Characterization of the central role of CD4⁺CD25⁺ T cells and IL-4 in mediating the IL-5 effect. (A) Pretreatment with NDS61, a mAb that depletes CD25⁺ cells,²³  eliminated any effect of rIL-5 treatment (●) because the clinical course was similar to controls treated with NDS61, but no IL-5 (○; n = 6/group). In contemporary controls not treated with NDS61 (n = 5/group), IL-5 inhibited EAN (data not shown). (B) The cauda equina nerves of animals treated with IL-5 (■) had reduced macrophage (**P = .002), CD4⁺ (**P = .002), and CD8⁺ (**P = .008) T-cell infiltrates than untreated (□) controls and IL-5/NDS61 (CD25⁺ depleted; ▩). Samples were taken from animals in panel A at day 31 (n = 5/group). (C) IL-5–treated animals (■) had a greater proportion of CD25⁺CD4⁺ T cells in nerves than untreated controls and NDS61/IL-5–treated animals (▩). Data are expressed as mean ± SD and number of cells counted as per high-power field (n = 5). (D) Blocking IL-4 prevented rIL-5 inhibition of EAN. Animals immunized to develop EAN were treated with MRCOX81, a monoclonal that blocks IL-4.²²  The disease course was similar in animals treated with anti–IL-4, with (▴) or without (▵) rIL-5, and untreated controls (○). Controls only treated with rIL-5 had a reduction in severity of EAN (●; n = 5/group).