Laboratory, Department of Clinical Pathology, Nagoya University Hospital, Nagoya, Japan

Transcription

1 REPRODUCTIVE ENDOCRINOLOGY Localization of angiotensin II, the AT1 receptor, angiotensin-converting enzyme, aminopeptidase A, adipocyte-derived leucine aminopeptidase, and vascular endothelial growth factor in the human ovary throughout the menstrual cycle Toko Harata, M.D., a Hisao Ando, M.D., a Akira Iwase, M.D., a Tetsuro Nagasaka, M.D., b Shigehiko Mizutani, M.D., a and Fumitaka Kikkawa, M.D. a a Department of Obstetrics and Gynecology, Nagoya University Graduate School of Medicine, Nagoya, Japan; and b Clinical Laboratory, Department of Clinical Pathology, Nagoya University Hospital, Nagoya, Japan Objective: To assess the expression and cellular distribution of angiotensin II (Ang II), angiotensin type 1 receptor (AT1R), angiotensin-converting enzyme (ACE), aminopeptidase A (APA), adipocyte-derived leucine aminopeptidase (A-LAP), and vascular endothelial growth factor (VEGF) in human ovarian tissue during the menstrual cycle. Design: Ovarian tissues (n 52) and corpora lutea (n 34) were obtained from patients undergoing hysterectomy/oophorectomy, and tissue sections were immunostained for each antigen. Setting: University hospital. Patient(s): Patients undergoing hysterectomy or oophorectomy for benign conditions. Intervention(s): Immunostaining of tissue sections using antibodies to each antigen. Main Outcome Measure(s): Microscopic evaluation to assess the presence, distribution, and cellular localization. Result(s): The luteal tissue is the major site of Ang II, ACE, AT1R, and VEGF, with highest staining intensity found during the midluteal phase and at pregnancy. The AT1R was found in theca cells. The APA was strongly immunolocalized in pericytes. Immunolocalization of AT1R was almost similar to that of VEGF including oocytes in the primordial and intermediate follicles. Conclusion(s): The expression and distinct pattern of the cellular localization of Ang II and its related proteins in human ovarian tissue during folliculogenesis and in the luteal tissue suggest their roles in the growth and differentiation of theca, granulose, and luteal cells. (Fertil Steril 2006;86: by American Society for Reproductive Medicine.) Key Words: Adipocyte-derived leucine aminopeptidase (ER-aminopeptidase 1), aminopeptidase A, angiotensin converting enzyme, angiotensin II, AT1 receptor, corpus luteum, immunohistochemistry, ovary, vascular endothelial growth factor Various stages of growing follicles are present in the ovarian cortex in women of reproductive ages. The follicle contains a developing oocyte surrounded by the granulosa cells, which are the main source of estrogen production. The granulosa cell layer of the mature follicle is surrounded by Received September 15, 2005; revised and accepted January 5, Supported in part by a Grant-in-Aid (C) for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan. Presented in part at the 20th Annual Meeting of the European Society of Human Reproduction and Embryology, Berlin, Germany, June 27 30, Reprint requests: Hisao Ando, M.D., Infertility Center, Toyohashi Municipal Hospital, 50, Hakkenn-Nishi, Aotake-cho, Toyohashi, Aichi , Japan (Fax: ; Kotobuki@se.starcat.ne.jp). the theca folliculi that merges with the ovarian stroma. Under the appropriate gonadotropic stimulation, the granulosa and the theca cells continue to grow until ovulation. During the folliculogenesis, the outer theca layer develops its own vascular network, while the granulosa inner layer remains avascular (1). Ovulation is followed by the remarkable changes that include new blood vessel formation and the differentiation of the granulosa and the theca cells into the large and the small luteal cells, respectively. Therefore, the ovulated follicle is converted into the corpus luteum to produce progesterone as well as estrogen. Although ovarian function is primarily controlled by the hypothalamus-pituitary axis, there is no doubt that a hormonal microenvironment specific for each individual follicle /06/$32.00 Fertility and Sterility Vol. 86, No. 2, August 2006 doi: /j.fertnstert Copyright 2006 American Society for Reproductive Medicine, Published by Elsevier Inc. 433

2 is established, which finally determines whether a follicle ovulates and becomes a corpus luteum or undergoes atresia. In this respect, autocrine and paracrine factors that act alone or modulate gonadotropin action are of paramount importance. It has been known for many years that there is an active ovarian version of the renin-angiotensin system (RAS) in multiple species, including rodents, domestic animals, and primates including humans (2). This system is relevant to clinical reproductive endocrinology for a variety of reasons, including proposed roles in ovulation, steroidogenesis, polycystic ovary syndrome (PCOS), and ovarian hyperstimulation syndrome (OHSS). Angiotensin II (Ang II) is the main biologically active octapeptide of the RAS. It is produced by the cleaving of dipeptides from the C-terminus of the decapeptide precursor known as Ang I. Angiotensin-converting enzyme (ACE), which is widely distributed on the luminal surface of the vascular endothelium, is a pivotal enzyme in the production of Ang II. On the other hand, the cleavage of Ang II has been complexed very recently, with the discovery of ACE-related carboxypeptidase 2 (ACE2), which can degrade Ang II into Ang 1-7 (3, 4). However, the noncardiovascular tissue distributions of Ang II degrading enzymes as well as its producing enzymes have not been elucidated, including male and female reproductive systems. Aminopeptidase A (APA) is a crucial factor that degrades Ang II into Ang III (Ang 2-8) (5). The cleavage of Ang II into Ang III is also catalyzed by adipocyte-derived leucine aminopeptidase (A-LAP)/ERaminopeptidase 1 (6). It is probable that the local concentration of Ang II in the ovary in cycling women is regulated by these proteolytic enzymes. High levels of Ang II immunoreactivity have been found in preovulatory follicular fluid from women with normal or gonadotropin-stimulated cycles (7), and Ang II is the main component of the Ang II immunoreactivity (8). The immunohistochemical localization of Ang II in human ovaries reveals densely stained theca and stromal cells (9). The ACE activity has been identified in the human ovary at lowermoderate levels (10). However, its cellular distribution in the ovary is not known, although enough Ang I is supplied via systemic circulation and ACE is possibly localized in the endothelial cells. Recently, locally produced vascular endothelial growth factor (VEGF) induced by Ang II has been highlighted in the neovascularization of tumor tissue (11) and in the retinal vasculature (12). The Ang II type 1 receptor (AT1R) is the main receptor for Ang II, although AT2R on granulosa cells is reportedly involved in ovulation in rabbits (13). More recently, we reported that the oral administration of ACE inhibitor and AT1 receptor in the luteal phase inhibited OHSS in patients at high risk, with the reduction of serum VEGF concentration (14). However, whether human ovarian cells express AT1 receptors is not yet clear (15). In the present study, we described the cellular distribution of immunoreactive Ang II and related proteolytic enzymes that are involved in the synthesis and degradation of Ang II. Furthermore, we investigated the distribution of AT1 receptors and VEGF to understand the role of Ang II in the human ovary. MATERIALS AND METHODS Tissue Collection We retrieved normal ovary specimens from the pathology files at Nagoya University Hospital, Nagoya, Japan. Patient clinical charts were reviewed, and cases were selected on the basis of a history of regular menstrual cycles and no use of any intrauterine device or hormone therapy for 6 months before the surgery. Histological slides of the ovary were subsequently reviewed, and cases were further selected on the basis of consistent histological findings. As a result of these reviews, the corpora lutea (n 34) or a portion of whole ovarian tissue (n 52) were obtained from women undergoing surgical procedures for a variety of benign gynecological conditions. Endometrial histological dating and the patient s last menstrual period were used as criteria for dating the corpora lutea. On the basis of these criteria, the luteal tissues were from the early (days 14 19; n 8), the middle (days 20 25; n 10), and the late (days 26 28; n 15) phases of the menstrual cycle. We used one sample of the corpus luteum from an ectopic pregnancy case (at 8 weeks). In this patient, the corpus luteum was resected due to active hemorrhage. These samples were fixed in 10% formalin, embedded in paraffin, and used for immunohistochemistry. Antibodies The antibodies for human ACE (a mouse monoclonal antibody; Chemicon International, Temecula, CA), for human AT1R (a rabbit polyclonal antibody, 306; Santa Cruz Biotechnology, Santa Cruz, CA), and for human VEGF (a rabbit polyclonal antibody, A-20; Santa Cruz Biotechnology) were selected on the basis of repeated preliminary immunohistochemical evaluation using various positive control specimens from candidates of two or three commercial antibodies for each antigen. We could not find a specific antibody to match human AT2R that was suitable for immunohistochemistry. The rabbit polyclonal antibody to Ang II (Peninsula Laboratories, San Carlos, CA) does not cross-react with Ang I (Ang 1-10) or Ang IV (Ang 3-8), but cross-reacts 100% with Ang III (Ang 2-8). The rabbit polyclonal antibody for APA and that for A-LAP are specific and described elsewhere (16, 17). Immunohistochemistry Formalin-fixed, paraffin-embedded tissue sections were cut to a thickness of 3 m. Deparaffinized sections in 0.01 M citrate buffer were treated three times for 5 minutes, at 90 C at 750 W, in an H2500 microwave oven (M&M, Tokyo, Japan) for heatinduced epitope retrieval. Endogenous peroxidase activity was 434 Harata et al. Ang II-related proteins in human ovary Vol. 86, No. 2, August 2006

3 blocked by incubation with 0.5% (w/v) hydrogen peroxide in methanol for 10 minutes, and nonspecific immunoglobulin binding was blocked by incubation with 10% normal goat serum in phosphate-buffered saline (PBS) for 10 minutes. Immunohistochemical staining was performed on the basis of the labeled streptavidin-biotin method. A Ventana Basic DAB Detection Kit (Ventana Medical Systems, Tucson, AZ), yielding a brown product from diaminobenzidine (DAB) copper sulfate, was used to detect each protein. Staining procedures were performed automatically using Ventana s BenchMark IHC Staining System (Ventana Medical Systems) according to the manufacturer s instructions; this system is a fully automatic computerized staining system that yields stable and reproducible data. The primary antibodies against Ang II, ACE, APA, A-LAP, AT1R, and VEGF were diluted 1:500, 1:300, 1:100, 1:100, 1:100, and 1:100 in PBS, respectively. In the negative control experiments, the primary antibody was replaced with rabbit or mouse IgG. The slides were counterstained with hematoxylin before mounting. Stained sections were observed under a BH2 microscope (Olympus, Tokyo, Japan) and photographed using a CCD Color Camera CS600 (Olympus). Evaluation for the staining intensity was undertaken in the criteria as follows: negative, occasional or weak, moderate, or strong. RESULTS Immunostaining of ovarian tissue sections with specific antibodies to Ang II, ACE, APA, A-LAP, and VEGF indicated the presence of these immunoreactive proteins in several ovarian cell types. Each protein exhibited a distinct distribution and variation of the staining intensity depending on the cell type, as well as the menstrual cycle (Table 1). TABLE 1 Immunostaining of ovarian structures with specific antibodies for Ang II and its related molecules. Structure Ang II ACE APA A-LAP AT1R VEGF Blood vessels Endothelial cells / / or / / / / Pericytes / / / / / Smooth muscle cells / / / / Primordial follicles Oocytes / / / Pregranulosa cells / / / Primary/intermediate follicles Oocytes / / / Granulosa cells / / / / / / Secondary/Graafian follicles Oocytes / / / / / Granulosa cells / / / / / / Theca cells / / / / Ovarian cortex Epithelial cells / or Stromal cells / / or / / / Corpus luteum (early luteal phase) Large luteal cells / Small luteal cells / / / / Corpus luteum (midluteal phase) Large luteal cells Small luteal cells / Corpus luteum (late luteal phase) Large luteal cells / / Small luteal cells / / / / Corpus luteum (in pregnancy) a Large luteal cells Small luteal cells / Corpus albicans / or Note: Staining intensity is expressed as negative, / occasional or weak, moderate, or strong. a Data from one patient at 8 weeks. Harata. Ang II-related proteins in human ovary. Fertil Steril Fertility and Sterility 435

4 FIGURE 1 Immunohistochemical localization of Ang II (A E), ACE (F J), APA (K N), and A-LAP (O) in human ovarian tissues displaying immunostaining with various intensities in association with (A) primordial follicles, (B, F, K) mature follicles, (C) luteal tissue from the early luteal phase, (D, G, L, O) luteal tissues from the midluteal phase, (E, H, M) luteal tissues from the late luteal phase, (I, N) luteal tissues at early pregnancy, and (J) corpus albicans. Po primordial follicle; St stroma of the ovarian cortex; Gr granulosa cells; Th theca cells; Ll large luteal cells; Sl small luteal cells; Pe pericytes. Original magnification; 200. Bar in (O) 100 m. Harata. Ang II-related proteins in human ovary. Fertil Steril Angiotensin II The Ang II immunoreactivity was found most strongly in the midluteal and pregnant corpora lutea (Fig. 1A E). The staining intensity was moderate in the early and late luteal phase. In such luteal tissues, both small and large luteal cells immunostained for Ang II, with more intense staining in large luteal cells. The corpus albicans also exhibited occasional staining. Each stage of follicles was stained for Ang II occasionally. Epithelial cells of the ovarian cortex were stained moderately. Angiotensin Converting Enzyme In addition to the corpora lutea, ACE was widely distributed in the stroma of the cortex and in the corpus albicans through- 436 Harata et al. Ang II-related proteins in human ovary Vol. 86, No. 2, August 2006

5 out the menstrual cycle (Fig. 1F J). The growing follicle exhibited occasional staining. Blood vessels exhibited weak or moderate stainings throughout the menstrual cycle. Aminopeptidase A The APA was found in pericytes with intense immunoreactivity (Fig 1K N). The large and small luteal cells were stained moderately and weakly, respectively, without any particular change according to the subphases. It was noted that well-grown pericytes made a clear contrast with the weakly stained luteal cells in the pregnant corpus luteum. Adipocyte-Derived Leucine Aminopeptidase Weak staining occurred in the pregranulosa and granulosa cells throughout folliculogenesis (Fig. 1O). The large and small luteal cells displayed weak or moderate staining with the peak intensity in the midluteal phase and during pregnancy. The AT1 receptors The theca cells were clearly stained conspicuously for the antibody specific for AT1R (Fig. 2A G). The receptors also displayed positive staining in the oocyte in the primordial and primary follicles. Immunostainings for pregranulosa and granulosa cells were limited. The luteal cells were strongly stained in the midluteal phase and during pregnancy. However, in the luteolytic phase, immunostaining for AT1R was moderate. Vascular smooth muscle cells were also stained moderately for AT1R. Vascular Endothelial Growth Factor The immunolocalization of VEGF was similar to that of AT1R (Fig. 2H M), except that the theca cells of secondary or Graafian follicles and the luteal cells of the late luteal phase displayed different intensities of staining. In contrast, strong immunostaining of luteal cells was observed in the midluteal phase and during pregnancy. DISCUSSION The present study provides additional evidence of localization of various components of the ovarian RAS within specific cell types of human ovarian tissues at different times of the menstrual cycle. In particular, this study represents the first immunohistochemical analysis for ACE, APA, A-LAP, and AT1R in human ovarian tissue at different cycle times of ovarian follicles and corpora lutea. Furthermore, immunolocalization of both Ang II and VEGF together in the cycling ovary was shown using the adjacent sections for the first time. The localization of VEGF was in agreement with the previous reports (18, 19). The Ang II immunoreactivity was clear in corpora lutea, using much larger numbers of samples with all subphases (n 34) than in the previous report by Palumbo et al. (n 4) (9). There were some controversial data in comparing the previous report (9) and our results. In the previous report, theca cells of large preovulatory follicles displayed intense granular staining for Ang II, while granulosa cells were unstained in most follicles. In our present study, the staining intensity of follicles varied with the samples. The accumulation of Ang II-like immunoreactivity in the corpus luteum appears to be the result of both the increased supply of Ang II precursors by active neovascularization and the balance between producing and degrading enzymes for Ang II. However, the immunoreactivity around the follicle is likely to be unstable due to the slow access of fixatives in the presence of follicular fluid. In such an environment, Ang II should be susceptible to degradation before it is fixed. The ACE was found in the luteal cells in our immunohistochemical analysis. However, localization of ACE in ovarian cells other than blood vessels has not been determined either in humans or in other animals (20). Although our present data are reasonable, further studies on a molecular basis, such as in situ hybridization, are needed. The APA was localized primarily in the ovarian pericytes. Pericytes surround the endothelial cells in the early neovascularization process (21). It has been demonstrated that APA is localized in the pericyte in pathological conditions associated with neovascularization (22). Neovascularization in the normal tissue under physiological conditions is limited to the ovary, the endometrium, and the placenta. In the human endometrium, APA is not localized in the pericytes (16). It is suggested that rich APA content in the pericyte maintains a high Ang II local concentration in the corpus luteum, keeping it away from circulation. The presence of APA and A-LAP in the luteal cells may also support the local regulation of Ang II binding to the receptor. More studies on local regulators of Ang II are needed, including the existence of other degeneration pathways, such as ACE2 (3, 4). We focused on the cellular distribution of AT1R, as well as the synthesis and the degradation of Ang II. Most of the studies performed so far have been based on ligand-binding techniques and the use of pharmacological receptor agonists and antagonists in nonprimate animal experiments (20). Furthermore, it appears that more studies have been conducted on AT2R, instead of AT1R, concerning the expression, the localization, and the function in nonprimate animals (20). Therefore, whether human ovarian cells express Ang II receptors has not yet been clarified, although we demonstrated Ang II-induced VEGF production using the luteingranulosa cells (23). There is no doubt that active neovascularization is essential for folliculogenesis and luteogenesis. The VEGF is known as a potent angiogenic factor with leaky vessel permeabilization, and it is involved in the pathogenesis of OHSS (24). We hypothesized that ovarian Ang II may induce VEGF in human ovaries under physiological conditions. Recently, we investigated the Ang II-induced VEGF production via AT1R in cultured lutein-granulosa cells (23). Fertility and Sterility 437

6 FIGURE 2 Immunohistochemical localization of AT1R (A G), VEGF (H M), and negative controls using mouse IgG (N) or rabbit IgG (O) in human ovarian tissues displaying immunostaining with various intensities in association with (A C, H) primordial follicles, (B, I) primary follicles, (C, D, J, N) secondary/mature follicles, (E, K, O) luteal tissues from the midluteal phase, (F, L) luteal tissues from the late luteal phase, and (G, M) luteal tissues from the late luteal phase. Po primordial follicle; Gr granulosa cells; Th theca cells; En endothelial cells; Ll large luteal cells; Sl small luteal cells. Original magnification; 200. Bar in (O) 100 m. Harata. Ang II-related proteins in human ovary. Fertil Steril Moreover, we have reported a decreased number of blood vessels in the corpus luteum on the Ang II-blocked rabbit OHSS model (23). These data are in accordance with the presented immunohistological data revealing the similar cellular localization of AT1R and VEGF in the corpus luteum. Sugino et al. (19) reported that VEGF receptors are localized in human luteal cells. Therefore, the action of Ang IIinduced VEGF might be involved in the formation and the regression of the corpus luteum, as well as the neovascularization in the corpus luteum. However, Ang II with other local factors in human corpora lutea should be further investigated. It was demonstrated that in vitro microdialysis infu- 438 Harata et al. Ang II-related proteins in human ovary Vol. 86, No. 2, August 2006

7 sion of Ang II to the bovine midcycle corpora lutea stimulated the release of endothelin-1 and oxytocin, and suppressed the release of progesterone (25). Folliculogenesis requires neovascularization of growing follicles. There was clear immunolocalization of AT1R in theca cells of the growing follicle. The VEGF was also immunolocalized in the theca layer. It should be noted that both AT1R and VEGF were detectable in the oocyte of the primordial and primary follicles but not in the oocyte of the mature follicles. These data suggest the possible roles of Ang II and VEGF during folliculogenesis and oocyte development. In conclusion, the immunolocalization of Ang II, its related peptidases, AT1R, and VEGF were studied in the human ovary and corpora lutea. Our data provided the possibility of the involvement of Ang II in folliculogenesis and luteal function in association with neovascularization by Ang II-induced VEGF. Further studies are required to show the functional evaluation of the ovarian RAS and Ang IIinduced VEGF in the cycling ovary. Acknowledgments: The authors thank Ms. Hiroko Sato for technical assistance in the immunohistochemistry analysis. REFERENCES 1. Suzuki T, Sasano H, Tanaka R, Fukaya T, Yajima A, Nagura H. Cyclic changes of vasculature and vascular phenotypes in normal human ovaries. Hum Reprod 1998;13: Nemeth G, Pepperell JR, Yamada Y, Palumbo A, Naftolin F. The basis and evidence of a role for the ovarian renin-angiotensin system in health and disease. J Soc Gynecol Investig 1994;1: Donoghue M, Hsieh F, Baronas E, Godbout K, Gosselin M, Stagliano N, et al. A novel angiotensin-converting enzyme-related carboxypeptidase (ACE2) converts angiotensin I to angiotensin 1-9. Circ Res 2000; 87:E Tipnis SR, Hooper NM, Hyde R, Karran E, Christie G, Turner AJ. A human homolog of angiotensin-converting enzyme. Cloning and functional expression as a captopril-insensitive carboxypeptidase. J Biol Chem 2000;275: Nagatsu I, Gillespie L, Folk JE, Glenner GG. Serum aminopeptidases, angiotensinase, and hypertension: I. Degradation of angiotensin II by human serum. Biochem Pharmacol 1965;14: Hattori A, Matsumoto H, Mizutani S, Tsujimoto M. Molecular cloning of adipocyte-derived leucine aminopeptidase highly related to placental leucine aminopeptidase/oxytocinase. J Biochem (Tokyo) 1999;125: Culler MD, Tarlatzis BC, Lightman A, Fernandez LA, DeCherney AH, Negro-Vilar AF, et al. Angiotensin II-like immunoreactivity in human ovarian follicular fluid. J Clin Endocrinol Metab 1986;62: Delbaere A, Englert Y, Bergmann PJ, De Prez E, Giot JM, Deschodt- Lanckman M. Bioactive angiotensin (1-8) is the main component of angiotensin II immunoreactivity in human follicular fluid. Peptides 1996;17: Palumbo A, Jones C, Lightman A, Carcangiu ML, DeCherney AH, Naftolin F. Immunohistochemical localization of renin and angiotensin II in human ovaries. Am J Obstet Gynecol 1989;160: van Sande ME, Scharpé SL, Neels HM, van Camp KO. Distribution of angiotensin converting enzyme in human tissues. Clin Chim Acta 1985;147: Richard DE, Berra E, Pouyssegur J. Nonhypoxic pathway mediates the induction of hypoxia-inducible factor 1 alpha in vascular smooth muscle cells. J Biol Chem 2000;275: Otani A, Takagi H, Suzuma K, Honda Y. Angiotensin II potentiates vascular endothelial growth factor-induced angiogenic activity in retinal microcapillary endothelial cells. Circ Res 1998;82: Yoshimura Y, Karube M, Aoki H, Oda T, Koyama N, Nagai A, et al. Angiotensin II induces ovulation and oocyte maturation in rabbit ovaries via the AT2 receptor subtype. Endocrinology 1996;137: Ando H, Furugori K, Shibata D, Harata T, Murata Y, Mizutani S. Dual renin-angiotensin blockade therapy in patients at high risk of early ovarian hyperstimulation syndrome receiving IVF and elective embryo cryopreservation: a case series. Hum Reprod 2003;18: Rainey WE, Bird JM, Byrd W, Carr BR. Effect of angiotensin II on human luteinized granulosa cells. Fertil Steril 1993;59: Ando H, Nagasaka T, Nomura M, Tsukahara S-I, Kotani Y, Toda S, et al. Premenstrual disappearance of aminopeptidase A in endometrial stromal cells around endometrial spiral arteries/arterioles during the decidual change. J Clin Endocrinol Metab 2002;87: Shibata D, Ando H, Iwase A, Nagasaka T, Hattori A, Tsujimoto M, et al. Distribution of adipocyte-derived leucine aminopeptidase (A- LAP)/ER-aminopeptidase (ERAP)-1 in human uterine endometrium. J Histochem Cytochem 2004;52: Otani N, Minami S, Yamoto M, Shikone T, Otani H, Nishiyama R, et al. The vascular endothelial growth factor/fms-like tyrosine kinase system in human ovary during the menstrual cycle and early pregnancy. J Clin Endocrinol Metab 1999;84: Sugino N, Kashida S, Takiguchi S, Karube A, Kato H. Expression of vascular endothelial growth factor and its receptors in the human corpus luteum during the menstrual cycle and in early pregnancy. J Clin Endocrinol Metab 2000;85: Hirsschi KK, D Amore PA. Pericytes in the microvasculature. Cardiovasc Res 1996;32: Schlingemann RO, Oosterwijk E, Wesseling P, Rietveld FJR, Ruiter DJ. Aminopeptidase A is a constituent of activated pericytes in angiogenesis. J Path 1996;179: Yoshimura Y. The ovarian renin-angiotensin system in reproductive physiology. Front Neuroendocrinol 1997;18: Ando H, Harata T, Fukui R, Shibata D, Goto M, Shimomura Y, et al. RAS blockade therapy to prevent early OHSS. Hum Reprod 2004; 19:i McClure N, Healy DL, Rogers PA, Sullivan J, Beaton L, Haning RV Jr, et al. Vascular endothelial growth factor as capillary permeability agent in ovarian hyperstimulation syndrome. Lancet 1994;344: Hayashi K, Miyamoto A. Angiotensin II interacts with prostaglandin F2 alpha and endothelin-1 as a local luteolytic factor in the bovine corpus luteum in vitro. Biol Reprod 1999;60: Fertility and Sterility 439