Inhibition of c-Jun N-Terminal Kinase Limits Lipopolysaccharide-induced Pulmonary Neutrophil Influx

The syndrome of acute lung injury (ALI), frequently associated with sepsis and endotoxemia in patients, causes significant morbidity and mortality (1-4). Although the neutrophil is an important beneficial component of the innate immune response to bacterial products, it is strongly implicated in the pathogenesis of ALI (1, 2, 5). The influx of neutrophils into the pulmonary parenchyma and sequentially into the alveolar space in ALI is a complex process that includes cell retention and margination within the pulmonary microvasculature, adhesion to the vascular endothelium, and migration into the alveolar compartment (6, 7). Common to all of these events is the requirement for assembly of the actin-containing cytoskeleton, which is induced in response to a variety of stimuli (7-10). We have hypothesized, accordingly, that inhibition of actin assembly could reduce neutrophil influx into the lung after an inflammatory insult, but the pathways that regulate neutrophil actin assembly after exposure to LPS or LPS-induced mediators remain unclear.

Major signaling pathways regulating cellular growth and response to cytokines and stress occur through the highly conserved mitogen-activated protein kinases (MAPKs) family, which consists of p42/44 extracellular signal-regulated kinase, p38 MAPK, and c-Jun N-terminal kinase (JNK). Although p38 is the only MAPK activated in LPS-stimulated neutrophils under suspended cell conditions (11, 12), we have recently shown the activation of JNK after LPS and tumor necrosis factor a (TNF-a) stimulation in adherent neutrophils, which enhances cell-to-cell and cell-to-substratum interactions, and more closely mirrors the complex microenvironment of the lung (10, 13).

In human neutrophils, although p38 MAPK regulates release of TNF-a and interleukin 8 (IL-8), superoxide production, chemotaxis, and adhesion after LPS stimulation, LPS-induced actin polymerization is p38 MAPK-independent (12). In addition, activation of Cdc42 has been shown to be necessary for actin assembly in neutrophils; however, the upstream signaling pathways have not been well characterized (14-16). Therefore, we hypothesized that MAPK pathways other than p38, such as JNK, may be important in LPS-induced actin assembly by acting through Cdc42, and therefore may diminish LPS-induced pulmonary neutrophil recruitment.

The role of JNK in ALI is not well understood. Recently, it was shown that inhibition of protein tyrosine kinases with genistein decreased LPS-induced alveolar neutrophil influx (17). The activation of JNK was also decreased in this study, thereby suggesting the importance of JNK activation in LPS-induced pulmonary neutrophil recruitment (17). In addition, the specific inhibitor of JNK kinase activity, SP600125 (18), was recently shown to decrease pulmonary neutrophil influx in a ventilator-induced lung injury model (19) and lung inflammation in a model of ischemia/reperfusion (20). These studies, taken together, suggested to us that systemic inhibition of JNK activation might decrease pulmonary neutrophil recruitment, through an action on actin assembly, in an LPS-induced model of acute lung inflammation.

Here, we show that systemic inhibition of JNK activation with SP600125 decreased neutrophil recruitment to the lung. We propose that inhibition of neutrophil capillary retention in the lung by a decrease in LPS- and TNF-a-induced actin polymerization, which proceeds through an inhibition of Cdc42 activation, is one mechanism for the decrease in lung neutrophil accumulation by systemic JNK inhibition after LPS exposure. Some of the results of these studies have been previously reported in the form of an abstract (21).

METHODS

Animals

Female C57BL/6 mice, aged 8-12 weeks, were used for all experiments. Animals were maintained in the animal care facility on a 12-hour light/ dark cycle with full access to food and water. Animal protocols were approved by the Animal Care and Use Committee at National Jewish Medical and Research Center.

Animal Studies

LPS-induced model of acute lung inflammation. The JNK inhibitor SP600125 was diluted in a buffer containing 30% polyethylene glycol (PEG)-400, 20% polypropylene glycol, 15% Cremophor EL, 5% ethanol, and 30% saline, as previously described (18). Mice were injected subcutaneously with SP600125 (30 mg/kg) or vehicle control, 3 hours before and 12 hours after (where applicable) LPS exposure. LPS (300

JNK Assay

Murine alveolar macrophages and bone marrow-derived band 3 neutrophils were obtained and resuspended in Roswell Park Memorial Institute (RPMI) 1640 with 1% murine heat-inactivated platelet-poor plasma, as previously described (23). Cells were placed in microfuge tubes with SP600125 (2 or 5 µM) or dimethyl sulfoxide (DMSO; 0.1%) as diluent control, and were incubated for 60 minutes at 37°C under nonsuspended conditions, undisturbed (adherent method), followed by stimulation with LPS (100 ng/ml) for 30 minutes. After stimulation, a JNK in vitro kinase assay was performed as previously described (10, 13). Radioactive bands were quantified using Phosphor-Imager (Storm 820; Molecular Dynamics, Sunnyvale, CA) analysis.

Neutrophil Functional Assays

Neutrophil isolation. Human neutrophils were isolated from healthy donors by the plasma Percoll method as previously described (11).

Actin polymerization and localization. Human neutrophils were incubated with the indicated concentration of SP600125 for 60 minutes under adherent conditions (see JNK ASSAY) and then stimulated with LPS (1 µg/ml, 40 minutes) or TNF-a (5 ng/ml, 15 minutes). Actin polymerization and localization for human or murine neutrophils were performed as previously described (24).

Retention. Isolated human neutrophils were labeled with ^sup 111^In and incubated with SP600125 (5 µm), or DMSO (0.1%) as control, for 60 minutes under adherent conditions before LPS (200 ng/ml, 40 minutes) or TNF-a (5 ng/ml, 15 minutes) exposure. LPS-induced neutrophil retention through 5-µm filters at 1- or 2.5-ml/minute flow rates was performed as previously described (25).

Cdc42 Activation

Cdc42 activation after LPS or TNF-a stimulation was assessed as previously described (26). Isolated human neutrophils were resuspended in Krebs-Ringer phosphate buffer (KRPD) with SP600125 (5 µM) or DMSO (0.1%) as control, and incubated for 60 minutes at 37°C under adherent conditions. Neutrophils were then stimulated with LPS (100 ng/ml) or TNF-a (5 ng/ml) for the indicated times. Immunoblotting for Cdc42 was performed as previously described (11).

Statistics

Statistical analysis was performed by one-way analysis of variance with Tukey-Kramer post hoc analysis or Student's t test, as indicated, using Prism software (GraphPad Software; San Diego, CA). A p value of less than 0.05 was considered significant.

Additional information on materials and methods can be found on the online supplement.

RESULTS

Exposure to LPS Activates JNK in Murine Neutrophils and Alveolar Macrophages, Which Is Inhibited by SP600125

Although we have previously demonstrated activation of JNK by LPS in human neutrophils (13), we wished to determine whether JNK was activated in murine neutrophils and alveolar macrophages after LPS exposure, and to assess the effectiveness of the JNK inhibitor SP600125 on inhibiting this LPS-induced JNK activation. Bone marrow-derived neutrophils and alveolar macrophages were isolated, incubated with SP600125 (2 or 5 µM) for 60 minutes under adherent conditions (see METHODS), and then exposed to LPS (100 ng/ml) for 30 minutes (peak in JNK activity, data not shown). After stimulation, cells were lysed, JNK-1 was isolated by immunoprecipitation, and an in vitro kinase assay was performed to quantify JNK activity using c-Jun^sub 1-29^ as an exogenous substrate. As shown in Figure 1, LPS increased JNK activity in both neutrophils and alveolar macrophages, which was inhibited by SP600125 at concentrations of 5 µM or less.

The JNK Inhibitor SP600125 Decreases LPS-induced

JNK Activation in the Lung In Vivo

Before examining the potential role of JNK in LPS-induced pulmonary neutrophil recruitment in vivo, we assessed if the JNK inhibitor SP600125 would decrease JNK activation in the whole lung after exposure to LPS. Mice administered SP600125 or vehicle alone were exposed to LPS by aerosolization. Eight hours after exposure, lungs were harvested, homogenized, and subjected to Western blotting to test for phosphorylation of c-Jun, a JNK substrate in which phosphorylation has been shown to be decreased by SP600125 in a separate model of acute lung inflammation (19). SP600125 pretreatment decreased c-Jun phosphorylation after LPS exposure to levels similar to control values (Figure 1C).

Inhibition of JNK Delays and Decreases LPS-induced Neutrophil Recruitment to the Lung

We have previously shown that aerosolized LPS causes an influx of inflammatory cells (predominately neutrophils) into the lung and alveolar space within 4 hours, with peak influx seen 8 hours after LPS exposure (Reference 22 and Figure 2). The JNK inhibitor SP600125, which decreased LPS-induced JNK activation in neutrophils, alveolar macrophages, and whole lung, appeared to be a suitable reagent with which to test the role of JNK inhibition on the neutrophil influx to the lung after LPS exposure. As seen in Figure 2, treatment of mice with SP600125 before LPS exposure significantly decreased the total number of cells (Figure 2A) and neutrophils (Figure 2B) found in the alveolar space compared with mice exposed only to LPS.

Because inhibition of JNK significantly decreased the number of neutrophils in the alveolar space after LPS exposure, we also tested whether systemic JNK inhibition decreased the accumulation of neutrophils in the whole lung, as assessed by lung myeloperoxidase (MPO) content, after LPS exposure. As seen in Figure 2C, mice treated with SP600125 before LPS exposure had significantly decreased MPO levels at 8, 24, and 48 hours after LPS exposure compared with mice treated with LPS alone. Importantly, MPO levels in mice treated with SP600125 before exposure to LPS were unchanged in comparison to control mice at the 8- and 48-hour time points.

Inhibition of JNK Decreases LPS-induced Microvascular Leak in the Lung

The degree of pulmonary microvascular leak in our LPS-induced model of ALI was assessed by lung wet-to-dry ratios. Lung wetto-dry ratios 8 hours after LPS exposure significantly increased compared with untreated mice and this increase was blocked by systemic JNK inhibition (Figure 3A). Comparison of the MPO level with the wet-to-dry ratio revealed a positive correlation, suggesting that, as reported previously (27, 28), microvascular leak is associated with neutrophil influx (Figure 3B).

Inhibition of JNK Activation Decreases Histologic Evidence of Acute Lung Inflammation

The pathologic features present in the lung in the exudative phase of ALI include interstitial thickening with fluid and cellular accumulation, intraalveolar filling with neutrophils, and hyaline membrane formation (29). To examine the effect of JNK inhibition on the histologie appearance of LPS-induced lung injury, mice were exposed to LPS with or without pretreatment with SP600125. At 8 hours after LPS exposure, the lungs were harvested from mice and examined microscopically. As seen in Figure 4, lungs from LPS-treated mice (Figure 4B) showed thickening of the interstitium, increased cellularity, the presence of alveolar exudates, and accumulation of neutrophils in the alveolar space compared with saline-exposed mice only (Figure 4A). Mice pretreated with the JNK inhibitor and then exposed to LPS showed an attenuated response to LPS (Figure 4C) compared with LPS exposure alone (Figure 4B).

Effect of JNK Inhibition on Chemokine and Cytokine Release After LPS Exposure

The ELR+ chemokines KC, macrophage inflammatory protein 2 (MIP-2), and IL-8 are potent neutrophil chemoattractants (3032). A decrease in the expression of KC or MIP-2 in the alveolar compartment would provide a mechanism for the decrease in LPS-induced neutrophil influx into the lung with SP600125 treatment. Therefore, chemokine release in the alveolar compartment was measured simultaneously with studies of leukocyte influx in mice pretreated with SP600125 and exposed to LPS compared with those exposed to LPS only. Surprisingly, BAL fluid levels of the neutrophil chemoattractants KC and MIP-2 were significantly elevated, and not decreased, in the setting of JNK inhibition after exposure to LPS at 4 hours, compared with the group exposed to LPS alone (Figure 5A). This increase in BAL fluid concentrations of KC and MIP-2 in mice treated with SP600125 is in contrast to the inhibition of neutrophil influx into the alveolar space seen in these mice (Figure 2B). In contrast to chemokine release, BAL fluid levels of the cytokines TNF-a and IL-1ß did not differ between treatment groups (Figure 5B). A recent study of ventilator-induced lung injury reported a decrease in BAL fluid MIP-2 levels in the setting of JNK inhibition, opposite to our findings in LPS-exposed lungs; however, BAL fluid cytokine levels were not reported in that study (19).

JNK Activation Regulates LPS- and TNF-a-induced Actin Polymerization

To examine potential mechanisms by which inhibition of JNK activation decreased neutrophil influx into the lungs of mice exposed to LPS, we initially investigated the effects of SP600125 on LPS-induced actin polymerization, because actin polymerization is involved in several pathways used by neutrophils for lung recruitment. Bone marrow-derived neutrophils were isolated from mice pretreated with SP600125 or vehicle alone as control, and were stimulated ex vivo with LPS or KC for various times. Actin polymerization after LPS exposure was decreased at all time points tested in neutrophils isolated from mice pretreated with the JNK inhibitor compared with those pretreated with vehicle (Figure 6A, left). In contrast to LPS stimulation, however, KC-induced actin polymerization was not altered in the presence of systemic JNK inhibition (Figure 6A, right).

Because mechanisms of actin assembly are better characterized in human neutrophils, compared with bone marrow-derived murine neutrophils, we elected to use human neutrophils to further examine neutrophil function in the setting of an impairment in actin assembly with JNK inhibition. We initially confirmed that inhibition of JNK diminished LPS-induced, but not IL-8-induced, actin polymerization in human neutrophils, similar to our findings in murine neutrophils. In addition, we examined the effects of JNK inhibition on TNF-a-induced actin polymerization, because the immune response to localized LPS exposure may occur through non-LPS mediators, such as TNF-a, and not directly through LPS (33). Human neutrophils were incubated with SP600125 under adherent conditions (see METHODS) and were then stimulated with LPS (1 µg/ml) for 40 minutes or TNF-a (5 ng/ml) for 15 minutes, which represent the peak time of actin polymerization for LPS and TNF-a, respectively (10, 34). Pretreatment with SP600125 decreased LPS- (Figure 6B, top) and TNF-a-induced (Figure 6B, bottom) F-actin polymerization in a concentration-dependent fashion.

Because inhibition of F-actin polymerization does not necessarily indicate that alterations in actin localization occurred in neutrophils pretreated with the JNK inhibitor, we assessed the effect of inhibition of JNK on LPS- and TNF-a-induced actin localization. Human neutrophils were incubated with or without SP600125 (5 µm) under adherent conditions for 60 minutes and then were exposed to LPS (1 µg/ml) or TNF-a (5 ng/ml) for 40 or 15 minutes, respectively. As seen in Figure 6C, LPS- (c) and TNF-a- (e) induced actin localization to lamellipodial extensions. Although pretreatment of neutrophils with SP600125 (5 µM) alone had no effect on cell morphology or actin localization, pretreatment with the JNK inhibitor before LPS exposure completely inhibited formation of lamellipodia seen with LPS stimulation and resulted in a diffuse staining pattern for actin similar to untreated cells (Figure 6C, d). Pretreatment with SP600125 also had a significant but less dramatic effect on TNF-a-induced actin localization, as compared with LPS stimulation (Figure 6C, f).

JNK Inhibition Alters LPS-induced Neutrophil Retention

Previously, we have reported that LPS induces an increase in neutrophil retention (8, 25). Neutrophil retention in the pulmonary microvasculature is related to increases in LPS-induced actin polymerization and neutrophil cell stiffness (8). Because SP600125 decreased LPS-induced actin polymerization, we next tested if inhibition of JNK activation would also decrease LPS-induced neutrophil retention in model capillaries, and thus provide a potential mechanism by which inhibition of JNK decreased the pulmonary neutrophil influx in our animal model of LPS-induced lung injury. Human neutrophils were labeled with ^sup 111^In, pretreated with SP600125 (5 µM) for 60 minutes under adherent conditions and then were exposed to LPS (1 µg/ml, 40 minutes) followed by perfusion of the ^sup 111^In-labeled neutrophils across 5-µM pores. Unstimulated neutrophils after SP600125 pretreatment showed a small increase in retention over baseline. Stimulation by LPS increased neutrophil retention, which was significantly decreased by pretreatment with SP600125 (Figure 7). A similar decrease in neutrophil retention with SP600125 was seen at a flow rate of 1 ml/minute (data not shown).

JNK Inhibition Decreases LPS- and TNF-a-induced Cdc42 Activation

The small G-proteins Rac and Cdc42 have been shown to be important for actin assembly (14, 15, 35). To determine whether Cdc42 might be a JNK effector, it was first important to know whether Cdc42 was activated by LPS or TNF-a in neutrophils. We initially examined the activation of Cdc42 on stimulation with LPS and TNF-a. Human neutrophils were exposed to LPS (100 ng/ml) or TNF-µ (5 ng/ml) under adherent conditions, with active Cdc42 pulled down after cell lysis by exposure to the Pakbinding domain of p21. As seen in Figure 8A, LPS induction of Cdc42 activation in human neutrophils peaks at 5 minutes, whereas Cdc42 activation after TNF-a peaks at 15 minutes. We next assessed if JNK inhibition altered LPS- and TNF-a-induced Cdc42 activation. Pretreatment of human neutrophils with SP600125 (5 µM) for 60 minutes under adherent conditions decreased both LPSand TNF-a-induced Cdc42 activation (Figure 8B).

DISCUSSION

This study implicates JNK activation as an important mediator of the inflammatory process induced by LPS inhalation-a model of acute lung inflammation. Recently, we have shown JNK activation in LPS- and TNF-a-stimulated human neutrophils (10, 13). We now extend those findings to show that JNK activation also occurs in LPS-stimulated murine neutrophils and alveolar macrophages, two cells implicated in sepsis-mediated lung injury, as well as in vivo in the lung after exposure to LPS. Furthermore, systemic inhibition of JNK, with the specific JNK inhibitor SP600125, decreased the early exudative phase of acute lung inflammation as evidenced by a decrease in the influx of neutrophils into the lung and inhibition of microvascular leak, supporting a role for JNK activation in lung inflammation after exposure to LPS. Although we have previously shown that inhibition of p38 activation in our LPS-induced model of acute lung inflammation decreased the neutrophil influx into the alveolar space, inhibition of p38 activation had no effect on neutrophil accumulation in the pulmonary parenchyma or microvasculature leak (Reference 22 and unpublished observations). These findings suggested that other, non-p38-dependent, pathways are also important in the LPS-induced regulation of neutrophil influx into the lung parenchyma and for microvascular leak. Our results suggest that a JNK signaling cascade is one such pathway. To our knowledge, this is the first report demonstrating a decrease in neutrophil influx to regions of LPS-induced inflammation by systemic JNK inhibition.

The use of chemical inhibitors for JNK is unavoidable in examining the role of JNK in LPS-induced acute lung inflammation because mice deficient in the two major isoforms of JNK (JNKl and JNK2) are embryonically lethal. Although mice deficient in one of the two JNK isoforms are viable, findings from these mice are difficult to interpret because of the potential confounding effect of the normally expressed JNK isoform on the results. For in vitro experiments in these studies, we attempted to use concentrations of the JNK inhibitor SP600125 specific for JNK. Although difficult to discern intracellularly, the IC^sub 50^ of each JNK isoform in vitro for SP600125 is less than 0.1 µM, whereas the IC^sub 50^ for most other proteins tested is greater than 10 µM (18). Therefore, we carefully used concentrations of SP600125 of 5 µM or less in our experiments in an attempt to exclude the potential inhibition of non-JNK proteins confounding our results. In addition, although pharmacokinetics are unavailable for SP600125, we used previously published concentrations of SP600125 in our in vivo experiments, which have been shown to inhibit the activation of JNK (18, 19).

The increase in microvascular permeability in LPS-induced ALI has been previously shown to be neutrophil-mediated (27, 28, 36). Our data suggest that one mechanism by which JNK inhibition may have prevented an LPS-induced increase in microvascular permeability was through the inhibition of neutrophil influx into the lung. This conclusion is supported by a positive correlation between lung MPO values and wet-to-dry ratios after LPS exposure in our model (Figure 3A). However, other potential mechanisms may also be altered by JNK inhibition after LPS exposure in the complex pulmonary environment, resulting in a decrease in pulmonary microvascular permeability. A complete analysis of the effect of JNK inhibition on the pulmonary circulation is likely of considerable importance and is the subject of further study.

Our group and others have reported that JNK is not involved in LPS-induced expression of the neutrophil chemoattractant IL-8 in neutrophils and monocytes (13, 18). Thus, it might be expected that JNK inhibition would affect neither the expression of CXCR2 ligands nor the early neutrophil influx into the lung after LPS exposure, which is partly a consequence of the expression of chemokines. Our data clearly show, however, that the neutrophil influx into the lung and alveolar space is significantly impaired with JNK inhibition (Figures 2B and 2C). Thus, JNK inhibition decreased pulmonary neutrophil influx after LPS exposure, even though KC and MIP-2 levels (the murine homologs of IL-8) in BAL fluid were increased (compared with those exposed to LPS only; see Figure 5). Thus, even in the setting of increased levels of chemoattractants in the alveolar compartment, inhibition of JNK prevented the LPS-induced influx of neutrophils into the lung. This decrease in LPS-induced neutrophil recruitment to the alveolar compartment in the setting of systemic JNK inhibition likely occurs secondary to a decrease in the number of neutrophils retained in the lung, as assessed by lung MPO and histology, such that with fewer neutrophils retained in the lung, fewer neutrophils are then able to respond to KC or MIP-2 and migrate into the alveolar compartment.

Although many potential mechanisms may underlie the inflammatory response to LPS in the lung, a recent report examining the importance of nonleukocyte components on LPS-mediated pulmonary neutrophil recruitment argues that the direct response of neutrophils to LPS is relatively unimportant regarding the mechanism of pulmonary neutrophil sequestration after systemic LPS exposure (33). The cellular response to LPS involves the release of a multitude of proinflammatory mediators, including TNF-a, IL-1ß, and IL-8. The relative importance of LPS or other inflammatory mediators in our model of lung inflammation is unknown. We therefore examined the effect of JNK inhibition on both LPS- and TNF-a-induced actin assembly as a mechanism for the decrease in pulmonary neutrophil influx and to further illustrate the importance of JNK in our LPS-induced model of acute lung inflammation.

We provide evidence in support of impairment in actin assembly, and the resulting downstream effects of this impairment, as a mechanism for the decrease in LPS-induced pulmonary neutrophil recruitment with JNK inhibition. Pretreatment of neutrophils with SP600125 decreased LPS- and TNF-a-induced actin polymerization, thereby suggesting that JNK inhibition would decrease LPS- and TNF-a-induced cell stiffness. This effect of JNK inhibition was supported by a decrease in LPSinduced neutrophil retention. Previous studies have shown that inhibition of actin polymerization decreases pulmonary neutrophil retention (8, 37, 38), and the decrease in MPO values seen in SP600125-treated/LPS-exposed mice compared with LPS-exposed only mice in our study illustrates this point (Figure 2C). Unlike LPS-stimulated neutrophils, however, we were unable to detect a difference in TNF-a-induced neutrophil retention. The reason for the lack of correlation between the significant inhibition of actin polymerization and the less apparent decrease in neutrophil retention with JNK inhibition after TNF-a stimulation is not clear. One possibility is that although inhibition of JNK decreased actin polymerization similarly in LPS- or TNF-a-stimulated neutrophils, the polarization of actin assembly was less impaired in TNF-a-exposed neutrophils, with approximately 50% of the SP600125-treated/TNF-a-stimulated cells still demonstrating cellular conformational changes indicating localized regions of polymerized actin (Figure 6B). These cells undoubtedly would possess increased stiffness and therefore would have an increased propensity for retention in pulmonary capillaries. Furthermore, modeling retention behavior of neutrophils in the lung remains complex. The time constant of deformation might be substantially different after treatment with SP600125, yet not necessarily resolved by retention in pores. These issues require considerable further study.

Although we observed attenuation in LPS- and TNF-a-induced actin polymerization, no decrease in KC-induced actin polymerization with SP600125 pretreatment was seen in murine neutrophils (Figure 6A) and suggests that signaling pathways leading to actin polymerization after LPS and TNF-a exposure are different than those after KC or IL-8 stimulation. Although signaling pathways resulting in actin polymerization after LPS and TNF-a exposure are JNK-dependent, those after KC and IL-8 are likely JNK-independent. Because KC-induced actin polymerization is more robust and occurs earlier, compared with actin polymerization induced by LPS or TNF-a (Figure 6A), the argument can be made that KC-induced actin polymerization, but not that induced by LPS or TNF-a, predominates in regulating neutrophil recruitment to the lung in our model. If this was the case, because KC levels were higher in the alveolar compartment in mice pretreated with SP600125 compared with those exposed to LPS alone, we would predict a corresponding increase in neutrophil actin polymerization and retention, and therefore pulmonary neutrophil recruitment in mice pretreated with SP600125, opposite to the findings presented here. The relative role of KC-induced actin polymerization and retention, compared with that induced by LPS or TNF-a, is difficult to examine in vivo, but it is unlikely that either are solely responsible for neutrophil retention in our model. Conceivably, KC may mediate neutrophil retention early on, whereas LPS or TNF-a induces neutrophil retention at later times points after exposure to LPS. Finally, the possibility exists that the higher levels of KC in the setting of systemic JNK inhibition in vivo may have induced neutrophil desensitization, which would have resulted in a corresponding inhibition of KC-induced actin polymerization. Such a mechanism would support our findings of a decrease in pulmonary neutrophil recruitment after exposure to LPS with systemic JNK inhibition.

In support of a potential mechanism by which inhibition of JNK activation decreases actin assembly after LPS or TNF-a stimulation, we have shown that JNK inhibition decreased LPS-and TNF-a-induced Cdc42 activation in nonsuspended neutrophils. No previous study, to our knowledge, has described a pathway by which inhibition of the activation of an MAPk leads to an inhibition of Cdc42 activation. Cdc42 activation is important for actin polymerization in neutrophils after several stimuli, including formylmethionyl-leucyl-phenylalanine (fMLP), IL-1ß, and osmotic stress (14, 16, 28). In addition, although Rac2 and Cdc42 have both been reported to be involved in actin polymerization in neutrophils, TNF-a-induced actin polymerization was not decreased in Rac2-deficient mice, suggesting a predominant role for Cdc42 in actin polymerization after TNF-a stimulation (35). The effect on LPS-induced actin polymerization was not investigated in that study. Therefore, on the basis of these studies, and our inability to detect Rac2 activation in neutrophils after LPS or TNF-a stimulation (P.O.A., unpublished observations, May 2004), we conclude that the observed decrease in LPS-and TNF-a-induced actin polymerization with JNK inhibition occurs through inhibition of Cdc42 activation.

A recent study investigating the role of JNK in ventilatorinduced ALI reported finding a diminution in pulmonary neutrophil influx by systemic JNK inhibition with SP600125, which was associated with a decrease in MIP-2 protein expression (19). This decrease in MIP-2 expression with JNK inhibition in ventilatorinduced ALI is in contrast to our findings in LPS-induced ALI, in which BAL fluid MIP-2 levels were increased, and highlights interesting potential differences in the importance of JNK in regulating MIP-2 expression in the lung after LPS exposure or cell stretch. Taken together, this suggests that the clinical importance of JNK inhibition might be even more apparent in subjects with LPS-induced ALI requiring mechanical ventilation.

Overall, our data indicate that inhibition of JNK decreases the number of neutrophils retained in the pulmonary microvasculature and therefore the number of neutrophils able to adhere and migrate into the alveolar compartment. On the basis of this study and our previous investigations into the role of p38 activation in LPS-induced acute lung inflammation, the p38 and JNK pathways regulate neutrophil recruitment to the lung by different mechanisms. Although a decrease in neutrophil chemotaxis was suggested as a mechanism for the decrease in alveolar neutrophils with p38 inhibition, decreased neutrophil retention in the microvasculature, by way of impairment in actin polymerization, is suggested with JNK inhibition. With apparently distinct mechanisms for the inhibition of neutrophil accumulation into the lung after LPS exposure, this suggests that inhibition of both p38 and JNK may exert powerful inhibitory effects on the number of neutrophils recruited to the alveolar space, and thereby on acute pulmonary inflammation.

Conflict of Interest Statement: P.G.A. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; S.K.Y. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; J.G.L does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; M.B.F. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; JAN. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; G.S.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.

Acknowledgment: The authors thank Dr. Ken Malcolm for his helpful suggestions and careful review of the manuscript.