Molecular Human Reproduction, Vol. 6, No. 8, 677-680,
August 2000
© 2000 European Society of Human Reproduction and Embryology
Endocrinology |
Improved FSH sensitisation and aromatase assay in human granulosa–lutein cells
University of Manchester, Department of Medicine (Clinical Biochemistry), Clinical Sciences Building, Hope Hospital, Stott Lane, Salford M6 8HD, UK
Abstract
Granulosa–lutein (GL) cells from follicular aspirates from women undergoing in-vitro fertilization (IVF) treatment are usually refractory to follicle stimulating hormone (FSH) regarding the induction and/or maintenance of aromatase activity which converts androgens (e.g. testosterone) to oestrogens. The normal method of assaying FSH-stimulated aromatase activity in GL cell cultures is to add exogenous testosterone throughout the cell culture period and measure the secreted oestradiol. Thus under the conditions usually employed for studying FSH-stimulated oestradiol secretion in GL cells, the `total' FSH effect is dependent both on the decay of the aromatase concentration in culture relative to its induction/maintenance by FSH and on changes in its activity in the face of a declining substrate concentration as the exogenous testosterone is converted over several days to oestradiol. We have therefore used a technique for challenging the cells with testosterone (10 µmol/l) for just 2 h at the end of the normal longer-term culture period such that its concentration was essentially unchanged, thus ensuring that there was no depletion of the aromatase substrate and that the FSH stimulation phase could be performed independently of exogenous testosterone. Consequently, GL cells were incubated for 0, 24 and 48 h prior to stimulation with FSH (100 IU/l) for 24 h after which they were washed and challenged with testosterone for 2 h and the secreted oestradiol was assayed. Freshly isolated GL cells from women undergoing IVF were refractory to FSH but after preincubation were responsive such that there was a 3–14-fold increase over basal activity depending on the cell preparation. In conclusion, we have developed a simple 48 h procedure for sensitizing GL cells to FSH and established the conditions for optimizing the assay of aromatase activity independently from the effect of FSH on its induction.
aromatase/FSH/granulosa cells/granulosa/lutein cells
Introduction
Granulosa–lutein (GL) cells have been used widely for studying the control of oestradiol secretion (Bergh et al., 1991
; Yong et al., 1992) by follicle stimulating hormone (FSH) and most experimental designs include androstenedione or testosterone (µmol/l range) in the culture medium throughout the FSH stimulation phase to act as a substrate for the aromatase complex which is induced and maintained by the hormone (Erickson et al., 1989, 1991; Bergh et al., 1991; Yong et al., 1992). However, there are potential problems with this experimental approach as such cell preparations are usually refractory to physiological concentrations of FSH alone although a synergistic effect with IGF-I with regard to oestradiol secretion has been demonstrated (Erickson et al., 1989, 1991; Bergh et al., 1991; Wood et al., 1994).
Aromatase activity also declines rapidly in GL cells in culture even in the presence of FSH (Wood et al., 1994) and as the testosterone is normally present throughout the culture period of days, its concentration can also decline markedly as it is converted to oestradiol. Thus under the conditions commonly employed for studying FSH-stimulated oestradiol secretion, the overall effect is dependent on the decay of the enzyme in culture relative to its induction/maintenance by FSH and on changes in its activity in the face of a declining and variable substrate concentration. This, coupled with the fact that there are large variations in the degree of luteinization of the cells from patient to patient, makes GL cell cultures a difficult experimental system with which to work. However, as preovulatory human granulosa cells, which are very sensitive to FSH (Bergh et al., 1997), are difficult to obtain routinely, the need for a simple FSH-sensitive GL cell system remains. Recent work has shown that such cells can be sensitized to the hormone at physiological concentrations following an FSH-free preincubation period and a long (up to 6 days) FSH stimulation phase (Foldesi et al., 1998). Our aim was to establish a simple, rapid GL cell culture which was always responsive to FSH and in which the assay of aromatase activity (as assessed by oestradiol secretion) was performed optimally and independently of the FSH stimulation phase.
Materials and methods
GL cell preparations
Ethical committee approval was obtained for the use of human GL cells which were isolated from follicles taken from women undergoing IVF. The cells were isolated by centrifugation of the follicular aspirate (350 g for 5 min) and resuspended in 1 ml of McCoy's 5A medium (Hyclone, Cramlington, Northumberland, UK), supplemented with HEPES (15 mmol/l; Flow Labs, Irvine, UK), penicillin/streptomycin (100 IU/ml; Gibco BRL, Paisley, UK), glutamine (2 mmol/l, Gibco), BSA (0.1%) and collagenase (0.1%, Worthington Biochemical Corporation, Twyford, Berks, UK). The cells were then layered onto a gradient of 50% Percoll, 50% supplemented McCoy's medium, centrifuged at 450 g for 20 min at room temperature before being aspirated from the Percoll interface and washed with supplemented McCoy's medium (without collagenase). Cell viability was 80–90% and the cells were plated at a density of 12 500 viable cells/well in 96-well plates. For the experiments on establishing the effect of exogenous testosterone (10–7 to 10–5 mol/l) on the GL cells they were incubated for 24 h in 50% DMEM/50% McCoy's medium prior to being assayed for oestradiol by radioimmunoassay. For the experiments on the effect of FSH on GL oestradiol secretion the cells were preincubated for 0–48 h before being stimulated with recombinant human FSH (100 IU/l) for 24 h (Organon, Oss, The Netherlands). The plates were then centrifuged at 400 g for 10 min and medium was removed. The GL cells were then incubated for a further 2 h in 50% DMEM/50% McCoy's medium containing testosterone (10–5 mol/l). After this time the plate was centrifuged, the medium removed and stored at –40°C until analysed for oestradiol [in-house radioimmunoassay (RIA)]. The cells were then fixed for estimation of cell number (crystal violet assay). Our experimental design is based on the approach outlined in the Introduction and in essence combines three crucial stages. (i) The cells are preincubated in vitro in medium lacking FSH for up to 48 h following aspiration; this procedure is thought to allow the cells time to recover from gonadotrophin desensitization following their in-vivo stimulation by human chorionic gonadotrophin (HCG) as part of the patient's IVF treatment and has been used by other workers (Jalkanen et al., 1986; Foldesi et al., 1998). (ii) A 24 h stimulation phase with FSH following the preincubation period then allows sufficient time for the aromatase complex to be specifically induced by the hormone as has been demonstrated for this and other cell systems in the past (Van Damme et al., 1979; Foldesi et al., 1998). This stimulation phase is performed in the absence of high and potentially variable concentrations of the aromatase substrate, testosterone. (iii) Finally the level of aromatase activity induced by the FSH is assessed by a short-term challenge protocol where testosterone is added for just 2 h. This design is to ensure that there is no substrate depletion and that oestradiol is secreted under as close to initial velocity conditions for the aromatase complex as is possible for a cell-based system. This experimental approach was originally developed by us for the assessment of 11ß reductase activity in GL cells using cortisone as the enzyme substrate (Knaggs et al., 1997).
All data were corrected for cell number and expressed as oestradiol secretion/50 000 cells.
Oestradiol assay
Oestradiol was measured by standard in-house RIA using oestradiol–6-CMO–[125I]histamine purchased from Amersham and sheep antiserum to 17ß-oestradiol purchased from Advanced Biotechnologies Ltd, Leatherhead, Surrey. The percentage cross-reactivities of this antiserum were oestriol: 0.08, cortisol: 0.02, androstenedione: 0.056, dehydroepiandrosterone: 0.12, testosterone: <0.02 and progesterone: <0.02. The limit of detection of the assay was 125 pmol/l and intra- and inter-assay coefficients of variation were <10% and <17% respectively over the range 125–32 000 pmol/l.
Crystal violet assay
After steroid challenge the plates were centrifuged for 15 min at 285 g at 4°C. The medium was removed from each well and stored at –40°C for oestradiol estimation. Granulosa cells were fixed with glutaraldehyde (1%), washed with distilled water and air-dried. Crystal Violet (Sigma Chemical Co.), 0.1% prepared in 200 mmol/l boric acid buffer, pH 9.0 (100 µl) was added to each well. The plate was agitated gently for 20 min, washed with distilled water and air-dried. Acetic acid, 10% (100 µl) was added to each well, the culture plate was agitated gently for 30 min and absorbance measured at 540 nm. The limit of detection was 2500 cells/well. The relationship between absorbance and cell number was linear between 10 000 and 50 000 cells/well with 50 000 cells/well giving an absorbance of 1 at 540 nm.
Results
Effect of testosterone
The effect of increasing concentrations of testosterone (10–7 to 10–5 mol/l) on oestradiol secretion from GL cells incubated for 24 h is illustrated in Table I. Oestradiol secretion increased similarly and dramatically with all testosterone concentrations from <0.125 nmol/l (basal, mean ± SD, n = 4) to 21.7 ± 1.35 nmol/l (10–7 mol/l testosterone; a 22% conversion of substrate to product).
|
Sensitization of GL cells to FSH by preincubation
In the experiments on the effects of FSH on GL oestradiol secretion, we employed 10–5 mol/l testosterone in order to minimize the conversion rate but maximize the absolute level of oestradiol secretion during the short (2 h) post-FSH stimulation phase testosterone challenge. In these experiments we preincubated GL cells for 0–48 h before stimulating them for 24 h with 100 IU/l FSH which was near maximal for this system (see Figure 1
). Aromatase activity, as assessed by the assay of oestradiol secretion, was measured by the short testosterone challenge protocol described in the Materials and methods section. As can be seen in Table II there was a fair degree of patient-to-patient variability both in the basal oestradiol secretion (44–208 pmol/50 000 cells) and FSH-stimulated secretion (176–924 pmol/50 000 cells). In the absence of a preincubation, the GL cells were refractory to FSH (data not shown). However, following the 24 h preincubation period, all cell preparations were FSH-sensitive (P < 0.05) and oestradiol secretion increased 3–14-fold (range 176–924 pmol/50 000 cells). No improvement in the response was observed by increasing the preincubation period to 48 h (data not shown). Further, there was no improvement in the sensitivity of the GL cells to FSH by increasing the FSH stimulation period from 24 to 48 or 72 h (data not shown).
|
|
Discussion
We have developed a simple, 48 h procedure for producing FSH sensitive GL cells. Our technique uses a preincubation step (24 h) as did another study (72 h) (Foldesi et al., 1998). However, we used a short FSH stimulation phase (1 day versus 4–6 days), far fewer cells per experimental point (12 500 versus 100 000) and an aromatase assay which was performed independently of the FSH stimulation phase in the presence of a saturating dose of the aromatase substrate.
In the absence of a preincubation period our GL cells were refractory to FSH, as previously described (Wood et al., 1994). However, all our GL preparations were sensitive to near-physiological doses of FSH (100 IU/l) following 24 h preincubation in FSH-free culture medium. The percentage (280–1390%) changes in oestradiol secretion induced by the hormone were comparable to those reported by Foldesi et al. (1998), who also experienced considerable patient-to-patient variability regarding the cell preparations. Although it is true that the FSH sensitivity seen with our GL cells is considerably less than that seen (1–10 IU/l) using preovulatory granulosa cells from follicles of natural cycles (Bergh et al., 1997; Harris et al., 1998) our system and that of Foldesi et al. have demonstrated that GL cells which had been preincubated to allow recovery from the HCG stimulation received in vivo do become reproducibly sensitive to near physiological concentrations of FSH.
The conditions employed in this study were chosen carefully. The testosterone substrate for the aromatase enzyme was not part of the medium incorporating the hormone but was included for a short time (2 h) at the end of the incubation to avoid problems associated with both substrate depletion and end-product inhibition of the enzyme. This experimental design has several advantages over that used in previously published work (Erickson et al., 1989, 1991; Yong et al., 1992). Under these conditions, the FSH stimulation of aromatase activity was unaffected by an otherwise large and changing level of testosterone. During the 2 h incubation, the conversion of substrate to product was <0.5%, i.e. the testosterone concentration essentially remained unchanged. Thus under the reaction conditions employed, the assay of aromatase was performed under optimal conditions and independent of the FSH stimulation. Using the protocol outlined in this paper for increasing the FSH sensitivity of GL cells, we have studied the variations in bioactivity of glycoforms of human recombinant FSH of differing acidity (unpublished data).
In summary, we have established the culture conditions required to sensitize GL cells to FSH in a simple and relatively rapid way for these types of cells. This technique will provide scientists with a reproducible model system to examine the effects of FSH on GL function.
Acknowledgments
The authors thank Organon, Oss, The Netherlands and Salford Royal Hospitals NHS Trust for financial support.
Notes
1,4 Present address: Nuffield Department of Obstetrics and Gynaecology, Women's Centre, John Radcliffe Hospital, Headley Way, Headington, Oxford OX3 9DU, UK 
2 Present address: Willink Biochemistry Genetics Unit, Royal Manchester Children's Hospital, Pendlebury, Salford M27 4HA, UK
3 Present address: Nurture, Queen's Medical Centre, Nottingham NG7 2UH, UK
5 To whom correspondence should be addressed. E-mail: ann.lambert{at}obs-gyn.ox.ac.uk
References
Bergh, C., Olsson, J.H. and Hillensjo, T. (1991) Effect of insulin-like growth factor-1 on steroidogenesis in cultured human granulosa cells. Acta Endocrinol., 125, 177–185.
Bergh, C., Selleskog, U. and Hillensjo, T. (1997) Recombinant human gonadotrophins stimulate steroid and inhibin production in human granulosa cells. Eur. J. Endocrinol., 136, 617–623.
Erickson, G.F., Garzo, V.G. and Magoffin, D.A. (1989) Insulin-like growth factor-I regulates aromatase activity in human granulosa and granulosa luteal cells. J. Clin. Endocrinol. Metab., 69, 716–724.
Erickson, G.F., Garzo, V.G. and Magoffin, D.A. (1991) Progesterone production by human granulosa cells cultured in serum free medium: effects of gonadotrophins and insulin-like growth factor-I (IGF-1). Hum. Reprod., 6, 1074–1081.
Foldesi, I., Breckwoldt, M. and Neulen, J. (1998) Oestradiol production by luteinized human granulosa cells: evidence of the stimulatory action of recombinant follicle-stimulating hormone. Hum. Reprod., 13, 1455–1460.
Harris, S.D., Lambert, A. and Robertson, W.R. (1998) In-vitro biopotency of two glycoform mixture of differing acidity derived from recombinant FSH measured using a human granulosa–lutein cell bioassay. Hum. Reprod., 13 (Abstract Book), O-202.
Jalkanen, J., Huhtaniemi, I, Koskimies, A. et al. (1986) In vitro recovery of human chorionic gonadotrophin-stimulated cyclic adenosine 3', 5'- monophosphate production in desensitised human granulosa-luteal cells. Fertil. Steril., 46, 920–924.[Web of Science][Medline]
Knaggs, P., Lambert A., Proudfoot, F. et al. (1997) A rapid (<8 h) method for the measurement of the oxoreductase activity of 11ß-hydroxysteroid dehydrogenase in granulosa–lutein cells from patients undergoing IVF. Mol. Hum. Reprod., 4, 147–151.
Wood, A.M., Lambert, A., Hooper, M.A.K. et al. (1994) Exogenous steroids and the control of oestradiol secretion by human granulosa–lutein cells by follicle stimulating hormone and insulin-like growth factor-I. Hum. Reprod., 9, 19–23.
Van Damme, M.P., Robertson, D.M., Marana, R. et al. (1979) A sensitive and specific in vitro bioassay method for the determination of FSH activity. Acta Endocrinol., 91, 224–237.
Yong, E.L., Baird D.T., Yates R. et al. (1992) Hormonal regulation of the growth and steroidogenic function of human granulosa cells. J. Clin. Endocrinol. Metab., 74, 842–849.[Abstract]
Submitted on January 11, 2000; accepted on May 24, 2000.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S.A. Whitehead and M. Lacey Phytoestrogens inhibit aromatase but not 17{beta}-hydroxysteroid dehydrogenase (HSD) type 1 in human granulosa-luteal cells: evidence for FSH induction of 17{beta}-HSD Hum. Reprod., March 1, 2003; 18(3): 487 - 494. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.Y. Andersen, L. Leonardsen, A. Ulloa-Aguirre, J. Barrios-De-Tomasi, K.S. Kristensen, and A.G. Byskov Effect of different FSH isoforms on cyclic-AMP production by mouse cumulus-oocyte-complexes: a time course study Mol. Hum. Reprod., February 1, 2001; 7(2): 129 - 135. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||