Critical Role of the Phosphatidylinositol 3-Kinase/Akt/Glycogen Synthase Kinase-3β Signaling Pathway in Recovery from Anthrax Lethal Toxin-induced Cell Cycle Arrest and MEK Cleavage in Macrophages*

Anthrax lethal toxin (LeTx) is a virulence factor causing immune suppression and toxic shock of Bacillus anthracis infected host. It inhibits cytokine production and cell proliferation/differentiation in various immune cells. This study showed that a brief exposure of LeTx caused a continual MEK1 cleavage and prevented tumor necrosis factor-α (TNF) production in response to lipopolysaccharide (LPS) in non-proliferating cells such as human peripheral blood mononuclear cells or mouse primary peritoneal macrophages. In human monocytic cell lines U-937 and THP-1, LeTx induced cell cycle arrest in G0-G1 phase by rapid down-regulation of cyclin D1/D2 and checkpoint kinase 1 through MEK1 inhibition. However, THP-1 cells adaptively adjusted to LeTx and overrode cell cycle arrest by activating the phosphatidylinositol 3-kinase/Akt signaling pathway. Inhibitory Ser-9 phosphorylation of glycogen synthase kinase 3β (GSK3β) by Akt prevented proteasome-mediated cyclin D1 degradation and induced cell cycle progress in LeTx-intoxicated THP-1 cells. Recovery from cell cycle arrest was required before recovering from on-going MEK1 cleavage and suppression of TNF production. Furthermore, pretreatment with LeTx or the GSK3-specific inhibitor SB-216763, or transfection with dominant active mutant Akt or degradation-defected mutant cyclin D1 protected cells from LeTx-induced cell cycle arrest, on-going MEK1 cleavage and suppression of TNF production. These results indicate that modulation of phosphatidylinositol 3-kinase/Akt/GSK3β signaling cascades can be beneficial for protecting or facilitating recovery from cellular LeTx intoxication in cells that depend on basal MEK1 activity for proliferation.

Bacillus anthracis is a Gram-positive spore forming bacterium and systemic infection of B. anthracis is often fatal when inhaled spores germinate inside the host (1). The virulence of B. anthracis is primarily attributed to two secreted exotoxins, lethal toxin (LeTx) 3 and edema toxin, and an exterior capsule, encoded in two large plasmids (2,3). The capsule comprises poly-D-glutamic acid, which protects bacteria from phagocytosis by the host immune cells and contributes to bacterial dissemination (4). LeTx and edema toxin are binary A:B toxins comprising protective antigen (PA) and lethal factor (LF) or edema factor (EF), respectively. PA is a molecular transporter allowing receptor-mediated entry and release of LF or EF into the cytosol. PA binds to either of two surface receptors: anthrax receptor 1 (also known as the tumor endothelial marker 8) and anthrax receptor 2 (also known as the capillary morphogenesis gene-2). Anthrax receptor 1 is expressed in high levels in macrophages, endothelial cells and several tumor cells (5-7); whereas, anthrax receptor 2 is widely distributed in human tissues (8). EF has adenylate cyclase activity (9), whereas LF is a zinc metalloprotease that cleaves and inactivates the N-terminal end of the mitogen-activated protein kinase (MAPK) kinases (MEK) 1 to 7, except MEK5, resulting in the inactivation of most their downstream signaling cascades. Both LF and EF cause severe defects in immune responses and contribute proliferation and dissemination of B. anthracis in the host (10,11).
Monocytes and macrophages are key innate immune cells and likely the first immune cells encountering the spores and germinating bacteria (12)(13)(14). LeTx induces rapid necrosis-like cell death of macrophages prepared from certain strains of inbred mice (15)(16)(17)(18). A genetic study investigating LeTx-sensitivity trait loci in different strains of mice identified the NACHT-leucine-rich repeat and pyrin domain-containing protein 1b (NALP1b) as a host factor that confers rapid LeTx cytotoxicity (19). Caspase-1 Ϫ/Ϫ macrophages with LeTx-susceptible genetic backgrounds were resistance to LeTx. LeTx also rapidly induces the release of interleukin-1␤ and -18 in LeTx-susceptible macrophages (20), indicating that LeTx induces Nalp1b-and caspase-1-mediated cell death. However, human macrophages lack NALP1b and are resistant to rapid necrotic cell death by LeTx. Instead, LeTx was shown to cause delayed apoptotic cell death of differentiated macrophages and inhibit cell proliferation/differentiation, which are probably mediated through MAPK inhibition (18,21,22). Because sustained MAPK activation is required for cell cycle progress of many different cells and the prevalence in expression of PA receptors, it is expected that LeTx causes cell cycle arrest in many cells and tissues of infected hosts. In fact, LeTx was shown to inhibit cell proliferation/differentiation in other cell types, including melanocytes (23), B lymphocytes (24), and T lymphocytes (25)(26)(27), which contribute to the host immune paralysis. To date the signaling mechanism of cell proliferation inhibition by LeTx and its contribution in the pathogenesis of anthrax remains largely unknown.
This study examined further details on the effects of LeTx on cell proliferation. LeTx induced cell cycle arrest in G 0 -G 1 phase via down-regulating key cell cycle progress proteins cyclin D1, D2, and checkpoint kinase 1 (Chk1) via MEK1 inhibition in human monocytic cell lines. LF incorporated inside cells had almost permanent MEK1-cleaving activity and completely blocked tumor necrosis factor-␣ (TNF) production in response to a bacterial component lipopolysaccharide (LPS). Recovery from the prolong MEK-cleaving LF activity and the TNF-suppressing effects required cell proliferation, which was mediated through adaptive response of THP-1 cells by inducing the phosphatidylinositol 3-kinase (PI3K)/Akt/GSK3␤ signaling pathway. Activation of PI3K/Akt by pretreating cells with LeTx or a chemical GSK3-inhibitor protected cell cycle arrest and facilitated recovery from cellular LeTx intoxication.

MATERIALS AND METHODS
Cell Culture and Reagents-The human monocytic leukemia cell line THP-1 and U937 cells were grown in complete RPMI 1640 medium containing 10% heat-inactivated fetal bovine serum (Sigma), 10 mM minimal essential medium non-essential amino acids solution, 100 units/ml penicillin G sodium, 100 g/ml streptomycin sulfate, and 1 mM sodium pyruvate. Cells were grown at 37°C in the humidified atmosphere of 5% CO 2 . LF and PA were prepared in the laboratory as previously described (28,29). LY294002, Akt inhibitor II, MG132, Ac-YVAD-CMK, SB202190, U0126, and JNK inhibitor I were purchased from Calbiochem (EMD Biosciences, La Jolla, CA). Chk1 inhibitor (SB218078) and GSK3 inhibitor (SB216763) was obtained from Sigma. Pan and phospho-site specific antibodies toward Akt, p38 MAPK, ERK1/2, MEK1 (COOH terminus of MEK1), and GSK3␣/␤ were obtained from Cell Signaling Technologies (Pickering, Ontario, Canada). The antibody raised against the N terminus of MEK1 was from QED Bioscience Inc. MEK3 and Chk1 antibodies were purchased form Santa Cruz Biotechnology, and cyclin D1 and LF antibodies were from Neo Markers (Fremont, CA) and List Biological Laboratories (Campbell, CA), respectively.
Preparation of Human and Murine Macrophages-Human peripheral blood mononuclear cells (PBMCs) were isolated from whole blood using Ficoll-Paque Plus (Amersham Bio-sciences). Blood obtained from volunteers were diluted with 3 volumes of phosphate-buffered saline and layered on top of the Ficoll at a 3 to 1 ratio. Blood cells were separated into different strata, and the lymphocyte layer was extracted. Isolated PBMCs were washed three times with phosphate-buffered saline and plated with complete RPM1 1640 media. For isolation of mouse peritoneal macrophages, mice received intraperitoneal injections of 3% thioglycollate 4 days before the preparation of peritoneal macrophages. Macrophages were harvested from mice by injecting 5 ml of saline into peritoneal cavity and re-extracted by withdrawal with a syringe. Cells were put through a cell strainer (40 m, BD Biosciences) and incubated at 37°C overnight in a humidified atmosphere of 5% CO 2. Bone marrow-derived macrophages were obtained from C57BL/6 mice as previously described (30), except culturing bone marrow cells with IL-3 (10 ng/ml) for 5 days.
Total Cell Lysate Preparation and Western Immunoblot Analysis-Total cell lysate extraction and Western blotting analysis were performed as previously described (17). Briefly, cells were lysed in ice-cold cell lysis buffer containing 20 mM MOPS, 15 mM EGTA, 2 mM EDTA, 1 mM Na 3 VO 4 , 1 mM dithiothreitol, 75 mM ␤-glycerophosphate, 0.1 mM phenylmethylsulfonyl fluoride, 1 g/ml aprotinin, 10 g/ml pepstatin A, 1 g/ml leupeptin, and 1% Triton X-100 and then sonicated on ice. Cell extracts were obtained by centrifuging the homogenate at 18,000 ϫ g for 10 min. These extracts were electrophoretically resolved in ready-made 10% SDS-PAGE gels (Bio-Rad), followed by transfer onto nitrocellulose membranes. Membranes were subsequently blocked with 5% skim milk for 30 min, immunoblotted with antibodies, and developed using an enhanced chemiluminescence detection system (ECL, Pierce Bioscience). Total cell lysates were obtained from THP-1 cells 48 h after LeTx (250 ng of LF and 500 ng of PA for 5 h) exposure and Kinetworks TM KCCP-1.0 Cell Cycle Protein multi-immunoblotting analysis was performed at Kinexus Bioinformatics Inc. (Vancouver, Canada) as described on the Kinexus website.
Cell Cycle and Proliferation Analysis-Analyses of intracellular fluorescence by carboxyfluorescein succinimidyl ester (CFSE) and DNA content using propidium iodide (PI) were performed using CellQuest software on a FACSCalibur flow cytometer (Becton Dickinson). For cell proliferation studies by CFSE, cells were incubated with CFSE of 1 M for 10 min at 37°C, and reaction was quenched on an ice bath for 2 min. Cells were washed with 3 volumes of complete media and plated. Cells were harvested daily for FACS analysis.
For DNA content analysis by PI, 1.0 ϫ 10 6 cells were harvested at the indicated time points, fixed by dropwise addition of ice-cold 70% ethanol after three times washing with 1ϫ phosphate-buffered saline containing 0.1% glucose, and were stored at 4°C. Subsequently, cell were pelleted by centrifugation and resuspended in staining solution containing 50 g of PI/ml and 100 units of RNase A/ml. After 60 min of incubation at room temperature, the cell was loaded in a FACSCalibur flow cytometer. The data were analyzed using CellQuest and ModFit LT 3.0 software (BD Biosciences). For the purpose of analysis, acquired events were gated to eliminate cell aggregates and debris.
Akt myr and Akt K179M Retroviruses Generation and Infection-The virus preparation and infection were performed as described (31). Briefly, constitutively active myristoylated Akt (Akt myr ) and dominant negative Akt (Akt K179M ) retroviruses were generated in Phoenix Amphotropic producer cells using the calcium phosphate method. Empty retroviral vector was transfected by same procedure for control virus. Cells were cultured for 24 h at 37°C and for another 24 h were incubated at 32°C for collection of viruses. Virus was collected and filtered through a 0.45-m filter. THP-1 cells of 1 ϫ 10 6 were suspended with viruses solution containing 4 g of Polybrene per milliliter and plated in 6-well plates. The plates were centrifuged at 2,500 rpm for 45 min at 32°C, and fresh RPMI 1640 media were added into the each well after 3-h incubation at 37°C. Viruses expressing cells were selected by green fluorescent protein 48 h after infection and treated with the LeTx for cell cycle analysis.
pCDNA-Cyclin D1 T286A Construction and Transfection-pCDNA3 containing human mutant cyclin D1-T286A was made by digesting a pFlex-D1-T286A (32) with the EcoR1 restriction enzyme and ligated the resulting fragment into the pCDNA3.1 vector (Invitrogen). The construct was verified by DNA sequencing. The stable transfection of constructed plasmids was performed by electrophoresis using Gene Pulser Xcell (Bio-Rad) in THP-1 cell line, and mutant cyclin D1 expression cells were selected by G418 antibiotics.

LeTx Induces G 1 Cell Cycle
Arrest-Previous studies have shown that LeTx inhibits proliferation and differentiation of immune cells (22,24,27). We further examined the mechanism of cell cycle arrest in the human monocytic cells lines U937 and THP-1. Proliferation of these cells after brief exposure of LeTx (250 ng/ml LF and 500 ng/ml PA for 5 h) was analyzed for 4 days by measuring the dilution of intracellular fluorescence dye CFSE in dividing cells. LeTx temporarily inhibited cell proliferation of both U937 and THP-1 cells for up to 2 and 3 days, respectively (Fig. 1A). Cell cycle analysis using PI chromosome staining showed that number of cells in G 0 -G 1 phase of cell cycle was increased, whereas numbers in S and G 2 -M phase were greatly diminished in LeTx-treated THP-1 and U937 cells (Fig. 1B).
LeTx Causes a Rapid Down-regulation of Cyclin D1, Cyclin D2, and ChK1 mRNA Levels through MEK1 Inhibition-To identify signaling molecules involved in the LeTx-induced cell cycle arrest, we examined 25 different cell cycle-related proteins using Kinetworks TM multi-immunoblotting analysis in LeTx-treated THP-1 cells. Among them, protein levels of cyclin D1 and Chk1 were most significantly diminished by LeTx ( Fig.  2A). Both protein and mRNA levels of cyclin D1 rapidly dropped in 3 h and maximally diminished within 6 h; whereas, the decrease in protein and mRNA levels of Chk1 were less dramatic and maximally decreased 24 -48 h after an LeTx treatment (Fig. 2, B and C). The rapid decrease of cyclin D1 in mRNA and protein levels was likely due to inhibition by LF at the transcriptional levels and the short half-life of the mRNA and protein. It is unlikely that cyclin D1 was directly targeted by LF, because cells treated with the proteasome inhibitor MG132 (data not shown) or stably transfected with proteasome-mediated degradation-defected mutant cyclin D1 (cyclin D1 T286A ) were resistant to LeTx-induced cyclin D1 down-regulation (Fig. 2D). Also cyclin D1 mRNA levels in cells treated with the transcription inhibitor actinomycin D were down-regulated with the same rate in both non-treated and LeTx-treated cells (data not shown).
Because LeTx cleaves the N-terminal end of MEKs and inhibits most downstream MAPKs, we examined whether inhibition of MAPKs could cause cell cycle arrest in THP-1 cells. Treatment of the MEK1 inhibitor U0126 (10 M) for 24 h caused accumulation of cells at G 0 -G 1 phase and prevented entering into S phase; whereas, the p38 MAPK inhibitor SB202190 (SB, 10 M) had no effects on S phase progress, but it significantly prevented cells entering G 2 phase (Fig. 3A). Treatment of both U0126 and SB additively blocked cell cycle progression. A JNK inhibitor (10 M) had no effect on cell cycle progress of THP-1 cells. These results indicate that LeTx prevents cells entering S phase and G 2 phase by inhibiting ERK1/2 and p38 MAPK, respectively. Similar results were detected in U937 cells. When THP-1 cells were treated with U0126, SB, or JNK inhibitor, only U0126 induced down-regulation of both cyclin D1 and Chk1, suggesting that the down-regulation of both cyclin D1 and ChK1 by LF was at least in part mediated through inhibiting MEK1 (Fig. 3B). In addition to cyclin D1, mRNA levels of cyclin D2 were similarly down-regulated by U0126 and LeTx, but not by SB or JNK inhibitor. LeTx had no effects on mRNA levels of cyclin B and cyclin E (Fig. 3C).

Inhibition of MAPKs and TNF Production in Response to LPS by LeTx
Were Permanent in Non-proliferating Cells-We have previously shown that LF reside inside cells and continuously cleaves MEK1 for prolong periods (ϳ4 -5 days) in murine macrophage cell line RAW246.7 cells (33). To examine the duration of LF activity in non-proliferating cells, PBMCs were briefly exposed to LeTx for 5 h, and cells were cultured in fresh media for up to 6 days. Each day, LeTx-exposed and non-exposed cells were treated with LPS (1 g/ml) for 6 h, and TNF production was measured using a TNF bioassay as previously described (34) and MEK-1 N-terminal cleavage was examined using Western blots. Throughout the cell culture period, a brief exposure of LeTx continuously cleaved MEK1 (Fig. 4A, upper panel) and prevented TNF production in responses to LPS (Fig. 4A, middle panel). Because PBMCs are a heterogeneous population of cells, including monocytes, T cells, and B cells, we further characterized TNFproducing cells in PBMCs and found that almost all TNFproducing cells in day 1 were CD13 ϩ myeloids and in day 6 were mostly CD3 ϩ lymphocytes (data not shown). Most of the CD13 ϩ cells were short-lived and did not survive after 24-h culture in both LeTx-treated and non-treated PBMCs (data not shown). To specifically examine in non-proliferating macrophages, mouse peritoneal macrophages were isolated from C57BL/6 mice harboring LeTx-resistant macrophages (17) and treated with LeTx. MEK1 was continuously cleaved by a brief exposure of LeTx for the next 6 days (Fig.  4A, lower panel) by when cells started dying by senescence. TNF production in response to LPS was also abolished for the next 6 days (data not shown). Cell Cycle Inhibition by LeTx Was Recovered in the Absence of ERK1/2 Activity in Monocytic Cells-The effect of LeTx is almost permanent in non-proliferating cells. If LeTx causes cell cycle arrest, the effect of LeTx should be permanent in these cells. However, when we examined MEK1 N-terminal cleavage and TNF production in response to LPS in LeTxtreated THP-1 cells, both MEK1 and TNF production were returned to normal levels in ϳ5 days after LeTx treatments (Fig. 4B). Based on Western blot analysis, LF inside cells was slowly degraded over Ͼ3 days (Fig. 4C). LF was no longer detectible after 4 days of a LeTx treatment. We then examined whether cell cycle arrest by LeTx was alleviated before MEK1/ERKs activities were recovered. In fact, THP-1 cells resumed cell cycle before the third day even when LF was readily detectible by Western blots (Fig. 4C), and TNF production in response to LPS was completely blocked (Fig. 4B,  lower panel). Cell cycle arrest was released after the second day and returned to almost at normal levels after 3 days of a LeTx treatment (Fig. 4D). These results indicate that LeTxexposed THP-1 cells underwent cell cycle progression in the absence of MAPK activities.
Cells Pre-exposed with LeTx Were Protected from Cell Cycle Arrest by Subsequent LeTx Treatment through Activation of PI3K-If cells adaptively adjusted themselves to LeTx and resumed cell cycle even in the absence of MEK activities, these cells should continue to proliferate even in the presence of further LeTx challenges. To examine whether these cells were indeed resistant to LeTx-induced cell cycle arrest, we briefly pre-treated cells for 5 h with LeTx and then treated them again 2 days after the pretreatment. Consistent with previous results, numbers of cells at S and G 2 -M phases were dramatically reduced by LeTx in 2 days, but LeTx-pretreated cells were resistant to cell cycle arrest by a subsequent LeTx-treatment (Fig.  5A, left panel). To examine whether the resistance to cell cycle arrest was due to inability of LeTx-exposed cells to incorporate LeTx into cells or MEK1-independent ERK1/2 activation, MEK-1 N-terminal cleavage and ERK1/2 phosphorylation were analyzed. Consistent with previous results in Fig. 4B, full-length MEK-1 and basal ERK1/2 phosphorylation were detected in 5 days after LeTxexposure, which were cleaved and inhibited, respectively, by further LeTx treatment (Fig. 5A, right panel). These results indicate no defects of LeTx pre-exposed cells in incorporating LF into the cytosol and preventing ERK1/2 activation. Previous studies in hematopoietic cells have shown that PI3K activation rescues G 1 cell cycle arrest (35). Thus, we examined whether PI3K inhibitors such as LY294002 (Ly) could prevent the adaptive response. Ly treatment alone in THP-1 cells had no effects on cell cycle progress. When LeTx-pretreated cells were treated with Ly (10 M), the recovery from cell cycle entering S phase was completely prevented even after 4 days of LeTx treatment (Fig. 5A). Similar results were detected by another PI3K inhibitor wortmannin (data not shown). We then examine whether Akt, which is a downstream protein of PI3K signaling, was activated by LeTx. Indeed, LeTx induced Akt phosphorylation at the activating Ser-473 site, in 2 days of LeTx treatment, and it maximally phosphorylated Akt in day 3, which was gradually diminished over Ͼ6 days (Fig. 5B, upper panel). The Akt phosphorylation appeared to be mediated through PI3K, because Ly or wortmannin (Wort) was able to block the Akt activation (Fig.  5B, lower panel). To further confirm that Akt activation is involved in cell cycle re-initiation in the presence of LeTx, we analyzed cell cycle progress in THP-1 cells treated with LeTx for 2 days, followed by treating or not treating with the Akt inhibitor II (Akti) or in cells transfected with retrovirus vectors

. An adaptive response of THP-1 cells to LeTx-induced cell cycle arrest.
A, non-proliferating human PBMCs or 3% thioglycollate-elicited primary peritoneal mouse macrophages were treated with LeTx (250 ng of LF and 500 ng of PA) for 5 h and cultured in the fresh medium for 6 days until cells became senescent and apoptotic. Top and bottom panels, N-terminal cleavage of MEK1 was examined for different time points as indicated in the panel. Western blots for Akt and p38 MAPK were used for loading controls. Middle panel, production of TNF in response to LPS (1 g/ml) was measured 5 h after LPS treatment of cells following 1, 3, and 6 days after brief LeTx exposures. The presence of TNF in the culture medium was measured using bioassays as previously described (34). Data show mean Ϯ S.D. (n ϭ 3). B, top panel, THP-1 cells were treated with LeTx (250 ng of LF and 500 ng of PA for 5 h), and inhibition of MAPKs by LeTx was examined using Western blotting (left panel) against N-terminal MEK1 and MEK3, phospho-ERK1/2 (pERK1/2) and phospho-p38 MAPK (p-p38 MAPK) at different time point as indicated in the panel. Western blotting for total p38 MAPK was used for loading control. Results are a representative data obtained at least two separated experiments. Bottom panel, production of TNF in response to LPS (1 g/ml) was measured 5 h after treating LPS in cells after 1, 2, 3, 5, and 6 days after brief LeTx exposures using a TNF bioassay. C, similarly, intracellular LF was visualized by Western blots after various time points in cells briefly exposed to LeTx. D, the percentage of cell cycle distribution was analyzed using Modfit 3.0 software according to the PI content in cells at different time points as indicated in the figure in LeTx-exposed THP-1 cells.
containing inactive Akt mutant Akt (Akt K179M , possessing an inhibitory mutation at the ATP-binding site) or membranetargeted constitutively-active mutant Akt (Akt myr , myristoylated Akt). The retroviral vectors also contained internal ribosome entry site sequences followed for green fluorescent protein, which allow sorting of only retrovirus-transfected cells (36). THP-1 cells treated with Akti or green fluorescent protein-positive Akt K179M -transfected cells showed much slower recovery than non-treated or vector control-transfected cells; whereas, green fluorescent protein-positive cells transfected with Akt myr recovered from cell cycle arrest much faster than vector control-transfected cells (Fig. 5C). Consistent with these results, protein levels of cyclin D1 and Chk1 were similarly recovered in 3 days after LeTx treatment, and Ly completely inhibited the spontaneous recovery of cyclin D1 and Chk1 in mRNA levels (data not shown) and protein levels (Fig. 5D). These results indicate that activation of PI3K and Akt in response to LeTx overrode cell cycle arrest induced by ERK1/2 MAPK inhibition, which is also required for recovery from cellular LeTx intoxication.

Inhibition of GSK3␤ by Akt Stabilized Cyclin D1 Degradation and Protected from LeTx-induced Cell
Cycle Arrest-Akt negatively regulates the GSK3␣ and -␤ isoforms by phosphorylating them on Ser-21 or Ser-9, respectively. We examined whether Akt activation by LeTx led phosphorylation of GSK3 and whether GSK3 inhibition protects LeTx-induced cell cycle arrest. Western blotting with a phosphospecific antibody for GSK3␣/␤ readily detected only 46-kDa GSK3␤ in THP-1 cells, and the phosphorylation was further enhanced 2 days after LeTx treatment (Fig. 6A). The phosphorylation was diminished by Ly or Akti (Fig. 6A). Furthermore, THP-1 cells treated with LeTx together with the GSK3-specific inhibitor SB216763 were protected cells from cell cycle arrest induced by LeTx (Fig.  6B). GSK3␤ was shown to regulate cell cycle through cyclin D1 proteolysis (37).
Because GSK3␤ phosphorylates cyclin D1 on Thr-286 and causes proteasome-mediated degradation, we examine whether GSK3␤ inhibition by Akt stabilizes cyclin D1. THP-1 cells were non-treated or pretreated with LeTx for 24 h and then treated with the GSK3-specific inhibitor SB216763 for up to 16 h. Protein levels of cyclin D1 started to increase as early as 1 h of the treatment, whereas MEK1 was cleaved by LF (Fig. 6C, top panel). Cyclin D1 protein was accumulated with the similar rate by a proteasome inhibitor MG132 (Fig. 6C,  bottom panel). However, cyclin D1 mRNA levels did not change by either treatment (data not shown). These results indicate that GSK3␤ inhibition by Akt at least in part stabilized cyclin D1 and overrode cell cycle arrest induced by LeTx in THP-1 cells.
Cell Cycle Progress Is Crucial for Recovery from Cellular LeTx Intoxication-LeTx intoxication was almost permanent in nonproliferating PBMCs and peritoneal macrophages (Fig. 4A). It took ϳ4 -5 days for THP-1 cells to recover from LeTx intoxication (Fig. 4B), and their cell cycle resumed after 2 days of LeTx treatment. Therefore, it is expected that preventing or facilitating cell cycle recovery will result in slow or fast recovery from LeTx intoxication. LeTx-exposed THP-1 cells were treated by Ly 3 days after LeTx treatment to prevent cell cycle recovery (Fig. 5A), and recovery of MEK1 cleavage by LF was examined. Full-length MEK1 was readily detectible after 6 days of LeTx exposure, whereas Ly-treated LeTx-exposed cells failed to recover from MEK1 cleavage (Fig. 7A). Expression of cyclin D1 T286A prevented LeTx-induced cell cycle arrest (Fig. 6D) and induced faster recovery from LeTx intoxication. THP-1 cells stably transfected with cyclin D1 T286A recovered from LeTxinduced MEK1 cleavage and suppression of TNF production in response to LPS much faster than vector transfected cells (Fig. 7B).

Inhibition of GSK3 Stabilized Cyclin D1 and Facilitated LeTx Recovery in Bone Marrow-derived
Macrophages-We further examined whether cell cycle progress also facilitated recovery from LeTx intoxication in non-transformed cells. Bone marrow-derive murine macrophages proliferate in vitro for several weeks in the presence of macrophage-colony stimulating factor or IL-3, which requires ERK activities (38). LeTx also cleaved MEK1, down-regulated cyclin D1 (Fig. 8A), and induced cell cycle arrest in these cells (data not shown). After 4 days of a LeTx treatment, these cells also spontaneously recovered from MEK1 cleavage and cyclin D1 down-regulation. Such recoveries were not detected in cells treated with Ly, whereas cells treated with GSK3-specific inhibitor SB216763 increased cyclin D1 protein levels and facilitated recovery from MEK1 cleavage (Fig. 8A). In consistence, TNF production also returned to ϳ70% of normal levels after 4 days of a LeTx treatment (Fig. 8B, right panel). In cells treated with Ly, TNF production was still completely blocked even after 4 days of a LeTx treatment. In contrast, cells treated with SB216763 produced TNF up to ϳ50% of normal levels even in 2 days after a LeTx treatment (Fig. 8B,  left panel). These results suggest that the PI3K/Akt/GSK3 signaling cascade is required for recovering from cellular LeTx intoxication and sustaining cell cycle progress by stabilizing cyclin D1 facilitated recovery from the intoxication.

DISCUSSION
LeTx was shown to inhibit cell cycle progress in monocytic cell lines (22), melanocytes (23), activated B cells (24), and activated T cells (25)(26)(27). This study further showed that LeTx prevented entry of S phase human monocytic cell lines through inhibition of ERK1/2 (Fig. 1). LeTx receptors have been shown to be widely expressed (8); therefore, cell cycle arrest by LeTx is likely to occur in a number of cell types that require basal ERK1/2 activ- . Two days later, phosphorylation of GSK3␤ on Ser-9 was analyzed using Western blotting. B, similarly, THP-1 cells were exposed to LeTx in the presence of absence of GSK3 inhibitor (GSK3i, 10 M), and cell cycle progress was analyzed using Modfit 3.0 software according to the PI content in cells. C, similarly, THP-1 cells were exposed to LeTx and, 1 day later, cells were treated with GSK3 inhibitor (GSK3i, 10 M) or proteasome inhibitor MG132 (10 M) for different periods as indicated in the panel. N-terminal degradation of MEK1 or presence of cyclin D1 was analyzed using Western blotting. D, proteasomemediated degradation-resistant mutant cyclin D1 (cyclin D1 T286A ) or control vector was stably transfected in THP-1 cells. These cells were then exposed to LeTx (250 ng of LF and 500 ng of PA for 5 h), and cell cycle progress was measured. Results shown in the figure are representative data of similar results observed in at least three independent experiments. ity for proliferation. ERK1/2 was shown to be required for cell cycle progress from G 1 to S phase through multiple mechanisms (39). Using Kinetworks TM multi-immunoblotting analysis and real-time PCR analysis, we found that LeTx down-regulated cyclin D1 and Chk1 protein levels Ͼ10-fold (Fig. 2, A-C), likely through inhibiting transcriptions of their genes (Fig. 2, C  and D). Because the cyclin D1 promoter contains a functional AP-1 binding site (40), it is possible that basal ERK1/2 activities were required for maintaining appropriate levels of cyclin D1 by phosphorylating AP-1 proteins such as c-Jun, JunB, JunD, and c-Fos in THP-1 cells. Indeed, the MEK1 inhibitor U0126, which specifically prevents ERK1/2 activation, also down-regulated mRNA and protein levels of cyclin D1 and ChK1 (Fig. 3, B and C). Regulation of cyclin D2 expression by ERK1/2 is less understood, but B cells were shown to require cyclin D2 accumulation through B-cell receptor-mediated ERK1/2 activation for proliferation (41). The essential role of cyclin D1 and D2 for escaping from G 1 and initiation and completion of S phase through activating cyclin-dependent kinases 2, 4, and 6 has been documented as well (42). The down-regulation of Chk1 by LeTx was also likely mediated through inhibiting ERK1/2 activity (Fig. 3, B and C). Chk1 was shown to play a key role in the G 2 /M checkpoint in DNA damage response (43). Activation of Chk1 upon DNA damage prevents G 2 /M transition by inhibiting activation of Cdc25, a dual phosphatase, resulting in inhibition of Cdc2 kinases (44). Chk1 was also shown to be required for cell cycle progress for S and G 2 /M phases in normal cell growth (45,46), but cellular requirement for Chk1 may be different at various developmental stages and cell types (47,48). The Chk1 inhibitor SB218078 (1 M) did not prevent cell cycle progress in THP-1 cells (data not shown), and expression of cyclin D1 T286A alone was able to override cell cycle arrest induced by LeTx (Fig. 6D), indicating that basal Chk1 activity is not required for normal cell cycle progress in THP-1 cells.
Previously we have shown that LF stays inside cell and continuously cleaves MEKs for 4 -5 days after a brief exposure of LeTx in murine macrophage cell line RAW246.7 cells (33). When LeTx was briefly exposed to non-proliferating human PBMCs or murine peritoneal macrophages, the MEK1 cleavage and inhibition of TNF production in response to LPS lasted as long as the cells survived in culture (Fig. 4A). If LeTx causes cell cyclearrest, it is expected that the LeTx effects are similarly prolonged as non-proliferating cells. However, intact MEK1 became visible, and TNF response to LPS was recovered in 4 -5 days after LeTx treatment in THP-1 cells (Fig. 4B). In fact, cell cycle of LeTx-treated THP-1 cells progressed before day 3 of LeTx treatment, even before MEK1 cleavage was alleviated (Fig. 4D). Furthermore, THP-1 cells pre-exposed to THP-1 cells were resistant to cell cycle arrest by subsequent LeTx challenges (Fig. 5A). This phenomenon is a reminiscent of the adaptive response to LeTx-induced cytolysis in murine macrophages (33,49). Certain murine macrophage cell lines or primary macrophages originated from inbred mice undergo rapid NALP1bdependent cytolysis by LeTx (16,17,19). These macrophages adaptively respond to a low non-cytolytic dose of LeTx and become resistant to subsequent cytolytic doses of LeTx; this is termed toxin-induced resistance (33,49). We have shown that toxin-induced resistance to cytolysis is mediated through down-regulation of two mitochondrial cell death proteins, Bcl-2/adenovirus E1B 19 kDa-interacting protein 3 (Bnip3) and Bnip3-like (33). Human macrophages do not express NALP1b and are resistant to LeTx-induced cytolysis (18,21,22). The adaptive response to LeTx-induced cell cycle arrest in THP-1 and U-937 cells was mediated through activating the PI3K/Akt signaling pathway (Fig. 5), and, more importantly, recovery from cell cycle arrest was a prerequisite for cellular recovery from the on-going MEK1 cleavage by LF (Figs. 5D and 7B). At this moment, it is unknown why LF can stay active for such a long period in the cytosol. A prolonged cellular intoxication of LF in non-proliferating cells indicates that the toxic effects of LeTx in a B. anthracis-infected host will last much longer than one may expect, considering that most tissues are terminally differentiated and not actively proliferating. One bolus administration of non-lethal dose of LeTx in mice was able to cleave MEK1 up to 7 days in various tissues such as heart, lung, liver, and kidney, although serum LeTx was cleared within 3 h after injection (data not shown).
Activation of the PI3K/Akt pathway was previously shown to regulate cell cycle progress through inducing cyclin D1 expression, cytoplasmic relocalization of the cyclin-dependent kinase inhibitor p27, and inhibiting transcriptional activation of a number of related forkhead transcriptional factors (FoxO1, FoxO3, and FoxO4). Several studies have also demonstrated that PI3K activation is sufficient to induce cyclin D1 expression in an ERK1/2-independent manner (50,51) and override G 1 arrest in hematopoietic cells (35). Given the importance of cyclin D1 in G 1 to S phase cell cycle progress, it is not surprising that the adaptive response of THP-1 cells in cell cycle arrest was at least in part mediated through reinstating cyclin D1 protein via the PI3K/Akt/GSK3␤ pathway (Fig. 9). The half-life of Bone marrow-derived macrophages from C57BL/6 mice were exposed to LeTx (250 ng of LF and 500 ng of PA for 5 h) in the presence or absence of GSK3 inhibitor (GSK3i, 10 M) or Ly294002 (Ly, 10uM). A, N-terminal degradation of MEK1 or presence of cyclin D1 was analyzed using Western blotting after 2 or 4 days of a LeTx treatment. Results shown in the figure are representative data of similar results observed in two independent experiments. B, production of TNF in response to LPS (1 g/ml) was measured 5 h after treating LPS in these cells using TNF bioassay. Data show mean Ϯ S.D. (n ϭ 3). DECEMBER 14, 2007 • VOLUME 282 • NUMBER 50 cyclin D1 in proliferating cells is ϳ20 min (52), and the rapid turnover is due to ubiquitination and proteasome-mediated degradation of phosphorylated cyclin D1 (32). Cyclin D1 phosphorylation on Thr-286 was mediated by the GSK3␤ (37), which is inhibited via phosphorylation on Ser-9 by Akt (53). Therefore, it can be predicted that reagents activating PI3K/ Akt or inhibiting GSK3␤ will protect LeTx-induced cell cycle arrest in susceptible cells and facilitate recovery from a prolonged LeTx intoxication, whereas reagents inhibiting the PI3K/Akt pathway, which can be found in several anti-inflammatory drugs, can be detrimental to some cells. At cellular levels, we showed that inhibition of GSK3␤ or expression of degradation mutant cyclin D1 indeed protected cell cycle arrest induced by LeTx and facilitated recovery from LeTx intoxication (Figs. 6 and 7). It should be noted that GSK3␣ and -␤ are phosphorylated at Tyr-279 and Tyr-216, respectively, by MEK1/2. 4 Therefore, it is possible that some of the effects of MEK inhibition by LeTx and other MEK inhibitors such as U0126 and PD98059 could also proceed by loss of GSK3 phosphotransferase activity. Fig. 9 illustrates signaling pathways involved in LeTx-induced cell cycle arrest and PI3K/Akt/ GSK3␤/cyclin D1-mediated recovery pathways.

LeTx-induced Cell Cycle Arrest and Recovery
LeTx has been regarded as a key virulence factor in the pathogenesis of anthrax, causing immune paralysis, cell cycle arrest, and cell death in host immune cells. These effects could contribute survival and proliferation of B. anthracis within the host. Furthermore, the majority of the infected hosts eventually succumb despite bacterial clearance by antibiotics. The fatal effects are attributed to production and release of anthrax toxins. This study showed that LF resides inside cells and contin-uously cleaves MEKs almost indefinitely in non-dividing cells, indicating an extensive period of LF intoxication in cells. The recovery from the effects of LeTx can be facilitated by activating PI3K/Akt/GSK3␤ signaling-mediated adaptive responses. Reagents activating the signaling pathways or inhibitors of GSK3␤ can be novel tools for treating for LeTx toxemia, whereas reagents that inhibit the adaptive signaling pathway can be detrimental to the host.

Cell cycle progress
Recovery from LeTx intoxication FIGURE 9. A diagram illustrating the LeTx-induced cell cycle arrest and recovery pathways. LeTx caused cell cycle arrest at least in part by inhibiting the MEK1 3 ERK1/2 3 cyclin D1 pathway, and possibly GSK3 activation, which is also catalyzed by MEK1/2. An adaptive response activating PI3K 3 Akt 3 GSK3␤ stabilizes cyclin D1, overrides cell cycle arrest, and facilitates recovery from LeTx intoxication.