|Home | About | Journals | Submit | Contact Us | Français|
The etiologic agent of inhalational anthrax, B. anthracis, produces virulence toxins that are important in the disease pathogenesis. Current studies suggest that mouse and human macrophages are susceptible to immunosuppressive effects of one of the virulence toxins, lethal toxin (LT). Thus a paradigm has emerged that holds that the alveolar macrophage (AM) does not play a significant role in the innate immune response to B. anthracis or defend against the pathogen as it is disabled by LT. This is inconsistent with animal models and autopsy studies that show minimal disease at the alveolar surface. We examined whether AM are immunosuppressed by LT. We found that human AM (HAM) were relatively resistant to LT-mediated innate immune cytokine suppression, MEK cleavage, and induction of apoptosis as compared to mouse RAW 264.7 macrophages. Mouse AM (MAM) and murine bone marrow derived macrophages (BMDM) were also relatively resistant to LT-mediated apoptosis despite intermediate sensitivity to MEK cleavage. The binding component of LT, protective antigen (PA), does not attach to HAM, although it did bind to MAM, murine BMDM and RAW 264.7 macrophages. HAM do not produce significant amounts of the PA receptors anthrax toxin receptor 1 (TEM8/ANTXR1) and anthrax toxin receptor 2 (CMG2/ANTXR2). Thus, mature and differentiated AM are relatively resistant to the effects of LT as compared to mouse RAW 264.7 macrophages. AM resistance to LT may enhance clearance of the pathogen from the alveolar surface and explain why this surface is relatively free of B. anthracis in animal models and autopsy studies.
Anthrax, a virulent and zoonotic disease recognized since early human history, is caused by a Gram-positive, aerobic, spore-forming, rod-shaped bacterium, Bacillus anthracis. Three primary forms of the disease are due to three different mechanisms of exposure: ingestion (gastrointestinal), contact (cutaneous) or inhalation (inhalational) (1). Although all forms of disease can lead to fatal system infection, inhalation is considered to be the most life-threatening pathway and was the most lethal primary form identified during the recent bioterrorism attack in 2001 and during the accidental exposure in the process of manufacture in Sverdlosk in 1979 (1, 2).
This disease occurs after inhalation of spores between 1 and 5 microns. This is the optimal size for delivery of the pathogen to the alveolar space, which is the site of entry of the pathogen (3). Inhalational anthrax is characterized by a rather unique finding in that the inhaled spores do not vegetate and cause disease at the site of entry (4-7). That is, pneumonia is not a typical feature of inhalational anthrax. Instead, spores are rapidly and efficiently phagocytosed by alveolar macrophages (AM) and dendritic cells (DC)(5, 8, 9), and subsequently carried through lung tissue and lymph ducts to the regional thoracic lymph nodes (TLN). It is from the lymph nodes that dissemination occurs.
B. anthracis produces two binary toxins and a capsule that appear to play some role in the pathogenesis of inhalational anthrax. It contains a plasmid, pX01, that encodes the three toxin components: an 83 kD lethal factor (LF), an 89 kD edema factor (EF), and an 85 kD protective antigen (PA). A second plasmid, pX02, encodes genes involved in synthesis of a poly-D-glutamyl capsule. Deletion of either plasmid attenuates virulence (10).
LF and EF each separately form a binary toxin with PA such that two different binary toxins are formed: lethal toxin (LT) consisting of PA plus LF, and edema toxin (ET) consisting of PA plus EF (11). The binary forms of the toxins are so named because of their biological effects in animal models. Intradermal injection of ET (PA+EF) induces edema, while injection of high concentrations of LT (PA+LF) causes severe hypotension and death (12).
There is evidence that B. anthracis toxins play a role in the pathogenesis of inhalational anthrax, but the exact role has been debated. In mice, at high exposure to spores intratracheally, specific mutations of the LF, EF, or PA genes did not appear to have a large effect on the LD50 or mean time to death (13). This work argues against toxins acting locally to facilitate the development of disease in animals exposed to B. anthracis. However, monoclonal antibodies to PA reduce dissemination from the lung by B. anthracis in rabbits, and it has long been known that vaccination with PA is protective (14-16).
B. anthracis LT has been shown to induce apoptosis in certain mouse macrophages through cleavage of MEK kinases (17). These cells, as well as HAM have also been demonstrated to express mRNA for the known anthrax toxin receptors (ATR) : anthrax toxin receptor 1 (TEM8\ANTXR1) and anthrax toxin receptor 2 (CMG2\ANTXR2)(18). These findings, as well as the experiments in rabbits, have led to the development of a paradigm that concludes that LT facilitates dissemination of B. anthracis through LT-mediated immunosuppression of the alveolar macrophage.
Inhibition of the main resident phagocyte of the lung, the alveolar macrophage, by B. anthracis toxins would seem to be in conflict with findings on human autopsies that the alveolar space is cleared of the pathogen. This also occurs in animal models of inhalational anthrax (4-7).
We have previously shown that human AM (HAM) rapidly and efficiently phagocytose B. anthracis spores. Spore exposure also induces production of several cytokines and chemokines through activation of MAPK signaling pathways (19). These results clearly showed that HAM have a robust innate immune response to B. anthracis. This led us to hypothesize that in contrast to the previous studies, HAM may play an active role in protection of the host against the pathogen. Here, we compared the immunosuppressive effect of B. anthracis LT on HAM, Mouse Alveolar Macrophages (MAM), murine Bone Marrow-Derived Macrophage (BMDM) and mouse RAW 264.7 macrophages.
We found that HAM, unlike mouse RAW 264.7 macrophages, are resistant to MEK cleavage induced by LT. HAM were also less sensitive to LT-mediated suppression of the innate immune cytokine response to spores than mouse RAW 264.7 macrophages. HAM, MAM and murine BMDM were resistant to the proapoptotic effects of LT as compared to RAW 264.7 macrophages. We found that binding of PA, the binding partner of LF and EF, was minimal in HAM, as was expression of the two known receptors of PA, TEM8\ANTXR1 and CMG2\ANTXR2. This occurred even though mRNA for both receptors was expressed in HAM.
The findings demonstrate that mature macrophages are resistant to the immunosuppressive effects of the anthrax toxin LT. The results suggest that, rather than a target for the virulence toxins and susceptible to their effects, AM are, in fact, resistant to the toxins and are likely an obstacle that must be overcome by the pathogen in order to establish disease in the host. The findings may explain why normal, non-edematous, alveolar surfaces are free of the pathogen, as alveolar macrophages are still able to respond vigorously to B. anthracis as they are resistant to its exotoxins.
RAW 264.7 mouse macrophage cells and B. anthracis Sterne strain 7702 (pX01+ pX02–) were kindly provided by Jimmy Ballard (University of Oklahoma Health Sciences Center, Oklahoma City, OK). RAW cells were cultured in complete media, RPMI-1640 with L-glutamine (Sigma-Aldrich, St. Louis, MO), 10% FBS (Mediatech, Manassas, VA), 1% penicillin-streptomycin solution (Mediatech). LF, PA63 and PA63-FITC were purchased from List Biological Laboratories (Campbell, CA). Antibodies and the vendors are listed below: anti-TEM8\ANTXR1 IgG, Abcam, Cambridge, MA; anti- rabbit IgG, GeneTex, San Antonio, TX or Cell Signaling Technology, Beverly, MA; anti-CMG2\ANTXR2 (human and mouse) IgG, R&D Systems, Minneapolis, MN; anti-goat IgG, R&D Systems; anti-CD11b, Abcam, San Diego, CA; anti-MEK 1 (Chemicon, Billerica, MA); anti-MEK 2 (Santa Cruz, Santa Cruz, CA), anti-MEK 4 (Santa Cruz); anti-MEK 6 (BioLegend, San Diego, CA); anti-MEK 7 (Life Span Biosciences, Seattle, WA).
For spore preparation, bacteria were grown overnight at 37°C with continuous shaking in LB medium and were then streaked onto Arret and Kirchbaum agar sporulating slants. Bacteria were incubated for 3 weeks at 30°C. The slants were washed with 10 ml of chilled, sterile, deionized water, spun at 10,000 × g for 10 min; and resuspended in 10 ml chilled water. The spore suspension was heated at 65°C for 30 min to kill vegetative bacteria. After heat treatment, the spores were centrifuged for 10 min at 10,000 × g. The pellet was washed five times to remove contaminating cell debris. The supernatant and the very top layer of the pellet were aspirated and discarded each time, and the spores were then resuspended in chilled sterile deionized water and centrifuged for 10 min. The titer of the spore preparation was determined by plate counts. The spores were diluted to 1 × 109 spores/ml and stored at 4°C. Titers were reconfirmed by plate counts before each use. There was no detectable endotoxin in the final spore dilutions used in the experiments as determined by a limulus amebocyte lysate assay (Cambrex, Walkersville, MD).
Human macrophages were obtained by bronchoscopy with the signed informed consent of human subjects according to a protocol approved by the Oklahoma University Health Sciences Center Institutional Review Board and the Institutional Biosafety Committee. These volunteers were healthy nonsmoking subjects, aged 18 to 35 years, with no history of pulmonary or cardiac disease or recent infections. Cells from human subjects were collected in sterile saline solution and centrifuged at 500 × g for 5 min. The supernatant was removed, and the pellets were washed in 10 ml of RPMI 1640 medium containing 50 μg/ml gentamicin and resuspended in 10 ml RPMI medium plus 2% FCS supplemented with 50 μg/ml gentamicin (RPMI-2). Cell counts were determined by a hemocytometer. The cell type was determined by morphology using Diff-Quick staining (Baxter, Miami, FL), as well as by flow cytometry using CD16. Cells were resuspended to a concentration of 1 × 106 macrophages/ml. There were >93% human macrophages in each BAL cell preparation by morphology and flow cytometry. One milliliter of cells per well was plated onto 24-well plates and allowed to incubate for 2 to 4 h to facilitate attachment. Subsequently, the medium was removed, and fresh medium containing 2% FCS and 50 μg/ml gentamicin was added. The cells were incubated overnight at 37°C with 5% CO2.
MAM were obtained from lavage fluids of mice. C57BL6 mice were deeply anesthetized and exsanguinated by cutting the distal aorta. The thorax was opened, and a 20-gauge blunt needle was tied into the proximal trachea for lung lavage. Lungs were then flushed with 1ml PBS for up to 15 times through the tracheal tube. The aliquots from each animal were pooled for analysis. Lavage fluids were immediately centrifuged for 15 min at 500 g. Total cells were stained with trypan blue for viability and counted. Differential cell counts were performed on cytospin preparations after staining with Diff-Quick. MAM were cultured in RPMI containing 5% penicillin-streptomycin and 10% FBS. 0.1×106 cells were plated per well in an untreated 6-well plate for overnight incubation at 37 °C and 5% CO2.
Murine BMDM were obtained from femur bone marrow of 4 to 8 weeks old C57BL6 mice. Cells were collected and resuspended in RPMI containing 5% penicillin-streptomycin, 10% FBS and 1/10 volume of CMG 14-12 culture supernatant containing Macrophage Colony-Stimulating Factor (M-CSF). Cells were incubated at 37 °C and 5% CO2 for up to 7 to 8 days. On day 8, cells were taken off the plate using cell dissociation solution (Sigma-Aldrich) and resuspended in the culture medium. One million cells were plated per well in a 6-well plate overnight at 37 °C. For all cell types, culture and cell preparation prior to flow cytometry were performed in phenol-red free media and solutions.
After incubation of the macrophages overnight, the medium was removed, and fresh RPMI-2 supplemented with 50 μg/ml gentamicin was added. Macrophages were pre-incubated with 5 μg/ml of lethal toxin for HAM and 100 or 500 ng/ml for RAW cells for 1 h prior to stimulation with Sterne spores at an MOI of 1 in triplicate wells of a 24-well plate, and allowed to incubate at 37°C for 7.5 h. Gentamicin remained in the medium during the entire incubation time in order to ensure that vegetative bacteria would be killed, and that no extracellular toxin would be produced. Absence of biologically active toxins in gentamicin containing culture media containing spores was confirmed by absence of cleavage of MEK in RAW264.7 cells (not shown). Spore diluent and spore diluent plus 5 μg/ml LT for HAM and 100 ng/ml LT for RAW cells were used as negative controls, and LPS (1 μg/ml) and LPS plus 5 μg/ml or 500 ng/ml LT were used as positive controls.
For intracellular cytokine measurements, 0.5 ml of 0.05% deoxycholate was added to lyse the cells. the mixture was incubated at 37°C for 10 min, and the supernatant was stored at −20°C. Cytokine ELISAs were performed as previously described by using anti-cytokine monoclonal primary capture antibodies and biotinylated anti-cytokine polyclonal secondary detection antibodies (R&D Systems). Plates were developed using TMB reagent (BD Biosciences).
RAW, murine BMDM, and HAM cells were plated on untreated 6 well plates at a concentration of 1 million cells per well. MAM were plated at a concentration of 0.1 × 106 cells per well. After overnight incubation, all cell types were washed once with PBS and 1 ml warm, nonenzymatic cell dissociation solution (Sigma-Aldrich, Cat# C5914-100ML) was added to each well for 15 minutes at 37 °C. The cells were collected, washed twice with room temperature Stain Buffer containing 2% FBS (BD Pharmingen, San Diego, CA) and adjusted to a concentration of 0.1 × 106 cells per 100 μl test volume in Stain Buffer. Single-use FITC labeled PA (List Biological Laboratories, Campbell, CA) was added at concentrations of 1, 10, 100, 500, and 1000 ng/ml. Negative controls were PA diluent and 1000 ng/ml of unlabeled PA. The cells were covered, then incubated at room temperature for 60 minutes, gently rocking. After incubation, cells were pelleted at 300 × g for 5 minutes, washed once with 4 ml cold Stain Buffer, resuspended in 100 ul Stain Buffer, and put on ice. For RAW, MAM, and murine BMDM cells, 2 ul of anti-mouse F4/80-APC conjugate (Pan Macrophage Marker, BM8) or the APC rat IgG2a, κ isotype control (eBiosciences, San Diego, CA) was added. Additional FcR blocking was not required. For HAM cells, 1 ul anti-human CD16-biotin or biotin mouse IgG1, κ isotype control, followed by 0.5 ul PE-Cy5 strepavidin (BD Pharmingen), were used. All incubations were for 15 minutes on ice. After cells were washed once and resuspended in 500μl of Stain Buffer, they were read on a BD FACSCalibur flow cytometer with CellQuest Pro software. Ten thousand events were counted. To identify the populations suitable for analysis, murine cells were gated on F4/80+ events and HAM on CD16+ events, followed in all cases by a scattergate. For every cell type, approximately 86% were members of both gates.
HAM, MAM, murine BMDM and RAW cells were harvested as described above, plated at a density of 1 × 106 cells/ml for HAM, murine BMDM, and RAW cells, and 0.1 × 106 cells/ml for MAM. They were maintained overnight at 37°C in 5% CO2 in 2 ml RPMI-2 or DMEM-10 medium and 50 μg/ml gentamicin. HAM, MAM and murine BMDM were treated with 100, 500 or 5000 ng/ml of LT, and RAW cells were treated with 1, 100, or 500 ng/ml of LT for three hours. Mock-infected negative control cells were exposed to an equivalent volume of LT-free diluent. After incubation at 37°C in 5% CO2 for the indicated time, the macrophages were harvested and lysed in 300 μl cold RIPA lysis buffer (150 mM NaCl, 50 mM Tris, pH 8.0, 10 mM each EDTA, NaF, Na-pyrophosphate, 1% NP-40, 0.5% Na deoxycholate, 0.1% SDS, 3 mM sodium vanadate, 10 μg/ml aprotinin, and 10 μg/ml leupeptin). For ATR protein determination, macrophages were directly harvested and lysed in cold RIPA lysis buffer. Macrophage homogenates were clarified by centrifugation at 4°C, and 20 to 30 μg of the resultant postnuclear lysates were mixed with SDS-polyacrylamide gel electrophoresis (PAGE) sample buffer (60 mM Tris, pH 6.8, 10% glycerol, 2.3% SDS) and heated to 95°C for 5 min. The samples were separated by 10% SDS-PAGE and electrophoretically transferred onto nitrocellulose membranes. To detect MEK's and ATR, the membranes were blocked by 5% powdered milk in Tris-buffered saline and probed overnight with the corresponding antibodies (described above) in 5% powdered milk in Tris-buffered saline. With the exception of CMG-2\ANTXR2, where species-specific antibodies were used, all MEK and ATR antibodies used cross-reacted with both RAW cells and HAM. The membranes were developed with horseradish peroxidase-conjugated specific secondary polyclonal antibodies and chemiluminescent reagents (Pierce Biotechnology, Rockford, IL). The developed membranes were exposed to X-ray film.
HAM, MAM, murine BMDM or RAW cells were collected into complete media and plated into untreated 6 well plates at 1.0 × 106/well for RAW, HAM, and murine BMDM, and at 0.1 × 106/well for MAM. Following overnight incubation at 37° C, the media in each well were aspirated and replaced with 2 ml containing 0, 10, 100, 1000, or 5000 ng LT per ml. The LT concentration consisted of 1:1 addition of single-use PA and LF solutions (List Biological Laboratories), aliquoted at −80° C. For the positive controls, 10 uM/ml staurosporine in DMSO (Life Sciences, Plymouth Meeting, PA) was added to HAM, MAM, and murine BMDM, while 5 uM/ml was added to RAW. These amounts were determined from the staurosporine dose curves for HAM and RAW (data not shown). After 3 hours of treatment at 37° C , media were removed and saved, and cells were harvested by addition of 1 ml warm Trypsin-EDTA Solution (Mediatech, Inc.) per well. None of the cells were scraped from the plastic.
Following a PBS wash, cells from each treatment well were washed then resuspended at 1 × 106/ml in cold 1X Annexin Buffer, according to the protocol provided in the Vybrant Apoptosis Assay Kit #3 (Invitrogen, Cat# V13242). Briefly, 100 μl of cell suspension was added to a Falcon tube, stained at room temperature with 5 μl Annexin V solution for 15 min (1 ul FcR blocker optional), washed twice with Annexin Buffer, then resuspended in 100 ul Annexin Buffer. Mouse cells were then stained with F4/80-APC or isotype, and HAM were stained with biotin CD16/strepavidin PE-Cy5 or isotype for 15 minutes on ice, covered. Tube volume was brought to 500 μl with Annexin Buffer. Samples were run on a BD FACSCalibur flow cytometer using CellQuest Pro software. For those samples being tested for necrosis, 2 μl of Vybrant PI solution was added just before the run. Compensation for spectral overlap in the fluorescence channels were adjusted with samples having cells incubated for 3 hours with staurosporine, then stained with either Annexin V, PI, F4/80-APC or biotin CD16/strepavidin PE-Cy5. RAW cells were also processed using the Vybrant FAM Poly Caspases Assay Kit as a countertest.
Total RNA from RAW and HAM cells was extracted using a modified TRIzol (Invitrogen, Carlsbad, CA) protocol, spectrophometrically quantitated, and the integrity verified by formaldehyde agarose gel electrophoresis.
Equal amounts (2 μg) of RNA from both cell types were used with oligo (dT) as primers for production of cDNA (SuperScript First-Strand Synthesis System for RT-PCR, Invitrogen) to produce cDNA. Gene specific primers for the receptors and the GAPDH housekeeping genes were used in standard PCR on a MJ Research DNA Engine thermal cycler with the following program: 1 cycle of 94°C for 2 min, 48°C for 1min, 72°C for 1min, followed by 32 cycles of 94°C for 30 sec, 48°C for 30 sec, 72°C for 1min, and ending with a 72°C for 7 min extension. Preliminary experiments showed the exponential amplification phase of all the amplicons to be between 24 and 31 cycles. The primers’ sequences are as follow: TEM8\ANTXR1 forward 5’-TGCTGCACCACTGGAATGAAATC-3’; TEM8\ANTXR1 reverse 5’-CTCCTCCTGGCAGAACTTTCTGG-3’; CMG2\ANTXR2 forward 5’-CTTTCATTGTGTTTTCTTCTCAAGCAAC-3’; CMG2\ANTXR2 reverse 5’-TGCATAAGATGGTACCAGGC -3’; mouse GAPDH forward 5’-AACTTTGGCATTGTGGAAGG-3’; mouse GAPDH reverse 5’-ACACATTGGGGGTAGGAACA-3’; Human GAPDH forward 5’-GGAAGGTGAAGGTCGGAGT-3’; Human GAPDH reverse 5’-GAAGATGGTGATGGGATTTC-3’; (20) Following PCR, samples were separated on a 1.5% agarose gel, then stained with ethidium bromide (Invitrogen) for imaging and the band volumes calculated by ImageQuant 5.0 software (Molecular Dynamics). Anthrax toxin receptor band densities from RAW and HAM were normalized to the corresponding GAPDH densities to correct for potential differences in input cDNA.
The current assumption that HAM are susceptible to innate immune suppression by LT is partially based on work using mouse RAW 264.7 macrophages and HAM independently (18). Our first objective was thus to determine the relative sensitivity of these cells by direct comparisons of the effects of LT on RAW 264.7 mouse macrophage cells and HAM.
HAM and RAW cells were exposed to various doses of LT (LF+EF) for 1 h prior to stimulation with B. anthracis spores (MOI=1) and incubated for 7.5h. Spore diluent (mock) and LT (mock + LT) were used as negative controls, and LPS (1μg/ml) and LPS + LT were used as positive controls. The effects of these stimuli and inhibitors on RAW and HAM cytokine release were measured by ELISA of the cell supernatants. Spores induced an innate immune cytokine response in both mouse RAW cells and HAM. Specifically, spores increased TNF-α and MCP-1 release by 35 fold and 9 fold, respectively, in RAW cells, and 19 fold and 3 fold, respectively, in HAM. Cytokine induction was greatly reduced by LT treatment in RAW cells, as 100 ng/ml LT decreased the induction of TNF-α 16 fold, and induction of MCP-1 by 6 fold (Fig. 1A). We did not test LT doses above 100ng/ml as doses of LT greater than this completely lysed RAW 264.7 cells. This sensitivity of RAW to LT immunosuppressive effects is consistent with effects seen in previous studies by others (21). In contrast, HAM were resistant to the immunosuppressive effects of LT. Addition of 5000ng/ml LT (50 times the RAW dose, enough to cause complete lysis of these cells) for the same time period only decreased the TNF-α and MCP-1 response to spores by 9 and 2 fold, respectively, and had a much smaller effect on the TNF response to LPS (Fig. 1B). HAM were also resistant to the immunosuppressive effects of LT (5000ng/ml) on spore induced induction of IL-1α, IL-1β, IL-6 and MIP-1α/β (Fig. 1B). These results demonstrate that, compared to RAW cells, HAM are very resistant to innate immune suppression by B. anthracis toxins, as measured by induction of cytokine release by B. anthracis spores.
We have previously demonstrated that MAPK activation is required for innate immune cytokine induction by B. anthracis spores in HAM. Activation of MAPK requires phosphorylation of the MAPK protein on both tyrosine and threonine (22, 23), a reaction catalyzed by the protein kinase referred to as MAPK kinase (MAPKK) or MEK (MAPK/ERK kinase). Previous work by other groups has shown that B. anthracis LF inhibits MAPK signaling in activated murine macrophages by cleaving the amino-terminal extension of MEKs that activate MAPKs (17, 24, 25). We compared the relative sensitivity of RAW 264.7 cells, murine BMDM, MAM and HAM to LT-mediated MEK cleavage. All cells were exposed to increasing doses of LT (PA+LF) for 3 h, cell lysates harvested, and MEK cleavage analyzed using anti-MEK antibodies specific for intact (non-cleaved) MEK's. The results show that in RAW cells, MEK's 1, 2, 6 and 7 are cleaved by doses of LT at or above 100ng/ml (Fig. 2). In contrast, even at doses that completely lyse RAW cells (5000 ng/ml) only MEK 1 is significantly cleaved in HAM (Fig. 2). Generally, murine BMDM and MAM have similar sensitivity to LT mediated-MEK cleavage as RAW 264.7 cells. In these cells, MEK's 1 and 2 are cleaved by 100ng/ml of LT, the same dose as in RAW 264.7 cells. In murine BMDM, MEK 6 is more resistant to LT cleavage than in RAW cells, but more susceptible than in HAM. MEK 7 in MAM is not cleaved by LT of 5000 ng/ml, as in HAM. The results confirm that HAM are resistant to B. anthracis virulence toxins. MEK's in murine BMDM and MAM are cleaved by LT as they are in RAW cells, but in some cases higher LT doses are required. In terms of the relative resistance of HAM to LT, this is consistent with the resistance of these cells to suppression of innate immune cytokine responses by LT.
MEK cleavage by LT in macrophages has been linked to induction of apoptosis in these cells (17). This finding is consistent with the current paradigm which holds that HAM are sensitive to immunosuppressive effects of LT (18, 26, 27). We measured the relative sensitivity of RAW, murine BMDM, MAM and HAM cells to LT-mediated apoptosis using flow cytometry. HAM and RAW cells were incubated for 3h with increasing doses of LT. Staurosporine (5-10 μM) was used as a positive control for apoptosis for all cell types. Cells exposed to spore and staurosporine diluent were used as negative controls. Apoptosis and necrosis were measured by staining with Annexin V (apoptosis) and propidium iodide (necrosis). The results (Fig. 3, and supplemental Fig. S1) confirmed that RAW cells are very sensitive to LT-mediated apoptosis, as 10 and 100ng/ml of LT induced apoptosis in 86% and 94% of the RAW cells, respectively. Approximately a third (LT10) or a half (LT100) of the apoptotic cells were also necrotic (Fig. 3, Panel A). Five percent of the total cell population had progressed to necrosis, without detectable apoptosis. Doses of LT greater than 100ng/ml completely lysed RAW cells and there were not enough events from intact cells to count. In contrast, at all doses of LT up to and including 5000ng/ml of LT, only 25% of the HAM were apoptotic and of those, only 10% had also become necrotic (Fig. 3, Panel D). Three percent or less of the HAM had progressed to necrosis without detectable apoptosis. There was also no dose response to LT in HAM. For MAM, 100ng/ml of LT only caused 11% of the cells to apoptose, with a third of these also being necrotic (Fig. 3, Panel C); the same dose caused 94% apoptosis for RAW cells. Murine BMDM were initially not as healthy as the other cell types as approximately 54% of untreated (mock) cells were apoptotic. However, LT, even at 500ng/ml., failed to further enhance apoptosis (Fig. 3, Panel B). Thus, MAM and murine BMDM were also resistant to LT-mediated apoptosis, similarly to HAM.
These results show that alveolar macrophages are resistant to LT-mediated apoptosis. Resistance of HAM to MEK cleavage may play a part in resistance of these cells to LT-mediated apoptosis. This can not completely account for this phenomenon, as MAM and murine BMDM are also resistant to LT-induced apoptosis, despite the fact the LT causes MEK cleavage in these cells.
We next sought to determine whether there could be any relationship between LT-mediated MEK cleavage and induction of apoptosis, and binding of the B. anthracis toxin component PA to our cells. PA is the cellular binding component of the A/B B. anthracis toxins LT and ET. We measured binding of PA to RAW cells, murine BMDM, MAM and HAM by flow cytometry using FITC-conjugated PA. All the cells were either exposed to various concentrations of FITC-PA or unlabeled PA, or the cells were exposed to PA diluent (negative control). The results show that there was minimal background fluorescence in the presence of PA diluent or 1000ng/ml unlabeled PA in all of the cells tested (Fig. 4, and supplemental Fig. S2). There was significant binding of FITC-PA to RAW, murine BMDM, and MAM cells that increased with increasing doses of PA, at all doses tested up to 1000 ng/ml PA (Fig. 4). In contrast, there was no significant binding of PA to HAM, even when 1000ng/ml FITC-PA was used (Fig. 4). These results show that HAM do not bind significant amounts of PA. The results also show that there is binding of PA to the other three cell types. These three cell types were also relatively sensitive to LT-mediated MEK cleavage. This suggests a mechanism for the resistance of HAM to LT-mediated MEK cleavage due to a lack of binding sites for the binding partner of LT, PA. The amount of PA binding did not directly correlate with sensitivity of the cell types to LT-mediated apoptosis.
CMG2\ANTXR2 and TEM8\ANTXR1 have been identified as the anthrax toxin receptors. Previous reports have suggested that ATR is expressed in HAM, but this has only been tested at the level of mRNA expression. We next sought to confirm whether ATR was expressed in HAM both at the mRNA level and protein level, as decreased ATR expression could contribute to the relative resistance of these cells to the immunosuppressive effects of LT, as well as the lack of binding of PA to the cells.
Endogenous mRNA levels of TEM8\ANTXR1 and CMG2\ANTXR2 were determined using relative end-point RT-PCR in HAM and RAW cells. RNA expression of both receptors was detected in both cell types. TEM8\ANTXR1 expression levels were similar in RAW and HAM cells at 9 and 8 percent of GAPDH levels, respectively (Fig. 5). CMG2\ANTXR2 expression was greater in RAW than HAM cells at 21 and 13 percent of GAPDH levels, respectively (Fig. 5).
We then attempted to measure expression of TEM8\ANTXR1 and CMG2\ANTXR2 receptor proteins in HAM, MAM, murine BMDM, and RAW cells by immunoblotting of cell extracts using anti-TEM8\ANTXR1 and anti-CMG2\ANTXR2 antibodies. Peptide and fusion protein standards were used to quantify the amounts of TEM8\ANTXR1 and CMG2\ANTXR2. Antibody to the αM component of the MAC1 (CD11b) receptor was used as a control for detection of cellular receptor expression in the two cell types studied, and GAPDH was used as a loading standard. The results (Fig. 6) show that there was minimal expression (<50ng/mg lysate protein) of TEM8\ANTXR1 in all of the cell types tested. For CMG2\ANTXR2 in HAM, we did not detect expression, though whole human lung tissue extracts, which would include capillaries, contained this protein (not shown). We were unable to confidently assess the level of CMG2 protein expression in any mouse tissue or cell type using the commercially available antibodies.
These results demonstrate that, in contrast to the mRNA levels detected by our laboratory and others, there is minimal expression of TEM8\ANTXR1, and no expression of CMG2\ANTXR2 in HAM cells. These results are consistent with our finding that there is minimal binding of PA to HAM. Also, the findings support a mechanism whereby a lack of expression of LT receptors in HAM may contribute to the resistance of these cells to the immunosuppressive effects of this virulence factor. The differences in binding of PA between HAM and the murine cell types may be due to differential protein expression of CMG2\ANTXR2 in these cells, although we were unable to assess this.
Anthrax toxins are thought to play a role in the pathogenesis of inhalational anthrax, but the exact role and cellular target of these toxins is unclear. It has been postulated that the target of the toxins is the alveolar macrophage and that innate immune suppression by toxins is the mechanism of virulence. In a sense, this hypothesis is logical because the innate immune response likely assists with early containment of the pathogen, as there is a significant delay between exposure to the pathogen and disease. Also, the alveolar macrophage is the main phagocytic cell of the lung and human and macaque AM as well as murine BMDM rapidly internalize the pathogen and kill spores after internalization (19, 28, 29). Furthermore, expression of the known receptors for LT, TEM8\ANTXR1 and CMG2\ANTXR2, was identified at the transcriptional level in cells from human bronchiolar lavage fluid, which contains primarily macrophages (18, 30, 31).
Additional support for the hypothesis that the alveolar macrophage is the cellular target of LT has been adduced from studies using other macrophage types. For example, LT induces apoptosis and necrosis in mouse macrophage cell lines and bone marrow derived macrophages (17, 27). LT has also been shown to induce apoptosis in activated human peripheral blood monocytes (32). LT also has other immunosuppressive effects. High doses of LT (1ug/ml) blunts LPS-mediated induction of TNFα, IL-1β, IL-6 and IL-8 by LPS in non-human primate alveolar macrophages (28). However, sensitivity of AM to B. anthracis toxins would not explain the finding in animal models, and in human autopsy studies, that the alveolar space in alveoli that are not flooded are cleared of the pathogen (5-7). In humans, even in end stage disease, the gradient of bacterial concentration is from the vascular space to the alveolar spaces (6). This suggests that the alveolar space is so hostile to B. anthracis, perhaps because the alveolar macrophage efficiently kills the pathogen (28, 29), that it can only be infected in a retrograde fashion, from the vascular space, during the terminal phase of disease when high grade bacteremia is present.
One explanation for these findings would be, in contrast to the current paradigm, that alveolar macrophages are, in fact, relatively resistant to the effects of B. anthracis toxins and can thus clear the pathogen from the alveolar space. Our findings show that this is indeed the case. By comparison to a commonly used surrogate for ex vivo murine macrophages, the RAW macrophage cell line, MAM, HAM, as well as murine BMDM, are resistant to some, or all, of the effects of B. anthracis lethal toxin. Of these cells HAM showed considerable resistant to all of the effects of LT, including inhibition of spore-mediated cytokine induction, LT-mediated MEK cleavage, and LT-mediated apoptosis. MAM and murine BMDM, while showing intermediate sensitivity to LT-mediated MEK cleavage, showed comparable resistance to LT-mediated apoptosis as did HAM.
Our findings demonstrate that the mechanism of HAM resistance relative to that of RAW cells is likely at least partly due to the relative lack of expression of functional anthrax toxin receptors in these cells. There was no significant binding of PA to HAM. Although there was minimal expression of TEM8\ANTXR1 on both cell types, there was no significant expression of CMG2\ANTXR2 on HAM. We were unable to assess CMG2\ANTXR2 protein expression in RAW cells. However, there was clearly a difference in functional PA binding between RAW and HAM cells. This difference is functionally significant, as downstream effects of the toxin, namely cytokine inhibition, MEK cleavage, and induction of apoptosis were dramatically decreased in HAM as compared to RAW cells. We do not think the difference in our assessment of the level of expression of PA binding sites in HAM is due to differences in our handling or processing of the cells used because we demonstrate similar expression of the known ATR at the level of transcription as others (18). Furthermore, it has been demonstrated that MEK3 is not cleaved during infection of HAM by B. anthracis, which is consistent with a lack of functional ATR in these cells (33). Thus, it is more likely that the ATR mRNA is either not translated, or that the expressed protein is rapidly degraded, and thus not detected in HAM.
It should be noted, however, that differential expression of functional PA binding sites, and even differences in susceptibility to LT-mediated MEK cleavage, do not fully account for the differences in the biological effects of LT in the various cells we tested. PA binding, and MEK cleavage was similar in RAW, murine BMDM, and MAM cells. However, murine BMDM and MAM were relatively resistant to LT-mediated induction of apoptosis, while RAW cells were sensitive to this effect. This along with the findings of others, suggests that MEK cleavage is necessary and sufficient for induction of apoptosis in immature macrophages (17), but MEK cleavage is not sufficient for apoptosis in mature alveolar macrophages (28). We suspect that this is due to the antiapoptotic machinery of the mature cell types (34-36).
Our results comparing the sensitivity of these cell types to LT may have other implications. Recently overexpression of secretory phospholipase A2 (sPLA2) has been shown to protect against B. anthracis infection in a mouse model (37). We demonstrate that RAW 264.7 cells, which are derived from the sPLA2-expressing Balb/C mouse strain, are more sensitive to LT than MAM from C57Bl/6 mice, which do not express sPLA2. This suggests the mechanism of sPLA2 activity against B. anthracis does not include enhancing macrophage resistance to LT.
The first conclusion that can be drawn from our findings is that the mature, resident alveolar macrophage is not a significant target of the anthrax lethal toxin in mouse and man. Although our study did not measure sensitivity of MAM or HAM to ET, the lack of functional PA binding in HAM makes it unlikely that these cells are a target for ET either, though sensitivity at high doses has been shown in macrophages from other species, namely guinea pigs (38). However, the fact remains that toxins are important in the pathogenesis of the disease as development of antibodies to PA after vaccination is predictive of protection (16, 39), and externally introduced PA antibodies are protective in rabbit models (15). Another possibility is that B. anthracis toxins are not important in the early phase of inhalational anthrax, but that is unlikely as antibodies to PA delay dissemination of the pathogen (15). More likely, the target of the toxin is another cell type in the lung, and we are investigating that possibility. That cell type will, of course, express a significant amount of the ATR.
Our findings also suggest that the alveolar macrophage is an unlikely candidate for the role of carrier of the pathogen to the TLN. That is, if, as we suggest, the alveolar macrophage is effective at clearing the pathogen from the alveolar space due to toxin resistance, then spore bearing macrophages would still be able to dispose of the pathogen while they are in transit to the TLN. Our results thus are consistent with the work of Cleret and colleagues who demonstrated in a mouse model that DC are likely the major transporter of B. anthracis spores to the thoracic lymph nodes (5), and the work of Brittingham and colleagues who demonstrated that B. anthracis may exploit human DC to facilitate infection (40). Mouse DC have also been shown to be vulnerable to B. anthracis LT (41, 42). Sensitivity of DC to LT may protect the pathogen from destruction during transit to the regional lymph nodes.
How then, does the pathogen overcome the sentinel cell of the lung, the alveolar macrophage? We suspect that the pathogen does not so much as overcome the alveolar macrophage as much as it escapes it by bypassing this cell, by weight of numbers. Although the exact number of spores required to cause an infection in humans is unknown, it is estimated that occupational exposure in goat hair mills is approximately 510 particles per 8 hour shift, and yet disease was uncommon in this population, even before vaccination was required (43, 44).
If it is assumed that anthrax victims of the 2001 postal bioterrorism incident were infected by a low number of spores, this might seem inconsistent with our conclusions. However, since the envelopes contained over 100 billion spores per gram of powder, a very small amount of powder could have delivered a huge dose of spores (45).
Taken together, the results here demonstrate that HAM are resistant to innate immune suppression by B. anthracis toxins, and may explain why B. anthracis is cleared from the alveolar space in human autopsy specimens. Resistance of the AM to the B. anthracis toxins may also determine the minimal infectious dose of spores that is required to cause disease in man.
We acknowledge the kind assistance of the nursing and bronchoscopy staff of the Veterans Administration Hospital of Oklahoma City, OK. We also acknowledge the assistance and expertise of the Oklahoma Medical Research Foundation Flow Cytometry core facility.
1The research described in this work was partially supported by the National Institute of Allergy and Infectious Diseases, project 1U19 AI62629 (to J.P.M., J.D.B. and K. M.C.), and by a Clinical Innovator Award from the Flight Attendant Medical Research Institute (to W. Wu).
The authors have no financial conflicts of interest.