Published ahead of print on July 13, 2007, doi:10.1165/rcmb.2007-0025OC
© 2007 American Thoracic Society DOI: 10.1165/rcmb.2007-0025OC Smoking Alters Alveolar Macrophage Recognition and Phagocytic AbilityImplications in Chronic Obstructive Pulmonary Disease1 Department of Thoracic Medicine, Royal Adelaide Hospital and Lung Research Laboratory, Hanson Institute, Adelaide, South Australia Correspondence and requests for reprints should be addressed to Dr. Sandra Hodge, Lung Research, Hanson Institute, Frome Rd, Adelaide, South Australia 5001. E-mail: sandy.hodge{at}imvs.sa.gov.au
Chronic obstructive pulmonary disease (COPD) is associated with defective efferocytosis (apoptosis and alveolar macrophage [AM] phagocytic function) that may lead to secondary necrosis and tissue damage. We investigated ex vivo AM phagocytic ability and recognition molecules (CD36, integrin  V 3, CD31, CD91, CD44) using flow cytometry. The transferrin receptor (CD71) was measured as an indicator of monocyte-macrophage differentiation in bronchoalveolar lavage (BAL). Proliferation was assessed with Ki-67. Based on evidence of systemic involvement in COPD, blood from 17 current smokers and 25 ex-smokers with COPD, 22 healthy smokers, and 20 never-smoking control subjects was also investigated. BAL was collected from 10 to 16 subjects in each group. Levels of recognition molecules and cAMP were assessed after exposure of AM to cigarette smoke in vitro. The phagocytic ability of AM was significantly decreased in both COPD groups and in healthy smokers compared with control subjects. However, phagocytic capacity was better in subjects with COPD who had ceased smoking, compared with those who were still smoking. AM from current smokers with COPD and healthy smokers exhibited reduced CD31, CD91, CD44, and CD71, and enhanced Ki-67 compared with healthy never-smoker control subjects. There were no differences in these markers in AM from ex-smokers with COPD compared with control subjects, or in blood monocytes from any group. Suppressive effects of cigarette smoke on AM recognition molecules associated with an increase in cAMP were confirmed in vitro. Our data indicates that a smoking-related reduction in AM phagocytic ability and expression of several important recognition molecules may be at least partially normalized in those subjects with COPD who have ceased smoking.
Key Words: cigarette smoke • chronic obstructive pulmonary disease • alveolar macrophage • phagocytosis • cAMP
Chronic obstructive pulmonary disease (COPD) is an incapacitating, highly prevalent airway disease that arises as a result of noxious injury to the lungs, most commonly due to cigarette smoking. Increasing evidence implicates alveolar macrophages (AM) as important mediators of the inflammatory response in COPD. These blood monocyte–derived effector cells can interact with resident cells and extracellular matrix to generate a proinflammatory response that includes increased oxidative stress, increased production of proinflammatory cytokines and up-regulation of antigen presentation and costimulatory molecules. These mediators have the potential to amplify the tissue injury and chronic inflammatory response that are characteristic of COPD (1). In this regard, our previous studies have also established the presence of failed efferocytosis (2) (defective AM phagocytic function with dysregulated apoptosis) in COPD (3, 4). The link between these changes, tissue damage, and inflammation in the lung has been previously demonstrated (5). However, the precise molecular basis of the defective phagocytic ability of AM, especially with regard to the effects of cigarette smoking, in COPD is unknown.
The interaction between the AM and apoptotic cells is mediated by a variety of macrophage membrane-associated proteins. Several of these recognition molecules on AM include thrombospondin receptor (CD36), integrin Cigarette smoking has been reported to cause changes to AM phenotype (12). Whether these changes are reversed by cessation of cigarette smoking in COPD remains unknown, although our recent evidence suggests that cessation of smoking does not prevent the increased rates of epithelial cell apoptosis (3) or increased T cell levels of granzyme b and perforin in COPD (13). In the present study, we therefore examined the effects of smoking on AM phagocytic ability and associated recognition molecules in COPD by investigating current smokers with COPD and ex-smokers with COPD, healthy smokers, and healthy never-smoker control subjects. Based on increasing evidence of systemic involvement in COPD (14), recognition molecules on peripheral blood–derived monocytes were also investigated.
Immunologic Reagents The following monoclonal antibodies (Mabs) and immunologic reagents were employed. Epithelial cell antigen (Dako, Glostrup, Denmark); CD51[FITC], CD36[FITC], CD14[PE-CY5], and CD33[PE-CY5] (Immunotech/Coulter, Marseille, France); CD33[FITC], CD45[PE], CD3[FITC], CD3[PE], CD44[FITC], CD71[FITC], CD31[PE], CD61[FITC], and Ki-67[FITC] (BD Biosciences [BD], Sydney, Australia); and CD91[PE] (Serotec, Oxford, UK). IgG1/IgG1[FITC/PE] (BD) was employed as a negative control. Red blood cell lysing agent (FACSlyse) and cell membrane permeabilizing agent (FACSperm) were obtained from BD. For the phagocytosis assay, mitotracker red (MTR; Molecular Probes, Eugene, OR) was employed. Annexin V (BD) and 7 aminoactinomycin D (7AAD) were used to investigate apoptosis.
Subject Population
Bronchoscopy Procedure BAL was obtained via flexible bronchoscopy, as we have previously described (16). Briefly, 50 ml of sterile normal saline (at room temperature) was instilled into the airway with a syringe then aspirated using low suction. Two further 50-ml aliquots of saline were instilled and aspirated in the same way. The first aliquot was processed for microbiological testing. For each collection from an individual patient the aspirated BAL specimens 2 and 3 were pooled, kept on ice, and processed within one hour of collection (the first aliquot was not included to avoid airway mucus contamination).
Preparation of Ex Vivo Samples Absolute numbers of lymphocytes, AM, and neutrophils were calculated from total leukocyte counts. Venous blood was collected into tubes containing 10 U/ml preservative-free sodium heparin (DBL, Sydney, Australia). Blood films were stained by May-Grünwald-Giemsa using an automated staining machine (Shandon Veristat Southern Products, Astmore, UK) and differential cell counts were performed using a CELL DYN 4,000 (Abbott Diagnostics, Sydney, Australia).
Phagocytosis Assay Staining with Annexin V (BD) was performed to ensure that greater than 80% of cells were apoptotic. Apoptotic epithelial cells were stained with 50 µl mitotracker red (MTR) (25 µg/ml), then 1 ml aliquots (4 x 106 cells/ml) were added to AM monolayers (cell count adjusted to 4 x 105 cells/ml) in 24-well plates for 45 minutes at 37°C/5% CO2. The fluid was gently removed from the adherent AM monolayer and cells exposed to 1 ml ice-cold 0.5 mM EDTA for 15 minutes. The cells were pelleted by centrifugation without further washing. AM were then stained with 5 µl CD33[PE-CY5] (Immunotech) for 15 minutes, then washed. Autofluorescence of AM and opsonized but not phagocytosed cells was quenched with crystal violet (0.8 mg/ml for 30 s) and CD33+/MTG+ events quantified using flow cytometry. In our experience the quenching procedure is near 100% efficient. Control tubes of AM only, labeled epithelial cells only, and AM + unlabeled epithelial cells were included, and their staining patterns were used to set quadrant markers for flow cytometric analysis of the percentage of AM that had ingested apoptotic cells.
Staining of Membrane-Associated Molecules on BAL AM
Intracellular Staining with Proliferation Marker Ki-67
Preparation of Cigarette Smoke Extract
In Vitro Investigation of the Effects of CSE on AM Recognition Molecules and Apoptosis
In Vitro Investigation of the Effects of CSE on AM cAMP Levels
Assessment of Functional Significance of Decreased AM Phagocytic Ability
Statistical Analysis
Patient Demographics and BAL Return The clinical characteristics of all subjects are presented in Table 1. The median age for ex-smokers with COPD was slightly but significantly higher (median, 63; range, 49–75) than never-smoker control subjects (median, 56; range, 30–75). There were no significant differences in age between the other groups (Table 1). There was significantly reduced BAL return in healthy smokers and in subjects with COPD (Table 2).
Differential Counts and Absolute Cell Numbers in BAL The total leukocyte count and absolute AM numbers were significantly increased in BAL from both COPD groups and healthy smokers compared with never-smoker control subjects (Table 2).
Differential Counts and Absolute Cell Numbers in Peripheral Blood
Phagocytosis of Apoptotic Bronchial Cells by AM Investigation of phagocytic capacity of adherent AM from both ex-smokers and current smokers with COPD and healthy smokers revealed a significant deficiency in their ability to phagocytose apoptotic airway epithelial cells compared with never-smoker control subjects (Figure 1). However, AM from ex-smokers with COPD had significantly improved phagocytic ability compared with current smokers with COPD (Figure 1).
Membrane-Associated Molecules on BAL AM There is no satisfactory means to determine the dilution factor during BAL (19). Therefore, to compensate for the reduced BAL return in subjects with COPD, the MFI of CD14-positive events, as an indicator of the amount of receptor expression per cell, was calculated. The MFI of staining for CD31, CD91, and CD44 were significantly decreased in current smokers with COPD and in healthy smokers compared with never-smoker control subjects (Figure 2). There were no significant decreases in levels of these markers in subjects with COPD who had ceased smoking, compared with never-smoker control subjects (Figure 2). Expression of CD36, CD51, or CD61 was low (MFI < 40), and no significant differences were noted between groups. The MFI of staining for CD71, a marker of mature AM, was significantly decreased in current smokers with COPD and in healthy smokers compared with never-smoker control subjects (Figure 2). The MFI of CD14 staining, a further indicator of AM differentiation (expressed at a greater level by peripheral blood monocytes compared with mature AM), was significantly increased in current smokers with COPD and in healthy smokers compared with never-smoker control subjects, but not significantly different in ex-smokers with COPD (Figure 2).
A small group of healthy ex-smokers was also investigated (n = 5). These subjects had phagocytic function and expression of recognition molecules that were similar to the never-smoker control subjects (CD31 healthy ex-smoker MFI 221 ± SD4 versus control MFI 200 ± SD101); CD91 healthy ex-smoker MFI 284 ± SD175 versus control MFI 230 ± SD142); CD44 healthy ex-smoker MFI 377 ± SD10 versus control MFI 354 ± SD14); CD71 healthy ex-smoker MFI 114 ± SD28 versus control MFI 123 ± SD74).
The current smokers with COPD were further stratified into two groups based on COPD severity (mild to moderate; Gold I/II [FEV1/FVC < 70% and FEV
Intracellular Staining for AM Proliferation with Ki-67
Membrane-Associated Molecules on Peripheral Blood–Derived Monocytes
CSE Inhibited Expression of Recognition Molecules on AM, In Vitro CSE decreased expression of CD44, CD71, CD31, and CD91 by AM. The inhibitory effects ranged from 2 to 14.5% (with 2.5% CSE) and 11 to 28% (with 10% CSE) (Figure 4). The inhibitory effect of 10% CSE was significant (P < 0.02) for all markers tested; however, the inhibitory effects of 2.5% CSE were significant only for CD31 (P = 0.025). Under the experimental conditions applied, there were no adverse effects of CSE on AM apoptosis, assessed by 7AAD staining (data not shown).
CSE Increased cAMP Levels in AM, In Vitro CSE increased levels of cAMP in AM. cAMP levels ranged from 3.5 to 72.5 pmol/ml (untreated AM), 22 to 147 pmol/ml (with 2.5% CSE), and 33 to 162 pmol/ml (with 10% CSE) (Figure 5). The effect was dose-dependent and significant only at a concentration of 10% CSE (Figure 5).
Assessment of Functional Significance of Decreased AM Phagocytic Ability There was a significant correlation (R-0.65) between the percentage of bronchial epithelial cells and AM phagocytic ability (Figure 6).
Preincubation of AM with Mabs to CD91 or CD31 resulted in a significant reduction in the percentage of AM that had ingested apoptotic epithelial cells (CD91 blockage 20% reduction ± SD 12%; CD31 blockage 21% reduction ± SD 11%).
Failed efferocytosis is being increasingly recognized as playing a significant role in the chronic inflammatory response and tissue damage in the lung in COPD (2). In this regard and consistent with our previous findings (4, 20), the present study found that the ability of AM to phagocytose apoptotic bronchial epithelial cells was significantly decreased in both current and ex-smokers with COPD and in healthy smokers compared with never-smoker control subjects. The use of airway epithelial cells as targets for phagocytosis enabled a physiologically relevant appraisal of phagocytosis in the airways. Interestingly, the present study also noted significantly greater AM phagocytic capacity in subjects with COPD who had ceased smoking, compared with those who were still smoking, suggesting a smoking-related effect on AM that may be partially resolved upon cessation of cigarette smoking. This is further supported by our findings of "normal" phagcoytic ability of AM from healthy ex-smokers. In contrast to these observed differences in AM function, we have previously found that T cells and bronchial epithelial cells from both current and ex- smokers with COPD have similar increased propensity to undergo apoptosis. In addition, we found a direct correlation between CD8+ T cell granzyme b levels and apoptosis of bronchial epithelial cells in the BAL from both current and ex-smokers with COPD (13). It thus appears that there is variability in the response to cigarette smoke by various inflammatory cells, and that other factors (e.g., the persistence of dysregulated apoptosis and secondary necrosis) may at least partially contribute to the reduced AM phagocytic ability that continues in subjects with COPD despite cessation of cigarette smoking. This is supported by a study by Kirkham and colleagues (21), which showed that cigarette smoke as well as carbonyl-modified extracellular matrix proteins decreased the phagocytosis of apoptotic neutrophils in vitro. To investigate whether the reduced phagocytic ability is related to cigarette smoke–induced changes in expression of AM recognition molecules, we analyzed the expression of key receptors on BAL AM. CD91 (LDL receptor–related protein) and CD31 (PECAM) levels were lower for current smokers with COPD and healthy smokers, but unchanged in subjects with COPD who had ceased smoking compared with never-smoker control subjects. Consistent with a study by Pons and coworkers (22), we found no significant changes in levels of CD36, CD51, or CD61 on AM from smokers or subjects with COPD, suggesting that this pathway may not significantly contribute to the deficient AM phagocytic ability noted in COPD. Studies have shown that lung collectins, including surfactant protein A and mannose-binding lectin, can stimulate phagocytosis of apoptotic cells by AM. CD91 is a component of the CD91-calreticulin mannose receptor complex that binds to the collagen-like region of the lung collectins (9). It is therefore likely that these reduced levels of CD91 on AM could lead to decreased recognition of apoptotic cells and subsequent phagocytosis. Similar effects would be expected with reduced levels of CD31, as this cell surface molecule has been reported to promote tethering of apoptotic cells to macrophages (8). The present study also found that AM from current smokers with COPD and from healthy smokers expressed significantly less CD44 than those from never-smoker control subjects. In contrast to our findings, the study by Pons and colleagues (22) found no significant differences in CD44 MFI between control subjects and subjects with COPD, although they did find that the percentage of AM with a CD44low+ phenotype was increased in patients with COPD. The anomalies in MFI between studies may be explained by differences in patient groups tested. The study by Pons and coworkers used control subjects that presented for bronchoscopic evaluation of pulmonary nodule, or hemoptysis. Further, their study did not attempt to divide the subjects with COPD into ex-smoker and current smoker groups. CD44 is the primary AM surface receptor responsible for efficient clearing of excess hyaluron, an extracellular matrix protein that has been shown to be increased in the lungs in COPD and associated with local inflammation and severity of disease (23). In addition, the importance of CD44 as a receptor involved in phagocytosis of apoptotic cells has been demonstrated. Cross-linking of CD44 markedly improved the ability of human monocyte–derived macrophages to phagocytose apoptotic neutrophils (10). It is therefore possible that deficient clearing of hyaluron and/or apoptotic cells as a result of the reduced expression of CD44 by AM from currently smoking subjects with COPD may contribute to the initiation of chronic inflammatory response and tissue damage that are characteristic of the disease.
Taken together, our results suggest that local changes to phagocytic recognition molecules may be directly related to cigarette smoke and that the changes may be normalized after smoking cessation in COPD. To further test this hypothesis, we studied the effects of cigarette smoke on relevant AM recognition molecules in an in vitro model. Interestingly, and consistent with our ex vivo findings, CSE at concentrations that are equivalent to smoking 0.5 to 2 packets of cigarettes in one day (18) significantly decreased expression of CD44, CD71, CD31, and CD91 by AM. As increased concentrations of the second messenger cAMP have been shown to reduce the phagocytic ability of MDM and AM (24, 25), we also assessed the effects of CSE on intracellular cAMP levels in AM. CSE increased AM cAMP levels in a dose-dependent manner, suggesting a smoking-related link between reduced recognition molecules, reduced phagocytic ability, and increased cAMP in AM from smokers with/without COPD. The precise mechanisms for cAMP suppression of AM phagocytic ability is unknown, although elevated cAMP levels have been shown to induce morphologic changes, including disassembly of cytoskeletal elements that are involved in cell/matrix interactions (25). Others have shown that the cAMP-dependent effects were induced by activation of Epac-1 (exchange protein directly activated by cAMP-1), an alternative target to PKA for cAMP signaling (24). Further studies of Epac-1 and the downstream events after Epac-1 activation in COPD are thus warranted. In addition to those investigated herein, the role of other factors that may control phagocytosis in the airway, including extracellular matrix proteins and Class A scavenger receptors, also remain to be elucidated, especially with regard to cigarette smoking status among those patients with COPD. A further possible mechanism may be through the release of TGF- Both monocytes and AM are members of the mononuclear phagocyte system. Blood monocytes mature into AM upon migration from the capillary bed to lung tissue as a response to inflammatory stimuli. During this terminal maturation, phenotypic and functional changes occur, including increased size and phagocytic capacity, decreased ability for T-lymphocyte stimulation, and the expression of maturation-associated markers, including the transferrin receptor, CD71. In the present study, the macrophage/monocyte population (CD14+) in BAL from current smokers with COPD and from healthy smokers exhibited significantly decreased expression of CD71 compared with that from healthy never-smokers. This finding is consistent with that of Lofdahl and colleagues (26); however, the decrease did not reach statistical significance in the latter study, possibly because of the inclusion of both current and ex-smoking subjects in the COPD group. We also found significantly greater expression of CD14 per se from these groups. As undifferentiated macrophages exhibit higher CD14 expression (and lower side scatter) and lower CD71 expression than mature AM, these findings suggest an increase in undifferentiated macrophages in the lungs of these subjects compared with never-smokers and subjects with COPD who had ceased smoking. These smoking-related changes may contribute to reducing the overall rate of phagocytosis in the lung, as undifferentiated macrophages have reduced phagocytic ability compared with AM. It remains unclear whether the increase in these immature cells results from increased influx of peripheral blood–derived monocytes, increased local AM proliferation, or a combination of both factors, although our findings of increased expression of Ki-67 by AM indicate that local AM proliferation may at least play a role. Interestingly, an extensive monocyte influx into the alveolar compartment has been reported in other pulmonary diseases, including acute respiratory distress syndrome and pulmonary histiocytosis (27, 28). The functional significance for COPD of our findings is difficult to fully assess. However, our observation of a significant negative correlation between apoptosis of bronchial epithelial cells and AM phagocytic ability ex vivo suggests at least some degree of functional link between reduced phagocytosis and increased apoptosis in COPD. This concept of defective efferocytosis and its relevance to chronic inflammatory lung disease is being increasingly recognized worldwide, and has been the subject of an interesting review by Vandivier and coworkers (2). To further assess the functional significance of reduced expression of specific AM recognition molecules in COPD, we performed blocking experiments with specific Mabs to the molecules of interest. We have previously reported that ligation of CD44 significantly increased the phagocytic ability of AM by up to 75% (20). In the present study we confirmed that blockage of CD91 and CD31 reduced the phagocytic ability of AM by 20% and 21%, respectively. It thus appears that all of these recognition molecules independently play at least some role in AM phagocytosis of apoptotic cells, and raises the possibility that decreased expression of any of these markers may have a negative effect on AM phagocytic ability. In peripheral blood, increased total leukocyte count in currently smoking subjects with COPD and increased neutrophil numbers in both COPD groups further supports current opinion and our previous data suggesting peripheral involvement in COPD. However, there were no changes in any recognition molecules on peripheral blood monocytes, indicating that the reduced levels of AM recognition molecules are a result of smoking-induced changes in the local lung environment in COPD. Taken together, our data indicate that a smoking-related reduction in AM phagocytic ability, and an increase in the number of undifferentiated monocytes in BAL coupled with reduced expression of several AM recognition molecules, may be at least partially normalized in those subjects with COPD who have ceased smoking. New treatment strategies aimed at increasing the expression of AM recognition molecules may have clinical significance for prevention and treatment of COPD in smokers.
This study was supported by an NHMRC Project Grant, an NHMRC Clinical Career Development Award, an Allen and Hanburys Respiratory Research Fellowship, and a Royal Adelaide Hospital Clinical Project Grant. Originally Published in Press as DOI: 10.1165/rcmb.2007-0025OC on July 13, 2007 Conflict of Interest Statement: S.H. has been previously financed by Boehringer Ingelheim (A$83,000) and Allen Hanbury's (A$125,000). H.J. has participated as a speaker in scientific meetings or courses organized and financed by various pharmaceutical companies (GlaxoSmithKline, Nicomed, AstraZeneca, Boehringer Ingelheim). M.H. has spoken at various educational meetings arranged by Astra, Altana, Pfizer, and GlaxoSmithKline, receiving $14,000 in 2006 and 2007. P.N.R. has received a $50,000 research grant from Pfizer in 2007. None of the other authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form January 28, 2007 Accepted in final form May 31, 2007
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
CLINICAL RELEVANCE
50%; n = 10] and severe; Gold III/IV [FEV1/FVC < 70% and FEV < 50%; n = 7]). There were no statistical differences between the groups for any of the markers tested (P > 0.05 for all markers, data not shown). 
