Abstract 028: BPH/5 Pre-eclamptic Mice Have Increased Uterine Interleukin (IL)-15 And This Is Associated with Decidual Natural Killer (dNK) Cell Loss At The Maternal-Fetal Interface Early In Pregnancy

Abstract

Preeclampsia (PE) is a hypertensive disease that affects 5-10% of pregnant women worldwide. Although a leading cause of maternal and perinatal morbidity/mortality, its origin is unknown. Abnormal placentation resulting in reduced placental perfusion is implicated. Early in pregnancy dNK cells are recruited to the maternal-fetal interface to promote uterine angiogenesis during decidualization. Uterine IL-15 activates dNK cells, which produce IFN-γ to orchestrate vessel remodeling. Using the BPH/5 mouse model that spontaneously develops the cardinal features of PE along with placental abnormalities, we tested the hypothesis that uterine inflammation is associated with dNK cell loss at the maternal-fetal interface prior to placenta formation. At the start of decidualization (e4.5-5.5), IL-15 mRNA was increased in BPH/5 implantation sites vs C57 controls (60-fold vs 20-fold respectively, p<0.05, n=4). mRNA expression of the dNK cell activation marker NKp46 was similar in BPH/5 implantation sites vs C57 at e4.5, but reduced by 10-fold at e5.5 (p<0.05, n=4). This finding was confirmed by flow cytometry (BPH/5: 0.7 x 10 3 cells vs C57: 9.7 x 10 3 cells, p<0.05, n=4) and immunohistochemistry. Moreover, there was a 3-fold reduction of IFNγ mRNA in BPH/5 implantation sites vs C57 at the time of maximal decidualization (e7.5, p<0.05). Since prolonged IL-15 stimulation may initiate loss of NK cell function, we tested if elevated IL-15 is responsible for dNK cell loss in C57. IL-15 was increased in the pregnant uterus of C57 by intrauterine injection of recombinant IL-15/IL-15 Receptor-α complex (rIL-15/IL-15Rα) at e2.5. This did not change the number of implantation sites at e5.5 (veh: 8.7 ± 0.8; rIL-15/IL15-Rα: 8 ± 1.2, p>0.05), but blunted NKp46 mRNA expression (30-fold vs veh, p<0.05) and decreased dNK cell numbers (veh: 9.6 x 10 3 vs rIL-15/IL-15Rα: 0.4 x 10 3 cells, n=3-5, p<0.05). This shows that BPH/5 females have increased uterine inflammation and dNK cell loss, and that artificial increases in intrauterine IL-15 in control mice cause dNK cell loss. This suggests that increased uterine expression of IL-15 contributes to dNK cell loss at the maternal-fetal interface in BPH/5 mice, which could lead to impaired placental perfusion and development of PE in this model.