| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Division of Reproductive Sciences, Oregon National Primate Research Center,3
Department of Physiology and Pharmacology,4 Oregon Health & Science University, Portland, Oregon 97239
Department of Environmental and Biomolecular Systems,5 OGI School of Science & Engineering, Oregon Health & Science University, Beaverton, Oregon 97006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
angiogenesis, angiopoietin-1, angiopoietin-2, corpus luteum, follicle, menstrual cycle, ovary, ovulation, preovulatory follicle, rhesus monkeys
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Recent studies indicate that angiopoietins play a critical role in vessel development and regression. To date, four different angiopoietins—ANGPT1 [10], ANGPT2 [4], and Angptl1 in mice and ANGPT4 [11] in humans—have been isolated. The angiopoietins appear to bind to one receptor tyrosine kinase, TEK. But a striking feature of the angiopoietin family is the opposing effects of these ligands when they bind to the receptor. ANGPT1 elicits an activation of TEK by increased tyrosine phosphorylation of TEK. ANGPT2 acts as an antagonist to inhibit receptor activation and can specifically block ANGPT1-dependent receptor-mediated processes [4]. Like ANGPT1, ANGPT4 appears to function as an agonist for the human TEK [12]. However, there is conflicting evidence as to whether ANGPTL1 acts as an agonist or antagonist to mouse TEK [11, 12]. Unlike VEGF, the ANGPT-receptor system does not promote endothelial cell proliferation or migration. Rather, the endogenous agonist ANGPT1 is believed to recruit periendothelial support cells to promote vessel maturation and maintains vessel integrity. Contrarily, as a natural antagonist, ANGPT2 acts to destabilize existing vessels, loosening the supporting cell matrix to allow angiogenic factors such as VEGF to stimulate endothelial cell proliferation and migration during early angiogenesis [13]. Alternatively, in the relative absence of VEGF or other angiogenic factors, ANGPT2 appears to destabilize vessels to the point of vessel degeneration [13] by the loss of periendothelial support cells [14]. Much less is known about the expression and actions of the ANGPT-TEK system in the ovary. There are few reports on ANGPT-receptor (TEK) expression or action in the ovary [6, 15, 16], but there are observations that ANGPT1 (but not ANGPT2) expression is elevated in luteinizing cells of the ovulatory follicle [5], whereas ANGPT2 expression is associated with follicle atresia [4] and corpus luteum regression [6].
The importance of increased vascularity for follicle growth and subsequent corpus luteum development has been implied for years [1, 2, 7]. If true, one would hypothesize that treatment with an antiangiogenic agent would disrupt vessel development and function (and perhaps cause vessel degeneration), thereby preventing the gametogenic (oocyte) and endocrine (steroid hormone) functions of the ovary. Recent experiments support this hypothesis [17], including evidence that systemic or local delivering of anti-VEGF compounds disrupts folliculogenesis [18, 19], ovulation [18], plus the development and maintenance of the corpus luteum [18, 20, 21]. However, to date, no reports have addressed the actions of ANGPT agonists or antagonists in the ovary. Therefore, studies were designed to test the hypothesis that exogenous ANGPT antagonist (ANGPT2), but not agonist (ANGPT1), would disrupt ovarian function when delivered directly into the preovulatory follicle of the ovary during the spontaneous menstrual cycle in the rhesus monkey.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The general care and housing of rhesus monkeys was provided by the Division of Animal Resources at the Oregon National Primate Research Center (ONPRC) [22]. Animal protocols and experiments were approved by the ONPRC Animal Care and Use Committee, and studies were conducted in accordance with the Guide for the Care and Use of Laboratory Animals [23].
Adult female rhesus monkeys exhibiting normal menstrual cycles of approximately 28 days were bled daily by saphenous venipuncture beginning 6–8 days after the onset of menses. Serum concentrations of estradiol (E2) and progesterone were determined by specific electrochemoluminescent assay using a Roche Elecsys 2010 Analyzer (Roche, Indianapolis, IN) in the Endocrine Services Laboratory of the ONPRC [24]. The interassay variations were 6.1% for E2 and 5.4% for progesterone, and the limits of sensitivity were 5 pg/ml for E2 and 0.03 ng/ml for progesterone. E2values were used to estimate the day of the midcycle gonadotropin surge and to select the day of intrafollicular injection as previously reported [18]. The day of the LH surge was subsequently confirmed using a mouse Leydig cell bioassay for serum LH [25], and verified that all animals were injected on the day before (Day –1) or the day of (Day 0) the midcycle gonadotropin surge. Day 1 of the luteal phase is designated as the day after the LH surge.
Intrafollicular Injection and Evaluation
Intrafollicular injection was performed on anesthetized animals as previously described [18, 26] during a laparotomy to expose the ovary bearing the dominant follicle. The needle on an insulin syringe containing 50 µl of solution was inserted through the stroma of the ovary before penetrating the follicular wall. Then, 50 µl of follicular fluid was aspirated into the syringe, diluting the injectable by half, before injecting 50 µl of this mixed solution into the follicle. Upon removal of the needle, the follicle was observed to make sure that deflation did not occur. Ovaries were viewed and photographed by laparoscopy at 3 days postinjection for evidence of follicle rupture and luteinization. Blood samples were collected on a daily basis throughout the expected luteal phase interval for 18 days or until menses, and analyzed for serum E2 and progesterone levels.
A sequential experimental design was typically employed whereby animals received an intrafollicular injection of vehicle (0.1% BSA in PBS, n = 7), and then ANGPT agonist or ANGPT antagonist (recombinant human ANGPT1, or recombinant human ANGPT2; R&D Systems, Minneapolis, MN) in subsequent treatment cycles. The final delivery dose of ANGPT1 was 20 µg (n = 4). The final delivery doses of ANGPT2 were 2 µg (n = 6) and 20 µg (n = 5). After menses in each protocol, animals were permitted to recover for at least one menstrual cycle before starting the next protocol.
Tissue Collection and Analysis
In further protocols, animals again acted as their own controls but were assigned randomly to receive an intrafollicular injection of either vehicle (n = 3), or the compounds (n = 3/treatment) on Day –1 or Day 0. Three days after the injection, the ovary bearing the injected follicle was removed by laparoscopy [18, 26] and fixed in formaldehyde overnight before embedding in paraffin. The paraffin-embedded tissues were serially sectioned at 5 µm in the Imaging & Morphology Core Laboratory at ONPRC using an American Optical (Southbridge, MA) microtome and mounted on Superfrost/Plus slides (Fisher, Santa Clara, CA). Slides were stained with hematoxylin-eosin and viewed via Zeiss Axioplan microscopy to evaluate ovulation (including oocyte release) and luteinization of the injected follicle. A CoolSNAP CCD Camera (Photometrics Inc. Tucson, AZ) was used for image capture.
Statistics
Estradiol and progesterone levels in serum were analyzed by analysis of variance with repeated measures to identify differences between the controls and each treatment group (SigmaStat statistical software version 2.0; SigmaStat, Chicago, IL). Differences were considered significant at P < 0.05 and values are presented as mean ± SEM.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The pattern and levels of serum E2 (not shown) and progesterone (Fig. 1, shaded region) during the luteal phase in vehicle-treated (controls; n = 7) animals were typically comparable to those observed in untreated animals during natural menstrual cycles in our colony [27]. Progesterone levels increased on Day 0 and continued to rise until the midluteal phase (Days 5–10), and then declined through the late luteal phase (Days 13–17) until menstruation. Compared with that of controls, injection of ANGPT1 (20 µg, n = 4) into the preovulatory follicle did not significantly alter progesterone levels except for 2 days in the late luteal phase (Day 10 and Day 13; Fig. 1). Reflecting this late difference, the average length of the luteal phase in the ANGPT1-treated group was significantly shorter than that of controls (14.5 ± 1.0 days vs. 17.4 ± 0.8 days, P < 0.05). There was no change in serum E2 levels after ANGPT1 treatment compared with that of controls (data not shown).
|
By 3 days after the intrafollicular injection, laparoscopic evaluation established that all vehicle- and ANGPT1-injected follicles (Fig. 2, A and B) displayed indices of ovulation (protruding stigmata) and luteinization (yellowish tissue with prominent capillaries). Subsequent evaluation of serial sections of ovaries (n = 3) collected 3 days after intrafollicular injection of vehicle revealed an ovulatory canal (Fig. 3A) that ran from within the luteinizing tissue to the surface of the ovary, as well as the absence of an oocyte. Developing vessels and enlarging granulosa luteal cells abounded in the luteinizing granulosa layer (Fig. 3B). Similar to the controls, the ovulatory canal bordered by luteinizing tissue was obvious in all three ANGPT1-treated ovaries (Fig. 3C), and no oocytes were detected in the injected follicles after serial section. Intact and well-developed vessels and enlarging luteal cells were present in the granulosa layer (Fig. 3D).
|
|
Intrafollicular Injection of Low-Dose ANGPT2
Injection of 2 µg of ANGPT2 into the preovulatory follicle of rhesus monkeys significantly decreased progesterone levels during the early luteal phase as a group (n = 6) compared with that of controls (Fig. 4). However, the progesterone levels were comparable to those of vehicle-controls by the mid to late luteal phase. When treated animals were examined individually, low-dose ANGPT2 injection suppressed progesterone concentrations throughout the luteal phase in three of six animals. However, the length of the luteal phase was not altered between controls (17.4 ± 0.8 days) and the low-dose ANGPT2 group (17.3 ± 1.8 days). Serum E2 levels in this ANGPT2 treatment group were comparable to those of the control group (not shown).
|
Laparoscopic evaluation at 3 days postinjection indicated that half (three of six) of the follicles in this ANGPT2 treatment group displayed indices of ovulation similar to those of vehicle-control animals (Fig. 2C). However, the other (three of six) low-dose ANGPT2 injected follicles did not exhibit any signs of follicular rupture.
Intrafollicular Injection of High-Dose ANGPT2
Unlike the typical pattern and levels of serum progesterone observed in vehicle-treated monkeys during the luteal phase, intrafollicular injection of 20 µg ANGPT2 (Fig. 5, n = 5) resulted in progesterone levels that never rose above baseline for 12–13 days. However, only one of five ANGPT2-treated animals mensed during this interval (Day 7 post-LH surge). Notably, a later rise in serum progesterone in ANGPT-treated monkey coincided with the decline of progesterone in control animals (Day 15–16). The pattern and levels of serum E2 also varied between controls and high-dose ANGPT2 treatment groups (Fig. 6) after the follicle was injected. After the expected decline in E2 levels following the LH surge, vehicle-treated animals exhibited typical low levels during the luteal phase (Fig. 6). In contrast, following the decline at midcycle, E2 levels in ANGPT2-treated animals increased to reach peak levels 11–12 days after the LH surge. Three of five animals displayed peak E2 levels that were comparable to those of prior preovulatory levels (>200 pg/ml), whereas the other two animals exhibited lower peak levels (Fig. 6). Notably, a subsequent LH surge and rise in serum progesterone were observed in the former, but not the latter, animals (Fig. 5).
|
|
Unlike in vehicle-injected controls, laparoscopy on Day 3 did not detect any sign of ovulation in high-dose ANGPT2-injected follicles (Fig. 2D). The ovaries bearing the injected follicle exhibited signs of distension with torn surface epithelium/tunica albuginea but a missing stigmata (Fig. 2D). Serial sections of ovaries (n = 3) collected 3 days after intrafollicular injection of high-dose ANGPT2 did not identify an ovulatory canal (Fig. 3E). Instead, the antrum of the unruptured follicle appeared filled with blood rather than follicular fluid. A degenerated oocyte was still in the injected follicle and surrounded by red and white blood cells (Fig. 3F).
Because it appeared unlikely that the injected follicle was the source of the subsequent rises in serum E2 (Fig. 6) and progesterone (Fig. 5) occurring 2 wk later in high-dose treated animals, another laparoscopy was performed on one animal at 10 days postinjection. No large structures (including the injected follicle) were visible on the injected ovary (Fig. 2E), but a large antral follicle was detected on the contralateral ovary (Fig. 2F).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Intrafollicular injection of 20 µg of ANGPT1 did not prevent ovulation or the timely development and function of the corpus luteum (as judged from circulating progesterone levels in the week following injection). As reported previously [18, 30], vehicle injection does not typically disrupt periovulatory events, and the results of local administration of ANGPT1 were similar to those of controls, at least until the late luteal phase of the cycle. Because the preovulatory and periovulatory follicle reportedly expresses ANGPT1 [5, 28, 29], the lack of acute effects of exogenous ANGPT1 may be due to ongoing actions of endogenous ANGPT1. Notably, transgenic mice overexpressing ANGPT1 appeared generally healthy and were fertile, although certain tissues such as the skin had many more and larger blood vessels [31].
However, some subtle changes were apparent in the current study following intrafollicular injection of ANGPT1: 1) although not quantified, the luteinizing tissues of the ovulatory follicle appear to have more extracellular space and larger, more organized vessels than the control follicles (Fig. 3D); and 2) the life span of the subsequent corpus luteum was consistently shortened by 2–3 days (Fig. 1, compared with controls). The mechanisms leading to these changes, and how events in the luteinizing ovulatory follicle relate to the luteal life span are unknown. Overexpression [32] or administration [33] of ANGPT1 promoted maturation of vessels by recruiting mural support cells (e.g., pericytes) and reducing vessel permeability (leakiness, leading to edema) in other animal models.
In contrast, intrafollicular injection of ANGPT2 resulted in a dose-dependent inhibition of ovulation and prevented the development and function of the subsequent corpus luteum. Although the effects varied between animals with the low dose (2 µg), the high dose (20 µg) consistently prevented follicle rupture and the expected rise in circulating progesterone levels in the 2 wk following injection. Serial sectioning of the ovary revealed a grossly degenerating follicle by 3 days post-ANGPT2 injection, with the remnants of the oocyte-cumulus complex in the antrum surrounded by red and white blood cells. These data are consistent with reports of an elevated ratio of ANGPT2:ANGPT1 expression in early atretic follicles in the cow [28] and pronounced ANGPT2 expression in a milieu of low VEGF during follicle atresia in the rat [4]. Thus, our results suggest that disrupting the balance toward angiolytic (i.e., ANGPT2) factors causes degeneration of the growing follicle, including the mature preovulatory follicle.
The effects of intrafollicular injection of ANGPT2 on follicle rupture and luteal development in the monkey were more pronounced than those following similar local administration of a general angiogenesis inhibitor (TNP-470) [30] or a specific VEGF antagonist (soluble VEGFR-1-immunoglobulin G chimera) [18]. Some of the differences may be related to the half-life of the agents (which is shorter for TNP-470 [30]) or the administered dose. However, dose-response studies with 1.5–30 µg sVEGFR-1 [18, 26] were able to block ovulation only in approximately half the animals, and to partially suppress serum progesterone levels during the subsequent luteal phase. Treatment with sVEGFR-1 did not prevent formation of the corpus luteum nor shorten the luteal life span in the menstrual cycle. Likewise, other investigators reported that systemic administration of VEGF antagonists to monkeys at midcycle or in the luteal phase (or both) suppressed serum progesterone levels but did not eliminate the luteal phase [21, 34]. In contrast, dose-response experiments in the current study indicate that exogenous ANGPT2 can completely block ovulation and eliminate the subsequent 2-wk luteal phase.
The differences in periovulatory effects of VEGF versus ANGPT antagonism may reflect differences in their cellular and molecular mechanisms of action in the follicle, as well as interaction with other microvascular factors. VEGF antagonism at midcycle/early luteal phase reduces endothelial cell area and microvessel development, resulting in increased extracellular space and smaller luteal cells in the corpus luteum [6, 34]. While these results support a critical role for the VEGF-R system in luteal development, at least in part via promotion of microvascular development, the lack of complete disruption may indicate important roles for other angiogenic factors such as ANGPT1 or endocrine gland-VEGF/prokineticin 1 [35, 36]. Also, despite evidence that VEGF antagonism at the midluteal phase can increase apoptosis in the primate corpus luteum [37], similar treatment at midcycle does not cause destruction of the preovulatory follicle to the extent that a corpus luteum fails to develop. In contrast, ANGPT antagonism at midcycle appears to initiate rapid dissolution of the preovulatory follicle; by 72 h the injected follicle displayed a grossly degenerating follicle wall and the entrapped oocyte was surrounded by antral fluid replete with blood cells. These results support a critical role for the ANGPT1 (agonist)-TEK system in promoting the health of the dominant preovulatory follicle, with antagonism by ANGPT2 leading to destruction of the follicle. Studies in other vascular beds suggest that ANGPT2 (depending on the relative level of VEGF activity) can either promote angiogenesis or angiolysis—the latter by causing loss of periendothelial support cells (e.g., pericytes [14]), endothelial cell death, and vessel regression [38]. However, other nonangiogenic actions of the ANGPT-TEK system in endothelial cells have been proposed [39]. Further studies are needed at earlier time points to evaluate the effects of ANGPT2 on follicular compartments [28], including the vasculature; this model could provide insight into the debate about whether follicular ischemia is a cause or consequence of atresia [40].
Notably, following intrafollicular ANGPT2 treatment at midcycle, the expected luteal phase was replaced by rising levels of estradiol, which in three of five animals reached preovulatory levels 11–12 days post-treatment that elicited an LH surge and a subsequent rise in serum progesterone levels. These data suggest that the treated follicle was replaced by another growing antral follicle that matured to the point of ovulation and luteinization approximately 2 wk later. Laparoscopy of one animal 10 days post-treatment confirmed the disappearance of the injected follicle and the appearance of another large antral follicle on the contralateral ovary. The hormonal patterns and emergence of another single ovulatory follicle within 12–14 days (i.e., the normal interval of the follicular phase) are remarkably similar to those observed in monkeys and women after ablation of the dominant structure (preovulatory follicle or corpus luteum) or removal of the ovary bearing the dominant structure during natural menstrual cycles [41–44]. The current data suggest that ANGPT2 treatment is comparable to physically removing the preovulatory follicle, thereby eliminating the dominant structure on the ovary and resetting the ovarian cycle to allow the selection and maturation of the next dominant follicle with timely ovulation and luteinization 2 wk later.
However, only three of five animals treated with high-dose ANGPT2 displayed subsequent follicular phases with expected estradiol levels and timely luteinization. Two of five animals displayed aborted follicular phases with elevated estradiol levels that did not reflect follicular maturation. In primates, it is generally recognized that the ovary bearing the dominant follicle in the subsequent cycle is not influenced by which ovary bears the ovulatory follicle in the preceding cycle [45]. Therefore, one would expect the dominant follicle in the subsequent cycle to develop on the ovary that received the ANGPT2 injection in 50% of the animals. It is tempting to speculate that in such instances, residual effects of ANGPT2 treatment, perhaps by disrupting the vasculature, prevented maturation of the next follicle selected soon after treatment.
In summary, these novel studies employing local administration of ANGPT agonist and antagonist into the preovulatory follicle indicate that ANGPT2 treatment 1) prevents the periovulatory events of ovulation and conversion of the follicle wall into a functional corpus luteum, and that these effects may be secondary to 2) causing degeneration of the dominant follicle, followed by resetting the ovarian cycle and in some instances the timely selection and maturation of the next ovulatory follicle. The data are consistent with the concept that the balance between ANGPT agonists (ANGPT1 and ANGPT4) and antagonists (ANGPT2 and ANGPTL1) in the developing antral follicle is critical for its growth and differentiation versus atresia. A tip in the balance toward angiolytic factors (i.e., ANGPT2) is not only associated with the results reported in references [4, 28], but may play a causal role (current data) in follicle degeneration, at least in the dominant antral follicle in primates.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Richard L. Stouffer, Division of Reproductive Sciences, Oregon National Primate Research Center, Oregon Health & Science University, 505 NW 185th Avenue, Beaverton, OR 97006. stouffri{at}ohsu.edu
Received: 17 October 2004.
First decision: 3 December 2004.
Accepted: 4 February 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Related articles in Biol Reprod:
This article has been cited by other articles:
|
|
 |
|
  J. Xu, F. Xu, J. D. Hennebold, T. A. Molskness, and R. L. Stouffer Expression and Role of the Corticotropin-Releasing Hormone/Urocortin-Receptor-Binding Protein System in the Primate Corpus Luteum during the Menstrual Cycle Endocrinology, November 1, 2007; 148(11): 5385 - 5395. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
