Experimental Gerontology 72 (2015) 8598

Contents lists available at ScienceDirect

Experimental Gerontology
journal homepage: www.elsevier.com/locate/expgero

Testosterone and estradiol treatments differently affect pituitary-thyroid

axis and liver deiodinase 1 activity in orchidectomized middle-aged rats
B. oi-Jurjevi a,,1, B. Filipovi a,1, K. Renko b, M. Miler a, S. Trifunovi a, V. Ajdanovi a, J. Khrle b, V. Miloevi a
a
b

Institute for Biological Research Sinia Stankovi, University of Belgrade, Despot Stefan Blvd. 142, 11000 Belgrade, Serbia
Institut fr Experimentelle Endokrinologie, Charit - Universittsmedizin Berlin, Augustenburger Platz 1, D-13353 Berlin, Germany

a r t i c l e

i n f o

Article history:
Received 7 April 2015
Received in revised form 9 September 2015
Accepted 10 September 2015
Available online 15 September 2015
Keywords:
Testosterone
Estradiol
Pituitary-thyroid axis
Middle age
Rat

a b s t r a c t
We previously reported that orchidectomy (Orx) of middle-aged rats (1516-month-old; MA) slightly affected
pituitary-thyroid axis, but decreased liver deiodinase (Dio) type 1 and pituitary Dio2 enzyme activities. At
present, we examined the effects of subsequent testosterone-propionate treatment (5 mg/kg; Orx + T), and
compared the effects of testosterone with the effects of estradiol-dipropionate (0.06 mg/kg; Orx + E) treatment.
Hormones were subcutaneously administered, daily, for three weeks, while Orx and sham-operated (SO)
controls received only the vehicle. The applied dose of T did not alter serum TSH, T4 and T3 concentrations in
Orx- MA, though it increased TSH when administrated to Orx young adults (2.5-month-old; Orx-YA). However,
pituitaries of OrxMA + T rats had higher relative intensity of immunouorescence (RIF) for TSH; in their thyroids we found increased volume and height of follicular epithelium, decreased volume of the colloid and higher
RIF for T4-bound to thyroglobulin (Tg-T4). Liver Dio1 activity was increased. E-treatment did not affect serum
hormone levels, pituitary RIF for TSH, or liver Dio1 activity in Orx-MA rats. Thyroids had decreased relative
volume and height of follicular epithelium, increased relative volume of the colloid, decreased volume of
sodium-iodide symporter-immunopositive epithelium and lower RIF for Tg-T4. Detected changes were
statistically signicant. In conclusion, androgenization enhanced pituitary TSH RIF, thyroid activation and
liver Dio1 enzyme activity in Orx-MA, without elevating serum TSH as in Orx-YA rats. Estrogenization induced
pituitary enlargement with no effect on pituitary TSH RIF, serum TSH or liver Dio1 activity. E also induced
alterations in thyroid histology that indicate mild suppression of its functioning, and contributed to thyroid
blood vessel enlargement in Orx-MA rats.
2015 Elsevier Inc. All rights reserved.

1. Introduction
Serum testosterone (T) levels in men decline progressively with age.
This is associated with numerous symptoms and poor health condition,
including type-2 diabetes, higher incidence of cardiovascular disease,
and increased mortality (Samaras et al., 2013).

Abbreviations: Dio1, deiodinase type 1 enzyme; Dio2, deiodinase type 2 enzyme;;

DAB, diaminobenzidine tetrahydrochloride; E, estradiol; HRP, horseradish peroxidase;
IHC, immunohistochemistry; IFC, immunouorescence; MA, middle-aged adults; NIS,
sodium-iodide symporter; PBS, phosphate buffer saline; RER, rough endoplasmic reticulum; RIF, relative intensity of uorescence; ROI, region of interest; SD, standard deviation;
Orx, orchidectomy; Orx-YA, orchidectomized young adult males; Orx-MA, orchidectomized
middle-aged males; OrxPBS, phosphate saline buffer; RIA, radioimmunoassay; SO, shamoperated testicle-intact males; TEM, transmission electron microscopy; T, testosterone;
TST, testosterone supplementation therapy; Tg, thyroglobulin; Tg-T4, T4bound to thyroglobulin; TH, thyroid hormones; T4, thyroxine; T3, triiodothyronine; TSH, thyroid-stimulating
hormone; TRH, thyrotropin-releasing hormone; YA, young-aged adults;; VEGF, vascular endothelial growth factor.
Corresponding author at: Department of Cytology, Institute for Biological Research,
Despot Stefan Blvd. 142, 11060 Belgrade, Serbia.
E-mail address: brankasj@ibiss.bg.ac.rs (B. oi-Jurjevi).
1
BJ and BF equally contributed to the paper.

http://dx.doi.org/10.1016/j.exger.2015.09.010
0531-5565/ 2015 Elsevier Inc. All rights reserved.

Testosterone supplementation therapy (TST) creates a wide range of

positive effects on most symptoms of andropause (Isidori et al., 2015).
Although there is a potential risk of this treatment (Samaras et al.,
2013), use of TST has increased over the past decade (Klotz, 2015). Estrogen (E) plays important role in male physiology. Some negative
andropause symptoms are reduced by E-treatment. In men with prostate carcinoma androgen ablation therapy increases the risk of osteoporosis and bone fractures (Lipton et al., 2012). Estrogen therapy can also
induce state of androgen deprivation and exert a benecial effect on
bone preservation (Eriksson et al., 1995), but is toxic and induce breast
growth in men (Cox and Crawford, 1995). However, parenteral administration of E in men with metastatic prostate cancer was reported to
reduce risk of its toxicity (Oh, 2002; Hedlund et al., 2008).
Thyroid cancer is more common in women than in men and it is
most common in women during their reproductive years. However,
among elderly population, men are at higher risk and the thyroid cancer
is more aggressive then in elderly women (Rukhman and Silverberg,
2011). Several aspects of the hypothalamic-pituitary-thyroid axis are
sexually dimorphic. This is more obvious in rodents than in humans
male rodents have higher thyrotropin (TSH) values in sera than females
(Chen, 1984). Sex steroids differently affect proliferation of follicular

86

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

epithelium and sodium-iodide symporter (NIS) gene expression in

the thyroid tissue (Banu et al., 2002; Stanley et al., 2010). Liver
iodothyronine deiodinase type 1 (Dio1) enzyme activity is higher in
males (Miyashita et al., 1995). Some of these sex-related differences
may change with advancing age (Chen, 1984; da Costa et al., 2001;
Schomburg et al., 2007). In rodents, aging is associated with mild central
hypothyroidism in both genders (Donda et al., 1987; Cizza et al., 1992).
Orchidectomy (Orx) of young male rats has been reported to decrease concentration of TSH in circulation, while subsequent T treatment increased TSH. However, the stimulatory effect of androgens
seems to cease with advancing age (Chen, 1984). When we tested
how pituitary-thyroid-periphery signaling responds to Orx in middleaged rats (MA), liver Dio1 and pituitary Dio2 enzyme activities were decreased. Minor changes in thyroid structure, besides unchanged serum
thyroxine (T4) and TSH in Orx-MA, were detected. Our testicle-intact
MA model was characterized by almost 50% and 30% lower serum T
and T4, respectively, without apparent changes in concentration of
TSH in comparison to young adult (YA) males (oi-Jurjevi et al.,
2012). Therefore, we hypothesized that the pituitarythyroid axis of
MA adults would be less sensitive to changes in the androgen milieu.
To further test this hypothesis, in the present study we examined if
the high dose of testosterone may provoke expected stimulation of
pituitary-thyroid axis and increase liver Dio1 activity in Orx-MA rats.
Keeping in mind that both T and E treatments are applied in therapy
of aging men, and that several points of HPT system regulation are sexually dimorphic, we also aimed to examine potential differences between the effects of androgenization and estrogenization. Effects of
both sex steroid treatments on serum concentration of TSH, total T4
and T3, which we used as markers of pituitary-thyroid axis functioning,
were evaluated in comparison with the age-matched Orx- and shamoperated (SO)- MA controls, as well as with equally-treated Orx
young-aged (Orx-YA) animals. Histological and liver Dio1 activity assessments were performed in MA rats. The special attention has been
paid to histostructural, ultrastructural, immunohistochemical (IHC)
and immunouorescence (IFC) examination of the thyroid tissue, due
to increased risk of thyroid carcinomas (La Vecchia et al., 1999;
Stanley et al., 2012; Zhang et al., 2015) development related to sex
steroid treatments.
2. Material and methods
2.1. Animals, treatments and organ processing
Wistar rats were housed in the unit for experimental animals at the
Institute for Biological Research Sinia Stankovi, Serbia. All animals
were fed ad libitum and were maintained at constant light (12 h light/
12 h dark) and temperature (21 2 C) conditions.
At the age of 2 (YA) and 15 (MA) months, rats were orchidectomized
(Orx-YA and Orx-MA, respectively) under ketamine anesthesia
(15 mg/kg body weight of Ketamidor10%, Richter Pharma, Wels,
Austria. Two weeks after the surgery, Orx animals were divided in the
following groups: (i) testosterone-treated Orx-YA (Orx-YA + T) and
Orx-MA (Orx-MA + T) groups were injected subcutaneously (s.c.)
with 5 mg/kg b.w. testosterone-propionate (Fluka Chemie AG, Buchs,
Switzerland); (ii) estradiol-treated Orx-YA (Orx-YA + E) and Orx-MA
(Orx-MA + E) group was injected s.c. with 0.06 mg/kg b.w. estradioldipropionate (ICN Galenika Pharmaceuticals, Belgrade, Serbia). All substances were administered daily for three weeks. Rats in the control
sham-operated (SO) and Orx group were s.c. administered the same
volume of vehicle (sterile olive oil) according to the same schedule.
Each group contained 6 animals.
The animals were decapitated 24 h after the last treatment. Blood
was collected from the trunk of each animal (both YA and MA groups)
and allowed to clot by leaving it at room temperature in glass tubes
(without any coagulant) for at least 30 min. The clot was removed by
centrifuging (room temperature; 10002000 g, 15 min). The sera

(supernatant) were transferred to eppendorf tubes and, together with

liver samples, stored at 70 C.
The pituitaries and one thyroid lobe only from MA group of animals
were xed in Bouin's solution, dehydrated and embedded in Histowax
(Histolab Product Ab, Gteborg, Sweden) for light or confocal microscopy. The other thyroid lobe was further processed for transmission
electron microscopy (TEM).
Weight loss/gain for each animal was calculated by subtracting nal
treatment weight from pretreatment weight (in case of orchidectomy
from precastration weight). The relative organ weights are expressed
as an absolute organ weight (mg)/nal body weight (g) 100.
All animal procedures were in agreement with the EEC Directive
(86/609/EEC) Directive 2010/63/EU on the protection of animals used
for experimental and other scientic purposes, and approved by the
Ethical Committee for the Use of Laboratory Animals of the Institute
for Biological Research Sinia Stankovi, University of Belgrade.
2.2. Novelli (acid fuchsin-light green) histochemical staining
Novelli histochemical staining (Novelli, 1959) was used for gaining
insight into the thyroid tissue vascular prole. In brief, deparafnized
and rehydrated thyroid sections were incubated in hot 1 N HCl (60 C,
3 min), followed by staining in 1% acid fuchsin (Fluka Chemie AG,
Buchs, Switzerland; 30 s) and 1% light green (Sigma-Aldrich, St. Louis,
MO, USA; 3 min), respectively. In between, the slides were washed in
distilled water, and after the last one, dehydration and mounting in
DPX (Sigma-Aldrich, Barcelona, Spain) were carried out. As the result,
acid fuchsin-stained purple erythrocytes within the blood vessels and
capillary network were clearly visible at the green background of the
thyroid tissue.
2.3. Immunohistochemistry
Procedures for IHC and IFC staining were performed as previously
described (Miler et al., 2014; oi-Jurjevi et al., 2014).
Pituitary thyrotrophs or lactotrophs were identied by incubation
with polyclonal rabbit antirat TSH or prolactin (donation from Dr.
A. F. Parlow, National Institute of Health, Bethesda, MD, USA; 1:500)
or monoclonal mouse anti-human prolactin (Prl; Abcam, Cambridge,
UK; 1:200) antibodies overnight at 4 C.
For double immune-detection, DAKO EnVision Doublestain System
was applied (Dako North America, Inc. Carpinteria, CA, USA) according
to the manufacturer's instruction.
For IHC and IFC characterization of thyroid tissue, the following primary antisera were applied overnight at 4 C: the rabbit antisera directed against rat NIS (Acris antibodies GmbH, Herford,Germany; 1:1200);
the rabbit antisera directed against human vascular endothelial growth
factor (VEGF; Abcam, Cambridge, UK; 1:100); the rabbit antisera directed against human thyroglobulin (Tg; Dakopatts, Glostrup, Denmark;
1:500); and the mouse antisera directed against human T4 bound to
thyroglobulin (Tg-T4 monoclonal antibody; QED Bioscience, San Diego,
CA, USA; 1:300).
For immunodetection of NIS and Tg, swine anti-rabbit IgG- horseradish peroxidase (HRP; Dakopatts, Glostrup, Denmark; 1:100) was
applied as a secondary antiserum for 1 h. For VEGF, Vectastain ABC Kit
(Vector Laboratories, Burlingame, USA) was applied according to
procedure suggested by the manufacturer.
Visualizations were performed using diaminobenzidine tetrahydrochloride (DAB; Dakopatts, Glostrup, Denmark) or fast red (Sigma-Aldrich,
Barcelona, Spain) chromogen substrates at concentrations suggested by
the manufacturer.
The sections were counter-stained with hematoxylin and mounted
in DPX medium (Sigma-Aldrich, Barcelona, Spain). Digital images
were made on a DM RB Photomicroscope with a DFC 320 CCD Camera
(Leica, Germany).

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

For IFC staining of pituitary TSH and thyroid Tg-T4, Alexa Fluor 488
donkey anti-rabbit IgG (Invitrogen Life technologies, CA, USA; 1:200)
was applied as secondary antiserum. The sections were mounted with
Mowiol 488 (Sigma-Aldrich, St. Louis, MO, USA). The IFC was detected
using a confocal laser scanning microscope Leica TCS SP5 II Basic (Leica
Microsystems CMS GmbH; Germany).
All incubation steps were performed at room temperature in a dark
humid chamber. For washes and antibody dilutions, 0.1 mol/l phosphate saline buffer (PBS; pH 7.4) was applied. Representative sections,
which were processed in the same way as described above using PBS
instead of the primary antibodies in the incubation, were used for the
control purposes.
2.4. Quantitative image analyses
A quantitative image analysis was performed on captured images
of IFC-stained sections taken from anterior, central, and posterior
part of pituitary or thyroid glands (23 slides per each tissue level/
organ/animal, n = 6). Relative intensity of uorescence (RIF) was
determined from measurements of one hundred TSH-IFC stained
pituitary thyrotroph cells or fty thyroid Tg-T4 IFC-stained follicles per
each tissue level/organ/animal. Measurements were performed using
the Quantify option in LAS AF Lite software (Leica Microsystems CMS
GmbH; Germany), as previously reported (Miler et al., 2014). In brief,
IFC-stained regions of interest (ROI) were encircled with drawing tool.
Two other immune negative spots in proximity to ROI were also rounded for background subtraction. The mean value of the two repeated
measurements for each pituitary thyrotroph cell or thyroid follicle
was calculated and then, after subtraction of background uorescence,
used for RIF determination.
2.5. Stereological and morphometrical analyses
Stereological analyses of thyroid sections were carried out by pointcounting method (Weibel, 1979) as previously described (Miler et al.,
2014). Briey, four to ve 5 m-thick sections from the anterior, central
and posterior parts of each animal's thyroid were analyzed. Sections
stained with hematoxylin-eosin (HE), Novelli, or NIS (IHC) were used
for morphometric study. The measurements were carried out using a
newCAST stereological software package (VIS Visiopharm Integrator
System, version 3.2.7.0; Visiopharm; Denmark), at objective magnication of 20 (for HE or Novelli stained sections) or 40 (for NIS IHC
stained sections).
The counting area was dened using a mask tool; test grid with uniformly spaced test points and lines was provided by the new-CAST software. Test points hitting the colloid, epithelium and interstitium were
determined. Relative volume densities (VV) were calculated as the
ratio of the number of points hitting each tissue component divided
by the number of points hitting the reference space, i.e. analyzed thyroid
section: VV (%) = Pp/Pt 100 (Pp, counted points hitting the tissue
component, Pt, total of points of the test system hitting reference
space). Relative volume densities (VV) were calculated for each tissue
component, representing VV of the colloid, epithelium and interstitium.
The relative volume density of blood vessels in interstitial tissue
fraction was determined on Novelli-stained sections as the ratio of the
number of points hitting blood vessels and capillary network divided
by the number of points hitting the interfollicular space.
NIS positivity was interpreted as membrane or cytoplasmic staining. The relative volume density of NIS-immunopositive epithelium
(membrane + cytoplasmic) was estimated as the ratio of the number of points hitting NIS-immunopositive membranes and/or NISimmunopositive cytoplasm divided by the number of points hitting
the reference space (epithelium + colloid + interstitium).
As to morphometry, the follicular epithelium cell height (; m) was
calculated according to (Bogataj et al., 1977) from the formula =
3Pe Lt/2Pt (Ie + Ii + (Ie + Ii)1/2 (Pe counted points on epithelium,

87

Lt total length of the test line, Pt total of points of the test system, Ie
and Ii external and internal intersections of the structures with the
test line, respectively). The activation index (AI) is the VV epithelium/
VV colloid ratio and indicates TSH-mediated activation of the thyroid
follicles (Kalinik et al., 1988).
2.6. Ultrastructural analyses
For electron microscopy, thyroid tissue samples (810 samples/
thyroid lobe/animal, n = 6) were immersed in 4% glutaraldehyde
24 h and post-xed in 1% osmium tetroxide for 1 h. Specimens were
then dehydrated through a graded series of ethanol (30100%) and
embedded in Araldite (Agar scientic, Cambridge, UK). Ultrathin sections were cut with a Diatome ultra 45 diamond knife (Diatome,
Switzerland). Five grids with ultrathin sections from 4 randomly chosen
tissue samples/thyroid lobe/animal were stained with uranyl acetate
and lead citrate (Chemapol, Prague, Czech Republic) and examined
under a MORGAGNI 268 (FEI Company, USA) transmission electron
microscope.
2.7. Hormone analyses
Concentration of TSH in sera was measured with radioimmunoassay
(RIA), using rat TSH kit (Immunodiagnostics systems GmbH, Germany),
and RIA T4 and triiodothyronine (T3) kits developed by INEP (Zemun,
Serbia). According to the manufacturer, the calibration range of TSH
RIA was up to 30 ng/ml and the lower limit of detection was
1.03 ng/ml. The calibration range of T4 RIA was up to 500 nmol/l, with
the lower limit of detection of 6 nmol/l. For T3 it was up to 10 nmol/l,
with the lower limit of detection of 0.3 nmol/l. All samples were measured in duplicate within a single assay, with intra-assay coefcient of
variation of 6.1% for TSH, 4.7% for T4 and 5.3% for T3.
2.8. 5- Iodothyronine deiodinase activity measurements
Liver protein samples were prepared (oi-Jurjevi et al., 2012)
and Dio1 enzyme activities measured precisely as previously described (Renko et al., 2012). In brief, reaction mixture contained
40 g of liver microsomal proteins, 10 l of water or 10 mmol 6-npropyl-2-thio-uracil (PTU) and 50 l of freshly prepared substrate
mix (10 mol rT3 (Sigma-Aldrich, MO, USA), 0.2 mol KPO4 (pH 6.8),
2 mmol ethylenediaminetetraacetic acid, and 80 mmol dithiothreitol).
Enzyme reaction lasted for two hours at 37 C. After centrifugation (4
C, 15.000 g, 15 min), supernatant was used for quantication of released iodide. Dowex W50-X2 resin columns served for separation of
intact rT3 and the deiodinated breakdown products from the released
iodide. The iodide content was determined by the Sandell-Kolthoff reaction, using cerium solution [22 mmol (NH4)4Ce(SO4)4 and 0.44 mol
H2SO4] and arsenite solution (25 mmol NaAsO2,0.8 mol NaCl, and
0.5 mol H2SO4). The changes in absorption (OD at 415 nm) were determined at reaction starting point and after 20 min. All protein samples
were assayed in triplicate. The tubes that contained PTU, Dio1 enzyme
inhibitor, were used for background subtraction. Calculation of the enzyme activities was performed as previously described (Renko et al.,
2012).
2.9. Statistical analyses
Statistical analyses of all obtained results were performed using
GraphPad Prism v.6 for Windows (San Diego, CA, USA). To test the normality of distribution, the sets of results obtained for all experimental
groups were rstly subjected to KolmogorovSmirnov normality test,
while the equality of variance was tested by Bartlett's test. After conrmation of Gaussian distribution and homogeneity of variance, results on
concentration of hormones in sera and body weight gain were subjected
to two-way analysis of variance (ANOVA) and then to Tukey's multiple

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

88

Fig. 1. Concentration of TSH (A), total T4 (B) and T3 (C) in sera of control sham-operated (SO), orchidectomized (Orx), testosterone (Orx + T) and estradiol (Orx + E) treated
orchidectomized young adult (YA) and middle-aged (MA) rats. The values are means SD (n = 6); statistics: two way ANOVA, Tukey's multiple comparisons post hoc test, p b 0.05 versus
Orx or SO rats of the corresponding age.

comparison test for post hoc comparisons. For all other results, we used
one-way ANOVA for comparative evaluations, followed by Tukey's multiple comparison tests (versus Orx or SO control groups). Statistical signicance was set at p b 0.05. The data are presented as the means SD.
3. Results
3.1. Concentration of hormones in sera
Age- and treatment- related changes in serum concentrations of
TSH, total T4 and T3, which served as markers of pituitary-thyroid axis
functioning, are summarized in Fig. 1.
In case of concentration of TSH in sera, the effect of treatment was
signicant (p = 0. 02, two way ANOVA). The interaction of treatment
and age almost reached the statistical signicance (p = 0.07, two way
ANOVA). In case of concentration of T4 and T3 in sera, the effects of animals' age were signicant (for T4 p = 0.0005; for T3 p = 0.0009, two
way ANOVA), while the effect of treatment, or the interaction between
and age and the treatment were not signicant.
In Orx-YA animals TSH was 35% (p = 0.03, Tukey's multiple comparisons post hoc test) lower in comparison with the value obtained for SOYA controls. Subsequent T-treatment reversed this change (p = 0.03,
Tukey's multiple comparisons post hoc test) in comparison to value obtained for Orx-YA group (Fig. 1A). Concentration of TH (T4 and T3)
remained unaltered when compared Orx-YA + T with the control
Orx- or SO-YA values.
However, in MA animals, orchidectomy or subsequent T- treatment
did not change hormone levels (Fig. 1A-C).
E-treatment at both examined age did not affect serum hormone
levels in comparison to Orx controls of the corresponding age. Serum
TSH level in Orx-YA + E group remained at the level of Orx-YA controls
(36% lower, p = 0.02, Tukey's multiple comparisons post hoc test, in
comparison to value obtained for SO-YA rats; Fig. 1).

3.2. Body and organ weights

Orchidectomy of MA rats resulted in weight loss (by 10%, p b 0.0001,
two way ANOVA, Tukey's multiple comparisons post hoc test; Table 1).
Subsequent T-treatment of Orx-MA rats reversed this change (p = 0.03,
two way ANOVA, Tukey's multiple comparisons post hoc test), with no
effect on pituitary or thyroid weights (Table 1). E-treatment did not
alter weight gain, but induced 2.6-fold increase of absolute and relative
pituitary weight (p b 0.0001, one way ANOVA, Tukey's multiple
comparisons post hoc test) in comparison with the values obtained for
Orx-MA and SO-MA controls. Thyroid weight was unchanged after
hormonal treatments (Table 1).

3.3. Pituitary gland-thyrotrophs and lactotrophs

To study if substantial increase in pituitary weight after E-treatment
of Orx-MA animals may be attributed to stimulation of prolactinproducing pituitary cell population, we performed IHC analysis of pituitary lactotrophs (Fig. 2AE). No signicant changes in the pituitary
lactotrophs of Orx-MA + T group could be distinguished (Fig. 2B) in
comparison with controls. Both hypertrophy and hyperplasia of
lactotrophs (Fig. 2D), as well as enlarged and dilated blood vessels,
were clearly evident in the pituitary of Orx-MA + animals (Fig. 2E).
Pituitary TSH immunouorescence was located within the cytoplasm of thyrotrophs, which were mainly polygonal or eccentric in
shape, often with tail-like cytoplasmic extensions (Fig. 3AD). After Ttreatment of Orx-MA rats, pituitary thyrotrophs were characterized by
more pronounced cytoplasmic TSHimmunopositivity; RIF was 44%
(p = 0.01, one way ANOVA, Tukey's multiple comparisons post
hoc test; Fig. 3E) higher in comparison with the value obtained for
Orx-MA controls (not signicantly increased in comparison with RIF
obtained for SO-MA rats), while in Orx-MA + E rats it did not change

Table 1
Effects of testosterone (T) and estradiol (E) treatments on body weight, body weight gain, pituitary and thyroid weight in orchidectomized middle-aged (Orx-MA) rats.
Groups

Precastration
body mass (g)

Pretreatment
body mass (g)

Final body
mass (g)

Body
weight (mg)

Absolute pituitary
weight (mg)

Relative pituitary weight

(mg 100/b.w)

SO-MA
Orx-MA

698 85
692 40

684 58
639 64

701 71
623 76

17.1 1.9
17.4 2.7

Orx-MA + T

707 54

664 60

697 67

Orx-MA + E

662 66

641 86

623 79

16 16
-(60 28)
(p b 0.0001)
37 10
(p = 0.03)
-(10 16)

Absolute thyroid
weight (mg)

Relative thyroid weight

(mg 100/b.w)

2.4 0.7
2.9 0.6

33.0 2.6
34.0 1.7

5.5 0.5
5.7 0.5

16.5 2.8

2.5 0.3

32.0 2.5

4.4 0.8

46.0 8.9
(p b 0.0001)

7.7 2.1
(p b 0.0001)

33.0 1.3

5.0 0.5

The values are means SD (n = 6); statistics: for weight gain two way ANOVA and Tukey's multiple comparisons post hoc test were applied (for Orx-MA p b 0.05 versus precastration
body mass, for Orx + T and Orx + E groups p b 0.05 versus pretreatment body mass); for organ weights one way ANOVA and Tukey's multiple comparisons post hoc test were applied, p b
0.05, versus Orx-MA and SO-MA rats.

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

89

Fig. 2. Prolactin (Prl; brown)- and thyrotropin (TSH; red)-immunopositive cells in the pars distalis of the pituitary gland from the control sham-operated (A) orchidectomized (B), testosterone (C) and estradiol (D, E) treated orchidectomized middle-aged rat. Light microscopy, double immunohistochemical staining of Prl and TSH.

signicantly in comparison to values obtained for control Orx- or

SO-MA groups (Fig. 3).
3.4. Thyroid glandhistostructural and functional characterization
The thyroid parenchyma of SO- and Orx-MA rats was composed of
follicles, which were generally characterized by cuboidal follicular epithelium and variable amount of luminal colloid (Figs. 4AB and 5AB).
Morphometric parameters obtained for the thyroid tissue are provided
in the Fig. 6. In comparison with SO-MA group, Orx-MA had more
frequent reabsorption vacuoles in the luminal colloid of the thyroid
follicles (Fig. 4A-B).
Both SO (Fig. 7A) and Orx (Fig. 7B) thyrocytes were characterized by
a typical polarization of organelles from basal to apical pole, with clearly
detectable microvilli at the apical and big euchromatic nuclei at the

basal pole. In the apical region a limited number of small dense exocytotic granules were visible. Dilated cisternae of rough endoplasmic reticulum (RER) were noted in Orx-MA but not in SO-MA group. A great
number of large dense highly osmiophilic bodies throughout the cytoplasm and colloid droplets of smaller size were noted in the thyrocytes
of both control MA groups.
In Orx-MA + T group, thyroid tissue was composed of follicles lined
by taller follicular epithelium (columnar epithelium) and smaller portion of luminal colloid in comparison to both Orx- and SO-MA controls
(Fig. 4C). The following parameters increased in comparison to the control Orx-MA values: The relative VV of the interstitium (by 9%, p =
0.008, one way ANOVA, Tukey's multiple comparisons post hoc test;
Fig. 6), and AI (by 11%, p = 0.02, and 30%, p = 0.002, respectively,
one way ANOVA, Tukey's multiple comparisons post hoc test; Fig. 6).
The relative VV of the colloid decreased (by 23%, p = 0.009; one way

90

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

Fig. 3. TSH-immunouorescence in the thyrotroph cells of pars distalis of the pituitary gland from the of control sham-operated (A; SO-MA), orchidectomized (B; Orx-MA), testosterone
(C; Orx-MA + T) and estradiol (D; Orx-MA + E) treated orchidectomized middle-aged rat; Confocal microscopy, immunouorescence for TSH. The values for relative intensity of
TSH immunouorescence (RIF; E) are means SD (n = 6); statistics: one way ANOVA, Tukey's multiple comparisons post hoc test, p b 0.05 versus Orx-MA rats.

ANOVA, Tukey's multiple comparisons post hoc test; Fig. 6). The same
parameters were changed in the same manner in comparison to the
values obtained for SO-MA controls (Fig. 6).
When examining blood vessels on Novelli-stained thyroid sections
(Fig. 5), we detected enlargement of interfollicular larger set of capillaries, as well as of smaller ones more closely adherent to the follicles
in the thyroids of Orx-MA + T group (Fig. 5C); Vv of blood vessels within
the interstitial tissue was 29% higher (p = 0.0008; one way ANOVA,
Tukey's multiple comparisons post hoc test; Fig. 6) in comparison to
the value obtained for Orx-MA animals. VV of blood vessels was similarly increased in comparison with the value obtained for SO-MA controls
(p = 0.0004; one way ANOVA, Tukey's multiple comparisons post hoc
test).
Increased height of thyrocytes in comparison to both Orx- and SOMA controls was easily distinguishable at TEM micrographs. Numerous
mitochondria between the dilated cisternae of RER were present in the

apical part of the cytoplasm, as well as small dense vesicles and quite big
colloid droplets, indicating intense reabsorption of colloid (Fig. 7C).
Histological analysis of thyroid tissue in Orx-MA + E rats revealed
distended follicles, composed of attened follicular epithelium and signicant amount of colloid in the lumen (Fig. 4D). The following parameters decreased in comparison with the values obtained for the control
Orx-MA group (Fig. 6): relative VV of the epithelium (by 9%, p = 0.03;
one way ANOVA, Tukey's multiple comparisons post hoc test) and
(by 13%, p = 0.003; one way ANOVA, Tukey's multiple comparisons
post hoc test). VV of the colloid increased (by 15%, p = 0.002, one way
ANOVA, Tukey's multiple comparisons post hoc test) in comparison
with the Orx-MA. The same morphometric parameters were similarly
changed in comparison with the values obtained for SO controls, except
VV of the colloid, which were not signicantly different when compared
with the values obtained for SO-MA controls (Fig. 6). In thyroid tissue of
Orx + E group we also detected enlargement of interfollicular capillars,

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

91

Fig. 4. Thyroid gland of control sham-operated (A), orchidectomized (B), testosterone (C)- and estradiol (D) - treated orchidectomized middle-aged rat. Light microscopy, hematoxylin-eosin.

Fig. 5. Thyroid gland of control sham-operated (A), orchidectomized (B), testosterone (C)- and estradiol (D) - treated orchidectomized middle-aged rat. Light microscopy, Novelli
histochemical staining. Black arrows indicate large blood vessels, and white arrows small vessels.

92

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

Fig. 6. The relative volume density (VV; %) of follicular epithelium (A), interstitium (B) and colloid (C), height of thyroid follicular cells (;m; D), activation index (AI; E), and VV of blood
vessels within the interstitium (%; F) of the thyroid tissue in sham-operated (SO-MA), orchidectomized (Orx-MA), testosterone (Orx-MA + T) and estradiol (Orx-MA + E) treated
orchidectomized middle-aged rats. The values are means SD (n = 6); statistics: one way ANOVA, Tukey's multiple comparisons post hoc test, p b 0.05 versus Orx-MA or SO-MA rats.

without effect on capillary network adjacent to the follicles (Fig. 5D). VV

of blood vessels was increased in comparison with Orx-MA (by 12%,
p = 0.03, one way ANOVA, Tukey's multiple comparisons post hoc
test; Fig. 6). When compared with SO-MA group, the difference almost
reached statistical signicance (p = 0.07; one way ANOVA, Tukey's
multiple comparisons post hoc test; Fig. 6). No microhemorrhage
could be distinguished when examined on semine sections (results
not shown).
Ultrastructurally, thyrocytes were of lowest height, with shortest
microvilli. Apical small dense vesicles and small colloid droplets indicated ongoing Tg synthesis and reabsorption of colloid. However, the presence of cystic dilation in close proximity to RER, lled with material of
lower electron density, may indicate degenerative changes (Fig. 7C).
Functional characterization of thyroid tissue was performed by a
comprehensive IHC and IFC analyses (Grard et al., 2003; Faggiano

et al., 2007). NIS-immunopositivity was localized in the cytoplasm

(weak intensity) and/or at the basolateral membrane (high intensity)
of thyrocytes. Within a single follicle only few thyrocytes expressed
membrane NIS-immunopositivity (Fig. 8). VV of NIS-immunopositive
epithelium (cytoplasmic + membrane) in Orx-MA group was 38
5%, of which basolateral membrane-positivity was 18 4%. These
values were not signicantly different from the values obtained for
SO-MA group (34 6%, basolateral NIS 14 5%). In Orx-MA + T
group VV of NIS-positive epithelium was not signicantly different
(38 3%, basolateral-immunopositivity represented 23 2%) from
the values obtained for control Orx- and SO-MA groups. In OrxMA + E rats, VV of NIS-positive cells was 23 2%, basolateral NISpositivity being 10 1%. Both parameters were decreased by 40%
(p = 0.02, one way ANOVA, Tukey's multiple comparisons post hoc
test) and 44% (p = 0.004, one way ANOVA, Tukey's multiple

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

93

Fig. 7. Thyroid follicular cells from control sham-operated (A), orchidectomized (B), testosterone (C)- and estradiol (D)- treated orchidectomized middle-aged rat. N, nucleus; RER, rough
endoplasmic reticulum; DB, dense bodies; CD, colloid droplets; C, colloid; AM, apical membrane; M, mitochondria; V, small dense vesicles; CC, thyroid C cell; Cy, cystic membrane-bound
dilatiation lled with lower electron-density material. Transmission electron microscopy.

Fig. 8. Immunohistochemical staining of sodium-iodide symporter (NIS) in the thyroid gland of control sham-operated (A), orchidectomized (B), testosterone (C)- and estradiol
(D)- treated orchidectomized middle-aged rat. Black arrows point to basolateral NIS immunostaining. Light microscopy.

94

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

Fig. 9. Immunohistochemical staining of vascular endothelial growth factor (VEGF) in the thyroid gland of control sham-operated (A), orchidectomized (B), testosterone (C)- and estradiol
(D)- treated orchidectomized middle-aged rat. Black arrows indicated occasional VEGF-immunopositive cells in blood vessels. Light microscopy.

comparisons post hoc test), respectively, in comparison to the control

Orx-MA values. However, the basolateral membrane NIS-positivity in
Orx-MA + E group was unaltered in comparison with SO-MA rats.
VEGF-immunopositivity was granular and uniformly distributed in
the cytoplasm of thyrocytes, being the most prominent in OrxMA + T group, in which the follicular epithelium was the tallest, and
weakest in Orx-MA + E group, which had the attened follicular epithelium. Occasional VEGF-immunopositive cells were also evident in blood
vessels (Fig. 9 AD).
Tg-immunostaining was most intense at the border between the
apical membrane of the thyrocytes and the colloidal lumen of OrxMA + T rats (Fig. 10). Quantitative IFC analysis revealed 40% higher
(p = 0.03, one way ANOVA, Tukey's multiple comparisons post hoc
test) and 30% lower (p = 0.001, one way ANOVA, Tukey's multiple comparisons post hoc test) RIF for Tg-T4, respectively, at the border of the
colloid lumen in follicles of Orx-MA + T and Orx-MA + E rats, in comparison with the value obtained for Orx-MA (Fig. 11). However, this
change was not signicant when Orx-MA + E were compared with
SO-MA (Fig. 11E).

3.5. Liver deiodinase type 1 activity

In the present study specic Dio1 enzyme activities in the liver were
determined by non-radioactive enzyme assay. The data for all experimental groups are summarized in the Fig. 12. Orchidectomy decreased
(by 18%, p = 0.03, one way ANOVA, Tukey's multiple comparisons
post hoc test), while subsequent T-treatment of Orx-MA rats reversed
liver Dio1 activity, which was 46% higher (p = 0.0004, one way
ANOVA, Tukey's multiple comparisons post hoc test) in comparison
with the value obtained for Orx-MA. E-treatment did not affect Dio 1 enzyme activity, which remained almost the same as in Orx-MA (Fig. 12).

4. Discussion
The results obtained in this study indicate direct and independent
stimulatory effect of T on: (i) pituitary TSH production, (ii) thyroid tissue activation, and (iii) increased Dio1 activity and metabolism of TH
in the liver of Orx-MA rats. The TSH levels in sera of Orx-MA + T
group did not increase as we detected in Orx-YA + T rats, in line with
our hypothesis on age-related decline in pituitary TSH cells responsiveness to T. Estrogenization caused pituitary enlargement, hypertrophy
and hyperplasia of lactotrophs, with no apparent effect on pituitary
TSH-immunostaining. Moreover, E had mild suppressive effect on
thyroid tissue, but it did not affect serum TSH and TH hormone levels,
or liver Dio1 enzyme activity in comparison with Orx-MA rats.
We previously reported that the doses of T and E applied in this experiment had bone-protective effect in our Orx-MA model (Filipovi
et al., 2013). Besides the benecial effect on bone tissue, the applied
doses of sex steroids, which induced approximately 10-fold increase
of serum T or E levels, respectively (Filipovi et al., 2013), were aimed
to challenge thyroid homeostasis of MA rats. The same dose of testosterone propionate was previously reported to increase serum TSH in young
male rats (Ross, 1990). Another estradiol ester, estradiol benzoate, at
dose of 0.25 mg/kg b.w. decreased the pituitary pool of TSH when administered to young euthyroid males, and increased serum TSH when
administered to hypothyroid rats (Boado et al., 1985).
-treatment induced mild weight gain and restored the body weight
to pre-castration level. The obtained result is in line with Woodward
(1993), who conrmed that anabolic effect of testosterone persists in
adrenalectomized or hypophysectomized animals. E-treatment did
not alter body weight, but induced substantial increase in pituitary
weight in comparison to the control value. This increase is probably
due to E-induced hypertrophy and hyperplasia of lactotrophs, as well
as hypertrophy of blood vessels, detected under our experimental

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

95

Fig. 10. Immunohistochemical staining of thyroglobulin (Tg) in the thyroid gland of sham-operated (A), orchidectomized (B), testosterone (C)- and estradiol (D)- treated orchidectomized
middle-aged rat. Light microscopy.

conditions, in accordance with other authors (Shukuwa et al., 2006;

Ishida et al., 2007).
Despite more intense TSHimmunouorescence within the cytoplasm of pituitary thyrotrophs in Orx-MA + T group, T-treatment did
not alter concentration of TSH in sera of Orx-MA. However, as expected,
the applied dose of T was sufcient to increase serum TSH levels, without signicant effect on T4 and T3 concentration in Orx-YA model. In
agreement with the present study results, castration and subsequent
testosterone replacement of young rodents changed serum TSH, while
E did not show such effect (Farbota et al., 1987). The obtained results
are in line with our initial hypothesis, and indicate decreased responsiveness of the pituitary thyrotrophs regulation to T with aging. Increased density of both thyrotropin-releasing hormone (TRH) and T3
receptors was detected in 2224-month-old male rats in comparison
to young ones, with no change in TRH receptor afnity, pituitary T3 or
plasma TSH concentrations (Donda et al., 1987, 1989). Therefore, detected age-related difference in TSH response to T may be at least in
part due to additional hormonal changes at the level of hypothalamus
and pituitary. It is well known that somatostatin, the major physiological inhibitor of GH secretion, also inhibits TSH secretion in rats and in
humans. This hormone expression also shows sexual dimorphism and
a gradual decline during aging (Martinoli et al., 1991). Further studies
of hypothalamic somatostatin expression are needed to examine this
possibility. Moreover, it is also possible that the sex steroids-mediated
paracrine regulation between different pituitary cell types (Denef,
2008) may be changed with advancing age. This may also affect
hypothalamic-pituitary TSH regulation.
Important nding of this study is that high doses of both sex steroid
treatments did not induce pathological changes in the thyroid tissue of
MA rodents. However, histological, morphometric and ultrastructural
changes of thyroid parenchyma were evident and clearly indicated

stimulatory effect of T, and mild suppressive effect of E. This differential

effect of sex steroids was probably directly mediated by androgen or estrogen receptors, which were identied in the thyroid tissue (Pelletier,
2000; Banu et al., 2002; Stanley et al., 2010).
Increased presence of reabsorption colloid vacuoles in the luminal
colloid in thyroid follicles of both Orx- and Orx + T-MA rats was evident
in comparison to SO-MA rats. Augmented colloid droplets were also
identied under TEM for Orx-MA + T rats, while no major difference
could be distinguished when comparing thyroid follicular cells of SOMA with Orx-MA.
Thyrocytes of Orx-MA, and even more of Orx-MA + T rats, were
characterized by the presence of RER dilated cisternae, which may indicate cytoplasmic vacuolization and the early stage of apoptosis. These
changes were not noted in the cytoplasm of SO- or Orx-MA + E rats.
In vitro studies of the effect of testosterone and estradiol on thyroid
papillary carcinoma cell line demonstrated that E, but not T, enhanced
anti-apoptotic signaling (Lee et al., 2005). As no other early signs of
apoptosis were evident at TEM (i.e. at the level of cell nuclei or mitochondria), it is more likely that these changes can be attributed to
functional activation of thyroid follicles in relation to androgen deprivation and subsequent T replacement, then to early stage of apoptosis.
Assessment of apoptosis by IHC is needed to further examine this
possibility.
In case of Orx-MA + T rats, IHC revealed enhanced VEGFimmunostaining in comparison to both Orx- and SO-MA rats. Together
with increased volume density of blood vessels and accumulation of
red blood cells in the capillaries, the obtained data point to vascular activation in the thyroid (Colin et al., 2013). In line with these, the detected increase in Tg-immunostaining and RIF for Tg-T4, the obtained
results indicated increased activation of angiofollicular units and biosynthesis of T4. Since TSH concentration was not increased under our

96

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

Fig. 11. Immunouorescence for thyroxine bound to thyroglobulin (Tg-T4) in the thyroid gland of control orchidectomized (A), testosterone (B)- and estradiol (C)- treated
orchidectomized middle-aged rat and the relative intensity of TSH (RIF; D). Confocal microscopy. The values for relative intensity of TSH immunouorescence (RIF; E) are means
SD (n = 6); statistics: one way ANOVA, Tukey's multiple comparisons post hoc test, p b 0.05 versus Orx-MA or SO-MA rats.

experimental conditions, activation of angiofollicular units in Orx + T

rats can be explained by androgen receptor-mediated enhancement of
TSH signaling in the thyroid tissue (Banu et al., 2001).
Vascular status of the thyroid gland may also be inuenced by estrogens (de Araujo et al., 2010) but in our case VEGF-immunostaining
and Tg-T4 IFC signal was weaker in Orx-MA + E rats, and may be a
consequence of impaired TSH action on the thyroid (Banu et al., 2001;
Colin et al., 2013). n the other hand, signicant enlargement of
interfollicular blood vessels may be attributed to a direct inuence of
estrogen (Banerjee et al., 1997).
In the thyroids of Orx-MA + E rats, we also detected decreased
immunopositivity for NIS both in the membrane and cytoplasm. In
line with this, in vitro studies in FTRL-5 cells demonstrated that E
down-regulated gene expression of NIS (Furlanetto et al., 1999), while
in vivo studies showed that it either did not affect or suppressed NIS expression in males (Stanley et al., 2010). Decreased NIS means decreased
iodination, and this is why TgT4 was also decreased in Orx-MA + E in

comparison to Orx-MA group. Our data are in line with hypothesis that
E-mediated inhibition of iodine transport may be one of the reasons
why women have a higher incidence of goiter than men (Furlanetto
et al., 1999).
However, results on colloid volume, basolateral NIS- and colloid TgT4-immunopositivity in the thyroids of Orx-MA + E group were more
comparable to SO-MA group. Taken together, our histological data,
besides differential effect of T and E, indicate that the thyroid gland of
MA males has a great capacity to adapt to manipulations with sex steroids, being more sensitive to Orx and subsequent androgenization,
then to estrogenization. Banu et al. (2002) found that sex steroids
stimulated thyrocyte proliferation in immature and young adult
rats in a gender specic manner: T stimulated thyrocyte proliferation
in males and immature females, while E had stimulatory effect in females but inhibited proliferation in male (immature and young
adult) rats. More recent research on human thyroid cancer tissue
showed that AR is posttrancriptionaly regulated by miR-124a, also

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

97

obtained changes did not signicantly affect pituitary TSH IFC expression or concentration of TSH in sera. In the thyroid, we did not nd histopathological changes. Structural and ultrastructural analyses, together
with IHC and IFC studies, indicate mild suppression of thyroid functioning, without the expected accompanying effect on concentration of TH
in sera or liver Dio1 enzyme activity. The obtained data are original
and indicate differential effect of E in comparison to T.
Declaration of Interest
The authors declare that there is no conict of interest.
Fig. 12. Enzyme activity of type 1 iodothyronine deiodinase (Dio1) in the liver of control
sham-operated (SO), orchidectomized (Orx), testosterone (Orx + T)- and estradiol
(Orx + E)- treated Orx middle-aged (MA) rats; enzyme activities were determined
using the nonradioactive iodide-release assay. The values are means SD (n = 5); statistics: one way ANOVA, Tukey's multiple comparisons post hoc test, p b 0.05 versus Orx-MA
or SO-MA rats.

in vitro demonstrated to be a potent inhibitor of AR that impairs cell

proliferation even in the presence of testosterone. We detected thyroid hyperplasia in 2 of 8 examined MA female rats upon administration of 10 higher dose of E during two weeks (oi-Jurjevi et al.,
2005). Therefore, it seems logical to assume that E- and T- mediated
signaling have a differential role in male and female thyroid tissue,
which may contribute to a well-known gender bias in thyroid cancer
incidence.
The unaltered T4 and T3 in serum of T-treated rats of both age (OrxYA and Orx-MA) can be at least in part explained by T-stimulated
increase of metabolic degradation of thyroid hormones in the liver, in
line with detected increase in liver Dio1 enzyme. Liver Dio1 enzyme, besides outer ring deiodination activity, which contributes to peripheral
bio-activation of T4 to T3, also has inner ring deiodination activity, and
shows high afnity towards sulfated thyroid hormones (Schneider
et al., 2006).
T3 is considered a major regulatory factor of Dio1 enzyme expression
and activity (Khrle, 2002), but sex steroids also contribute to its regulation (Miyashita et al., 1995; da Costa et al., 2001). Orchidectomy and
subsequent T-treatment of young adult rats affected liver Dio 1 activity
(Lisba et al., 2001) in the same direction as we demonstrated in the
present study for MA rats. In line with Silvestri et al. (2008), we also detected decreasing trend in Dio1 enzyme expression and activity with
advancing age in male rats. However, this decrease in Dio1 activity became signicant only in old-aged males (24-month-old rats; results
not shown). Our previous (oi-Jurjevi et al., 2012) and current data
on liver Dio1 enzyme activity, strongly indicate sexual dimorphism
and preserved sensitivity to T (and not to E) in direct regulation of
this enzyme in MA males.
5. Conclusion
In summary, we showed that systemic androgenization of Orx-MA
moderately stimulated pituitary-thyroid axis. The effects of stimulation
were visible at the level of: (i) pituitary, as increased TSH RIF; (ii) thyroid, as increased IHC and IFC expression of relevant thyroid-specic
proteins, in line with the corresponding structural and ultrastructural
changes in thyroid tissue; and (iii) in the liver, as increased activity of
Dio1 enzyme. However, serum TSH and TH levels were unchanged,
while the same dose of T increased serum TSH in Orx-YA model. The obtained data are in line with our initial hypothesis that the pituitarythyroid axis response to T is lower in MA in comparison to YA rats,
and point to altered pituitary TSH regulation as the center of this
change. Subcutaneous administration of estradiol to Orx-MA males
exerted the adverse effect by increasing pituitary weight and inducing
hypertrophy and hyperplasia of pituitary lactotrophs. However, the

Acknowledgments
This research was supported by grants from the Ministry of Education and Science of the Republic of Serbia (No. 173009) and the
Deutsche Forschungsgemeinschaft (DFG-GK 1208, TP3, RE3038/1-1),
Charit Universittsmedizin Berlin. B. . Jurjevi was supported by
European Society for Endocrinology (ESE) Short-Term Fellowship. The
authors would like to thank Mr. Kristijan Jurjevic for his assistance
with English proofreading.
References
de Araujo, L.F., Grozovsky, R., dos Santos Pereira, M.J., de Carvalho, J.J., Vaisman, M.,
Carvalho, D.P., 2010. Expressions of vascular endothelial growth factor and nitric
oxide synthase III in the thyroid gland of ovariectomized rats are upregulated by
estrogen and selective estrogen receptor. Thyroid 20, 8592.
Banerjee, S.K., Sarkar, D.K., Weston, A.P., De, A., Campbell, D.R., 1997. Over expression of
vascular endothelial growth factor and its receptor during the development of
estrogen-induced rat pituitary tumors may mediate estrogen-initiated tumor angiogenesis. Carcinogenesis 18, 11551161.
Banu, K.S., Govindarajulu, P., Aruldhas, M.M., 2001. Testosterone and estradiol modulate
TSH-binding in the thyrocytes of wistar rats: inuence of age and sex. J. Steroid
Biochem. Mol. Biol. 78, 329342.
Banu, S.K., Govindarajulu, P., Aruldhas, M.M., 2002. Testosterone and estradiol upregulate androgen and estrogen receptors in immature and adult rat thyroid glands
in vivo. Steroids 67, 10071014.
Boado, R., Deza, S., Zaninovich, A.A., 1985. Effects of oestradiol benzoate on the pituitary
secretion and peripheral kinetics of thyrotrophin in the rat. Acta Endocrinol.
(Copenh) 109, 232236.
Bogataj, M., Kalinik, M., Poredo, P., 1977. A simplied indirect method of the average
shell thickness of the hollow spheres from samples insections. Folia Anat. Yugosl. 6,
4549.
Chen, H.J., 1984. Age and sex difference in serum and pituitary thyrotropin concentrations
in the rat: inuence by pituitary adenoma. Exp. Gerontol. 19, 16.
Cizza, G., Brady, L.S., Calogero, A.E., Bagdy, G., Lynn, A.B., Kling, M.A., Blackman, M.R.,
Chrousos, G.P., Gold, P.W., 1992. Central hypothyroidism is associated with advanced
age in male Fischer 344/N rats: in vivo and in vitro studies. Endocrinology 131,
26722680.
Colin, I.M., Denef, J.F., Lengel, B., Many, M.C., Grard, A.C., 2013. Recent insights into the
cell biology of thyroid angiofollicular units. Endocr. Rev. 34, 209238.
da Costa, V.M., Moreira, D.G., Rosenthal, D., 2001. Thyroid function and aging: genderrelated differences. J. Endocrinol. 171, 193198.
Cox, R.L., Crawford, E.D., 1995. Estrogens in the treatment of prostate cancer. J. Urol. 154,
19911998.
Denef, C., 2008. Paracrinicity: the story of 30 years of cellular pituitary crosstalk.
J. Neuroendocrinol. 20, 170.
Donda, A., Reymond, M.J., Lemarchand-Braud, T., 1989. Inuence of age on the control of
thyrotropin secretion by thyrotropin-releasing hormone in the male rat. Neuroendocrinology 49, 389394.
Donda, A., Reymond, M.J., Zrich, M.G., Lemarchand-Braud, T., 1987. Inuence of sex and
age on T3 receptors and T3 concentration in the pituitary gland of the rat: consequences on TSH secretion. Mol. Cell. Endocrinol. 54, 2934.
Eriksson, S., Eriksson, A., Stege, R., Carlstrm, K., 1995. Bone mineral density in patients
with prostatic cancer treated with orchidectomy and with estrogens. Calcif. Tissue
Int. 57, 9799.
Faggiano, A., Caillou, B., Lacroix, L., Talbot, M., Filetti, S., Bidart, J.M., Schlumberger, M.,
2007. Functional characterization of human thyroid tissue with immunohistochemistry. Thyroid 17, 203211.
Farbota, L., Hofmann, C., Oslapas, R., Paloyan, E., 1987. Sex hormone modulation of serum
TSH levels. Surgery 102, 10811087.
Filipovi, B., Soi-Jurjevi, B., Ajdanovi, V., Panteli, J., Nestorovi, N., Miloevi, V.,
Sekuli, M., 2013. The effects of sex steroids on thyroid C cells and trabecular bone
structure in the rat model of male osteoporosis. J. Anat. 222, 313320.
Furlanetto, T.W., Nguyen, L.Q., Jameson, J.L., 1999. Estradiol increases proliferation and
down-regulates the sodium/iodide symporter gene in FRTL-5 cells. Endocrinology
140, 57055711.

98

B. oi-Jurjevi et al. / Experimental Gerontology 72 (2015) 8598

Grard, A.C., Daumerie, C., Mestdagh, C., Gohy, S., De Burbure, C., Costagliola, S., Miot, F.,
Nollevaux, M.C., Denef, J.F., Rahier, J., Franc, B., De Vijlder, J.J., Colin, I.M., Many, M.C.,
2003. Correlation between the loss of thyroglobulin iodination and the expression
of thyroid-specic proteins involved in iodine metabolism in thyroid carcinomas.
J. Clin. Endocrinol. Metab. 88, 49774983.
Hedlund, P.O., Damber, J.E., Hagerman, I., Haukaas, S., Henriksson, P., Iversen, P.,
Johansson, R., Klarskov, P., Lundbeck, F., Rasmussen, F., Varenhorst, E., Viitanen, J.,
2008. SPCG-5 study group. Parenteral estrogen versus combined androgen deprivation in the treatment of metastatic prostatic cancer: part 2. Final evaluation of the
Scandinavian prostatic cancer group (SPCG) study no. 5. Scand. J. Urol. Nephrol. 42,
220229.
Ishida, M., Mitsui, T., Yamakawa, K., Sugiyama, N., Takahashi, W., Shimura, H., Endo, T.,
Kobayashi, T., Arita, J., 2007. Involvement of cAMP response element-binding protein
in the regulation of cell proliferation and the prolactin promoter of lactotrophs in primary culture. Am. J. Physiol. Endocrinol. Metab. 293, E1529E1537.
Isidori, A.M., Balercia, G., Calogero, A.E., Corona, G., Ferlin, A., Francavilla, S., Santi, D.,
Maggi, M., 2015. Outcomes of androgen replacement therapy in adult male
hypogonadism: recommendations from the Italian society of endocrinology.
J. Endocrinol. Investig. 38, 103112.
Kalinik, M., Vraspir-Porenta, O., Kham-Lindtner, T., Logonder-Mlinsek, M., Pajer, Z.,
Stiblar-Martincic, D., Zorc-Pleskovic, R., Trobina, M., 1988. The interdependence of
the follicular, parafollicular, and mast cells in the mammalian thyroid gland: a review
and a synthesis. Am. J. Anat. 183, 148157.
Klotz, L., 2015. Testosterone therapy and prostate cancer-safety concerns are well
founded. Nat. Rev. Urol. 12, 4854.
Khrle, J., 2002. Iodothyronine deiodinases. Methods 347, 125167.
La Vecchia, C., Ron, E., Franceschi, S., Dal Maso, L., Mark, S.D., Chatenoud, L., Braga, C., Preston-Martin, S., McTiernan, A., Kolonel, L., Mabuchi, K., Jin, F., Wingren, G., Galanti,
M.R., Hallquist, A., Lund, E., Levi, F., Linos, D., Negri, E., 1999. A pooled analysis of
casecontrol studies of thyroid cancer. III. Oral contraceptives, menopausal replacement therapy and other female hormones. Cancer Causes Control 10, 157166.
Lee, M.L., Chen, G.G., Vlantis, A.C., Tse, G.M., Leung, B.C., van Hasselt, C.A., 2005. Induction
of thyroid papillary carcinoma cell proliferation by estrogen is associated with an altered expression of Bcl-xL. Cancer J. 11, 113121.
Lipton, A., Smith, M.R., Ellis, G.K., Goessl, C., 2012. Treatment-induced bone loss and fractures in cancer patients undergoing hormone ablation therapy: efcacy and safety of
denosumab. Clin. Med. Insights. Oncol. 6, 287303.
Lisba, P.C., Curty, F.H., Moreira, R.M., Oliveira, K.J., Pazos-Moura, C.C., 2001. Sex steroids
modulate rat anterior pituitary and liver iodothyronine deiodinase activities. Horm.
Metab. Res. 33, 532535.
Martinoli, M.G., Ouellet, J., Rhaume, E., Pelletier, G., 1991. Growth hormone and somatostatin gene expression in adult and aging rats as measured by quantitative in situ
hybridization. Neuroendocrinology 54, 607615.
Miler, M., Soi-Jurjevi, B., Nestorovi, N., Risti, N., Medigovi, I., Savin, S., Miloevi, V.,
2014. Morphological and functional changes in pituitary-thyroid axis following
prolonged exposure of female rats to constant light. J. Morphol. 275, 11611172.
Miyashita, K., Murakami, M., Iriuchijima, T., Takeuchi, T., Mori, M., 1995. Regulation of rat
liver type 1 iodothyronine deiodinase mRNA levels by testosterone. Mol. Cell.
Endocrinol. 115, 161167.
Novelli, A., 1959. Rapid staining of mitochondria. Experientia 15, 274.
Oh, W.K., 2002. The evolving role of estrogen therapy in prostate cancer. Clin. Prostate
Cancer 1, 8189.
Pelletier, G., 2000. Localization of androgen and estrogen receptors in rat and primate
tissues. Histol. Histopathol. 15, 12611270.

Renko, K., Hoeg, C.S., Hiller, F., Schomburg, L., Khrle, J., 2012. Identication of iopanoic
acid as substrate of type 1 deiodinase by a novel nonradioactive iodide-release
assay. Endocrinology 153, 25062513.
Ross, D.S., 1990. Testosterone increases TSH-beta mRNA, and modulates alpha-subunit
mRNA differentially in mouse thyrotropic tumor and castrate rat pituitary. Horm.
Metab. Res. 22, 163169.
Rukhman, N., Silverberg, A., 2011. Thyroid cancer in older men. Aging Male 14, 9198.
Samaras, N., Samaras, D., Lang, P.O., Forster, A., Pichard, C., Frangos, E., Meyer, P., 2013. A
view of geriatrics through hormones. What is the relation between andropause and
well-known geriatric syndromes? Maturitas 74, 213219.
Schneider, M.J., Fiering, S.N., Thai, B., Wu, S.Y., St Germain, E., Parlow, A.F., St Germain, D.L.,
Galton, V.A., 2006. Targeted disruption of the type 1 selenodeiodinase gene (Dio 1) results in marked changes in thyroid hormone economy in mice. Endocrinology 147,
580589.
Schomburg, L., Riese, C., Renko, K., Schweizer, U., 2007. Effect of age on sexually dimorphic
selenoprotein expression in mice. Biol. Chem. 388, 10351041.
Shukuwa, K., Izumi, S., Hishikawa, Y., Ejima, K., Inoue, S., Muramatsu, M., Ouchi, Y., Kitaoka,
T., Koji, T., 2006. Diethylstilbestrol increases the density of prolactin cells in male
mouse pituitary by inducing proliferation of prolactin cells and transdifferentiation
of gonadotropic cells. Histochem. Cell Biol. 126, 111123.
Silvestri, E., Lombardi, A., de Lange, P., Schiavo, L., Lanni, A., Goglia, F., Visser, T.J., Moreno,
M., 2008. Age-related changes in renal and hepatic cellular mechanisms associated
with variations in rat serum thyroid hormone levels. Am. J. Physiol. Endocrinol.
Metab. 294, E1160E1168.
oi-Jurjevi, B., Filipovi, B., Milosevi, V., Nestorovi, N., Manojlovi-Stojanoski, M.,
Brki, B., Sekuli, M., 2005. Chronic estradiol exposure modulates thyroid structure
and decreases T4 and T3 serum levels in middle-aged female rats. Horm. Res. 63,
4854.
oi-Jurjevi, B., Filipovi, B., Renko, K., Ajdanovi, V., Manojlovi-Stojanoski, M.,
Miloevi, V., Khrle, J., 2012. Orchidectomy of middle-aged rats decreases
liver deiodinase 1 and pituitary deiodinase 2 activity. J. Endocrinol. 215,
247256.
oi-Jurjevi, B., Filipovi, B., Wirth, E.K., ivanovi, J., Radulovi, N., Jankovi, S.,
Miloevi, V., Khrle, J., 2014. Soy isoavones interfere with thyroid hormone
homeostasis in orchidectomized middle-aged rats. Toxicol. Appl. Pharmacol.
278, 124134.
Stanley, J.A., Aruldhas, M.M., Chandrasekaran, M., Neelamohan, R., Suthagar, E.,
Annapoorna, K., Sharmila, S., Jayakumar, J., Jayaraman, G., Srinivasan, N., Banu, S.K.,
2012. Androgen receptor expression in human thyroid cancer tissues: a potential
mechanism underlying the gender bias in the incidence of thyroid cancers.
J. Steroid Biochem. Mol. Biol. 130, 105124.
Stanley, J.A., Aruldhas, M.M., Yuvaraju, P.B., Banu, S.K., Anbalagan, J., Neelamohan, R.,
Annapoorna, K., Jayaraman, G., 2010. Is gender difference in postnatal thyroid growth
associated with specic expression patterns of androgen and estrogen receptors?
Steroids 75, 10581066.
Weibel, E.R., 1979. Practical methods for biological morphometry. Stereological Methods.
Academic Press, London, pp. 1415.
Woodward, C.J., 1993. A re-evaluation of the anabolic effect of testosterone in rats: interactions with gonadectomy, adrenalectomy and hypophysectomy. Acta Endocrinol.
(Copenh) 128, 473477.
Zhang, L.J., Xiong, Y., Nilubol, N., He, M., Bommareddi, S., Zhu, X., Jia, L., Xiao, Z., Park, J.W.,
Xu, X., Patel, D., Willingham, M.C., Cheng, S.Y., Kebebew, E., 2015. Testosterone regulates thyroid cancer progression by modifying tumor suppressor genes and tumor
immunity. Carcinogenesis 36, 420428.