Skin sections of clinically involved and clinically normal-looking skin from patients with atopic dermatitis were incubated with anti-human IgE antibodies using the indirect immunoperoxidase technique. Apart from positive dermal anti-IgE staining, positive epidermal anti-IgE staining was also observed. The morphology of the epidermal staining cells suggested the involvement of dendritic cells. This was confirmed by positive immuno-double labelling with OKT6 and anti-IgE. This phenomenon seemed to be specific for atopic dermatitis since skin sections from normal non-atopic controls, patients with allergic asthma, contact dermatitis, and schistosomiasis showed no epidermal anti-IgE staining. To further elucidate the nature of the epidermal anti-IgE staining cells, epidermal cell suspensions were prepared from clinically involved skin from patients with atopic dermatitis. These cell suspensions also showed positive anti-IgE staining cells and positive immuno-double labelling with OKT6 and anti-IgE. Immunogold electron microscopy with anti-IgE on epidermal cell suspensions from patients with atopic dermatitis showed gold particles on the cell membranes of cells containing Birbeck granules, being Langerhans' cells. Epidermal cell suspensions from normal non-atopic controls were negative. The presence of IgE molecules on epidermal Langerhans' cells, which seems to be specific for patients with atopic dermatitis, provides an explanation for the high frequency of positive patch test reactions to inhalant allergens.

Intracutaneous testing and patch tests with house dust mite and grass pollen allergens were performed in patients with atopic dermatitis. Only patients with an immediate type skin reaction to house dust mite or grass pollen allergens showed a positive patch test reaction to these allergens 24-48 h after testing. Occasionally positive patch test reactions at 20 min, 2 h and 6 h were also observed. Patch test reactions were not found in normal controls or atopic patients without atopic dermatitis. Analysis of the cellular infiltrate demonstrated an influx of eosinophils into the dermis, starting from 2-6 h after patch testing. Immunostaining with antibodies against granular constituents of the eosinophils revealed that the infiltrating eosinophils were in an activated state and had lost part of their granular contents. At 24 h eosinophils also appeared in the epidermis. Electron microscopy showed that in the epidermis, some eosinophils were in close contact with Langerhans cells, suggesting a cell-cell interaction. Taken together, these results strongly suggest an active role for eosinophils in patch test reactions to inhalant allergens in atopic dermatitis patients.

Human granulocytes isolated from peripheral blood have been described to synthesize both LTB4 and LTC4 from arachidonic acid. We have observed that the amount of LTC4 produced by human granulocyte preparations is strongly dependent on the relative amount of eosmophils. To investigate a possibly significant difference in leukotriene synthesis of the eosinophilic and neutrophilic granulocytes, we developed a purification method to isolate both cell types from granulocytes obtained from the blood of healthy donors. Leukotrlenes were generated by incubation of the purified cells with arachidonic acid, calcium lonophore A23187, calcium-chlonde and reduced glutatbione. Surprisingly, eosinophils were found to produce almost exclusively the spasmogenic LTC4. In contrast, neutropbils produce almost exclusively the chemotactic LT&, its w-hydroxylated metabolite 20-hydroxy-LTB4 and two nonenzymically formed LTB4 isomers. Leukotnene Granulocyte Eosmophil

Healthy volunteers underwent bronchial challenge with increasing doses of nebulized leukotriene D4 (0.007 - 200 nmol) at 15 min intervals. Total amounts of 200 nmol (females) and 400 nmol (males) were inhaled, corresponding to approximately 100 nmol and 200 nmol deposited in the lung, respectively. Of the latter amounts 3 +/- 1% (mean +/- S.E.M., n = 5) was found to be excreted as leukotriene E4 into the urine within 12 h. No further excretion after this period was observed. Approximately 50% of the total urinary leukotriene E4 was excreted during the first 2 h. These results suggest that a possible formation of sulfidopeptide leukotrienes in the lung in vivo can be monitored by measuring leukotriene E4 excretion into the urine.

No abstract