been demonstrated to associate with TLR2 in response to heat killed Staphylococcus aureus (36). Whether PI3K and/or the small G-proteins might be involved in TLR4-dependent activation of the MAPKs after LPS stimulation, and what the functional consequences of such activation, is unknown.
The JNK subfamily of MAPK has been shown to be involved in a variety of cellular functions. Activation of JNK has been linked to induction of apoptosis, as seen after TNF-␣ (37) and IL-1␤ stimulation (38), and has been suggested to be important in the transcriptional regulation of several inflammatory mediators, including IL-2 (39), Cox-2 (40), metalloproteinase 1 (39), vascular endothelial growth factor (41), and monocyte chemoattractant protein 1 (MCP-1) (39). JNK is activated upon exposure to osmotic stress (42), irradiation (43), growth factors (44), heat shock (42), IL-1␤ (39), and TNF-␣ (37,42,45). In addition, JNK activation has been shown to occur in macrophages upon LPS stimulation (46 -48), however, the activation of JNK in LPS-stimulated neutrophils has not been previously described. We (49,50) and others (51) have reported that LPS stimulation of suspended neutrophils results in the activation of p38 MAPK, but not ERK or JNK. The absence of JNK activation in our previous studies may relate to the use of suspended neutrophils, which might not replicate events that occur in a complex microenvironment, because we have recently shown the activation of JNK in adherent neutrophils upon TNF-␣ stimulation (37).
Several pathways leading to JNK activation have been described, which, although stimulus-dependent, display common features. Pathways have been described that depend on the small G-proteins Cdc42 and Rac1 (30,31), PI3K (34,43,44), and recently we have identified Syk as an upstream activator of JNK in adherent neutrophils after TNF-␣ stimulation (37). Activation of JNK after LPS stimulation in macrophages requires PI3K (46), although upstream components in the JNK pathway in neutrophils are unknown. We queried whether activation of Syk and PI3K may also occur in LPS-stimulated neutrophils, and if so whether Syk and PI3K participate in LPS-induced JNK activation.
We show here that LPS stimulation activates JNK in nonsuspended neutrophils and that this activation occurs through Syk-dependent and both PI3K-dependent and -independent pathways. Additionally, we show that LPS stimulates PI3K and Syk activation, that PI3K is involved in the activation of Syk, and that Syk associates with TLR4 upon LPS stimulation. This pathway leading to the activation of JNK in non-suspended neutrophils appears independent of that which leads to p38 activation after LPS stimulation. Finally, we show that inhibition of JNK leads to a relatively specific decrease in LPS-induced MCP-1 expression in neutrophils. These studies provide new insight into mechanisms by which neutrophils respond to LPS.
JNK Kinase Assay-After immunoprecipitation with anti-JNK-1, the beads were washed once in JNK lysis buffer and twice in JNK kinase buffer (20 mM HEPES (pH 7.5), 20 mM ␤-glycerophosphate, 10 mM p-nitrophenyl phosphate, 10 mM MgCl 2 , 1 mM dithiothreitol, and 50 mM Na 3 VO 4 ). To each sample was then added 10 Ci of [␥-32 P]ATP and 500 ng of c-Jun 1-79 in 40 l of the JNK kinase buffer and incubated for 30 min at 30°C. The reaction was terminated by the addition of 15 l 5ϫ Laemmli sample buffer, and the samples were boiled for 8 min. Proteins were separated on 10% SDS-PAGE gels and transferred to nitrocellulose, and radioactive bands were identified by autoradiography and quantified by PhosphorImager analysis (Amersham Biosciences).
Syk and Lyn Kinase Assays-After immunoprecipitation the beads were collected by centrifugation and washed twice in wash buffer A (25 mM HEPES (pH 7.5), 150 mM NaCl, 0.1% Brij 97, 1 mM Na 3 VO 4 , 10 g/ml aprotinin, and 10 g/ml leupeptin) and twice in wash buffer B (25 mM HEPES (pH 7.5), 150 mM NaCl, 1 mM Na 3 VO 4 , 10 g/ml aprotinin, and 10 g/ml leupeptin). The beads and kinase mix (25 mM HEPES (pH 7.5), 2 mM MnCl 2 , 20 mM p-nitrophenyl phosphate, and 25 Ci of [␥-32 P]ATP) were incubated separately at 30°C for 5 min. After incubation, 40 l of the kinase mix was added to the beads, which were then incubated at 30°C for 1 min. Proteins were then separated with radioactive bands identified and analyzed as described above for the JNK kinase assay.
p38 Kinase Assay-The beads were collected by centrifugation after immunoprecipitation with anti-p38 and were washed once in radioimmune precipitation assay lysis buffer and twice in PAN (10 mM HEPES, 100 mM NaCl (pH 7.0), and 21 g/ml aprotinin). To the beads were added 40 l of kinase mix (20 mM HEPES (pH 7.6), 20 mM MgCl 2 , 20 mM ATP, 2 mM dithiothreitol, 100 M Na 3 VO 4 , 25 mM ␤-glycerophosphate, 20 Ci of [␥-32 P]ATP, and 500 ng of recombinant ATF-2 1-110 ), which were incubated for 30 min at 30°C. As per the methods for the JNK kinase assay, proteins were separated with radioactive bands identified and analyzed.
PI3K Activity Assay-After immunoprecipitation the samples were centrifuged and the beads were washed once in PI3K lysis buffer and three times in PAN. After washing, the beads were resuspended in PAN with PI3K activity measured as described by Kazlauskas and Cooper (53). Briefly, the beads and kinase mix (20 mM HEPES (pH 7.2), 5 Ci of [␥-32 P]ATP, and 10 M ATP with 0.2 mg/ml phosphatidylinositol 4,5-bisphosphate (Avanti Polar Lipids; Alabaster, AL)) were incubated separately for 10 min at room temperature. The beads were mixed with 10 l of the kinase mix for 15 min at 30°C. The reactions were terminated by the addition of 100 l of 1 M HCl, and the phospholipids were extracted with 200 l of CH 3 Cl/CH 3 OH (1:1). The organic phase was washed with 80 l of HCl/CH 3 OH (1:1), and the solution was lyophilized to dryness. Phospholipids were resuspended in 10 l of CH 3 Cl/CH 3 OH (1:1), spotted onto silica gel 60 plate (20 ϫ 20 cm, VWR Scientific) that had been pre-soaked with 2% sodium-potassium tartrate, and were resolved by ascending thin-layer chromatography in a buffer containing CH 3 Cl/CH 3 OH/4 M NH 4 OH (9:7:2). The radiolabeled lipids were detected by autoradiography and quantified with a PhosphorImager.
Inhibitor Studies-After resuspension in KRPD or RPMI 1640 as described above, neutrophils were incubated with or without piceatannol (10 or 30 M), wortmannin (50 nM), or SP600125 (2-20 M) as indicated at 37°C for the indicated times prior to LPS stimulation. Me 2 SO (0.1%) was used as a diluent control for each experiment.
Statistical Analysis-Data are expressed as means Ϯ S.E. Multiple comparisons were performed by one-way ANOVA with Tukey (post hoc) test for determination of differences between groups. Statistical analysis was also performed by Student's paired t test where indicated. A p value less than 0.05 was considered significant.

LPS Induces the Activation of c-Jun NH 2 -terminal Kinase
(JNK) in Neutrophils-Because we have recently observed increased expression of MCP-1 mRNA upon LPS stimulation in neutrophils (54), and the induction of AP-1, of which c-Jun is one component, is important for the expression of MCP-1 mRNA (55,56), we hypothesized that LPS may induce the activation of JNK. Isolated human neutrophils were incubated under non-suspended conditions for 60 min, then exposed to LPS, followed by immunoprecipitation of JNK1 and an in vitro kinase assay using recombinant c-Jun 1-79 as an exogenous substrate. As can be seen in Fig. 1, LPS induced an increase in JNK activity, as indicated by increased phosphorylation of the c-Jun fragment c-Jun 1-79 as early as 10 min after exposure to LPS, which peaked at 15 min, and was maintained for at least 30 min after LPS stimulation. The LPS-induced activation of JNK occurs more rapidly than compared with the activation of p38 seen previously, which peaks 30 min after stimulation by LPS (49).
The JNK Inhibitor SP600125 Inhibits LPS-induced JNK Activation-To assess the effectiveness of the JNK inhibitor SP600125 on LPS-induced JNK activation in neutrophils, human neutrophils were pretreated for 60 min with increasing concentrations of SP600125 prior to LPS stimulation, followed by an in vitro JNK kinase assay. As can be seen in Fig. 2, prior treatment with SP600125 diminished JNK activation in a dosedependent manner with a complete inhibition of the LPS-induced increase in JNK activation at a concentration of 5 M (Fig. 2). To assess the specificity of SP600125 on LPS-induced JNK activation, and in particular to confirm that SP600125 did not affect activation of p38, the other MAPK activated in LPSstimulated neutrophils, we assessed the LPS-induced activation of p38 in neutrophils pretreated with SP600125 (2-10 M) prior to LPS stimulation. Activation of p38 after LPS stimulation was not diminished by pretreatment with SP600125 at any concentration utilized (data not shown).
Effect of JNK Inhibition on LPS-induced Cytokine and Chemokine Expression-Neutrophils respond to LPS with increased expression of TNF-␣, IL-8, IL-1␤, IL-6, and macrophage inflammatory protein 1␣ (2,3,57). Recently we have shown that in addition to the up-regulation of these pro-inflammatory mediators, neutrophils have increased expression and release of MCP-1 after LPS stimulation (54). The transcriptional regulation of MCP-1 expression has been shown to be controlled primarily through the activation of AP-1, of which c-Jun is a prototypical component (55,56). To examine the downstream effects of JNK inhibition on LPS-induced cytokine and chemokine expression, neutrophils were incubated for 60 min with the JNK inhibitor SP600125 (2-20 M), stimulated with LPS for 4 h, followed by assessment of mRNA expression by RT-PCR and protein release by ELISA. As shown in SP600125 exuded less effect on LPS-induced TNF-␣ and IL-8 mRNA expression than on MCP-1. Pre-treatment with SP600125 prior to LPS exposure completely inhibited TNF-␣ mRNA expression only at the highest concentration utilized (20 M), whereas the expression of IL-8 was even more resistant to the effects of SP600125, with only a partial decrease in the LPS-induced IL-8 mRNA expression with pretreatment with SP600125 (Fig. 3A). Concentrations of SP600125 greater than 10 M have been previously shown to be less specific for JNK (40), thereby suggesting that MCP-1, but neither TNF-␣ or IL-8, is dependent on JNK activation for increased expression after LPS stimulation. As was seen with mRNA expression, the LPS-induced increase in MCP-1 protein expression was also dose dependently decreased upon pretreatment with SP600125 with a return to baseline levels of MCP-1 expression by preincubation with 5 M (IC 50 of ϳ4 M) (Fig. 3B). The LPS-induced protein expression of TNF-␣ and IL-8 were less affected by preincubation with SP600125, where levels of TNF-␣ were significantly inhibited only at SP600125 concentrations of 10 M and higher (IC 50 of ϳ10 M) and IL-8 levels were not significantly changed, even with pretreatment with 20 M of SP600125 (IC 50 of Ͼ20 M) (Fig. 3B).
Syk and PI3K Activation Increases in Neutrophils Exposed to LPS-The upstream mechanism by which JNK is activated in neutrophils remains unclear, but we have previously shown that Syk responds to TNF-␣ (37) and PI3K is activated in neutrophils after several stimuli (58 -60). To determine if Syk is activated after LPS exposure in neutrophils, neutrophils were incubated with LPS, Syk was immunoprecipitated, and autophosphorylation was assessed. As can be seen in Fig. 4A, LPS-stimulated neutrophils demonstrated a time-dependent increase in autophosphorylation of a 40-kDa species, representing a proteolytic cleavage product of Syk (37, 57). Syk activity increased within 5 min after exposure to LPS with a return to baseline by 10 min after LPS stimulation. The LPS-induced activity of Syk was confirmed (Fig. 4B) by assessing Syk phosphorylation in lysates, in which full-length Syk was identified, by immunoblotting with a phospho-Syk (Tyr-352)-specific antibody.
We next investigated if LPS induces PI3K activity in neutrophils. Neutrophils were stimulated with LPS followed by the immunoprecipitation of p85 and an in vitro lipid kinase assay utilizing phosphatidylinositol as substrate. As seen in Fig. 4C, PI3K is rapidly activated after LPS stimulation in neutrophils with increased activity seen as earlier as 2 min after LPS stimulation, peak activation at 5 min, followed by a gradual decline in activation back to baseline over the next 35

. Syk and PI3K activity increases in LPS stimulated neutrophils.
A, human neutrophils were incubated for 60 min at 37°C and stimulated with LPS (100 ng/ml) for the times shown. Syk was immunoprecipitated from cell lysates followed by assessment of Syk autophosphorylation by an in vitro kinase assay. Proteins were separated on SDS-PAGE gels, transferred to nitrocellulose, with radiolabeled bands identified by autoradiography (upper panel). Immunoblotting for Syk was then performed to ensure that equal amounts of Syk were immunoprecipitated from each sample (lower panel). Blots shown represent one of three experiments with similar results. B, human neutrophils were incubated at 37°C for 60 min and then were stimulated with LPS (100 ng/ml) for the indicated times. Cells were then lysed in boiling Laemmli sample buffer with proteins separated by SDS-PAGE followed by immunoblotting for phospho-Syk (Tyr352). C, isolated neutrophils were stimulated with LPS (100 ng/ml) for the indicated time points or were left unstimulated. PI3K activity was assessed per "Experimental Procedures." The graph represents mean Ϯ S.E. of three separate experiments. *, p Ͻ 0.05 by one-way ANOVA. Membranes were probed to show that equal amounts of JNK-1 were immunoprecipitated (lower panel). The representative autoradiograph and immunoblot of JNK-1 are from one of three experiments with similar results. Quantification of phosphorylated c-Jun band intensity was performed by PhosphorImager analysis. Activity of c-Jun after LPS stimulation alone was assigned an intensity of 100% with intensity after piceatannol or wortmannin pretreatment normalized to LPS stimulation alone. Results shown are mean Ϯ S.E. of three separate experiments. *, p Ͻ 0.001 by one-way ANOVA. C, isolated neutrophils were incubated with or without piceatannol (left) or wortmannin (50 nM) (right), with Me 2 SO (0.1%) as diluent control, for 60 min at 37°C. Cells were then stimulated with LPS (100 ng/ml) for 30 min followed by immunoprecipitation for p38 and assessment of p38 kinase activity utilizing the recombinant ATF fragment ATF  . Proteins were separated by SDS-PAGE, transferred to nitrocellulose, with radiolabeled proteins detected by autoradiography. Autoradiographs of one of three experiments with similar results are shown. D, neutrophils were incubated under non-suspended conditions for 60 min and then stimulated with LPS (100 ng/ml) (left) or incubated with piceatannol (or Me 2 SO (0.1%) as diluent control) for 60 min then stimulated with LPS (100 ng/ml) (right). Immunoprecipitation for Lyn or pTyr was performed as indicated, proteins were separated by SDS-PAGE, followed by assessment of Lyn autophosphorylation from Lyn immunoprecipitations by an in vitro kinase assay (as per Syk autophosphorylation in Fig. 4) or immunoblotting for Hck for pTyr immunoprecipitations, as indicated. Immunoblotting for Lyn was performed on Lyn immunoprecipitations to ensure that equal amounts of Lyn were immunoprecipitated. Autoradiographs and immunoblots of one of three experiments with similar results are shown.
were incubated with piceatannol (10 M) to inhibit Syk or with wortmannin (50 nM) to inhibit PI3K prior to stimulation with LPS. Preincubation with piceatannol led to a significant, but incomplete, decrease in LPS-induced JNK activation (Fig. 5A) at 15 and at 30 min, whereas wortmannin inhibited the activation of JNK only at 15 min after LPS stimulation (Figs. 5B). However, inhibition of LPS-induced JNK activation 15 min after LPS stimulation was more pronounced with PI3K inhibition compared with Syk inhibition. To assess if Syk and/or PI3K were involved in other LPS-induced signaling pathways leading to MAPK activation, or were specific to the JNK pathway, we assessed if Syk and/or PI3K were involved in the LPS-induced activation of p38 MAPK, because only p38 MAPK, but not ERK, has been shown to be activated in LPS-stimulated neutrophils. As can be seen in Fig. 5C, neither piceatannol nor wortmannin altered LPS-stimulated p38 activation.
To confirm the specificity of piceatannol inhibition of Syk in our model, we examined the effect of piceatannol on LPSinduced Lyn and Hck activation. Exposure of non-suspended neutrophils to LPS increased Lyn activation, as assessed by autophosphorylation of immunoprecipitated Lyn, and Hck phosphorylation, which peaked 5 and 15 min after LPS exposure, respectively (Fig. 5D, left). Incubation with piceatannol prior to LPS exposure did not diminish LPS-induced Lyn activation or Hck phosphorylation in non-suspended neutrophils (Fig. 5D, right).

Inhibition of Either Syk or PI3 Kinase Decreases LPS-induced Expression of MCP-1-Because inhibition of Syk and
PI3K led to a decrease in JNK activation, and inhibition of JNK decreased MCP-1 expression, we hypothesized that inhibition of Syk and PI3K would also decrease MCP-1 expression. Neutrophils pretreated with piceatannol showed a decrease of ϳ50% in the secretion of MCP-1 protein after LPS stimulation, whereas those pretreated with wortmannin showed a near complete inhibition of LPS-induced MCP-1 expression (Fig.  6A). To determine if transcriptional or translational mechanisms were responsible for the decrease in LPS-induced MCP-1 protein expression with Syk or PI3K inhibition, we assessed the expression of MCP-1 mRNA in LPS-stimulated neutrophils pretreated with piceatannol or wortmannin. As can be seen in Fig. 6, pretreatment with piceatannol prior to LPS stimulation reduced the expression of MCP-1 mRNA by ϳ50% compared with LPS stimulation alone, whereas pretreatment with wortmannin completely inhibited the LPS-induced increase in MCP-1 expression, thereby suggesting a pre-translational mechanism for the decrease in MCP-1 protein expression with Syk or PI3K inhibition.
SP600125 Does Not Inhibit LPS-induced Syk Phosphorylation-In these experiments we utilized concentrations of SP600125 previously shown to be specific for JNK (40). To assist in confirming that our results of a decrease in LPSinduced MCP-1 expression with SP600125 were the result of JNK inhibition and not the result of SP600125 inhibiting Syk, we assessed LPS-induced Syk phosphorylation in the presence of SP600125. As seen in Fig. 7, pretreatment with SP600125 did not affect LPS-induced Syk phosphorylation at any concentration of SP600125 utilized.
PI3 Kinase Is Upstream of Syk-Our findings of a more rapid activation of PI3K compared with Syk after LPS stimulation (Fig. 4) combined with the more significant decrease in LPSinduced JNK activation and MCP-1 expression with PI3K inhibition, compared with Syk inhibition, suggested to us that PI3K may reside upstream of Syk in LPS-stimulated neutrophils. To investigate this possibility, neutrophils were preincubated with wortmannin and LPS-induced Syk autophosphorylation was assessed. The inhibition of PI3K with wortmannin attenuated Syk activation after LPS exposure (Fig. 8). This suggests that PI3K is upstream of Syk in the LPS-induced pathway leading to JNK activation in neutrophils, and that PI3K also activates a Syk-independent pathway leading to JNK.
Syk Associates with TLR4 in Neutrophils after LPS Exposure-To strengthen our findings of the role of Syk in LPSinduced signaling pathways in neutrophils, we investigated the possibility that Syk is a member of the TLR4 signaling complex. Recently, associations of PI3K and Rac1 with TLR2 (36); CXCR4, CD11b, HSP70, and HSP90 with TLR4 (61, 61); and Bruton's tyrosine kinase with TLR4 (63) have been described. To investigate this possibility, neutrophils were stimulated with LPS, TLR4 was immunoprecipitated from cell lysates, and the presence of Syk was assessed by immunoblotting. As can be seen in Fig. 9, Syk co-precipitates with TLR4 at baseline, but the association increases within 5 min after LPS stimulation, peaks at 10 min, and returns to baseline by 20 min. The kinetics of the increased association of Syk with TLR4 after LPS stimulation correlates with the increase in Syk activity shown previously (Fig. 4, A and B). To confirm that equivalent amounts of TLR4 were immunoprecipitated from each sample, the membranes were stripped and reblotted for TLR4. Although full-length TLR4 was identified in neutrophil cell lysates (data not shown), no full-length TLR4 could be detected in the TLR4 immunoprecipitants even with the use of three separate TLR4 antibodies (data not shown).
To determine whether proteolysis of TLR4 may explain these findings, human neutrophils were incubated with DFP prior to LPS exposure to prevent degradation of TLR4 and Syk upon cell lysis. As seen in Fig. 9B, Syk and TLR4 clearly associate with each other in TLR4 or Syk immunoprecipitates, respectively, and use of DFP allows visualization of full-length TLR4 and Syk in such immunoprecipitations. Addition of DFP, however, augments the baseline association of TLR4 with Syk and eliminates LPS-dependent changes in association. Similar results were seen if DFP was added after stimulation but immediately prior to cell lysis (data not shown). DISCUSSION We have identified a novel signaling pathway in LPS-stimulated neutrophils inducing the activation of the MAPK JNK. Although the activation of JNK after LPS stimulation has been described in macrophages (48,64), to our knowledge this is the first description of the activation of JNK in neutrophils stimulated with LPS. LPS-stimulated JNK activation is not observed in human neutrophils in suspension. The activation of JNK after LPS stimulation is only seen in neutrophils under non-suspended conditions, as was the case in analyzing the activation of JNK after TNF-␣ stimulation (37).
The focus of this work was to examine the upstream signaling events leading to the LPS-induced activation of JNK and the downstream consequences thereof. In this study we have identified Syk and PI3K as two new components of LPS signaling and demonstrated that both Syk and PI3K participate in JNK activation. The activation of Syk has been shown to occur in neutrophils upon initiation of phagocytosis (65) and after TNF-␣ stimulation (37), and PI3K activation has been shown in neutrophils upon exposure to IL-8, leukotriene B4, and formylmethionylleucylphenylalanine (58 -60); however, the activation of Syk or PI3K after LPS stimulation has not been previously described. In addition, although Syk has been shown to play a role in JNK activation in T cells (66) and in neutrophils FIG. 8. LPS stimulated Syk activation is PI3K dependent in neutrophils. Neutrophils were isolated and incubated with wortmannin (50 nM), or with Me 2 SO (0.1%) as diluent control, for 60 min at 37°C and stimulated with LPS (100 ng/ml) for 5 min. Syk was immunoprecipitated from cells followed by performance of an in vitro kinase assay to assess Syk autophosphorylation. Proteins were resolved by SDS-PAGE, transferred to nitrocellulose, with radiolabeled bands determined by autoradiography and quantified by PhosphorImager analysis. Immunoblotting for Syk was performed to show that equal amounts of Syk were immunoprecipitated. A, representative autoradiograph (upper) and Syk immunoblot (lower) of one of three experiments with similar results. B, quantification of Syk phosphorylation by Phospho-rImager analysis. Intensity of Syk phosphorylation after LPS stimulation was assigned 100% with intensity after wortmannin pre-treatment normalized to LPS stimulation alone. Results shown are mean Ϯ S.E. of three separate experiments. *, p Ͻ 0.005 by Student's paired t test.
FIG. 9. Syk increases association with TLR4 upon LPS stimulation. A, neutrophils were stimulated with LPS (100 ng/ml) followed by immunoprecipitation (IP) of TLR4. Immunoprecipitated proteins were separated by SDS-PAGE, transferred to nitrocellulose, and immunoblotted (IB) with an anti-Syk antibody (upper panel). The arrows represent the 42-and 40-kDa proteolytic products of Syk seen in neutrophil lysates. The blots shown are representative of one of three experiments with similar results. B, in separate experiments, neutrophils were incubated with DFP (1 mM) for 10 min prior to LPS stimulation (100 ng/ml) for the indicated times. Cells were lysed, TLR4 or Syk was immunoprecipitated with proteins separated by SDS-PAGE, and immunoblotting for Syk or TLR4 performed, respectively. After stripping, the immunoblots were reprobed for TLR4 or Syk to ensure that equal amounts of protein were immunoprecipitated. after TNF-␣ stimulation (37); and PI3K has been shown to be important in JNK activation in other cell systems after shear stress (26), Fc receptor cross-linking (34) or upon exposure to platelet-derived growth factor (32), angiotensin II (67), or hydrogen peroxide (33), only limited information is available regarding the role of Syk and/or PI3K in LPS-induced JNK activation (48). Collectively, our data show a functional link between Syk, PI3K, and JNK in LPS-stimulated neutrophils not seen in other cell systems.
We have shown a relationship between Syk and PI3K activation in the pathway leading to LPS-induced JNK activation in neutrophils. Although PI3K has been shown to reside downstream of Syk in neutrophils after phagocytosis of IgG-coated erythrocytes (65), no role of JNK was determined. In contrast to those findings, our results suggest that PI3K is upstream of Syk. Not only is PI3K activation induced more rapidly than Syk, but also pretreatment with wortmannin inhibited the increased activation of Syk found after LPS exposure (Fig. 8).
In addition our results suggest that neither Syk nor PI3K are involved in signaling pathways leading to p38 activation after LPS stimulation in neutrophils (Fig. 5C), and because we have not been able to detect the activation of ERK MAPK after LPS stimulation in suspended or non-suspended neutrophils, 2 this suggests that both Syk and PI3K are specific to the activation of JNK MAPK in neutrophils after stimulation by LPS.
Our findings further extend the repertoire of neutrophil signaling after exposure to LPS and suggest that up-regulation of genes dependent on JNK increase in non-suspended neutrophils. We show that the activation of JNK led to an increase in the expression and release of MCP-1, as assessed by a decrease in MCP-1 expression with pretreatment with the JNK inhibitor SP600125. Additionally, because pretreatment with piceatannol and wortmannin also decreased LPS-induced MCP-1 expression, this further implicates Syk and PI3K in the pathway leading to JNK activation after LPS stimulation in neutrophils. In contrast, our data suggest that JNK activation is much less important in the LPS-induced regulation of TNF-␣ and IL-8 mRNA expression, because inhibition of their expression after LPS stimulation only occurred with concentrations of the JNK inhibitor SP600125 previously shown to be less specific for JNK (40).
The use of chemical inhibitors for Syk and PI3K in examining signaling pathways in human neutrophils is unavoidable due to the inability to reliably transfect neutrophils. In an attempt to overcome limitations of inhibitor use, we have utilized concentrations of piceatannol or wortmannin, which have been shown to be specific for Syk or PI3K respectively (37,65,68,69). In particular, although piceatannol inhibits several tyrosine kinases, concentrations of piceatannol of Յ10 M have been shown to only inhibit the activation of Syk without inhibition of Lyn, Hck, or Fgr activation (65). To further confirm these previous findings in our model system, we have shown that piceatannol does not inhibit LPS-induced Lyn or Hck activation in non-suspended neutrophils (Fig. 5D). Therefore, piceatannol is relatively specific for Syk in this system. Despite these limitations, the use of piceatannol and wortmannin provides strong evidence to suggest that Syk and PI3K are involved in the LPS-induced signaling pathway leading to JNK activation. Likewise for the JNK inhibitor SP600125, we attempted to use concentrations that are specific for JNK. Although difficult to discern intracellularly, the IC 50 of each JNK isoform in vitro for SP600125 is Ͻ0.1 M, whereas the IC 50 for most other proteins tested is Ͼ10 M (40). In our series of experiments we carefully utilized concentrations of SP600125 Ͻ 10 M in an attempt to exclude the potential inhibition of non-JNK proteins confounding our results. In addition, to confirm the specificity of SP600125 in our system we have shown that SP600125 did not affect Syk phosphorylation after exposure to LPS. Taken together, our findings of an inhibition of LPS-induced JNK activation, as well as MCP-1 expression, with concentrations of SP600125 of 2 M support the specificity of our findings of the role of JNK in LPS-induced MCP-1 expression. Finally, in further support of these inhibitor studies is the close correlation between inhibition of each component and the inhibition of MCP-1.
Although both Syk and PI3K regulate JNK activity in LPSstimulated neutrophils, other pathways undoubtedly exist. This is evidenced by the incomplete inhibition of LPS-induced JNK activation by inhibition of Syk or PI3K (Fig. 5, A and B). The activation of JNK proceeds through the activation of MKK4 and MKK7 (70 -73). Although the specific MKK (i.e. MKK4 and/or MKK7) utilized in LPS-induced activation of JNK in neutrophils is unknown, and is the subject of current studies, the possibility exists that the Syk-and PI3K-dependent and -independent pathways may preferentially utilize different MKKs leading to JNK activation after LPS stimulation (70).
We show that the protein-tyrosine kinase Syk is not only activated after LPS stimulation but is associated with TLR4 in quiescent neutrophils and can increase its association with TLR4 after LPS stimulation. This interaction between Syk and TLR4 seen in neutrophils at baseline and upon LPS stimulation may be direct or indirect. Previously Syk has been shown to associate with CD18 and the Fc␥ receptor, both of which interact with TLR4 upon LPS stimulation in macrophages, which suggests an indirect association of Syk with TLR4 (61,74,75). In support of a direct association of Syk with TLR4, TLR4 contains a motif in the intracytoplasmic portion that is a putative Syk-SH2 domain recognition site, suggesting that Syk may associate with TLR4 directly through this domain (76). The elucidation of the mechanisms by which TLR4 and Syk associate in LPS-stimulated neutrophils requires further investigation.
The activation of JNK after LPS stimulation only in nonsuspended neutrophils suggests the potential involvement of the integrins CD11b and CD18 in this pathway. Two points strengthen the possibility that CD11b regulates/modulates this pathway. First, we have recently shown the involvement of CD11b in TNF-␣-induced JNK activation in adherent neutrophils (37), and second, CD11b has been recently shown to increase association with TLR4, and with CD14, in LPS-stimulated macrophages suggesting that a similar process may occur in neutrophils (61). We are currently examining the potential role of CD11b in the LPS-induced activation of JNK in neutrophils.
Although here implicated in the response of neutrophils to LPS, the role of JNK in the inflammatory response in vivo is unknown. JNK activation was shown to increase in an LPSinduced model of acute lung injury, which was inhibited by the protein-tyrosine kinase inhibitor genistein (77). Markers of acute lung injury were also decreased in this model. In addition, MCP-1 is responsible for the recruitment of monocytes to sites of infection to further the inflammatory response (78,79). Based on our findings of the release of MCP-1 in non-suspended neutrophils in a JNK-dependent manner, we speculate that JNK activation in neutrophils induced to become adherent at sites of inflammation would increase monocyte recruitment by the release of MCP-1. Taken together these suggest that inhibition of JNK may have significant effects on the underlying pathophysiology of acute lung injury. 2 P. G. Arndt, unpublished observations.