The effects of sex steroids on thyroid C cells and trabecular bone structure in the rat model of male osteoporosis



Branko Filipović, Department of Cytology, Institute for Biological Research, University of Belgrade, 142 Despot Stefan Boulevard, 11060 Belgrade, Serbia. T: + 381 11 2078322; F: + 381 11 2761433; E:


Androgen deficiency is one of the major factors leading to the development of osteoporosis in men. Since calcitonin (CT) is a potent antiresorptive agent, in the present study we investigated the effects of androgen deficiency and subsequent testosterone and estradiol treatment on CT-producing thyroid C cells, skeletal and hormonal changes in middle-aged orchidectomized (Orx) rats. Fifteen-month-old male Wistar rats were either Orx or sham-operated (SO). One group of Orx rats received 5 mg kg−1 b.w. testosterone propionate (TP) subcutaneously, while another group was injected with 0.06 mg kg−1 b.w. estradiol dipropionate (EDP) once a day for 3 weeks. A peroxidase–antiperoxidase method was applied for localization of CT in the C cells. The studies included ultrastructural microscopic observation of these cells. The metaphyseal region of the proximal tibia was measured histomorphometrically using an imagej public domain image processing program. TP or EDP treatment significantly increased C cell volume (Vc), volume densities (Vv) and serum CT concentration compared with the Orx animals. Administration of both TP and EDP significantly enhanced cancellous bone area (B.Ar), trabecular thickness (Tb.Th) and trabecular number (Tb.N) and reduced trabecular separation (Tb.Sp). Serum osteocalcin (OC) and urinary Ca concentrations were significantly lower after these treatments in comparison with Orx rats. These data suggest that testosterone and estradiol treatment in Orx middle-aged rats affect calcitonin-producing thyroid C cells, which may contribute to the bone protective effects of sex hormones in the rat model of male osteoporosis.


Sex steroids are important for the maintenance of the female and male skeletons. In both sexes, age-related decline in circulating sex hormone levels is associated with modifications of bone remodeling and this is the main cause of osteoporosis in older people. Osteoporosis research is rarely undertaken in men. While there is no dramatic decline in sex hormones in men as occurs in menopausal women, the continuous slow reduction in circulating androgens is associated with bone loss (Swerdloff & Wang, 2002). However, although testosterone is the major male sex hormone, the principal female hormone, estradiol, is also formed in men in significant amounts. While a small fraction of estradiol production is provided by testicular secretion (Nitta et al. 1993), the remaining estradiol is derived from aromatization of circulating androgens by the enzyme aromatase in adipose, skin, muscle, bone and brain tissue (Simpson & Dowsett, 2002).

Another important hormone involved in bone metabolism is calcitonin (CT). This potent hypocalcemic peptide contributes to calcium (Ca) homeostasis by direct inhibition of osteoclast-mediated bone resorption and output of Ca from skeletal tissues (Warshawsky et al. 1980). CT is produced and secreted by thyroid C cells, the function of which is affected by gonadal steroids (Foresta et al. 1985; Greenberg et al. 1986). As with circulating levels of gonadal steroids, the number of C cells and plasma CT decline with age in humans. However, in aged rats, together with lower levels of gonadal steroids, physiological hyperplasia of C cells and hypercalcitoninemia have been noticed (Delverdier et al. 1990; Sekulić et al. 1998; Lu et al. 2000). This hypersecretion of CT in old rats may be partly caused by hyperprolactinemia (Lu et al. 2000). The lack of sex hormones in gonadectomized rats of both sexes reduces the synthesis and release of CT from thyroid C cells (Lu et al. 2000; Sakai et al. 2000). The mechanisms by which sex hormones affect CT production, may include their receptors, as already demonstrated in thyroid C cells (Naveh-Many et al. 1992; Zhai et al. 2003).

In humans, hypogonadism accelerates bone loss and is associated with osteoporosis (Hoff & Gagel, 2005). In male rats, orchidectomy (Orx) leads to androgen deficiency, causing increased bone resorption (Tuukkanen et al. 1994; Filipović et al. 2007). For this reason, as ovariectomized female rats are widely used to mimic postmenopausal osteoporosis, Orx rats represent an adequate animal model for simulation of male osteoporosis (Vanderschueren et al. 1992). Such animal models may be useful in studies examining bone remodeling rates and for assessing the efficiency of bone-sparing treatments.

In this study, we evaluated the effects of chronic testosterone and estradiol treatment on calcitonin-producing thyroid C cells, bone structure and bone function in Orx middle-aged rats. This animal model of male osteoporosis mimics the hormonal changes that are an important contributing factor to overall loss of bone in the aging population.

Materials and methods

Animals and experimental design

The experiment involved 32 male Wistar rats, bred at the Institute for Biological Research ‘Siniša Stanković’ (IBISS), Belgrade, Serbia, and maintained under constant laboratory conditions (22 °C, 12 h light–dark cycle) with free access to food and water. The animals were randomly bilaterally orchidectomized (Orx) or sham-operated (SO) at the age of 15 months under ketamine anesthesia (15 mg kg−1 b.w.). At 2 weeks post-operation, the rats were divided into different groups and treated once a day for 3 weeks. One group (n = 8) of Orx rats received 5 mg kg−1 b.w. testosterone propionate (TP; Fluka Chemie AG, Buchs, Switzerland) subcutaneously (s.c.). Another group (n = 8) of Orx animals was injected s.c. with 0.06 mg kg−1 b.w. estradiol dipropionate (EDP; ICN Galenika Pharmaceuticals, Belgrade, Serbia). The SO group (n = 8) and the third Orx group (n = 8) received equivalent volumes of sterile olive oil and served as controls. All animals were sacrificed 24 h after the last injection. Before killing, 24-h urine samples were collected for Ca assay. Sera were separated from trunk blood after decapitation and stored at −70 °C until analyzed biochemically. The experimental protocols were approved by the Local Animal Care Committee of IBISS in conformity with recommendations provided in the European Convention for the Protection of Vertebrate Animals Used for Experimental and Other Scientific Purposes (ETS no. 123, Appendix A).

Immunohistochemical studies

The thyroid lobes were excised and fixed in Bouin's solution for 48 h. The peroxidase-antiperoxidase (PAP) method used for detecting CT in C cells was conducted on 5-μm-thick longitudinal paraplast sections of the thyroid glands. In this procedure, anti-human CT antiserum (Dakopatts, Copenhagen, Denmark) diluted 1 : 500 served as the primary antibodies.

Stereological analysis of thyroid C cells

Immunocytochemically stained sections of the thyroid glands were used for morphometric examination of specifically labeled C cells. These cells were stereologically analyzed by Weibel's method (Weibel, 1979). The C cell volume (Vc) and volume density (volume of C cells per unit volume of thyroid gland; Vv) were measured under a light microscope using 50 test fields at 1000× magnification with the multipurpose M42 test system.

Transmission electron microscopy studies of thyroid C cells

For electron microscopic observation a thyroid gland lobe was excised, fixed in 4% glutaraldehyde solution in 0.1 m phosphate buffer (PB) (pH 7.4) for 24 h and postfixed in 1% osmium tetroxide in the same buffer. Tissue slices were dehydrated through a graded series of ethanol and embedded in Araldite resin. Ultrathin sections of thyroid gland were stained with uranyl acetate and lead citrate, and examined with a transmission electron microscope (MORGAGNI 268; FEI Company, USA).

Bone histomorphometry of the tibia

Right tibiae were used for bone histomorphometric analysis of the tibial proximal metaphysis. The bones were fixed in Bouin's solution, decalcified with 20% ethylenediaminetetraacetic acid disodium salt, routinely processed, embedded in paraplast and sectioned longitudinally. Sections were stained by the Azan method, as previously described (Filipović et al. 2007).

An imagej public domain image processing program was used to measure bone histomorphometric parameters of the tibial specimens. A standard sampling site was established in the secondary spongiosa of the proximal tibial metaphysis, 1 mm distal to the epiphyseal growth plate. All parameters were expressed as recommended by the American Society for Bone and Mineral Research histomorphometry nomenclature (Parfitt et al. 1987). These data were used to calculate cancellous bone area (B.Ar) and cancellous bone perimeter (B.Pm) (Evans et al. 1994). Trabecular thickness (Tb.Th), trabecular number (Tb.N), and trabecular separation (Tb.Sp) reflect the spatial distribution of trabeculae and are derived from B.Ar and B.Pm (Parfitt et al. 1983; Chappard et al. 1999) as described earlier (Filipović et al. 2007).

Serum and urine biochemical analyses

Blood serum was separated for estimation of CT, osteocalcin (OC), testosterone, estradiol, Ca and phosphorus (P). Urine was collected for determination of Ca concentration. Serum CT levels were measured in an immunochemiluminometric assay (Nichols, USA), using a mouse monoclonal antihuman CT antibody marker with acridium ester. The luminescence was quantified with a semiautomated MLA 1 chemiluminescence analyzer (Ciba-Corning). Serum OC was determined on a Rochle Elecsys 2010 immunoassay analyzer (Roche Diagnostics GmbH, Mannheim, Germany). Serum testosterone and estradiol were measured by competitive ACS 180 Plus (Bayer) and ADVIA immunoassay using direct chemiluminescent technology and polyclonal rabbit anti-testosterone or anti-estradiol antibody, respectively, bound to monoclonal mouse anti-rabbit antibody. Serum Ca and P and urinary Ca were determined on a Hitachi 912 analyzer (Roche Diagnostics GmbH).

Statistical analysis

The data were analyzed using statistica 6.0 software (Statsoft, Tulsa, OK, USA). The Kolmogorov–Smirnov procedure was used to test for deviation from normal distribution, followed by one-way analysis of variance (anova). Duncan's multiple range test was employed for post hoc comparisons between groups. The confidence level of P < 0.05 was considered statistically significant. The data are presented as means ± standard error of the mean (SEM).


Immunohistochemical, morphometric and ultrastructural assessment of thyroid C cells after sex hormone treatment

In the SO group, CT-producing thyroid C cells formed groups, were numerous and developed an intense immunocytochemical reaction for CT (Fig. 1A). At the subcellular level, most granules of these cells had low density contents. The mitochondria, round to oval in shape, were dispersed in the cytoplasm (Fig. 2A). After Orx, C cells were more individual, smaller and of darker intensity for the CT reaction compared with the same cells in SO animals (Fig. 1B). At the ultrastructural level we observed a clear difference in the density of granular content. There were numerous electron opaque granules and fewer cellular organelles were present than in SO rats (Fig. 2B). In animals treated with TP or EDP, C cells were similar to the SO control, with a lighter granular cytoplasm compared with C cells in the Orx group (Fig. 1C,D). After TP treatment, the cytoplasm of these cells contained numerous granules with low density content. The Golgi apparatus was prominent and composed of large profiles of smooth membranes. More mitochondria were dispersed in the cytoplasm than in Orx rats (Fig. 3A). In the Orx rats treated with EDP, secretory granules were of low density and few in number when compared with Orx rats. The cytoplasmic area contained profiles of rough endoplasmic reticulum, aggregated into long lamellar arrays (Fig. 3B).

Figure 1.

Calcitonin-immunostained thyroid C cells in (A) sham-operated (SO), (B) orchidectomized (Orx), (C) orchidectomized rats treated with testosterone propionate (Orx + TP) and (D) orchidectomized rats treated with estradiol dipropionate (Orx + EDP). Note the less voluminous cells with strong immunocytochemical reaction after Orx and voluminous cells with light cytoplasm, grouped in clusters after TP or EDP treatment.

Figure 2.

Electron micrographs showing the ultrastructure of thyroid C cells in (A) sham-operated (SO) and (B) orchidectomized (Orx). N, nucleus; M, mitochondria; sg, secretory granules; L, lysosomes; FC, follicular cells.

Figure 3.

Electron micrographs showing the ultrastructure of thyroid C cells in (A) orchidectomized rats treated with testosterone propionate (Orx + TP) and (B) orchidectomized rats treated with estradiol dipropionate (Orx + EDP).

Previously, we had found a significant decrease in the Vc and Vv in Orx rats compared with the SO controls (Filipović et al. 2007). The treatment of Orx rats with TP or EDP increased Vc by 22 and 16% (P < 0.05), and Vv by 21 and 16% (P < 0.05), respectively, in relation to the Orx animals. No significant differences in the morphometric parameters of C cells were detected after treatment with either sex hormone when compared with SO control rats (Fig. 4A,B).

Figure 4.

(A) Cellular volume (Vc) and (B) volume density (Vv) of calcitonin immunoreactive thyroid C cells in sham-operated rats (SO), orchidectomized rats (Orx), orchidectomized rats treated with testosterone propionate (Orx + TP) or estradiol dipropionate (Orx + EDP). All values are mean ± SEM. ●P < 0.05 vs. SO and *P < 0.05 vs. Orx.

Trabecular structure parameters after sex hormone treatment

Analysis of trabecular structure parameters of the proximal tibia metaphysis has already shown that Orx induced cancellous bone loss and marked decreases of B.Ar, Tb.Th and Tb.N, whereas Tb.Sp was significantly increased in comparison with SO rats (Filipović et al. 2007).

In Orx rats treated with TP, numerous trabecular spicules were observed as in SO rats (Fig. 5C). When compared with the control Orx group we found that B.Ar, Tb.Th and Tb.N had increased by 132, 24 and 26% (P < 0.05), respectively, whereas Tb.Sp had decreased by 23% (P < 0.05) after TP treatment. No significant changes in the bone histomorphometric parameters were detected after treatment with TP when compared with the SO group (Fig. 6A–C). EDP treatment restored the trabecular structure and the morphological appearance of tibiae was similar to that observed in SO rat bone (Fig. 5D). Treatment with EDP enhanced B.Ar, Tb.Th and Tb.N by 175, 31 and 44% (P < 0.05), respectively, and reduced Tb.Sp by 37% (P < 0.05) in comparison with the Orx rats. Also, EDP treatment increased Tb.N by 18% (P < 0.05) and decreased Tb.Sp by 19% (P < 0.05) when compared with the SO animals (Fig. 6A–C).

Figure 5.

Trabecular bone of the proximal tibia in (A) sham-operated, (B) orchidectomized (Orx), (C) orchidectomized rats treated with testosterone propionate (Orx + TP) and (D) orchidectomized rats treated with estradiol dipropionate (Orx + EDP). Note reduced mass of blue-stained trabecular bone in Orx rats. TP and EDP treatment prevented loss of trabecular bone; 5-μm sections from the center of the specimen; Azan method stain.

Figure 6.

Histomorphometric analysis of trabecular bone in proximal tibia metaphysis of sham-operated rats (SO), orchidectomized rats (Orx), orchidectomized rats treated with testosterone propionate (Orx + TP) or estradiol dipropionate (Orx + EDP). Graphs provide data for (A) cancellous bone area (B.Ar), (B) trabecular thickness (Tb.Th), (C) trabecular number (Tb.N) and (D) trabecular separation (Tb.Sp). All values are mean ± SEM. ●P < 0.05 vs. SO and *P < 0.05 vs. Orx.

Serum and urine parameters after sex hormone treatment

We have shown previously that Orx induced a significant reduction of serum CT, Ca and P levels, whereas serum OC and urinary Ca concentration in Orx rats were significantly higher than for the SO group (Filipović et al. 2007). In the current study, Orx led to a 93% decrease in serum concentrations of testosterone (P < 0.05) but had no effect on serum estradiol when compared with SO rats (Table 1).

Table 1. Serum calcitonin (CT), testosterone, estradiol, osteocalcin (OC), calcium (Ca), phosphorus (P) and urine Ca concentrations in sham-operated rats (SO), orchidectomized rats (Orx), testosterone propionate- (Orx + TP) and estradiol dipropionate-treated (Orx + EDP) orchidectomized rats
 SOOrxOrx + TPOrx + EDP
  1. Data expresed as mean ± SEM (n = 8).

  2. a

    P < 0.05 vs. Orx.

  3. b

    P < 0.05 vs. SO.

CT (ng L−1)2.60 ± 0.101.20 ± 0.07b2.37 ± 0.02a1.92 ± 0.08b,a
Testosterone (ng mL−1)1.51 ± 0.170.11 ± 0.02b19.36 ± 0.73b,a0.20 ± 0.02b
Estradiol (pg mL−1)36.15 ± 3.4431.00 ± 2.7283.15 ± 7.57b,a399.65 ± 34.72b,a
OC (ng L−1)9.02 ± 0.3119.74 ± 0.81b3.60 ± 0.20b,a2.57 ± 062b,a
Ca (mm L−1)2.46 ± 0.012.32 ± 0.02b2.36 ± 0.012.38 ± 0.03
P (mm L−1)2.35 ± 0.012.03 ± 0.01b2.20 ± 0.071.93 ± 0.01b
Urine Ca (mm L−1)3.49 ± 0.097.98 ± 0.50b1.25 ± 0.22b,a2.81 ± 0.21a

After TP administration, serum CT concentration was elevated by 98% (P < 0.05) in comparison with Orx rats. This treatment increased serum testosterone by 99 and 92% (P < 0.05) in comparison with the Orx and SO groups, respectively. After TP treatment, serum estradiol was also raised by 63 and 57% (P < 0.05) when compared with Orx and SO animals, respectively. Serum OC and urinary Ca levels were respectively 82 and 84% lower (P < 0.05) than in the control Orx group, and 60 and 64% lower (P < 0.05) than in the SO group (Table 1).

After EDP treatment, the concentration of serum CT was 60% (P < 0.05) higher than in the Orx animals, but 26% (P < 0.05) lower than in the SO rats. Administration of EDP decreased the serum concentration of testosterone by 87% (P < 0.05) compared with SO, whereas serum estradiol was 13 and 11 times higher than in the Orx and SO animals (P < 0.05), respectively. In the EDP-treated group, serum OC concentration was decreased by 87 and 72% (P < 0.05) in comparison with Orx and SO rats, respectively. Serum P was 18% (P < 0.05) lower after EDP administration compared with values for the SO group, whereas urinary Ca concentration was 65% lower than in the Orx animals (P < 0.05) (Table 1).


Thyroid C cells produce CT, a hormone that lowers plasma Ca concentration by suppressing osteoclast activity. Although both estrogen and testosterone may influence CT secretion, little information is available regarding possible sex hormonal regulation of these cells. We previously reported that estradiol deficiency after Ovx or androgen deficiency after Orx modulated the structure of rat thyroid C cells and decreased CT synthesis and release (Filipović et al. 2003, 2007). Other authors have also suggested that lack of gonadal steroids negatively affect thyroid C cells in Ovx and Orx rats (Lu et al. 2000; Sakai et al. 2000), as well as in both female and male human subjects (Isaia et al. 1992; Lu et al. 2000).

In the present study we demonstrated that Orx in male middle-aged rats induced a marked decrease in the peripheral circulating concentration of testosterone, while TP application to Orx animals elevated not only serum testosterone but also the concentration of serum estradiol. The increase in the estradiol levels may have been due to aromatization of exogenous testosterone in extratesticular tissue. On the other hand, serum estradiol concentrations in Orx rats can be elevated by supplying EDP, without altering testosterone concentrations. Compared with Orx animals, treatment of Orx rats with TP or EDP significantly increased the volume and volume density of thyroid C cells and raised serum CT concentrations. Also, low density granule contents after gonadal steroid treatments as well as prominent Golgi apparatus in C cells of testosterone-treated rats, indicated secretory activity similar to the SO control. Additionally, long profiles of rough endoplasmic reticulum after estradiol treatment implied active protein synthesis in these cells. These findings clearly indicate stimulatory effects of male and female sex steroids on the structure and function of CT-producing thyroid C cells in Orx middle-aged rats. To our knowledge, in addition to our morphological and hormonal data, only one report has described the effects of sex steroids on C cells in male rats (Sekulić & Lovren, 1993). These authors also demonstrated stimulatory effects of testosterone and estradiol on the morphology of thyroid C cells in aged male rats. On the other hand, several studies have shown that estrogen treatment stimulates CT secretion in Ovx rats (Catherwood et al. 1983; Filipović et al. 2003) and women (Isaia et al. 1992). Regulation of the activity of thyroid C cells by sex steroids apparently involves activation of their receptors. Namely, receptors for estrogens (ER) have been detected in both normal and hyperplastic C cells (Naveh-Many et al. 1992; Blechet et al. 2007) and binding of estradiol to ERβ in these cells can stimulate their activity. However, as androgen receptors (AR) have been detected in hyperplasia of C cells and medullary thyroid carcinoma (Zhai et al. 2003; Blechet et al. 2007), we cannot be sure that C cells normally have a functional AR. Further investigations are needed to confirm any presence of a functional androgen receptor in C cells. Also, in-depth studies of the molecular events, especially determination of mRNA levels for CT in C cells might distinguish between different rates of steroid effects on synthesis, storage and secretion of CT in these animals.

Evidence is now available that gonadal hormone deficiency is associated with reduced bone mass in both sexes, and there is general acceptance that androgens maintain bone structure in the male as do estrogens in the female. In addition to the influence of gonadal steroids on CT-producing C cells, we followed their effects on cancellous bone in proximal tibia of male rats. Orx rats represent an excellent animal model of male osteoporosis (Vanderschueren et al. 1992). We have already shown that androgen withdrawal induced by Orx in middle-aged rats results in reduction of trabecular bone mass and increased cancellous bone turnover (Filipović et al. 2007). In our experiment, we simulated a dose of testosterone typically applied in substitution therapy in andropausal men, and the dose of estradiol corresponded to one shown to have an osteoprotective effect in male orchidectomized mice and rats (Vandenput et al. 2001; Fitts et al. 2004). After treatment with both TP and EDP in this study, we detected significant increases of B.Ar, Tb.Th and Tb.N, whereas Tb.Sp was decreased markedly compared with Orx rats. Changes in bone histomorphometric parameters suggest that these treatments increased trabecular bone mass. The effect of EDP on these parameters was more pronounced than that of TP, but this may have been due to a different dosage rather than the nature of the sex steroid hormone and its receptor. Also, the decline of serum OC and urinary Ca concentrations indicates that both TP and EDP reduced cancellous bone turnover and urinary Ca excretion, which were raised after Orx. Our findings confirm previous reports that testosterone and estrogen improve trabecular structure in Orx rats (Wuttke et al. 2005; Stuermer et al. 2009).

Much evidence suggests that sex steroids have a protective effect on bone directly through their receptors. AR expression has been detected in bone cells, including osteoblasts and osteoclasts (Colvard et al. 1989; Turner et al. 2008). Thus, androgens may maintain trabecular bone volume directly via the osteoblasts (Notini et al. 2007) or inhibit bone resorption by suppressing the formation and activity of osteoclasts (Huber et al. 2001; Michael et al. 2005). Also, the antiresorptive effect of estrogen is due to direct action on bone, by inducing apoptosis in osteoclasts (Kousteni et al. 2002), probably through activating ERα in these bone-resorbing cells (Nakamura et al. 2007) and/or expression of both ERα and ERβ in osteoblasts (Bord et al. 2001). This effect of estrogen may be mediated by inhibiting the synthesis and secretion of some cytokines in osteoblasts that act as paracrine mediators in osteoclasts and increase their activity, as well as stimulating the secretion of alkaline phosphatase associated with increased bone formation (Turner et al. 1994; Harris et al. 1996).

In addition to binding to AR receptors in bone cells, testosterone may also exert significant effects on bone indirectly through aromatization to estradiol and subsequent activation of the ER (Syed & Khosla, 2005). The skeleton is a site for aromatase activity and osteoblasts possess aromatase and other enzymes necessary for the biosynthesis of estrogen (Vanderschueren et al. 1996). The increased bone turnover and decreased bone mass in aromatase-deficient male mice demonstrate the importance of estrogen in protecting the male skeleton (Miyaura et al. 2001). However, although data from cultured rat osteoblasts indicate the presence of aromatase activity in bone cells, Orx in aromatase-knockout mice resulted in bone loss beyond the adverse effects of aromatase deficiency alone. This suggests that androgens significantly attenuate bone turnover and preserve bone mass independently of aromatization (Matsumoto et al. 2006). Testosterone and the non-aromatizable androgen, dihydrotestosterone, can independently prevent osteopenia in rats after Orx, which indicates the importance of the AR-mediated pathway (Wakley et al. 1991).

In summary, this study showed that both testosterone and estradiol play important roles in maintaining bone mass and are included in the potential mechanism(s) of age-related bone loss in the male sex. Our observations of the stimulatory effects of sex steroids, at the cellular and subcellular levels of thyroid C cells in male rats, are pioneering and represent a solid basis for further molecular studies. The increase of calcitonin levels after sex hormone application in this experimental model of male osteoporosis indicates a possible additional mechanism by which these hormones may influence bone metabolism.


This work was supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia, Grant No. 173009. The authors wish to express their gratitude to Prof. Dr. Steve Quarrie for language correction of the manuscript.