| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLE |
Molecular Reproduction Research Laboratory (N.D., M.R.S.), Clinical Research Institute of Montréal, Montréal, Québec H2W 1R7, Canada; Department of Medicine (N.D., M.R.S.), Division of Experimental Medicine, McGill University, and Department of Pathology (I.R.), St. Mary’s Hospital of McGill University, Montréal, Canada H3T 1M5; and Department of Medicine (M.R.S.), Université de Montréal, Montréal, Canada H3T 1J4
Address all correspondence and requests for reprints to: M. Ram Sairam, Ph.D., Director, Molecular Reproduction Research Laboratory, Clinical Research Institute of Montréal, 110 Pine Avenue West, Montréal, Québec, Canada H2W 1R7. E-mail: sairamm{at}ircm.qc.ca.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In examining the reproductive capacity of FSH receptor knockout (FORKO) mice recently generated in our laboratory (6, 7), we observed progressive declines in fertility in the heterozygous females. Although null female (FORKO) mutants are sterile, heterozygous female mice display reduced fertility from the very beginning and experience early reproductive senescence (7, 8). In addition, numerous conditions such as ovarian/uterine atrophy/pathology, obesity, and skeletal abnormality that appear in young 3-month-old null FORKOs also develop in the heterozygous females on aging (7, 9). The disappearance of fertility in Fshr-haplo-insufficient females by about 9 months suggested premature biological aging, and we have ascribed this phenomenon to an accelerated loss of ovarian function/oocytes (8).
The present study was undertaken to investigate the age-related changes in the uterus of the Fshr-haploinsufficient mice and evaluate whether the ensuing endocrine imbalances might contribute to the pregnancy failure and induction of pathology in this tissue. Our investigation reveals that uterine masses indicating pathology occur with relatively high frequency in these mice that undergo accelerated biological aging because of hormonal imbalances. Thus, this model might provide an experimental system to understand and manipulate the uterine conditions that afflict many women of reproductive age.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Steroid hormone and gonadotropin RIA
For characterization of hormonal profiles, blood samples obtained by the intracardiac method under ether anesthesia from wild-type and heterozygous mice were collected into centrifuge tubes containing 0.05% EDTA. Plasma was obtained by centrifugation and stored at -20 C until assay. Plasma concentrations of estradiol and testosterone were each determined in half of the samples. For measurement of progesterone, we used different samples from 3 and 12 months +/- and wild-type female mice. Samples from the 3-month-old females were collected at proestrus. For the 12-month-old mice, we ignored the stage of the cycle for both genotypes because this was either irregular or nonexistent for most +/- mice. Thus, these represented random collections. To perform RIAs (7), we used Coat-a-Count kits (Diagnostic Products Corp., Los Angeles, CA) following the manufacturer’s instructions. FSH and LH levels were estimated (9) by using RIA kits obtained from the National Hormone and Peptide Program (National Institute of Diabetes and Digestive and Kidney Diseases, courtesy of Dr. A. F. Parlow, University of California, Los Angeles, CA). Values are expressed as equivalents of mouse reference preparations (AFP-5308D for FSH and AFP-5306A for LH).
Breeding performance
Because we already knew the reduced breeding performance of 3-month-old +/- females (7), we extended this study to older mice. Groups of 8- to 10-month and 11- to 13-month animals were examined for their reproductive efficiency. In each case virgin Fshr-haploinsufficient and wild-type females were mated with proven wild-type males. Copulation plugs were identified on the morning following mating when the beginning of pregnancy was referred to as "1 dpc" (1 d post coitus). Females were killed at 10, 14, and 18 dpc, and postimplantation pregnancy sites were scored based on the presence of normal, resorbed, or dead embryos in the uterus (10). The appearances of the fetuses at different stages were noted.
Uterine histology and immunocytochemistry
Virgin females were killed at different ages, and the fresh mass of the dissected uterus was recorded. The uteri were immediately fixed overnight in 10% formalin, processed as serial paraffin histological sections of 5-µm thickness, and stained with hematoxylin and eosin. Some sections of animals with enlarged uterine masses were also examined. Uterine sections of older virgin mice were processed for immunohistochemistry. Sections were incubated overnight at 4 C with polyclonal antibodies generated against a synthetic peptide of rat LH-R (11, 12) and used at a dilution of 1:500. Actin antibody was obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). A rabbit polyclonal estrogen receptor 
peptide antibody no. 1280 was donated by Dr. Pierre Chambon (Universite de Louis Pasteur, Strasbourg, France). Binding of antibodies was visualized by the Immunosystem staining kit (Santa Cruz Biotechnology, Inc.).
Western blot
Western blotting of progesterone receptor (PR) in individual samples for the two genotypes at 3 and 12 months was performed as described previously (7). Fresh tissues were extracted with lysis buffer containing detergent and protease inhibitor cocktail (50 mM Tris-HCl, pH 7.2; 1% Nonidet P-40; 50 mM glycerophosphate; 5 mM dithiothreitol; 1 mM sodium vanadate; 0.05 mM NaF; 0.1 mM phenylmethylsulfonyl fluoride; and 5 µg/ml leupeptin). Then 50 µg protein were run on SDS-PAGE gels and transferred to nitrocellulose for reaction with the antibody at appropriate dilutions. Each experiment was performed two to three times with different extracts. For PR detection, we used the monoclonal antibody JZB39 from Dr. G. L. Greene (University of Chicago, Chicago, IL). As previously shown (7), this antibody recognizes both the PR A and B forms that arise from the same gene. After treatment of the blots with (1:2000) second antibody (Santa Cruz Biotechnology, Inc.), the complex was finally detected by the Amersham enhanced chemiluminescence kit (Amersham Pharmacia Biotech, Beaconsfield, Québec, Canada) and compared with the reported values for molecular weight. Band intensities were compared by densitometry for determining ratios at different ages.
Statistics
All data were expressed as mean ± SEM and analyzed by one-way ANOVA. When a significant effect was obtained with one-way ANOVA, a t test was used for analyzing the significance of the difference between two means. A P value < 0.05 was considered to be statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). There was a tendency for increase in the secretion of testosterone in the +/- ovaries as early as 3 months, compared with wild-type females of the same age, but this was not significant. However, this hormone increased in 12-month-old females of both genotypes with the levels being much higher in +/- (P < 0.05) (Table 1). The plasma concentration of progesterone was reduced in the +/- ovaries at both ages studied. There were no significant differences in the plasma concentrations of either LH or FSH between the two genotypes at 3 months of age. However, plasma LH was increased (by 77%) more than FSH (by 33%) in 12-month-old +/- mice in comparison with +/+ controls of the same age (data not shown).
|
). Starting from 14 dpc, pregnancy resorption rates in Fshr +/- females rose to 57% and reached 90% on d 18, whereas in wild-type this increase was low attaining a mean level of 18%. Thus, in this age group, the Fshr +/- animals suffered highly significant rates of loss. The pattern was exaggerated in the older age group (Fig. 1B).
|
Changes in uterine growth and histology
Mean uterine weights increased significantly with advancing age in both the wild-type and Fshr +/- mice (Fig. 2). For this experiment, we collected tissues from 3-month virgin females of both genotypes killed at proestrus, and at this age there was no difference. In calculating the data for the 7- and 12-month groups, no distinction was made between virgin and breeders, and these tissues were collected at random. By 1 yr of age, there were changes in both genotypes (Fig. 2A). The uteri from aged +/- mice were always heavier (by 77%), compared with that of age-matched littermates, and the pooled data shown in this figure are for animals that had no visible tumors. Some of them might have had microscopic tumors not yet apparent externally. Figure 3A shows the presence of a unilateral uterine mass (tumor) in a 20-month-old virgin +/- female. This animal also had an ovarian cyst on the contralateral side that we have described previously (9). However, many retired Fshr +/- mice at 12+ months had even developed a large mass in one uterine horn with a representative example shown in Fig. 3B.
|
|
, C and F). The endometrial glands were also enlarged and the epithelia were hypertrophied. The 12-month-old wild-type mice had normal appearance of all three layers of uterus (Fig. 2B). There were numerous blood vessels in the uterine stroma of the 12-month-old Fshr heterozygous mice (Fig. 2E), indicating evidence of probable acceleration of angiogenesis. These structures were clearly less abundant in the wild-type uterus (Fig. 2D).
Emergence of uterine nodular structure in Fshr +/- mutants
With increasing age, uterine masses were apparent in the Fshr-haploinsufficient mice. Interestingly, by 12 months of age, 10 of 43 (23%) virgin heterozygous females examined and most (38 of 47, 82%) retired (failed) heterozygous female breeders developed visible uterine abnormalities. There was a tendency for them to be localized predominantly (in 67% of cases) in the right uterine horn. Uterine abnormality was not observed in wild-type females, either virgin or retired breeders at any age studied. In some of the animals that were analyzed, the weight of abnormal structure alone in different +/- females varied from 205 to more than 4000 mg. Figure 3B illustrates an extreme example of a uterine tumor found in the right horn of the +/- uterus of retired breeder at 14 months of age with tumor weight exceeding 4 g (
8% of the body weight). Because we were interested only in illustrating its appearance, we did not preserve this specimen for histology. To the naked eye, this mass appeared to be engorged with numerous blood vessels and the left horn remained relatively pale in color. When the mass was cut, bundles of fibers could be seen. Sometimes uterine bleeding in the older +/- mice was the first sign that prompted us to check the animal for evidence of abnormalities.
Pattern of uterine pathology in Fshr +/- mice
The histological appearances of the uterus in both the virgin and retired breeders were different from the wild type at the corresponding ages. Examples of pathology in virgin +/- females are shown in Fig. 2, C–G. These included cystlike structures and enlarged lumen (Fig. 2C) and prominent population of small vessels in the endometrium (Fig. 2E). Some uteri showed endometrial glands deeply penetrating the myometrium, a picture that is highly consistent with adenomyosis (Fig. 2F). Other uteri contained dilated vessels in the endometrium without thrombosis but associated with hemosiderin containing macrophages (Fig. 2G). In some virgin +/- females, the changes had progressed further to create visible masses (example in Fig. 3A). Structures like this had large organizing thrombi in ectatic venous type channels located in the myometrium (Fig. 3C). Increased angiogenesis is apparent in the stroma (Fig. 3D) next to the thrombus. The enlargement in Fig. 3E revealed the red blood cells adjacent to the mass in the center. The masses in general of the retired +/- breeders were larger (Fig. 3B), and they also presented additional pathology. Like the older +/- virgins, many showed varying degrees of thrombi in the outer myometrium of the affected horn. Evidence of organization included presence of macrophages in growth of fibroblasts and calcification (Fig. 3F). This was also an indication of increased angiogenesis in the uterus. The larger masses that were always unilateral were circumscribed and pushing the uterine walls. These showed trophoblastic tissues and decidua (Fig. 3G) that presumably arose from a previous pregnancy that for some reason did not undergo resorption but persisted. In many females this condition prevailed 6–9 months after cessation of breeding. Similar microscopic examination of the uteri of older wild-type (virgin or retired) females showed no evidence of significant vascular dilatation or thrombosis. It is interesting to note that in null FORKO females that experience chronic estrogen deficiency (7), no such uterine abnormalities are seen at any age (3–12 months). Instead, they all develop profound unilateral ovarian pathology on aging; coincidentally, these tumors predominate on the right ovary (9).
Uterine marker genes in aging
In addition to the larger size of the uterus, one of the distinct features of the abnormality appeared to be changes in the vasculature of the uterus in older +/- Fshr mice. When we examined ER by using a specific antibody (Fig. 4, A–D), we could find higher expression in the +/- uterus in both the endometrium and stroma (Fig. 4, A vs. B). More significantly, the vessel size and expression of ER in endothelial cells appear to be higher in +/- mice (compare Fig. 4, C and D). Immunohistochemistry performed with an LH-R peptide antibody that was previously characterized (11, 12) revealed immunostaining in the luminal and glandular epithelium as well as stroma (Fig. 4, E and F). Although we have not quantitated the change, there appears to be a slight but consistent increase in the expression of LH-R in the nonaffected horn of the 12-month-old +/- uterus (Fig. 4F). Higher LH-R expression in uterine vessels of +/- mice were also apparent (Fig. 4, G and H). As a marker of smooth muscle cells, actin staining was verified. This assessment in older and virgin Fshr +/- mice that had not yet shown visible masses revealed an apparent higher expression of this protein, suggesting enhanced proliferation of smooth muscle cells (Fig. 4, I and J).
|
). This pattern was consistent and reproducible in all animals. The 3-month-old heterozygous uterus showed reduction in the two forms of PR, compared with wild-type littermates with the PR-A form being more affected. By 12 months of age, in the nonaffected horn of +/- mouse, there was a dramatic decrease in PR-B form and up-regulation of form A, compared with wild-type controls. The ratio of these two forms appeared to be completely reversed between the two genotypes at this age. In its trend, receptor expression in the 3-month-old +/- uterus looked more like the 1-year-old wild-type uterus, although the ratio was still different. At 12 months PR-B:PR-A ratio was about 5 in the +/+ uterus and 0.3 in the +/-; thus, PR-A was clearly more dominant in the latter.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, A and C). The occurrence of a similar phenotype described in mice lacking the PR (13) along with our data on perturbations in the ratio of the A and B forms of the PR at both 3 and 12 months of age (Fig. 5) suggests that maintenance of an optimal ratio is critical for preserving the uterine milieu that can support pregnancy. Our results demonstrating increased fetal loss in +/- females might also be due to the alteration in PR expression in the uterus, elevated LH, or/and androgen:estrogen ratio that could affect implantation and development of the embryos (14). Mice that chronically overexpress the LHß show pregnancy failure at midgestation because of the high LH concentrations, resulting in elevation of estradiol and testosterone (10). Although abnormalities in reproductive and other tissues are noted, no uterine tumors have been reported in these transgenic animals (10).
Considering the magnitude of uterine abnormality (Figs. 2 and 3) in the Fshr +/- mice, we surmised that the burden of mass observed in the example shown must be very significant. Based on histological and other staining characteristics, we have initially classified the abnormality as a spectrum of changes that include dilated vessels. These undergo thrombosis to different extents in both the old Fshr +/- virgin and retired breeders producing a mass. Although we cannot discount at this time the presence of other uterine pathologies in the aging Fshr mice, additional careful evaluations are needed. Perhaps the imbalance between two forms of PR along with other factors may also contribute to the high incidence of uterine abnormality in the Fshr-haploinsufficient mice, some of which become huge (Figs. 2 and 3). This observation may be consistent with other reports, which show that an imbalance in the ratio of the two isoforms can lead to alterations in PR signaling affecting critical cell fate decisions during normal or abnormal mammary development (15) or other receptor-dependent physiological states (16). The PR is unusual in the nuclear receptor superfamily in that two isoforms exist; these result from translation of different mRNAs from within the same gene and are estrogen regulated, performing tissue/cell specific functions (16). In women excessive tissue fibrosis and increased smooth muscle proliferation characterize several diseased states such as uterine leiomyomas. The higher concentrations of PR-A and PR-B in the affected tissue than in normal myometrium and in particular the dominance of form A over B in leiomyomata (17, 18) appears to be reproducible in our aging Fshr +/- females. Thus, a strong link between progression of the abnormality and PR imbalance in our Fshr +/- mice is currently favored as a working hypothesis. However, more studies are needed to explore the various abnormalities.
LH receptors that were previously thought to be expressed exclusively in specific gonadal cells may also be functional in other tissues as shown by many recent studies (19). Interestingly, our studies in agreement with other emerging reports in women (19, 20) and rodents (21) revealed that LH receptors are present in the endometrial epithelial and stromal cells as well as myometrial muscle of wild-type and Fshr +/- mouse uteri (Fig. 4, E–H). As also previously shown in human endometrium (19, 20), immunostaining for LH-R was more intense in glandular than stromal cells of the Fshr +/- mice (Fig. 4F). Comparison of LH-R immunostaining between the two genotypes, although not quantitative, suggested a slight but perceptible increase in immunopositive binding in +/- uterus (Fig. 4F). Whether this translates into a functional parameter of consequence to uterine dysfunction and pathology in the present context remains to be investigated. Thus, the overexpression of LH-R protein in the +/- uterus rising together with our finding of higher LH levels in 1-yr-old mutants may result in an enlarged uterus in aged +/- females, compared with wild-type littermates (Fig. 2, A and C). This would seem to support previous assumptions that the hormone may indeed have direct actions that may contribute to uterine hypertrophy (19, 20).
Among the various effects ascribed to LH (human chorionic gonadotropin) action on the uterus, two events in particular, i.e. effect on increasing uterine blood flow (22) and modulation of cyclooxygenase and prostaglandin production (23), may be relevant to the pathology observed in our animals. An apparent higher LH-R expression in the vessels (Fig. 4, G and H) appears to be consistent with literature reports and an action by elevated levels of the hormone in older Fshr females (24). Although we have not yet performed analysis of vascular endothelial growth factor/receptor expression in relation to uterine abnormality, it might be speculated that these events could also be contributing to the pathology because the masses appeared to have more blood vessels. The larger increase in uterine weight of 12-month-old +/- females, compared with wild-type littermates, can also be due to the high circulating testosterone that might be acting through the androgen receptor in the uterus (25). However, because testosterone also increased in the 1-yr wild-type mice (Table 1) and these mice did not develop uterine tumors, factors (or imbalances) in addition to androgens must conspire to induce the abnormalities we have observed in the Fshr +/- uterus. As mentioned earlier, LH ß transgenic mice that show high androgen levels do not develop uterine tumors (10).
Although our histological examination revealed an abundance of blood vessels in the uterine stroma of heterozygous females (Fig. 2E, arrows) that was distinctly different from the pattern in the wild-type uterus, additional studies will be required to demonstrate enhanced angiogenesis and molecular events associated with this process. It has been postulated that such changes in the uterine vasculature might reflect diminished tissue function during aging (26), leading to acceleration of uterine tumorigenesis in women. Recent studies addressing the relative roles of estrogen and progesterone (via their respective nuclear receptors) in mice have indicated different functions with respect to vascular permeability and angiogenesis (27). This report suggests that progesterone but not estrogen stimulates angiogenesis in the uterus, and, based on our present findings, we may hypothesize that the imbalance of its receptor forms is likely to drive the growth process on its path to neoplasia.
The appearance of spontaneous uterine pathology in the majority of the aging Fshr-haploinsufficient retired breeders and a significant number (23%) of virgins of this genotype indicate that hormonal imbalance does play an important role in the induction of uterine abnormality in mice. The higher incidence noted in Fshr-haploinsufficient mice that we have classified as retired breeders because they could no longer reproduce suggests that pregnancy or events associated with resorption/scarring may have exacerbated the induction of pathology. However, it is also plausible that hormone changes brought on by pregnancy and lactation in the Fshr +/- could contribute to their development. Because these were not reported in wild-type or control mice of any strain, we believe that this genetic model might provide useful insights into the hormonal control of apparent uterine tumorigenesis at a molecular level. Interestingly, why structures (Fig. 4G) resembling fetal tissue persist (or reappear) several months after a failed pregnancy and classification as retired breeder remains a mystery at this time. Their appearance in some older virgin mice is also very intriguing (data not shown). Interestingly, we have not observed uterine tumors in our FORKO null mutants at any age. The 3-month-old FORKO -/- females that are chronically deprived of estrogen and showing low levels of progesterone have uteri that remain infantile and underdeveloped (7). In these null mice, it appears that the presence of a 10-fold higher testosterone of ovarian origin had no influence in committing the uterus toward its progression to tumorigenesis. These observations suggest that estrogen action is indeed required for the induction of uterine pathology, a finding that is also in agreement with numerous clinical data (3, 4). In addition to steroids, a variety of growth factors such as increased expression of members of the TGF ß family and altered responses have been implicated in diseases such as human leiomyoma (28), but the contribution of these growth modulators in our mice and during aging remains to be explored.
Angiogenesis is apparently important to cyclic regeneration of the endometrium and maintenance of microvasculature, processes that appear to be altered in the Fshr +/- mice. These mice showed evidence of adenomyosis (Fig. 2F). Several clinical abnormalities such as adenomyosis and fibroids are associated with abnormal vascularity in the uterus (29). The former condition apparently coexists commonly with endometrial carcinoma (29). Uterine leiomyomas (fibroids) are the most common type of benign neoplasm of the reproductive system in premenopausal women with obesity contributing to increased risk (30, 31). Interestingly, the aging Fshr +/- females are also obese (8). Although the true incidence of fibroids in this group of women has not been determined, some studies suggest that it could vary from a low of 20% to as high as 77% with more than 40% above age 35 yr having leiomyoma (28, 29). Premenopause is also a period associated with the greatest frequency of hysterectomies, with many for symptomatic or enlarging fibroids. Hysterectomy is the second most frequently performed major operation (60%) in American women with nearly 600,000 procedures performed annually (32). Because about one third of these are performed for uterine fibroids and other conditions, there is a critical need to determine how hormones control tissue patterning and tumor biology at the molecular level to develop specific and nonsurgical treatments with minimal side effects for their elimination.
In addition to our present discovery of uterine dysfunction and subsequent pathology in Fshr-haploinsufficient mice, we are aware of at least two other reports that link haploinsufficiency of other genes causing similar if not identical phenomena. For example, about 30% of 1-yr-old Eker female rats that are heterozygous for a germ line mutation of the tuberous sclerosis (Tsc-2) tumor suppressor gene (33) develop hormone-responsive uterine leiomyomas. Mice that are heterozygous for Acyl-coenzyme A synthetase 4 deficiency that have reduced fertility also exhibit abnormal uterus with polycysts that are attributed to alteration of uterine prostaglandins (34).
In conclusion, this report documents the development of some uterine vascular pathology and adenomyosis in haploinsufficient Fshr females that have undergone premature biological aging. Understanding the endocrine and molecular basis of these abnormalities and loss of control of cell/tissue patterning in an experimental setting that accelerates reproductive aging might further the development of useful strategies for diagnosis and treatment. We believe that the Fshr-haploinsufficient mice might allow the exploration of genes involved in tissue patterning under conditions that produce hormonal/growth factor/receptor imbalances.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: dpc, Days post coitus; Fshr, FSH receptor; FORKO, FSH receptor knockout; PR, progesterone receptor.
Received December 12, 2002.
Accepted for publication May 27, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  T R. Kumar What have we learned about gonadotropin function from gonadotropin subunit and receptor knockout mice? Reproduction, September 1, 2005; 130(3): 293 - 302. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. P. Mayer, P. J. Devine, C. A. Dyer, and P. B. Hoyer The Follicle-Deplete Mouse Ovary Produces Androgen Biol Reprod, July 1, 2004; 71(1): 130 - 138. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
