The antiapoptotic molecule Bcl-xL has been implicated in the differentiation and survival of activated macrophages in inflammatory conditions. In this report, the role of Bcl-xL in LPS-induced cytokine gene expression and secretion was studied. Bcl-xL-transfected RAW 264 macrophages were protected from gliotoxin-induced apoptosis, indicating the presence of functional Bcl-xL. Overexpression of Bcl-xL in this macrophage cell line was also associated with a marked inhibition of LPS-induced TNF-α, JE/monocyte chemoattractant protein 1, and macrophage inflammatory protein 2 secretion. Inhibition of LPS-induced cytokine secretion was paralleled by a decrease in levels of steady-state mRNA for the above cytokines and for IL-1β. Decreased production of TNF-α in Bcl-xL transfectants was not due to increased mRNA degradation, as the mRNA half-lives were the same in Bcl-xL transfectants and control macrophages. Although the composition of NF-κB complexes detected by EMSA and supershift analysis in nuclear lysates derived from Bcl-xL transfectants and control cells was indistinguishable, LPS-induced inhibitory κBα degradation, as well as NF-κB binding and AP-1 activation, were slightly decreased by ectopic expression of Bcl-xL. More strikingly, LPS-induced phosphorylation of p38 mitogen-activated protein kinase and c-Jun N-terminal kinase was strongly repressed by Bcl-xL overexpression, offering a possible mechanism for the inhibition of LPS-induced cytokine production. These data provide the first evidence for a novel role for Bcl-xL as an anti-inflammatory mediator in macrophages.
Cells of the monocyte/macrophage lineage play a crucial role in the initiation and maintenance of inflammation (1). In inflammatory situations, macrophages quickly become activated by bacterial LPS and/or cytokines. Although the molecular mechanisms of LPS-induced effects are not yet fully described, some of these pathways are well-known and include various kinases such as protein kinase C (2), Src-related kinases (3), and the three subclasses of mitogen-activated protein (MAP)4 kinases including the extracellular signal-regulated kinases (ERKs or p42, p44 kinases), p38 MAP kinases, and c-Jun N-terminal kinases (JNKs) (4). Transcription factors (e.g., NF-κB (5) and AP-1 (4)) and G-proteins (6) are also involved in LPS signal transduction. Activated macrophages secrete cytokines and release free radicals, such as NO, which contribute to the elimination of pathogens and tumor cells. NO also induces programmed cell death in a variety of cells, including macrophages themselves (7, 8). To avoid programmed cell death and maintain their defense potential, macrophages have developed protective mechanisms against NO-induced damage (9). In pathologic situations, overstimulation by bacterial LPS and the enhanced survival of macrophages could lead to excessive release of cytokines, resulting in uncontrolled systemic inflammation and septic shock (10). To avoid this condition, cytokine release is regulated by both transcriptional and posttranscriptional mechanisms (11, 12). In this study, we present a novel mechanism that may attenuate excessive cytokine release and protect macrophages against NO toxicity while maintaining NO secretion.
The Bcl family of proteins plays a central role in the regulation of apoptotic cell death induced by a wide variety of stimuli, including NO (13). Bcl-xL, a Bcl-2 homologue, has been reported to be up-regulated selectively during macrophage differentiation (14, 15) and has recently been described as a molecule which protects macrophages from NO-induced injury (16). However, ectopic expression of Bcl-xL does not affect the ability of the RAW 264 macrophage cell line to produce NO in response to LPS and IFN-γ activation (16). Bcl-2 and Bcl-xL have both been reported to protect endothelial cells against TNF-α-induced apoptosis through inhibition of NF-κB translocation (17). Inhibition of NF-κB activation has the potential to prevent the up-regulation of genes that are important in inflammatory conditions (18, 19). In fact, overexpression of Bcl family members has been demonstrated to inhibit E-selectin gene expression induced by TNF-α, LPS, and PMA (17). The role of Bcl-xL as an antiapoptotic molecule has been well established in various in vitro and in vivo systems, but little is known about other possible regulatory functions of this protein. In this study, the question of whether Bcl-xL can regulate LPS-induced cytokine expression and release in macrophages was studied. Using RAW 264 macrophage cell line, stable transfectants that overexpress Bcl-xL, we demonstrated that overexpression of Bcl-xL attenuates secretion of TNF-α, JE/monocyte chemoattractant protein (MCP)-1, and macrophage inflammatory protein (MIP)-2 following LPS treatment. This inhibition is also reflected at the level of steady-state mRNA for the above cytokines and for IL-1β. In the case of TNF-α, the decrease in steady-state mRNA levels in Bcl-xL transfectants is not due to an increased rate of mRNA degradation. To delineate possible mechanisms for suppressed cytokine expression, the effect of Bcl-xL overexpression on transcription factors, NF-κB and AP-1, as well as MAP kinases, important mediators of LPS-induced effects, were examined. Ectopic expression of Bcl-xL interfered slightly with the degradation of the inhibitory protein IκBα, consistent with a modest decrease in NF-κB nuclear translocation. More strikingly, however, phosphorylation of p38 MAP kinase and JNK were strongly attenuated in Bcl-xL transfectants. Moreover, as a possible consequence of decreased JNK phosphorylation, LPS-induced AP-1 transactivation was also attenuated in RAW 264 cells transiently transfected with Bcl-xL compared with vector-transfected control macrophages. To our knowledge, this is the first report demonstrating that Bcl-xL can inhibit LPS-induced expression of various cytokines in macrophages. These data suggest that in addition to its antiapoptotic function, Bcl-xL may also play a role in the resolution of inflammation by preventing macrophage release of excess and, therefore, potentially dangerous levels of inflammatory cytokines.
Materials and Methods
Escherichia coli K235 LPS (<0.008% protein) was prepared using the method of McIntire et al. (20). Propidium iodide was obtained from Boeh-ringer-Mannheim (Indianapolis, IN). Ab pairs for MIP-2 ELISA were purchased from R&D Systems (Minneapolis, MN). DuoSet ELISA kits for TNF-α and JE/MCP-1 were from PharMingen (San Diego, CA). All other chemicals were obtained from Sigma (St. Louis, MO) and were the highest analytical grade possible.
Culture of RAW 264 macrophage cell line
The RAW 264 mouse macrophage cell line was obtained from Riken Cell Bank (Tsukuba, Japan). A description of the RAW 264-Bcl-xL transfectants and the appropriate neomycin-vector control transfectants used in this paper was previously published by Okada et al. (16). Briefly, Bcl-xL transfectants were generated by cotransfection of the plasmid Bcl-xL-pEF-BOS (21) and the plasmid pST-neoB (22), which carries a neomycin-resistance gene. The following clones were used in this paper: 33-1, 35-1, and 35-3 Bcl-xL transfectants, and 33-3 pST-neoB transfectant. Among the Bcl-xL transfectants, clone 35-1 and clone 35-3 represent clones with the highest and lowest levels of Bcl-xL expression, respectively. Parental RAW 264 cells and/or clone 33-3 (pST-neoB transfectant) were used as controls throughout the experiment. In addition to clone 33-3, two other pST-neoB transfectants were tested for cytokine production with similar results (data not shown). Cell morphology and proliferative behavior of the transfectants and the parental RAW 264 cells was comparable.
RAW 264 macrophages were grown in RPMI 1640 supplemented with 10% FCS (HyClone, Logan, UT), 2 mM l-glutamine, 30 mM HEPES, 0.4% sodium bicarbonate and penicillin-streptomycin (100 IU/ml and 100 μg/ml, respectively). The medium was supplemented by 300 μg/ml G418 sulfate (Calbiochem, La Jolla, CA) for the transfected clones. Macrophages were plated in six-well plates (Corning Glass, Corning, NY) at a density of 1.5 million cells per well in a volume of 1.5 ml. After 3 h of adherence, parental RAW 264 macrophages and the transfectants were treated with various agents for the indicated time periods.
Lactate dehydrogenase assay
In cytokine release studies and transient transfection experiments, the results were standardized by the total lactic dehydrogenase (LDH) activity measured after detergent treatment of the macrophages in the same wells. LDH activity, which reflects the total cellular mass in a well, was measured according to the method of Wroblewski and LaDue (23) with minor modification (24).
DNA analysis by flow cytometry
For selective and quantitative determination of apoptosis, a modification of the flow cytometric DNA analysis published by Nicoletti et al. was utilized (25) as described previously (26). This method quantifies the percentage of apoptotic cells whose DNA content is lower than that of diploid cells. Due to DNA loss, apoptotic cells are represented by a distinct and quantifiable subdiploid peak in the fluorescence histogram. All measurements were performed with the same instrument settings, and at least 10,000 cells were measured in every sample.
Measurement of cytokines in cell culture supernatants or lysates
Macrophage supernatants were harvested at indicated time points and stored at −70°C. The levels of TNF-α, JE/MCP-1, and MIP-2 released by macrophages were measured using ELISA kits or Ab pairs according to the manufacturer’s instructions. TNF bioactivity in culture supernatants was also measured by a previously described, standard TNF cytotoxicity assay (27).
Analysis of macrophage mRNA by RT-PCR and Northern blot analysis
Total RNA was isolated from macrophage cultures, and the relative quantities of mRNA for hypoxanthine-guanine phosphoribosyltransferase (HPRT), TNF-α, IL-1β, JE/MCP-1, and MIP-2 were determined by RT-PCR, as described previously in detail (28). The optimal cycle number for each gene under nonsaturating conditions was determined empirically. Amplified products were electrophoresed and transferred to Hybond N+ membranes (Amersham, Arlington Heights, IL) by standard Southern blotting techniques. DNA was cross-linked by exposure to UV light, baked onto a nylon membrane, and hybridized with an internal oligonucleotide probe. Labeling of the probe and subsequent detection of bound probe was conducted using an enhanced chemiluminescence system (Amersham). The probes and primer combinations for JE/MCP-1, MIP-2, TNF-α, and IL-1β have been described elsewhere (29, 30).
To quantify TNF-α mRNA stability in RAW 264 and the transfectants, the relative amounts of the RT-PCR products were determined by a semiquantitative approach, detailed previously (31).
The levels of Bcl-xL expression in various transfectants were assessed by Northern blot analysis, as previously described (16).
Preparation of nuclear extracts and EMSA
Nuclear extracts were prepared according to Dignam et al. (32) with small modifications, as published previously (33). Briefly, the NF-κB-specific oligonucleotide 5′-AGCTCAGAGGGGGACTTTCCGAGAG-3′ from the murine Igκ-light chain gene enhancer and the AP-1-specific oligonucleotide 5′-CGCTTGATGAGTCAGCCCGAA-3′ were synthesized by the Biomedical Instrumentation Center Synthesis and Sequencing Facility (Uniformed Services University of the Health Sciences, Bethesda, MD), and 32P labeled with Klenow fragment using the oligolabeling kit (Amersham Pharmacia Biotech, Piscataway, NJ) according to the manufacturer’s instructions. Protein concentration was determined using the Bio-Rad assay kit (Bio-Rad, Hercules, CA). Nuclear extracts (4 μg) were incubated with 0.2 ng radiolabeled DNA probe in a binding buffer (final volume, 20 μl) containing 2 μg poly(dI-dC) (Amersham Pharmacia Biotech), 10 mM Tris-HCl (pH 7.9), 50 mM KCl, 4% glycerol, 1 mM DTT, and 0.25 mg/ml BSA for 30 min at room temperature. For supershift analyses, nuclear extracts were first incubated with 1 μl of the respective antiserum against members of the NF-κB family for 5 min at room temperature in the binding buffer, followed by the addition of a radiolabeled probe. After incubation, a portion of each reaction (18 μl) was loaded onto a 5% nondenaturing polyacrylamide gel, and the DNA-protein complexes were resolved from free oligonucleotide by electrophoresis (0.25 × Tris borate/EDTA, 150 V, 2 h). The gels were dried (80°C, 2 h) and exposed to x-ray films (X-OMAT AR, Eastman Kodak, Rochester, NY).
Preparation of cellular extracts and Western blot analysis
Cellular extracts were prepared as previously described (34). Twenty micrograms of total protein in Laemmli buffer was boiled for 5 min, resolved by electrophoresis in SDS-12% polyacrylamide gels in Tris/glycine/SDS buffer (25 mM Tris, 250 mM glycine, 0.1% SDS), and blotted onto Immunobilon P transfer membranes (100 V, 1.5 h, 4°C). After blocking for 2 h in TBS-T (20 mM Tris-HCl, 150 mM NaCl, 0.1% Tween 20) containing 1% gelatin and 5% nonfat milk, membranes were washed three times in TBS-T and probed for 1.5 h with the respective Abs diluted in TBS-T/0.5% nonfat milk. Rabbit anti-phospho-p38, anti-phospho-ERK1,2, anti-phospho-JNK1,2 Abs (anti-active Ab, Promega, Madison, WI), and rabbit pAb against IκB-α/MAD3 (Santa Cruz Biotechnology, Santa Cruz, CA) were used at a dilution of 1:3,000 and 1:1,000, respectively. Rabbit anti-total-p38 Ab (Santa Cruz Biotechnology, Santa Cruz, CA) was used at a dilution of 1:900. Following washing three times (15 min each time) in TBS-T, membranes were incubated for 1 h with secondary HRP-conjugated donkey anti-rabbit IgG (Promega, 1:10,000 dilution), washed 5 times (10 min each time) in TBS-T, and bands were visualized using enhanced chemiluminescence reagents (Amersham Pharmacia Biotech) according to manufacturer’s instructions.
Transient transfection experiments
The AP-1 luciferase reporter plasmid (p(AP-1)3LdLuc) was constructed as described previously (35). For transient transfections, RAW 264 cells were seeded onto 24-well plates (Costar, Cambridge, MA) at a density of 1 × 105 cells/well in RPMI 1640/10% FCS, incubated overnight, and cotransfected with the AP-1 reporter (0.3 μg/well) for 3 h and either 0.3 μg/well pEF-BOS (vector-transfected group) or 0.3 μg/well Bcl-xL-pEF-BOS-(Bcl-xL-transfected group) constructs using 5 μl/well of SuperFect transfection reagent (Qiagen, Chatsworth, CA). The total amount of plasmid DNA was equalized to 0.9 μg/well by adding corresponding amounts of pBluescript II SK+/+ phagemid (Stratagene, La Jolla, CA). Following transfections, cells were stimulated with LPS or medium for 18 h, washed twice with ice-cold PBS, and lysed in a lysis buffer (Analytical Luminescence Laboratory, Sparks, MD) for 30 min with constant shaking. The supernatant (20 μl) was assayed in 250 μl of assay buffer (25 mM glycylglycine, 15 mM MgSO4, 1% Triton X-100, 1 mM ATP) using a Monolight 2010 luminometer (Analytical Luminescence Laboratory). Following the injection of 100 μl of luciferin (0.3 mg/ml, Analytical Luminescence Laboratory) into the test tube, light emission was measured in 10 s intervals. Luciferase activities were normalized for LDH activity values measured from the very same cellular extracts. Results are expressed as mean values ± SEM of the corrected luciferase activities in medium- and LPS-treated cells, for both vector- and Bcl-xL-transfected groups. The relative fold-stimulation parameter, which reflects the transactivation potential of AP-1, was calculated by dividing the mean values of LPS-treated samples by the mean values of medium-treated samples.
Unless otherwise stated, results were expressed as arithmetic means of triplicate samples ± SEM, obtained from a representative experiment. All experiments were repeated two to three times with similar results. Wherever statistically significant differences are shown or mentioned in the text, a one-way ANOVA combined with a Tukey test was used at the significance level of p < 0.05. For regression analysis and densitometry, SigmaStat 2.0 for Windows (SPSS, Chicago, IL) and National Institutes of Health Image for Windows software (Scion, Frederick, MD) was used, respectively.
Overexpression of Bcl-xL protects macrophages against gliotoxin-induced apoptosis
The ectopic expression of Bcl-xL in the transfectants used throughout this study was confirmed by Northern blot analyses (Fig. 1,A). Clones 33-1, 35-1, and 35-3 are Bcl-xL transfectants and indeed, express significantly more Bcl-xL mRNA than either the control transfectant (33-3Neo) or the RAW 264 parental cell line from which all the clones were originally derived. To ensure that these transfectants express levels of ectopic Bcl-xL sufficient to exert an antiapoptotic state, parental RAW 264 cells, neomycin-vector control, and a Bcl-xL-overexpressing clone were exposed to gliotoxin, a potent inducer of apoptosis in macrophages (26, 36). Gliotoxin (5 μg/ml) evoked ∼75–80% apoptosis in parental RAW 264 macrophages and in neomycin-vector control cells as measured by DNA analysis. In contrast, Bcl-xL-overexpressing macrophages were less sensitive to gliotoxin-induced apoptotic cell death as evidenced by a significantly greater number of diploid cells (M1) and a marked reduction of cells in the subdiploid (M2) fraction (Fig. 1 B).
LPS-induced cytokine mRNA expression and protein secretion is decreased in RAW 264 clones that overexpress Bcl-xL
Because Bcl-xL has been reported to inhibit NF-κB-dependent gene expression in endothelial cells (17), we investigated whether Bcl-xL can regulate the levels of cytokines released into culture supernatants after LPS stimulation. To test this hypothesis, control macrophages and clones that overexpress Bcl-xL were compared. After 3 and 6 h of treatment with 10 ng/ml LPS, RAW 264 cells and neomycin-vector controls secreted significantly higher levels of TNF-α, JE/MCP-1, and MIP-2 compared with Bcl-xL-overexpressing transfectants, as measured by ELISA (Fig. 2,A). Moreover, the level of Bcl-xL mRNA expression correlated with the ability to inhibit cytokine secretion from macrophages. Clone 35-1, which expresses higher levels of Bcl-xL than clone 35-3 (Fig. 1,A), secreted significantly lower levels of TNF-α than clone 35-3 (Fig. 2,B). TNF-α-induced cytolytic activity was measured in this experiment by bioassay (Fig. 2,B), rather then by ELISA (Fig. 2 A), supporting the finding that overexpression of Bcl-xL affects the level of bioactive TNF-α induced by LPS in macrophage cultures.
To investigate further how overexpression of Bcl-xL results in a decrease in the levels of cytokines secreted by LPS-treated macrophages, we analyzed possible changes in the steady-state levels of different cytokine mRNA species. Expression of four cytokine genes was studied by RT-PCR in RAW 264 cells, neomycin-vector control, and Bcl-xL transfectants. Bcl-xL-transfected clones showed decreased expression of TNF-α, IL-1β, JE/MCP-1, MIP-2 (Fig. 3,A), and cyclooxygenase-2 (data not shown) compared with neomycin-vector control cells or the parental RAW 264 macrophages, in contrast to the stable expression of the housekeeping gene, HPRT, which did not change as a result of addition of LPS and/or ectopic expression of Bcl-xL. Moreover, we have recently reported that LPS treatment of primary macrophages results in a modest down-regulation of Toll-like receptor (TLR)-4 steady-state mRNA, while steady-state TLR-2 mRNA is substantially up-regulated (37). This regulation of steady-state TLR-2 (Fig. 3 A) and TLR-4 mRNA (data not shown) was also confirmed in RAW 264 macrophages, whether Bcl-xL was overexpressed. Thus, ectopic expression of Bcl-xL does not exert a global suppressive effect on transcription.
Inhibition of LPS-induced TNF-α mRNA levels in Bcl-xL-overexpressing clones is not due to decreased mRNA stability
One mechanism that could account for decreased steady-state levels of different cytokine mRNAs in Bcl-xL-overexpressing cells might involve increased rates of degradation of those mRNA species. Posttranscriptional mechanisms involving mRNA stabilization have been postulated to play a significant role in TNF-α mRNA expression (12, 38). To determine whether Bcl-xL overexpression results in a more rapid turnover of TNF-α mRNA, we compared rates of mRNA degradation in LPS-treated RAW 264 cells and transfectants after inhibiting the transcription with actinomycin D. The t1/2 of TNF-α mRNA proved to be almost exactly the same in parental macrophages and control and Bcl-xL transfectants (Fig. 3,B), and was consistent with a previous report, which measured the t1/2 for TNF-α mRNA to be 1.1 h in RAW 264 macrophages (39). In Fig. 3 B, it is also shown that the steady-state levels of TNF-α mRNA in Bcl-xL-transfected cells are clearly lower at any time point than in parental cells or in neomycin-vector controls. These data demonstrate that at least in the case of TNF-α mRNA, decreased expression of this cytokine in Bcl-xL transfectants cannot be attributed to increased mRNA degradation.
Stabilization of IκBα and decreased NF-κB translocation in Bcl-xL-overexpressing macrophages stimulated with LPS
Many inflammatory cytokine genes are under the control of a key transcription factor, NF-κB (40). Because a previous study demonstrated an interaction of Bcl-xL with NF-κB in endothelial cells (17), the possible inhibitory effect of Bcl-xL on the activation of this transcription factor was next studied by EMSA. In RAW 264 macrophages (data not shown), neomycin-vector controls, and Bcl-xL transfectants, 10 ng/ml LPS induced NF-κB translocation and binding to the NF-κB-specific consensus oligonucleotide by 15 min (Fig. 4,A). At 30 and 60 min following LPS stimulation, NF-κB translocation was only slightly diminished in Bcl-xL transfectants compared with that observed in neomycin-vector controls (Fig. 4,A). NF-κB binding was concentration dependent and reached a plateau at a concentration range of 10–100 ng/ml LPS in both controls and Bcl-xL transfectants (data not shown). However, this relatively modest degree of inhibition was highly reproducible. Therefore, to confirm this effect of overexpressed Bcl-xL on NF-κB translocation in macrophages, the kinetics of LPS-induced IκBα degradation was also measured by Western analysis of total cellular extracts. Stimulation with 10 ng/ml LPS for 15 min was sufficient to induce approximately the same degree of IκBα degradation in vector-control cells and Bcl-xL-overexpressing macrophages (Fig. 4,B) consistent with NF-κB binding data (Fig. 4,A). However, in Bcl-xL transfectants, the amount of IκBα returned to prestimulation levels by 45 min after LPS stimulation, in contrast to neomycin-vector control cells, where IκBα levels remained much lower, even after 60 min, consistent with the observed differences between control and Bcl-xL macrophages in the EMSA (Fig. 4 A). Thus, the effect of Bcl-xL appears to be upstream of NF-κB translocation to the nucleus.
To investigate the possibility that ectopic expression of Bcl-xL affects the composition of NF-κB complexes and therefore, their transactivating potency, supershift experiments were also conducted. Preincubation with Abs against members of the Rel family proteins p50, p65, p52, c-Rel, and Rel B caused the appearance of slower migrating bands compared with the major LPS-inducible band, indicating the presence of these factors in the complex (Fig. 4 C). However, no differences in the composition of LPS-induced NF-κB complexes from Bcl-xL transfectants and neomycin-vector control cells were discernable.
Decreased LPS-induced phosphorylation of p38 MAP kinase and JNK1/2 in Bcl-xL transfectants
In addition to NF-κB activation, other intracellular pathways are also important in mediating LPS-induced effects. Phosphorylation of MAP kinases has been reported to play an essential role in the activation of these enzymes and for the induction of many cytokines (4). Western analyses were conducted to examine the potential involvement of different MAP kinases in the inhibition of cytokine gene expression in Bcl-xL-transfected macrophages. After stimulation with 10 ng/ml LPS, a clear diminution in the extent of p38 MAP kinase and JNK1/2 phosphorylation was detected in cell lysates prepared from Bcl-xL transfectants, compared with neomycin-vector control cells, at all time points examined (Fig. 5). Moreover, the phosphorylation of p38 and JNK1/2 in Bcl-xL transfectants was more transient, compared with control macrophages. Although subtle differences between control- and Bcl-xL-transfected cells were also observed in the case of ERK1/2, phosphorylation of these MAP kinases were affected to a lesser extent than p38 kinases and JNK1/2 (Fig. 5). Finally, total p38 MAP kinase levels measured in the very same cellular extracts did not differ significantly, indicating that differences in phosphorylated MAP kinase levels cannot be attributed to loading variations. These data indicate that overexpression of Bcl-xL interferes with the LPS-induced p38 and JNK1/2 phosphorylation in RAW 264 macrophages.
Inhibition of LPS-induced AP-1 transactivation and DNA-binding by Bcl-xL overexpression
In addition to NF-κB, transcription factor AP-1 has also been shown to be important in the induction of cytokines by LPS in macrophages (4). To determine whether a discernible decrease in AP-1 binding could be detected in the Bcl-xL-overexpressing stable transfectants, EMSAs were performed. Consistent with the diminution in NF-κB translocation in Bcl-xL transfectants, Fig. 6 shows that in response to LPS stimulation, binding of nuclear AP-1 was decreased in the Bcl-xL-overexpressing clone compared with that observed in the case of the neomycin-vector control clone.
It is also known that JNK phosphorylates c-Jun, a component of the AP-1 complex, which results in a significant increase in the transactivating potency of AP-1. Because LPS-induced JNK activation is strongly attenuated by Bcl-xL overexpression (Fig. 5), we investigated whether Bcl-xL overexpression has a possible inhibitory effect on AP-1 transactivation in RAW 264 macrophages. To address this question, RAW 264 cells were transiently cotransfected with an AP-1 luciferase reporter and either the Bcl-xL-containing expression vector or the empty vector. After stimulation with 10 ng/ml LPS, AP-1 transactivation was significantly less in RAW 264 cells cotransfected with Bcl-xL-overexpressing construct compared with vector-transfected cells (Table I). These data are consistent with the observed inhibition of LPS-mediated JNK phosphorylation and AP-1 DNA binding in cells overexpressing Bcl-xL.
|Plasmid Constructs .||AP-1 Transactivationa (Relative Fold Stimulation) .||.|
|.||Expt. I .||Expt. II .|
|pEF-BOS+ p(AP-1)3-LdLuc||2.6 ± 0.2||2.8 ± 0.3|
|Bcl-xL-pEF-BOS+ p(AP-1)3-LdLuc||1.6 ± 0.2*||2.0 ± 0.1*|
|Plasmid Constructs .||AP-1 Transactivationa (Relative Fold Stimulation) .||.|
|.||Expt. I .||Expt. II .|
|pEF-BOS+ p(AP-1)3-LdLuc||2.6 ± 0.2||2.8 ± 0.3|
|Bcl-xL-pEF-BOS+ p(AP-1)3-LdLuc||1.6 ± 0.2*||2.0 ± 0.1*|
RAW 264 cells were transiently cotransfected with 0.3 μg of p(AP-1)3-LdLuc and either pEF-BOS or Bcl-xL-pEF-BOS (0.3 μg each) as described in Materials and Methods. After stimulation with medium or 10 ng/ml LPS for 18 h, lysates were assayed for luciferase and LDH activity. Luciferase activity in each well was normalized for LDH activity, which represents the total cellular mass in the same well. The data are presented as relative fold stimulation ± SEM compared with medium-treated control cells.
, Statistically significant difference (n = 6, p < 0.01) compared to vector-transfected cells.
In addition to its well characterized developmental and pathological role in the immune system (41), apoptosis has also been proposed to be important in the resolution of inflammation by eliminating activated macrophages when there is no longer a need for them (42). It is equally essential, though, that macrophages possess mechanisms that afford protection against the toxic substances they produce to combat pathogens. These events are under strict control by pro- and antiapoptotic regulatory factors, such as the members of the Bcl family. In fact, overexpression of Bcl-2 (8) and Bcl-xL (16) have been reported to protect macrophages against NO-mediated apoptosis following LPS treatment. Although Bcl-2 and Bcl-xL share many common functions in different cell types, their expression patterns and possibly their functions are cell type specific (13). In RAW 264 cells and also in primary macrophages, steady-state levels of Bcl-2 mRNA are unaffected by LPS stimulation, in contrast to Bcl-xL mRNA, which is rapidly up-regulated and maintained at high levels for up to 48 h (16). These data suggest that Bcl-xL, rather than Bcl-2, may have a physiologically relevant regulatory function in macrophages.
In this study, we report that in addition to its antiapoptotic role, Bcl-xL may also regulate gene expression and secretion of proinflammatory and chemotactic cytokines in macrophages, a process that could potentially attenuate inflammation. Cytokine production by macrophages is regulated both transcriptionally and/or translationally (11, 12). Our data support the hypothesis that inhibition of cytokine expression by Bcl-xL occurs, in part, at the level of transcription, as evidenced by diminished steady-state mRNA for the four LPS-inducible genes examined. This inhibition is observed in the absence of an effect on the expression of the housekeeping gene, HPRT. In addition, not all LPS-regulated genes were affected in Bcl-xL-overexpressing clones; LPS-induced modulation of mouse TLR-2 (Fig. 3 A) and TLR-4 (data not shown) gene expression by LPS treatment of RAW 264 cells was also unaffected by ectopic expression of Bcl-xL. Furthermore, no differences in LPS plus IFN-γ-induced NO production were observed among RAW 264 cells, Bcl-xL transfectants, and neomycin-vector control macrophages (16), supporting the conclusion that the effect of Bcl-xL is gene specific and cannot be attributed to a general inhibition of transcription or due to a toxic effect caused by transfection.
In macrophages, NF-κB is a principle transactivating factor involved in the induction of many LPS-inducible cytokines (5). A substantial body of evidence suggests that members of the Bcl family, such as Bcl-2, interact with NF-κB to regulate apoptosis in various cell types (43, 44, 45). In primary endothelial cells, both Bcl-2 and Bcl-xL were shown to decrease NF-κB-dependent gene expression (17). In contrast, there are also numerous reports that show no effect of Bcl-2 on NF-κB activation in L929 cells, MCF7 breast carcinoma cells, or Jurkat T cells (46, 47, 48). Similarly, Bcl-xL has no effect on NF-κB activation in MCF7 cells (49). These contradictory data, which suggest cell-type-specific functions for Bcl family members, prompted us to investigate the effect of Bcl-xL on NF-κB in macrophages. Overexpression of Bcl-xL only modestly decreased the amount of translocated NF-κB in macrophages (Fig. 4,A). Because it has been reported that Bcl-2 overexpression decreases the nuclear levels of transactivator p50/p65 heterodimers in favor of the repressive p50/50 homodimers in a T cell hybridoma (43), supershift analyses were employed to determine the subunit composition of NF-κB complexes in Bcl-xL- and vector-transfected macrophages. In contrast to the above-mentioned report, we failed to find any discernible differences in the composition of NF-κB complexes among the clones analyzed (Fig. 4 C).
Because the translocation of NF-κB is dependent upon the phosphorylation and subsequent degradation of IκB species, we also studied the effect of ectopic expression of Bcl-xL on IκBα degradation. Although the kinetics and degree of IκBα degradation was similar in both vector-control and Bcl-xL macrophages at 5 and 15 min after LPS stimulation, at 45 and 60 min, IκBα levels had already returned to prestimulation levels in the Bcl-xL clone, in contrast to the control, where it was much less abundant. Overexpression of Bcl-2 and Bcl-xL in endothelial cells has been reported to result in a decrease in NF-κB-dependent gene expression, and in the case of Bcl-2, the stabilization of a slower migrating, presumably hyperphosphorylated form of IκBα (17). However, we did not detect a slower migrating form, indicating that in murine macrophages, attenuated IκBα degradation may not be a result of hyperphosphorylation.
One possible explanation for the inhibitory action of Bcl-xL on IκBα degradation, and subsequently on translocation of NF-κB, might be related to its reported protease-inhibiting effect. IκBα contains a conserved IL-1 converting enzyme-like protease consensus site, which can be cleaved in vitro by caspase-3, but not by caspases-1 or -2 (50). Bcl-xL has been postulated to be a negative regulator of caspase-3 activation through its binding to Apaf-1 (51), thereby inhibiting the formation of a complex that activates caspase-3 (52). Alternatively, the failure to activate caspase-3 may prevent the proteolytic activation of mitogen-activated kinase kinase kinase 1 (MEKK1) (53, 54), one of the upstream kinases that regulates the IκB kinase complex (55). Inhibition of MEKK1 activity would also lead to decreased JNK activation, consistent with our results demonstrating diminished JNK1/2 phosphorylation in LPS-stimulated Bcl-xL transfectants (Fig. 5).
Another possible explanation for the effect of Bcl-xL on IκBα degradation could likely be the indirect antioxidant function of Bcl family members (56). Overexpression of Bcl-xL in astrocytes has been reported to increase levels of glutathione, one of the most important antioxidant molecules in mammalian cells (57). In addition, overexpression of γ-glutamylcystein synthetase, the rate-limiting enzyme for glutathione synthesis, suppressed IκBα degradation and NF-κB activation completely (58). Other antioxidants, such as pyrrolidone dithiocarbamate, are also potent inhibitors of NF-κB translocation (59, 60).
MAP kinases are also particularly important mediators of LPS-induced effects and are involved in the regulation of expression of many macrophage-derived cytokines (4). However, evidence suggests that although MAP kinases can regulate the transactivating potency of NF-κB, LPS-induced nuclear translocation of NF-κB may not be a direct consequence of MAP kinase activation (61). This prompted us to investigate the three subclasses of MAP kinases as potential targets of the inhibitory effect of Bcl-xL overexpression on LPS-induced cytokine production. Ectopic expression of Bcl-xL significantly inhibited the phosphorylation of p38 kinase and JNK1/2 (Fig. 5). These MAP kinases are often activated by stressors such as osmotic shock, inflammation, ionizing radiation, DNA-damaging agents, or various chemical inducers of apoptosis (4). Recent evidence suggests that Bcl-xL inhibits JNK-mediated apoptosis, and acts upstream of JNK and p38 phosphorylation (62, 63). Because both p38 and JNK activation are caspase-3 dependent (64) and Bcl-xL inhibits caspase-3 activation (51), which has been implicated in the activation of MEKK1 (53, 54), the interaction of these pathways could lead to mitigated JNK phosphorylation. Alternatively, as has been shown in methylmethane sulfonate-induced apoptosis, Bcl-xL blocks JNK activation by inhibiting phosphorylation of related adhesion focal tyrosine kinase/proline-rich tyrosine kinase 2 in a caspase-independent fashion (65).
The above mentioned antioxidant properties of Bcl-xL (56) might also contribute to its inhibitory effect on p38 and JNK phosphorylation. Reactive oxygen species have been demonstrated to induce p38 MAP kinase phosphorylation, and in astrocytes, antioxidants were able to block this effect through an increase in phosphatase activity (66). MEKK1-dependent JNK activation has also been found to be blocked by antioxidants (62).
Consistent with the marked inhibition of JNK phosphorylation observed in Bcl-xL transfectants compared with neomycin-vector control cells, LPS-induced AP-1 binding (Fig. 6) and transactivation (Table I) were also attenuated by Bcl-xL overexpression. These data suggest that the strong reduction of LPS-induced cytokine production in Bcl-xL transfectants is likely to be the result of the inhibition of JNK by Bcl-xL.
Despite the observation that steady-state levels of TNF-α mRNA were decreased to a lesser extent compared with the change in mRNA levels of the other cytokine genes examined, secreted levels of TNF-α were markedly inhibited (Fig. 2 A), raising the additional possibility of posttranscriptional regulation. Although transcriptional control of TNF-α expression by p38 has been reported (67), there is a sequence motif in the 3′ untranslated region of the TNF-α mRNA that represses translation in unstimulated macrophages but not in LPS-stimulated cells (12), and that involves the activation of both p38 and JNK (68, 69).
The involvement of p38 in the induction of IL-1β has also been reported. Treatment of macrophages with SB203580, a specific p38 inhibitor, reduced LPS-induced IL-1β mRNA synthesis in RAW 264 cells by 80% (70). Although we know of no data in macrophages that directly supports the previously reported involvement of p38 in LPS-induced JE/MCP-1 in human umbilical vein endothelial cells (71) and mesangial cells (72), induction of this cytokine by IL-1β and TNF-α involves p38 kinase and is independent of ERK or JNK. It has also been reported that in RAW 264 cells, p38 MAP kinase targets the CAAT/enhancer-binding protein (NFIL-6) (70), which has been reported to be important in the induction of JE/MCP-1 in myeloid cells (73, 74) and the regulation of IL-1β in macrophages (70). MIP-2 expression, also attenuated by Bcl-xL overexpression in macrophages, has been demonstrated to be ERK- and p38-dependent in peritoneal neutrophils stimulated by staurosporine (75). All four of the Bcl-xL-sensitive genes examined are dependent, to varying extents, on NF-κB (5). It has been reported that p38 is required for NF-κB-dependent gene expression through phosphorylation of a basal transcription factor, the TATA-binding protein that in turn, interacts with the p65 subunit of NF-κB to promote NF-κB-dependent gene expression (61). Finally, the activation of a variety of other transcription factors (e.g., NFIL-6, cAMP response element binding protein, etc.) are also MAP-kinase dependent (76). Thus, diminished availability of such factors in Bcl-xL transfectants could well contribute to diminished cytokine expression after LPS stimulation.
Thus, inhibition of NF-κB translocation and AP-1 activation, combined with decreased p38 kinase and JNK1/2 activation by Bcl-xL, could explain the mitigated cytokine production in Bcl-xL transfectants. Taken collectively, our data suggest that LPS-induced up-regulation of Bcl-xL in macrophages may not only protect against NO-mediated self-destruction (16), but also can contribute to the resolution of inflammation by mitigating excessive and potentially harmful cytokine release.
This work was supported by National Institutes of Health Grants AI-18797 and AI-44936 (to S.N.V.). The opinions or assertions contained within are the private views of the authors and should not be construed as official or necessarily reflecting the views of the Uniformed Services University of the Health Sciences or the Department of Defense.
Abbreviations used in this paper: MAP, mitogen-activated protein; ERK, extracellular signal-regulated kinase; JNK, c-Jun N-terminal kinase; LDH, lactic dehydrogenase; MCP, monocyte chemoattractant protein; MIP, macrophage inflammatory protein; IκBα, inhibitory κBα; MEKK1, mitogen-activated kinase kinase kinase 1; TLR, Toll-like receptor; HPRT, hypoxanthine-guanine phosphoribosyltransferase.