help button home button
AJRCMB
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Published ahead of print on June 28, 2007, doi:10.1165/rcmb.2007-0087OC
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Online Supplement
Right arrow All Versions of this Article:
2007-0087OCv1
37/5/553    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Veres, T. Z.
Right arrow Articles by Braun, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Veres, T. Z.
Right arrow Articles by Braun, A.
American Journal of Respiratory Cell and Molecular Biology. Vol. 37, pp. 553-561, 2007
© 2007 American Thoracic Society
DOI: 10.1165/rcmb.2007-0087OC

Spatial Interactions between Dendritic Cells and Sensory Nerves in Allergic Airway Inflammation

Tibor Z. Veres1, Sabine Rochlitzer1, Marina Shevchenko1,4, Barbara Fuchs1, Frauke Prenzler1, Christina Nassenstein1, Axel Fischer2, Lutz Welker3, Olaf Holz3, Meike Müller1, Norbert Krug1 and Armin Braun1

1 Department of Immunology, Allergology and Immunotoxicology, Fraunhofer Institute of Toxicology and Experimental Medicine, Hannover, Germany; 2 Clinical Research Unit of Allergy, Charité Campus-Virchow, Charité School of Medicine, Humboldt University, Berlin, Germany; 3 Cytological Laboratory, Hospital Grosshansdorf, Centre for Pneumology and Thoracic Surgery, Grosshansdorf, Germany; and 4 Shemyakin and Ovchinnikov Institute of Bioorganic Chemistry RAS, Moscow, Russia

Correspondence and requests for reprints should be addressed to Armin Braun, Ph.D., Dept. of Immunology, Allergology and Immunotoxicology, Fraunhofer Institute of Toxicology and Experimental Medicine, 30625 Hannover, Germany. E-mail: braun{at}item.fraunhofer.de


   Abstract
Top
 Abstract
CLINICAL RELEVANCE
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Neuroimmune interactions play a critical role in the pathogenesis of asthma. Symptoms like wheezing and cough have been attributed to neural dysregulation, whereas sensitization and the induction of allergic inflammation have been linked with the activity of dendritic cells. Neuropeptides were previously shown to control dendritic cell function in vitro, suggesting interactions between dendritic cells and sensory nerves. Here we characterized the anatomical basis of the interactions between dendritic cells and nerves in the airways of mice and monitored the changes during allergic inflammation. Airway microdissection, whole-mount immunohistology, and confocal microscopy were used for the three-dimensional quantitative mapping of airway nerves and dendritic cells along the main axial pathway of nonsensitized versus ovalbumin-sensitized and -challenged CD11c-enhanced yellow fluorescent protein (CD11c-EYFP) transgenic mice. CD11c-EYFP–positive airway mucosal dendritic cells were contacted by calcitonin gene-related peptide–immunoreactive sensory fibers and their co-localization increased in allergic inflammation. Moreover, protein gene product 9.5–positive neuroepithelial bodies and airway ganglia were associated with dendritic cells. In human airways, human leukocyte antigen DR–positive mucosal dendritic cells were found in the close proximity of sensory nerves and neuroepithelial cells. These results provide morphologic evidence of the interactions between dendritic cells and the neural network of the airways at multiple anatomical sites.

Key Words: airway nerves • dendritic cells • neuroimmune interactions • asthma



   CLINICAL RELEVANCE
Top
 Abstract
 CLINICAL RELEVANCE
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Here we identify anatomical sites at which airway dendritic cells are contacted by various neural structures in health and in allergic inflammation, providing a link between the neural dysregulation and the immunological changes associated with asthma.

 
The mucosal surface of the airways is directly exposed to potentially harmful airborne agents. An efficient detection of these hazards is crucial in order to induce an appropriate reaction to eliminate them and to maintain tissue integrity. This important task is carried out by two "sensory networks." (1) Physical and chemical irritants are detected by a network of sensory nerves projecting to the epithelium (1, 2). Upon stimulation, nonmyelinated sensory C-fibers release pro-inflammatory neuropeptides like calcitonin gene-related peptide (CGRP) and substance-P (SP) via an axon-reflex mechanism resulting in bronchoconstriction, mucus hypersecretion, vasodilatation, and recruitment of leukocytes, a process also referred to as "neurogenic inflammation" (3, 4). (2) Antigenic information is processed by the network of dendritic cells (DCs) located beneath the epithelium (5). Given their ability to detect microbes using their pattern recognition receptors (6), they act as "sensory cells" of the immune system recognizing danger signals. Both of these sensory networks play an important role in the development of allergic asthma. Neural dysregulation explains many symptoms of the disease, such as wheezing, cough, and shortness of breath (4), possibly caused by a hyperactive state and increased neuropeptide production of sensory nerves (7). Then again, DCs act as inducers of the immunologic mechanisms leading to allergic inflammation (8). In addition to their central role in the sensitization against harmless environmental antigens (9), their ability to activate T-cells in the airway mucosa enables them to maintain a prolonged inflammation (10). An efficient response to environmental hazards could benefit from a coordinated action of these two networks. Indeed, neuropeptides released by sensory nerves can influence the activity of DCs (reviewed in Refs. 11, 12) in terms of modulating their chemotaxis (13), maturation, and T-cell stimulatory capacity (14–16). Conversely, little is known about how DCs may affect neural activity. Either way, interaction between DCs and nerves requires their anatomical proximity. Direct contact between DCs of the periphery and sensory nerves has been described before in the skin (15) and the liver (17). In the airways, DCs are contacted by SP-immunoreactive nerve fibers, and their recruitment to the airways after allergen challenge is dependent on SP (18).

Confocal microscopy of whole-mount preparations stained for various neuronal markers has been widely used to study the distribution of airway nerves with different phenotypes (1, 19) and neuroepithelial cells (20) in various species, including humans (21). However, little is known about the anatomical location of DC-nerve contact sites in the airways and even less about its importance in humans. Here we provide information about the three-dimensional distribution of the interaction sites of DCs with nerves, neuroepithelial cells, and parasympathetic ganglia in the conducting airways of mice in normal, healthy conditions and in an ongoing airway inflammation. Moreover, we show evidence for a close anatomical relationship between DCs and neural cells in the human airway mucosa. Some of the results of these studies have been previously reported in the form of an abstract (22).


   MATERIALS AND METHODS
Top
 Abstract
 CLINICAL RELEVANCE
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
Animals
Heterozygous CD11c-EYFP-transgenic mice (23) (kindly provided by Michel C. Nussenzweig, The Rockefeller University, New York) on a C57BL/6 background were used at 10 to 15 weeks of age. The animals were fed with ovalbumin (OVA)-free laboratory food and tap water ad libitum, and held in regular 12-hour dark:light cycles at a temperature of 22°C. All animal experiments were performed in concordance with the German animal protection law under a protocol approved by the appropriate governmental authority (Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit).

Treatment Protocol
CD11c-EYFP mice (n = 9) were sensitized with 10 µg OVA (Grade VI; Sigma, St. Louis, MO) adsorbed to 1.5 mg Al(OH)3 diluted in 0.9% NaCl on Days 0, 14, and 21 via intraperitoneal injection or sham-sensitized with 1.5 mg Al(OH)3 in 0.9% NaCl intraperitoneally, respectively (n = 9). OVA-sensitized animals were exposed to 1% OVA aerosol in 0.9% NaCl for 20 minutes on Days 27, 28, and 35. Sham-sensitized controls were exposed to 0.9% NaCl aerosol on Days 27 and 28, and to 1% OVA aerosol in 0.9% NaCl on Day 35. This treatment resulted in an eosinophilic airway inflammation in OVA-sensitized animals according to the cytological analysis of the bronchoalveolar lavage fluid (see Figure E1 in the online supplement) that was comparable to the inflammation in wild-type animals (24) (data not shown).

Tissue Processing
Twenty-four hours after the last allergen provocation the animals were killed with an overdose of intraperitoneally administered pentobarbital. The chest cavity was opened, the trachea was cannulated, and the lungs were inflated in situ with Zamboni's solution (2% paraformaldehyde and 15% picric acid in 0.1 M phosphate buffer, pH 7.3) at a pressure of 20 mm H2O. After ligation of the trachea the lungs were removed and fixed overnight in the same fixative at 4°C. Next day, the left and the right apical lobes were separated, and each lobe was pinned to the bottom of a petri dish coated with Sylgard (Dow Corning, Midland, MI). The main axial pathways were then carefully dissected under a stereomicroscope (Wild Heerbrugg, Switzerland).

Whole-Mount Immunostaining
The dissected airways were washed in 0.1 M phosphate buffer on a 24-well culture plate until they were clear of fixative, then washed with PBS for 1 hour. Next, the tissue was permeabilized with 0.3% Triton X-100 in PBS for 2 hours and washed three times for 10 minutes each with PBS. To reduce nonspecific antibody binding, the samples were incubated with 1% BSA in PBS for 30 minutes, and the same solution was used for the dilution of all antibodies. After blocking, the bronchus from the left lobe was incubated overnight at 4°C with a mixture of the following primary antibodies: chicken polyclonal to green fluorescent protein (GFP) (Abcam, Cambridge, UK) at a dilution of 1:500, rabbit polyclonal to protein gene product (PGP) 9.5 (1:200; Abcam), and guinea pig polyclonal to CGRP (1:400; Acris, Hiddenhausen, Germany). Since enhanced yellow fluorescent protein (EYFP) signal was very weak after the fixation and tissue processing, it was amplified by labeling with the anti-GFP antibody that cross-reacts with EYFP. The bronchus from the right apical lobe was treated with a mixture of the appropriate isotype controls. Next day, the samples were washed for 6 hours with PBS at room temperature, whereas the buffer was changed every hour. After blocking with 1% BSA in PBS for 30 minutes, the specimens were incubated overnight at 4°C with a mixture of F(ab')2-fragments of the following secondary antibodies: donkey anti-chicken Cy2 (1:200), donkey anti-rabbit Cy3 (1:400), and donkey anti–guinea pig Cy5 (1:200) (all with minimal cross-reactivity, from Jackson ImmunoResearch, West Grove, PA). Next day, the samples were washed again for 6 hours with PBS at room temperature, with a change of the buffer every hour. Additional samples from untreated animals were stained with rat monoclonal antibodies against major histocompatibility complex class II (MHC-II) (clone M5/114.15.2; BD Pharmingen, San Diego, CA) at a dilution of 1:50 or SP (clone NC 1; Chemicon, Temecula, CA) at a dilution of 1:100 instead of the PGP 9.5 staining. These primary antibodies were detected with a donkey anti-rat Cy3 secondary antibody F(ab')2-fragment (1:200; Jackson ImmunoResearch). Samples stained against GFP and MHC-II were finally incubated with Phalloidin conjugated to Alexa Fluor 680 (Molecular Probes, Eugene, OR) at a dilution of 1:40 for 30 minutes and washed twice with PBS. All incubation and washing steps were performed on a rotatory shaker at 150 rpm. After the last washing step, the samples were mounted on a glass slide using Prolong Gold mounting medium (Molecular Probes). To prevent the compression of the tissue, custom-made, slim coverslip pieces were used as spacers between the glass slide and the main coverslip. After the mounting medium had cured, the samples were sealed with nail polish.

Human Bronchial Tissue
Human samples originated from four patients with bronchial carcinoma undergoing lobectomy. The study was approved by the local ethics committee (Ethikkommision der Ärztekammer Schleswig-Holstein) and all patients gave their written informed consent. One-centimeter-long pieces of bronchi with an internal diameter of 3 to 5 mm were taken from parts of the resected lobe that was not associated with the carcinoma. The samples were immediately immersed into Zamboni's solution and fixed overnight at 4°C. Next day, the bronchi were opened with a lengthwise cut and pinned to a petri dish coated with Sylgard with the luminal side upwards. The mucosa was then carefully peeled off, divided into pieces of 20 to 50 mm2, and washed on a 24-well culture plate until the tissue was clear of the fixative. Then the same permeabilization and staining protocol was used as for the mouse tissue samples detailed above. The mixture of primary antibodies also contained the same antibodies against PGP 9.5 and CGRP that showed cross-reactivity with human tissue. Human airway mucosal DCs were detected with a mouse monoclonal antibody against human leukocyte antigen DR (HLA-DR) (clone LN-3; Novocastra, Newcastle, UK) at a dilution of 1:50. One piece of tissue was incubated with the appropriate isotype controls. The mixture of secondary antibodies contained the following: donkey anti-rabbit Alexa Fluor 488 (Molecular Probes), donkey anti-mouse Cy3, and donkey anti–guinea pig Cy5 (both from Jackson ImmunoResearch). The human samples were mounted the same way as described before, with the epithelium upwards.

Confocal Microscopy and Image Analysis
Images were acquired using an LSM 510 META (Carl Zeiss, Jena, Germany) confocal microscope with x10, x20, and x40 (water immersion) objectives, and the laser wavelengths 488 nm, 543 nm and 633 nm were used for the excitation of the fluorochromes Cy2 (or Alexa Fluor 488), Cy3, and Cy5 (or Alexa Fluor 680), respectively. Alexa Fluor 680 was efficiently excited at 633 nm. Triple-stained specimens were scanned in two steps: first, Cy2 and Cy5 channels were acquired simultaneously with the appropriate emission filters, since their emission spectra showed only minimal overlap. Cy3 channel was scanned in a second step. With this acquisition method, no channel cross-talk was observed.

Image stacks for the quantitative analysis were scanned with an XY resolution of 1024 x 1024 that covered an area of 325.8 µm x 325.8 µm. The first optical slice was taken at the luminal surface of the epithelium, and 69 more slices were scanned at an interval of 0.55 µm, resulting in a scan depth of 38.1 µm that contained the epithelium, the smooth muscle layer, and all airway nerves. Two image stacks were taken at each airway generation level, one on the ventral and one on the dorsal side of the main axial pathway, at locations free of large nerve bundles and neuroepithelial bodies. Image stacks were then analyzed using Imaris 4.5.2. (Bitplane, Zurich, Switzerland). First, surface objects were generated in each channel to measure the volumes of fluorescently labeled structures. These virtual surfaces encompassed voxels with a certain minimum fluorescence intensity (see Figure 3C). To find the optimal intensity threshold value for the generation of a surface object, the three-dimensional rendered image was visually compared with the maximum-intensity two-dimensional projection of the same dataset. Once thresholds were set for each channel, they were used for all datasets throughout the analysis except for the Cy5 (CGRP)-channel, where the values were occasionally modified because of higher variations of background fluorescence according to the mean intensity projections. After surface objects of each channel were generated, a quantitative co-localization analysis was performed to calculate the volumes of overlapping surfaces. Two fluorescence channels (e.g., channel A and channel B) were always analyzed with this method. Since volume comprises voxels with a certain fluorescence intensity in channel A and channel B, overlapping surfaces are regarded as a group of co-localized voxels (i.e., voxels with intensities above the threshold in both channels). The number of these co-localized voxels was then calculated for each two channels, and it is expressed as "% of voxels in channel A above threshold co-localized with channel B" or "% of voxels in channel B above threshold co-localized with channel A." Both values were calculated and compared in order to detect differences in co-localization.

Confocal images are shown as two-dimensional maximum-intensity projections or three-dimensional surface objects. The GIMP 2.2 (http://www.gimp.org) software was used for final image processing.

Statistical Analysis
Data are expressed as mean ± SEM. Statistical significance of differences in the volume of nerves and dendritic cells as well as in the percentage of co-localization between OVA-sensitized and nonsensitized animals were analyzed with unpaired t test using GraphPad Prism 4.03. Differences with P < 0.05 were considered as statistically significant.


   RESULTS
Top
 Abstract
 CLINICAL RELEVANCE
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
Visualization of Airway Nerves in Mice
For the determination of airway innervation, whole-mount immunostaining for the pan-neuronalmarker PGP 9.5 was performed and the extensive network of nerves along the whole length of the main axial pathway (Figure 1A) could be visualized. Some large nerve trunks ran lengthwise, giving rise to branches running in all directions, while the smallest nerves provided a crosswise innervation. At each level of PGP 9.5–positive (PGP 9.5+) nerves, thin sensory fibers could be identified showing immunoreactivity against the neuropeptide CGRP (Figures 1B and 1C). These fibers, although at a lower density, followed the distribution of PGP 9.5+ nerves and contained the neuropeptide SP as well. Even though the large nerve bundles were located beneath the epithelium, the smallest, PGP 9.5+CGRP+ varicose fibers ran in the epithelium (data not shown). PGP 9.5 and CGRP staining revealed pulmonary neuroepithelial bodies (NEBs), groups of neuroepithelial cells preferentially located at airway branching points but found elsewhere too (Figures 1B and 1C, arrows).


Figure 1
View larger version (74K):
[in this window]
[in a new window]

  		Figure 1. Identification of airway nerves. Whole-mount immunostaining of the main axial pathway of the left lung (of a nonsensitized mouse) stained for the pan-neuronalmarker protein gene product (PGP) 9.5 and the neuropeptide calcitonin gene-related peptide (CGRP). (A) Montage of 25 confocal images scanned from the luminal side of a whole-mount showing the neural network of the entire mucosal surface area. The specimen was divided into four regions (0, 1–3, 4–6, 7–9) further referred to as "airway generations" according to the bronchiolar branching points (asterisks). These regions were compared during the quantitative analysis (scale bar = 1 mm). (B and C) Inset in A scanned at higher resolution comparing the distribution of all PGP 9.5+ nerves (B) with CGRP+ sensory fibers (C, scale bar = 100 µm). Arrow points to a neuroepithelial body consisting of a group of neuroendocrine cells expressing both PGP 9.5 and CGRP.

 
Identification of Airway Mucosal DCs
CD11c-EYFP transgenic mice (23) were used to visualize CD11c+ airway DCs. Figure 2A shows the distribution of CD11c-EYFP+ DCs in the airway epithelium (see also Video E3). These cells had long dendritic processes and expressed MHC-II (Figure 2B), in contrast to CD11c+ alveolar macrophages that did not express detectable amounts of MHC-II (Figure E2). On the other hand, not all MHC-II+ cells expressed CD11c-EYFP. CD11c-EYFP+MHC-II+ DCs were located on the luminal side of the smooth muscle layer, whereas MHC-II+CD11c-EYFP– cells were found on the abluminal side (Figure 2C).


Figure 2
View larger version (79K):
[in this window]
[in a new window]

  		Figure 2. Identification of airway mucosal dendritic cells (DCs). Triple-staining of a whole-mount specimen (from a naive animal) for enhanced yellow fluorescent protein (EYFP), MHC-II, and F-actin. (A) Network of CD11c-EYFP+ cells in the airway epithelium showing typical DC morphology (maximum-intensity projection of 10 optical sections scanned at region 0 with an interval of 1.2 µm, beginning from the airway lumen, scale bar = 100 µm). (B) Inset in A scanned at higher resolution (60 optical slices with an interval of 0.5 µm) comparing the distribution of CD11c-EYFP+ (upper left image) and MHC-II+ cells (upper right image). Most cells expressed both markers, however some single-positive cells could be identified (merged lower image, scale bar = 40 µm). (C) Three-dimensional reconstruction of the image stack shown in B. The airway epithelium and the smooth muscle were visualized by phalloidin-staining of F-actin filaments. The upper left, upper right, and the lower left images show the luminal side of the smooth muscle layer, the lower right image shows the abluminal side. In the upper right and lower left images the epithelium was virtually removed for a better view of DCs (grid spacing = 40 µm). Note that in the epithelial compartment (luminal side of the smooth muscle layer), all cells express both CD11c-EYFP and MHC-II. On the abluminal side of the smooth muscle layer (lower right image), many cells express only MHC-II but no CD11c (MHC-II staining of the CD11c-EYFP cells is covered here by surface rendering of the CD11c-EYFP channel).

 
The Spatial Relationship between Airway Mucosal DCs and Sensory Nerves
The simultaneous determination of CD11c-EYFP+ DCs and CGRP+ fibers revealed many contact points between the network of DCs and the network of sensory nerves. Figure 3A displays a larger area of the airway wall showing DCs in the close proximity of CGRP+ fibers. At higher resolution (Figure 3B), many DCs were found to be in direct contact with sensory nerve fibers containing both neuropeptides CGRP and SP. This finding was further confirmed by the three-dimensional reconstruction in Figure 3C (see also Video E4).


Figure 3
View larger version (67K):
[in this window]
[in a new window]

  		Figure 3. The spatial relationship between airway mucosal DCs and sensory nerves. Triple-staining of a whole-mount specimen (from a nonsensitized animal) for EYFP, substance P (SP), and CGRP. (A) Overview of a larger area at region 0 of the airway wall showing the distribution of CD11c-EYFP+ DCs and CGRP+ sensory nerves (projection of 9 optical slices with an interval of 1.6 µm, scale bar = 100 µm). Many, but not all DCs are contacted by CGRP+ nerve fibers. (B) Inset in A scanned at higher resolution (Z-interval = 0.5 µm) showing three DCs contacted by nerve fibers containing both neuropeptides CGRP and SP (scale bar = 40 µm). (C) Three-dimensional reconstruction of the image stack in B showing that the CGRP+ fibers run in the immediate proximity of DCs (grid spacing = 20 µm).

 
The Association of DCs with Pulmonary Neuroepithelial Bodies and Airway Ganglia
PGP 9.5+CGRP+ NEBs situated at the branching points of airways were at many locations associated with a group of CD11c-EYFP+ DCs with their cell bodies and dendritic processes "wedged" in the gaps between the neuroepithelial cells (Figures 4A and 4B; Video E5). Other NEBs found at random locations along the main axial pathway were not associated with DCs.


Figure 4
View larger version (71K):
[in this window]
[in a new window]

  		Figure 4. Association of airway DCs with neuroepithelial bodies and airway ganglia. Triple-staining of a whole-mount specimen (from a non-sensitized animal) against EYFP, PGP 9.5 and CGRP. A: projection of 17 confocal images (Z-interval = 0.5 µm) of a pulmonary neuroepithelial body located at an airway branching point. Most PGP 9.5+ cells (upper left image) expressed the neuropeptide CGRP (upper right image). Some CD11c-EYFP+ DCs are "embedded" into the group of neuroendocrine cells (lower images, scale bar = 50 µm). (B) Three-dimensional reconstruction of the image stack in A showing that DCs and neuroendocrine cells are situated in the same level (grid spacing = 20 µm). (C) Three-dimensional reconstruction of an airway ganglion. In the upper image, PGP 9.5+ ganglionic neurons were visualized by volume rendering. In the gaps between the neuronal cell bodies some CD11c-EYFP+ cells can be identified. In the lower image, PGP9.5+ cells were reconstructed using surface rendering and made transparent to reveal the distribution of CD11c-EYFP+ cells. Note that some CGRP+ fibers enter the ganglion and contact CD11c-EYFP+ cells.

 
Large PGP 9.5+ airway ganglia located at the entrance of the left intralobar bronchus contained a few CD11c-EYFP+ DCs in between the neuronal cell bodies (Figure 4C; Video E6).

Quantitative Mapping of Airway Nerves and Mucosal DCs in Allergic Airway Inflammation
To find out whether the contact between DCs and sensory nerves is altered during the development of an allergic airway inflammation, a three-dimensional quantitative mapping of DCs and nerves along the main axial pathways of OVA-sensitized and -challenged CD11c-EYFP transgenic mice was performed and compared with nonsensitized controls. High-resolution confocal image stacks were taken at four different airway generation levels of the main axial pathway specified by the bifurcation number of side-branches according to the method of Peake and coworkers (25) (see Figure 1A) from the proximal to the distal part. Locations containing neuroepithelial cells and large nerve bundles that would bring inhomogeneity to the sampling procedure were discarded. Three-dimensional surface objects were then generated from the confocal image stacks (see dendritic cells and nerves as surface objects in Figure 3C) and analyzed regarding the full volume of DCs and nerves in the airway wall. The extent of DC-nerve contact was quantified via three-dimensional co-localization analysis.

Figure 5 shows the volume of PGP 9.5+ nerves, CGRP+ sensory fibers and CD11c-EYFP+ DCs, expressed as µm3 beneath 1 mm2 of airway epithelium. The volume of PGP 9.5+ nerves (Figure 5A) showed a decreasing centre-to-periphery distribution reflecting their decreasing density toward the small airways.


Figure 5
View larger version (11K):
[in this window]
[in a new window]

  		Figure 5. Quantitative analysis of the three-dimensional distribution of CD11c-EYFP+ DCs, PGP 9.5+ nerves, and CGRP+ sensory fibers in the airway wall in course of an allergic airway inflammation. Mice were sensitized to OVA/Alum intraperitoneally on Days 0, 14, 21 and exposed to OVA aerosol on Days 27, 28 and 35. On Day 36, animals were killed, the lungs were fixed and the main axial pathway of the left lobe was stained as a whole-mount against EYFP, PGP 9.5, and CGRP. Two confocal image stacks were taken at each airway generation from (0) to (7–9) and surface objects of each fluorescence channel were generated to measure the volume of PGP 9.5+ nerves (A), CGRP+ sensory fibers (B), and CD11c-EYFP+ DC (C). Volumes are calculated as µm3 under 1 mm2 area of airway epithelium. Data are shown as mean ± SEM (*P < 0.05 versus NaCl/OVA, n = 9). Open bars, NaCl/OVA; solid bars, OVA/OVA.

 
Allergic airway inflammation had no major effect on the density of nerves. Although there was a small but significant decrease in volume in the most proximal region, there were no further changes between the OVA-sensitized animals and nonsensitized controls. The volume of CGRP+ sensory nerves was not significantly altered (Figure 5B).

In the observed confluent network of DCs the boundaries of single cells were not always clearly defined. Instead of counting the cells, the full volume of CD11c-EYFP+ structures in each stack of images was quantified. No differences in the size and morphology of individual DCs between sensitized animals and nonsensitized controls were seen (data not shown). Twenty-four hours after the last allergen provocation, the full volume of DCs in the airway wall of OVA-sensitized animals did not change at the airway generations (0), (1–3), and (4–6) compared with nonsensitized controls, whereas at the most distal region (7–9) it was significantly increased (Figure 5C, P < 0.05).

Because the proximity of DCs and nerves exceeded the resolution power of light microscopy, the contacting areas appeared as co-localized, double-positive voxels in the confocal image stack. To determine the extent of DC–nerve contact, the percentage of co-localized voxels in each fluorescence channel was calculated.

Figure 6 shows the results of a three-dimensional co-localization analysis that was used to quantify the co-localized voxels of CD11c-EYFP+ DCs and CGRP+ sensory nerves. The percentage of CGRP+ voxels that were co-localized with CD11c-EYFP+ voxels increased in OVA-sensitized animals at the airway generations (0), (4–6), and (7–9) (Figure 6A); however, this increase reached statistical significance only in the most peripheral region (P < 0.05). Measuring contact as the percentage of CD11c-EYFP+ voxels that were co-localized with CGRP+ voxels (Figure 6B) revealed an increasing tendency in case of the OVA-sensitized animals at the same airway generations (0), (4–6), and (7–9), although these differences were not statistically significant. The increasing co-localization in OVA-sensitized animals expressed both as a percentage of CD11c-EYFP+ voxels as well as a percentage of CGRP+ voxels suggests an increasing area of close contact between sensory nerves and DCs during the development of an allergic airway inflammation.


Figure 6
View larger version (11K):
[in this window]
[in a new window]

  		Figure 6. Quantitative co-localization analysis of DCs and sensory nerves. Overlapping areas are calculated as the percentage of CGRP+ voxels of the image stack co-localized with CD11c-EYFP+ voxels (A) or as the percentage of CD11c-EYFP+ voxels co-localized with CGRP+ voxels (B). Data are shown as mean ± SEM (*P < 0.05 versus NaCl/OVA, n = 9). Open bars, NaCl/OVA; solid bars, OVA/OVA.

 
The Spatial Relationship between DCs and Sensory Nerves in the Human Airway Mucosa
The importance of sensory innervation of the human airways and its role in allergic airway disease is still not clearly understood. There is only a limited information available on the three-dimensional distribution of nerves (1), neuroepithelial cells (21), and DCs in the human airway mucosa. It was interesting to see whether the close proximity of sensory nerves and DCs observed in mice exists in the human airways as well.

Human samples were processed according to Weichselbaum and colleagues (21) to study the anatomical relationship of DCs and nerves in the human airways. Staining for PGP 9.5 and CGRP revealed a dense network of nerves and neuroepithelial cells in the human airway mucosa (Figure 7A). Large PGP 9.5+ nerve bundles running in the lamina propria gave rise to small varicose fibers projecting into the epithelium, many of which contained CGRP (Figure 7A). HLA-DR staining was used to visualize human airway DCs, showing a morphology very similar to those of mice (Figure 7B).


Figure 7
View larger version (104K):
[in this window]
[in a new window]

  		Figure 7. The distribution and spatial interaction of nerves, neuroendocrine cells and DCs in the human airway mucosa. Bronchi with an internal diameter of 3 to 5 mm taken from surgical samples of patients undergoing lung resection were fixed and the mucosa was peeled away from the cartilage and the rest of the airway wall. Pieces of 20 to 50 mm2 size were stained as whole-mounts against HLA-DR, PGP 9.5, and CGRP. (A) Staining for PGP 9.5 and CGRP reveals a network of airway nerves. Large panel: Z-projection ("top-view") of a stack of 120 optical slices scanned at an interval of 0.5 µm starting from the luminal side. Small panel: X-projection ("side-view") of the same dataset. A thick nerve bundle traveling in the lamina propria (arrow) penetrates the basement membrane and gives rise to PGP 9.5+ and/or CGRP+ varicose fibers terminating in the epithelium. Asterisk shows a PGP 9.5+ neuroendocrine cell. (B) Staining against HLA-DR reveals a network of DCs showing typical dendritic morphology. Z-projection of 25 optical slices ("top-view"), Z-interval = 0.5 µm (scale bar = 50 µm). (C) Double-staining for HLA-DR and CGRP reveals DCs contacted by sensory nerve fibers. Large panel: Z-projection ("top-view") of 56 slices (interval = 0.5 µm), arrows showing contact points. Small panel: Y-projection ("side-view") of the same dataset (scale bar = 50 µm). (D) Three-dimensional reconstruction of the image stack shown in C (grid spacing = 50 µm). (E and F) Triple-staining for HLA-DR, PGP 9.5, and CGRP identifies DCs contacting PGP 9.5+ neuroepithelial cells, some of which also express CGRP. Large panels: Projections along the Z axis ("top-view") of a stack of 67 slices with an interval of 0.5 µm (scale bar = 50 µm). Small panels: Y-projections ("side-view") of the same dataset.

 
In the human specimens several HLA-DR+ DCs were found to be contacted by CGRP+ sensory fibers (Figures 7C and 7D; Video E7). Furthermore, human airway mucosal DCs were frequently associated with PGP 9.5+ neuroepithelial cells, many of which also expressed CGRP (Figures 7E and 7F).


   DISCUSSION
Top
 Abstract
 CLINICAL RELEVANCE
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
In the present study we performed a three-dimensional quantitative mapping of the distribution of nerves and dendritic cells along the main axial pathway of mice to study their spatial interactions in normal conditions and in the course of an allergic airway inflammation.

The pan-neuronal marker PGP 9.5 outlined the complex network of all nerves and neuroepithelial cells in the airway wall, and staining for the neuropeptide CGRP revealed small, varicose fibers of sensory nerves similar like in previous works (1, 20). The determined density of PGP 9.5+ nerves showed a decreasing center-to-periphery distribution as expected. The distribution of CGRP+ sensory fibers was more homogeneous along the whole length of the main axial pathway. In contrast with an other study using thin sections and conventional quantification methods (26), allergic airway inflammation had no effect on the density of CGRP+ sensory nerves. However, it is difficult to compare our data based on the measurement of CGRP+ nerve volumes with that previous work measuring CGRP-immunoreactive surface areas.

The distribution of DCs was studied using the markers CD11c (which was detected as EYFP in a CD11c-EYFP transgenic mouse) and MHC-II. In the epithelium, a contiguous network of CD11c+MHC-II+ cells with dendritic processes was identified, which in concordance with studies on the phenotype of DCs from lung digests (27, 28) suggest that these cells are immature myeloid DCs. On the abluminal side of the smooth muscle layer, a network of MHC-II+CD11c– cells, possibly plasmacytoid DCs (29, 30) or lung B-cells (30), were identified. The full volume of CD11c+ structures, reflecting the amount of DCs in the airway wall was only moderately affected by allergic inflammation. In a previous study by Osterholzer and coworkers, the number of MHC-II+CD11cmod DCs from lung digests increased after antigen challenge of primed animals in a CCR2- and CCR6-dependent manner (28). Other groups found a robust increase in the number of MHC-II+ cells in the tracheal mucosa after allergen challenge (10, 31). However, we detected only a small increase in the amount of CD11c-EYFP+ cells at the most peripheral level of the main axial pathway. Since some of the DCs with a surface expression of CD11c might not have an active CD11c promoter, it is possible that not all airway mucosal DCs were detected in this system. This might explain why the data presented here differ from those previous studies. DCs were frequently contacted by sensory nerves containing both neuropeptides CGRP and SP along the whole length of the main axial pathway, providing an anatomical basis for the modulation of DC activity by these neuropeptides as described earlier (13–15, 32). To rule out the possibility of observing a coincidental phenomenon, we performed a three-dimensional quantitative co-localization analysis in a dynamic system (i.e., in an experimental allergic airway inflammation) in which functional interactions between these networks have previously been described (18). When the extent of DC–sensory nerve contact was quantified, the percentage of co-localized volumes of both networks showed an increase in the inflamed airways, especially at the distal part of the main axial pathway. In case of CGRP+ nerves, the increase of co-localization at the most distal region might have been caused by the recruitment of DCs. However, this increase in co-localization was measured at other locations as well with unchanged DC volumes, and more interestingly, the increase in the percentage of co-localized DC volume occurred exactly at the same regions. This altogether suggests that contact between DCs and sensory nerves increases during allergic inflammation. CGRP has been shown to induce the chemotaxis of immature DCs but inhibits their migration after maturation (13). Our data suggest the possibility that CGRP could be a chemoattractant for DCs during allergic airway inflammation. However, the potiental role of CGRP in attracting DCs during allergic airway inflammation need to be examined in future studies. CGRP suppresses DC activity (15) by down-regulating the expression of HLA-DR and the co-stimulatory molecule CD86 (14). In contrast, SP promotes DC function by activating the transcription factor NF-{kappa}B (32), which is essential for effective antigen presentation (33). Thereby, the activity of airway mucosal DCs (and accordingly local T-cell activation) could be delicately fine-tuned by contacting sensory nerves that release CGRP and SP upon nonspecific stimulation.

Since PGP 9.5 staining revealed not only nerves but also NEBs and airway ganglia, we studied their possible association with DCs. Strikingly, NEBs located at airway branching points and ganglia located at the entrance of the intrapulmonary airways frequently contained a few DCs. NEBs have been previously suggested to function as airway chemoreceptors (34, 35), detecting changes in the oxygen concentration of inhaled air. Upon hypoxia, they react with the release of the content of their dense-core vesicles containing several neuropeptides, also CGRP. Thus, the contact between CGRP-containing NEBs and DCs may suggest a mechanism by which the immune system would be notified if the ventilation of a certain part of the lung is compromised. DCs located inside parasympathetic airway ganglia might modulate the activity of ganglia neurons like mast cells do that have been described earlier to be located near them (36). Interestingly, these DCs were also contacted by CGRP+ nerve fibers innervating the ganglia themselves.

In the human airway mucosa, the abundance of HLA-DR+ DCs with cytoplasmic processes has been demonstrated earlier (5, 37, 38). These cells showed low autofluorescence and were extremely potent stimulators in mixed leukocyte reactions compared to alveolar macrophages (39).

We found HLA-DR+ DCs to be frequently contacted by CGRP+ sensory fibers. In addition, some DCs in the epithelium were associated with solitary neuroepithelial cells. Because both DCs and nerves were abundantly distributed in the lamina propria, we can not fully exclude the possibility of coincidental contacts. Still, our findings in the human airways, together with the data obtained in mice and functional studies of other groups, suggest the importance of DC–nerve interactions in humans as well. Cross-talk between DCs and various neural structures of the airways offers one potential connection between sensitization and allergic inflammation on the one hand and neural dysregulation on the other hand causing asthmatic symptoms like excessive mucus production, wheezing, and cough.


   Acknowledgments
 
The authors thank Michel C. Nussenzweig for providing them with the CD11c-EYFP-transgenic mice.


   Footnotes
 
This work was funded by the Deutsche Gesellschaft für Pneumologie (DGP), the German Academic Exchange Service (DAAD), DFG BR 2126/1-1 and SFB 587 B4.

This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org

Originally Published in Press as DOI: 10.1165/rcmb.2007-0087OC on June 28, 2007

Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.

Received in original form March 14, 2007

Accepted in final form May 30, 2007


   References
Top
 Abstract
 CLINICAL RELEVANCE
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Lamb JP, Sparrow MP. Three-dimensional mapping of sensory innervation with substance p in porcine bronchial mucosa: comparison with human airways. Am J Respir Crit Care Med 2002;166:1269–1281.[Abstract/Free Full Text]
  2. Belvisi MG. Sensory nerves and airway inflammation: role of A delta and C-fibres. Pulm Pharmacol Ther 2003;16:1–7.[CrossRef][Medline]
  3. Barnes PJ. Asthma as an axon reflex. Lancet 1986;1:242–245.[CrossRef][Medline]
  4. Barnes PJ. Neurogenic inflammation in the airways. Respir Physiol 2001;125:145–154.[CrossRef][Medline]
  5. Schon-Hegrad MA, Oliver J, McMenamin PG, Holt PG. Studies on the density, distribution, and surface phenotype of intraepithelial class II major histocompatibility complex antigen (Ia)-bearing dendritic cells (DC) in the conducting airways. J Exp Med 1991;173:1345–1356.[Abstract/Free Full Text]
  6. Iwasaki A, Medzhitov R. Toll-like receptor control of the adaptive immune responses. Nat Immunol 2004;5:987–995.[CrossRef][Medline]
  7. Fischer A, McGregor GP, Saria A, Philippin B, Kummer W. Induction of tachykinin gene and peptide expression in guinea pig nodose primary afferent neurons by allergic airway inflammation. J Clin Invest 1996;98:2284–2291.[Medline]
  8. Lambrecht BN, Hammad H. Taking our breath away: dendritic cells in the pathogenesis of asthma. Nat Rev Immunol 2003;3:994–1003.[CrossRef][Medline]
  9. Holt PG, Stumbles PA. Regulation of immunologic homeostasis in peripheral tissues by dendritic cells: the respiratory tract as a paradigm. J Allergy Clin Immunol 2000;105:421–429.[CrossRef][Medline]
  10. Huh JC, Strickland DH, Jahnsen FL, Turner DJ, Thomas JA, Napoli S, Tobagus I, Stumbles PA, Sly PD, Holt PG. Bidirectional interactions between antigen-bearing respiratory tract dendritic cells (DCs) and T cells precede the late phase reaction in experimental asthma: DC activation occurs in the airway mucosa but not in the lung parenchyma. J Exp Med 2003;198:19–30.[Abstract/Free Full Text]
  11. Lambrecht BN. Immunologists getting nervous: neuropeptides, dendritic cells and T cell activation. Respir Res 2001;2:133–138.[CrossRef][Medline]
  12. Dunzendorfer S, Wiedermann CJ. Neuropeptides and the immune system: focus on dendritic cells. Crit Rev Immunol 2001;21:523–557.[Medline]
  13. Dunzendorfer S, Kaser A, Meierhofer C, Tilg H, Wiedermann CJ. Cutting edge: peripheral neuropeptides attract immature and arrest mature blood-derived dendritic cells. J Immunol 2001;166:2167–2172.[Abstract/Free Full Text]
  14. Carucci JA, Ignatius R, Wei Y, Cypess AM, Schaer DA, Pope M, Steinman RM, Mojsov S. Calcitonin gene-related peptide decreases expression of HLA-DR and CD86 by human dendritic cells and dampens dendritic cell-driven T cell-proliferative responses via the type I calcitonin gene-related peptide receptor. J Immunol 2000;164:3494–3499.[Abstract/Free Full Text]
  15. Hosoi J, Murphy GF, Egan CL, Lerner EA, Grabbe S, Asahina A, Granstein RD. Regulation of langerhans cell function by nerves containing calcitonin gene-related peptide. Nature 1993;363:159–163.[CrossRef][Medline]
  16. Delgado M, Gonzalez-Rey E, Ganea D. The neuropeptide vasoactive intestinal peptide generates tolerogenic dendritic cells. J Immunol 2005;175:7311–7324.[Abstract/Free Full Text]
  17. Neuhuber WL, Tiegs G. Innervation of immune cells: evidence for neuroimmunomodulation in the liver. Anat Rec A Discov Mol Cell Evol Biol 2004;280:884–892.[Medline]
  18. Kradin R, MacLean J, Duckett S, Schneeberger EE, Waeber C, Pinto C. Pulmonary response to inhaled antigen: neuroimmune interactions promote the recruitment of dendritic cells to the lung and the cellular immune response to inhaled antigen. Am J Pathol 1997;150:1735–1743.[Abstract]
  19. Weichselbaum M, Everett AW, Sparrow MP. Mapping the innervation of the bronchial tree in fetal and postnatal pig lung using antibodies to PGP 9.5 and SV2. Am J Respir Cell Mol Biol 1996;15:703–710.[Abstract]
  20. Larson SD, Schelegle ES, Hyde DM, Plopper CG. The three-dimensional distribution of nerves along the entire intrapulmonary airway tree of the adult rat and the anatomical relationship between nerves and neuroepithelial bodies. Am J Respir Cell Mol Biol 2003;28:592–599.[Abstract/Free Full Text]
  21. Weichselbaum M, Sparrow MP, Hamilton EJ, Thompson PJ, Knight DA. A confocal microscopic study of solitary pulmonary neuroendocrine cells in human airway epithelium. Respir Res 2005;6:115.[CrossRef][Medline]
  22. Veres T, Rochlitzer S, Krauel K, Krug N, Braun A. Three-dimensional quantitative mapping of the spatial relationship between respiratory tract dendritic cells and sensory nerves in allergen sensitized and challenged mice [abstract]. ATS 2006 San Diego International Conference, Poster Discussion Session [B22].
  23. Lindquist RL, Shakhar G, Dudziak D, Wardemann H, Eisenreich T, Dustin ML, Nussenzweig MC. Visualizing dendritic cell networks in vivo. Nat Immunol 2004;5:1243–1250.[CrossRef][Medline]
  24. Path G, Braun A, Meents N, Kerzel S, Quarcoo D, Raap U, Hoyle GW, Nockher WA, Renz H. Augmentation of allergic early-phase reaction by nerve growth factor. Am J Respir Crit Care Med 2002;166:818–826.[Abstract/Free Full Text]
  25. Peake JL, Reynolds SD, Stripp BR, Stephens KE, Pinkerton KE. Alteration of pulmonary neuroendocrine cells during epithelial repair of naphthalene-induced airway injury. Am J Pathol 2000;156:279–286.[Abstract/Free Full Text]
  26. Dakhama A, Kanehiro A, Makela MJ, Loader JE, Larsen GL, Gelfand EW. Regulation of airway hyperresponsiveness by calcitonin gene-related peptide in allergen sensitized and challenged mice. Am J Respir Crit Care Med 2002;165:1137–1144.[Abstract/Free Full Text]
  27. von Garnier C, Filgueira L, Wikstrom M, Smith M, Thomas JA, Strickland DH, Holt PG, Stumbles PA. Anatomical location determines the distribution and function of dendritic cells and other APCs in the respiratory tract. J Immunol 2005;175:1609–1618.[Abstract/Free Full Text]
  28. Osterholzer JJ, Ames T, Polak T, Sonstein J, Moore BB, Chensue SW, Toews GB, Curtis JL. CCR2 and CCR6, but not endothelial selectins, mediate the accumulation of immature dendritic cells within the lungs of mice in response to particulate antigen. J Immunol 2005;175:874–883.[Abstract/Free Full Text]
  29. de Heer HJ, Hammad H, Soullie T, Hijdra D, Vos N, Willart MA, Hoogsteden HC, Lambrecht BN. Essential role of lung plasmacytoid dendritic cells in preventing asthmatic reactions to harmless inhaled antigen. J Exp Med 2004;200:89–98.[Abstract/Free Full Text]
  30. Masten BJ, Lipscomb MF. Comparison of lung dendritic cells and B cells in stimulating naive antigen-specific T cells. J Immunol 1999;162:1310–1317.[Abstract/Free Full Text]
  31. van Rijt LS, Prins JB, Leenen PJ, Thielemans K, de Vries VC, Hoogsteden HC, Lambrecht BN. Allergen-induced accumulation of airway dendritic cells is supported by an increase in CD31(Hi)Ly-6C(Neg) bone marrow precursors in a mouse model of asthma. Blood 2002;100:3663–3671.[Abstract/Free Full Text]
  32. Marriott I, Mason MJ, Elhofy A, Bost KL. Substance P activates NF-KappaB independent of elevations in intracellular calcium in murine macrophages and dendritic cells. J Neuroimmunol 2000;102:163–171.[CrossRef][Medline]
  33. Yoshimura S, Bondeson J, Foxwell BM, Brennan FM, Feldmann M. Effective antigen presentation by dendritic cells is NF-KappaB dependent: coordinate regulation of MHC, co-stimulatory molecules and cytokines. Int Immunol 2001;13:675–683.[Abstract/Free Full Text]
  34. Peers C, Kemp PJ. Acute oxygen sensing: diverse but convergent mechanisms in airway and arterial chemoreceptors. Respir Res 2001;2:145–149.[Medline]
  35. Adriaensen D, Brouns I, Van Genechten J, Timmermans JP. Functional morphology of pulmonary neuroepithelial bodies: extremely complex airway receptors. Anat Rec A Discov Mol Cell Evol Biol 2003;270:25–40.[CrossRef][Medline]
  36. Myers AC, Undem BJ, Weinreich D. Influence of antigen on membrane properties of guinea pig bronchial ganglion neurons. J Appl Physiol 1991;71:970–976.[Abstract/Free Full Text]
  37. Tschernig T, de Vries VC, Debertin AS, Braun A, Walles T, Traub F, Pabst R. Density of dendritic cells in the human tracheal mucosa is age dependent and site specific. Thorax 2006;61:986–991.[Abstract/Free Full Text]
  38. Vermaelen K, Pauwels R. Pulmonary dendritic cells. Am J Respir Crit Care Med 2005;172:530–551.[Abstract/Free Full Text]
  39. Nicod LP, Lipscomb MF, Toews GB, Weissler JC. Separation of potent and poorly functional human lung accessory cells based on autofluorescence. J Leukoc Biol 1989;45:458–465.[Abstract]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Online Supplement
Right arrow All Versions of this Article:
2007-0087OCv1
37/5/553    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Veres, T. Z.
Right arrow Articles by Braun, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Veres, T. Z.
Right arrow Articles by Braun, A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Proc. Am. Thorac. Soc. Am. J. Respir. Crit. Care Med.
Copyright © 2007 American Thoracic Society.