Carmen Mendoza 1,2, Nieves Cremades 3, Estrella Ruiz-Requena 1, Francisco Martinez 1, Esperanza Ortega 1, Sergio Bernabeu 3 and Jan Tesarik 4,5

Transcription

1 Human Reproduction vol.14 no.3 pp , 1999 Relationship between fertilization results after intracytoplasmic sperm injection, and intrafollicular steroid, pituitary hormone and cytokine concentrations Carmen Mendoza 1,2, Nieves Cremades 3, Estrella Ruiz-Requena 1, Francisco Martinez 1, Esperanza Ortega 1, Sergio Bernabeu 3 and Jan Tesarik 4,5 1 Department of Biochemistry and Molecular Biology, University of Granada, Granada, Spain, 2 MAR &Gen, Molecular Assisted Reproduction and Genetics, Granada, Spain, 3 Bernabeu Institute of Fertility and Gynaecology, Alicante, Spain and 4 Laboratoire d Eylau, Paris, France 5 To whom correspondence should be addressed at the Laboratoire d Eylau, 55 Rue Saint-Didier, Paris, France Previous studies relating hormone and cytokine concentrations in follicular fluid to oocyte fertilizability were flawed by the uncertainty about the actual oocyte maturity status at the time of recovery and by the possible contribution of the male factor to failures of conventional in-vitro fertilization. This is the first study in which oocyte maturity was assessed immediately after recovery and only mature oocytes were selected for treatment by intracytoplasmic sperm injection. Fertilization outcomes were related to follicular fluid concentrations of 17β-oestradiol, progesterone, follicle stimulating hormone, luteinizing hormone (LH), growth hormone (GH), prolactin (PRL), interleukin-1 (IL-1) and tumour necrosis factor-α (TNFα). Those oocytes that subsequently showed normal fertilization were harvested from follicles with higher concentrations of progesterone, GH, PRL, IL-1 and TNFα as compared with those of oocytes that failed to fertilize. Among the normally fertilized oocytes, low GH concentrations were associated with the failure of cleavage and with poor morphology of cleaving embryos, whereas rapidly cleaving embryos developed from oocytes recovered from follicles with high concentrations of LH and IL-1. These data suggest important roles for GH, IL-1 and TNFα, and of residual LH after pituitary suppression, as positive regulators of the final phase of oocyte intrafollicular development. Key words: cytokines/fertilization results/follicular fluid/icsi/ pituitary hormones Introduction Fertilization results in human assisted reproduction are influenced by a combination of male and female factors. Thus, fertilization failure in an in-vitro fertilization (IVF) attempt can be due to a sperm abnormality, poor oocyte quality or both. As compared with standard IVF, intracytoplasmic sperm injection (ICSI) into oocytes is mostly indicated in cases of poorer sperm quality, in which the high risk of failure of classical IVF would be expected. Paradoxically, even spermatozoa from such poor quality samples yield high fertilization and pregnancy rates when used in ICSI (Nagy et al., 1995). Fertilization failure after ICSI is thus likely to be mainly due to abnormalities of the oocyte. During the final phase of ovarian follicular development, the oocyte resides in an antral follicle where it is initially associated with specialized granulosa cells (cumulus oophorus and corona radiata cells) and where it is exposed to a particular humoral microenvironment (follicular fluid) whose composition differs from that of blood plasma. The final phase of oocyte meiotic and cytoplasmic maturation, coinciding with the development and growth of antral follicles, is subject to a complex interplay of endocrine, paracrine and autocrine control mechanisms (reviewed in Gougeon, 1996). Hormones and other regulatory substances involved in these mechanisms are either locally secreted within the ovary (steroid hormones, cytokines) or are produced outside and enter the follicles secondarily. The intrafollicular concentration of some of these agents at specific times of antral follicle development is likely to be related to the success or failure of various developmental processes in the oocyte that are necessary for its fertilizability and further developmental competence. A number of studies have attempted to find a relationship between the concentration of steroids (McNatty et al., 1975, 1979; Fowler et al., 1978; McNatty and Baird, 1978; Bomsel- Helmreich et al., 1979; Brailly et al., 1981; Wramsby et al., 1981; Carson et al., 1982; Fishel et al., 1983; Botero-Ruiz et al., 1984; Elsworth et al., 1984; Lobo et al., 1985; Chabab et al., 1986; Westergaard et al., 1986; Kreiner et al., 1987; Nayudu et al., 1987; Franchimont et al., 1989; Hartshorne, 1989; Seibel et al., 1989; Yding Andersen, 1990, 1993; Tavmergen et al., 1992; Yding Andersen et al., 1992; Enien et al., 1995; Cianci et al., 1996), pituitary hormones (McNatty et al., 1975; McNatty and Baird, 1978; Huyser et al., 1994) and cytokines (Barak et al., 1992; Huyser et al., 1994; Cianci et al., 1996; Branisteanu et al., 1997; Bili et al., 1998) in the follicular fluid, on the one hand, and different parameters of oocyte quality on the other hand. However, all of these studies dealt with classical IVF attempts and were thus unable to determine exactly the oocyte maturity status at the time of recovery. Therefore, oocyte maturity and developmental potential were estimated indirectly, by evaluating the cumulus oophorus and corona radiata morphology and by analysing fertilization results on the day following in-vitro insemination, when the somatic cells surrounding the oocyte were removed. In those conditions, the analysis of fertilization results may be biased by the uncertainty as to the oocyte maturity at the time 628 European Society of Human Reproduction and Embryology

2 Intrafollicular factors and ICSI results of in-vitro insemination. In fact, some of the oocytes showing the first polar body on the day after in-vitro insemination may still have been immature at the time when they were exposed to spermatozoa. Moreover, the conditions of ICSI restrict the multifactorial nature of fertilization success and failure, putting ahead those factors that are responsible for oocyte activation and the ensuing developmental processes culminating in the completion of oocyte meiosis and in the development of pronuclei. This study was undertaken to examine whether fertilization results and parameters of embryo quality and viability, namely the cleavage speed and embryo morphology, after ICSI are related to the concentration of selected steroids, pituitary hormones and cytokines, or to their ratios, detected in the fluid aspirated from the follicles from which the respective oocytes were harvested. Materials and methods Patients This study involves 15 infertile couples enrolled in the ICSI programme of the Bernabeu Institute of Fertility and Gynaecology. All these couples were undergoing the ICSI treatment for male factor infertility, without any apparent female contribution. Informed consent was obtained from the couples for the use of the follicular fluid samples, obtained during oocyte recovery for the ICSI treatment, for the analyses described in this study. Ovarian stimulation, follicular fluid sampling and oocyte collection Pituitary desensitization was performed with leuprolide acetate (Procrin; Abbot Laboratories, Madrid, Spain), beginning in the mid-luteal phase of the previous cycle. When complete ovarian suppression (serum oestradiol concentration 80 pmol/l and no follicles seen on ultrasound) was obtained, follicle growth was stimulated by combined administration of urinary follicle stimulating hormone (FSH) (Neofertinorm; Serono, Madrid, Spain) and human menopausal gonadotrophin (HMG) (Lepori; Pharma-Lepori, Barcelona, Spain), in a daily dose of 150 IU each, followed by injection of 5000 IU of human menopausal gonadotrophin (HCG) (Profasi HP; Serono) h before follicular aspiration. Follicular fluid was sampled by transvaginal ultrasound-guided puncture and aspiration of large antral follicles. Each follicle was aspirated separately and collected in a different dish. Fluid aspirated from each individual follicle was maintained isolated from that coming from other follicles. Flushing follicles with medium upon aspiration was avoided. Oocyte cumulus complexes were identified in the fluid samples coming from the individual follicles and used for ICSI. Those follicular fluid samples in which an expanded cumulus oocyte complex had been identified and that did not contain any visible blood contamination were used in this study. Determination of hormone and cytokine concentrations Commercial enzyme immunoassay kits (Boehringer, Mannheim, Germany) were used to determine follicular fluid concentrations of 17β-oestradiol, progesterone, FSH, luteinizing hormone (LH), prolactin (PRL), interleukin-1β (IL-1) and tumour necrosis factor-α (TNFα). The lowest standard concentration that could be measured with the IL-1 and TNFα kits was 0.05 pg/ml in both cases. Growth hormone (GH) was determined with the use of direct radioimmunoassay (Sorin Biomedica, Vercelli, Italy). In addition to absolute values, ratios of each hormone concentration to all other hormones tested were determined for each follicular fluid sample. From these ratios, only those showing significant differences between different types of oocytes and embryos (see below) are presented. ICSI, assessment of oocyte maturity, evaluation of fertilization and embryo grading ICSI and subsequent culture of injected oocytes and embryos were performed using a standard laboratory set-up and methodology (Tesarik and Sousa, 1995). Within 1 h after the recovery of oocyte cumulus complexes, the cumulus oophorus and corona radiata cells surrounding the oocytes were removed, and the oocytes were inspected for the signs of meiotic maturation. Mature (metaphase II), immature (germinal vesicle or metaphase I) and degenerative oocytes were distinguished. Only mature oocytes were processed by ICSI. Fertilization results were assessed h after ICSI. Fertilization was considered normal when the oocytes contained 2PN. Abnormally fertilized oocytes contained either 1PN or 3PN. Oocytes lacking pronuclei at this time of observation were considered unfertilized. Embryo cleavage was evaluated h after ICSI. The number of blastomeres and the morphological grade were noted for each embryo. Embryos with 10% anuclear fragments, with 10 20% fragments and with 20% fragments are referred to as grade 1, grade 2 and grade 3 respectively. Quantitative evaluation and statistics Follicular fluid concentrations of individual hormones and cytokines associated with each maturation and viability status of oocytes, with each fertilization result, and with each category of cleaving embryos are expressed as mean SD. Data were analysed for normality of distribution using Kolmogorov Smirnov test. Distribution was normal for oestradiol, progesterone, FSH, LH, PRL and GH, for which differences between individual groups of oocytes and embryos were evaluated by analysis of variance (ANOVA) and Student s t-test. For IL-1 and TNFα, for which distribution was not normal, Kruskal Wallis test was used. Variability of each hormone and cytokine value between individual women and between individual follicles was assessed by ANOVA (after logarithmic transformation where adequate). Because there was consistently more variability between individual follicles than between individual women, each follicular fluid sample was treated as a single event. Results General observations on oocytes and embryos Out of 164 oocytes from 15 treatment cycles, 114 oocytes have reached meiotic maturity (metaphase II), 27 were immature (20 at metaphase I and seven at the germinal vesicle stage), and 23 oocytes were degenerative. ICSI was performed with the 114 mature oocytes and resulted in the development of 76 two-pronucleated (2PN) zygotes. Of the remaining 38 oocytes injected, 18 developed an abnormal fertilization pattern (either 1PN or 3PN), whereas 20 other oocytes failed to be fertilized. Of the 76 2PN zygotes, nine remained undivided, 37 had two blastomeres, 14 had three blastomeres, and 16 had four blastomeres when observed h after ICSI. As to embryo morphology, 15 cleaving embryos were grade 1, 38 were grade 2, and 14 were grade 3 at the same time of observation. 629

3 C.Mendoza et al. Table I. Concentrations of selected steroids, pituitary hormones and cytokines in follicles yielding mature, immature and degenerative oocytes Oocyte category Oestradiol (pg/ml) Progesterone (pg/ml) FSH (IU/l) LH (IU/l) PRL (ng/ml) GH (ng/ml) IL-1 (pg/ml) TNFα (pg/ml) Mature (n 114) a c Immature (n 27) ab c Degenerative (n 23) b Values are means SD. a,b For values sharing at least one superscript letter within each column, P 0.05 (ANOVA and Student s t-test). c For values sharing the same superscript letter within each column, P 0.01 (ANOVA and Student s t-test). FSH follicle stimulating hormone; LH luteinizing hormone; GH growth hormone; PRL prolactin, IL-1 interleukin-1; TNFα tumour necrosis factor-α. Table II. Concentrations of selected steroids, pituitary hormones and cytokines in follicles yielding mature oocytes that developed a normal fertilization pattern (2PN), an abnormal fertilization pattern (1PN or 3PN) or failed to fertilize after intracytoplasmic sperm injection Fertilization result Oestradiol Progesterone FSH (IU/l) LH (IU/l) PRL (ng/ml) GH (ng/ml) IL-1 (pg/ml) TNFα (pg/ml) (pg/ml) (pg/ml) 2PN (n 76) c a a cd ef e 1PN or 3PN (n 18) d ab b c e f Unfertilized (n 20) cd b ab d f ef Values are means SD. a,b For values sharing at least one superscript letter within each column, P 0.05 (ANOVA and Student s t-test). c,d For values sharing at least one superscript letter within each column, P 0.01 (ANOVA and Student s t-test). e,f For values sharing at least one superscript letter within each column, P 0.05 (Kruskal Wallis test). Relationship between follicular fluid steroid, gonadotrophin and cytokine concentrations and oocyte maturity and viability With the exception of GH and PRL, concentrations of none of the hormones and cytokines under study was significantly different between follicles yielding mature, immature and degenerative oocytes (Table I). Also, follicles yielding mature oocytes had significantly lower (P 0.05) PRL/progesterone ratios as compared with those from which immature oocytes were harvested. Differences in the other hormone and cytokine concentrations measured and between their ratios were not significant for the three types of follicles. Relationship between follicular fluid steroid, gonadotrophin and cytokine concentrations and fertilization results Oocytes that subsequently developed into 2PN zygotes after ICSI were harvested from follicles with significantly higher concentrations of progesterone, GH, PRL, IL-1 and TNFα in follicular fluid than those of oocytes that failed to fertilize (Table II). Moreover, follicles yielding oocytes that showed abnormal forms of fertilization (one pronucleus or three pronuclei) after ICSI had lower concentrations of LH than those of oocytes developing into 2PN zygotes (Table II). These differences in the absolute values resulted in significant differences in all ratios involving one of the respective hormones and cytokines, on the one hand, and one of the other hormones and cytokines measured whose absolute values did not differ, on the other hand (data not shown). The other ratios showing significant differences between individual groups of oocytes are presented in Table III. These data show an association of high oestradiol/fsh and low PRL/TNFα, GH/ TNFα and IL-1/TNFα ratios with normally fertilized oocytes as compared with those of oocytes that failed to fertilize. 630 Moreover, follicles yielding oocytes that developed abnormal fertilization patterns had lower GH/TNFα ratios than normally fertilized oocytes (Table III). Relationship between follicular fluid steroid, gonadotrophin and cytokine concentrations and early embryonic development When absolute values of hormone and cytokine concentrations measured in follicular fluid were related to the cleavage speed of normally fertilized oocytes (2PN zygotes) harvested from the respective follicles, those oocytes that had reached the 4- cell stage by the time of observation (40 44 h after ICSI) came from follicles with higher concentrations of LH, PRL and IL-1 than those of less rapidly developing oocytes that only reached the 3-cell or the 2-cell stage by the time of observation (Table IV). Moreover, 2PN zygotes that failed to undergo cleavage developed from oocytes that were harvested from follicles with lower concentrations of progesterone, PRL and GH than those of other oocytes (Table IV). No additional differences between individual groups of oocytes were revealed by obtaining ratios of the absolute values of hormones and cytokines displayed in Table IV (data not shown). When absolute values of hormone and cytokine concentrations were related to embryo morphology (Table V), the only significant differences concerned LH and GH, whose concentrations were markedly reduced in follicles yielding oocytes that developed into embryos with poor morphology (grade 3) as compared with those of better-morphology embryos (grades 1 and 2). Moreover, follicles yielding oocytes that developed into grade 3 embryos had higher FSH/PRL ratios (P 0.01) and lower PRL/IL-1 ratios (P 0.05) than those of the two other categories of follicles (data not shown).

4 Intrafollicular factors and ICSI results Table III. Selected ratios of follicular fluid steroid, gonadotrophin and cytokine concentrations in follicles yielding mature oocytes that developed a normal fertilization pattern (2PN), an abnormal fertilization pattern (1PN or 3PN) or failed to fertilize after ICSI Fertilization result Oestradiol/FSH PRL/TNFα GH/TNFα IL-1/TNFα 2PN (n 76) a c c c 1PN or 3PN (n 18) b d c d Unfertilized (n 20) ab cd c cd Values are means SD. Each ratio is calculated using absolute values expressed in the same units as in Table II. a,b For values sharing at least one superscript letter within each column, P 0.05 (ANOVA and Student s t- test). c,d For values sharing at least one superscript letter within each column, P 0.05 (Kruskal Wallis test). For abbreviations, see Table I; PN pronucleus. Table IV. Concentrations of selected steroids, pituitary hormones and cytokines in follicles yielding mature oocytes that developed different numbers of cells by h after normal fertilization (two pronuclei) by intracytoplasmic sperm injection Number of cells Oestradiol Progesterone FSH (IU/l) LH (IU/l) PRL (ng/ml) GH (ng/ml) IL-1 (pg/ml) TNFα (pg/ml) (pg/ml) (pg/ml) 1(n 9) abc a ab abc d (n 37) a b a a e (n 14) b c b b f (n 16) c abc ab c def Values are means SD. a,b,c For values sharing at least one superscript letter within each column, P 0.05 (ANOVA and Student s t-test). d,e,f For values sharing at least one superscript letter within each column, P 0.05 (Kruskal Wallis test). For abbreviations, see Table I. Table V. Concentrations of selected steroids, pituitary hormones and cytokines in follicles yielding mature oocytes that developed into embryos of different quality by h after normal fertilization (two pronuclei) by intracytoplasmic sperm injection Embryo quality Oestradiol Progesterone FSH (IU/l) LH (IU/l) PRL (ng/ml) GH (ng/ml) IL-1 (pg/ml) TNFα (pg/ml) (pg/ml) (pg/ml) Grade 1 (n 15) a a Grade 2 (n 38) b b Grade 3 (n 14) ab ab Values are means SD. a,b For values sharing at least one superscript letter within each column, P 0.05 (ANOVA and Student s t-test). For abbreviations, see Table I. Discussion After the extensive application of in-vitro fertilization (IVF) in the treatment of human infertility in the early 1980s, many studies tried to relate hormone and cytokine concentrations in follicular fluid to oocyte maturity and fertilization results (see Introduction). However, this is the first study in which oocyte maturity could be evaluated shortly after oocyte recovery. It is well known that many oocytes that are immature shortly after recovery can reach metaphase II during culture, and that some cases reported as failures of fertilization may be, in fact, failures of oocyte maturation by the time of in-vitro insemination. In this study, the respective associations of follicular fluid hormone and cytokine concentrations with oocyte maturity, fertilizability and further developmental potential of the early embryo were dissected from each other, so that fertilization results were evaluated only for those oocytes that were at metaphase II at the time of ICSI, and further developmental progression was evaluated only in embryos resulting from normally fertilized zygotes. Moreover, the speed of early cleavage divisions could be better interpreted in terms of embryo quality in these conditions in which embryos were evaluated within a narrow time window after ICSI, as compared with conventional IVF where the exact time of gamete union, and thus the time zero of the beginning of embryonic development, is impossible to assess. This study dealt with follicular fluid concentrations of two pituitary gonadotrophins (FSH and LH), two non-gonadotrophin pituitary hormones (GH and PRL), two steroids (oestradiol and progesterone) and two cytokines (IL-1 and TNFα). These distinct groups of substances are further discussed separately. Follicle stimulating hormone and luteinizing hormone The follicular fluid concentrations of FSH and LH measured in this study must be interpreted with caution because of the previous suppression of gonadotrophin secretion by gonadotrophin-releasing hormone (GnRH) agonist administration and the application of exogeneous hormones during ovarian stimulation. Assuming that both gonadotrophins were sup- 631

5 C.Mendoza et al. pressed to the same degree, the measured LH concentrations reflect the residual pituitary secretion escaping the inhibitory action of GnRH agonist as well as the LH activity of the administered HMG, whereas the measured FSH/LH ratios basically reflect the amount of externally administered FSH used for ovarian stimulation. Of course, both the FSH and LH concentrations can be influenced by putative local mechanisms regulating the retention of each gonadotrophin within the follicle (McNatty et al., 1975; Mason et al., 1994; Jakimiuk et al., 1998). The remaining circulating concentrations of LH after profound pituitary suppression are 1 2 IU/l (Devroey et al., 1994), which is 2 4 times more than the values measured in this study in the follicular fluid of follicles harbouring mature oocytes. The present data did not show any relationship between the absolute concentrations of FSH and LH or the FSH/LH ratio, on the one hand, and oocyte meiotic maturity on the other hand. However, there were unexpectedly low absolute concentrations of LH in follicles yielding oocytes that developed abnormal fertilization patterns. Moreover, low LH values signalled developmental impairment well beyond fertilization, since oocytes harvested from follicles with higher LH concentrations cleaved more rapidly after fertilization and, on the other hand, high FSH/LH ratios were associated with poor embryo morphology. These differences are unlikely to be due to the administered dosage of FSH and LH, derived from HMG, because, if this were the case, the FSH/LH ratio would be constant. These findings thus suggest that residual pituitary LH production, persisting after GnRH agonist desensitization treatment, may be important for some intrafollicular events necessary for the optimal cytoplasmic maturation of the oocyte. Because LH is responsible for the production of aromatizable substrate by theca interna cells (Karnitis et al., 1994; Nahum et al., 1995), it is possible that, for instance, follicles exposed to extremely low concentrations of LH in the mid-follicular phase fail to produce sufficient oestradiol to support further normal development. Further study is needed to explain these observations. Growth hormone and prolactin Of the hormones and cytokines measured in this study, GH was the only one whose values were significantly higher for follicles yielding developmentally competent oocytes according to all parameters tested. GH is known to enhance several processes that are necessary for the pre-ovulatory development of follicles and oocytes, including the in-vivo stimulation of FSH-induced oestradiol production in humans (Lanzone et al., 1996), FSH-independent in-vitro stimulation of oestradiol production by isolated human granulosa cells (Mason et al., 1990), the enhancement of FSH-induced formation of LH and HCG receptors in rat granulosa cells (Jia et al., 1986), the amplification of FSH-induced progesterone production in rat (Hutchinson et al., 1988) and pig (Hsu and Hammond, 1987) granulosa cells, or the stimulation of androgen synthesis by rat thecal cells (Apa et al., 1996). Gonadotrophin-independent facilitation of progesterone synthesis by granulosa cells and acceleration of oocyte meiotic maturation have also been reported in the rat (Apa et al., 1994a,b). The present observation 632 that, among metaphase II oocytes, those with the best fertilizability and postfertilization developmental potential have been harvested from follicles with relatively high GH concentrations is consistent with the above findings. It can be speculated that an early rise of GH in small antral follicles is beneficial for oocyte quality by enhancing, or acting in synergy with, gonadotrophin-controlled developmental processes, whereas a delayed rise can accelerate the growth of those small antral follicles whose oocytes cannot achieve full cytoplasmic maturity. This would explain the present finding of high GH concentrations measured also in those follicles yielding meiotically immature oocytes. A tentative mechanism by which follicles regulate the intrafollicular GH concentration implies the local action of cytokines modifying the follicular wall permeability for circulating hormones (see below). This possibility is supported by the observation that the follicles yielding fertilizable oocytes also contained the highest concentration of PRL and LH, the other two hormones of extrafollicular origin evaluated in this study. The intrafollicular concentrations of FSH are irrelevant in this context because they mainly reflect the externally applied hormone. Oestradiol and progesterone Oestradiol concentrations in follicular fluid are known to increase before the pre-ovulatory LH surge, whereas the period after the LH surge is characterized by a dramatic increase in the concentration of progesterone (Sanyal et al., 1974; Bomsel-Helmreich et al., 1979; McNatty et al., 1979; Edwards et al., 1980). Published data about the relationship between oestradiol and progesterone in human follicular fluid and oocyte maturity and fertilizability are not consistent. High oestradiol concentrations were reported to be associated with oocyte maturity (Kreiner et al., 1987), fertilizability by conventional IVF (Wramsby et al., 1981) and the potential of rapid postfertilization development (Botero-Ruiz et al., 1984), whereas others did not find any relationship between oestradiol and progesterone concentrations in follicular fluid, on the one hand, and fertilization results, on the other hand (Uehara et al., 1985; Hartshorne, 1989). More recently, progesterone/ oestradiol ratios were shown to be higher in follicles whose oocytes fertilized (Enien et al., 1995). None of these studies dealt with ICSI treatment cycles, and the data could thus be biased by the unknown oocyte maturity status at the time of in-vitro insemination (see above). Differences in ovarian stimulation protocols may also have added some variability. In this study, we only found an association between higher progesterone values and oocytes capable of fertilization by ICSI. Interleukin-1 and tumour necrosis factor-α Both IL-1 and TNFα have previously been detected in human follicular fluid (Khan et al., 1988; Roby et al., 1990; Zolti et al., 1990). In addition to leukocytes and activated tissue macrophages that are well-known cellular sources of the two cytokines, human granulosa cells have been found to express both IL-1β and TNFα (Machelon and Emilie, 1997). The ovarian expression of both cytokines is hormonally regulated

6 Intrafollicular factors and ICSI results and reaches a peak in the peri-ovulatory period (Zolti et al., 1990; Hurwitz et al., 1992). Concentrations of IL-1 and TNFα in human follicular fluid, in relation to IVF results, were analysed previously by two independent studies (Barak et al., 1992; Bili et al., 1998) in which a positive correlation between IL-1 and TNFα concentrations, but no correlation between the concentration of either cytokine and oocyte fertilization or embryo quality, was found. Our data support only the former observation because the highest IL-1 and TNFα concentrations were associated with normally fertilizable oocytes whose follicles also had the highest progesterone. Moreover, we also found a difference in the concentration of IL-1 between follicles yielding normally fertilizable oocytes and those whose oocytes failed to fertilize or developed an abnormal fertilization pattern. The fact that follicles yielding normally fertilizable oocytes had the highest concentrations of IL-1 and TNFα supports the hypothesis that both cytokines are involved in the regulation of processes influencing oocyte quality. Because TNFα can stimulate angiogenesis (Leibovich et al., 1987), and IL-1, in addition to stimulating TNFα secretion, is known to enhance directly vascular permeability (Dinarello, 1988), both cytokines may act in synergy to ensure enhanced entry of circulating pituitary hormones into small antral follicles. In fact, the category of follicles yielding normally fertilizable oocytes in this study, which showed the highest concentrations of IL-1 and TNFα, was also that with the highest concentrations of LH, GH and PRL (see above). In later phases of antral follicle development, IL-1 and TNFα may be instrumental in operating the peri-ovulatory switch of the follicular steroid production from oestradiol to progesterone predominance (Adashi, 1990; Machelon and Emilie, 1997) through their inhibitory effect on FSH-induced oestradiol production by granulosa cells (Fukuoka et al., 1992; Watanabe et al., 1994; Best and Hill, 1995; Rice et al., 1996). The present findings corroborate this hypothesis by demonstrating the coincidence between high IL-1, TNFα and progesterone concentrations in that category of follicles from which fertilizable oocytes were harvested. However, other mechanisms, such as the possible contribution of IL-1 to nitric oxide generation (Tao et al., 1997), may also come into play. It remains to be elucidated whether IL-1 and TNFα can directly influence oocyte quality or whether they merely reflect the activity of an independent mechanism that influences at the same time oocyte quality and cytokine production. The demonstration of the type 1 receptor for IL-1 in mouse oocytes (Simon et al., 1994), the simulation by TNFα of the HCG effect on oocyte maturation in in-vitro perfused rabbit ovaries (Takehara et al., 1994) and the demonstration that TNFα can protect isolated mouse oocytes against spontaneous fragmentation (Sato et al., 1995) argue in favour of the former possibility. The present demonstration of the association of higher IL-1/ TNFα ratios with fertilization failure may thus reflect an impaired oocyte quality due to inadequate stimulation of TNFα production by IL-1. Clinical implications The association of high IL-1 and TNFα concentrations in follicular fluid with successful oocyte fertilization and post- fertilization development, as demonstrated in this study, suggests that these, and probably other, locally acting factors play an important role in determining oocyte viability and developmental competence. This action appears to be mediated by an enhancement of the access of circulating pituitary hormones, namely FSH, LH and GH, into early antral follicles, by changing the pattern of steroid secretion in late preovulatory follicles and, possibly, by directly influencing the oocyte throughout the antral follicle development. The local character of these actions may explain the well-known variability of ovarian response to the systemic administration of pituitary hormones. For instance, if the entry of circulating FSH and GH into early antral follicles is compromised by inadequate local secretion of vasoactive cytokines, the follicle can fail to proceed from androgen dominance to oestrogen dominance with all possible negative influences resulting from the high androgen-to-oestrogen ratio on oocyte quality (Tesarik and Mendoza, 1997). In these conditions, an early oocyte retrieval with subsequent oocyte in-vitro exposure to high oestradiol concentrations is likely to be more efficient in rescuing the endangered oocyte than any systemic hormonal treatment (Tesarik and Mendoza, 1995, 1997). This policy has been used with success to treat patients with anovulatory polycystic ovary syndrome (Trounson et al., 1994). It remains to be determined whether low concentrations of intrafollicular IL-1 and TNFα, similar to those described in this study in follicles yielding developmentally incompetent oocytes, are a dominant feature of patients with repeated IVF or ICSI failure due to a poor oocyte quality and, if so, whether this condition is at least partly caused by the lack of a direct stimulation of the oocyte by intrafollicular TNFα. If this is the case, early oocyte recovery with subsequent incubation with recombinant human TNFα may be of help. Acknowledgements The authors wish to thank Biomed S.L. and Ingelheim Diagnostica y Tecnologia S.A. for providing free samples of cytokine determination kits. References Adashi, E.Y. (1990) The potential relevance of cytokines to ovarian physiology: the emerging role of resident ovarian cells of the white blood cell series. Endocr. Rev., 11, Apa, R., Lanzone, A., Miceli, F. et al. (1994a) Growth hormone induction of rat granulosa cell tissue-plasminogen activator expression and progesterone synthesis. Mol. Cell. Endocrinol., 99, Apa, R., Lanzone, A., Miceli, F. et al. (1994b) Growth hormone induces in vitro maturation of follicle- and cumulus-enclosed rat oocytes. Mol. Cell. Endocrinol., 104, Apa, R., Caruso, A., Andreani, C.L. et al. (1996) Growth hormone stimulates androsterone synthesis by rat theca-interstitial cells. Mol. Cell. Endocrinol., 118, Barak, V., Mordel, N., Holzer, H. et al. (1992) The correlation of interleukin 1 and tumour necrosis factor to oestradiol, progesterone and testosterone levels in periovulatory follicular fluid of in-vitro fertilization patients. Hum. Reprod., 7, Best, C.L. and Hill, J.A. (1995) Interleukin-1α and -β modulation of luteinized human granulosa cell oestrogen and progesterone biosynthesis. Hum. Reprod., 10, Bili, H., Tarlatzis, B.C., Daniilidis, M. et al. (1998) Cytokines in the human ovary: presence in follicular fluid and correlation with leukotriene B 4. J. Assist. Reprod. Genet., 15,

7 C.Mendoza et al. Bomsel-Helmreich, O., Gougeon, A., Thébault, A. et al. (1979) Healthy and atretic human follicles in the preovulatory phase; differences in evolution of follicular morphology and steroid content of follicular fluid. J. Clin. Endocrinol. Metab., 48, Botero-Ruiz, W., Laufer, N., DeCherney, A.H. et al. (1984) The relationship between follicular fluid steroid concentration and successful fertilisation of human oocytes in vitro. Fertil. Steril., 41, Brailly, S., Gougeon, A., Milgrom, E. et al. (1981) Androgens and progestins in the human ovarian follicle: differences in the evolution of preovulatory, healthy nonovulatory, and atretic follicles. J. Clin. Endocrinol. Metab., 53, Branisteanu, I., Pijnenborg, R., Spiessens, C. et al. (1997) Detection of immunoreactive interleukin-11 in human follicular fluid: correlations with ovarian steroid, insulin-like growth factor I levels, and follicular maturity. Fertil. Steril., 67, Carson, R.S., Trounson, A.O. and Findlay, J.K. (1982) Successful fertilisation of human oocytes in vitro: concentration of estradiol-17β, progesterone and androstenedione in the antral fluid of donor follicles. J. Clin. Endocrinol. Metab., 55, Chabab, A., Hedon, B., Arnal, F. et al (1986) Follicular steroids in relation to oocyte development and human ovarian stimulation protocols. Hum. Reprod., 1, Cianci, A., Calogero, A.E., Palumbo, M.A. et al. (1996) Relationship between tumour necrosis factor α and sex steroid concentrations in the follicular fluid of women with immunological infertility. Hum. Reprod., 11, Devroey, P., Mannaerts, B., Smitz, J. et al. (1994) Clinical outcome of a pilot efficacy study on recombinant human FSH (Org 32489) combined with various GnRH agonist regimens. Hum. Reprod., 9, Dinarello, C.A. (1988) Biology of interleukin-1. FASEB J., 2, Edwards, R.G., Steptoe, P.C., Fowler, R.E. and Baillie, J. (1980) Observations on preovulatory human ovarian follicles and their aspirates. Br. J. Obstet. Gynaecol., 87, Elsworth, L.R., Balmaceda, J.P., Schenken, R.S. et al. (1984) Human chorionic gonadotrophin and steroid concentrations in human follicular fluid in relation to follicle size and oocyte maturity in stimulated ovarian cycles. Acta Eur. Fertil., 15, Enien, W.M., El Sahwy, S., Harris, C.P. et al. (1995) Human chorionic gonadotrophin and steroid concentrations in follicular fluid: the relationship to oocyte maturity and fertilization rates in stimulated and natural in-vitro fertilization cycles. Hum. Reprod., 10, Fishel, S.B., Edwards, R.G. and Walters, D.E. (1983) Follicular steroids as a prognosticator of successful fertilisation of human oocytes in vitro. J. Endocrinol., 99, Fowler, R.E., Edwards, R.G., Walters, D.E. et al. (1978) Steroidogenesis in preovulatory follicles of patients given human menopausal and chorionic gonadotrophins as judged by the radioimmunoassay of steroids in follicular fluid. J. Endocrinol., 77, Franchimont, P., Hazee-Hagelstein, M.-T., Hazout, A. et al. (1989) Correlation between follicular fluid content and the results of in vitro fertilization and embryo transfer. I. Sex steroids. Fertil. Steril., 52, Fukuoka, M., Yasuda, K., Fujiwara, H. et al. (1992) Interactions between interferon γ, tumour necrosis factor α, and interleukin-1 in modulating progesterone and oestradiol production by human luteinized granulosa cells in culture. Hum. Reprod., 7, Gougeon, A. (1996) Regulation of ovarian follicular development in primates: facts and hypotheses. Endocr. Rev., 17, Hartshorne, G.M. (1989) Preovulatory follicular fluid: relationships to ovarian stimulation protocol, fertilization and sperm penetration in vitro. Fertil. Steril., 52, Hsu, C.J. and Hammond, J.M. (1987) Concomitant effects of growth hormone on secretion of insulin-like growth factor I and progesterone by cultured porcine granulosa cells. Endocrinology, 121, Hurwitz, A., Loukides, J., Ricciarelli, E. et al Human intraovarian interleukin-1 (IL-1) system: highly compartmentalized and hormonally dependent regulation of the genes encoding IL-1, its receptor, and its receptor antagonist. J. Clin. Invest., 89, Hutchinson, L.A., Findlay, J.K and Herington, A.C. (1988) Growth hormone and insulin-like growth factor-i accelerate PMSG-induced differentiation of granulosa cells. Mol. Cell. Endocrinol., 55, Huyser, C., Fourie, F.L.R., Bosmans, E. and Levay, P.F. (1994) Interleukin- 1β, interleukin-6, and growth hormone levels in human follicular fluid. J. Assist. Reprod. Genet., 11, Jakimiuk, A.J., Weitsman, S.R., Brzechffa, P.R. and Magoffin, D.A. (1998) Aromatase mrna expression in individual follicles from polycystic ovaries. Mol. Hum. Reprod., 4, Jia, X., Kalmijn, J. and Hsueh, A.J.W. (1986) Growth hormone enhances follicle-stimulating hormone-induced differentiation of cultured rat granulosa cells. Endocrinology, 118, Karnitis, V.J., Townson, D.H., Friedman, C.I. and Danforth, D.R. (1994) Recombinant human follicle-stimulating hormone stimulates multiple follicular growth, but minimal estrogen production in gonadotropin-releasing hormone antagonist-treated monkeys: examining the role of luteinizing hormone in follicular development and steroidogenesis. J. Clin. Endocrinol. Metab., 79, Khan, S.A., Schmid, K., Hallin, P. et al. (1988) Human testis cytosol and ovarian follicular fluid contain high amounts of interleukin-1-like factor(s). Mol. Cell. Endocrinol., 58, Kreiner, D., Hung-Ching, L., Itskovitz, J. et al. (1987) Follicular fluid estradiol and progesterone are markers of preovulatory oocyte quality. Fertil. Steril., 48, Lanzone, A., Fortini, A., Fulghesu, A.M. et al. (1996) Growth hormone enhances estradiol production follicle-stimulating hormone-induced in the early stage of the follicular maturation. Fertil. Steril., 66, Leibovich, S.J., Ploverini, P.J., Shepard, H.M. et al. (1987) Macrophageinduced angiogenesis is mediated by tumor necrosis factor-alpha. Nature, 329, Lobo, R.A, dizerega, G.S. and Marrs, R.P. (1985) Follicular fluid steroid levels in dysmature and mature follicles from spontaneous and hyperstimulated cycles in normal and anovulatory women. J. Clin. Endocrinol. Metab., 60, Machelon, V. and Emilie, D. (1997) Production of ovarian cytokines and their role in ovulation in the mammalian ovary. Eur. Cytokine Netw., 8, Mason, H.D., Martikainen, H., Beard, R.W. et al. (1990) Direct gonadotrophic effect of growth hormone on oestradiol production by human granulosa cells in vitro. J. Endocrinol., 126, R1 R4. Mason, H.D., Willis, D.S., Beard, R.W. et al (1994) Estradiol production by granulosa cells of normal and polycystic ovaries: relationship to menstrual cycle history and concentrations of gonadotropins and sex steroids in follicular fluid. J. Clin. Endocrinol. Metab., 79, McNatty, K.P. and Baird, D.T. (1978) Relationship between follicle-stimulating hormone, androstenedione and oestradiol in human follicular fluid. J. Endocrinol., 76, McNatty, K.P., Hunter, W.M., McNeilly, A.C. and Sawers, R.S. (1975) Changes in the concentrations of pituitary and steroid hormones in the follicular fluid of human Graafian follicles throughout the menstrual cycle. J. Endocrinol., 64, McNatty, K.P., Smith, D.M., Makris, A. et al. (1979) The microenvironment of the human antral follicle: interrelationships among the steroid levels in antral fluid, the population of granulosa cells, and the status of the oocyte in vivo and in vitro. J. Clin. Endocrinol. Metab., 49, Nagy, Z.P., Liu, J., Joris, H. et al (1995) The result of intracytoplasmic sperm injection is not related to any of the three basic sperm parameters. Hum. Reprod., 10, Nahum, R., Thong, K.J. and Hillier, S.G. (1995) Metabolic regulation of androgen production by human thecal cells in vitro. Hum. Reprod., 10, Nayudu, P.L., Gook, D.A., Lopata, A. et al. (1987) Follicular characteristics associated with viable pregnancy after in vitro fertilization in humans. Gamete Res., 18, Rice, V.M., Williams, V.R., Limback, S.D. and Terranova, P.F. (1996) Tumour necrosis factor-α inhibits follicle stimulating hormone-induced granulosa cell oestradiol secretion in the human: dependence on size of follicle. Hum. Reprod., 11, Roby, K.F., Weed, J., Lyles, L. and Terranova, P.F. (1990) Immunological evidence for a human ovarian tumor necrosis factor. J. Clin. Endocrinol. Metab., 71, Sanyal, M.K., Berger, M.J., Thompson, I.E. et al. (1974) Development of Graafian follicles in adult human ovary. I. Correlation of estrogen and progesterone concentration in antral fluid with growth of follicles. J. Clin. Endocrinol. Metab., 38, Sato, E., Nakayama, T., Kamio, K. et al. (1995) Immunohistochemical localization and possible roles of tumor necrosis factor-α in mouse cumulusoocyte complexes. Develop. Growth Differ., 37, Seibel, M.M., Smith, D., Dlugi, A.M. and Levesque, L. (1989) Periovulatory follicular fluid hormone levels in spontaneous human cycles. J. Clin. Endocrinol. Metab., 68, Simon, C., Frances, A., Piquette, G. and Polan, M.L. (1994) Immunohistochemical localization of the interleukin-1 system in the mouse ovary during follicular growth, ovulation and luteinization. Biol. Reprod., 50,

8 Intrafollicular factors and ICSI results Takehara, Y., Dharmarajan, A.M., Kaufman, G. and Wallach, E.E. (1994) Effect of interleukin-1β on ovulation in the in vitro perfused rabbit ovary. Endocrinology, 134, Tao, M., Kodama, H., Kagabu, S. et al. (1997) Possible contribution of follicular interleukin-1β to nitric oxide generation in human pre-ovulatory follicles. Hum. Reprod., 12, Tavmergen, E., Tavmergen, E.N. and Capanoglu, R. (1992) Do analogues of gonadotrophin-releasing hormone influence follicular fluid steroids, oocyte maturity and fertilization rates? Hum. Reprod., 7, Tesarik, J. and Mendoza, C. (1995) Nongenomic effects of 17β-estradiol on maturing human oocytes: relationship to oocyte developmental potential. J. Clin. Endocrinol. Metab., 80, Tesarik, J. and Mendoza, C. (1997) Direct non-genomic effects of follicular steroids on maturing human oocytes: oestrogen versus androgen antagonism. Hum. Reprod. Update, 3, Tesarik, J. and Sousa, M. (1995) Key elements of a highly efficient intracytoplasmic sperm injection technique: Ca 2 fluxes and oocyte cytoplasmic dislocation. Fertil. Steril., 64, Trounson, A., Wood, C. and Kausche, A. (1994) In vitro maturation and the fertilization and developmental competence of oocytes recovered from untreated polycystic ovarian patients. Fertil. Steril., 62, Uehara, S., Naganuma, T., Tsuiki, A. et al. (1985) Relationship between follicular fluid steroid concentrations and in vitro fertilization. Obstet. Gynecol., 66, Watanabe, H., Nagai, K., Yamaguchi, M. et al. (1994) Concentration of interleukin-1β correlates with prostaglandin E2 and F2α in human preovulatory follicular fluid. Hum. Reprod., 9, Westergaard, L.G., Christensen, I.J. and McNatty, K.P. (1986) Steroid levels in ovarian follicular fluid related to follicle size and and health status during the normal menstrual cycle in women. Hum. Reprod., 1, Wramsby, H., Kullander, S., Liedholm, P. et al. (1981) The success rate of in-vitro fertilization of human oocytes in relation to the concentration of different hormones in follicular fluid and peripheral plasma. Fertil. Steril., 36, Yding Andersen, C. (1990) Levels of steroid-binding proteins and steroids in human preovulatory follicle fluid and serum as predictors of success in in vitro fertilization-embryo transfer treatment. J. Clin. Endocrinol. Metab., 71, Yding Andersen, C. (1993) Characteristics of human follicular fluid associated with successful conception after in vitro fertilization. J. Clin. Endocrinol. Metab., 77, Yding Andersen, C., Westergaard, L.G., Sinosich, M.J. and Byskov, A.G. (1992) Human preovulatory follicular fluid: inhibin and free steroids related to optimal follicular maturation in ovarian stimulation regimes and possible function in ovulation. Hum. Reprod., 7, Zolti, M., Meirom, R., Shemesh, M. et al. (1990) Granulosa cells as a source and target organ for tumor necrosis factor-α. FEBS Lett., 261, Received on July 27, 1998; accepted on November 17,