ABSTRACT
Intranasal vaccination stimulates formation of cyclooxygenases (COX) and release of prostaglandin E2 (PGE2) by lung cells, including alveolar macrophages. PGE2 plays complex pro- or anti-inflammatory roles in facilitating mucosal immune responses, but the relative contributions of COX-1 and COX-2 remain unclear. Previously, we found that Mycobacterium bovis BCG, a human tuberculosis vaccine, stimulated increased release of PGE2 by macrophages activated in vitro; in contrast, intranasal BCG activated no PGE2 release in the lungs, because COX-1 and COX-2 in alveolar macrophages were subcellularly dissociated from the nuclear envelope (NE) and catalytically inactive. This study tested the hypothesis that intranasal administration of BCG with cholera toxin (CT), a mucosal vaccine component, would shift the inactive, NE-dissociated COX-1/COX-2 to active, NE-associated forms. The results showed increased PGE2 release in the lungs and NE-associated COX-2 in the majority of COX-2+ macrophages. These COX-2+ macrophages were the primary source of PGE2 release in the lungs, since there was only slight enhancement of NE-associated COX-1 and there was no change in COX-1/COX-2 levels in alveolar epithelial cells following treatment with CT and/or BCG. To further understand the effect of CT, we investigated the timing of BCG versus CT administration for in vivo and in vitro macrophage activations. When CT followed BCG treatment, macrophages in vitro had elevated COX-2-mediated PGE2 release, but macrophages in vivo exhibited less activation of NE-associated COX-2. Our results indicate that inclusion of CT in the intranasal BCG vaccination enhances COX-2-mediated PGE2 release by alveolar macrophages and further suggest that the effect of CT in vivo is mediated by other lung cells.
INTRODUCTION
The development of effective mucosal vaccines has been hindered by the lack of useful adjuvants and by our limited knowledge of their modes of action (1). The cyclooxygenase (COX) product prostaglandin E2 (PGE2), pharmacologically targeted by nonsteroidal anti-inflammatory drugs, is commonly considered a potent proinflammatory mediator; PGE2 shifts T cells to Th2, Th17, and regulatory T cell responses and shifts macrophages (Mϕ) to alternatively activated Mϕ (M2) in autoimmune diseases, cancer, and other chronic inflammatory diseases (2). In contrast, PGE2 in the lungs has complex pro- and anti-inflammatory roles modulating not only the immune-inflammatory responses, but also mucosal protection from inflammatory injuries and tissue repair processes (3–5). More specifically, lung PGE2 is reported to play paradoxical roles, including upregulation of apoptotic death of bactericidal Mϕ, thereby inhibiting replication of intracellular Mycobacterium tuberculosis (6), and inhibition of Th2 differentiation and allergic inflammation (5, 7). This complexity is at least in part explained by multiple pulmonary mucosal and inflammatory cells differentially expressing four distinct PGE2 receptors, termed E-prostanoids 1 to 4, which are pharmacologically targeted for chronic inflammatory diseases (5, 8–11). However, despite recognition of the indispensable roles of constitutive COX-1 and inducible COX-2 for PGE2-mediated mucosal inflammation, there is still insufficient information regarding the activities of these two rate-limiting enzymes for PGE2 release during the course of mucosal vaccination.
Because it promotes local immune responses in the lung, intranasal (i.n.) vaccination of mice with Mycobacterium bovis bacillus Calmette-Guérin (BCG) offered better protection against M. tuberculosis than did systemic vaccination (12). Our previous studies (13, 14) indicate that alveolar Mϕ, activated by i.n. heat-killed BCG, develop bactericidal Mϕ (M1) that facilitate Th1 immunity. However, in these M1 Mϕ, both COX-1 and COX-2 are dissociated from the nuclear envelop (NE), accumulate in aggregates in the endoplasmic reticulum (ER), and are catalytically inactive. Although PGE synthase, which converts PGH2 to PGE2, appears to be active (15), these COX-2+ Mϕ release no PGE2 (13). Furthermore, the impairment of PGE2 release seems to be independent of degradation of PGE2 driven by 15-hydroxyprostaglandin dehydrogenase (16).
Intranasal Mycobacterium tuberculosis, Mycobacterium butyricum, and Propionibacterium acnes, like BCG, also induce M1 activation of alveolar Mϕ that is characterized by the inactive forms of COX-2 and COX-1 (13). In contrast, i.n. administration of heat-killed Escherichia coli or lipopolysaccharide (LPS) induces NE-associated COX-2, which can be inactivated by subsequent in vivo stimulation by BCG, suggesting that selected microbes regulate whether alveolar Mϕ express active or inactive COX-2 (13). Mycobacterial inactivation of COX will limit PGE2-mediated mucosal protection and immune responses. Therefore, it would be important to understand the mechanism(s) regulating inactive/NE-dissociated COX formation in response to mucosal mycobacterial vaccination in vivo and NE-associated, active COX-2 in murine alveolar Mϕ (or other tissue Mϕ) cultured with BCG in vitro (13, 17, 18).
It is known that the protection against pulmonary tuberculosis is significantly enhanced when recombinant BCG producing cholera toxin (CT) or its subunits or a mixture of BCG and CT (BCG/CT) is used as an intranasal vaccine (19–21). CT is the enterotoxin of Vibrio cholerae and a mucosal Th2 adjuvant in animal models (22–25). CT holotoxin consists of a receptor-binding homopentameric B subunit (CTB) that is noncovalently associated with a single catalytic A subunit (CTA) that modifies a G-protein associated with adenylate cyclase, thus stimulating cyclic AMP (cAMP) production (26–28). CT is known to enhance mucosal COX-2 expression and PGE2 synthesis in intestines in vivo, at least in part regulating water and electrolyte transport in the intestine without noticeable influx of leukocytes or tissue damage (29–31). We therefore hypothesized that intranasal CT alone would induce alveolar Mϕ expressing NE-associated COX-2 and elevated PGE2 release and, furthermore, that CT would prevent the induction of inactive, NE-dissociated COX-2 in alveolar Mϕ observed following intranasal vaccination of BCG alone. Based on the enhanced COX-2 mRNA expression and PGE2 release observed by Hinz et al. (32) following treatment of human monocytes in vitro with CT and LPS, we hypothesized that BCG with CT in vitro would also enhance active COX-2 expression and PGE2 release.
MATERIALS AND METHODS
Mice.Nonpregnant female C57BL/6 mice, 8 to 10 weeks old, were obtained from Jackson Laboratory (Bar Harbor, ME), maintained in barrier-filtered cages, and fed Purina laboratory chow and tap water ad libitum. The experimental protocols employed in this study were approved by the IACUC at Florida Atlantic University.
Administration of BCG and CT.The M. bovis BCG Tokyo 172 strain (Japan BCG Laboratory, Tokyo, Japan) was prepared as described previously (33). CT and CTB were purchased from List Biological Lab, Campbell, CA. Groups of mice (6 mice/group) were given 50 μl of saline containing 500 μg of BCG and 1 μg of CT, 500 μg of BCG, 1 μg of CT, or saline (controls) intranasally at 0 h, and samples were collected at 24 h. The intranasal doses of BCG and CT were based on previous studies (14, 34).
Alveolar Mϕ preparation.Lungs were perfused through the right ventricle and pulmonary artery with 10 ml of 37°C 30 mM EDTA in Hanks balanced salt solution (HBSS). The trachea was cannulated, and bronchoalveolar lavage (BAL) fluid with 1 ml saline was recovered immediately as a source of alveolar Mϕ (13). Total cell counts were determined with a Coulter counter (model Z1; Beckman Coulter, Hialeah, FL). To enrich plastic-adherent Mϕ, cells in bronchoalveolar lavage (BAL) fluid at 106 cells/ml in RPMI 1640 containing 5% heat-inactivated fetal bovine serum (FBS), 100 U/ml penicillin, 100 μg/ml streptomycin, and 0.25 μg/ml amphotericin B were incubated in culture dishes (Falcon, Oxnard, CA) for 1 h at 5% CO2 and 37°C in a 100% humidified incubator. Nonadherent cells were removed by washing with media. BAL Mϕ and other inflammatory cells were enumerated by differential counts following cytospin preparations and staining with Diff-Quik (Baxter Healthcare, Miami, FL). Adherent BAL cell differentials from all animals used in experiments revealed >90% Mϕ.
Alveolar epithelial cell preparation.Immediately after collection of BAL specimens, pulmonary alveolar epithelial cells (EC) were isolated as previously described with modifications (35). RPMI 1640 plus 2% FBS and 1 mg/ml dispase (Invitrogen, Grand Island, NY) (1 ml) were perfused into the airways and alveoli. The lungs were removed and incubated at 37°C for 1 h in RPMI 1640 containing 2% FBS, 1 mg/ml dispase, and 2 mg/ml collagenase type II (Invitrogen), followed by mincing. Cells were filtered through Teflon mesh with 40-μm pores and washed with the medium with no centrifugation at 0°C. To remove Mϕ, cells were incubated on plastic dishes. After 1 h, nonadherent EC were resuspended in RPMI 1640 plus 2% FBS and removed. The percentage of CD32+ EC was >90% (data not shown). COX-1, COX-2, and peroxisome proliferator-activated receptor gamma (PPARγ) levels were characterized by Western blotting and immunohistochemistry using specific antibodies (Cayman Chemical, Santa Cruz, CA).
RAW 264.7 cells.RAW 264.7 Mϕ have been used to model murine tissue Mϕ, including alveolar Mϕ, for study of M1 activation and COX-2 and PGE2 expression (36, 37). As described previously (36), Mϕ were incubated in RPMI 1640 plus 5% FBS at 37°C for 6 h to 24 h with BCG and/or CT at selected doses.
Western blotting.Protein concentrations in the lysates of Mϕ and EC were measured with a bicinchoninic acid assay (Pierce, Rockford, IL). Following SDS-PAGE and transfer to polyvinylidene difluoride (PVDF) membrane, the membrane was blocked with 7.5% nonfat dry milk and incubated with antibodies (anti-COX-1, 1:1,000 [Cayman Chemical]; anti-COX-2, 1:200 [Novus Biologicals, Littleton, CO]; anti-PPARγ, 1:200 [Santa Cruz Biotechnology, Santa Cruz, CA]; anti-β-actin, 1:5,000 [Sigma-Aldrich, St. Louis, MO]) in 7.5% nonfat dry milk overnight at 4°C. Following incubation with peroxidase-conjugated secondary antibody (1:10,000 [Jackson ImmunoResearch, West Grove, PA]), proteins were detected by chemiluminescence (ECL Plus; Amersham, Piscataway, NJ) (13).
Subcellular COX-1 and COX-2 by confocal microscopy analysis.Alveolar Mϕ were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS), pH 7.5, for 20 min, permeabilized with 0.1% Triton X-100 in PBS for 7 min, and incubated in blocking buffer consisting of PBS with 10% FBS overnight at 4°C prior to incubation with anti-COX-2 (1:500) or anti-COX-1 (1:500) antibody overnight at 4°C. Subsequently, cells were washed with PBS and incubated with Alexa Fluor 488 goat anti-rabbit IgG (1:500 [Invitrogen]) for COX-1 and COX-2 for 1 h at 22°C. Nuclei and BCG were stained with propidium iodide (PI) (Sigma-Aldrich) at 10 μg/ml (18). Cells were examined with a laser scanning confocal microscope (LSM700; Zeiss, Thornwood, NY). Images were processed with ZEN software. For each mouse, at least 100 COX-2+ alveolar Mϕ were counted and scored as having NE-associated or NE-dissociated enzyme. Data are reported as the mean percentage of COX-2+ Mϕ that had NE-associated COX-2.
ELISA.Levels of PGE2 from the cell cultures and from the BAL specimens were measured by enzyme-linked immunosorbent assay (ELISA), as previously described (13).
Statistical analysis.Results are expressed as means ± standard errors of the means (SEM). P < 0.05 is considered statistically significant. All statistical analyses were performed with StatView version 5.0 (SAS Institute, Inc., Cary, NC).
RESULTS
Coadministration of BCG and CT induces active COX-2 in alveolar Mϕ.Murine alveolar Mϕ were isolated 24 h after i.n. administration of a mixture of BCG and CT (BCG/CT), CT alone, BCG alone, or saline. As shown in Fig. 1A, CT alone caused a slight increase in Mϕ COX-1 and COX-2; BCG alone caused a greater increase in levels of both enzymes. BCG/CT induced COX-1 and COX-2 at levels comparable to those induced by BCG alone (Fig. 1B).
COX-1 and COX-2 expression by alveolar Mϕ isolated from mice given i.n. BCG and/or CT. Mice were given 500 μg of BCG and/or 1 μg of CT intranasally. After 24 h, alveolar Mϕ were isolated from groups of mice. COX-1, COX-2, and PPARγ levels were determined by Western blotting with 5 μg protein; β-actin bands show equivalent loading of samples. (A) Representative Western blot. (B) The ratio of each molecule to β-actin obtained by densitometric analysis is expressed as arbitrary units (A.U.). Values are means ± SEM (n = 3). *, P < 0.05 for comparisons between control saline and BCG groups. ns, not significant for comparisons between the BCG and BCT/CT groups.
PPARγ, consisting of the 53-kDa PPARγ1 subunit and the 57-kDa PPARγ2 subunit, is constitutively expressed by alveolar Mϕ but not alveolar epithelial cells; our previous study (14) indicated that the major PPARγ bands were undetectable in the BCG-treated cells. While clearly visualized in the BCG/CT-treated cells, levels were still less than 15% of those seen in the cells treated with CT alone (Fig. 1B).
Confocal microscopy revealed that COX-1 and COX-2 in alveolar Mϕ isolated from mice given i.n. CT alone were not clearly stained; less than 5% of alveolar Mϕ exhibited detectable COX-2, which was NE associated (data not shown). As shown previously (13, 17, 18), i.n. BCG induced COX-1 and COX-2 in more than 95% of alveolar Mϕ and the enzymes were localized in densely stained cytoplasmic structures dissociated from the NE. Following i.n. BCG/CT, COX-1 remained in the NE-dissociated form in >90% of COX-1+ Mϕ (Fig. 2). In sharp contrast, over 80% of the COX-2+ Mϕ in the BCG/CT group exhibited NE-associated COX-2 (Fig. 2). Since BCG/CT did not increase the number of COX-2+ Mϕ (data not shown) or total COX-2 expression (Fig. 1) compared to those treated with BCG alone, our results suggest that CT may posttranslationally modulate the NE association of COX-2. Furthermore, CTB was not an effector substitute for CT (data not shown).
Subcellular localization of COX-1 and COX-2 in alveolar Mϕ by i.n. BCG and/or CT. (A) Mice were treated as indicated in Fig. 1. Alveolar Mϕ were examined by confocal microscopy following immunofluorescent staining with anti-COX-1 (green) or anti-COX-2 (green) and PI (red) for the nucleus and BCG. Bars, 5 μm. Results are representative of three separate experiments. (B) The percentages of COX-1+ Mϕ and COX-2+ Mϕ that expressed NE-associated (white bars) and NE-dissociated (black bars) forms were calculated. Values are means ± SEM (n = 3). ns, not significant. *, P < 0.001, for comparisons of NE association between the BCG and BCG/CT groups.
PGE2 levels in BAL (bronchoalveolar lavage) specimens were almost unchanged by the treatments with either BCG alone or CT alone in comparison to those in the saline control (Fig. 3). Relatively high levels of PGE2 in BAL were detected following BCG/CT treatment, an observation consistent with the observation of NE-associated COX-2 in alveolar Mϕ from mice treated with BCG/CT. Our data indicate that coadministration of BCG and CT enhanced mucosal PGE2 production.
PGE2 levels in BAL specimens. Mice were treated as indicated in Fig. 1. After 24 h, levels of PGE2 in BAL specimens were measured by ELISA. Values are means ± SEM (n = 3). *, P < 0.05 for comparisons between the BCG and BCG/CT groups.
Beubler et al. (29) demonstrate that intestinal CT induces PGE2 synthesis via posttranslational activation of COX-2 in the rat jejunum. In contrast, we found that CT alone induced no increase in mucosal PGE2 level, although the level of COX-2 in alveolar Mϕ was slightly increased. Therefore, we conducted additional experiments to determine whether CT enhancement of COX-2 occurred predominantly in alveolar epithelial cells. In support of previous findings by Lama et al. (38), we found that normal alveolar epithelial cells exhibited COX-1 and COX-2 (Fig. 4). Furthermore, neither the levels of COX-1 and COX-2 (Fig. 4) nor their NE association (data not shown) was changed by intranasal CT, BCG, or BCG/CT. Taken together, our results suggest that, unlike intestinal CT, intranasal CT does not enhance COX-2-mediated PGE2 release in alveolar Mϕ or epithelial cells, further implying that COX-2 in intestinal Mϕ and/or epithelial cells was enhanced by the mixture of CT and enteric bacteria.
COX-1 and COX-2 expression by alveolar epithelial cells. Shown is a representative Western blot showing protein levels in alveolar epithelial cells isolated 24 h after mice were given 500 μg of BCG and/or 1 μg of CT intranasally. COX-1 and COX-2 levels were determined by Western blotting with 5 μg protein, as indicated in Materials and Methods; the β-actin bands show equivalent loading of samples. Unlike alveolar Mϕ, PPARγ was not detected in all 4 groups (data not shown).
Differential effect of CT on COX-2 expression and PGE2 release in RAW 264.7 cells in vitro.When various sources of Mϕ, including alveolar, peritoneal, and splenic Mϕ isolated from normal mice as well as the RAW 264.7 Mϕ-like cell line, are activated by BCG in vitro, they express only NE-associated and catalytically active COX-2, mediating increased PGE2 release (13, 17, 18, 36). To determine the effect of CT on increased BCG-induced COX-2 expression in Mϕ in vitro, we used RAW 264.7 cells. We found that CT alone had no effect on COX-2 expression or PGE2 production. Furthermore, induction of active COX-2 expression by BCG was unaffected by stimulation with a mixture of BCG and CT (BCG/CT paradigm) (data not shown).
Intriguingly, the effects of CT are dependent on the timing of the treatments; different schedules produced opposite effects on COX-2 expression and PGE2 release. As shown in Fig. 5, stimulation of RAW 264.7 cells with CT followed by BCG (CT-BCG paradigm) resulted in lower levels of both COX-2 (Fig. 5A and B) and PGE2 release (Fig. 5C) than were found in cells stimulated by BCG alone. On the other hand, when cells were initially stimulated with BCG, followed by CT (BCG-CT paradigm), COX-2 expression (Fig. 5A and B) and PGE2 release (Fig. 5C) were increased over the levels induced by BCG alone. RAW 264.7 cells expressed no detectable level of COX-1 with or without the treatments listed in this study (data not shown). These results suggested that differential effects of CT on BCG-induced COX-2 expression depend on its presence during a critical time in the production of COX-2. Furthermore, CTB did not enhance COX-2 in the BCG-CTB paradigm, suggesting that the holotoxin is required for the enhancement of COX-2 in vitro (Fig. 6).
Effects of post- or pretreatment with CT on COX-2 expression and PGE2 release in BCG-treated RAW 264.7 cells. For CT pretreatment (CT-BCG), RAW 264.7 cells were incubated with CT at the indicated doses, and 18 h after CT treatment, BCG was added at the indicated doses. For CT posttreatment (BCG-CT), RAW 264.7 cells were incubated with BCG, and 18 h after BCG treatment, CT was added. Controls received saline. Cells were harvested at 24 h. Results are representative of three separate experiments. (A) Levels of COX-2 following CT-BCG and BCG-CT treatments, respectively, were measured by Western blotting after loading of 1.5 μg protein. (B) The ratio of COX-2 to β-actin obtained by densitometric analysis is expressed as arbitrary units (A.U.). Values are means ± SEM of three separate experiments. *, P < 0.005, and **, P < 0.0001, compared with the saline group; †, P < 0.05, and ††, P < 0.0001, compared with Mϕ treated with BCG alone. (C) To measure PGE2 release, adherent RAW 264.7 cells were resuspended in serum-free RPMI 1640 and further cultured for 2 h. PGE2 levels in the culture fluids were measured by ELISA. Values are means ± SEM (n = 3). *, P < 0.01 for comparisons between the BCG groups and the corresponding groups containing CT and BCG.
Effects of CTB on COX-2 expression in BCG-treated RAW 264.7 cells. For CT pretreatment (CTB-BCG), RAW 264.7 cells were incubated with CTB or CT at the indicated doses (μg/ml), and 18 h after the treatment, 5 μg/ml BCG was added. For CTB posttreatment (BCG-CTB), RAW 264.7 cells were incubated with BCG, and 18 h later, CTB or CT was added. Controls received saline. Cells were harvested at 24 h. (A) Levels of COX-2 following CTB-BCG and BCG-CTB treatments, respectively, were measured by Western blotting with 5 μg protein; β-actin bands show equivalent loading of samples. (B) The ratio of COX-2 to β-actin obtained by densitometric analysis is expressed as arbitrary units (A.U.). Values are means ± SEM (n = 3). *, P < 0.0001 compared with the saline group; †, P < 0.0001 compared with Mϕ treated with BCG alone.
Intranasal CT-BCG and BCG-CT treatments and NE association of COX-2.To determine whether the differential schedules of CT treatments contribute to better recovery of active COX-2 in vivo, intranasal CT and BCG were given separately in schedules mimicking in vitro treatments (CT-BCG and BCG-CT). COX-1 and COX-2 levels in alveolar Mϕ from mice given BCG alone at 6 h measured by Western blotting were comparable to those at 24 h (data not shown). Furthermore, COX-1/COX-2 levels in the sequential CT-BCG or BCG-CT paradigms were comparable to those given the simultaneous BCG/CT paradigm (data not shown). As shown in Fig. 7A, confocal analysis revealed NE-associated COX-2 in around 80% of COX-2+ Mϕ from mice treated with the CT-BCG or BCG/CT paradigm. In contrast, in the BCG-CT paradigm, only 30% of Mϕ had NE-associated COX-2; less than 5% of cells had NE-associated COX-2 in the mice treated with BCG alone (Fig. 7A). Mϕ from the CT-BCG group released significantly more PGE2 ex vivo than either the BCG/CT or BCG-CT group (Fig. 7B). Interestingly, less than 10% of the COX-1+ Mϕ in any groups had NE association (data not shown). Therefore, CT in vivo did not alter the NE-dissociated form of COX-1 nor contribute to increased COX-1-mediated PGE2 release in the schedules we examined. Finally, our results indicated that increased COX-2 expression seen in vitro with the BCG-CT paradigm was not optimal for NE-associated COX-2 formation in vivo.
Subcellular localization of COX-2 in alveolar Mϕ by i.n. CT-BCG and BCG-CT. Mice were given i.n. 500 μg of BCG and/or 1 μg of CT on the following schedules: CT at 0 h followed by BCG at 18 h (CT-BCG), a mixture of BCG and CT at 0 h (BCG/CT), and BCG at 0 h followed by CT at 18 h (BCG-CT). BCG alone and CT alone were given at 18 h. At 24 h, alveolar Mϕ were isolated in all groups. (A) Alveolar Mϕ were examined by confocal microscopy following immunofluorescent staining with anti-COX-2 and PI for the nucleus and BCG. The percentages of COX-2+ Mϕ that expressed NE-associated (white bars) and NE-dissociated (black bars) forms were calculated. Values are means ± SEM (n = 3). *, P < 0.05, and **, P < 0.001, compared with NE association of BCG alone. (B) Alveolar Mϕ were cultured in serum-free RPMI 1640 for 2 h. PGE2 in the supernatant was measured by ELISA. Values are means ± SEM (n = 3). *, P < 0.05, and **, P < 0.01, compared with BCG alone.
DISCUSSION
Our previous study (13) indicated that intranasal administration of BCG activated alveolar Mϕ; however, these Mϕ persistently expressed inactive forms of COX-1 and COX-2, both of which were dissociated from the NE and therefore released no PGE2. The present study clearly demonstrates that although intranasal CT alone does not increase mucosal PGE2 levels in the lungs, coadministration of intranasal BCG and CT promotes a shift from NE dissociation to NE association of COX-2 in alveolar Mϕ and a significant increase in mucosal release of PGE2 compared to administration of BCG alone.
In our preliminary studies prior to the present one, we used selected mouse models to determine if potential factors producing selective in vivo or in vitro Mϕ activation would contribute to the NE localization of COX-2. These included (i) mice deficient in methionine sulfoxide reductase A, an enzyme protecting hosts from oxidation, (ii) mice deficient in Flt-1 tyrosine kinase, an enzyme regulating monocyte infiltration, (iii) mice deficient in CRTH2 (chemoattractant receptor-homologous molecule expressed on Th2 cells), a receptor of PGD2 derived from activated mast cells, and (iv) mice that were treated with GW9662, a PPARγ antagonist. Our published (14) and unpublished (data not shown) results, however, showed that alveolar Mϕ in these animals still exhibited NE-dissociated COX-2 when activated by i.n. BCG and NE-associated COX-2 when activated in vitro. Therefore, the present study demonstrates for the first time that enzyme activity during intranasal vaccination with mycobacteria is up- or downregulated by the presence or absence of CT, respectively.
Another provocative finding is that CT holotoxin, but not CTB, is required for the shift of NE localization of COX-2 in alveolar Mϕ in vivo. To intoxicate a cell and therefore to regulate Mϕ activation as a mucosal adjuvant, CTB binds the ganglioside GM1, a receptor present on the membrane of all nucleated cells, and the holotoxin is transported to the ER lumen (39). In the ER, CTA is reduced to generate CTA1, and protein disulfide isomerase (PDI) unfolds and dissociates CTA1 from the holotoxin (40). The unfolded CTA1 toxin is subsequently transported back across the ER membrane by ER-associated degradation (ERAD) (41–44). CTA1 refolds in the cytosol, where it stimulates basolateral adenylate cyclase for cAMP production (26, 27). It is possible, therefore, that the CT-induced processes that facilitate refolding and export of CTA1 from the ER may also refold COX-2 into the NE-associated, active form. This would explain the presence of NE-associated COX-2 in alveolar Mϕ from mice treated intranasally with a mixture of BCG/CT or CT followed by BCG, but not when BCG precedes CT.
However, we also found that the effect of CT under the in vivo condition is distinct from that in vitro on increased COX-2 expression. Therefore, another possibility is that CT may induce a characteristic pattern of mucosal immunity through activation of other mucosal cell types, such as dendritic cells (DC) and mast cells. Available evidence suggests that CT can induce migration and maturation of DC from mesenteric lymph nodes and upregulate their Th2 cytokines via upregulation of Jagged-2 and OX40L (45) but impair the differentiation of monocytes into DC (46). CT induces IL-4 and IL-6 release from mast cells (47). Neurological effects of CT are also considered: following intranasal administration, retrograde transport of CT to the brain via the olfactory system resulted in increased proinflammatory cytokine production and COX-2 expression in murine hypothalamus (48, 49). The potential for vagal stimulation of the enteric nervous system to modulate the immune response cannot be dismissed. These cell populations and systems, which are not components of our Mϕ cultures, potentially contribute to the mucosal Mϕ response to intranasal BCG/CT treatment.
In this regard, our preliminary study provides further support for a contribution of other cells in the CT-stimulated shift to NE-localized COX-2 formation. Previously, we demonstrated that intraperitoneal administration of BCG induces NE-dissociated inactive forms of COX-1 and COX-2 in peritoneal and splenic Mϕ within 24 h (17, 18). In the preliminary study, intraperitoneal administration of a mixture of BCG/CT resulted in no shift of NE localization of COX-2 in either Mϕ population (data not shown). Therefore, the effect of CT on the shift seems to be selectively invoked in mucosal vaccination; mucosal cells in addition to alveolar Mϕ would play a tissue-specific role.
In conclusion, the present study begins to unveil the mechanism of COX-2 inactivation caused by BCG. This newly discovered phenotype regulated by CT, a mucosal adjuvant, would significantly contribute to the regulation of PGE2-mediated pulmonary inflammation and facilitate host mucosal immune responses against mycobacterial vaccinations. Further studies are needed to confirm whether the increase in mucosal PGE2 following intranasal vaccination with a BCG/CT mixture promotes protective immunity against M. tuberculosis, including mucosal IgA, systemic cell-mediated immunity, and/or apoptotic death of bactericidal alveolar Mϕ, which was previously proposed (6, 20, 50).
ACKNOWLEDGMENTS
We thank Traci Pantuso for technical assistance.
This work was supported by NIH RO1 HL71711, DOD DAMD 17-03-1-0004, Bankhead-Coley Cancer Research Program 06BS-04-9615, and the Florida Atlantic University Research Program (Y.S.).
FOOTNOTES
- Received 24 September 2012.
- Returned for modification 23 October 2012.
- Accepted 3 November 2012.
- Accepted manuscript posted online 12 November 2012.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.
