The chronic inflammatory skin disease atopic dermatitis, now referred to as atopic eczema/dermatitis syndrome (AEDS) (Johansson et al, 2001), is increasing in prevalence, and AEDS symptoms have been recorded for as many as 20% of children throughout the world (ISAAC, 1998;Leung, 2000). A complex interplay between genetic predisposition, immune dysregulation, skin barrier dysfunction, lifestyle, and environmental factors is thought to be important for the development of AEDS (Cooper, 1994;Werfel and Kapp, 1998;Alm et al, 1999). Evidence is accumulating that the opportunistic yeast Malassezia, formerly known as Pityrosporum, can contribute to the inflammatory reaction in AEDS (Faergemann, 1999;Scheynius et al, 2002). The presence of IgE specific to Malassezia has been reported in 32%-68% of AEDS patients, and positive atopy patch test (APT) reactions in more than half of the patients (Tengvall Linder et al, 2000;Scheynius et al, 2002). A correlation between positive APT and Th2-like response in peripheral blood mononuclear cells has been described (Johansson et al, 2002).
Dendritic cells (DCs) are recognized as the most potent of the antigen-presenting cells due to their ability to activate naive T cells, and they act as conductors of immune responses by functioning as a link between the innate and acquired immune systems (Banchereau et al, 2000;Granucci et al, 2001). Immature DCs can efficiently take up antigens in the periphery and, following activation, differentiate into professional antigen-presenting cells expressing CD83 and high levels of costimulatory molecules. These mature DCs have strongly reduced ability to take up antigen, but are instead excellent at presenting antigen (Banchereau et al, 2000). Langerhans cells are a subset of immature CD1a+ DCs found in the epidermis of the skin. DCs with a Langerhans-cell-like phenotype can be generated in vitro by culturing the CD14+ cell fraction of peripheral blood mononuclear cells (Romani et al, 1994;Sallusto and Lanzavecchia, 1994) to gain monocyte-derived dendritic cells (MDDCs). As DCs initiate adaptive immune responses, it is likely that they play an important role in skewing the reaction to allergens towards a Th2-like response in atopic individuals. We have shown previously that interaction of Malassezia and human immature MDDCs leads to uptake, maturation, and cytokine production, possibly promoting a Th2-like immune response (Buentke et al, 2000;2001).
The regulation of DCs by other cells is far from fully understood. Natural killer (NK) cells are part of the innate immune system; their ability to induce cell death in tumor cells and virus-infected cells has been studied intensely. More recently a potential role in affecting DC regulation has been addressed. Several studies have demonstrated an interaction between DCs and NK cells, indicating an importance of cross-talk for their mutual regulation (Chambers et al, 1996;Geldhof et al, 1998;Carbone et al, 1999;Fernandez et al, 1999;Wilson et al, 1999;Spaggiari et al, 2001;Ferlazzo et al, 2002;Gerosa et al, 2002;Piccioli et al, 2002). It is not known if this interaction takes place also in vivo, or if a potential interaction of NK cells and DCs would be affected by allergen exposure of the DCs. In this study, we addressed these questions by investigating the presence of NK cells in the skin from AEDS patients and healthy individuals. Specifically, we studied the effect of Malassezia uptake by DCs on subsequent interactions with NK cells.
Materials and methods
Immunohistochemical and immunofluorescence staining of skin biopsy specimens
Skin specimens (4 mm punch biopsies) were taken under local anesthesia from three healthy individuals, and from nonlesional, lesional, and Malassezia extract APT-positive skin (at 24 h and 72 h after provocation) from four AEDS patients who had serum IgE antibodies specific for Malassezia (Tengvall Linder et al, 2000). The biopsies were snap frozen on dry ice and stored at -80°C until further analysis. Acetone-fixed, 6 
m thick, cryostat sections were immuno histochemically stained with an anti-CD56 (MY31, Becton Dickinson, Sweden) monoclonal antibody (MoAb), using Vectastain ABC-Elite (Vector Laboratories, Burlingame, CA), according to the manufacturer's protocol, and then developed with 3-amino-9-ethylcarbazole (0.02% wt/vol, Aldrich-Chemie, Germany), followed by counterstaining with Mayer's hematoxylin (Sigma, Sweden). CD56+ cells were identified by their definite brownish-red staining and visible nuclei. For double immunofluorescence staining, the sections were incubated, first, with normal donkey or goat serum, followed by anti-CD56 MoAb overnight at 4°C. Thereafter either donkey antimouse IgG Rhodamine Red-X (Jackson Immunoresearch, Göteborgs Termometerfabrik, Sweden) or goat antimouse IgG Alexa 546 (Molecular Probes, The Netherlands) conjugated antibody was applied for 1.5 h at room temperature. Next, normal mouse serum was added to the sections, followed by either a fluorescein isothiocyanate (FITC) conjugated anti-CD3 MoAb (SK4, Becton Dickinson) or an FITC-conjugated anti-CD1a MoAb (NA1/34, Dakopatts, Denmark) incubated overnight at 4°C. The sections were evaluated using a Leica TCS SP2 confocal laser scanning microscope system, equipped with an inverted Leica DM IRBE microscope, an argon laser, and two HeNe lasers (Leica Microsystems, Germany). Oil was used as immersion medium, and paraphenylenediamine in glycerol as mounting medium to reduce fading (Johnson et al, 1982). Leica confocal software was used to acquire and visualize the data. Staining was not observed when irrelevant isotype-matched antibodies, mouse IgG1 (DAK-GO1, Dakopatts) or mouse IgG2a FITC (X39, Becton Dickinson), were used or when primary antibodies were omitted.
Culturing of Malassezia
M. sympodialis strain no. 42132, previously designated M. furfur (Mayser & Gross, 2000;Scheynius et al, 2002), was obtained from the American Type Culture Collection, and is referred to here as Malassezia. The yeast was cultured at 37°C for 4 d as previously described (Buentke et al, 2001), and then harvested in sterile water and counted under a light microscope before incubation with MDDCs. The endotoxin content was determined in yeast culture supernatants from three independent experiments, and was found to be less than 0.3 EU per ml (Limulus test, performed by Apoteket, Stockholm, Sweden).
Generation of MDDCs from peripheral blood
Buffy coats were obtained from healthy donors from the Karolinska Hospital blood bank. All the test samples were ImmunoCapTM, m70, negative to Malassezia (= 0.35 kU per l) with a median total serum IgE level of 16 kU per l (range 2.6–140 kU per l, reference ranges 1.6–122 kU per l), and Phadiatop negative, i.e., did not show IgE reactivity to 11 common airborne allergens (Pharmacia Diagnostics, Sweden). Monocytes were isolated and MDDCs generated as previously described (Romani et al, 1994;Buentke et al, 2001). Briefly, the CD14+ cells were diluted to 4 
105 cells per ml in complete culture medium, RPMIc [RPMI 1640 medium supplemented with 25 g per ml gentamicin, 2 mM L-glutamine, 100 IU per ml penicillin, 100 g per ml streptomycin (Gibco BRL, Life Technologies, U.K.), and 10% (vol/vol) fetal bovine serum (FBS; Hyclone, Logan, UT), denoted "RPMI 1640 with 10% FBS", supplemented with 50 M 2-
-mercaptoethanol (Sigma), 550 U per ml granulocyte-macrophage colony-stimulating factor (GM-CSF; Schering Plough, Kenilworth, NJ), and 800 U per ml of recombinant human interleukin-4 (rIL-4; Nordic BioSite, Sweden)], and cultured in 25 or 75 cm2 culture flasks (Costar, Cambridge, MA). To generate immature MDDCs the CD14+ cells were cultured in a humidified incubator with 6% CO2 in air for 6 d with refeeding on day 3. To mature the MDDCs, RPMIc supplemented with cytokines (550 U per ml GM-CSF and 800 U per ml rIL-4) and 100 ng per ml of lipopolysaccharide (LPS) (L8274, Escherichia coli serotype 026-B6, Sigma) was added to the cells on day 6 of culture for an additional 46 h. The cells were harvested by gentle resuspension. This study was approved by the local ethics committee.
Interaction of MDDCs with Malassezia
After 6 d of culture, the MDDCs appeared as loosely adherent cell aggregates with typical dendritic morphology. At this time point, the immature MDDCs were harvested and cocultured with Malassezia, at a 1:5 ratio, in 25 cm2 culture flasks (Costar) at 37°C, in RPMIc supplemented with cytokines (550 U per ml GM-CSF and 800 U per ml rIL-4) or in complete culture medium alone, at a cell density of 4 105 MDDCs per ml, for approximately 46 h. Malassezia was also incubated in RPMIc culture medium alone for 46 h. Upon harvest, the cells were centrifuged, and cell viability was assessed by trypan blue exclusion. Some MDDCs, cocultured with or without Malassezia were air-dried onto three-well (
internal 14 mm) microscope glass-slides (
2 104 cells per well, Novakemi, Sweden), and stored at -80°C until use. In some experiments, the culture supernatants were sterile filtered (0.2 m) and used to stimulate immature DCs or NK cells.
Characterization of the MDDCs
At day 8 of culture, the MDDCs' surface phenotype was assessed by flow cytometry. The following FITC- or phycoerythrin- (PE) conjugated mouse MoAbs were used: anti-CD1a PE (T6-RD1, Coulter, Beckman-Coulter, Sweden), anti-CD14 FITC (Leu-M3), anti-CD40 FITC (5C3), anti-CD54 PE (Leu-54), anti-CD80 FITC (L307.4), anti-CD83 FITC (HB15e), anti-CD86 FITC (2331 FUN-1), anti-HLA-DR FITC (L243), and anti-HLA-A, B, C FITC (G46-2.6) from PharMingen/Becton Dickinson. Isotype-matched antibodies, mouse IgG1 FITC- or PE-conjugated (X40), and mouse IgG2a FITC-conjugated (X39) from Becton Dickinson were used as negative controls. Approximately 5 104 cells were incubated for 30 min on ice with MoAbs. Gates were set according to a characteristic forward-scatter pattern (Romani et al, 1996). The flow cytometer was calibrated according to the manufacturer's instructions before each acquisition, and a minimum of 104 cells was acquired on a FACSCalibur flow cytometer and analyzed using CellQuest software (Becton Dickinson). A typical immature DC phenotype was present in the MDDCs cultured only in medium (Table I) (Romani et al, 1994;1996), and the LPS-matured MDDCs showed a characteristic mature DC phenotype (Table I) (Banchereau et al, 2000). Owing to auto-fluorescence of the yeast, flow cytometry could not be used to assess the phenotype of MDDCs cultured with Malassezia. Instead the expression of CD83, the DC maturation marker (Zhou and Tedder, 1996), was studied by immunocytochemistry. Frozen and acetone-fixed MDDCs were stained with anti-CD83 (HB15e, PharMingen) MoAb, using Vectastain ABC-Elite (Vector Laboratories, Burlingame, CA) according to the manufacturer's protocol and as previously described (Buentke et al, 2001). The cells were evaluated with a Leiz microscope. A CD83+ cell was defined as one with a definite brownish-red staining and a visible nucleus. A minimum of 500 MDDCs were counted, and data were compared using the nonparametric Wilcoxon matched pairs test. The percentage of MDDCs cultured with Malassezia-expressing CD83 was significantly higher (p <0.05) than for MDDCs cultured in medium, i.e., 77% (median, range 46%-87%, n = 7) and 7.2% (1.4%-16%, n = 7), respectively. The median purity of the MDDC cultures, as determined by cell size and cell granularity using flow cytometry, was 95% (median, range 78%-98%, n = 12). The median viability of MDDCs incubated with Malassezia for 46 h was 92% (range 84%-100%, n = 12), with LPS 93% (range 88%-97%, n = 5), and with medium only 96% (range 92%-98%, n = 12), as determined by trypan blue exclusion.
Generation of short-term activated polyclonal NK cells
Peripheral blood mononuclear cells, depleted of CD14+ cells, were frozen in RPMI 1640 (Gibco BRL, Life Technologies) supplemented with 50% heat-inactivated FBS and 10% dimethyl sulfoxide, and stored at -150°C. To generate autologous short-term activated polyclonal NK cells, peripheral blood mononuclear cells were thawed and diluted to 2 106 cells per ml in RPMI 1640 with 10% FBS supplemented with 1000 U per ml of rIL-2 (Pepro Tech EC, U.K.), and then cultured in 25 cm2 culture flasks at 37°C for 48 h. K562 cells, a human erythroleukemia line (a kind gift from Professor Giorgio Trinchieri, Schering-Plough Research Institute, Laboratory for Immunological Research, Dardilly, France), were cultured in RPMI 1640 with 10% FBS.
Time-lapse study
MDDCs were seeded on glass coverslips in a 24-well plate (Costar) at a cell density of 2 105 per ml in RPMIc supplemented with GM-CSF and rIL-4 (550 U per ml GM-CSF and 800 U per ml rIL-4), and incubated at 37°C overnight to allow adhesion. The coverslips were transferred to an inert aluminum chamber containing RPMIc medium, and NK cells and Malassezia were added at a 1:5 ratio before the chamber was sealed with silicon immediately before the onset of the time-lapse photography as previously described (Buentke et al, 2001).
NK-cell-mediated cytotoxicity
The susceptibility of MDDCs to autologous short-term activated polyclonal NK-cell-induced cell death was measured in a standard 4 h 51Cr-release assay, at 37°C, using Na251CrO4-labeled target cells in triplicate at various effector:target ratios. 4 105 MDDCs and K562 cells were labeled with 40 l Na251CrO4 (Amersham, Sweden) for 1 h at 37°C, washed, and resuspended in complete culture medium. The amount of specific 51Cr released was expressed as a percentage and calculated as: % release =[(experimental release – spontaneous release)/(maximum release – spontaneous release)] 100. The spontaneous release was usually less than 50% of the maximum. K562 cells, which do not express major histocompatibility complex (MHC) class I molecules, were used as control targets to assess the cytotoxic capacity of NK cells.
Effect of soluble factors from MDDC-Malassezia cocultures on the interaction of NK cells and MDDCs
In some experiments NK cells and MDDCs were incubated in a 1:2 dilution of culture supernatant from MDDCs, cultured with or without Malassezia, during the 4 h 51Cr-release assay. To assess the effect of soluble factors on the phenotype of the MDDCs, 4 105 immature MDDCs per ml were resuspended in RPMIc and incubated in 5 ml tubes (Falcon, Becton Dickinson) for 4 h in a 1:2 dilution with sterile filtered supernatants from MDDCs cultured either in medium or with Malassezia, and with supernatants from Malassezia alone. The cells were harvested, labeled with MoAbs for 30 min at 4°C, and analyzed by flow cytometry. Data were compared using the nonparametric Wilcoxon matched pairs test. NK cells were also preincubated for 4 h at 37°C, in a 1:2 dilution of sterile filtered culture supernatant from autologous MDDC cultures, with or without Malassezia, and with supernatant from Malassezia cultured alone. The NK cells were then incubated with Na251CrO4-labeled K562 cells, and their cytotoxic capacity was measured in the standard 4 h 51Cr-release assay.
Results
NK cells are in close proximity to CD1a+ DCs in Malassezia APT-positive skin from AEDS patients
Immunohistochemical staining revealed few scattered CD56+ cells in the dermis, close to the epidermis, in skin biopsy specimens from healthy individuals (Figure 1a) and in nonlesional skin from AEDS patients (Figure 1b). In lesional and Malassezia APT-positive skin from the AEDS patients, CD56+ cells were also found in the epidermis (Figure 1c) and numerous CD56+ cells were observed in the dermal cell infiltrates (Figure 1d,e). To exclude the possibility that these CD56+ cells could be NK T cells, double immunofluorescence staining with anti-CD3 and anti-CD56 antibodies was performed. In healthy skin, nonlesional, lesional, and Malassezia APT-positive skin at 24 h and 72 h, all the CD56+ cells detected were CD3– and therefore most likely NK cells (Figure 1f). We next asked the question whether interaction between NK cells and DCs might take place in the skin during the inflammatory disorder AEDS associated with Malassezia. To address this question we performed double immuno fluorescence staining using antibodies to CD56 and CD1a. In the dermis, close to the epidermis, we were able to detect CD56+ NK cells in close contact with CD1a+ DCs in Malassezia APT-positive skin from AEDS patients (Figure 2).
Figure 1.
Different localization of NK cells in nonlesional compared to lesional and Malassezia APT-positive skin from AEDS patients. Using immunohistochemistry, CD56+ cells were found in the dermis close to the epidermis in healthy individuals (A), and in nonlesional skin from AEDS patients (B). In lesional (C, D) and Malassezia APT-positive skin at 72 h (E) from AEDS patients, CD56+ cells were found in the epidermis (C) and dermal cell infiltrates (D, E). Epidermis is upwards in the pictures. Double immunofluorescence staining, and confocal laser scanning microscopy, showed that the CD56+ cells were CD3– (lesional skin, F). Scale bar: 10 m.
Figure 2.
CD56+ NK cells and CD1a+ DCs in close contact in the skin. Double immunofluorescence staining was performed with MoAbs against CD56 and CD1a in skin biopsy specimens from AEDS patients with Malassezia APT-positive reactions (72 h). The dendrites from the CD1a+ DC can be seen in close proximity to the CD56+ NK cells. The dotted line indicates the border between epidermis and dermis, with dermis to the lower right in the picture. The specimen was analyzed using confocal laser scanning microscopy (Leica Microsystems). Scale bar: 10 m.
NK-cell-mediated lysis of autologous MDDCs is lower after preincubation with Malassezia
By using time-lapse photography, the interactions between NK cells and MDDCs in the presence of Malassezia were followed for up to 6 h (Figure 3). One consequence of the interaction between NK cells and DCs is the NK-cell-mediated induction of cell death in DCs (Wilson et al, 1999). We observed that the interaction of several NK cells with an MDDC resulted in death of the MDDC after approximately 2–3 h, as apparent by the cell's rounding, swelling, loss of membrane integrity, and release of small vesicles (Figure 3f–i). The MDDC's dendrite comes into contact with NK cell no. 1 (Figure 3b) in a manner like that shown in the skin, pictured in Figure 2. The effect of Malassezia on the interaction between NK cells and DCs was further studied using an in vitro assay for NK-cell-induced cytotoxicity. MDDCs, preincubated with the yeast Malassezia for 46 h at a 1:5 ratio, were relatively less susceptible to NK-cell-induced cell death compared to the control MDDCs incubated in culture medium for the same time period (Figure 4). LPS-matured MDDCs were less susceptible to NK-cell-induced cell death than MDDCs incubated in medium and MDDCs cocultured with Malassezia (Figure 4). A shorter coculture period (20 h) for MDDCs and Malassezia did not result in reduced susceptibility to NK-cell-mediated lysis (data not shown).
Figure 3.
Time-lapse photography of short-term activated NK cells inducing cell death in an immature MDDC. Time-lapse photography was used to study the NK-cell-mediated lysis of an immature MDDC. One picture was taken every third second. The nine different time points, after coincubation of 2 105 immature MDDCs per ml with NK cells and Malassezia at a 1:5 ratio, are indicated in the upper right corner of each figure. The numbers 1 and 2 are used to indicate the movement of two different NK cells. Contact between NK cell no. 1 and the MDDC is seen in (B) and between NK cell no. 2 and the MDDC in (D), (E). Scale bar: 10 m. The photographs are from one experiment representative of two.
Figure 4.
Immature MDDCs are less susceptible to autologous NK-cell-induced cell death if preincubated with the yeast Malassezia. The percentage lysis of K562 (circles), of MDDCs (4 103 per well) incubated in medium alone (squares), of MDDCs incubated for 46 h with the yeast Malassezia at a 1:5 ratio (filled squares), and of MDDCs cultured with LPS (100 ng per ml) for 46 h (triangles), by autologous short-term activated polyclonal NK cells, was detected in a 4 h standard 51Cr-release assay. The results shown are from one of three representative experiments, using cells from different blood donors.
Soluble factors from cocultures of MDDCs and Malassezia affect the cytotoxic capacity of NK cells
The reduced susceptibility of Malassezia-pretreated MDDCs to NK-cell-mediated cytotoxicity might be a consequence of induction of a more mature phenotype in the MDDCs, or an effect of soluble mediators produced by the MDDCs when activated with Malassezia. To investigate whether soluble factors produced by MDDCs preincubated with Malassezia affected NK-cell-induced death in immature MDDCs, sterile filtered culture supernatant was added during the 4 h 51Cr-release assay. This rendered the MDDCs less susceptible to NK-cell-induced cell death (Figure 5).
Figure 5.
Soluble factors in supernatants from cocultures of MDDCs and Malassezia yeast cells lead to decreased induction of autologous NK-cell-mediated cell death in immature MDDCs. A 4 h standard 51Cr-release assay was used to detect the percentage lysis by NK cells of K562 (circles), of MDDCs (4 103 per well) cultured in medium alone with no addition of culture supernatants (squares), of MDDCs with addition of culture supernatant from autologous MDDCs cultured in medium for 46 h (triangles), and of MDDCs with addition of culture supernatant from autologous MDDCs cocultured with the yeast Malassezia at a 1:5 ratio for 46 h (filled triangles). The results shown are from one of two representative experiments, using cells from different blood donors.
Next we investigated whether the soluble factors produced by MDDCs preincubated with Malassezia affected the phenotype of the immature MDDCs. Using flow cytometry, we found a significant increase of CD86+ MDDCs after addition of the MDDC-Malassezia supernatant as well as after addition of MDDC-LPS supernatant (data not shown). Heating of the supernatants to 56°C for 30 min, or addition of culture supernatants from yeast alone, did not affect the CD86 expression (data not shown). No differences were detected in the levels of CD1a, CD80, CD83, HLA-DR, or MHC class I expression after addition of MDDC-Malassezia supernatant (data not shown).
To further investigate the effect of soluble factors on the cytotoxic capacity of NK cells, NK cells were preincubated with culture supernatants. NK cells preincubated with MDDC-Malassezia coculture supernatant killed K562 MHC class I negative target cells to a lower degree than did NK cells incubated in medium alone (Figure 6). Preincubation in supernatants from Malassezia alone also resulted in less killing of the K562 cells, indicating that factors derived directly from the yeast might affect the activity of NK cells (Figure 6). Thus in addition to cell-cell contact, soluble factors might also play a role in the interaction between NK cells and DCs.
Figure 6.
The activity of autologous NK cells is inhibited by soluble factors in supernatants from cocultures of MDDCs and M. sympodialis. Short-term activated polyclonal NK cells were preincubated with culture supernatant from MDDCs cultured with or without Malassezia or from Malassezia alone, for 4 h, before the cytotoxic capacity of the NK cells were assessed on the MHC class I negative target cell line K562. Shown is the percentage lysis of K562 by NK cells preincubated in medium (circles), of K562 by NK cells preincubated in MDDC-Malassezia coculture supernatant (1:5 ratio for 46 h, filled triangles), and of K562 by NK cells preincubated in Malassezia culture supernatant (46 h, triangles). These results are from one of four representative experiments, using cells from different blood donors in a 4 h standard 51Cr-release assay. Using supernatants from MDDCs incubated with medium resulted in a background induction of NK-cell-mediated cell death of median 93% of the K562 cells at the 100:1 ratio (range 91%-95%, n = 2).
Full figure and legend (12K)Discussion
In this study we show that CD56+/CD3– NK cells are located close to CD1a+ DCs in Malassezia APT-positive skin from AEDS patients and that uptake of the allergenic yeast Malassezia by immature MDDCs affects the subsequent interaction of MDDCs and NK cells.
Despite increasing interest in NK cell–DC interactions, little evidence indicates where this interaction takes place in vivo. It has been shown, however, that cell-cell contact is important in the cross-talk (Fernandez et al, 1999). Therefore we questioned whether NK cells were present in the skin of patients with the chronic inflammatory skin disease AEDS and, if so, might be distributed in a way that they could interact with DCs. As described here, we have demonstrated that NK cells and CD1a+ DCs can appear in close contact in Malassezia APT-positive skin, suggesting a role for NK cell-DC cross-talk. To our knowledge this is the first time that human NK cells and DCs have been shown to be in actual contact in vivo. NK cells in human skin have been associated with skin tumors, and occasional NK cells have been described in lesional skin from patients with AEDS (Zachary et al, 1985;Nakamura et al, 1995). The possible role of NK cells in AEDS has been studied with somewhat contradictory results. The main focus has been on the number of NK cells in peripheral blood and their level of activation. In AEDS patients, both a numerical decrease of NK cells and lowered NK cell activity compared to individuals without AEDS has been reported (Wehrmann et al, 1990;Hashiro and Okumura, 1998). Other groups found no difference in the number of NK cells (Bouloc et al, 2000), or higher NK cell activity than in healthy individuals (Strannegård and Strannegård, 1980). Such conflicting results may reflect the possibility that AEDS is a heterogeneous disease (Johansson et al, 2001), or that treatments, like topical steroid treatment, may have a systemic effect on NK cells, thus influencing experimental outcomes (Lesko et al, 1989). Additionally, NK cells leaving the circulation could account for the lower number reported in most studies. Support for this seems to come from the numerous CD56+ cells we observed in dermal infiltrates in lesional and Malassezia APT-positive skin from AEDS patients (Figure 1d,e) and the report of increased levels of the NK cell chemoattractant MCP-1/CCL2 in patients with AEDS (Kaburagi et al, 2001). CD1a+ cells in lesional skin of AEDS patients can produced MDC/CCL22, another chemokine that attracts NK cells, as well as DCs and Th2 cells (Vulcano et al, 2001).
We showed that NK cells were differentially distributed in lesional and Malassezia APT-positive skin, with NK cells in the epidermis and dermal infiltrates, compared to nonlesional skin or skin from healthy individuals. Besides, an inverse relation between the NK cell activity and severity of eczema has been demonstrated (Chiarelli et al, 1987). This, together with our findings, points to a role for NK cells in AEDS. Moreover, NK cells have been implicated in the development of allergen-induced airway inflammation in a mouse model, further supporting the role of NK cells in allergic diseases (Korsgren et al, 1999).
We then pursued our interest in studying the interaction of human NK cells and DCs preincubated with an allergen. That the opportunistic yeast Malassezia can act as an allergen and thereby contribute to the inflammatory reaction in AEDS has been shown in many studies (Faergemann, 1999;Scheynius et al, 2002), and we have previously demonstrated that Malassesia can be taken up by immature MDDCs, leading to their maturation and production of cytokines with a potential to skew the immune reaction towards a Th2-like response (Buentke et al, 2000;2001). Using time-lapse photography, we visualized the interaction between MDDCs and several NK cells, in the presence of Malassezia. Similar events, like swelling and loss of membrane integrity in the dying cell, were recorded elsewhere by the same technique, although target cells other than MDDCs were used (Eriksson et al, 1999). Spaggiari et al showed the involvement of two activating receptors in the NK-cell-mediated induction of cell death in DCs, and several recent in vitro studies have demonstrated functional effects upon interaction between NK cells and DCs, implying an importance for their mutual regulation (Spaggiari et al, 2001;Ferlazzo et al, 2002;Gerosa et al, 2002;Piccioli et al, 2002).
Using a 51Cr-release assay, we found that immature MDDCs cocultured with Malassezia for only 20 h were as susceptible to NK-cell-induced cell death as MDDCs incubated in medium. When immature MDDCs were cocultured with the yeast for 46 h, however, they became less susceptible to NK-cell-induced cell death, as were MDDCs matured with LPS. In one donor, an autologous NK cell clone was also produced and the results were the same as when using the short-term activated polyclonal NK cells (data not shown). Additionally, Malassezia induced maturation of MDDCs resulting in an increased number of CD80-, CD83-, and CD86-positive cells after coculture for approximately 46 h (Buentke et al, 2001). Evidently, MDDCs pretreated with Malassezia require time to fully mature for the reduced susceptibility to NK-cell-induced cell death to occur. DCs in different maturation stages have been reported to be differently susceptibility to NK cell lysis. Human immature DCs were found to be more susceptible to autologous NK-cell-induced cell death than DCs matured by LPS, tumor necrosis factor 
, monocyte-conditioned medium, or CD40L engagement (Carbone et al, 1999;Wilson et al, 1999). Viral infection also rendered the DCs less susceptible to NK-cell-induced cell death (Wilson et al, 1999), a conclusion that is in agreement with our observations. A balance of activating and inhibiting signals determines whether NK cells induce cell death in their target cells or not (Moretta et al, 2001). This indicates that maturation of DCs by microorganisms, like yeast or virus, or products from microorganisms, like LPS, influence the surface phenotype of DCs or the factors they secrete in such a way that the DCs become less susceptible to NK-cell-mediated lysis.
In addition to cell-cell contact, our data show that soluble factors may play a role in the NK cell-DC interaction. Supernatant from cocultures of MDDCs and Malassezia rendered untreated immature MDDCs less susceptible to NK-cell-induced cell death, possibly through soluble factors with an inhibitory effect on the NK cell. Cytokines produced by DCs, such as IL-10 and IL-12, have been shown to modulate NK cell activity (Goodier and Londei, 2000;Grufman and Kärre, 2000). We reported previously, however, that DC interaction with Malassezia induced the production of IL-1, tumor necrosis factor , and IL-18, but little or no IL-10 or IL-12p70 (Buentke et al, 2001). Soluble factors in the coculture supernatant did result in an increased number of CD86+ MDDCs, which might indicate that the soluble factors only indirectly affected the NK cell–DC interaction. Upregulation of MHC class I has been suggested to be of importance in the interaction between mature DCs and NK cells (Ferlazzo et al, 2001), but we could not detect any increase in MHC class I mean fluorescence intensity values for MDDCs stimulated with coculture supernatant. Our results also indicate that factors from Malassezia itself might have an inhibitory effect on the NK cell cytotoxicity. Possibly the yeast cells produce some substance similar to the reported viral IL-10 (Minter et al, 2001). In vitro studies have shown that the ratio between NK cells and DCs is important for the outcome of the cell interaction. At high ratios NK-cell-induced cell death is the preferential outcome, whereas mutual activation occurs at lower ratios (Piccioli et al, 2002; own unpublished results). We have shown NK cell and DC contact in Malassezia APT-positive skin, and that Malassezia can influence the interaction of NK cells and DCs in vitro, leading to lower NK-cell-induced cell death. What the preferential effect of Malassezia on the cross-talk in vivo is needs to be further investigated. Our findings that Malassesia directly or indirectly affects the interaction of NK cells and CD1a+ DCs, in concert with previous reports on the involvement of Malassezia in AEDS (Faergemann, 1999;Scheynius et al, 2002), suggest that the allergenic yeast may interfere and/or affect the cross-talk between NK cells and DCs in this disease.
In conclusion, our data indicate that CD56+ CD3– NK cells and CD1a+ DCs can interact in allergen-provoked APT-positive skin of AEDS patients. The differential distribution of NK cells in nonlesional compared to lesional and Malassezia APT-positive skin from AEDS patients suggests that NK cells may play a role in regulating DCs in AEDS. These data also imply that the allergenic yeast Malassezia affects the interaction of NK cells and DCs, and that soluble factors from the DCs or the yeast may be of importance in the NK cell-DC cross-talk in addition to cell-cell contact.
References
| 1. | Alm JS, Swartz J, Lilja G, Scheynius A & Pershagen G. Atopy in children of families with an anthroposophic lifestyle. Lancet (1999) 353: 1485–1488. | Article | PubMed | ISI | ChemPort | |
| 2. | Banchereau J, Briere F & Caux C et al. Immunobiology of dendritic cells. Annu Rev Immunol (2000) 18: 767–811. | Article | PubMed | ISI | ChemPort | |
| 3. | Bouloc A, Charue D, Nikolova M & Bensussan A. No modulation of circulating natural killer cell and natural killer receptor bearing memory T cell subsets in patients with atopic dermatitis. J Invest Dermatol (2000) 115: 1160–1162. | Article | PubMed | ISI | ChemPort | |
| 4. | Buentke E, Zargari A, Heffler LC, Avila-Carino J, Savolainen J & Scheynius A. Uptake of the yeast Malassezia furfur and its allergenic components by human immature CD1a+ dendritic cells. Clin Exp Allergy (2000) 30: 1759–1770. | Article | PubMed | ISI | ChemPort | |
| 5. | Buentke E, Heffler LC, Wallin RP, Lofman C, Ljunggren HG & Scheynius A. The allergenic yeast Malassezia furfur induces maturation of human dendritic cells. Clin Exp Allergy (2001) 31: 1583–1593. | Article | PubMed | ISI | ChemPort | |
| 6. | Carbone E, Terrazzano G & Ruggiero G et al. Recognition of autologous dendritic cells by human NK cells. Eur J Immunol (1999) 29: 4022–4029. | Article | PubMed | ISI | ChemPort | |
| 7. | Chambers BJ, Salcedo M & Ljunggren HG. Triggering of natural killer cells by the costimulatory molecule CD80 (B7-1). Immunity (1996) 5: 311–317. | Article | PubMed | ISI | ChemPort | |
| 8. | Chiarelli F, Canfora G, Verrotti A, Amerio P & Morgese G. Humoral and cellular immunity in children with active and quiescent atopic dermatitis. Br J Dermatol (1987) 116: 651–660. | PubMed | ISI | ChemPort | |
| 9. | Cooper KD. Atopic dermatitis: recent trends in pathogenesis and therapy. J Invest Dermatol (1994) 102: 128–137. | Article | PubMed | ISI | ChemPort | |
| 10. | Eriksson M, Leitz G, Fallman E, Axner O, Ryan JC, Nakamura MC & Sentman CL. Inhibitory receptors alter natural killer cell interactions with target cells yet allow simultaneous killing of susceptible targets. J Exp Med (1999) 190: 1005–1012. | Article | PubMed | ISI | ChemPort | |
| 11. | Faergemann J. Pityrosporum species as a cause of allergy and infection. Allergy (1999) 54: 413–419. | Article | PubMed | ISI | ChemPort | |
| 12. | Ferlazzo G, Semino C & Melioli G. HLA class I molecule expression is up-regulated during maturation of dendritic cells, protecting them from natural killer cell-mediated lysis. Immunol Lett (2001) 76: 37–41. | Article | PubMed | ISI | ChemPort | |
| 13. | Ferlazzo G, Tsang ML, Moretta L, Melioli G, Steinman RM & Munz C. Human dendritic cells activate resting natural killer (NK) cells and are recognized via the NKp30 receptor by activated NK cells. J Exp Med (2002) 195: 343–351. | Article | PubMed | ISI | ChemPort | |
| 14. | Fernandez NC, Lozier A & Flament C et al. Dendritic cells directly trigger NK cell functions: cross-talk relevant in innate anti-tumor immune responses in vivo. Nat Med (1999) 5: 405–411. | Article | PubMed | ISI | ChemPort | |
| 15. | Geldhof AB, Moser M, Lespagnard L, Thielemans K & De Baetselier P. Interleukin-12-activated natural killer cells recognize B7 costimulatory molecules on tumor cells and autologous dendritic cells. Blood (1998) 91: 196–206. | PubMed | ISI | ChemPort | |
| 16. | Gerosa F, Baldani-Guerra B, Nisii C, Marchesini V, Carra G & Trinchieri G. Reciprocal activating interaction between natural killer cells and dendritic cells. J Exp Med (2002) 195: 327–333. | Article | PubMed | ISI | ChemPort | |
| 17. | Goodier MR & Londei M. Lipopolysaccharide stimulates the proliferation of human CD56+CD3– NK cells: a regulatory role of monocytes and IL-10. J Immunol (2000) 165: 139–147. | PubMed | ISI | ChemPort | |
| 18. | Granucci F, Vizzardelli C & Pavelka N et al. Inducible IL-2 production by dendritic cells revealed by global gene expression analysis. Nat Immunol (2001) 2: 882–888. | Article | PubMed | ISI | ChemPort | |
| 19. | Grufman P & Kärre K. Innate and adaptive immunity to tumors: IL-12 is required for optimal responses. Eur J Immunol (2000) 30: 1088–1093. | Article | PubMed | ISI | ChemPort | |
| 20. | Hashiro M & Okumura M. The relationship between the psychological and immunological state in patients with atopic dermatitis. J Dermatol Sci (1998) 16: 231–235. | Article | PubMed | ISI | ChemPort | |
| 21. | ISAAC:. Worldwide variation in prevalence of symptoms of asthma, allergic rhinoconjunctivitis, and atopic eczema: ISAAC. The International Study of Asthma and Allergies in Childhood (ISAAC) Steering Committee. Lancet (1998) 351: 1225–1232. | Article | PubMed | ISI | |
| 22. | Johansson C, Eshaghi H, Tengvall Linder M, Jacobson E & Scheynius A. Positive atopy patch test reaction to Malassezia furfur in atopic dermatitis correlates with a Th2-like PBMC response. J Invest Dermatol (2002) 118: 1044–1051. | Article | PubMed | ISI | ChemPort | |
| 23. | Johansson SGO, Hourihane JO & Bousquet J et al. A revised nomenclature for allergy. An EAACI position statement from the EAACI nomenclature task force. Allergy (2001) 56: 813–824. | Article | PubMed | ISI | ChemPort | |
| 24. | Johnson GD, Davidson RS, McNamee KC, Russell G, Goodwin D & Holborow EJ. Fading of immunofluorescence during microscopy: a study of the phenomenon and its remedy. J Immunol Meth (1982) 55: 231–242. | Article | ISI | ChemPort | |
| 25. | Kaburagi Y, Shimada Y, Nagaoka T, Hasegawa M, Takehara K & Sato S. Enhanced production of CC-chemokines (RANTES, MCP-1, MIP-1alpha, MIP- 1beta, and eotaxin) in patients with atopic dermatitis. Arch Dermatol Res (2001) 293: 350–355. | Article | PubMed | ISI | ChemPort | |
| 26. | Korsgren M, Persson CG & Sundler F et al. Natural killer cells determine development of allergen-induced eosinophilic airway inflammation in mice. J Exp Med (1999) 189: 553–562. | Article | PubMed | ISI | ChemPort | |
| 27. | Lesko MJ, Lever RS, Mackie RM & Parrott DM. The effect of topical steroid application on natural killer cell activity. Clin Exp Allergy (1989) 19: 633–636. | PubMed | ISI | ChemPort | |
| 28. | Leung DY. Atopic dermatitis: new insights and opportunities for therapeutic intervention. J Allergy Clin Immunol (2000) 105: 860–876. | Article | PubMed | ISI | ChemPort | |
| 29. | Mayser P & Gross A. IgE antibodies to Malassezia furfur, M. sympodialis and Pityrosporum orbiculare in patients with atopic dermatitis, seborrheic eczema or pityriasis versicolor, and identification of respective allergens. Acta Derm Venereol (2000) 80: 357–361. | PubMed | ISI | ChemPort | |
| 30. | Minter RM, Ferry MA & Rectenwald JE et al. Extended lung expression and increased tissue localization of viral IL-10 with adenoviral gene therapy. Proc Natl Acad Sci USA (2001) 98: 277–282. | Article | PubMed | ChemPort | |
| 31. | Moretta A, Bottino C & Vitale M et al. Activating receptors and coreceptors involved in human natural killer cell-mediated cytolysis. Annu Rev Immunol (2001) 19: 197–223. | Article | PubMed | ISI | ChemPort | |
| 32. | Nakamura S, Suchi T & Koshikawa T et al. Clinicopathologic study of CD56 (NCAM) -positive angiocentric lymphoma occurring in sites other than the upper and lower respiratory tract. Am J Surg Pathol (1995) 19: 284–296. | PubMed | ISI | ChemPort | |
| 33. | Piccioli D, Sbrana S, Melandri E & Valiante NM. Contact-dependent stimulation and inhibition of dendritic cells by natural killer cells. J Exp Med (2002) 195: 335–341. | Article | PubMed | ISI | ChemPort | |
| 34. | Romani N, Gruner S & Brang D et al. Proliferating dendritic cell progenitors in human blood. J Exp Med (1994) 180: 83–93. | Article | PubMed | ISI | ChemPort | |
| 35. | Romani N, Reider D & Heuer M et al. Generation of mature dendritic cells from human blood. An improved method with special regard to clinical applicability. J Immunol Meth (1996) 196: 137–151. | Article | ISI | ChemPort | |
| 36. | Sallusto F & Lanzavecchia A. Efficient presentation of soluble antigen by cultured human dendritic cells is maintained by granulocyte/macrophage colony-stimulating factor plus interleukin 4 and downregulated by tumor necrosis factor alpha. J Exp Med (1994) 179: 1109–1118. | Article | PubMed | ISI | ChemPort | |
| 37. | Scheynius A, Johansson C, Buentke E, Zargari A & Tengvall Linder M. Atopic eczema/dermatitis syndrome and Malassezia. Int Arch Dermatol (2002) 127: 161–169. | ChemPort | |
| 38. | Spaggiari GM, Carosio R & Pende D et al. NK cell-mediated lysis of autologous antigen-presenting cells is triggered by the engagement of the phosphatidylinositol 3-kinase upon ligation of the natural cytotoxicity receptors NKp30 and NKp46. Eur J Immunol (2001) 31: 1656–1665. | Article | PubMed | ISI | ChemPort | |
| 39. | Strannegård IL & Strannegård Ö. Natural killer cells and interferon production in atopic dermatitis. Acta Dermato-Venereologica Suppl (1980) 92: 48. |
| 40. | Tengvall Linder M, Johansson C, Scheynius A & Wahlgren C. Positive atopy patch test reactions to Pityrosporum orbiculare in atopic dermatitis patients. Clin Exp Allergy (2000) 30: 122–131. | Article | PubMed | ChemPort | |
| 41. | Vulcano M, Albanesi C & Stoppacciaro A et al. Dendritic cells as a major source of macrophage-derived chemokine/CCL22 in vitro and in vivo. Eur J Immunol (2001) 31: 812–822. | Article | PubMed | ISI | ChemPort | |
| 42. | Wehrmann W, Reinhold U, Kukel S, Franke N, Uerlich M & Kreysel HW. Selective alterations in natural killer cell subsets in patients with atopic dermatitis. Int Arch Allergy Appl Immunol (1990) 92: 318–322. | PubMed | ISI | ChemPort | |
| 43. | Werfel T & Kapp A. Environmental and other major provocation factors in atopic dermatitis. Allergy (1998) 53: 731–739. | PubMed | ISI | ChemPort | |
| 44. | Wilson JL, Heffler LC, Charo J, Scheynius A, Bejarano MT & Ljunggren HG. Targeting of human dendritic cells by autologous NK cells. J Immunol (1999) 163: 6365–6370. | PubMed | ISI | ChemPort | |
| 45. | Zachary CB, Allen MH & MacDonald DM. In situ quantification of T-lymphocyte subsets and Langerhans cells in the inflammatory infiltrate of atopic eczema. Br J Dermatol (1985) 112: 149–156. | PubMed | ISI | ChemPort | |
| 46. | Zhou LJ & Tedder TF. CD14+ blood monocytes can differentiate into functionally mature CD83+ dendritic cells. Proc Natl Acad Sci USA (1996) 93: 2588–2592. | Article | PubMed | ChemPort | |
Acknowledgments
We wish to thank Gunilla Jacobsson Ekman and Hojjatollah Eshaghi for valuable technical support. This work was supported by grants from the Swedish Medical Research Council (project no. 16x-7924), the Swedish Council for Work Life Sciences, the Swedish Asthma and Allergy Association's Research Foundation, the Swedish Foundation for Health Care Sciences and Allergy Research, the Queen Silvia Jubilee Foundation, the Hesselman Foundation, and Karolinska Institutet.
