Abstract
Recent evidence suggests that the previously enigmatic cell type designated plasmacytoid monocytes can function as dendritic cells and contribute substantially to both innate and adaptive immunity. This cell type has previously been described only in bone marrow, blood, and organized lymphoid tissue, but not at effector sites with direct Ag exposure such as the mucosae. Plasmacytoid dendritic cells (P-DCs) matured in vitro can induce T cells to produce allergy-promoting Th2 cytokines; therefore, their possible occurrence in nasal mucosa during experimentally elicited allergic rhinitis was examined. Patients with silent nasal allergy were challenged topically with relevant allergen daily for 7 days. Biopsy specimens as well as blood samples were obtained before and during such provocation, and P-DCs were identified by their high expression of CD123 (IL-3R α-chain), together with CD45RA. Our results showed that P-DCs were present in low and variable numbers in normal nasal mucosa but increased dramatically during the allergic reaction. This accumulation concurred with the expression of the L-selectin ligand peripheral lymph node addressin on the mucosal vascular endothelium. The latter observation was particularly interesting in view of the high levels of L-selectin on circulating P-DC precursors and of previous reports suggesting that these cells can enter organized lymphoid tissue via high endothelial venules (which express peripheral lymph node addressin constitutively). Together, our findings suggested that P-DCs are involved in the triggering of airway allergy and that they are directed to allergic lesions by adhesion molecules that normally mediate leukocyte extravasation in organized lymphoid tissue.
More than 40 years ago, Lennert and Remmele (1) identified a previously unrecognized rare cell type in the paracortical areas of reactive lymph nodes with a morphology similar to that of plasma cells. They were later found to express T cell and monocyte markers (2, 3) and were therefore designated plasmacytoid T cells or plasmacytoid monocytes. This distinct subset has been observed in high numbers in organized lymphoid tissue afflicted with reactive or neoplastic disorders (2, 4, 5, 6, 7, 8), but its functional significance remains elusive. Recently, however, novel properties have been reported for this cell type, suggesting important immunological roles. Notably, it was shown that plasmacytoid monocytes are identical with immature dendritic cells (DCs)3 (3) present in human peripheral blood and tonsils (9, 10). When these cells were cultured with IL-3 and CD40 ligand, they developed dendritic morphology and acquired the ability to activate naive T cells (9, 10). Moreover, such circulating immature DCs were found to represent a small leukocyte population that produces large amounts of IFN-α in the presence of certain viruses (11). Two other reports have confirmed that plasmacytoid monocytes, natural IFN-producing cells, and immature DCs in peripheral blood and tonsils are the same cell type (12, 13). Therefore, this distinct leukocyte subset, termed plasmacytoid DC (P-DC), most likely has an important role both in innate defense against pathogens and as APCs in the adaptive immune system.
Despite such functional data obtained with isolated P-DCs, little is known about the properties of this cell type in vivo. P-DCs have been identified in bone marrow, blood, and organized lymphoid tissue, but not at effector sites with direct Ag exposure, such as the skin and mucosae. Therefore, in the present study, we examined the possible occurrence of P-DCs in human nasal mucosa, particularly during experimentally induced allergic rhinitis. This model should have special relevance to the function of P-DCs in vivo, because when matured in vitro, these cells can induce naive T cells to produce allergy-promoting Th2 cytokines (14). In our experimental model, extracts of grass or birch pollen were administered intranasally to allergic rhinitis patients out of season as well as to nonallergic volunteers. We found that P-DCs, identified by their high levels of CD123 (IL-3R α-chain) (10), were few in normal nasal mucosa but increased dramatically during the allergic reaction. This event coincided with the expression of the L-selectin ligand peripheral lymph node addressin (PNAd) on the nasal vascular endothelium, which was interesting because circulating P-DC precursors express high levels of L-selectin (10, 12). Also, it has been reported that P-DCs enter organized lymphoid tissue via high endothelial venules (HEVs), where PNAd is constitutively expressed. Together, our results suggested that P-DCs are involved in the triggering of airway allergy and might be directed to allergic lesions by adhesion molecules that normally mediate leukocyte extravasation in organized lymphoid tissue.
Materials and Methods
Test subjects
Medical students (n = 120) were tested by skin prick test with a panel of 8 common aeroallergens and examined for allergic symptoms; 13 with allergy (age 22–29 years, 8 men and 5 women) and 7 without allergy (age 22–27 years, 5 men and 2 women) volunteered to participate in the study. Of the allergic subjects, 8 reacted with a strong positive skin response against grass and 5 reacted against birch; all had experienced typical nasal symptoms during the pollen season for >3 years. The nonatopic subjects had a negative skin prick test and no history of allergic symptoms. None of the subjects had a history of upper respiratory tract infection during the preceding 4 wk, nasal polyps, nasal surgery, or nasal deformities; neither had they taken any medication for at least 1 month before the study. All volunteers were nonsmokers and free of nasal symptoms before allergen challenge, which was performed outside the pollen season.
Study design
Study I. On day 0, a mucosal biopsy specimen was obtained from the lower turbinate of one nostril. Thereafter, the opposite nostril was challenged with relevant allergen, and clinical symptoms were subsequently recorded (see below). On day 1, another biopsy specimen was obtained from the lower turbinate of the challenged side. All subjects (patients and controls) sprayed the same nostril in a similar manner once daily for 7 days (see below). On day 8, both nostrils were subjected to biopsy of the lower turbinate.
Study II. One year later, the same test subjects (11 allergic subjects and 5 controls) participated in a study with a similar challenge protocol as described above. All subjects sprayed both nostrils daily for 7 days and peripheral blood samples were obtained on days 0, 2, 5, and 8. No biopsy was performed in study II. Both studies were approved by the National Ethics Committee, and informed written consent was obtained from each participating subject.
Nasal allergen challenge
A hand-driven pump spray was used to deliver a defined volume (50 μl) of relevant allergen solution (Aquagen timothy or Aquagen birch, ALK, Horsholm, Denmark) in the nostrils. On the first day of challenge in study I, the allergic patients were exposed unilaterally to increasing concentrations of allergen (10,000–100,000 SQ-U/ml) until a threshold dose that elicited typical acute rhinitis symptoms was established. This dose was applied in the same nostril daily for 7 days in study I and in both nostrils in study II (see above). The control subjects were exposed to the highest allergen concentration used for patients receiving Aquagen timothy. The severity of symptoms (nasal blockage, nasal discharge, and sneezing) was recorded daily on a 4-point scale (0–3) with a maximum score of 9. The median sum of individual symptom scores on each day of challenge was >6 in the patient group, and <1 in the controls.
Preparation of nasal biopsy specimens
Mucosal specimens were obtained from the lower edge of the inferior turbinate, 1.5–2 cm posterior to the front edge, by means of a Gerritsma forceps with a cup diameter of 2.5 mm. Local anesthesia was induced by placing a cotton wool carrier with 7 mg/ml tetracain/0.3 mg/ml adrenalin under the inferior turbinate adjacent to the biopsy site. Tissue samples from the same nostril were taken 0.5 cm apart. All specimens were immediately placed on a thin slice of carrot for appropriate orientation and handling, embedded in OCT (Tissue-Tek, Miles Laboratories, Elkhart, IN), and snap-frozen in liquid nitrogen as detailed elsewhere (15).
Preparation of tonsillar tissue specimens
Samples of palatine tonsils were obtained from patients operated for recurrent tonsillitis; handling and freezing were performed as described above.
Preparation of blood samples
Peripheral blood was collected by vein puncture with Vacutainers containing EDTA. Whole blood was within 2 h mixed with a NH4Cl solution (16) at room temperature for 2–5 min to lyse the erythrocytes. The leukocytes were sedimented by centrifugation, washed once with PBS containing 1% FCS, and then subjected to multicolor immunofluorescence staining (see below). Differential counts were performed at the Department of Clinical Chemistry with an automated counting system (Abbott Diagnostic Division, Irving, TX).
In vitro stimulation of PBMCs
Peripheral blood from normal donors was collected as described above. PBMCs were separated by Lymphoprep (Nycomed Pharma, Oslo, Norway), washed twice in PBS, and resuspended at a concentration of 1 × 106 cells/ml in RPMI-10% FCS. The cells were then incubated overnight with or without 1000 IU/ml recombinant human IFN-α-2b (Introna, Schering-Plough, Madison, NJ). The cells were then washed twice in PBS, and cytospins (100 μl) were prepared (1 × 106 cells/ml). These preparations were air-dried overnight and acetone-fixed for 10 min at room temperature.
Multicolor immunofluorescence staining
To determine tissue density, phenotype, and proliferation of CD123high cells and the expression of PNAd by mucosal vessel walls in situ, we applied a multicolor immunostaining technique to acetone-fixed serial cryosections (4 μm) as described elsewhere (15). Briefly, a mouse mAb of the IgG2a subclass specific for human CD123 (clone 7G3, 2 μg/ml; PharMingen, San Diego, CA) was combined with mouse mAb to either: CD4 (clone SK3 + SK4, IgG1, 10 μg/ml; Becton Dickinson Immunocytometry Systems (BDIS), San Jose, CA); CD11c (clone KB90, IgG1, 1/10; gift from Dr. K. Pulford, Oxford, U.K.); CD45RA (clone L48, IgG1, 1/10; BDIS); CD68 (clone KP1, IgG1, 1/500, Dako, Glostrup, Denmark); FcεRI (clone 15-1, IgG1, 1/500; gift from Dr. J. P. Kinet, Bethesda, MD); or HLA-DR (clone HL39, IgG3, 1/100; Sanbio, Uden, The Netherlands). In some experiments, we also combined a mAb of the IgG1 subclass specific for human CD123 (clone 9F5, 2 μg/ml; PharMingen) with mAbs to either: CD1a (Clone NA1/34, IgG2a, 1/10; Dako); CD14 (clone RM052, IgG2a, 3 μg/ml; Biosys, Compeigne, France); or CD45R0 (clone UCHL-1, IgG2a, 1/10; gift from Dr. P. C. L. Beverly, London, U.K.). All these paired mAb mixtures were applied for 1 h at room temperature to serial sections. Combinations of Cy3-labeled (“red”) goat anti-mouse IgG2a (1. 5 μg/ml) and biotinylated subclass-specific goat anti-mouse IgG1 (10 μg/ml) or IgG3 (10 μg/ml) (all from Southern Biotechnology Associates, Birmingham, AL), mixed with rabbit antiserum to human cytokeratin (1/100; authors’ laboratory), were next applied for 1.5 h and followed by 7-amino-4-methylcoumarin-3-acetic acid (AMCA)-labeled (“blue”) goat anti-rabbit IgG (7.5 μg/ml; Vector Laboratories, Burlingame, CA) mixed with “green” Cy2-streptavidin (1 μg/ml; Amersham, Galesbury, U.K.) for 30 min. The blue cytokeratin staining was included for delineation of epithelial elements (17). Immunostaining of T cells was performed as detailed elsewhere (15).
To examine cell proliferation in situ, we performed immunostaining for the nuclear proliferation marker Ki-67 Ag. A mixture of mAb to CD123 (IgG2a) and mAb Ki-67 (IgG1, 1/50; DAKO) was incubated on cryosections for 1 h, followed by incubation with FITC-labeled goat anti-mouse IgG2a (10 μg/ml) and Cy3-labeled goat anti-mouse IgG1 (2 μg/ml; both from Southern Biotechnology Associates).
In situ IFN-α production was evaluated by immunostaining for MxA, which is an IFN-α-inducible intracellular protein well established as a “surrogate” marker for local IFN-α production (18, 19, 20, 21). Acetone-fixed serial cryosections were immersed in 4% paraformaldehyde for 5 min at 4°C to reduce protein leaching during immunohistochemistry. After a brief rinse, the sections were incubated with a combination of mouse mAb to MxA (clone M143, IgG2a, 1.5 μg/ml; courtesy Dr. O. Haller, Freiburg, Germany) and rabbit antiserum to human cytokeratin overnight at room temperature, followed by Cy3-labeled goat anti-mouse IgG (0.8 μg/ml, Jackson ImmunoResearch Laboratories, West Grove, PA) and AMCA-labeled goat anti-rabbit IgG. Adjacent sections were costained for CD123 and CD45RA (see above) to reveal a possible correlation between the density of MxA-expressing cells and P-DCs in nasal mucosa. As a positive control for MxA expression, cytospins of IFN-α-stimulated PBMCs (see above) were fixed and immunostained as described above. For photographic documentation, paraformaldehyde-fixed tissue sections and cytospins were incubated with a combination of anti-MxA (IgG2a) and anti-CD123 (IgG1) followed by Cy3-labeled goat anti-mouse IgG2a and FITC-labeled goat anti-mouse IgG1.
Paired fluorescence determination of the proportion and staining intensity of mucosal vessels expressing PNAd was performed with rat mAb MECA-79 applied for 1 h (IgM, 1/30; courtesy of Dr. E. C. Butcher, Stanford, CA) followed by Cy-3-conjugated (“red”) goat anti-rat IgM (2 μg/ml, Jackson ImmunoResearch Laboratories) mixed with FITC-labeled (“green”) Ulex europaeus lectin-1 (2 μg/ml; Vector Laboratories) for 30 min. By this approach, all vessels were decorated green, and those reactive with MECA-79 showed a mixed (yellow) color.
In all staining experiments, negative controls were obtained both by omission of primary mAbs and by incubation with irrelevant isotype- and concentration-matched primary mAbs.
Immunofluorescence microscopy of P-DCs and MECA-79-reactive vessels
The immunostained tissue sections were blindly examined by the same investigator (F.L.J.) at ×400 magnification in a fluorescence microscope (Model E800, Nikon, Tokyo, Japan). To determine relevant cell density, all immunostained stromal cells (positive for one or both of the individual markers) were counted to a mucosal depth of 242 μm by superimposing a grid (242 × 242 μm) parallel to the basement membrane of the surface epithelium. At least two sections from each specimen were examined with a combination of mAbs to CD123 and CD45RA (usually a total area of more than 1 mm2) to determine the tissue density of P-DCs. Selected specimens were examined with the other marker combinations.
To evaluate endothelial MECA-79 reactivity, a scoring system was established to combine the extent and staining intensity of immunoreactive vessels. The vessels were divided into two groups according to their smallest outer diameter (<15 μm or ≥15 μm, respectively) outlined by U. europaeus lectin-1 staining. All detectable vessels, to a stromal depth of 484 μm, were counted and graded with regard to the staining intensity for MECA-79 on an arbitrary scale from nil (−) to strong (++). For the percentage of MECA-79-reactive vessels, the following scores were assigned: 0, no positive vessels; 1, 1–9% positive vessels; 2, 10–24% positive vessels; 3, 25–40% positive vessels; and 4, >40% positive vessels. The score for staining intensity was related to the percentage of U. europaeus lectin-1-positive vessels deemed to react strongly (++) with MECA-79: 1, < 2% vessels; 2, 2–20% vessels; and 3, >20% vessels. The product of these two values for numbers and intensity of immunoreactive vessels was then used as the finally assigned score for each specimen, which thus could have a possible range from 0 to 12. More than 100 vessels with a diameter of <15 μm, and 50 with a diameter of ≥15 μm, were counted in every specimen.
Flow cytometric analysis
P-DC precursors in peripheral blood were identified by multicolor immunofluorescence as previously described (10). After lysis of RBC, leukocytes were labeled with PE-conjugated anti-CD123, peridinin chlorophyll protein-conjugated anti-HLA-DR, and a mixture of FITC-conjugated Abs to lineage markers that are weakly expressed on CD123high cells (CD3, CD14, CD16, CD20, CD56). These reagents (part of a kit designed to identify circulating DCs) were applied as recommended by the manufacturer (BDIS). Ab-tagged cells were examined in a FACScalibur flow cytometer (BDIS), and the data obtained were blindly analyzed with Paint-a-Gate software (BDIS). The size of the CD123high cell population was determined as the percentage of PBMCs identified by light scatter parameters.
Statistics
A Wilcoxon matched pairs sign rank sum test was performed to compare the two test groups with regard to the number of immunostained cells and the score for MECA-79-reactive vessels in nasal mucosa, as well as the number of leukocyte subsets in peripheral blood at various time points.
Results
Phenotypic characterization of P-DCs in tonsils
To verify our ability to identify reliably P-DCs in situ, we applied various combinations of mAbs to tonsillar cryosections. As previously reported (10), a population of CD123high cells was specifically decorated in the T cell areas. Paired immunofluorescence staining showed that virtually all of these cells distinctly coexpressed CD4, CD45RA, CD68, and HLA-DR (Fig. 1⇓, a–d). Thus, the decorated cells had the same phenotype as previously described for plasmacytoid monocytes (2, 9) or immature CD123high DCs (10) in tonsils. Also, as previously reported for extravasated plasmacytoid monocytes (9, 12), the identified P-DCs were found in close proximity to HEVs (Fig. 1⇓, a and b).
In situ phenotypic characterization of CD123high P-DCs. Paired immunofluorescence staining for CD123 (Cy3, red) and CD4, CD45RA, CD68, or HLA-DR (FITC, green) in serial cryosections of tonsil (a–d) or allergic nasal mucosa (e–h). Virtually all CD123high cells coexpress CD4, CD45RA, CD68, and HLA-DR in both types of tissue as shown by mixed (yellow) color. Note strongly CD123+ high endothelial venules (a and b) and weakly positive vessels in nasal mucosa (e and f) revealed by their pure red color (∗, vessel lumen). Magnification, ×630.
A few P-DCs normally occur in nasal mucosa
Variable numbers of CD123highCD45RA+ cells were identified in the lamina propria of most samples of histologically normal nasal mucosa from unchallenged allergic as well as nonallergic individuals (Fig. 2⇓). Interestingly, occasional allergic patients had quite high numbers of these cells (Fig. 2⇓). Conversely, the epithelium always contained only few cells of this phenotype in both test groups (median, 1 cell/mm basement membrane; range, 0–5).
Density of lamina propria CD123high P-DCs (numbers/mm2) in cryosections of nasal mucosa from controls (A) and patients with nasal allergy (B) before and after allergen challenge as indicated. Cell counts from the same individual are connected by lines. See Materials and Methods for detailed description of experimental protocol. Medians are indicated by heavy broken horizontal lines. Statistical comparison was performed by Wilcoxon matched pairs sign rank sum test.
Numerous P-DCs accumulate in nasal mucosa of allergic subjects after topical allergen challenge
To investigate how the CD123highCD45RA+ P-DCs respond to allergen, we used an established human experimental model for atopic hypersensitivity (22). On day 1 after topical challenge, the density of these cells was unchanged in both test groups (Fig. 2⇑). However, after daily provocations for an additional 6 days, their density increased significantly in the lamina propria of the allergic subjects, whereas it remained unchanged in the controls (Fig. 2⇑). Large individual variations were observed, some allergic subjects exhibiting a very striking increase in the number of P-DCs. Parallel histological examination by hematoxylin and eosin staining revealed accumulation of additional mononuclear cells and eosinophils in challenged nasal mucosa. Further staining experiments showed that most of the former were CD4+ T cells of the memory (CD45R0+) phenotype (data not shown). Interestingly, however, when we immunostained for the CD123highCD45RA+ P-DCs, the density of such double-positive cells coincided with abundant accumulation of mononuclear cells positive only for CD45RA (Fig. 1⇑f). The latter turned out to be naive (CD3+CD45RA+) T cells (data not shown), a phenotype that hardly occurs in normal nasal mucosa (15).
To ensure that the recruited CD123highCD45RA+ cells indeed were P-DCs, we performed additional immunostaining experiments that enabled us to show that >98% of them coexpressed CD4, CD68, and HLA-DR as described above for the tonsillar counterparts (Fig. 1⇑, e–h). Moreover, they did not express markers such as CD11c, CD14, CD20, and FcεRI, thus confirming that they were distinct from other types of immature DCs, monocytes, B cells, and basophils (data not shown). Although P-DCs were present intraepithelially in relatively small numbers, even after 7 days of allergen challenge, such challenge resulted in a significant increase (p = 0.003) of this subset intraepithelially in the allergic test group (median, 8; range, 1–15) compared with the situation on day 1 (median, 0; range, 0–7).
P-DCs are distinct from CD1a+ cells in nasal mucosa
A previous study (22) based on a similar experimental model for allergic rhinitis showed that CD1a+ DCs accumulated in the lesion after allergen challenge with kinetics similar to that shown here for the P-DCs. However, paired immunostaining revealed that the CD123high DC subset always was negative for CD1a (Fig. 3⇓a). Interestingly, the CD1a+ cells and the CD123high cells were distributed differently in nasal mucosa; the former subset predominantly occurred in the surface epithelium as previously reported (22), whereas P-DCs were mainly located in the lamina propria. Our finding suggested that these two phenotypically distinct DC subsets accumulate within different tissue compartments of allergic airway mucosa in response to allergen challenge, which might reflect possible functional differences.
In situ phenotypic characterization of CD123high P-DCs and related PNAd and MxA expression. Multicolor immunofluorescence staining: for CD123 (Cy3, red), CD1a (Cy2, green), and epithelial cytokeratin (AMCA, blue) (a); for CD123 (FITC, green) and Ki-67 Ag (Cy3, red) (b); with mAb MECA-79 (Cy3, red) for PNAd and Ulex europaeus lectin-1 (FITC, green for endothelium) (c); and for CD123 (FITC, green) and MxA (Cy3, red) (d) in serial cryosections of nasal mucosa from patient with nasal allergy after 7 days of topical allergen challenge. e and f, Immunofluorescence staining for MxA on cytospins of PBMCs that first had been incubated overnight with (e) or without (f) 1000 IU/ml recombinant human IFN-α-2b. a, Note intra- and subepithelial CD1a+ DCs that do not express CD123. a and b, Weakly CD123+ vessels are indicated by luminal asterisks. c, Numerous medium-sized vessels (diameter, ≥15 μm) are strongly PNAd expressing as shown by their mixed yellow color, whereas virtually all vessels with diameters of <15 μm are negative (green only). Basement membrane of surface epithelium is indicated by broken line. Some epithelial cells react with MECA-79 as also noted by others (26 ). Inset, High magnification shows that PNAd is predominantly localized at the luminal face of the endothelial cells. d–f, Immunostaining for MxA was performed in parallel with the same protocol, and the three pictures were taken with the same exposure time for red emission. Magnification: a and b, × 630; c, ×400. Inset, ×1200); e and f, ×400.
Accumulation of P-DCs in nasal mucosa does not reflect local proliferation
The observed accumulation of P-DCs in allergen-challenged nasal mucosa could be due either to increased precursor release from the bone marrow, increased extravasation, or local proliferation. To examine the latter possibility, we determined cellular expression of the nuclear proliferation marker Ki-67 Ag. This marker was strongly expressed in some epithelial cells and scattered cells in lamina propria, but no CD123high cells were deemed to be positive (Fig. 3⇑b). Their counterparts in tonsils were also negative for Ki-67 Ag, as previously shown (8). Therefore, it is unlikely that accumulation of P-DCs at either tissue site was caused by local proliferation.
No increase of circulating P-DC precursors in allergic subjects during allergen challenge
Recent information suggests active participation of the bone marrow and the hematopoietic processes in response to allergen challenge of the airways, resulting in increased numbers of circulating inflammatory cells such as eosinophils (23). Proliferative precursors of P-DCs are found in the bone marrow (10). Therefore, we investigated whether an increase of circulating CD123high precursors could be detected in our experimental allergy model. In allergic patients, the number of eosinophils was significantly higher (p = 0.002) at all time points during the provocation compared with day 0, whereas it remained unchanged in controls. By contrast, no significant increase was observed in the number of circulating CD123high precursors during the challenge period. Before challenge, CD123high cells constituted on average 0.40 ± 0.17 and 0.47 ± 0.22% of the total number of circulating PBMCs in allergic patients and control subjects, respectively, and remained lower than 0.5% in both groups throughout the experiment (without any change in the absolute number of PBMCs). Thus, the frequency of P-DC precursors in peripheral blood agreed with those of previous reports (10, 24), and our finding suggested that the accumulation of this immature DC subset in allergen-challenged nasal mucosa primarily depended on local mechanisms.
Increased endothelial PNAd expression in allergen-challenged nasal mucosa of allergic patients
Little is known about the mechanisms directing the emigration of circulating DC precursors to various tissue sites. However, accumulation of the immature CD123high DC subset in and around HEVs in lymphoid organs (and especially in inflamed lymph nodes), suggests that P-DCs extravasate through the specialized high endothelium (9, 12). In support of this hypothesis, circulating P-DC precursors express high levels of L-selectin (10, 12), an adhesion molecule that together with PNAd form a homing receptor-endothelial ligand pair involved in lymphocyte trafficking via HEVs. Therefore, we examined vascular PNAd expression in nasal mucosa before and after allergen challenge by immunostaining for MECA-79 reactivity (25, 26).
The total score of MECA-79 reactivity was significantly increased in nasal mucosa challenged for 7 days compared with biopsy specimens obtained from the same nostril on day 1 (p = 0.0017), as well as from the control nostril (p = 0.03), in all allergic patients. The reactivity was confined to vessels with a diameter of ≥15 μm, and in some samples from challenged tissue >40% of such vessels were strongly positive (Fig. 3⇑c). Interestingly, most samples of nasal mucosa from controls displayed some MECA-79-positive medium-sized vessels similar to that shown for unchallenged allergic patients. This finding was unexpected because MECA-79 reactivity is normally confined to HEVs in organized lymphoid tissue (25, 26).
P-DC accumulation and MxA expression are unrelated in allergen-challenged nasal mucosa
Recent studies have shown that circulating P-DC precursors correspond to the so-called natural IFN-producing cells which produce high levels of IFN-α in response to viruses and bacteria (12, 13). Therefore, we wanted to examine whether these cells reacted in a similar manner when occurring in the allergic lesion. MxA is an IFN-α-inducible protein (18, 19, 20, 21), and we used the expression of this intracellular molecule as a surrogate marker for IFN-α production in situ. Immunostaining for MxA was performed on paraformaldehyde-fixed sections of tonsils (n = 3) and nasal biopsy specimens obtained on days 0 and 7 (challenged nostril) from allergic (n = 4) and nonallergic (n = 3) subjects. Parallel sections from both groups showed representative densities of P-DCs. Only tonsillar tissue sections and nasal mucosa from one allergic patient (on both day 0 and day 7) were deemed to be weakly positive for MxA, whereas the other nasal specimens were negative for this marker (Fig. 3⇑d, and data not shown). IFN-α-stimulated PBMCs served as a strong positive control for MxA expression (Fig. 3⇑e), whereas unstimulated PBMCs were negative (Fig. 3⇑f). This finding suggested that P-DCs did not produce substantial amounts of IFN-α in allergen-challenged nasal mucosa of allergic patients.
Discussion
This study is apparently the first to show that P-DCs, identified as CD123 high CD45RA+ cells, can occur at a mucosal effector site. Recent information has suggested that this enigmatic cell type plays an important role in innate as well as adaptive immunity. The precursors isolated from peripheral blood are able to produce large amounts of IFN-α in response to viruses and bacteria (13, 27), whereas they, after differentiation to DCs in vitro, effectively can stimulate naive T cells to produce immunoregulatory cytokines (14, 27, 28). However, little is known about the function of P-DCs in vivo. Contrary to other DC types, P-DCs have previously been identified only in bone marrow, peripheral blood, and organized lymphoid tissue but not at epithelial surfaces where immature DCs are presumed to be involved in surveillance against microbial Ags (29). It was therefore of considerable interest that these cells accumulated abundantly in nasal mucosa of experimentally challenged allergic rhinitis patients. This finding clearly demonstrated that P-DCs can migrate to a site of inflammation outside organized lymphoid tissue.
Our results suggested that P-DCs are involved in the pathogenesis of nasal allergy. As differentiated P-DCs, these cells appear to have an inherent capacity to induce naive T cells to produce allergy-promoting Th2 cytokines (14), which are believed to play a key role in allergic rhinitis (30). Alternatively, the P-DCs can produce large amounts of IFN-α (11, 12, 13), a cytokine associated with Th1-type immunity (31). Production of IFN-α could thus play a role in counteracting a Th2-polarized allergic microenvironment. Therefore, we examined whether P-DCs in the allergic lesion had produced IFN-α in situ. MxA is an IFN-α-inducible cytoplasmic protein that mediates resistance to viruses, and its detection is frequently used as a surrogate marker for local IFN-α production (19, 20). However, only low and inconsistent levels of nasal MxA expression were detected. Thus, it is unlikely that P-DCs produced IFN-α locally to limit the allergic reaction.
Although little is known about the in vivo properties of P-DCs, recent in vitro data strongly suggest that these cells may participate in various types of inflammatory disorders besides allergy. Indeed, we have recently identified P-DCs in certain chronic inflammatory skin diseases known to be associated with local MxA expression, such as lupus erythematosus and lichen planus (L. Farkas, K. Beiske, F. Lund-Johansen, P. Brandtzaeg, and F. L. Jahnsen, manuscript in preparation), suggesting that they also can produce IFN-α in vivo. Interestingly, however, P-DCs were virtually absent from various nonallergic chronic inflammatory disorders occurring in nasal or intestinal mucosa, such as nasal polyps, celiac disease, and inflammatory bowel disease (F. L. Jahnsen, L. Farkas, H. S. Carlsen, and P. Brandtzaeg, unpublished observations). Thus, tissue accumulation of P-DCs appears to be restricted to certain lesions and is neither an exclusive nor a common feature of mucosal inflammation. To gain further knowledge about the possible role of P-DCs in human pathology, we are currently examining the distribution of this cell type and the local expression of MxA in a variety of inflammatory lesions.
HEVs express L-selectin ligands (also called PNAd, identifiable by mAb MECA-79) that mediate L-selectin-dependent rolling of naive lymphocytes. Circulating P-DC precursors express high levels of L-selectin (10, 12), which most likely explains their extravasation preference in HEV-containing organized lymphoid tissue. Interestingly, therefore, we found that challenged allergic nasal mucosa, in which P-DCs accumulated abundantly, contained many medium-sized strongly MECA-79-reactive blood vessels. This finding was intriguing because MECA-79 reactivity is normally restricted to organized lymphoid tissue (26). A role for PNAd in leukocyte recruitment to the allergic nasal lesion was further supported by the finding that CD45RA+ T cells were abundantly present in the same allergen-challenged biopsy specimens (data not shown). Leukocyte extravasation is a multistep process, involving several adhesion molecules and chemoattractants (32). Previous studies have shown that endothelial expression of both ICAM-1 and VCAM-1 are increased in a similar human model for allergic rhinitis (33, 34). P-DC precursors express ligands for both endothelial receptors (10, 12), suggesting that they, together with PNAd, could be involved in the extravasation of P-DCs in nasal allergy. The likely additional requirements for selected chemokines is currently under investigation in our laboratory. Endothelial PNAd expression has likewise been demonstrated in various chronic inflammatory skin disorders (26, 35), which supports the notion that this addressin is involved in leukocyte recruitment to sites of inflammation. To further elucidate this possibility, we are currently examining whether accumulation of P-DCs in certain inflammatory skin lesions, as discussed above, might be related to local endothelial MECA-79 reactivity.
We also tested the possibility that increased precursor release from the bone marrow, or local proliferation, could have contributed to the increased number of P-DCs in nasal mucosa after allergen challenge. A significant elevation of circulating eosinophils suggested that the bone marrow was stimulated by this challenge, but the number of circulating P-DC precursors remained unaltered (0.4–0.5% of all PBMCs) at a level similar to that previously reported (10, 24, 36). This observation, together with the fact that the nasal P-DCs did not express the proliferation marker Ki-67 Ag, suggested that neither precursor release from the bone marrow nor local proliferation contributed significantly to their accumulation in the allergic lesion.
In summary, P-DCs that previously have been observed only in bone marrow, peripheral blood, and organized lymphoid tissue were found to extravasate abundantly into nasal mucosa during an induced allergic reaction. Because this DC type can promote a Th2 response in vitro (14), our results suggested the interesting possibility that they contribute to airway allergy.
Acknowledgments
We thank the technical staff at Laboratory for Immunohistochemistry and Immunopathology for excellent assistance and Hege E. Bryne for help with the manuscript revision.
Footnotes
-
↵1 This work was supported by the Norwegian Cancer Society, the Research Council of Norway, the Research Fund for Asthma and Allergy, the Red Cross Research Fund for Children with Asthma and Allergy, and the Norwegian Foundation for Health and Rehabilitation.
-
↵2 Address correspondence and reprint requests to Dr. Per Brandtzaeg, Laboratory for Immunohistochemistry and Immunopathology, Institute of Pathology, Rikshospitalet, N-0027 Oslo, Norway. E-mail address: per. brandtzaeg{at}labmed.uio.no
-
↵3 Abbreviations used in this paper: DC, dendritic cell; P-DC, plasmacytoid DC; PNAd, peripheral lymph node addressin; HEV, high endothelial venule; AMCA, 7-amino-4-methylcoumarin-3-acetic acid; BDIS, Becton Dickinson Immunocytometry Systems.
- Received January 10, 2000.
- Accepted July 18, 2000.
- Copyright © 2000 by The American Association of Immunologists
References
In this issue
