A Mycoplasma fermentans -derived Synthetic Lipopeptide Induces AP-1 and NF- k B Activity and Cytokine Secretion in Macrophages via the Activation of Mitogen-activated Protein Kinase Pathways*

Mycoplasma lipoproteins have been demonstrated to stimulate monocytic cells and induce proinflammatory cytokine secretion. In this paper, we show that a synthetic analog of the Mycoplasma fermentans membrane-associated lipopeptide macrophage-activating lipopep-tide-2 (MALP-2) induces mRNA synthesis and protein secretion of interleukin-1 b and tumor necrosis factor- a in human monocytes/macrophages and the murine macrophage cell line RAW 264.7, whereas the nonlipidated counterpart lacks this effect, underscoring the importance of protein acylation for cell activation. Synthetic MALP-2 (sMALP-2) induced the activation of MAPK family members extracellular signal regulated kinases 1 and 2, c-Jun NH 2 -terminal kinase, and p38 and induced NF- k B and AP-1 transactivation in macrophages. Whereas the specific p38 inhibitor SB203580 ab-rogated both cytokine synthesis and NF- k B and AP-1 transactivation in response to MALP-2, the selective MAPK/extracellular signal-regulated kinase-1 inhibitor PD-98059 decreased interleukin-1 b and tumor necrosis

Mycoplasma fermentans, a human pathogen, is a potent activator of monocytes/macrophages. Macrophage activation by M. fermentans results in the secretion of numerous cytokines including interleukin (IL) 1 -1␤, IL-6, IL-8, IL-10, and tumor necrosis factor-␣ (TNF␣) (1)(2)(3). Other inflammation mediators such as NO have also been shown to be produced by macrophages in response to mycoplasmas (3). Mycoplasmas, the smallest self-replicating bacteria, are characterized by a wallless envelope (4); thus, they are LPS-free pathogens. We have previously demonstrated that mycoplasma membrane lipoproteins (lipid-associated membrane proteins; LAMPs) are responsible for human and murine macrophage activation (1,2,5). Although the signaling pathways triggered in macrophages by both M. fermentans LAMP and Gram-negative bacteria LPS are comparable, unlike LPS, M. fermentans LAMP activity does not require binding to CD14 or serum proteins (5). M. fermentans LAMP activates mitogen-activated protein kinases (MAPKs), including extracellular signal-regulated kinases 1 and 2 (ERK1/2), c-Jun NH 2 -terminal kinase (JNK), and p38 (5). MAPKs have been shown to be involved in M. fermentans LAMP-mediated cytokine induction, since selective inhibitors of these pathways have been shown to impair M. fermentans LAMP-mediated cytokine production by human and murine macrophages.
Initial partial biochemical characterization of an M. fermentans-derived high molecular weight material that activates macrophages suggested that it was an amphiphilic lipid-like molecule (3). Further studies led to the identification of an acylated S- (2,3-dihydroxypropyl)cysteine, typical for Braun's murein lipoprotein, in a lipopeptide purified from M. fermentans-derived high molecular weight material (6). This suggested that this lipopeptide could be the macrophage-activating agent of M. fermentans-derived high molecular weight material. Very recently, the same group has elucidated the structure of this lipopeptide, named MALP-2 (for macrophage-activating lipopeptide) (7). MALP-2 is a 2-kDa free NH 2 terminus lipopeptide with the following structure: S- (2,3-bisacyloxyprolyl)cysteine-GNNDESNISFKEK. Interestingly, M. fermentans-derived MALP-2 as well as its synthetic analog induced NO release from murine macrophages and showed an identical dose-dependent effect (7).
In the present study, we have tested the capacity of synthetic MALP-2 (sMALP-2) to stimulate cytokine production by macrophages and activate MAPK pathways. Additionally, we have investigated the activation of NF-B and AP-1, two transcription factors playing a role in the induction of proinflammatory cytokines, in macrophages challenged with sMALP-2. sMALP-2 stimulatory effects on macrophages were fully comparable with those previously found with M. fermentans lipoproteins, suggesting that lipopeptides derived from mycoplasmas might constitute excellent tools for better understanding the immunomodulation me-diated by mycoplasmas and the role of this property in the pathogeny of these microorganisms.
Synthetic Lipopeptide-Peptide (CGNNDESNISFKEK) and acylated peptide corresponding to the previously published MALP-2 sequence (7) were synthesized as described by Metzger et al. (8,9). Synthesis was performed by Synt:em (Nîme, France). Tripalmitoyl S-glycerylcysteine was kindly provided by Dr. Radolf. Acylated peptide was resuspended in 25 mM octyl glucoside in phosphate-buffered saline. The detergent had no effects on cells at the final used concentration. Lipopeptide preparation was tested for absence of endotoxin (Ͻ60 pg/ml) by Limulus amoebocyte lysate assay (Hemachem, St. Louis, MO).
Cell Cultures, Stimulation, and Lysate Preparation-The murine macrophage cell line RAW 264.7 (American Type Culture Collection, Rockville, MD) was cultured (37°C, 5% CO 2 ) in Dulbecco's modified Eagle's medium culture medium containing 10% fetal calf serum, 2 mM L-glutamine, and antibiotics. The human monocytic cell line THP-1 (American Type Culture Collection) was cultured (37°C, 5% CO 2 ) in RPMI culture medium containing 10% fetal calf serum, 2 mM L-glutamine, and antibiotics. Cell lines were tested every 2 weeks by a polymerase chain reaction-based detection assay for mycoplasma contamination (10). For stimulation experiments, cells were seeded at 10 6 cells/ml density and then allowed to cultivate overnight. Cells were stimulated with sMALP-2 (at 200 nM) for appropriate time intervals. LPS was used in control experiments at 1 g/ml to stimulate RAW 264.7 cells and at 100 ng/ml to stimulate THP-1 cells. For phosphotransferase assays, cells were washed twice with ice-cold phosphate-buffered saline containing 1 mM Na 3 VO 4 . For each 10 6 cells initially seeded, 100 l of the following lysis buffer was added: 20 mM MOPS, pH 7.2, 5 mM EDTA, 1% (w/v) Nonidet P-40, 1 mM dithiothreitol, 75 mM ␤-glycerol phosphate, 1 mM Na 3 VO 4 , and protease inhibitor mixture (Boehringer GmbH, Mannheim, Germany). Lysis was performed at 4°C for 20 min with continuous shaking. Cell lysates were centrifuged (10.000 ϫ g for 10 min at 4°C), and supernatant was aliquoted and stored at Ϫ80°C. Protein concentration in cell lysates was determined by a micro-BCA assay (Pierce).
Assay for Cytokines-Cells were stimulated as indicated above for a 24-h period and were lysed by two consecutive cycles of freezing/thawing. Thus, the samples represented the total amount of cytokines produced (both intracellular and released into the supernatant). The cytokine concentration from murine or human macrophages was measured by a corresponding IL-1␤ and TNF␣ ELISA kit. The assays were performed according to the manufacturer's instructions.
RNA Isolation and Analysis-Cells were stimulated as described above at the indicated times, and total RNA was extracted from 10 7 cells using the total RNA isolation kit from Bioprobe (Montreuil, France) according to the manufacturer's instructions. Using Hybond-N ϩ filters (Amersham Pharmacia Biotech), slot RNA were performed with 5-10 g of RNA as described by Sambrook et al. (12), and slots were UV-cross-linked. Specific murine IL-1␤ and TNF␣ oligonucleotide probes were from CLONTECH Laboratories (Heidelberg, Germany). The amount of RNA in each sample was normalized by probing for ␤-actin. All probes were labeled by a 5Ј-DNA labeling system with [␥-32 P]dATP (Amersham Pharmacia Biotech). Slots were prehybridized in 6ϫ SSC, 0.5% SDS, 5ϫ Denhardt's solution, and 100 g/ml of denatured salmon sperm DNA for 2 h at 42°C. An overnight hybridization was performed at 42°C in the same buffer containing the probe at 2.5 ϫ 10 5 cpm/ml. The membrane was then washed twice for 10 min in 2ϫ SSC, 0.05% SDS at room temperature and once in 1ϫ SSC, 0.1% SDS at 42°C. Slots were exposed to a PhosphorImager screen and quantitatively assessed by means of ImageQuant software (Molecular Dynamics, Sunnyvale, CA).
Immunoprecipitation of ERK, JNK, and p38 from Stimulated Cell Lysates-ERK1/2 and JNK were immunoprecipitated by incubating 500 g of cell lysates with 5 g of anti-ERK1-CT antibody or 2 g of anti-JNK1 antibody, respectively, at 4°C for 4 h with continuous rotation. Then 30 l of protein A-Sepharose was added, and the incubation was extended for an additional 2 h. The mixtures were then centrifuged (7000 ϫ g for 2 min at room temperature), and protein A-Sepharose beads were washed three times with buffer B (12.5 mM MOPS, pH 7.2, 0.5 mM EGTA, 12.5 mM ␤-glycerol phosphate, 7.5 mM MgCl 2 , 1 mM dithiothreitol, 1% Nonidet P-40) containing 250 mM NaCl. The beads were resuspended in 50 l of buffer B containing 10 mM MgCl 2 and 1 mM MnCl 2 for phosphotransferase.
For p38 immunoprecipitation, the protocol was slightly modified. 10 g of anti-p38 antibody was first coupled to A/G-Sepharose beads for 2 h at 4°C and then washed with buffer B and added to 500 g of cell lysates. Immunoprecipitation was allowed overnight at 4°C with continuous rotation, and immunoprecipitates were analyzed as described above.
Measurement of Phosphotransferase Activity-ERK and p38 activation was determined in immunoprecipitates by means of measuring radioactively their respective phosphotransferase activities toward a peptide substrate using p42/44 MAPK or p38/Mpk2 detection kits. Assays were performed according to the manufacturer's instructions. Activity was expressed as [ 32 P]ATP cpm.
To measure JNK activation, 2 g of GST-c-Jun was added to stressactivated protein kinase/JNK immunoprecipitates in the presence 50 M [␥-32 P]ATP. The reactions were conducted at 30°C for 30 min and then terminated by adding SDS sampler buffer to 1ϫ final concentration. Samples were analyzed by SDS-PAGE using 12% gels. Gels were fixed in 10% acetic acid and 50% methanol and then embedded in cellophane sheets, dried, and autoradiographed.
Nuclear Extract Preparation and Electrophoretic Mobility Shift Assay-Cells were stimulated with sMALP-2 at indicated time intervals, nuclear extracts were prepared as described in Ref. 13, and protein content was determined by micro-BCA assay (Pierce). NF-B and AP-1 binding activity in nuclei of uninduced and induced cells was determined by an electrophoretic mobility gel shift assay (EMSA) as described in Ref. 14 using 2-4 g of nuclear proteins. EMSA gels were exposed to a PhosphorImager screen and quantitatively assessed by means of ImageQuant software (Molecular Dynamics). Supershift experiments were performed by preincubating nuclear extracts with anti-Rel family antibodies for 1 h at 4°C before carrying out EMSA.
Plasmids, Cell Transfection, Activation, and Assay for Luciferase Activity-The NF-B, AP-1, and NF-AT (15,16) luciferase reporter constructs were kindly provide by Dr. A. Acuto (Institut Pasteur, Paris, France). RAW 264.7 cells were grown up to 80% confluence and then transfected with the indicated plasmids by the electroporation method as described by Stacey et al. (17). CHO-CD14 and CHO-RSV cells were grown at 80% confluence and transfected with 15 g of the indicated reporter plasmid by electroporation as described in Ref. 18. Transfected cells were cultured in growth medium for an overnight period and then stimulated or not stimulated with sMALP-2 (200 nM), LPS (1 g/ml) for 6 h. In control experiments, transfected cells were stimulated with PMA (50 ng/ml) or PMA (50 ng/ml) plus A23178 (2 g/ml). Cells were then harvested, and luciferase activity was determined as described by Brasier (19) using an automated luminometer (Lumat LB 9501, EG1G Berthod, Wilbad, Germany). Cell lysates were analyzed for protein content using the micro-BCA assay (Pierce), and luminescence units were normalized for total protein content. Luciferase activities are reported as means of values from four independent experiments, each performed in duplicate.

sMALP-2 Induces Transcription and Secretion of IL-1␤ and TNF␣ by Macrophages-M. fermentans
sMALP-2 has been demonstrated to induce NO release from mouse macrophages (7). First, we have investigated the potential ability of sMALP-2 to stimulate cytokine production by monocytic cells. The murine macrophage cell line RAW 264.7 was challenged with increased concentration of sMALP-2, and TNF␣ production level was determined by ELISA. TNF␣ could be detected in RAW 264.7 supernatant when cells were stimulated with sMALP-2 at concentrations higher then 50 nM; the lowest doses did not stimulate TNF␣ production (Fig. 1A). In a 100 -600 nM concentration range of sMALP-2, the TNF␣ level secreted by RAW 264.7 cells was found to be dose-dependent. A similar dose-response curve was obtained when measuring IL-1␤ production by murine macrophages under identical conditions (data not shown). In following experiments, sMALP-2 was used at 200 nM concentration; however, in some inhibition experiments the highest concentrations were applied.
sMALP-2 preincubation with polymyxin B (1000 units/ml) for 2 h before its addition to cell cultures had no effect on TNF␣ production, whereas polymyxin B completely abolished TNF␣ induction by LPS (data not shown). These data and the results of the Limulus amoebocyte lysate assay (see "Experimental procedures") clearly indicate that the stimulatory effect of sMALP-2 preparations are not to be ascribed to endotoxin contamination.
As shown in Fig. 1, B and C, sMALP-2 at 200 nM induced the production of a considerable amount of IL-1␤ and TNF␣ by murine macrophage RAW 264.7 (Fig. 1B) and by human monocytes/macrophages as well (Fig. 1C). Similar results were obtained using the human myelomonocytic cell line THP-1 (data not shown). When human monocytes/macrophages or murine macrophages RAW 264.7 were challenged with LPS, TNF␣, and IL-1␤ secretion levels were comparable with those obtained in response to sMALP-2 ( Fig. 1, B and C). The nonlipidated form of MALP-2 failed to induce cytokine production by both human monocytes/macrophages and RAW 264.7 cells. In addition, tripalmitoyl S-glycerylcysteine was unable to stimulate cytokine production by these cells (Fig. 1, B and C). These findings demonstrate that lipid modification of MALP-2 peptide is required for macrophage stimulation and cytokine production.
We have further examined the levels of cytokine mRNA in RAW 264.7 cells stimulated with sMALP-2. Data from RNA slot hybridization, presented in Fig. 1D, clearly show that sMALP-2 stimulated IL-1␤ and TNF␣ mRNA synthesis, with a peak accumulation found at about 4 h after challenge.
Previous reports have strongly suggested that mycoplasmaderived membrane fractions stimulate monocytes/macrophages by a mechanism distinct from that of LPS and independent of CD14 and/or serum (1,5,20). We have therefore investigated whether sMALP-2 was able to induce NF-B and AP-1 transactivation in CHO cells expressing human CD14 (11). CHO cells expressing CD14 (CHO-CD14) were transiently transfected with NF-B or AP-1 luciferase reporter plasmid, and then cells were stimulated with LPS and sMALP-2. CHO cells harboring the empty expression vector (CHO-RSV) were transiently transfected with reporter plasmids and stimulated as described above. As expected, LPS induced an important luciferase activity (4 -5-fold increases) in CHO-CD14 cells transfected with either NF-B or AP-1 reporter plasmids, while no activation could be observed with CHO-RSV under similar conditions (Fig. 2). On the contrary, sMALP-2 induced neither NF-B nor AP-1 transactivation in transfected CHO-CD14 cells (Fig. 2), even when it was used at higher concentrations (data not shown). Control experiments were performed by stim-  (Fig. 2). These data suggest that, unlike LPS, sMALP-2 does not require CD14 for cell stimulation.

sMALP-2 Activates MAPK Pathways and NF-B and AP-1 Transcription Factors in Macrophages-MAPKs
are key molecules involved in macrophage activation by different bacterial products. We have therefore investigated the activation of the well characterized MAPKs, ERK1/2, p38, and JNK, in RAW 264.7 cells challenged with sMALP-2. Data presented in Fig. 3 clearly indicate that sMALP-2 significantly induced activation of all of the tested kinases. LPS has also been shown to induce activation of MAPKs in murine macrophages (21). MAPK activation kinetics in RAW 264.7 stimulated by LPS and sMALP-2 were compared. As shown in Fig. 3, LPS peak activation of ERK1/2, p38, and JNK was found at 15 min after stimulation, while peak activation of these kinases occurred at 30 min after challenge with sMALP-2. The kinetics of MAPK activation by sMALP-2 were comparable with those obtained when cells were stimulated with a crude extract of M. fermentans membrane lipoproteins (5). A similar pattern of MAPK activation was observed in human monocytes/macrophages and the THP-1 cell line stimulated with sMALP-2 (data not shown).
Transcription factors NF-B and AP-1 are involved in the transcription of a variety of genes, including proinflammatory cytokines, during the immune response (22)(23)(24). Thus, we have addressed the ability of sMALP-2 to activate nuclear translocation of the transcription factors NF-B and AP-1. Using 32 P-labeled oligonucleotides containing consensus NF-B or AP-1, we have performed EMSAs on nuclear extracts from RAW 264.7 cells stimulated with sMALP-2. As demonstrated in the Fig. 4, sMALP-2 was capable of inducing both NF-B and AP-1 translocation. The specificity of AP-1 and NF-B DNA biding were verified by competition analysis with an excess of nonradiolabeled probes. NF-B and AP-1 activation by sMALP-2 was found to be dose-dependent, and transcription factors were detectable starting from 100 nM of sMALP-2 concentration, confirming data presented in Fig. 1 (data not shown).
As depicted in Fig. 4A, similarly to the MAPK activation, LPS-mediated NF-B translocation was more rapid (with a peak found at 2 h of stimulation) than that observed with sMALP-2 (peak at 4 h). The kinetics of NF-B activation mediated by LPS reported herein were comparable with those previously described by Tebo et al. (25). We have further investigated by means of EMSA supershift the NF-B isoforms induced by sMALP-2. The NF-B p65 and p50 antibodies supershifted the protein complex, whereas c-Rel and RelB antibodies did not modify the pattern (data not shown), indicating that a p65/p50 heterodimer accounts for the NF-B-translocated form.
sMALP-2 also activated AP-1 translocation in RAW 264.7 cells, and peak activity was observed at 4 h of stimulation. Although a basal AP-1 activation could be detected in these cells, PhosphorImager quantification of EMSA gels clearly shows that sMALP-2 induced a 4 -5-fold increase in AP-1 DNA binding activity compared with control level (Fig. 4B). Once again, AP-1 translocation in response to LPS occurred earlier (peak at 2 h) than that induced by sMALP-2 (data not shown).
Pursuing the investigation of AP-1 and NF-B induction by sMALP-2, we have studied the transactivation ability of these . Immunoprecipitation (IP) and phosphotransferase activity (PA) measurement were performed as described in "Experimental Procedures." A, ERK1/2 (left panel) were immunoprecipitated from cell lysates using anti-ERK1-CT antibody, and their activation was quantified using the p42/44 MAPK detection kit (Amersham Pharmacia Biotech) according to the manufacturer's instructions. p38 (right panel) was immunoprecipitated from cell lysates using anti-p38 antibody, and its activation was quantified using the p38/RK/Mpk2 detection kit (Upstate Biotechnology, Inc.) according to the manufacturer's instructions. Untreated cells were used as control (CTRL). Results are means Ϯ S.D. of three independent assays, each performed in duplicate. B, JNK was immunoprecipitated from cell lysates using JNK1 antibody (anti-JNK1 antibody). Immunoprecipitates were incubated with GSTc-Jun under phosphorylating conditions, and phosphorylated GST-c-Jun (arrow) was visualized by gel electrophoresis and autoradiography. Untreated cells were used as control (0). The presented gels are representative of three distinct experiments. factors by means of luciferase reporter plasmids. RAW 264.7 cells were transiently transfected with plasmids containing the AP-1-or NF-B-dependent luciferase reporter gene and then stimulated with sMALP-2, and luciferase activity was determined in cell lysates. As demonstrated in Fig. 5, sMALP-2 induced a 3-fold increase in luciferase activity in cells transfected by either AP-1 or NF-B reporter plasmid, indicating that the translocated factors can mediate transactivation. sMALP-2 failed to induce transactivation of the T cell-specific transcription factor NF-AT-dependent luciferase reporter gene. As a positive control, stimulation of NF-AT-transfected RAW 264.7 cells with PMA/A23187 (26) induced a 4-fold increase in specific luciferase activity (data not shown). Comparable levels of NF-B-and AP-1-mediated luciferase induction were observed in LPS stimulation experiments (Fig. 5). sMALP-2 also induced AP-1-and NF-B-dependent luciferase expression in THP-1 cells (data not shown).

Involvement of MAPK Pathways in MALP-2-induced Cytokine Production and AP-1 and NF-B Activation-
To examine the involvement of MAPK activation in sMALP-2-mediated cytokine production, we have used the specific inhibitors of MAPK pathways, SB203580 and PD-98059 (27,28). At the highest used concentration, these two inhibitors, PD-98059 and SB203580, did not induce any cell toxicity as determined by microscopic observation and trypan blue uptake (data not shown).
SB203580 is a bicyclic imidazole compound able to specifically inhibit p38 (28). This inhibitor selectively blocked the sMALP-2-mediated activation of p38 in RAW 264.7 cells without significantly affecting the stimulation of ERK1/2 or JNK (data not shown). Preincubation of RAW 264.7 with SB203580 for 1 h prior to challenging with sMALP-2 blocked in a dosedependent manner both IL-1␤ and TNF␣ production (Table I). Comparable inhibition effects of SB203580 were obtained when cells were stimulated with higher concentrations of sMALP-2 (data not shown).
PD-98059 is a synthetic compound that specifically inhibits the ERK-activating MAPK kinase MEK-1 (27,29). This compound selectively inhibited the sMALP-2-mediated activation of ERK1/2 in murine macrophages without significantly affecting p38 or JNK stimulation (data not shown). The inhibition of MEK-1 by PD-98059 treatment partially inhibited, in a dosedependent manner, cytokine production by RAW 265.7 in response to sMALP-2. Unlike SB203580, PD-98059 was unable to completely block sMALP-2 stimulation, and only 50% inhibition was obtained at 30 M concentration (Table I) 1, 2, 4, and 6 h), and untreated cells were used as control (0). Nuclear extracts and EMSA were performed as indicated under "Experimental Procedures." A, NF-B EMSA. Nuclear extracts from cells stimulated with sMALP-2 (left part) or with LPS (right part) were probed with 32 P-labeled NF-B consensus oligonucleotide. The specificity of DNA binding was assessed by preincubating extracts with unlabeled specific (NF-B) or nonspecific (AP-1) competitor oligonucleotide at a 50-fold molar excess. The arrow indicates a specific NF-B band. B, AP-1 EMSA. Nuclear extracts from cells stimulated with sMALP-2 were probed with 32 P-labeled AP-1 consensus oligonucleotide. The specificity of DNA binding was assessed by preincubating extracts with unlabeled specific (AP-1) or nonspecific (AP-2) competitor oligonucleotide at 50-fold molar excess. The right part presents AP-1 DNA binding activity from EMSA gel presented herein quantified by PhosphorImager. EMSA experiments were repeated four times, and gels presented in the figure are from a representative set.
Interestingly, treatment of RAW 264.7 cells with either PD-98059 or SB203580 yielded a similar inhibition pattern when LPS was used as stimulating agent (Table I). In addition, ERK1/2 and p38 pathway inhibitors, respectively, partially or completely inhibited cytokine production in the human cell line THP-1 stimulated with sMALP-2 or LPS (data not shown).
We have also assessed the effect of MAPK pathway inhibitors on sMALP-2-mediated AP-1 and NF-B transactivation. RAW 264.7 cells were transiently transfected with either AP-1 or NF-B reporter plasmids, and then cells were treated with PD-98059 or SB203580 prior to stimulation with sMALP-2. Whereas the MEK-1 inhibitor, PD-98059, at a concentration that inhibited 50% of cytokine production (see Table I) or higher, affected neither AP-1-nor NF-B-dependent luciferase transactivation (Fig. 6), SB203580 significantly reduced both AP-1-and NF-B-mediated transactivation (Fig. 6). The effect of SB303580 was dose-dependent and at 30 M completely blocked AP-1 and NF-B activities. Similar inhibitions were obtained when cells were stimulated with higher sMALP-2 concentrations (600 M; data not shown). We have additionally assessed the effect of these inhibitors on LPS-induced NF-B or AP-1 transactivation. In these experiments, inhibition profiles comparable with that reported with sMALP-2 were observed (data not shown). These data underscore the involvement of p38 pathway in the nuclear response to bacterial modulin in macrophages, while the ERK1/2 pathway appears to control distinct events.

DISCUSSION
M. fermentans is associated with arthritis in humans (30,31) and it has been proposed as a putative co-factor in AIDS progression (30,32). The ability of mycoplasmas to induce proinflammatory cytokines and to favor Th1 3 Th2 switch might participate in the pathogenesis mechanisms of these bacteria (1,2). Previous findings have clearly indicated that membrane lipoprotein fractions from M. fermentans are responsible of cytokine induction in macrophages. The precise characterization of the biochemical entities involved in such activation is an important step toward the understanding of the interaction of mycoplasmas and immune cells and necessary to delineate their contribution to the mycoplasma pathogenesis and clinical manifestations. MALP-2, a lipopeptide characterized from M. fermentans, has been shown to strongly induce NO synthesis by macrophages (7). Very recently, it has been reported that MALP-2 is a mycoplasma membrane-associated entity resulting from the processing of a precursor lipoprotein (33,34). In the present study, we show that a synthetic analog of MALP-2 is capable of inducing the activation of NF-B and AP-1 and the cytokine production by murine macrophages. We have also demonstrated the involvement of MAPK pathways in the signaling events triggered by MALP-2.
Our results confirmed and extended earlier findings concerning the macrophage activation by bacterial lipoproteins and lipopeptides. Lipoproteins from the outer membrane of Escherichia coli and its synthetically prepared NH 2 -terminal lipopeptide segments have been demonstrated to stimulate both human and murine macrophages and to induce the secretion of IL-1, TNF␣, and IL-6 (35,36). Spirochetal lipoproteins have also been shown to stimulate macrophages (37), and the lipid modification determines the ability to stimulate the production of cytokines by these cells (38). The importance of protein acylation to cell activation was further underscored by the fact that synthetic lipohexapeptides corresponding to the NH 2 termini of the 47-kDa lipoprotein of Treponema pallidum and the acylated outer surface protein A (OspA) of Borrelia burdoferi also activated macrophages in terms of cytokine secretion, whereas the nonlipidated hexapeptides were without effect (39). In the present report, we also provided evidence showing that macrophage activation by MALP-2 is dependent upon acylation. Whereas both native MALP-2 and its synthetic an-  alog have been previously shown to induce NO release from murine macrophages at a 1-100 pM range of concentrations (7), the effect of sMALP-2 on cytokine production by macrophages occurs at 100 -600 nM concentration. For NO release assay, the simultaneous stimulation with IFN-␥ and MALP-2 (7) could at least in part account for the significant high sensitivity of macrophages and the requirement of lower concentration of lipopeptide to induce NO activity with respect to cytokine production. By using both electrophoretic mobility shift and transactivation assays, we have clearly shown that sMALP-2 stimulates the transcription factors NF-B and AP-1. It has been previously reported that T. pallidum and B. burgdorferi lipoproteins and respective synthetic lipopeptides induce NF-B translocation in different cell types (40,41) with kinetics more rapid (maximal stimulation from 1 to 2 h) than that obtained with sMALP-2 (peak at 4 h). Interestingly, the kinetics of NF-B and AP-1 induction by LPS were similar to that reported for the spirochetal products. It would be interesting to test whether the observed differences can be ascribed to the different lipopeptide structures.
Findings presented herein indicate that signaling pathways triggered in response to sMALP-2 and involved in cytokine production include ERK1/2 and p38 MAPKs. These same pathways with identical activation kinetics have been also implicated in macrophage response to M. fermentans membrane fractions (5). NF-B and AP-1 triggering by sMALP-2 also involved MAPK pathways. In agreement with previously published reports indicating that p38 MAPK is required for the activation of several transcription factors, including cAMP response element-binding protein (42), c-Fos, and c-Jun (30 -32, 43-45), we have shown that p38 pathway activation is required for AP-1 induction in response to sMALP-2. Interestingly, the p38 pathway was also found to be required for NF-B response in sMALP-2-stimulated macrophages. Very recently, Bergmann et al. have shown that p38 inhibitor, SB203580, blocked NF-B-mediated luciferase transactivation in response to TNF␣ without affecting NF-B translocation (46). Accordingly, in our cells, SB203580 was unable to inhibit NF-B translocation in response to sMALP-2 (data not shown). At present, how p38 MAPK may affect, directly or indirectly, NF-B function is unclear, and the question remains open; however, our data underscore the important contribution of this pathway in nuclear response to bacterial lipopeptide stimuli. ERK1/2 has been clearly shown to contribute to the cell signaling resulting in cytokine production, but unlike the p38 MAPK pathway, it is not involved in sMALP-2-mediated NF-B or AP-1 transactivation. Several other studies have shown that although the p42/44 MAP kinase pathway is activated in response to distinct stimuli including anisomycin, fibroblastic growth factor, epidermal growth factor, 12-O-tetradecanoylphorbol-13-acetate, and UV irradiation, it is not involved in nuclear signaling (42,44,47). The signaling events controlled by ERK1/2 pathway remains to be clearly defined.
Our data demonstrate that the intracellular events leading to cytokine synthesis induced by both sMALP-2 or LPS are very similar, suggesting that common upstream pathways are triggered by these stimuli. In contrast to LPS, little is known about the mechanism by which bacterial lipoproteins and lipopeptides activate macrophages. The requirement for peptide or protein acylation strongly suggest that lipopeptides interact via the lipid moiety with currently unknown membrane molecule(s). Two recent papers have reported evidence suggesting that T. pallidum and B. burdoferi lipoproteins and synthetic lipopeptides activate monocytes by a CD14-dependent pathway (48,49). Actually, mouse anti-human CD14 antibodies were shown to block the activation of monocytic cells by these spirochetal products (48) and to significantly decrease the sensitivity of endothelial cells to Borrelia lipoproteins (49). Very interestingly, Wooten et al. (49) have demonstrated that OspA and CD14 can form stable complexes. Yet, spirochetal lipoproteins and lipopeptides were shown to activate monocytes by a CD14dependent pathway that fundamentally differs from that of LPS (48). We and others have previously demonstrated that M. fermentans lipoproteins induce monocyte activation by mechanisms distinct from that of LPS (1,19). In this way, and in contrast to the results reported with spirochetal products, anti-CD14 antibodies have been shown to be inefficient in blocking the effects of mycoplasma lipoproteins on monocytic cells (5,20). In addition, human monocytic cell line THP-1 does not require vitamin D 3 treatment, which increases CD14 expression on cell membrane, to efficiently produce cytokine in response to mycoplasma lipoproteins. 2 Further experiments are necessary to evaluate whether mycoplasma and spirochetal derived lipoproteins and/or lipopeptides actually display some differences in triggering the monocyte activation.
As indicated above, it has been recently reported that MALP-2 derives from a precursor larger lipoprotein by posttranslational processing (33,34). Relative amount of MALP-2 and its precursor lipoprotein varied from one M. fermentans isolate to another. MALP-2 was initially characterized by selecting for M. fermentans isolates that are strong activators of macrophages (7). Interestingly, the selected strain II-29/1 abundantly expresses MALP-2 on the cell surface, whereas the precursor lipoprotein was undetectable (33,34). To our knowledge, MALP-2 is the first membrane-associated lipopeptide so far described, and the nature of MALP-2 precursor lipoprotein processing is presently not understood. It remains to be determined whether the precursor lipoprotein displays an activity comparable with that of the lipopeptide. Preliminary data from our laboratory suggest that the recombinant precursor lipoprotein is capable of activating macrophages, but unlike MALP-2, acylation is not required. The confirmation of these results awaits the production of antibodies against different epitopes of the precursor lipoprotein and MALP-2. Given that MALP-2 is a surface-associated molecule able to activate macrophages, a synthetic analog of this lipopeptide constitutes a valuable tool to address the issue of the contribution of immunomodulation by mycoplasmas to the pathogenesis.