Twenty-four asthmatic current smokers (Sm) with a history of smoking of at least 5 pack-years, 21 asthmatic never-smokers (NSm), and 10 healthy volunteer never-smokers (control), all aged 18–65 years, were recruited from hospital clinics, general practice, and by advertisement. Asthma was diagnosed according to American Thoracic Society criteria (23). All patients with asthma showed bronchodilator reversibility of FEV₁ exceeding 15% or peak expiratory flow rate variability greater than 20%. Subject demography is shown in Table 1. Patients with asthma had not received medication other than β-agonists alone or β-agonists and inhaled corticosteroids for at least the last 3 months. The two groups of patients with asthma were subdivided into four subgroups: smoker asthmatics, steroid-naive (SmC^–, n = 9), smoker asthmatics, steroid-treated (SmC⁺; n = 15), never-smoker asthmatics, steroid naive (NSmC^–, n = 6), and never-smoker asthmatics, steroid-treated (NSmC⁺, n = 15) (Table 1). All patients were clinically stable. No subject had reported a respiratory infection or an exacerbation for at least 6 weeks before the study. All subjects gave written informed consent and the study was approved by the local ethics committee.

Spirometry and tests of airway reversibility were first performed, exhaled carbon monoxide (CO) was measured, and medical history was taken. Total IgE blood levels were measured. On a subsequent visit bronchoscopy was performed. Smokers were advised to refrain from smoking for at least 8 hours before bronchoscopy. At least two biopsy specimens were taken from the subsegmental carinae of the right lower lobe, immediately fixed, and processed to paraffin wax. For further details, see the online supplement.

Immunohistochemistry was applied to identify Langerhans cells (CD1a⁺), mature DCs (CD83⁺), B cells (CD20⁺), the Th1 cytokine IFN-γ and the Th2 cytokine IL-4. Validation experiments were performed on paraffin-embedded lung surgical resection specimens, on lavaged lung macrophages, and on a lung digest–derived cell population enriched for DCs by adherence (24). For further details see the online supplement.

Sections were coded and counted by a blinded observer (M.T.). At least two bronchial biopsies for each subject were measured and counted and the average of the counts for each subject was used. Data for cell counts were expressed as the number of cut cell profiles with a nucleus visible (i.e., positive cells) per square millimeter of the epithelium or subepithelium. For further details see the online supplement.

The unpaired Student t test was used for analyses of normally distributed variables and the Mann-Whitney U test was used for nonnormally distributed variables. The Spearman rho correlation coefficient was used for correlations. p ⩽ 0.05 was accepted as significantly different. A statistical software package (StatsDirect; Camcode, Cambridge, UK) was used. For further details see the online supplement.

No significant differences were found for CD83⁺ cells, CD20⁺ cells, or IL-4⁺ cells between steroid-treated and steroid-naive patients with asthma (Table 2). Steroid-treated never-smoker patients with asthma had significantly fewer CD1a⁺ cells and a lower IFN-γ:IL-4 ratio compared with steroid-naive never-smoker patients with asthma (Figure 4). The difference was not observed when steroid-treated and steroid-naive smoker patients with asthma were compared (Figure 4).

Figure 4. Significantly decreased CD1a⁺ Langerhans cell numbers in the epithelium and decreased IFN-γ–positive cell numbers and IFN-γ:IL-4 ratio in the subepithelium of steroid-treated versus steroid-naive never-smoker patients with asthma (NSmC⁺ vs. NSmC^–), but not in steroid-treated versus steroid-naive smoker patients with asthma (SmC⁺ vs. SmC^–). Results are expressed as numbers of cells per square millimeter of epithelial or subepithelial area and for IFN-γ:IL-4 ratio as absolute numbers.

Download Figure | Download Powerpoint

[More] [Minimize]

A significant correlation was observed between the numbers of CD83⁺ mature DCs and CD20⁺ B lymphocytes in steroid-naive patients with asthma (r = 0.58, p = 0.024) but not in steroid-treated patients with asthma. No correlation was detected between CD83⁺ mature DCs and Th1 or Th2 cytokine–expressing cells.

There are two likely explanations for the observed reduction of CD83⁺ mature DC numbers in the airways of asthmatic smokers: (1) downregulation of maturation markers with consequent loss of their detection by immunostaining and/or (2) chemokine receptor switching and resultant migration of mature DCs away from the mucosa to local lymph nodes. It has been shown that “conditioning” of DCs with cigarette smoke extract in vitro suppresses induction of maturation-associated markers and DC-induced T-cell priming (15). Although nicotine per se can suppress the capacity of DCs to induce T-lymphocyte responses (13, 14), it appears that the effects of cigarette smoke extract on DC maturation are not mediated by nicotine exclusively (15). Further, Robbins and colleagues showed that exposure of mice to cigarette smoke results in decreased numbers of lung DCs and decreased expression of the maturation marker B7.1 (19). Yet, others demonstrate that pulmonary DCs increase in number when mice are exposed to cigarette smoke (16, 17). These divergent results could be attributed to differences in experimental design or dose and duration of cigarette smoke exposure. The present study provides novel findings concerning pulmonary DCs in human smoker and never-smoker patients with asthma and our results support the hypothesis that, in asthma, cigarette smoke reduces the numbers of cells expressing the CD83 maturation marker.

DC-related alterations are increasingly recognized as an important mechanism associated with decreased immunity in a number of diseases (7, 8). Although the consequences of reduced bronchial CD83⁺ mature DC numbers on airway immunity are unclear we may speculate that this may alter the response to respiratory tract infection. This, in patients with asthma who smoke, may partly explain the observed association between smoking and increases in exacerbation frequency and the tendency of asthmatic smokers to accelerated decline in lung function (1). In support of this suggestion, frequent lower respiratory tract infections in current smokers with chronic obstructive pulmonary disease promote long-term decline of lung function (25). Loss of DCs might also alter the balance of inflammatory cell phenotypes, resulting in a pattern more similar to that of chronic obstructive pulmonary disease and one that is less responsive to steroid treatment. Indeed, increased sputum neutrophils and decreased eosinophils have been reported in patients with asthma who smoke (6). We are currently trying to determine whether such changes in the patterns of inflammation occur in airway tissue of asthmatic smokers.

In contrast to CD83⁺ mature DCs, the number of Langerhans cells did not differ between smoker and never-smoker patients with asthma. In this regard, two previous studies have shown greater numbers of Langerhans cells in the small airways and lung parenchyma, and in the bronchoalveolar lavage (BAL) fluid of healthy smokers or smokers undergoing lung resection for tumor (26, 27). This apparent discrepancy between our and earlier results could be due to differences in the comparator subject groups or to the lesser frequency of Langerhans cells in the bronchial epithelium of the large airways as compared with the small airways, sampled by BAL.

Mature DCs have major effects on B-cell growth and immunoglobulin secretion (8) whereas B cells, in turn, are critically involved in the process of DC maturation (28). Previous studies have reported lower than normal serum levels of immunoglobulin in smokers (7). Moreover, experimental models have shown that cigarette smoke may cause B-cell dysfunction (7). Accordingly, we have shown in situ that there are fewer B lymphocytes in the airways of asthmatic smokers compared with either asthmatic never-smokers or healthy never-smokers. A significant positive correlation was also observed between numbers of B cells and numbers of CD83⁺ mature DCs in steroid-naive smoker and never-smoker patients with asthma, which could be biologically relevant. However, we acknowledge that such a correlation is not, of itself, proof of this.

During the development of adaptive immunity, the phenotype and function of DCs play an important role in influencing Th1 or Th2 type differentiation. Having found fewer mature DCs and B cells in asthmatic smokers, it might also be expected that the balance of Th1 and Th2 cytokine expression would be altered. We showed herein a trend to reduced Th1 cytokine expression in steroid-naive smokers. The trend compares favorably with reports from cell studies that cigarette smoke exposure inhibits DC-mediated Th1 priming (13–15). IL-18, another cytokine known to be important in the development of the Th1 response, is also lowered in the sputum of asthmatic smokers (4). Accordingly, epidemiologic studies suggest that the incidence of hypersensitivity pneumonitis, a Th1-mediated hypersensitivity reaction to inhaled allergens, is rare in cigarette smokers (29). These findings suggest that smoking may alter Th-1 immunity in asthma, which may further increase susceptibility to infection (14).

Although Th2 cytokines have been long identified as major contributors to allergy and asthma, studies have shown that Th1 cytokines act concurrently with Th2 cytokines to favor the chronicity of the inflammatory response in asthmatic airways (30). By measuring IFN-γ expression in the airways of never-smoker patients with asthma and healthy subjects we have shown greater Th1 expression and a higher Th1:Th2 ratio in never-smoker patients with asthma as compared with healthy subjects. These data are compatible with the hypothesis that Th1 cytokines act in concert with Th2 to support the chronic immune response in asthma.

Airway inflammation in mild to moderate allergic asthma is characteristically steroid responsive, with inhaled or oral steroids usually resulting in reduction of both airway inflammation and symptoms. Langerhans cells have been shown previously to decrease after inhaled steroid treatment (12) and our findings of fewer Langerhans cells in steroid-treated versus steroid-naive never-smoker patients with asthma are consistent with this. However, in a subgroup analysis a similar reduction was not observed in CD83⁺ mature DCs in steroid-treated versus steroid-naive never-smoker patients with asthma of our study. More steroid-naive and steroid-treated never-smokers need to be examined before definitive conclusions about the effect of steroids on bronchial mature DC numbers can be made.

IFN-γ expression and the IFN-γ:IL-4 ratio were also found to be reduced in steroid-treated never-smoker patients with asthma, adding to the in vitro evidence that glucocorticoids inhibit the production of IFN-γ (31). Thus both reduced Langerhans cells and Th1 immunity may contribute to the therapeutic effects of inhaled corticosteroids in asthma.

In contrast to steroid-treated versus steroid-naive never-smoker patients with asthma, Langerhans cells, IFN-γ–expressing cells, and the IFN-γ:IL-4 ratio were not reduced in steroid-treated versus steroid-naive smoker patients with asthma. Several studies have suggested that the efficacy of corticosteroids is reduced in patients with asthma who are active cigarette smokers (1, 2, 6). A reduced ratio of glucocorticoid active receptor-α to its nonfunctioning β isoform and decreased histone deacetylase activity have been highlighted as mechanisms of smoking-induced steroid resistance (32, 33). Considering that Langerhans cells and Th1 immunity have important roles in asthma pathogenesis, resistance to reduction in airway Langerhans cells and Th1 immune response might be a contributing mechanism that explains the reduced response to the antiasthma effects of inhaled corticosteroid therapy in smokers with asthma. Moreover, as steroids normally act, in part, via suppression of DC maturation, the reduced numbers of CD83⁺ mature DCs that we report in patients with asthma who smoke could fundamentally alter their response to steroid treatment (11).

We acknowledge there are limitations to the interpretation of our results and that our study is descriptive. It is, however, quantitative and provides novel observational data that address an important, clinically relevant question concerning which limited data are available. First, our findings provide a basis for a potential mechanism to explain the clinical observations that have been made in patients with asthma who smoke. Second, there may have been confounding effects due to steroid therapy because a significant number of our subjects with asthma were on regular inhaled steroid-treatment. However, when steroid-treated patients with asthma were compared with steroid-naive patients, no significant differences were demonstrated in either CD83⁺ mature DCs or B-cell numbers. Moreover, when steroid-naive asthmatic smokers were compared independently with steroid-naive asthmatic never-smokers, again CD83⁺ mature DCs and B lymphocytes were found to be fewer in smokers. Third, differences in atopic status between our subjects might have influenced the numbers of mature DCs and B cells. Although total IgE blood levels were lower in the control group relative to the asthma groups, they did not differ between smoker and never-smoker patients with asthma (Table 1). Moreover, no difference was detected in CD83⁺ mature DCs and B cells between patients with asthma and control subjects with high or normal total IgE blood levels and no correlation was found between total IgE blood levels and CD83⁺ mature DC and B-cell numbers in either group (data not shown). Fourth, it could be argued that smoking immediately before bronchoscopy might have affected our findings (34). However, this is unlikely because all our smokers were advised to refrain from smoking for at least 8 hours before bronchoscopy. Finally, although our results showed a reduction in CD83⁺ mature DCs and B cells in asthmatic smokers, it is unclear whether the reduction was due to smoking alone or to an interaction between smoking and asthma. As smoking is implicated in the pathogenesis of a number of pulmonary conditions, further studies are needed to investigate whether smoke alone can induce such changes or whether its association with an asthma phenotype is critical to our observations.

In conclusion, we provide the first in vivo evidence that cigarette smoke exposure alters bronchial mucosal immunity in the human asthmatic airway. We speculate that cigarette smoking may mediate some of its adverse effects in asthma by altering airway immunity by (1) reduction of the numbers of CD83⁺ mature DCs in the bronchial mucosa or (2) reduction of B-cell numbers, which may impact on B-lymphocyte–mediated humoral immunity, or by (3) reduction of Th1 cytokine expression, which may affect cell-mediated immune responses.