Expression of hypoxia-inducible factors in human endometrium and suppression of matrix metalloproteinases under hypoxic conditions do not support a major role for hypoxia in regulating tissue breakdown at menstruation

J. Zhang and L.A. Salamonsen,1

Prince Henry's Institute of Medical Research, P.O.Box 5152, Clayton, Victoria 3168, Australia


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: Classical studies in monkeys suggested that menstruation results from vasoconstriction, hypoxia and necrosis. The heterodimeric hypoxia-inducible factor (HIF) complex is critical in oxygen homeostasis via increased stability of HIF-1{alpha}/2{alpha} monomers, and these act as markers of hypoxia. We hypothesized that focal hypoxia in perimenstrual endometrium results in locally increased matrix metalloproteinases (MMP), leading to tissue destruction. METHODS: HIF-1{alpha}, HIF-2{alpha} and HIF-1ß were immunolocalized in cycling endometrium. Endometrial stromal cells were cultured under hypoxic and normoxic conditions and MMP measured by zymography and Western blots. RESULTS: HIF-1{alpha} and HIF-2{alpha} were detected in only some endometrial samples, and not confined to the perimenstrual tissue. Where present, they were primarily cytoplasmic, not nuclear. HIF-1ß was localized in epithelium, leukocytes and some decidual cells. Cultured endometrial stromal cells responded to hypoxia with increased cellular HIF-1{alpha} and secreted vascular endothelial growth factor. ProMMP-1 and proMMP-3 production was reduced in response to hypoxia regardless of the steroidal milieu (no added steroids, estrogen or estrogen plus progesterone). Active MMP-2 and membrane type 1 MMP but not proMMP-2 or proMMP-9 production were also inhibited by hypoxia. CONCLUSIONS: These results do not support a role for hypoxia in the focally increased production and activation of MMP observed prior to and during menstruation.

Key words: endometrium/HIF-1/hypoxia/matrix metalloproteinase/menstruation


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The classical work on menstruation used direct observation of the vascular and tissue changes occurring in intraocular implants of endometrium in rhesus monkeys (Markee, 1940Go). This suggested that menstruation was initiated by increased coiling of the spiral arterioles and vascular stasis, followed by a period of vasodilatation and perivascular bleeding from the wall of a capillary or arteriole. Some 24 h later, intense vasoconstriction and tissue necrosis was observed. It has been assumed that menstruation in the human likewise results from vasoconstriction of spiral arterioles with resultant hypoxia leading to necrosis (Speroff et al., 1994Go), an assumption not well supported by recent data. It is still not clear whether short-term vasoconstriction followed by ischaemia–reperfusion injury does occur in the human endometrium. Endometrial blood flow, as measured by thermal and 133Xe clearance techniques, is not reduced just prior to or during menstruation (Fraser and Peek, 1992Go) whereas no endometrial ischaemia/reperfusion episodes were detected perimenstrually or at any other time during the menstrual cycle by a Doppler technique (Gannon et al., 1997Go). It is now clear that major factors likely to be important in initiating menstruation are the influx of leukocytes into the endometrium and the expression and activation of matrix metalloproteinases (MMP) (Salamonsen and Woolley, 1999Go). The extent to which vascular events, that might result in hypoxia and necrosis, are also involved in initiation of human menstruation and how they might influence matrix metalloproteinases still remains to be determined (Salamonsen et al., 1999Go).

Hypoxia-inducible factor (HIF)-1 is a nuclear protein that activates gene transcription specifically in response to reduced cellular O2 concentration. Its presence in a tissue therefore acts as a marker for hypoxia. HIF-1 is a heterodimer composed of HIF-1{alpha} and HIF-1ß subunits, all members of the basic helix–loop–helix superfamily of transcription factors. HIF-1{alpha} is the specific hypoxia-regulated subunit whereas HIF-1ß, also known as ARNT, the aryl hydrocarbon receptor nuclear translocator (Wang et al., 1995Go), has a broad spectrum of expression. An alternative dimerization partner for HIF-1ß has recently been identified, and, in keeping with its functional homology with HIF-1{alpha}, has been called HIF-2{alpha} (Flamme et al., 1997Go). HIF-1 appears to have a central role in mediating transcriptional responses to hypoxia and a large number of genes have already been identified as targets of regulation by HIF-1 (Semenza, 1998Go).

HIF-1 activity is regulated primarily by oxygen-dependent modulation of steady-state HIF-1{alpha} or HIF-2{alpha} protein levels. Provided sufficient HIF-{alpha} and HIF-1ß mRNA are present in the cells to serve as templates for the translation of protein, there is no need for mRNA induction by hypoxia, and protein levels are independent of the regulation of mRNA expression (Semenza, 1998Go). The three proteins are differently regulated. In non-hypoxic cells, HIF1-ß protein is detected in both the nucleus and cytoplasm but when cells are exposed to 1% O2 there is progressive nuclear translocation (Wang et al., 1995Go). By contrast, HIF-1{alpha} protein is not detected in the absence of hypoxia but is considerably increased in the nucleus following 4–8 h of continuous hypoxia. On return to 20% O2, the HIF-1ß subunit reappears in the cytoplasm while HIF-1{alpha} disappears completely. Interestingly, both HIF-1{alpha} and HIF-2{alpha} are increased in placenta during the early first trimester, consistent with the physiological hypoxia of the human placenta at that gestational age (Rajakumar and Conrad, 2000Go). However, this is not the case in the pre-eclamptic placenta in which only HIF-2{alpha} is increased (Rajakumar et al., 2001Go).

Among the genes either up- or down-regulated by hypoxia in a range of cell types (Bandyopadhyay et al., 1995Go) are a number that are of potential importance at menstruation, including endothelin-1 (Marsh et al., 1996Go), MMP-3 and -9 (Jeziorska et al., 1995Go, 1996Go) and nitric oxide synthase (Tschugguel et al., 1999Go). However, data now available on the regulation of MMP by hypoxia suggest that both cell-type and MMP-type specific regulation occurs.

The present studies had two specific aims. To establish whether hypoxia can be detected in human endometrium, particularly during the perimenstrual phase, we examined endometrial tissue across the normal menstrual cycle for the presence of the HIF-1{alpha}, HIF-2{alpha} and HIF-1ß proteins and determined their cellular source and intracellular location. Secondly, given that hypoxia can either positively or negatively regulate MMP depending on the cell type, and that MMP are produced at focal points within the endometrial stroma during the perimenstrual and menstrual phases of the cycle when hypoxia would be expected to occur, we determined the effects of hypoxia on the production and activation of MMP from endometrial stromal cells in primary culture. The results do not support the notion that HIF proteins are substantially up-regulated within the endometrium prior to menstruation. Further, they demonstrate that in endometrial stromal cells, hypoxia overall down-regulates MMP rather than increasing them as occurs at menstruation. Thus hypoxia is not likely to have a central role in regulating the increased MMP expression associated with tissue breakdown during human menstruation.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Tissue collection and processing
Endometrial tissue was obtained at curettage from women with regular menstrual cycles and no apparent endometrial dysfunction, who gave informed consent for collection of tissue. Approval was given by the Human Ethics Committee at Monash Medical Centre, Melbourne. The women were either of proven fertility and were scheduled for tubal ligation or were undergoing testing for patency of the Fallopian tubes. Patients with leiomyomas or endometriosis or those who had received any steroid treatments during the past 12 months were specifically excluded from the study. Full thickness biopsies were obtained from women undergoing hysterectomy for reasons unrelated to endometrial dysfunction. Tissue samples were transferred to buffered formalin immediately after removal from the patient, fixed overnight at 4°C, washed in Tris-buffered saline (pH 7.6, TBS) and processed to paraffin wax blocks. Tissue was dated initially from the patient's testimony and confirmed histologically according to a published method (Noyes et al., 1950Go). Sections from a similarly processed sample of placental tissue from a woman undergoing termination of pregnancy at 8 weeks, from a pre-eclamptic woman at 36–38 weeks and from two full term normal pregnancies were generously provided by Dr Euan Wallace. Tissue for cell culture was separately collected into culture medium [Dulbecco's modified Eagle's medium (DMEM)/Ham's F-12 with antibiotics] immediately after removal and transported to the laboratory at room temperature.

Immunohistochemistry
Mouse anti-human HIF-1{alpha} was clone 54, from Transduction Laboratories (Lexington, KY, USA), diluted to 1.25 µg/ml in 10% fetal calf serum (FCS)/TBS. Mouse anti-human HIF-2{alpha} (a generous gift from Dr Helen Turley, Oxford, UK) was clone EP190b directed against amino acids 535–631 of the molecule. Anti-HIF-2{alpha} showed no cross-reaction with HIF-1{alpha} and was used at a final concentration of 5 µg/ml in 10% FCS/TBS. Rabbit anti-human HIF-1ß was from Novus Biologicals Ltd (Littleton, CO, USA) and was used at a concentration of 3.3 µg/ml in 10% normal goat serum (NGS)/TBS. Secondary antibodies (biotinylated horse anti-mouse IgG and biotinylated goat anti-rabbit IgG) were from Vector Laboratories Inc., Burlingame, CA, USA.

Sections of endometrial tissue were cut at 5 µm, dewaxed and rehydrated prior to immunohistochemistry. Placental tissue from the mid first trimester was used as a positive control. One section from the same tissue block was included in every staining run to provide quality control. For every tissue a second section on the same slide was used as a negative control with normal mouse IgG (for HIF-1{alpha} and HIF-2{alpha}) or normal rabbit serum (for HIF-1ß) diluted to the same protein concentration as the primary antibody and substituted for the primary antibody. Prior to HIF-2{alpha} staining, the sections were microwaved twice in 0.1 mol/l citric buffer (pH 6) on high power for 5 min twice and then cooled to room temperature.

Immunohistochemical detection for HIF-1{alpha} used the EnVision kit with horse-radish peroxidase (HRP; Dako, Glostrup, Denmark) and for HIF-2{alpha} and HIF-1ß used the StreptABC complex/HRP kit (Dako) along with the liquid diaminobenzidine (DAB) kit from Zymed Laboratories Inc. (San Francisco, CA, USA). All procedures were performed at room temperature unless otherwise stated. The sections were initially washed with 0.6% Tween 20/TBS (TBS–Tween) for 15 min with shaking, then with 3% H2O2 in methanol for 10 min. Anti-HIF-1{alpha}, was applied for 30 min. The sections for staining with HIF-1ß were treated with 10% NGS/TBS before incubation with primary antibody for 60 min. For HIF-2{alpha} staining, the sections were treated with 20% horse serum/TBS for 10 min before primary antibody was applied for 40 min at 37°C. All sections were then washed with TBS, TBS–Tween, then TBS before application of the appropriate secondary antibody and the detection system according to the manufacturer's instructions. All sections were counterstained with Harris's haematoxylin (1:10), dehydrated and mounted. Microscopy was performed using an Olympus BX50 microscope fitted with a Fujix HC-2000 high resolution digital camera.

Isolation and culture of human endometrial stromal cells under normoxic and hypoxic conditions
Cells were prepared from endometrial tissue as described previously (Zhang et al., 1998Go). Briefly, chopped tissue was digested with bacterial collagenase type III (Worthington Biochemical Corp., Freehold, NJ, USA) at a concentration of 45 IU/ml, in the presence of 3.5 µg/ml deoxyribonuclease (Boehringer Mannheim Biochemica, Mannheim, Germany) in calcium- and magnesium-free phosphate-buffered saline for 40 min at 37°C. The resultant digest was then filtered sequentially through 45 and 10 µm nylon filters to remove glands, and erythrocytes were removed by centrifugation on Ficoll-Paque (Pharmacia, Uppsala, Sweden). The resulting cells were resuspended in a 1:1 mixture of DMEM and Ham's F-12 medium (Trace Biosciences, Sydney, Australia) with 10% charcoal-stripped FCS and antibiotics (penicillin, streptomycin and fungizone) and plated in 48-well dishes (2x105 cells/well). All experiments were performed on primary cultures under three hormonal regimes: without added hormones, with estradiol-17ß (E2: 10 nmol/l; Sigma) alone and with E2 plus progesterone (100 nmol/l;Sigma). These were added from the start of culture and maintained for the duration of the experiment. Cells were changed to serum-free conditions at 48 h when they were nearly confluent and cultured for an additional 48 h with normal oxygenation (21% O2). Half of the cells were then transferred to a sealed chamber (modular incubator chamber; Billups–Rothenberg Inc., DelMar, CA, USA) which was flushed with a mixture of 2% O2, 5% CO2 in 93% N2 and maintained at 37°C, while the control wells were maintained in 21% O2. All cells were cultured under these conditions for a further 48 h, then the medium was rapidly harvested and replaced. Cells were allowed to recover for an additional 48 h under normoxic conditions and medium was again collected. All harvested medium was centrifuged to remove cellular debris and stored at –20°C for subsequent analysis. To test the level of O2 at the end of the first phase of the experiment, the medium in the culture wells, under hypoxic or normoxic conditions, was subjected to gas analysis using an Omni modular system. Each experiment was performed with five individual cell cultures and in triplicate or quadruplicate wells for each group within each cell culture.

Western blot analysis
To test whether the cells had responded to the hypoxic conditions, cells which had been subjected to either normoxic or hypoxic conditions were lysed rapidly with lysis buffer [1% sodium dodecyl sulphate (SDS), 10% glycerol in 0.5 mol/l Tris buffer pH 6.8] and tested for the induction of HIF-1{alpha}. Medium was examined for the induction of vascular endothelial growth factor (VEGF). The cell lysates were also examined for MT1-MMP, the membrane-bound MMP known to activate proMMP-2. Both cell lysates and culture medium (equal volumes from paired cultures) were subjected to SDS–polyacrylamide gel electrophoresis under reducing conditions and transferred to polyvinylidene difluoride membranes. Non-specific binding sites were blocked with 10% skim milk powder in TBS with 0.1% Tween 20 (SM/TBS/Tween) for 30 min. Blots were then incubated overnight at 4°C with either rabbit anti-mouse VEGF [raised against recombinant mouse VEGF164 (Gargett et al., 1999Go)] diluted 1:3000 in 5% SM/TBS/Tween, or with anti-HIF-1{alpha}, diluted 1:500 in 5% SM/TBS/Tween. Each blot was then incubated with either donkey anti-rabbit IgG or sheep anti-mouse IgG, each conjugated to HRP, and the ECL kit (Amersham Int, Sydney, Australia). Blots were exposed to Kodak X-Omat film for 10 min and developed. The positive control for HIF-1{alpha}, was 10 µg of a cell lysate prepared from HeLa cells treated with CoCl2 (Transduction Laboratories). The Western blot for VEGF contained one lane with pre-stained molecular weight standards that included bands at 44 and 32 kDa (BioRad, North Ryde, Australia). Each of these experiments was performed on two different cell preparations. HIF-2{alpha} was not examined in these experiments due to insufficient availability of antibody.

To establish whether MT1-MMP, which activates proMMP-2, was regulated by hypoxia, Western blots of cell lysates taken after 24 or 48 h of hypoxia, were incubated with mouse anti-human MT1-MMP (clone 114–6G6, 10 µg/ml; Oncogene Research Products, Cambridge, MA, USA) using conditions as described above for HIF-1{alpha}.

Gelatin and casein zymography
MMP-2 and MMP-9 in samples of culture medium were analysed by zymography on 10% SDS–polyacrylamide gels containing 1 mg/ml gelatin; MMP-1 and MMP-3 were analysed on similar gels but containing 1 mg/ml casein (all reagents from Bio-Rad) under non-reducing conditions as described previously (Zhang et al., 1998Go, 2000Go). Equal volumes of medium from each culture well in an experiment were loaded. MMP activity was visualized by negative staining and bands were identified by comparison with samples containing previously identified latent and active MMP and by their mol. wts determined using mol. wt markers (Bio-Rad). MMP identity of bands was also confirmed by incubation of parallel gels in the presence of EDTA (5 mmol/l). Relative activities of band intensity between lanes were semiquantified by densitometric analysis of zymograms using the Hewlett–Packard Scanjet IIp with Deskscan software (Hewlett–Packard, Palo Alto, CA, USA) set on a black and white photo with 256 gray shades. The area of the bands was analysed using the NIH (Bethesda, MD, USA) Image Version 1.54 equipped with gel-plotting macros, by measuring the area beneath the peaks plotted through the lane profile for the appropriate enzyme band. Comparisons were made only between samples run on the same gel. Previously we have shown that analysis of MMP by this method provides comparable results to those obtained by enzyme-linked immunosorbent assay (Salamonsen et al., 1997Go) but zymography has the advantage of providing assessment of both latent and active forms of each enzyme.

Statistical analysis
Data from five different cell culture experiments were combined and expressed as mean percentage of the appropriate control (100%) ± SEM. Analysis was by paired Student's t-test. Differences were considered significant at the 0.05 level.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Immunolocalization of HIF
HIF-1 {alpha}
Strong positive staining for HIF-1{alpha} was observed in both nuclei and cytoplasm in the chorionic villi in a sample of placenta taken at 8 weeks of pregnancy, a time when hypoxia is well recognized (Caniggia et al., 2000Go) (Figure 1aGo); such staining was not observed when an equivalent concentration of mouse IgG replaced the primary antibody in the staining procedure (Figure 1aGo, insert). Very little staining for HIF-1{alpha} was detected in the endometrium. A total of 34 endometrial samples were tested; six in the menstrual phase, eight in the proliferative phase, eight in the early secretory, five in the mid secretory and seven in the late secretory phases. Twenty samples had no staining in any compartment. Of the other 14 samples, staining was seen intermittently in glands (10 samples) decidual cells (three samples, Figure 1cGo) or blood vessels (seven samples, Figure 1cGo, insert). In each of these, there were areas within the tissue section where staining was detected but the majority of the section was unstained (Figure 1dGo). In the two perimenstrual samples of full thickness tissue (luminal epithelium–myometrium), no staining was detected. One of these was a day 1 tissue, the other immediately pre-menstrual (Figure 1eGo) when hypoxia could have been expected. Where glandular staining was seen, this was of a diffuse nature throughout the cytoplasm, with no apparent nuclear staining and was mostly of very low intensity. Pale glandular staining was mostly apparent only after day 23. Stromal staining was only in patches within any tissue, of very low intensity and only in the tissues after day 25. Staining in blood vessels was more intense, and was first observed immediately after ovulation (day 14), until day 20 (Figure 1cGo insert). This blood vessel staining appeared to be in endothelial cells. After day 20, vessel staining disappeared except for two of 12 samples, one day 25 and one day 28 (Figure 1cGo). Staining was also very pale in a few decidual cells late in the cycle (Figure 1cGo). In all cases, staining was cytoplasmic rather then nuclear.



View larger version (108K):
[in this window]
[in a new window]
 
Figure 1. Immunohistochemical staining. (a) Chorionic villus at 8 weeks of gestation, positively stained for hypoxia-inducible factor(HIF)-1{alpha}; insert shows negative control, IgG substituted for primary antibody on same tissue. (b) 36–38 week placenta from a woman with pre-eclampsia positively stained for HIF-2{alpha}; negative control on upper insert; lower insert shows nuclear staining. (c) Positive staining for HIF-1{alpha} in decidual cells in a day 27 sample of endometrium. (d) Decidual cells from same tissue as in C, but with no visible staining for HIF-1{alpha}. (e) Full thickness endometrium on day 28, showing lack of staining for HIF-1{alpha}. (f) Gland with cytoplasmic staining for HIF-2{alpha}in day 25 tissue. (g) HIF-1ß in leukocytes and glandular epithelium in a day 1 endometrial sample. (h) HIF-1ß staining in an endometrial gland on day 16. (i) HIF-1ß staining in decidual cells in day 1 endometrium. Bars: (b and e) 100 µm, (f) 25 µm, (b, inserts) 10 and 20 µm respectively. All other bars: 50 µm.

 
HIF-2{alpha}
As shown previously (Rajakumar and Conrad, 2000Go), HIF-2{alpha} was strongly stained in the pre-eclamptic villous placenta at 36–38 weeks of pregnancy (Figure 1bGo). The staining was both nuclear (Figure 1bGo, insert) and cytoplasmic. Such staining was not apparent when an equivalent concentration of mouse IgG replaced the primary antibody in the staining procedure (Figure 1bGo, insert). No staining was seen in the 8 week placenta used as a positive control for HIF-1{alpha} or in placenta from two full term normal pregnancies. Twenty samples of human endometrium, the same as those used for HIF-1{alpha} staining, were examined for immunoreactive HIF-2{alpha}, with emphasis on the pre-menstrual (days 25–28, n = 8) or menstrual (n = 2) phases. Interestingly, positive staining, where present, was in completely different cellular locations to that of HIF-1{alpha}. For example, in the tissue showing positive but patchy decidual staining for HIF-1{alpha} (Figure 1cGo), HIF-2{alpha} was undetectable. Where staining was observed it was either present in individual cells thought to be macrophages, or in the glandular cytoplasmic epithelium. In the latter, all staining was cytoplasmic, not nuclear. In particular, pale glandular staining was apparent predominantly during the proliferative phase (days 7–13) and somewhat stronger glandular staining was observed on days 25 and 26 (Figure 1fGo). It was absent on other days. During the late secretory phase, most samples displayed no staining, except for very pale decidual staining in one tissue. Of particular relevance to this study, no HIF-2{alpha} was detectable around blood vessels or in the stroma in the perimenstrual period.

HIF-1ß
HIF-1ß staining was immunolocalized primarily in two cell types in the endometrium, the epithelial cells of the glands and in leukocytes. In leukocytes (probably macrophages), staining was both nuclear and cytoplasmic and this was the strongest staining seen in the tissue (Figure 1gGo). A little positive staining was seen in the luminal epithelial cells but this was always patchy. Epithelial cell staining was generally diffuse and cytoplasmic, but, after ovulation, often appeared in bright patches either apically or basally within the cytoplasm of the epithelial cells and appeared to be associated with vacuoles (Figure 1hGo). Some staining was detected in decidual cells remaining during menstruation (Figure 1iGo).

Cell culture experiments
Demonstration of hypoxic conditions in the incubation chambers
Measurement of O2 tension in the gaseous environment of the cells was performed at the end of each incubation (n = 5 separate experiments). The mean (± SD) pO2 at the end of the low O2 incubation was 78 ± 14 mmHg whereas that under control (normoxic) conditions was 155 ± 1.5 mmHg.

HIF-1{alpha} but not HIF-1ß is induced in endometrial stromal cells under hypoxic conditions
Western blot analysis performed separately in two experiments showed an induction of HIF-1{alpha} protein (~120 kDa) in cell lysates from cells cultured under hypoxic conditions (Figure 2aGo), whereas HIF-1ß was low in cells cultured under normoxic conditions. HIF-1ß (~92 kDa) was present in both normoxic and hypoxic conditions (Figure 2bGo) although an increase was observed at 48 h of hypoxia.



View larger version (37K):
[in this window]
[in a new window]
 
Figure 2. Western blot for hypoxia-inducible factor (HIF)-1{alpha} (a) and HIF-1ß (b) in lysed endometrial stromal cells following culture under normoxic or hypoxic conditions, for 5, 24 and 48 h. The positive control (Std) was cell lysate from HeLa cells treated with CoCl2. Representative of two complete experiments using different cells.

 
VEGF is induced in endometrial stromal cells under hypoxic conditions
The anti-VEGF used has previously been shown to detect VEGF in culture medium from human endometrial stromal cells (Gargett et al., 1999Go). To confirm that the cells in our study responded to hypoxia, the culture medium was analysed for VEGF. Two isoforms of VEGF, (~38 and ~32 kDa) were detected by Western blot analysis in culture medium from endometrial cells cultured under hypoxic conditions. These were barely detectable in culture medium from the same cells cultured under normoxic conditions (Figure 3Go).



View larger version (50K):
[in this window]
[in a new window]
 
Figure 3. Western blot for vascular endothelial cell growth factor in culture medium from endometrial stromal cells after 48 h of either normoxic or hypoxic conditions. Molecular weights were established by comparison with prestained mol. wt markers run on the same gel. Representative of two experiments using different cells.

 
MT1-MMP is decreased in endometrial stromal cells under hypoxic conditions
The cell lysates were also examined by Western blot for the presence of MT1-MMP (Figure 4Go). Bands were seen at ~64, ~54 and ~125 kDa, which would correspond to latent, active and dimeric forms of the protein. All three bands were decreased at both 24 and at 48 h in cells cultured under hypoxic compared with normoxic conditions. Densitometric analysis revealed this decrease to be to ~50% of control for the latent form and to 75% of control for the active form of the enzyme.



View larger version (69K):
[in this window]
[in a new window]
 
Figure 4. Western blot for membrane-type 1 matrix metalloproteinase in cell lysates from endometrial stromal cells after 24 and 48 h of either normoxic (N) or hypoxic (H) conditions. Molecular weights were established by comparison with prestained mol. wt markers run on the same gel.

 
Effect of hypoxia on MMP-2 and MMP-9 production by endometrial stromal cells
Three bands of gelatinase activity could be detected on gelatin zymograms corresponding to proMMP-9 (92 kDa), proMMP-2 (72 kDa) and active MMP-2 (62–65 kDa) (Figure 5aGo). As shown previously (Rawdanowicz et al., 1994Go), all bands disappeared when EDTA was included in the incubation medium, confirming that the proteolytic activity was from the MMP class of proteases. Culture medium was assayed following cell culture under normoxic (control) conditions and under hypoxic conditions and subsequently after 48 h recovery in normoxic conditions.



View larger version (36K):
[in this window]
[in a new window]
 
Figure 5. Gelatin zymography performed on culture medium from endometrial cells. (a) Zymogram of medium from triplicate wells cultured in the presence of estradiol (E2) alone, under control normoxic (N) or hypoxic (H) conditions and from the same wells following 48 h recovery in normoxic conditions (RN and RH).(b) Combined data from five separate experiments. The three bands on the gels representing pro-matrix metalloproteinase (proMMP)-2, active MMP-2 and proMMP-9 were separately analysed by densitometry. In each case, N or RN were designated as 100% and the corresponding H and RH expressed as a percentage of these. The histograms represent three separate steroid hormone treatment groups. {square}, nil , E2 alone {blacksquare}, E2 + progesterone. *P < 0.05,**P < 0.01 compared with control.

 
Hormone treatment of endometrial stromal cells has previously been demonstrated to regulate the production of active MMP-2 (Zhang et al., 2000Go). This was the case also in the present study (data not shown). Under normoxic conditions, neither proMMP-2 nor proMMP-9 secretion were different between the three treatment groups (no steroids, E2, E2 + progesterone) but the levels of active MMP-2 were reduced to 86 ± 5 and 75 ± 4% in the E2 and E2 + progesterone groups (P < 0.01 and P < 0.05) respectively, compared with the group without steroids (100%).

In cultures without added hormones (controls), a decrease in active MMP-2 levels (to 75% of the normoxic group) was seen under hypoxic conditions (P = 0.026) but active enzyme levels were restored when the hypoxic cells were transferred back to normoxic conditions. In contrast, no changes in proMMP-2 or proMMP-9 were detected (Figure 5bGo) during hypoxia or recovery from hypoxia. Similar effects on active MMP-2 were observed when either E2 (P = 0.011) or E2 + progesterone (P = 0.047) were added to the cultures although the greatest reduction of active MMP-2 was seen in the presence of E2 (to 58% of normoxic results): recovery was also not detected in the 48 h of the experiment in the presence of E2 alone (Figure 5a and bGo). Previous studies (Zhang et al., 2000Go) have demonstrated that the activation of proMMP-2 in endometrial cell cultures results from the action of membrane-type (MT)1-MMP, the expression of which is reduced in the presence of progesterone. However, the effect of hypoxia was not different in the cells cultured with E2 alone or with E2 and progesterone although recovery was not seen in the cultures with E2 alone.

Effect of hypoxia on MMP-1 and MMP-3 production by endometrial stromal cells
Three bands of activity were detected when culture medium was subjected to zymography on casein gels (Figure 6aGo) and these were identified as having molecular weights identical to those of proMMP-3 (59/57 kDa) and proMMP-1 (56 and 52 kDa forms). As for the gelatinases, no bands were seen when duplicate gels were incubated in the presence of EDTA. In most cultures, very little of the lower mol. wt active forms of the two enzymes (28–49 kDa) was seen.



View larger version (44K):
[in this window]
[in a new window]
 
Figure 6. Casein zymography performed on culture medium from endometrial cells. (a) Zymogram of medium from triplicate wells cultured in the presence of estradiol (E2) alone, under control normoxic (N) or hypoxic (H) conditions and from the same wells following 48 recovery in normoxic conditions (RN and RH).(b) Combined data from five separate experiments. The three bands on the gels representing pro-matrix metalloproteinase (proMMP)-3 and two mol. wt bands representing proMMP-1 were separately analysed by densitometry. The data for the two proMMP-1 bands showed identical responses and were combined. In each case, N or RN were designated as 100% and the corresponding H and RH expressed as a percentage of these. The histograms represent three separate steroid hormone treatment groups. {square}, nil , E2 alone; {blacksquare}, E2 + progesterone. *P < 0.05, ***P < 0.001 compared with control.

 
Both proMMP-1 and proMMP-3 have previously been shown to be reduced by ovarian steroid hormones in similar cultures (Salamonsen et al., 1997Go). This was the case also in the present study (data not shown). Under normoxic conditions, the levels of proMMP-1 were reduced to 87.3 ± 1.5 and 69.3 ± 6.8% in the E2 and E2 + progesterone groups (P < 0.001 and P < 0.05) respectively compared with the group without steroids (100%). Levels of proMMP-3 were reduced to 82.8 ± 5.3 and 71 ± 3.6% in the E2 and E2 + progesterone groups (P < 0.05 and P < 0.001) respectively, compared with the group without steroids (100%).

Both proMMP-1 and proMMP-3 were significantly decreased when the cells were incubated under hypoxic compared with normoxic conditions (Figure 6bGo), although, unlike the active MMP-2, levels remained low (P < 0.01 in each case) within the 48 h recovery phase. The effects of hypoxia were more marked in the presence of E2 than in its absence and were even greater in the presence of E2 + progesterone.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
This study examined the hypothesis that hypoxia occurs in the endometrium late in the menstrual cycle, and subsequently stimulates the focal production of MMP which are necessary for the tissue breakdown at menstruation. Using components of the HIF transcription factor complex as cellular markers for hypoxia in the tissue, we demonstrated that HIF-1{alpha} and HIF-2{alpha} are detectable in only a very small proportion of endometrial samples taken during the perimenstrual phase. These data suggest that if hypoxia does occur at this time, it is not widespread, and is likely to be very short-lived. We also demonstrated that hypoxia inhibits the production of proMMP-1, proMMP-3 and MT1-MMP, and inhibits the activation of proMMP-2 in endometrial stromal cells. Thus, hypoxia is unlikely to provide a stimulus for increased MMP transcription and activation immediately prior to menstruation.

Since the studies of Markee in the 1940s, it has been assumed that vasoconstriction followed by hypoxia is the underlying cause of menstruation in women. This assumption was not supported by published data using measurements of blood flow in the human uterus (Fraser and Peek, 1992Go; Gannon et al., 1997Go). Furthermore, other observations (Hisaw and Hisaw, 1961Go) indicated that bleeding will occur following progesterone withdrawal from atrophied endometrium which has no spiral arterioles, suggesting that vasoconstriction of these vessels may not be an essential component of the cascade of events culminating in menstruation. Until recently, no molecular markers of hypoxia were available.

HIF-1 is a ubiquitously expressed transcription factor, activated within a physiological range of O2 concentrations (Jiang et al., 1996Go). It is a heterodimeric DNA-binding complex, composed of the HIF-1ß component dimerized with one of the HIF-{alpha} components. HIF-1ß is generally constitutively expressed and is involved in the induction of more than one pathway (Reyes et al., 1992Go) although its nuclear localization appears to represent protein that is heterodimerized with HIF-{alpha} (Wang et al., 1995Go). The HIF-{alpha} subunits are regulated predominantly via stability of the protein. HIF-1{alpha} is highly unstable under normotoxic conditions and hypoxia significantly prolongs its half-life, thus allowing accumulation and formation of the complex (Huang et al., 1996Go). Indeed, 31% of all amino acids are proline, glutamic acid, serine or threonine residues which have been implicated in protein instability (Rogers et al., 1986Go).

Although both HIF-1{alpha} and HIF-2{alpha} were undetectable in the majority of endometrial samples examined, one or both proteins were present in a small number of samples, variously in the glands, decidual cells or blood vessels. Importantly, in any endometrial tissue where positive immunostaining was observed, this was only in a small focus within the tissue (a single gland, blood vessel or area of decidua). There was also no correlation between immunostaining and the time of the menstrual cycle. By contrast, in the placental tissue taken at 8 weeks of pregnancy, a time when hypoxia is known to occur and HIF-1{alpha} to be present (Caniggia et al., 2000Go; Rajakumar and Conrad, 2000Go), staining for HIF-1{alpha} was consistent throughout the tissue section. Likewise, substantial HIF-2{alpha} staining was demonstrated in placenta from pre-eclamptic women at term (Rajakumar et al., 2001Go). The staining pattern for these proteins in the endometrium suggests that in normal cycling endometrium, there are occasional microenvironmental circumstances in which one or more blood vessels undergoes vasoconstriction, limiting the O2 supply to the tissue at a focal point. However, there is no substantial increase perimenstrually. It was somewhat surprising that HIF-1ß was not ubiquitous within the tissue. However, the strong immunoreactivity in leukocytes within the tissue [as shown previously in tumour-associated macrophages (Talks et al., 2000Go)] validated the lack of expression within surrounding cells or the endometrium.

There are a number of possible explanations for the lack of immunostaining for hypoxia-inducible transcription factors in the vicinity of blood vessels within human perimenstrual endometrium, only one of which is that hypoxia does not occur at this time. Any hypoxia is likely to be local rather than global and thus conclusions based on limited sampling may be erroneous. Certainly, our previous experience in examining MMP expression in perimenstrual endometrial samples shows that these focally produced enzymes are not detected in all samples taken at this time (Hampton and Salamonsen, 1994Go). Hypoxia may also occur within a very narrow time frame, limiting the chance of accurate sampling, and in many areas the O2 levels may not fall sufficiently to induce HIF proteins. Finally, as it is not possible to examine endometrial tissues for HIF-1 in vivo and as HIF-{alpha} subunits are labile in oxygenated cells, it is possible that loss occurred during the operative procedure or before the tissue was fixed. Nevertheless, all tissues including the positive controls were handled in the same way, providing security that the overall contrast in HIF-{alpha} expression represents an important biological difference between the tissues studied.

The staining for HIF-1{alpha} and HIF-2{alpha} in the endometrial epithelium is of interest in light of the recent demonstration that agents besides hypoxia can activate HIF under normoxic conditions. In particular, insulin-like growth factor (IGF)-I and IGF-II treatment of cells results in induction of HIF-1{alpha} protein (Zelzer et al., 1998Go; Feldser et al., 1999Go). In the human endometrium, IGF-I mRNA is most pronounced in the late proliferative phase, whereas IGF-II mRNA is highly expressed in the mid-secretory phase (Giudice and Irwin 1999Go). While the IGF are predominantly of stromal origin, their receptors are in highest abundance on epithelial cells (Rutanen 1998Go), providing a possible explanation for the HIF proteins in the latter cells.

Hypoxia is known to induce a cascade of physiological responses that includes erythropoiesis, glycolysis, angiogenesis and vascular cellular proliferation, and a number of hypoxia-inducible genes have been identified (Blancher and Harris 1998Go; Semenza, 1998Go). Previous studies have demonstrated that hypoxia can stimulate the release of factors important for angiogenesis in human endometrial cells in culture; these include both VEGF (Popovici et al., 1999Go; Sharkey et al., 2000Go) and angiopoietin-2 (Krikun et al., 2000Go). The stimulation of VEGF, along with demonstration of an increase of HIF-1{alpha} protein, was used in the present study to ensure that the endometrial cell cultures were indeed responsive to hypoxia and confirms the earlier findings. It must be borne in mind, however, that in such cell culture models, the extent of hypoxia (78 mmHg in this case) is less than would be seen in a truly hypoxic tissue and that cell culture in 21% O2 is hyperoxic.

Strong evidence supports a key role for the MMP in the tissue breakdown at menstruation (Rodgers et al., 1994Go; Marbaix et al., 1996Go; Salamonsen and Woolley, 1999Go). The sharp rise in expression of a number of these enzymes in the perimenstrual phase of the cycle is no doubt regulated by a multiplicity of factors, including withdrawal of progesterone (Marbaix et al., 1992Go; Schatz et al., 1994Go; Salamonsen et al., 1997Go; Zhang et al., 2000Go) and locally produced cytokines (Rawdanowicz et al., 1994Go; Bruner et al., 1995Go; Singer et al., 1999Go). However, the very focal nature of the expression, particularly of MMP-1 and MMP-3 at menstruation, suggests regulation by factors present predominantly at such foci. Leukocytes could contribute to such focal regulation, and indeed, leukocyte products have marked effects on the expression and activation of the MMP produced by endometrial stromal cells. This has been demonstrated for both mast cells (Zhang et al., 1998Go) and neutrophils (Lathbury and Salamonsen 2000Go); the subject has been reviewed (Salamonsen and Lathbury, 2000Go). The present study sought to extend our understanding of MMP regulation at menstruation by examining the role of hypoxia.

Hypoxia has been shown to regulate a number of MMP but a wide variety of documented responses suggest that these are cell type specific and differ between MMP family members. For example, MMP-9 is increased in response to hypoxia in human umbilical vein endothelial cells (Bandyopadhyay et al., 1995Go), neurons (Heo et al., 1999Go), cardiac myocytes (Romanicet al., 2001Go) and MDA-MB-231 breast cancer cells (Canning et al., 2001Go) whereas MMP-1 is increased in dermal fibroblasts (Yamanaka and Ishikawa 2000Go) and MMP-3 in cardiac myocytes (Romanicet al., 2001Go). Conversely, hypoxia is inhibitory to MMP-3 in dermal fibroblasts (Steinbrech et al., 1999Go) and to MMP-9 in MCF-7 breast cancer cells (Canning et al., 2001Go). Hypoxia occurs in the placenta at ~6–8 weeks of gestation, and in villous explants, antisense inhibition of HIF-1{alpha} expression inhibited expression of transforming growth factor ß3 and MMP-2 while triggering markers of an invasive trophoblast phenotype such as {alpha}1-integrin and MMP-9 (Caniggia et al., 2000Go).

The results of the present study clearly indicate that hypoxia is overall inhibitory to MMP production by endometrial stromal cells. In particular, proMMP-1, proMMP-3, and the active form of MMP-2 were all decreased in the culture medium, regardless of the steroidal milieu (no steroids, E2 alone or E2 + progesterone). Previous studies from our laboratory have demonstrated that activation of proMMP-2 in such cultures is probably mediated by the MT1-MMP (MMP-14) (Zhang et al., 2000Go) and we have demonstrated here that MT1-MMP is also reduced by hypoxia in endometrial stromal cells. Progesterone is a well-established inhibitor of these same enzymes (Marbaix et al., 1992Go; Salamonsen et al., 1997Go; Zhang et al., 2000Go) and its withdrawal prior to menstruation appears to be conducive to, but not necessary for, the stimulation of MMP production by locally produced factors (Osteen et al., 1997Go). Whether or not the inhibition of MMP by hypoxia can be over-ridden by stimulatory molecules such as cytokines remains to be established.

In conclusion, these studies do not support the hypothesis that hypoxia in the endometrium, in the late secretory phase of the menstrual cycle, stimulates MMP production and activation. However, hypoxia may play a role once uterine bleeding and shedding is established, by inhibiting MMP production and stimulating VEGF and other angiogenic factors to support endometrial regeneration.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Work in the authors' laboratory is supported by the NH&MRC of Australia (Grant 971292). We would like to thank members of the Department of Obstetrics and Gynaecology at Monash University and in particular Professor G.Kovacs and his patients for providing the endometrial tissue used in these studies, Dr E.Wallace for providing us with placental tissue and Dr P.Rogers for providing us with VEGF antibody. We also thank Dr Helen Turley, (Oxford, UK), for providing the antibody against HIF-2{alpha} and Dr Margaret Rees (Oxford, UK) and Dr Rebecca Jones for helpful discussions about the data. Sue Panckridge provided assistance in the preparation of illustrations and Samantha Park assisted in preparing the manuscript.


    Notes
 
1 To whom correspondence should be addressed. E-mail: lois.salamonsen{at}med.monash.edu.au Back

Submitted on July 6, 2001


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Bandyopadhyay, R.S., Phelan, M. and Faller, D.V. (1995) Hypoxia induces AP-1-regulated genes and AP-1 transcription factor binding in human endothelial and other cell types. Biochim. Biophys. Acta, 1264, 72–78.[ISI][Medline]

Blancher, C. and Harris, A.L. (1998) The molecular basis of the hypoxia response pathway: tumour hypoxia as a therapy target. Cancer Met. Rev., 17, 187–194.[ISI][Medline]

Bruner, K.L., Rodgers, W.H., Gold, L.I., Korc, M., Hargrove, J.T., Matrisian, L.M. and Osteen, K.G. (1995) Transforming growth factor-beta mediates the progesterone suppression of an epithelial metalloproteinase by adjacent stroma in the human endometrium. Proc. Natl Acad. Sci. USA, 92, 7362–7366.[Abstract]

Caniggia, I., Mostachfi, H., Winter, J., Gassmann, M., Lye, S.J., Kuliszewski, M. and Post, M. (2000) Hypoxia-inducible factor-1 mediates the biological effects of oxygen on human trophoblast differentiation through TGFbeta. J. Clin. Invest., 105, 577–587.[Abstract/Free Full Text]

Canning, M.T., Postovit, L.-M., Clarke, S.H. and Graham, C.H. (2001) Oxygen-mediated regulation of gelatinase and tissue inhibitor of metalloproteinases-1 expression by invasive cells. Exp. Cell Res., 267, 88–94.[ISI][Medline]

Feldser, D., Agani, F., Iyer, N., Pak, B., Ferreirs, G. and Semenza, G.L. (1999) Reciprocal positive regulation of hypoxia-inducible factor 1alpha and insulin-like growth factor 2. Cancer Res., 59, 3915–3918.[Abstract/Free Full Text]

Flamme, I., Frohlich, T., vonReutern, M., Kappel, A., Damert, A. and Ritson, A. (1997) HRF, a putative basic helix–loop–helix–PAS–domain transcription factor is closely related to hypoxia-inducible factor-1 alpha and developmentally expressed in blood vessels. Mechan. Dev., 63, 51–60.[ISI][Medline]

Fraser, I.S. and Peek, M.J. (1992) Effects of exogenous hormones on endometrial capillaries. In Alexander, N.J. and d'Arcangues, C. (eds), Steroid Hormones and Uterine Bleeding. AAAS Publications, Washington, pp. 65–79.

Gannon, B.J., Carati, C.J. and Verco, C.J. (1997) Endometrial perfusion across the normal human menstrual cycle assessed by laser Doppler fluxmetry. Hum. Reprod., 12, 132–139.[ISI][Medline]

Gargett, C.E., Lederman, F.L., Lau, T.M., Taylor, N.H. and Rogers, P.A.W. (1999) Lack of correlation between vascular endothelial growth factor production and endothelial cell proliferation in the human endometrium. Hum. Reprod., 14, 2080–2088.[Abstract/Free Full Text]

Giudice, L.C. and Irwin, J.C. (1999) Roles of the insulin-like growth factor family in nonpregnant human endometrium and at the decidual: trophoblast interface. Semin. Reprod. Endocrinol., 17, 13–21.[ISI][Medline]

Hampton, A.L. and Salamonsen, L.A. (1994) Endometrial expression of messenger ribonucleic acid encoding matrix metalloproteinases and their tissue inhibitors coincides with menstruation. J. Endocrinol., 141, R1–R3.[Abstract]

Heo, J.H., Lucero, J., Abumiya, T., Koziol, J.A., Copeland, B.R. and del Zoppo, G.J. (1999) Matrix metalloproteinases increase very early during experimental focal cerebral ischemia. J. Cerebral Blood Flow Metab., 19, 624–633.[ISI][Medline]

Hisaw, F.L. and Hisaw, F.L.J. (1961) Action of estrogen and progesterone on the reproductive tract of lower primates. In Young, W.C. (ed.), Sex and Internal Secretions. Williams and Wilkins, Baltimore, Vol. 1, pp. 556–568.

Huang, L.E., Arany, Z., Livingston, D.M. and Bunn, H.F. (1996) Activation of hypoxia-inducible transcription factor depends primarily upon redox-sensitive stabilization of its alpha subunit. J. Biol. Chem., 271, 32253–32259.[Abstract/Free Full Text]

Jeziorska, M., Nagase, H., Salamonsen, L.A. and Woolley, D.E. (1996) Immunolocalization of the matrix metalloproteinases, gelatinase B and stromelysin-1 in human endometrium throughout the menstrual cycle. J. Reprod. Fertil., 107, 43–51.[Abstract]

Jeziorska, M., Salamonsen, L.A. and Woolley, D.E. (1995) Mast cell and eosinophil distribution and activation in human endometrium throughout the menstrual cycle. Biol. Reprod., 53, 312–320.[Abstract]

Jiang, B.J., Semenza, G.L., Bauer, C. and Marti, H.H. (1996) Hypoxia-inducible factor 1 levels vary exponentially over a physiologically relevant range of O2 tension. Am. J. Physiol., 271, C1172–C1180.[Abstract/Free Full Text]

Krikun, G., Schatz, F., Finlay, T., Kadner, S., Mesia, A., Gerrets, R.P. and Lockwood, C.J. (2000) Expression of angiopoietin-2 by human endometrial endothelial cells: regulation by hypoxia and inflammation. Biochem. Biophys. Res. Commun., 275, 159–163.[ISI][Medline]

Lathbury, L.J. and Salamonsen, L.A. (2000) In-vitro studies of the potential role of neutrophils in the process of menstruation. Mol. Hum. Reprod., 6, 899–906.[Abstract/Free Full Text]

Marbaix, E., Donnez, J., Courtoy, P.J. and Eeckhout, Y. (1992) Progesterone regulates the activity of collagenase and related gelatinases A and B in human endometrial explants. Proc. Natl Acad. Sci. USA, 89, 11789–11793.[Abstract]

Marbaix, E., Kokorine, I., Moulin, P., Donnez, J., Eeckhout, Y. and Courtoy, P.J. (1996) Menstrual breakdown of human endometrium can be mimicked in vitro and is selectively and reversibly blocked by inhibitors of matrix metalloproteinases. Proc. Natl Acad. Sci. USA, 93, 9120–9125.[Abstract/Free Full Text]

Markee, J.E. (1940) Menstruation in intraocular endometrial transplants in the rhesus monkey. Contrib. Embryol., 177, 220–230.

Marsh, M.M., Findlay, J.K. and Salamonsen, L.A. (1996) Endothelin and menstruation. Hum. Reprod., 11, 83–89.[ISI][Medline]

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

Osteen, K.G., Sierra-Rivera, E., Keller, N.R. and Fox, D.B. (1997) Interleukin-1 alpha opposes progesterone-mediated suppression of MMP-7. A possible role of this cytokine during human implantation. Ann. NY Acad. Sci., 828, 137–145.[ISI][Medline]

Popovici, R.M., Irwin, J.C., Giaccia, A.J. and Giudice, L.C. (1999) Hypoxia and cAMP stimulate vascular endothelial growth factor (VEGF) in human endometrial stromal cells: potential relevance to menstruation and endometrial regeneration. J. Clin. Endocrinol. Metab., 84, 2245–2248.[Abstract/Free Full Text]

Rajakumar, A. and Conrad, K.P. (2000) Expression, ontogeny and regulation of hypoxia inducible transcription factors in the human placenta. Biol. Reprod., 63, 559–569.[Abstract/Free Full Text]

Rajakumar, A., Whitelock, K.A., Weissfeld, L.A., Daftary, A.R., Markovic, N. and Conrad, K.P. (2001) Selective overexpression of the hypoxia-inducible transcription factor, HIF-2alpha, in placentas from women with preeclampsia. Biol. Reprod., 64, 499–506.[Abstract/Free Full Text]

Rawdanowicz, T.J., Hampton, A.L., Nagase, H., Woolley, D.E. and Salamonsen, L.A. (1994) Matrix metalloproteinase secretion by cultured human endometrial stromal cells: identification of interstitial collagenase, gelatinase A, gelatinase B and stromelysin 1. Differential regulation by interleukin-1alpha and tumor necrosis factor beta. J. Clin. Endocrinol. Metab., 79, 530–536.[Abstract]

Reyes, H., Reisz-Porszasz, S. and Hankinson, O. (1992) Ah receptor nuclear translocator protein (ARNT) as a component of the DNA binding form of the Ah receptor. Science, 256, 1193–1195.[ISI][Medline]

Rodgers, W.H., Matrisian, L.M., Guidice, L.C., Dsupin, B., Cannon, P., Sviter, C., Gorstein, F. and Osteen, K.G. (1994) Patterns of matrix metalloproteinase expression in cycling endometrium imply differential functions and regulation by steroid hormones. J. Clin. Invest., 94, 946–953.[ISI][Medline]

Rogers, S., Wells, R. and Rechsteiner, M. (1986) Amino acid sequences common to rapidly degraded proteins: the PEST hypothesis. Science, 234, 364–368.[ISI][Medline]

Romanic, A.M., Burns-Kurtis, C.L., Gout, G., Berrebi-Bertrand, I. and Ohlstein, E.H. (2001) Matrix metalloproteinase expression in cardiac myocytes following myocardial infarction in the rabbit. Life Sci., 68, 799–814.[ISI][Medline]

Rutanen, E.-M. (1998) Insulin-like growth factors in endometrial function. Gynecol. Endocrinol., 12, 399–406.[ISI][Medline]

Salamonsen, L.A. and Lathbury, L.J. (2000) Endometrial leukocytes and menstruation. Hum. Reprod. Update, 6, 16–27.[Abstract/Free Full Text]

Salamonsen, L.A., Butt, A.R., Hammond, F.R., Garcia, S. and Zhang, J. (1997) Production of endometrial matrix metalloproteinases but not their tissue inhibitors is modulated by progesterone withdrawal in an in vitro model for menstruation. J. Clin. Endocrinol. Metab., 82, 1409–1415.[Abstract/Free Full Text]

Salamonsen, L.A., Kovacs, G. and Findlay, J.K. (1999) Current concepts of the mechanisms of menstruation. In Smith, S.K. (ed.), Baillière's Clinical Obstetrics and Gynaecology, Dysfunctional Uterine Bleeding. Baillière Tyndall, London, pp. 161–179.

Salamonsen, L.A. and Woolley, D.E. (1999) Menstruation: induction by matrix metalloproteinases and inflammatory cells. J. Reprod. Immunol., 44, 1–27.[ISI][Medline]

Schatz, F., Papp, C., Toth-Pal, E. and Lockwood, C.J. (1994) Ovarian steroid-modulated stromelysin-1 expression in human endometrial stromal and decidual cells. J. Clin. Endocrinol. Metab., 78, 1467–1472.[Abstract]

Semenza, G.L. (1998) Hypoxia-inducible factor-1: master regulator of O2 homeostasis. Curr. Opin. Genet. Dev., 8, 588–594.[ISI][Medline]

Sharkey, A.M., Day, K., McPherson, A., Malik, S., Licence, S., Smith, S.K. and Charnock-Jones, D.S. (2000) Vascular endothelial growth factor expression in human endometrium is regulated by hypoxia. J. Clin. Endocrinol. Metab., 85, 402–409.[Abstract/Free Full Text]

Singer, C.F., Marbaix, E., Lemoine, P., Courtoy, P.J. and Eeckhout, Y. (1999) Local cytokines induce differential expression of matrix metalloproteinases but not their tissue inhibitors in human endometrial fibroblasts. Eur. J. Biochem., 259, 40–45.[Abstract/Free Full Text]

Speroff, L., Glass, R.H. and Kase, N.G. (1994) The uterus. In Speroff, L., Glass, R.H. and Kase, N.G. (eds), Clinical Gynecologic Endocrinology and Infertility. Williams and Wilkins, Sydney, pp. 109–139.

Steinbrech, D.S., Longtaker, M.T., Mehrara, B.J., Saadeh, P.B., Chin, G.S., Gerrets, R.P., Chau, D.C., Rowe, N.M. and Gittes, G.K. (1999) Fibroblast response to hypoxia: the relationship between angiogenesis and matrix regulation. J. Surg. Res., 84, 127–133.[ISI][Medline]

Talks, K.L., Turley, H., Gatter, K.C., Maxwell, P.H., Pugh, C.W., Ratcliffe, P.J. and Harris, A.L. (2000) The expression and distribution of the hypoxia-inducible factors HIF-1alpha and HIF-2alpha in normal human tissues, cancers, and tumor-associated macrophages. Am. J. Pathol., 157, 411–421.[Abstract/Free Full Text]

Tschugguel, W., Schneeberger, C., Unfried, G., Brautigam, G., Stonek, F., Wieser, F., Vytiska-Binstorfer, E., Czerwenka, K., Weninger, W., Kaider, A. et al. (1999) Elevation of inducible nitric oxide synthase activity in human endometrium during menstruation. Biol. Reprod., 60, 297–304.[Abstract/Free Full Text]

Wang, G.L., Jiang, B.-H., Rue, E.A. and Semenza, G.L. (1995) Hypoxia-inducible factor 1 is a basic-helix–loop–helix-PAS heterodimer regulated by cellular O2 tension. Proc. Natl Acad. Sci. USA, 92, 5510–5514.[Abstract]

Yamanaka, M. and Ishikawa, O. (2000) Hypoxic conditions decrease the mRNA expression of proalpha1 (I) and (III) collagens and increase matrix metalloproteinases-1 of dermal fibroblasts in three-dimensional cultures. J. Dermatol. Sci., 24, 99–104.[ISI][Medline]

Zelzer, E., Levy, Y., Kahana, C., Shilo, B.Z., Rubinstein, M. and Cohen, B. (1998) Insulin induces transcription of target genes through the hypoxia-incuible factor HIF-1alpha/ARNT. EMBO J., 17, 5085–5094.[Abstract/Free Full Text]

Zhang, J., Nie, G.-Y., Wang, J., Woolley, D.E. and Salamonsen, L.A. (1998) Mast cell regulation of endometrial stromal cell matrix metalloproteinases: a mechanism underlying menstruation. Biol. Reprod., 59, 693–703.[Abstract/Free Full Text]

Zhang, J., Hampton, A.L., Nie. G.-Y. and Salamonsen, L.A. (2000) Progesterone inhibits activation of latent matrix metalloproteinase (MMP)-2 by membrane-type 1 MMP: enzymes coordinately expressed in human endometrium. Biol. Reprod., 62, 85–94.[Abstract/Free Full Text]

accepted on October 8, 2001.