Search tips
Search criteria 


Logo of ajrcmbIssue Featuring ArticlePublisher's Version of ArticleSubmissionsAmerican Thoracic SocietyAmerican Thoracic SocietyAmerican Journal of Respiratory Cell and Molecular Biology
Am J Respir Cell Mol Biol. 2009 August; 41(2): 146–154.
Published online 2008 December 18. doi:  10.1165/rcmb.2008-0230OC
PMCID: PMC2715904

Persistent Inactivation of Macrophage Cyclooxygenase-2 in Mycobacterial Pulmonary Inflammation


The induction of cyclooxygenase-2 (COX-2) in tissue macrophages (M[empty]) increases prostaglandin E2 (PGE2) release, potentially down-regulating granulomatous inflammation. In response to Mycobacteria, local M[empty] express COX-2, which is either nuclear envelope (NE)-associated or NE-dissociated. Persistent mycobacterial pulmonary inflammation is characterized by alveolar M[empty] expressing NE-dissociated (inactive) COX-2 without release of PGE2. In this study, we examined COX-2 in alveolar M[empty] after intranasal exposure to heat-killed Mycobacterium bovis BCG (HK-BCG). After administration, whole lungs of C57Bl/6 mice were lavaged with saline; COX-2 expression and PGE2 release by alveolar M[empty] and tumor necrosis factor (TNF)-α and nitric oxide levels in the lung lavage were monitored. Normal alveolar M[empty] had undetectable levels of COX-2 on Western blots. However, 1 day after intranasal administration, almost all alveolar M[empty] had phagocytosed HK-BCG and expressed NE-dissociated COX-2 without any increase in the release of PGE2. At 28 days after intranasal administration, 68% of alveolar M[empty] still contained both BCG and the NE-dissociated form of COX-2. NE-associated (active) COX-2 was not observed in alveolar M[empty]. In contrast, 7 days after intraperitoneal injection of HK-BCG, peritoneal M[empty] containing HK-BCG were no longer detected. At 28 days after intranasal administration, TNF-α and nitrite levels in the lung lavage fluid were significantly higher than those in controls. Our results indicate that mycobacterial pulmonary inflammation is associated with suppressed PGE2 production by alveolar M[empty], with expression of COX-2 dissociated from the NE.

Keywords: alveolar macrophages, PGE2, COX-2, mycobacteria


This study investigates the pathophysiology associated with mycobacterial pulmonary inflammation. A novel aspect of alveolar macrophage activation by mycobacteria is the persistent presence of inactive cyclooxygenase isoforms with suppression of prostaglandin E2 production.

Alveolar macrophages (M[empty]) secrete numerous chemical mediators in response to Mycobacterium tuberculosis and Mycobacterium bovis bacillus Calmette-Guérin (BCG), a tuberculosis vaccine and adjuvant, thereby regulating persistent inflammatory responses in the lungs and causing granuloma formation. Granulomatous inflammation represents the pathological correlate of both protective immunity and inflammatory tissue destruction and repair (1, 2). Immunoregulatory effects of mycobacteria are demonstrated by increased resistance to infections and tumor growth, attenuation of the allergic response, and induction of autoimmune diseases in animal models (36). The pleiotropic effects of mycobacteria could be mediated by induction in M[empty] of not only Th1-mediated microbicidal M[empty] responses but also pathogenic Th2, Th17, and Treg responses (79). The exact mechanisms regulating persistent mycobacterial inflammation in alveolar M[empty] still remain to be elucidated.

Previous studies with M[empty] demonstrated that mycobacteria induce expression of cyclooxygenase-2 (COX-2), an inducible enzyme, which increases prostaglandin E2 (PGE2) biosynthesis in vitro (10, 11). PGE2 suppresses granuloma formation (12) by down-regulating bactericidal M[empty] activities, including iNOS-mediated synthesis of nitric oxide (NO), NADPH oxidase-mediated superoxide anion release, and synthesis of interleukin (IL)-12/IL-18/tumor necrosis factor (TNF)-α (1315). In contrast, PGE2 promotes production of IL-10, IL-23, and matrix metalloproteinase 9 by M[empty] (7, 8, 16, 17). Therefore, regulation of these responses at various stages of pulmonary inflammation may depend on the presence of COX-2+ M[empty] at appropriate locations.

Catalytically active COX-2 (and COX-1) are localized subcellularly in the nuclear envelope (NE) and endoplasmic reticulum (ER), where they mediate PGE2 biosynthesis (18, 19). COX-2 dissociated from the NE is catalytically inactive (2022). These forms of COX-2 have microscopically distinct cellular appearances and their associated activity in producing PGE2 has been established in several studies including this one. Therefore, we have used the designations active or inactive interchangeably with NE associated or NE dissociated, respectively, Previously, we found that normal splenic and peritoneal M[empty] treated in vitro with heat-killed BCG (HK-BCG) express active COX-2 and release increased amounts of PGE2 within 24 hours (20, 21). In sharp contrast, within 24 hours after intraperitoneal administration, HK-BCG induces inactive COX-2 transiently in splenic and peritoneal M[empty], followed by active COX-2 in splenic M[empty] at 7 to 14 days after administration (20, 21). Splenic M[empty] with active COX-2 are derived from bone marrow (23). The formation of these COX-2+ M[empty] subsets seems to be dependent on the microorganism and its phagocytosis, the tissue, and the presence of bone marrow–derived precursors which become mature M[empty] expressing active COX-2 at inflammatory sites (20, 21, 23, 24). Thus, COX-2+ M[empty] could potentially either inhibit or promote PGE2-sensitive antimicrobial immunity (7, 20, 21).

We have hypothesized that pulmonary inflammation in response to Mycobacteria is directly related to the activity of COX-2. To elucidate the role of M[empty] COX-2 in mycobacterial pulmonary inflammation, we administered HK-BCG intranasally and determined whether this resulted in expression of catalytically active or inactive COX-2, and the persistence of the effects on alveolar M[empty] for up to 28 days. The pulmonary inflammation induced by mycobacteria was also compared with that induced by other selected microbes and fractionated mycobacterial components.



Cultured M. bovis BCG Tokyo 172 strain or Propionibacterium acnes (24) were washed, autoclaved, and lyophilized. This HK-BCG or HK–P. acnes powder was suspended in pyrogen-free saline at 10 mg/ml and dispersed by brief (10 s) sonication immediately before use. These preparations contained undetectable levels of endotoxin (< 0.03 EU/ml), as determined by the Limulus amebocyte lysate assay (Sigma Aldrich, St. Louis, MO) (25). Listeria monocytogenes 10403S serotype 1/2a and Escherichia coli MG1655 (obtained from Dr. Chris Burns, Florida Atlantic University) were grown in brain-heart infusion broth (Difco Laboratories, Detroit, MI) at 37°C. The bacteria were harvested in the logarithmic phase of growth, treated at 60°C for 60 minutes, washed three times with cold saline, re-suspended in saline at 10 mg/ml (5 × 108 bacilli/mg), and stored in aliquots at −80°C. HK–M. tuberculosis and HK–Mycobacterium butyricum were purchased from Difco Laboratories. Lysate of HK-BCG was prepared by sonication (20 s, 15 times) followed by filtration with a 0.22-μm Millipore filter. Alternatively, HK-BCG was treated for 6 hours at 37°C with either 1 M hydrochloric acid or 1 M sodium hydroxide (26, 27). After this treatment, HK-BCG was washed six times with phosphate-buffered saline, pH 7.0 (PBS), and suspended in saline at 10 mg/ml. Delipidated HK-BCG was prepared by extraction with chloform/methanol (3:1, vol/vol) 1 hour at 37°C followed by centrifugation to remove the supernatant. This process was repeated once. The pellet was washed with 70% ethanol six times followed by saline three times and resuspended in saline (26, 27).


Nonpregnant female C57Bl/6 mice, 8 to 10 weeks old, were obtained from Harlan Laboratory (Indianapolis, IN). Mice were maintained in barrier-filtered cages and fed Purina laboratory chow and tap water ad libitum. Experimental protocols employed in this study were approved by the IACUC at Florida Atlantic University.

Administration of HK Microbes

Groups of mice (3–5/group) received 0.5 mg of HK-BCG, HK–P. acnes, HK–L. monocytogenes, or HK–E. coli in 0.05 ml saline intranasally or 1.0 mg in 0.1 ml saline intraperitoneally on Day 0. Controls received 0.05 ml (intranasally) or 0.1 ml (intraperitoneally) saline. Mice were killed on Days 1, 7, 14, and 28.

Macrophage Preparations

To obtain alveolar M[empty] from mice, the trachea was cannulated and the whole lung lavaged with 1 ml saline. For peritoneal M[empty], peritoneal lavage was performed with 5 ml of saline. M[empty] were prepared as previously described (24). Total cell counts were determined with a Coulter counter (Model Z1; Beckman Coulter, Hialeah, FL). Cell classification was performed on cytocentrifuged preparations stained with Diff-Quik (Baxter Healthcare, Miami, FL). To enrich plastic-adherent M[empty] for in vitro experiments, peritoneal or alveolar cells at 106 cells/ml suspended in RPMI 1640 plus 5% FBS were incubated in culture dishes (Falcon, Oxnard, CA) for 2 hours. Nonadherent cells were removed by washing with media. In some experiments, adherent cells were cultured with 20 μg/ml HK-BCG, HK–P. acnes, HK–L. monocytogenes, or HK–E. coli for an additional 24 hours.

Western Blotting

Washed alveolar or peritoneal M[empty] were resuspended in lysis buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 4 mM EDTA, 0.1% SDS, 1:500 protease inhibitor cocktail [P8340; Sigma Aldrich], 1% NP-40, and 1% sodium deoxycholate). After 10 minutes on ice, debris was eliminated by centrifugation (10 min, 10,000 × g). Protein concentration in the lysate was measured with a bicinchoninic acid assay (Pierce, Thermo Fisher Scientific, Rockford, IL) and bovine serum albumin as standard. For each sample, 1.5 μg of lysate protein, unless stated otherwise, was separated by SDS-PAGE, and then transferred to PVDF membrane (Millipore, Bedford, MA). The membrane was blocked with 10% nonfat dry milk, and incubated with antibodies (anti–COX-2, 1:2,000; anti–COX-1, 1:2,000 [Cayman Chemical]; anti–β-actin, 1:8,000 [Sigma Aldrich] for the detection of β-actin as constitutively expressed protein control) in 5% nonfat dry milk, overnight at 4°C. After incubation with peroxidase-conjugated secondary antibody (1:20,000; Jackson ImmunoResearch, West Glove, PA), proteins were detected by chemiluminescence (ECL plus; Amersham, Piscataway, NJ) following the manufacturer's instructions.

Subcellular Localization of COX-2 by Confocal Microscopy Analysis

Alveolar and peritoneal M[empty] prepared as described above were fixed in 4% paraformaldehyde in PBS, pH 7.5, for 30 minutes. The fixed cells were permeabilized with 0.1% Triton X-100 in PBS for 5 minutes and incubated in blocking buffer consisting of PBS with 10% FBS overnight at 4°C before incubation with anti–COX-2 or anti–COX-1 antibody (Cayman Chemical), 1:500 in blocking buffer, overnight at 4°C. Subsequently, cells were washed with PBS three times and incubated with fluorescein isothiocyanate (FITC)-conjugated donkey anti-rabbit IgG (1:500; Jackson ImmunoReseach) for 1 hour at 22°C. For detection of nucleus and HK microbes, propidium iodide (PI; Sigma Aldrich) was mixed at 10 μg/ml with the secondary antibody solution. After washing three times, cells were examined with a laser scanning confocal microscope (Bio-Rad Radiance 2100; Bio-Rad Laboratories, Hercules, CA). Images were processed with Adobe Photoshop software. We have demonstrated previously (20) that HK-BCG is co-stained by PI and anti-BCG anti-sera.

Evaluation of Apoptosis

The apoptosis of M[empty] was examined microscopically. After removal of the culture media from culture dishes, a fluorescent dye solution containing acridine orange (Sigma Aldrich) and PI at 10 and 20 μg/ml, respectively, in culture media was added to the dishes (28). A coverslip was placed over the cells, and they were immediately examined by fluorescence microscopy.


Levels of TNF-α and IL-10 in supernatants from lung lavage were measured by two-site ELISA specific for the respective cytokines, according to the manufacturer's instructions (BD Pharmingen, San Diego, CA).

Measurement of Nitrite Concentration

To evaluate the concentration of NO present in supernatants of lung lavage, the stable conversion product of NO, nitrite (NO2), was measured by the Griess assay (29). Supernatants were added to an equal volume of Griess reagent (1% sulfanilamide and 0.1% naphthylethylenediamine dihydrochloride in 2.5% H3PO4) in a 96-well plate and incubated for 10 minutes at 22°C. Nitrite levels were determined colorimetrically at 540 nm from a standard curve prepared with known concentrations of sodium nitrite.

PGE2 Assay

Plastic adherent M[empty] (106 cells/ml) were cultured in RPMI 1640 for 2 hours. PGE2 levels in the culture supernatants were measured by competitive ELISA (Cayman Chemical, Ann Arbor, MI).

Immunodetection of PGE2 at its Sites of Production

M[empty] were treated with 1% water-soluble 1-ethyl-3(3-dimethylamino-propyl) carbodiimide (EDAC; Sigma Aldrich) to cross-link eicosanoid carboxyl groups to amines in adjacent proteins (30). After 30 minutes of incubation with EDAC at 37°C to promote both cell fixation and permeabilization, cells were stained with mouse anti-PGE2 at 1:50 (Cayman Chemical). Presence of the primary antibody was determined with FITC-conjugated donkey anti-mouse IgG at 1:500 (Jackson ImmunoResearch). Cellular localization was determined by confocal microscopy.

Statistical Analysis

Lung lavage fluid isolated from at least three mice was pooled. Cytokines and nitrite in supernatants were measured in at least triplicate in each group. Differences between mean values were analyzed by Student's t test. P < 0.05 is considered statistically significant.


Catalytically Inactive Forms of COX in Alveolar M[empty]

Both peritoneal and splenic M[empty] isolated 24 hours after intraperitoneal injection of 1 mg HK-BCG contain phagocytosed BCG and express increased levels of COX-2, which is catalytically inactive (20, 21). However, we noted in our earlier investigation that alveolar M[empty] isolated from these mice contained no BCG and expressed no COX-2. In the present studies, we have determined the effect of intranasal administration of HK-BCG on expression of COX-2 by alveolar M[empty] isolated 24 hours after administration of HK-BCG. A dose of 0.5 mg HK-BCG was selected to achieve an inflammatory response comparable to that associated with mycobacterial infections and models of autoimmune disease employing Freund's complete adjuvant (31, 32). In addition, pulmonary granulomatous inflammation with a polarized Th1 cytokine profile has been induced in mice treated with 0.5 mg HK–P. acnes (33).

One day after intranasal administration of HK-BCG, alveolar M[empty] had increased COX-2 protein (Figure 1A). Confocal microscopy revealed that M[empty] containing intracellular HK-BCG were positive for COX-2. Catalytically inactive COX-2 was indicated by its localization in densely stained cytoplasmic structures dissociated from the NE. No NE associated COX-2 was detected (Figure 1B).

Figure 1.Figure 1.
Cyclooxygenase (COX)-1 and COX-2 in alveolar macrophages (M[empty]) after in vivo treatment with heat-killed Mycobacterium bovis BCG (HK-BCG) and other microbes. Alveolar M[empty] were isolated from groups of mice 24 hours after intranasal administration ...

The response to HK-BCG was compared with that induced by intranasal administration of other microbes. As shown in Figures 1A and 1B, in vivo exposure to HK–P. acnes resulted in expression of NE-dissociated COX-2 at a lower level than that induced by HK-BCG. Active (NE-associated) COX-2 was not detected after exposure to HK–P. acnes. In contrast, HK–E. coli induced NE-associated COX-2 whereas HK–L. monocytogenes induced little or no COX-2 (Figures 1A and 1B).

The expression level of COX-1 by normal alveolar M[empty] was significantly lower than for peritoneal M[empty]. By Western blotting of 1.5 μg M[empty] protein, COX-1 was not detected (Figures 1A and and2A).2A). When protein loading was increased to 12.5 or 25 μg, smaller amounts of COX-1 were detected in alveolar M[empty] compared with those from normal peritoneal M[empty] (Figure 2A). COX-1 was not detected in alveolar MØ examined by immunofluorescent confocal microscopy, but was seen in periotoneal M[empty] (Figures 1 and and2).2). Unexpectedly, NE-dissociated COX-1 was induced in alveolar M[empty] by intranasal administration of HK-BCG or HK–P. acnes (Figures 1A and and1B1B).

Figure 2.Figure 2.Figure 2.
COX-1 and COX-2 in alveolar M[empty] treated in vitro with HK-BCG and other microbes. (A) Constitutive COX-1 detection. COX-1 and β-actin levels in normal alveolar M[empty] (AM) were determined by Western blot at 1.5, 12.5, and 25 μg ...

These studies demonstrate that in vivo phagocytosis of HK-BCG by alveolar M[empty] induces the NE-dissociated forms of both COX-2 and COX-1. The formation of NE-dissociated COX is dependent on specific microbe properties, since E. coli and L. monocytogenes did not induce the inactive COX isoforms.

COX Expression In Vitro

We further examined the effects of in vitro exposure to HK–P. acnes, E. coli, and L. monocytogenes compared with the effects of HK-BCG on normal alveolar M[empty]. All of these microbes induced similar amounts of COX-2 protein in alveolar M[empty] (Figure 2A), which by confocal microscopy was NE-associated (Figure 2B). Treatment with lower (1 μg/ml) or higher (100 μg/ml) doses of HK-BCG further confirmed the induction of NE-associated COX-2 (data not shown). COX-2 levels induced by 20 μg/ml of HK-BCG were comparable to those induced by 100 μg/ml of HK-BCG (data not shown). In contrast to COX-2, however, COX-1 was not induced in vitro by HK-BCG or HK-P. acnes (Figures 2B and 2C).

PGE2 Release by Alveolar M[empty] with NE-Dissociated COX-1 and COX-2

Assays of cellular supernatants demonstrated that normal alveolar M[empty] produced less PGE2 than normal peritoneal M[empty] (Figure 3). This constitutive production of PGE2 is probably dependent on the constitutive presence of COX-1 (21). When alveolar M[empty] were treated in vitro with HK-BCG, PGE2 production increased (Figure 3), consistent with the localization of COX-2 at the NE as seen in Figure 2C. In contrast, for alveolar M[empty] treated with HK-BCG in vivo, intracellular PGE2 and cellular PGE2 production were unchanged from the saline control (Figure 3).

Figure 3.
PGE2 release by alveolar M[empty]. For in vitro HK-BCG treatment, normal alveolar (AM) and peritoneal M[empty] (PM) as comparison controls were incubated with 20 μg/ml HK-BCG for 24 hours. For in vivo HK-BCG treatment, mice received 0.5 mg ...

NE-Dissociated COX-1 and COX-2 Formation Persists for 28 Days In Vivo

Previously we found that splenic M[empty] expressed the inactive form of COX-2 1 day after intraperitoneal administration of HK-BCG, and then switched to active COX-2 after an additional 6 to 13 days (20). We also noted that intracellular BCG is preferentially associated with inactive COX-2; conversely active COX-2 was less likely to be associated with intracellular BCG. Therefore, after intranasal administration of HK-BCG we monitored the COX-2 localization and presence of intracellular BCG in alveolar M[empty] for up to 28 days.

PI-stained intracellular BCG were detected in 79%, 72%, and 68% of alveolar M[empty] at Days 7, 14, and 28, respectively, after intranasal HK-BCG administration (Figure 4). The M[empty] with intracellular BCG exclusively expressed the inactive forms of COX-1 and COX-2. In sharp contrast, intracellular HK-BCG were observed in only 5%, 1%, and less than 0.5% of peritoneal M[empty] isolated 7, 14, or 28 days, respectively, after mice were given intraperitoneal HK-BCG (Figure 4). Only the inactive form of COX-2 was observed in the presence of intracellular HK-BCG in these peritoneal M[empty] (Figure 4). Unlike splenic M[empty] at 7 to 14 days, in the experiments described here neither alveolar nor peritoneal M[empty] exhibited the active form of COX-2 even after 28 days (data not shown). Moreover, M[empty] from these two tissues did not undergo apoptosis after phagocytosis of HK-BCG (data not shown).

Figure 4.
COX-1 and COX-2 expression in alveolar M[empty] after intranasal HK-BCG administration. Alveolar M[empty] (AM) were harvested at indicated periods from mice treated intranasally with 0.5 mg HK-BCG. As comparison controls, peritoneal M[empty] (PM) ...

HK-BCG Promotes Production of TNF-α and Nitric Oxide in the Lungs

To confirm the immunomodulatory activity of HK-BCG and HK–P. acnes in vivo, we measured TNF-α in lung lavage. TNF-α is a critical contributor to the granulomatous inflammation of tuberculosis and sarcoidosis (34, 35). As shown in Figure 5A, TNF-α production increased on Day 1 after intranasal HK-BCG, and an even larger increase was seen after treatment with HK–P. acnes. NO, which is released by activated M[empty] and measured as nitrite (36), was also increased after the treatments (Figure 5B). In contrast, IL-10 was produced constitutively and IL-10 levels in the lung lavage were not increased by treatment with either microbe (Figure 5C). In alveolar M[empty], cell surface receptors, including the TLRs, play differential roles in activation by these two bacteria (37, 38), which may explain the differential responses in production of COX-2 and inflammatory mediators.

Figure 5.
TNF-α, IL-10, and NO levels in whole lung lavage. TNF-α/IL-10 and nitrite were measured by ELISA and Griess assay, respectively, in the supernatant obtained from a 1-ml whole lung lavage from mice given 0.5 mg HK-BCG or HK–P. acnes ...

As seen in Figure 5D, TNF-α and nitrite levels in the lungs were significantly higher than those in controls even 28 days after HK-BCG administration. PGE2 levels (Day 0, 120 ± 42; Day 1, 138 ± 46; Day 7, 141 ± 54; Day 14, 54 ± 43; Day 28, 43 ± 33 pg/ml [mean ± SE, n = 4]) did not increase during these 28 days after intranasal HK-BCG). In contrast, TNF-α and NO were not detected in the lung lavage 7 days after HK–E. coli or L. monocytogenes administration, although comparable levels of the mediators were detected on Day 1 (data not shown).

Other HK-Mycobacteria and HK-BCG Components

To further explore the production of COX-2 in alveolar M[empty], we determined whether other HK-mycobacteria and fractionated components of HK-BCG induced NE-dissociated COX-2 after intranasal administration. As summarized in Table 1, NE-dissociated COX-2 was induced not only by HK–M. tuberculosis and M. butyricum, but also by a sonicated lysate of HK-BCG, as well as components of HK-BCG obtained by treatment with chloroform/methanol or 1 M hydrochloric acid. However, HK-BCG pre-treated with 1 M sodium hydroxide did not induce either NE-associated or -dissociated COX-2. These data suggest that mycobacterial proteins and/or acid-resistant carbohydrates are needed for the in vivo induction of NE-dissociated COX-2. In this regard purified protein derivative (PPD), a HK-mixture of secretory mycobacterial proteins, which induced M[empty] COX-2 in vitro, did not induce COX-2 in vivo.



Our previous studies clearly demonstrated that the route of administration of mycobacteria determines whether activated M[empty] express catalytically active or inactive COX-2. The type of COX-2 expressed depended on the tissue type, phagocytosis of the mycobacteria in vivo, and the nature of M[empty] precursors (20, 21, 23). The data presented here indicate that alveolar M[empty] activated by in vivo phagocytosis of mycobacteria express COX-2 that is NE-dissociated and catalytically inactive, failing to increase release of PGE2. Moreover, this type of M[empty] activation remains dominant for at least 28 days after intranasal administration of HK-BCG. Even at this late time point, we were unable to detect either M[empty] expressing the NE-associated form of COX-2, or increased PGE2 release. Our results clearly demonstrate that a single administration of mycobacteria induces a persistent state of M[empty] activation, pulmonary inflammation and anti-microbial activities (e.g., suppressed PGE2 release, with elevation of TNF-α and NO) that continue for 28 days, implying granulomatous inflammation. Based on this and our previous studies (7, 20, 21, 23), Figure 6 summarizes HK-BCG-induced COX-2 expression among alveolar, splenic, and peritoneal M[empty].

Figure 6.
HK-BCG induces tissue-dependent COX-2+ M[empty] formation. Results obtained from this and previous studies (7, 20, 21, 23) are summarized. HK-BCG at 0.5 mg or 1 mg doses was given C57Bl/6 mice intranasally or intraperitoneally, respectively. ...

The respiratory system is exposed to inhaled particulates including various microbes that must be cleared constantly from the airways. Alveolar M[empty] contribute to this clearance largely by phagocytosis of microbes and particulate matter. Our studies of these M[empty] indicate that the induction of a catalytically inactive form of COX-2 by potential pathogens administered in vivo is microbe dependent. However, organisms that produce Th1-mediated granulomatous inflammation, M. tuberculosis, M. butyricum, and P. acnes, as well as M. bovis BCG (2, 33, 39), all induce NE-dissociated COX-2 in alveolar M[empty]. In sharp contrast, in vivo administration of HK-microbes of E. coli or L. monocytogenes induces NE-associated COX-2 or no COX-2, respectively.

Phagocytosis and internalization of HK-BCG by alveolar M[empty] in vivo is associated with formation of NE-dissociated COX-2, which is detected even 28 days after HK-BCG administration. Mycobacteria and P. acnes have complex cell wall structures resistant to host anti-microbial activity and lysosomal digestion, compared with those of E. coli or L. monocytogenes. The altered response of alveolar M[empty] may also relate to the fact that their clearance pathway to the draining lymph node is different from that of peritoneal M[empty] (40, 41); peritoneal M[empty] phagocytosing HK-BCG are cleared within 7 days. We determined that the clearance of these peritoneal M[empty] is not associated with apoptosis (data not shown).

Our previous studies demonstrated that NE-dissociated COX-2 is isolated in the microsomal fraction but is not co-localized with phagosomal HK-BCG or lysosome-associated membrane protein 1 (21), and that an endoplasmic reticulum protein, calreticulin, was not co-localized with NE-dissociated COX-2 (unpublished observation). It is of particular significance that the BCG lysate still induces NE-dissociated COX-2 (Table 1), demonstrating that the soluble molecules are sufficient to induce this altered M[empty] response, and particulate forms of mycobacteria are not required. Our results (Table 1) suggest that mycobacterial proteins and/or carbohydrates play a key role in the induction of inactive COX-2. In this regard, despite induction of COX-2 in vitro, PPD did not induce COX-2 in vivo (Table 1). Further studies are needed to identify the mycobacterial molecules which activate alveolar M[empty] with expression of inactive forms of COX-2.

This study further confirms that in vivo factors associated with HK-BCG administration are obligatory for establishing the catalytically inactive form of COX-2 and consequently the regulation of prostaglandin production. Our previous studies with peritoneal M[empty] in culture (21), however, do not support the hypothesis that extracellular factors produced in response to intraperitoneal HK-BCG administration in the peritoneal cavity or present in sera regulate the subcellular localization of COX-2. Thus, it appears that other endogenous factors present in vivo but not in vitro are important for the induction of catalytically inactive COX-2. From our results, these factors are organism dependent and appear to be effective for at least 28 days after intranasal administration of HK-BCG. We are not aware of previous reports indicating that mycobacteria localize in the peri-NE region of infected M[empty].

COX-1 is the constitutive isoform of COX, but may be inducible in rare cases. For example, developmental induction of COX-1 in ovine pulmonary tissue, especially pulmonary artery, has been reported (42). Our results indicate that normal alveolar M[empty] exhibit a relatively small amount of COX-1 compared with peritoneal M[empty], and that COX-1 is induced by mycobacteria in vivo but not in vitro. Similar differential COX-1 expression between the two M[empty] was shown in normal rats (43). Alveolar M[empty] in alveolar space are exposed to high oxygen tension and have direct contact with alveolar epithelial cells that constitutively produce surfactant and soluble mediators including IL-10 and PGE2. Although no significant increases in these mediators were noted in the lavage fluid assays in this study, it is possible that increases in these mediators might have occurred, but binding to cells or tissues may have prevented their detection in lavage supernatants. This unique environment may down-regulate constitutive COX-1 expression. It is unknown whether the mechanism for induction of COX-1 after in vivo phagocytosis of mycobacteria is similar to that for induction of NE-dissociated COX-2 under the same conditions.

Recent studies with mice deficient in COX-1 and/or COX-2 indicate that COX-1 deficiency is detrimental, whereas COX-2 deficiency is beneficial to the host during pulmonary infection with influenza virus (44) and in airway allergic responses (45). Experiments using specific COX inhibitors and genetic knock-down approaches in selenium-supplemented M[empty] indicate that COX-1, but not COX-2, is responsible for the increased synthesis of 15-deoxy-Δ12,14-PGJ2, an endogenous anti-inflammatory eicosanoid (46). This eicosanoid (prostaglandin) stimulates M[empty] through peroxisome proliferator–activated receptor (PPARγ) activation, and down-regulates M[empty] anti-microbial activities including IL-12 production, oxidative burst, and NO release (47). Unlike peritoneal M[empty], however, alveolar M[empty] isolated from normal mice express PPARγ constitutively (47). Furthermore, PPARγ is also detected in alveolar M[empty] from healthy humans, but is deficient in alveolar M[empty] from patients with pulmonary sarcoidosis, a granulomatous disease with Th1 inflammation (48). Therefore, it remains to be studied whether PPARγ activity in alveolar M[empty] is reduced in response to mycobacteria or P. acnes in a COX-dependent manner.

PGE2, constitutively produced mainly by airway epithelial cells and fibroblasts, influences physiologic functions including increases in surfactant secretion and wound closure (49, 50). Our preliminary immunohistochemical analysis of lung sections indicated that expression of constitutive NE-associated COX-1 and the absence of COX-2 in type II pneumocytes were unchanged by the intranasal administration of HK-BCG (data not shown). This is consistent with our observation that PGE2 release into lung lavage is not increased by exposure to mycobacteria.

In conclusion, we have observed a relatively slower clearance of mycobacterial components by airway alveolar M[empty] compared with that for peritoneal M[empty]. After the phagocytosis of mycobacteria, alveolar M[empty] produce an inactive form of COX-2. This results in “hyperactivated” and persistent inflammation with production of the inflammatory mediators, TNF-α and NO, apparently without a contribution from PGE2 production. Taken together, our series of studies (7, 20, 21, 23) has demonstrated that the kinetics and mechanisms of alveolar M[empty] activation by mycobacteria are distinct from those of peritoneal and splenic M[empty], suggesting tissue-dependent development of chronic inflammation. Additional work is needed to elucidate whether the mechanism of NE-dissociated COX-2+ alveolar M[empty] formation is dependent on stages of M[empty] maturation at persistent mycobacterial pulmonary inflammation. Finally, heat-resistant, filterable molecules, which may be common among mycobacteria and P. acnes, are essential for the expression of the NE-dissociated (inactive) COX-2 and for the up-regulation of host immune responses. Whether this is a direct consequence of suppressed production of PGE2 remains to be established.


The authors thank Dr. Ikuro Honda, Japan BCG Laboratory, Tokyo, for providing HK-BCG and PPD for this study and Dr. C. Kathleen Dorey, Florida Atlantic University, for critical review of the manuscript.


This work was supported by NIH RO1 HL71711, DOD DAMD 17-03-1-0004, Bankhead-Coley Cancer Research Program 06BS-04-9615, the Charles E. Schmidt Biomedical Foundation (Y.S.), and Florida Atlantic University.

Originally Published in Press as DOI: 10.1165/rcmb.2008-0230OC on December 18, 2008

Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.


1. Ehlers S, Benini J, Held HD, Roeck C, Alber G, Uhlig S. Alphabeta T cell receptor-positive cells and interferon-gamma, but not inducible nitric oxide synthase, are critical for granuloma necrosis in a mouse model of mycobacteria-induced pulmonary immunopathology. J Exp Med 2001;194:1847–1859. [PMC free article] [PubMed]
2. Fulton SA, Martin TD, Redline RW, Henry Boom W. Pulmonary immune responses during primary Mycobacterium bovis–Calmette-Guerin bacillus infection in C57Bl/6 mice. Am J Respir Cell Mol Biol 2000;22:333–343. [PubMed]
3. Erb KJ, Holloway JW, Sobeck A, Moll H, Le Gros G. Infection of mice with Mycobacterium bovis-Bacillus Calmette-Guerin (BCG) suppresses allergen-induced airway eosinophilia. J Exp Med 1998;187:561–569. [PMC free article] [PubMed]
4. Sakuma T, Suenaga T, Yoshida I, Azuma M. Mechanisms of enhanced resistance of Mycobacterium bovis BCG-treated mice to ectromelia virus infection. Infect Immun 1983;42:567–573. [PMC free article] [PubMed]
5. Davies M, Sabbadini E. Effect of different doses of BCG on an allogeneic cell-mediated antitumor response in mice. J Natl Cancer Inst 1977;59:263–265. [PubMed]
6. Engleman EG, Sonnenfeld G, Dauphinee M, Greenspan JS, Talal N, McDevitt HO, Merigan TC. Treatment of NZB/NZW F1 hybrid mice with Mycobacterium bovis strain BCG or type II interferon preparations accelerates autoimmune disease. Arthritis Rheum 1981;24:1396–1402. [PubMed]
7. Shibata Y, Henriksen RA, Honda I, Nakamura RM, Myrvik QN. Splenic PGE2-releasing macrophages regulate Th1 and Th2 immune responses in mice treated with heat-killed BCG. J Leukoc Biol 2005;78:1281–1290. [PubMed]
8. Sheibanie AF, Tadmori I, Jing H, Vassiliou E, Ganea D. Prostaglandin E2 induces IL-23 production in bone marrow-derived dendritic cells. FASEB J 2004;18:1318–1320. [PubMed]
9. Garg A, Barnes PF, Roy S, Quiroga MF, Wu S, Garcia VE, Krutzik SR, Weis SE, Vankayalapati R. Mannose-capped lipoarabinomannan- and prostaglandin E2-dependent expansion of regulatory T cells in human Mycobacterium tuberculosis infection. Eur J Immunol 2008;38:459–469. [PMC free article] [PubMed]
10. Cross ML, Slobbe LJ, Buchan GS, Griffin JF. In vitro responses of cervine macrophages to bacterial stimulants. Vet Immunol Immunopathol 1996;53:249–256. [PubMed]
11. Keller R, Keist R, van der Meide PH. Modulation of tumoricidal activity, induced in bone-marrow-derived mononuclear phagocytes by interferon gamma or Corynebacterium parvum, by interferon beta, tumor necrosis factor, prostaglandin E2, and transforming growth factor beta. Int J Cancer 1991;49:796–800. [PubMed]
12. Rangel Moreno J, Estrada Garcia I, De La Luz Garcia Hernandez M, Aguilar Leon D, Marquez R, Hernandez Pando R. The role of prostaglandin E2 in the immunopathogenesis of experimental pulmonary tuberculosis. Immunology 2002;106:257–266. [PubMed]
13. van der Pouw Kraan TC, Boeije LC, Smeenk RJ, Wijdenes J, Aarden LA. Prostaglandin-E2 is a potent inhibitor of human interleukin 12 production. J Exp Med 1995;181:775–779. [PMC free article] [PubMed]
14. Harbrecht BG, Kim YM, Wirant EA, Simmons RL, Billiar TR. Timing of prostaglandin exposure is critical for the inhibition of LPS- or IFN-gamma-induced macrophage NO synthesis by PGE2. J Leukoc Biol 1997;61:712–720. [PubMed]
15. Pillinger MH, Philips MR, Feoktistov A, Weissmann G. Crosstalk in signal transduction via EP receptors: prostaglandin E1 inhibits chemoattractant-induced mitogen-activated protein kinase activity in human neutrophils. Adv Prostaglandin Thromboxane Leukot Res 1995;23:311–316. [PubMed]
16. Shinomiya S, Naraba H, Ueno A, Utsunomiya I, Maruyama T, Ohuchida S, Ushikubi F, Yuki K, Narumiya S, Sugimoto Y, et al. Regulation of TNFalpha and interleukin-10 production by prostaglandins I(2) and E(2): studies with prostaglandin receptor-deficient mice and prostaglandin E-receptor subtype-selective synthetic agonists. Biochem Pharmacol 2001;61:1153–1160. [PubMed]
17. Pavlovic S, Du B, Sakamoto K, Khan KM, Natarajan C, Breyer RM, Dannenberg AJ, Falcone DJ. Targeting prostaglandin E2 receptors as an alternative strategy to block cyclooxygenase-2-dependent extracellular matrix-induced matrix metalloproteinase-9 expression by macrophages. J Biol Chem 2006;281:3321–3328. [PubMed]
18. Otto JC, Smith WL. The orientation of prostaglandin endoperoxide synthases-1 and -2 in the endoplasmic reticulum. J Biol Chem 1994;269:19868–19875. [PubMed]
19. Morita I, Schindler M, Regier MK, Otto JC, Hori T, DeWitt DL, Smith WL. Different intracellular locations for prostaglandin endoperoxide H synthase-1 and -2. J Biol Chem 1995;270:10902–10908. [PubMed]
20. Yamashita M, Tsuji S, Nishiyama A, Myrvik QN, Henriksen RA, Shibata Y. Differential subcellular localization of COX-2 in macrophages phagocytosing heat-killed Mycobacterium bovis BCG. Am J Physiol Cell Physiol 2007;293:C184–C190. [PubMed]
21. Yamashita M, Shinohara T, Tsuji S, Myrvik QN, Nishiyama A, Henriksen RA, Shibata Y. Catalytically inactive cyclooxygenase 2 and absence of prostaglandin E2 biosynthesis in murine peritoneal macrophages following in vivo phagocytosis of heat-killed Mycobacterium bovis bacillus Calmette-Guerin. J Immunol 2007;179:7072–7078. [PubMed]
22. Spencer AG, Thuresson E, Otto JC, Song I, Smith T, DeWitt DL, Garavito RM, Smith WL. The membrane binding domains of prostaglandin endoperoxide H synthases 1 and 2: peptide mapping and mutational analysis. J Biol Chem 1999;274:32936–32942. [PubMed]
23. Shibata Y, Gabbard J, Yamashita M, Tsuji S, Smith M, Nishiyama A, Henriksen RA, Myrvik QN. Heat-killed BCG induces biphasic cyclooxygenase 2+ splenic macrophage formation–role of IL-10 and bone marrow precursors. J Leukoc Biol 2006;80:590–598. [PubMed]
24. Shibata Y, Bautista AP, Pennington SN, Humes JL, Volkman A. Eicosanoid production by peritoneal and splenic macrophages in mice depleted of bone marrow by 89Sr. Am J Pathol 1987;127:75–82. [PubMed]
25. Shibata Y, Honda I, Justice JP, Van Scott MR, Nakamura RM, Myrvik QN. Th1 adjuvant N-acetyl-D-glucosamine polymer up-regulates Th1 immunity but down-regulates Th2 immunity against a mycobacterial protein (MPB-59) in interleukin-10-knockout and wild-type mice. Infect Immun 2001;69:6123–6130. [PMC free article] [PubMed]
26. Melancon-Kaplan J, Hunter SW, McNeil M, Stewart C, Modlin RL, Rea TH, Convit J, Salgame P, Mehra V, Bloom BR, et al. Immunological significance of Mycobacterium leprae cell walls. Proc Natl Acad Sci USA 1988;85:1917–1921. [PubMed]
27. Wang L, Slayden RA, Barry CE III, Liu J. Cell wall structure of a mutant of Mycobacterium smegmatis defective in the biosynthesis of mycolic acids. J Biol Chem 2000;275:7224–7229. [PubMed]
28. Lizon C, Fritsch P. Chemical toxicity of some actinides and lanthanides towards alveolar macrophages: an in vitro study. Int J Radiat Biol 1999;75:1459–1471. [PubMed]
29. Ding AH, Nathan CF, Stuehr DJ. Release of reactive nitrogen intermediates and reactive oxygen intermediates from mouse peritoneal macrophages: comparison of activating cytokines and evidence for independent production. J Immunol 1988;141:2407–2412. [PubMed]
30. D'Avila H, Melo RC, Parreira GG, Werneck-Barroso E, Castro-Faria-Neto HC, Bozza PT. Mycobacterium bovis Bacillus Calmette-Guerin induces TLR2-mediated formation of lipid bodies: intracellular domains for eicosanoid synthesis in vivo. J Immunol 2006;176:3087–3097. [PubMed]
31. Billiau A, Matthys P. Modes of action of Freund's adjuvants in experimental models of autoimmune diseases. J Leukoc Biol 2001;70:849–860. [PubMed]
32. Chillingworth NL, Donaldson LF. Characterisation of a Freund's complete adjuvant-induced model of chronic arthritis in mice. J Neurosci Methods 2003;128:45–52. [PubMed]
33. McCaskill JG, Chason KD, Hua X, Neuringer IP, Ghio AJ, Funkhouser WK, Tilley SL. Pulmonary immune responses to Propionibacterium acnes in C57BL/6 and BALB/c mice. Am J Respir Cell Mol Biol 2006;35:347–356. [PMC free article] [PubMed]
34. Bean AG, Roach DR, Briscoe H, France MP, Korner H, Sedgwick JD, Britton WJ. Structural deficiencies in granuloma formation in TNF gene-targeted mice underlie the heightened susceptibility to aerosol Mycobacterium tuberculosis infection, which is not compensated for by lymphotoxin. J Immunol 1999;162:3504–3511. [PubMed]
35. Medica I, Kastrin A, Maver A, Peterlin B. Role of genetic polymorphisms in ACE and TNF-alpha gene in sarcoidosis: a meta-analysis. J Hum Genet 2007;52:836–847. [PubMed]
36. Nathan CF. Secretory products of macrophages. J Clin Invest 1987;79:319–326. [PMC free article] [PubMed]
37. Yadav M, Schorey JS. The beta-glucan receptor dectin-1 functions together with TLR2 to mediate macrophage activation by mycobacteria. Blood 2006;108:3168–3175. [PubMed]
38. Kalis C, Gumenscheimer M, Freudenberg N, Tchaptchet S, Fejer G, Heit A, Akira S, Galanos C, Freudenberg MA. Requirement for TLR9 in the immunomodulatory activity of Propionibacterium acnes. J Immunol 2005;174:4295–4300. [PubMed]
39. Nishiwaki T, Yoneyama H, Eishi Y, Matsuo N, Tatsumi K, Kimura H, Kuriyama T, Matsushima K. Indigenous pulmonary Propionibacterium acnes primes the host in the development of sarcoid-like pulmonary granulomatosis in mice. Am J Pathol 2004;165:631–639. [PubMed]
40. Bellingan GJ, Caldwell H, Howie SE, Dransfield I, Haslett C. In vivo fate of the inflammatory macrophage during the resolution of inflammation: inflammatory macrophages do not die locally, but emigrate to the draining lymph nodes. J Immunol 1996;157:2577–2585. [PubMed]
41. Jakubzick C, Tacke F, Llodra J, van Rooijen N, Randolph GJ. Modulation of dendritic cell trafficking to and from the airways. J Immunol 2006;176:3578–3584. [PubMed]
42. Brannon TS, North AJ, Wells LB, Shaul PW. Prostacyclin synthesis in ovine pulmonary artery is developmentally regulated by changes in cyclooxygenase-1 gene expression. J Clin Invest 1994;93:2230–2235. [PMC free article] [PubMed]
43. Wilborn J, DeWitt DL, Peters-Golden M. Expression and role cyclooxygenase isoforms in alveolar and peritoneal macrophages. Am J Physiol Lung Cell Mol Physiol 1995;268:L294–L301. [PubMed]
44. Carey MA, Bradbury JA, Seubert JM, Langenbach R, Zeldin DC, Germolec DR. Contrasting effects of cyclooxygenase-1 (COX-1) and COX-2 deficiency on the host response to influenza A viral infection. J Immunol 2005;175:6878–6884. [PubMed]
45. Carey MA, Germolec DR, Bradbury JA, Gooch RA, Moorman MP, Flake GP, Langenbach R, Zeldin DC. Accentuated T helper type 2 airway response after allergen challenge in cyclooxygenase-1−/− but not cyclooxygenase-2−/− mice. Am J Respir Crit Care Med 2003;167:1509–1515. [PubMed]
46. Vunta H, Davis F, Palempalli UD, Bhat D, Arner RJ, Thompson JT, Peterson DG, Reddy CC, Prabhu KS. The anti-inflammatory effects of selenium are mediated through 15-deoxy-Delta12,14-prostaglandin J2 in macrophages. J Biol Chem 2007;282:17964–17973. [PubMed]
47. Reddy RC, Keshamouni VG, Jaigirdar SH, Zeng X, Leff T, Thannickal VJ, Standiford TJ. Deactivation of murine alveolar macrophages by peroxisome proliferator-activated receptor-gamma ligands. Am J Physiol Lung Cell Mol Physiol 2004;286:L613–L619. [PubMed]
48. Culver DA, Barna BP, Raychaudhuri B, Bonfield TL, Abraham S, Malur A, Farver CF, Kavuru MS, Thomassen MJ. Peroxisome proliferator-activated receptor gamma activity is deficient in alveolar macrophages in pulmonary sarcoidosis. Am J Respir Cell Mol Biol 2004;30:1–5. [PubMed]
49. Morsy MA, Isohama Y, Miyata T. Prostaglandin E(2) increases surfactant secretion via the EP(1) receptor in rat alveolar type II cells. Eur J Pharmacol 2001;426:21–24. [PubMed]
50. Savla U, Appel HJ, Sporn PH, Waters CM. Prostaglandin E(2) regulates wound closure in airway epithelium. Am J Physiol Lung Cell Mol Physiol 2001;280:L421–L431. [PubMed]

Articles from American Journal of Respiratory Cell and Molecular Biology are provided here courtesy of American Thoracic Society