Gamze Bildik1, Nazli Akin1, Ayse Seyhan2, Yashar Esmaeilian1, Kayhan Yakin2,3, Ipek Keles2, Basak Balaban2, Baris Ata2,3, Bulent Urman2,3, Ozgur Oktem1,2,3. 1. Graduate School of Health Sciences, Koc University, Davutpasa, Topkapi, Istanbul, Turkey. 2. American Hospital Women's Health Center, Assisted Reproduction Unit, Istanbul, Turkey. 3. Department of Obstetrics and Gynecology, Division Reproductive Endocrinology and Infertility, School of Medicine, Koc University, Davutpasa, Topkapi, Istanbul, Turkey.
Abstract
STUDY QUESTION: Are there any differences in the molecular characteristics of the luteal granulosa cells (GC) obtained from stimulated versus non-stimulated (natural) IVF cycles that may help explain the defective luteal phase in the former? SUMMARY ANSWER: Luteal GC of stimulated IVF cycles, particularly those of agonist-triggered antagonist cycles, are less viable ex vivo, express LH receptor and anti-apoptotic genes at lower levels, undergo apoptosis earlier and fail to maintain their estradiol (E2) and progesterone (P4) production in comparison to natural cycle GC. WHAT IS KNOWN ALREADY: Luteal function is defective in stimulated IVF cycles, which necessitates P4 and/or hCG administration (known as luteal phase support) in order to improve clinical pregnancy rates and prevent miscarriage. The luteal phase becomes shorter and menstruation begins earlier than a natural cycle if a pregnancy cannot be achieved, indicative of early demise of corpus luteum (premature luteolysis). Supra-physiological levels of steroids produced by multiple corpora luteae in the stimulated IVF cycles are believed to inhibit LH release directly via negative feedback actions on the hypothalamic-pituitary-ovarian axis resulting in low circulating levels of LH and a defective luteal phase. We hypothesized that some defects in the viability and steroidogenic activity of the luteal GC of the stimulated IVF cycles might contribute to this defective luteal phase in comparison to natural cycle GC. This issue has not been studied in human before. STUDY DESIGN, SIZE, DURATION: A comparative translational research study of ex vivo and in vitro models of luteal GC recovered from IVF patients undergoing natural versus stimulated IVF cycles was carried out. Luteinized GC were obtained from 154 IVF patients undergoing either natural (n = 22) or stimulated IVF cycles with recombinant FSH and GnRH agonist (long) (n = 44), or antagonist protocol triggered conventionally either with recombinant hCG (n = 46) or with a GnRH agonist (n = 42). GC were maintained in vitro for up to 6 days. PARTICIPANTS/MATERIALS, SETTING, METHODS: Cellular viability (YO-PRO-1 staining), the expression of the steroidogenic enzymes, pro-apoptotic genes [Bcl-2-associated death promoter (BAD), Bcl-2-associated X protein (BAX) and Caspase-3 (CASP3)], anti-apoptotic genes [RAC-alpha serine/threonine-protein kinase (AKT-1) and Bcl-2-like protein 2 (BCL2-L2)], LH receptor, vascular endothelial growth factor (VEGF) (using real-time quantitative PCR at mRNA level and western blot immunoprecipitation assay at protein level) and in vitro E2 and P4 production (electrochemiluminescence immunoassay) were compared in GC among the groups. MAIN RESULTS AND THE ROLE OF CHANCE: Natural cycle GC were significantly more viable ex vivo (88%) compared to their counterparts of the stimulated IVF cycles (66, 64 and 37% for agonist and antagonist cycles triggered with hCG and GnRH agonist respectively, P < 0.01). They were also more capable of maintaining their vitality in culture compared to their counterparts from the stimulated IVF cycles: at the end of the 6-day culture period, 74% of the cells were still viable whereas only 48, 43 and 22% of the cells from the agonist and antagonist cycles triggered with hCG and agonist respectively, were viable (P < 0.01). The mRNA expression of anti-apoptotic genes (AKT-1 and BCL2-L2) was significantly lower, while that of pro-apoptotic genes (BAD, BAX and CASP3) was significantly higher in the stimulated cycles, particularly in the agonist-triggered antagonist cycles, compared to natural cycle GC (P < 0.01 for long protocol and antagonist hCG trigger, P < 0.001 for agonist trigger). The expression of steroidogenic enzymes (stAR, SCC, 3β-HSD and aromatase) and VEGF was significantly higher in the agonist and hCG-triggered antagonist cycles compared to natural cycle GC. Therefore, in vitro E2 and P4 production in cells from the stimulated IVF cycles was significantly higher than their counterparts obtained from the natural cycles in the first 2 days of culture. However, after Day 2, their viability and hormone production began to decline very rapidly with the most drastic decrease being observed in the agonist-triggered cycles. By contrast, natural cycle GC maintained their viability and produced E2 and P4 in increasing amounts in culture up to 6 days. In vitro P production and the mRNA and protein expression of LH receptor, VEGF and 3β-HSD were most defective in the agonist-triggered antagonist cycles compared to natural and agonist and hCG-triggered antagonist cycles. In vitro hCG treatment of a subset of the cells from the agonist-triggered cycles improved their viability, increased E2 and P4 production in vitro and up-regulated the mRNA expression of anti-apoptotic gene BCL-L2 together with steroidogenic enzymes stAR, SCC, 3B-HSD, LH receptor and VEGF. LARGE SCALE DATA: Not applicable. LIMITATIONS, REASONS FOR CAUTION: The limitations include analysis of luteinized GC only might not reflect the in vivo mechanisms involved in survival and function of the whole corpus luteum; GC recovered during oocyte retrieval belong to a very early stage of the luteal phase and might not be representative; effects of ovulation triggered with hCG may not equate to the endogenous LH trigger; the clinical characteristics of the patients may vary among the different groups and it was not possible to correlate stimulation-related molecular alterations in luteal GC with the clinical outcome, as no oocytes have been utilized yet. Therefore, our findings do not conclusively rule out the possibility that some other mechanisms in vivo may also account for defective luteal function observed in stimulated IVF cycles. WIDER IMPLICATIONS OF THE FINDINGS: Ovarian stimulation is associated with significant alterations in the viability and steroidogenic activity of luteal GC depending on the stimulation protocol and mode of ovulation trigger. Reduced survival and down-regulated expression of 3B-HSD, LH receptor and VEGF leading to compromised steroid production in stimulated cycles, and particularly in the agonist-triggered cycles, may at least in part help explain why the luteal phase is defective and requires exogenous support in these cycles. STUDY FUNDING/COMPETING INTEREST(S): This study was funded by the School of Medicine, the Graduate School of Health Sciences of Koc University and Koç University Research Center for Translational Medicine (KUTTAM), equally funded by the Republic of Turkey Ministry of Development Research Infrastructure Support Program. All authors declare no conflict of interest.
STUDY QUESTION: Are there any differences in the molecular characteristics of the luteal granulosa cells (GC) obtained from stimulated versus non-stimulated (natural) IVF cycles that may help explain the defective luteal phase in the former? SUMMARY ANSWER: Luteal GC of stimulated IVF cycles, particularly those of agonist-triggered antagonist cycles, are less viable ex vivo, express LH receptor and anti-apoptotic genes at lower levels, undergo apoptosis earlier and fail to maintain their estradiol (E2) and progesterone (P4) production in comparison to natural cycle GC. WHAT IS KNOWN ALREADY: Luteal function is defective in stimulated IVF cycles, which necessitates P4 and/or hCG administration (known as luteal phase support) in order to improve clinical pregnancy rates and prevent miscarriage. The luteal phase becomes shorter and menstruation begins earlier than a natural cycle if a pregnancy cannot be achieved, indicative of early demise of corpus luteum (premature luteolysis). Supra-physiological levels of steroids produced by multiple corpora luteae in the stimulated IVF cycles are believed to inhibit LH release directly via negative feedback actions on the hypothalamic-pituitary-ovarian axis resulting in low circulating levels of LH and a defective luteal phase. We hypothesized that some defects in the viability and steroidogenic activity of the luteal GC of the stimulated IVF cycles might contribute to this defective luteal phase in comparison to natural cycle GC. This issue has not been studied in human before. STUDY DESIGN, SIZE, DURATION: A comparative translational research study of ex vivo and in vitro models of luteal GC recovered from IVFpatients undergoing natural versus stimulated IVF cycles was carried out. Luteinized GC were obtained from 154 IVFpatients undergoing either natural (n = 22) or stimulated IVF cycles with recombinant FSH and GnRH agonist (long) (n = 44), or antagonist protocol triggered conventionally either with recombinant hCG (n = 46) or with a GnRH agonist (n = 42). GC were maintained in vitro for up to 6 days. PARTICIPANTS/MATERIALS, SETTING, METHODS: Cellular viability (YO-PRO-1 staining), the expression of the steroidogenic enzymes, pro-apoptotic genes [Bcl-2-associated death promoter (BAD), Bcl-2-associated X protein (BAX) and Caspase-3 (CASP3)], anti-apoptotic genes [RAC-alpha serine/threonine-protein kinase (AKT-1) and Bcl-2-like protein 2 (BCL2-L2)], LH receptor, vascular endothelial growth factor (VEGF) (using real-time quantitative PCR at mRNA level and western blot immunoprecipitation assay at protein level) and in vitro E2 and P4 production (electrochemiluminescence immunoassay) were compared in GC among the groups. MAIN RESULTS AND THE ROLE OF CHANCE: Natural cycle GC were significantly more viable ex vivo (88%) compared to their counterparts of the stimulated IVF cycles (66, 64 and 37% for agonist and antagonist cycles triggered with hCG and GnRH agonist respectively, P < 0.01). They were also more capable of maintaining their vitality in culture compared to their counterparts from the stimulated IVF cycles: at the end of the 6-day culture period, 74% of the cells were still viable whereas only 48, 43 and 22% of the cells from the agonist and antagonist cycles triggered with hCG and agonist respectively, were viable (P < 0.01). The mRNA expression of anti-apoptotic genes (AKT-1 and BCL2-L2) was significantly lower, while that of pro-apoptotic genes (BAD, BAX and CASP3) was significantly higher in the stimulated cycles, particularly in the agonist-triggered antagonist cycles, compared to natural cycle GC (P < 0.01 for long protocol and antagonist hCG trigger, P < 0.001 for agonist trigger). The expression of steroidogenic enzymes (stAR, SCC, 3β-HSD and aromatase) and VEGF was significantly higher in the agonist and hCG-triggered antagonist cycles compared to natural cycle GC. Therefore, in vitro E2 and P4 production in cells from the stimulated IVF cycles was significantly higher than their counterparts obtained from the natural cycles in the first 2 days of culture. However, after Day 2, their viability and hormone production began to decline very rapidly with the most drastic decrease being observed in the agonist-triggered cycles. By contrast, natural cycle GC maintained their viability and produced E2 and P4 in increasing amounts in culture up to 6 days. In vitro P production and the mRNA and protein expression of LH receptor, VEGF and 3β-HSD were most defective in the agonist-triggered antagonist cycles compared to natural and agonist and hCG-triggered antagonist cycles. In vitro hCG treatment of a subset of the cells from the agonist-triggered cycles improved their viability, increased E2 and P4 production in vitro and up-regulated the mRNA expression of anti-apoptotic gene BCL-L2 together with steroidogenic enzymes stAR, SCC, 3B-HSD, LH receptor and VEGF. LARGE SCALE DATA: Not applicable. LIMITATIONS, REASONS FOR CAUTION: The limitations include analysis of luteinized GC only might not reflect the in vivo mechanisms involved in survival and function of the whole corpus luteum; GC recovered during oocyte retrieval belong to a very early stage of the luteal phase and might not be representative; effects of ovulation triggered with hCG may not equate to the endogenous LH trigger; the clinical characteristics of the patients may vary among the different groups and it was not possible to correlate stimulation-related molecular alterations in luteal GC with the clinical outcome, as no oocytes have been utilized yet. Therefore, our findings do not conclusively rule out the possibility that some other mechanisms in vivo may also account for defective luteal function observed in stimulated IVF cycles. WIDER IMPLICATIONS OF THE FINDINGS: Ovarian stimulation is associated with significant alterations in the viability and steroidogenic activity of luteal GC depending on the stimulation protocol and mode of ovulation trigger. Reduced survival and down-regulated expression of 3B-HSD, LH receptor and VEGF leading to compromised steroid production in stimulated cycles, and particularly in the agonist-triggered cycles, may at least in part help explain why the luteal phase is defective and requires exogenous support in these cycles. STUDY FUNDING/COMPETING INTEREST(S): This study was funded by the School of Medicine, the Graduate School of Health Sciences of Koc University and Koç University Research Center for Translational Medicine (KUTTAM), equally funded by the Republic of Turkey Ministry of Development Research Infrastructure Support Program. All authors declare no conflict of interest.