Emergence of Uterine Pathology during Accelerated Biological Aging ...

0013-7227/02/$15.00/0 Printed in U.S.A.

Endocrinology 143(9):3618 –3627 Copyright © 2002 by The Endocrine Society doi: 10.1210/en.2001-211402

Emergence of Uterine Pathology during Accelerated Biological Aging in FSH Receptor-Haploinsufficient Mice NATALIA DANILOVICH, INDROJIT ROY,

AND

M. RAM SAIRAM

Molecular Reproduction Research Laboratory (N.D., M.R.S.), Clinical Research Institute of Montreál, Montreál, Que´bec 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, Montreál, Canada H3T 1M5; and Department of Medicine (M.R.S.), Universite´ de Montreál, Montreál, Canada H3T 1J4 A fully functional FSH receptor (Fshr) is required for ovarian follicular development and fertility. Fshr null females are sterile because of failure of follicular maturation, ovulation, and estrogen deficiency. Because Fshr-haploinsufficient females also begin to show age-dependent reproductive deficits that mimic biological aging, we have investigated the changes that occur in the uterus of these mice. The uterine weight in 12-month-old Fshr ⴙ/ⴚ mice increased 2-fold, and most retired breeders (those that stopped breeding earlier than our wildtype females) developed unilateral uterine masses that appeared similar to several abnormalities that also occur in women and associated with infertility. Curiously, there was a tendency for most of the abnormality to occur in the right horn. Up to 25% of the virgin Fshr-haploinsufficient mice also

O

VARIAN STEROIDS SECRETED under the influence of pituitary gonadotropins regulate the normal development and function of the rest of reproductive tract in females (uterus, vagina, and mammary gland) facilitating cyclic changes and sequential events required for maintaining pregnancy and lactation. In the uterus, the growth stimulatory actions of ovarian steroids are mediated primarily by interaction with respective nuclear receptors (1). With aging, altered ovarian steroidogenesis brings about significant changes to uterine functions affecting implantation and reduced capability to sustain pregnancy (2). In addition, many uterine abnormalities such as leiomyomas or most commonly called fibroids that usually occur in many women also contribute to infertility and other health problems (3). Although the diverse factors that cause the development of such pathologies are not fully understood, there is overwhelming evidence indicating that ovarian steroids, estrogen and progesterone in particular, are among the important hormones that influence tumor growth (4, 5). 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

Abbreviations: dpc, Days post coitus; Fshr, FSH receptor; FORKO, FSH receptor knockout; PR, progesterone receptor.

developed pathology. These transformations were not present in either wild-type mice or the estrogen-deficient Fshr null females at any age. In haploinsufficient females, estrogen and progesterone were reduced and testosterone was elevated in circulation by 1 yr. Fshr-haploinsufficient mice developed an imbalance of progesterone receptor isoforms A and B in the uterus. This alteration of progesterone receptors along with an increase in LH receptors in the uterus may contribute to the induction of high frequency of uterine pathology. Angiogenesis, vascular abnormality, and adenomyosis appeared to be increased in the uterine horn bearing pathological mass. The Fshr-haploinsufficient mice might help in understanding the molecular basis of induction of uterine pathology and tissue patterning. (Endocrinology 143: 3618 –3627, 2002)

null FORKOs also develop in the heterozygous females on aging (7, 9). The disappearance of fertility in Fshr-haploinsufficient 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 agerelated 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 Animals The study was performed with the approval of an institutional ethics committee, and all experiments were conducted according to standard stipulated guidelines of animal care. The animals of the required genotype were produced (7) by breeding 129T2/SV EmsJ Fshr ⫹/⫺ male and females of 3–5 months. A total of 99 ⫹/⫺ Fshr and 75 wild-type mice of different ages were used in this study. There were 27 mice of each genotype at 3 months; at 12 months, we compared 25 ⫹/⫺ and 20 wild-type mice. In addition, those females that had been mated two or more times for colony maintenance and classified as breeders were also considered. In this category, there were 47 ⫹/⫺ and 28 ⫹/⫹ females. Those ⫹/⫺ Fshr females that failed to reproduce after 8 –9 months of age were classified as retired breeders. In contrast to the Fshr ⫹/⫺ females, the wild-type mice continued to breed beyond 1 yr. The animals were housed (n ⫽ 5 or less per cage) under controlled lighting conditions (12

3618

Danilovich et al. • Uterine Pathology and Biological Aging in Mice

Endocrinology, September 2002, 143(9):3618 –3627 3619

h light, 12 h dark) with commercial food pellets (rodent laboratory chow 5001/Harlan Teklad S-2335 diet 1:1 mixture) and tap water provided ad libitum. No particular attention was paid to the presence of ingredients in the diet with potential phytoestrogenic activity.

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 ␮g/ml leupeptin). Then 50 ␮g 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, Que´ bec, Canada) and compared with the reported values for molecular weight. Band intensities were compared by densitometry for determining ratios at different ages.

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).

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 Ovarian steroids in aging Fshr ⫹/⫺ female mice

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.

The level of estradiol-17␤ in 3-month-old ⫹/⫺ females on the morning of proestrus tended to be decreased, compared with ⫹/⫹ mice of the same age. Estrogen levels decreased significantly in both genotypes at 12 months. The difference between the 12-month-old ⫹/⫺ and ⫹/⫹ females was also significant (Table 1). 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).

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-␮m 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.).

Higher rates of fetal resorption in haploinsufficient mice

We have already reported that Fshr ⫹/⫺ mice at 3 months produce fewer pups (average 5.6), compared with wild-type females (average 9.8) of the same age (6, 7). Because aging also affects the function of the uterus by altering the endometrium and reducing its ability to support the growth of the fetus (2), it was important to evaluate older mice in this

Western blot Western blotting of progesterone receptor (PR) in individual samples for the two genotypes at 3 and 12 months was performed as described

TABLE 1. Steroid hormone levels during aging and Fshr haploinsufficiency 3-month-old

Estradiol 17-␤ (pg/ml) Progesterone (ng/ml) Testosterone (ng/dl)

12-month-old

⫹/⫹

⫹/⫺

⫹/⫹

⫹/⫺

15.7 ⫾ 0.7 (n⫽12) 2.9 ⫾ 0.7 (n⫽10) 3.8 ⫾ 0.2 (n⫽ 6)

8 ⫾ 0.4 (n⫽14) 1.8 ⫾ 0.8 (n⫽10) 5.0 ⫾ 0.4 (n⫽15)

6.6 ⫾ 1.4a (n⫽13) 2.3 ⫾ 0.1 (n⫽ 6) 14.3 ⫾ 1.6a (n⫽ 6)

3 ⫾ 0.6b,a (n⫽14) 0.5 ⫾ 0.1c (n⫽ 6) 21.4 ⫾ 0.6b (n⫽14)

The levels of hormones determined by respective RIAs are shown. Values represent the mean ⫾ SEM. Letters denote statistical significance: indicates within a genotype between age groups, and b and c indicate within an age group across the genotypes. a, P ⬍ 0.005 vs. the appropriate control; b,c, P ⬍ 0.05 vs. the appropriate control.

a

3620

Endocrinology, September 2002, 143(9):3618 –3627

study. For this purpose, we considered breeding data for ⫹/⫺ and ⫹/⫹ mice between 8 and 10 and 11 and 13 months of age because the ⫹/⫺ females showed reduced productivity by about 8 months. To examine whether hormonal and/or receptor imbalances contribute to disturbances or pregnancy failure in ⫹/⫺mice, we determined the percentage of resorbed pups by counting normal, resorbed, or dead embryos in the uterus on 3 different days of gestation. In the 8to 10-month group, the percentage of resorbed pups per pregnancy in ⫹/⫺ females (12%) appeared to be elevated, compared with wild-type females (7%) at 10 dpc, a difference that was significant (Fig. 1A). 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). At this age the wild-type mice also suffer higher losses of up to 40% at 18 dpc. However, at this age 73% of the ⫹/⫹ females delivered an average of 4.5 ⫾ 2 healthy-looking viable pups. In the older ⫹/⫺ females, losses occurred early and at significantly higher rates. Thus, 14 dpc resorption had already reached 92%. Of the 36 ⫹/⫺ Fshr females that failed to give birth after mating, 21 were killed 2 d after the expected date of parturition. In these females, there was evidence of a pregnancy that had begun but later failed. Uterine implantation scars and/or remnants of resorbing fetoplacental tissue were clear indicators of this phenomenon. When we pooled the data for the 8- to 13-month-old ⫹/⫺ females, the overall number of pups per female was 1.0 ⫾ 0.4 (range 0 – 4). When four ⫹/⫺ Fshr females were killed on the day they


gave birth to a single stillborn pup, all were carrying one or two additional pups that were at various stages of resorption. These data indicate that there was a higher rate of pregnancy failure in the Fshr-haploinsufficient mice in comparison with ⫹/⫹ females. 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-monthold 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. We reported previously that at 3 months of age, the uterine histology of the Fshr heterozygous female is not distinguished from a wild-type littermate (7). However, histological evaluation of transverse sections from the uterine horn of 12-month-old ⫹/⫺ mice showed abnormally enlarged uterus with hyperplastic and disorganized luminal epithelium (Fig. 2, C and F). The endometrial glands were also enlarged and the epithelia were hypertrophied. The 12month-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

FIG. 1. Differences in pregnancy maintenance. Score of postimplantation pregnancies in wild-type and heterozygous mice between 8 and 10 (A) and 10 and 13 (B) months of age. Scoring was done as described in Materials and Methods. In both A and B, n ⫽ 4 – 6 per genotype and per interval. *, P ⬍ 0.05 vs. the appropriate control. **, P ⬍ 0.005 vs. the appropriate control.

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 his-



FIG. 2. Uterine abnormality in Fshr ⫹/⫺ mice. A, Uterine weight (wet weight) in wild-type and Fshr heterozygous mice at different ages. For this evaluation we selected older Fshr ⫹/⫺ mice that did not show visible appearance of tumors. Letters denote statistical significance: b indicates within an age group across the genotypes, and c and d indicate within a genotype between age groups. b,c,d, P ⬍ 0.05 vs. the appropriate control. B, 12-month-old wild-type uterine transverse section (⫻25). C, 12-month-old ⫹/⫺ uterine transverse section that is much larger and contain cysts (Cy) and hypertrophied epithelium (⫻25). D and E represent a higher magnification of the boxed areas in B and C (⫻200). There appears to be an increased angiogenesis in the stroma of ⫹/⫺ uterus (arrows indicate blood vessels). F, Example of adenomyosis in older virgin ⫹/⫺ uteri with endometrial glands (*) and stroma present within the myometrium (My) with no connection to the endometrial cavity. G, Hemosiderin-laden macrophages (arrows) were present abundantly within the uteri of ⫹/⫺ females. Hemosiderin is a hemoglobin-derived pigment containing iron. In areas of hemorrhage, hemosiderin, a goldenyellow to brown granular pigment, is phagocytized by macrophages.

tology. 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 cyst-

3622



FIG. 3. Patterns of uterine abnormalities in Fshr heterozygous mice. A, Female genital organs of virgin 18-month-old Fshr ⫹/⫺ mouse showing a nodular uterine neoplasm in the right uterine horn (arrowhead). Note the ovarian cyst on the contralateral side (9) B, Representative picture of a very large uterine mass present in the right horn of a Fshr heterozygous 14-month-old mouse. As indicated in Materials and Methods, we classify 8- to 9-month ⫹/⫺ females as retired (failed) breeders because they no longer reproduce. The diameter of tumor is illustrated by a double-ended arrow and equals approximately 3.5 cm. The left, nonaffected uterine horn is shown by an arrowhead. C, Histology of the affected right uterine horn in a virgin ⫹/⫺ mouse (⫻6.25). Note the part of uterine horn (Ut) filled with large mass (Tu) resembling an organized thrombus.


like 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). Because progesterone action is also important for uterine growth and function, we also compared the expression of this nuclear receptor in the uterus as a marker of estrogenic influence and aging-related modifications. Western blot of uterine extracts from 3- and 12-month-old virgin Fshrhaploinsufficient mice showed an imbalance in the two forms of PR, A and B, that arise from the same gene (Fig. 5). 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

The phenomenon of reduced fertility during the condition of gene haploinsufficiency in mutant animals derived from homologous recombination studies is of considerable experimental and clinical interest because such animals allow us to investigate gene dose-related effects because they become apparent on aging. As reported previously, histological analyses of the uteri of 3-month-old mutants with haploinsufficiency of the Fshr did not show any significant morphological alteration, compared with age-matched wild-type females (7). However, they produced smaller litters (40% less), even when they were mated with wild-type males, indicating that females had reduced fertility. By about 9 months, they had stopped breeding altogether. We have now characterized this phenomenon as accelerated biological aging following dramatic alterations occurring in the ovarian structural compartments of Fshr-haploinsufficient mice (8). In this investigation, we have extended this work to understand some of the aging changes that manifest in the uteri of these mice. It is remarkable that by 12 months of age, the Fshr ⫹/⫺ female is characterized by the presence of an abnormally enlarged uterus (Fig. 2, A and C). The occurrence of a similar phenotype described in mice lacking the PR (13)

D, An enlargement of the top boxed square area in C showing endometrium (En) and myometrium (My) containing many dilated vessels (arrows). E, High-power magnification of the small inset in C demonstrating classical features of a venous thrombus composed of red blood cells (Rbc) and fibrin (Fib). F, Transversal uterine section of retired ⫹/⫺ breeder has regions of calcification localized in vessels near the myometrium (My) (outlined by arrowheads). En, endometrium. G, Uterine mass localized in the endometrium of a ⫹/⫺ retired breeder has a circumscribed structure pushing the borders. H, Higher magnification of the inset in G. Note an admixture of different cells with eosinophilic cytoplasm: giant multinucleated cells similar to syncytiotrophoblasts (arrowheads) and mononuclear cells resembling cytotrophoblasts (arrows). This pathology was accompanied by abundant blood vessels containing red blood cells.

3624



FIG. 4. Immunohistochemistry of the uterine tissue at different ages. A–D demonstrate immunostaining of uterine sections with estrogen receptor ␣ anitbody. Heterozygous uterus (B) expresses a higher level of protein in the endometrium (En). Note nuclear staining confined to endometrial glands (*) and stromal cells (arrows). Arrows in C and D indicate immunopositive staining in endothelial cells of uterine blood vessels. E–H, Immunohistochemistry of the uterus with LH-R antibody. There is immunopositive staining localized to luminal (Lu) and glandular epithelium (*) as well as stroma (St). Arrows indicate blood vessels. Note intense staining for LH-R in the endothelial cells of blood vessels in the ⫹/⫺ representative mouse (H). I and J are immunohistochemistry for actin.


FIG. 5. Status of PR in the mouse uterus. Western blot of uterine extracts was performed with PR antibody at 3 and 12 months of age. For the Fshr ⫹/⫺, the left horn without tumor was used. Probing with ␤ actin antibody confirmed equivalent protein loading. The bottom panel depicts densitometric scanning of the PR-A and PR-B isoforms. Asterisks denote significance, compared with respective values in the ⫹/⫹ uterus, P ⬍ 0.05. The following PR-B/PR-A ratios were calculated for the genotypes at the two different ages. ⫹/⫹, 1.4 and 5; ⫹/⫺, 2.5 and 0.3.

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 LH␤ 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 spec-


trum 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 Fshrhaploinsufficient 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

3626


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-monthold ⫹/⫺ 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 ␤ 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-monthold 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 ␤ 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 We thank Drs. N. R. Moudgal, G. L. Greene, P. Chambon, and A. F. Parlow for providing us the various reagents used in this study. The assistance of Ms. Agneta Balla in immunohistochemistry and that of Ms. Yinzi Yang in animal management is greatly appreciated. Received December 12, 2002. Accepted May 27, 2002. Address all correspondence and requests for reprints to: M. Ram Sairam, Ph.D., Director, Molecular Reproduction Research Laboratory, Clinical Research Institute of Montre´ al, 110 Pine Avenue West, Montre´ al, Que´ bec, Canada H2W 1R7. E-mail: [email protected]. This work was supported by grants from the Canadian Institute of Health Research. N.D. is the holder of a doctoral study award from the Canadian Institute of Health Research.

References 1. Couse JF, Korach KS 1999 Estrogen receptor null mice: what we have learned and where will they lead us? Endocr Rev 20:358 – 417 2. Von Saal FS, Finch CF, Nelson JF 1994 Natural history and mechanisms of reproductive aging in humans, laboratory rodents, and other selected vertebrates. In: Knobil E, Neill JD, eds. The physiology of reproduction. 2nd ed. New York: Raven Press; 1213–1314 3. Buttram V, Reiter R 1981 Uterine leiomyomata: etiology, symptomology, and management. Fertil Steril 36:433– 445 4. Wilson LS, Yang F, Rees ED 1980 Estradiol and progesterone binding in uterine leiomyoma and in normal uterine tissues. Obstet Gynecol 55:20 –24 5. Rein MS, Barbieri RL, Friedman AJ 1995 Progesterone: a critical role in the pathogenesis of uterine myomas. Am J Obstet Gynecol 172:14 –18 6. Dierich A, Sairam MR, Monaco L, Fimia GM, Gansmuller A, LeMeur M, Sassone-Corsi P 1998 Impairing follicle-stimulating hormone (FSH) signaling in vivo: targeted disruption of the FSH receptor leads to aberrant gametogenesis and hormonal imbalance. Proc Natl Acad Sci USA 95:13612–13617 7. Danilovich N, Babu PS, Xing W, Gerdes M, Krishnamurthy H, Sairam MR 2000 Estrogen deficiency, obesity and skeletal abnormalities in follicle-stimulating hormone receptor knockout (FORKO) female mice. Endocrinology 141:4295– 4308 8. Danilovich N, Sairam MR 2002 Haploinsufficiency of the follicle-stimulating hormone receptor accelerates oocyte loss inducing early reproductive senescence and biological aging in mice. Biol Reprod 67:361–369 9. Danilovich N, Roy I, Sairam MR 2001 Ovarian pathology and high incidence of sex cord tumors in follitropin receptor knockout (FORKO) mice. Endocrinology 142:3673–3684 10. Mann RJ, Keri RA, Nilson JH 1999 Transgenic mice with chronically elevated luteinizing hormone are infertile due to anovulation, defects in uterine receptivity, and midgestation pregnancy failure. Endocrinology 140:2592–2601 11. Moudgal NR, Krishnamurthy HN, Surekha S, Krishnamurthy H, Dhople VK, Nagaraj R, Sairam MR 2001 Immunobiology of a synthetic luteinizing hormone receptor peptide 21– 41. J Androl 22:992–998 12. Krishnamurthy H, Kats R, Danilovich N, Sairam MR 2001 Intercellular communication between the Sertoli and Leydig cells in the absence of folliclestimulating hormone. Biol Reprod 65:1201–1207 13. Lydon JP, DeMayo FJ, Funk CR, Mani SK, Hughes AR, Montgomery CA, Shyamala G, Conneely OM, O’Malley BW 1995 Mice lacking progesterone receptor exhibit pleiotropic reproductive abnormalities. Genes Dev 9:2266 – 2278 14. Brzyski RG, Grow DR, Sims J, Seltman H 1995 Increase in androgen:estrogen ratio specifically during low-dose follicle-stimulating hormone therapy for polycystic ovary syndrome. Fertil Steril 64:693– 697 15. Shyamala G, Yang X, Cardiff RD, Dale E 2000 Impact of progesterone receptor

16. 17.

18.

19. 20. 21. 22. 23.

24.

25. 26. 27. 28.

29. 30. 31. 32. 33.

34.

on cell-fate decisions during mammary gland development. Proc Natl Acad Sci USA 97:3044 –3049 Conneely OM, Mulac-Jericevic B, Lydon JP, De Mayo FJ 2001 Reproductive functions of the progesterone receptor isoforms: lessons from knock-out mice. Mol Cell Endocrinol 179:97–103 Viville B, Charnock-Jones DS, Sharkey AM, Wetzka B, Smith SK 1997 Distribution of the A and B forms of the progesterone receptor messenger ribonucleic acid and protein in uterine leiomyomata and adjacent myometrium. Hum Reprod 12:815– 822 Nisolle M, Gillerot S, Casanas-Roux F, Squifflet J, Berliere M, Donnez J 1999 Immunohistochemical study of the proliferation index, oestrogen receptors and progesterone receptors A and B in leiomyomata and normal myometrium during the menstrual cycle and under gonadotrophin-releasing hormone agonist therapy. Hum Reprod 14:2844 –2850 Rao CV 2001 An overview of the past, present and future of nongonadal LH/hCG actions in reproductive biology and medicine. Semin Reprod Med 19:7–17 Reshef E, Lei ZM, Rao CV, Pridham DD, Chegini N, Luborsky JL 1990 The presence of gonadotropin receptors in nonpregnant human uterus, human placenta, fetal membranes, and decidua. J Clin Endocrinol Metab 70:421– 430 Zheng M, Shi H, Segaloff DL, Van Voorhis BJ, Zheng M 2001 Expression and localization of luteinizing hormone receptor in the female mouse reproductive tract. Biol Reprod 64:179 –187 Rao CV, Alsip NL 2001 Use of the rat model to study hCG/LH effects on uterine blood flow. Semin Reprod Med 19:75– 85 Shemesh M, Gurevich M, Mizrachi D 1997 Expression of functional luteinizing hormone (LH) receptor and its messenger ribonucleic acid in bovine uterine veins: LH induction of cyclooxygenase and augmentation of prostaglandin production in bovine uterine veins. Endocrinology 138:4844 – 4851 Danilovich N, Javeshghani D, Xing W, Sairam MR 2002 Endocrine alterations and signaling changes associated with declining ovarian function and advanced biological aging in follicle-stimulating hormone receptor haploinsufficient mice. Biol Reprod 67:370 –378 Lindzey J, Curtis S, Washburn T, Korach K 1996 Uterotropic effects of dihydrotestosterone in estrogen receptor knockout and wild-type mice. Proc 10th International Congress of Endocrinology, San Francisco, CA, vol I, OR18-4 Fitzgerald C, Zimon AE, Jones EE 1998 Aging and reproductive potential in women. Yale J Biol Med 71:367–381 Ma W, Tan J, Matsumoto H, Robert B, Abrahamson DR, Das SK, Dey SK 2001 Adult tissue angiogenesis: evidence for negative regulation by estrogen in the uterus. Mol Endocrinol 15:1983–1992 Lee BS, Nowak RA 2001 Human leiomyoma smooth muscle cells show increased expression of transforming growth factor-␤3 (TGF␤3) and altered responses to the antiproliferative effects of TGF␤. J Clin Endocrinol Metab 86:913–920 Hickey M, Fraser IS 2000 Clinical implications of disturbances of uterine vascular morphology and function. Baillieres Clin Obstet Gynaecol 14:937–951 Cramer SF, Patel A 1990 The frequency of uterine leiomyomas. Am J Clin Pathol 94:435– 438 Moorehead ME, Conrad CJ 2001 Uterine leiomyoma: a treatable condition. Ann NY Acad Sci 948:121–129 Wilcox LS, Koonin LM, Pokras R, Struss LT, Xia Z, Peterson HB 1994 Hysterectomy in the United States, 1988 –1990. Obstet Gynecol 83:549 –555 Burroughs KD, Fuchs-Young R, Davis B, Walker CL 2000 Altered hormonal responsiveness of proliferation and apoptosis during myometrial maturation and the development of uterine leiomyomas in the rat. Biol Reprod 63:1322– 1330 Cho YY, Kang MJ, Sone H, Suzuki T, Abe M, Igarashi M, Tokunaga T, Ogawa S, Takei YA, Miyazawa T, Sasano H, Fujino T, Yamamoto TT 2001 Abnormal uterus with polycysts, accumulation of uterine prostaglandins, and reduced fertility in mice heterozygous for acyl-coA synthetase 4 deficiency. Bichem Biophys Res Commun 282:993–997