This Article Abstract Full Text (PDF) All Versions of this Article: 200410-1384SOv1 172/5/530    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Vermaelen, K. Articles by Pauwels, R. Search for Related Content PubMed PubMed Citation Articles by Vermaelen, K. Articles by Pauwels, R.

Pulmonary Dendritic Cells

Department of Respiratory Diseases, Ghent University Hospital, Ghent, Belgium

Correspondence and requests for reprints should be addressed to K. Vermaelen, M.D., PhD., Department of Respiratory Diseases, Ghent University Hospital, 7K12ie, De Pintelaan 185, Ghent B-9000, Belgium. E-mail: karim.vermaelen{at}ugent.be

	ABSTRACT

TOP ABSTRACT CONTENTS INTRODUCTION DCs OF THE LUNG:... LUNG DCs AND THE... DCs IN HUMAN LUNG... CONCLUSIONS AND PERSPECTIVES REFERENCES

 
Dendritic cells (DCs) are leukocytes that are emerging as chief orchestrators of immune responses. The crucial task of DCs is the continuous surveillance of antigen-exposed sites throughout the body, and their unique responsibility is to decide whether to present sampled antigen in an immunogenic or tolerogenic way. Any misstep can either lead to a flawed immune defense or to allergy, even autoimmunity. It comes as no surprise that the lungs become increasingly the subject of DC-related investigations, as they represent a vast interface between the body and the outer world. This constitutes an enormous challenge for the immune system: "firing up" immune responses inappropriately could have devastating results for the fragile gas exchange structures. Evidence accumulates that DCs play a pivotal role in maintaining the delicate balance between tolerance and active immune response in our respiratory system. The exponentially growing body of DC-related publications is a big challenge. This article aims to provide researchers and clinicians with an up-to-date view on DC biology and its relevance to pulmonary medicine. A developing trend in the field of DCs is the shift from fundamental immunologic research toward exciting clinical insights and applications. For the pulmonary clinician, this heralds the dawn of promising therapies in various domains such as infections, allergy, and cancer.

Key Words: airway • chemokine • chronic obstructive pulmonary disease • dendritic cell • lymph node

	CONTENTS

TOP ABSTRACT CONTENTS INTRODUCTION DCs OF THE LUNG:... LUNG DCs AND THE... DCs IN HUMAN LUNG... CONCLUSIONS AND PERSPECTIVES REFERENCES

 
Introduction

The "Sentinel" Paradigm of Dendritic Cell Biology

Defining Features of the Human DC Family

DCs of the Lung: Localization and Phenotype

Lung DCs and the Shaping of the Pulmonary Immune Response

DC Recruitment to the Lung Is Driven by Environmental

Stimuli

Lung DCs Continuously Report Antigenic Information from

the Airways to Pulmonary LNs

DCs Translate Signals from the Pulmonary Environment into

a Specific Immune Response

DCs in Human Lung Disease: Friend or Foe?

Allergic Asthma

Chronic Obstructive Pulmonary Disease

Pulmonary LC Histiocytosis

Lung Cancer

Lung Transplant Rejection

Conclusions and Perspectives

	INTRODUCTION

TOP ABSTRACT CONTENTS INTRODUCTION DCs OF THE LUNG:... LUNG DCs AND THE... DCs IN HUMAN LUNG... CONCLUSIONS AND PERSPECTIVES REFERENCES

 
The "Sentinel" Paradigm of Dendritic Cell Biology
The first dendritic cells (DCs) were described by Paul Langerhans in 1868 in the basal layers of the epidermis (Über die Nerven der menschlichen Haut. Virchows Archiv für pathologische Anatomie und Physiologie, und für klinische Medicin 1868;44:325–337). These "Langerhans' cells" displayed a typical dendritic morphology, with long, branched arms interdigitating between surrounding epithelial cells. This peculiar shape prompted Dr. Langerhans to consider these cells as a type of neuron. It took more than a century to generate insights through which DCs were correctly identified as white blood cells related to macrophages and monocytes. Subsequently, DCs were described in all lymphoid organs (1), in the blood (2), in the bone marrow (3), and in several other organs, including the lung (4), gut (5), liver (6), heart (7), kidney (8), and urogenital tract (9). Meanwhile, the crucial role of DCs in the control of immunity started to emerge slowly, thanks in part to a series of seminal articles by Steinman and coworkers in the 1970s (1, 10–13). It was known that T lymphocytes, the effector "muscles" of the immune response, cannot recognize antigens (e.g., microbes, tumor fragments, or allergens) in their native form. So-called antigen-presenting cells (APCs) are required to first sample and then process antigens into short fragments, which are presented to T cells on major histocompatibility complex (MHC) molecules. In addition, APCs provide "costimulatory" molecular signals that are required to raise T-cell activation above a threshold allowing an active response. Before the era of DCs, the only known professional APCs were the macrophages, monocytes, and B lymphocytes. With time, however, it became clear that the antigen-presenting power of DCs greatly exceeded what was documented so far (14, 15). This appeared to rely on some unique features in the biology of these cells (for a comprehensive review, see Reference 16).

First, DCs are strategically mobilized to anatomic sites with high antigen exposure (e.g., skin, mucosal surfaces, spleen), thanks in part to their sensitivity to a whole array of chemoattracting inflammatory signals. At this stage, DCs are termed "immature," which implies a high capacity to sense, sample, and process incoming antigen, but a poor ability to stimulate T cells. When antigen exposure is accompanied by so-called danger signals (typically, molecular signatures from pathogens or tissue destruction), DCs undergo a series of dramatic changes, which are defined as "maturation" or "activation" (reviewed in Reference 17). From this point on, the antigen-uptake and antigen-processing function is shut down while large amounts of processed antigen are displayed on cell surface MHC molecules, together with a whole battery of T-cell costimulatory factors. At the same time, DCs become specifically attracted to chemokines emanating from the T-cell–rich areas of local lymphoid organs (this is in sharp contrast to more sedentary sentinel cells, such as macrophages). After migration into the lymph node (LN), DCs induce the proliferation of antigen-specific T lymphocytes, which differentiate into cytokine-producing effector T cells capable of recirculating to the endangered tissues. A summary of this "sentinel" paradigm of DC biology and the molecules generally involved is shown in Figure 1.

View larger version (45K): [in this window] [in a new window]   Figure 1. The "sentinel" paradigm of dendritic cell (DC) biology. Injury to body surfaces induces the release of numerous DC-attracting mediators, such as inflammatory CC-chemokines, ß-defensins, bacterial peptides, or complement cleavage products. Immature DCs are recruited out of the blood circulation and migrate through interstitial tissues to the site of antigen entry and tissue damage. At this stage, DCs are in an antigen-sampling mode: they express a whole array of endocytic receptors, including Fc receptors, receptors for opsonizing complement, and lectinlike receptors (e.g., mannose receptor, DC-specific intercellular adhesion molecule–grabbing nonintegrin [DC-SIGN], DEC-205, Langerin, blood DC antigen [BDCA]-2; reviewed in Reference 97). In addition, immature DCs display a panel of Toll-like receptors (TLRs), which are primitive sensors for a number of "danger signals." These include typical microbial signatures (LPS, CpG oligonucleotides, viral dsRNA, mycobacterial and fungal wall components [260]), as well as endogenous danger signals released secondary to damage or stress inflicted on the local tissue environment (e.g., cytokines such as tumor necrosis factor [TNF-] and interleukin [IL] 1ß, ß-defensins, heat-shock proteins released from necrotic cells [35, 261, 262], and extracellular matrix degradation products [263–265]). After antigen uptake and activation, DCs lose all chemokine receptors except those for chemokine gradients emanating from secondary lymphoid organs (e.g., CCR7, the receptor for the lymphoid organ chemokine CCL21). Meanwhile, the antigen-capturing machinery is shut down while high amounts of peptide-loaded major histocompatibility complex (MHC) molecules and T-cell costimulatory receptors appear on the surface. These immunogenic DCs leave the site of antigen exposure and migrate specifically to the T-cell zones of draining of lymph nodes, where they recruit and activate antigen-specific T cells.

In summary, the DC system of leukocytes has evolved into a fine-tuned, highly sensitive sentinel network capable of both igniting as well as shutting down immune responses as appropriate. This article first provides general notions on DC subsets, where these subsets are found in the lung, and how DCs control the pulmonary immune response. In a second part, we summarize current insights on the possible role of DCs in several major pulmonary pathologies.

Some of the results of the studies described have been previously reported in the form of abstracts (25, 26).

Defining Features of the Human DC Family
Despite the arsenal of techniques available to modern-day researchers, DCs are still difficult to study due to a number of issues, as follows, all of which also apply to the lung:

1. DCs are rare cells in situ: they represent at most a few percent of the total cell population in a given organ.
   
2. Isolating DCs from tissue samples easily induces activation artifacts: DCs are exquisitely sensitive to stress signals arising from the environment. Classical methods for extraction of these cells involve a whole sequence of procedures, including enzymatic organ digestion, gradient centrifugations, and/or overnight incubation steps. All these interventions can cloud the original state of the DC in situ.
   
3. In contrast to other leukocytes, DCs cannot be immunophenotypically identified using one single antigenic marker. Human DCs are commonly identified by the abundant expression of MHC class II (HLA-DR is commonly used as a marker) and the absence of T, B, and natural killer cell, monocytic, and granulocytic lineage markers. Mature or activated human DCs are characterized by a further increase of HLA-DR and costimulatory molecules (e.g., CD40, B7-1, B7-2) on their surface. Additional specific markers of mature human DCs are the immunoglobulin superfamily member CD83 (27), the 55-kD actin-bundling protein Fascin (28, 29), and the DC-specific lysosome-associated membrane protein DC-LAMP (30).
   

An overview of the human DC family tree, together with typical identifying markers, is shown in Figure 2. Currently, an important distinction is made between two main DC subtypes: myeloid DCs (mDCs) and plasmacytoid DCs (pDCs). Human mDCs develop from bone marrow–derived monocytic precursors, whereas pDCs are developmentally related to the lymphoid lineage (31). In addition to differences in identifying markers (32, 33), the most relevant distinction between mDCs and pDCs is functional because these cells are activated by a different set of pathogenic stimuli. This reflects a differential expression of specific Toll-like receptors (TLRs) on both subsets (34).

View larger version (43K): [in this window] [in a new window]   Figure 2. Current, simplified view on the origin and development of the human DC family. This diagram is constructed mostly from in vitro culture studies. Granulocyte-macrophage colony–stimulating factor (GM-CSF) recurs as an important ingredient for the generation of DCs from bone marrow precursors. Additional cytokine combinations, presumably reflecting different tissue environments, induce a differentiation into distinct DC subsets. Specific markers are listed for each DC "end-product" (this is in addition to common identifying features, such as the high expression of MHC class II and the absence of T-cell, B-cell, natural killer–cell, granulocyte, and monocyte markers). The subsets shown here are those found in peripheral tissues (i.e., before activation by pathogenic stimuli and migration into lymphoid organs). Note: strictly speaking, the differences between myeloid and plasmacytoid DCs (in terms of BDCA markers and TLRs) are more prominent at the pre-DC stage.

	DCs OF THE LUNG: LOCALIZATION AND PHENOTYPE

TOP ABSTRACT CONTENTS INTRODUCTION DCs OF THE LUNG:... LUNG DCs AND THE... DCs IN HUMAN LUNG... CONCLUSIONS AND PERSPECTIVES REFERENCES

 
Sertl and colleagues (4) was one of the first to describe the presence of DCs in the airway epithelium, lung parenchyma, and visceral pleura of human and mouse specimens. These cells displayed a typical dendritic morphology, expressed copious amounts of MHCII on their surface, and acted as potent T-cell stimulators in vitro. Later studies have provided more detail into the features and localization of DCs in the different pulmonary compartments. From a methodologic point of view, immunofluorescent techniques allow identification of pulmonary DCs among low-autofluorescent leukocytes (next to lymphocytes and granulocytes), in contrast to the highly autofluorescent pulmonary macrophages (45–47). Using this approach, a small number of DCs can be detected in the bronchoalveolar lavage fluid (BALF). A fraction of these cells express LC markers (CD1a, S100) and exhibit potent T-cell stimulatory capabilities in an allogeneic mixed leukocyte reaction (45, 48). A recent study succeeded in distinguishing different DC subsets in human BALF (i.e., an mDC subset; 0.06% of BALF cells) and a pDC subset (0.02%) (49). Within the bronchial epithelium, DCs are found with typical features of LCs, such as the presence of Birbeck granules, the expression of CD1a, and Langerin (Figure 3) (43, 50, 51). Specific tissue preparation techniques applied to rodent specimens have succeeded in visualizing this superficial airway DC network (Figure 4) (52–54). In the absence of inflammation, DCs are distributed at an average density of several hundred cells per millimeter squared in the large airways, decreasing to less than a hundred DCs per millimeter squared within smaller intrapulmonary airways (52). DCs in the lung parenchyma are present in larger numbers. They are mostly found within interalveolar septa (55). Cochand and coworkers (56) have clearly described the immature character of human lung parenchymal DCs, in terms of low T-cell costimulary molecule expression, the presence of receptors for inflammatory chemokines, and a strong antigen-uptake activity. Recently, our group succeeded in identifying three previously undescribed types of DCs in normal human lung parenchyma: two subsets of mDCs (one, CD11c⁺/BDCA[blood DC antigen]-1⁺; the other, CD11c⁺/BDCA-3⁺) and one pDC subset (CD11c^–/BDCA-2⁺) (51). We are currently trying to characterize the functional differences between these subsets, and the impact of lung disease on these cells.

View larger version (145K): [in this window] [in a new window]   Figure 3. Bronchial epithelium showing Langerin-positive DCs. (Reproduced by permission from Reference 43.)

View larger version (123K): [in this window] [in a new window]   Figure 4. DC network in the airway epithelium ("whole mount" preparation of mouse trachea stained for MHC class II).

	LUNG DCs AND THE SHAPING OF THE PULMONARY IMMUNE RESPONSE

TOP ABSTRACT CONTENTS INTRODUCTION DCs OF THE LUNG:... LUNG DCs AND THE... DCs IN HUMAN LUNG... CONCLUSIONS AND PERSPECTIVES REFERENCES

DC Recruitment to the Lung Is Driven by Environmental Stimuli
Even in the absence of overt inflammation, DCs or their precursors are constantly recruited from the blood into the lung. Holt and coworkers (57) were the first to demonstrate that respiratory tract DCs are continuously replenished by a steady-state bone marrow output. Further studies in animals have shown that the steady-state deployment of the respiratory tract DC network is clearly correlated with age (58): although MHCII⁺ DCs already colonize the fetal lung mesenchyme, airway mucosal DCs only appear a few days after birth, reaching adultlike density and distribution after weaning. Interestingly, DCs spread in a "wave," which starts at the nasal mucosa and progressively spreads deeper into the trachea, all the way down to the alveolar walls. Also, the degree of maturation or activation seems to correlate with the proximity to the outside environment.

Not surprisingly, inflammatory stimuli in the inhaled air have a profound impact on these steady-state dynamics. In a series of hallmark experiments, McWilliam and colleagues (59) described how the inhalation of pathogenic material, such as bacteria or viral particles, induced a very rapid influx of DCs into the airways of rats, in some cases peaking already 2 hours after challenge. Remarkably, this DC recruitment was as fast or sometimes ahead of the prototypic neutrophil influx. It was an isolated phenomenon among mononuclear cells because macrophage numbers were not increased and lymphocytes were not present at that time point. This surprising discovery overthrows the classical view of host defense in which a first phase of neutrophilic influx is gradually taken over by a mononuclear infiltrate. It indicates that DC recruitment is an integral part of the early phase of the innate immune response, with the potential to progress toward a powerful adaptive defense.

Resolution of the inflammation usually restores baseline numbers of pulmonary DCs. However, in some cases, this normalization can be considerably delayed, as observed after Sendai virus infection in rat airways (60), and in a mouse model of respiratory syncytial virus infection in which pulmonary DC numbers were still increased weeks after remission of the acute inflammatory phase (61).

On a molecular level, isolated respiratory tract DCs were shown to be attracted by a whole array of stimuli, including chemokines, complement cleavage products, and bacterial peptides (62). By far the most studied among these mediators are the chemokines (reviewed in Reference 63). Chemokine concentration gradients elicit a directed cellular movement, a phenomenon known as chemotaxis. In vitro studies show that many cells of the lung produce a wide array of chemokines with a known effect on DCs in general (see Table 1 and References 64–81), although only a few have been studied in detail with respect to pulmonary DC mobilization. In vivo, different chemokines orchestrate the recruitment of DCs into the lung depending on the inflammatory stimulus present. In a rat model of inhaled heat-killed Moraxella catarrhalis, the phenomenon seemed dependent on the expression of CCR1 and CCR5, which are receptors for the chemokines CCL5 (regulated on activation, normal T-cell expressed and secreted [RANTES]) and/or CCL3 (macrophage inflammatory protein [MIP]-1) produced at the airway level (82). Interestingly, freshly isolated human pulmonary DCs were found to specifically express both CCR1 and CCR5. Meanwhile, chemokines of the monocyte chemotactic protein (MCP) family appear to be important for the recruitment of DCs in a murine model of pulmonary Mycobacterium tuberculosis infection: mice genetically deficient for the corresponding chemokine receptor, CCR2, have a defect in the progressive accumulation of pulmonary DCs after inoculation with M. tuberculosis. This results in a secondary shortage in pulmonary T-cell recruitment, correlating with premature death, compared with wild-type animals (83, 84). In addition to chemokines, the airway epithelium could attract DCs by means of defensins. These are small cationic peptides with bactericidal activity, produced by epithelial cells after contact with pathogenic stimuli (85). Remarkably, human ß-defensin 2 (which is produced by airway epithelial cells [86]) was shown to engage CCR6, a chemokine receptor typically expressed by immature DCs homing to epithelia (87).

It is not clear yet whether lung DCs are recruited from the blood in a differentiated form or as early precursors. Suda and coworkers (90) have shown that the pulmonary vasculature is enriched with direct DC precursors compared with the peripheral blood circulation. This pool of cells combines a close relationship to the lung tissue environment as well as the potential to differentiate rapidly in strongly immunogenic DCs. Also, the possibility of lung DC population renewal through local proliferation of intrapulmonary progenitors cannot be excluded. It is already known that steady-state pulmonary alveolar macrophage populations are predominantly maintained by in situ cell proliferation, and to a lesser extent by recruitment of monocytic precursors from the circulation (91). Moreover, in the skin, local cell division is the sole mechanism by which LCs appear to regenerate under noninflammatory conditions (92). It is important to note that granulocyte-monocyte colony–stimulating factor (GM-CSF), a key growth factor for "DC-poiesis," is expressed in significant amounts in the lung. GM-CSF production in the airways is increased in an inflammatory context (e.g., respiratory viral infection or exposure to environmental pollutants). Wang and colleagues (93) demonstrated how local transgenic overexpression of GM-CSF in the airways can result in a dramatic increase of pulmonary DC numbers. Another important hematopoietic growth factor for DCs is Fms-like tyrosine kinase 3 ligand (Flt-3L) (94). In contrast to GM-CSF, the potential source of Flt-3L within pulmonary tissues has not been mapped. Nevertheless, repeated injection of this cytokine greatly expands DC numbers in the lung parenchyma (95), and conversely, Flt-3L–deficient animals have a severe reduction in the number of lung interalveolar DCs, whereas airway mucosal DCs remain unaffected (96). These observations suggest that, similar to DCs in lymphoid organs, lung DCs do not constitute a homogenous population. Rather, lung DCs could fall into separate subsets with possibly different hematopoietic origins.

In summary, DCs can populate the lung through different mechanisms (Figure 5). DCs present in the blood circulation could be directly recruited by means of appropriate pulmonary chemokine signals. Alternatively, monocytic DC precursors could first be chemoattracted from the blood into the lung and subsequently differentiate into DCs under the influence of cytokines secreted by resident pulmonary cells (e.g., GM-CSF produced by respiratory epithelium). In addition, a hypothetical mechanism for maintaining lung DC homeostasis might be proliferation of putative intrapulmonary DC progenitors, induced by growth factors secreted by resident pulmonary cells (GM-CSF or Flt-3L). Finally, indirect evidence points to the possibility of transdifferentation of pulmonary macrophages into DCs under the influence of GM-CSF (93).

View larger version (35K): [in this window] [in a new window]   Figure 5. Different ways by which DCs could (re)populate the lung: (1) chemokine-driven recruitment of differentiated DCs from the blood circulation; (2) recruitment of direct DC precursors, which then (3) differentiate into DCs after exposure to local cytokines. Less well documented are (4) putative proliferation intrapulmonary DC progenitors and (5) transdifferentiation of pulmonary macrophages into DCs.

Using a mouse model, we have shown that antigen uptake by airway DCs can proceed without any breach to the mucosal barrier, and in the absence of any inflammatory stimulus (101). This is surprising given that DCs of mouse intrapulmonary airways are predominantly located underneath the airway epithelial basement membrane, and that these experiments involved administration of large (i.e., nonpermeant) macromolecules in the airways using a rigorously nontraumatic and sterile instillation technique. This could point to an antigen capture mechanism as described in a recent study on the gut mucosa: indeed, despite their subepithelial location, DCs in the intestinal lamina propria are perfectly capable of capturing noninvasive antigens by extending interepithelial pseudopods toward the lumen without disrupting the epithelial barrier's integrity (102).

The instillation of inert fluorescent macromolecules into the airways enabled us to track DCs migrating and transporting antigen from the airway mucosa toward pulmonary LNs (Figure 6) (101). This transport was rapid, occurred constitutively (i.e., in the absence of inflammation), and the migratory DCs were seen to specifically penetrate the T-cell zones of the mediastinal lymph nodes (MLNs). These observations confirm earlier studies suggesting that the steady-state recruitment of DCs into the lung is balanced by a continuous emigration of these cells to the MLNs (57). Also, intratracheal administration of soluble protein antigen results in the appearance of strong antigen-presenting activity in the draining MLNs, with similar kinetics (103). Analogous conclusions were reached using intratracheal transfer of in vitro–derived DCs (104, 105), although these cultured DCs might not reflect the exact nature of the endogenous pulmonary DC populations in terms of phenotype, activation level, and antigen-uptake capacity.

View larger version (301K): [in this window] [in a new window]   Figure 6. (A) After deposition of a fluorescent macromolecule in the airways, migratory DCs can be found transporting their fluorescent cargo to the mediastinal lymph nodes (red stain: MHC class II; yellow-green: dextran-fluorescein). (B) In less than 24 hours, airway-derived DCs containing tracer antigen have penetrated deep into the T-cell areas of the mediastinal lymph nodes. (Reproduced by permission from Reference 101.)

As mentioned in the INTRODUCTION, the presence of an inflammatory process at antigen-exposed surfaces is a strong stimulus for the migration of antigen-transporting DCs toward LNs. Transposed to the lung, this would result in an increased flow of antigenic information from the airways toward the MLNs. We have studied this phenomenon in a model of allergic airway disease, and found that ongoing airway inflammation causes a massive and accelerated flux of allergen-transporting DCs from the airway mucosa to the MLNs (117). This is probably due to the intense release of DC-activating inflammatory molecules, such as tumor necrosis factor (TNF-) and prostaglandins, during allergic airway inflammation. Whether this phenomenon contributes to the amplification of the allergen-driven immune response remains to be investigated.

DCs Translate Signals from the Pulmonary Environment into a Specific Immune Response
After arriving into the LNs, DCs face their last and most important task: that is, instruct T cells to respond to presented antigen in the most adequate way. The type and activation state of the DC, the dose of antigen, as well as the nature of concomitant environmental factors present at the time of antigen encounter determine the nature of the resulting T-cell response. Currently, three different possible outcomes for effector T cells are distinguished. A T-helper 1 (Th1) response is characterized by the production of IFN- and TNF by T cells. It is the normal outcome after DC exposure to viruses or bacteria and is crucial for the control of intracellular pathogens such as Mycobacterium spp. It is also the basis of the delayed type hypersensitivity reaction. Th2 differentiation usually occurs after contact with extracellular parasites (e.g., helminths) and involves production of cytokines such as interleukin 4 (IL-4), IL-5, IL-9, and IL-13, resulting in IgE production as well as accumulation of eosinophils and mast cells. Allergens are nonpathogenic environmental antigens that elicit an inappropriate Th2 response. A third main outcome is the induction of tolerogenic or regulatory T cells producing immunosuppressive cytokines, such as IL-10 or transforming growth factor ß (TGF-ß). This is probably the most prevalent response in steady-state conditions. It forms a constant safeguard against the emergence of inappropriate inflammatory reactions to harmless antigen. The way this paradigm can be transposed to the lung's immune homeostasis is summarized in Figure 7.

View larger version (49K): [in this window] [in a new window]   Figure 7. Schematic diagram shows how DCs translate stimuli detected in the inhaled air into a specific immune response. (A) Inert antigen is continuously sampled by DCs, which go through a constitutive program of migration and maturation. In the absence of inflammation, expression of tolerogenic receptors and soluble mediators is predominant. This leads to the generation of regulatory T lymphocytes capable of repressing any emerging immune response to this specific antigen. (B) Invasion by pathogens activates DCs through their TLRs. This leads to enhanced migration to the mediastinal lymph nodes, where antigen is presented in the context of Th1-inducing signals. The Th1 lymphocytes thus generated are crucial for a resolution of the infection. (C) Allergens can also cause DC activation, albeit through poorly defined molecular signals. Some allergens, such as house dust mite extracts, can proteolytically disrupt the integrity of the epithelial layer and activate underlying DCs. Alternatively, inert antigen can become allergenic because of the concomitant presence of a Th2-inducing environmental adjuvant such as diesel exhaust particles (not shown). Activated DCs transport allergen to the draining lymph nodes, where it is presented along with putative Th2-biasing signals. ICOS-L = inducible costimulator ligand; PD-L1 = programmed death ligand-1; PGE2 = prostaglandin E2; TSLP = thymic stromal lymphopoietin.

Nevertheless, induction of tolerance in the lung is unlikely to be the exclusive function of pDCs. Mice deficient for Flt-3L, a key growth factor for pDCs, show no impairment in the development of tolerance to harmless inhaled antigen. In addition, the mDCs that transport protein antigen to the MLNs produce IL-10 and express high amounts of inducible costimulator ligand (ICOS-L), both of which lead to the induction of antigen-specific regulatory T cells with the power to inhibit airway inflammation (123, 124).

As expected, the presence of microbial organisms in the lung modifies the outcome of antigen presentation by DCs in the pulmonary LNs. As an example, a single pulmonary delivery of influenza virus leads to a transient amplified flux of DCs toward the MLNs, generating influenza-specific CD8⁺ T cells that produce large amounts of IFN- (125). Heat-killed Listeria monocytogenes (126) or Aspergillus fumigatus (127) delivered into the airways are also rapidly transported to the MLNs by DCs, leading to the priming of CD4⁺ T cells. Remarkably, the pulmonary DC is able to discriminate among different fungal components: phagocytosis and processing of Aspergillus conidia induces a Th1 response, whereas uptake of hyphae results in Th2 polarization.

The immune response initiated by pulmonary DCs is not only preconditioned by stimuli from the inhaled air. The lung's own cells provide numerous molecular signals that are known to affect DC function in some way. Significant insight has been gained regarding the coexistence of pulmonary DCs and macrophages. In the lung parenchyma, macrophages in the alveolar lumen and DCs within interalveolar septa are separated by a fraction of a micrometer. It has repeatedly been reported that alveolar macrophages inhibit DC functions through the production of soluble mediators (128). Alveolar macrophage–derived nitric oxide was shown to inhibit MHCII expression on lung DCs and suppress their T-cell stimulatory activity (129). Other DC-inhibiting factors secreted by alveolar macrophages include prostaglandins, H₂O₂, TGF-ß (130), and IL-10 (131), as well as decoy receptors for IL-1 and TNF (132). Together, these data suggest that pulmonary macrophages exert their well-known immunomodulatory function through an important restraining influence on DCs. In this view, noninvasive aeroantigens could be taken care of by scavenging macrophages crawling over the airway surface, and these cells would produce mediators that prevent the triggering of DC activation. However, damage to the macrophage–surfactant–epithelium barrier would allow antigen to reach deeper sentinel DCs and would shift the local cytokine environment in favor of DC activation and the initiation of an adaptive immune response. Intriguingly, interstitial macrophages can exert a supportive influence on pulmonary DC immune function by preprocessing particulate antigen into smaller peptides that are then loaded on the surface of neighboring DCs (133, 134).

In contrast to DC–macrophage dialogs, the interaction of pulmonary DCs with other prominent cells of the lung has been less studied so far. Nevertheless, airway and alveolar epithelial cells, fibroblasts, mast cells, nerve endings, and lymphocytes all express mediators that are generally known to affect the immunomodulatory function of DCs. The putative interactions of all these pulmonary cell types with DCs (summarized in Figure 8) are a vast and largely unexplored research terrain with important therapeutic implications.

View larger version (49K): [in this window] [in a new window]   Figure 8. Pulmonary DCs can potentially interact with many cell types in the conducting airways and lung parenchyma. Top: (1) Airway epithelium, (2) type II alveolar epithelial cells, (3) pulmonary alveolar macrophages, (4) pulmonary interstitial macrophages, (5) (myo)fibroblasts, (6) bronchus-associated lymphoid tissue (BALT) lymphocytes, (7) nonadrenergic, noncholinergic (NANC) nerve endings, (8) capillary endothelium, and (9) mast cells. Bottom: The molecules involved are listed, together with the expected effect on DCs (chemokines are listed separately in Table 1). APC = antigen-presenting cell; CGRP = calcitonin gene-related peptide; 5-HT = serotonin; ICAM = intercellular adhesion molecule; SP = substance P; TGF = transforming growth factor; VCAM = vascular cell adhesion molecule; VIP = vasointestinal peptide; VLA = very late antigen.

	DCs IN HUMAN LUNG DISEASE: FRIEND OR FOE?

TOP ABSTRACT CONTENTS INTRODUCTION DCs OF THE LUNG:... LUNG DCs AND THE... DCs IN HUMAN LUNG... CONCLUSIONS AND PERSPECTIVES REFERENCES

Allergic Asthma
Asthma is a chronic inflammatory disorder of the airways featuring bronchial smooth muscle hyperreactivity, congestion of the airway mucosa, and excessive mucus secretion. This results in characteristic symptoms of wheezing, coughing, and dyspnea. These manifestations are typically variable and the airflow limitation is often reversible. Importantly, there is a relentless increase in asthma incidence throughout the world (www.ginasthma.org).

The last decades have seen a growing insight in the cellular and molecular mechanisms underlying allergic asthma. Cutting-edge experimental research has progressively shifted the focus from downstream effectors, such as mast cells, IgE, and eosinophils, up to the Th2 lymphocytes and their proallergic cytokine products (e.g., IL-4, IL-5, IL-13) (135). More recent research by our group and others has implicated the DC as an even more upstream instigator in the allergic inflammatory cascade. Introduction of antigen-loaded DCs into the airways of healthy animals is sufficient to induce allergic sensitization to that antigen (104). Conversely, selective elimination of DCs during ongoing allergic airway inflammation results in the virtual disappearance of the inflammatory symptoms (54). Importantly, other APCs in the lung, such as macrophages or B cells, are unable to take over the fundamental role of DCs in the initiation or maintenance of the eosinophilic inflammation (136, 137).

Experimental studies point to several properties of lung DCs that could account for their crucial role in the maintenance of Th2-based inflammation to inhaled antigen. First, DCs are recruited in massive amounts to the lung during allergic airway inflammation. At the plateau of eosinophilic inflammation, recovery of DCs from BALF is increased 30- to 100-fold (117, 138–140). This increase is supported by a proliferation of early DC precursors in the bone marrow (140), suggesting possible bone marrow–stimulating signals emanating from the inflamed airways, as was already reported for eosinophils (141). Second, DCs within inflamed airways are unusually activated compared with DCs from healthy lungs: cell surface expression of MHCII and T-cell costimulatory molecules is markedly enhanced, whereas levels of ICOS-L, a molecule involved in the induction of inhalation tolerance (142), are decreased (117). In addition, DCs are the chief producers of the chemokine CCL17 in the lung (139). CCL17 ("TARC" [thymus- and activation-regulated chemokine] in the old nomenclature) is a molecule that specifically attracts Th2 lymphocytes toward APCs (143). The interaction of airway DCs with these allergen-specific T cells leads to reciprocal activation and precedes the appearance of the late-phase asthmatic response (144). Seen as a whole, the local DC activation within the airways, together with the immediate exposure to aeroallergen and the concomitant recruitment of airway T and B cells, could provide an adequate microenvironment for an in situ perpetuation of the chronic airway inflammation. In one report, it was even suggested that the chronicity of allergic airway disease could be attributed to the existence of an exceptionally long-lived subset of allergen-presenting DCs in the BALF (145).

As would be expected from these data, molecular defects that impair the mobilization of DCs into the lung during allergen exposure should lead to diminished airway inflammation. We and others have identified matrix metalloproteinase 9 (MMP-9) as a critical molecule enabling the entry of DCs into the airways (139, 146). MMP-9 is an MMP that breaks down basement membranes and allows migratory cells to pass from one tissue compartment to another. MMP-9 deficiency impairs DC accumulation after allergen exposure, resulting in a local shortage of CCL17 and ultimately a collapse of the allergic airway inflammation. A summary of experimental insights concerning the role of DCs in the sensitization and maintenance of allergic airway inflammation is provided in Figure 9.

View larger version (76K): [in this window] [in a new window]   Figure 9. DCs in the pathogenesis of allergic asthma. (1) In the sensitization phase, DCs sample inhaled antigen and transport it to the draining mediastinal lymph nodes (MLNs). (2) In the MLNs, incoming DCs present processed antigen to naive T cells that continuously leave the bloodstream in search of a suitable APC. Under the influence of specific environmental and/or inherited factors, the DC induces proliferation and differentiation of T-helper lymphocytes with a proallergic profile. (3) These allergen-specific "Th2" cells leave the lymph node and enter the circulation en route to antigen-exposed organs. (4) On reexposure to the same antigen, airway DCs producing the Th2-attracting chemokine CCL17 interact locally with the antigen-specific Th2 cells. (5) This initiates the release of Th2 cytokines, including IL-5, which leads to the recruitment of eosinophils. (6) As inflammation intensifies, the expression of vascular adhesion molecules and chemokines is increased, leading to an amplified recruitment of inflammatory cells, including mast cells and T and B lymphocytes. At the same time, the inflamed airways emit molecular signals that increase the output of these cells from the bone marrow (not shown). (7) DCs, attracted to the respiratory epithelium by means of chemokines such as CCL20, start to transmigrate into the airway lumen using the metalloprotease MMP-9. (8) The DCs thus recruited show strong signs of activation, with high levels of MHC class II and T-cell costimulatory molecules. The concomitant recruitment of lymphocytes and the immediate availabilty of aeroantigen create the stage for a local perpetuation of the inflammatory response focused in and around the airways.

Research on human subjects has provided some important insights on the function of DCs in asthma as well. In patients with asthma, allergen challenge causes a sharp decline in the number of circulating blood mDCs within a few hours (151). During the same timeframe, a rapid accumulation of mDCs occurs within the bronchial mucosa (152). This suggests that, in subjects with asthma, an alteration occurs in the lung's chemokine networks, allowing a rapid entry of blood-borne mDCs into the allergen-exposed lung (pDCs do not seem to be affected in the same way in these studies). Recently, the acute drop in circulating mDCs after allergen challenge was shown to be inhibited by treatment with the cysteinyl leukotriene 1 (CysLT1) receptor antagonist pranlukast (153). This was associated with a decreased production of the DC-attracting chemokine CCL20 (measured in induced sputum), and consistent with the expression of the CysLT1 receptor on circulating mDCs. Earlier studies using bronchial biopsies had already documented an increase in the number of intraepithelial airway DCs in subjects with asthma compared with healthy individuals, and this number returned to control levels after inhaled corticosteroid treatment (it is known from animal studies that glucocorticosteroids induce apoptotic depletion of the airway DC network [53]) (154). A functional comparison between bronchial epithelial DCs isolated from healthy atopic versus atopic donors with asthma yielded interesting results: the respiratory epithelium of subjects with asthma was strongly enriched with CD1a⁺ DCs, and these cells preferentially induced the production of IL-4 and IL-5 from sensitized autologous Th cells (50). Also, CD1a⁺ airway mucosal DCs from patients with asthma displayed increased surface levels of FcRI, the high-affinity receptor for IgE, which could help DCs collect IgE-coated allergen on their surface (155).

Additional studies using human blood as a starting material further confirm that DCs from patients with asthma or allergy are "altered" in a way that secures a perpetuation of the inappropriate Th2 response. The possible role of environmental agents in this phenomenon is the subject of intense research. Diesel exhaust particles have often been proposed as Th2-inducing "adjuvants," and it was recently shown that these pollutants can establish a Th2 milieu by influencing both the DC as well as the responder T cell (156). In addition, diesel exhaust particles cause inflammation at the level of the airways by means of reactive oxygen species (157). This would result in DC activation and thus amplify DC-mediated antigen transport toward MLNs, where antigen would be presented in the context of the Th2-biasing signals. More recently, pollen-derived isoprostanes were shown to activate DCs and turn them into Th2 inducers (158). Additional indoor sources of proallergic factors have also been identifed, such as house dust mites thriving in overinsulated houses or cockroach allergens in inner city areas. Der p-1, a protein found in fecal pellets of the house dust mite Dermatophagoides pteronyssinus, has interesting properties with respect to DC biology. This molecule exhibits proteolytic activity and was shown to cleave epithelial tight junction proteins (159), thereby facilitating the delivery of antigen to intra- or subepithelial DCs. Meanwhile, the disturbance in the epithelial integrity would constitute a danger signal and lead to DC activation. Also, bronchial epithelial cells produce several DC-attracting chemokines after exposure to Der p-1 (160). In several studies, human blood–derived DCs were exposed to house dust mite–derived proteins in vitro. From these investigations, it appears that only DCs obtained from atopic subjects exhibit a proallergic function on exposure to house dust mite protein, whereas DCs from healthy subjects induce no T-cell polarization or even maintain tolerogenic properties (110, 161, 162). Intriguingly, only DCs from atopic individuals secrete high levels of the Th2-attracting chemokines CCL17 and CCL22 after Der p-1 exposure (163). The reason why DCs from atopic individuals are "hard-wired" to preferentially induce and attract proallergic T cells is still obscure. To date there is virtually no evidence for genetic polymorphisms determining the T-cell–polarizing function of DCs in atopy. Much more research is addressing the question of how environmental influences can help establish an inappropriately Th2-oriented immune climate, sometimes for life. Further exploration of this issue from a DC standpoint might provide a mechanistic basis to the so-called hygiene hypothesis for the increasing prevalence of allergy. This theory proposes that the immune system of infants needs Th1-inducing stimuli (typically provided by microbial exposure) to evolve from an immature, Th2-biased immune response. The more "aseptic" Westernized urban lifestyle, where use of antibiotics and vaccines is widespread and orofecal pathogen burden is low, implies a shortage of such Th1-promoting triggers, thus creating an allergy-prone situation. Alternatively, inappropriate Th2 responses might arise from a failure of tolerogenic mechanisms, rather than from a lack of Th1 stimuli: Wills-Karp and coworkers (164) recently proposed a refinement to the hygiene hypothesis, in which a continuous low level of microbial stimulation (presumably from gut commensals) would ensure a tolerogenic climate dominated by the production of the immunosuppressive cytokine IL-10. Absence of such "counterregulatory" stimulus would thus remove an important inhibition on inappropriate Th2 (allergic) or even Th1 (autoimmune) immune responses.

There are some interesting elements pointing to the immediate relevance of DCs to the hygiene hypothesis, although many aspects of this complex relationship still need to be resolved. The "immature" character of mucosal immunity in early life might be attributed to the state of the DC system during that period: compared with the adult situation, DCs within neonatal airways are not only less numerous and less mature in terms of MHC expression (58) but also hyporesponsive to inflammatory stimuli (165). In addition, DCs from neonates show an impaired production of IL-12, leading to a defective capacity to induce Th1 responses (166). Holt (167) suggested that a delay in the postnatal acquisition of Th1 competence could explain the difference between atopic and nonatopic individuals. However, in line with the ideas put forth by Wills-Karp and colleagues, we argue that chronic low-level exposure to microbial stimuli would help mucosal DCs evolve toward a tolerogenic rather than a Th1-inducing function in the first place. From that point on, tolerogenic DC networks at mucosal surfaces would constitute the prime sensors for any additional Th1- or Th2-inducing environmental signals, acting alone or in combination. As an experimental illustration of this concept, Eisenbarth and coworkers (168) showed how inert protein antigen coinhaled with low levels of LPS endotoxin triggers airway DCs to induce a Th2 rather than a tolerogenic antigen-specific response. In contrast, coinhalation of high doses of LPS led to a shift from a Th2- to a Th1-biased reaction (168). The modulation of DC-induced airway allergic responses by endotoxin was also confirmed in a report by Kuipers and colleagues (169). These experimental findings are in line with epidemiologic data in which exposure to house dust endotoxin was found to decrease allergic sensitization in infants at high risk of developing asthma (170). In another study, instillation of Dermatophagoides farinae protein–exposed DCs into the airways of healthy mice led to allergic sensitization to that antigen. However, prior coexposure of the DCs to D. farinae extract together with respiratory syncytial virus shifted the pulmonary immune response in vivo from Th2 toward Th1 (171). It should be noted that this experiment involved exposure of mature DCs to respiratory syncytial virus. Primo-infection with respiratory syncytial virus during early infancy, a time when airway DCs are defective in Th1 polarization, is known to establish a Th2-biased climate in the airways, increasing the risk for the development of asthma in later life.

Next to environmental stimuli, a number of studies have stressed the importance of tissue-derived factors in the conditioning of DCs toward a proallergic function. PGE₂, which can be produced by airway epithelial cells, suppresses IL-12 production by DCs, leading to the development of IL-4– and IL-5–producing T cells (172). Histamine, the prototypic mediator released by mast cells, triggers human DCs to induce Th2 immune responses as well (173), thus creating the possibility of a positive feedback loop involving IgE and activated mast cells (174). This finding sheds new light on the results of a large prospective study, in which intake of an antihistamine during infancy significantly decreased the risk of developing asthma in a subset of allergen-sensitized children (ETAC [Early Treatment of the Atopic Child] study [175]).

GM-CSF, another product of the airway epithelium, is not only a DC differentiation factor but also a strong Th2-polarizing cytokine in vivo (176). This might explain why in vitro–generated DCs (using an abundance of GM-CSF in the culture medium) induce allergic sensitization after intratracheal transfer, whereas endogenous airway DCs are tolerogenic by default. Thymic stromal lymphopoietin (TSLP), an IL-7–related cytokine produced by epithelial cells and mast cells, is one of the most powerful triggers to date for DC-dependent allergic immune responses in humans (177). TSLP-exposed DCs secrete high amounts of the Th2-attracting chemokines CCL17 (TARC) and prime naive Th cells to produce proallergic cytokines (IL-4, IL-5, IL-13). Although TSLP is expressed in the normal lung and overexpressed in the skin of patients with atopic dermatitis, TSLP levels in asthmatic lungs have not been investigated yet, something that could have important therapeutic implications. Equally important is knowing which environmental factors might trigger an overproduction of TLSP by pulmonary cells.

In summary, there is compelling evidence to conclude that DCs play a pivotal role in the pathogenesis of allergic asthma. Experimental data indicate that DCs are obligatory APCs for the initiation and maintenance of allergic airway inflammation. In humans, the presence of asthma is associated with an increase in DC numbers in the bronchial mucosa. This increase can be caused by a strikingly rapid movement of circulating DCs into the lung tissue after allergen exposure. Meanwhile, DCs from allergic individuals seem to lose their tolerogenic character and instead perpetuate the allergic state through the expression of costimulatory molecules, cytokines, and chemokines that favor Th2 responses.

Chronic Obstructive Pulmonary Disease
In contrast to allergic asthma, the role of DCs in chronic obstructive pulmonary disease (COPD) has been much less investigated. COPD encompasses two clinical entities, each of which correlates with specific histopathologic features. On one hand, there is chronic obstructive bronchiolitis characterized by inflammatory infiltrates surrounding peripheral airways, leading to poorly reversible airflow obstruction (www.goldcopd.com). On the other hand, there is emphysema, which involves a destruction of interalveolar septa with a progressive decrease of the gas exchange surface. The inflammatory infiltrate in COPD contains mainly neutrophils, CD8⁺ (cytotoxic) T cells, and monocytes/macrophages (reviewed in Reference 178), a combination of cells that has not been fitted into a clear pathogenetic mechanism yet (in contrast to the Th2 lymphocyte/eosinophil presence in asthma).

Recently, we documented a relentless increase in lung DCs in a murine model of tobacco smoke–induced pulmonary emphysema. After 6 months of smoke inhalation, DC numbers in the BALF were 10-fold greater than those observed in control animals (25). An earlier study by Zeid and Muller (179) documented a similar increase in pulmonary Langerhans-like cells on tissue sections of tobacco smoke–exposed mice. In contrast to these studies, Robbins and coworkers (180) reported a decrease in pulmonary DC numbers after smoke inhalation. The reason for these divergent observations can be attributed in part to the methodology used for the detection of DCs and also to large differences in the smoke exposure regimen (a relatively low dose was used in the latter study).

In humans, increased numbers of CD1a⁺ DCs were detected in the airway mucosa of patients with COPD and these numbers declined after fluticasone proprionate inhalation (181). It should be noted that smoking by itself induces an increase in DC numbers in the airways, regardless of the presence of COPD (182, 183). The association of COPD with smoking suggests an innate inflammatory response to inhaled toxic components. It is very likely that the airway DC network is acutely sensitive to smoke inhalation: tobacco smoke contains many potential DC-activating components such as reactive oxygen species and endotoxin. With respect to the latter, we have recently observed a TLR4-dependent activation of pulmonary DCs after smoke inhalation (TLR4 is the signaling receptor for LPS; Maes and colleagues, unpublished manuscript, and Reference 26). In addition, tobacco smoke could trigger airway epithelial cells to release DC-activating factors, such as TNF-, GM-CSF, or heat-shock proteins. It is tempting to speculate that the smoke-activated DCs would lose their default tolerogenic character and induce a sustained immune response to otherwise harmless antigens present in the tobacco smoke (e.g., tobacco glycoprotein). Alternatively, intracellular components released through toxic damage could be chemically altered by smoke and become immunogenic. DCs would then take up, process, and present these modified intracellular antigens to CD8⁺ cytotoxic T cells. The dangerous combination of activated DCs and an abundance of these "new" antigens could set the stage for a self-perpetuating, tissue-directed immune response with devastating consequences in the long term. This acquired immune response could form the basis for the relentless progression of inflammation observed in patients with severe COPD even after smoking cessation. This disturbing phenomenon was revealed in a study by Retamales and coworkers (184), but was rather attributed to latent adenoviral infection of lung epithelial cells, the degree of which was correlated with emphysema severity and associated with persistent inflammatory cell recruitment despite cessation of smoking. Chronic viral infection could indeed provide a plausible explanation for the relative increase in CD8⁺ T cells in the airway infiltrates of patients with COPD. Interestingly, smokers with airway obstruction have elevated levels of a specific latent adenoviral DNA sequence compared with smokers without obstructive disease (185). In addition, CD8⁺ T cells were shown to mediate cytotoxic damage to airway epithelial cells when these were expressing a viral antigen (186). Here again, DCs are known to be required for the initial sensitization of T cells to the viral antigens, either through direct infection of the DCs themselves or after uptake and processing of fragments from infected epithelium (187).

There are some tantalizing hints suggesting a possible role of DCs in the development of emphysema itself. An indirect indication is that LC histiocytosis (LCH) of the lung, which is also associated with smoking and involves a pathologic accumulation of LCs, is accompanied by destructive lesions in the alveolar zones. On a cellular and molecular level, it has been shown that, during the differentiation from monocytes, DCs show an impressive upregulation of transcripts for human macrophage elastase or MMP-12 (188). MMP-12 appears to be crucial for the induction of emphysematous changes in an animal model of tobacco smoke inhalation (189). In addition, transgenic animals with inducible TNF- have an increase in pulmonary MMP-12 levels and spontaneously develop emphysema (190). TNF- is increased in the sputum of patients with COPD (191) and is a strong activator of DC maturation and trafficking (192, 193). Recently, BALF macrophages from patients with COPD were found to display an increased elastolytic activity in vitro, together with increased expression of MMP-12 mRNA. However, as often in similar studies, BALF cell differentiation is performed on a simple morphologic basis, which precludes any accurate distinction between macrophages and DCs. An interesting observation is that elastin fragments generated through the proteolytic activity of MMP-12 are strongly chemotactic for monocytes (194). An interesting hypothesis would be that the continuous inflammatory transit of DCs through elastin-rich alveolar regions maintains or even amplifies itself by means of chemotactic degradation products. As an indication for such a phenomenon, DC migration into inflamed lung interstitium and BAL compartment is decreased in MMP-12–deficient mice (our unpublished observations).

Further studies in animal models will be necessary to determine whether the pathogenesis of COPD relies in substantial amount on a pathologic response of DCs to tobacco smoke components. These investigations might lead to the discovery of several subentities within COPD, each based on the relative contribution of innate versus adaptive pathologic immune responses in the lung. It is hoped that these insights will help to extend the field of research toward novel targets in COPD treatment.

Pulmonary LCH
LCH is a disease characterized by an abnormal accumulation of Langerhans-like cells in one or several organs. Nowadays, the term LCH is used to encompass several clinical entities distinguished on the basis of organ involvement (solitary bone lesions in eosinophilic granuloma, as opposed to systemic involvement in Hand-Schüller-Christian disease, Letterer-Siwe disease, and histiocytosis X). The link between histiocytosis X and Langerhans cells was suggested decades ago by the finding of Birbeck granules in the "histiocytes" accumulating in these lesions (195). The first detailed account of pulmonary involvement was published by Auld (196) in 1957. Classical LC markers, such as CD1a, S-100, and Langerin, are used for the immunohistologic diagnosis of LCH (Figure 10). Langerin has even been suggested as a diagnostic tool for the detection of LCs in the BAL of patients with pulmonary LCH (197). In addition to LCs, LCH granulomas contain macrophages, lymphocytes, and eosinophils (hence the occasional name of "eosinophilic granuloma").

View larger version (130K): [in this window] [in a new window]   Figure 10. Tissue sample from a patient with pulmonary Langerhans cell histiocytosis, stained for Langerin (I. Demedts, unpublished data).

There is still debate concerning the activation/maturation status of LCs in LCH lesions. In one study, LCH cells were described as having low levels of CD83 and B7-2, and weak T-cell stimulatory capacity (204). Combined with overexpression of CCR6, these findings suggest an immature phenotype, except in self-healing LCH lesions where the LCs appeared mature (204). The immature character of LCH cells seemed to correlate with the presence of numerous IL-10–secreting cells in the direct tissue environment. However, these observations were made in bone and skin LCH lesions; similar investigations performed in the lung show that LCs in pulmonary LCH are clearly mature, with very high expression of B7-1 and B7-2 (205). The mature/activated state of pulmonary LCH cells was clearly supported by the local presence of a DC-activating cytokine milieu with elevated IL-1ß and low levels of IL-10. A robust increase in GM-CSF production has been documented in LCH lesions (206), whereas increased TNF- has been described in extrapulmonary LCH (207). Notably, both cytokines are known to be critical for the hematopoietic development of LCs (208). A recent study reported elevated levels of Flt-3L and stem cell factor (SCF) in the serum of patients with LCH (209), a combination of hematopoietic growth factors that was shown to sustain long-term expansion of primitive DC precursors (210). In addition, the abundant production of TGF-ß around LCH lesions (211), combined with GM-CSF, might allow freshly recruited monocytes to differentiate locally into LCs and thus contribute to lesion growth (212).

It is still unclear why granulomatous LCH lesions at the level of the bronchioles evolve into fibrocystic scars in the alveolar zones. Additional studies on MMP production by LCH cells might shed some light on the process and possibly provide a link with the development of emphysema in smokers. Hayashi and colleagues (213) described overexpression of the collagenolytic MMP-2 in pulmonary LCH cells. Also, the cytokines that are overexpressed in LCH lesions are known to be strong inducers of MMPs in cells