Human amniotic fluid lacks interleukin-2 and interleukin-15 but can interact with the beta-chain of the interleukin-2 receptor

The present study investigated whether an explanation for the conflicting r eports on the interleukin-2 (IL-2) status of amniotic fluid is due to the p resence of IL-15 which shares biological activities with IL-2 and utilizes the IL-2 receptor beta-chain. Amniotic fluids from 45 normally progressing pregnancies between 14 and 16 weeks after the last menstrual period were as sayed for IL-2 and IL-15 by bioassay and enzyme-linked immunosorbent assay (ELISA). The ability of amniotic fluids to induce cytotoxic T lymphoblastoi d line-2 (CTLL-2) cell proliferation was demonstrated to be dependent upon bioassay culture conditions. In serum-free medium each amniotic fluid stimu lated CTLL-2 proliferation with a mean level of IL-2-like bioactivity of 14 .7 +/- 2.3 ng/ml but amniotic fluids failed to induce CTLL-2 proliferation in serum-supplemented medium. Treatment with neutralizing anti-IL-2 or anti -IL-15 antibodies failed to inhibit amniotic fluid-induced CTLL cell prolif eration in serum-free medium, indicating a lack of IL-2 and IL-15 bioactivi ty. In contrast, treatment with anti-IL-2 receptor beta-chain antibody sign ificantly reduced amniotic fluid-induced proliferation. The lack of IL-2 an d IL-15 activity in amniotic fluids was confirmed using ELISA. Although hig h levels of IL-15 immunoactivity were detected in all samples, specificity controls showed a lack of specific IL-15 immunoactivity in amniotic fluid. Pretreatment of amniotic fluids with 100-500 ng/ml mouse immunoglobulin G a brogated IL-15 immunoactivity, indicating that amniotic fluid contains mole cules binding to Fc regions of immunoglobulins and responsible for false EL ISA positivity. These studies unequivocally show that amniotic fluid lacks IL-2 and IL-15 but can stimulate CTLL-2 cell proliferation via the IL-2 rec eptor beta-chain. The absence of IL-2 and IL-15 in normal mid-trimester amn iotic fluid suggests that the cytokine profile of human pregnancy appears t o be associated with a bias against type 1 cytokines within the feto-placen tal unit.