{"paper_id":"68c59222-2b68-46da-a58c-f0baeff1d316","body_text":"REVIEW\nEstrogen biosynthesis in endometriosis: molecular basis\nand clinical relevance\nS E Bulun 1, K M Zeitoun 2, K Takayama 3 and H Sasano3\n1Departments of Obstetrics and Gynecology and Molecular Genetics, University of Illinois at Chicago,\n820 S. Wood St. M/C 808, Illinois 60612, USA\n2Department of Obstetrics and Gynecology, Columbia University College of Physicians and Surgeons,\n622 W. 168th St., New York, New York 10032–3702, USA\n3Departments of Pathology and Obstetrics and Gynecology, Tohoku University School of Medicine,\n2–1 Seiryo Machi, Sendai-Shi 980, Japan\n(Requests for o ffprints should be addressed t o S E Bulun; Email: sbulun @uic.edu)\nABSTRACT\nConversion of C 19 steroids to estrogens is catalyzed\nby aromatase in human ovary, placenta and\nextraglandular tissues such as adipose tissue, skin\nand the brain. Aromatase activity is not detectable\nin normal endometrium. In contrast, aromatase is\nexpressed aberrantly in endometriosis and is\nstimulated by prostaglandin E\n2 (PGE2). This results\nin local production of estrogen, which induces\nPGE\n2 formation and establishes a positive feedback\ncycle. Another abnormality in endometriosis, i.e.\ndeficient hydroxysteroid dehydrogenase (17 /afii9826-HSD)\ntype 2 expression, impairs the inactivation of\nestradiol to estrone. These molecular aberrations\ncollectively favor accumulation of increasing quan-\ntities of estradiol and PGE\n2 in endometriosis. The\nclinical relevance of these findings was exemplified\nby the successful treatment of an unusually\naggressive case of postmenopausal endometriosis\nusing an aromatase inhibitor.\nJournal of Molecular Endocrinology (2000) 25, 35–42\nINTRODUCTION\nEndometriosis is a chronic disease manifested by\npelvic pain and infertility and defined as the\npresence of endometrial glands and stroma within\nthe pelvic peritoneum and other extra-uterine sites.\nIt is estimated to a ffect 2–10% of women in the\nreproductive age group (Vessey et al. 1993, Kjerulff\net al. 1996). Endometriosis is viewed to be a\npolygenically inherited disease of complex multi-\nfactorial etiology (Olive & Schwartz 1993).\nSampson’s theory of transplantation of endometrial\ntissue on the pelvic peritoneum via retrograde men-\nstruation is the most widely accepted explanation for\nthe development of pelvic endometriosis because\nof convincing circumstantial and experimental\nevidence (Sampson 1927). Since retrograde men-\nstruation is observed in almost all cycling women,\nendometriosis is postulated to develop as a result\nof the coexistence of a defect in clearance of the\nmenstrual e ffl ux from pelvic peritoneal surfaces,\npossibly involving the immune system (Halme et al.\n1988). Alternatively, intrinsic molecular aberrations\nin pelvic endometriotic implants were proposed to\ncontribute significantly to development of endo-\nmetriosis. Aberrant expression of aromatase, certain\ncytokines and tissue metalloproteinases, deficiency\nof 17 /afii9826-hydroxysteroid dehydrogenase (17 /afii9826-HSD)\ntype 2 and resistance to the protective action of\nprogesterone are some of these molecular abnormali-\nties (Khorram et al. 1993, Sharpe-Timms et al. 1995,\nNoble et al. 1996, Osteen et al. 1996, Bruner et al.\n1997, Zeitoun et al. 1998, 1999). Since endometriosis\nis an estrogen-dependent disorder, aromatase ex-\npression and 17 /afii9826-HSD type 2 deficiency are of\nparamount importance in the pathophysiology of\nendometriosis. In this article, aberrant mechanisms\nof estrogen biosynthesis and metabolism in women\nwith endometriosis are reviewed, with emphasis on\nidentifying targets for new treatment strategies.\n35\nJournal of Molecular Endocrinology (2000) 25, 35–42\n0952–5041/00/025–035 /p302000 Society for Endocrinology Printed in Great Britain\nOnline version via http://www.endocrinology.org\nDownloaded from Bioscientifica.com at 06/12/2026 04:02:00PM\nvia free access\n\n\nDISCUSSION\nEstrogen biosynthesis and metabolism in\nhumans\nThe conversion of androstenedione and testosterone\nto estrone and estradiol is catalyzed by aromatase,\nwhich is expressed in a number of human tissues\nand cells such as ovarian granulosa cells, placental\nsyncytiotrophoblast, adipose tissue and skin fibro-\nblasts, and the brain. In the reproductive-age\nwoman, the ovary is the most important site of\nestrogen biosynthesis, and this takes place in a cyclic\nfashion. Upon binding of follicle-stimulating hor-\nmone (FSH) to its G-protein-coupled receptor in\nthe granulosa cell membrane, intracellular cAMP\nlevels rise and enhance binding of two critical\ntranscription factors, i.e. steroidogenic factor-1\n(SF-1) and cAMP response element binding protein\n(CREB), to the classically located proximal pro-\nmoter II of the aromatase gene (Michael et al. 1995,\n1997). This, in turn, activates aromatase expression\nand consequently estrogen secretion from the\npre-ovulatory follicle (Simpson et al. 1994, Michael\net al. 1995).\nOn the other hand, in postmenopausal women,\nestrogen formation takes place in extra-ovarian\ntissues such as the adipose tissue and skin\n(MacDonald et al. 1967, 1978, Ackerman et al.\n1981) (Fig. 1). In contrast to cAMP regulation of\naromatase expression in the ovary, this is controlled\nprimarily by cytokines (IL-6, IL-11, TNF /afii9825) and\nglucocorticoids via the alternative use of promoter\nI.4 in adipose tissue and skin fibroblasts (Simpson\net al. 1994). The major substrate for aromatase in\nadipose tissue and skin is androstenedione of\nadrenal origin. In postmenopausal women, approxi-\nmately 2% of circulating androstenedione is con-\nverted to estrone, which is further converted to\nestradiol in these extra-ovarian tissues. This may\ngive rise to signi ficant serum levels of estradiol\ncapable of causing endometrial hyperplasia or even\ncarcinoma (MacDonald et al. 1967, 1978).\nAromatase expression in Müllerian-derived\ntissues\nMüllerian tissues are known targets of estrogen\naction. Until recently, estrogen action has been\nclassically viewed to occur only via an ‘endocrine’\nmechanism: in other words, it was thought that only\ncirculating estradiol, whether secreted by the ovary\nor formed in the adipose tissue, could exert an\nestrogenic e ffect after delivery to target tissues via\nthe bloodstream. Studies on aromatase expression\nin breast cancer demonstrated that paracrine\nmechanisms play an important role in estrogen\naction in this tissue (Bulun et al. 1993a). Estrogen\nproduced by aromatase activity in breast adipose\ntissue fibroblasts was demonstrated to promote the\ngrowth of adjacent malignant breast epithelial cells\n(Yue et al. 1998). Finally, we demonstrated an\n‘intracrine’ effect of estrogen in uterine leiomyomas\nand endometriosis: estrogen produced by aromatase\nactivity in the cytoplasm of leiomyoma smooth\nmuscle cells or endometriotic stromal cells can exert\nits effects by readily binding to its nuclear receptor\nwithin the same cell (Bulun et al. 1994, Noble et al.\n1996, 1997). Disease-free endometrium and myo-\nmetrium, on the other hand, lack aromatase\nexpression (Bulun et al. 1993b, Noble et al. 1997).\nThe significance of aromatase expression in\nendometriosis\nAmong estrogen-responsive pelvic disorders, aro-\nmatase expression was studied in greatest detail in\nendometriosis (Bulun et al. 1993b, Noble et al. 1996,\n1997, Zeitoun et al. 1999). Firstly, extremely high\n1. Extra-ovarian estrogen formation in women.\nEstradiol (E2) in women is either directly secreted by\nthe ovary or produced in extra-ovarian sites (adipose\ntissue and skin). The principal substrate for extra-\novarian aromatase activity in women is androstenedione\n(A) of adrenal and ovarian origins. Androstenedione is\nconverted by aromatase to estrone (E\n1) in adipose tissue\nand skin fibroblasts. Estrone is further converted to E 2\nby 17/afii9826-HSD type 1 activity in these peripheral tissues.\nThus, extra-ovarian aromatization is the major source\nfor circulating E\n2 in the postmenopausal period or\nduring ovarian suppression.\n   and others · Aromatase and endometriosis36\nwww.endocrinology.orgJournal of Molecular Endocrinology (2000) 25, 35–42\nDownloaded from Bioscientifica.com at 06/12/2026 04:02:00PM\nvia free access\n\n\nlevels of aromatase mRNA were found in extra-\novarian endometriotic implants and endometriomas.\nSecondly, endometriosis-derived stromal cells in\nculture incubated with a cAMP analog displayed\nextraordinarily high levels of aromatase activity\ncomparable to that in placental syncytiotrophoblast\n(Noble et al. 1997). These exciting findings led us to\ntest a battery of growth factors, cytokines and other\nsubstances that might induce aromatase activity via\na cAMP-dependent pathway in endometriosis.\nProstaglandin E\n2 (PGE2) was found to be the most\npotent known inducer of aromatase activity in\nendometriotic stromal cells (Noble et al. 1997). In\nfact, this PGE\n2 effect was found to be mediated via\nthe cAMP-inducing EP 2 receptor subtype. More-\nover, estrogen was reported to increase PGE 2\nformation by stimulating cyclo-oxygenase type\n2 (COX-2) enzyme in endometrial stromal cells\nin culture (Huang et al. 1996). Thus, a positive\nfeedback loop for continuous local productions of\nestrogen and PGs is established, favoring the\nproliferative and in flammatory characteristics of\nendometriosis (Fig. 2). Additionally, aromatase\nmRNA was also detected in the eutopic endometrial\nsamples of women with moderate to severe endo-\nmetriosis (but not in those of disease-free women)\nalbeit in much smaller quantities compared with\nendometriotic implants (Noble et al. 1996). This\nmay be suggestive of a genetic defect in women with\nendometriosis, which is manifested by this subtle\nfinding in the eutopic endometrium. We propose\nthat when defective endometrium with low levels of\naberrant aromatase expression reaches the pelvic\nperitoneum by retrograde menstruation, it causes an\ninflammatory reaction that exponentially increases\nlocal aromatase activity, i.e. estrogen formation,\ninduced directly or indirectly by PGs and cytokines\n(Noble et al. 1997). It would be rather naive to\npropose that aberrant aromatase expression is the\nonly important molecular mechanism in the devel-\nopment and growth of pelvic endometriosis. There\nmay be many other molecular mechanisms that\nfavor the development of endometriosis: abnormal\nexpression of proteinase type enzymes that remodel\ntissues or their inhibitors (matrix metalloprotein-\nases, tissue inhibitor of metalloproteinase-1), certain\ncytokines (IL-6, RANTES) and growth factors\n(EGF) represent some of mechanisms (Khorram\net al. 1993, Sharpe-Timms et al. 1995, Osteen et al.\n1996, Bruner et al. 1997). Alternatively, a defective\nimmune system that fails to clear peritoneal surfaces\nof the retrograde menstrual e ffl ux has been\nproposed in the development of endometriosis\n(Halme et al. 1988, Hill 1992). The development of\nendometriosis in an individual woman probably\nrequires the coexistence of a threshold number of\nthese aberrations. Nonetheless, aberrant aromatase\nexpression is clinically relevant, since aromatase\ninhibitors suppress postmenopausal endometriosis\n(Takayama et al. 1998).\nRegulation of aromatase expression in\nendometriotic stromal cells\nAs emphasized earlier, PGE\n2 was found to be the\nmost potent known inducer of aromatase activity\nby increasing cAMP levels via cell surface EP\n2\nreceptors in endometriotic stromal cells (Noble\n2. Origin of estrogen in endometriotic lesions:\nestradiol (E2) that a ffects an endometriotic lesion arises\nfrom several body sites. In an ovulatory woman, E 2 is\nsecreted directly from the ovary in a cyclic fashion. In\nthe early follicular phase and after menopause, extra-\novarian tissues (adipose and skin) are the most\nimportant sources to account for the circulating E\n2.\nEstradiol is also produced locally in the endometriotic\nimplant itself in both ovulatory and postmenopausal\nwomen. The most important precursor, androstenedione\n(A) of adrenal and ovarian origins, becomes converted to\nestrone (E\n1) that is in turn reduced to E 2 in these tissues\nand endometriotic implants. We demonstrated\nsignificant levels of 17 /afii9826-hydroxysteroid dehydrogenase\ntype 1 expression in endometriosis, which catalyzes the\nconversion of E\n1 to E2 (Zeitoun et al. 1998). Estradiol\nand cytokines (IL-l /afii9825, TNF/afii9826), which are increased in\nendometriosis, induce cyclo-oxygenase-2 (COX-2)\ngiving rise to elevated concentrations of PGE\n2 in this\ntissue (Huang et al.). PGE2 in turn, is the most potent\nknown stimulator of aromatase in endometriotic stromal\ncells (Noble et al. 1997). This establishes a positive\nfeedback loop in favor of continuous estrogen formation\nin endometriosis.\nAromatase and endometriosis ·    and others\n37\nwww.endocrinology.org Journal of Molecular Endocrinology (2000) 25, 35–42\nDownloaded from Bioscientifica.com at 06/12/2026 04:02:00PM\nvia free access\n\n\net al. 1997). On the other hand, neither cAMP\nanalogs nor PGE 2 was capable of stimulating any\ndetectable aromatase activity in eutopic endometrial\nstromal cells in culture. The obvious question\nbecame: what are the molecular differences that give\nrise to aromatase expression in endometriosis and\nits inhibition in eutopic endometrium? To address\nthis, we first determined that the cAMP-inducible\npromoter II was used for in vivo aromatase\nexpression in endometriotic tissue (Zeitoun et al.\n1999). Then, a stimulatory transcription factor,\nSF-1, and an inhibitory factor, chicken ovalbumin\nupstream promoter transcription factor (COUP-\nTF), were found to compete for the same binding\nsite in aromatase promoter II. COUP-TF was\nubiquitously expressed in both eutopic endo-\nmetrium and endometriosis, whereas SF-1 was\nexpressed, speci fically in endometriosis but not in\neutopic endometrium, and binds to aromatase\npromoter more avidly than COUP-TF (Zeitoun\net al. 1999). Thus, SF-1 and other transcription\nfactors (e.g. CREB) activate transcription in\nendometriosis, whereas COUP-TF, which occupies\nthe same DNA site in eutopic endometrium,\ninhibits this process (Zeitoun et al. 1999) (Fig. 3).\nIn summary, one of the molecular alterations\nleading to local aromatase expression in endo-\nmetriosis but not in normal endometrium is the\naberrant production of SF-1 in endometriotic\nstromal cells, which overcomes the protective\ninhibition maintained normally by COUP-TF in\nthe eutopic endometrium.\nInterconversions of estrone and estradiol in\nendometriosis\nThe primary substrate for aromatase activity in\nendometriosis is androstenedione of adrenal and\novarian origins in premenopausal women. The\nmajor product of aromatase activity in endometri-\nosis, namely estrone, is only weakly estrogenic and\nmust be converted to the potent estrogen estradiol\nto exert a full estrogenic e ffect. We demonstrated\nthat the enzyme 17 /afii9826-HSD type 1, which catalyzes\nthe conversion of estrone to estradiol, is expressed\nin endometriosis (Andersson & Moghrabi 1997,\nZeitoun et al. 1998). In contrast, the enzyme\n17/afii9826-HSD type 2 (encoded by a separate gene)\n3. Proposed mechanism of regulation of aromatase (P450arom)\nexpression by SF-1 and CREB in endometriosis. Upon binding of PGE 2 to\nits cell surface EP 2 receptor, intracellular cAMP levels increase. This gives\nrise to binding of cAMP response element binding protein (CREB) and SF-1\nto specific motifs upstream of aromatase promoter II. The stimulatory type\ntranscription factor SF-1 binds as a monomer to a nuclear receptor half-site\nwith a higher a ffi nity compared with that of the inhibitory factor COUP-TF\n(not shown in the figure), which binds to the same site relatively loosely as a\ndimer. SF-1 then synergizes with CREB (bound to upstream CRE) and\npossibly other factors to activate transcription of the P450arom gene in\nresponse to cAMP (Zeitoun et al. 1999).\n   and others · Aromatase and endometriosis\n38\nwww.endocrinology.orgJournal of Molecular Endocrinology (2000) 25, 35–42\nDownloaded from Bioscientifica.com at 06/12/2026 04:02:00PM\nvia free access\n\n\ninactivates estradiol by catalyzing its conversion to\nestrone in eutopic endometrial glandular cells\nduring the luteal phase (Andersson & Moghrabi\n1997). Progesterone actually induces the activity of\nthis enzyme in endometrial glandular cells in\nculture, making inactivation of estradiol to estrone\none of the anti-estrogenic properties of progesterone\n(Satyaswaroop et al. 1982). The expression of\n17/afii9826-HSD type 2 is absent from endometriotic\nglandular cells, as demonstrated in paired samples\nof eutopic endometrium and pelvic endometriosis\nobtained simultaneously during the luteal phase\n(Zeitoun et al. 1998). Consequently, this protective\nmechanism that lowers estradiol levels is lost in\nendometriotic tissue (Zeitoun et al. 1998). The\naberrant expression of aromatase, the presence of\n17/afii9826-HSD type 1 and the absence of 17 /afii9826-HSD type\n2 from endometriosis collectively give rise to\nelevated local levels of estradiol compared with\neutopic endometrium. Additionally, 17 /afii9826-HSD type\n2d eficiency may also be viewed as a defective action\nof progesterone, which fails to induce this enzyme\nin endometriotic tissue (Fig. 4).\nRationale for using aromatase inhibitors to\ntreat endometriosis\nEndometriosis is successfully suppressed by estro-\ngen deprivation with GnRH analogs or the\ninduction of surgical menopause. Control of pelvic\npain with GnRH agonists is usually successful\nduring and immediately after the treatment,\nwhereas pain associated with endometriosis returns\nin up to 75% of these women (Henzl et al. 1988,\nWaller & Shaw 1993). There may be multiple\nreasons for the failure of GnRH agonist treatment\nof endometriosis. One likely explanation is the\npresence of signi ficant estradiol production that\ncontinues in the adipose tissue, skin and endometri-\notic implant per se during the GnRH agonist\ntreatment. Therefore, blockage of aromatase activity\nin these extra-ovarian sites with an aromatase\ninhibitor may keep larger number of patients in\nremission for longer periods of time (Fig. 5). The\nmost striking evidence for the signi ficance of\nextra-ovarian estrogen production is the recurrence\nof endometriosis after successfully completed hys-\nterectomy and bilateral salpingo-oophorectomy in a\n4. Defective inactivation of estradiol (E 2) in endometriosis: E 2\nreaches the endometriotic lesions via the blood stream. Additionally, estrone\n(E1) is produced in the stromal cell via aberrant aromatase activity. Estrone is\nfurther reduced to E 2 by 17/afii9826-HSD type 1 in the endometriotic tissue.\nEstradiol is normally inactivated by conversion to E 1 by 17/afii9826-HSD type 2 in\nepithelial cells of the eutopic endometrium in response to progesterone\nduring the secretory phase. In endometriotic tissue, however, E\n2 is not\nmetabolized owing to the lack of 17 /afii9826-HSD type 2, giving rise to increased\nlocal concentration of this potent estrogen. The absence of 17 /afii9826-HSD type 2\nexpression in endometriosis despite high levels of progesterone during the\nsecretory phase is indicative of selective progesterone resistance in this tissue.\nAromatase and endometriosis ·    and others\n39\nwww.endocrinology.org Journal of Molecular Endocrinology (2000) 25, 35–42\nDownloaded from Bioscientifica.com at 06/12/2026 04:02:00PM\nvia free access\n\n\nnumber of women (Metzger et al. 1991, Takayama\net al. 1998). Endometriotic tissue in one such\naggressive case was found to express much higher\nlevels of aromatase mRNA compared with pre-\nmenopausal endometriosis (Takayama et al. 1998).\nWe recently reported the treatment of a 57-year-old\noverweight woman who had recurrence of severe\nendometriosis after hysterectomy and bilateral\nsalpingo-oophorectomy. Two additional laparoto-\nmies were performed owing to persistent severe\npelvic pain and bilateral ureteral obstruction leading\nto left renal atrophy and right hydronephrosis.\nTreatment with megestrol acetate was ine ffective. A\nlarge (3 cm) vaginal endometriotic lesion contained\nunusually high levels of aromatase mRNA. The\npatient was given anastrozole (an aromatase in-\nhibitor) for 9 months. Despite the addition of\ncalcium and alendronate (a nonsteroidal inhibitor\nof bone resorption), bone density in the lumbar\nspine decreased by 6 ·2%. The occurrence of\nsignificant bone loss in this particular case should\nbe studied further. Dramatic relief of the pain\nand regression of the vaginal endometriotic lesion\nwere observed within the first month of treatment.\nAt the same time, circulating estradiol levels were\nreduced to 50% of the baseline value. Markedly\nhigh pretreatment levels of aromatase mRNA in\nthe endometriotic tissue became undetectable in\na repeat biopsy 6 months later, and the lesion\nnearly disappeared after 9 months of therapy.\nTwo potential mechanisms may have accounted\nfor this strikingly successful result. Firstly, there\nwas evidence of suppression of extra-ovarian (i.e.\nskin and adipose tissue) aromatase activity, giving\nrise to a signi ficant decrease in serum estradiol\nlevel (Fig. 5). Secondly, unusually high levels of\naromatase expression in the endometriotic lesion\ndisappeared after treatment with the aromatase\ninhibitor, anastrozole (Fig. 5). Besides the\nexpected direct inhibition of aromatase activity\nin endometriosis by anastrozole, the disappear-\nance of aromatase mRNA expression in the lesion\nmay be explicable by denial of estrogen that is\nknown to stimulate local biosynthesis of PGE\n2,\nwhich, in turn, stimulates aromatase expression\n(Fig. 2).\nIn summary, the recently developed potent\naromatase inhibitors are candidate drugs in the\ntreatment of endometriosis that is resistant to\nstandard regimens. In fact, the use of aromatase\ninhibitors may be the only available treatment\nfor aggressive postmenopausal endometriosis. It\nremains to be seen whether aromatase inhibitors\nalone or together with present lines of therapy in\npremenopausal women will increase the pain-free\ninterval and time to recurrence after discontinuation\n(Fig. 5). Studies are under way to address these\nquestions.\n5. Sites of action of aromatase inhibitors to treat endometriosis. In\ncases resistant to treatment with GnRH agonists or in postmenopausal\nendometriosis, the use of aromatase inhibitors to block estrogen formation in\nthe skin and adipose tissue as well as in endometriotic tissue may be critical\nin inhibiting the growth of endometriotic tissue. Recurrent endometriosis,\nespecially after surgical removal of the ovaries, may represent lesions that are\nsensitive to extremely low levels of estradiol (E\n2). Thus, suppression of E 2\nproduction in the extra-ovarian sites (adipose tissue/skin) and in\nendometriotic tissue may be mandatory for successful treatment of\nendometriosis.\n   and others · Aromatase and endometriosis\n40\nwww.endocrinology.orgJournal of Molecular Endocrinology (2000) 25, 35–42\nDownloaded from Bioscientifica.com at 06/12/2026 04:02:00PM\nvia free access\n\n\nSUMMARY\nThe development and growth of endometriotic\nlesions are estrogen-dependent. The mechanisms\nand e ffectiveness of hormonal treatments for\nendometriosis should be re-evaluated in view of the\nnew advances that increased our understanding of\nthe body sites of estrogen production in a woman\nwith endometriosis. In addition to ovarian secre-\ntion, estradiol is also produced in peripheral sites\nsuch as skin, adipose tissue and endometriotic\nlesions per se . We suggest that the intracrine and\nparacrine effects of estradiol produced in the target\ntissue amplify the estrogenic action of steroid\nhormones delivered via the circulation. Addition-\nally, defective inactivation of estradiol in endo-\nmetriosis in contrast to eutopic endometrium may\nfurther enhance this local effect. Aberrant aromatase\nactivity and defective estradiol metabolism in\nendometriosis are consequences of speci fic molecu-\nlar aberrations such as inappropriate expression of\na stimulatory transcription factor or progester-\none resistance in this tissue. The clinical relevance\nof these findings was recently exempli fied by the\nsuccessful treatment of a severe case of recurrent\npostmenopausal endometriosis with an aromatase\ninhibitor. Future treatment strategies may be\ndesigned to target the signal transduction for\naromatase expression in endometriosis or to\nenhance progesterone action in this tissue.\nACKNOWLEDGEMENTS\nWe thank Margarita Guerrero and Dee Alexander\nfor providing expert editorial assistance. This\nresearch work was supported, in part, by the NIH\ngrants HD38691 and CA67167 (Serdar Bulun) and\nan AAOGF Fellowship Award (Khaled Zeitoun).\nREFERENCES\nAckerman GE, Smith ME, Mendelson CR, MacDonald PC &\nSimpson ER 1981 Aromatization of androstenedione by\nhuman adipose tissue stromal cells in monolayer culture.\nJournal of Clinical Endocrinology and Metabolism 53 412–417.\nAndersson S & Moghrabi N 1997 Physiology and molecular\ngenetics of 17 /afii9826-hydroxysteroid dehydrogenases. Steroids 62\n143–147.\nBruner KL, Matrisian LM, Rodgers WH, Gorstein F &\nOsteen KG 1997 Suppression of matrix metalloproteinases\ninhibits establishment of ectopic lesions by human\nendometrium in nude mice. Journal of Clinical Investigation\n99 2851–2857.\nBulun SE, Mahendroo MS & Simpson ER 1993 a Polymerase\nchain reaction ampli fication fails to detect aromatase\ncytochrome P450 transcripts in normal human endometrium\nor decidua. Journal of Clinical Endocrinology and Metabolism\n76 1458–1463.\nBulun SE, Price TM, Mahendroo MS, Aitke nJ&S i m p s o n\nER 1993b A link between breast cancer and local estrogen\nbiosynthesis suggested by quanti fication of breast adipose\ntissue aromatase cytochrome P450 transcripts using\ncompetitive polymerase chain reaction after reverse\ntranscription. Journal of Clinical Endocrinology and\nMetabolism 77 1622–1628.\nBulun SE, Simpson ER & Word RA 1994 Expression of the\nCYP19 gene and its product aromatase cytochrome P450 in\nhuman leiomyoma tissues and cells in culture. Journal of\nClinical Endocrinology and Metabolism 78 736–743.\nH a l m eJ ,W h i t eC ,K a u m aS ,E s t e sJ&H a s k i l lS1 9 8 8\nPeritoneal macrophages from patients with endometriosis\nrelease growth factor activity in vitro. Journal of Clinical\nEndocrinology and Metabolism 66 1044–1049.\nHenzl MR, Corson SL, Moghissi K, Buttram VC, Berqvist C\n& Jacobson J 1988 Administration of nasal nafarelin as\ncompared with oral danazol for endometriosis. New England\nJournal of Medicine 318 485–489.\nHill JA 1992 Immunology and endometriosis. Fertility and\nSterility 58 262–264.\nHuang JC, Dawood MY & Wu KK 1996 Regulation of\ncyclooxygenase-2 gene in cultured endometrial stromal cells\nby sex steroids. Proceedings of the American Society of\nReproductive Medicine Meeting S5 (abstract).\nKhorram O, Taylor RN, Ryan IP, Schall TJ & Landers DV\n1993 Peritoneal fluid concentrations of the cytokine\nRANTES correlate with the severity of endometriosis .\nAmerican Journal of Obstetrics and Gynecology 169\n1545–1549.\nKjerulff KH, Erickson BA & Langenberg PW 1996 Chronic\ngynecological conditions reported by US women: findings\nfrom the National Health Information Survey 1984 to 1992.\nAmerican Journal of Public Health 86 195–199.\nMacDonald PC, Rombaut RP & Siiteri PK 1967 Plasma\nprecursors of estrogen I. Extent of conversion of plasma\n/afii9797\n4-androstenedione to estrone in normal males and\nnon-pregnant normal, castrate and adrenalectomized females .\nJournal of Clinical Endocrinology and Metabolism 27\n1103–1111.\nMacDonald PC, Edman CD, Hemsell DL, Porte rJ&S iiteri\nPK 1978 E ffect of obesity on conversion of plasma\nandrostenedione to estrone in postmenopausal women with\nand without endometrial cancer. American Journal of\nObstetrics and Gynecology 130 448–455.\nMetzger DA, Lessey BA, Soper JT, McCarty KS, Jr & Haney\nAF 1991 Hormone-resistant endometriosis following\ntotal abdominal hysterectomy and bilateral salpingo-\noophorectomy: correlation with histology and steroid\nreceptor content. Obstetrics and Gynecology 78 946–950.\nMichael MD, Kilgore MW, Morohashi KI & Simpson ER\n1995 Ad4 BP/SF-1 regulates cyclic AMP-induced\ntranscription from the proximal promoter (PII) of the human\naromatase P450 (CYP19) gene in the ovary. Journal of\nBiological Chemistry 270 13561–13566.\nMichael MD, Michael LF & Simpson ER 1997 A CRE-like\nsequence that binds CREB and contributes to cAMP-\ndependent regulation of the proximal promoter of the human\naromatase P450 (CYP19) gene. Molecular and Cellular\nEndocrinology 134 147–156.\nNoble LS, Simpson ER, John s A & Bulun SE 1996 Aromatase\nexpression in endometriosis . Journal of Clinical Endocrinology\nand Metabolism 81 174–179.\nNoble LS, Takayama K, Putman JM, Johns DA, Hinshelwood\nMM, Agarwal VR, Zhao Y, Carr BR & Bulun SE 1997\nProstaglandin E\n2 stimulates aromatase expression in\nendometriosis-derived stromal cells. Journal of Clinical\nEndocrinology and Metabolism 82 600–606.\nAromatase and endometriosis ·    and others 41\nwww.endocrinology.org Journal of Molecular Endocrinology (2000) 25, 35–42\nDownloaded from Bioscientifica.com at 06/12/2026 04:02:00PM\nvia free access\n\n\nOlive DL & Schwartz LB 1993 Endometriosis. New England\nJournal of Medicine 328 1759–1769.\nOsteen KG, Bruner KL & Sharpe-Timms KL 1996 Steroid\nand growth factor regulation of matrix metalloproteinase\nexpression and endometriosis. Seminars in Reproductive\nEndocrinology 15 301–308.\nSampson JA 1927 Peritoneal endometriosis due to the\nmenstrual dissemination of endometrial tissue into the\nperitoneal cavity. American Journal of Obstetrics and\nGynecology 14 422–425.\nSatyaswaroop PG, Wartell DJ & Mortel R 1982 Distribution\nof progesterone receptor, estradiol dehydrogenase, and 20 /afii9825-\ndihydroprogesterone dehydrogenase activities in human\nendometrial glands and stroma: progestin induction of\nsteroid dehydrogenase activities in vitro is restricted to the\nglandular epithelium. Endocrinology 111 743–749.\nSharpe-Timms KL, Penney LL, Zimmer RL, Wright JA,\nZhang Y & Surewicz K 1995 Partial puri fication and amino\nacid sequence analysis of endometriosis protein-II (ENDO-II)\nreveals homology with tissue inhibitor of metalloproteinases-1\n(TIMP-1). Journal of Clinical Endocrinology and Metabolism 80\n3784–3787.\nSimpson ER, Mahendroo MS, Means GD, Kilgore MW,\nHinshelwood MM, Graham-Lorence S, Amarneh B, Ito Y,\nFisher CR, Michael MD, Mendelson CR & Bulun SE 1994\nAromatase cytochrome P450, the enzyme responsible for\nestrogen biosynthesis. Endocrine Reviews 15 342–355.\nTakayama K, Zeitoun K, Gunby RT, Sasano H, Carr BR &\nBulun SE 1998 Treatment of severe postmenopausal\nendometriosis with an aromatase inhibitor. Fertility and\nSterility 69 709–713.\nVessey MP, Villard-Mackintos hL&P a i n t e rR1993\nEpidemiology of endometriosis in women attending family\nplanning clinics. British Medical Journal 306 182–184.\nWaller KG & Shaw RW 1993 Gonadotropin-releasing hormone\nanalogues for the treatment of endometriosis: long-term\nfollow-up. Fertility and Sterility 59 511–515.\nYue W, Wang JP, Hamilton CJ, Demers LM & Santen RJ\n1998 In situ aromatization enhances breast tumor estradiol\nlevels and cellular proliferation. Cancer Research 58\n927–932.\nZeitoun K, Takayama K, Sasano H, Suzuki T, Moghrabi N,\nAndersson S, Johns A, Meng L, Putman M, Car r B & Bulun\nSE 1998 De ficient 17/afii9826-hydroxysteroid dehydrogenase type 2\nexpression in endometriosis: failure to metabolize estradiol-17/afii9826.\nJournal of Clinical Endocrinology and Metabolism 83 4474–4480.\nZeitoun K, Takayama K, Michael MD & Bulun SE 1999\nStimulation of aromatase P450 promoter (II) activity in\nendometriosis and its inhibition in endometrium are\nregulated by competitive binding of SF-1 and COUP-TF to\nthe same cis-acting element. Molecular Endocrinology 13\n239–253.\n17 December 1999\n   and others · Aromatase and endometriosis42\nwww.endocrinology.orgJournal of Molecular Endocrinology (2000) 25, 35–42\nDownloaded from Bioscientifica.com at 06/12/2026 04:02:00PM\nvia free access","source_license":"CC0","license_restricted":false}