High levels of apoptosis in ejaculated spermatozoa from infertile men

Design: Randomized, double-blinded, controlled trial. Materials/Methods: Forty-two women diagnosed with PCOS according to the 1990 NICHD consensus trial (oligomenorrhea, clinical or laboratory hyperandrogenism, absence of other causes of anovulation) were randomized to receive metformin 1000 mg bid, rosiglitazone 4 mg bid, or placebo. Diabetes was excluded using a 2 hour glucose tolerance test. Subjects underwent a frequently sampled intravenous glucose tolerance test (FSIGT) pre-treatment, and after 3 completed months of therapy. Results were analyzed using the minimal model, MINMOD. Insulin levels were measured using a specific and sensitive double antibody radioimmunoassay. Data was analyzed using ANOVA. Post hoc differences were analyzed using Tukey’s Honestly Significantly Different (HSD) test. Results: Seventeen subjects were randomized to metformin, 18 to rosiglitazone, and 7 to placebo. There was no difference in baseline BMI, hirsutism score, total testosterone, free testosterone, DHEAS, FSH, or LH among groups. The prevalence of impaired glucose tolerance at baseline was not significantly different among groups: 35% in the metformin group (6 of 17), 44% in the rosiglitazone group (8 of 18), and 29% in the placebo group (2 of 7). Post treatment, insulin sensitivity (Si) was significantly improved in the rosiglitazone group compared to the other groups (p ⬍0.05). Si increased from 2.27 ⫾ 0.69 to 3.63 ⫾ 0.89 min-1/(pmol/L) in the rosiglitazone group. Si did not improve in the metformin group, 2.41 ⫾ 0.60 to 2.36 ⫾ 0.82 min-1/(pmol/L), or the placebo group, 2.46 ⫾ 1.18 to 1.97 ⫾ 0.89 min-1/(pmol/L). First phase insulin secretion (AIRg) improved in the metformin group 393 ⫾ 69 to 690 ⫾ 157 pmol/L (p ⫽ 0.05), but not in the rosiglitazone 668 ⫾ 123 to 629 ⫾ 129 pmol/L or placebo groups 763 ⫾ 257 to 541 ⫾ 157 pmol/L. The disposition index (DI, the product of Si and AIRg) improved in the metformin and rosiglitazone groups (733 ⫾ 150 to 846 ⫾ 191, and 846 ⫾ 152 to 1261 ⫾ 273, respectively). DI did not improve in the placebo group (1531 ⫾ 898 to 768 ⫾ 277). Mean BMI was 36 ⫾ 1.4 (range 21– 63). Women lost an average of 2.7 ⫾ 1.2 lbs in the metformin group and 0.7 ⫾ 0.5 pounds in the placebo group. Women in the rosiglitazone group gained an average of 5.1 ⫾ 2.0 lbs (p ⫽ 0.001 for metformin and rosiglitazone). There was no significant changes in any other parameters including fasting glucose and insulin, total and free testosterone, DHEAS, FSH and LH among the groups after 3 months of treatment. Conclusions: Rosiglitazone improves insulin sensitivity and metformin improves the first phase insulin secretion in women with PCOS. Supported by: General Clinical Research Center (GCRC), supported by NIH Grant M01-RR-01346.

Wednesday, October 16, 2002 4:00 P.M. O-279 Extracellular matrix-activin interactions in granulosa cell growth and apoptosis. Erkan Buyuk, Maja Oktay, Filippo Giancotti, Zev Rosenwaks, Kutluk H. Oktay. Ctr for Reproductive Medicine and Infertility, Weill Medical Coll of Cornell Univ, New York, NY; Montefiore Medical Ctr, Bronx, NY; Memorial Sloan Kettering Cancer Ctr, New York, NY. Objective: The precise mechanism that regulates the granulosa cell growth in preantral follicles is unknown. We have previously shown in mouse ovarian organ culture studies that extracellular matrix (ECM) affects preantral follicle growth, in possible interaction with activin-a (1). The objective of this study was to determine the role of ECM and activin-a in preantral/undifferentiated granulosa cell growth and survival. Design: Controlled in vitro study. Materials/Methods: A spontaneously immortalized rat granulosa cell line (SIGC) was used. SIGC carry the characteristics of preantral/undifferentiated granulosa cells(2) and express integrins that bind collagen(C), laminin(L), and fibronectin(F)(3). They are responsive to activin-a as determined by luciferase assay using p3TP-lux reporter gene containing activin-responsive PAI-1 promoter(4). SIGC were cultured in triplicates in serum free media on C, L, F or polylysine(P) coated (negative control) coverslips for 24 – 48 hours with or without 30ng/ml activin-a. Cell proliferation was determined by BrdU uptake after 24 hours, and the apoptosis was determined by TUNEL assay at 24 and 48 hours using immunofluorescence microscopy in a blinded fashion. Results: All three types of ECM molecules increased granulosa cell proliferation significantly compared to polylysine (p ⫽ 0.0001)(Figure). However, the proliferation rate was higher if cells were grown on C and L

S106

Abstracts

than on F (p ⫽ 0.0085). Activin-a suppressed cell proliferation on C (82%) more significantly than L (42%) (p ⫽ 0.05) but did not significantly influence cell proliferation on polylysine. Likewise, all three types of ECM molecules prevented apoptosis compared to polylysine (p ⫽ 0.0001)(Table). Among the three ECM tested, C seemed to have the strongest antiapoptotic effect. Cells grown on C, regardless of activin-a treatment, showed significantly less apoptosis compared to F at 24 hours (p ⫽ 0.03) or to L&F in the absence of activin-a at 48 hours (p ⫽ 0.045, p ⫽ 0.04, respectively). However addition of activin-a increased apoptosis of cells grown on C after 48 hours (p ⫽ 0.006).

Effects of ECM and Activin-A on mean percentage of apoptosis in SIGC

C L F P

24 h

Activin48 h

P

24 h

Activin⫹ 48 h

P

2.5 3.3 4.7 51.3

4.5 10.9 11.4 67

n.s. 0.0001 0.04 n.s

1.7 2.3 4.3 62.9

7.5 12.3 9.5 61

0.006 0.002 0.02 n.s

Conclusions: 1) ECM, especially C and L stimulate granulosa cell proliferation; 2) Activin-a suppresses granulosa cell proliferation especially on C, indicating an interaction between ECM and activin-a; 3) ECM, especially C prevents apoptosis but this protective effect may be reduced in the presence of activin-a. Overall, these results indicate that ECM and activin-a regulate survival and growth of granulosa cells of preantral follicles through complex interactions. The mechanism of these interactions is being investigated at the signaling level in both SIGC and primary granulosa cells. 1. Oktay et al. Biol Reprod 2000;63:457– 61. 2. Stein et al. Cancer Res 1991;51:696 –706. 3. Oktay et al. J Soc Gynecol Investig 2002;9(Supp): 338A. 4. Oktay et al. ASRM 2002 abstract. Supported by: ASRM-Serono Research Grant to K.O., and by the Center for Reproductive Medicine & Infertility.

Wednesday, October 16, 2002 4:15 P.M. O-280 High levels of apoptosis in ejaculated spermatozoa from infertile men. Mohamed H. Moustafa, Ramadan A. Saleh, Mehmet Oder, Anthony J. Thomas Jr., Mohammed A. Abdel-Hafez, Ashok Agarwal. Cleveland Clin Fdn, Cleveland, OH; Cairo Univ, Cairo, Egypt. Objective: Animal studies have suggested that apoptosis, i.e. programmed cell death, is a key regulator of spermatogenesis in normal and pathological states. In this study we compared levels of apoptosis in ejaculated spermatozoa from a group of infertile men and normal sperm donors. Design: Prospective study in a male infertility clinic. Materials/Methods: Our study included a randomly selected group of infertile men (n ⫽ 28) with a history of infertility of more than one year. Twenty-two donors with normal standard semen parameters served as a control. Semen samples were obtained after 2 to 3 days of sexual abstinence

Vol. 78, No. 3, Suppl. 1, September 2002

and examined according to World Health Organization guidelines (WHO, 1999). Semen smears were prepared for assessment of sperm morphology by WHO method and by Kruger strict criteria. Levels of apoptosis were detected using Annexin-V staining assay that detects externalization of phosphatidylserine to the outer surface of the plasma membrane of apoptotic spermatozoa. Propidium iodide (PI) stain was used to exclude necrotic spermatozoa. The percent apoptosis and necrosis were determined by epifluorescent microscopy. Two-hundred spermatozoa were randomly examined and classified as: normal (negative annexin-V and PI), apoptotic (positive annexin-V and negative PI) and necrotic (positive annexin-V and PI). Results: Comparisons of study parameters between patients and donors are shown in the table. Levels of apoptosis and necrosis were inversely correlated with sperm motility (r ⫽ ⫺0.32, P ⫽ 0.03 & r ⫽ ⫺0.53, P ⬍0.0001; respectively) and normal sperm forms, using strict criteria, (r ⫽ ⫺0.31, P ⫽ 0.03 & r ⫽ ⫺0.31, P ⫽ 0.03; respectively). Standard sperm parameters failed to predict high levels of apoptosis, using different cut-off values of apoptosis. Normal Donors (n ⫽ 22)

Parameters

Concentration (1 ⫻ 106/mL) 69 (52, 125) Motility (%) 68 (57, 77) Morphology by WHO (%) 34 (27, 36) Morphology by strict criteria (%) 10 (8, 13) Apoptosis (%) 6.7 (5, 8) Necrosis (%) 34 (27, 46)

Infertile Men (n ⫽ 28)

Pvalue

40 (20, 77) 0.02 45 (36, 62) 0.003 22 (15, 36) 0.02 6 (3, 11) 0.005 12 (9, 16) ⬍0.001 56 (46, 63) ⬍0.001

Values are median and intequartile range (25th and 75th percentiles). Wilcoxon rank-sum test was used for the analysis. P ⬍ 0.05 was significant. Conclusions: Our results indicate a significant increase in the levels of apoptosis in ejaculated spermatozoa from infertile men. Standard sperm parameters are not predictive of high levels of apoptosis. Apoptosis may be an independent phenomenon, which plays an important role in the pathophysiology of male infertility. Further research is required to elucidate the exact mechanism(s) of association of apoptosis with male infertility. Supported by: None.

Wednesday, October 16, 2002 4:30 P.M. O-281 Synchronization of recipients for embryo transfer: A novel method using a gonadotropin releasing hormone (GnRH) antagonist in rhesus monkeys. Sherri L. Thormahlen, Richard L. Stouffer, Don P. Wolf, Mary B. Zelinski-Wooten. Oregon National Primate Research Ctr, Beaverton, OR; Oregon National Primate Research Ctr, Dept of Physiol and Pharm, Oregon Health & Science Univ, Beaverton, OR; Oregon National Primate Research Ctr, Dept of Obstet and Gynec, Oregon Health & Science Univ, Beaverton, OR. Objective: Artificial ovarian cycles using exogenous estrogen and progesterone are currently used to synchronize embryo recipients, i.e. prepare the endometrium in donor oocyte cycles. In macaques and women, surgical ablation of the dominant follicle or corpus luteum during spontaneous cycles is followed by development of a new follicle that ovulates 12–14 days later. Therefore, the utility of chemical ablation of the dominant ovarian structure to control the timing of subsequent ovulation and synchronize recipients for embryo transfer was evaluated in rhesus monkeys. Design: Prospective assignment of macaques to receive GnRH antagonist during midfollicular or midluteal phases of the menstrual cycle. Materials/Methods: Adult, female rhesus monkeys exhibiting normal cyclicity (n⫽3/group) received either a single (0800 h) or double (0800 and 1600 h) sc injection of 0.25 mg cetrorelix acetate (Cetrotide, Serono, Inc.) for 3 consecutive days during the midfollicular (days 7–9 post-menses) or midluteal (days 7–9 post-ovulation) phase of spontaneous menstrual cycles. The success of follicle or corpus luteum ablation and the interval to subsequent ovulation were determined by daily steroid levels obtained prior to, during, and post-treatment. Patterns of steroid hormones were compared using analyses of variance with one repeated measure followed by Neuman

FERTILITY & STERILITY威

Keuls or t-tests. The interval to ovulation during the preceding spontaneous cycle and the subsequent post-treatment cycle was compared within animals using paired t-tests. Results: Neither dose of GnRH antagonist given at midfollicular phase consistently ablated follicular function. Estradiol levels were reduced to ⬍30 pg/ml within 24 h post-injection in only 1/3 animals in both midfollicular groups. Interestingly, successful ablation of the dominant follicle was achieved when estradiol levels on the first day of injection were ⬍100 pg/ml. Ovulation occurred 12 and 18 days after follicular ablation in these 2 animals. The single daily dose of GnRH antagonist during midluteal phase did not induce premature luteal regression. In contrast, two daily injections at midluteal phase suppressed (p ⬍0.05) progesterone to levels ⬍1.0 ng/ml within 48 h and maintained inadequate luteal function thereafter. The subsequent interval to ovulation (15 ⫾ 1 days) was similar to the pretreatment interval (15 ⫾ 3). Conclusions: Acute exposure to a GnRH antagonist at the doses used was more successful in ablating luteal than follicular function during spontaneous menstrual cycles in macaques. The stage of follicle maturation at which treatment is initiated or the dose may impact the ability of GnRH antagonist to suppress further development of the dominant follicle. Nonetheless, chemical ablation of the dominant ovarian structure during spontaneous menstrual cycles with a GnRH antagonist can potentially control the timing of ovulation for synchronization of oocyte donors with recipients or multiple recipients for embryo transfer. Supported by: NIH grants HD31633, RR12804 and RR00163.

Wednesday, October 16, 2002 4:45 P.M. O-282 Association of basal follicle stimulating hormone (FSH) levels with FSH receptor variants in subfertile women with normal menstrual cycle: Implications for ovarian sensitivity to FSH. Cornelis B. Lambalk, Tamar Benjamins, Pieter Harms, Joris M. van Montfrans, Jorg Gromoll, Manuela Simoni. Div of Reproductive Medicine, Vrije Univ, Amsterdam, Netherlands; Institute of reproductive Medicine of the Univ of Muenster, Muenster, Germany. Objective: Recently a frequently occurring polymorphic variant of the FSH receptor gene has been described in which the amino acid asparagine (Asn) is replaced by serine (Ser) at position 680. The Ser variant is associated with higher FSH levels in the early follicular phase and more FSH is needed to obtain normal follicular response in ovulatory patients undergoing IVF treatment.(Perez Mayorga et al. J Clin Endocrinol Metab 85: 3365–3369, 2000). Aim of our study was to test whether this receptor variant occurs more often in regularly menstruating patients with elevated basal FSH, a condition which is associated with a higher ovarian FSH threshold (de Koning et al Hum Reprod 16 abstract book 1: 207, 2001). Design: Retrospective cohort study. Materials/Methods: DNA was obtained from EDTA blood samples from subfertile patients(⬎1 year) with a regular menstrual cycle and elevated (FSH10 IU/L) or normal early follicular phase FSH who participated to an earlier case control study to investigate chance of natural or induced pregnancy (van Montfrans et al. Fertil Steril 74:97–103, 2000). Determination of the 680 FSH receptor variant was performed using the allelic discrimination technique with fluorescent dyes specific for each variant. The PCR reactions were run on ABI 7000 sequence detection system (Applera, USA). Results: See Table Distribution of Wildtype (Asn/Asn680) and variants for FSH receptor in patients with/without elevated FSH.*

FSH10 IU/L (n ⫽ 38) FSH ⬍ 10 IU/L (n ⫽ 40)

Asn/Asn n ⫽ 22

Asn/Ser n ⫽ 39

Ser/Ser n ⫽ 13

8 (21%) 18 (45%)

22 (58%) 17 (42.5%)

8 (21%) 5 (12.5%)

* : p ⬍ 0.05 (linear-by-linear association). Conclusions: The 680Asn variant of the FSH receptor is significantly associated with FSH levels ⬍10 IU/L in more than 80 % (based on allelic

S107