Contrasting the antigen-presenting dendritic cells (DCs) in the conducting airways, the alveolar DC populations in human lungs have remained poorly investigated. Consequently, little is known about how alveolar DCs are altered in diseases such as chronic obstructive pulmonary disease (COPD). This study maps multiple tissue DC categories in the distal lung across COPD severities. Specifically, single-multiplex immunohistochemistry was applied to quantify langerin/CD207+, CD1a+, BDCA2+, and CD11c+ subsets in distal lung compartments from patients with COPD (GOLD stage I–IV) and never-smoking and smoking controls. In the alveolar parenchyma, increased numbers of CD1a+langerin− (p < 0.05) and BDCA-2+ DCs (p < 0.001) were observed in advanced COPD compared with controls. Alveolar CD11c+ DCs also increased in advanced COPD (p < 0.01). In small airways, langerin+ and BDCA-2+ DCs were also significantly increased. Contrasting the small airway DCs, most alveolar DC subsets frequently extended luminal protrusions. Importantly, alveolar and small airway langerin+ DCs in COPD lungs displayed site-specific marker profiles. Further, multiplex immunohistochemistry with single-cell quantification was used to specifically profile langerin DCs and reveal site-specific expression patterns of the maturation and activation markers S100, fascin, MHC2, and B7. Taken together, our results show that clinically advanced COPD is associated with increased levels of multiple alveolar DC populations exhibiting features of both adaptive and innate immunity phenotypes. This expansion is likely to contribute to the distal lung immunopathology in COPD patients.

© 2022 The Author(s). Published by S. Karger AG, Basel

Introduction

Chronic obstructive pulmonary disease (COPD) is a chronic inflammatory disease that affects the conducting airways and lung parenchyma. The inflammation, which typically results from long-term exposure to smoke [1, 2], involves innate immune mechanisms. However, adaptive immune mechanisms also participate, as indicated by the expansion of ectopic lymphoid tissue in COPD lungs [3, 4]. Consequently, there is growing attention towards the role of dendritic cells (DCs) in COPD since these cells link innate and adaptive immunity [5, 6].

Although significant amount of work has been carried out on airway DCs [7, 8], the vast majority of the studies have been carried out in murine models where significant species differences make the translation to human lungs complicated. In addition, the large heterogeneity of DCs within each species poses additional challenges regarding their exploration. As a result, relatively little is known about the DC populations in human lungs, especially in the poorly accessible distal regions. However, a handful of important studies have been made and at least three major subsets of lung DCs have been described: myeloid DCs (CD11c+; type 1 and 2), plasmacytoid DCs (BDCA-2+), and Langerhans-type DCs (langerin+, CD1a+) [6, 9].

Different sampling techniques have been employed to study lung DCs such as lung homogenates, bronchoalveolar lavage fluid, and induced sputum [10-12]. Only few studies have used histology to examine COPD lung DCs in their lung anatomical tissue context. The results so far suggest that in the bronchioles (i.e., small airways), langerin+ DCs but not CD1a+ or BDCA2+ DCs accumulate in COPD [13-17]. However, bronchial BDCA-1+ DCs (type 1 myeloid DC) have been reported to decrease in COPD [15].

We have previously demonstrated that the vast majority of lymphoid aggregates in COPD lungs have direct interfaces towards the alveolar lumen and that the epithelium at alveolar-lymphoid interface is selectively infiltrated by langerin+ DCs [18]. This observation, and parallel findings in mice showing a particularly efficient antigen uptake by alveolar DCs [19], suggests that the alveolar parenchyma should be regarded as an important arena for antigen uptake in COPD. However, contrasting the situation in the conducting airways, very little is known about DCs in human alveolar tissues. The few studies that have been performed indicate that CD1a+ and few langerin+ DCs are present in the alveolar parenchyma in COPD patients and smoking subjects [13, 14]. In addition, both myeloid and plasmacytoid DCs are present in homogenated lung parenchyma of COPD patients [20, 21]. Hence, while the accumulated data suggest that the alveolar septa in COPD lungs are likely to be populated with multiple DC populations, basic description and quantification of major alveolar DC populations are missing, especially in relation to disease severity.

The present study performs a detailed quantitative histological assessment of DC populations in the small airways and the alveolar parenchyma from patients with different severities of COPD. In terms of marker selection, and keeping in mind the current lack of consensus for DC sub-classification in humans, the following DC phenotypes were quantified: langerin+ and CD1a+, langerin− DCs, BDCA2+ plasmacytoid DCs as well as myeloid DCs. Myeloid DCs were identified by CD11c positivity after a novel physical exclusion of confounding CD68high and/or CD163high monocytes/macrophages. Since langerin-positivity is a characteristic feature of the DCs patrolling alveolar-lymphoid tissue interfaces [18], langerin+ cells were further subjected to in-depth histology-based profiling to reveal langerin cell phenotypes specific to certain lung anatomical locations (i.e., small airway vs. alveolar parenchyma), specifically in terms of their co-expression of CD11c and/or CD68. We also apply a new single-cell quantification approach to show that langerin+ cells in COPD lungs have an increased extent of mutually exclusive expression of the DC fascin and S100, maturation markers whose expression in the alveolar compartment correlated with other activation markers like B7 and MHC2.

MethodsPatients

Distal lung tissue samples were obtained from never-smokers, smokers without COPD, and patients with GOLD stage I–III COPD undergoing lung resection surgery for lung cancer. Lung tissue samples from GOLD stage IV COPD patients, who lacked history of lung cancer, were collected during lung transplantation surgery. For subject characteristics, see Table 1 and the online supplementary data (see www.karger.com/doi/10.1159/000526080 for all online suppl. material). The tissue samples used in this study have, in part, been described previously [18, 22]. All subjects signed informed consent to participate in the study.

Table 1.

Characteristics of the study subjects

Immunohistochemistry and Rational for DC Classification

All tissue samples were subjected to standardized fixation and processing for generating paraffin sections. Immunohistochemistry (IHC) procedures on 4-μm dewaxed sections were performed in an automated slide staining robot (DakoCytomation, Glostrup, Denmark) after appropriate antigen retrieval procedures. All primary antibodies and associated antigen retrieval protocols (see Table 2) have been validated extensively for use on paraffin sections. Bright-field single and double IHC staining was performed using the Dako EnVision single and double stain kits, respectively (for details, see below). For the bulk quantification across the COPD severities, DCs were sorted into subsets as follows.

Table 2.

Primary antibodies used for IHC

Classical Langerin-Positive and CD1a Langerin-Negative DCs

A modified double staining IHC protocol (Dako EnVision Doublestain kit) was applied to simultaneously identify classical langerin (i.e., CD207)-positive cells as well CD1a+ cells lacking langerin expression. Double-staining IHC was performed using the EnVision Double stain System kit (K5361; Dako) as previously described [18, 23]. Briefly, endogenous peroxidase and alkaline phosphatase activity was blocked with H2O2 and levamisole, respectively. Sections were then incubated with antibodies directed against langerin for 1 h. After subsequent incubation with polymer/HRP-linked secondary antibodies for 30 min, the immunoreactivity was visualized using 3,3′-diaminobenzidine (DAB) chromogen (resulting in a brown opaque staining) or VinaGreen chromogen (resulting in a green staining, for visualization purposes). Next, sections were incubated with Dako Double Stain Blocking Reagent to prevent additional binding of secondary antibodies to the first primary antibody. Sections were then incubated with the anti-CD1a antibody for 1 h. The CD1a immunoreactivity was visualized with a polymer/AP-linked secondary antibody. Finally, sections were incubated with Dako Liquid Permanent Red substrate solution (red stain), counterstained with Mayer’s haematoxylin, air-dried and mounted with Pertex. From the resulting staining, it is possible to detect langerin+ cells (brown cells; with and without CD1a) as well as langerin-negative CD1a+ DCs, i.e., cells that were not stained brown by the prior langerin staining (these CD1a+langerin− DCs appeared as bright red cells).

BDCA-2-Positive Plasmacytoid DCs

IHC for the plasmacytoid DC marker BDCA-2 was performed as described previously [18]. Briefly, endogenous peroxidase activity was blocked with 0.3% hydrogen peroxide for 10 min. Sections were then incubated with mouse anti-human BDCA-2 monoclonal antibodies for 1 h, EnVision Flex+ Mouse (K8021) for 30 min, and polymer/HRP-linked secondary antibodies (EnVision Peroxidase/Dab Detection System kit, K5007; Dako) for 30 min. Next, sections were incubated with DAB substrate-chromogen solution for 40 min (10 min, four times), counterstained with Mayer’s haematoxylin, dehydrated through ethanol series, cleared in xylene, and mounted with Pertex (HistoLab, Gothenburg, Sweden).

Immunohistochemical Identification of CD11c+CD68−CD163− DCs after Physical Exclusion of Confounding CD11c+ Monocytes/Macrophages

Since the DC marker CD11c is also abundantly expressed on lung monocytes/macrophages [24, 25], we used a double staining protocol where the confounding CD11c-expressing monocytes and/or macrophages were obscured and encapsulated with opaque DAB complexes [26] prior to staining for CD11c (online suppl. Fig. S1). Double-staining IHC was performed as described above. Briefly, after blocking endogenous enzymes, sections were incubated with a cocktail of antibodies against CD68 (which recognize lysosomal glycoproteins and plasma membrane proteins of monocytes and macrophages) and antibodies against CD163 (which recognize membrane proteins of some monocytes and most macrophages [27]). The sites of CD68 and CD163 immunoreactivity were identified with a secondary antibody linked to HRP-conjugated polymers (Dako EnVision system) to form extensive opaque brown DAB (or Vina green for visualization purposes) complexes at the sites of any confounding CD68 and/or CD163 macrophage. Next, after a D-block step, sections were stained for CD11c that was visualized with a polymer/AP-linked secondary antibody and Dako Liquid Permanent Red substrate solution (red stain). Sections were then counterstained with Mayer’s haematoxylin, air-dried and mounted with Pertex. Bright red cells in the resulting staining are consequently CD11c+CD68−CD163− since they never were labelled with opaque brown DAB or Vina green (see Fig. 1d, 2b, c).

Fig. 1.

Identification and quantification of DC subsets in small airways. a Bright-field micrograph exemplifying langerin+ (brown DAB stain) DCs in a COPD lung small airway. b A langerin+ DC with cytoplasmic protrusion towards the airway lumen (left, arrowhead), and the typical dendritic morphology of a tangentially cross-sectioned intraepithelial langerin+ DC (right panel). c The typical round-shaped morphology of BDCA-2+ DCs. d CD11c+ DCs (red stain) which lacked immunoreactivity for CD68/CD163 (which here is indicated by Vina green HRP chromogen). Cell nuclei were counterstained with Mayer’s haematoxylin (blue stain). Ep, small airway epithelium; Lu, small airway lumen. Scale bars, a = 35 μm; b = 70 μm; c = 40 μm; d = 15 μm. e–l Scattergrams with quantitative data on epithelial and sub-epithelial small airway DC populations across the control and COPD patient severity groups. Colour codes reveal individual patients within each patient group (e.g., red designates the same patient among the never-smokers regardless of measurement, but red among GOLD I patients designates a different patient). e–j Whereas epithelial langerin+, CD1a+, langerin–, and BDCA2 cells are expressed as number of cells per millimetre epithelial (i.e., BM) length, sub-epithelial cell numbers were expressed per square millimetre sub-epithelial airway wall area. Note, due to their irregular shape and frequent clustering, CD11c+ CD68–, CD163– DCs were expressed as area units of CD11c immunoreactivity per millimetre epithelial length (k) and as proportion of CD11c-positive pixels per total sub-epithelial tissue (l). Black symbols represent individual mean values, and horizontal lines indicate medians for each study cohort. Statistical analysis was performed using Kruskal-Wallis non-parametric test followed by Dunn’s multiple comparison post-test. *: p < 0.05; **: p < 0.01.

Fig. 2.

Identification and quantification of DC subsets in the alveolar parenchyma. a Alveolar langerin+CD1a+ DCs extending spike-like podosomes, indicative of an active migration (arrowhead) [41, 42]. b, c Examples of alveolar CD11c+ DC negative for CD68 or CD163 (red stain) together with CD68 and/or CD163-expressing macrophages (green stain). Cell nuclei were counterstained with Mayer’s haematoxylin (blue stain). Lu, alveolar lumen. Scale bars, a = 9 μm; b = 20 μm; c = 13 μm. d–g Scattergrams with quantitative data on alveolar DC populations across the control and COPD patient severity groups. Colour codes reveal individual patients within each patient group (e.g., red designates the same patient among the never-smokers regardless of measurement, but red among GOLD I patients designates a different patient). Whereas epithelial langerin+, CD1a+, and langerin−, and BDCA2 cells are expressed as cells per square millimetre firm alveolar tissue (i.e., after the air spaces had been excluded). g CD11c+ CD68−, CD163− DCs were expressed as proportion of red CD11c positive pixels of the total firm alveolar tissue area. Black symbols represent individual mean values, and horizontal lines indicate medians for each study cohort. Statistical analysis was performed using Kruskal-Wallis non-parametric test followed by Dunn’s multiple comparison post-test. *: p < 0.05; **: p < 0.01; ***: p < 0.001.

In-Depth Phenotyping of Langerin+ DCs with Immunofluorescence

Langerin-positive DCs were subjected to co-expression analysis by triple immunofluorescence IHC. The aim was to explore if co-expression of CD68 and/or CD11c among langerin+ DCs varied across lung anatomical locations. A parallel triple-staining approach was used to perform a single-cell quantification of langerin DC expression of the activation and maturation markers S100 and fascin. In brief, sections were first blocked with Dako Protein Block Serum Free (X0909) for 10 min. Next, sections were incubated with a mixture of primary antibodies directed against langerin (IgG2b), CD68 (IgG3), and CD11c (IgG2a) for 1 h. Sections were then incubated for 1 h with a mixture of fluorescence-labelled secondary antibodies diluted at 1:200 (Alexa Fluor 488-conjugated goat anti-mouse IgG3, A21151; Alexa Fluor 647-conjugated goat anti-mouse IgG2a, A21241; Alexa Fluor 555-conjugated goat anti-mouse IgG2b, A21147; Life Technologies/Molecular Probe). Hoechst (H33342) was used as nuclear counterstaining. A similar staining was also performed for the combination of langerin, S100, fascin, MHC2, and B7 (see Table 2 for antibody details), albeit with an elution step added, in 7 controls and 7 GOLD IV samples. Langerin, S100, and fascin were stained and imaged before being eluted at 55°C for 30 min in elution buffer (2% wt/vol of SDS and 0.8% wt/vol beta-mercaptoethanol) and then at 50°C for 30 min in elution buffer without beta-mercaptoethanol. Staining and imaging were continued with MHC2 and B7.

Quantitative Assessments of DCs

High-resolution digital images of the entire tissue areas were generated from all bright-field sections using a slide-scanning robot (ScanScope Slide Scanner; Aperio Technologies, Vista, CA, USA [28]). On each section, DCs were analysed and quantified in all small airways and areas of alveolar tissue that were well separated from any visible small airways, pulmonary arteries, arterioles, or veins. Morphometric measurements were performed on the generated digital images using Aperio ImageScope V.10.0 Image Analysis Software (Aperio Technologies).

Small Airway Epithelial DCs (Langerin+, CD1a+Langerin−, and BDCA-2+ Subsets)

In each small airway, the number of DCs within each marker combination was manually counted in a blinded manner, and data were expressed as number of DCs per millimetre epithelial basement membrane (BM). The BM perimeter was determined via manual delineation by cursor tracing.

Small Airway Sub-Epithelial DCs (Langerin+, CD1a+Langerin−, and BDCA-2+ Subsets)

The sub-epithelial tissue was defined as the entire wall area beneath the small airway epithelium and was measured by cursor tracing extending from the BM to the adventitia border facing the alveolar parenchyma. The number of DCs within each marker category was manually counted (blinded), and the data were expressed as number of DCs per square millimetre of sub-epithelial tissue.

Alveolar Tissue DCs (Langerin+, CD1a+Langerin−, and BDCA-2+ Subsets)

In each section, the alveolar tissue was analysed after manually excluding (curser tracing) any visible conducting airways, pulmonary arteries, or veins. Next, the non-tissue background pixels corresponding to the air spaces were automatically filtered away by computerized image analysis, and then the area of alveolar tissue was determined. Patchy regions with emphysematous and fibrotic tissues were excluded from the analysis. The number of DCs within each marker category was counted, and the numbers were expressed as cells per square millimetre of alveolar tissue.

Quantification of Small Airway and Alveolar CD11c+CD68−CD163− Immunoreactivity

CD11c immunoreactivity (red staining) that remained after prior physical blocking of confounding and abundant monocytes and macrophages (with brown, where applicable, or green chromogen representing CD68 and CD163) was quantified using computerized image analysis (VisiomorphDP; Visiopharm, Hoersholm, Denmark). The epithelium was manually delineated for each small airway, and the area positively stained for CD11c immunoreactivity was quantified within the selected region. Data were expressed as CD11c immunoreactivity per millimetre BM. The perimeter of the BM was measured as described above. The sub-epithelial tissue was delineated as described above, and the area positively stained for CD11c was normalized to the total area of the sub-epithelial tissue. Similarly, in the alveolar parenchyma, the area positively stained for CD11c was normalized to the ana­lysed area of the alveolar tissue with airspaces excluded.

Profiling of CD68 and CD11c Expression Profiles for Langerin+ DCs in Multiple Lung Anatomical Compartments

An in-depth mapping of langerin+ cells was performed for the triple langerin-CD11c-CD68 immunofluorescence stained sections from patients with GOLD stage IV COPD. Images were captured using a Nikon Eclipse 80i microscope (Nikon Instruments, Tokyo, Japan) and analysed using NIS-Elements AR 3.0 software (Nikon). All langerin+ DCs within each lung compartment (small airway epithelium, alveolar parenchyma, and the newly identified DC-rich alveolar-lymphoid interface epithelium [18]) were counted, and the proportions of langerin+ DCs that co-expressed any detectable expression of either CD68 or CD11c, or both, were quantified.

Quantification of S100, Fascin, MHC2, and B7 Intensity Levels in Individual Langerin-Positive DCs

Langerin-positive cells in controls and GOLD 4 COPD were also subjected to a type of single-cell quantification of expression levels of S100 and fascin. Triple-stained sections underwent high-resolution whole-slide scanning by an Olympus VS120 slide scanning microscope to yield high-resolution images (with a 16 bit; 0–16,000 dynamic range) of each marker channel. The langerin-positive cells were segmented out by intensity threshold values to generate langerin mask objects (i.e., region of interests [ROIs]). The langerin-specific ROIs were then virtually placed over the original S100, fascin, MHC2, and B7 channel images to measure the mean fluorescence intensity for each marker in each individual langerin cell ROI (CCV software, Medetect AB, QuPath).

Statistics

Kruskal-Wallis nonparametric test (or Wilcoxon where indicated) followed by Dunn’s multiple comparisons post-test was used for comparison between all study groups (Prism V.5.0; GraphPad software, San Diego, CA, USA). Values are given as median (range), unless otherwise stated. Spearman’s rank method was used to calculate correlation coefficients. For 5-plex fluorescence staining of DCs, Wilcoxon nonparametric signed rank tests were used for intensity means of all cells per patient in alveolar parenchyma versus small airways.

ResultsAssessment and Quantification of Small Airway/Bronchiolar DCsIdentification and Characteristics of Small Airway DCs

Langerin+ DCs (exemplified in Fig. 1a, b; online suppl. Fig. 1) were frequently observed in the airway epithelium. The langerin+ DCs were typically localized to the basal part of the airway epithelium and extended cytoplasmic protrusions, which in some rare cases reached the epithelial surface (Fig. 1a, b). Langerin+ cells were also found in the lamina propria and the adventitia layer. Although langerin and CD1a were commonly co-expressed, a subset of CD1a+ DCs expressed no, or very low, immunoreactivity for langerin. Such CD1a-expressing DCs were mainly found in the airway epithelium and to a lesser extent in the lamina propria and the adventitia of small airways. BDCA-2+ DCs, all of which displayed a round, non-typical dendritic morphology, were found both in the small airway epithelium and in the sub-epithelial tissue where they were distributed mainly in the airway adventitia layer (Fig. 1c). Exclusion of confounding CD11c-expressing monocytes and macrophages by simultaneous staining for CD68 and/or CD163 with one chromogen (brown or green, where applicable) and CD11c with a different chromogen (red) revealed the presence of abundant CD11c+CD68−CD163− DCs in the epithelium, lamina propria, and adventitia of small airways (exemplified in Fig. 1d). After the exclusion strategy, no, or exceedingly few, alveolar macrophages or monocytes had detectable CD11c immunoreactivity (data not shown). The CD11c+CD68−CD163− DCs typically had a dendritic morphology, but more round-shaped cells were also observed (Fig. 1d), and frequently occurred in clusters demanding area quantification instead of cell number quantification.

Increased Numbers of Small Airway Langerin+ and BDCA-2+ DCs in COPD Groups

Langerin+ DCs were increased in the airway epithelium in GOLD IV COPD compared with never-smoking controls (p = 0.032; Fig. 1e). The number of langerin+ DCs was, however, unchanged in the sub-epithelial tissue between the different study cohorts (Fig. 1f). In patients with COPD, a weak correlation was observed between the number of epithelial langerin+ DCs and FEV1% of predicted (Fig. 1e; Table 1, r = −0.45; p = 0.018). With COPD groups pooled and compared to never-smokers (smokers without COPD excluded), the increase of langerin + DCs in epithelial tissue among COPD patients persisted (p = 0.0152; online suppl. Fig. 2a). There were no changes in the number of small airway CD1a+langerin− DCs between the study cohorts, neither in the airway epithelium nor in the sub-epithelial tissue (Fig. 1g, h). However, with COPD groups pooled and compared to never-smokers (smokers without COPD excluded), CD1a+langerin− DCs are increased in epithelial tissue of COPD patients (p = 0.0409; online suppl. Fig. 2c).

The number of both epithelial and sub-epithelial small airway BDCA-2+ DCs was increased in patients with GOLD IV COPD compared with smokers without COPD (p = 0.0068 and 0.004, respectively). Differences were also found in GOLD stage I COPD versus stage IV in both epithelial and sub-epithelial tissues (p = 0.05 and 0.0034, respectively; Fig. 1i, j). However, a negative correlation between FEV1% of predicted and the number of BDCA-2+ DCs in the airway epithelium (Fig. 1i; Table 1, r = −0.49; p = 0.008) and sub-epithelial tissue (Fig. 1j; Table 1, r = −0.60; p = 0.0008) was observed in patients with COPD.

In relation to the area of CD11c immunoreactivity after monocyte/macrophage exclusion (via CD68 and/or CD163 staining), no significant changes were observed between the study groups, neither in the airway epithelium nor in the sub-epithelial tissue (Fig. 1k, l).

Assessment and Quantification of Alveolar DCsIdentification and Marker Characteristics of Alveolar DCs

Langerin+ DCs were found in the alveolar septa in all study cohorts. Interestingly, contrasting the small airway langerin+ DCs, the majority of the cells had a direct physical contact with the alveolar lumen, often via spike-like protrusions (Fig. 2a). CD1a+langerin− DCs (Fig. 2e) were also identified within the alveolar parenchyma, albeit in rare numbers compared to the other DC populations. BDCA-2+ DCs were identified in the alveolar parenchyma in all study cohorts. Similar to small airway BDCA-2+ DCs, the alveolar BDCA-2+ DCs had a round non-dendritic (i.e., unbranched) morphology. All study subjects investigated also had identifiable CD11c+CD68−CD163− and CD11c+CD68lowCD163low DCs within the alveolar walls (Fig. 2b, c). These cells frequently had a dendritic (i.e., branched) morphology where some of the protrusions extended into the alveolar lumen.

Multiple DC Subsets Accumulate in the Alveolar Parenchyma in COPD Lungs

Alveolar langerin+ DCs had higher numeric values in advanced COPD compared to controls, although no statistically significant difference was reached with samples unpooled (Fig. 2d). However, with all COPD samples pooled and compared to never-smokers (i.e., excluding smokers without COPD; online suppl. Fig. 3a), langerin + DCs appeared to increase in COPD samples (p = 0.027). The number of alveolar CD1a+langerin− DCs was increased in patients with GOLD stage IV COPD compared with never-smoking controls (p = 0.014; Fig. 2e), and this difference remained with COPD samples pooled (p = 0.043; online suppl. Fig. 3b). Alveolar BDCA-2+ DCs also had a higher cell density in patients with GOLD IV COPD compared to controls (p = 0.0007; Fig. 2f), albeit with COPD samples pooled, this difference did not persist (p = 0.24; online suppl. Fig. 3c). A weak negative correlation was observed between the number of alveolar BDCA-2+ DCs and FEV1% of predicted (Fig. 2f; Table 1, r = −0.45; p = 0.017). The area of CD11c immunoreactivity after CD68 and CD163 exclusion was increased in the alveolar parenchyma of patients with GOLD stage IV COPD compared with never-smoking controls (p = 0.013; Fig. 2g). This difference remained when pooling COPD samples (p = 0.006; online suppl. Fig. 3d).

Identification of Site-Specific Langerin+ DCs in Patients with Advanced COPD

Previously published data suggest that langerin+ DCs may express markers that traditionally are viewed as non-langerin DC markers (e.g., CD68 and CD11c). In order to study if this phenomenon is differentially manifested among distal lung tissue compartments, triple immunofluorescence staining was applied to investigate the co-expression of CD68 and CD11c on langerin+ DCs in the small airway epithelium, the alveolar parenchyma, and in the quite newly identified DC-rich interface epithelium between lymphoid aggregates and alveolar lumen [18] (Fig. 3). This in-depth analysis was carried out in patients with GOLD stage IV COPD. The co-expression analysis revealed that langerin+ DCs either stained negative for both CD68 and CD11c, or stained positive for CD68 and/or CD11c. The relative occurrence of the four marker combinations is presented in Figure 3. The marker combinations differed within each compartment (Fig. 3a). For example, the alveolar parenchyma and alveolar-lymphoid interfaces contained a high proportion of single positive langerin-expressing DCs, and this proportion was significantly higher compared to triple positive DCs (langerin+CD68+CD11c+) and langerin+CD68+CD11c− DCs in the same compartment. Importantly, the proportion of triple-positive langerin+ DCs was significantly higher in the airway epithelium compared with both the alveolar parenchyma and alveolar-lymphoid interfaces (p < 0.05).

Fig. 3.

Identification of site-specific langerin+ DCs in patients with advanced COPD. a Scattergram plots displaying the quantifications of langerin+ DCs and their co-expression of CD68 and/or CD11c. Data are presented as percentage of the total number of langerin+ DCs in airway epithelium, alveolar parenchyma, and lymphoid aggregate (LA) interfaces of patients with GOLD stage IV COPD. Statistical analysis was performed using Kruskal-Wallis non-parametric test followed by Dunn’s multiple comparison post-test. Horizontal lines indicate medians for each study cohort. *: p < 0.05; **: p < 0.01; ***: p < 0.001. #: p < 0.05 compared to langerin+CD68+CD11c+ cells in the airway epithelium. The right pie charts represent the proportions of the mean percentage of each marker combination in each compartment. b–e Photomicrographs of sections triple-stained for langerin (yellow), CD11c (red), and CD68 (green). b, c Small airways. d Alveolar-lymphoid aggregate interface epithelium. e Alveolar parenchyma. White arrowheads indicate langerin+CD68−CD11c+ DCs. Grey arrowheads indicate langerin+CD68+CD11c− DCs. Black arrowheads indicate langerin+CD68+CD11c+ DCs. White arrows indicate langerin+CD68−CD11c− cells. Sections were counterstained with DNA-binding fluorochrome (Hoechst 3332, blue). Ep, small airway epithelium; LA, lymphoid aggregate; Lu, small airway lumen.

Alveolar Langerin DCs Display an Increased but Mutually Exclusive S100 and Fascin Expression in Very Severe COPD

Our in-depth analysis of S100 and fascin immunoreactivity across individual langerin DCs revealed low fascin and a very low S100 expression among langerin residing in the bronchiolar epithelium in control subjects (Fig. 4a). In very severe COPD, the bronchiolar epithelial langerin cells had moderately increased S100 staining (Fig. 4a, b, e, f; p = 0.054) but no elevation of fascin staining intensity (p > 0.6, Fig. 4a, b). In the alveolar parenchyma, langerin+ DCs had an up-regulation of fascin (p = 0.018, Fig. 4c, d) in GOLD IV COPD compared to controls. Further, the alveolar langerin DC expression of S100 showed increased values, although this was just below the threshold for statistical significance (p = 0.12). Interestingly, in all groups, S100 and fascin-high cells displayed a pattern of mutually exclusive expression (i.e., langerin DC with high fascin levels rarely expressed high S100, and vice versa). Density histograms revealed increased S100 and fascin in advanced COPD compared to control situations also among the DCs that had intensity levels under the arbitrary staining intensity threshold values (Fig. 4e, f). The fact that cells with increased S100 or fascin were foremost observed in the alveolar parenchyma, rather than the DC-rich bronchi and bronchioles, was exemplified with combined single-cell expression and tissue background maps (Fig. 4g, h).

Fig. 4.

Langerin+ DCs display location-dependent expression of fascin and S100 in GOLD IV COPD. a–d Scatterplots with single cell data on fascin and S100 staining intensity among individual langerin+ DCs in the bronchial epithelium (a, b) and the alveolar parenchyma (c, d) of lungs from controls and advanced COPD. d Data are presented as scatterplots overlaid with colour-coded bivariate smoothed quantile density contours from pooled single cells from 7 controls and 8 GOLD IV patients. Note the increased but mutually exclusive positivity for S100 and fascin in advanced COPD. Percentage values for fascin+, S100, or double-positive cells are indicated. e, f Density histograms of S100 and fascin intensity among single cells that are below the arbitrarily set threshold values for S100 (i.e., intensity level <45) and facsin (<30). g, h Low-power lung tissue overview images from a representative COPD case where x, y coordinates for langerin-positive cells have been marked with circular symbols that are colour-coded for single cell staining intensity of S100 (g) or fascin (h). Note the generally higher expression in the alveolar parenchyma compared to bronchiolar regions (marked with arrow heads). Scale bar, g, h = 2 mm.

An expanded 5-plex staining was used to also assess the DC maturation markers MHC2 and B7. When comparing pooled cells within patient categories and anatomical sites, B7 and MHC2 correlated with each other, but with a low r value (p < 0.001, r = 0.2). MHC2 and B7 correlation to fascin and S100 was generally higher in the alveolar compartment than in small airways (Fig. 5). Strongest correlation was in COPD between B7 and fascin (r = 0.64, p < 0.001) and B7-S100 (r = 0.79, p < 0.001). B7-S100 correlation was also strong in control alveolar tissue (r = 0.69, p < 0.001). Interestingly, despite the overall correlation, as for fascin and S100, in the alveolar region, there was a trend towards mutually exclusive expression pattern for the high-expressing cells in the B7-fascin, B7-S100, and MHC-fascin comparisons. Contrasting the picture in the airway epithelium, whereas most markers in alveolar DCs had slightly higher values in COPD compared to controls, a tendency towards a reversed pattern was seen in the airway wall.

Fig. 5.

Relationships between expression levels of the DC activation markers B7 and MHC2 and fascin (a, red colour) and S100 (b, blue) across langerin-positive DCs residing in non-diseased control tissue and GOLD IV stage COPD lungs. Data set with pooled and log-transformed DCs values from respective study group. Correlation levels were determined with a justified non-parametric Spearman’s Rho test (r denotes correlation level and p probability of correlation). The highest correlations are marked in bold. Despite the overall correlation, note the trend among the high-expressing cells in the alveolar B7-fascin, B7-S100, MHC-fascin comparisons to have a mutually exclusive expression.

Discussion

The present study provides the first detailed histology-based characterization of multiple DC populations in the distal lung of patients with different severities of COPD. Importantly, our data reveal that at advanced stages of COPD, the most marked increase in DC numbers, including CD11c+CD68−CD163−, BDCA-2+, and CD1a+langerin− populations, takes place in the alveolar parenchyma. Moreover, our data reveal a significant heterogeneity among the langerin+ DCs, which appear to display distinct site-specific phenotypes. Hence, although more research is needed to define the complex heterogeneity of human lung DC populations, this study provides critical support to the recent notion that the alveolar parenchyma should be regarded as an important site for antigen uptake and DC-mediated immune responses in COPD lungs [18, 19]. Moreover, DCs may also contribute to distal lung innate responses, as indicated by the present finding of up-regulated plasmacytoid as well as S100high langerin DCs in the alveolar tissue.

Previous studies performing in situ two-photon microscopy in mice have demonstrated that alveolar DCs more frequently have luminal protrusions and are more efficient in capturing surface antigens than their bronchial counterparts [19]. It is likely that a similar relationship is also present in human lungs since both the recently identified alveolar-lymphoid DCs [18], and the alveolar DCs identified in this study frequently extended luminal protrusions, a phenomenon that was infrequent among small airway DCs. As previously suggested [18, 19], these observations may be a reflection of a need for a particularly rapid immune reaction in the vulnerable alveolar tissue.

The increase in airway langerin+ DCs in COPD patients is in accordance with previous studies [13, 15]. However, contrasting the previous explorations, langerin+ DCs in this study were relatively more abundant in the alveolar parenchyma. The present observation that CD1a is expressed on a large portion of langerin+ lung DCs is consistent with the findings by Van Pottelberge et al. [15]. In bronchoalveolar lavage fluid, approximately 70% of langerin+ DCs are also positive for CD1a [29]. Furthermore and in agreement with other studies, CD1a+ DC frequency was not changed in small airways across severities of COPD patients [14, 15, 17], albeit a difference was found when comparing never-smokers to pooled COPD samples. As also revealed in this study, the population of alveolar CD1a+ DCs seems to be increased in COPD both across unpooled and pooled COPD severities. Interestingly, and in contrast to a study by Van Pottelberge et al. [16], the present study did reveal an increase in the number of BDCA-2+ DCs in small airways of patients with advanced COPD. Importantly, in advanced COPD, this was parallelled by a clear increase in alveolar BDCA-2+ DCs also. The expansion of BDCA-2 plasmacytoid DCs in the lung may have relevant pro-inflammatory implications since these cells are regarded as an important source of IFN-gamma and can execute a variety of other pro-inflammatory and innate immunity mechanisms [30].

The level of CD11c expression, which can be regarded as a crude marker of total myeloid DCs [11], was in this study not changed in small airways of COPD patients. Yet, it cannot be excluded that a sub-population of small airway myeloid DCs is altered. An important example of this possibility comes from previous elegant demonstrations of decreased levels of BDCA-1+ type 1 myeloid DCs in small airways in COPD [15]. Unfortunately, a lack of commercial anti-BDCA-1 antibodies compatible with paraffin-embedded tissue samples prevented us to explore if similar down-regulation of type 1 myeloid DCs also takes place in our COPD patients. In any case, the available data suggest that the generally more abundant type 2 myeloid DCs may not be markedly changed in small airways of COPD patients [20]. In light of this, the novel demonstration in this study that the level of CD11c immunoreactivity is also increased in the alveolar parenchyma is important.

A complicating factor when studying DCs is their heterogeneous and plastic nature. Consequently, the division of DCs into well-established subsets based on surface markers is to a large extent incomplete. For example, a subset of lung langerin+ DCs has in previous studies been found to also express myeloid DC markers [15]. This is consistent with studies suggesting that langerin+ DCs originate from blood myeloid CD11c+CD1a+ DCs [31]. Other studies forward the notion that langerin+ DCs originate from monocytes [32]. In this study, the langerin+ DCs in the airway epithelium commonly co-expressed CD68 and CD11c. Importantly, this was in contrast to the alveolar parenchyma and the alveolar-lymphoid interfaces where the majority of the langerin-expressing DCs were CD68−CD11c−. These novel findings underscore the heterogeneous nature of lung langerin+ DCs. Moreover, our data show that this heterogeneity can be manifested as marker expression profiles specific to certain parts of the lung anatomy. Previous studies suggest that langerin+CD68+ cells, which in this study were more common in small airways, represent a more immature subset of Langerhans cells [33]. It remains to elucidate whether the difference in co-expression of CD68 and CD11c between airway and alveolar langerin+ DCs reflect separate stages of maturation. In any case, the concept of site-specific phenotypes within each DC subset has important methodological implications. For example, unless laser capture techniques are used, detailed information about spatial localization is lost when exploring DC functions in vitro or ex vivo with techniques like FACS analysis or high-end single-cell analysis. Consequently, further studies with a high-plex single cell in situ phenotyping seem highly warranted.

Langerin-expressing cells were in this study also subjected to a new type of histology-based single-cell analysis regarding their co-expression of S100, fascin, MHC2, and B7. Fascin is generally regarded to have critical maturation-associated DC functions such as formation of veil-like membrane protrusions (podosomes) and migration into lymphatic tissues [34]. S100 proteins, on the other hand, have alarmin-like functions and are more associated with several innate immunity responses [35, 36]. The MHC2 protein is expressed on antigen-presenting cells, and B7 needs to be co-expressed on these cells in order to co-stimulate successful T-cell activation. Intriguingly, our single-cell approach revealed that the expressions of S100 and fascin do not correlate among cells pooled across samples. Speculatively, the fascinhigh phenotypes are involved in adaptive immune responses and antigen presentation, whereas the S100high langerin DCs execute innate immune responses. However, in 3 patients, S100 and fascin correlated, entailing further studies. The maturation markers MHC2 and B7 showed non-significant tendencies of higher absolute expression in DCs of small airways than in the alveolar parenchyma. However, the correlations between B7 expression to fascin and S100 levels were both higher in the alveolar parenchyma than in small airways.

The clinical significance of accumulated DC populations in COPD lungs remains unknown. It is possible that the accumulation of alveolar DCs may reflect an adaptation for improved and faster antigen recognition as a result of an increased antigen load. In COPD, alveolar antigens may come from the recurrent infections in the lower respiratory tract [37]. The defective alveolar macrophage phagocytosis in COPD [38] may increase the alveolar antigen load further. Although it is reasonable to speculate that the alveolar DCs in COPD may be involved in the defence against distal lung pathogens, DCs may also be involved in other COPD-related events. For example, CD1a+ cells from COPD lungs produce CCL3 and CXCL9, both of which are CD8+ T-cell attracting chemokines [39]. Furthermore, studies have shown that BDCA-2+ DCs isolated from COPD patients produce higher levels of TNF-α and IL-8 upon maturation compared to BDCA-2+ DCs from control subjects [16]. Moreover, lung DCs isolated from smoke-exposed mice produced increased levels of MMP-12 [40]. It is clear, however, that the function of each DC subset in the lung needs further investigation.

In histology-based detection of DCs, CD11c expression and a “dendritic” morphology is commonly used as a marker of myeloid DCs. However, the wide-spread CD11c expression on macrophages also, many of which display an irregular morphology, calls for cautiousness when using this approach. Through our methodological innovation, which takes advantage of the physical blocking properties by chromogenic DAB complexes, we could in the present study exclude the vast majority of confounding CD68 and/or CD163-expressing macrophages. The possibility to exclude confounding cells by negative selection offers significant advantages compared to standard methods using single marker detection. However, it is of note that with the current approach, there is still a possibility that minor staining for the confounding cell marker is hidden behind the red chromogen used to detect the cell of interest. For example, some of the red cells in Figure 1d could theoretically have minor immunoreactivity for CD68 and/or CD163. In any case, the level of immunoreactivity would be under the threshold to justify the cells as being CD68+ and/or CD163+. The cells with red CD11c-representative staining can thus be designated to be CD11c+CD68lowCD163low, to adopt the same cell phenotyping nomenclature system as is used in FACS analysis.

A potential limitation with the present study, and most COPD studies involving surgical resections, is the inclusion of patients with lung cancer, which could theoretically influence the number of DCs. Importantly, only tissue far away from the tumour was included, and patients in the GOLD stage IV group lacked any history of cancer. Furthermore, the relatively small number of subjects investigated was partly compensated by analysing several lung regions in each subject. It should also be noted that CD11c, which we use as a pan-marker for myeloid DCs, can also be expressed by cells co-expressing CD69 (i.e., NK-cells) CD25 (i.e., activated T-cells/B-cells), MPO (i.e., neutrophils), and CD20 (i.e., B-cells). In our hands, typically 1–4% of CD11c-positive cells are also positive for either of these other markers (data not shown). While we argue CD11c is the most suitable pan-marker for myeloid DCs and that our conclusions still hold, it is a drawback of this study.

In conclusion, this study reveals several novel aspects of the constellation of human alveolar DCs in their spatial lung tissue context and, in addition, demonstrates that severe stages of COPD are associated with an expansion of multiple alveolar DC populations. Although further phenotyping and functional alveolar DC studies are needed, the present type of expansion is likely to lead to an increased capacity for alveolar antigen uptake, which in turn may contribute to the aggravated inflammation and immunopathology of COPD.

Acknowledgements

We thank Karin Jansner and Britt-Marie Nilsson, Unit of Airway Inflammation, Department of Experimental Medicine, for skilful technical assistance with tissue processing and serial sectioning.

Statement of Ethics

The study was approved by the local Swedish Research Human Ethics Committee in Lund, Sweden (Dnr LU-40203). All subjects signed informed consent to participate in the study.

Conflict of Interest Statement

Caroline Sanden and Jimmie Jönsson are employees of Medetect AB. Jonas S. Erjefält is the founder of Medetect AB. The other authors have no competing interests and potential conflict of interest in relation to the present article.

Funding Sources

This work was supported by the Swedish Heart and Lung Foundation (20190508), the Swedish Medical Research Council (VR2017-19), and an unrestricted research grant from Glaxo­SmithKline, UK. None of the funding sources took any part in the execution and publication of the present study.

Author Contributions

Michiko Mori collected tissue samples, performed laboratory work, analysed the major data, and wrote the manuscript. Carl-Magnus Clausson contributed to staining, image digitalization, feedback, and revision of the manuscript. Caroline Sanden, Jimmie Jönsson, Karolina Åkesson, and Premkumar Siddhuraj contributed to the multi-plex DC analysis. Cecilia K. Andersson contributed with tissue handling and sample characterization and revised the manuscript. Medya Shikhagaie, Anders Bergqvist, and Premkumar Siddhuraj revised the manuscript. Claes-Göran Löfdahl contributed to the clinical characterisation. Jonas S. Erjefält designed and supervised the study, participated in the manuscript writing, and critically revised the manuscript. All authors read and approved the final version of the manuscript.

Data Availability Statement

All data generated or analysed during this study are included in this article and its online supplementary material. Other detailed raw data are available on request directed to the corresponding author.

This article is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC). Usage and distribution for commercial purposes requires written permission. Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug. Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.