| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
Subunit Promoter-Driven Simian Virus 40 T-Antigen: Ontogeny, Functional Characteristics, and Endocrine Effects1
a Department of Physiology, University of Turku, 20520 Turku, Finland
ABSTRACT
We previously reported formation of ovarian granulosa cell tumors with 100% penetration in a transgenic mouse model with murine inhibin
subunit promoter-driven (inh)/Simian Virus 40 T-antigen (Tag). The tumor-bearing inh/Tag mice showed highly elevated serum levels of immunoreactive inhibin. To investigate the onset of tumorigenesis and related endocrine consequences, 6–8 female mice of two inh/Tag lines and their mating control littermates were killed monthly between 1 and 6 mo of age. We also investigated tumorigenesis-related fertility aspects of these two mouse lines. The ontogeny and progression of tumors could be monitored in both inh/Tag lines by alterations of ovarian weights and serum hormone levels. Serum progesterone levels increased in both inh/Tag lines in an age-dependent manner as ovarian tumorigenesis progressed, and a reciprocal decrease occurred in serum LH and FSH. Neither serum estradiol (E2) nor uterine weights were significantly altered during tumorigenesis, suggesting that the ovarian tumors represented late stages of granulosa cell differentiation. In conclusion, the present findings show in the inh/Tag TG mice a relation between endocrine consequences of granulosa cell tumorigenesis, and a connection of onset of tumor formation with aberrant steroidogenesis and gonadotropin secretion. These findings indicate that tumors are endocrinologically active and able to exert enhanced negative feedback effects on pituitary function. The tumors provide a good model for endocrinologically active hormone-dependent tumors.
ovary
INTRODUCTION
Ovarian granulosa cell tumors are rare in women, accounting for 3.0%–7.6% of primary ovarian tumors, although the prospects of their treatment, in comparison to other ovarian cancers, are still poor [1, 2]. The tumor-related mortality rate with granulosa cell tumors is only 37.3% [3], whereas approximately 80% of patients die of recurrent disease [3, 4]. It is therefore of great importance to find circulating markers as early indicators of recurrent disease.
The molecular mechanisms of ovarian granulosa cell tumor formation are poorly understood. The granulosa cell tumors generally produce estradiol (E2), but at least 30% of them are steroidogenically inactive [5, 6]. Therefore, E2 is not an optimal biochemical marker candidate for granulosa cell tumors [5, 7]. It is postulated that inhibin acts in the normal ovary as a defense mechanism against proliferative effects of elevated gonadotropins [8, 9]. This is supported by the fact that ovarian granulosa cell tumors are often associated with elevated levels of inhibin and reduced levels of serum FSH [5, 7, 10]. It was reported that inhibin is produced by human granulosa cell tumors, and it was proposed as a marker of primary and recurrent granulosa cell tumors [5, 7, 11, 12]. We have earlier shown in our inh/Tag female mice with ovarian tumors up to 20-fold increased levels of serum immunoreactive inhibin [13]. The majority of the immunoreactive inhibin secreted by the inh/Tag ovarian tumors represented a free inhibin subunit, and the levels of chimeric inhibin A and inhibin B remained rather low [13].
As models for human malignancies, spontaneous granulosa cell tumors have been reported in mice [10], rats [14], and several domestic animal species, including cat, dog, sow, sheep, cow, and horse [15, 16]. Transgenic (TG) mouse models are useful by facilitating the understanding of hormonal and paracrine mechanisms involved in genesis, growth, and functions of the gonadal endocrine tumors [17, 18]. We have established a novel TG mouse model by "targeted gonadal tumorigenesis," using a 6-kilobase (kb) fragment of the murine inh promoter fused with the Simian Virus 40 T antigen (Tag) coding sequences [19, 20]. The gonadal tumors, originating from granulosa or Leydig cells, appeared in two established TG mouse lines (IT6-M and IT6-F) with 100% penetration at the age of 5–7 mo [19, 20]. We have further characterized the mechanisms of this gonadal tumorigenesis and found that the tumor growth is gonadotropin-dependent [21], resembling in this way the inhibin knockout mice, which also developed gonadotropin-dependent gonadal tumors [8, 22].
This study characterizes the ovarian tumorigenesis of two murine inh/Tag TG lines. The onset and progression of the ovarian tumors, the tumorigenesis-related fertility aspects, and the endocrine regulation of the granulosa cell tumors are examined. We also studied the pathophysiological and endocrine consequences of granulosa cell tumorigenesis, and a putative connection of onset of tumor formation with aberrant steroidogenesis and gonadotropin secretion.
MATERIALS AND METHODS
Experimental Animals
We used TG female mice from IT6-F and IT6-M lines, expressing the inh/Tag transgene, as earlier described [19]. As descendants of the two founder animals (IT6-F and IT6-M), the TG males were fertile even with advanced tumorigenesis, whereas the females of both lines were sterile [19]. Therefore, conventional breeding to produce homozygous IT6 lines was not possible. Mating TG males with C57Bl/6 (IT6-F line) or DBA/2J (IT6-M line) females maintained the IT6 lines. As control female mice, mating littermates of the C57Bl/6 (C57Bl) and DBA/2J (DBA) strains were used. Genotyping of the mice was performed from tail DNA by polymerase chain reaction (PCR) as earlier described [19, 20]. The mice were specific pathogen-free and housed two to four per cage in controlled conditions of light (12L:12D) and temperature (21°C ± 1°C). They were fed with mouse chow SDS RM-3 (Special Diet Service; E, soy free; Whitham Essex, UK) and tap water ad libitum. The University of Turku Ethical Committee on Use and Care of Animals approved all the procedures using mice for our experiments.
Ovulation Induction and Fertility Assessment
In order to induce superovulation, 6-wk-old female mice of the IT6-F (n = 9) and IT6-M (n = 9) lines, and for non-TG controls, C57BL (n = 5) and DBA (n = 5) mice, were treated with 5 IU of equine chorionic gonadotropin (eCG; Sigma, St. Louis, MO), followed after 47 h by 5 IU of hCG (Organon, Oss, The Netherlands) [23]. The occurrence of vaginal plugs and stage of estrous cycle were checked from vaginal smears. The fertilized ova were surgically removed from oviducts of live animals. Surviving 1- or 2-cell-stage embryos were transferred into the oviducts of Day 0.5 pseudopregnant non-TG F1 foster mice [23].
Tumor Ontogeny and Progression
Six to eight female mice of the IT6-F, IT6-M, C57BL, and DBA lines were killed monthly between 1 and 6 mo of age. The mice were anesthetized with i.p. administration of Avertin [23], body weights were recorded, and a longitudinal laparotomy was performed within a few minutes after induction of anesthesia. Blood samples were collected into heparinized syringes by cardiac puncture. Blood was allowed to clot overnight at 4°C, and centrifuged (300 x g) at room temperature to separate serum. The sera were stored at -20°C until analyzed. The ovaries were dissected out, weighed, and snap-frozen in liquid nitrogen, or fixed either in Bouin's solution or in 4% paraformaldehyde (PFA). Weights of the uteri and adrenal glands were also measured. Various other tissues (adrenal, pituitary, liver, lungs, kidney, spleen, thymus, heart, and brain) were also taken for RNA or histological analysis. Any signs of macroscopically observed metastases were always noted and confirmed by histopathological analysis.
Bouin's-fixed or 4% PFA-fixed paraffin sections (5-µm-thick) of ovaries were stained with hematoxylin/eosin for histological analysis.
Hormone Measurements
Serum FSH was measured by a double-antibody RIA method (National Institute of Diabetes and Digestive and Kidney Diseases; Bethesda, MD), where the hormone preparation was radioiodinated with sodium [125I]iodide (IMS 300; Amersham, Buckinghamshire, UK) using the chloramine-T method, as described earlier [24]. Serum LH was measured by a supersensitive immunofluorometric assay (Delfia; Wallac OY, Turku, Finland), developed in our laboratory for rat LH [25]. Progesterone and E2 were measured from diethyl ether extracts of the sera by RIAs as described earlier [26, 27].
The approximate assay sensitivities for FSH and LH were 250 and 0.75 pg/tube, respectively, for progesterone and E2, 50 fmol/tube, and 20 fmol/tube, respectively. The intraassay and interassay coefficients of variation for these assays were below 10%–15%, respectively.
In Vitro Production of Progesterone by Granulosa Tumor Cells
We used established, immortalized granulosa tumor cell lines, KK-1, derived from an IT6-M founder female mouse ovarian tumor [19] and NT-1 [28], derived from an IT6-F female ovarian tumor, and characterized in detail by us. The cells were plated 1 day before stimulation on 24-well culture dishes (Greiner Labortechnik, Frickenhausen, Germany), at a density of 105 cells/well in 500 µl of culture medium. NT-1 and KK-1 cells were cultured for 8 h in the presence and absence of 50 IU/L of recombinant (rc) FSH (Org 32489; Organon, OSS, The Netherlands) or 50 µg/L of highly purified hCG (NIH CR-121) or 10 µmol/L of forskolin (Sigma). All experiments were done in quadruplicate and repeated twice. Progesterone was measured from diethyl ether extracts of the incubation media by RIA as described before [27].
Statistical Analysis
A Macintosh version of the Super ANOVA program (Abacus Concepts, Inc., Berkeley, CA) was used for a paired t-test, followed by factorial tests, Duncan's new multiple range test, and Fisher's protected least significant difference (LSD) post hoc tests. The values are presented as mean ± SEM. P values of < 0.05 were regarded as statistically significant.
RESULTS
Assessment of Fertility of the IT6-F and IT6-M Females
We earlier reported that the TG females of the IT6-F and IT6-M lines never sired live litters, although we frequently observed vaginal plugs, indicating mating [19]. Now, as was judged by vaginal smears, the IT6-F females showed normal estrus cycle, but the IT6-M TG females had persistent estrus (Table 1). The IT6-M females did not respond to PMSG + hCG-induced superovulation, although vaginal plugs were present in most (7 out of 9) of them, whereas the IT6-F females responded normally to the ovulation induction (Table 1). The fertilized ova from oviducts were surgically removed from live animals, and those derived from IT6-F TG females developed normally to blastocysts in vitro (10/10 vs. non-TG embryos 10/10). They also developed to term (9 pups out of 24 embryos transferred) after embryo transfer into oviducts of pseudopregnant non-TG foster mothers.
|
Tumor Ontogeny and Progression/Ovarian Weights
The tumor progression was more aggressive in IT6-F TG females compared with IT6-M females, and as was suggested by increasing ovarian weights, the beginning of tumorigenesis was already seen between 3–4 mo (IT6-F) and 4–5 mo (IT6-M; Fig. 1). The IT6-M ovarian tumor weights were about half of the IT6-F TG female ovarian tumor weights at the age of 6 mo (Fig. 1). In the IT6-F group, the ovarian weight increase was statistically significant at the age of 5 mo, in the IT6-M mice it was 1 mo later. We found macroscopically visible changes (metastases) in only 6-mo-old IT6-F mouse livers (2/8 mice) or in both lungs and liver (1/8 mice), confirmed later by histopathological analysis (data not shown). The ovarian tumor formation was bilateral in 37.5% of IT6-F and 50% of IT6-M mice, at the age of 6 mo.
|
Serum Hormones
Progesterone levels were significantly elevated in the IT6-F mice between 4–5 mo of age (Fig. 2A), and in IT6-M mice, the elevation of serum progesterone occurred 1 mo later (Fig. 2B). The elevated levels of serum progesterone in both lines closely paralleled the age-related increases of ovarian weights (Fig. 1).
|
A reciprocal suppression occurred in serum LH and FSH levels, as compared with those of serum progesterone or ovarian weights. In the IT6-F females, serum LH levels decreased significantly between the 4th and 5th mo of age (as compared with C57BL controls; Fig. 3A). In the IT6-M animals, the same decrease occurred again 1 mo later (Fig. 3B). Almost the same phenomenon was seen in serum FSH levels, which was only less drastic than with LH (Fig. 4, A and B).
|
|
Serum E2 levels and those of in the IT6-F and IT6-M TG females showed an age-dependent increase, but there were no significant differences between them and the respective control mice (data not shown).
In Vitro Production of Progesterone by the Granulosa Tumor Cell Lines
Basal progesterone production of KK-1 cell lines (passage 7) and NT-1 cells from passage 6 onward, were high (4.1–5.1 nmol/L/105 cells x 8 h, respectively; Fig. 2C). This progesterone production was stimulated by hCG (P < 0.01) and forskolin (P < 0.001). Progesterone production of the NT-1 cells was slightly higher that of KK-1 cells. FSH did not significantly stimulate progesterone production of either of cell line (data not shown).
Uterine Weights
No statistically significant differences were observed between uterine weights of the IT6-F and C57BL, or IT6-M and DBA females (data not shown). This finding is in keeping with the serum E2 data.
Adrenal Weights
No statistically significant differences were observed between the adrenal weights of IT6-F and C57BL or IT6-M and DBA females (data not shown).
Ovarian Histopathology
Histopathological analysis of the ovarian tumors of both lines showed individual variation, as is typical for SV40 Tag-induced tumors [29, 30]. The histopathological picture of inh/Tag ovarian tumors was typical for granulosa cell tumors, with a mostly mixed pattern of microfollicular, traebecular, and diffuse areas; Call-Exner bodies were present in all tumors [19]. The ovarian tumors from IT6-F females showed a greater degree of transformation, consisting of tumor cells completely invading the ovarian parenchyme. Folliculogenesis was ongoing, showing multiple corpora lutea and developing follicles, until advanced tumorigenesis even at the age of 6 mo (Fig. 5). On the other hand, in IT6-M females, the ovaries occasionally contained multiple corpora lutea invading almost the entire ovarian structure, which is typical of a pseudopregnant ovary [31]. The normal follicles were presumably compressed to the outside of the ovary and appeared atretic (Fig. 5). No signs of ongoing folliculogenesis were observed in these IT6-M female ovaries, although primordial follicles were present even at the age of 6 mo. Call-Exner body, a characteristic rosette-like structure common to granulosa cell tumors, could be found earliest in 5-mo-old IT6-F and 6-mo-old IT6-M females (Fig. 6). Some multiple atypical mitotic granulosa cells appeared to be more luteinized, being large with eosinophylic or vacuolized cytoplasm, and prominent cell borders were already seen at the age of 4 mo in IT6-F and at 5 mo in IT6-M female ovaries (Fig. 6). Mitotic activity was pronounced in these cases, ranging from 6 per 10 high-power (HP) fields to 16 per 10 HP fields, compared with 2–4 per 10 HP fields in the control ovaries (Fig. 6).
|
|
DISCUSSION
In the present study, for the first time we monitored the serum levels of gonadotropins, progesterone, and E2 during postnatal life of the inh/Tag TG mice, thus covering the time before and after appearance of tumors. The serum progesterone levels were inversely correlated with those of LH and FSH, and directly with ovarian weight gain prior to appearance of tumors in an age-dependent manner. In addition to inhibin and FSH, the appearance of tumors in the TG mice can thus also be monitored by measurements of serum progesterone and LH. These findings support the contention that granulosa cell tumorigenesis is associated with elevated levels of inhibin and reduced levels of serum FSH gonadotropins [7, 10, 32].
The role of ovarian inhibins in the regulation of LH secretion has remained obscure. Some studies have reported no changes in LH gene expression, secretion, or pituitary content after inhibin immunoneutralization or injection of rh-inhibin A [33–36], whereas others have detected increases in all parameters of pulsatile LH secretion after inhibin immunoneutralization [37]. In connection with testicular Sertoli cell tumors, in response to the high levels of serum inhibin, both LH and FSH levels decreased [32, 38].
In our study, the high levels of serum progesterone, along with tumorigenesis, likely reflect the increased number of progesterone-secreting cells in the tumors. The present in vitro studies, in agreement with our earlier reports [19, 28], show that the granulosa tumor cells produced high amounts (4–5 nmol/105 cells x 8 h) of progesterone, which reached 20-fold higher levels than reported on forskolin-stimulated primary cultures of rat luteinizing granulosa cells (0.2 nmol/105 cells x 48 h) [39]. This high progesterone production may be the main reason of the inversely proportional and decreased serum LH concentrations in the tumor mice, although the role of inhibin cannot be ruled out. These granulosa tumor cells showed a low rate of E2 formation from exogenous androstendione substrate in response to recombinant FSH [19]. Likewise, the serum E2 levels were not increased. The lack of E2 response of the apparent granulosa cell tumors can be explained by the two-cell hypothesis of E2 production [40], which states that granulosa cells can produce E2 only if the precursor, testosterone, is presented to them by adjacent theca cells. Many granulosa cell tumors contain few theca cells, and extraovarian tumors are generally devoid of them [40]. The other possibility is that the tumors represent granulosa cells about to become luteinized, and are therefore low in aromatase.
The successful ovulation induction, embryo transfer, and development of them to term may indicate impaired embryonic development in uteri of IT6-F females. Expression of the inhibin gene has been reported in the mouse placenta [41–43]. Concomitant expression of the SV40 T-ag transgene could impair placental function in the IT6-F TG females, although we need to show in the future that the transgene is not expressed in the placenta of the litters sired by transgenic males before we can go further with this hypothesis. Anyhow, the possibly impaired placental function in the IT6-F TG females remains an open question. Our data show that IT6-F females are susceptible to ovulation induction and are not sterile, but infertile. This information opens up the possibility that IT6-F females could be mated (with embryo transfer) with other TG mice, when females are sterile. The persistent estrous cycle and the failure of ovulation induction in the IT6-M females may be due to the different copy numbers of the transgene [19, 20]. We found neither macroscopically nor histologically any PCOS (polycystic ovarian syndrome)-like changes in the IT6-M mice ovaries, although they occasionally contained multiple corpora lutea consisting of almost entire ovarian structure, which is typical of the pseudopregnancy [31]. Estrous cycle disorders and failure of ovulation induction attempts in the IT6-M females have to be studied further. The IT6-M line can be helpful in studying chronic anovulation-related ovarian granulosa cell tumorigenesis and ovulation-related gene expressions.
There are some reports on spontaneous occurrence of very rare juvenile-type granulosa cell tumors in SWR inbred (2%–5%) [10, 44] or in SWRx SJL (SWXJ) F1 (10%–25%) [45] recombinant inbred strains of mice, appearing at puberty between 3 and 6 wk of age. Mice bearing these granulosa cell tumors showed high serum inhibin, low serum FSH, unaltered serum E2 and LH (in SWR moderately suppressed), and reduced levels of progesterone. Metastases occurred in SWR mice 5–9 mo after onset of tumorigenesis, whereas in IT6-F females, after 1–2 mo. In SWR mice, tumor growth was demonstrated to be under steroid control. Treatment of SWXJ-9 recombinant inbred mice with the steroid hormone precursor dihydroepiandrosterone (DHEA) resulted in a 2- to 3-fold increase in tumor incidence, whereas E2 inhibited tumor onset [44]. In contrast, in IT6 females, the tumor growth is gonadotropin-dependent [21] and the granulosa tumor cells express both gonadotropin receptors [19]. There are conflicting reports as to the presence or absence of gonadotropin receptors in human ovarian cancer specimens [46–49]. The controversy exists as to whether ovarian epithelial neoplasms are sensitive to gonadotropins or not, although recent studies have demonstrated that gonadotropin receptors are expressed even in ovarian surface epithelium of ovarian epithelial carcinoma and gonadotropins stimulating cell proliferation of ovarian cancer cells (for a review, [50]). In this context, our IT6 females may be useful for further elucidating the specific roles of gonadotropins and their receptors in ovarian carcinogenesis.
There are numerous models for genetically targeted tumorigenesis in TG mice [51], but only few of them have successfully developed gonadal tumors (for a review, [18]). These studies have mainly concentrated on description of the tumors and tumor formation, as well as on establishment of cell lines. Less has been learned about the tumor ontogeny and regulatory aspects of tumorigenesis. The IT6 inh/Tag TG model resembles the human ovarian granulosa cell tumorigenesis (i.e., the development of primary tumors in the ovary occurs in a short period of time with intact oocytes and follicles), ongoing folliculogenesis until advanced tumorigenesis (IT6-F line), depressed serum gonadotropins, elevated inhibin levels, aberrant steroidogenesis, similar histopathologic features, and finally, malignant potential. This combination makes IT6 inh/Tag female mice a good model for endocrine cancer research, including testing of different types of cancer therapy.
In conclusion, the IT6 inh/Tag tumor model offers an opportunity to determine the host endocrine status and thus to gain insight into possible endocrine and paracrine consequences of the ovarian granulosa cell tumorigenesis. The results obtained here provide information about the hypothalamic-pituitary-gonadal relationships in ovarian granulosa cell tumorigenesis and for the possible inductive or supportive role of gonadotropins and about the endocrine effects of the tumor-produced inhibins and steroid hormones. In view of the present data, gonadotropins, and possibly also progesterone, should be tested as potential endocrine tumor markers upon human granulosa cell tumorigenesis, along with the previously established inhibin.
ACKNOWLEDGMENTS
We thank Dr. Hans Bunschoten for constructive remarks, Dr. Talal El-Hefnawy and Tiina Lensu, B.Med. for their help, and Ms. Tarja Laiho and Ms. Johanna Vesa for their technical assistance.
FOOTNOTES
First decision: 15 August 2000.
1 This work was supported by a grant from The Academy of Finland and The Finnish Cancer Society. 
2 Correspondence: Ilpo Huhtaniemi, Department of Physiology, University of Turku, Kiinamyllynkatu 10, FIN-20520 Turku, Finland. FAX: 358 2 2502610; ilpo.huhtaniemi{at}utu.fi
Accepted: November 15, 2000.
Received: July 14, 2000.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  R S Bhaskaran and M Ascoli The post-endocytotic fate of the gonadotropin receptors is an important determinant of the desensitization of gonadotropin responses J. Mol. Endocrinol., April 1, 2005; 34(2): 447 - 457. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C.L. Schuh Trials, Tribulations, and Trends in Tumor Modeling in Mice Toxicol Pathol, January 1, 2004; 32(1_suppl): 53 - 66. [Abstract] [PDF]
|
|
|
|
 |
|
 H. Krishnamurthy, H. Kishi, M. Shi, C. Galet, R. S. Bhaskaran, T. Hirakawa, and M. Ascoli Postendocytotic Trafficking of the Follicle-Stimulating Hormone (FSH)-FSH Receptor Complex Mol. Endocrinol., November 1, 2003; 17(11): 2162 - 2176. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. H. Burns and M. M. Matzuk Minireview: Genetic Models for the Study of Gonadotropin Actions Endocrinology, August 1, 2002; 143(8): 2823 - 2835. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
