Published ahead of print on March 4, 2004, doi:10.1164/rccm.200312-1659OC
© 2004 American Thoracic Society
Enhanced Generation of Helper T Type 1 and 2 Chemokines in Allergen-induced AsthmaAllergy and Immunology Section, and Pulmonary and Critical Care Section, Department of Medicine, University of Wisconsin, Madison, Wisconsin Correspondence and requests for reprints should be addressed to Elizabeth A. (Becky) Kelly, Ph.D., Section of Pulmonary and Critical Care Medicine, 600 Highland Avenue, CSC K4/928, University of Wisconsin School of Medicine, Madison, WI 53792-9988. E-mail: eak{at}medicine.wisc.edu
Allergen-induced asthma is characterized by airway eosinophilia and recruitment of helper T (Th) Type 2 lymphocytes. We hypothesized that lymphocyte-associated chemokines contribute to allergen-induced airway inflammation. Sixteen subjects with asthma were phenotyped according to their response to inhaled antigen as single- or dual-phase responders, and then underwent bronchoscopy and segmental allergen bronchoprovocation. Bronchoalveolar lavage fluids were obtained before and 48 hours after segmental challenge with allergen to determine the cellular response and patterns of Th1 and Th2 chemokines and cytokines. Airway cells, cytokines, and lymphocyte-associated chemokines increased after segmental challenge. Th2 chemokines (thymus and activation-regulated chemokine, macrophage-derived chemokine) correlated with airway eosinophils and concentrations of interleukin-5 and -13. In contrast, airway lymphocytes correlated with both Th2 and Th1 (monokine-induced by IFN-  , IFN-–inducible protein-10) chemokines. Notably, when subjects were analyzed according to the presence of a late-phase response, concentrations of both types of lymphocyte-associated chemokines were significantly greater in subjects with a dual-response phenotype. Our findings suggest that both Th2 and Th1 chemokines may be involved in allergen-induced airway inflammation. However, asthma subjects with a dual-responder phenotype have greater generation of chemokines that may lead to enhanced airway inflammation and obstruction after allergen exposure.
Key Words: bronchoalveolar lavage • chemokines • eosinophils • late-phase response • helper T Type 1/helper T Type 2 cells Allergic airway inflammation in asthma involves eosinophils, lymphocytes, cytokines, and chemokines, which contribute to airway obstruction and hyperresponsiveness. Multiple molecules are generated in response to allergen challenge and direct trafficking of inflammatory cells, including activated T cells, into the airway (1). Chemokines can contribute to the extravasation of leukocytes by inducing expression of integrin and adhesion molecules (2). These actions suggest that chemokines and their receptors are important in the generation and regulation of allergic airway inflammation in asthma.
In vitro studies have demonstrated that human lymphocytes highly polarized toward a helper T Type 1 (Th1) or helper T Type 2 (Th2) phenotype selectively express distinct sets of chemokine receptors (3–5). CC chemokine receptor (CCR) 5 and CXC receptor (CXCR) 3 are expressed primarily on Th1 cells. In contrast, CCR3, CCR4, CCR8, and CXCR4 are expressed primarily on Th2 cells. Although expression of these receptors may not be mutually exclusive, it is clear that the majority of polarized Th1 cells express CXCR3 and the majority of Th2 cells express CCR4 (6). Both the Th2 chemokine receptor, CCR4, and the Th1 chemokine receptor, CXCR3, are upregulated in allergic airway disease (1, 7–10). CCR4 recognizes thymus and activation-regulated chemokine (TARC, CCL-17) and macrophage-derived chemokine (MDC, CCL-22) and its expression is elevated in bronchial biopsies, serum, and induced sputum in patients with allergic asthma (8, 11–13) and is further increased after airway antigen challenge (14). However, the expression of CXCR3 ligands monokine-induced by IFN- Airway allergen challenge is associated with eosinophilic inflammation and airway obstruction and is useful as a model to study mechanisms of airway inflammation in asthma. Two approaches have been used to evaluate the effect of antigen on allergic inflammation in the lower airway: aerosol inhalation challenge and localized segmental bronchoprovocation with antigen (SBP-Ag). SBP-Ag provokes a marked localized airway inflammatory response that facilitates analysis of recruited cells, mediator generation, and ex vivo cell function. Localized antigen challenge, however, does not induce measurable changes in lung function. In contrast, after inhaled antigen challenge, the immediate (within minutes) drop in lung function (early-phase response) and the later fall in forced expiratory volume in 1 second (FEV1) between 3 and 8 hours (late-phase response, LPR) can be evaluated. The early response most likely represents the release of mast cell mediators, whereas the later response coincides with an initial influx of inflammatory cells (15). In this study, we combined both approaches to take advantage of their individual strengths. Inhalation challenge was used first to "phenotype" the subjects as single responders (early) or dual responders (early and late). SBP-Ag was then performed to define the characteristics of selective features of airway inflammation. Specifically, SBP-Ag was used to determine the pattern of Th1 and Th2 chemokines in bronchoalveolar lavage (BAL) fluid and to ascertain the relationship between these factors and airway cellular recruitment. Finally, the airway inflammatory response to SBP-Ag was compared between the single and dual responders on the basis of the previously defined pulmonary physiologic reaction to inhaled antigen. (Airway physiology and cellular data from 10 of the 16 subjects have been previously reported as part of a study to evaluate the expression of chemokine receptors on blood and airway eosinophils [7].)
Subjects Sixteen subjects with allergic asthma were studied (Table 1) . Each subject underwent a medical history, physical examination, allergen skin prick test, and pulmonary function testing including spirometry to determine FEV1, reversibility of FEV1 to ß-agonist, and airway hyperresponsiveness to methacholine at the screening visit. Exclusion criteria are included in the online supplement. Informed consent was obtained from each subject before participation. The study was approved by the University of Wisconsin-Madison (Madison, WI) Center for Health Sciences Human Subjects Committee.
Determination of Early- and Late-phase Response to Inhaled Allergen Challenge At least 4 weeks before bronchoscopy, a graded inhaled antigen challenge was performed to determine Ag-PD20 (the dose that caused a 20% fall in FEV1), and to document the presence of early- and late-phase responses. Detailed methods are provided in the online supplement. Briefly, after determining the response to diluent (normal saline), five breaths of antigen were inhaled and spirometry was repeated 10 minutes later. Consecutively higher concentrations of antigen were given until the FEV1 fell by 20% from baseline. The maximum immediate (within 15–30 minutes) drop in FEV1 was determined and then subjects were monitored every 15 minutes until their FEV1 returned to within 10% of their baseline. Thereafter, they were monitored at 1-hour intervals for up to 8 hours to determine the LPR (a sustained 15% fall in FEV1 lasting at least 15 minutes and occurring 3 to 8 hours after challenge).
Segmental Bronchoprovocation and BAL
Chemokine and Cytokine Measurements
Statistical Analysis
Subject Characteristics (See Table 1.) All subjects met American Thoracic Society criteria for the definition of asthma. Twelve of the 16 subjects with asthma had a provocative concentration of methacholine producing a 20% fall in FEV1 (PC20) of less than 8 mg/ml. Of the other four subjects with asthma, one (Subject 8) had episodic asthma symptoms and 14% ß2-agonist reversibility. Subjects 2, 7, and 9 had had physician-diagnosed asthma for more than 10 years, were taking inhaled albuterol as needed, and had previous exercise- or virus-induced asthma symptoms that required treatment with inhaled corticosteroid. All subjects had an immediate (15–30 minutes) fall in FEV1 exceeding 20% and eight subjects had a second fall (LPR) in FEV1 of 15% or more 3–8 hours later.
BAL Cells, Cytokines, and Chemokines after SBP-Ag
Relationship of Th1 and Th2 Chemokines to BAL Eosinophils and Lymphocytes Forty-eight hours after SBP-Ag, strong correlations were found between absolute numbers of airway eosinophils and concentrations of Th2-type chemokines TARC (rs = 0.74, p < 0.001; Figure 1A) and MDC (rs = 0.67, p = 0.004; Figure 1B). TARC and MDC also correlated with the percentage of eosinophils in the BALF (TARC: rs = 0.72, p = 0.001; MDC: rs = 0.58, p = 0.012). In contrast, the presence of BAL lymphocytes was associated not only with Th2-type chemokines (Figures 1C and 1D), but also with Th1-type chemokines IP-10 and MIG (Figures 1E and 1F) (rs > 0.6, p < 0.005 for all chemokines). No significant relationships were noted between BAL lymphocytes, eosinophils, and chemokine concentrations before SBP-Ag.
Relationship between Th2 Chemokines and Th2 Cytokines Forty-eight hours after SBP-Ag, the BALF concentrations of Th2-type chemokines (TARC and MDC), but not Th1-type chemokines, had striking correlations with IL-5 (Figure 2A : rs = 0.90, p < 0.001; and Figure 2B: rs = 0.76, p < 0.001) and IL-13 (Figure 2C: rs = 0.82, p < 0.001; and Figure 2D: rs = 0.67, p = 0.004). These correlations were not observed before SBP-Ag.
Relationship of Lymphocyte-associated Chemokines and the Pattern of Airway Response to Allergen Inhalation To further assess the potential clinical relevance of lymphocyte-associated chemokine generation, we determined BALF chemokine concentrations 48 hours after SBP-Ag in subjects with different patterns of airway response (single versus dual responders) to whole lung inhalation of antigen. Although IP-10, MIG, TARC, and MDC were augmented in BALF obtained from both groups 48 hours after SBP-Ag, the increase in both Th1- and Th2-type chemokines was remarkably greater in asthma subjects with a dual-responder phenotype (Figure 3) . The concentrations of I-TAC were not significantly different between the two groups. After SBP-Ag, the dual-responder group also had increased BAL lymphocytes (7.9 [4.2, 12.1] x 104/ml BALF versus 2.2 [1.7, 2.7] x 104/ml BALF, p < 0.001), a trend toward increased BAL eosinophils (48.0 [20.6, 111.5] x 104/ml BALF versus 7.4 [4.7, 32.3] x 104/ml BALF, p = 0.065), and significantly increased concentrations of IL-5 (56.2 [22.7, 339.9] pg/ml BALF versus 2.9 [0.0, 31.8] pg/ml BALF, p < 0.05). Subjects with the single- or dual-responder phenotype had similar concentrations of BAL cells, cytokines, and chemokines before SBP-Ag.
Using a model of allergen-induced asthma, we have demonstrated that lymphocyte-associated chemokines are associated with eosinophilic airway inflammation. Forty-eight hours after SBP-Ag, the Th2-type chemokines TARC and MDC had a highly significant correlation with the numbers of BAL eosinophils and concentrations of the Th2-associated cytokines IL-5 and IL-13. In contrast, airway lymphocytes correlated not only with Th2-type chemokines, but with the Th1-associated chemokines IP-10 and MIG as well. The potential clinical relevance of these chemokines was demonstrated by the fact that subjects with allergic asthma who had been phenotyped as dual responders on the basis of presentation of an LPR to inhaled antigen challenge had more intense airway inflammation and greater production of Th2- and Th1-type chemokines after localized antigen challenge. Th2-type chemokines may contribute to the persistence of eosinophilic airway inflammation through lymphocyte recruitment and their production of Th2-associated cytokines. These data suggest that both Th2- and Th1-type chemokines are potentially involved in antigen-induced allergic airway inflammation and obstruction.
Although lymphocytes, in particular Th2-type cells, are considered to be an essential component of antigen-induced airway inflammation, little is known about how these cells are recruited to the airway. From in vitro studies, TARC and MDC are known to selectively recruit CCR4-expressing Th2 lymphocytes (3, 4). In vivo evidence that TARC plays a crucial role in antigen-induced airway eosinophilia was provided in a murine model as the administration of a specific antibody against TARC attenuated airway eosinophilia and decreased Th2-type cytokines (18). Our observations that BALF levels of TARC and MDC were significantly correlated with IL-5 and IL-13, as well as with numbers of airway eosinophils, further support a role for these chemokines in antigen-induced eosinophilic airway inflammation. It was, however, somewhat surprising that these chemokines were present in the airway as late as 48 hours after antigen challenge. In general, chemokines, including eotaxin (15, 19), regulated upon activation, normal T cell expressed and secreted (15, 20, 21), macrophage inflammatory protein-1
Th1-associated chemokines IP-10 and MIG are produced by a variety of cell types and could play either a positive or negative role in the regulation of airway inflammation. There is compelling evidence in animal models that Th1 cytokines, and by implication Th1-associated chemokines, downregulate allergic inflammation. For example, IFN-
Despite the evidence that Th1-associated chemokines have the potential to downregulate allergic inflammation, there is also evidence that these same factors can promote airway eosinophilia. For example, overexpression of IP-10 in transgenic mice increased airway hyperresponsiveness, eosinophilia, and IL-4 in an IFN- Our observations corroborate a report by Bochner and colleagues (14) that demonstrated increased BALF concentrations of TARC, MDC, and IP-10, but not of I-309, in subjects with allergic asthma 20 hours after SBP-Ag. In their study, the Th2-associated chemokine response was confirmed to be antigen specific as levels remained at or below the level of detection in fluid from a corresponding saline-challenged segment. In contrast to their study, our data show that, when BAL is performed 48 hours after an SBP challenge with physiologic doses of allergen, TARC and MDC are strongly correlated with BAL eosinophilia and Th2-associated cytokines (IL-5 and IL-13). The lack of correlation in the study by Bochner and coworkers may reflect the timing of the BAL (20 hours) or the dose of antigen given (all subjects received 500 units of either ragweed or house dust mite). Our data, demonstrating that subjects with a dual-phase phenotype have higher concentrations of Th2- and Th1-associated chemokines after SBP-Ag, are in accord with previous reports (15, 28–30) that subjects with a documented LPR have greater eosinophilic airway inflammation and increased numbers of IL-5+ T cells in the bone marrow (30) between 24 and 48 hours after inhaled antigen challenge. The relationship between airflow obstruction seen between 3 and 8 hours after antigen challenge and eosinophilic airway inflammation present 48 hours later is not clearly understood. It has been implied, however, that the inflammatory events occurring at 48 hours reflect, to some extent, the initial airway inflammation associated with the antigen-induced LPR. Thus, it is tempting to speculate that the Th2-associated chemokines in the airway at 48 hours are also involved in, or are reflective of, the late asthmatic response. However, to prove a cause-and-effect relationship, it will be necessary to measure the LPR in the presence of inhibitors of these chemokines. Another possibility, and perhaps a more likely scenario, is that the immune response that occurs during the LPR initiates a cascade of events resulting in a more intense eosinophilic airway inflammation seen between 24 and 48 hours after antigen challenge. Thus, the magnitude of the later inflammatory response could reflect the intensity of the LPR and may be reflective and possibly dependent on the events associated with this response. Indeed, we found that subjects, who had previous evidence for greater airway obstruction at 3–8 hours postallergen challenge, had more airway inflammation (eosinophils, lymphocytes, Th2 cytokines, and lymphocyte-associated chemokines) 48 hours after SBP-Ag. We recognize that there are certain limitations to our human studies. First, it is not feasible to perform detailed kinetic studies using bronchoscopy and BAL after SBP-Ag. Consequently, our data were limited to a single time point (48 hours) after antigen challenge. To more fully understand and define the roles of Th1- and Th2-associated chemokines in allergic inflammation, it would be of interest to study these patterns during the initiation (3–8 hours), peak (12 to 48 hours), persistence (7–10 days), and resolution (14–21 days) of the antigen-induced eosinophilic airway inflammation. Second, the use of segmental challenge did not allow us to determine simultaneous changes in pulmonary function in response to the locally administered antigen. Physiologic changes induced by SBP-Ag are generally limited because the antigen is deposited only in two bronchopulmonary segments. The previous characterization of the pulmonary function responses of these subjects to whole lung inhalation of antigen gives us a historical physiologic pattern by which to evaluate local airway inflammatory responses. Moreover, we have previously demonstrated that the cellular and protein changes after SBP-Ag are qualitatively similar to those induced by inhaled antigen, a method by which LPR can be measured (31). Thus, we feel that this combined approach provides a unique method to assess airway inflammatory responses in relationship to previously defined physiology. Third, our bronchial lavage study may not fully reflect the presence of chemokines in the bronchial mucosa. However, a report showing that TARC and MDC were strongly expressed in the airway epithelium 24 hours after local allergen challenge (10) suggests that lymphocyte-associated chemokines, particularly Th2-associated chemokines, are enhanced throughout the airway after exposure to antigen and thus supports our findings. Fourth, it should be emphasized that although the majority of in vitro polarized Th2 lymphocytes are CCR4+/CXCR3– whereas Th1 lymphocytes are CXCR3+/CCR4–, there is likely to be a more diverse expression profile on both polarized and nonpolarized cells in vivo (6). Indeed, studies of these receptors in human blood revealed a complex pattern that included a high percentage of CCR4+/CXCR3+ T cells, particularly within populations of Th0 and nonpolarized T cells (6). Clearly, single-cell analysis of BAL (and mucosal) T cells for expression of chemokine receptors and Th1/Th2 cytokines is necessary to fully understand and appreciate the contribution of lymphocyte-associated chemokines to allergen-driven recruitment Th1- and Th2-like lymphocytes. Finally, we recognize the possible contribution of lipopolysaccharides (LPS) to our findings. Although the Ag extract used for SBP contains only trace amounts of LPS (as determined by the manufacturer), it is possible that LPS is introduced into the airway when the bronchoscope is passed through the nose. The fact that there were minimal changes in numbers of neutrophils after antigen challenge (medians with 25th and 75th percentiles increased from 0.1 [0.1, 0.2] to 1.7 [1.0, 3.2] x 104 cells/ml BALF) suggests that contamination by LPS was minimal. However, we cannot totally exclude the possibility that even trace amounts of LPS had an effect on chemokine generation. In conclusion, our studies demonstrate that both Th2- and Th1-type chemokines are involved in antigen-induced airway inflammation. Asthma subjects with the dual-responder phenotype developed a more intense airway inflammation that was associated with enhanced levels of Th1- and Th2-associated chemokines. Our observations also suggest that TARC and MDC are involved in the eosinophilic airway inflammation through their recruitment of lymphocytes producing Th2 cytokines. The presence of these chemokines 48 hours after airway antigen challenge also suggests that they contribute to the persistence of allergic airway inflammation in asthma.
The authors acknowledge research nurses Ann Dodge, B.S.N., Mary Jo Jackson, B.S.N., and Andrea Tweedie, B.S.N., for patient recruitment, screening, and assistance with bronchoscopies; Keith Meyer, M.D., for assistance with bronchoscopies; Sarah Panzer, B.S., and Rebecca Lawniczak, B.S., for BAL processing and help with the ELISA; and Michael Evans, M.A., for assistance with data analysis.
Supported by an institutional Specialized Center of Research grant (NIH HL56396) and by a University of Wisconsin General Clinical Research Center grant (NIH M01RR03186). This article has an online supplement, which is accessible from this issue's table of contents online at www.atsjournals.org Conflict of Interest Statement: L.L. has no declared conflict of interest; N.N.J. has no declared conflict of interest; W.W.B. has no declared conflict of interest; E.A.B.K. has no declared conflict of interest. Received in original form December 5, 2003; accepted in final form March 3, 2004
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
chemoattractant (I-TAC, CXCL-11) has not been well characterized during allergic inflammation in asthma. 
p < 0.05, between-group comparisons (postantigen challenge values of single-responder [n = 8] versus dual-responder [n = 8] subjects).
