Serum levels of insulin-like growth factor I and insulin-like growth factor-binding protein-l and -3 in women with regular menstrual cycles*

Transcription

1 FERTILITY AND STERILITY Vol. 63, No.6, June 1995 Copyright 1995 American Society for Reproductive Medicine Printed on acid-free papf'r in U. s. A. Serum levels of insulin-like growth factor I and insulin-like growth factor-binding protein-l and -3 in women with regular menstrual cycles* Hsin-Shih Wang, M.D., Ph.D.t Jing-Der Lee, M.D. Yung-Kuei Soong, M.D., M.Ph. Department of Obstetrics and Gynecology, Chang Gung Medical School, Chang-Gung Memorial Hospital, Lin-Kou Medical Center, Taipei, Taiwan, Republic of China Objective: To investigate the cyclic changes of serum insulin-like growth factor I (IGF-I), IGF-binding protein-1 (lgfbp-1) and IGFBP-3 levels during ovulatory menstrual cycle. Design: A prospective study following a preset protocol. Setting: A tertiary-care academic medical center. Participants: Thirty young female adults with regular menstrual cycles were recruited (18 with normal luteal phase and 12 with inadequate luteal function confirmed by serum P levels). Main Outcome Measures: Serum concentrations of IGF-I, IGFBP-1, and IGFBP-3 from women with regular menstrual cycles were assayed. Circulating E2 and P levels also were determined to certify the ovulatory cycles. Results: In women with normal luteal function, there was a peak of serum IGFBP-11evels before ovulation concomitant with the preovulatory E2 peak. The nadir of serum IGFBP-11evels was in the midluteal phase. Circulating IGFBP-1 elevated rapidly during late luteal phase and reached a peak on the 1st day of menstruation then declined slightly until a preovulatory IGFBP-1 peak occurred. In women with inadequate luteal function (midluteal serum levels of P < 10 ng/ml [conversion factor to SI unit, 3.180]), the preovulatory increase in serum IGFBP- 1 was not significant and the circulating IGFBP-11evels fluctuated throughout the menstrual cycle except for a unique peak of serum IGFBP-1 on the 1st day of menstruation. By contrast, there were no cyclic changes of serum IGF-I and IGFBP-3 levels in women with regular menstrual cycles, including both normal and inadequate luteal functions. Conclusions: The preovulatory increase in serum IGFBP-11evels may be offollicular origin and associated with the subsequent luteal function in females with ovulatory cycles. However, the involvement ofigfbp-1 in the process of follicular maturation and luteogenesis, as well as the regulation of luteal function, needs to be explored further. Fertil Steril 1995;63: Key Words: IGF-I, IGFBP-1, IGFBP-3, menstrual cycle Insulin-like growth factors (IGF-I and IGF-II) are polypeptides that act as a second messenger of GH and exert the growth-promoting effects of GH (1). Received June 20, 1994; revised and accepted January 13, * Supported by Medical Research grant CMRP-0426 from Chang-Gung Memorial Hospital, Lin-Kou Medical Center, Taipei, Taiwan, Republic of China. t Reprint requests: Hsin-Shih Wang, M.D., Ph.D., Department of Obstetrics and Gynecology, Chang-Gung Medical School, Chang-Gung Memorial Hospital, Lin-Kou Medical Center, Taipei, Taiwan, Republic of China (FAX: ). Insulin-like growth factor I is thought to be GH dependent and to be produced mainly by the liver (1). At the cellular level, IGFs stimulate cell proliferation and differentiation through an autocrine or paracrine mechanism (2). In the circulation, IGFs are carried by specific IGF-binding proteins (IGFBP- 1 to -6). In postnatal life, IGFBP-3 is the major binding protein for circulating IGFs (3). By contrast, IGFBP-l appears to regulate acute changes of serum IGFs (4) and additionally modulates the actions ofigfs at the cellular level (5). As with IGF I, IGFBP-3 is GH dependent (6). However, IGFBP Wang et a1. IGF-IGFBPs during the menstrual cycle Fertility and Sterility

2 1 is not GH dependent, but is P (7) and insulin dependent (8). In women with normal menstrual cycles, the endometrium of the uterus in the follicular phase is believed to be stimulated by E2, the main sex steroid at this stage, to produce IGF-I locally, which in turn accelerates the cell proliferation of the endometrium (9). After ovulation, P is prominent and stimulates the endometrium to secrete IGFBP-1, which modulates the action ofigf-i by shifting cell proliferation to cell differentiation of the endometrium (10). However, circulating IGFBP-1 levels appear to be unchanged in the luteal phase though an increase in local production ofigfbp-1 is observed by the endometrium in the milieu of high P levels (11). There is evidence that IGFBP-1 is synthesized by granulosa cells in dominant follicles and secreted into follicular fluid (FF) (12, 13). By contrast, IGF-I in FF is believed to be of serum origin (14). In ovarian follicles, insufficient production ofigfbp-1 may enhance the bioactivity oflocal IGF-I, which in turn stimulates androgen production of thecal origin and then results in follicular atresia and anovulation (15). In the present study, we explored the cyclic changes of circulating IGF-I, IGFBP-1, and IGFBP- 3 levels during the menstrual cycle. The relationship between serum IGFBP-1 levels and luteal function (interpreted by serum concentration of P) also was investigated. Subjects MATERIALS AND METHODS Blood samples were collected from 30 apparently healthy women who had at least three regular menstrual cycles (28 to 30 days per cycle) before the present study. Age of subjects ranged from 21 to 28 years. Informed consent was obtained from each woman. All samples were collected between 10:00 A.M. and 12:00 A.M. under nonfasting conditions on every other day throughout the menstrual cycle starting on the 1st day of menstruation. Serum was stored at -20 C until assay. Radioimmunoassay for IGFBP-l Serum concentrations ofigfbp-1 were measured by an RIA as described previously (16) labeled IGFBP-1 was prepared by a modification of the chloramine T method and used as a tracer. A polyclonal antiserum to IGFBP-1 (lot no. S515; see reference 16) was used at a final dilution of 1:400,000, which bound approximately 50% of 1251-labeled IGFBP-1. The detection limit of the assay was 5 f-lgil. The Vol. 63, No.6, June 1995 intra-assay coefficients of variation were 4.5% at 50 f-lg/l and 3.7% at 105 f-lgil (n = 16). The interassay coefficients of variation were 8.8% at 50 f-lgil and 7.4% at 105 f-lg/l (n = 12). Radioimmunoassay for IGFBP-3 Serum levels of IGFBP-3 were determined by an RIA kit (Nichols Institute Diagnostics B.V., Wijchen, The Netherlands). The batch of recombinant human IGFBP-3 (the f3 subunit of IGFBP-3 complex; containing 264 amino acids with a molecular weight of 28,500 d) used for radiolabeling and as standard in the RIA was of the same degree of purity as the material used for immunization of rabbits. Before RIA, serum samples were diluted 1:100 with assay buffer. The minimum detection limit of assay was 1.1 mgil. The intra-assay coefficients of variation were 6.4% at 7.4 mgil and 7.9% at 19.2 mgil (n = 40). The interassay coefficients of variation were 9.3% at 7.4 mg/l and 10.8% at 19.2 mgil (n = 20). Immunoradiometric Assay (IRMA) for IGF-I Serum levels ofigf-i were assayed by an IRMA kit (Diagnostic Systems Laboratories Inc., Webster, TX). Before assay for IGF-I, serum samples were treated by acid-ethanol extraction to dissociate IGF I from IGF -binding proteins. The detection limit of the assay was 0.83 f-lgil. The intra-assay coefficients of variation were 7.2% at 40 f-lg/l and 4.2% at 157 f-lg/l (n = 12). The interassay coefficients of variation were 10.1% at 40 f-lg/l and 6.4% at 157 f-lgil (n = 10). Immunofluorometric assays for E2 and P To certify ovulatory cycles, serum levels of ovarian hormones (E2 and P) were determined by immunofluorometric assays (Pharmacia, Turku, Finland). The detection limit of the E2 assay was 13.6 pg/ml (conversion factor to SI unit, 3.671). The intra-assay coefficients of variation were 5.7% at 136 pglml and 3.1 % at 409 pg/ml (n = 10). The interassay coefficients of variation were 8.4% at 136 pglml and 5.6% at 409 pglml (n = 6). The minimum detection limit of the assay for P was 0.1 ng/ml (conversion factor to SI unit, 3.18). The intra-assay coefficients of variation were 2.5% at 1.5 nglml and 6.8% at 11 ng/ml (n = 10). The interassay coefficients of variation were 8.9% at 1.5 nglml and 10.2% at 11 nglml (n = 6). Statistical Analysis Significant difference in serum IGFBP-1 and E2 levels on days 11, 13, and 15 of the menstrual cycle Wang et al. IGF-IGFBPs during the menstrual cycle 1205

3 IGFBP-1 (!1g1L) 60, , Progesterone (nglml) , o crease in serum concentration of E2 was not significant (Fig. 2); however, there was a relatively high level of circulating E2 in the luteal phase. In women with normal luteal function, there was a peak of serum IGFBP-l before ovulation (on day 13 of the menstrual cycle) concomitant with a preovulatory E2 peak (Fig. 1). The nadir of serum IGFBP-l levels was in the midluteal phase when serum P reached a maximum level Circulating IGFBP-l elevated rapidly during late luteal phase and reached a peak on the 1st day of menstruation, and then declined slightly until a preovulatory IGFBP-l peak appeared. In women with inadequate luteal function, the preovulatory increase in serum IGFBP-l was not significant and the circulating IGFBP-llevels fluctuated throughout the menstrual IGFBP-1 (!1g/L) 40 Estradiol (pglml) # P<O.0001 (1=16.3, df=34) # P<o.o001 (1=15.7, df=34) 100 Progesterone (ng/ml) 20 Figure 1 Serum levels ofigfbp-1, P, and E2 in women having ovulation with a subsequently normal luteal function during the menstrual cycle (n = 18). Values are means ± SD. Conversion factors to SI unit, 3.18 for P, and for E2 was evaluated by Student's paired t-test. A P value < 0.05 was considered significant. RESULTS Of 30 females recruited in the present study who had regular menstrual cycles ranging from 28 to 30 days per cycle, 18 subjects had ovulation with a subsequently normal luteal function (a preovulatory peak of serum E2 in association with serum P levels > 10 ng/ml on days 19, 21, and 23 of menstrual cycle) (Fig. 1) whereas 12 women were thought to have inadequate luteal function (as interpreted by midluteal serum levels of P < 10 nglml on days 19, 21 and 23 of ~enstrual cycle) (Fig. 2). In the group of inadequate. luteal function, the preovulatory in Wang et al. IGF-IGFBPs during the menstrual cycle Estradiol (pg/ml) 200 Figure 2 Serum levels of IGFBP-1, P, and E2 in women with inadequate luteal function during the menstrual cycle (n = 12). Values are means ± SD. Conversion factors to SI unit, 3.18 for P, and for E2 Fertility and Sterility

4 IGF-I (~g/l) 600, IGFBP-3 (mg/l) Figure 3 Serum levels of IGF-I and IGFBP-3 in women with normal and inadequate luteal functions during the menstrual cycle (n = 30). Values are means:!: SD. cycle except a unique peak of serum IGFBP-l on the 1st day of menstruation (Fig. 2). By contrast, there were no cyclic changes of serum IGF-I and IGFBP-3 levels in women with ovulation followed by either a normal or an inadequate luteal function (Fig. 3). DISCUSSION In preovulatory follicles, IGFBP-l is de novo synthesized by granulosa cells and may regulate the growth and differentiation of granulosa cells (17). Our previous studies also have shown that, at the preovulatory stage, IGFBP-l levels in FF are approximately 4.5-fold higher than those in serum (14). In addition, in situ hybridization studies have shown that messenger RNA (mrna) expression for IGFBP- 1 appears to be presented only in granulosa cells of dominant follicles (13). Furthermore, physiological significance of IGFBP-l in the ovary during follicular phase is believed to inhibit IGF-I-induced androgen production by theca cells and to prevent further follicular atresia and anovulation (15). These observations indicate that local production ofigfbp-l in dominant follicles at preovulatory stage appears to Vol. 63, No.6, June 1995 be substantial and may be important to the regulation of follicular growth and maturation through modulating actions ofigf-i. In the present study, a preovulatory increase in serum IGFBP-llevels concomitant with high levels of circulating E2 was observed in women with ovulation followed by a normal luteal function (Fig. 1). This is consistent with previous results that serum IGFBP-l concentration appears to be elevated at preovulatory stage in controlled ovarian hyperstimulation cycles (18). By contrast, in women with inadequate luteal function, the preovulatory increase in serum concentration of IGFBP-l and E2 was not significant and the circulating IGFBP-llevels fluctuated throughout the menstrual cycle (Fig. 2). These findings suggest that preovulatorily elevated serum IGFBP-l levels in females with ovulation and a normal luteal function may result from the substantial production of IGFBP-l in dominant follicles of good quality. The mechanism would be the same as the production and release of E2 from granulosa cells in dominant follicles. There is evidence that circulating IGFBP-l is thought to be produced mainly by the liver (19) and to be regulated by insulin (8). However, there is no cyclical change in serum insulin levels during ovarian stimulation (20). In postmenopausal women, circulating levels ofigfbp-l are not affected by administration of estrogen with or without P (21). Thus, the preovulatory increase in serum IGFBP-llevels neither result from the changes of production by the liver under the influence of insulin nor reflect the response of the liver secondary to the elevated estrogen levels. There is a nocturnal peak of serum IGFBP-llevels that rises 10- to 20-fold between midnight and 6:00 A.M. (8, 22). This diurnal variation of circulating IGFBP-l is thought to be regulated by insulin. In addition, there is evidence that IGFBP-l is regulated inversely by insulin (8) and also by metabolic and dietary factors (22). Ingestion of a meal causes an immediate lowering of plasma IGFBP-llevels (22). Circulating IGFBP-llevels also are suppressed after oral glucose administration (23). In the present study, blood samples were collected under nonfasting conditions. Thus, the preovulatory rise of serum IGFBP-l may not be associated with diurnal variation and food intake because both effects may lower circulating IGFBP-llevels. Insulin-like growth factor-binding protein-3 concentrations in FF decrease during folliculogenesis (24). Using in situ hybridization techniques, mrna for IGFBP-3 is detected in granulosa and theca cells of dominant follicles (13). Insulin-like growth factorbinding protein-3, a 150-kd complex, is thought to be retained in the circulation whereas smaller IGFBPs Wang et a1. IGF IGFBPs during the menstrual cycle 1207

5 (e.g., IGFBP-l) can cross the capillary (3). Thus, it is unlikely that serum IGFBP-3 would be a contributor for FF IGFBP-3. Unlike IGFBP-l and IGFBP-3, FF IGF-I is thought to be of serum origin (14). In dominant follicles at preovulatory stage, IGF-I concentration is substantially lower in FF than in serum (14). However, in the present study, no cyclical changes were observed in circulating IGF-I and IGFBP-3 levels (Fig. 3). Because both IGF-I and IGFBP-3 are GH dependent and synthesized mainly in the liver (1), subtle changes in ovarian follicles may not affect the circulating IGF-I and IGFBP-3 levels. These observations suggest that the follicular growth in the ovary may not be reflected by serum IGF-I and IGFBP-3 levels, though local concentrations of IGF-I and IGFBP-3 appear to vary during folliculogenesis (13, 14, 24). Insulin-like growth factor-binding protein-l appears to be P dependent and synthesized in the stromal cells ofthe endometrium and decidualized endometrium in the presence of P (11). However, the present results showed that serum IGFBP-l levels decreased during early luteal phase and rose rapidly shortly before menstruation in women with normal luteal function (Fig. 1). These findings suggest that local IGFBP-l production by the endometrium during luteal phase may be retained until breakdown of the endometrium before menstruation, when IGFBP-l may leak into the circulation. Alternatively, the endometrial production of IGFBP-l in vivo may not be controlled or even inhibited by P. Previous data have shown that IGF-I stimulates the growth of granulosa-lutein cells and in turn the production of P by these cells (25). However, this IGF I-induced P synthesis appears to be inhibited by IGFBP-l (17) but unaffected by LH or FSH (25). It is possible that the decline in serum levels ofigfbp- 1 during early luteal phase might reflect a decrease in the production of IGFBP-l by luteinized granulosa cells in women with normal luteal function. This decreased IGFBP-l in luteinized granulosa cells might enhance the bioactivity oflocal IGF-I and further stimulate the production of P. In conclusion, the preovulatory increase in serum IGFBP-llevels may be offollicular origin and associated with the subsequent luteal function in females with ovulatory cycles. However, the involvement of IGFBP-l in the process offollicular maturation and luteogenesis, as well as the regulation ofluteal function, needs to be investigated further. REFERENCES 1. Schwander JC, Hauri C, Zapf J, Froecsh ER. Synthesis and secretion of insulin-like growth factor and its binding protein by the perfused rat liver: dependence on growth hormone status. Endocrinology 1983; 113: Holly JMP, Wass JAH. Insulin-like growth factors: autocrine, paracrine or endocrine? New prospectives ofthe somatome din hypothesis in the light of recent developments. J Endocrinol 1989; 122: Binoux M, Hossenlopp P. Insulin-like growth factor (lgf) and IGF-binding proteins: comparison of human serum and lymph. J Clin Endocrinol Metab 1988;67: Wang HS, Chard T. Chromatographic characterization of insulin-like growth factor-binding proteins in human pregnancy serum. J Endocrinol 1992; 133: Straus DS, Ooi GT, Orlowski CC, Rechler MM. Expression of the genes for insulin-like growth factor-i (lgf-i), IGF-II, and IGF-binding proteins-1 and -2 in fetal rat under conditions of intrauterine growth retardation caused by maternal fasting. Endocrinology 1991; 128: Baxter RC, Martin JL. Radioimmunoassay of growth hormone-dependent insulin-like growth factor binding protein in human plasma. J Clin Invest 1986;78: Rutanen E-M, Koistinen R, Sjoberg J, Julkunen M, Wahlstrom T, Bohn H, et al. Synthesis of placental protein 12 by human endometrium. Endocrinology 1986; 118: Suikkari A-M, Koivisto VA, Rutanen E-M, Yki-Jarvinen H, Karonen S-L, Seppala M. Insulin regulates the serum levels of low molecular weight insulin-like growth factor-binding protein. J Clin Endocrinol Metab 1988;66: Giudice LC, Dsupin BA, Jin IH, Vu TH, Hoffman AR. Differential expression of messenger ribonucleic acids encoding insulin-like growth factors and their receptors in human uterine endometrium and decidua. J Clin Endocrinol Metab 1993;76: Rutanen E-M, Pekonen F. Insulin-like growth factor binding protein-1 in human endometrial function. In: Spencer EM, editor. Modern concepts of insulin-like growth factors. New York: Elsevier, 1991: Wahlstrom T, Seppala M. Placental protein 12 (PP12) is induced in the endometrium by progesterone. Fertil Steril 1984;41: Seppala M, Wahlstrom T, Koskimies AI, Tenhunen A, Rutanen E-M, Koistinen R, et al. Human preovulatory follicular fluid, luteinized cells of hyperstimulated preovulatory follicles, and corpus luteum contain placental protein 12. J Clin Endocrinol Metab 1984;58: EI-Roeiy A, Chen X, Roberts VJ, Shimasaki S, Ling N, LeRoith D, et al. Expression of the genes encoding the insulinlike growth factors (lgf-1 and 11), the IGF and insulin receptors, and IGF-binding proteins-1-6 and the localization of their gene products in normal and polycystic ovary syndrome ovaries. J Clin Endocrinol Metab 1994; 78: Chang SY, Hiesh KC, Wang HS, Soong YK. Follicular fluid levels of insulin-like growth factor-i, insulin-like growth factor binding protein-1 and ovarian steroids collected during ovum pick-up. Fertil SteriI1994;62: Nobels F, Dewailly D. Puberty and polycystic ovarian syndrome: the insulin/insulin-like growth factor I hypothesis. Fertil Steril 1992;58: Wang HS, Perry LA, Kanisius J, lies RK, Holly JMP, Chard T. Purification and assay of insulin-like growth factor-binding protein-i: measurement of circulating levels throughout pregnancy. J Endocrinol 1991; 128: Angervo M, Koistinen R, Suikkari A-M, Seppala M. Insulinlike growth factor binding protein-1 inhibits the DNA amplification induced by Insulin-like growth factor I in human granulosa-luteal cells. Hum Reprod 1991;6: Martikainen H, Tapanainen J, Ronnberg L, Kauppila A, Selenius P, Seppala M. Insulin-like growth factor binding protein- 1 and ovarian stimulation. Hum Reprod 1991;6: Wang et al. IGF-IGFBPs during the menstrual cycle Fertility and Sterility

6 19. Hossenlopp P, Seurin D, Segovia B, Portolan G, Binoux M. Heterogeneity of insulin-like growth factor binding proteins between structure and affinity. 2. Forms released by human and rat liver in culture. Eur J Biochem 1987; 170: Suikkari A-M, Koistinen R, Yki-Jarvinen H, Koivisto VA, Koskimies AI, Seppala M. Association of insulin-like growth factor-binding protein (IGF-BP25) with serum insulin and reproductive functions. In: Drops SLS, Hintz RL, editors. Insulin-like growth factor binding proteins. Amsterdam: Elsevier, 1989: Ren S-G, Braunstein GD, Cedars M, Gambone J, Judd H. Serum concentrations of two endometrial proteins are not useful for monitoring postmenopausal estrogen/progesterone therapy. Fertil SteriI1990;54: Busby WH, Snyder DK, Clemmons DR. Radioimmunoassay of a 26,000-dalton plasma insulin-like growth factor binding protein: control by nutritional variables. J Clin Endocrinol Metab 1988;67: Suikkari A-M, Koivisto VA, Koistinen R, Seppala M, Yki Jarvinen H. Dose-response characteristics for suppression of low molecular weight plasma insulin-like growth factor-binding protein by insulin. J Clin Endocrinol Metab 1989;68: San Roman GA, Magoffin DA. Insulin-like growth factorbinding proteins in healthy and atretic follicles during natural menstrual cycles. J Clin Endocrinol Metab 1992;76: Wood AM, Lambert A, Higham V, Hilton B, Mitchell GG, Hooper MAK, et al. Insulin-like growth factor-i stimulates growth and progesterone production by granulosa-lutein cells. Lack of interaction with physiological concentrations of luteinizing hormone and follicle stimulating hormone. Hum Reprod 1993;8: Vol. 63, No.6, June 1995 Wang et a1. IGF-IGFBPs during the menstrual cycle 1209