Defective Apoptotic Cell Phagocytosis Attenuates Prostaglandin E^sub 2^ and 15-Hydroxyeicosatetraenoic Acid in Severe Asthma Alveolar Macrophages

Clearance of apoptotic cells by phagocytes is crucial in maintaining tissue homeostasis. This process prevents dying cells from releasing proinflammatory contents into the environment. In animal and in vitro studies, apoptotic cell clearance induces secretion of antiinflammatory mediators, such as transforming growth factor ß1 (TGF-ß1) and prostaglandin E^sub 2^ (PGE^sub 2^), with associated suppression of proinflammatory cytokines, chemokines, and eicosanoids (1-3). The physiologic significance of this phenomenon has been confirmed in a murine model in which in vivo clearance of apoptotic cells accelerates the resolution of LPS-induced lung inflammation (2), mediated partly by induction of TGF-ß1. Both unstimulated murine (4-6) and human monocyte-derived macrophages (1) have been demonstrated to phagocytose apoptotic cells, which is markedly enhanced after activation with LPS, zymosan (1, 2), glucocorticoids (7), and lipoxins (8).

Regulation of apoptotic cell clearance in human lungs, either in the normal or inflamed/diseased state, remains poorly understood, as most previous studies have been limited to animal models, cell lines, and human monocyte-derived macrophages. We sought to examine clearance in human alveolar macrophages from normal lungs and from lungs with asthma (9). Apoptotic cell uptake was examined in freshly harvested human bronchoalveolar lavage (BAL) macrophages (AMfs) from normal lungs and lungs of patients with mild-moderate and severe asthma. We hypothesized that defects in apoptotic cell uptake by AMfs from asthmatic lungs could contribute to the chronic inflammation seen in this disease.

METHODS

Cells

Human T-lymphocyte Jurkat cells (American Type Tissue Culture Collection, Manassas, VA) were cultured in Roswell Park Memorial Institute 1640 medium (10% fetal calf serum, 2 mM L-glutamine, 100 U/ml penicillin, 100

Reagents

See online supplement for details.

Subjects

Patients with severe asthma were referred to the National Jewish Medical and Research Center for severe oral corticosteroid-dependent asthma. These subjects had frequent hospitalizations and/or emergency room visits, evidence for ongoing severe airflow limitation (FEV^sub 1^ < 70% predicted), and oral or high-dose inhaled corticosteroid use (10). Patients with moderate asthma had FEV^sub 1^ less than 80% predicted on 400-1,000 µg of inhaled corticosteroids and ß-agonists. Patients with mild asthma had an FEV^sub 1^ greater than 80% predicted on ß-agonists alone. Normal control subjects had normal pulmonary function, no bronchodilator response, and no history of respiratory illness. Some patients with severe asthma were treated in vivo with a high-dose corticosteroid burst (prednisone equivalent of at least 40 mg/d × 7-10 d), and BAL was again performed.

Isolation of AMfs

BAL cells were isolated and cytospins made (11 ). AMfs were isolated by adhesion (60 min) and cultured in serum-free X-vivo 10 + gentamycin media (37°C, 5% CO2) for all experiments (see online supplement for details).

Viability Assays

Suspended, unfixed AMfs were stained with trypan blue (0.08%, 1 min) and examined with light microscopy. DNA fragmentation of apoptotic cells were detected in cytospins using the enzyme terminal deoxynucleotidyl transferase with the DeadEnd colorimetric transferasemediated deoxyuridine triphosphate nick end labeling (TUNEL) system (Promega, Madison, WI; see online supplement for details).

Phagocytosis Assays

The AMfs (5 × 10^sup 4^ AMfs/12-mm glass coverslip) were stimulated with 10 µg/ml LPS for 2-8 h, and/or dexamethasone (Dex, 0.1-50.0 µl) for 1-6 h. Because no difference was found between addition of Dex before, during, or after introduction of LPS, Dex was added after LPS for all subsequent experiments. Apoptotic Jurkats were added to AMf cultures (ratio of 20:1 in 250 µl media) for 60 min. Phagocytic bodies were assessed by light microscopy in a blinded fashion; phagocytosis was quantified by the phagocytic index (PI, %), which represents the number of phagocytic bodies per 200 AMfs × 100 (12). Controls included cell-free media, viable Jurkats, and opsonized apoptotic Jurkats.

Cytokine Assays

A total of 0.3-0.4 × 10^sup 6^ AMfs/well were stimulated with LPS (10 µg/ml, 8 h), followed by Dex (1 µM, 1 h). Apoptotic Jurkats were incubated with AMfs (ratio of 10:1 in 250 µl medium) for 2 h, and then replaced with 500 µl of fresh media. The AMf cultures were incubated for 18 h, and the supernatants were assayed for cytokine and eicosanoid concentrations using ELISAs. Similar controls were performed as described above.

ELISAs

Human tumor necrosis factor a (hTNF-a), human granulocyte-macrophage colony-stimulating factor (hGM-CSF), hPGE^sub 2^, human 15-hydroxyeicosatetraenoic acid (h15-HETE), and hTGF-ß1 and hTGF-ß2 were measured from the 18-h supernatants of cultured AMfs (see online supplement for details).

Statistical Analysis

Analysis of variance with Student's t test and Dunnett's method was performed using JMP software (SAS Institute, Cary, NC)

RESULTS

Patient Selection and Profiles

The profiles of normal control subjects and subjects with asthma included in the retrospective review of BAL cytospins (Figure 1) are presented in Table 1. A second group of subjects, used primarily for the in vitro experiments, is represented in Table 2. In both Tables 1 and 2, the FEV^sub 1^ was lower in all patients with asthma, whereas the FVC was lower in patients with severe asthma only. The percent change in FEV^sub 1^ after bronchodilator was increased and the FEV^sub 1^/FVC ratio was decreased in all patients with asthma compared with normal subjects. All patients with severe asthma were on oral corticosteroids (prednisone or equivalent 36.89 ± 8.34 mg/d [Table 1] and 23 ± 0.03 mg/d [Table 2]), whereas patients with moderate asthma were on inhaled steroids and only patients with mild asthma were on no steroids. All subjects had less than a 5 pack-yr history of cigarette smoking and are currently nonsmokers.

In Table 1, normal control subjects and patients with asthma were similar with regard to age, sex ratio, and BAL total cell counts and differentials. In the second population represented in Table 2, normal control subjects and patients with asthma were similar with regard to age, sex, and BAL total cell counts. The combined mild and moderate asthma group was smaller in size, more likely to be female (8/12), and had less total cells. Patients with severe asthma had a significantly lower percentage of macrophages, and trends for both higher lymphocyte and neutrophil percentages.

AMfs from Patients with Severe Asthma Had Reduced PIs in BAL Cytospins That Normalized after an In Vivo Corticosteroid Burst

Indirect evidence of human AMf phagocytosis was obtained by retrospectively comparing the phagocylic bodies present in BAL cytospin AMfs obtained from normal subjects and patients with mild-moderate and severe asthma. AMfs from normal subjects and patients with mild-moderate asthma had similar numbers of phagocytic bodies (quantified by PIs; Figure 1). In contrast, AMfs from patients with severe asthma had a reduced PI compared with all other subjects. Twelve patients with severe asthma were treated with a high-dose oral corticosteroid burst (prednisone equivalent of at least 40 mg/d × 7-10 d), followed by a BAL. The PI was significantly increased in the postburst cytospin AMfs compared with preburst.

Human AMfs from Normal Subjects Phagocytosed Apoptotic Cells In Vitro

To begin to address these discrepancies between AMfs from normal subjects and those from patients with severe asthma in vivo, the phagocytic activities of normal AMfs were assessed in vitro. Freshly harvested AMfs from normal control subjects were cultured and exposed to cell-free media, live human T-lymphocyte Jurkat cells (ViableJ), apoptotic Jurkats (ApoJ), or opsonized apoptotic Jurkats (OpsJ) for 60 min. The phagocytic efficiency (quantified by the PI) was increased in AMfs exposed to ApoJ compared with cells exposed to cell-free media, ViableJ, and OpsJ (Figure 2). Thus, unstimulated normal human AMfs are able to phagocytose apoptotic cells, but not viable or opsonized cells.

LPS- and Dex-induced Uptake of Apoptotic Cells by Normal Human AMfs

The regulation of apoptotic cell uptake by LPS and the corticosteroid Dex in normal control AMfs was then examined. Normal control AMfs were stimulated with LPS and/or Dex, followed by exposure to cell-free media, ViableJ, ApoJ and OpsJ. LPS increased the phagocytosis of both ApoJ and OpsJ, suggesting a global activation of the AMfs (Figure 2B). In contrast, pre-exposure to Dex increased the uptake of only ApoJ, and had no effect on ViableJ or OpsJ (Figure 2C). The LPS + Dex combination had no additional effect over either stimulus alone (Figure 2D). LPS induction of uptake of ApoJ increased linearly with increased stimulation time, and 8 h of LPS stimulation was subsequently chosen for the remaining experiments to minimize AMfs ex vivo time (Figure 3A). Maximal effects of Dex on uptake of ApoJ were observed within 30 min and at 1 µM concentration (Figures 3B and 3C).

AMfs from Patients with Severe Asthma Had Defective In Vitro Apoptotic Cell Uptake in Unstimulated and LPS-stimulated States, which Was Restored by Corticosteroids

The reduced in vivo PI in BAL cytospin AMfs from patients with severe asthma (Figure 1) may have resulted from either decreased or increased phagocytosis efficiency, as well as reduced availability of phagocytic targets. Phagocytosis by AMfs from patients with mild-moderate and severe asthma was thus examined in vitro using freshly isolated AMfs. AMfs from normal control subjects and patients with asthma were cultured and exposed to Jurkats as described above. In patients with mild-moderate asthma, unstimulated AMfs could phagocytose ApoJ (Figure 4) as in normal AMfs. LPS further enhanced uptake of both ApoJ and OpsJ, whereas Dex enhanced specific uptake of ApoJ. Thus, phagocytosis of Jurkats cells by AMfs from patients with mild-moderate asthma was similar to that in AMfs from normal control subjects.

In contrast to the other groups, unstimulated AMfs from patients with severe asthma were unable to phagocytose ApoJ (Figures 4 and 5A), confirming the defect found in vivo in BAL cytospin AMfs. Furthermore, LPS failed to induce ApoJ uptake in the AMfs from patients with severe asthma (Figures 4 and 5B), yet LPS was able to enhance phagocytosis of OpsJ. Conversely, the upregulation of uptake of ApoJ by Dex in AMfs from patients with severe asthma was similar to that of the other subject groups (Figure 5C), consistent with the increased PI found in vivo in the postcorticosteroid BAL cytospin AMfs. Thus, AMfs from patients with severe asthma appear to have a defect in basal and LPS-stimulated uptake of apoptotic cells.

LPS Induction of Inflammatory Mediators Was Reduced in AMfs from Patients with Severe Asthma

Because the defect in uptake of apoptotic cells by AMfs from patients with severe asthma was linked to LPS responsiveness, the impact of LPS on inflammatory mediator secretion was also examined. AMfs from normal subjects and those with asthma were stimulated with LPS or Dex. Protein concentrations of TNF-a, GM-CSF, transforming growth factor ß1 (TGF-ß1), TNF-ß2, PGE^sub 2^, and 15-HETE were measured in the supernatants at 18 h. Mediator levels were minimally detected in unstimulated and Dex-stimulated AMfs from all groups (data not shown). LPS stimulation significantly increased the secretion of TNF-a, GM-CSF, PGE^sub 2^, and 15-HETE in all subject groups (Figure 6). However, the magnitude of the LPS induction of TNF-a and GM-CSF was significantly lower in all AMfs from patients with asthma compared with those from normal subjects. PGE^sub 2^ and 15-HETE levels were significantly reduced in AMfs from patients with severe asthma compared with normal control subjects and patients with mild-moderate asthma. In contrast to other mediators, supernatant TGF-ß1 and ß2 proteins were minimally detectable in all subjects even after LPS stimulation (ELISAs for TGF-ß1 and ß2 had lower detection limits of 10 pg/ml).

Apoptotic Cell Exposure Failed to Induce PGE^sub 2^ and 15-HETE in AMfs from Patients with Severe Asthma

To determine the functional consequence of defective uptake of apoptotic cells by AMfs from patients with severe asthma, mediator secretion was measured after exposure to ApoJ in the presence and absence of LPS. In unstimulated AMfs from all subject groups, ApoJ had no effect on baseline mediator levels (data not shown). After LPS stimulation, the addition of ApoJ to AMfs from normal subjects and patients with mild-moderate asthma further increased PGE^sub 2^ and 15-HETE secretions over levels from LPS stimulation alone (Figure 7). There was no effect on TNF-a, GM-CSF, TGF-ß1, or TGF-ß2 secretion. In contrast, the addition of ApoJ to AMfs from patients with severe asthma (after LPS stimulation) had no effect on PGE^sub 2^ or 15-HETE secretions. The lack of effect of ApoJ on TNF-a, GM-CSF, TGF-ß1, or TGF-ß2 secretion in patients with severe asthma was similar to control subjects. Exposure to ViableJ and OpsJ had no effect on mediator secretion.

AMf Viabilities Were Similar between Normal Control Subjects and Patients with Asthma

To ensure that the above findings did not result from differences in AMf viability between subject groups, the viability of the harvested AMfs was assessed by trypan blue dye exclusion and DNA fragmentation. Low levels of positively stained trypan blue dye were demonstrated in all subject groups, consistent with high viability (Figure 8A). An earlier marker of apoptosis, chromosomal DNA fragmentation, was detected by the TUNEL assay. Low levels of positively stained cells by TUNEL reaction (DNA fragmentation) were seen in both AMfs from normal control subjects and those from patients with severe asthma (Figure 8B).

DISCUSSION

In this study, differences in uptake of apoptotic cells and secretion of inflammatory mediators were observed between primary human AMfs from normal control subjects and patients with asthma. Phagocytosis of apoptotic cells by AMfs from normal control subjects and patients with milder asthma was associated with an induction of the antiinflammatory and/or antifibrotic eicosanoids, 15-HETE and PGE^sub 2^. In contrast, AMfs from patients with severe asthma had defective basal and LPS-mediated uptake of apoptotic cells, with associated failure to induce PGE^sub 2^ and 15-HETE.

Although unstimulated normal human AMfs inefficiently phagocytose apoptotic cells (as reported for other cell types) (4-6), this process could be enhanced by preactivation with LPS or Dex (1, 7, 13). LPS-mediated phagocytosis of apoptotic cells induced secretion of the antiinflammatory and/or antifibrotic eicosanoids, PGE^sub 2^ and 15-HETE (Figure 7). LPS stimulation resulted in a global activation of the AMfs, as it also stimulated secretion of inflammatory mediators and phagocytosis of opsonized cells. In contrast, corticosteroid-mediated phagocytosis was specific to apoptotic cells, with associated suppression of inflammatory mediators.

In this first study of apoptotic cell uptake in asthma, AMfs from patients with severe asthma, but not from patients with mild-moderate asthma, appeared to have decreased in vivo phagocytic bodies, which then increased after an in vivo corticosteroid burst (Figure 1). Although this in vivo finding may have resulted from decreased phagocytosis, it can also be seen in cells with increased phagosome degradation efficiency or reduced availability of phagocytic targets. We agree that this is only a measure of intracellular phagocytic bodies. However, using phagocytosis assays, we proceeded to show that, in vitro, AMfs from patients with severe asthma demonstrated defective phagocytosis of apoptotic cells in both unstimulated and LPS-stimulated states that was restored by corticosteroid treatment (Figures 4 and 5). This suggests that similar defects in clearance may exist in vivo (Figure 1).

The LPS-mediated defect was specific for apoptotic cell uptake, as phagocytosis of cells via other mechanisms was intact (e.g., opsonized apoptotic cells via Fc? receptor ligation). However, a decreased LPS response was also observed in the finding that LPS stimulated lower levels of inflammatory mediators in all patients with asthma compared with normal control subjects, and, in patients with severe asthma, the decreased secretion of PGE^sub 2^ and 15-HETE. Finally, the LPS-related defective apoptotic cell uptake was associated with failure to induce PGE^sub 2^ and 15-HETE in AMfs from patients with severe asthma.

This suppressed LPS inflammatory mediator response was also observed by several authors. Chanez and colleagues demonstrated that LPS stimulated secretion of interleukin 1, TNF-a, and interleukin 6 in AMfs from patients with asthma was decreased when compared with nonatopic normal control subjects (14). Data from one study showing increased LPS-stimulated GM-CSF and TNF-a levels in AMfs from patients with asthma could not be compared with our data because the control group consisted mostly of atopic and smoking subjects, which may influence AMfs activation state (15). The decreased production of PGE^sub 2^ was consistent with findings in airway smooth muscle, suggesting decreased PGE^sub 2^ and cyclooxygenase-2 expression in asthma (16). Defective phagocytosis of apoptotic cells was also demonstrated in unstimulated AMfs from patients with chronic obstructive pulmonary disease (17). This may suggest that defective phagocytosis of apoptotic cells is a common feature of chronic lung diseases.

Some of our findings were surprising. The previously described increase in 15-HETE levels in severe asthma lavage fluid and tissue (18, 19) may suggest that other cell types (epithelial cells and eosinophils) are the primary source in the lung tissue. Surprisingly, in contrast to other models (1, 13, 20-23), TGF-ß1 and TGF-ß2 were only minimally secreted by human AMf1 and unaffected by apoptotic cell clearance, further supporting the necessity of using primary human cells in research. In fact, actual TGF-ß protein secretion has been low in other studies using primary normal human AMfs (24-26), in spite of being detectable by immunohistochemistry and polymerase chain reaction (27, 28). One report of increased TGF-ß1 secretion in AMfs from patients with asthma (24) was confounded by the presence of fetal bovine serum, which may have influenced the measured TGF-ß1 level. Finally, previously reported defective Fc receptor-mediated phagocytosis (29, 30) in asthma was not observed in this study, which may be due to different patient and cell population (sputum/peripheral blood macrophage) and activation state of the obtained cells (inhaled LPS was used to recruit leukocytes).

The mechanism behind the defective apoptotic cell uptake with LPS in AMfs from patients with severe asthma is most likely complex, involving both intrinsic and extrinsic factors. An intrinsic defect in apoptotic cell uptake mechanisms may be present in AMfs from patients with asthma, but the observation that LPS-stimulated mediator secretion was suppressed and apoptotic cells could be phagocytosed in response to corticosteroids both suggested defects in LPS signaling pathways. Viabilities of the AMfs were similar between all subject groups (Figure 8), so other mechanisms are required to explain the defective apoptotic cell phagocytosis and LPS signaling.

Although an effect of chronic systemic corticosteroid use to suppress the LPS response (Figure 6) is difficult to exclude, the low levels of TNF-a and GM-CSF after LPS stimulation in nonsteroid-treated patients with mild asthma suggest that other factors may contribute to the suppressed LPS response. Interestingly, an acute treatment with corticosteroids enhanced uptake of apoptotic cells, but chronic steroid use was insufficient to restore the defective phagocytosis found exclusively in AMfs from patients with severe asthma. Whether chronic versus acute steroid use could have opposite effects can only be answered by direct comparison of chronic versus acute steroid exposure in normal subjects. These studies are difficult to execute. Other environmental and intrinsic factors involved in suppressed LPS response are currently being investigated in our laboratory.

In conclusion, this study expands previous studies in human monocyte-derived macrophages, murine models, and cell lines to human airway cells in the inflammatory disease of asthma. First, in normal AMfs, classically opposing inflammatory (LPS) and antiinflammatory (Dex) mediators were both shown to upregulate apoptotic cell uptake, enabling resolution of inflammation. One appears to work through specific secretion of antiinflammatory/ antifibrotic eicosanoids, whereas the other is associated with more generalized suppression of inflammatory mediators. Corticosteroid induction of apoptotic cell clearance in vivo may be yet another mechanism for the therapeutic effect of corticosteroids. Finally, AMfs from patients with asthma may have a suppressed response to LPS, manifested by suppressed secretions of inflammatory mediators in all patients with asthma and the defective apoptotic cell uptake seen in patients with severe asthma. The functional consequence of these suppressed responses in AMfs from patients with asthma may be disadvantageous to resolution of inflammation (31, 32). By failing to remove dying cells, with the adjunctive loss of 15-HETE and PGE^sub 2^ responses, AMfs could potentially contribute to chronic inflammation, abnormal remodeling, and bronchospasm in asthma.

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

© 2005 American Thoracic Society Provided by ProQuest LLC. All Rights Reserved.