Subpopulations of human granulosa-luteal cells in natural and stimulated in vitro fertilization-embryo transfer cycles*

Vol. 65, No.3, March 1996

FERTILITY AND STERILITY Copyright

Printed on acid-free paper in U. S. A.

1996 American Society for Reproductive Medicine

Subpopulations of human granulosa-luteal cells in natural and stimulated in vitro fertilization-embryo transfer cycles*

Ksenija Gersak, M.Sc.t Tomaz Tomazevic, Ph.D. Department of Obstetrics and Gynecology, University Medical Center, Ljubljana, Slovenia

Objective: To find out the differences between human granulosa-luteal cells derived from natural and stimulated IVF-ET cycles. Design: Cells were obtained from dominant follicles of 52 women with natural cycles in whom preovulatory hCG was given when the follicle was mature and from 50 dominant follicles of women undergoing IVF-ET with hMG and hCG. Setting: In Vitro Fertilization Unit, University Department of Obstetrics and Gynecology Ljubljana, Slovenia. Main Outcome Measure: Four subpopulations of cells were observed by computerized image analysis in which hCG was localized using immunoperoxidase staining. Results: The nonluteinized granulosa cells from natural cycles were larger than those from the stimulated ones. In luteinized cell types, there was no difference in cell area between natural and stimulated cycles. Three staining types of hCG localization were found: on the surface membrane, on the surface membrane and within the cytoplasm, and within the cytoplasm alone. The hCG stained cells from natural cycles were larger than the ones from stimulated cycles. The natural developing follicles contained more hCG stained cells than the stimulated ones. The follicles with fertilizable oocytes had more cells with cytoplasmic hCG localization. Only in natural cycles was there was a correlation between follicular fluid hCG levels and the percentage of the hCG stained cells. Conclusion: We found differences in morphometric characteristics and hCG localization between human granulosa and granulosa-luteal cells obtained from natural and stimulated IVF-ET cycles. Fertil SteriI1996;65:608-13 Key Words: Human granulosa-luteal cells, human chorionic gonadotropin localization, natural and stimulated IVF-ET cycle

Human corpus luteum (CL) is composed of at least two morphologically distinct luteal cell types, granulosa-derived large luteal cells and theca interna-derived small luteal cells (1). The functional capacity of CL is determined by the quality of growth and development in the follicle that is its progenitor. However, the follicular development and maturation are affected by the hormonal microenvironment dur-

Received April 28, 1995; revised and accepted August 24, 1995. * Preliminary results of this study have been presented at the 9th World Congress on In Vitro Fertilization and Assisted Reproduction, Vienna, Austria, April 3 to 8, 1995. t Reprint requests: Ksenija Gersak, M.Sc., University Medical Center, Department of Obstetrics and Gynecology, Slajmerjeva 3, 61000 Ljubljana, Slovenia (FAX: 386-61-1401110).

608

ing the follicular phase (2-4). Human luteal cells have been studied using enzymatically dispersed mature CL, derived from surgical or autopsy specimens (5). With introduction ofIVF-ET into the clinical practice, granulosa-luteal cells have become abundantly available as a by-product of the procedures. The "abnormal" endocrinologic environment ofIVF-ET stimulated cycles in which granulosa cells are developing may impact upon normal luteal cell function (1). Exogenous gonadotropins have the potential to alter the steroidogenic, biochemical, and physical properties of developing follicles and, consequently, corpora lute a (6). The aim of our study was to evaluate the differences between human granulosa cells derived from natural IVF-ET cycles and granulosa-luteal cells derived from the stimulated

Gersak and Tomazevic Subpopulations of granulosa-luteal cells

Fertility and Sterility

J

cycles by comparing the morphometric characteristics of the aspirated cells and the presence of hCG on or within the cells. MATERIALS AND METHODS Patients

The granulosa-luteal cells were obtained from patients undergoing an IVF-ET program at the University Department of Obstetrics and Gynecology, Ljubljana, Slovenia. All patients, aged 25 to 35 years (2.3 ± 3.7 years; mean ± SD), had tubal factor infertility with regular ovulatory menstrual cycles verified by laparoscopy and by transvaginal ultrasonography. Their partners had normal semen analyses according to the World Health Organization criteria (7). They were divided randomly into two groups: natural and stimulated IVF-ET cycles. In women with natural cycles, follicular development was monitored by daily determinations of serum E2 levels starting on day 8 of the cycle. Serial transvaginal ultrasound examinations (transducer type 1849, 7.0 MHz; Bruel & Kjaer Medical, Gentofte, Denmark) were begun when serum E2 levels reached 110 pg/mL (conversion factor to SI unit, 3.671). Human chorionic gonadotropin (5,000 IU, Pregnyl; Organon, Oss, The Netherlands) was administrated 1M when the dominant follicle was> 17 mm in diameter, adjusted to individual duration of menstrual cycle and absence of LH in urine samples using a commercial test (Ovulations test; Epignost, Linz, Austria) on day 12.5 ± 0.9 of the cycle. In women with stimulated cycles, ovarian stimulation was performed with two ampules of hMG (150 IU 1M hMG per day) (Pergonal; Serono, Freiburg im Breisgau, Germany) starting on day 2 of the cycle. Human menopausal gonadotropin doses were regulated individually after the 5th day of stimulation (14.8 ± 1.6 ampules hCG per cycle). As in natural cycles, follicular development was monitored by daily determinations of serum E2 level and with transvaginal ultrasonography. When at least one follicle was> 17 mm in diameter and another was> 15 mm, ovulation was induced with 5,000 IU hCG between 11 and 12 P.M. on the same day (day 9.3 ± 1.1 of the cycle). Transvaginal follicular aspiration was performed 34 to 36 hours after hCG administration in both groups. Isolation of Granulosa-Luteal Cells

Fifty-two dominant follicles, containing only mature oocytes in metaphase I (8) and clear FF, were evaluated in natural cycles compared with 50 follicles in stimulated IVF-ET cycles. Mter oocytes were

Vol. 65, No.3, March 1996

removed and identified, the visible granulosa-luteal cell clumps were collected separately before individual aspirates were centrifuged (200 X g, 5 minutes). The volume of supernatant was measured and stored at -20°C until assayed for steroids and hCG. The sedimented cells and clumps were resuspended in 1.0 mL Ham's F-10 medium (pH 7.35, Seromed; Biochrom KG, Berlin, Germany). Cell viability was assessed by Trypan blue staining (87.3% ± 5.1%). The cells were fixed with 10% buffered Formalin for 30 minutes. Cell suspensions were placed on glass slides, air-dried, and stained with either hematoxylin and eosin for morphometric analysis or immunocytochemically for localization ofhCG. Oocytes were examined 24 to 26 hours after insemination. Morphometric Analysis

A total of 100 cell areas and cytoplasmic:nuclear ratios were determined for each individual follicle, using a computerized image analysis (a light microscope coupled with CCD camera, a color monitor, PC 486 computer, and the software package Horizon, Institute of Pathophysiology, School of Medicine Ljubljana, Slovenia). Granulosa-luteal cell subpopulations were identified by hematoxylin and eosin staining, cell area, and cytoplasmic:nuclear ratio, as described by Whitman et al. (9). Human Chorionic Gonadotropin Localization

Human chorionic gonadotropin was localized using immunocytochemical staining procedures (peroxidase-antiperoxidase method on Formalin-fixed cells with polyclonal antibodies to hCG, rabbit antihuman hCG, code no. A 231, lot no. 019; Dakopatts, Glostrup, Denmark) (10, 11). We considered hCG reaction as positive when brown precipitates were found on the large area of the membrane and/or within the cytoplasm. Hormone Assays

All samples were measured in duplicate. Serum E2 levels were measured by fluoroimmunoassay using commercial kits (DELFIA; Wallac Oy, Turku, Finland). The intra-assay and interassay coefficients of variation were 6.1% and 5.6%, respectively. Human chorionic gonadotropin was determined immunoluminometrically using commercial kits (LIA-mat HCG; Byk-Sangtec Diagnostica, Dietzenbach, Germany). The intra-assay and interassay coefficients of variation were 4.8% and 7.5%, respectively. Statistical Analysis

Statistical analysis was performed using analysis of variance (F-test) and Student's t-test for indepen-

Gersak and Tomazevic Subpopulations of granulosa-luteal cells

609

I

t

Table 1 Cell Areas and Cytoplasmic:Nuclear Ratios of Human Granulosa-Luteal Cell Subpopulations Obtained From Natural and Stimulated IVF-ET Cycles*

Cyclesoocyte fertilizability

Nonluteinized cells

Natural Fertilized Natural N onfertilized Stimulated Fertilized Stimulated N onfertilized

43.83 :!: 1.56 :!: 43.31 :!: 1.70 :!: 39.42 :!: 1.58 :!: 39.89:!: 1.58 :!:

6.18t 0.24 5.67t 0.41 6.09 0.27 7.29 0.29

Small luteinized 59.84 1.76 58.07 1.86 60.22 1.85 59.50 1.91

:!: :!: :!: :!: :!: :!: :!: :!:

9.63 0.34 10.43 0.37 10.46 0.36 10.86 0.41

Medium luteinized 86.98 :!: 2.03 :!: 85.02 :!: 2.01 :!: 87.16 :!: 2.01 :!: 86.49 :!: 2.04:!:

6.70 0.34 6.64 0.29 7.48 0.38 6.71 0.42

Large luteinized 119.51 :!: 2.21 :!: 121.97 :!: 2.08 :!: 118.74:!: 2.14 :!: 121.49 :!: 2.23 :!:

16.41 0.34 21.36 0.33 21.42 0.41 22.32 0.49

Membrane hCG stained cells 53.66 1.72 50.91 1.93 51.67 1.76 50.05 1.73

:!: :!: :!: :!: :!: :!: :!: :!:

8.85t 0.28 9.54:j: 0.41 10.11 0.33 7.35 0.28

Membranecytoplasmic hCG stained cells 71.09 :!: 1.86 :!: 73.93:!: 2.00 :!: 65.67 :!: 1.93 :!: 66.56 :!: 1.86 :!:

13.91 t 0.42 14.30t 0.36 11.51 0.39 13.10 0.32

Cytoplasmic hCG stained cells 94.76 2.04 96.99 1.99 86.24 2.06 87.15 1.96

:!: :!: :!: :!: :!: :!: :!: :!:

17.60t 0.36 19.32t 0.27 15.01 0.38 12.27 0.42

* Values are means:!: SD. t P < 0.001.

:j:P < 0.05.

dent variables. Correlations were examined by Pearson correlation coefficient for linear relationships. Significance was defined as P < 0.05 and all values were expressed as means ± SD.

within the cytoplasm. The average cell areas and cytoplasmic:nuclear ratios of each hCG stained cell type are presented in Table 1. The hCG stained cells from naturally developing follicles were larger than those from the stimulated ones (P < 0.001, P < 0.05), except for the cells with membrane hCG localization obtained from naturally developing follicles and nonfertilizing oocytes. Cells > 120 p,m 2 in naturally developing follicles and cells > 100 p,m 2 in the stimulated ones did not localize hCG on the surface membrane. Cells > 150 p,m 2 did not stain for hCG at all. The relative number of each hCG stained cell type is shown in Figure 1B. The naturally developing follicles contained more hCG stained cells than the stimulated ones (P < 0.001). The follicles with fertilizable oocytes had more granulosa-luteal cells with cytoplasmic hCG localization and fewer cells with membrane hCG localization (P < 0.001). The ability ofthe oocyte to fertilize in vitro was not affected by follicular fluid (FF) hCG concentration in stimulated IVF-ET cycles (28.8 ± 10.7 mIU/mL with fertilized oocytes and 24.3 ± 7.6 mIU/mL with nonfertilized oocytes; P >0.05 [conversion factor to SI unit, 1.00]). In naturally developing follicles with fertilized oocytes, FF had higher hCG levels (32.0 ± 8.8 mIUlmL) than with nonfertilized oocytes (17.2 ± 3.2 mIU/mL) (P < 0.001). There was a correlation between FF hCG levels and the percentage of hCG stained cells but only in natural IVF-ET cycles: the higher the FF hCG level, the fewer cells with membrane hCG localization (r = -0.85, P < 0.001), and more cells with cytoplasmic hCG (r = 0.94, P < 0.001) were observed in the same aspirate (Fig. 2).

RESULTS

The volume of aspirated naturally developing follicles was 4.0 ± 0.7 mL and the volume of dominant follicles from stimulated cycles was 4.0 ± 0.9 mL. Fertilization rate was 71.1% in natural IVF -ET cycles and 60.0% in the stimulated ones (only dominant follicles). The average number of blastomers was 3.5 in natural and 3.6 in stimulated IVF-ET cycles. We identified four granulosa-luteal cell subpopulations. Nonluteinized and luteinized cells were identified by the presence of eosinophilic cytoplasm, which characterizes all luteal cells. Luteinized granulosa-luteal cells were classified into three subpopulations according to total cell area and cytoplasmic:nuclear ratio: small, medium, and large cells. The nonluteinized cells were all small (Table 1). The nonluteinized granulosa cells from natural developing follicles were larger than those from the stimulated ones (P < 0.001). The average cell areas of each luteinized cell type was influenced neither by the two IVF protocols nor by fertilizability of the oocytes. The relative number of each ofthe four subpopulations within the follicular aspirates is shown in Figure 1A. No differences were detected between natural developing follicles. In stimulated cycles, follicles with fertilized oocytes had more nonluteinized cells and fewer large luteinized cells (P < 0.05). Three staining types of hCG localization were found: on the surface membrane, on the surface membrane and within the cytoplasm, and only

610

DISCUSSION

We found differences in morphometric characteristics and hCG localization between human granu-

Gersak and Tomazevic Subpopulations of granulosa-luteal cells

Fertility and Sterility

A

(f)

w

'0::<

......

0...

(f)

«

z ...... (f)

-.J -.J W

U

NONLUT B

(f)

w

'0::<

SMALL

MEDIUM

LARGE

50

**

40

**

...... 0...

(f)

«

2S (f)

-.J -.J W

U

x

MEM

MEMjCYTO

CYTO

TOTAL

Figure 1 Distribution of human granulosa and granulosa-luteal cell subpopulations within the follicular aspirates obtained from natural and stimulated IVF-ET cycles with fertilized or nonfertilized oocytes. (A), nonlut, hematoxylin and eosin stained nonluteinized cells, small, medium, and large luteinized cells. *p < 0.05. (B), Human chorionic gonadotropin stained cells. mem, membrane hCG localization; memlcyto, membrane and cytoplasmic hCG localization; cyto, cytoplasmic hCG localization. *p < 0.001 (fertilized-nonfertilized), **p < 0.001 (natural-stimulated). iIlI, stimulated fertilized; iIlI, stimulated nonfertilized; 0, natural fertilized; ~, natural nonfertilized.

the exogenous administration of gonadotropins (12). Therefore, it is possible that even after exposure to FSH such follicles could not generate their full complement of granulosa cells, which remain cytoplasm deficient. As granulosa-luteal cells from stimulated follicles differentiate into luteal cells, they possibly compensate for their volume. However, we found that the average cell area of each luteinized cell subpopulation was not influenced by either the two different IVF -ET protocols or the fertilizability of the oocytes (Table 1). In our study, four subpopulations of granulosaluteal cells were obtained in all 102 follicles (Fig. 1). Only in stimulated cycles were there differences between the distribution in aspirates with fertilized and nonfertilized oocytes. Whitman et al. (9) analyzed only stimulated cycles. They also found four granulosa-luteal cell subpopulations in each of the follicular aspirates, but follicles with nonfertilizable oocytes had fewer nonluteinized cells and a higher number of large cells. A higher relative number of large granulosa-luteal cells may reflect changes in the stimulated follicles that began to luteinize before the heG injection (9). The process of luteinization is associated with plasma membrane LH and heG receptor content (13). By binding to the surface membrane receptors, LH or heG stimulates luteinization and LH and heG receptor complex is internalized into the cytoplasm (14-16). Positive heG staining was evaluated semiquantitatively by the presence of brown precipitates on surface membrane and/or within cytoplasm viewed under bright-field optics. Therefore the lu-

50 r=0.94*

(f)

w 40 fo« 0:: ......

0...

losa cells obtained from natural IVF-ET cycles and granulosa-luteal cells obtained from the stimulated ones. The nonluteinized granulosa cells from naturally developing follicles were larger than those from stimulated cycles. This result may reflect a longer follicular phase and a slower cell development in natural cycles, before heG administration. In stimulated cycles, the follicular phase was approximately 3 days shorter than in natural ones. The second explanation could be found in the observation that antral follicles (:54 mm in diameter) with 50% to 95% of their functionally active granulosa cells could be recruited into the developing follicle pool only after

Vol. 65, No.3, March 1996

(f)

30

(f)

20

« z ...... -.J -.J W U

x

••

r=-0.85*

••

10

0 0

• 10

20

40

50

60

70

hCG (IUjL) Figure 2 Correlation between the FF hCG levels and the percentage of hCG stained cells obtained from natural developing follicles, 6, cells with membrane hCG localization; ., cells with cytoplasmic hCG localization. *p < 0.001.

Gersak and Tomazevic

Subpopulations of granulosa-luteal cells

611

teinized cells with lower LH and hCG receptor densities could be identified as "negative." We found three types of hCG localization in all 102 follicular aspirates. The follicular aspirates from natural NF-ET cycles had significantly more hCG stained cells (Fig. 1B). Their LH and hCG receptor density on the cell membrane was probably also higher. The more hCG bound to functionally active receptors on the surface membrane, the more extensive the process of luteinization. However, our results show that all three hCG stained cell types from natural cycles were larger than the cells obtained from the stimulated ones (Table 1). There was no particular cell type, the presence or absence of which would be characteristic offertilizability of the oocyte. But there were more cells with cytoplasmic hCG localization and fewer with membrane hCG localization in the follicles containing fertilizable oocytes from natural and stimulated IVFET cycles. Our results confirm earlier findings (17, 18). Whitman et al. (19) demonstrated similar observations in stimulated cycles. They also found a higher relative number of cytoplasmic hCG stained cells in follicles with fertilizable oocytes but no differences in the other two hCG stained cell types. These results suggest that granulosa-luteal cells from those follicles have a higher LH and hCG receptor content before hCG administration than those containing nonfertilizable oocytes. The LH and hCG receptor complex in the cytoplasm may be detectable up to 42 hours (1, 9) and eventually is degraded within the lysosomes. In human granulosa-luteal cells, the process of degradation could be completed earlier. In our study, 36 hours after the hCG administration we could not find cells >120 f-Lm 2 with membrane hCG localization, but only with cytoplasmic localization. When the volume of cytoplasm increased the cell area > 150 f-Lm 2 , cells were not stained for hCG at all. Probably in those cells the process of luteinization has been completed, resulting in the formation of large luteal cells with increased volume of cytoplasm and structural elements needed for active steroidogenesis. Follicular fluid hCG levels correlated with the percentage of hCG stained cells only in natural NF cycles regardless of oocyte fertilizability (Fig. 2). In follicles with lower hCG levels, granulosa cells with membrane hCG localization were predominant, whereas the percentage of cells with cytoplasmic hCG localization was higher in follicles with higher hCG level. The relationship may not be governed only by the endocrine drive ofhCG but also by vascular supply, the specific enzyme content of the cells, and autocrine and paracrine intraovarian regulation

612

(20). It also could be influenced by the absolute number of different follicular cell types. Differences in morphometric characteristics and hCG localization could be associated with the quality of oocytes and the corpus luteum function. The relative number of hCG stained granulosa-luteal cells might explain different steroidogenic characteristics of individual corpus luteum. It also could be involved in the mechanism ofluteal phase deficiency and, consequently, in endometrial receptivity. Naturally developing granulosa cells are more responsive to the hCG and therefore could provide an optimal environment for the oocyte maturation. Follicles in stimulated cycles may not reflect the physiology of natural cycles and likewise corpora lutea developing from them. The luteal phase support has a beneficial effect in NF-ET cycles. Our results from a previous study (21) and reports by Hutchinson-Williams et al. (22, 23) show that timing of luteal supplementation with hCG significantly influences clinical pregnancy rate ofhMG and hCG-induced NF-ET cycles. Therefore, we presume that the lower number of hCG stained granulosa-luteal cells obtained from stimulated follicles influence functional characteristics of corpus luteum during early luteal phase, before theca-derived luteal cells in midluteal phase respond to hCG with a significant increase in P production. Further studies are needed to improve the understanding of differences found in luteal cell subpopulations indicating various regulatory mechanisms ofluteogenesis in IVF-ET cycles. Acknowledgments. We thank Srecko Rainer, Ph.D., Department of Obstetrics and Gynecology, University Medical Center, Ljubljana, Slovenia for supervision of this study; Institute of Pathophysiology, School of Medicine, University Medical Center, Ljubljana, Slovenia for using the computerized image analysis system; Mrs. Marusa Zalar for her technical assistance; and the IVF-laboratory team: Mrs. Liljana Bacer-Kermavner, Mrs. Jozica Mivsek, Mrs. Brigita Valentincic, Mrs. Alenka Veble, and Mrs. Irma Virant-Klun for assistance with the collection of materials.

REFERENCES 1. Fritz MA, Fritz TA. The functional Microscopic anatomy of

2.

3.

4. 5.

the corpus luteum: the "small cell"-"large cell" controversy. Clin Obstet Gynecol1991;24:144-56. Rotmensch S, Dor J, Furman A, Rudak E, Mashiach S, Amsterdam A. Ultrastructural characterization of human granulosa cells in stimulated cycles: correlation with oocyte fertilizability. Fertil Steril1986;45:671-9. Hill GA, Herbert CM, III, Colston Wentz AC, Osteen KG. Use of individual human follicles to compare oocyte in vitro fertilization to granulosa cell in vitro luteinization. Fertil Steril1987;48:258-64. Hartshorne GM. Subpopulation of granulosa cells within the human ovarian follicle. J Reprod Fertil 1990;89:773-82. Ohara A, Mori T, Taii S, Ban C, Narimoto K. Functional differentiation in steroidogenesis of two types of luteal cells

Gersak and Tomazevic Subpopulations of granulosa-luteal cells

Fertility and Sterility

6.

7.

8.

9.

10. 11.

12.

13.

14.

isolated from mature human corpora lutea of menstrual cycle. J Clin Endocrinol Metab 1987;65:1192-200. Stouffer RL, Hodgen GD, Graves PE, Danforth DR, Eyster KM, Ottobre JS. Characterization of corpora lutea in monkeys after superovulation with human menopausal gonadotropin or follicle-stimulating hormone. J Clin Endocrinol Metab 1986;62:833-9. World Health Organization. Laboratory manual for the examination of human semen and semen-cervical mucus interaction. 2nd ed. Cambridge: The Press Syndicate of the University of Cambridge, 1987. Veeck LL. Atlas of human oocyte and early conceptus. Human oocytes at the time of follicular harvest. Baltimore: Williams & Wilkins, 1986:5-131. Whitman GF, Luciano AA, Maier DB, Peluso JJ. Human chorionic gonadotropin localization and morphometric characterization of human granulosa-luteal cells obtained during in vitro fertilization cycles. Fertil SteriI1989;51:475-9. Baurne JA. Handbook of immunoperoxidase staining methods. Santa Barbara: Dako Corporation, 1983:1-37. Robinson G. Immunohistochemistry. In: Bancroft JD, Stevens A, editors. Theory and practice of histological techniques. Edinburgh: Churchill Livingstone, 1982:406-27. McNatty KP, Moore Smith D, Makris A, Osathanondh R, Ryan K. 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 1979;49:85160. Rajaniemi HJ, Ronnberg L, Kauppila A, Ylostalo P, Julkanen M, Saastamoinen J, et al. Luteinizing hormone receptors in human ovarian follicles and corpora lutea during menstrual cycle and pregnancy. J Clin Endocrinol Metab 1981; 52:30713. Jaaskelainen K, Markkanen S, Rajaniemi H. Internalization of receptor-bound human chorionic gonadotropin in preovulatory rat granulosa cells in vivo. Acta Endocrinol (Copenh) 1983; 103:406-9.

Vol. 65, No.3, March 1996

15. Amsterdam A, Nimrod A, Lamprecht SA, Burstein Y, Lindner HR. Internalization and degradation of receptor-bound hCH in granulosa cell cultures. Am J PhysioI1979;236:129. 16. Lakkakorpi JT, Yang M, Rajaniemi HJ. Processing of the LHI hCG receptor and bound hormone in rat luteal cells after hCG-induced down-regulation as studied by a double immunofluorescence technique in conjunction with confocal laser scanning microscopy. J Histochem Cytochem 1994;42:72732. 17. Gersak K. Subpopulations of human granulosa-luteal cells in natural and stimulated IVF cycles. J Assist Reprod Genet 1995; 12 Suppl:51. 18. Gersak K, Tomazevic T. Subpopulations of human granulosaluteal cells in natural and stimulated IVF-ET cycles. In: Aburumieh A, Bernat E, Dohr G, Feichtinger W, Fischl F, Huber J, et aI., editors. IX. World congress on in vitro fertilization and assisted reproduction. Bologna: Manduzzi Editore, 1995:191-4. 19. Whitman GF, Luciano AA, Maier DB, Peluso JJ. Influence of human chorionic gonadotropin (hCG) and hCG internalization by granulosa cells on the rate of vitro fertilization and embryonic development of human oocytes. Fertil Steril 1988; 50:607 -11. 20. Dawood YM. Corpus luteal insufficiency. Curr Opin Obstet Gynecol 1994;6:121-7. 21. Tomazevic T, Meden-Vrtovec H, Drobnic S, Bernjak I, Kastelic D, Veble A. Mid-luteal hCG support ofIVF cycles stimulated by menotrophins. Hum Reprod 1992; 7 Suppl 2:61. 22. Hutchinson-Williams KA, Lunenfeld B, Diamond MP, Lavy G, Boyers SP, DeCherney AH. Human chorionic gonadotropin, estradiol, and progesterone profiles in conception and nonconception cycles in an in vitro fertilization program. Fertil Steril 1989;52:441-5. 23. Hutchinson-Williams KA, DeCherney AH, Lavy G, Diamond MP, Naftolin F, Lunenfeld B. Luteal rescue in in vitro fertilization-embryo transfer. Fertil Steril 1990;53:495-501.

Gersak and Tomazevic

Subpopulations of granulosa-luteal cells

613