Lipopolysaccharide (LPS) is a powerful stimulator of macrophages and induces apoptosis in these cells. Using primary cultures of bone marrow-derived macrophages, we found that the autocrine production of tumor necrosis factor-α (TNF-α) has a major function in LPS-induced apoptosis. LPS activates PKC and regulates the different mitogen-activated protein kinases (MAPK). We aimed to determine its involvement either in the secretion of TNF-α or in the induction of apoptosis. Using specific inhibitors and mice with the gene for PKCϵ disrupted, we found that LPS-induced TNF-α-dependent apoptosis is mostly mediated by PKCϵ, which is not directly involved in the signaling mechanism of apoptosis but rather in the process of TNF-α secretion. In our cell model, all three MAPKs were involved in the regulation of TNF-α secretion, but at different levels. JNK mainly regulates TNF-α transcription and apoptosis, whereas ERK and p38 contribute to the regulation of TNF-α production, probably through posttranscriptional mechanisms. Only JNK activity is mediated by PKCϵ in response to LPS and so plays a major role in TNF-α secretion and LPS-induced apoptosis. We demonstrated in macrophages that LPS involving PKCϵ regulates JNK activity and produces TNF-α, which induces apoptosis.
- cellular activation
- protein kinases/phosphatases
- signal transduction
lipopolysaccharide, isolated from the outer membrane of Gram-negative bacteria, is one of the strongest stimulators of macrophages and leads to the secretion of nitrogen intermediates, prostaglandins, and cytokines. The secretion of tumor necrosis factor-α (TNF-α), interleukin 1β (IL-1β), IL-6, and IL-12 leads to rapid induction and amplification of the host response to infection (10, 37, 42, 48). Lipopolysaccharide (LPS) has a clearly inflammatory effect and is also thought to induce apoptosis in several cell types, including macrophages (1, 20, 60). Whereas the inflammatory response is mediated through several secreted factors, the cytotoxic effects of LPS on macrophages are ascribed to the production of TNF-α or nitric oxide (NO) (3, 43, 53). Using bone marrow-derived macrophages from mice with disrupted genes for the TNF receptors or the inducible NO synthase, we determined that LPS-induced apoptosis results from two independent mechanisms: first and mainly, through the autocrine secretion of TNF-α (early apoptotic events) and second, through the production of NO (late apoptosis) (69).
LPS interacts with LPS-binding protein (LBP), thus allowing binding to CD14 and association with the Toll-like receptor 4 (Tlr4) containing an intracellular signaling domain (8). Binding of LPS to these receptors results in the activation of a number of signaling cascades, such as PKC and mitogen-activated protein kinase (MAPK). However, the precise mechanism by which LPS triggers apoptosis or the release of cytokines such as TNF-α is unclear (46).
The PKC family consists of several isoforms that fall into three main groups based on their primary structure and activation requirements. In previous studies, we described how macrophages derived from bone marrow express only three PKC isoforms: PKCβ1 (conventional), PKCϵ (novel), and PKCξ (atypical) (63). PKCϵ is involved in the regulation of important aspects of macrophage biology, in particular proliferation and macrophage activation (18, 62). However, LPS also activates the three major MAP kinase cascades in macrophages, namely the ERK, p38, and JNK pathways. All three pathways are linked to activation by LPS and subsequent cytokine gene expression (23, 26, 27, 50, 65).
Because of the importance of TNF-α-dependent apoptosis induced by LPS in macrophage biology, we attempted to identify the signaling pathways involved in this process. Here, we provide evidence that TNF-α-dependent apoptosis induced by LPS in macrophages is mediated by PKCϵ. Moreover, although all three MAP kinases are necessary for correct TNF-α regulation, only the activation of JNK is clearly mediated by PKCϵ. Therefore, JNK is mainly responsible for the TNF-α-dependent apoptosis induced by LPS. Both PKCϵ and MAPK are involved in TNF-α secretion induced by LPS and, thus, in the apoptosis induced by LPS. However, none of the three MAPKs modified the apoptotic events triggered after interaction of TNF-α with its receptor, as we showed by using recombinant TNF-α.
MATERIAL AND METHODS
Reagents. LPS was obtained from Sigma Chemical (St. Louis, MO). In several experiments, the results obtained with commercial LPS were compared with those from purified LPS kindly donated by Dr. C. Galanos (Max Planck Institute, Freiburg, Germany) (40), and no differences were found. Murine recombinant TNF-α was purchased from PrepoTech EC (London, UK). Bisindolymaleimide I (GF 109203X), PD98059, SB203580, and curcumin were obtained from Calbiochem (San Diego, CA). Gö6976 was a kind gift from Dr. A. García de Herreros (Institut Municipal d'Investigació Mèdica, Barcelona, Spain). The SP600125 was obtained from Tocris Cookson (Ellisville, MO). The phosphop38 MAP kinase (Thr180/Tyr182) antibody was obtained from Cell Signaling (Beverly, MA). The p38 MAP kinase antibody (sc-535) was purchased from Santa Cruz Biotechnology, (Santa Cruz, CA). Use of all reagents followed the manufacturer's recommendations.
Cell culture. Bone marrow-derived macrophages were isolated as previously described (13). Six-week-old BALB/C mice (Charles River Laboratories, Wilmington, MA) were killed by cervical dislocation, and both femurs were dissected free of adherent tissue. The ends of the bones were cut off, and the marrow tissue was flushed by irrigation with culture medium. The marrow plugs were dispersed by passing a 25-gauge needle through them, and the cells were suspended by vigorous pipetting and washed by centrifugation. Cells were cultured in plastic tissue culture dishes (150 mm) in 40 ml of DMEM containing 20% FBS and 30% L-cell-conditioned medium as a source of macrophage colony-stimulating factor (M-CSF). Once macrophages were 80% confluent (i.e., after 6 days of culture), they were deprived of L-cell-conditioned medium for 16-18 h and treated with LPS (Sigma) in the presence and absence of selective inhibitors. The cells were incubated at 37°C in a humidified 5% CO2 atmosphere. Macrophages from PKCϵ-knockout (KO) mice (11) were obtained in the same way. The use of animals was approved by the Comitè d'experimentació animal of the University of Barcelona with the number 2523.
Analysis of apoptosis. DNA fragmentation due to internucleosomal cleavage was determined as described elsewhere (67). Low molecular weight apoptotic DNA was measured by an ELISA technique (Cell Death Detection ELISA Plus, Boehringer Mannheim), which was directed against cytoplasmic histone-associated DNA fragments. Each stage was performed in triplicate, and the results were expressed as means ± SD. In some cases, apoptosis was determined by flow cytometry with 4',6'-diamidino-2-phenylindole (DAPI) staining or by DNA laddering as previously described (67).
Determination of TNF-α production. TNF-α secretion was measured by using a commercial murine TNF-α ELISA kit (Biosource, Camarillo, CA). Cells (5 × 105) were cultured in 24-well plates and stimulated with LPS (10 ng/ml) in the presence and absence of the indicated specific inhibitors. Supernatant samples were obtained 12 h later and frozen until subjected to ELISA analysis following the manufacturer's protocol.
Protein extraction and Western blot analysis. Western blots were performed as described elsewhere (63). For phosphop38 Western blot analysis, equal amounts of protein (60 μg) were separated on 10% SDS-PAGE. The proteins were then electrotransferred to nitrocellulose membranes (Hybond-ECL; Amersham, Arlington Heights, IL). The membranes were blocked for at least 1 h at room temperature in Tris-buffered saline-0.1% Tween-20 (TBS-T) containing 5% (wt/vol) nonfat dry milk and then incubated with TBS-T containing 5% BSA and the primary antibody (1:1,000) overnight at 4°C. After three washes of 5 min each with TBS-T, the membranes were incubated with peroxidase-conjugated anti-rabbit IgG (Cappel, Durham, NC) antibody for 1 h. After three washes of 5 min with TBS-T, enhanced chemiluminescence (ECL) detection was performed (Amersham), and the membranes were exposed to X-ray films (Amersham). The bands of interest were quantified by densitometry. The Western blot of p38 MAP kinase was performed as described for the phosphorylated form, with a minor modification: the incubation with primary antibody (1:2,000) was performed with TBS-T without 5% BSA.
Northern blot analysis. Total cellular RNA (20 μg), extracted with TRIZOL reagent (Life Technologies, Grand Island, NY), was separated in 1% agarose with 5 mM MOPS [3-(N-morpholino) propanesulfonic acid], pH 7.0/1 M formaldehyde buffer. The RNA was transferred overnight to a Hybond-XL nitrocellulose membrane (Amersham) and fixed by ultraviolet irradiation (150 mJ) (62). For TNF-α mRNA detection, we used the EcoRI/HindIII fragment of pSP65/TNFa kindly supplied by Dr. M. Nabholz (Institut Suisse de Recherches Expérimentales sur le Cancer, Epalinges, Switzerland). To study the expression of IL-1β, we obtained a probe by digesting the construct pGEM/IL-1β provided by Dr. R. Wilson (Glaxo Research and Development, Greenford, UK) with EcoRI/PstI. All probes were labeled with [32Pα]dCTP (Amersham) with the oligolabeling kit method (Pharmacia Biotech, Uppsala, Sweden). To check for differences in RNA loading, the expression of the 18S rRNA transcript was analyzed by using an 18S probe, as previously described (11). After the membranes were incubated for 18 h at 65°C in hybridization solution [5 × standard sodium citrate (SSC), ×5 Denhart's, 1% SDS, and 106 cpm/ml of 32P-labeled probe], they were exposed to Kodak X-AR films (Kodak, Rochester, NY). The pertinent bands were quantified with a molecular analyst system (Bio-Rad).
Determination of MAPK activity. ERK activity was analyzed by in-gel-kinase assay, as described (63). After electrophoresis, the gel was dried, exposed to X-ray films (Kodak), and quantified with a Bio-Rad molecular analyst.
JNK activity was assayed as described elsewhere (61) with minor modifications. Briefly, the cells were washed with PBS and lysed in cold lysis buffer (1% NP-40, 20 mM HEPES-Na, pH 7.5, 10 mM EGTA, 40 mM β-glycerophosphate, 25 mM MgCl2, 2 mM sodium orthovanadate, 1 mM DTT, 0.5 mM PMSF, 1 μg/ml aprotinin, 1 μg/ml leupeptin, and 1 μg/ml iodacetamide). Total protein (150 μg) was mixed with 75 μlof 20% protein A-Sepharose and 1 μl of anti-JNK1 antibody (sc-474, Santa Cruz Biotechnology) in a total volume of 500 μl. The samples were rotated for 2 h at 4°C. The immunocomplexes were washed three times with cold PBS supplemented with 1% NP-40 and 2 mM sodium orthovanadate, once with cold JNK buffer (20 mM HEPES-Na, pH 7.5, 20 mM β-glycerophosphate, 20 mM MgCl2, 0.1 mM sodium orthovanadate, 2 mM DTT) and resuspended in JNK reaction buffer [JNK buffer supplemented with 1 μg GST-c-Jun (1-169) (Calbiochem)] and as a substrate, 20 μM ATP, 1 μCi γ-[32P]ATP. The reaction was run for 30 min at 30°C and was then stopped by adding 12 μl of ×5 Laemmli buffer. The samples were incubated for 3 min at 100°C and separated by 10% SDS-PAGE. After electrophoresis, the gels were fixed in isopropanol:water:acetic acid (25:65:10), dried, and exposed to Kodak X-AR films.
In these studies, we used bone marrow-derived macrophages, which are a homogeneous population of primary and quiescent cells (13). Bone marrow macrophages growing in the presence of M-CSF are unevenly distributed in the different phases of the cell cycle. Upon activation with LPS, macrophages stop at the G0/G1 phase of the cell cycle and die through the induction of apoptosis (11, 25). LPS-induced apoptosis was quantified by using an ELISA kit that measures the presence of histone-associated low molecular weight DNA fragments (Fig. 1). This technique, used for the quantitative measurement of apoptosis in this cellular model, correlates closely with other methods (67-69). In some experiments, apoptosis was also determined by flow cytometry with DAPI staining or by DNA laddering with qualitatively identical results. The induction of macrophage apoptosis by LPS was time and dose dependent (Fig. 1). The kinetics of induction of apoptosis was very fast with a maximal induction as early as 6 h after the start of LPS treatment. The levels of apoptosis did not increase any further after 6 h and up to 24 h of stimulation (Fig. 1A).
Maximal induction of apoptosis was obtained at a concentration of 100 ng/ml of LPS (Fig. 1B), a dose that saturates binding of LPS to its high-affinity receptor, CD14/TLR-4 (46). It has been reported that other molecules may also transfer signals at very high doses of LPS (1-10 μg/ml) (49). To avoid signaling through other LPS-receptors independent of CD14, in this study we used subsaturating amounts of LPS (1-10 ng/ml) that already induce apoptosis in macrophages (Fig. 1B). Moreover, since we were interested in the TNFα-dependent early apoptotic events in response to LPS, we performed all the studies 6-12 h after LPS treatment. Because inducible nitric oxide synthase (iNOS) protein expression and NO production after LPS treatment can only be detected after 12-24 h of treatment, any interference of NO in our results can be excluded (69). This is also confirmed by the fact that an inhibitor of NO production, SMT, does not reduce LPS-induced apoptosis at 12 h of treatment (69). To exclude any possibility of contaminants in the commercial LPS preparation, experiments were repeated using purified LPS (40), and identical results were obtained.
Because a high number of LPS-induced processes are dependent on PKC activation in macrophages (21, 33, 54), we studied the effect of PKC inhibition in macrophages. Our previous studies with specific antibodies had revealed that only PKCβ1, ϵ, and ξ were present in bone marrow-derived macrophages, whereas the rest of the isozymes were not detected (18). Because these cells cannot be transfected efficiently, we used specific chemical inhibitors (12). When the cells were preincubated with the PKC inhibitor calphostin C (10-100 nM) before the addition of LPS, we observed a dose-dependent inhibition of the induction of apoptosis in response to LPS (Fig. 2A). The next step was to establish which of the PKC isoforms present in macrophages was involved in LPS-induced apoptosis. To achieve this, we used Bisindolylmaleimyde I (GF 109203X), which is a specific inhibitor of conventional PKC (β1), and new PKC (ϵ). As shown in Fig. 2B, GF 109203X inhibits LPS-induced apoptosis, which suggests that one GF-sensitive PKC isoform is involved in the regulation of apoptosis by LPS and also excludes the involvement of atypical PKCζ in LPS-induced apoptosis because the GF 109203X doses used in our experiments do not inhibit this isoform (63).
Macrophages were also treated with Gö6976, a selective inhibitor of conventional PKC isoforms, including PKCβ1 (38). Gö6976 did not modify the levels of apoptosis induced by LPS (Fig. 2C). Taken together, these results suggest that PKC regulates LPS-induced apoptosis and that PKCϵ is the main PKC isoform involved in this process.
Finally, to check our results further, we analyzed the apoptosis induced by LPS in bone marrow-derived macrophages obtained from mice in which the PKCϵ gene had been disrupted by homologous recombination (PKCϵ KO). Macrophages from these mice were resistant to LPS induction of apoptosis (Fig. 2D), confirming the results obtained with drug inhibitors.
Because we had found in previous studies that TNF-α secretion induced by LPS is mainly responsible for LPS-induced apoptosis (69), we wanted now to determine whether the effect of PKCϵ on LPS-induced apoptosis is mediated through the induction of TNF-α expression or whether it is a mechanism directly involved with the apoptosis machinery.
Bone marrow macrophages express high levels of TNF-α mRNA after 3 h of LPS stimulation (Fig. 3A). LPS also induces the expression of IL-1β mRNA, partly as a consequence of TNF-α secretion. The inhibition of PKCϵ by pretreatment of cells with GF 109203X before LPS stimulation resulted in almost complete inhibition of the expression of TNF-α and IL-1β mRNA (Fig. 3A). Stimulation with LPS also led to high levels of TNF-α protein secretion in the cell culture supernatant, which rose to concentrations close to 4.5 ng/ml (Fig. 3B). These amounts of TNF-α are clearly sufficient to induce apoptosis in macrophages, because doses between 1 and 10 ng/ml of murine recombinant TNF-α induced significant levels of apoptosis (69). Again, inhibition of PKCϵ activity with GF 109203X inhibited the secretion of TNF-α (Fig. 3B).
Our results suggest that PKCϵ regulates LPS-induced apoptosis and that there is a correlation with TNF-α production. To determine whether LPS induces PKCϵ-dependent TNF-α secretion or TNF-α induces the activation of PKCϵ required for apoptosis, we measured the apoptosis induced by recombinant TNF-α in the presence of GF 109203X. GF 109203X did not affect the apoptosis induced by recombinant TNF-α (Fig. 3C), which suggests that PKCϵ is involved in the apoptosis induced by LPS through the induction of TNF-α production but not in the TNF-α signaling pathway that induces apoptosis.
Of macrophages from wild-type animals, macrophages from PKCϵ KO mice treated with LPS showed lower levels of TNF-α mRNA (Fig. 3D). However, recombinant TNF-α induced apoptosis in macrophages from PKCϵ KO mice at similar levels to control macrophages.
LPS also activated the three major MAP kinase cascades in macrophages, namely ERK, p38, and JNK pathways. All three pathways have been linked to activation by LPS and subsequent cytokine gene expression (24, 57). As well as by LPS, the ERK pathway is activated by growth factors and differentiation signals (51). Activation of the p38 and JNK pathways is often linked to cell stress (5, 25). Because LPS can activate different signal transduction pathways that result in the activation of MAP kinases, we decided to analyze the involvement of these MAP kinases in apoptosis induced by LPS in macrophages.
Although JNK is activated preferentially by cellular stress signals such as irradiation, heat shock, osmotic stress, and protein synthesis inhibitors, stimulation by growth factors has also been reported (28). LPS induced activation of JNK was blocked by the preincubation of macrophages with two JNK inhibitors, SP600125 (25 μM) and curcumin (diferulolymethane, 5 μg/ml), which inhibit JNK activation by different mechanisms: SP600125 acts directly on JNK kinase and curcumin inhibits JNK activation, affecting other components upstream of the JNK kinases (7, 28) (Fig. 4A). These drugs also inhibited the apoptosis induced by LPS (Fig. 4B). Moreover, when we analyzed TNF-α mRNA expression (Fig. 4C) and TNF-α secretion (Fig. 4D), we found that the expression of TNF-α also decreased when cells were incubated with the JNK inhibitors. These results suggest that the JNK pathway can play a role in the expression of TNF-α and the apoptosis induced in response to LPS.
To determine the role of the other two MAP kinases in TNF-α secretion, specific inhibitors of p38 and ERK-1/2 were used. At concentrations of 50 μM of PD98059 and 5 μM of SB203580, activation of ERK and p38, respectively, was blocked in bone marrow macrophages (Fig. 5A). PD98059 slightly decreased the levels of the TNF-α mRNA after LPS treatment, whereas SB203580 did not modify them (Fig. 5B). However, the inhibition of the ERK and p38 MAP kinases produced a minor decrease of TNF-α secretion (Fig. 5C). Because IL-1β expression is induced in response to secreted TNF-α, the SB and PD pretreatments also reduced the levels of LPS-induced IL-1β mRNA (Fig. 5, B and C). These results suggest that ERK and p38 MAPK regulate TNF-α expression mainly through posttranscriptional mechanisms, as has been reported elsewhere (32, 52, 70). Obviously, the lower TNF-α secretion derived from the response to these inhibitors also decreased LPS-induced apoptosis (Fig. 5D).
Taken together, these results suggested an involvement of PKCϵ and MAP kinases in the apoptosis induced by LPS through the regulation of TNF-α. Thus we also wanted to determine whether PKCϵ was involved in the regulation of MAP kinases, as has already been reported in peritoneal macrophages (11). As the differential time course of ERK is crucial in relation to proliferation or differentiation (61), we performed a selected MAPK analysis over time. We observed that JNK activity induced by LPS was clearly inhibited by GF 109203X at all times (Fig. 6A), but not by Gö6976 (Fig. 6B), which suggested that PKCϵ mediates the LPS-induced activation of the JNK pathway. This was confirmed by a large reduction in JNK activity after LPS treatment in macrophages from KO PKCϵ mice (Fig. 6C), whereas ERK and p38 were not inhibited.
We further extended our analysis to determine the involvement of PKCϵ in the activation of ERK and p38 MAP kinase pathways by LPS. GF 109203X slightly decreased the levels of ERK or p38 activity over short time periods but led to a significant extension of activity during the response of macrophages to LPS (Fig. 6, D and E). Again, Gö6976 did not modify the activation of ERK or p38 kinases induced by LPS. Thus PKCϵ inhibition did not block the activation of ERK and p38 kinases but promoted a long activation pattern.
Figure 3 showed that the reduction of apoptosis mediated by the inhibition of PKCϵ by GF 109203X was mediated by inhibition of TNF-α expression and not through a direct effect of PKCϵ on apoptosis signaling machinery. We extended this observation to the three MAP kinases. None of the three inhibitors (i.e., curcumin, PD, or SB) blocked the apoptosis induced by recombinant TNF-α (Fig. 7A). This suggested again that the role of the three MAP kinases in apoptosis induced by LPS is not mediated through modulation of the mechanisms involved in TNF-α-induced apoptosis signaling but through the direct regulation of TNF-α expression induced by LPS.
In this context, GF 109203X did not inhibit JNK activation (Fig. 7B) or TNF-α-induced IL-1β expression (Fig. 7C). TNF-α (10 ng/ml) did not induce the activation of ERK or p38, either. Moreover, the fact that curcumin or SP600125 did not affect apoptosis induced by recombinant TNF-α indicates that JNK activation in macrophages in response to TNF-α did not modulate the signaling cascade leading to apoptosis in response to TNF-α. Our results suggest that in response to LPS, there is a new signaling pathway that regulates TNF-α production in macrophages; LPS activates PKCϵ, which in turn activates the c-jun kinase and induces the production of TNF-α with the subsequent triggering of apoptosis.
In a previous study, we established that bone marrow-derived macrophages die through apoptosis in response to LPS, predominantly through autocrine secretion of TNF-α (69). The present paper reports the signal transduction pathway leading to the secretion of TNF-α and subsequent apoptosis. It describes the role of PKCϵ and JNK in the cascade of events that start after interaction of LPS with the cell membrane and finish with autocrine secretion of TNF-α, which induces apoptosis in macrophages.
Several processes induced by LPS in macrophages are dependent on PKC activation (21, 33, 54), such as the induction of MAPK phosphatase (MKP)-1 (62), NO production (19, 21), or the expression of TNF-α (54). As a cell model for macrophages, we used bone marrow-derived macrophages, which are primary cell cultures that can proliferate, activate, differentiate, or suffer apoptosis when induced with different stimuli. However, transfection efficiency is very low in this cell model (as is the case in other primary cell cultures) (13). Therefore, we had to use specific chemical inhibitors to study the signaling pathways involved in macrophage biology. Our results suggest that LPS induces TNF-α secretion and apoptosis through a pathway that involves the activation of PKC. The activation of this kinase is sufficient to induce apoptosis in several cell types (22, 47, 55). By using two unrelated PKC inhibitors, we showed that PKC is also involved in the induction of apoptosis in macrophages by LPS. Although PKC participates in the control of several LPS-induced events (21, 33, 54), it is still unclear which isoform(s) is involved in each of these effects. In fact, the comparison of different macrophage cell lines, or even different primary monocytic/macrophagic populations, shows significant variations in the expression of PKC isozymes (15, 39, 41), perhaps as a consequence of their specific state of differentiation/maturation. We had previously found that bone marrow macrophages express PKC isoforms β1, ϵ, and ζ (18). Although LPS shows great structural similarity with ceramide (66), a second messenger that activates PKCζ (36), it is unlikely that this isoform mediates the induction of apoptosis by LPS for several reasons. First, the PKC inhibitor GF 109203X that blocks apoptosis induction was used at doses that inhibit conventional and novel PKCs, but not atypical isoforms such as PKCζ (38). Second, we detected no PKCζ translocation in response to LPS (62). Third, although PI-3K is activated by LPS and mediates the synthesis of PIP3, a second messenger that activates PKCζ (44), wortmannin (a specific inhibitor of PI-3K), does not protect macrophages from apoptosis induced by LPS (data not shown). Finally, it has been shown that LPS and ceramide use divergent signaling pathways in macrophages to induce cell death (31). These observations suggest that PKCζ is not involved in the LPS-signaling pathway that leads to apoptosis induction.
Several observations support the involvement of PKCϵ rather than PKCβ1 in the induction of TNF-α and apoptosis by LPS. First, GF 109203X inhibits conventional PKC isoforms better than novel ones (38). Concentrations of GF 109203X up to 1 μM completely inhibit the activation of conventional PKCs, including β1, whereas concentrations up to 5 μM are needed to completely block novel isoforms, including ϵ. In our experiments, maximal inhibition of apoptosis or TNF-α production were reached at concentrations of 3-5 μM of GF 109203X. Second, Gö6976, a selective inhibitor of conventional PKCs, does not block induction of TNF-α or apoptosis by LPS. Finally, LPS does not induce apoptosis or TNF-α gene expression in bone marrow-derived macrophages from mice with an inactivated PKCϵ gene. All these results demonstrate that PKCϵ is specifically involved in the LPS-induction of TNF-α and apoptosis in bone marrow macrophages. Our results corroborate a recent report showing a critical role of PKCϵ for LPS-induced IL-12 synthesis in monocyte-derived dendritic cells (2).
Finally, we have demonstrated using inhibitors that PKCϵ plays a major role in the LPS-induced activation of JNK and in the TNF-α secretion, but in the macrophages of PKCϵ-knockout mice there is a residual JNK activity and TNF-α expression, which is not sufficient to induce apoptosis in these cells. This residual JNK activity and TNF-α expression can be blocked completely by a pretreatment with Gö6976 (data not shown), suggesting that PKCβ1 could play a small compensatory activity in the LPS signaling pathways in this PKCϵ-null mice.
The signaling events that occur after the interaction of LPS with CD14 are not fully understood, but several studies have suggested the involvement of MAP kinase activation (23, 26, 27, 50, 65). Upon receptor engagement, TNF-associated factor (TRAF)2 is recruited to CD40 and transferred to lipid rafts in a RING (really important new gene) finger-dependent process, which enables the activation of downstream signaling cascades, including JNK and nuclear factor-κB (NF-κB) (4). A small inhibition of LPS-induced apoptosis was found when macrophages were incubated with the MAP kinase inhibitors for ERK and p38. However, JNK is the kinase pathway mainly involved in the TNF-α production induced by LPS that involves activation of PKCϵ. JNK is involved in the transcriptional regulation of TNF-α through the regulation of the formation of the AP-1 transcription factor complex and the binding to the TNF-α promoter (56).
The ultimate fate of a cell exposed to TNF-α is determined by signal integration between its different effectors, including IκB kinase (IKK), c-Jun NH2-terminal protein kinase (JNK), and caspases (6). Activation of caspases is required for apoptosis (59), whereas activation of IKK inhibits apoptosis through the transcription factor NF-κB, whose target genes include inhibitors of caspases (64). JNK activates the transcription factor c-Jun/AP-1, as well as other targets (14). However, the function of JNK activation in apoptosis induced by TNF-α is less clear (34, 45).
In our experiments, the fact that PKC and MAPK inhibitors blocked apoptosis induced by LPS and TNF-α expression, but did not inhibit apoptosis induced by recombinant TNF-α treatment, indicates that PKCϵ and the three MAP kinases can modulate LPS-induced apoptosis through the production of TNF-α but not through a direct effect on the apoptotic mechanism. In endothelial cells, JNK was involved in TNF-α-induced apoptosis (30). However, in other cell models, no involvement of MAP kinases or PKCϵ has been described in the TNF-α-induced pathway leading to apoptosis (9, 16, 29). Recently, research conducted on fibroblasts has found that the NF-κB pathway negatively modulates JNK activation mediated by TNF-α and contributes to the inhibition of apoptosis (17, 58). However, our results with curcumin do not confirm these observations. The differences could be due to different responses of macrophages and fibroblasts to TNF-α (35, 69) because TNF-α induces apoptosis in macrophages without inhibiting protein synthesis or the NF-κB pathway.
Our results suggest that only JNK is downstream of the PKCϵ activity. This supports the existence of a new pathway in macrophages where LPS induces PKCϵ, which activates JNK, which in turn modulates TNF-α expression at the transcriptional level and thus mediates apoptosis induced by LPS. Kinase activity experiments support our hypothesis because JNK activity induced by LPS is inhibited when macrophages are pretreated with the PKC inhibitor GF 109203X. Our results were confirmed by the use of mice with a disrupted PKCϵ gene (40). These mice showed lower resistance to infection by Gram-positive or -negative bacteria, and the peritoneal macrophages showed a dramatic reduction of LPS-induced TNF-α production with a partial reduction of LPS-induced ERK or p38 MAP kinase activities (40). Our results confirm and extend these observations. In our experiments, the bone marrow-derived macrophages from these mice were resistant to LPS-induced apoptosis, and both TNF-α secretion and JNK activity were lower. This suggests preferential induction of JNK activity over the other two MAP kinases.
This work was supported by Ministerio de Ciencia y Tecnología Grant BMC2001-3040 (to A. Celada).
We thank Drs. Lisardo Boscá from the Instituto de Bioquimíca, Consejo Superior de Investigaciones Científicas, Madrid, and Peter Parker of the Imperial Cancer Research Fund, London, United Kingdom, for help with PKCϵ KO mice.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2003 the American Physiological Society