1

Thyroid Hormone Interacts with the Sympathetic Nervous 1

System to Modulate Bone Mass and Structure in Young-Adult 2

Mice 3

Tatiana L. Fonseca1, Marilia B.C.G. Teixeira1, Manuela Rodrigues-Miranda1, 4

Marcos V Silva1, Gisele M. Martins1, Cristiane C. Costa1, Danielle Y. Arita2, 5

Juliana D Perez2, Dulce E. Casarini2, Patricia C. Brum3 and Cecilia H.A. 6

Gouveia1 7

8

1University of São Paulo, Department of Anatomy, Institute of Biomedical 9

Sciences, Av. Prof Lineu Prestes, 2415, São Paulo, Brazil, 05508-000. 10

2Federal University of São Paulo School of Medicine, Department of Internal 11

Medicine, Renal Division, Rua Botucatu, 740, São Paulo, Brazil, 04023-900. 12

3University of São Paulo, School of Physical Education and Sport, Av. Prof. 13

Mello Moraes, 65, São Paulo, Brazil, 05508-030. 14

15

Running title: TH-SNS interaction modulates bone structure. 16

17

Corresponding author: 18

Cecilia H. A. Gouveia, PhD 19

Assistant Professor 20

Department of Anatomy 21

Institute of Biomedical Sciences 22

University of Sao Paulo 23

Av. Prof. Lineu Prestes, 2415, Sao Paulo, SP, Brazil, 05508-900 24

Phone: 55-11-3091-7737, FAX: 55-11-3091-7366 25

e-mail: cgouveia@usp.br 26

Articles in PresS. Am J Physiol Endocrinol Metab (July 8, 2014). doi:10.1152/ajpendo.00643.2013

Copyright © 2014 by the American Physiological Society.

2

ABSTRACT 27

To investigate whether thyroid hormone (TH) interacts with the 28

sympathetic nervous system (SNS) to modulate bone mass and structure, we 29

studied the effects of daily T3 treatment, in a supraphysiological-dose, for 12 30

weeks, on the bone of young-adult mice with chronic sympathetic hyperactivity 31

owing to double-gene disruption of adrenoceptors that negatively regulate 32

norepinephrine release, 2A-AR and 2C-AR (2A/2C-AR-/- mice). As expected, T3 33

treatment caused a generalized decrease in the areal bone mineral density 34

(aBMD) of WT mice (determined by DXA), followed by deleterious effects on the 35

trabecular and cortical bone microstructural parameters (determined by µCT) of 36

the femur and vertebra, and on the biomechanical properties (maximum load, 37

ultimate load and stiffness) of the femur. Surprisingly, α2A/2C-AR-/- mice were 38

resistant to most of these T3-induced negative effects. Interestingly, the mRNA 39

expression of osteoprotegerin, a protein that limits osteoclast activity, was 40

upregulated and downregulated by T3 in the bone of α2A/2C-AR-/- and WT mice, 41

respectively. β1-AR mRNA expression and IGF-1 serum levels, which exert 42

bone anabolic effects, were increased by T3 treatment only in α2A/2C-AR-/- mice. 43

As expected, T3 inhibited the cell growth of calvaria-derived osteoblasts 44

isolated from WT mice, but this effect was abolished or reverted in cells isolated 45

from KO mice. Collectively, these findings support the hypothesis of a TH-SNS 46

interaction to control bone mass and structure of young-adult mice, and 47

suggests that this interaction may involve α2-AR signaling. Finally, the present 48

findings offer new insights into the mechanisms through which TH regulates 49

bone mass, structure and physiology. 50

3

Key words: thyroid hormone, sympathetic nervous system, bone, α-51

adrenoceptors, β-adrenoceptors.52

4

INTRODUCTION 53

Thyroid hormone (TH) is essential for normal bone development, 54

maturation and metabolism. Hypothyroidism retards bone growth and slows 55

bone turnover (3), whereas thyrotoxicosis is frequently associated with 56

accelerated bone development and metabolism and decreased bone mass (43). 57

Histomorphometric studies indicate that thyrotoxicosis increases both 58

osteoblastic and osteoclastic activities, but the latter predominates. As a result, 59

bone turnover is accelerated, favoring bone resorption, a negative balance of 60

calcium, and bone loss (26, 41). Although the negative effects of overt 61

thyrotoxicosis to bone mass and turnover are well known, the mechanisms by 62

which TH promotes its effects are less clear. Thyroid hormone can affect the 63

skeleton indirectly through changes in other hormones, e.g., growth hormone 64

and IGF-I (31, 33), but in vitro studies show that T3 also acts directly in bone 65

cells, modifying their proliferation, differentiation and/or expression of several 66

bone-related genes (27, 39, 69, 70). It is generally accepted that most T3 67

actions are mediated by its nuclear receptors (TRs), which were shown to be 68

expressed in osteoblasts (70), osteoclasts (1), and chondroblasts (4). 69

Recently, the sympathetic nervous system (SNS) was identified as an 70

important and potent regulator of bone metabolism (17, 61). A series of studies 71

suggest that the SNS negatively regulates bone formation and positively 72

regulates bone resorption via 2-adrenoceptor (2-AR) signaling, which is 73

expressed in osteoblasts. Administration of propranolol, a beta blocker, and 74

isoproterenol, a beta agonist, were demonstrated to increase and decrease 75

bone mass in adult animals, respectively (8, 10, 60). The role of the SNS in 76

controlling bone mass was supported by a high-bone-mass (HBM) phenotype in 77

5

an animal model of low SNS activity, a mouse deficient for dopamine -78

hydroxylase (DbH–/–), the step-limiting enzyme responsible for catecholamine 79

synthesis (10, 46). A more precise role of -adrenergic signaling on bone mass 80

was supported by the analysis of mice with global gene deletion of 2-AR (2-81

AR–/–). These animals do not exhibit endocrine and metabolic abnormalities, 82

have a normal body weight and exhibit a HBM phenotype beginning at 6 83

months of age due to an increase in bone formation and a decrease in bone 84

resorption (18). 85

However, a recent study by our group demonstrated that young adult 86

mice with global double gene inactivation of 2A and 2C adrenoceptors (2A/2C -87

AR-/-) also exhibit a phenotype of HBM, that becomes more evident as the 88

animals age (from 40 to 120 days of age), despite exhibiting chronic 89

sympathetic hyperactivity and intact 2-AR (19). The α2-adrenergic receptors 90

(α2-ARs) were initially characterized as presynaptic autoreceptors that 91

negatively regulate catecholamine release (13, 29). Later, it was shown that 2-92

ARs are not restricted to presynaptic sites but can also have postsynaptic 93

locations and functions, which include the regulation of body temperature, 94

intraocular pressure, lipolysis, and pain perception (12, 28, 47). Three subtypes 95

of α2-ARs have been identified: α2A-AR, α2B-AR and α2C-AR (15, 32, 47). Hein et 96

al. reported that the double blockage of α2A-AR and α2C-ARs in mice results in 97

increased serum levels of norepinephrine (NE) (29). In addition to the HBM 98

phenotype, we found that α2A/2C-AR-/- mice exhibit increased bone formation and 99

decreased bone resorption (19). Consistent with a previous study that identified 100

the presence of 2-AR isoforms in human-derived bone cells (61), we 101

demonstrated the mRNA expression of 2A-, 2B- and 2C-AR in the bone of 102

6

mice and in osteoblast-like cells derived from the mouse calvarium (MC3T3-E1 103

cells). By immunohistochemistry, we found that 2A-, 2B-, 2C-, and 2-AR are 104

expressed in osteoblasts, osteocytes, osteoclasts, and chondrocytes from fetal 105

and adult mice (19). Collectively, these findings raise the hypothesis that 2-AR 106

may also mediate SNS signaling in the skeleton. 107

There is evidence that TH interacts with the SNS to control several 108

physiological processes. It is known that this interaction is required for maximal 109

thermogenesis, lipolysis and glycogenolysis (45). In addition, many of the 110

clinical features of hyperthyroidism, such as palpitations, tremor, sweating, 111

nervousness, weight loss, heat intolerance, stare, tachycardia, and increased 112

pulse pressure, mimic the manifestations of excess -adrenergic activity and 113

indicate that thyrotoxicosis results in a state of hyperadrenergic stimulation (22, 114

49). -adrenergic blockage profoundly modifies the severity of these symptoms, 115

and their use in the management of hyperthyroidism is well established (34, 35, 116

50, 63, 64). 117

Considering the negative effects of thyrotoxicosis and SNS activation on 118

bone mass, we raised the hypothesis that TH may also interact with the SNS to 119

regulate bone metabolism. A piece of evidence supporting this possible 120

interaction is the fact that the treatment of hyperthyroid patients with propranolol 121

corrects their hypercalcemia (53) and decreases their urinary excretion of 122

hydroxyproline, a biochemical marker of bone resorption (6). 123

The aim of the present study was to investigate whether TH interacts with 124

the SNS to regulate bone mass and structure during the young-adulthood of 125

mice, a phase of important bone mass accrual and growth. We observed that 126

7

young-adult α2A/2C-AR-/- mice are resistant to the deleterious effects of 127

thyrotoxicosis on bone. These findings strongly suggest that adrenergic 128

signaling modulates TH actions in the bone. 129

MATERIALS AND METHODS 130

Animal maintenance and manipulation. A cohort of female congenic α2A/2C-131

AR KO mice in a C57BL/6J (B6) background and their wild-type controls (WT) 132

were studied from 40 to 123 days of age, a phase of important bone mass 133

accrual and growth (25). The animals were considered young-adults, since the 134

pubertal maturation in B6 female mice begins when serum estradiol increases, 135

by day 26 after birth, and is complete when vaginal opening occurs, by day 31 136

(2). Animals received intraperitoneal injections of T3 (3,3´,5-Triiodo-L-thyronine; 137

Sigma-Aldrich, Munich, Germany) in a daily dose that corresponds to 138

approximately 10 times its physiological dose (10xT3 = 3.5 µg/100 g BW/day) 139

and that is known to induce bone loss (21). T3 injections were administered 140

every day, at the same time each day. Animals (n = 7 animals/group) were 141

treated for 12 weeks. Body weight was measured every week, and the amount 142

of hormone administered was adjusted according to the changes in body weight 143

to maintain the proper dosage. All experimental procedures were performed in 144

accordance with the guidelines of the Standing Committee on Animal Research 145

of the University of Sao Paulo (protocol # 065-17/02). 146

Serum levels of T3, T4 and IGF-I. At the end of the experimental period, the 147

animals were sacrificed, and the blood was collected. Serum levels of total T4 148

and T3 were measured using radioimmunoassay (RIE) commercial kits (RIA-149

gnost T4 and RIA-gnost T3, CIS bio international- Schering, Paris, France). 150

8

Serum IGF-I levels were measured by RIA (Diagnostic system laboratories, 151

Webster, USA). 152

Femur and Plasma levels of Cathecholamines. For cathecolamine extraction, 153

the two whole femurs of each animal were crushed in a steel mortar and pestle 154

set (Fisher Scientific International, Inc, Hampton, NH) precooled in dry ice. The 155

crushed bones were weighed and homogenized in 1.5 ml of cold 0.1 M 156

perchloric acid containing 0.02% sodium pyrosulfite (Na2S2O5) and 0.7 nM 3,4-157

dihydroxybenzylamine hydromide (DHBA) as an internal standard. The 158

homogenates were kept at 4°C overnight and centrifuged at 10,000 rpm for 50 159

min at 4°C. The supernatants were stored at -80°C until they were assayed for 160

norepinephrine (NE) and epinephrine (EP). Femur and plasma levels of 161

cathecholamines were measured by HPLC using ion-pair reverse-phase 162

chromatography coupled with electrochemical detection (0.5 V), as described 163

previously (40, 42). 164

Determination of Areal BMD by DXA. Areal BMD (aBMD) was measured 165

using the pDXA Sabre Bone Densitometer and the pDXA Sabre Software 166

version 3.9.4 (Norland Medical Systems, Inc., Fort Atkinson, WI, U.S.A.), both 167

especially designed for small animals. The research mode scan option was 168

used for the measurements. Pixel spacing for the scan was set to 0.5 × 0.5 mm 169

and the scan speed to 4 mm/s. To limit the scan area, which allows the scans to 170

be performed in a higher resolution mode, the scans were performed from the 171

first lumbar vertebra to the hind limbs. The regions of interest (ROIs) for 172

analysis were the (1) hind body (HB), which includes the lumbar spine, pelvic 173

bones and hind limbs; (2) lumbar vertebrae (L1–L6); (3) both femurs; and (4) 174

both tibias. Taking into account the possibility of bone mass differences 175

9

between the left and right limbs as a result of functional bilateral asymmetry (20) 176

the aBMD of femur and tibia were expressed as the mean of the left and right 177

limbs for each animal. For the scans, the animals were anesthetized with a 178

ketamine and xylazine cocktail (10 and 30 mg/kg BW) and scanned in the prone 179

position. The animals were subjected to a basal scan and to scans after 4 and 180

12 weeks of treatment. 181

Micro-CT (CT) analysis. The micro-architecture of the femur and lumbar 182

vertebra (L6) was investigated using a micro-computed tomography unit 183

(SkyScan 1172; SkyScan, Aartselaar, Belgium), in which the samples were 184

rotated through 360° at a rotation step of 0.5°. The x-ray settings were 185

standardized to 100 kV for the baseline bone specimens, with an exposure time 186

of 590 (ms). A 0.5-mm-thick aluminum filter and a beam-hardening algorithm 187

were used to minimize beam-hardening artifacts. The registered data sets were 188

segmented into binary images. The two- and three-dimensional (2D and 3D) 189

measurements were obtained with the CTAn version 1.5 (SkyScan). The distal 190

metaphysis of the femur and the vertebral body of L6 were selected as the 191

regions of interest (ROI) for the trabecular bone analysis. The following 3D 192

structural parameters were determined from approximately 180 and 210 193

transverse femoral and vertebral Ct slices, respectively: trabecular bone volume 194

(BV/TV), trabecular thickness (Tb.Th), trabecular number (Tb.N), trabecular 195

separation (Tb.Sp), trabecular porosity (Tb.Po), structure model index (SMI), 196

trabecular pattern factor (Tb.PF) and volumetric bone mineral density (vBMD). 197

Morphometric variables of cortical bone were measured in the femoral diaphysis 198

and in the vertebral body (approximately 170 and 70 transverse CT slices), from 199

the distal limit of the pedicles until the distal growth plate. The cortical thickness 200

10

(Ct.Th), tissue area (T.Ar), cortical area (Ct.Ar), medullary area (Ma.Ar), cortical 201

bone volume (Ct.BV), and the polar moment of inertial (MMI) were measured in 202

2D cross-sectional images. The cortical porosity (Ct.Po) and vBMD were 203

measured in 3D. To guarantee that the same bone regions were measured in 204

every sample, the number of transverse CT slices per sample were calculated 205

accordingly to the respective bone length (femur or vertebra). 206

Three-point bending test. The right femurs were tested in an Instron testing 207

machine (Model 3344, Instron Corporation,Norwood, MA, USA). Fresh-frozen 208

bones were thawed to room temperature (22o C). The anterior cortex at the mid-209

diaphysis of the femur was placed in compression and the posterior cortex in 210

tension during the test. The lower support points were separated by an span of 211

50% of the femoral length (femur length/2). A constant displacement rate of 212

0,03 mm/s was applied until the bone fractured. Fracture was taken as the 213

complete loss of loadcarrying ability. To stabilize the specimen, a small preload 214

(5% of the average maximal load) was applied before actual testing. During the 215

bending test, load-displacement data were collected by a computerized data-216

acquisition system at a sampling rate of 80 Hz. The biomechanical properties 217

evaluated were the maximum load [a measure of the maximum force that the 218

bone withstood before fracture (N)], ultimate load [the load at the fracture point 219

(N)], resilience [a measure of the ability of a bone to suffer elastic deformity (J)], 220

Young's modulus (mPa) and stiffness (N/mm) [which are measures of the 221

extrinsic rigidity of the bone tissue]. 222

Real-time PCR. Expression of α2A-AR, α2C-AR, β2-AR, β1-AR, receptor activator 223

of nuclear factor κB (RANK), RANK ligand (RANK-L) and osteoprotegerin 224

(OPG) were determined by real-time PCR as previously described (14) in the 225

11

whole tibia. Briefly, the total RNA was extracted using Trizol (Invitrogen, 226

Calbard, CA, USA) following the manufacturer’s instructions. Total RNA was 227

reverse transcribed using RevertAid-H-Minus M-MuLV Reverse Transcriptase 228

(Fermentas, Hanover, MD, USA) to synthesize the first strand cDNA, used as a 229

template. SYBR® Green Super Mix (Applied Biosystems, Warrington, UK) was 230

used for the real-time PCR using the ABI Prism 7500 sequence detector 231

(Applied Biosystems, Foster City, CA, USA). All primers used in this study [α2A-232

AR = Forward (F): CGC AGG CCA TCG AGT ACA A and reverse (R): GAT 233

GAC CCA CAC GGT GAC AA (NM_007417.3); 2C-AR = F: CAT GGG CGT 234

GTT CGT ACT GT and R: CAG GCC TCA CGG CAG ATG (NM_007418.3); β1-235

AR= F: TCG TCC GTC GTC TCC TTC TAC and R: ACA CCC GCA GGT ACA 236

CGA A (NM_007419); β2-AR= F: GCC ACG ACA TCA CTC AGG AAC and R: 237

CGA TAA CCG CAC TGA GGA TGG (NM_007420); RANK= F: TCT GCA GCT 238

CTT CCA TGA CAC T and R: CGA TGA GAC TGG GCA GGT AAG 239

(NM_009399); RANKL= F: GGC CAC AGC GCT TCT CAG and R: GAG TGA 240

CTT TAT GGG AAC CCG AT (NM_011613.2); OPG= F: AGT CCG TGA AGC 241

AGG AGT G and R: CCA TCT GGA CAT TTT TTG CAA A (NM_U94331); 18 242

S= F: GTA ACC CGT TGA ACC CCA TT and R: CCA TCC AAT CGG TAG 243

TAG CG (M_11188); -actin= F: AAG ATT TGG ACC ACA CTT TCT ACA and 244

R:CGG TGA GCA GCA CAG GGT (NM_031144)] were synthesized (Integrated 245

DNA Technologies, Coralville, IA, USA) specifically for real-time PCR using the 246

Primer Express software (Applied Biosystems). All Ct values were normalized 247

using an internal control: 18s or -actin mRNA. Both internal controls were 248

validated for this study, showing to be stable (their expression did not vary due 249

to mice lineage or T3 treatment). Relative gene expression quantification was 250

12

assessed by the Ct method, as previously described by Livak, (36). The final 251

values for samples are reported as fold-induction relative to the expression of 252

the control, with the mean control value being arbitrarily set to 1. 253

Calvaria-derived osteoblastic cell growth. Primary mouse osteoblastic cells 254

were obtained by sequential enzyme digestion of excised calvarial bones from 255

neonatal WT and α2A/2C-AR-/- mice, using 1% trypsin and 1% collagenase in α-256

MEM (Gibco BRL, Paesley, UK). Osteoblasts were initially seeded at a density 257

of 104 cells/well in 24-well plates and cultured in growth media containing α-258

MEM (Invitrogen) supplemented with 10% FBS (Gibco) and 1% 259

penicillin/streptomycin (Gibco). After 72 hours (day 0), the culture medium was 260

changed to the differentiation medium, growth media supplemented with 50 261

μg/ml ascorbic acid (Sigma, Saint Louis, MO, USA) and 10 mM β-262

glycerophosphate (Sigma), and the treatment with T3 (10-8 M) and/or clonidine 263

(CLO, 10-6 M) was initiated. Medium was changed every three days. The cells 264

were collected on day 0 and after 3, 6, and 9 days of treatment, to determine 265

the cell number and viability by direct counting with a Neubauer Chamber 266

(Improve Neubaur, Germany). Cell viability was determined by the Trypan blue 267

exclusion method, with cells suspended in 0.4% Trypan blue dye (1:2 dilution, 268

Gibco-BRL, Grand Island, NY, USA). Cells were assayed in quadruplicates. 269

Statistical analysis. Results are presented as the mean ± standard error of the 270

mean (SEM). Two-way analysis of variance was used to compare TH effects in 271

WT and KO mice and cells, and was always followed by the Tukey’s test to 272

detect differences between groups. The unpaired Student’s t-test was used for 273

pairwise comparisons of the groups. For all tests, significance was defined as 274

13

p<0.05. For statistical analyses, we used the GraphPad Instat software package 275

(GraphPad Software Inc., San Diego, CA, USA). 276

RESULTS 277

Sympathetic hyperactivity and induction of a thyrotoxic state in 278

α2A/2C-AR-/- mice. We evaluated the effect of daily administration of 10 times the 279

physiological dose of T3 for 12 weeks on bone of WT and α2A/2C-AR-/- mice, 280

which is an animal model of chronic elevated sympathetic tone. T3 treatment 281

increased the serum concentration of T3 19-fold and, consequently, decreased 282

T4 serum levels 10- and 11-fold in WT and KO mice, respectively (Table 1), 283

which reflects TSH inhibition by negative feedback (68). Serum T3 levels were 284

not different between T3-treated and untreated WT and KO mice; however, 285

untreated KO mice presented 31% higher serum T4 levels than untreated WT 286

mice. As expected, we observed that plasma levels of NE and EP were higher 287

in α2A/2C-AR-/- mice compared with WT mice. (Fig. 1A-B). Accordingly, the femur 288

concentration of NE was 96% higher in the bones of KO mice compared with 289

WT mice (Fig. 1C), whereas EP bone concentration was not different between 290

WT and KO mice. Interestingly, T3 treatment decreased NE and EP plasma 291

levels (by 37% and 62%, respectively) only in WT mice, whereas T3 decreased 292

NE concentration (by 53 %) only in the bones of α2A/2C-AR-/- mice (Fig. 1C), 293

bringing the bone NE concentration of KO mice to the level of WT mice (KO+T3 294

= WT). Corroborating previous studies that reported cardiac hypertrophy in 295

conditions of thyrotoxicosis, T3 treatment significantly and equally increased 296

heart mass in WT (WT= 1,07 ± 0.02 vs. WT+T3= 1.35 ± 0.11, p<0.05) and KO 297

(KO= 1.16 ± 0.02 vs. KO+T3= 1.46 ± 0.01; p<0.05) animals, which confirms the 298

toxicity of the regimen of T3 treatment used in this study. Accordingly to 299

14

previous studies that applied the same dose (66) or 4-fold (57) the dose of T3 300

used in the present study, body weight and body length were not affected by TH 301

administration in WT and KO animals. However, BW was significantly higher 302

(about 13%) in KO mice during the whole experimental period (p<0.01). 303

Thyroid hormone effects on serum levels of IGF-1. Since there is 304

evidence that the growth hormone/IGF-1 axis may mediate the adrenergic 305

signaling (9, 48) and TH actions in bone (31, 33), we measured serum levels of 306

IGF-1, which were significantly increased (24 %) by T3 treatment only in α2A/2C-307

AR-/- mice, and not in WT mice (Table 1). 308

Thyroid hormone effects on the aBMD of α2A/2C-AR-/- mice. As 309

predicted, thyrotoxicosis was detrimental to the aBMD of WT mice, significantly 310

impairing bone mass acquisition in all skeletal sites evaluated (Fig. 2). The 311

aBMD of the hind body (lumbar spine, pelvis and hind limbs), lumbar vertebrae 312

(L1-L6), femur and tibia of WT mice treated with T3 for 12 weeks was 11%, 313

14%, 13% and 9% lower in comparison with euthyroid WT mice, respectively. 314

Surprisingly, the bones of α2A/2C-AR-/- mice were found to be resistant to the 315

osteopenic effects of TH excess, as the supraphysiological dose of T3 had no 316

effect on the aBMD of KO mice. This critical finding strongly suggests that TH 317

interacts with the SNS to regulated bone mass. 318

Thyroid hormone effects on the bone microarchitecture of α2A/2C-AR-319

/- mice. The µCT data of the trabecular bone structure of the distal methaphysis 320

of the femur (Fig. 3A-H) support the DXA findings. T3 treatment significantly 321

decreased BV/TV (72%), Tb.N (62%), Tb.Th (19%) and vBMD (8%) and 322

increased Tb.Sp (68%), Tb.Po (9%) and Tb.Pf (77%) in WT mice. On the other 323

hand, these parameters were not affected by T3 treatment in α2A/2C-AR-/- mice, 324

15

except Tb.N (Fig. 3B), which was also decreased by T3 in KO mice (30%). The 325

comparison of WT versus KO mice, shows higher femoral Tb.N in α2A/2C-AR-/- 326

animals (Fig. 3B). Considering the T3 treated animals, we found higher BV/TV, 327

Tb.N, Tb.Th and vBMD and lower Tb.Sp, Tb.Po and Tb.Pf in KO versus WT 328

mice, which strengthens the resistance of the femoral trabecular bone of KO 329

mice to the deleterious effects of thyrotoxicosis. The trabecular bone of the 330

vertebra was less sensitive to the TH effects (Fig. 3I-P). T3 treatment 331

significantly decreased trabecular vBMD (14%) only in WT (Fig. 3P) and 332

decreased BV/TV (29%) only in KO mice (Fig. 3I). Accordingly to previous 333

findings (19), vertebral BV/TV (Fig. 3I) was higher and vertebral Tb.Po (Fig. 3M) 334

and SMI (Fig. 3O) were lower in untreated WT versus untreated KO mice. The 335

µCT analysis of the cortical bone of the femoral diaphysis demonstrated that T3 336

treatment decreased Ct.Th (22%), T.Ar (25%), Ct.Ar (25%), Ma.Ar (49%), Ct.BV 337

(25%) and increased Ct.Po (9%) only in WT mice (Fig. 4A-F). The comparison 338

of WT versus KO mice shows higher cortical vBMD (Fig. 4H) in α2A/2C-AR-/- 339

animals, which corroborates the DXA findings. Taking into account the T3 340

treated animals, we found higher femoral Ct.Th, T.Ar, Ct.Ar, Ma.Ar, Ct.BV and 341

lower Ct.Po in KO versus WT mice, which supports the resistance of the 342

femoral cortical bone of KO mice to the deleterious effects of thyrotoxicosis. The 343

cortical bone of vertebra (L6) was negatively affected by thyrotoxicosis both in 344

WT and KO mice. T3 treatment reduced Ct.Th (22% in WT and 32% in KO), 345

Ct.Ar (46% in WT and 35% in KO), Ct.BV (46% in WT and 38% in KO) and MMI 346

(45% in WT and 38% in KO) in WT and α2A/2C-AR-/- animals (Fig. 4I-K, 4M and 347

4O). However, T3 treatment reduced T.Ar (45%) and increased Ct.Po (14%) 348

only in the vertebra of WT animals. Corroborating the DXA findings, cortical 349

16

vBMD and Ct.BV of L6 were higher in the KO mice (Fig. 4P and 4M, 350

respectively). 351

Thyroid hormone effects on the bone biomechanical parameters of 352

α2A/2C-AR-/- mice. In accordance with DXA and µCT findings, the three-point 353

bending test showed that the femoral maximum load, ultimate load and stiffness 354

were significantly decreased by T3 treatment only in WT mice and not in α2A/2C-355

AR-/- animals (Fig. 5A-C). In addition, we found lower femoral maximum load, 356

ultimate load, stiffness, young’s modulus and resilience in WT+T3 than in 357

KO+T3 mice (Fig. 5). No significant differences were found in any of the 358

biomechanical parameters tested between untreated WT and KO mice. 359

Altogether, these data indicate a resistance of the KO mice to the deleterious 360

effects of thyrotoxicosis also on the biomechanical features of the femur. 361

Effect of T3 treatment on the gene expression of the 362

RANKL/RANK/OPG system. To further investigate a possible interaction 363

between thyroid hormone and the SNS in the bone, we evaluated the effect of 364

T3 on the gene expression of the RANKL/RANK/OPG system, which plays a 365

pivotal role in bone remodeling by regulating osteoclast formation and activity 366

(67). T3 treatment had no effect on the expression of RANK and RANK-L (Fig. 367

6A-B), both in WT and KO animals. On the other hand, OPG mRNA expression 368

was decreased (66 %) by T3 treatment in WT mice and increased (two-fold) by 369

T3 in KO animals (Fig. 6C), which may partially explain the osteopenic effect of 370

T3 in WT mice but not in KO mice. Surprisingly, OPG mRNA expression 371

showed to be lower in KO mice versus WT mice. 372

Effect of T3 treatment on the expression of adrenoceptors in bone. 373

T3 treatment increased the mRNA expression of β1-AR in the bone of α2A/2C-AR-374

17

/- mice by 40 % (p<0.05) but did not increase expression in the bone of WT 375

animals (Fig. 7A). On the other hand, T3 treatment decreased β2-AR mRNA 376

expression by 33 % in WT but not in KO mice (Fig. 7B). In WT animals, T3 377

treatment had no effect on the bone expression of α2A-AR mRNA (Fig. 7C), but 378

significantly decreased α2C-AR mRNA expression by 30 % (Fig. 7D). There 379

were no significant differences between WT and KO animals regarding to the 380

mRNA expression of adrenoceptors in the bone. 381

Calvaria-derived Osteoblastic Cells from WT and α2A/2C-AR-/- respond 382

differently to T3 and Clonidine. We tested the effect of T3 treatment on cell 383

growth of osteoblasts derived from WT and α2A/2C-AR-/- mice (Fig. 8A). T3 (10-8 384

M) significantly decreased the growth of WT cells after 3, 6 and 9 days of 385

treatment, which corroborates previous studies (5, 65, 70). On the other hand, 386

in KO cells, T3 treatment for 3 days had no effect on cell growth, whereas 6 and 387

9 days of treatment increased this parameter. Compared with WT cells, 388

untreated KO cells grew significantly less during the whole experimental period 389

(Fig. 8A), suggesting that α2A-AR and/or α2C-AR have a function on the 390

regulation of this process. Next, WT and KO cells were treated with clonidine 391

(CLO), an unspecific α2-AR agonist, alone or combined with T3, to test if α2-392

adrenoceptors are functional in osteoblastic cells, and if their activation modifies 393

the responsiveness of these cells to T3. In WT cells (Fig. 8B), CLO or T3 394

treatment for 3, 6 and 9 days significantly decreased cell growth, while the 395

combination of both agents promoted a higher decrease than T3 alone, in an 396

additive manner. In contrast, in KO cells (Fig. 8C), CLO increased cell growth 397

after 3, 6 and 9 days of treatment. T3 also increased KO cell growth, but not as 398

much as CLO and only after 6 and 9 days of treatment. The combination of both 399

18

agents showed that T3 effect was not modified by CLO in KO cells; however, T3 400

significantly attenuated the effect of CLO on cell growth after 3 and 6 days of 401

treatment. Cell viability, assessed by Trypan Blue exclusion, was always ≥ 90% 402

and was not affected by T3 or CLO treatment at any time point tested in either 403

cell lineage. Since cell viability was not affected by both agents, it is likely that 404

their effects in cell growth were mainly the result of an effect in cell proliferation, 405

rather than in cell death. 406

DISCUSSION: 407

To investigate if the osteopenic effects of TH excess could depend on 408

SNS activation, we evaluated whether the thyrotoxicosis-induced osteopenia is 409

intensified in α2A/2C-AR-/- mice, which exhibit elevated plasma NE and intact β2-410

AR (29). Interestingly, these mice exhibit a phenotype of HBM, despite their 411

elevated sympathetic tone, which is strong evidence that β2-AR is not the 412

unique adrenoceptor involved in the control of bone turnover and raises the 413

hypothesis that 2-ARs may also mediate SNS signaling in the skeleton (19). 414

Treatment with the supraphysiological dose of T3 for 12 weeks promoted 415

the characteristic effects of thyrotoxicosis (30, 54, 55), including cardiac 416

hypertrophy both in WT and KO mice, and a generalized decrease in aBMD in 417

WT mice, determined by DXA. In contrast, a surprising and very interesting 418

finding was that α2A/2C-AR-/- mice were resistant to the thyrotoxicosis-induced 419

osteopenia, as the supraphysiological dose of T3 had no effect on their bone 420

mass. The µCt analysis substantiates the resistance of α2A/2C-AR-/- mice to the 421

osteopenic effects of TH excess. As expected, thyrotoxicosis negatively 422

affected most of the microarchitectural features of the distal metaphysis 423

trabecular bone and midshaft cortical bone of the femur in WT but not in α2A/2C-424

19

AR-/- mice. These findings were accompanied by T3-induced declines in the 425

biomechanical properties (maximum load, ultimate load and stiffness) of the 426

femur only in WT, but not in KO mice. Corroborating previous studies that show 427

that TH preferentially affects femoral but not vertebral bone (26, 44, 51, 59), the 428

microstructural parameters of the trabecular bone of L6 were barely affected by 429

T3 treatment. However, the trabecular vBMD of L6, determined by µCT, was 430

significantly decreased by TH treatment only in WT mice and not in KO mice, 431

which is in agreement with the aBMD findings determined by DXA. On the other 432

hand, the cortical bone of L6 was negatively affected by thyrotoxicosis both in 433

WT and KO mice. Nevertheless, it was slightly more affected by TH in WT than 434

in KO mice, since T3-treatment decreased T.Ar and increased Ct.Po of L6 in 435

WT but not in KO mice. Altogether, these results strongly suggest that TH 436

excess depends on intact α2-AR signaling to negatively regulate bone mass, 437

structure and biomechanical resistance, especially in the femur. The fact that 438

the double gene deletion of α2A- and α2C-AR prevented the negative effects of 439

TH in the femoral cortical bone, but not in the vertebral cortical bone suggests 440

that the TH-SNS interaction may vary depending on the skeletal site. In fact, 441

several studies show that the TH-adrenergic interrelationships is by no means 442

uniform from tissue to tissue, organ to organ, or in different regions of a same 443

organ (56). 444

We also evaluated the effect of TH on the RANK/RANK-L/OPG system. 445

RANK-L, expressed by the osteoblasts, is the ligand of RANK (an osteoclast 446

plasma membrane receptor) and the main inducer of osteoclast formation, 447

function and survival (67). OPG, also expressed by the osteoblasts, is the 448

natural inhibitor of osteoclastic activity, since it binds RANK-L and thereby 449

20

impairs RANK-L/RANK association. T3 significantly increased OPG mRNA 450

expression in α2A/2C-AR-/- mice and decreased OPG mRNA expression in WT 451

mice. This finding is consistent with the osteopenic effect of T3 in WT but not in 452

α2A/2C-AR-/- mice, and suggests that the RANK/RANK-L/OPG system may 453

partially mediate TH actions in the bone. Additionally, it suggests that this 454

mediation is affected by modifications in SNS signaling, which in turn reinforces 455

a TH-SNS interaction to regulate bone physiology. 456

The interaction of TH with the SNS is normally complex and may occur at 457

multiple levels (56). TH generally potentiates β-AR signals and it occurs in 458

different ways, depending on the tissue or organ (56). Certainly, one 459

mechanism whereby TH can alter these signals is by modifying the expression 460

of adrenoceptors (7). Considering this mechanism, we first evaluated if TH 461

could modify the mRNA expression of β1- and β2-AR in the bone. We found 462

that TH decreased β2-AR mRNA expression only in the bone of WT mice. In 463

contrast, T3 treatment increased β1-AR mRNA expression only in the bone of 464

α2A/2C-AR-/- mice. These findings once more show that the absence of α2A-AR 465

and/or α2C-AR alters the way that bone responds to thyrotoxicosis, supporting 466

the hypothesis of a TH-SNS interaction to regulate bone physiology. It is well 467

known that TH can modify the number of β-ARs, but, again, this effect is not the 468

same in all tissues, nor does it include all isoforms (7, 52). As a consequence of 469

thytoxicosis, for example, the number of β-AR has been shown to be increased 470

in rat heart muscle and brown adipose tissue, decreased in rat liver, and 471

increased or not changed in rat whyte adipose tissue and human leukocytes (7). 472

It is noteworthy that β1-AR KO mice exhibit low bone mass, which suggests that 473

this receptor mediates the anabolic actions of the SNS in the bone (11). In fact, 474

21

a recent study indicates that β1-AR exerts predominant and opposite effects of 475

β2-AR on bone remodeling through both systemic and local factors (48). 476

Moreover, evidence indicates that the anabolic effects of β1-AR signaling are 477

partially mediated by systemic IGF-1, which should be upregulated by β1-AR 478

activation (48). Interestingly, in the present study, we found that TH treatment 479

increased IGF-1 serum levels only in α2A/2C-AR-/- mice, which exhibited elevated 480

β1-AR mRNA expression. Thus, the positive effect of T3 treatment on β1-AR 481

mRNA expression and on IGF-1 serum levels in α2A/2C-AR-/- mice, but not in WT 482

mice, may partially explain the resistance of KO mice to the TH-induced 483

osteopenia. 484

We also found that T3 treatment reduced α2C-AR mRNA expression, but 485

not α2A-AR mRNA expression in the bone of WT mice. The effect of TH on α-486

ARs has been much less investigated than on β-ARs. In general, it seems that 487

TH decreases the expression of α-ARs (7), which corroborates our findings. 488

Besides bone, it was observed in the heart, liver and white adipose tissue (7). 489

The physiological implication of the T3-induced reduction in α2C-AR mRNA 490

expression, observed in the present study, cannot be inferred at this point. 491

However, the thyrotoxicosis-induced modulation of β1-AR, β2-AR and α2C-AR 492

mRNA expression, observed in the bone, suggests that these adrenoceptors 493

may be sites of interaction between thyroid hormone and the SNS to regulate 494

bone structure and physiology. Nevertheless, it is important to consider that the 495

adrenoceptors themselves are only two of many sites of TH-SNS interaction. 496

Furthermore, T3-induced modifications in adrenoceptors are usually modest or 497

inexistent and, thus, generally cannot account for the T3-induced modifications 498

in the SNS responsiveness (56). Usually, modulations of more distal cellular 499

22

effectors show to be responsible for the TH-induced changes in the 500

catecholamine responsiveness (56). It is possible that these downstream 501

effectors may have been modified by the disruption of α2-AR signaling in bone 502

cells, which remains to be investigated. 503

Another point of TH-SNS interaction is the central sympathetic outflow. 504

TH regulates the sympathetic nerve activity and, therefore, modulates serum 505

and tissue levels of catecholamines (38). Regardless of a hyperadrenergic state 506

in thyrotoxicosis, there is a reduction of the central sympathetic outflow, which 507

results in reduced serum levels of NE (38, 62). This paradox is also observed in 508

hypothyroid patients, who clinically manifest signs of decreased adrenergic 509

stimulation but can be expected to have increased serum levels of EP, NE, and 510

its metabolites (35). We, therefore, evaluated plasma and bone levels of 511

cathecolamines. As predicted, NE and EP were significantly elevated in the 512

plasma of α2A/2C-AR-/- mice, which correlates with the fact that α2A- and α2C-AR 513

are the main autoreceptors responsible for limiting NE release from the 514

sympathetic nerves and from chromaffin cells in the adrenal medulla (32), 515

respectively. As expected, T3 decreased plasma NE in WT mice. In contrast, 516

this effect was not observed in α2A/2C-AR-/- mice, suggesting that α2-ARs are 517

important to mediate the TH-effects on the sympathetic outflow. Interestingly, 518

bone concentration of NE was significantly elevated in α2A/2C-AR-/- mice and was 519

only reduced by T3 treatment in α2A/2C-AR-/- mice. The effect of T3 on NE tissue 520

levels also varies depending on the tissue. As examples, thyrotoxicosis was 521

shown to increase or not change NE concentration in the heart and to 522

decreased NE levels in the cerebral cortex (37). In general, the 523

sympathomimetic features of thyrotoxicosis cannot be explained by increased 524

23

sympathetic outflow, turnover or cathecolamine release. On the other hand, TH 525

generally modifies the cellular responsiveness to SNS signals, which overrides 526

the inhibitory or stimulatory effect of thyrotoxicosis or hypothyroidism, 527

respectively, on the central SNS outflow (22, 56). This may also occur in the 528

bone, since plasma and bone levels of NE could not explain the osteopenic 529

effects of thyrotoxicosis in WT mice. 530

In order to get insights if α2A-AR signaling can directly modify bone 531

responsiveness to TH, we evaluated the effect of T3 on cell growth of calvaria-532

derived osteoblasts isolated from WT and α2A/2C-AR-/- mice. We first observed a 533

lower cell growth in α2A/2C-AR-/- cells than in WT cells, suggesting that α2A-AR 534

and/or α2C-AR signaling in osteoblasts may be important for this process. 535

Accordingly, growth of WT cells was decreased by CLO, an unspecific α2-AR 536

agonist, whereas growth of KO cells was increased by this agent. The response 537

of KO cells to CLO may be the result of a CLO-mediated activation of α2B-AR, 538

which was shown to serve functions in peripheral tissues as well as in the CNS 539

(32), and to be expressed in bone cells, including osteoblasts (19). As expected 540

and corroborating previous studies (5, 65, 70), T3 significantly decreased the 541

growth of WT cells. However, following the pattern of T3 effects observed in 542

vivo, this negative effect of T3 was completely absent or was reversed in α2A/2C-543

AR-/- cells. Additionally, the combination of T3 with CLO had an additive effect 544

on the inhibition of WT cell growth, whereas T3 attenuated the positive effect of 545

CLO on KO cell growth. Altogether, these findings suggest that a TH-SNS 546

interaction, involving α2-AR signaling, may occur in osteoblasts to locally 547

regulate bone physiology. 548

24

These in vivo results also suggest that non-neuronal α2-ARs may be 549

important to directly mediate SNS actions in the bone. As mentioned before, α2-550

AR are classic autoreceptors (presynaptic receptors that inhibit the exocytosis 551

of their own neurotransmitter) that inhibit NE or EP release from adrenergic 552

neurons (of the SNS and CNS), and from chromaffin cells (of the adrenal 553

medulla) to control the sympathetic tonus (58). However, α2-ARs are also 554

expressed as heteroceptors on non-adrenergic neurons in the peripheral and 555

CNS. Heteroreceptors are presynaptic release-modulating receptors that are 556

activated by neurotransmitters from neighboring neurons (58). As 557

heteroceptors, α2A-ARs can inhibit the release of several neurotransmitters, 558

including serotonin, GABA and dopamine (23). In addition to these presynaptic 559

neuronal receptors (auto and heteroreceptors), α2-ARs were also identified in 560

many non-neuronal cell types of the body (23), including osteoblasts, 561

osteocytes, osteoclasts and chondrocytes (19, 61). Recently, the dopamine β-562

hydroxylase (Dbh) promoter was used to drive expression of α2A-AR exclusively 563

in noradrenergic and adrenergic neurons. Dbh-α2A transgenic mice were 564

crossed with α2A/2C-AR-/- mice to generate animals with selective expression of 565

α2A-autoreceptors in adrenergic cells (24). The analysis of these mice confirmed 566

the importance of α2A-autoreceptors as presynaptic inhibitory feedback 567

receptors to control NE release, but, more importantly, showed that the majority 568

of α2-AR effects were mediated by α2-ARs in non-adrenergic neurons 569

(heteroceptors) or non-neuronal cells. In addition, studies have shown that non-570

neuronal α2-ARs have essential roles in the body (23). Considering all these 571

features of α2-ARs, the TH-SNS interactions to control bone physiology may be 572

extremely complex and may involve α2-autoreceptors, α2-heteroceptors and/or 573

25

α2-ARs present in bone cells. Further studies are required to identify the 574

involvement of these α2-ARs subtypes in bone physiology. 575

It is important to consider that our data is based on a global gene 576

deletion model, which has its strengths and limitations (16). The resistance of 577

α2A/2C-AR-/- mice to the thyrotoxicosis-induced alterations in the bone mass, 578

microstructure and biomechanical parameters provides strong evidence that TH 579

depends on the integrity of the SNS to control bone physiology, which is an 580

original and important finding. On the other hand, global KO models make it 581

difficult to discriminate direct from indirect effects and the roles of a particular 582

gene (16). Therefore, our data do not clearly elucidate if TH interacts with the 583

SNS at the CNS and/or at the local (bone) level. Given the complexity of TH-584

SNS interaction to control other physiological functions (56), it is much likely 585

that the diverse bone responses of WT and KO mice to TH, observed in the 586

present study, are the result of both central and skeletal SNS-TH interaction. 587

The sites and mechanisms of this interaction are important points for future 588

investigation. In vitro and in vivo studies involving mice with specific deletions of 589

α2-AR subtypes in bone cells and also in adrenergic cells (autoreceptors and 590

heteroceptors) will be valuable tools to elucidate these issues. 591

Another point to be considered is the life stage of the animals. This study 592

shows that TH depends on the integrity of the SNS to regulate bone mass and 593

structure during the young-adulthood of mice, a phase of important bone mass 594

accrual. The TH-SNS interaction during later stages of life may be different and 595

promote different effects on bone mass and structure than those observed in 596

the present study. This matter is supported by evidence that the SNS signaling 597

may vary depending on the life stage. The study of Pierroz at (48), for instance, 598

26

suggests that β1-adrenergic signaling exerts a predominant anabolic stimulus 599

on bone during growth, whereas β2-adrenergic signaling predominantly 600

regulates bone turnover during aging. 601

In summary, the present study provides strong evidence that thyroid 602

hormone interacts with the SNS to control bone mass and structure in young-603

adult mice. Nevertheless, the mechanisms underlying this interaction and 604

whether it occurs at the central and/or skeletal level, or whether the TH-605

adrenergic interrelationships depend on α2-autoreceptors, α2-heteroceptors 606

and/or non-neuronal α2-ARs are important issues raised by this study that need 607

to be clarified in future studies. Finally, the present findings offer new insights 608

into the mechanisms through which TH regulates bone mass, structure and 609

physiology. 610

ACKNOWLEDGMENTS: 611

We thank the Institute of Biomedical Sciences, University of Sao Paulo, for 612

providing technical support to this study. The present address of T.L.F. is RUSH 613

University Medical Center, Department of Internal Medicine, Division of 614

Endocrinology and Metabolism, 1653 W Congress Parkway, Suite 301, 60612, 615

Chicago, IL, USA. 616

GRANTS: 617

This work was supported by grants (C.H.A.G.) from FAPESP (Process number 618

2010/06409-0 and 2013/02247-3), Brazil. T.L.F., M.B.C.G.T., M.V.S., G.M.M. 619

and C.C.C. were recipients of fellowships from FAPESP (Process Numbers 620

05/59557-8, 10/50068-2, 2012/11858-3, 10/04911-0 and 09/52485-2, 621

respectively). M.R.M. was recipient of a fellowship from CAPES. 622

AUTHORS CONTRIBUTIONS: 623

T.L.F. performed all the experiments; collected, processed and analyzed most 624

of the samples; interpreted data and co-wrote the manuscript; M.B.C.G.T. and 625

C.C.C. helped T.L.F. to perform all the experiments and to collect, process and 626

27

analyze the samples; M.V.S. and G.M.M. performed the CT analysis; M.R.M. 627

performed the cell culture experiments. D.Y.A., J.D.P. and D.E.C. performed the 628

NE plasma and bone measurements; P.C.B. contributed to the conception and 629

design of the research; C.H.A.G. conception and design of research and co-630

wrote the manuscript. All authors approved the final version of the manuscript. 631

DISCLOSURES: 632

No conflicts of interest, financial or otherwise, are declared by the authors. 633

REFERENCES: 634

1. Abu EO, Bord S, Horner A, Chatterjee VK, and Compston JE. The 635

expression of thyroid hormone receptors in human bone. Bone 21: 137-142, 1997. 636

2. Ahima RS, Dushay J, Flier SN, Prabakaran D, and Flier JS. Leptin 637

accelerates the onset of puberty in normal female mice. J Clin Invest 99: 391-395, 638

1997. 639

3. Allain TJ, and McGregor AM. Thyroid hormones and bone. J Endocrinol 640

139: 9-18, 1993. 641

4. Ballock R, Mita BC, Zhou X, Chen DH, and Mink LM. Expression of 642

thyroid hormone receptor isoforms in rat growth plate cartilage in vivo. J Bone 643

Miner Res 14: 1550-1556, 1999. 644

5. Beber EH, Capelo LP, Fonseca TL, Costa CC, Lotfi CF, Scanlan TS, and 645

Gouveia CH. The thyroid hormone receptor (TR) beta-selective agonist GC-1 646

inhibits proliferation but induces differentiation and TR beta mRNA expression in 647

mouse and rat osteoblast-like cells. Calcif Tissue Int 84: 324-333, 2009. 648

6. Beylot M, Vincent M, Benzoni D, Bodson A, Riou JP, and Mornex R. 649

[Effects of propranolol and indomethacin upon urinary hydroxyproline in 650

hyperthyroid patients (author's transl)]. Nouv Presse Med 11: 989-991, 1982. 651

7. Bilezikian JP, and Loeb JN. The influence of hyperthyroidism and 652

hypothyroidism on alpha- and beta-adrenergic receptor systems and adrenergic 653

responsiveness. Endocr Rev 4: 378-388, 1983. 654

8. Bonnet N, Beaupied H, Vico L, Dolleans E, Laroche N, Courteix D, and 655

Benhamou CL. Combined effects of exercise and propranolol on bone tissue in 656

ovariectomized rats. J Bone Miner Res 22: 578-588, 2007. 657

9. Bonnet N, Benhamou CL, Malaval L, Goncalves C, Vico L, Eder V, Pichon 658

C, and Courteix D. Low dose beta-blocker prevents ovariectomy-induced bone loss 659

in rats without affecting heart functions. J Cell Physiol 217: 819-827, 2008. 660

10. Bonnet N, Brunet-Imbault B, Arlettaz A, Horcajada MN, Collomp K, 661

Benhamou CL, and Courteix D. Alteration of trabecular bone under chronic beta2 662

agonists treatment. Med Sci Sports Exerc 37: 1493-1501, 2005. 663

11. Bonnet N, Pierroz DD, and Ferrari SL. Adrenergic control of bone 664

remodeling and its implications for the treatment of osteoporosis. J Musculoskelet 665

Neuronal Interact 8: 94-104, 2008. 666

12. Brede M, Philipp M, Knaus A, Muthig V, and Hein L. alpha2-adrenergic 667

receptor subtypes - novel functions uncovered in gene-targeted mouse models. Biol 668

Cell 96: 343-348, 2004. 669

28

13. Brum PC, Kosek J, Patterson A, Bernstein D, and Kobilka B. Abnormal 670

cardiac function associated with sympathetic nervous system hyperactivity in mice. 671

Am J Physiol Heart Circ Physiol 283: H1838-1845, 2002. 672

14. Capelo LP, Beber EH, Huang SA, Zorn TM, Bianco AC, and Gouveia CH. 673

Deiodinase-mediated thyroid hormone inactivation minimizes thyroid hormone 674

signaling in the early development of fetal skeleton. Bone 43: 921-930, 2008. 675

15. Civantos Calzada B, and Aleixandre de Artinano A. Alpha-adrenoceptor 676

subtypes. Pharmacol Res 44: 195-208, 2001. 677

16. Davey RA, MacLean HE, McManus JF, Findlay DM, and Zajac JD. 678

Genetically modified animal models as tools for studying bone and mineral 679

metabolism. J Bone Miner Res 19: 882-892, 2004. 680

17. Ducy P, Amling M, Takeda S, Priemel M, Schilling AF, Beil FT, Shen J, 681

Vinson C, Rueger JM, and Karsenty G. Leptin inhibits bone formation through a 682

hypothalamic relay: a central control of bone mass. Cell 100: 197-207, 2000. 683

18. Elefteriou F, Ahn JD, Takeda S, Starbuck M, Yang X, Liu X, Kondo H, 684

Richards WG, Bannon TW, Noda M, Clement K, Vaisse C, and Karsenty G. 685

Leptin regulation of bone resorption by the sympathetic nervous system and 686

CART. Nature 434: 514-520, 2005. 687

19. Fonseca TL, Jorgetti V, Costa CC, Capelo LP, Covarrubias AE, Moulatlet 688

AC, Teixeira MB, Hesse E, Morethson P, Beber EH, Freitas FR, Wang CC, 689

Nonaka KO, Oliveira R, Casarini DE, Zorn TM, Brum PC, and Gouveia CH. 690

Double disruption of alpha2A- and alpha2C-adrenoceptors results in sympathetic 691

hyperactivity and high-bone-mass phenotype. J Bone Miner Res 26: 591-603, 2011. 692

20. Fox KM, Kimura S, Plato CC, and Kitagawa T. Bilateral asymmetry in 693

bone weight at various skeletal sites of the rat. Anat Rec 241: 284-287, 1995. 694

21. Freitas FR, Moriscot AS, Jorgetti V, Soares AG, Passarelli M, Scanlan TS, 695

Brent GA, Bianco AC, and Gouveia CH. Spared bone mass in rats treated with 696

thyroid hormone receptor TR beta-selective compound GC-1. Am J Physiol 697

Endocrinol Metab 285: E1135-1141, 2003. 698

22. Geffner DL, and Hershman JM. Beta-adrenergic blockade for the 699

treatment of hyperthyroidism. Am J Med 93: 61-68, 1992. 700

23. Gilsbach R, and Hein L. Are the pharmacology and physiology of alpha(2) 701

adrenoceptors determined by alpha(2)-heteroreceptors and autoreceptors 702

respectively? Br J Pharmacol 165: 90-102, 2012. 703

24. Gilsbach R, Roser C, Beetz N, Brede M, Hadamek K, Haubold M, 704

Leemhuis J, Philipp M, Schneider J, Urbanski M, Szabo B, Weinshenker D, and 705

Hein L. Genetic dissection of alpha2-adrenoceptor functions in adrenergic versus 706

nonadrenergic cells. Mol Pharmacol 75: 1160-1170, 2009. 707

25. Glatt V, Canalis E, Stadmeyer L, and Bouxsein ML. Age-related changes in 708

trabecular architecture differ in female and male C57BL/6J mice. J Bone Miner 709

Res 22: 1197-1207, 2007. 710

26. Gouveia CH, Jorgetti V, and Bianco AC. Effects of thyroid hormone 711

administration and estrogen deficiency on bone mass of female rats. J Bone Miner 712

Res 12: 2098-2107, 1997. 713

27. Gouveia CH, Schultz JJ, Bianco AC, and Brent GA. Thyroid hormone 714

stimulation of osteocalcin gene expression in ROS 17/2.8 cells is mediated by 715

transcriptional and post-transcriptional mechanisms. J Endocrinol 170: 667-675, 716

2001. 717

28. Hein L. Adrenoceptors and signal transduction in neurons. Cell Tissue Res 718

326: 541-551, 2006. 719

29

29. Hein L, Altman JD, and Kobilka BK. Two functionally distinct alpha2-720

adrenergic receptors regulate sympathetic neurotransmission. Nature 402: 181-721

184, 1999. 722

30. Hu LW, Liberti EA, and Barreto-Chaves ML. Myocardial ultrastructure in 723

cardiac hypertrophy induced by thyroid hormone--an acute study in rats. 724

Virchows Arch 446: 265-269, 2005. 725

31. Kindblom JM, Gothe S, Forrest D, Tornell J, Vennstrom B, and Ohlsson C. 726

GH substitution reverses the growth phenotype but not the defective ossification in 727

thyroid hormone receptor alpha 1-/-beta-/- mice. J Endocrinol 171: 15-22, 2001. 728

32. Knaus AE, Muthig V, Schickinger S, Moura E, Beetz N, Gilsbach R, and 729

Hein L. Alpha2-adrenoceptor subtypes--unexpected functions for receptors and 730

ligands derived from gene-targeted mouse models. Neurochem Int 51: 277-281, 731

2007. 732

33. Lakatos P, Foldes J, Nagy Z, Takacs I, Speer G, Horvath C, Mohan S, 733

Baylink DJ, and Stern PH. Serum insulin-like growth factor-I, insulin-like growth 734

factor binding proteins, and bone mineral content in hyperthyroidism. Thyroid 10: 735

417-423, 2000. 736

34. Landsberg L. Catecholamines and hyperthyroidism. Clin Endocrinol Metab 737

6: 697-718, 1977. 738

35. Levey GS, and Klein I. Catecholamine-thyroid hormone interactions and 739

the cardiovascular manifestations of hyperthyroidism. Am J Med 88: 642-646, 740

1990. 741

36. Livak K. ABI Prism 7700 Sequence Detection system, User bulletin 2. PE 742

Applied Biosystems 1997. 743

37. Mano T, Sakamoto H, Fujita K, Makino M, Kakizawa H, Nagata M, 744

Kotake M, Hamada M, Uchimura K, Hayakawa N, Hayashi R, Nakai A, Itoh M, 745

Kuzuya H, and Nagasaka A. Effects of thyroid hormone on catecholamine and its 746

metabolite concentrations in rat cardiac muscle and cerebral cortex. Thyroid 8: 747

353-358, 1998. 748

38. Matsukawa T, Mano T, Gotoh E, Minamisawa K, and Ishii M. Altered 749

muscle sympathetic nerve activity in hyperthyroidism and hypothyroidism. J 750

Auton Nerv Syst 42: 171-175, 1993. 751

39. Milne M, Kang MI, Quail JM, and Baran DT. Thyroid hormone excess 752

increases insulin-like growth factor I transcripts in bone marrow cell cultures: 753

divergent effects on vertebral and femoral cell cultures. Endocrinology 139: 2527-754

2534, 1998. 755

40. Monte JC, Casarini D, Parise E, Schor N, and dos Santos OF. 756

Neurohumoral systems in patients with cirrhosis. Ren Fail 19: 335-342, 1997. 757

41. Mosekilde L, and Melsen F. A tetracycline-based histomorphometric 758

evaluation of bone resorption and bone turnover in hyperthyroidism and 759

hyperparathyroidism. Acta Med Scand 204: 97-102, 1978. 760

42. Naffah-mazzacoratti M FM, Cavalheiro EA. Serum catecholamine levels 761

determined by high performance liquid chromatography coupled with 762

electrochemical detection 763

Arquivos Brasileiros de Endocrinologia e Metabolismo 36: 119-122, 1992. 764

43. O'Shea PJ, Harvey CB, Suzuki H, Kaneshige M, Kaneshige K, Cheng SY, 765

and Williams GR. A thyrotoxic skeletal phenotype of advanced bone formation in 766

mice with resistance to thyroid hormone. Mol Endocrinol 17: 1410-1424, 2003. 767

30

44. Ongphiphadhanakul B, Alex S, Braverman LE, and Baran DT. Excessive 768

L-thyroxine therapy decreases femoral bone mineral densities in the male rat: 769

effect of hypogonadism and calcitonin. J Bone Miner Res 7: 1227-1231, 1992. 770

45. Oppenheimer JH, Schwartz HL, Lane JT, and Thompson MP. Functional 771

relationship of thyroid hormone-induced lipogenesis, lipolysis, and thermogenesis 772

in the rat. J Clin Invest 87: 125-132, 1991. 773

46. Pataki A, Muller, K, Bilbe, G, Green, JR, Glatt M. Anabolic Effect of beta2-774

agonists, formoterol and sulbutamol on cancellous bone ovariectomized. Bone 9: 775

1996. 776

47. Philipp M, and Hein L. Adrenergic receptor knockout mice: distinct 777

functions of 9 receptor subtypes. Pharmacol Ther 101: 65-74, 2004. 778

48. Pierroz DD, Bonnet N, Bianchi EN, Bouxsein ML, Baldock PA, Rizzoli R, 779

and Ferrari SL. Deletion of beta-adrenergic receptor 1, 2, or both leads to different 780

bone phenotypes and response to mechanical stimulation. J Bone Miner Res 27: 781

1252-1262, 2012. 782

49. Pijl H, de Meijer PH, Langius J, Coenegracht CI, van den Berk AH, 783

Chandie Shaw PK, Boom H, Schoemaker RC, Cohen AF, Burggraaf J, and 784

Meinders AE. Food choice in hyperthyroidism: potential influence of the 785

autonomic nervous system and brain serotonin precursor availability. J Clin 786

Endocrinol Metab 86: 5848-5853, 2001. 787

50. Ramsay I. Adrenergic beta-receptor blockade in hyperthyroidism. Br J Clin 788

Pharmacol 2: 385-388, 1975. 789

51. Rosen HN, Sullivan EK, Middlebrooks VL, Zeind AJ, Gundberg C, 790

Dresner-Pollak R, Maitland LA, Hock JM, Moses AC, and Greenspan SL. 791

Parenteral pamidronate prevents thyroid hormone-induced bone loss in rats. J 792

Bone Miner Res 8: 1255-1261, 1993. 793

52. Rubio A, Raasmaja A, Maia AL, Kim KR, and Silva JE. Effects of thyroid 794

hormone on norepinephrine signaling in brown adipose tissue. I. Beta 1- and beta 795

2-adrenergic receptors and cyclic adenosine 3',5'-monophosphate generation. 796

Endocrinology 136: 3267-3276, 1995. 797

53. Rude RK, Oldham SB, Singer FR, and Nicoloff JT. Treatment of thyrotoxic 798

hypercalcemia with propranolol. N Engl J Med 294: 431-433, 1976. 799

54. Sanford CF, Griffin EE, and Wildenthal K. Synthesis and degradation of 800

myocardial protein during the development and regression of thyroxine-induced 801

cardiac hypertrophy in rats. Circ Res 43: 688-694, 1978. 802

55. Siehl D, Chua BH, Lautensack-Belser N, and Morgan HE. Faster protein 803

and ribosome synthesis in thyroxine-induced hypertrophy of rat heart. Am J 804

Physiol 248: C309-319, 1985. 805

56. Silva JE, and Bianco SD. Thyroid-adrenergic interactions: physiological 806

and clinical implications. Thyroid 18: 157-165, 2008. 807

57. Soukup T, Zacharova G, Smerdu V, and Jirmanova I. Body, heart, thyroid 808

gland and skeletal muscle weight changes in rats with altered thyroid status. 809

Physiol Res 50: 619-626, 2001. 810

58. Starke K. Presynaptic autoreceptors in the third decade: focus on alpha2-811

adrenoceptors. J Neurochem 78: 685-693, 2001. 812

59. Suwanwalaikorn S, Ongphiphadhanakul B, Braverman LE, and Baran DT. 813

Differential responses of femoral and vertebral bones to long-term excessive L-814

thyroxine administration in adult rats. Eur J Endocrinol 134: 655-659, 1996. 815

31

60. Takeda S, Elefteriou F, Levasseur R, Liu X, Zhao L, Parker KL, 816

Armstrong D, Ducy P, and Karsenty G. Leptin regulates bone formation via the 817

sympathetic nervous system. Cell 111: 305-317, 2002. 818

61. Togari A. Adrenergic regulation of bone metabolism: possible involvement 819

of sympathetic innervation of osteoblastic and osteoclastic cells. Microsc Res Tech 820

58: 77-84, 2002. 821

62. Tu T, and Nash CW. The influence of prolonged hyper- and hypothyroid 822

states on the noradrenaline content of rat tissues and on the accumulation and 823

efflux rates of tritiated noradrenaline. Can J Physiol Pharmacol 53: 74-80, 1975. 824

63. Turner P. Beta-adrenergic receptor blocking drugs in hyperthyroidism. 825

Drugs 7: 48-54, 1974. 826

64. Utiger RD. Beta-adrenergic-antagonist therapy for hyperthyroid Graves' 827

disease. N Engl J Med 310: 1597-1598, 1984. 828

65. Varga F, Rumpler M, Luegmayr E, Fratzl-Zelman N, Glantschnig H, and 829

Klaushofer K. Triiodothyronine, a regulator of osteoblastic differentiation: 830

depression of histone H4, attenuation of c-fos/c-jun, and induction of osteocalcin 831

expression. Calcif Tissue Int 61: 404-411, 1997. 832

66. Villicev CM, Freitas FR, Aoki MS, Taffarel C, Scanlan TS, Moriscot AS, 833

Ribeiro MO, Bianco AC, and Gouveia CH. Thyroid hormone receptor beta-834

specific agonist GC-1 increases energy expenditure and prevents fat-mass 835

accumulation in rats. J Endocrinol 193: 21-29, 2007. 836

67. Wada T, Nakashima T, Hiroshi N, and Penninger JM. RANKL-RANK 837

signaling in osteoclastogenesis and bone disease. Trends Mol Med 12: 17-25, 2006. 838

68. Waung JA, Bassett JH, and Williams GR. Thyroid hormone metabolism in 839

skeletal development and adult bone maintenance. Trends Endocrinol Metab 23: 840

155-162, 2012. 841

69. Williams GR. Actions of thyroid hormones in bone. Endokrynol Pol 60: 380-842

388, 2009. 843

70. Williams GR, Bland R, and Sheppard MC. Characterization of thyroid 844

hormone (T3) receptors in three osteosarcoma cell lines of distinct osteoblast 845

phenotype: interactions among T3, vitamin D3, and retinoid signaling. 846

Endocrinology 135: 2375-2385, 1994. 847

848

849

32

FIGURE LEGENDS 850

Fig. 1. Effect of thyroid hormone excess on plasma and bone levels of NE and 851

EP. 852

(A-B) Plasma levels. (C-D) Femoral levels. (A and C) NE (norepinephrine). (B 853

and D) EP (epinephrine). Animals were treated with a supraphysiological dose 854

of T3 (3.5 g/100 g BW/day) or saline for 12 weeks. Treatment started when the 855

animals were 40 days old. Values are expressed as the mean ± SEM (n = 7 per 856

group). *p<0.05 vs. the respective saline-treated animals (WT vs. WT+T3 or KO 857

vs. KO+T3). The numbers above the bars indicate the p values for differences 858

between WT vs. KO mice, as indicated. 859

Fig. 2. Effect of thyroid hormone excess on the aBMD of WT and α2A/2C-AR-/- 860

mice as determined by DXA. 861

(A) Hind body (includes L1-L6, pelvic bones and hind limbs). (B) Lumbar 862

vertebrae (L1-L6). (C) Femur. (D) Tibia. Animals were treated with a 863

supraphysiological dose of T3 (3.5 g/100 g BW/day) or saline for 12 weeks. 864

Treatment started when the animals were 40 days old. Values are expressed as 865

the mean ± SEM (n = 7 per WT and KO group, respectively). *p<0.05 and 866

**p<0.01 WT vs. WT+T3; + p<0.05 and ++p<0.01 WT vs. KO and WT+T3 vs. 867

KO+T3. 868

Fig. 3. Effect of thyroid hormone excess on the structural parameters of the 869

trabecular bone of the distal metaphysis of the femur and vertebral body of L6 in 870

WT and α2A/2C-AR-/- mice assessed by CT. 871

(A-H) Distal metaphysis of the femur. (I-P) Vertebral body of L6. Animals were 872

treated with a supraphysiological dose of T3 (3.5 g/100 g BW/day) or saline for 873

33

12 weeks. Values are expressed as the mean ± SEM (n = 6 per group). 874

*p<0.05, **p<0.01 and *** p<0.001 vs. the respective saline-treated animals 875

(WT vs. WT+T3 or KO vs. KO+T3). The numbers above the bars indicate the p 876

values for differences between WT vs. KO mice, as indicated. 877

Fig. 4. Effect of thyroid hormone excess on the structural parameters of the 878

cortical bone of the femoral midshaft and vertebral body of L6 in WT and α2A/2C-879

AR-/- mice assessed by CT. 880

(A-H) Femoral midshaft. (I-P) Vertebral body of L6. Animals were treated with a 881

supraphysiological dose of T3 (3.5 g/100 g BW/day) or saline for 12 weeks. 882

Values are expressed as the mean ± SEM (n = 6 per group). *p<0.05 and 883

**p<0.01 vs. the respective saline-treated animals (WT vs. WT+T3 or KO vs. 884

KO+T3). The numbers above the bars indicate the p values for differences 885

between WT vs. KO mice, as indicated. 886

Fig. 5. Effect of thyroid hormone excess on the biomechanical parameters of 887

the femur in WT and α2A/2C-AR-/- mice. 888

Data were assessed by means of the three-point bending test. Animals were 889

treated with a supraphysiological dose of T3 (3.5 g/100 g BW/day) or saline for 890

12 weeks. Values are expressed as the mean ± SEM (n = 6 per group). *p<0.05 891

and **p<0.01 vs. the respective saline-treated animals (WT vs. WT+T3). The 892

numbers above the bars indicate the p values for differences between WT+T3 893

vs. KO+T3 mice. 894

Fig. 6. Effect of thyroid hormone excess on the relative mRNA expression of the 895

RANK/RANK-L/OPG system in WT and α2A/2C-AR-/- mice as determined by real-896

time PCR analysis. 897

34

Animals were treated with a supraphysiological dose of T3 (3.5 g/100 g 898

BW/day) or saline (-) for 12 weeks. mRNA expression was determined by real-899

time PCR analysis. Values are expressed as the mean ± SEM (n = 6 per 900

group). **p<0.01 vs. the respective saline-treated animals (WT vs. WT+T3 or 901

KO vs. KO+T3). The numbers above the bars indicate the p values for 902

differences between WT vs. KO, as indicated. 903

Fig. 7. Effect of excess thyroid hormone on the relative mRNA expression of 904

adrenoceptors. 905

(A-B) WT vs. KO mice. (C-D) WT mice. Animals were treated with a 906

supraphysiological dose of T3 (3.5 g/100 g BW/day) or saline (-) for 12 weeks. 907

mRNA expression was determined by real-time PCR analysis. Values are 908

expressed as the mean ± SEM (n = 6 per group). *p<0.05 and ***p<0.001 vs. 909

the respective saline-treated animals (WT vs. WT+T3 or KO vs. KO+T3). The 910

numbers above the bars indicate the p values for differences between WT vs. 911

KO, as indicated. 912

Fig. 8. Effect of T3 and Clonidine (CLO) on osteoblastic cell growth. 913

(A) Effect of T3 on WT and α2A/2C-AR-/- cells. (B) Effect of T3 and/or CLO on WT 914

cells. (C) Effect of T3 and/or CLO on α2A/2C-AR-/- cells. Cells were cultured in 915

osteogenic media and treated with 10-8 M T3 and/10-6 M CLO for 3, 6 or 9 days. 916

Each point represents the mean ± SEM (n = 4). *p < 0.01 and **p<0.001 vs. 917

untreated WT cells; +p<0.05 vs. untreated α2A/2C-AR-/- cells; Xp<0.01 vs. all 918

groups. The numbers above the bars indicate the p values for differences 919

between the indicated groups. 920

921

35

TABLE 1. T3, T4 and IGF-1 serum levels in WT and 922 α2A/2C-AR-/- mice. 923

Treatment - T3 P

ng/ml Animal

T3

WT 0.370.07 7.000.6 <0,0001

KO 0.390.02 7.320.5 <0,0001

P ns ns

T4

WT 27.63.1 2.70,4 <0.0001

KO 36.22.8 3.20.5 <0,0001

P <0.05 ns

IGF-1

WT 57,82,1 56,43,4 ns

KO 53,52,3 66,62,7 <0.05

P ns ns

Data are presented as the mean ± SEM (n = 7 per group). The 924 significance between the groups was determined by Two-way 925 ANOVA, followed by Tukey test. ns = non-significant. 926

927