Published ahead of print on January 6, 2006, doi:10.1164/rccm.200504-594OC
© 2006 American Thoracic Society doi: 10.1164/rccm.200504-594OC
Cigarette Smoke–induced EmphysemaA Role for the B Cell?Departments of Respiratory Medicine and Pathology and Laboratory Medicine, University Medical Center Groningen, Groningen, The Netherlands Correspondence and requests for reprints should be addressed to Prof. Huib A. M. Kerstjens, M.D., Ph.D., Department of Respiratory Medicine, University Medical Center Groningen, Postbox 30.001, NL-9700-RB Groningen, The Netherlands. E-mail: h.a.m.kerstjens{at}int.umcg.nl
Rationale: Little is known about what drives the inflammatory reaction in the development of chronic obstructive lung disease. B cells have been found. Objective: To study the involvement of B cells in the development of emphysema. Methods: The presence of B-cell follicles and their interaction with other cells were investigated in lungs of patients with chronic obstructive pulmonary disease and of smoking mice. B cells were isolated from lymphoid follicles by laser microdissection and analyzed for the presence of immunoglobulin rearrangements and somatic mutations. Main Results: Lymphoid follicles consisting of B cells and follicular dendritic cells with adjacent T cells were demonstrated both in the parenchyma and in bronchial walls of patients with emphysema. A clonal process was observed in all follicles and the presence of ongoing somatic mutations was observed in 75% of the follicles, indicating oligoclonal, antigen-specific proliferation. Similar lymphoid follicles were detected in mice that had developed pulmonary inflammation and progressive alveolar airspace enlargement after smoking. The increase in the number of B-cell follicles was progressive with time and correlated with the increase in mean linear intercept. Specific bacterial or viral nucleic acids could not be detected. Conclusions: B-cell follicles with an oligoclonal, antigen-specific reaction were found in men and mice with emphysema. In mice, the development was progressive with time and correlated with the increase in airspace enlargement. We hypothesize that these B cells contribute to the inflammatory process and/or the development and perpetuation of emphysema by producing antibodies against either tobacco smoke residues or extracellular matrix components.
Key Words: autoimmunity • B cell • chronic obstructive pulmonary disease • emphysema • inflammation Although the pathogenesis of chronic obstructive pulmonary disease (COPD) is still not clear, it is obvious that smoking is a major etiologic factor in the Western world. More than 90% of patients with COPD actively smoke or have done so. However, only 15 to 20% of all smokers eventually develop COPD, suggesting an interindividual variability in susceptibility to cigarette smoke (1). Imbalances in proteases/antiproteases, oxidants/antioxidants, or more generally inflammatory/antiinflammatory effects have been postulated to play a role in the development of COPD, but none has been proven to play a pivotal independent role (2, 3). It is widely accepted that the proteolytic potential of neutrophils and macrophages is important for the destruction of the extracellular matrix in emphysema. This is supported by increased numbers of neutrophils and macrophages in both airways and parenchyma of patients with COPD (4, 5). Moreover, animal studies have demonstrated that macrophages and their proteolytic activity are a prerequisite for the development of cigarette smoke–induced emphysema (6, 7). In addition, Churg and coworkers (8) demonstrated that neutrophil elastase, and thus the neutrophil, is essential for cigarette smoke–induced emphysema in mice. Little is known about the role lymphocytes play in the development of COPD. Several studies in human tissue sections have demonstrated increased numbers of CD8+ and CD4+ T cells in the circulation, airways, and parenchyma of patients with COPD (9–12). Even less is known about the role B cells might have in the development of COPD. Bosken and coworkers (13) detected B cells organized in follicle-like structures that were present in airway adventitia of smokers. Hogg and coworkers (12) assessed the inflammation in small airways of surgically resected lung tissue from patients with stage 0–4 COPD according to the severity classification of the Global Initiative for Chronic Obstructive Lung Disease (GOLD) (1). They described the presence of increased numbers of small airways containing CD4+ and CD8+ lymphocytes, and B cells. As in the earlier work from their group (13), they found a marked increase in lymphoid follicles in the small airways of patients with GOLD stage 3–4 COPD compared with stage 0–2. The role, and thus importance, of these lymphoid follicles remained unclear (12). We hypothesize that B cells contribute to the development of COPD by means of a specific antigen-driven process. Therefore, we studied the presence, organization, and clonality of these cells both in human lung tissue and in our mouse model for cigarette smoke–induced emphysema. Some of the results of these studies have been previously reported in the form of an abstract (14, 15).
Subjects Human lung tissue from surgical resection material of eight patients with COPD (two lung volume reduction surgery, the others lung transplantation) was used (16). COPD was certified by the GOLD lung function criteria (1) and the presence of emphysema was based on histologic examination of lung tissue. The mean FEV1 was 23% predicted, the mean FEV1/VC was 35%. For further details, see the online supplement.
Characterization of Lymphoid Follicles in Human Lung Sections
Immunoglobulin Gene Analyses of B-Cell Infiltrates in Human Lungs Immunoglobulin gene analyses were performed by a nested PCR as described previously (17). FR1 and FR3 primers were used. The PCR products were excised from the gel and cloned. The presence of inserts was checked by PCR, using M13 primers. The PCR products were purified, sequenced, and subsequently subjected to alignment analyses with SeqManII and MegAlign software (both from DNASTAR, Madison, WI), and where possible phylogenetic trees were constructed. Twenty-five to 30 clones were analyzed per follicle. For details, see the online supplement.
Mouse Smoke Exposure
Evaluation of Emphysema and Inflammation in Mice After 2, 4, and 6 mo of smoke or sham exposure, the mice were killed. The right lung was snap frozen. The left lung was inflated and fixed with formalin with 25 cm H2O for 24 h. The mean linear intercept was determined on 25 to 30 photomicroscopic images per animal as a measure of alveolar airspace enlargement by two independent individuals in a blinded manner (20). Inflammatory infiltrates were assessed semiquantitatively on hematoxylin- and eosin-stained paraffin-embedded lung sections. Neutrophils, macrophages, and B cells were detected with anti–GR-1, anti–MAC-3, and anti-B220, respectively, in lung material treated with Nakane's fixative and double stainings were performed on frozen lung sections. The presence of follicular dendritic cells was investigated with FDC-M1 and FDC-M2.
Cytokine Analysis in Murine Lung Homogenates
Statistics
Characterization of Lymphoid Follicles in Human Lung Tissue Lymphoid follicles were detected both in the airway walls and in the parenchyma in lung sections of patients with emphysema (Figure 1). The majority of these follicles (> 80%) were detected in the parenchyma of the lung rather than in the airway walls. As indicated, follicles in the airway wall were found, but only a minority were within the subepithelial area and thus compatible with bronchus-associated lymphoid tissue (BALT). Changes in the epithelium overlaying the follicles that are indicative of BALT were rarely seen. The majority of cells in the follicles were B cells and follicular dendritic cells were additionally present (Figures 1A–1D). The predominant immunoglobulin expressed on the surface of the B cells was IgM (Figure 1E), whereas cells with other immunoglobulin isotypes were present in low numbers (not shown).
The B cells in the infiltrates were predominantly CD27 positive, suggestive of memory B cells (Figure 2A) (21). CD138+ plasma cells were detected in the near vicinity of the follicles (data not shown). Almost all cells within follicles expressed the costimulatory molecule CD40 (Figure 2B) on their membrane. In addition, the proliferation marker Ki-67 was found in the lymphoid follicles (Figure 2C) (22). Both CD4+ and CD8+ T cells were found directly adjacent to the B cells, the majority (80–90%) being CD4 positive.
Immunoglobulin Gene Analyses of B Cells from Infiltrates from Humans Ten lymphoid follicles were isolated by laser microdissection from the lung tissue of eight patients with COPD. Sequence analysis of the immunoglobulin rearrangements revealed the presence of 12 different B-cell contigs with one or more sequence variations in 10 of the contigs. In seven of eight patients clonal B-cell populations were observed, whereas in one patient only unrelated sequences were observed. The seven patients with related immunoglobulin sequences demonstrated the presence of ongoing mutations (Table 1).
Mouse Model of Cigarette Smoke–induced Emphysema: Airspace Enlargement A 7% increase in mean linear intercept was observed in smokers as compared with nonsmoking control subjects (p = 0.03) after 4 mo of cigarette smoke exposure, indicating the onset of emphysema (Figure 3). After 6 mo of smoke exposure the mean linear intercept was increased by 13% (p = 0.007).
Characterization of Lymphoid Follicles in Murine Lung Tissue Histologic evaluation of hematoxylin and eosin staining of paraffin-embedded tissue revealed the presence of pigmented macrophages in lung parenchyma of smoking mice (Figure 4). Inflammatory infiltrates, with a predominantly lymphoid character, were found to be present both in the parenchyma and around terminal bronchioles (Figure 4). Semiquantitative scoring on tissue sections of smoking mice showed that both the number and the size of these inflammatory infiltrates increased during prolonged exposure to cigarette smoke. The number of infiltrates was significantly higher after 6 mo of smoke exposure than after 2 and 4 mo of smoke exposure (Figure 5).
Double staining for CD3 and B220 indicated that the inflammatory infiltrates consisted mainly of B cells surrounded by T cells (Figure 6A). IgM was present on the membranes of these B cells (Figure 6B), but not IgD, IgG, IgE, and IgA. Macrophages and neutrophils were present in the perimeter of the infiltrates. Only scattered plasma cells were seen in the parenchyma; most were IgM positive and generally not related to the follicular aggregates.
Similar to human emphysematous tissue, the murine B-cell aggregates were lymphoid follicles, given the presence of follicular dendritic cells (Figure 6C). Both CD4+ and CD8+ T cells were present around the lymphoid follicles, the majority (> 80%) being CD4 positive (data not shown).
Cytokine Analyses on Murine Lung Homogenates
Our data show the presence of lymphoid aggregates in the airways and in the lung parenchyma of humans with emphysema. The aggregates consisted mainly of B cells interspaced by follicular dendritic cells and surrounded by T cells. Sequence variations in the individual B-cell contigs were demonstrated in the B cells, indicating oligoclonal B-cell proliferation in response to stimulation with as yet unknown antigen(s) (22). B-cell follicles have been demonstrated previously in the small airways of patients with COPD by the group of Hogg and coworkers, both more recently (12) and in earlier work (13). Our findings are novel in several ways. First, we assessed not only the small airways but also the parenchyma. We found that more than 80% of the follicles were located in the parenchyma. This is much more than can be attributed only to effects of airways being adjacent to (either directly above or below) the sampled section. Furthermore, the composition of the B-cell follicles was analyzed: the B-cell follicle core was surrounded by T cells, the majority being CD4 positive. The B cells were interspaced by follicular dendritic cells that are necessary for antigen presentation and affinity maturation. Expression of Ki-67 antigen and the costimulatory molecule CD40 indicates local blast transformation compatible with an (early) germinal center reaction (22). In addition, sequence analysis of rearranged immunoglobulin genes in individual B-cell clones harvested from follicles by laser microdissection revealed the presence of ongoing mutations in clonally related B cells. Together, these data indicate (oligo)clonal B-cell proliferation and support a true germinal center reaction. Interestingly, analogous to our findings in humans with COPD, similar lymphoid follicles were observed in lung tissue in our smoking mouse model. This model produces progressive emphysema over time. Starting at 4 mo of exposure, a progressive increase in both size and number of these follicles was found. As in humans, the follicles contained B cells surrounded by CD4+ and CD8+ T cells and additionally centrally located follicular dendritic cells. We found associations between follicle formation and increased levels of cytokine that are involved in B-cell proliferation. IL-4, IL-6, KC (IL-8), and IL-13 were observed in the lung tissue homogenates of our smoking mice. These cytokines are essential for the formation and differentiation of germinal centers (23). Overexpression of IL-6 has been shown to give rise to areas of proliferating B-cell follicles in mice (24). In addition, overexpression of IL-13 in mice results in severe emphysema (25). Thus, the presence of B-cell follicles has now been demonstrated in human and murine lung tissue in association with emphysema, and in both the airway wall and lung parenchyma. The important question that arises from these findings is the potential role of these B cells in the development of emphysema. The observed ongoing mutations in clonally related B cells within the lymphoid follicles suggest an antigen-driven selection process. At present, it is unclear against which antigen(s) this B-cell proliferation is directed. We would like to put forward the following speculative but tempting considerations. At least three potential sources of antigens should be considered: microbial, cigarette smoke components or derivatives, and degradation products of extracellular matrix. Part of the reactivity of the observed B cells may be directed against bacterial or viral antigens. Indeed, most patients with COPD are colonized with bacteria (26). In particular, adenovirus has been implied in COPD (27, 28) and B cells may play a role in this respect, even though the major part of an antiviral response is of a cellular (cytotoxic) nature (29). Hogg and coworkers have demonstrated B-cell follicles in the small airways of humans with emphysema that were labeled as BALT because of their subepithelial localization. The authors expressed the expectation that the B-cell follicles represented a response against microbial antigens (12, 30), but did not include further support. In our study, the majority of B-cell follicles were not seen within the subepithelial area. The few follicles we observed in relation to the airways were not covered by recognizable lymphoepithelium or M cells that transport antigens across the epithelium and are specific to BALT (31). Moreover, the littermate control mice we used in our studies were exposed to the same housing and (sham) handling, but did not develop these follicles. Analysis of fresh frozen tissue samples by real-time PCR did not indicate any presence of mycoplasma, chlamydia, adenovirus, or Pneumocystis jiroveci. Real-time PCR analysis of these samples using broad-range primers reactive with 16S RNAs from prokaryotes did not show any reactivity above baseline levels, thus providing no evidence for the presence of specific bacterial pathogens. A second source of antigens that could theoretically cause specific B-cell proliferation is cigarette smoke. Cigarette smoke contains approximately 4,500 different compounds (31), of which some are proteins and therefore potentially immunogenic (32). Some of these compounds will precipitate in the lung, possibly bind to the extracellular matrix, and may elicit an antibody response. Alternatively, reactive components from smoke can react with proteins in the tissue to form new, immunogenic protein adducts (33). Subsequently, immune complex formation may occur, eliciting an inflammatory response and subsequently tissue degradation. Finally, the extracellular matrix (ECM) itself may be a source of antigens. Earlier studies have demonstrated that breakdown products of several ECM proteins such as hyaluronic acid, elastin, and collagen have chemotactic and activating effects on neutrophils and macrophages, resulting in the release of oxidants and proteases that are detrimental to the ECM (34, 35). In addition, hyaluronic acid causes activation and proliferation of B cells (36). Apart from this general chemotactic and activating role, we hypothesize that cigarette smoke–induced breakdown products of the ECM might additionally be immunogenic and trigger a specific B-cell reaction. The induced anti-ECM antibodies may subsequently bind to fragments of, or to intact, ECM, causing further degradation of ECM by phagocytes. This would also be compatible with our observation that the immunoglobulins we found were of the IgM isotype because these can be involved in antibody reaction to extracellular matrix products, especially polysaccharides (37). The response to polysaccharides is often T-cell independent in mice and related in particular to the B1 B-cell (IgM+ and IgD–) CD5+ subset, whereas in humans this involves the (spleen-based) marginal zone B-cell subset (IgM+ and IgD–, strongly CD21+, and CD5–). The lung B cells in our study were CD5– and weakly CD21+ in humans, and also CD5– in mice (our unpublished results). The absence of IgD may thus indicate an activated state of lung B cells, but does not indicate a specific origin or nature. Moreover, the fact that it was already known that CD8+ and CD4+ T cells are prominent in COPD enables the suggestion that CD4+ cells in particular may result from nonspecific activation and provide specific help to cytotoxic CD8+ cells and for T-cell dependent B-cell responses. Our reasoning regarding B cells in the lung is speculative, yet it is supported by the presence of similar B-cell follicles in the inflamed synovia and a humoral response against ECM fragments that has been documented in rheumatoid arthritis (38–40). It is conceivable that a viral or bacterial infection or colonization, as frequently seen in COPD, could lead to breakdown of tolerance, facilitating such a reaction against self-antigens. Such events are thought to play a role also in the initial phase and during exacerbations of several autoimmune diseases (41). Both mechanisms described above, that is, smoke- or ECM-derived antigens as inductors of an inflammatory process, may also explain the observation that pulmonary inflammation continues in patients with COPD after smoking cessation (42). We did not investigate the effects of smoking cessation in this mouse model. It is well known that at least some tobacco residues remain in the lung for a long time and thus may maintain inflammation of the airways and lung parenchyma. Direct or indirect antibody-mediated degradation of ECM results in newly formed ECM fragments, which in turn may contribute to perpetuation of the inflammatory reaction. These findings and considerations regarding the antigen specificity of the B-cell response are also compatible with a hypothesis that an autoimmune component may play a role in the development or perpetuation of COPD (43, 44), in which, therefore, besides CD4+ and CD8+ T cells, B cells may also be implicated. In summary, we demonstrate the presence of B-cell follicles in the parenchyma of patients with emphysema. Immunoglobulin gene analysis revealed an oligoclonal process with ongoing mutations, suggesting an antigen-driven process that is probably not of microbial origin. Putative roles for these B cells include a reaction to cigarette smoke components or to extracellular matrix degradation products. The presence of similar B-cell follicles in a smoking mouse model opens additional avenues for further research into the exact role of B cells in the pathogenesis of emphysema.
The authors thank Pieter Klok for skillful animal handling, and also thank Raimond Heukers and Klaas Kooistra for technical assistance. Nicole Stowell, Anuk Das, and Ron Griswold (Centocor, Malvern, PA) performed the Luminex assays. Many parts of the smoking machine setup were a kind gift from Paul O'Byrne, with whom the authors also had fruitful discussions in the initial phases of setting up a smoking animal model. Ron Hendrix at the laboratory for microbiology in Enschede, the Netherlands, kindly performed the PCRs on the mouse material for viral and bacterial RNA. Jody Wright helped greatly with setting up the mean linear intercept measurements.
Supported by Stichting Astma Bestrijding, the Dutch government (as part of the projects enabled by the Spinoza premium granted to Prof. D. S. Postma), and Centocor (Malvern, PA). This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Originally Published in Press as DOI: 10.1164/rccm.200504-594OC on January 6, 2006 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. Received in original form April 17, 2005; accepted in final form January 6, 2006
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
(IL-1
, IL-6, and KC (murine homolog of IL-8) and cytokines involved in the differentiation and proliferation of B cells, such as IL-4, IL-6, and IL-13, were measured in a multiplex ELISA system on whole lung homogenates. 
production in response to in vitro stimulation with collagen type II in rheumatoid arthritis is associated with HLA-DRB1(*)0401 and HLA-DQ8. Arthritis Res 2000;2:75–84.
