Local levonorgestrel regulation of androgen receptor and 17{beta}-hydroxysteroid dehydrogenase type 2 expression in human endometrium

Kevin A. Burton1, Teresa A. Henderson1, Stephen G. Hillier1, J.Ian Mason1, Fouad Habib2, Robert M. Brenner3 and Hilary O.D. Critchley1,4

1 Centre for Reproductive Biology, Chancellor’s Building, 49 Little France Crescent, Edinburgh EH16 4SB, 2 Department of Oncology, Western General Hospital, Crewe Road South, Edinburgh EH4 2XU, UK and 3 Division of Reproductive Sciences, Oregon Regional Primate Centre, 505 NW 185th Avenue, Beaverton, OR 97006, USA

4 To whom correspondence should be addressed. e-mail: Hilary.Critchley{at}ed.ac.uk


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: The levonorgestrel-releasing intrauterine system (LNG-IUS) is a highly effective contraceptive. However, unscheduled breakthrough bleeding (BTB), leads to discontinuation in a proportion of users. The LNG-IUS down-regulates endometrial progesterone and estrogen receptors and this may play a role in the mechanism responsible for BTB. LNG is an androgenic progestogen and so we examined the regulation of the androgen receptor (AR) in endometrium exposed to intrauterine LNG. Furthermore, as the enzyme 17{beta}-hydroxysteroid dehydrogenase type 2 (17{beta}HSD2) regulates intracellular levels of estrogens, progestins and androgens, we evaluated the changes in expression of 17{beta}HSD2 in the same tissue endometrial samples. METHODS: Immunohistochemistry and real time quantitative RT–PCR were used to compare protein and mRNA expression of AR and 17{beta}HSD2 in endometrial biopsies from women with normal menstrual cycles and those using a LNG-IUS. RESULTS: Immunohistochemistry showed that AR and 17{beta}HSD2, which were immunolocalized to the stroma and glands of endometrium respectively, were both suppressed by LNG-IUS treatment, though moderate staining of 17{beta}HSD2 was evident 1 month after insertion of the LNG-IUS. AR mRNA expression was down-regulated in LNG-exposed endometrium when compared with the proliferative phase of the menstrual cycle. 17{beta}HSD2 mRNA was significantly increased 3 months (but not 6–12 months) after LNG-IUS insertion. CONCLUSIONS: Endometrial intracellular estradiol levels would have been suppressed by 17{beta}HSD2 during the first few, but not the later, months of LNG-IUS action, and the lowered endometrial estradiol level may contribute to the frequent BTB evident in the early months of LNG-IUS use. The subsequent decline in 17{beta}HSD2 would lead to elevated local intracellular estradiol in the later months, when the BTB tends to subside. The suppression of AR by the LNG-IUS may also play a role in BTB, as elevated AR has been associated with amenorrhoea.

Key words: androgen receptor/endometrium/17{beta}-hydroxysteroid dehydrogenase type 2/levonorgestrel


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Levonorgestrel (LNG) is a potent progestagen that also has considerable binding affinity for the androgen receptor (AR). The levonorgestrel-releasing intrauterine system (LNG-IUS; Schering HC, UK) releases 20 µg LNG locally to the endometrium over 24 h (Luukkainen et al., 1990Go). Patient acceptability is high but a proportion of patients will discontinue therapy due to unscheduled breakthrough bleeding (BTB) (Findlay, 1996Go). The mechanism responsible for BTB remains poorly understood but is likely to involve angiogenic factors, matrix metalloproteinase activation, and altered haemostatic factors that result in increased endometrial vessel fragility (Hickey et al., 2000Go). Such a pathogenesis may be directly related to the high dose of local LNG exposure. Alternatively, indirect actions of LNG may be responsible, such as its influence upon endometrial sex steroid receptor expression, or ligand availability in the endometrium.

The temporal and spatial distributions of sex steroid receptors such as estrogen receptor (ER), progestin receptor (PR) and AR in endometrium have been established. AR has been spatially localized to the endometrial stroma with up-regulation in the estrogen-dominated proliferative phase followed by down-regulation in the late secretory phase (Mertens et al., 2001Go; Slayden et al., 2001Go). ER and PR are also up-regulated in glandular and stromal cells in the proliferative phase. In the secretory phase, a down-regulation of ER is noted in both glands and stroma. PR is also down-regulated in the glands but receptor expression persists in stromal tissues (Garcia et al., 1988Go; Lessey et al., 1988Go; Snijders et al., 1992Go; Critchley et al., 1993Go).

LNG, delivered locally by the LNG-IUS, is known to down-regulate ER and PR expression in the human endometrium (Critchley et al., 1998Go). This alteration in sex steroid receptor expression may affect endometrial cytokine release (Jones et al., 2000Go). The cellular and molecular expression of the androgen receptor in LNG-IUS-treated endometrium has not been previously reported and AR may play an important role in endometrial physiology (Slayden et al., 2001Go; Brenner et al., 2002Go).

In reproductive tissues, the local actions of sex steroids are modulated by hydroxysteroid dehydrogenases (HSD). The human 17{beta}HSD family has six known members, each being a separate gene from a different chromosome with distinct properties in terms of substrates and redox direction. The type 2 isoform (17{beta}HSD2) has a major role in the inactivation of estradiol to estrone. 17{beta}HSD2 is also the enzyme responsible for converting androgens to less potent forms while also activating progesterone. The 17{beta}HSD2 isoform is expressed in endometrial glandular epithelium, and is up-regulated by progesterone (Maentausta et al., 1993Go). Its activity decreases when progesterone concentrations decline pre-menstrually or following antiprogestin administration (Maentausta et al., 1993Go; Mustonen et al., 1998Go). LNG binds avidly to the AR (Kloosterboer et al., 1988Go). The availability of other androgenic ligands to bind to the AR may be influenced by the availability of the enzyme 17{beta}HSD2.

We have thus evaluated expression of both AR and 17{beta}HSD2 in a longitudinal study of endometrial samples from naturally cycling women and from women treated with LNG-IUS over a 12 month time frame. The results provide new data on the expression patterns of these two molecular components that may be relevant to the bleeding patterns seen during intrauterine treatment with the LNG-IUS.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Human endometrial samples were obtained from three different patient groups. Institutional ethical approval was obtained and all women provided written informed consent.

The first group (‘control’ group) consisted of fertile women (n = 24) with regular menstrual cycles lasting between 25 and 35 days who had not been using hormonal preparations for the preceding 3 months. Samples were obtained at the time of hysterectomy for benign indications. Subjects with significant uterine pathology (for example, fibroids) were excluded. Endometrial tissue was collected from women during the proliferative phase (n = 6), early secretory (n = 6), mid secretory (n = 6) and late secretory phase (n = 6) of the menstrual cycle. After the uterus had been removed, a full thickness biopsy of the uterus was obtained, extending from the lumen to the myometrial layer that included the superficial and basal endometrium in addition to the myometrium. The stage of the menstrual cycle was consistent with the patient’s reported last menstrual period and was confirmed by histological dating, according to the criteria of Noyes et al. (1950Go). Furthermore, serum samples collected at the time of endometrial tissue collection for determination of circulating estradiol and progesterone concentrations by radioimmunoassay were consistent with the designated cycle stage, and a significant decline in progesterone concentrations was apparent between biopsies taken in the mid and late secretory phase of the cycle (P < 0.01).

Endometrial biopsies were fixed overnight in 4% paraformaldehyde, then embedded in paraffin for immunohistochemical analysis. Additional biopsies from the endometrium of the same patients were frozen immediately at the point of tissue collection in liquid nitrogen, then stored at –80°C. Frozen biopsies were later used to obtain RNA (see below).

The second group consisted of fertile women (n = 12) with regular menstrual cycles lasting between 25 and 35 days who had not been using hormonal preparations for the preceding 6 months. The patients were to have a LNG-IUS inserted as a method of long-term hormonal contraception. All patients underwent a pre-insertion biopsy in either the proliferative (n = 5) or secretory (n = 5) phase of the cycle, immediately after which a LNG-IUS (Mirena; Schering HC, UK) was inserted as an outpatient procedure. The stage of the cycle prior to insertion of the LNG-IUS was determined by reference to serum sex steroid concentrations and histological dating by the criteria of Noyes et al. (1950Go). Further endometrial samples were collected at 1 (n = 2), 3 (n = 5), 6 (n = 5) and 12 months (n = 5) following insertion of the LNG-IUS. All specimens were fixed in 10% neutral-buffered formalin at 4°C and embedded in paraffin for immunohistochemical analysis.

The third patient group included women attending a menstrual problems clinic for management of unscheduled bleeding with the LNG-IUS. Endometrial samples were collected from patients attending the clinic with an LNG-IUS in utero for 3 months, 6 months, or >1 year. Samples were frozen immediately upon collection in liquid nitrogen and stored at –80°C.

RNA extraction and real time quantitative RT–PCR
Frozen samples of endometrium stored at –80°C were first homogenized, and then total RNA was extracted using the commercially available product Trizol (Invitrogen Life technologies Ltd, UK) according to the manufacturer’s instructions. Following extraction, total RNA was resuspended in 50 µl of RNA storage buffer (Ambion, USA) and stored at –80°C.

Samples of total RNA to be used for quantification of AR mRNA required pre-treatment with RNase-free DNase (Invitrogen) as this primer probe preparation did not span an intron. Samples of RNA to be used for quantification of 17{beta}HSD2 mRNA did not require pre-treatment with RNase-free DNase, as this primer probe preparation spanned an intron.

The RT–PCR reaction was performed in a 10 µl volume of reaction solution containing 1xTaqman RT buffer, MgCl2 (5.5 mmol/l), deoxyNTP, random hexamers (2.5 µmol/l), multiscribe reverse transcriptase (1.25 IU/µl), RNase inhibitor (0.40 IU/µl), and nuclease-free water (all reagents from Applied Biosystems, UK). The mix was divided into aliquots in individual tubes (8 µl/tube) and template RNA 200 ng (100 µg/l) was added. The samples were over-layered with mineral oil and RT reaction was conducted at 25°C for 60 min, 48°C for 45 min, and 95°C for 5 min for 1 cycle. Thereafter samples were stored at –20°C.

Real time quantitative PCR was performed in an ABI 7700 Sequence Detection System (Perkin–Elmer, USA). Oligonucleotide forward and reverse primers and oligonucleotide Taqman probes for AR and 17{beta}HSD2 were designed, with the use of Primer Express version 1.0 (Perkin–Elmer Applied Biosystems, UK), from sequences entered in the GenBank database.

The sequences for the AR and 17{beta}HSD2 primers and probe are shown in Table I. Quantitative PCR reaction mixtures were made containing Sure Start Taq DNA polymerase (0.025 IU/µl; Stratagene, UK), dNTP (all at 200 µmol/l) and specific forward and reverse primers (300 nmol/l) and probe (200 nmol/l) for either AR or 17{beta}HSD2. Primers and VIC-labelled probe (all at 50 nmol/l; Applied Biosystems, UK) for ribosomal 18s RNA were also added to the same reaction mixture.


View this table:
[in this window]
[in a new window]
 
Table I. Design specifications for androgen receptor (AR) and 17{beta}-hydroxysteroid dehydrogenase type 2 (17{beta}HSD2) real time quantitative primers and probe for use in the Perkin–Elmer ABI Prism 7700 analyser
 
All values are given relative to an internal standard endometrial sample (comparator) within the Taqman run to show temporal differences in mRNA expression across the menstrual cycle, i.e. comparing endometrium with endometrium. To allow ease of computation, the lowest 2{Delta}{Delta}Ct was used as a comparator and all other values were relative to this.

Copy numbers were not used as we performed relative quantification as described above and as outlined in the Perkin–Elmer guidelines. No DNA or RNA standard was used.

The sizes of the amplicons that are generated by the primer probe set for AR and 17{beta}HSD2 are 65 and 72 bp respectively (see Table I).

The accuracy of the quantitative PCR was tested by serial dilution (to x32 dilution) of a pool of AR cDNA and a pool of 17{beta}HSD2 cDNA. The slope of the line plotting the cycle number at which the curve crossed a threshold (Ct) against dilution had a gradient of <0.14 and 0.13 {Delta}Ct units/x32 dilution respectively. All menstrual cycle data and LNG-IUS-treated endometrium data were obtained from separate single PCR runs and related to standard AR cDNA and 17{beta}HSD2 cDNA preparations using the formula 2{Delta}{Delta}Ct that relates the ratio of 18S and specific amplicon in the sample cDNA with that of the standard preparation. Intra-assay variability for AR was calculated to 1% and for 17{beta}HSD2 to 3.3%.

Statistical analysis
Numeric data from the real time quantitative RT–PCR were analysed by one-way analysis of variance, followed by Fisher’s protected least significant difference test.

Antibody specificity
The properties of the antibodies for the immunolocalization of the AR (F39.4; Slayden et al., 2001Go) and 17{beta}HSD2 (Mab C2-12; Moghrabi et al., 1997Go) have been previously described and validated.

Immunohistochemistry
Paraffin sections (5 µm) were dewaxed and rehydrated through a series of alcohols, then washed in distilled water and 0.01 mol/l phosphate-buffered saline (PBS; Sigma–Aldrich Ltd, UK). The slides were then subjected to antigen retrieval by pressure-cooking in 0.01 mol/l sodium citrate at pH 6 for 5 min. After cooling for 20 min, the slides were washed in PBS. Endogenous peroxidase activity was quenched with immersion in 3% hydrogen peroxide (Merck, UK) in distilled water for 10 min at room temperature. Non-specific binding of the antibodies was blocked by incubating the sections for 20–30 min at room temperature in non-immune horse serum (Vectastain; Vector Laboratories, Inc., UK).

For localization of the AR, sections were incubated with the mouse monoclonal antibody F39.4 (IgG1; Biogenex Laboratories, Inc., USA) overnight at 4°C at a 1:480 dilution in PBS/bovine serum albumin gelatin or similarly with a control mouse IgG1 antibody at a matched IgG1 concentration to the F39.4 antibody (1:600 dilution). For localization of 17{beta}HSD2, sections were incubated with the mouse monoclonal antibody C2-12 (gift from Dr Stefan Andersson) overnight at 4°C at a 1:20 dilution in normal horse serum or similarly with a control mouse IgG antibody at a matched IgG concentration to the C2-12 antibody (1:400 dilution). Following a wash in PBS with Tween 20, the slides were incubated in biotinylated horse anti-mouse secondary antibody (Vectastain) in normal horse serum for 60 min at room temperature, reacted with the avidin–biotin peroxidase complex (Vectastain Elite) for 60 min at room temperature and visualized with the substrate and chromagen 3,3'-diaminobenzidine (Dako Liquid, Dako Corp., USA).

Scoring and analysis of immunoreactivity
The immunostaining intensity of epitopes in all tissue sections was assessed in a semi-quantitative manner on a 4-point scale: 0 = no immunostaining, 1 = mild immunostaining, 2 = moderate immunostaining and 3 = intense immunostaining. All tissue sections were scored blind by two observers. We had previously validated this scoring system in a subset of tissue sections in which immunoreactivity was measured with a computerized image analysis system, and a strong correlation between quantitative data derived from image analysis and subjective scores by a trained observer was obtained (Wang et al., 1998Go). Semi-quantitative scoring results were analysed by a non-parametric method, the Kruskall–Wallis test, followed by Dunn’s post-hoc multiple comparison test.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Temporal expression of AR mRNA in endometrium during the menstrual cycle and following treatment with LNG-IUS
The menstrual cycle samples were divided into proliferative (n = 12) and secretory (n = 15) phases with further classification of samples as early (n = 4), mid (n = 3), and late proliferative (n = 5) and early (n = 5), mid (n = 5) and late secretory (n = 5) phases.

Significantly higher levels of AR mRNA were observed in the early and mid proliferative phase endometrial tissue samples compared with the remainder of the menstrual cycle stages (Figure 1).



View larger version (34K):
[in this window]
[in a new window]
 
Figure 1. Quantitative evaluation of androgen receptor (AR) mRNA in endometrium across the stages of the menstrual cycle and following intrauterine delivery of levonorgestrel (LNG-IUS). All endometrial tissue samples were compared with an internal control (comparator) obtained during the early proliferative stage of the menstrual cycle. AR mRNA levels were high in the early and mid proliferative samples but fell significantly in samples taken from the late proliferative and early–late secretory samples (P < 0.01). Levels of AR mRNA were again low in the LNG-IUS-treated samples when compared with the (early and mid) proliferative stages of the cycle (P < 0.01). EP = early proliferative; MP = mid proliferative; LP = late proliferative; ES = early secretory; MS = mid secretory; LS = late secretory.

 
The LNG-IUS-treated endometrium samples were analysed following 3, 6 or >12 months of intrauterine treatment. To allow comparison with the menstrual cycle, samples were compared statistically with an early proliferative phase sample as a control.

Comparison revealed a significantly greater production of AR mRNA in the proliferative phase when compared with any duration of treatment with the LNG-IUS (P < 0.01) (Figure 1). No significant differences were noted between the secretory phase and the LNG-IUS-treated endometrium.

Temporal expression of 17{beta}HSD2 messenger RNA in endometrium through the menstrual cycle and following treatment with LNG-IUS
The same samples analysed for AR were used for the 17{beta}HSD2 assays. The highest levels of 17{beta}HSD2 mRNA expression was noted in the early proliferative phase, with the levels being significantly greater than the other stages of the menstrual cycle apart from the late secretory phase (P < 0.008) (Figure 2).



View larger version (24K):
[in this window]
[in a new window]
 
Figure 2. Quantitative evaluation of 17{beta}-hydroxysteroid dehydrogenase type 2 (17{beta}HSD2) mRNA in endometrium across the stages of the menstrual cycle and following intrauterine delivery of levonorgestrel (LNG-IUS). All endometrial tissue samples were compared with an internal control (comparator) obtained during the early proliferative stage of the cycle. 17{beta}HSD2 mRNA levels were significantly higher in the early proliferative phase of the menstrual cycle when compared with the mid–late proliferative and the early–mid secretory phase samples (P < 0.008). In LNG-IUS-treated samples, relatively higher levels of 17{beta}HSD2 mRNA were detected following 3 months. This was significantly greater than levels noted during the menstrual cycle and the other LNG-IUS-treated samples (P < 0.01). EP = early proliferative; MP = mid proliferative; LP = late proliferative; ES = early secretory; MS = mid secretory; LS = late secretory.

 
17{beta}HSD2 mRNA was also detected in endometrium exposed to intrauterine LNG. Samples were again analysed after 3, 6 or >12 months of in-vivo treatment with LNG-IUS. Relatively higher levels of 17{beta}HSD2 mRNA were detected following 3 months treatment with LNG-IUS. This was significantly greater than levels noted during the menstrual cycle stages and the other LNG-IUS treatment groups (P < 0.01; Figure 2).

AR immunostaining in levonorgestrel-treated endometrium
Presence of LNG-IUS resulted in a typically pseudo-decidualized stroma with atrophic glands and these histological features were observed in all post-insertion endometrial biopsies, as previously shown (Silverberg et al., 1986Go). Prior to insertion, all proliferative phase endometrial tissue samples displayed strongly positive nuclear immunoreactivity for AR in the endometrial stroma with minimal glandular nuclear immunoreactivity. As we previously noted (Slayden et al., 2001Go), AR immunoreactivity in secretory phase biopsies was also localized to stromal nuclei and was clearly decreased (Figure 3).



View larger version (15K):
[in this window]
[in a new window]
 
Figure 3. Androgen receptor (AR) immunoreactivity scores in endometrial stromal compartments throughout the menstrual cycle and following high dose intrauterine levonorgestrel exposure (1–12 months). Box-and-whisker plots: box represents the 25th and 75th percentiles and the heavy bar represents the median. Symbols (O) represent outliers. P = proliferative; S = secretory.

 
Intrauterine delivery of levonorgestrel via LNG-IUS was associated with minimal AR immunoreactivity in the stromal compartment (Figures 3 and 4).



View larger version (108K):
[in this window]
[in a new window]
 
Figure 4. Immunohistochemical localization of androgen receptor (AR) in human endometrium before and after treatment with levonorgestrel-releasing intrauterine system (LNG-IUS) at x40 magnification. (A) Proliferative phase: high level of stromal immunoreactivity. (B) Secretory endometrium: decreased stromal immunostaining. (C) Negative: no immunoreactivity. Endometrial biopsies at 3 months (D), 6 months (E) and 12 months (F) following insertion of LNG-IUS. Note persistent down-regulation of stromal immunoreactivity. Scale bar = 50 µm.

 
17{beta}HSD2 immunostaining in levonorgestrel-treated endometrium
In full-thickness endometrial biopsies, 17{beta}HSD2 was immunolocalized to the cytoplasm of glandular epithelium with no stromal immunoreactivity observed (Figure 5). Maximal immunoreactivity was noted in the secretory phase, during which both superficial and basal layers displayed strong immunoreactivity. Basal layer immunoreactivity was first evident in the early secretory phase. Immunostaining in both the early and mid secretory phase biopsies was significantly greater than in the proliferative phase (P < 0.05). Superficial layer immunoreactivity for 17{beta}HSD2 was also first evident in the early secretory phase with the mid secretory phase biopsies displaying significantly greater levels of immunostaining than the proliferative phase (P < 0.01). In the late secretory phase following the mid secretory peak, there was persistent superficial and basal immunoreactivity.



View larger version (106K):
[in this window]
[in a new window]
 
Figure 5. Immunohistochemical localization of 17{beta}-hydroxysteroid dehydrogenase type 2 (17{beta}HSD2) in human endometrium before and after high dose intrauterine levonorgestrel (LNG) exposure (x20 magnification). (A) Proliferative endometrium: no immunoreactivity in epithelial cells. Inset is negative control. (B) Secretory endometrium: immunoreactivity localized to glandular epithelium and absent stromal immunoreactivity. (C) Biopsy performed 1 month after LNG–intrauterine system insertion: definite, though reduced glandular immunoreactivity. Endometrium 3 months (D), 6 months (E) and 12 months (F) after exposure to high-dose intrauterine LNG. Negligible/absent immunostaining observed at these time-points. Scale bar = 50 µm.

 
One month after insertion of the LNG-IUS there was still detectable 17{beta}HSD2 immunostaining, and the staining intensity was greater than that normally found during the proliferative phase. Thereafter, at 3 months after insertion there was a significant decline in 17{beta}HSD2 expression compared with the level of 17{beta}HSD2 immunoreactivity in the secretory phase (P < 0.01; Figures 5 and 6). This low level of immunoreactivity was maintained in the 12 month biopsy specimens. The low degree of immunostaining for 17{beta}HSD2 between 3 and 12 months post-insertion was not significantly different from that seen during the proliferative phase (Figure 6).



View larger version (21K):
[in this window]
[in a new window]
 
Figure 6. 17{beta}-Hydroxysteroid dehydrogenase type 2 (17{beta}HSD2) immunoreactivity scores in endometrial glandular compartments of superficial endometrium across the menstrual cycle and following levonorgestrel-releasing intrauterine system (LNG-IUS) exposure (1–12 months). Box-and-whisker plots: boxes represent the 25th and 75th percentiles and the heavy bar represents the median. The whiskers represent the 10th and 90th percentiles. P = proliferative; S = secretory.

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The present studies have identified high levels of endometrial 17{beta}HSD2 protein in the first month after insertion of the LNG-IUS associated with high levels of 17{beta}HSD2 mRNA expression in endometrial tissue 3 months post-insertion. Since 17{beta}HSD2 converts estradiol to the less potent estrogen, estrone, the endometrial glands are exposed to more estrone than estradiol during the first 3 months after LNG-IUS insertion. Any estradiol-dependent products of the glands that may have paracrine actions throughout the endometrium would thus be suppressed or perturbed at this time. It is during the first months after LNG-IUS insertion that a proportion of users report unscheduled BTB. We propose that the unscheduled BTB reported may be due in part to an intracellular ‘estrogen deficiency’ that either directly or indirectly leads to vascular fragility. 17{beta}HSD2 protein expression is significantly reduced following 3 months treatment, coincident with the onset in improvement in menstrual bleeding patterns. However, the decline in 17{beta}HSD2 mRNA expression appears to be staggered, as these mRNA levels do not become undetectable until 6 months of LNG-IUS treatment.

Further, the present data demonstrate that endometrium exposed to high-dose intrauterine delivery of levonorgestrel exhibits a significant decrease in stromal AR protein immunoreactivity when compared with control proliferative endometrium. This significant down-regulation is in keeping with levels exhibited in the secretory phase and is maintained for >=12 months following insertion of the LNG-IUS. In addition, there is a down-regulation of AR messenger RNA in LNG-IUS-exposed endometrium, which, in this case, is closely reflected by the pattern of expression of the AR protein.

Levonorgestrel binds to both the AR and PR (Kloosterboer et al., 1988Go; Lemus et al., 1992Go). The role of androgens in the regulation of the AR protein and mRNA is not yet known, with some studies suggesting an up-regulation in response to androgens and others suggesting a down-regulation (Chada et al., 1994Go; Fujimoto et al., 1994Go; Iwai et al., 1995Go; Apparao et al., 2002Go). In-situ hybridization studies in the primate uterus suggested that testosterone can synergize with estradiol to up-regulate AR (Adesanya-Famuyiwa et al., 1999Go). Levonorgestrel has dual progestogenic and androgenic activity shown by its relative binding affinity for the PR when compared with AR (Kloosterboer et al., 1988Go). It is therefore unclear as to the mechanism of AR down-regulation in levonorgestrel-exposed endometrium, but it could be a consequence of both progestogenic and androgenic action.

In human endometrium, steroidal regulation of receptor action is dependent upon ligand availability. The 17{beta}HSD enzyme family comprises eight members, of which six human isoforms have been characterized. Its primary role is the inactivation of estradiol to estrone as well as testosterone to androstenedione, but in addition it converts the inactive 20{alpha}-dihydroprogesterone to active progesterone (Peltoketo et al., 1999Go; Labrie et al., 2000Go). Previous publications also report an up-regulation of 17{beta}HSD2 in the progesterone-dominated secretory phase (Casey et al., 1994Go). With intrauterine delivery of levonorgestrel, we found an increase in the expression of 17{beta}HSD2 protein when compared with the proliferative phase of the menstrual cycle followed by a down-regulation of the 17{beta}HSD2 protein in glandular epithelium after 3 months of treatment that was maintained up to the 12 month time-point. A significant up-regulation of the 17{beta}HSD2 mRNA in the initial 3 months was also noted followed by a down-regulation in patients treated for >6 months. This is consistent with the down-regulation of the PR receptor over this same time frame (Critchley et al., 1998Go).

Of interest, the decline in 17{beta}HSD2 mRNA expression appears to be staggered from that of the protein with a decrease in protein expression being noted prior to mRNA decline. This may be due to enhanced stability of 17{beta}HSD2 mRNA possibly resulting in a persistence of the mRNA for long periods without translation into protein. It has been shown that mRNA can be stable for days at a time and this may be regulated in some cases by sex steroids (Day et al., 1998Go; Staton et al., 2000Go). Analysis of 17{beta}HSD2 mRNA through the menstrual cycle also shows a lack of correlation with protein expression, because the highest levels of mRNA noted in the early proliferative phase occurred when no protein expression was noted. A similar pattern of 17{beta}HSD2 mRNA expression is seen in the data reported by Casey et al. (1994Go), Mustonen et al. (1998Go) and Kitawaki et al. (2000Go). Northern blot analysis revealed 17{beta}HSD2 mRNA in the early proliferative phase (Casey et al., 1994Go) and Kitawaki et al. found comparable levels of 17{beta}HSD2 mRNA in proliferative and secretory phase tissue specimens (Kitawaki et al., 2000Go). Quantitative analysis of 17{beta}HSD2 mRNA in human endometrium reported by Mustonen et al. (1998Go) also revealed a lack of correlation with protein levels, higher levels of mRNA expression occurring during the late secretory phase after the peaks in both serum progesterone and 17{beta}HSD2 protein expression.

Another consistent feature in the endometrium of long-term progestin-only contraceptive users is the abnormalities reported in the structure of endometrial microvessels (Hickey et al., 2000Go). There is an associated increase in blood vessel fragility that may contribute to the pathogenesis of endometrial BTB. In the endometrium, progesterone receptors are absent in the endothelium but prominent in the perivascular cells (Koji et al., 1994Go; Critchley et al., 2001Go). Levonorgestrel is reported to induce changes in vessel fragility that may be mediated through perivascular cells rather than the endothelium (Roberts et al., 1992Go).

In summary, exposure of the endometrium to intrauterine LNG results in decidualization and atrophy. The BTB problems reported by a proportion of users with this method of long-term progestin delivery are maximal during the first few months following insertion of an LNG-IUS. During this period, the 17{beta}HSD2 enzyme is elevated and steroid receptors, including AR, are suppressed. As a consequence, intracellular levels of the estrone are raised while those of estradiol, the more potent estrogen, are lowered. Problematic BTB usually improves after 3–6 months of LNG-IUS usage. At that time, levels of 17{beta}HSD2 protein are low and thus intracellular concentrations of estradiol would be raised and estrone lowered. The levels of both AR mRNA and protein remain suppressed. We suggest that these gradual changes in endocrine relationships during treatment with the LNG-IUS mitigate the tendency for vascular fragility in the LNG-IUS-treated endometrium. However, it is likely that the mechanisms underlying the BTB associated with the use of LNG-IUS, and indeed other progestin-only contraceptive methods, are complex and involve multiple factors. Hopefully, data such as those presented here will contribute further to our understanding of local progestin action and associated tendency for BTB.


    Acknowledgements
 
We wish to acknowledge the helpful advice provided by Professor Rodney Kelly (MRC HRSU, Edinburgh) on the quantitative RT–PCR studies. This work has been supported by RCOG/Wellbeing Grant support to H.O.D.C., S.G.H., J.I.M. and F.H. (Ref. C2/99) and MRC Programme Grant support to H.O.D.C., S.G.H., J.I.M. (grant no. 0000066).


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Adesanya-Famuyiwa, O.O., Zhou, J., Wu, G. and Bondy, C. (1999) Localization and sex steroid regulation of androgen receptor gene expression in rhesus monkey uterus. Obstet. Gynecol., 93, 265–270.[Abstract/Free Full Text]

Apparao, K. B., Lovely, L.P., Gui, Y., Lininger, R.A. and Lessey, B.A. (2002) Elevated endometrial androgen receptor expression in women with polycystic ovarian syndrome. Biol. Reprod., 66, 297–304.[Abstract/Free Full Text]

Brenner, R.M., Slayden, O.D. and Critchley, H.O. (2002) Anti-proliferative effects of progesterone antagonists in the primate endometrium: a potential role for the androgen receptor. Reproduction, 124, 167–172.[Abstract/Free Full Text]

Casey, M.L., MacDonald, P.C. and Andersson, S. (1994) 17 {beta}-Hydroxysteroid dehydrogenase type 2: chromosomal assignment and progestin regulation of gene expression in human endometrium. J. Clin. Invest., 94, 2135–2141.[ISI][Medline]

Chada, S., Pache, T.D., Huikeshoven, F.J.M., Brinkmann, A.O. and Kwast, V.D. (1994) Androgen receptor expression in human ovarian and uterine tissue of long term androgen-treated transsexual women. Hum. Pathol., 25, 1198–1204.[ISI][Medline]

Critchley, H.O., Bailey, D.A., Au, C.L., Affandi, B. and Rogers, P.A. (1993) Immunohistochemical sex steroid receptor distribution in endometrium from long-term subdermal levonorgestrel users and during the normal menstrual cycle. Hum. Reprod., 8, 1632–1639.[Abstract]

Critchley, H.O., Wang, H., Kelly, R.W., Gebbie, A.E. and Glasier, A.F. (1998) Progestin receptor isoforms and prostaglandin dehydrogenase in the endometrium of women using a levonorgestrel-releasing intrauterine system. Hum. Reprod., 13, 1210–1217.[Abstract]

Critchley, H.O., Brenner, R.M., Henderson, T.A., Williams, K., Nayak, N.R., Slayden, O.D., Millar, M.R. and Saunders, P.T. (2001) Estrogen receptor {beta}, but not estrogen receptor alpha, is present in the vascular endothelium of the human and nonhuman primate endometrium. J. Clin. Endocrinol. Metab., 86, 1370–1378.[Abstract/Free Full Text]

Day, D.A. and Tuite, M.F. (1998) Post-transcriptional gene regulatory mechanisms in eukaryotes: an overview. J. Endocrinol., 157, 361–371.[Abstract/Free Full Text]

Findlay, J.K. (1996) Future directions for research on endometrial bleeding. Hum. Reprod., 11 (Suppl. 2), 179–183.[ISI][Medline]

Fujimoto, J., Nishigaki, M., Hori, M., Ichigo, S., Itoh, T. and Tamaya, T. (1994) The effect of estrogen and androgen on androgen receptors and mRNA levels in uterine leiomyoma, myometrium and endometrium of human subjects. J. Steroid Biochem. Mol. Biol., 50, 137–143.[CrossRef][ISI][Medline]

Garcia, E., Bouchard, P., De Brux, J., Berdah, J., Frydman, R., Schaison, G., Milgrom, E. and Perrot-Applanat, M. (1988) Use of immunocytochemistry of progesterone and estrogen receptors for endometrial dating. J. Clin. Endocrinol. Metab., 67, 80–87.[Abstract]

Hickey, M. and Fraser, I.S. (2000) A functional model for progestogen-induced breakthrough bleeding. Hum. Reprod., 15, 1–6.[Free Full Text]

Iwai, M., Kanzaki, H., Fujimoto, M., Kojima, K., Hatayama, H., Inoue, T., Higuchi, T., Nakayama, H., Mori, T. and Fujita, J. (1995) Regulation of sex steroid receptor gene expression by progesterone and testosterone in cultured human endometrial stromal cells. J. Clin. Endocrinol. Metab., 80, 450–454.[Abstract]

Jones, R.L. and Critchley, H.O. (2000) Morphological and functional changes in human endometrium following intrauterine levonorgestrel delivery. Hum. Reprod., 15 (Suppl. 3), 162–172.[Medline]

Kitawaki, J., Koshiba, H., Ishihara, H., Kusuki, I., Tsukamoto, K. and Honjo, H. (2000) Progesterone induction of 17{beta}-hydroxysteroid dehydrogenase type 2 during the secretory phase occurs in the endometrium of estrogen-dependent benign disease but not in normal endometrium. J. Clin. Endocrinol. Metab., 85, 3292–3296.[Abstract/Free Full Text]

Kloosterboer, H.J., Vonk-Noordegraaf, C.A. and Turpijn, E.W. (1988) Selectivity in progesterone and androgen receptor binding of progestagens used in oral contraceptives. Contraception, 38, 325–332.[ISI][Medline]

Koji, T., Chedid, M., Rubin, J.S., Slayden, O.D., Csaky, K.G., Aaronson, S.A. and Brenner, R.M. (1994) Progesterone-dependent expression of keratinocyte growth factor mRNA in stromal cells of the primate endometrium: keratinocyte growth factor as a progestomedin. J. Cell Biol., 125, 393–401.[Abstract]

Labrie, F., Luu-The, V., Lin, X.S., Simard, J., Labrie, C., El-Alfy, M., Pelletier, G. and Belanger, A. (2000) Intracrinology: role of the family of 17{beta} hydroxysteroid dehyrogenases in human physiology and disease. J. Mol. Endocrinol., 25, 1–16.[Abstract/Free Full Text]

Lemus, A.E., Vilchis, F., Damsky, R., Chavez, B.A., Garcia, G.A., Grillasca, I. and Perez-Palacios, G. (1992) Mechanism of action of levonorgestrel: in vitro metabolism and specific interactions with steroid receptors in target organs. J. Steroid Biochem. Mol. Biol., 41, 881–890.[CrossRef][ISI][Medline]

Lessey, B.A., Killam, A.P., Metzger, D.A., Haney, A.F., Greene, G.L. and McCarty, Jr, K.S. (1988) Immunohistochemical analysis of human uterine estrogen and progesterone receptors throughout the menstrual cycle. J. Clin. Endocrinol. Metab., 67, 334–340.[Abstract]

Luukkainen, T., Lahteenmaki, P. and Toivonen, J. (1990) Levonorgestrel-releasing intrauterine device. Ann. Med., 22, 85–90.[ISI][Medline]

Maentausta, O., Svalander, P., Danielsson, K.G., Bygdeman, M. and Vihko, R. (1993) The effects of an antiprogestin, mifepristone, and an antiestrogen, tamoxifen, on endometrial 17{beta}-hydroxysteroid dehydrogenase and progestin and estrogen receptors during the luteal phase of the menstrual cycle: an immunohistochemical study. J. Clin. Endocrinol. Metab., 77, 913–918.[Abstract]

Mertens, H.J., Heineman, M.J., Theunissen, P.H., de Jong, F.H. and Evers, J.L. (2001) Androgen, estrogen and progesterone receptor expression in the human uterus during the menstrual cycle. Eur. J. Obstet. Gynecol. Reprod. Biol., 98, 58–65.[CrossRef][ISI][Medline]

Moghrabi, N., Head, J.R. and Andersson, S. (1997) Cell type-specific expression of 17{beta}-hydroxysteroid dehydrogenase type 2 in human placenta and fetal liver. J. Clin. Endocrinol. Metab., 82, 3872–3878.[Abstract/Free Full Text]

Mustonen, M.V., Isomaa, V.V., Vaskivuo, T., Tapanainen, J., Poutanen, M.H., Stenback, F., Vihko, R.K. and Vihko, P.T. (1998) Human 17{beta}-hydroxysteroid dehydrogenase type 2 messenger ribonucleic acid expression and localization in term placenta and in endometrium during the menstrual cycle. J. Clin. Endocrinol. Metab., 83, 1319–1324.[Abstract/Free Full Text]

Noyes, R.W. and Hertig, A.T. (1950) Dating the endometrial biopsy. Fertil. Steril., 1, 3–25.[ISI][Medline]

Peltoketo, E.H., Luu-The, V., Simard, J. and Adamski, J. (1999) 17{beta} hydroxysteroid dehydrogenase (HSD)/17-ketosteroid reductase (KSR) family; nomenclature and main characteristics of the 17HSD/KSR enzymes. J. Mol. Endocrinol., 23, 1–11.[Abstract/Free Full Text]

Roberts, D.K., Parmley, T.H., Walker, N.J. and Horbelt, D.V. (1992) Ultrastructure of the microvasculature in the human endometrium throughout the normal menstrual cycle. Am. J. Obstet. Gynecol., 166, 1393–1406.[ISI][Medline]

Silverberg, S.G., Haukkamaa, M., Arko, H., Nilsson, C.G. and Luukkainen, T. (1986) Endometrial morphology during long-term use of levonorgestrel-releasing intrauterine devices. Int. J. Gynaecol. Pathol., 5, 235–241.[ISI][Medline]

Slayden, O.D., Nayak, N.R., Burton, K.A., Chwalisz, K., Cameron, S.T., Critchley, H.O., Baird, D.T. and Brenner, R.M. (2001) Progesterone antagonists increase androgen receptor expression in the rhesus macaque and human endometrium. J. Clin. Endocrinol. Metab., 86, 2668–2679.[Abstract/Free Full Text]

Snijders, M.P., de Goeij, A.F., Debets-Te Baerts, M.J., Rousch, M.J., Koudstaal, J. and Bosman, F.T. (1992) Immunocytochemical analysis of oestrogen receptors and progesterone receptors in the human uterus throughout the menstrual cycle and after the menopause. J. Reprod. Fertil., 94, 363–371.[Abstract]

Staton, J.M., Thomson, A.M. and Leedman, P.J. (2000) Hormonal regulation of mRNA stability and RNA–protein interactions in the pituitary. J. Mol. Endocrinol., 25, 17–34.[Abstract/Free Full Text]

Wang, H., Critchley, H.O., Kelly, R.W., Shen, D. and Baird, D.T. (1998) Progesterone receptor subtype B is differentially regulated in human endometrial stroma. Mol. Hum. Reprod., 4, 407–412.[Abstract]

Submitted on March 11, 2003; resubmitted on July 4, 2003; accepted on September 4, 2003.