High levels of luteinizing hormone analog stimulate gonadal and adrenal tumorigenesis in mice transgenic for the mouse inhibin-α-subunit promoter/Simian virus 40 T-antigen fusion gene

Abstract

Transgenic (TG) mice expressing the Simian virus 40 T-antigen under the control of the murine inhibin-α promoter (Inhα/Tag) develop granulosa and Leydig cell tumors at the age of 5–6 months, with 100% penetrance. When these mice are gonadectomized, they develop adrenocortical tumors. Suppression of gonadotropin secretion inhibits the tumorigenesis in the gonads of intact animals and in the adrenals after gonadectomy. To study further the role of luteinizing hormone (LH) in gonadal and adrenal tumorigenesis, a double TG mouse model was generated by crossing the Inhα/Tag mice with mice producing constitutively elevated levels of LH (bLHβ-CTP mice). Our results show that in double TG mice (bLHβ-CTP/Inhα/Tag), gonadal tumorigenesis starts earlier and progresses faster than in Inhα/Tag mice. Both ovarian and testicular tumors were histologically comparable with the tumors found in Inhα/Tag mice. In addition, adrenal tumorigenesis was found in intact double TG females, but not in Inhα/Tag females. Inhibin-α and LH receptor (LHR) were highly expressed in tumorigenic gonadal tissues, and the elevated LH levels were shown to be associated with ectopic LHR and high inhibin-α expression in the female adrenals. We conclude that in the Inhα/Tag tumor mouse model, elevated LH levels act as a tumor promoter, advancing gonadal and adrenal tumorigenesis.

Introduction

The gonadotropins, luteinizing hormone (LH) and follicle-stimulating hormone (FSH), are the main regulators of gonadal differentiation, growth and steroidogenesis. It is therefore to be expected that they are involved in the appearance and/or maintenance of gonadal tumors. Among human gonadal malignancies, ovarian tumors are the commonest, with the poorest prognosis. In attempts to improve the diagnostics and treatment of ovarian cancers, it is essential to learn more about their pathogenesis, including the regulation of their aggressive growth and invasion, and to develop better predictive markers. It has often been suggested that LH could promote the development of certain types of ovarian and testicular cancer. This is supported by epidemiological findings showing an increased incidence of ovarian cancer after menopause, when circulating gonadotropin levels are elevated (Wise et al., 1996). There are also data showing that ovarian stimulation with gonadotropins is associated with an increased incidence of granulosa cell tumors (Willemsen et al., 1993). In addition, certain activating mutations of the LH receptor (LHR) gene have been shown to cause Leydig cell tumors in humans (Liu et al., 1999), while the results of in vitro studies have demonstrated that LH and FSH can stimulate the growth of human ovarian carcinoma cells (Simon et al., 1983; Wimalasena et al., 1992). In line with these results, a modest clinical response has been observed in the treatment of certain types of ovarian cancer with gonadotropin-releasing hormone (GnRH) analogs (Adelson and Reece, 1993). Studies on experimental animals have also supported the gonadotropin theory. In inhibin-α-deficient mice, gonadotropins are essential for gonadal and adrenal tumorigenesis (Kumar et al., 1996), and chronically elevated circulating levels of LH or human chorionic gonadotropin (hCG) can cause ovarian tumors in certain strains of mice (Risma et al., 1995; Rulli et al., 2002). Furthermore, in male rats, chronic treatment with LH produces Leydig cell hyperplasia (Neumann, 1991).

We have studied gonadal and adrenal tumorigenesis in transgenic (TG) mice expressing the Simian virus 40 T-antigen under the control of the murine inhibin-α-subunit promoter (Inhα/Tag). As shown previously (Kananen et al., 1995,1996a), these mice develop gonadal tumors originating from Leydig and granulosa cells, with 100% penetrance, at the age of 5–6 months. To find out whether the extragonadal tumors frequently found in these mice were primary or because of metastases of gonadal tumors, we gonadectomized neonatal mice (Kananen et al., 1996b). To our surprise, gonadectomy led to the development of adrenal gland tumors, with 100% penetrance. It was first hypothesized that these tumors were formed as a result of post-gonadectomy elimination of inhibin production, since inhibin-α knockout mice have provided evidence that this peptide may be a tumor suppressor (Matzuk et al., 1992). Moreover, inhibin-α knockout mice also develop adrenal tumors if they are gonadectomized, to save them from death caused by gonadal tumors (Matzuk et al., 1994). The finding of high LHR expression in adrenal tumors of Inhα/Tag mice evoked a question about the role of gonadotropins in adrenal tumorigenesis. We later showed that either pharmacologically or genetically induced hypogonadotropic hypogonadism prevented both gonadal and adrenal tumor development in gonad-intact Inhα/Tag mice (Kananen et al., 1997), thus supporting a role for gonadotropins in tumor development. In addition, the effect of testosterone on tumorigenesis was excluded (Rilianawati et al., 2000).

To study further the role of LH in this gonadal tumor model, we produced double TG mice with constitutively elevated LH levels and Inhα/Tag expression by crossbreeding Inhα/Tag mice with bLHβ-CTP mice (Risma et al., 1995). Under the LH α-subunit promoter, bLHβ-CTP mice express a modified form of the bovine LH β-subunit gene, containing a 24-amino-acid C-terminal peptide (CTP) of hCGβ, which prolongs its circulatory half-life. Upon coupling with the endogenous α-subunit it forms dimers with high LH bioactivity. The chronically increased levels of serum bioactive LH in bLHβ-CTP mice induce increased steroid levels and a polycystic appearance in the ovaries, and cause infertility. In some mouse strains also ovarian tumors appear. In males, the only phenotype found is reduced testis size (Risma et al., 1995). The double TG mice obtained allowed us to determine the extent to which elevated LH levels play a specific role in gonadal and adrenal tumorigenesis, and to analyse the influence of elevated LH levels on the expression of the inhibin-α gene.

Results

Gonadal tumorigenesis in double TG bLHβ-CTP/Inhα/Tag mice

To find out whether constitutively elevated LH levels promote gonadal and adrenal tumorigenesis in Inhα/Tag mice, gonadal and adrenal sizes, as well as histology, were compared among wild-type (WT), Inhα/Tag, bLHβ-CTP and double TG bL-CTP/Inhα/Tag mice at the ages of 3 and 5 months. At these ages, no significant differences were seen in the weights of the ovaries of WT and Inhα/Tag mice, while the female bLHβ-CTP mice had multicystic and enlarged ovaries with numerous luteinized follicles (Figure 1a, b), as reported previously (Risma et al., 1995). The weights of the ovaries of double TG and bLHβ-CTP mice showed no significant difference at 3 months, but the former were markedly larger at the age of 5 months (Figure 1a). At the age of 3 months double TG mice presented with highly proliferating tumors with mitotic figures (Figure 1b), akin to those seen in Inhα/Tag mice at an older age and shown to be of granulosa cell origin (Kananen et al., 1995), and they had luteinized follicles, as found in bLHβ-CTP mice (Risma et al., 1995) (Figure 1b). The granulosa cell origin of these tumors was reconfirmed by aromatase-specific immunostaining (Figure 3b). Female bLHβ-CTP mice never form granulosa cell tumors in the C57Bl background (Keri et al., 2000), but the effect of the bLHβ-CTP transgene has not been studied in a DBA/2J background. To exclude the possibility that the bLHβ-CTP transgene induces tumorigenesis in a DBA/2J background, we always used littermates for comparison between the groups. As expected, some of the ovaries of the Inhα/Tag mice also showed evidence of granulosa cell tumorigenesis in its initial stage, which, however, was slight at the age of 3 months. No overgrowth of luteal tissue was found in these ovaries (data not shown).

Figure 1
figure1

(a) Ovarian weights of 3- and 5-month-old control (WT), Inhα/Tag, bLHβ-CTP and bLHβ-CTP/Inhα/Tag mice (***P<0.001 compared with bLHβ-CTP mice, tested after logarithmic transformation of the data, n=5 or 6 mice per group). (b) Histological pictures of ovaries of 3-month-old control, Inhα/Tag, bLHβ-CTP and bLHβ-CTP/Inhα/Tag mice (bar=125 μm). The sections are stained with hematoxylin/eosin. C, cyst; CL, corpus luteum; DF, developing follicle; GCT, granulosa cell tumor; L, luteinized tissue. The bottom right panel is a higher power magnification from the framed area of the adjacent bLHβ-CTP/Inhα/Tag picture. The arrows indicate late mitotic anaphases (bar=31 μm)

Figure 3
figure3

Immunocytochemical staining of aromatase in WT (a) and bLHβ-CTP/Inhα/Tag ovaries (b), and 3β-HSD I in WT (c) and bLHβ-CTP/Inhα/Tag (d) testes (Bar=62.5 μm). Granulosa cells of developing follicle (a) and Leydig cells (c) are intensively stained (brown color) in WT gonads as well as in the corresponding cells of the tumor tissues (b, d)

At the age of 3 months, the weights of the bLHβ-CTP/Inhα/Tag mouse testes were significantly increased as compared with all the other groups (Figure 2a). Histologically, the testes of bLHβ-CTP/Inhα/Tag mice showed massive hemorrhagic and invasive tumors with mitotic figures originating from Leydig cells (Figure 2b), the cellular origin of which has been verified by us before (Kananen et al., 1996a) and was reconfirmed by 3β-HSD I-specific immunostaining (Figure 3d). In contrast, no macroscopic tumors were present at 3 months in the single TG Inhα/Tag mice. However, histological analysis revealed initiation of tumor formation in Inhα/Tag mice (Figure 2b), as shown by focal enlargement of the interstitial Leydig cell mass. The mean size of the bLHβ-CTP mouse testes was slightly, although not significantly, smaller (Figure 2a) than that of the WT mice, as also reported previously (Risma et al., 1995). Slight hyperplasia was evident in the interstitial tissue of these mice.

Figure 2
figure2

(a) Testis weights of 3-month-old control (WT) (n=6), Inhα/Tag (n=6), bLHβ-CTP (n=3) and bLHβ-CTP/Inhα/Tag mice (n=5) (*P<0.05 compared with controls). (b) Histological pictures of testes of 3-month-old control, Inhα/Tag, bLHβ-CTP and bLHβ-CTP/Inhα/Tag mice (bar=125 μm). The sections are stained with hematoxylin/eosin. LC, Leydig cell; LCT, Leydig cell tumor; ST, seminiferous tubule. The bottom right panel is a higher power magnification from the framed area of adjacent bLHβ-CTP/Inhα/Tag picture. The arrows indicate late mitotic anaphases (bar=31 μm)

Adrenal gland tumorigenesis in intact bLHβ-CTP/Inhα/Tag female mice

Morphological and histological analyses were also performed to test whether the elevated LH levels affected adrenal tumorigenesis in Inhα/Tag mice. The data revealed that at the age of 3 months the adrenal weights of the bLHβ-CTP/Inhα/Tag female mice were similar to those of the bLHβ-CTP mice (Figure 4a). However, histological signs of tumorigenesis as well as mitotic figures were found in the bLHβ-CTP/Inhα/Tag mice (Figure 4b), but not in the Inhα/Tag mice, suggesting a role for LH as a promoter of adrenal tumorigenesis in the Inhα/Tag background. As expected, no adrenal tumorigenesis was found in either the bLHβ-CTP or WT (Figure 4b) mice. Furthermore, no signs of adrenal tumorigenesis were found in the bL-CTP/Inhα/Tag male mice (data not shown), which had LH levels much lower than in the corresponding females (see below).

Figure 4
figure4

(a) Adrenal gland weights of 3-month-old control (WT), Inhα/Tag, bLHβ-CTP and bLHβ-CTP/Inhα/Tag female mice (n=5 or 6 mice per group). (b) Histological pictures of adrenal glands of 5-month-old control, Inhα/Tag, bLHβ-CTP and bLHβ-CTP/Inhα/Tag females (bar=125 μm). The sections are stained with hematoxylin/eosin. ZG, zona glomerulosa; ZF, zona fasciculata; ZR, zona reticularis; m, medulla; AGT, adrenal gland tumor. The bottom right panel is a higher power magnification from the bordered area of adjacent bLHβ-CTP/Inhα/Tag picture. The arrows indicate late mitotic anaphases (bar=31 μm)

Serum hormone levels in bLHβ-CTP/Inhα/Tag mice

According to previous reports (Risma et al., 1995), circulating LH levels in bLHβ-CTP female mice increased by about 10-fold compared with WT females. The present results indicate that male bLHβ-CTP mice also have elevated LH levels, at the age of 9–12 weeks, compared with WT males (5.18±0.45 (n=11) vs 1.35±0.34 μg/ml (n=16); P<0.0001, respectively), although the increase is less marked than in females. Similar results were found when LH bioactivity was analysed. The mean level in mice overexpressing LH (bLHβ-CTP and bLHβ-CTP/Inhα/Tag) was 4.1±1.0 IU/l in females, and 1.3±0.5 IU/l in males. In the WT and Inhα/Tag mice the bioactivity levels in both sexes were below the limit of detection of the bioassay (<0.5 IU/l).

In the Inhα/Tag mice, serum LH levels were comparable to those in WT mice before the appearance of macroscopically visible tumors. After the progression of Leydig and granulosa cell tumors, serum LH levels decrease (Rahman and Huhtaniemi, 2001), because of increased steroidogenesis in the tumorous gonads. In double TG mice, where the phenotype is a combination of high serum LH levels and steroid-producing tumors, both the serum steroid (Figure 5) and LH levels are high compared with those in Inhα/Tag mice having no tumors at this age.

Figure 5
figure5

Serum concentrations of FSH (a), inhibin (b), progesterone (c) and testosterone (d) in 3-month-old control, Inhα/Tag, bLHβ-CTP and bLHβ-CTP/Inhα/Tag male and female mice (*P<0.05; **P<0.01 vs controls; n=5 or 6, apart from three mice in the male bLHβ-CTP group)

Serum inhibin is considered to be a marker of gonadal tumorigenesis and, therefore, it was also assayed in the present TG gonadal tumor models. Even though the granulosa and Sertoli cells are the main inhibin producers, increased inhibin levels have been measured with other types of gonadal cancers as well (Healy et al., 1993, Toppari et al., 1998, Herath et al., 2001). The large tumor masses in the double TG mice correlated with highly elevated inhibin levels in both male (inhibin B) and female (inhibin A) bLHβ-CTP/Inhα/Tag mice, and with reciprocally suppressed FSH levels (Figure 5). Similar increases in inhibin levels were not detected in the other mouse models analysed (Figure 5), indicating that the granulosa and Leydig cell tumors were the sources of high amounts of this peptide. In bLHβ-CTP males, inhibin B levels were below the limit of detection of the assay (15 ng/l).

The progression of gonadal tumorigenesis is associated with elevated LHR and inhibin-α gene expression

Northern blot hybridization analyses were carried out to study LHR and inhibin-α gene expression in the gonadal tumor models used. In females, strongest LHR hybridization signal was found in the bLHβ-CTP mice. Expression of LHR also appeared to be increased in tumorigenic ovaries of the double TG mice (Figure 6), which is in keeping with our earlier observations on the luteinizing stage of tumorous granulosa cells (Rahman and Huhtaniemi, 2001). However, the LHR mRNA level in the double TG mice was lower than in the bLHβ-CTP mice, whose ovarian tissue consisted almost entirely of luteinized follicles with a high LHR expression. Levels of inhibin-α mRNA in females were highest in the bLHβ-CTP/Inhα/Tag mice, but the bLHβ-CTP females also presented with elevated inhibin-α expression (Figure 6). These findings corroborate the elevated serum inhibin A levels (Figure 5) found in these groups.

Figure 6
figure6

Northern blot hybridization analysis of LHR and inhibin-α gene expression in gonadal tissues of 3-month-old control, bLHβ-CTP, Inhα/Tag and bLHβ-CTP/Inhα/Tag mice. The expected sizes of the hybridizing mRNAs, 6.8, 4.2, 2.7 and 1.8 kb (LHR) and 1.5 kb (inh-α) are indicated on the right. Ribosomal RNAs are indicated on the left. The bottom panel shows ethidium bromide staining of 28S ribosomal RNA as a loading control. Some background hybridization to the 18S and 28S rRNAs is visible in the LHR blot

Testicular tissues of all mouse models expressed LHR and inhibin-α mRNA. Most models showed comparable low levels of LHR mRNA in the testes, the exception being the bLHβ-CTP/Inhα/Tag mice, which had far higher levels of LHR expression (Figure 6). The high amount of LHR provided further evidence for the Leydig cell origin of the testicular tumors in double TG males. In addition, the mRNA splice variants (4.2 and 2.7 kb mRNA) present at low abundance in the testes of the other mouse lines were clearly visible in the bLHβ-CTP/Inhα/Tag testes. In line with the high serum inhibin B levels, a highly elevated hybridization signal for the inhibin-α-subunit was found in the testicular tissues of bLHβ-CTP/Inhα/Tag mice, compared with bLHβ-CTP, Inhα/Tag and control animals (Figure 6).

Elevated LH levels in the regulation of inhibin-α expression in mouse adrenal gland

To find out whether inhibin-α and LHR coexpression is also present in adrenal gland hyperplasia and tumorigenesis, the mRNA expression patterns of the two genes were analysed by Northern blot hybridization. In female mouse adrenals, the expression of both LHR and inhibin-α was similarly increased in bLHβ-CTP mice and in bLHβ-CTP/Inhα/Tag double TG mice (Figure 7a), while in male adrenals, no expression of LHR or inhibin-α was found (Figure 7a). The LHR mRNA expression was confirmed by RT–PCR, which yielded the same finding (Figure 7b). To localize LHR mRNA and inhibin-α mRNA in the adrenal glands of bLHβ-CTP females, in situ hybridization was performed. Intense hybridization signals with the LHR and inhibin-α antisense probes were observed radially across the whole cortex, including the zona glomerulosa, zona fasciculata and zona reticularis (Figure 8). Surprisingly, the expression was not even, but patchy in certain areas of the adrenal gland. Both inhibin-α and LHR expression were localized in the same regions (Figure 8), suggesting that LH may upregulate the expression of LHR and inhibin-α in the same cells.

Figure 7
figure7

(a) Northern blot hybridization analysis of LHR and inhibin-α gene expression in adrenal tissues of 5-month-old control, bLHβ-CTP, Inhα/Tag and bLHβ-CTP/Inhα/Tag mice. The expected sizes of the hybridizing mRNAs, 6.8, 4.2, 2.7 and 1.8 kb (LHR) and 1.5 kb (inh-α) are indicated on the right. Ribosomal RNAs are indicated on the left. The lowest panel shows ethidium bromide staining of 28S ribosomal RNA as a loading control. Faint background hybridization to the 18S and 28S rRNAs is visible. (b) Ethidium bromide staining of the agarose gel of RT–PCR analysis of LHR (250 bp) and β-actin (245 bp) mRNA

Figure 8
figure8

Localization of LHR and inhibin-α gene expression by in situ hybridization in adrenal gland sections from 5-month-old bLHβ-CTP female mice. Panels (a) and (c) show sections hybridized with the LHR and inhibin α sense probes, respectively, showing no signal (bar=200 μm), and panels (b) and (d) show higher power magnification from sections hybridized with the respective antisense probes (bar=125 m). C, cortex; m, medulla; ZF, zona fasciculata; ZG, zona glomerulosa

Discussion

We have previously shown that tumorigenesis in Inhα/Tag mice is gonadotropin dependent (Kananen et al., 1995,1996a,1996b,1997), but it has not been clear whether LH or FSH is mainly responsible for the tumorigenesis. In this study, we demonstrated that Inhα/Tag TG mice develop gonadal tumors much faster when they have persistently elevated serum LH levels, in the face of suppressed FSH secretion, because of expression of the bLHβ-CTP transgene. Hence, the data on the double TG bLHβ-CTP/Inhα/Tag mice emphasize either the role of LH as a main tumor promoter in the Inhα/Tag mice or the importance of altered gonadotropin ratios in gonadal tumorigenesis. These theories are supported by studies on inhibin-α knockout mice, where it has been shown that gonadal tumorigenesis is prevented in inhibin/GnRH-deficient mouse model (Kumar et al., 1996), while inhibin/FSH-deficient mice do develop gonadal tumors, although the tumor progression is delayed compared with mice deficient in inhibin alone (Kumar et al., 1999). In contrast to previous data (Risma et al., 1995) indicating that there is no difference in circulating LH levels between the TG and WT males, we show in the present report that LH levels are elevated in male bLHβ-CTP mice, at least at the age of 9–12 weeks, but the elevation is less marked than in females and could be explained by the different age groups analysed. There are also signs of LH stimulation in the bLHβ-CTP males, such as Leydig cell hyperplasia and a tendency towards elevated serum and testicular testosterone levels, compared with the controls in the present study. As shown previously (Gaytan et al., 1994), elevated LH concentrations with concomitantly low FSH levels before puberty in rodents lead to early differentiation of testicular cells and to reduced testis size in the adult, a feature also present in the bLHβ-CTP males.

Female bLHβ-CTP mice have previously been shown to have hypertrophic adrenal glands, to produce high amounts of corticosterone and to express LHR (Kero et al., 2000). In this study, we found adrenal tumorigenesis in the intact double TG bLHβ-CTP/Inhα/Tag female mice, while in single TG Inhα/Tag mice, adrenal tumors develop only after gonadectomy (Kananen et al., 1996b). Owing to the rapid growth of gonadal tumors, in the present study we could not follow the development of adrenal tumors beyond the age of 4–5 months. Up to this age, we did not find a significant increase in adrenal weights of the bLHβ-CTP/Inhα/Tag females, but histological analyses clearly demonstrated tumorigenesis in the adrenal cortex.

The presence of high LH levels in the double TG bLHβ-CTP/Inhα/Tag mice also enhanced the aggressiveness of ovarian tumors, with occasional metastases being sent to the lungs and liver (data not shown). In TG males, with lower LH levels, metastases were not found. Adrenal tumors were found in all the double TG females, even without macroscopic signs of metastases in other tissues, indicating that the tumors were primary and not of metastatic origin. The same conclusion was made earlier as regards adrenal tumors developing in gonadectomized inhα/Tag mice (Kananen et al., 1996b).

The mechanism behind the progression of tumor growth in bLHβ-CTP/Inhα/Tag mice is probably mitogenic stimulation by LH on target cells. Chronic activation of LHR function could also act in a proto-oncogenic fashion, as shown with some other G-protein-coupled receptors (Allen et al., 1991; Parma et al., 1993). Activation of LHR typically leads to stimulation of cAMP production via Gs-protein and adenylyl cyclase. However, high concentrations of LH have also been shown to lead to stimulation of the phospholipase C pathway, probably via Gi2-protein (Kuhn and Gudermann, 1999). Although the physiological significance of this pathway remains unclear, it could be an important pathway leading to LH/hCG-dependent growth. Interestingly, activating mutations of Gi2 have been found in a subset of ovarian sex cord stromal tumors and adrenal cortical tumors (Lyons et al., 1990). In addition, an activating somatic mutation of the LHR gene, with preferential stimulation of inositol trisphosphate production, has been shown to be associated with human Leydig cell tumors (Liu et al., 1999).

We have shown earlier in studies carried out in vitro that ovarian, testicular and adrenal tumor cell lines originating from Inhα/Tag mice display a dose-dependent response of cAMP production to hCG stimulation (Kananen et al., 1995, 1996; Rilianawati et al., 1998). In Inhα/Tag mice, LH stimulation is likely to increase the cAMP levels of target cells and also to stimulate the inhibin-α promoter, leading to increased T-antigen expression. It has been shown that an important component of gonadotropin-regulated, cAMP-dependent, inhibin-α-subunit gene expression is the synthesis or activation of cAMP response element (CRE) binding protein (CREB) (Pei et al., 1991). Both gonadotropins can induce the phosphorylation of CREB, leading to its activation (Mukherjee et al., 1996). The extent to which the enhanced tumorigenesis can be explained by increased expression of T-antigen, thus causing faster progression of the tumors in bLHβ-CTP/Inhα/Tag mice, is not currently known.

According to the present results it is evident that clear LHR expression occurs in the granulosa tumor cells of Inhα/Tag mice. This is in contrast to bLHβ-CTP mice of the CF-1 strain developing granulosa cell tumors, where the expression of LHR is low vs controls and other bLHβ-CTP strains developing luteomas (Owens et al., 2002). In addition to different causes of tumorigenesis in these two granulosa cell tumor models (Inhα/Tag and bLHβ-CTP), the level of differentiation of the tumorous granulosa cells is probably different. The low level of LHR in the granulosa cell tumors of bLHβ-CTP mice suggests a low degree of differentiation, whereas in Inhα/Tag mice the tumorous granulosa cells approach the stage of luteinization, as has been shown in vitro with immortalized cells derived from these tumors (Kananen et al., 1995). High serum LH levels, instead of downregulating LHR expression, upregulate it in both bLHβ-CTP and bLHβ-CTP/Inhα/Tag mice, leading to high expression of LHR in the gonadal and adrenal tumors of bLHβ-CTP/Inhα/Tag mice. In the normal adult rodent testis and ovary, a high level of stimulation of LH leads to rapid downregulation of the receptor (Dufau, 1998). Hence, the normal downregulation of LHR by LH is altered in bLHβ-CTP mice. This is a functional feature that is typical of LHR expression in fetal Leydig cells (Huhtaniemi et al., 1981, Huhtaniemi, 1994). There may be differences in receptor regulation or differences in the regulation of gene transcription between normal healthy and tumorigenic gonads, and the phenomenon needs to be studied further.

The high expression of inhibin-α in the gonadal tumors indicates that it could be used as a marker of proliferation. The same finding was made with mouse 1.2 cDNA arrays (CLONTECH Laboratories, Inc., East Meadow Circle, Palo Alto, CA), with no other significant changes (data not shown). Elevated inhibin expression has been found in certain types of human gonadal cancer. As granulosa cells synthesize inhibin, elevated serum inhibin levels were first detected in patients with granulosa cell tumors (Lappohn et al., 1989). In these patients, all forms of inhibin have been shown to be elevated (Healy et al., 1993). The α-subunit, rather than dimeric forms, seems to be increased in epithelial ovarian carcinomas (Lambert-Messerlian et al., 1997; Ala-Fossi et al., 2000). In human testicular tumors, inhibin β-subunit has been shown to be expressed in germ cell tumors (De Jong et al., 1990). Although the major circulating dimeric form in males in inhibin B (Illingworth et al., 1996), also the role of inhibin A as a tumor marker in males, is supported by studies showing that Leydig cell and testicular sex cord-stromal tumors produce inhibin A (Herath et al., 2001, Iczkowski et al., 1998). There is a discrepancy between human sex cord stromal tumors secreting large amounts of inhibins and the inhibin-α knockout mouse model, alluding to a tumor suppressor role of inhibin. Inhibin resistance of human tumors has been offered as an explanation, caused, for example, by mutations in the inhibin signaling pathway or in the inhibin receptor (Matzuk et al., 1996). It is also possible that the control of inhibin secretion from normal and tumorous granulosa cells is different (Gocze et al., 1997). In conclusion, the current study provides further evidence for a tumor promoter effect of prolonged high levels of LH in a TG mouse model expressing the viral oncogene Tag in gonadal somatic cells and the adrenal cortex. The relevance of these observations in regard to human conditions with prolonged high level exposure to gonadotropins, that is, after menopause, in hypergonadotropic hypogonadism, during gonadotropin treatment and in polycystic ovarian syndrome, remains an interesting challenge for further investigations.

Materials and methods

Generation and identification of TG mice

Four different TG mouse lines were used in the present study: (1) mice TG for the Simian virus 40 T-antigen expressed under control of the mouse inhibin-α-subunit promoter (Inhα/Tag mice) (Kananen et al., 1995); (2) mice TG for the bovine LH β-subunit gene fused with the 24-amino-acid CTP extension of the human chorionic gonadotropin β-subunit, expressed under control of the bovine LH α-subunit promoter (bLHβ-CTP) (Risma et al., 1995); (3) double TG bLHβ-CTP/Inhα/Tag mice and (4) age-matched WT littermates as controls. The genetic background of the bLHβ-CTP mice was C57Bl, and that of the Inhα/Tag mice DBA/2J. Since the female mice from both bLHβ-CTP and Inhα/Tag lines were infertile after normal mating, the following procedure was used to generate double TG bLHβ-CTP/Inhα/Tag mice: Superovulation in 4–6-week-old bLHβ-CTP females was induced by hormonal treatment with PMSG and hCG, as previously described (Hogan et al., 1994). The hormonally treated females were bred with Inhα/Tag males, and oocytes were collected from the oviducts 1 day after mating. Fertilized oocytes were then transferred to pseudopregnant foster mothers using standard methods. Genotyping of the bLHβ-CTP and Inhα/Tag transgenes was performed on tail DNA of the offspring, using PCR methods described previously (Kananen et al., 1995; Risma et al., 1995). The double TG bLHβ-CTP/Inhα/Tag males were fertile, and were bred with WT C57Bl/6J females. The experiments were carried out using F1 and F2 generations of the mouse lines generated from these crossings. In all, 3–6 mice per group were used in each experiment, with age-matched WT littermates as controls.

After weaning at the age of 21 days, the mice were housed 4–6 per cage in a room with controlled light (12 h light, 12 h darkness) and temperature (21±1°C). For sample collection, the 3- and 5-month-old mice were anesthetized in the morning (07.30–08.30 h) with Avertin (tribromoethanol) (Hogan et al., 1994), blood was collected by cardiac puncture and the mice were then sacrificed by cervical dislocation. All mice were handled in accordance with the institutional animal care policies of the University of Turku (Turku, Finland).

mRNA analysis

Snap-frozen tissues were kept at −70°C until RNA was isolated, using a single-step acid guanidinium thiocyanate-phenol chloroform extraction method, as previously described (Chomczynski and Sacchi, 1987). A measure of 20 μg of total RNA was resolved on 1% denaturing agarose gel and transferred onto nylon membranes (Hybond-N, Amersham, Buckinghamshire, England). The membranes were hybridized with [32P]cRNA, corresponding to nucleotides 441–881 of rat LHR cDNA, using standard techniques. Hybridization results were visualized by autoradiography on X-ray film (XAR 5, Eastman Kodak, Rochester, NY, USA). For detecting inhibin-α-subunit mRNAs, specific antisense cRNA probes were generated using cDNA templates for in vitro transcription. The template for the inhibin-α probe contained the corresponding full-length rat cDNA.

In order to amplify small amounts of RNA, the DNA Engine Opticon™ system (MJ Research™, Inc., Waltham, MA, USA) and QuantiTect SYBR Green RT–PCR Kit (Qiagen, Valencia, CA, USA) were used to confirm the LHR expression in adrenals. In all, 50 ng DNase-treated (Gibco, Paisley, Scotland) RNA were transcribed and amplified according to manufacturers' instructions. The primers used for LHR were (5′-IndexTermTTGCCGAAGAAAGAACAGAAT-3′, sense) and (5′-IndexTermAGCCAAATCAACACCCTAAG-3′, antisense), amplifying a 250 bp fragment corresponding to nucleotides 870–1120 (Genbank accession number NM_013582) of the LHR cDNA. The same primer pair for β-actin was used as described earlier (Mikola et al., 2001). PCR products were resolved on 1% agarose gel.

Histological analysis, in situ hybridization and immunocytochemistry

Tissues for histological and in situ hybridization analyses were fixed in freshly prepared 4% paraformaldehyde, and embedded in paraffin. For histological analysis, 5-μm paraffin sections were stained with hematoxylin/eosin.In situ hybridization analyses for LHR and inhibin-α were carried out on paraffin sections using a method described previously (Kero et al., 2000) and using the same probes as in Northern blot hybridization (see above). Sections from WT and bLHβ-CTP/Inhα/Tag ovaries and testes were used for immunocytochemical staining with a mouse anti-aromatase antibody (1 : 50 in PBS) (kindly donated by Dr Philippa Saunders, MRC Reproductive Biology Unit, Edinburgh, UK) or with mouse anti-3β-HSD I antibody (1 : 2000 in PBS) (kindly donated by Dr Anita H Payne, Stanford University, CA, USA), respectively. The antigen–antibody complexes were visualized with the immunoperoxidase technique (Vectastain Elite ABC Kit, Vector, Burlingame, CA, USA). The ovarian sections were counterstained with Mayer's hematoxylin.

Hormone measurements

The bioactivity of serum LH was determined by using the mouse interstitial cell in vitro bioassay (Ding and Huhtaniemi, 1989). Testosterone production, determined by RIA, was used as an index of the response to LH. Human LH (code 80/552, WHO International Laboratory for Biological Standards, N.I.B.S.C., Hertfordshire, England) was used as standard. Serum LH from 9–12-week-old bLHβ-CTP males was measured by RIA as previously described (Niswender et al., 1968). Serum FSH concentrations were measured by immunofluorometric assay as described earlier (van Casteren et al., 2000). Progesterone and testosterone concentrations were measured by RIA from diethyl ether extracts as described earlier (Huhtaniemi et al., 1985; Vuorento et al., 1989). Concentrations of inhibins A and B were measured by dimeric inhibins A and B immunoassays according to the instructions of the manufacturer (Oxford Bio-Innovation, Serotec, Oxford, UK).

Statistical analysis

Statistical analyses were performed by using a SigmaStat program (version 2.0 for Windows 95, SPSS Inc., Chicago, IL, USA). Kruskal–Wallis one-way analysis, or one-way ANOVA was performed for analyzing statistical significance (P<0.05). For pairwise multiple comparisons, the Student–Newman–Keuls test was used. Values are presented as mean±s.e.m.

References

  1. Adelson MD and Reece MT . (1993). Clin. Obstet. Gynecol., 36, 690–700.

  2. Ala-Fossi SL, Aine R, Punnonen R and Mäenpää J . (2000). Eur. J. Gynaecol. Oncol., 21, 187–189.

  3. Allen LF, Lefkowitz RJ, Caron MG and Cotecchia S . (1991). Proc. Natl. Acad. Sci. USA, 88, 11354–11358.

  4. Chomczynski P and Sacchi N . (1987). Anal. Biochem., 162, 156–159.

  5. de Jong FH, Grootenhuis AJ, Steenbergen J, van Sluijs FJ, Foekens JA, ten Kate FJ, Oosterhuis JW, Lamberts SW and Klijn JG . (1990). J. Steroid Biochem. Mol. Biol., 37, 863–866.

  6. Ding YQ and Huhtaniemi I . (1989). Acta Endocrinol. (Copenh) 121, 46–54.

  7. Dufau ML . (1998). Annu. Rev. Physiol., 60, 461–496.

  8. Gaytan F, Pinilla L, Romero JL and Aguilar E . (1994). J. Endocrinol., 142, 527–534.

  9. Gocze PM, Beamer WG, de Jong FH and Freeman DA . (1997). Gynecol. Oncol., 65, 143–148.

  10. Healy DL, Burger HG, Mamers P, Jobling T, Bangah M, Quinn M, Grant P, Day AJ, Rome R and Campbell JJ . (1993). N. Engl. J. Med., 329, 1539–1542.

  11. Herath CB, Watanabe G, Wanzhu J, Noguchi J, Akiyama K, Kuramoto K, Groome NP and Taya K . (2001). J. Androl., 22, 838–846.

  12. Hogan B, Beddington R, Constantini F and Lacy E . (1994). Manipulating the Mouse Embryo. A Laboratory Manual. 2nd edn. Cold Spring Harbor Laboratory Press: New York.

  13. Huhtaniemi I . (1994). Eur. J. Endocrinol., 130, 25–31.

  14. Huhtaniemi IT, Katikineni M and Catt KJ . (1981). Endocrinology, 109, 588–595.

  15. Huhtaniemi I, Nikula H and Rannikko S . (1985). J. Clin. Endocrinol. Metab., 61, 698–704.

  16. Iczkowski KA, Bostwick DG, Roche PC and Cheville JC . (1998). Mod. Pathol., 11, 774–779.

  17. Illingworth PJ, Groome NP, Byrd W, Rainey WE, McNeilly AS, Mather JP and Bremner WJ . (1996). J. Clin. Endocrinol. Metab., 81, 1321–1325.

  18. Kananen K, Markkula M, el-Hefnawy T, Zhang FP, Paukku T, Su JG, Hsueh AJ and Huhtaniemi I . (1996a). Mol. Cell. Endocrinol., 119, 135–146.

  19. Kananen K, Markkula M, Mikola M, Rainio EM, McNeilly A and Huhtaniemi I . (1996b). Mol. Endocrinol., 10, 1667–1677.

  20. Kananen K, Markkula M, Rainio E, Su JG, Hsueh AJ and Huhtaniemi IT . (1995). Mol. Endocrinol., 9, 616–627.

  21. Kananen K, Rilianawati, Paukku T, Markkula M, Rainio EM and Huhtanemi I . (1997). Endocrinology, 138, 3521–3531.

  22. Keri RA, Lozada KL, Abdul-Karim FW, Nadeau JH and Nilson JH . (2000). Proc. Natl. Acad. Sci. USA, 97, 383–387.

  23. Kero J, Poutanen M, Zhang FP, Rahman N, McNicol AM, Nilson JH, Keri RA and Huhtaniemi IT . (2000). J. Clin. Invest., 105, 633–641.

  24. Kuhn B and Gudermann T . (1999). Biochemistry, 38, 12490–12498.

  25. Kumar TR, Palapattu G, Wang P, Woodruff TK, Boime I, Byrne MC and Matzuk MM . (1999). Mol. Endocrinol., 13, 851–865.

  26. Kumar TR, Wang Y and Matzuk MM . (1996). Endocrinology, 137, 4210–4216.

  27. Lambert-Messerlian GM, Steinhoff M, Zheng W, Canick JA, Gajewski WH, Seifer DB and Schneyer AL . (1997). Gynecol. Oncol., 65, 512–516.

  28. Lappohn RE, Burger HG, Bouma J, Bangah M, Krans M and de Bruijn HW . (1989). N. Engl. J. Med., 321, 790–793.

  29. Liu G, Duranteau L, Carel JC, Monroe J, Doyle DA and Shenker A . (1999). N. Engl. J. Med., 341, 1731–1736.

  30. Lyons J, Landis CA, Harsh G, Vallar L, Grunewald K, Feichtinger H, Duh QY, Clark OH, Kawasaki E and Bourne HR et al. (1990). Science, 249, 655–659.

  31. Matzuk MM, Finegold MJ, Su J-G, Hsueh AJW and Bradley A . (1992). Nature, 360, 313–319.

  32. Matzuk MM, Finegold MJ, Mather JP, Krummen L, Lu H and Bradley A . (1994). Proc. Natl. Acad. Sci. USA, 91, 8817–8821.

  33. Matzuk MM, Kumar TR, Shou W, Coerver KA, Lau AL, Behringer RR and Finegold MJ . (1996). Recent Prog. Horm. Res., 51, 123–157.

  34. Mikola MK, Rahman NA, Paukku TH, Ahtiainen PM, Vaskivuo TE, Tapanainen JS, Poutanen M and Huhtaniemi IT . (2001). J. Endocrinol., 170, 79–90.

  35. Mukherjee A, Park-Sarge OK and Mayo K . (1996). Endocrinology, 137, 3234–3245.

  36. Neumann F . (1991). Mutat. Res., 248, 341–356.

  37. Niswender GD, Midgley Jr AR, Monroe SE and Reichert Jr LE . (1968). Proc. Soc. Exp. Biol. Med., 128, 807–811.

  38. Owens GE, Ken RA and Nilson JH . (2002). Mol. Endocrinol., 16, 1230–1242.

  39. Parma J, Duprez L, Van Sande J, Cochaux P, Gervy C, Mockel J, Dumont J and Vassart G . (1993). Nature, 365, 649–651.

  40. Pei L, Dodson R, Scoderbek W, Maurer R and Mayo K . (1991). Mol. Endocrinol., 5, 521–534.

  41. Rahman NA and Huhtaniemi IT . (2001). Biol. Reprod., 64, 1122–1130.

  42. Rilianawati, Paukku T, Kero J, Zhang FP, Rahman N, Kananen K and Huhtaniemi I . (1998). Mol. Endocrinol., 6, 801–809.

  43. Rilianawati, Kero J, Paukku T and Huhtaniemi I . (2000). J. Endocrinol., 166, 77–85.

  44. Risma KA, Clay CM, Nett TM, Wagner T, Yun J and Nilson JH . (1995). Proc. Natl. Acad. Sci. USA, 92, 1322–1326.

  45. Rulli SB, Kuorelahti A, Karaer O, Pelliniemi LJ, Poutanen M and Huhtaniemi I . (2002). Endocrinology, 143, 4085–4095.

  46. Simon WE, Albrecht M, Hänsel M, Dietel M and Hölzel F . (1983). J. Natl. Cancer Inst., 70, 839–845.

  47. Toppari J, Kaipia A, Kaleva M, Laato M, de Kretser DM, Krummen LA, Mather JP and Salmi TT . (1998). APMIS, 106, 101–112 discussion 112–113.

  48. van Casteren JI, Schoonen WG and Kloosterboer HJ . (2000). Biol. Reprod., 62, 886–894.

  49. Vuorento T, Lahti A, Hovatta O and Huhtaniemi I . (1989). Scand. J. Clin. Lab. Invest., 49, 395–401.

  50. Willemsen W, Kruitwagen R, Bastiaans B, Hanselaar T and Rolland R . (1993). Lancet, 341, 986–988.

  51. Wimalasena J, Dostal R and Meehan D . (1992). Gynecol. Oncol., 46, 345–350.

  52. Wise PM, Krajnak KM and Kashon ML . (1996). Science, 273, 67–70.

Download references

Acknowledgements

The skillful technical assistance of N Messner, J Vesa and T Laiho is gratefully acknowledged. This study was supported by grants from The Fund of the Finnish Cancer Societies, the Sigrid Jusélius Foundation, the Emil Aaltonen Foundation, the Ida Montin Foundation and the Academy of Finland.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Ilpo Huhtaniemi.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Mikola, M., Kero, J., Nilson, J. et al. High levels of luteinizing hormone analog stimulate gonadal and adrenal tumorigenesis in mice transgenic for the mouse inhibin-α-subunit promoter/Simian virus 40 T-antigen fusion gene. Oncogene 22, 3269–3278 (2003). https://doi.org/10.1038/sj.onc.1206518

Download citation

Keywords

  • LH
  • adrenal and gonadal tumorigenesis
  • inhibin

Further reading