Long term co-transplantation of human ovarian tissue with AMH-producing endothelial cells increase productivity and longevity of the graft

Abstract

A major obstacle to salvaging a robust pool of functional follicles from cryopreserved ovarian tissue following auto-transplantation is the global activation of the follicular pool that occurs during the ischemic window. AMH has been suggested to exert a repressive input on activation and/or growth of follicles during normal folliculogenesis. We used lentivirus to engineer exogenous endothelial cells (exECs) to constitutively express and secrete AMH, aiming to reduce premature follicular mobilization and growth upon co-transplantation in long-term grafts. Xenograft of human ovarian tissue into NSG mice with co-transplantation of ECs, and ECs ectopically expressing human AMH. In order to consolidate the phenomenon of activation, despite co-transplantation with ECs, we co-transplanted multiple grafts from a single patient and harvested at 2 (n=3), 3 (n=4), and 14 (n=2) weeks. Also, ECs were transduced with lentiviral vectors linked to a fluorescent reporter gene (mCherry) expressing AMH. Human ovarian tissue was co-transplanted with AMH-ECs into NSG oophorectomized mice (n=2); as controls, we co-transplanted patient-matched ovarian tissue with ECs (n=2). Engrafted tissue was harvested 14 weeks after transplantation. The ratio of follicles in each treatment was assessed in histologic sections using light and confocal microscopy. Control grafts harvested at 2, 3 and 14 weeks revealed a shift in the follicular pool away from quiescence with a decreased percentage of primordial follicles upon increased graft length. In contrast, grafts co-transplanted with AMH-ECs showed a significant retention of primordial follicles at 14 weeks. In these long-term grafts, we found a 4.6-fold increase in primordial follicles percentage with AMH-ECs; 18.39±0.75 vs. blank-ECs 3.99±1.61, P=0.01. For primary follicles, we found a reverse ratio, an approximately 0.8-fold decrease compared to blank-ECs; 33.40±3.28 vs. 41.94±1.30 respectively, P=0.05. Engineered exECs that constitutively express AMH induced a greater proportion of quiescent primordial follicles relative to control exECs, indicating suppression of premature mobilization that has been noted in the context of ovarian tissue transplantation. These findings present a cell-based strategy that combines accelerated perfusion with a direct paracrine delivery of a bioactive payload to transplanted ovarian tissue. Improved tissue viability and enforced retention of a quiescent primordial follicle pool combine to increase productivity and longevity of ovarian tissue grafts.