n., submandibular lymph nodes, but not NALTs, were consistently find more clearly stained (Fig. 3, inset). These results taken together demonstrate that the submandibular lymph nodes are the main organ that responds to i.n. injected allergens. To explore the mechanisms of IgG Ab production and compare them with those of IgE Ab production in submandibular lymph nodes, we injected the allergen with or without complete Freund’s adjuvant i.n. once into BALB/c mice (Fig. 4). A significant amount of serum IgE (465.4 ±111.6 ng/mL; mean ± SD; n =9) was induced by one i.n. injection of allergen alone. In contrast, one i.n. injection

of the allergen with adjuvant induced a much smaller amount of serum IgE (172.5 ± 74.7ng/mL; mean ± SD; n =9). This was greater than that (57.6 ± 32.2 ng/mL; mean ± SD; n =9) in mice treated with adjuvant alone or that (40.8 ± 14.8 ng/mL; mean ± SD; n =9) in PBS-injected mice. In contrast, a large amount of serum IgG (1585.4 ± 161.0 μg/mL; mean ± SD; n =9) was induced by one i.n. injection

of the allergen with adjuvant into mice; this amount of serum IgG was greater than R788 cost that obtained after one i.n. injection of adjuvant (1018.2 ±33.2 μg/mL; mean ± SD; n =9) or allergen (904.9 ± 51.2 μg/mL; mean ± SD; n =9) alone, both of which were greater than that (514.7 ± 161.8 μg/mL; mean ± SD; n =9) in PBS-injected mice. These results indicate that one i.n. injection of allergen alone or with adjuvant is suitable for induction of serum IgE or IgG Ab, respectively. To explore which population of cells in the submandibular lymph nodes is involved in the production of IgE Ab in response to the allergen, we separated the cells into macrophage-, lymphocyte-, and granulocyte-rich populations by Percoll density-gradient centrifugation. The yield of cells from the submandibular lymph nodes

was 78–89% (n =9). Fraction 3 (rich in lymphocytes) was the major (93.5 ± 7.2%; mean ± SD; n =9) population, 3-oxoacyl-(acyl-carrier-protein) reductase followed by fraction 2 (rich in macrophages; 1.2 ± 0.1%; mean ± SD; n =9) and fraction 4 (rich in granulocytes; 0.3 ± 0.1%; mean ± SD; n =9) in that order. Fraction 1 (rich in somewhat damaged cells) contained a small number of cells. As we obtained the macrophage-, lymphocyte-, and granulocyte-rich fractions, we incubated various combinations of these cells for 6 days and then assessed the amounts of IgE Ab in the culture media (Fig. 5). Bulk submandibular lymph node cells from mice that had been treated with allergen once i.n. produced a significant amount of IgE Abs (6.2 ± 3.4 ng/mL; mean ± SD; n =9); whereas the lymphocyte-rich (fraction 3) fraction of the lymph node cells did not (1.5 ± 0.8 ng/mL; mean ± SD; n =9). The macrophage-rich (fraction 2) fraction was also inactive (1.1 ± 0.9 ng/mL; mean ± SD; n =9). Of particular interest, IgE Ab production (4.6 ± 2.8 ng/mL; mean ± SD; n =9) was restored by addition of the macrophage-rich fraction to the lymphocyte-rich fraction.