Insulin-like growth factor-II (IGF-II), IGF-binding protein-3 (IGFBP-3), and IGFBP-4 in follicular fluid are associated with oocyte maturation and embryo development

Insulin-like growth factor-II (IGF-II), IGF-binding protein-3 (IGFBP-3), and IGFBP-4 in follicular fluid are associated with oocyte maturation and embryo development Tzu-Hao Wang, M.D., Ph.D.,a,b Chia-Lin Chang, M.D.,a Hsien-Ming Wu, M.D.,a Ya-Ming Chiu, M.Sc.,a Chun-Kai Chen, M.D.,a and Hsin-Shih Wang, M.D., Ph.D.a,c a

Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital, Lin-Kou Medical Center, b College of Medicine, and c Graduate Institute of Clinical Medical Sciences, Chang Gung University, Taoyuan, Taiwan

Objective: The objective of this study was to investigate the association between follicular fluid (FF) levels of insulin-like growth factors (IGFs), IGF binding proteins (IGFBPs), and pregnancy-associated plasma protein-A (PAPP-A), which is a protease for IGFBP-4, and the quality of subsequent embryo development from in vitro fertilized oocytes aspirated from the same follicle. Design: Prospective study. Setting: University infertility clinic and academic research laboratory. Patient(s): One hundred sixty-two infertile women undergoing controlled ovarian hyperstimulation for IVF and embryo transfer were recruited in a university hospital. Intervention(s): During oocyte retrieval, samples of 225 FFs and matched mature oocytes were collected and studied. Main Outcome Measure(s): Concentrations of FF IGF-I, IGF-II, IGFBP-1, IGFBP-3, IGFBP-4, and PAPP-A were determined using ELISA. Progesterone secretion by cultured granulosa cells (GC) was measured by RIA. Result(s): Levels of IGF-II, IGFBP-3, and IGFBP-4 in FF on the day of oocyte retrieval were significantly correlated with embryo scores on day 3 (72 hours after oocyte retrieval). Levels of IGF-II, IGFBP-3, and IGFBP-4 in FF from follicles in which oocytes developed into day 2 embryos (48 hours after oocyte retrieval; 20 ⱖ embryo score ⱖ 6) after fertilization were significantly higher than those from follicles in which oocytes were unable to be fertilized and were arrested in embryo development on day 2, whereas the levels of PAPP-A were significantly lower in the former than the latter group. Using multiple regression analysis, we found that high levels of IGFBP-3 and IGFBP-4 combined with low levels of PAPP-A in FF were significantly correlated with successful fertilization and early development into day 2 embryos. In contrast, high FF IGFBP-1 and IGFBP-4 in combination with low FF IGF-I were significantly correlated with a later (day 2– day 3) embryo development. A significant stimulation of P secretion in cultured GCs by the combination of recombinant IGF-II, IGFBP-3, and IGFBP-4 further strengthened these proteins’ functional roles in promoting late follicular development. Conclusion(s): High IGF-II, IGFBP-3, IGFBP-4, and low PAPP-A levels in FF at the time of oocyte retrieval suggest better oocyte maturation and early embryo development (within 48 hours after oocyte retrieval), whereas high IGFBP-1, IGFBP-4, and low FF IGF-I levels may favor later embryo development (between 48 and 72 hours after oocyte retrieval). (Fertil Steril威 2006;86:1392–1401. ©2006 by American Society for Reproductive Medicine.) Key Words: Insulin-like growth factors, IGF-binding proteins, pregnancy-associated plasma protein-A, follicular fluid, granulosa cell, oocyte, embryo

Insulin-like growth factors (IGF-I and IGF-II), which are proinsulin-like small peptides with an approximate molecular mass of 7.5 kDa, are important mitogens that profoundly

Received November 25, 2005; revised and accepted March 29, 2006. Supported by the Chang Gung Memorial Hospital to T. H. Wang (CMRPG-1008) and to H. S. Wang (CMRPG-1011-II, CMRPG-32049-II), as well as the National Science Council, Taiwan, to T. H. Wang (NSC912314-B-182A-109,NSC92-2314-B-182A-054)andtoH.S.Wang(NSC922314-B-182-037, NSC93-2314-B-182-019). Reprint requests: Hsin-Shih Wang, M.D., Ph.D., Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital, Lin-Kou Medical Center, Taoyuan, Taiwan (FAX: 886-3-328-8252; E-mail: hswang@adm. cgmh.org.tw).

1392

affect cell growth, differentiation, and metabolism (1–3). In addition to the endocrine effects exerted by circulating IGFs, locally produced IGFs exert paracrine and autocrine effects on cell proliferation (4). The mitogenic effects of IGF are mediated through interactions with specific cell surface receptors, known as type I and type II IGF receptors (5). The most specific characteristic of IGFs, which distinguishes them from other growth factors, is that they are regulated and modulated by a heterogeneous group of binding proteins (IGFBPs) (6). These high-affinity IGFBPs compete with the IGF receptor for binding. To date, six binding proteins (IGFBP-1– IGFBP-6) have been identified and characterized. In biological

Fertility and Sterility姞 Vol. 86, No. 5, November 2006 Copyright ©2006 American Society for Reproductive Medicine, Published by Elsevier Inc.

0015-0282/06/$32.00 doi:10.1016/j.fertnstert.2006.03.064

fluids, IGFs are frequently bound to IGFBPs. In blood circulation, IGFBPs act as IGF carriers and prolong the half-life of IGFs (7). At the cellular level, IGFBPs function as modulators of IGF availability and activity (8, 9). In addition to modulating IGF bioactivity, IGFBPs are capable of some biological actions that are independent of their abilities to bind IGFs (10, 11). For instance, the direct association of IGFBPs with a variety of extracellular and cell surface molecules results in stimulation of bone cell proliferation and arrest of breast cancer cell growth (12, 13). In the human ovary, the IGF system, which includes IGFs, IGF receptors, IGFBPs, and IGFBP proteases, plays an important role in follicular growth, development, and atresia, as well as in steroidogenesis (14). Unlike IGF-I present in ovarian follicles of other species, IGF-II in the human ovarian follicle is the primary IGF acting as a mediator of FSH action, stimulating steroidogenesis in granulosa cells (GCs) (15, 16). In contrast, follicular fluid (FF) IGF-I in humans is believed to be from blood circulation rather than de novo production (17). In situ hybridization studies have shown that, in human dominant follicles, IGFBP-1 mRNA and IGFBP-4 mRNA are localized solely in GCs, whereas IGFBP-2 and IGFBP-3 mRNA are in both GC and theca cells (15, 18). The IGF-II bioavailability is shown to be inhibited by IGFBP-4 (19). In selected human follicles, IGFBP-4, an inhibitor of IGF-II action that is abundant within androgendominant follicles (15, 18), is proteolyzed by pregnancyassociated plasma protein-A (PAPP-A), a specific IGFBP-4 protease, resulting in a decrease in binding affinity of IGFBP-4 for IGF-II and an increase in IGF-II bioavailability (14, 20). In addition, PAPP-A is localized in GCs with an increased intensity as luteinization progresses (21). As with PAPP-A, IGFBP-4 is consistently expressed in healthy theca interna and in luteinized GCs under LH regulation (22). In late follicular phase, however, PAPP-A mRNA is abundant in the GCs of most follicles without apparent selectivity for the IGFBP-4-expressing follicles or for the dominant follicles (22). These observations indicate that IGFBP-4 or an IGFBP-4 proteolytic product may be involved in LHinduced steroidogenesis and luteinization. To date, the ovarian IGF/IGFBP system is shown to be associated with follicular growth, development, and atresia. However, the roles of the IGF/IGFBP system in FF and the events after ovulation, such as subsequent embryo development, have not been explored. In this study we simultaneously analyzed individual FF and matched oocyte outcomes to identify the relationship among the ovarian IGF/ IGFBP system, oocyte quality, and subsequent embryo development. MATERIALS AND METHODS Subjects of Ovulation Induction A total of 162 infertile women undergoing controlled ovarian hyperstimulation (COH) for IVF and embryo transfer Fertility and Sterility姞

were recruited in this study. Indications for IVF and embryo transfer included unexplained infertility, male factor infertility, endometriosis, ovulation disorders, and tubal absence or occlusion. Ages of patients ranged from 23– 40 years (mean, 33.7 years). Body mass indices, weight (in kilograms) divided by height squared (in meters), were 22.5 ⫾ 3.1 kg/m2 (range, 19.2–27.3 kg/m2). All patients gave written informed consent to participate in the study, and the experimental design and involvement of human subjects in this study were approved by the ethical committees of Chang Gung Memorial Hospital. The protocol for COH included pituitary down-regulation with SC injection of GnRH agonist (GnRH-a) (1 mg/d; Lupron, Abbott Laboratories, North Chicago, IL) that was started on day 21 of the previous menstrual cycle. Subcutaneous administration of recombinant hFSH (Gonal-F, Laboratories Serono S.A., Aubonne, Switzerland) at dosages from 150 – 450 IU/d was initiated on the third day of menstruation. The dosage was adjusted according to the women’s age and previous response of ovulation induction. Human chorionic gonadotropin (10,000 IU; Pregnyl, N.V. Organon, Oss, Netherlands) was applied IM 34 –36 hours before oocyte retrieval, and ovulation induction was monitored by both transvaginal sonography and serum E2 levels. Collection of FF and Granulosa Cells Follicular fluid without blood contamination and its matched oocyte from each single follicle were collected individually during oocyte retrieval. Follicle aspirates that were contaminated with blood were excluded. Immediately after retrieval, oocytes were evaluated for nuclear maturity and graded as metaphase II, metaphase I, or prophase I. The FF was stored at ⫺70°C until subsequent assays for IGF-I, IGF-II, IGFBP-1, IGFBP-3, IGFBP-4, and PAPP-A. For in vitro treatment assays, three sets of pooled GC were collected from randomly selected IVF patients with an average of 15 follicles (range, 10 –20 follicles) and cultured in M199 culture medium with 10% fetal bovine serum as previously reported (23). Pooled GC from each patient were seeded in 12-well plates at the density of 2.5 ⫻ 104 cells/well and cultured for 24 hours, serum starved in M199 for 16 hours, and treated with one of the following recombinant protein combinations at the indicated final concentrations: 400 ng/mL IGF-II (Sigma Chemicals, St. Louis, MO); 4 ␮g/mL IGFBP-3 (Sigma); 400 ng/mL IGFBP-4 (R & D Systems, Inc., Minneapolis, MN); 400 ng/mL IGF-II plus 400 ng/mL IGFBP-4; and all of three reagents. After a 24-hour treatment, culture media from each well was collected, diluted 100-fold with normal saline, and assayed with P RIA. We normalized the P production of each treatment with the total cell number in each well. Laboratory Evaluation of Embryo Oocytes were cultured in petri dishes in IVF-20 (Vitrolife, Göteborg, Sweden) at 37°C in a humidified 5% CO2/95% air 1393

environment. The semen was centrifuged through a 80% Percoll (Sigma) discontinuous gradient at 800 ⫻ g for 15 minutes. Insemination with 6,000 –10,000 progressively motile spermatozoa for each oocyte was performed 4 – 6 hours after oocyte pick-up. After IVF, the resulting embryos were cultured in IVF-20 at 37°C under 5% CO2 in air. Quality of embryo was assessed by the embryo score described previously (24). Briefly, a quality score of each embryo on days 2 and 3 (48 and 72 hours after oocyte retrieval) was derived from the morphological grade of the embryo multiplied by the number of blastomeres. Morphologically, embryos were graded as follows: grade 4, equalsized symmetrical blastomeres; grade 3, uneven blastomeres with ⬍10% extracellular fragmentation; grade 2, uneven blastomeres with 10%–50% extracellular fragmentation; grade 1, ⬎50% blastomeric fragmentation with uneven blastomeres. ELISA for IGF-I, IGF-II, IGFBP-1, IGFBP-3, IGFBP-4, PAPP-A, and RIA for P Human IGF-I levels in FF were measured using a commercially available ELISA kit (Quantikine human IGF-I Immunoassay, R & D Systems). Before measurement, samples of FF were pretreated and diluted to 1:100 with pretreatment buffers enclosed in the kit. The minimum detection limit of the assay was 0.094 ng/mL. The intraassay coefficients of variation (CVs) were 3.5% at 0.5 ng/mL, 4.3% at 1.2 ng/mL, and 4.3% at 2.4 ng/mL (n ⫽ 8). The interassay CVs were 8.1% at 0.4 ng/mL, 8.3% at 1.1 ng/mL, and 7.5% at 2.3 ng/mL (n ⫽ 6). Human IGF-II levels in FF were determined using an ELISA kit (Active IGF-II; Diagnostic Systems Laboratories, Inc., Webster, TX). The minimum detection limit of the assay was 50 ng/mL. The intraassay CVs were 7.4% at 281 ng/mL, 4.7% at 423 ng/mL, and 3.6% at 685 ng/mL (n ⫽ 10). The interassay CVs were 10.1% at 295 ng/mL, 8.6% at 446 ng/mL, and 8.8% at 735 ng/mL (n ⫽ 6). Human IGFBP-1 levels in FF were assayed using a commercial ELISA kit (DuoSet ELISA Development System for human IGFBP-1; R & D Systems). Before assay, samples of FF were diluted to 1:50 with phosphate-buffered saline (PBS). The minimum detection limit of the assay was 62.5 pg/mL. The intraassay CVs were 5.8% at 550 pg/mL and 6.3% at 1,350 pg/mL (n ⫽ 12). The interassay CVs were 7.9% at 535 pg/mL and 9.2% at 1420 pg/mL (n ⫽ 10). Human IGFBP-3 levels in FF were measured using a commercial ELISA kit (Quantikine human IGFBP-3 Immunoassay; R & D Systems). Samples of FF were diluted to 1:100 in the calibrator diluent buffer contained in the kit before measurement. The minimum detection limit of the assay was 0.78 ng/mL. The intraassay CVs were 4.8% at 4.1 ng/mL, 5.0% at 12.7 ng/mL, and 2.3% at 31.1 ng/mL (n ⫽ 10). The interassay CVs were 5.4% at 4.1 ng/mL, 6.4% at 11.8 ng/mL, and 8.0% at 28.5 ng/mL (n ⫽ 6). 1394

Wang et al.

IGFs/IGFBPs and embryo development

Human IGFBP-4 levels in FF were assayed using a commercial ELISA kit (DuoSet ELISA Development System for human IGFBP-4; R & D Systems). Before measurement, FF samples were diluted to 1:50 in PBS. The minimum detection limit of the assay was 250 pg/mL. The intraassay CVs were 6.9% at 650 pg/mL and 7.1% at 1,650 pg/mL (n ⫽ 12). The interassay CVs were 8.2% at 670 pg/mL and 8.6% at 1,730 pg/mL (n ⫽ 10). Human PAPP-A levels in FF were determined using a commercial ELISA kit (PAPP-A ELISA; Immuno-Biological Laboratories, Hamburg, Germany). The minimum detection limit of the assay was 1 ␮g/mL. The intraassay CVs were 4.3% at 2.8 ␮g/mL, 6.1% at 5.9 ␮g/mL, and 5.7% at 9.2 ␮g/mL (n ⫽ 10). The interassay CVs were 6.5% at 2.9 ␮g/mL, 8.3% at 5.6 ␮g/mL, and 8.9% at 9.4 ␮g/mL (n ⫽ 8). Progesterone levels were determined by radioimmunoassay using an I125-based RIA kit (Diagnostic Systems Laboratories). The intraassay CVs were 6.4% at 0.77 ng/mL, 3.3% at 5.75 ng/mL, and 5.6% at 14.55 ng/mL (n ⫽ 12). The interassay CVs were 2.4% at 0.81 ng/mL, 1.7% at 1.17 ng/mL, and 3.3% at 11.4 ng/mL (n ⫽ 10). Western Blotting Analysis By the Bradford assay, protein concentrations of FF ranged from 60 –90 mg/mL (mean 70.9, SD 9.9, from 10 random samples). Sixty micrograms of protein from pooled FF (n ⫽ 5) was electrophoresed on 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) under reducing conditions and was transferred onto a nitrocellulose membrane (pore size 0.45 ␮m; Amersham Bioscience, Amersham, UK). After the transfer, the nitrocellulose membranes were incubated with 5% (w/v) nonfat milk in TBST (5 mM Tris and 0.9% NaCl at pH 7.5, 0.1% Tween-20) at room temperature for 1 hour. The membranes were then treated three times with Tris buffer saline containing 0.1% (v/v) Tween-20 (10 minutes each) at room temperature. The membranes were incubated with an anti-IGFBP-4 antibody (1:500, goat antihuman polyclonal antibody; Santa Cruz Biotechnology, Santa Cruz, CA) in TBST containing 3% BSA for 2 hours at room temperature. At the end of incubation, the membranes were washed three times with TBST (10 minutes each) and incubated in a secondary antibody (donkey antigoat IgG conjugated with horseradish peroxidase; Santa Cruz Biotechnology) diluted in TBST containing 3% bovine serum albumin (BSA) (1:1,500) for 2 hours at room temperature. The membranes were then washed three times (each for 10 minutes) with TBST and detected using an enhanced chemiluminescent kit (ECL Western Blotting Analysis System RPN 2109; Amersham Bioscience, Amersham). The membranes were stripped and reprobed stepwise with anti-␤-actin (1:500, mouse antihuman polyclonal antibody; Santa Cruz Biotechnology) and a secondary antibody (1:1,500, goat antimouse IgG conjugated with horseradish peroxidase; Santa Cruz Biotechnology) to verify that the amount of protein was the same in each lane. Vol. 86, No. 5, November 2006

Fertility and Sterility姞

TABLE 1 Association between FF levels of IGFs (IGF-I and IGF-II), IGFBPs (IGFBP-1, IGFBP-3, and IGFBP-4), PAPP-A (a protease for IGFBP-4), and the embryo development evaluated on days 2 and 3 (48 and 72 hours after oocyte retrieval). Embryo development Nonfertilized oocytes and arrested embryos on day 2 Nonfertilized oocytes Embryos of developmental arrest on day 2 Embryos of early development up to day 2 (20 ⱖ D2 score ⱖ 6) Embryos of developmental arrest and impending arrest on day 3 Embryos of ongoing development from day 2 to day 3 (40 ⱖ D3 score ⱖ 10) Note: Values are means ⫾ SD. a Student’s t-test: P⬍.003. b Student’s t-test: P⬍.00001. c Student’s t-test: P⬍.00001. d Student’s t-test: P⬍.003. e Student’s t-test: P⬍.0002. f Student’s t-test: P⬍.00005. g Student’s t-test: P⬍.002. Wang. IGFs/IGFBPs and embryo development. Fertil Steril 2006.

No.

IGF-I (ng/mL)

IGF-II (ng/mL)

IGFBP-1 (ng/mL)

IGFBP-3 (ng/mL)

IGFBP-4 (ng/mL)

PAPP-A (␮g/mL)

58

106.1 ⫾ 32.7

76.2 ⫾ 13.2a

62.4 ⫾ 40.7

1,366.5 ⫾ 418.6b

51.5 ⫾ 47.5c

8.2 ⫾ 2.9d

46 12 167

102.8 ⫾ 33.9 118.7 ⫾ 24.9 110.9 ⫾ 29.0

75.1 ⫾ 13.5 80.7 ⫾ 11.4 82.9 ⫾ 14.6a

61.7 ⫾ 41.6 65.4 ⫾ 38.4 66.2 ⫾ 39.1

1,319.3 ⫾ 402.4 1,546.5 ⫾ 448.1 1,736.4 ⫾ 546.1b

57.7 ⫾ 49.3 27.9 ⫾ 31.0 82.1 ⫾ 41.1c

7.9 ⫾ 2.3 9.3 ⫾ 4.4 6.9 ⫾ 2.8d

47

122.1 ⫾ 33.1g

80.7 ⫾ 15.0

48.3 ⫾ 23.5e

1,612.5 ⫾ 597.2

61.7 ⫾ 39.6f

7.4 ⫾ 4.2

120

106.5 ⫾ 26.2g

83.7 ⫾ 14.4

73.2 ⫾ 41.7e

1,784.9 ⫾ 519.3

90.2 ⫾ 39.0f

6.7 ⫾ 1.9

1395

Statistical Analysis Linear regression was used to explore the relationships among the FF levels of IGF-I, IGF-II, IGFBP-1, IGFBP-3, IGFBP-4, and PAPP-A, as well as embryo scores on day 3 (72 hours after oocyte retrieval). A Student’s t-test was used to compare the levels of IGFs and IGFBPs in FF between the groups according to embryo scores on days 2 and 3. Multiple regression analysis using the Statistical Analysis System (SAS) software (version 8.1 for windows; SAS Institute Inc., Cary, NC) was used to identify the effects of each variable in a combination of IGFs, IGFBPs, and PAPP-A in FF on the outcome of embryo development. Values of P⬍.05 were considered statistically significant. RESULTS A total of 225 FF samples and the corresponding oocytes with mature nuclei at metaphase II were collected. After IVF, 179 became fertilized with two pronuclei observed (fertilization rate 80%), whereas 46 were not. Thereafter, 12 fertilized oocytes failed to develop into day 2 embryos (48 hours after oocyte retrieval) and another 47 embryos had

arrested growth or impending arrest of embryo development between days 2 and 3 (48 –72 hours after oocyte retrieval), resulting in a final count of 167 and 120 embryos studied on days 2 and 3, respectively (Table 1). The FF levels of IGF-II, IGFBP-3, and IGFBP-4 from the follicles in which oocytes developed into day 2 embryos after fertilization (20 ⱖ day 2 embryo score ⱖ 6, n ⫽ 167) were significantly higher than those from the follicles in which oocytes were unable to be fertilized and demonstrated developmental arrest of embryos on day 2 (n ⫽ 58) (P⬍.003, P⬍.00001, and P⬍.00001, respectively) (Table 1). In contrast, the levels of PAPP-A in FF from follicles in which fertilized oocytes developed to day 2 embryos (20 ⱖ day 2 embryo score ⱖ 6, n ⫽ 167) were significantly lower than those from follicles in which oocytes were unable to be fertilized and demonstrated developmental arrest of embryos on day 2 (n ⫽ 58) (P⬍.003) (Table 1). In accordance with the favorable role of IGF-II, IGFBP-3, and IGFBP-4 for embryo development, the FF levels of these proteins on the day of oocyte retrieval were positively correlated with embryo scores on day 3 (72 hours after oocyte

FIGURE 1 The correlation between embryo score on day 3 (72 hours after oocyte retrieval) and levels of IGFs and IGFBPs in follicular fluid (n ⫽ 120): (A) IGF-II, (B) IGFBP-3, (C) IGFBP-4, and (D) PAPP-A. d.f. ⫽ degree of freedom.

Wang. IGFs/IGFBPs and embryo development. Fertil Steril 2006.

1396

Wang et al.

IGFs/IGFBPs and embryo development

Vol. 86, No. 5, November 2006

FIGURE 2 The relationship between FF levels of (A) IGF-II and IGFBP-4, and (B) PAPP-A and IGFBP-4, analyzed by linear regression (n ⫽ 120). d.f. ⫽ degree of freedom.

PAPP-A) on the outcome of embryo development, we further identified high levels of IGFBP-3, high levels of IGFBP-4, and low levels of PAPP-A in FF to be significantly correlated with a better outcome of day 2 embryos (P⫽.008, P⫽.0008, and P⫽.013, respectively). In the group of oocytes that were successfully fertilized and that subsequently developed into day 2 embryos (20 ⱖ day 2 embryo score ⱖ 6, n ⫽ 167), the levels of IGFBP-1 and IGFBP-4 from follicles where day 2 embryos continued to develop into day 3 embryos (40 ⱖ day 3 embryo score ⱖ 10, n ⫽ 120) were significantly higher than those from follicles in which embryos were unable to grow further after day 2 (day 3 embryo score ⬍10) (n ⫽ 47) (P⬍.0002, and P⬍.00005, respectively) (Table 1). On the contrary, IGF-I levels were significantly (P⬍.002) lower in the follicles from which corresponding embryos developed better from day 2– day 3 (n ⫽ 120) than those in the follicles resulting in poorer outcomes (n ⫽ 47). Multiple regression analysis further confirmed high FF levels of IGFBP-1 and IGFBP-4 in combination with low FF IGF-I to be significantly correlated with successful embryo development from day 2– day 3 (P⫽.018, P⫽.003, and P⫽.001, respectively). In vitro treatment of GCs with IGF-II, IGFBP-3, IGFBP-4, or combinations of these reagents also confirmed their functional roles in FF. Compared with the basal P secretion at 24.3 ⫾ 10.0 pg/cell (mean ⫾ SD), treatment with each of

FIGURE 3

Wang. IGFs/IGFBPs and embryo development. Fertil Steril 2006.

retrieval; P⬍.001, P⬍.005, and P⬍.001, respectively) (Fig. 1A–C). On the contrary, the FF levels of IGF-I and IGFBP-1 were not significantly correlated with day 3 embryo scores (data not shown). There was a positive association between the FF levels of IGF-II and IGFBP-4 (P⬍.0005) (Fig. 2A). Although no significant correlation was detected between the FF levels of PAPP-A and embryo scores on day 3 (Fig. 1D), the FF levels of PAPP-A exhibited a significantly negative correlation with FF levels of IGFBP-4 (P⬍.005) (Fig. 2B).

Western blot analysis of pooled FF collected from follicles where oocytes were fertilized and developed into embryos using anti-IGFBP-4 (left) and anti-␤-actin (right) antibodies. Lane 1, embryo score on day 2 (48 hours after oocyte retrieval) ⱖ8, representing good embryo development; lane 2, embryo score on day 2 ⱕ2, indicating an arrest of embryo development. Shown herein is a representative figure from three independent experiments with similar results.

To identify whether the IGFBP-4 measured with ELISA was the 32-kDa intact form or the 16.5-kDa proteolyzed form, we applied Western blot analysis and determined the IGFBP-4 in FF to be the 32-kDa intact form (Fig. 3). Consistent with the ELISA results (Table 1), the concentration of IGFBP-4 in pooled FF from follicles in which oocytes developed into day 2 embryos after fertilization (Fig. 3, lane 1) was greater than that from follicles with nonfertilized oocytes (Fig. 3, lane 2). Using multiple regression analysis to dissect the effect of each individual variable (IGF-II, IGFBP-3, IGFBP-4, and Fertility and Sterility姞

Wang. IGFs/IGFBPs and embryo development. Fertil Steril 2006.

1397

FIGURE 4 Stimulation of P secretion from pooled granulosa cells by in vitro treatment with recombinant IGF-II, IGFBP-3, IGFBP-4, or indicated combinations for 24 hours. Data shown are means ⫾ SD from three independent experiments. Progesterone secretion in each treatment is normalized with the cell number. Abbreviations: II ⫽ IGF-II, 3 ⫽ IGFBP-3, 4 ⫽ IGFBP-4. *P⫽.013.

Wang. IGFs/IGFBPs and embryo development. Fertil Steril 2006.

IGF-II, IGFBP-3, and IGFBP-4 increased P secretion to 52.1 ⫾ 34.6 pg/cell (P⫽.18), 50.5 ⫾ 35.2 pg/cell (P⫽.19), and 34.4 ⫾ 7.6 pg/cell (P⫽.24), respectively (n ⫽ 3) (Fig. 4). The combined treatment with IGF-II and IGFBP-4 increased P secretion to 44.3 ⫾ 27.3 pg/cell (P⫽.3), whereas the triple treatment of IGF-II, IGFBP-3, and IGFBP-4 significantly increased P secretion to 72.0 ⫾ 16.4 pg/cell (P⫽.013) (Fig. 4). In three independent experiments, the 16-hour serum starvation and treatment of the aforementioned combination of IGF-II and IGFBPs did not result in any recognizable patterns of an increase or decrease of cell number when compared to the cell number in the control wells (data not shown). Therefore, the results of P production were not affected by the changes of cell number in each well. DISCUSSION Results of this study demonstrate, for the first time, a significant association between FF IGF and IGFBP levels and embryo development (Fig. 1A–C). In addition, FF concentrations of IGFs and IGFBPs (IGF-II, IGFBP-3, and IGFBP-4) in follicles with a better early outcome, that is, oocytes that developed into day 2 embryos after fertilization (20 ⱖ day 2 embryo score ⱖ 6), were significantly higher than those from follicles with poor early outcome, that is, oocytes were unable to be fertilized and were arrested in embryo development on day 2 (Table 1). These observations suggest a close relationship between intraovarian IGF and IGFBP levels and embryo quality. The autocrine and paracrine roles of IGFs and their receptors are involved in the pathophysiology of follicle recruit1398

Wang et al.

IGFs/IGFBPs and embryo development

ment, oocyte maturation, and, potentially, embryo development (25). The concentrations of IGF-II (but not IGF-I) are significantly higher in ovarian venous effluents than that of the peripheral circulation (26). In addition, levels of FF IGF-II, but not IGF-I, are found to be positively associated with follicular development (27). Moreover, in vitro studies have shown that IGF-II increases the release of E2 from cultured human GCs in a dose-dependent manner (28). In contrast, FF levels of IGF-I at the preovulatory stage are considerably lower than serum levels, indicating that IGF-I in FF is probably derived from peripheral circulation diffusion (17, 29). Although IGF-I mRNA is not detected in GCs of the mature human ovary, IGF-I receptor mRNA is abundant in GCs of both Graafian and atretic follicles (15, 16). Collectively, development of human GCs may be regulated by IGFs (IGF-II synthesized de novo and IGF-I from the circulation) through IGF receptors (2). In human oocytes and preimplantation embryos, transcripts for IGF-II, IGF-I receptor, IGF-II receptor, and insulin receptor are detected (30). High gene expressions of IGF-II, IGF-I receptor, IGF-II receptor, and insulin receptor are associated with high growth potential in day 3– day 6 human embryos (shortly before implantation) (31). These results suggest that human embryonic development at the preimplantation stage is regulated by IGFs from both embryonic (IGF-II) and maternal (insulin and IGF-I) sources (30). In the present study, concentrations of IGF-II in FF from follicles in which oocytes successfully developed into day 2 embryos were significantly higher than those from follicles in which oocytes were unable to be fertilized or arrested in embryo development on day 2 (Table 1). In addition, levels of IGF-II in FF on the day of oocyte retrieval were significantly correlated with embryo scores on day 3 (Fig. 1A). On the other hand, in the group of embryos that survived to day 2, the 47 that were in arrest or impending arrest on day 3 exhibited significantly higher FF IGF-I levels than those successfully developed into day 3 embryos (n ⫽ 120) (Table 1). Multiple regression analysis also verified high FF IGF-I to be significantly associated with failed embryo development from day 2– day 3 (P⫽.018). Collectively, IGF-II in FF reflects a better probability of oocytes for successful fertilization and its early embryo development, whereas FF IGF-I may play a negative role on later embryo development. Because human oocytes and preimplantation embryos express IGF-II, IGF-I receptor, and IGF-II receptor (30), it is appealing for us to speculate that high FF IGF-II might directly promote oocyte maturation through IGF receptors and stimulate the subsequent de novo production of IGF-II in preimplantation embryos through autocrine regulation. In combination with LH, IGF-I enhances androgen biosynthesis in theca interstitial cells through up-regulation of LH receptors and an augmentation of LH-stimulated cAMP (32, 33). In the ovary during its follicular phase, IGFBP-1 is believed to inhibit IGF-I-induced androgen production by theca cells and in turn inhibit/prevent follicular atresia and Vol. 86, No. 5, November 2006

FIGURE 5 The effect of the FF levels of IGFs (IGF-I and IGF-II), IGFBPs (IGFBP-1, IGFBP-3, and IGFBP-4), and PAPP-A on the outcomes of oocyte fertilization and subsequent embryo development.

Wang. IGFs/IGFBPs and embryo development. Fertil Steril 2006.

anovulation by enhancing E2 formation in GCs (a shift from androgen-dominant to estrogen-dominant follicles) (34). If IGFBP-1 produced by GCs is insufficient to antagonize the IGF-I bioactivity in theca stromal cells, the consequent overproduction of androgens may cause a defective follicular maturation (2). Collectively, high IGFBP-1 and low IGF-I in the ovary may play an important role in follicular growth and maturation. This condition may also favor early embryo development, as we have observed high IGFBP-1 levels and low IGF-I levels in follicles in which oocytes developed into good embryos on day 3 (Table 1 and Fig. 5). The IGFBP-3 exerts dual functions in modulating IGF action in vivo. Soluble IGFBP-3 inhibits IGF-I action by sequestering and preventing IGF-I receptor binding, whereas surface-associated IGFBP-3 enhances the growth-promoting effects of IGF-I in bovine fibroblasts (35). In the human ovary, IGFBP-3 mirrors the increase of IGF-II in estrogen (E)-dominant follicles and acts as a regulator of IGF-II action within the E-dominant follicle (27). Our in vitro treatment experiments also demonstrated that IGFBP-3 reversed the inhibitory effect of IGFBP-4 on the IGF-II stimulation of GCs (Fig. 4). Collectively, these results suggest that the increased IGFBP-3 levels in FF may compete with IGFBP-4 and preserve the bioavailability and bioactivity of Fertility and Sterility姞

IGF-II in stimulating luteal formation, late oocyte maturation, and embryo development. Decreased levels of IGFBP-4 and increased levels of its protease PAPP-A (20) have been associated with the selection of dominant follicles (36 –38). The protease effect of PAPP-A on IGFBP-4 has been shown by an increase in levels of fragmented IGFBP-4 (16.5 kDa) in dominant follicles (39). However, at later follicular maturation, the role of a dynamic change between increased PAPP-A and decreased IGFBP-4 may be diminished, as demonstrated by the report that the ovarian IGFBP-4 mRNA levels are markedly increased after treatment with hCG when IGF-II and PAPP-A mRNAs are not significantly altered (22). In addition, our Western blot analysis detected only an intact form of IGFBP-4 in the FF collected in this study (Fig. 3). This is in agreement with the detection of low PAPP-A levels in dominant follicles with good quality oocytes (Table 1). Thus, instead of the promoting role in early follicular growth as previously suggested (39, 40), proteolysis of IGFBP-4 by PAPP-A may only play a minor role, if any, in oocyte maturation and subsequent embryo development. Our results suggest that low FF concentrations of PAPP-A and high levels of IGF-II, IGFBP-3, and IGFBP-4 in ovarian follicles may be used for predicting which oocytes would be 1399

successfully fertilized and develop into the day 2 embryos (Table 1). In support of this proposal, the use of high levels of FF IGFBP-3 and IGFBP-4, as well as low levels of FF PAPP-A in multiple regression analysis, could identify the majority of fertilized oocytes and embryo development up to day 2 (Fig. 5). Treatment with the combination of high IGF-II, IGFBP-3, and IGFBP-4 activated the functions of GCs, as exemplified by the increased secretion of P (Fig. 4), further strengthening the promoting roles of that specific combination in late follicular maturation, sufficient luteal function, and subsequent embryo development. Results from the present study raise the possibility that FF IGFs and IGFBPs exert their biological functions synergetically but differentially at distinct stages of follicular development. Most significant changes in the FF IGF/IGFBP system may be the high PAPP-A and low IGFBP-4 during the selection of dominant follicles (20, 35–37), which shift to low PAPP-A and high IGFBP-4 in the late and more mature follicles shortly before oocyte maturation and ovulation (Table 1). We speculate that, to modulate human oocyte maturation and later embryo development, FF IGF-II may act only under the regulation or assistance of various IGFBPs (e.g., higher levels of FF IGFBP-3 and IGFBP-4 in the present study) (Fig. 5). In contrast, FF IGF-I may exhibit an inhibitory action on later embryo development, especially when FF IGFBP-1 levels are insufficient to counteract the detrimental effects of IGF-I on the overproduction of androgens in the follicle (2). In conclusion, the combination of high IGF-II, high IGFBP-3, high IGFBP-4, and low PAPP-A levels in FF at the time of oocyte retrieval correlates with better oocyte maturation and early embryo development (within 48 hours after oocyte retrieval). In addition, high IGFBP-1, high IGFBP-4, and low IGF-I levels in FF may be favorable factors in later embryo development between 48 and 72 hours after oocyte retrieval. Acknowledgment: The authors thank Chee-Jen Chang, Ph.D., and Hsin-Yee Chen, B.Sc., for statistical advice; Yung-Kuei Soong, M.D., Hong-Yuan Huang, M.D., Chia-Wei Wang, M.D., and Chi-Long Lee, M.D., for providing clinical samples; Mei-Li Wang, B.Sc., Chieh-Yu Lin, M.Sc., and Ming-Li Wei, B.Sc., for their excellent technical assistance; and Yinin Hu, B.Sc. (Northwestern University, Evanston, IL), for English editing.

REFERENCES 1. Wang HS, Chard T. The role of insulin-like growth factor-I and insulin-like growth factor-binding protein-1 in the control of human fetal growth. J Endocrinol 1992;132:11–9. 2. Wang HS, Chard T. IGFs and IGF-binding proteins in the regulation of human ovarian and endometrial function. J Endocrinol 1999;161:1–13. 3. Hwa V, Oh Y, Rosenfeld RG, Rajaram S, Baylink DJ, Mohan S. The insulin-like growth factor-binding protein (IGFBP) superfamily. Insulinlike growth factor-binding proteins in serum and other biological fluids: regulation and functions. Endocrinol Rev 1999;20:761– 87. 4. Jones JI, Clemmons DR. Insulin-like growth factors and their binding proteins: biological actions. Endocrinol Rev 1995;16:3–34. 5. Nissley P, Lopaczynski W. Insulin-like growth factor receptors. Growth Factors 1991;5:29 – 43.

1400

Wang et al.

IGFs/IGFBPs and embryo development

6. Cataldo NA. Insulin-like growth factor binding proteins: do they play a role in polycystic ovary syndrome? Semin Reprod Endocrinol 1997; 15:123–36. 7. Binoux M, Roghani M, Hossenlopp P, Hardouin S, Gourmelen M. Molecular forms of human IGF binding proteins: physiological implications. Acta Endocrinol (Copenh) 1991;124:41–7. 8. Busby WH Jr, Klapper DG, Clemmons DR. Purification of a 31,000dalton insulin-like growth factor binding protein from human amniotic fluid. Isolation of two forms with different biologic actions. J Biol Chem 1988;263:14203–10. 9. Bar RS, Booth BA, Boes M, Dake BL. Insulin-like growth factorbinding proteins from vascular endothelial cells: purification, characterization, and intrinsic biological activities. Endocrinology 1989;125: 1910 –20. 10. Jones JI, Gockerman A, Busby WH Jr, Wright G, Clemmons DR. Insulin-like growth factor binding protein 1 stimulates cell migration and binds to the alpha 5 beta 1 integrin by means of its Arg-Gly-Asp sequence. Proc Natl Acad Sci USA 1993;90:10553–7. 11. Oh Y, Muller HL, Pham H, Rosenfeld RG, Lamson G. Demonstration of receptors for insulin-like growth factor binding protein-3 on Hs578T human breast cancer cells. J Biol Chem 1993;268:26045– 8. 12. Mohan S, Nakao Y, Honda Y, Landale E, Leser U, Dony C, et al. Studies on the mechanisms by which insulin-like growth factor (IGF) binding protein-4 (IGFBP-4) and IGFBP-5 modulate IGF actions in bone cells. J Biol Chem 1995;270:20424 –31. 13. Gill ZP, Perks CM, Newcomb PV, Holly JM. Insulin-like growth factor-binding protein (IGFBP-3) predisposes breast cancer cells to programmed cell death in a non-IGF-dependent manner. J Biol Chem 1997;272:25602–7. 14. Giudice LC. Insulin-like growth factor family in Graafian follicle development and function. J Soc Gynecol Investig 2001;8:S26 –9. 15. Zhou J, Bondy C. Anatomy of the human ovarian insulin-like growth factor system. Biol Reprod 1993;48:467– 82. 16. el-Roeiy A, Chen X, Roberts VJ, LeRoith D, Roberts CT Jr, Yen SS. Expression of insulin-like growth factor-I (IGF-I) and IGF-II and the IGF-I, IGF-II, and insulin receptor genes and localization of the gene products in the human ovary. J Clin Endocrinol Metab 1993;77:1411– 8. 17. Chang SY, Hsieh KC, Wang HS, Soong YK. Follicular fluid levels of insulin-like growth factor I, insulin-like growth factor binding protein I, and ovarian steroids collected during ovum pick-up. Fertil Steril 1994; 62:1162–7. 18. el-Roeiy A, Chen X, Roberts VJ, Shimasakai S, Ling N, LeRoith D, et al. Expression of the genes encoding the insulin-like growth factors (IGF-I and II), the IGF and insulin receptors, and IGF-binding proteins1-6 and the localization of their gene products in normal and polycystic ovary syndrome ovaries. J Clin Endocrinol Metab 1994;78:1488 –96. 19. Zhou R, Diehl D, Hoeflich A, Lahm H, Wolf E. IGF-binding protein-4: biochemical characteristics and functional consequences. J Endocrinol 2003;178:177–93. 20. Conover CA, Oxvig C, Overgaard MT, Christiansen M, Giudice LC. Evidence that the insulin-like growth factor binding protein-4 protease in human ovarian follicular fluid is pregnancy associated plasma protein-A. J Clin Endocrinol Metab 1999;84:4742–5. 21. Rhoton-Vlasak A, Gleich GJ, Bischof P, Chegini N. Localization and cellular distribution of pregnancy-associated plasma protein-a and major basic protein in human ovary and corpora lutea throughout the menstrual cycle. Fertil Steril 2003;79:1149 –53. 22. Zhou J, Wang J, Penny D, Monget P, Arraztoa JA, Fogelson LJ, et al. Insulin-like growth factor binding protein 4 expression parallels luteinizing hormone receptor expression and follicular luteinization in the primate ovary. Biol Reprod 2003;69:22–9. 23. Wang TH, Horng SG, Chang CL, Wu HM, Tsai YJ, Wang HS, et al. Human chorionic gonadotropin-induced ovarian hyperstimulation syndrome is associated with up-regulation of vascular endothelial growth factor. J Clin Endocrinol Metab 2002;87:3300 – 8. 24. Chang CL, Wang TH, Horng SG, Wu HM, Wang HS, Soong YK. The concentration of inhibin B in follicular fluid: relation to oocyte maturation and embryo development. Hum Reprod 2002;17:1724 – 8.

Vol. 86, No. 5, November 2006

25. Hernandez ER. Regulation of the genes for insulin-like growth factor (IGF) I and II and their receptors by steroids and gonadotropins in the ovary. J Steroid Biochem Mol Biol 1995;53:219 –21. 26. Jesionowska H, Hemmings R, Guyda HJ, Posner BI. Determination of insulin and insulin-like growth factors in the ovarian circulation. Fertil Steril 1990;53:88 –91. 27. van Dessel HJT, Chandrasekher Y, Yap OW, Lee PD, Hintz RL, Faessen GH, et al. Serum and follicular fluid levels of insulin-like growth factor I (IGF-I), IGF-II, and IGF-binding protein-1 and -3 during the normal menstrual cycle. J Clin Endocrinol Metab 1996;81: 1224 –31. 28. Kubota T, Kamada S, Ohara M, Taguchi M, Sakamoto S, Shimizu Y, et al. Insulin-like growth factor II in follicular fluid of the patients with in vitro fertilization and embryo transfer. Fertil Steril 1993;59:844 –9. 29. Pellegrini S, Fuzzi B, Pratesi S, Mannelli M, Criscuoli L, Messeri G, et al. In-vivo studies on ovarian insulin-like growth factor I concentrations in human preovulatory follicles and human ovarian circulation. Hum Reprod 1995;10:1341–5. 30. Lighten AD, Hardy K, Winston RM, Moore GE. Expression of mRNA for the insulin-like growth factors and their receptors in human preimplantation embryos. Mol Reprod Dev 1997;47:134 –9. 31. Liu HC, He ZY, Mele CA, Veeck LL, Davis OK, Rosenwaks Z. Expression of IGFs and their receptors is a potential marker for embryo quality. Am J Reprod Immunol 1997;38:237– 45. 32. Bergh C, Carlsson B, Olsson JH, Selleskog U, Hillensjo T. Regulation of androgen production in cultured human thecal cells by insulin-like growth factor I and insulin. Fertil Steril 1993;59:323–31.

Fertility and Sterility姞

33. Nahum R, Thong KJ, Hillier SG. Metabolic regulation of androgen production by human thecal cells in vitro. Hum Reprod 1995;10:75– 81. 34. Nobels F, Dewailly D. Puberty and polycystic ovarian syndrome: the insulin/insulin-like growth factor I hypothesis. Fertil Steril 1992;58: 655– 66. 35. Conover CA, Ronk M, Lombana F, Powell DR. Structural and biological characterization of bovine insulin-like growth factor binding protein-3. Endocrinology 1990;127:2795– 803. 36. San Roman GA, Magoffin DA. Insulin-like growth factor-binding proteins in healthy and atretic follicles during natural menstrual cycles. J Clin Endocrinol Metab 1993;76:625–32. 37. Hourvitz A, Widger AE, Filho FL, Chang RJ, Adashi EY, Erickson GF. Pregnancy-associated plasma protein-A gene expression in human ovaries is restricted to healthy follicles and corpora lutea. J Clin Endocrinol Metab 2000;85:4916 –20. 38. Mazerbourg S, Overgaard MT, Oxvig C, Christiansen M, Conover CA, Laurendeau I, et al. Pregnancy-associated plasma protein-A (PAPP-A) in ovine, bovine, porcine, and equine ovarian follicles: involvement in IGF binding protein-4 proteolytic degradation and mRNA expression during follicular development. Endocrinology 2001;142:5243–53. 39. Cwyfan Hughes S, Mason HD, Franks S, Holly JM. Modulation of the insulin-like growth factor-binding proteins by follicle size in the human ovary. J Endocrinol 1997;154:35– 43. 40. Iwashita M, Kudo Y, Yoshimura Y, Adachi T, Katayama E, Takeda Y. Physiological role of insulin-like-growth-factor-binding protein-4 in human folliculogenesis. Horm Res 1996;1:31– 6.

1401