Different Basic Fibroblast Growth Factor and Fibroblast Growth Factor ...

0013-7227/03/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 88(6):2853–2859 Copyright © 2003 by The Endocrine Society doi: 10.1210/jc.2002-021434

Different Basic Fibroblast Growth Factor and Fibroblast Growth Factor-Antisense Expression in Eutopic Endometrial Stromal Cells Derived from Women with and without Endometriosis ` , M. VIGNALI, A. MIHALICH, M. REINA, S. MANGIONI, E. PONTI, L. ALBERTI, P. VIGANO A. M. DI BLASIO

AND

Molecular Biology Laboratory, Istituto Auxologico Italiano (A.M., M.R., S.M., E.P., L.A., P.V., A.M.D.B.), 20135 Milan, Italy; Second Department of Obstetrics and Gynecology, University of Milan School of Medicine (M.V.), 20122 Milan, Italy In all species studied, the basic fibroblast growth factor (bFGF) gene is transcribed into multiple mRNAs, one of which is an antisense RNA (1B FGF-AS) probably involved in regulating the stability of the sense transcript. In this study we investigated whether the regulatory mechanisms of bFGF expression might be altered in endometrial stromal cells derived from women with endometriosis. bFGF and 1B FGF-AS mRNA levels were quantified in primary cultures of eutopic endometrial stromal cells derived from 29 women without endometriosis and 24 patients affected by the disease. When the data were analyzed according to the phase of the menstrual cycle, endometrial stromal cells derived from patients in the late proliferative phase showed significantly higher bFGF

E

NDOMETRIOSIS IS A chronic disease defined as the presence of both endometrial glands and stroma outside the uterine cavity. The most widely accepted theory for the development of pelvic endometriosis is that it is a consequence of implantation of viable endometrial tissues in the pelvis via retrograde menstruation (1). As during menses, dissemination of the endometrial cells from the uterus into ectopic locations is a common phenomenon (2), it remains unknown why misplaced endometrial cells implant only in 2–10% of women, which represents the frequency of this disease in the general female population. Evidence suggests that endometriosis is a polygenically inherited disease of complex multifactorial etiology (3). It is certainly an estrogen-dependent pathology, because it has not been reported in prepubertal or premenarchal women (4). Moreover, in most patients, the induction of hypoestrogenism normally results in an involution, although not complete regression, of the endometriotic implants (5). The female reproduction system undergoes profound cyclic changes involving both cell proliferation and angiogenesis. The molecular and cellular mechanisms that mediate these processes in the endometrium are still poorly understood, but these modifications are related to the cyclical change in circulatory levels of ovarian steroids, growth facAbbreviations: bFGF, Basic fibroblast growth factor; EP, early proliferative; FGF-AS, antisense mRNA of basic fibroblast growth factor; HGPRT, hypoxanthine guanine phosphoribosyltransferase; L, luteal; LP, late proliferative.

mRNA values and significantly lower 1B FGF-AS mRNA levels compared with control samples. Furthermore, the mean bFGF/1B FGF-AS mRNA ratio was significantly higher in endometrial stromal cells derived from patients compared with that in controls (mean ⴞ SEM, 2.31 ⴞ 0,55 and 0.77 ⴞ 0.14, respectively; P ⴝ 0.009). Moreover, for bFGF expression the differences existing at the mRNA level were maintained at the protein level. These findings support the hypothesis that 1B FGF-AS mRNA could regulate the expression of the sense transcript and suggest that in endometrial cells derived from patients, the presence of higher bFGF levels could improve their ability to proliferate at the ectopic site. (J Clin Endocrinol Metab 88: 2853–2859, 2003)

tors (6), and cytokines, such as IL-6 (7). Basic fibroblast growth factor (bFGF) is a heparin-binding cationic protein with mitogenic and angiogenic properties (8). The presence of bFGF has been demonstrated in endometriotic tissue by immunohistochemistry (9), and we have previously demonstrated the presence of bFGF and its receptor mRNAs in purified primary cultures of stromal cells isolated from eutopic and ectopic endometrial samples (10). Dysregulation of bFGF expression has been implicated in a variety of pathological conditions involving angiogenesis, vascular smooth muscle cell proliferation, and solid tumor growth (11–17). However, the transcriptional and posttranscriptional mechanisms regulating bFGF expression are complex and still largely unknown. The presence of an endogenous antisense mRNA of bFGF (FGF-AS) has been demonstrated in different mammalian cells (18 –20), including human, and indirect evidence supports the hypothesis that this antisense RNA may regulate bFGF expression. Indeed, steady state levels of the antisense RNA are inversely related to the level of bFGF mRNA during rat embryonic development and in a variety of tumor cell lines (20 –23). In Jurkatt cells, it has been demonstrated that as a result of alternative splicing, two forms of human FGF-AS mRNA are present, each consisting of five exons. Both share exons 2–5 and differ only in the first exon, either 1A or 1B (1A FGF-AS and 1B FGF-AS) (23). The sense and the two antisense transcripts share a 641-bp region of sequence homology at their 3⬘ ends (Fig 1). The FGF-AS gene that codifies for the two transcripts has been localized on the same chromosomal site as the bFGF gene, so it could be

2853

2854

J Clin Endocrinol Metab, June 2003, 88(6):2853–2859

Mihalich et al. • bFGF and FGF-AS in Endometriosis

FIG. 1. Schematic representation of bFGF and 1A/1B FGF-AS transcripts. Exons 4 and 5 of 1A/1B FGF-AS mRNA and complementary regions of bFGF mRNA are indicated by dashed boxes. Dotted lines between antisense exons indicate splicing of intron RNA. The black box indicates the 1B exon.

defined as a human endogenous antisense gene (24). Only the 1B FGF-AS contains an open reading frame of 316 amino acids, predicting a translation product of 35 kDa (GFG) which is 75% and 83% homologous with Xenopus and rat clones, respectively. GFG contains a conserved region, the MutT domain, that is present in a family of proteins involved in preventing mutations and in cleaning the cells of potentially deleterious endogenous metabolites (22, 25). In the present study we have investigated whether an altered expression of bFGF and its antisense transcript could be among the etiopathogenetic mechanisms of endometriosis. To this aim, we set up experiments to evaluate, in different phases of the menstrual cycle, bFGF and 1B FGF-AS mRNA levels in eutopic endometrial stromal cells derived from patients affected by endometriosis at different stages and from women in which the disease was laparoscopically ruled out. The results obtained indicate that in the late proliferative (LP) phase endometrial stromal cells derived from patients had significantly higher bFGF mRNA levels and lower 1B FGF-AS mRNA levels compared with controls. Materials and Methods Tissue collection Endometrium was collected from 53 women scheduled for laparoscopy for ovarian cysts or infertility or pelvic pain. Laparoscopic examination demonstrated the presence of endometriosis in 24 patients. The extent of endometriosis was staged, according to the Revised American Society for Reproductive Medicine system (26), as stage I in 3 patients, stage II in 2, stage III in 15, and stage IV in 4 cases. Twenty-nine women without laparoscopic evidence of endometriosis served as a control group. Intraoperative diagnosis demonstrated benign ovarian cysts in 9 cases, adhesions in 3, myomas in 8, hydrosalpingitis in 2, and normal pelvis in 7 cases. Based on the data of the last and following menstrual periods and of the histological examination of the samples, specimens of endometrium were assigned to the early proliferative (EP) phase in 10 controls and 6 patients, to the LP phase in 12 controls and 10 patients, and to the luteal (L) phase in 7 controls and 8 patients. This research project was granted by the local human institutional investigation committee. All patients were informed in detail about the aims and procedures of the study and subsequently gave their written consent to sample collection. No participant received any hormonal medications in the 3 months before surgery.

Cell preparation and culture We established stromal cell monolayer from eutopic endometrial samples as previously described (10). Briefly, the tissues were gently minced into small pieces (1–2 mm3) and washed in fresh medium to remove mucus or debris. Thereafter, they were incubated for 2 h at 37 C in a shaking water bath in 10 ml Ham’s F-10 containing 0.2% colla-

genase. At the end of the incubation, cell clumps were mechanically dispersed by aspiration through a Pasteur pipette. Single stromal cells were separated from large clumps of epithelium during a 10-min period of differential sedimentation at single gravity. The top 8 ml medium, containing predominantly stromal cells, were then slowly removed, and the cells were collected by centrifugation (200 ⫻ g). The stromal-enriched fraction was washed twice in Ham’s F-10 supplemented with 10% fetal calf serum and antibiotics and was allowed to adhere selectively to 25-cm2 tissue culture dishes for 15 min at 37 C in a 95% air and 5% CO2 incubator. Thereafter, nonattached epithelial cells still present were removed, and a purified stromal preparation was obtained. Total RNA was extracted when the cultures became subconfluent; this was generally achieved within 10 –15 d of cell culture.

RNA extraction and RT Total RNA was extracted from human endometrial stromal cells according to the method of Chomczynski and Sacchi (27). Three micrograms of total RNA were heated for 10 min at 70 C in presence of 1 ␮g random primers. Then the RNA and random primers were incubated for 60 min at 37 C in presence of RT buffer, 200 U Moloney murine leukemia virus reverse transcriptase, 0.5 mm of each deoxynucleotide triphosphate, and 25 U recombinant RNasin ribonuclease inhibitor (Promega Corp., Madison, WI) in a total volume of 25 ␮l.

cDNA normalization Amplification of the hypoxanthine guanine phosphoribosyltransferase (HGPRT) first strand cDNA was performed using 200 ng total cDNA in a 50-␮l volume containing 1⫻ Taq polymerase buffer (1.5 mm magnesium chloride), 0.2 mm of each deoxynucleotide triphosphate, 12.5 ␮m of each HGPRT primer, and 2.5 U Taq polymerase (Promega Corp.). The HGPRT PCR conditions were: initial denaturation at 94 C for 5 min, followed by 28 cycles consisting of denaturation for 30 sec at 94 C, annealing for 30 sec at 52 C, and extension for 30 sec at 72 C. Final extension was performed for 5 min at 72 C. To exclude the possible genomic DNA contamination, HGPRT-specific primers were designed to amplify, after RT, a 97-bp product from mRNA and a 267-bp product from DNA. HGPRT primers are reported in Table 1. PCR products were electrophoresed through 3% agarose gels and visualized with ethidium bromide. In every HGPRT amplification, a constant amount of cDNA (resulting from RT of 120 ng total RNA) derived from K562 cells was used. The fluorescence of HGPRT amplification products was analyzed by a densitometer, and the DNA fragment derived from K562 cells was used as the cDNA external standard for cDNA normalization as previously described (28). We performed this analysis to assure that subsequent amplification reactions used the same amount of cDNA for each sample.

Competitive-PCR Competitor construction. We constructed specific competitors for bFGF, 1A FGF-AS, and 1B FGF-AS cDNAs that were amplified under the same conditions as those used for the human correspondent cDNA, but could be distinguished from the targets. To construct the bFGF competitor, we


J Clin Endocrinol Metab, June 2003, 88(6):2853–2859 2855

TABLE 1. Oligonucleotide primer sequences and GenBank accession numbers Primers

Sequences

HGPRT forward HGPRT reverse bFGF forward bFGF reverse 1A FGF-AS forward 1A FGF-AS reverse 1B FGF-AS forward a 1B FGF-AS reverse b 1B FGF-AS forward c 1B FGF-AS reverse d

5⬘-GCTTGCTGGTGAAAAGGACC-3⬘ 5⬘-GTCAAGGGCATATCCTACAAC-3⬘ 5⬘-GGCTTCTTCCTCCTGCGCATCCA-3⬘ 5⬘-GCTCTTAGCAGACATTGGAAGA-3⬘ 5⬘-GCTGTAACGGCATCTGTGAAAG-3⬘ 5⬘-CCTAGCAACTCTGCTGGTGATGGGAG-3⬘ 5⬘-CTTCGGCGGGTTACCGCTGGGCCTCGGGCGCAC-3⬘ 5⬘-GCCTAGCAACTCTGCTGGTGATGGGAG-3⬘ 5⬘-CATTTGGGCAACGGACGAATC-3⬘ 5⬘-CAGTCAGATGTTTAATCAATC-3⬘

Location

nt nt nt nt nt nt nt nt nt nt

39815-39834 40083-40063 578-597 931-910 17-38 738-712 99-131 841-815 1-21 1075-1055

Accession no.

M26434 J04513 AF019632 AF019633 AF019633

nt, Nucleotide.

amplified the rat bFGF cDNA fragment included between the specific primers indicated in Table 1. The human and rat bFGF cDNA have a complete homology in these primers’ sequences, the length of the amplification product is the same, and the overall homology of the entire two fragments is 89%. In the rat fragment the 35 nucleotide is different from the human one, and thus an HincII restriction site is missing. We cloned the DNA fragment obtained from the rat cDNA amplification into the pTarget vector (Promega Corp.) and subsequently amplified it by PCR. The purified DNA was quantified by a spectrophotometer, and different competitor dilutions were prepared. To construct the 1A FGF-AS and 1B FGF-AS competitors, we amplified both human cDNA sequences included between the couples of primers for 1A FGF-AS and 1B FGF-AS (1B FGF-AS forward a and reverse b; Table 1). We cloned the two products of amplification into the pTarget vector (Promega Corp.). There was an ApaI restriction site in the sequences of both 1A FGF-AS and 1B FGF-AS cloned fragments. At this site we inserted a small DNA fragment of 160 bp. Thereafter, the two cloned fragments were amplified by PCR and quantified using a spectrophotometer. Different competitor dilutions were then prepared. Competitive PCR analysis. A constant amount of cDNA was coamplified with known concentrations of the specific competitors in 50 ␮l PCR mixture containing 1⫻ Taq polymerase buffer (1.5 mm magnesium chloride), 0.2 mm of each deoxynucleotide triphosphate, 25 ␮m of each specific primer, and 2.5 U Taq polymerase (Promega Corp.). Oligonucleotide primers specific for bFGF, 1A FGF-AS, and 1B FGF-AS are reported in Table 1. After an initial denaturation step at 95 C for 5 min, PCR conditions were: 1) for bFGF amplification, 27 cycles of 94 C for 30 sec, 60 C for 30 sec, and 72 C for 30 sec; 2) for 1A FGF-AS amplification, 33 cycles of 94 C for 45 sec, 64 C for 30 sec, and 72 C for 1 min; and 3) for 1B FGF-AS amplification, 30 cycles of 94 C for 30 sec, 65 C for 30 sec, and 72 C for 1 min. Ten microliters of bFGF PCR products were digested with 10 U of the restriction enzyme HincII at 37 C overnight and then resolved on 4% agarose gel stained with ethidium bromide. PCR products of 1A and 1B FGF-AS amplifications were directly resolved on 2% agarose stained with ethidium bromide. In each experiment the fluorescence of both target and competitor products was quantified by a densitometer as absorbance. The competitor/target absorbance ratio was plotted against known concentrations of the competitor. The amount of each mRNA was determined by the equation obtained from the regression curve of the plotted values. The results were expressed as femtograms of the specific mRNA per nanogram of total RNA.

Semiquantitative PCR for bFGF A constant amount of cDNA was amplified in 50 ␮l PCR mixture containing 1⫻ Taq polymerase buffer (1.5 mm magnesium chloride), 0.2 mm of each deoxynucleotide triphosphate, 25 ␮m of each specific primer, and 2.5 U Taq polymerase (Promega Corp.). Oligonucleotide primers specific for bFGF or HGPRT are reported in Table 1. PCR conditions were the same as those used in the competitive PCR and for the cDNA normalization. bFGF and HGPRT amplification products were resolved on a 3% agarose stained with ethidium bromide. The fluorescence of both fragments was quantified by a densitometer, and the bFGF/ HGPRT absorbance ratio was used to estimate bFGF mRNA levels in each sample.

Sequence analysis The identities of all PCR DNA fragments were confirmed by sequence analysis. Sequencing reactions were prepared employing ABI Big Dye terminator chemistry (PE Applied Biosystems, Foster City, CA) and were analyzed using the ABI PRISM 310 genetic analyzer (PE Applied Biosystems).

Protein extraction Human endometrial stromal cells were resuspended in 5% sodium dodecyl sulfate and 0.125 m Tris (pH 6.8) and boiled for 5 min. The cell lysis products were then sonicated, and protein concentrations were determined by the bicinchoninic acid protein assay (Pierce Chemical Co., Rockford, IL).

Western blot analysis Equal amounts of protein (40 ␮g) from cell lysates were separated on a sodium dodecyl sulfate-15% polyacrylamide gel, electrophoretically transferred to nitrocellulose membranes, and allowed to stick to the filters for 1 h at room temperature in 5% nonfat milk in Tris-buffered saline with 0.5% Tween 20. Membranes were incubated with dilutions of primary antibodies in 5% nonfat milk in Tris-buffered saline with 0.5% Tween 20. The antibodies used were a monoclonal antihuman bFGF antiserum (Sigma-Aldrich Corp., St. Louis, MO) and a polyclonal antiserum that recognizes the MutT domain of the rat FGF-AS protein (gift from Dr. P. R. Murphy). Bound antibody was detected using sheep antimouse or antirabbit IgG conjugated to horseradish peroxidase (Amersham Pharmacia Biotech, Arlington Heights, IL) and the enhanced chemiluminescence reagent (Amersham Pharmacia Biotech). The stained immunoblots were placed against x-ray film (Kodak, Rochester, NY) to generate chemiluminographs.

Transfection and expression of 1B FGF-AS cDNA The full-length 1B FGF-AS cDNA was amplified using primers 1B FGF-AS forward (c) and reverse (d) and was subcloned into the eukaryotic expression vector pTarget vector (Promega Corp.). The clone was transfected into wild-type HEK 293T cells using the calcium phosphate method (29). The expression of the construct was confirmed by Western blot.

Statistical analysis Data are expressed as the mean ⫾ sem. Differences between groups were determined, as appropriate, by unpaired t test; ANOVA, followed by Fisher’s least significant difference test, was used as the posttest. The correlation between bFGF sense and antisense mRNA levels was evaluated by linear regression analysis. Statistical significance was considered when P ⬍ 0.05.

2856


Results bFGF mRNA expression in eutopic endometrial stromal cells in different phases of the menstrual cycle

Levels of bFGF mRNA were examined in eutopic endometrial stromal cells derived from 28 controls and 22 patients affected by endometriosis using competitive PCR. Figure 2 shows the results of bFGF mRNA quantification in the different phases of the menstrual cycle. In both groups of women, levels of bFGF mRNA were between 0.5–5.9 fg/ng total RNA. In controls, bFGF mRNA did not significantly change during the entire menstrual cycle. In contrast, in women with endometriosis, bFGF mRNA levels were higher in the LP phase (mean ⫾ sem, 3.8 ⫾ 0.40 fg/ng total RNA) compared with both the EP and the L phases (mean ⫾ sem, 1.9 ⫾ 0.38 and 2.3 ⫾ 0.40 fg/ng total RNA, respectively). These differences were statistically significant (P ⫽ 0.006 for LP vs. EP; P ⫽ 0.01 for LP vs. L). 1B FGF-AS mRNA levels in eutopic endometrial stromal cells in different phases of the menstrual cycle

The results of competitive PCR quantification of 1B FGF-AS mRNA obtained from 27 controls and 23 patients are summarized in Fig. 3. 1B FGF-AS mRNA levels in endome-

FIG. 2. Levels of bFGF mRNA in eutopic endometrial stromal cells derived from control women (white circles) and women with endometriosis (dashed circles). The graph represents the results obtained in the different phases of the menstrual cycle. The lines represent the mean values. The mean value in the LP phase is significantly higher than those in the other two phases (P ⫽ 0.006, LP vs. EP; P ⫽ 0.01, LP vs. L).

FIG. 3. Levels of 1B FGF-AS mRNA in eutopic endometrial stromal cells derived from control women (white circles) and women with endometriosis (dashed circles). The graph represents the results obtained in the different phases of the menstrual cycle. The lines represent the mean values. The mean value in the LP phase is significantly lower than that in the EP phase (P ⫽ 0.03).


trial cells derived from controls did not vary significantly as a function of the different phases of the menstrual cycle. The mean values were 2.9 ⫾ 0.7 3.9 ⫾ 0.51, and 2.9 ⫾ 0.75 fg/ng total RNA in the EP, LP, and L phases, respectively. As observed for the sense transcript, 1B FGF-AS mRNA levels present in endometrial cell samples derived from women with endometriosis showed a different pattern of expression compared with those derived from the control group. Indeed, during the LP phase, 1B FGF-AS mRNA levels were significantly lower than those in the EP phase (P ⫽ 0.03) and were lower, although not significantly, compared with those in the L phase. In all samples examined 1A FGF-AS mRNA levels were extremely low and hardly comparable (data not shown). Expression of bFGF and 1B FGF-AS in the LP phase

Linear regression analysis indicated that bFGF and 1B FGF-AS mRNA levels were significantly correlated in controls (P ⫽ 0.04), whereas in women with endometriosis this correlation was not present (data not shown). Taken together, the results obtained suggested that transcription of bFGF and 1B FGF-AS in endometrial stromal cells could be differently regulated in the two groups of women studied,



FIG. 4. Levels of bFGF mRNA (A) and 1B FGF-AS mRNA (B) present during the LP phase in eutopic endometrial stromal cells derived from control women (white circles) and women with endometriosis (dashed circles). The lines indicate the mean values. *, P ⫽ 0.04; **, P ⫽ 0.004.

FIG. 5. bFGF mRNA/1B FGF-AS mRNA ratio in endometrial samples obtained during the LP phase from control women (white circles) and women with endometriosis (dashed circles). The lines indicate the mean values. *, P ⫽ 0.009 vs. controls.

and the more relevant differences were present during the late proliferative phase. As shown in Fig. 4A, in this phase bFGF mRNA levels were significantly higher in cells derived from women with endometriosis than in those from controls (mean ⫾ sem, 3.8 ⫾ 0.40 and 2.7 ⫾ 0.33 fg/ng total RNA, respectively; P ⴝ 0.04). The opposite pattern was observed for 1B FGF-AS mRNA, as its levels were significantly lower in cells derived from patients compared with those from controls (mean ⫾ sem, 1.9 ⫾ 0.24 and 3.9 ⫾ 0.51 fg/ng total RNA, respectively; P ⫽ 0.004; Fig. 4B). We then evaluated in each sample derived from women with and without endometriosis in the LP phase the ratio between the amounts of bFGF and 1B FGF-AS mRNAs. Figure 5 shows that the bFGF/1B FGF-AS mRNA ratios were 0.77 ⫾ 0.14 and 2.31 ⫾ 0.55 in controls and patients, respectively, and the difference was statistically significant (P ⫽ 0.009). These data indicate that in endometrial stromal cells, bFGF mRNA levels were more than 2-fold higher than 1B FGF-AS mRNA levels in women with endometriosis, but were slightly lower in controls.

FIG. 6. Representative experiments on bFGF mRNA and protein determinations in endometrial stromal cells derived from two control women (white bar) and two patients with endometriosis (dashed bar). A, bFGF/HGPRT ratio, expressed in arbitrary units. B, Western blot analysis of bFGF protein performed in the same samples. The molecular weights indicate the three different forms of bFGF protein.

were also present at the protein level, we determined, by semiquantitative PCR and Western blot, bFGF mRNA and protein levels present in the same sample derived from patients and controls. In each experiment we evaluated one control and one patient with endometriosis, and a total of eight pairs of women were examined. In every experiment the differences between patients with endometriosis and controls at the mRNA level were also maintained at the protein level (Fig. 6, A and B). To perform the same study on 1B FGF-AS mRNA and protein levels, we first studied GFG protein expression in 10 patients and 11 controls by Western blot, using as a positive control proteins derived from HEK 293T cells transfected with the full-length 1B FGF-AS cDNA. We could not detect GFG protein in any of the samples examined (data not shown).

bFGF and GFG protein in eutopic endometrial stromal cells

1B FGF-AS promoter sequences

To investigate whether the differences found between women with and without endometriosis at the mRNA level

We hypothesized that the different 1B FGF-AS expression observed in the two groups of women could be caused by

2858


mutations in the antisense promoter sequence. For this reason we sequenced the two regions upstream of the 1A (409 bp upstream) and 1B (187 bp upstream) exons (23) in eight patients and five controls. In these regions different potential binding sites for transcription factors, two half estrogenresponsive element (GGTCA) (30), and one dioxin-responsive enhancer (GCGTG) (31) were present at nucleotides –261, –114, and –33 from the transcription start site, respectively. We did not find any mutations. Discussion

The most widely accepted theory for the development of pelvic endometriosis is that it is a consequence of implantation of viable endometrial tissue in the pelvis via retrograde menstruation. As this is a common phenomenon, it remains unknown why the disease arises in only a limited number of women. Experimental evidence indicates that endometriosis is a multifactorial disease in which genetic and environmental components are involved. bFGF is constitutively present in the human endometrium and, due to its mitogenic and angiogenic activities, is most likely involved in determining endometrial tissue modifications during the menstrual cycle (9, 10, 32, 33). We as well as other investigators have also demonstrated the presence of bFGF and its receptor in eutopic and ectopic human endometrium (9, 10). Recent evidence suggests that the expression of bFGF could be regulated by the presence of a naturally occurring antisense transcript, FGF-AS. Indeed, several examples of natural endogenous antisense transcripts have been reported in eukaryotic systems (34 –37), but in the majority of the cases the functions of these molecules are still largely unknown. To date, the expression of the FGF-AS transcript in the human species has been investigated in few tissues, including normal and neoplastic pituitaries (19, 24). Asa and co-workers (24) demonstrated that in normal pituitary, FGF-AS is expressed, whereas bFGF is undetectable. Instead, pituitary adenomas expressed bFGF, but had reduced levels of FGF-AS. These data support the hypothesis that the antisense transcript could regulate the expression of bFGF (24). Based on these findings it is possible to postulate that different expressions of bFGF and FGF-AS could also modulate the proliferative capacity of human endometrial cells. The results of the present study are in line with this hypothesis. In eutopic endometrial cells derived from control women there is a significant correlation between bFGF sense and antisense mRNA levels, whereas this is not present in cells derived from women with endometriosis. Indeed, during the LP phase of the menstrual cycle, patients had bFGF mRNA levels significantly higher and 1B FGF-AS mRNA levels significantly lower than those detected in controls. As a consequence, the mean bFGF/1B FGF-AS ratio was higher in patients than in controls, and the difference was statistically significant. These data strongly suggest that the antisense transcript could negatively regulate bFGF expression by inducing degradation of its mRNA. Similar to what observed in the Xenopus oocyte (18), in human endometrial stromal cells higher levels of 1B FGF-AS mRNA corresponded to lower levels of both bFGF mRNA and protein. In


contrast, in rat glioma cells the presence of the antisense transcript influenced bFGF protein levels, but not mRNA levels (38). It is worth noting that the different expressions of the sense and antisense bFGF transcripts in endometrial cells of women with and without endometriosis are evident during the LP phase of the menstrual cycle, when estrogen levels reach their peak. The effect of estrogen on bFGF expression is still a matter of controversy. Some studies indicate that bFGF expression correlates with the expected estrogen levels throughout the menstrual cycle (33, 39), whereas others, performed in both humans and rats, have shown no correlation (32, 40). In contrast to these previous reports, in the present study we evaluated bFGF expression in highly purified primary cultures of endometrial stromal cells. In cells derived from controls, we did not find relevant differences in bFGF mRNA levels as a function of the menstrual cycle, thus confirming an absence of the estrogen effects. In contrast, in cells derived from patients, bFGF mRNA levels were significantly higher in the LP phase, and this is most likely a consequence of the concomitantly lower expression of the 1B FGF-AS transcript. These results lead us to speculate that in women with endometriosis, estrogens could indirectly influence bFGF expression by modulating the transcription of its antisense RNA. It should be noted that in the promoter region of the 1B FGF-AS gene there are two half estrogen-responsive elements (30) and two dioxin-responsive elements (31). However, no mutations in the nucleotide sequences of these regions were observed in any of the patients examined. Thus, the possible role of estrogens in the different expressions of bFGF and 1B FGF-AS needs to be further clarified. Besides the effects on the regulation of bFGF expression, the 1B FGF-AS transcript could be directly involved in the pathogenesis of endometriosis acting through its protein product GFG. This protein contains a conserved region, the MutT domain, that is present in a family of proteins that are involved in preventing mutations and in cleaning the cells of potentially deleterious endogenous metabolites caused by environmental contaminants and oxidative damage (25, 41). Several studies have highlighted that oxidative stress and, in particular, exposure to dioxin might induce the development of endometriosis (41– 44). Based on these two observations, it is tempting to speculate that a decreased synthesis of GFG protein could be among the causes of increased endometrial cell sensibility to the effects of oxidative stress. At present this remains a speculative, as in this study we could not detect GFG protein in endometrial stromal cells. We cannot rule out that this is due to the antibody used, which may not have been sensitive enough to detect low levels of the protein. Few data are currently available on the activity of the GFG protein in humans; thus, we are undertaking further studies to shed new light on its presence and function in endometrial cells. In conclusion, the results presented herein clearly indicate that in women with endometriosis, eutopic endometrial stromal cells are characterized by an increased expression of bFGF, which physiologically is a mitogenic and angiogenic peptide. Recently, Nisolle et al. (45) reported that in nude mice during the initial phases of development of endometriosis, endometrial stromal cells are the cells that attach to the mesothelium, proliferate, and then induce the early an-


giogenesis. The present findings in association with these experimental data, allow us to postulate that an altered expression of bFGF could be one of the genetic determinants that contributes to the development of endometriosis lesions. Acknowledgments We are grateful to Dr. P. R. Murphy for the generous gift of the antibody for the rat GFG protein MutT domain. Received September 11, 2002. Accepted March 4, 2003. Address all correspondence and requests for reprints to: Dr. A. M. Di Blasio, Molecular Biology Laboratory, Istituto Auxologico Italiano, Viale Montenero 32, 20135 Milan, Italy. E-mail: [email protected].

References 1. Sampson JA 1927 Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 14:422– 425 2. Halme J, Hammond MG, Hulka JF, Raj SG, Talbert LM 1984 Retrograde menstruation in healthy women and in patients with endometriosis. Obstet Gynecol 64:151–155 3. Olive DL, Schwartz LB 1993 Endometriosis. N Engl J Med 328:1759 –1769 4. Gleicher N, el-Roeiy A, Confino E, Friberg J 1987 Is endometriosis an autoimmune disease? Obstet Gynecol 70:115–122 5. Shaw RW 1992 Treatment of endometriosis. Lancet 340:1267–1271 6. Smith SK 1995 The aetiology of endometriosis. Hum Reprod 10:1274 7. Bergqvist A, Bruse C, Carlberg M, Carlstrom K 2001 Interleukin 1␤, interleukin-6, and tumor necrosis factor-␣ in endometriotic tissue and in endometrium. Fertil Steril 75:489 – 495 8. Gospodarowicz D, Ferrara N, Schweigerer L, Neufeld G 1987 Structural characterisation and biological functions of fibroblast growth factor. Endocr Rev 8:95–114 9. Ferriani RA, Charnock-Jones DS, Prentice A, Thomas EJ, Smith SK 1993 Immunohistochemical localization of acidic and basic fibroblast growth factors in normal human endometrium and endometriosis and the detection of their mRNA by polymerase chain reaction. Hum Reprod 8:11–16 10. Di Blasio A M, Centinaio G, Carniti C, Somigliana E, Vigano` P, Vignali M 1995 Basic fibroblast growth factor messenger ribonucleic acid levels in eutopic and ectopic human endometrial stromal cells as assessed by competitive polymerase chain reaction amplification. Mol Cell Endocrinol 115:169 –175 11. Becker D, Meier CB, Herlyn M 1989 Proliferation of human malignant melanomas is inhibited by antisense oligodeoxynucleotides targeted against basic fibroblast growth factor. EMBO J 8:3685–3691 12. Morrison RS 1991 Suppression of basic fibroblast growth factor expression by antisense oligodeoxynucleotides inhibits the growth of transformed human astrocytes. J Biol Chem 266:728 –734 13. Sato Y, Hamanaka R, Ono J, Kuwano M, Rifkin DB, Takaki R 1991 The stimulatory effect of PDGF an vascular smooth muscle cell migration is mediated by the induction of endogenous basic FGF. Biochem Biophys Res Commun 174:1260 –1266 14. Yamanishi DT, Graham MJ, Florkiewicz RZ, Buckmeier JA, Meyskens Jr FL 1992 Differences in basic fibroblast growth factor RNA and protein levels in primary melanocytes and metastatic melanoma cells. Cancer Res 52:5024 –5029 15. Murphy PR, Sato Y, Knee RS 1992 Phosphorothioate antisense oligonucleotides against basic fibroblast growth factor inhibit anchorage-dependent and anchorage-independent growth of malignant glioblastoma cell line. Mol Endocrinol 6:877– 884 16. Ross R 1993 The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature 362:801– 809 17. Graeven U, Rodeck U, Karpinski S, Jost M, Philippou S, Schmiegel W 2001 Modulation of angiogenesis and tumorigenicity of human melanocytic cells by vascular endothelial growth factor and basic fibroblast growth factor. Cancer Res 61:7282–7290 18. Kimelman D, Kirschner MW 1989 An antisense mRNA directs the covalent modification of the transcript encoding fibroblast growth factor in Xenopus oocytes. Cell 59:687– 696 19. Knee R, Pitcher SE, Murphy PR 1994 Basic fibroblast growth factor sense (FGF) and antisense (GFG) RNA transcripts are expressed in unfertilized human oocytes and in differentiated adult tissues. Biochem Biophys Res Commun 205:577–583 20. Murphy PR, Knee RS 1994 Identification and characterization of an antisense


21. 22. 23. 24. 25. 26. 27. 28.

29. 30. 31. 32. 33.

34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45.

RNA transcript (gfg) from the human basic fibroblast growth factor gene. Mol Endocrinol 8:852– 859 Li AW, Too CKL, Murphy PR 1996 The basic fibroblast growth factor (FGF-2) antisense RNA (GFG) is translated into a MutT-related protein in vivo. Biochem Biophys Res Commun 223:19 –23 Knee R, Li A, Murphy PR 1997 Characterization and tissue-specific expression of the rat basic fibroblast growth factor antisense mRNA and protein. Proc Natl Acad Sci USA 94:4943– 4947 Gagnon ML, Moy GK, Klagsbrun M 1999 Characterization of the promoter for the human antisense fibroblast growth factor-2 gene: regulation by Ets in Jurkat cells. J Cell Biochem 72:492–506 Asa S L, Ramayar L, Murphy PR, Li AW, Ezzat S 2001 The endogenous fibroblast growth factor-2 gene product regulates pituitary cell growth and hormone production. Mol Endocrinol 15:589 –599 Bessman MJ, Frick DN, O’Handley SF 1996 The MutT proteins or “Nudix” hydrolases, a family of versatile, widely distributed, “housecleaning” enzymes. J Biol Chem 271:25059 –25062 American Society for Reproductive Medicine 1997 Revised American Society for Reproductive Medicine classification of endometriosis: 1996. Fertil Steril 67:817– 821 Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidinium thiocynate-phenol-chloroform extraction. Anal Biochem 162: 156 –159 Fugazzola L, Mihalich A, Persani L, Cerutti N, Reina M, Bonomi M, Ponti E, Mannavola D, Giammona E, Vannucchi G, Di Blasio AM, Beck-Peccoz P 2002 Highly sensitive serum thyroglobulin and circulating thyroglobulin mRNA evaluations in the management of patients with differentiated thyroid cancer in apparent remission. J Clin Endocrinol Metab 87:3201–3208 Graham FL, van der Eb AJ 1973 A new technique for the assay of infectivity of human adenovirus 5 DNA. Virol 52:456 Klein-Hitpass L, Schorpp M, Wagner U, Ryffel GU 1986 An estrogenresponsive element derived from the 5⬘ flanking region of the Xenopus vitellogenin A2 gene functions in transfected human cells. Cell 46:1053–1061 Wang F, Hoivik D, Pollenz R, Safe S 1998 Functional and physical interactions between the estrogen receptor Sp1 and nuclear aryl hydrocarbon receptor complexes. Nucleic Acids Res 26:3044 –3052 Rusnati M, Casarotti G, Pecorelli S, Ragnotti G, Presta M 1990 Basic fibroblast growth factor in ovulatory cycle and postmenopausal human endometrium. Growth Factors 3:299 –307 Sangha RK, Li XF, Shams M, Ahmed A 1997 Fibroblast growth factor receptor-1 is a critical component for endometrial remodeling: localization and expression of basic fibroblast growth factor and FGF-R1 in human endometrium during the menstrual cycle and decreased FGF-R1 expression in menorrhagia. Lab Invest 77:389 – 402 Adelman JP, Bond CT, Douglass J, Herbert E 1987 Two mammalian genes transcribed from opposite strands of the same DNA locus. Science 235:1514 – 1517 Lazar MA, Hodin RA, Darling DS, Chin WW 1989 A novel member of the thyroid/steroid hormone receptor family is encoded by the opposite strand of the rat c-erbA␣ transcriptional unit. Mol Cell Biol 9:1128 –1136 Krystal GW, Armstrong BC, Battey JF 1990 N-myc mRNA forms an RNA-RNA duplex with endogenous antisense transcripts. Mol Cell Biol 10:4180 – 4191 Rivkin M, Rosen KM, Villa-Komaroff L 1993 Identification of an antisense transcript from the IGF-II locus in mouse. Mol Reprod Dev 35:394 –397 Li AW, Murphy PR 2000 Expression of alternatively spliced FGF-2 antisense RNA transcripts in the central nervous system: regulation of FGF-2 mRNA translation. Mol Cell Endocrinol 162:69 –78 Fujimoto J, Hori M, Ichigo S, Tamaya T 1996 Expression of basic fibroblast growth factor and its mRNA in uterine endometrium during the menstrual cycle. Gynecol Endocrinol 10 :193–197 Wordinger RJ, Moss AE, Lockard T, Gray D, Chang I-FC, Jackson TL 1992 Immunohistochemical localization of basic fibroblast growth factor within the mouse uterus. J Reprod Fertil 96:141–152 Taddei F, Hayakawa H, Bouton MF, Cirinesi AM, Matic I, Sekiguchi M, Radman M 1997 Counteraction by MutT protein of transcriptional errors caused by oxidative damage. Science 278:128 –130 Shertzer HG, Nebert DW, Puga A, Ary M, Sonntag D, Dixon K, Robinson LJ, Cianciolo E, Dalton TP 1998 Dioxin causes a sustained oxidative stress response in the mouse. Biochem Biophys Res Commun 253:44 – 48 Yang JZ, Agarwal SK, Foster WG 2000 Subchronic exposure to 2, 3, 7, 8tetrachlorodibenzo-p-dioxin modulates the pathophysiology of endometriosis in the cynomolgus monkey. Toxicol Sci 56:374 –381 Van Langendonckt A, Casanas-Roux F, Donnez J 2002 Oxidative stress and peritoneal endometriosis. Fertil Steril 77:861– 870 Nisolle M, Casanas-Roux F, Donnez J 2000 Early-stage endometriosis: adhesion and growth of human menstrual endometrium in nude mice. Fertil Steril 74:306 –312