Defective macrophage phagocytosis of bacteria in COPD

Subject selection

Healthy subjects and smokers were recruited from the National Heart and Lung Institute, Royal Brompton Hospital, London (UK). COPD subjects were recruited from the Royal Free Hospital, London or Wexham Park Hospital, Slough (UK). All subjects gave written informed consent as approved by the Royal Brompton and Harefield NHS Trust Ethics Committee (London, UK). BALF was obtained from consenting patients at St. Mary's Hospital, London (UK) or Wexham Park Hospital. Demographic data are presented in table 1⇓. COPD patients were significantly older than the control groups but there were no differences in smoking history with the smoking controls (table 1⇓).

Cell culture

Monocytes were isolated from PBMC using a Monocyte Isolation kit II (Miltenyi Biotec, Surrey, UK) and cultured in the presence of 2 ng·mL⁻¹ granulocyte-macrophage colony-stimulating factor (GM-CSF) for 12 days to generate MDM as described previously 15. Alveolar macrophages were isolated from BALF as described previously 16.

Phagocytosis assays

Non-typeable H. influenzae strain 1479 and S. pneumoniae serotype 9V strain 10692 were cultured and heat killed at 60°C for 2 h. Bacteria were fluorescently labelled using Alexafluor 488 dye (2 mg·mL⁻¹ in dimethyl sulfoxide) in the dark, at room temperature overnight. The labelled bacteria were washed repeatedly in PBS to remove unbound label and resuspended in PBS. For alternative experiments, live bacteria were resuspended in broth containing CellTracker Red CMPTX dye (12.5 μM) (Molecular Probes, Invitrogen Ltd, Paisley, UK) and incubated in the dark at 37°C for 45 min. The labelled bacteria were washed in Dulbecco’s-PBS (D-PBS), until the free dye was removed, and stored at -20°C. Fluorescently labelled polystyrene beads or bacteria were added to cells and incubated for the times indicated. Cells were washed with D-PBS and fluorescence of extracellular particles was quenched by adding Trypan blue (2% (volume/volume)) for 1 min. Excess fluid was removed and fluorescence determined for beads and heat-killed bacteria using an excitation λ 480 nm and emission λ 520 nm. For experiments using live bacteria, MDM were exposed for 4 h prior to measurement in a plate-reader using an excitation λ 570 nm and emission λ 610 nm. Initial experiments and pharmacology experiments with E. coli were performed using the Vybrant Phagocytosis kit (Invitrogen Ltd) according to the manufacturer's instructions. Data are presented as phagocytosis relative to the fluorescence of each specific bacterium to account for differences in labelling. In order to confirm the labelling procedure was not altering the ability of cells to phagocytose, E. coli (strain K12; Sigma, Poole, UK) were labelled under identical conditions and used in comparative experiments. Alternatively, following exposure of cells to bacteria as described above, cells were removed from the plate by agitation and fluorescence measured using a flow cytometer.

Confocal microscopy

Macrophages were cultured on Lab-tek Permanox (Nunc, Loughborough, UK) chamber slides and fixed with 4% (weight/volume) paraformaldehyde. Nuclei were stained with DAPI (4',6-diamidino-2-phenylindole) and cell cytoplasm by incubation with Evans blue dye (0.1% (w/v)). Slides were viewed on a Leica TCS 4D Confocal microscope with a Krypton-Argon laser to detect fluorescence of the FITC (fluorescein isothiocyanate)/yellow–green/Alexafluor 488 fluorochromes. Images of the three stains (DAPI (blue), Evans Blue (red) and the fluorochromes (green)) were overlaid.

Electron microscopy

MDM were incubated with H. influenzae (1 mg·mL⁻¹, for 1 h) and fixed by addition of glutaraldehyde. Secondary fixation was performed with osmium tetroxide. Samples were then suspended in molten 2% (w/v) agar and dehydrated by incubation with increasing concentrations of methanol. Cells were infiltrated with propylene oxide and Araldite resin, followed by incubation and embedding in Araldite alone. Ultrathin sections were then cut, mounted on copper support grids and stained using uranyl acetate and Sato's lead citrate. Grids were transferred to the TEM for visualisation.

Flow cytometric analysis of macrophage receptor expression

MDM were removed from the cell culture plates using non-enzymatic cell dissociation solution (Sigma), washed in PBS containing 0.5% (w/v) bovine serum albumin (BSA) and 0.1% (w/v) sodium azide (PAB) and then resuspended at a concentration of 1×10⁶ mL⁻¹. MDM suspension (180 μL) was incubated for 1 h on ice with 20 μL of either the appropriate mouse isotype control antibody, or mouse monoclonal antibody against cluster of differentiation (CD)14, CD163, CD36, mannose receptor, macrophage receptor with collagenous structure (MARCO), phosphatidylserine (PS) receptor, human leukocyte antigen-DR (HLA-DR), toll-like receptor (TLR)2 and TLR4, all at 50 μg·mL⁻¹. The cells were washed twice with PAB and a F(ab)'₂ fragment of a phycoerythrin (PE)-labelled goat anti-mouse immunoglobulin (Ig)G (20 μg·mL⁻¹) (Dako Ltd, Ely, UK) was incubated with the cells for 30 min on ice. The cells were washed with PAB and finally resuspended in 300 μL of FACSFlow^TM (Becton Dickinson, Oxon, UK) containing 0.5% (v/v) formaldehyde prior to analysis on a FACScan cytometer (Becton Dickinson). 5,000 events were acquired and the fluorescence staining of the MDM was assessed at 575 nm. The specific mean fluorescence (SMF) values were calculated as the fold difference of the mean fluorescence intensity (MFI) of the isotype control to the MFI of the test antibody.

Statistical analysis

Data are presented as mean±sem for n observations. Comparisons between subject groups or cell types were performed using Kruskal–Wallis analysis using GraphPad Prism software followed by Dunn's multiple comparison test (GraphPad Software Inc., San Diego, CA, USA) or a Mann–Whitney test where appropriate. Differences were considered significant where p<0.05.

Comparison of alveolar macrophage and MDM phagocytosis

Initial experiments demonstrated that phagocytosis of polystyrene beads by both alveolar macrophages and MDM were similar (fig. 1a⇓). Moreover, MDM from nonsmokers, smokers and COPD patients phagocytosed beads equally (fig. 1b⇓) which was confirmed as internalisation by confocal microscopy (fig. 1c⇓). This would indicate that the MDM model reflects alveolar macrophage phagocytosis and that inert particle removal is not altered in COPD. In order to ascertain whether similar responses were observed when macrophages were exposed to bacteria, a series of experiments were devised. Alveolar macrophages from COPD patients phagocytosed less E. coli compared with cells from nonsmokers and smokers (fig. 2a⇓). By contrast, monocytes from each of the subject groups had low phagocytic responses that did not differ with either smoking or disease status (fig. 2b⇓). However, MDM from COPD patients exhibited a reduced capacity to ingest bacteria similar to that of alveolar macrophages (fig. 2c⇓). We confirmed this observation using fluorescence-activated cell sorting (FACS) analysis. Phagocytosis by MDM from nonsmokers and smokers were similar (figs 2d⇓ and e) but cells from COPD patients exhibited a significant curve shift to the left (fig. 2f⇓), indicating that fewer cells had engulfed bacteria. These comparative phagocytosis experiments suggest that, in COPD, defective macrophage phagocytosis is acquired during differentiation and that MDM may be useful for studying the underlying defective mechanism.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1—

Comparison of alveolar macrophage and monocyte-derived macrophages (MDM) phagocytosis of beads. a) Alveolar macrophages (▓), n = 7 and MDM (▪) n = 7, were exposed to fluorescent beads for 6 h. Phagocytosis was measured using a fluorometric plate reader. b) Comparison of phagocytic capacity of MDM from nonsmokers (NS; □; n = 9), smokers (S; ░; n = 13) and chronic obstructive pulmonary disease (COPD) patients (▒; n = 10), following exposure to 50×10⁶ beads·mL⁻¹ for 6 h. Data are presented as mean±sem. c) Confocal micrograph of MDM engulfing fluorescent beads (50×10⁶ beads·mL⁻¹).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2—

Phagocytosis of Escherichia coli by a) alveolar macrophages, b) monocytes and c) monocyte-derived macrophages (MDM) from nonsmokers (NS; □), smokers (S; ░) and chronic obstructive pulmonary disease (COPD) patients (▒) following exposure to 1 mg·mL⁻¹ FITC-E. coli for 1 h. a) NS: n = 5; S: n = 5; COPD: n = 4. b) NS: n = 4; S: n = 4; COPD: n = 6. c) NS: n = 7; S: n = 6; COPD: n = 8. Data are presented as median±interquartile range. *: p<0.05. The phagocytic response of MDM to fluorescently labelled E. coli was also determined by flow cytometry for d) NS, e) S and f) COPD patients. Data are presented as a representative fluorescence-activated cell sorting histogram, in which the grey lines are cells in the absence of bacteria and the purple lines are cells in the presence of bacteria for at least four independent experiments.

MDM phagocytosis of H. influenzae and S. pneumoniae

To investigate whether this reduced phagocytic response to bacteria was restricted to E. coli the phagocytic responses of MDM to fluorescently-labelled H. influenzae, S. pneumoniae and E. coli were evaluated. All MDM phagocytosed the three bacterial strains in a concentration-dependent manner (figs 3a–c⇓) but the response of COPD MDM was attenuated, not only towards E. coli (fig. 3a⇓) but also for S. pneumoniae (fig. 3b⇓) and H. influenzae (fig. 2c⇑). To investigate whether defective phagocytosis in COPD MDM was an artefact of using heat-killed bacteria, experiments were devised using labelled, live bacteria. Under these conditions, MDM from control subjects phagocytosed increased H. influenzae and S. pneumoniae (p<0.05) compared with cells from COPD patients (H. influenzae: 3587±390 versus 2170±166 fluorescence units for control and COPD, respectively, n = 9; S. pneumoniae: 3885±344 versus 2514±288 fluorescence units for control and COPD, respectively, n = 9). Confocal microscopy confirmed internalisation of S. pneumoniae (fig. 3d⇓) and H. influenzae (fig. 3e⇓), and electron microscopy confirmed bacterial ingestion and formation of phagosomes (fig. 4⇓) indicating that MDM from COPD patients have the capacity to mount correct phagocytic responses, albeit attenuated.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3—

Phagocytic responses of monocyte-derived macrophages (MDM) from nonsmokers (□), smokers (░) and chronic obstructive pulmonary disease (COPD) patients (▒) to pathogenic bacteria. MDM were generated from nonsmokers (n = 15–16), smokers (n = 13–14), and COPD patients (n = 16–17), and exposed to increasing concentrations of fluorescently labelled a) Escherichia coli, b) Streptococcus pneumoniae and c) Haemophillus influenzae for 1 h. Phagocytosis was measured using a fluorometric plate reader. Data are presented as median±interquartile range. *: p<0.05; **: p<0.01. Internalisation of particles was confirmed using confocal microscopy. Representative orthogonal Z-stack views of phagocytosed d) S. pneumoniae and e) H. influenzae respectively.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4—

Electron microscopy images of monocyte-derived macrophages (MDM) phagocytosis of Haemophillus influenzae. MDM were generated from chronic obstructive pulmonary disease (COPD) patients and incubated with 1 mg·mL⁻¹ H. influenzae for 1 h. Cells were then fixed, dehydrated and embedded in Araldite resin. Ultrathin slices were cut and stained, then viewed on a transmission electron microscope. Arrows indicate bacteria. a) A bacterium on the surface of an MDM; b and c) formation of phagocytic cups around bacteria; d and e) this is proceeded by complete encapsulation by pseudopodia; and f) the bacteria are then transported into the cell cytoplasm within a phagosome. Scale bars = 500 nm.

This reduced clearance of bacterial pathogens by macrophages may account for increased infections and the concomitant decline in lung function observed in COPD. Therefore, we next examined the relationship between lung function parameters and phagocytosis. Phagocytic responses of MDM to E. coli, H. influenzae or S.pneumoniae did not correlate with forced expiratory volume in 1 s (FEV₁) % predicted (r = 0.07, r =  -0.05 and r = 0.12, respectively). Similarly, the FEV₁/FVC ratio did not correlate with phagocytosis of E. coli, H. influenzae or S.pneumoniae (r =  -0.12, r =  -0.10 and r = 0.32, respectively). In addition, there was no relationship between phagocytic response and current smoking status. Despite COPD patients having smoked more cigarettes than smokers and being older (see table 1⇑) there were no correlations between either age or number of cigarettes smoked (pack-yrs) and phagocytosis of any bacteria examined.

Cell surface expression of receptors involved in bacterial recognition

There are numerous receptors involved in bacterial recognition by macrophages 17, 18, thus reduced receptor expression in COPD could account for attenuated phagocytosis. However, there were no differences in the expression of TLR2, CD14, TLR4, or CD163 on MDM from COPD patients compared with cells from the other subject groups (table 2⇓). There were also no differences in expression of non-specific scavenger receptors including the mannose receptor, MARCO or CD36 (table 2⇓). Nor were there any differences in expression of the PS receptor or HLA-DR on these cells (table 2⇓).

MDM from COPD patients phagocytosed beads normally (fig. 1b⇑) limiting defective phagocytosis to engulfment of pathogenic bacteria. Since there were no differences in expression of a number of receptors considered important in recognition of non-opsonised particles (table 2⇑), the downstream mechanisms emanating from receptor ligation that regulate phagocytic responses were examined. Pharmacological modulation of these pathways was utlilised to investigate whether altered signal transduction could account for these observations. The phosphoinositol-3-kinase inhibitor, LY-294002 and the p38 inhbitior, SB 203580, inhibited MDM phagocytosis of E. coli from all subjects similarly (figs S1a and c in the supplementary material). By contrast, the Rho kinase inhibitor, Y-27632, had no effect on phagocytosis at any of the concentrations tested (fig. S1b in the supplementary material).

Effect of pharmacological agents on MDM phagocytosis

COPD patients are currently taking a variety of medicaments to ameliorate their symptoms. These include glucocorticosteroids and bronchodilators. Therefore, it was possible that systemic effects of these drugs could alter the responses of blood-derived macrophages. To test this possibility, MDM from nonsmokers, smokers and patients with COPD were pre-treated with various pharmacological agents prior to phagocytosis assay (fig. 5⇓).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5—

Effects of pharmacological agents on phagocytosis of bacteria by monocyte-derived macrophages (MDM). MDM from nonsmokers (▪), smokers (▴) and COPD patients (•) were pre-treated with budesonide (a and d) or formoterol (b and e) for 1 h, or azithromycin (c and f) for 24 h, prior to exposure of the cells to either 1 mg·mL⁻¹ Haemophillus influenzae (a–c) or Streptococcus pneumoniae (d–f) for 1 h. Data were normalised to the phagocytosis response of each cell type in the absence of drug (100%). Data are presented as mean±sem (n = 5–6).

Exposure of MDM to budesonide improved the phagocytic responses of cells from COPD patients and smokers towards both H. influenzae and S. pneumoniae (figs 5a⇑ and d). Agents that elevate cAMP are considered to be inhibitory for phagocytosis 19; however, formoterol did not inhibit phagocytosis of the bacteria in this system (figs 5b⇑ and e) but stimulated the response of cells from smokers. There was no effect on the responses of cells derived from patients with COPD. The macrolide, azithromycin, is reported to restore the phagocytic response of alveolar macrophages from COPD patients and enable removal of apoptotic cells 20; however, in this system, azithromycin had no effect on the response of cells from patients with COPD but improved the responses of cells from smokers (figs 5c⇑ and f). Other drugs (tiotropium bromide (10⁻⁹–10⁻⁶ M) or theophylline (10⁻⁸–10⁻⁵M)) prescribed commonly to patients with COPD had no effect on the phagocytic response (data not shown).