Leptin suppresses anti-Mullerian hormone gene expression through the JAK2/STAT3 pathway in luteinized granulosa cells of women undergoing IVF

Abstract

Do the adipocytokines, leptin and adiponectin affect the granulosa cell expression of anti-Mullerian hormone (AMH) and its receptor (AMHR-II)? Leptin suppresses AMH mRNA levels in human luteinized granulosa cells through the JAK2/STAT3 pathway, while adiponectin has no such effect. AMH is one of the most reliable markers of ovarian reserve. Serum AMH levels decline with obesity. Obesity is associated with elevated leptin and reduced adiponectin levels. This prospective study included 60 infertile women undergoing fresh IVF and ICSI cycles utilizing autologous oocytes at Montefiore's Institute for Reproductive Medicine and Health between July 2010 and April 2012. Follicular fluid was collected from small (SFs; <14 mm) and large follicles (LFs; ≥14 mm) from 38 participants. Total RNA was extracted separately from mural and cumulus granulosa cells and mRNA levels were measured by RT-PCR. In an additional group of participants (N = 22), primary cumulus and mural granulosa cells (pooled SFs and LFs) were cultured in media alone or with addition of either leptin (N = 7), adiponectin (N = 8) or JAK2/STAT3 inhibitor + leptin (N = 7), and AMH and AMHR-II mRNA levels measured. Levels of AMH, leptin and adiponectin protein were measured in follicular fluid. AMH and AMHR-II mRNA and follicular fluid AMH protein levels were inversely correlated with age. AMH mRNA expression was six times higher in cumulus compared with mural granulosa cells in SFs (P< 0.05) and eight times higher in cumulus compared with mural granulosa cells in LFs (P < 0.001). In follicular fluid, leptin protein level positively correlated (r = 0.7, P = 0.03), while adiponectin protein level inversely correlated (r = -0.46, P = 0.02) with BMI. Leptin treatment suppressed AMH and AMHR-II mRNA in both cumulus and mural granulosa cells (all P < 0.05). In the presence of JAK2/STAT3 inhibitor, leptin treatment did not alter AMH but continued to suppress AMHR-II mRNA in cumulus cells (P = 0.02). Adiponectin treatment did not alter AMH or AMHR-II mRNA levels. This study included a luteinized granulosa cell model as these cells were collected from women who were hyperstimulated with gonadotrophins. The results obtained may not fully extrapolate to non-luteinized granulosa cells. Leptin may program abnormal AMH signaling, thereby resulting in ovarian dysfunction. This study opens a new perspective for understanding the low ovarian reserve seen in obese women and provides new insights into potential mechanisms that explain the lower AMH seen in obese women. Whether our findings explain the worse response to ovulation induction observed in obese women needs to be further elucidated.