Ovarian modulators of type 1 11{beta}-hydroxysteroid dehydrogenase (11{beta}HSD) activity and intra-follicular cortisol:cortisone ratios correlate with the clinical outcome of IVF

L.M. Thurston1, D.P. Norgate1, K.C. Jonas1, L. Gregory2, P.J. Wood3, B.A. Cooke1 and A.E. Michael1,4

1 Department of Biochemistry and Molecular Biology, Royal Free and University College Medical School (Royal Free Campus), University College London, Rowland Hill Street, London NW3 2PF; 2 Cardiff Assisted Reproduction Unit, University Hospital of Wales, Heath Park, Cardiff CF4 4XN; 3 Regional Endocrine Unit, Southampton General Hospital, Southampton SO16 6YD, UK

4 To whom correspondence should be addressed. e-mail: t.michael{at}rfc.ucl.ac.uk


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: Follicular fluid (FF) contains compounds that can modulate NADP+-dependent oxidation of cortisol by type 1 11{beta}-hydroxysteroid dehydrogenase (11{beta}HSD). The objective of this study was to investigate the relationships between levels of the ovarian modulators of type 1 11{beta}HSD, intra-follicular cortisol:cortisone ratios and the clinical outcome of IVF cycles. METHODS: A single random sample of FF was aspirated from each of 132 patients undergoing gonadotrophin-stimulated IVF. Components of FF, resolved using C18 column chromatography, were evaluated for effects on NADP+-dependent cortisol oxidation in rat kidney homogenates. Intra- follicular steroid concentrations were measured by radioimmunoassays. Clinical pregnancies were confirmed by ultrasonography at 6 weeks post-embryo transfer. RESULTS: Levels of the hydrophilic ovarian 11{beta}HSD stimuli were significantly lower (P < 0.0001) and levels of the hydrophobic ovarian 11{beta}HSD inhibitors were significantly higher (P < 0.002) in conception versus non-conception cycles. Intra-follicular cortisol:cortisone ratios increased with the degree of inhibition of 11{beta}HSD by the hydrophobic FF fractions. FF obtained from conception cycles had significantly higher cortisol:cortisone ratios than samples from non-conception cycles (12.9 ± 0.3 versus 8.5 ± 0.2, respectively; P < 0.0001). CONLCUSIONS: Conception by IVF is associated with elevated intra-follicular cortisol:cortisone ratios, which reflect low levels of ovarian stimuli and/or high levels of ovarian inhibitors of type 1 11{beta}HSD.

Key words: 11{beta}HSD/cortisol/cortisone/hydroxysteroid dehydrogenase/IVF


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Within target cells, access of glucocorticoids to corticosteroid receptors is regulated by 11{beta}-hydroxysteroid dehydrogenase (11{beta}HSD) (reviewed by Monder, 1991Go; White et al., 1997Go; Koteletsev et al., 1999Go; Stewart and Krozowski, 1999Go; Quinkler et al., 2001Go; Seckl and Walker, 2001Go). To date, two isoforms of this enzyme have been cloned. Type 1 11{beta}HSD is an NADP(H)-dependent enzyme that can act in vitro as either a low affinity 11{beta}-dehydrogenase enzyme or as an 11-ketosteroid reductase (11KSR). In vivo, type 1 11{beta}HSD is thought to act predominantly as an 11KSR enzyme, generating cortisol and corticosterone from their inert 11-ketosteroid counter parts (cortisone and 11-dehydrocorticosterone respectively) (Lakshmi and Monder, 1988Go; Agarwal et al., 1989Go; Tannin et al., 1991Go). In contrast, type 2 11{beta}HSD acts almost exclusively as a high affinity 11{beta}-dehydrogenase enzyme, oxidizing physiological glucocorticoids to inert 11-ketosteroid metabolites (Brown et al., 1993Go; Rusvai and Naray-Fejes-Toth, 1993Go; Agarwal et al., 1994Go; Albiston et al., 1994Go; Zhou et al., 1995Go).

Over the past decade, expression of both cloned isoforms of 11{beta}HSD has been confirmed in several ovarian cell types, including the oocyte, cumulus cells, granulosa cells, theca cells, granulosa-lutein cells, corpus luteum and ovarian surface epithelium (Benediktsson et al., 1992Go; Waddell et al., 1996Go; Michael et al., 1997Go; Smith et al., 1997Go; 2000Go; Tetsuka et al., 1997Go; 1999aGo,bGo; Ricketts et al., 1998Go; Yong et al., 2000Go; 2002Go). Within granulosa cells, the expression of these enzymes is developmentally regulated. Specifically, granulosa cells that have not experienced an ovulatory LH surge express type 2 11{beta}HSD, whereas luteinizing granulosa cells express type 1 11{beta}HSD with no detectable expression of the type 2 enzyme isoform (Michael et al., 1997Go; Smith et al., 1997Go; Tetsuka et al., 1997Go; 1999aGo,bGo; Ricketts et al., 1998Go; Thurston et al., 2003Go).

In initial studies, we reported an inverse correlation between ovarian 11{beta}HSD activities (as measured in human granulosa-lutein cells) and the probability of conception by gonadotrophin-stimulated IVF (Michael et al., 1993Go; 1995Go). However, two independent studies found no such relationship (O’Shaughnessy et al., 1997Go; Thomas et al., 1998Go). Attention subsequently turned to cortisol:cortisone ratios in follicular fluid (FF) as an in-vivo reflection of ovarian 11{beta}HSD activities. While Yding Andersen et al. (1999)Go found no association between intra-follicular cortisol:cortisone ratios and conception by IVF, two studies have reported high intra-follicular ratios of cortisol:cortisone (consistent with low ovarian 11{beta}HSD activities) to be associated with an increased probability of conceiving in either gonadotrophin-stimulated or natural cycle IVF (Michael et al., 1999Go; Keay et al., 2002Go).

Recently, we have reported that FF aspirated from women undergoing IVF contains compounds that can modulate the NADP(H)-dependent activities of type 1 11{beta}HSD (Thurston et al., 2002Go). Using C18 column chromatography with a linear methanol gradient to resolve FF components, we have established that human FF contains both hydrophilic compounds, eluted at 0% (v/v) methanol, that can acutely stimulate NADP(H)-dependent type 1 11{beta}HSD activities up to 2.7-fold and hydrophobic compounds, eluted at 80% (v/v) methanol, that can acutely inhibit type 1 11{beta}HSD by up to 78%. In light of our previous findings, the primary aim of the present study was to establish whether intra-follicular levels of the hydrophilic ovarian stimuli and/or the hydrophobic ovarian inhibitors of type 1 11{beta}HSD might correlate with the clinical outcome of IVF. The secondary aims were to test whether the intra-follicular cortisol:cortisone ratios correlated with (i) the levels of these paracrine modulators of type 1 11{beta}HSD activity or (ii) the clinical outcome of IVF.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Sample collection and cell culture
Patients (aged between 23 and 43 years) attending the Cardiff Assisted Reproduction Unit (University Hospital of Wales, Cardiff, UK) underwent controlled ovarian hyper-stimulation as described previously (Michael et al., 1995Go; 1999Go). In brief, patients were treated for at least 14 days with a GnRH analogue nasal spray (Synarel: Nafarelin, Searle, Monsanto plc., High Wycombe, UK; or Suprecur: Buserelin, Shire Pharmaceutical Ltd, Andover, UK) until pituitary desensitization was achieved. Multiple follicluar development was then induced by administration of either recombinant (r)FSH (Puregon; Follitropin {beta}, Organon Laboratories Ltd., Cambridge, UK) or hMG (Menopure; Ferring Pharmaceuticals Ltd, Langley, UK). Follicular development was monitored by vaginal ultrasound accompanied by measurements of plasma estradiol (E2) and LH (ACS180; Automated Chemiluminescence System, Bayer Corporation, Tarrytown, NY, USA). Ultrasound measurements and hormone assays were carried out on days 1–8 and day 11 of stimulation, and thereafter on alternate days.

When a minimum of three follicles had attained a maximum diameter >18 mm, 10 000 IU hCG (Pregnyl, Organon, UK or Profasi, Serono, UK) was administered by subcutaneous injection. At 36–37 h post-hCG administration, samples of FF were aspirated transvaginally in the course of oocyte retrieval. On collection, samples were stored for up to 3 days at 4°C prior to transport (on wet ice) to the Royal Free Campus of the Royal Free and University College Medical School of Medicine (RF&UCMS). On arrival, samples of FF were divided into aliquots (1–5 ml) and frozen at –20°C pending analysis.

FF samples were only used for study if confirmed to be free of blood and/or granulosa cells (assessed by visual inspection and light microscopy), and also free of Earle’s balanced salt solution (EBSS) (the presence of which was indicated by the deposition of phenol red pH inidicator when FF samples were loaded onto C18 columns). As a precautionary measure, all FF samples were centrifuged for 10 min at 250 g and the supernatants were transferred to fresh sample tubes before freezing. Acellular samples of FF were stored frozen at –20°C for up to 1 month prior to measurement of intra-follicular steroid concentrations. In total, 132 samples of FF were analysed in this study. Each FF sample was a single, random FF sample (i.e. not necessarily the FF aspirated from the follicle giving rise to one of the replaced embryos) from each of 132 different patients, 44 of whom conceived.

All samples were obtained in accordance with the Declaration of Helsinki and with the approval of the University Hospital of Wales Cardiff & Vale NHS Trust local ethics committee.

Assay for ovarian modulators of type 1 11{beta}HSD
Effects of each of the 132 FF samples, and of the hydrophilic and hydrophobic components thereof, on NADP+-dependent oxidation of cortisol to cortisone were assessed using rat kidney homogenates as described previously (Thurston et al., 2002Go). Immediately after thawing, a 1 ml aliquot of each FF sample was fractionated using a modification of the method described by Morris and colleagues for the partial purification of endogenous 11{beta}HSD inhibitors in urine (Morris et al., 1992Go; Lo et al., 1997Go; Thurston et al., 2002Go). After conditioning with 20 ml methanol and washing with double-distilled water (DDW), separate C18 Sepak cartridges (Waters, Elstree, UK) were each loaded with a 1 ml volume of a single FF sample. After discarding the loading eluent, each column was sequentially eluted with 1 ml volumes of a stepwise gradient of 0–100% (v/v) methanol (Merck, UK) in DDW. Each sample fraction was collected into a separate borosilicate tube, and those fractions eluted at methanol concentrations >20% (v/v) were evaporated to dryness under nitrogen at 40°C before being re-suspended in 1 ml volumes of 20% (v/v) methanol in DDW.

For the next phase of the assay, homogenates of male rat kidneys, prepared as described previously in our laboratory (Sewell et al., 1998Go; Thompson et al., 2000Go), were used as a consistent source of type 1 11{beta}HSD activity. For each assay, 1 g rat kidney was homogenized in 18 ml hypotonic Tris-EDTA lysis buffer (Rusvai and Naray-Fejes-Toth, 1993Go; Sewell et al., 1998Go; Thompson et al., 2000Go). After addition of 2 ml 1.5 mol/l KCl (Merck) to restore isotonicity, the homogenate was centrifuged at 250 g and the supernatant decanted to a fresh glass tube. From this, 100 µl volumes were transferred to screw-cap glass culture tubes, to each of which was added 600 µl phosphate-buffered saline (PBS) (Life Technologies, UK). In each assay, a triplicate set of tubes was also prepared as assay blanks containing 100 µl of bovine serum albumen solution (1 g/l prepared in PBS) in place of renal homogenate. Each triplicate set of tubes then received: (i) 100 µl PBS (controls and blanks only); (ii) 100 µl of the unfractionated FF sample; (iii) 100 µl of the hydrophilic FF fraction, eluted at 0% (v/v) methanol; or (iv) 100 µl of the hydrophobic FF fraction eluted at 80% (v/v) methanol. After a 30 min pre-incubation at 37°C in a gyratory waterbath, 11{beta}HSD assays were initiated by addition to each tube of 100 µl NADP+ (4 mmol/l in PBS) (Sigma, UK) and 100 µl PBS containing 0.5 µCi [3H]cortisol plus unlabelled cortisol (to a final steroid concentration of 100 nmol/l). Tubes were then incubated in the waterbath for 1 h, after which reactions were terminated by the addition of 2 ml ice-cold chloroform to each tube.

After resolving [3H]cortisol enzyme substrate from the [3H]cortisone by thin-layer chromatography, 11{beta}HSD activities were quantified using a Bioscan 200 TLC radiochromatogram scanner (LabLogic, Sheffield, UK), as described previously (Michael et al., 1997Go; Thurston et al., 2002Go). Type 1 11{beta}HSD activities in the presence of the unfractionated FF samples, the hydrophilic fractions of each FF sample, and the hydrophobic fractions of each FF sample were each expressed relative to the control enzyme activities in the absence of FF or fractions thereof, standardized within each assay to 100%.

Assay of intra-follicular cortisol:cortisone ratios
For 79 FF samples, selected at random from the series of 132 FF samples, total intra-follicular cortisol and cortisone concentrations (i.e. free + protein-bound steroid concentrations) were each measured using previously described radioimmunoassays (RIA) (Moore et al, 1985Go; Wood et al, 1996Go). The cortisol RIA, which had a range of 30–2000 nmol/l, had intra- and inter-assay coefficients of variation (COV) <7 and <8% respectively, over the working range. The cortisone RIA, which had a range of 4–500 nmol/l, had intra- and inter-assay COVs <8 and <10% respectively. The antibodies employed in each assay showed <0.1% cross-reactivity with progesterone and <1% for 17{alpha}-hydroxyprogesterone. The EBSS used to flush follicles contained no detectable cortisol or cortisone. Moreover, EBSS spiked with known concentrations of added cortisol and cortisone (Sigma Chemical Co., Poole, UK) showed parallelism to the respective standard curves and the creation of known cortisol:cortisone ratios in EBSS (by addition of exogenous cortisol and cortisone) produced the expected ratios in the cortisol and cortisone assays (R2 = 0.998, P < 0.0001).

Determination of IVF clinical outcome
Dependent on the success of IVF, between 1 and 3 embryos were transferred to each patient 2 days post-fertilization. (For seven of the 132 treatment cycles, none of the aspirated oocytes were fertilized in vitro and so no embryos were available for transfer.) This study was conducted in a double-blind fashion in so far as levels of the ovarian modulators of type 1 11{beta}HSD and the follicular steroid concentrations were measured without prior knowledge of IVF outcome, and no data were exchanged between the collaborating centres until completion of the study. IVF cycles were classified retrospectively as either conception or non-conception cycles, where clinical pregnancies had been confirmed by observation of a yolk sac plus fetal pole (with evidence of a fetal heart beat) on ultrasonography at 6 weeks post-embryo transfer.

Statistical analyses
Initial statistical testing confirmed that levels of enzyme modulators and follicular steroid concentrations/ratios did not differ significantly between those cycles in which rFSH (Menopure) or hMG (Puregon) had been administered (Table I). Hence, all data were included in subsequent analyses irrespective of the gonaodotrophin used to stimulate follicular development.


View this table:
[in this window]
[in a new window]
 
Table I. Biochemical measurements for IVF cycles stimulated with rFSH (Puregon) or hMG (Menopure)
 
Each data set was subjected to Kolmogorov–Smirnov tests to establish whether or not the data conformed to Gaussian (normal) frequency distributions. The larger data sets, which included all observations irrespective of IVF outcome, did not meet this criterion and could not be normalized by logarithmic transformation. Hence, these data were analysed using non-parametric tests. When data were segregated according to clinical outcome of IVF, the smaller data sets did not deviate significantly from normal data distributions. Hence, mean values for each of the clinical and biochemical parameters were compared between conception versus non-conception cycles using unpaired t-tests.

Correlations between each of the clinical and biochemical parameters were assessed by calculation of Spearman’s rank correlation coefficient (r) or by linear regression analyses [including calculations of Pearson’s correlation coefficients (R2) with a post hoc test for deviation of each linear gradient from zero] as appropriate. Associations between clinical/biochemical parameters and the probability of conception in all 132 IVF cycles, and in the subset of 79 cycles for which follicular cortisol:cortisone ratios had been determined were subsequently assessed by logistic regression analyses.

All statistical analyses were performed using GraphPad Prism2 statistical software (San Diego, CA, USA), and probabilities of <5% were accepted to indicate statistical significance. Data are generally presented as mean ± SEM values, and where data are presented graphically as percentage values, statistical tests were performed on the absolute (non-referenced) data.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Association between clinical parameters and ovarian modulators of type 1 11{beta}HSD
The degree of inhibition of NADP+-dependent 11{beta}HSD activity in rat kidney homogenates exerted by FF samples prior to C18 fractionation (‘whole FF’) was independent of every tested patient/cycle parameter apart from the proportion of oocytes successfully fertilized in vitro (Table II). Similarly, the stimulation of type 1 11{beta}HSD activity by the hydrophilic FF fraction eluted at 0% (v/v) methanol was generally independent of the tested patient/cycle parameters, but was weakly inversely correlated to the plasma LH concentration (Table II). In contrast, the degree of inhibition of renal NADP+-dependent cortisol oxidation by the hydrophobic FF fraction eluted at 80% (v/v) methanol was strongly inversely correlated to the number of rFSH/hMG ampoules required to stimulate follicular development (P < 0.01) and positively correlated with the plasma estradiol concentration 2 days prior to oocyte collection (P < 0.01) (Table II).


View this table:
[in this window]
[in a new window]
 
Table II. Relationships of clinical parameters for 132 IVF cycles to the in-vitro modulation of NADP+-dependent 11{beta}HSD activities by FF prior to fractionation (‘whole FF’), by the hydrophilic fraction of the FF samples [eluted at 0% (v/v) methanol] and by the hydrophobic fraction of the FF samples [eluted at 80% (v/v) methanol]
 
Within the clinical parameters tested, both plasma LH and E2 concentrations were inversely correlated with the number of rFSH/hMG ampoules required to stimulate adequate follicular development (r = –0.198, P < 0.03 and r = –0.421, P < 0.0001, respectively). However, the direct correlation between plasma LH and E2 concentrations was not statistically significant (r = 0.155, P > 0.07).

The effects of whole FF samples and of the hydrophilic and hydrophobic fractions thereof on NADP+-dependent 11{beta}HSD activities did not differ significantly between FF samples obtained from patients with differing primary causes of infertility (ANOVA: P > 0.5 for whole FF, P > 0.2 for the hydrophilic FF fraction and P > 0.3 for the hydrophobic FF fraction).

The inhibition of enzyme activity as exerted by the whole FF prior to fractionation was directly correlated with the inhibition exerted by the hydrophobic fractions of those same FF samples (r = 0.455, P < 0.001) but was inversely correlated with the stimulatory action of the corresponding hydrophilic FF fractions (r = –0.523, P < 0.001). In addition, there was a significant inverse correlation between the degree to which the hydrophobic FF fractions inhibited and hydrophilic FF fractions stimulated the NADP+-dependent oxidation of cortisol, respectively (r = –0.366, P < 0.001).

Association between IVF outcome and ovarian modulators of type 1 11{beta}HSD
The numbers of oocytes recovered for IVF were independent of the effects of FF samples prior to fractionation and of the hydrophilic and hydrophobic fractions of the FF samples on the NADP+-dependent oxidation of cortisol in rat kidney homogenates (Table II). While the proportion of oocytes successfully fertilized in vitro was inversely correlated with the degree of enzyme inhibition exerted by the FF samples prior to fractionation, fertilization rates did not correlate with the degree of modulation of 11{beta}HSD activities by the hydrophilic or the hydrophobic fractions of these FF samples (Table II).

In the logistic regression analysis for all 132 IVF cycles, the effects of whole FF and of the resolved hydrophilic and hydrophobic fractions on type 1 11{beta}HSD activity were each strongly associated with the probability of conception (Table III). Specifically, increasing enzyme inhibition of NADP+-dependent cortisol oxidation by either whole FF or the hydrophobic fraction thereof was associated with increased probability of conception (P < 0.01), whereas increasing stimulation by the hydrophilic FF fraction eluted at 0% (v/v) methanol was associated with a significant decrease in the probability of conception (P < 0.001) (Table III). Both the logistic regression analysis (Table III) and the comparisons of each patient/cycle parameter between conception versus non-conception cycles (Table IV) demonstrated that the probability of conception was inversely correlated to the number of rFSH/hMG ampoules administered (P < 0.01), but increased with the plasma E2 concentration (P < 0.02).


View this table:
[in this window]
[in a new window]
 
Table III. Logistic regression analysis for the probability of conception in 132 IVF treatment cycles
 

View this table:
[in this window]
[in a new window]
 
Table IV. Clinical parameters for 132 IVF cycles segregated by IVF outcome
 
In the second logistic regression analysis (performed on the subset of 79 IVF cycles for which follicular cortisol and cortisone concentrations were assessed), only the associations with the effects of the resolved hydrophilic and hydrophobic fractions of FF on type 1 11{beta}HSD activity achieved statistical significance (Table V).


View this table:
[in this window]
[in a new window]
 
Table V. Logistic regression analysis for the probability of conception in 79 IVF treatment cycles
 
In accordance with the findings of both logistic regression analyses, samples of whole FF obtained from conception cycles exerted a significantly greater inhibition of type 1 11{beta}HSD activity than did samples from non-conception cycles (P < 0.002; Figure 1A). While the hydrophilic fractions of FF from conception cycles exerted less stimulation of NADP+-dependent cortisol metabolism than did non-conception samples (P < 0.0001; Figure 1B), the hydrophobic fractions of FF from conception cycles had a more pronounced inhibitory action than did the corresponding fractions from FF aspirated in non-conception cycles (P < 0.002; Figure 1C).





View larger version (98K):
[in this window]
[in a new window]
 
Figure 1. Relationship of clinical outcome for 132 IVF cycles to (A) the inhibition of NADP+-dependent 11{beta}HSD activities by whole FF samples, (B) the stimulation of NADP+-dependent 11{beta}HSD activities by the hydrophilic FF fractions [eluted at 0% (v/v) methanol], and (C) the inhibition of NADP+-dependent 11{beta}HSD activities by the hydrophobic FF fractions [eluted at 80% (v/v) methanol]. In each panel, the vertical bars represent the mean (± SEM) inhibition exerted by 88 FF samples obtained from non-conception cycles (open bar) and 44 FF samples obtained from conception cycles (filled bar).

 
Correlation between intra-follicular cortisol:cortisone ratios and ovarian modulators of type 1 11{beta}HSD
The ratios of cortisol:cortisone in each of the 79 FF samples tested were independent of the inhibition of NADP+-dependent cortisol oxidation by those FF samples prior to fractionation (P > 0.05) and of the stimulation of type 1 11{beta}HSD activity exerted by the hydrophilic fractions of those samples (P > 0.05) (Table VI). However, the intra-follicular cortisol:cortisone ratios were directly correlated with the inhibition of type 1 11{beta}HSD activity exerted by the hydrophobic fractions of each sample (P < 0.01; Table VI; Figure 2).


View this table:
[in this window]
[in a new window]
 
Table VI. Relationships of clinical parameters for 79 IVF cycles to the intra-follicular cortisol concentration, the intra-follicular cortisone concentration and the ratio of cortisol:cortisone in FF
 


View larger version (33K):
[in this window]
[in a new window]
 
Figure 2. Correlation between intra-follicular cortisol:cortisone ratios and the degree of inhibition of NADP+-dependent 11{beta}HSD activities in renal homogenates by the hydrophobic FF fractions [eluted at 80% (v/v) methanol]. Open symbols indicate FF samples obtained from 61 non-conception cycles; filled symbols indicate FF samples obtained from 18 conception cycles.

 
The total concentrations of both cortisol and cortisone in FF were independent of any of the patient and cycle parameters assessed in the current study, including the levels of enzyme modulators in FF (Table VI).

Association between IVF outcome and intra-follicular cortisol:cortisone ratios
The numbers of oocytes retrieved and the proportions of those oocytes successfully fertilized in vitro were each independent of the intra-follicular cortisol:cortisone ratios and of the total concentrations of cortisol or cortisone in FF (Table VI). In the second logistic regression analysis, both the intra-follicular cortisol:cortisone ratios and the concentrations of cortisol were associated significantly with the probability of conception (Table V). Intra-follicular cortisol:cortisone ratios and total cortisol concentrations were both significantly elevated in FF obtained from conceptions cycles relative to samples from non-conception cycles (P < 0.0001 and P < 0.006 respectively; Figure 3). All 18 FF samples obtained from patients that achieved pregnancy through IVF had cortisol:cortisone ratios >=11.4, whereas all 61 samples with cortisol:cortisone ratios <11.4 were obtained from non-conception cycles.





View larger version (88K):
[in this window]
[in a new window]
 
Figure 3. Relationship of IVF cycle outcome to (A) intra-follicular cortisol concentrations, (B) intra-follicular cortisone concentrations and (C) intra-follicular cortisol:cortisone ratios. In each panel, the vertical bars represent the mean (± SEM) inhibition exerted by 61 FF samples obtained from non-conception cycles (open bar) and 18 FF samples obtained from conception cycles (filled bar).

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This study has confirmed that ovarian FF contains compounds that can modulate inactivation of cortisol by the NADP+-dependent, type 1 isoform of 11{beta}HSD. The principal finding of this study was that conception following IVF is associated with low levels of the hydrophilic ovarian stimuli to type 1 11{beta}HSD and with high levels of the hydrophobic ovarian enzyme inhibitors. Pregnancy is apparently favoured by a profile of enzyme modulators in FF that would be expected to decrease ovarian 11{beta}HSD activities, whereas conception through IVF appears to be precluded by a profile of enzyme modulators that would be expected to increase ovarian 11{beta}HSD activities. These data are consistent with our previous findings linking high ovarian 11{beta}HSD activities with failure to conceive by IVF (Michael et al., 1993Go; 1995Go).

The type 1 isoform of 11{beta}HSD appears to be the major enzyme isoform expressed in luteinizing granulosa cells isolated from pre-ovulatory ovarian follicles (Michael et al., 1997Go; Smith et al., 1997Go; Tetsuka et al., 1997Go; 1999aGo,bGo). This has lead us to speculate that the modulators of type 1 11{beta}HSD activity might regulate cortisol–cortisone inter-conversion within the follicle (Thurston et al., 2002Go). In support of this proposal, we now report a significant, direct correlation between intra-follicular cortisol:cortisone ratios and levels of the hydrophobic ovarian inhibitors of type 1 11{beta}HSD activity. This correlation suggests that the net oxidation of cortisol to cortisone in ovarian follicles may be influenced by the extent to which hydrophobic components of FF inhibit NADP+-dependent 11{beta}HSD activity within the ovarian follicle in vivo.

Inhibition of type 1 11{beta}HSD activity by the hydrophobic FF fractions was inversely correlated with the number of rFSH/hMG ampoules required for controlled ovarian hyperstimulation, and was directly proportional to the plasma E2 concentration 2 days prior to follicular aspiration. These findings could indicate that levels of enzyme inhibitors in FF are dependent on exposure to gonadotrophins or E2. An alternative explanation would be that the level of 11{beta}HSD inhibitors in the hydrophobic fraction of FF is an index of the ovarian response to pharmacological stimulation, in much the same way as the number of rFSH/hMG ampoules administered or the plasma E2 concentration. Since these cycle parameters are also associated with the clinical outcome of IVF outcome, levels of the ovarian modulators of NADP+-dependent 11{beta}HSD activity may not be independent predictors of IVF outcome. However, the fact that levels of enzyme modulators are more strongly associated with cycle outcome than any of the examined clinical/endocrine variables does suggest that their correlation with conception is not solely reliant on established clinical relationships.

The findings reported herein raise the possibility that in making direct measurements of ovarian 11{beta}HSD activities in vitro, potentially predictive relationships to the clinical outcome of IVF may depend on the ability of a particular experimental protocol to reflect levels of enzyme modulators in FF. In all of our studies to date, human granulosa-lutein cells have been routinely stored at 4°C for up to 3 days in FF prior to isolation, hence maximizing exposure to the ovarian modulators of type 1 11{beta}HSD accumulated in FF. Moreover, after isolation from the FF, granulosa-lutein cells have been routinely maintained in serum-supplemented culture for 3 days prior to the measurement of ovarian 11{beta}HSD activities. This provides cells with the opportunity to synthesize a paracrine inhibitor of 11{beta}HSD in vitro (Michael et al., 1996Go; Thurston et al., 2003Go) such that ovarian 11{beta}HSD activities would be lowest in those granulosa-lutein cells with the greatest capacity to synthesize the endogenous 11{beta}HSD inhibitor(s). In contrast to our own studies, others have assessed ovarian 11{beta}HSD activities in human granulosa cells that had been neither stored in FF nor cultured prior to measurement of enzyme activities. Based on our current findings, we suggest that O’Shaughnessy et al. (1997)Go and Thomas et al. (1998)Go found no significant correlation between ovarian 11{beta}HSD activities and IVF outcome due to procedural differences that prevented the enzyme assays from reflecting production of the ovarian 11{beta}HSD modulators either in FF or in the granulosa cells as they underwent functional luteinization in vitro.

In previous studies, statistical assessments of the relationship between IVF outcome and the ratio of cortisol:cortisone in FF have been influenced by relatively low sample numbers (Michael et al., 1999Go; Yding Andersen et al., 1999Go; Keay et al., 2002Go). To the best of our knowledge, the present study is the first to feature data for FF samples from more than 50 different patients. In this study, we have found a highly significant association between high intra-follicular cortisol:cortisone ratios and conception following IVF. While this finding is consistent with our previous study, intra-follicular cortisol:cortisone ratios were higher than in our previous report (Michael et al., 1999Go). Moreover, in our prior publication, intra-follicular cortisol concentrations were lower in FF from conception than non-conception cycles, a finding diametrically opposed to the data reported herein. The resolution to this apparent contradiction probably lies in the fact that in the previous study, low sample numbers limited the power of statistical tests to <80%. In the present study, assessment of FF samples from a total of 79 patients increased the statistical power for all comparisons to >99%. Hence, we would propose that our previous data (obtained using the same RIAs as the current study, and featuring patient samples from the same assisted reproduction unit) may have been prone to sampling error, which is not a significant consideration for the study reported herein.

Excluding a random association, there are three possible interpretations of the direct correlation between the intra-follicular cortisol:cortisone ratio and the probability of conception by IVF. First, increased concentrations of cortisol may be required within the ovarian follicle for proper follicular maturation and/or developmental competence of the oocyte. Certainly, the higher cortisol:cortisone ratios in FF samples from conception cycles were attributable to significantly increased concentrations of cortisol in the face of relatively constant cortisone concentrations. Previous reports have commented upon the significant increase in intra-follicular cortisol concentrations associated with follicular and oocyte maturation (Fateh et al., 1989Go; Yding Andersen and Hornness, 1994Go; Harlow et al., 1997Go). Hence, high intra-follicular cortisol concentrations, reflected in an elevated cortisol:cortisone ratio, may simply indicate those IVF cycles in which appropriate follicular and/or oocyte maturation has been successfully induced.

The second possible explanation, recently suggested by Yding Andersen (2002)Go, is that the cortisol present in FF may be required to exert positive effects on events within the oviduct. While effects of cortisol derived from FF on the oviduct may have a bearing on natural conception and in such techniques as gamete intra-Fallopian transfer and intra-uterine insemination, this cannot account for our present findings since the FF was not available to act in the Fallopian tube.

The third explanation of the association between high intra-follicular cortisol:cortisone ratios and conception is that this ratio provides an in-vivo index of ovarian 11{beta}HSD activities, which, in turn, reflect the levels of paracrine/autocrine modulators of type 1 11{beta}HSD activity in FF. A high ratio of cortisol:cortisone in FF could reflect either a decrease in the net oxidation of cortisol to cortisone, or an increase in the net reduction of cortisone to cortisol. Previous studies by Hillier and colleagues have assumed that oxidation of cortisol in follicular granulosa cells is catalysed by type 2 11{beta}HSD, whereas reduction of cortisone to cortisol is catalysed in luteinizing cells by type 1 11{beta}HSD (Tetsuka et al., 1997Go; Hillier and Tetsuka, 1998Go; Yong et al., 2000Go). In the current study, the cortisol:cortisone ratios increased with the extent to which the hydrophobic fractions of FF were able to inhibit NADP+-dependent oxidation of cortisol by type 1 11{beta}HSD in rat kidney homogenates. This observation is consistent with the view that in ovarian follicles, cortisol is the major substrate for ovarian isoforms of 11{beta}HSD and cortisone is the major product (Michael et al., 1997Go; Yding Andersen et al., 1999Go). While we would concur with previous reports that luteinizing human granulosa cells express type 1 11{beta}HSD with no detectable expression of type 2 11{beta}HSD (Michael et al., 1997Go; Smith et al., 1997Go; Tetsuka et al., 1997Go; 1999aGo,bGo; Ricketts et al., 1998Go), our current data suggest that either type 1 11{beta}HSD acts predominantly as an 11{beta}-dehydrogenase enzyme in mature ovarian follicles or that the activity of an alternative, as yet uncloned, oxidative isoform of 11{beta}HSD predominates in the latter stage of follicular development.

In conclusion, an increased probability of conception following IVF appears to be associated with low levels of hydrophilic ovarian stimuli to type 1 11{beta}HSD, high levels of ovarian 11{beta}HSD inhibitors and high intra-follicular cortisol:cortisone ratios, each of which would predict low levels of ovarian cortisol oxidation, previously associated with the success of IVF (Michael et al., 1993Go; 1995Go). Conversely, high levels of the hydrophilic ovarian 11{beta}HSD stimuli, low levels of hydrophobic 11{beta}HSD inhibitors and decreased intra-follicular cortisol:cortisone ratios, each indicative of increased ovarian 11{beta}-dehydrogenase activities, are characteristic of unsuccessful IVF cycles. We are currently pursuing the molecular nature of these paracrine/autocrine ovarian modulators of cortisol–cortisone inter-conversion by type 1 11{beta}HSD.


    Acknowledgements
 
The authors wish to thank staff at the Cardiff Assisted Reproduction Unit (University Hospital of Wales, Cardiff, UK) for providing the clinical samples used in this study, Mrs C.Glenn (Regional Endocrine Unit, Southampton, UK) for performing the cortisol and cortisone assays, and Dr N.Prathalingam for help with statistical analyses. This work was supported by project grant 052970 from the Wellcome Trust and by funding from Freemedic plc.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Agarwal, A.K., Monder, C., Eckstein, B. and White, P.C. (1989) Cloning and expression of rat cDNA encoding corticosteroid 11{beta}-dehydrogenase. J. Biol. Chem., 264, 18939–18943.[Abstract/Free Full Text]

Agarwal, A.K., Mune, T., Monder, C. and White, P.C. (1994) NAD+-dependent isoform of 11{beta}-hydroxysteroid dehydrogenase. Cloning and characterization cDNA from sheep kidney. J. Biol. Chem., 269, 25959–25962.[Abstract/Free Full Text]

Albiston, A.L., Obeyesekere, V.R., Smith, R.E. and Krozowski, Z.S. (1994) Cloning and tissue distribution of the human 11{beta}-hydroxysteroid dehydrogenase type 2 enzyme. Mol. Cell. Endocrinol., 105, R11–R17.[CrossRef][ISI][Medline]

Benediktsson, R., Yau, J.L.W., Low, S.C., Brett, L.P., Cooke, B.E., Edwards, C.R.W and Seckl, J.R. (1992) 11{beta}-hydroxysteroid dehydrogenase in the rat ovary: high expression in the oocyte. J. Endocrinol., 135, 53–58.[Abstract]

Brown, R.W., Chapman, K.E., Edwards, C.R.W. and Seckl, J.R. (1993) Human placental 11{beta}-hydroxysteroid dehydrogenase: evidence for and partial purification of a distinct NAD-dependent iosform. Endocrinology, 132, 2614–2621.[Abstract]

Fateh, M., Ben-Rafael, Z., Benadiva, C.A., Mastroianni, L., Jr and Flickinger, G.L. (1989) Cortisol levels in human follicular fluid. Fertil. Steril., 51, 538–541.[ISI][Medline]

Harlow, C.R., Jenkins, J.M. and Winston, R.M.L. (1997) Increased follicular fluid total and free cortisol during the luteinizing hormone surge. Fertil. Steril., 68, 48–53.[CrossRef][ISI][Medline]

Hillier, S.G. and Tetsuka, M. (1998) An anti-inflammatory role for glucocorticoids in the ovaries? J. Reprod. Immunol., 39, 21–27.[CrossRef][ISI][Medline]

Keay, S.D., Harlow, C.R., Wood, P.J., Jenkins, J.M. and Cahill, D.J. (2002) Higher cortisol:cortisone ratios in the preovulatory follicle of completely unstimulated IVF cycles indicate oocytes with increased pregnancy potential. Hum. Reprod., 17, 2410–2414.[Abstract/Free Full Text]

Koteletsev, Y., Seckl, J.R. and Mullins, J.J. (1999) 11{beta}-hydroxysteroid dehydrogenases: key modulators of glucocorticoid action in vivo. Curr. Opin. Endocrinol. Diabetes, 6, 191–198.[CrossRef]

Lakshmi, V. and Monder, C. (1988) Purification and characterization of the corticosteroid 11{beta}-dehydrogenase component of the rat liver 11{beta}-hydroxysteroid dehydrogenase complex. Endocrinology, 123, 2390–2398.[Abstract]

Lo, Y.H., Sheff, M.F., Latif, S.A., Ribiero, C., Silver, H., Brem, A. and Morris, D.J. (1997) Kidney 11{beta}-HSD2 is inhibited by glycyrrhetinic acid-like factors in human urine. Hypertension, 29, 500–505.[Abstract/Free Full Text]

Michael, A.E., Gregory, L., Walker, S.M., Antoniw, J.W., Shaw, R.W., Edwards,C.R.W. and Cooke, B.A. (1993) Ovarian 11{beta}-hydroxysteroid dehydrogenase: potential predictor of conception by in-vitro fertilization and embryo transfer. Lancet, 342, 711–712.[ISI][Medline]

Michael, A.E., Gregory, L., Piercy, E.C., Walker, S.M., Shaw, R.W. and Cooke, B.A. (1995) Ovarian 11{beta}-hydroxysteroid dehydrogenase activity is inversely related to the outcome of in vitro fertilization-embryo transfer treatment cycles. Fertil. Steril., 64, 590–598.[ISI][Medline]

Michael, A.E., Gregory, L., Thaventhiran, L., Antoniw, J.W. and Cooke, B.A. (1996) Follicular variation in ovarian 11{beta}-hydroxysteroid dehydrogenase (11{beta}HSD) activities: evidence for the paracrine inhibition of 11{beta}HSD in human granulosa-lutein cells. J. Endocrinol., 148, 419–425.[Abstract]

Michael, A.E., Evagelatou, M., Norgate, D.P., Clarke, R.J., Antoniw, J.W., Stedman, B., Brennan, A., Welsby, R., Bujalska, I., Stewart, P.M. et al. (1997) Isoforms of 11{beta}-hydroxysteroid dehydrogenase in human granulosa-lutein cells. Mol. Cell. Endocrinol., 132, 43–52.[CrossRef][ISI][Medline]

Michael, A.E., Collins, T.D., Norgate, D.P., Gregory, L., Wood, P.J. and Cooke, B.A. (1999) Relationship between ovarian cortisol:cortisone ratios and the clinical outcome of in vitro fertilization and embryo transfer (IVF-ET). Clin. Endocrinol., 51, 535–540.[CrossRef][ISI][Medline]

Monder, C. (1991) Corticosteroids, receptors, and the organ-specific functions of 11{beta}-hydroxysteroid dehydrogenase. FASEB J., 5, 3047–3054.[Abstract/Free Full Text]

Moore, A., Aitken, R., Burke, C., Gaskell, S., Groom, G., Holder, G., Selby, C. and Wood, P. (1985) Cortisol assay: guidelines for the provision of a clinical biochemistry service. Ann. Clin. Biochem., 22, 435–454.[ISI][Medline]

Morris, D.J., Semafuko, W.E.B., Latif, S.A., Vogel, B., Grimes, C.A. and Sheff, M.F. (1992) Detection of glycyrrhetinic acid-like factors (GALFs) in human urine. Hypertension, 20, 356–360.[Abstract]

O’Shaughnessy, A., Muneyyirci-Delale, O., Nacharaju, V.L., Baker, A., Benfer, K., Summers, D., Hoover, L., Carlson, J., Hourani, C. & Lurie, D. (1997) 11{beta}-Hydroxysteroid dehydrogenase enzyme activity in granulosa cells derived from ovarian follicles of in vitro fertilization patients. Early Pregnancy, 3, 183–189.[Medline]

Quinkler, M., Oelkers, W. and Diederich, S. (2001) Clinical implications of glucocorticoid metabolism by 11{beta}-hydroxysteroid dehydrogenase in target tissues. Eur. J. Endocrinol., 144, 87–97.[ISI][Medline]

Ricketts, M.L., Verhaeg, J.M., Bujalska, I., Howie, A.J., Rainey, W.E. and Stewart, P.M. (1998) Immunohistochemical localization of type 1 11{beta}-hydroxysteroid dehydrogenase in human tissues. J. Clin. Endocrinol. Metab., 83, 1325–1335.[Abstract/Free Full Text]

Rusvai, E and Naray-Fejes-Toth, A. (1993) A new isoform of 11{beta}-hydroxysteroid dehydrogenase in aldosterone target cells. J. Biol. Chem., 268, 10717–10720.[Abstract/Free Full Text]

Seckl, J.R. and Walker, B.R. (2001) 11{beta}-hydroxysteroid dehydrogenase type 1 – a tissue-specific amplifier of glucocorticoid action. Endocrinology, 142, 1371–1376.[Abstract/Free Full Text]

Sewell, K.J., Shirley, D.G., Michael, A.E., Thompson, A., Norgate, D.P. and Unwin, R.J. (1998) Inhibition of renal 11{beta}-hydroxysteroid dehydrogenase in vivo by carbenoxolone in the rat and its relationship to sodium excretion. Clin. Sci., 95, 435–443.[ISI][Medline]

Smith, M.P., Keay, S.D., Hall, L., Harlow, C.R. and Jenkins, J.M. (1997) The detection and confirmation of 11{beta}-hydroxysteroid dehydrogenase type 1 transcripts in human luteinized granulosa cells using RT–PCR and plasmid pUC18. Mol. Hum. Reprod., 3, 651–654.[Abstract]

Smith, M.P., Mathur, R.S., Keay, S.D., Hall, L., Hull, M.G.R. and Jenkins, J.M. (2000) Periovulatory human oocytes, cumulus cells, and ovarian leukocytes express type 1 but not type 2 11{beta}-hydroxysteroid dehydrogenase RNA. Fertil. Steril., 73, 825–830.[CrossRef][ISI][Medline]

Stewart, P.M. and Krozowski, Z.S. (1999) 11{beta}-hydroxysteroid dehydrogenase. Vitam. Horm., 57, 249–324.[ISI][Medline]

Tannin, G.M., Agarwal, A.K., Monder, C., New, M.I. and White, P.C. (1991) The human gene for 11{beta}-hydroxysteroid dehydrogenase. Structure, tissue distribution and chromosomal localization. J. Biol. Chem., 266, 16653–16658.[Abstract/Free Full Text]

Tetsuka, M., Thomas, F.J., Thomas, M.J., Anderson, R.A., Mason, J.I. and Hillier, S.G. (1997) Differential expression of messenger ribonucleic acids encoding 11{beta}-hydroxysteroid dehydrogenase types 1 and 2 in human granulosa cells. J. Clin. Endocrinol. Metab., 82, 2006–2009.[ISI][Medline]

Tetsuka, M., Haines, L.C., Milne, M., Simpson, G.E. and Hillier, S.G. (1999a) Regulation of 11{beta}-hydroxysteroid dehydrogenase type 1 gene expression by LH and interleukin-1{beta} in cultured rat granulosa cells. J. Endocrinol., 163, 417–423.[Abstract/Free Full Text]

Tetsuka, M., Milne, M., Simpson, G.E. and Hillier, S.G. (1999b) Expression of 11{beta}-hydroxysteroid dehydrogenase, glucocorticoid receptor, and mineralocorticoid receptor genes in rat ovary. Biol. Reprod., 60, 330–335.[Abstract/Free Full Text]

Thomas, F.J., Thomas, M.J., Tetsuka, M., Mason, J.I. and Hillier, S.G. (1998) Corticosteroid metabolism in human granulosa-lutein cells. Clin. Endocrinol., 48, 509–513.[CrossRef][ISI][Medline]

Thompson, A., Bailey, M.A., Michael, A.E. and Unwin, R.J. (2000) Effects of changes in dietary intake of sodium and potassium and of metabolic acidosis on 11{beta}-hydroxysteroid dehydrogenase activities in rat kidney. Exp. Nephrol., 8, 44–51.[CrossRef][ISI][Medline]

Thurston, L.M., Norgate, D.P., Jonas, K.M., Chandras, C., Kloosterboer, H.J., Cooke, B.A. and Michael, A.E. (2002) Ovarian modulators of 11{beta}-hydroxysteroid dehydrogenase (11{beta}HSD) activity in follicular fluid from gonadotrophin-stimulated assisted conception cycles. Reproduction, 124, 801–812.[Abstract/Free Full Text]

Thurston, L.M., Chin, E., Jonas, K.C., Bujalska, I.J., Stewart, P.M., Abayasekara, D.R.E. and Michael, A.E. (2003) Expression of 11{beta}-hydroxysteroid dehydrogenase (11{beta}HSD) proteins in luteinizing human granulosa-lutein cells. J. Endocrinol., in press.

Waddell, B.J., Benediktsson, R. and Seckl, J.R. (1996) 11{beta}-hydroxysteroid dehydrogenase type 2 in the rat corpus luteum: induction of messenger ribonucleic acid expression and bioactivity coincident with luteal regression. Endocrinology, 137, 5386–5391.[Abstract]

White, P.C., Mune, T. and Agarwal, A.K. (1997) 11{beta}-Hydroxysteroid dehydrogenase and the syndrome of apparent mineralcorticoid excess. Endocr. Rev., 8, 135–156.[CrossRef]

Wood, P.J., Glenn, C. and Donovan, S.J. (1996) A simple RIA for serum cortisone without preliminary steroid extraction. J. Endocrinol., 148 (Suppl.), P319.

Yding Andersen, C. (2002) Possible new mechanism of cortisol action in female reproductive organs: physiological implications of the free hormone hypothesis. J. Endocrinol., 173, 211–217.[Abstract/Free Full Text]

Yding Andersen, C. and Hornnes, P. (1994) Intrafollicular concentrations of free cortisol close to follicular rupture. Hum. Reprod., 9, 1944–1949.[Abstract]

Yding Andersen, C., Morineau, G., Fukuda, M., Westergaard, L.G., Ingerslev, H.J., Fiet, J. and Byskov, A.G. (1999) Assessment of the follicular cortisol:cortisone ratio. Hum. Reprod., 14, 1563–1568.[Abstract/Free Full Text]

Yong, P.Y.K., Thong, K.J., Andrew, R., Walker, B.R. and Hillier, S.G. (2000) Development-related increase in cortisol biosynthesis by human granulosa cells. J. Clin. Endocrinol. Metab., 85, 4728–4733.[Abstract/Free Full Text]

Yong, P.Y.K., Harlow, C., Thong, K.J. and Hillier, S.G. (2002) Regulation of 11{beta}-hydroxysteroid dehydrogenase type 1 gene expression in human ovarian surface epithelial cells by interleukin-1. Hum. Reprod., 17, 2300–2306.[Abstract/Free Full Text]

Zhou, M.-Y., Gomez-Sanchez, E.P., Cox, D.L., Cosby, D. and Gomez-Sanchez, C.E. (1995) Cloning, expression and tissue distribution of the rat nicotinamide adenine dinucleotide-dependent isoform of 11{beta}-hydroxysteroid dehydrogenase. Endocrinology, 136, 3729–3734.[Abstract]

Submitted on September 19, 2002; accepted on May 1, 2003.





This Article
Abstract
FREE Full Text (PDF )
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (3)
Request Permissions
Google Scholar
Articles by Thurston, L.M.
Articles by Michael, A.E.
PubMed
PubMed Citation
Articles by Thurston, L.M.
Articles by Michael, A.E.