Introduction

Thyroid hormones (THs) are synthesized by the thyroid gland, including thyroxine (T4) and triiodothyronine (T3). T4 is the major secreted hormone, yet, T3 is classically considered as the active and more potent hormone since it binds to thyroid hormone receptors (TRs) with higher affinity than T4 [1].The clinical effects of THs on bone have been interested for more than a century. THs are required for skeletal development and establishment of peak bone mass [2, 3]. In adults, T3 regulates bone turnover and bone mineral density [4], and normal euthyroid status is essential to maintain optimal bone strength [5, 6]. In addition, alterations in THs levels have a profound effect on the cardiovascular system [7], which includes changes in myocardial contractility, heart rate, and resistance of peripheral vasculature [8–10]. Recently, mRNAs for TR isoforms were identified in aortic and coronary smooth muscle cells [11], suggesting that THs may play a direct role in vascular smooth muscle. Tatar et al. reported that serum-free T3 levels were inversely associated with arterial stiffness in hemodialysis patients without diabetes [12]. These findings hinted that THs could involve in calcium homeostasis regulating and correlate with vascular calcification (VC).

VC is a well-recognized pathological entity associated with cardiovascular morbidity and mortality [13]. Recent studies have shown that VC is an active, cell-regulated process, with many similarities to bone formation [14, 15]. Vascular smooth muscle cell (VSMC) can lose the expression of smooth muscle lineage markers and begin to express osteogenic markers and deposit a mineralized bone-like matrix [14–19]. During the mineralization of VSMC, runt-related transcription factor2 (Runx2), a transcription factor expressed at early stages of osteoblast differentiation, is upregulation [20], and in concert with osterix by regulating the expression of osteoblast-related gene [21], such as osteopontin (OPN) contributes to VC [22]. Recent study suggested that klotho knockdown can accelerate VSMC calcification through a Runx2-dependent pathway. Klotho, a cofactor for fibroblast growth factor 23 (FGF-23) signaling, is locally expressed in VSMC, and protects against VC by mediating FGF23 inhibitory effects on matrix mineralization [23]. However, the direct target genes of THs in VSMCs are poorly understood.

In our work, we investigated the change of THs concentration in plasma and the effect of T3 on phenotype transition in vitamin D3 plus nicotine-(VDN) induced VC rats. We observed the effect of T3 on Runx2, Osterix, OPN and klotho expression, and lineage markers of VSMC to investigate the mechanisms of T3 on VC.

Materials and Methods

Materials

Male Sprague–Dawley (SD) rats (200 ± 10 g) were purchased from the Animal Center, Peking University Health Sciences Center. All animal care and experimental protocols complied with the Animal Management Rule of the Ministry of Health, People’s Republic of China (Document No. 55, 2001). Rats were housed under standard conditions (room temperature 20 ± 8 °C, humidity 60 ± 10 %, lights from 6:00 to 18:00). T3 was from Merck (Darmstadt, German). Methimazole (MMI) and Vitamin D3 were from Sigma (St Louis, MO, USA). The calcium assay kit was from Zhongsheng Biosino Bio-technology and Science (Beijing, China). The alkaline phosphatase (ALP) activity kit was from Beihuakangtai Clinical Reagent Co., (Beijing, China). Radioimmunoassay kits for rat serum T3 and T4 were provided by Beijing Free Biological Technology (Beijing, China). Anti-SM22α antibody was from abcam (Cambridge, UK). Antibodies against α-actin, β-actin, klotho, and OPN antibody were from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Nitrocellulose membrane was from Hybond-C (Amersham Life Science, England), and an enhanced chemiluminescence kit was from Beijing Applygen Technologies (Beijing, China). Trizol was from GIBCOL (BRL, Rockville, MD), and dNTP, M-MuLV RT, Oligo (dT) 15 primer, and Taq DNA polymerase were from Promega (Madison, WI). Oligonucleotides were synthesized by Sai Baisheng Biotechnology (Beijing, China). All other chemicals and reagents were of analytical grade.

Animal Model

Male SD rats were randomly assigned to 5 groups (n = 8 in each group).VC was induced by VDN as described previously [24]. In brief, the rats were injected with vitamin D3 (300,000 IU/kg) intramuscularly and given nicotine intragastrically with peanut oil (5 ml/kg, 5 mg nicotine dissolved in 1 ml peanut oil) at 8 am, and nicotine in peanut oil was administered again at 8 pm on the same day. Rats in control group received normal saline intramuscularly and two gavages of peanut oil without nicotine (5 ml/kg) instead. The rats in Cal+T3 group were the VDN rats daily injection of T3 (0.2 mg/kg body weight IP) for 10 days. MMI (400 mg/L) was added to the drinking water for 4 weeks for VDN alone or plus T3 injection. After 4 weeks, all rats fasted overnight but had free access to water.

At the end of the experiment, hemodynamic variables were measured with use of PowerLab of BL-420F instruments (TaiMeng Ltd., Chengdu, China). After hemodynamic parameters were measured, a blood sample was drawn and mixed with 1 mg/ml EDTA-Na2 and 500 KIU/ml aprotinin. Plasma was obtained by centrifugation at 3,000×g for 15 min at 4 °C and stored at −70 °C. The intact aorta was then stored at −70 °C.

Detection of VC

The calcium content was measured by use of an atomic absorption spectrophotometer (novAA 300, Analytik Jena AG, Germany) at 422.7 nm as we described. Calcium content of the aorta was normalized to tissue dry weight. The ALP activity in aortic tissue was measured with the use of an ALP assay kit and results were normalized to the total protein concentration.

Radioimmunoassay of Free T3 and Free T4 Levels

The plasma samples were prepared and pre-treated with aprotinin (500 KIU/ml). Plasma was loaded onto a Sep-Pak 18C cartridge after equilibration with normal saline. The cartridge was washed with 2.5 ml normal saline and 10 % acetonitrile in 0.1 % trifluoroacetic acid, and then eluted with 2 ml of 50 % acetonitrile in 0.1 % trifluoroacetic acid. The elution was lyophilized and subjected to radioimmunoassay. The range of the standard curve of T3 was 0.50–8.0 ng/ml, in which a double-logarithmic scale showed a linear relationship with r 2 = 0.999. All samples were within the range of the standard curve. The detection limit was calculated to be 0.25 ng/ml. The intra- and interassay coefficients of variation were validated within our work and were 10 and 15 %, respectively. There is no cross-reactivity found with rat T4. The range of the standard curve of T4 was 20–320 ng/ml, in which a double-logarithmic scale showed a linear relationship with r 2 = 0.999. All samples were within the range of the standard curve. The detection limit was calculated to be 3.0 ng/ml. The intra- and interassay coefficients of variation were validated within our work and were 10 and 15 %, respectively. There is no cross-reactivity found with rat T3.

Real-Time PCR Analysis

Total RNA from aorta tissues was isolated and reverse transcribed by use of a reverse-transcription system (Promega, Madison, WI, USA). In total, 20 μL of the reaction mixture underwent real-time PCR, and then evaluated by use of SYBR Green I fluorescence. The rat primers were for Runx2 (149 bp), forward, 5′-AGT GAT TTA GGG CGC ATT CC-3′ and reverse, 5′-GGT GGG GAG GAT TGT GTC TG-3′; Osterix (148 bp), forward, 5′-TTG GTT AGG TGG TGG GCA GG-3′ and reverse, 3′-GAG GTG GGG TGC TGG ATA GG-5′; and β-actin (207 bp), forward, 5′-CAC CCG CGA GTA CAA CCT TC-3′ and reverse, 5′-CCC ATA CCC ACC ATC ACA CC-3′ as a normalization control. All amplification reactions involved use of the Mx3000 Multiplex Quantitative PCR System (Stratagene, La Jolla, CA, USA). After being denatured at 95 °C for 7 min, the solution underwent PCR for Runx2 at 95 °C for 30 s, 60 °C for 30 s, and 72 °C for 40 s for 40 cycles and for Osterix at 95 °C for 30 s, 60 °C for 30 s, and 72 °C for 40 s for 40 cycles.

Western Blot Analysis

Protein extracts from aortas were resuspended in sample buffer that contained 2 % SDS, 2 % mercaptoethanol, 50 mmol/L Tris–HCl (pH 6.8), 10 % glycerol, and 0.05 % bromophenol blue. The protein mixture was then placed in boiling water for 10 min and briefly centrifuged at a low speed. The denatured protein samples were resolved on a 10 % SDS–polyacrylamide gel in running buffer containing 25 mmol/L Tris, 192 mmol/L glycine, and 0.1 % SDS. The proteins were then transferred to a nitrocellulose membrane for 3 h at 4 °C at 200 mA in a transfer buffer containing 25 mmol/L Tris, 192 mmol/L glycine, and 20 % methanol. Non-specific protein binding was blocked by incubating the membrane with 5 % non-fat dry milk in TBS-T (20 mmol/L Tris–HCl (pH 7.6), 150 mmol/L NaCl, and 0.02 % Tween 20) for 1 h at room temperature with agitation. A goat anti-rat OPN primary antibody (1:500), anti-klotho (1:200), anti-SM22α (1:1,000), or anti-α-actin (1:2,000) were added to the membrane in TBS-T and incubated at 4 °C overnight with agitation. The secondary antibody was diluted in TBS-T (1:2,000 dilution) and applied to the membrane, and the reaction was incubated at room temperature for 1 h with agitation and underwent enhanced chemiluminescence detection. Between each of the steps, the membrane was washed 3 times for 10 min each with TBS-T at room temperature. Protein expression was analyzed by NIH image software and normalized to β-actin expression.

Statistical Analysis

All statistical analyses were carried out using the Graphpad Prism 5.0 software. Data are expressed as mean ± SEM. Student’s t test was used to compare results between 2 groups. One-way ANOVA, followed by a Student–Newman–Keuls test, was used to compare results between more than 2 groups. p < 0.05 was considered statistically significant.

Results

Plasma T3 and T4 was Decreased in Rats with VC

Compared with the control group, VDN rats showed significantly decreased immunoreactive-T3 (ir-T3) and ir-T4 content in plasma by 25 and 12 %, respectively, (both p < 0.05) (Fig. 1).

Fig. 1
figure 1

Changes of triiodothyronine (T3) (a) and thyroxine (T4) (b) plasma concentrations in rat with vascular calcification (VC). Ctrl control, Cal calcification. Data are shown as mean ± SEM, n = 6 at least in each group. *p < 0.05 versus Con

Full size image

T3 Treatment Improved Hemodynamic Function and Attenuated VC

In the rat model of VC, heart weight (HW)/body weight (BW), SBP, LVSP, and +LV dp/dt max were significantly elevated, by 11.6, 14.6, 23.0, and 113.5 %, respectively (all p < 0.05), as compared with controls (Table 1). Compared with VC alone, T3 treatment reduced HW/BW, SBP, and +LV dp/dt max by 11.4 %, 10.4, and 27.6 % (p < 0.05) (Table 1).

Table 1 Blood pressure, cardiac function, body weight, and left ventricle ratio of rats
Full size table

In VDN rats, aortic ALP activity and calcium level were significantly increased by 223.4 and 318.8 % (both p < 0.01), respectively, as compared with controls. Compared with VC alone, T3 treatment decreased aortic ALP activity and calcium content by 65.2 and 68.3 %, respectively (both p < 0.01) (Fig. 2). MMI, the most potent antithyroid drug, treatment abolished the protective effects of T3 on VC in VDN rats. MMI treated alone did not affect the aortic ALP activity and calcium content in VDN rats.

Fig. 2
figure 2

Triiodothyronine (T3) treatment decreased aortic calcium content (a) and alkaline phosphatase (ALP) activity (b) in vitamin D3 plus nicotine (VDN) rats. Ctrl control, Cal calcificiation, MMI methimazole. Data are expressed as mean ± SEM, n = 6 at least in each group. *p < 0.05, **p < 0.01 versus Con, # p < 0.05 versus Cal; & p < 0.05 versus Cal + T3

Full size image

T3 Decreased the Aortic mRNA Level of Runx2 and Osterix and Protein Level of OPN and Increased Protein Level of klotho

Runx2, Osterix mRNA levels, and OPN protein level were higher by 180, 43.9, and 257 %, and klotho protein level was lower by 53 % in VDN rats aortic tissue than in control rats. T3 treatment prevented the enhanced expression of Runx2, Osterix mRNA, and OPN protein by 45.2, 38.6, and 25.2 %, respectively (all p < 0.01) (Figs. 3, 4), and increased the klotho protein level by 92 % (p < 0.01) compared with that in rat aortas with VC alone. MMI treatment withdrew the effect of T3 on VDN rats. There is no significant change between VDN rats and MMI-treated VDN rats.

Fig. 3
figure 3

Triiodothyronine (T3) treatment decreased the aortic mRNA level of Osterix and runt-related transcription factor2 (Runx2). Quantitative real-time PCR of mRNA levels of a Osterix and b Runx2 in rat aortas. Data are expressed as mean ± SEM, n = 4. *p < 0.05 versus Con, # p < 0.05 versus Cal; & p < 0.05 versus Cal + T3

Full size image
Fig. 4
figure 4

Triiodothyronine (T3) treatment decreased protein level of osteopontin (OPN) and increased protein level of klotho. Western blot analysis of protein level of a osteogenic marker OPN and b klotho in rat aortas. β-actin was a loading control. Data are expressed as mean ± SEM, n = 4. *p < 0.05, **p < 0.01 versus Con; # p < 0.05 versus Cal; & p < 0.05 versus Cal + T3

Full size image

T3 Retarded Loss of Lineage Markers in VSMCs

Because the phenotype transition of VSMCs is associated with VC in vitro and in vivo, we further investigated the protein expression of smooth muscle lineage markers. Compared with normal vessels, VDN vessels showed decreased protein level of α-actin and SM-22α by 76.4 and 63.5 % (both p < 0.01), respectively (Fig. 5). T3 treatment prevented the reduction in α-actin and SM-22α by 409 and 130 % (both p < 0.01), respectively (Fig. 5). MMI cancels the effects of T3 in VDN VSMCs. There is no significant change between VDN rats and MMI-treated VDN rats.

Fig. 5
figure 5

Triiodothyronine (T3) treatment retarded smooth muscle cell phenotypic transition in rat aortas with vascular calcification. Western blot analysis of protein level of a α- actin and b lineage marker α-smooth muscle actin (α-SMA) in rat aortas. Data are expressed as mean ± SEM, n = 4. *p < 0.05, **p < 0.01 versus Con, # p < 0.05, ## p < 0.01 versus Cal, && p < 0.01 versus Cal + T3

Full size image

Discussion

In the present study, we found that administration of vitamin D3 and nicotine to rats resulted in serious VC, with systolic hypertension and left ventricular concentric hypertrophy, which suggests the reduced arterial compliance and the compensatory potentiation of cardiac contractile function. T3 treatment attenuated calcification and prevented the loss of lineage markers in rat aortas. Furthermore, T3 restored klotho expression and downregulated Runx2, Osterix, and OPN expression in calcified aortas. MMI is known to interfere with the synthesis and release of THs. MMI treatment significantly decreased plasma T3 content in VDN rats and abolished the effects of T3 on VC in VDN rats. These results suggested that T3 could be a key endogenous protective factor in VC.

Lots of clinical studies showed that ectopic calcification in blood vessels is often accompanied by either decreased bone mineral density or disturbed bone turnover. This contradictory association has been observed in the general population [25], in osteoporosis and chronic kidney disease patients [26, 27]. The term calcification paradox has been coined to refer to this counterintuitive observation [28]. It is urgent to find an endogenous factor, which not only can be used to treat osteoporosis, but also have an inhibiting effect on VC. The profound effects of THs on bone mineralization have been confirmed in clinical and pharmacological studies. Recently study showed that serum-free T3 levels were inversely associated with arterial stiffness in hemodialysis patients [12]. Analysis performed among health subjects showed that free T4 level was inversely associated with coronary artery calcification [29] and independently associated with carotid intima-media thickness [30] in euthyroid subjects. Sato et al. reported that aortic smooth muscle tissues from methimazole-induced hypothyroid rats showed increasing calcium content compared with that from the control euthyroid animals, whereas T3 treatment reduced calcium content [31]. These results indicated that normal THs level involved in the prevention of VC; however, the alteration of endogenous T3 content and the effect of T3 in VC were still unclear. In the present study, plasma T3 levels were significantly decreased in VDN-induced rat VC. T3 treatment attenuated the calcium content and ALP activity in rat aortas with VC, improved the hemodynamic function. MMI blocked the therapeutic effect of T3 on VC. These results indicated that T3 could be an endogenous anti-VC factor and might be a novel target in keeping calcium homeostasis. However, the physiological and direct target genes of T3 in VSMCs are not known.

VC is an active and complex process, with VSMC phenotype transformation. Once the osteogenic phenotype is induced, cells of the vascular wall begin to express osteogenic markers [14–19]. An increasing number of studies have implicated the upregulation of Runx2 expression in the osteogenic differentiation of VSMCs [32–34].The transcription factor Runx2 is one of the three mammalian members of the Runt-related transcription family [35]. Runx2 has originally been identified as an essential transcription factor in osteoblast differentiation, bone matrix gene expression, and consequently bone mineralization [20]. Runx2 and osterix, which is activated downstream of Runx2 [21, 36], regulate the expressions of various bone marker genes, including ALP, osteocalcin (OCN), OPN and finally, mineralization of bone nodules [37]. Recent study showed that suppression of klotho was associated with upregulation of Runx2 and induced VSMC phenotype transformation [23]. Klotho, known as α-Klotho, a 130-kDa single-pass transmembrane protein, expresses at high levels in renal distal tubular epithelial cells, and to a lesser extent in VSMC. Vascular-produced Klotho acted as an endogenous inhibitor of calcification and as a cofactor for FGF-23 signaling, and mediated vascular resistance to FGF23 in VC [24]. We showed that T3 treatment restored the expression of klotho and decreased the expression of Runx2, Osterix, and OPN, prevented the loss of the lineage markers, SM22α and α-actin. Furthermore, MMI completely abolished the effect of T3 on kotho, Runx2, Osterix, and OPN expression, which demonstrated that the effect was mediated by T3. Therefore, T3 attenuated VC in part by balancing stimulation and inhibition of VSMC calcification, thus maintaining phenotype stabilization.

In conclusion, the local endogenous THs may be impaired during VC. T3 attenuated calcification, prevented the loss of lineage markers, restored the expression of klotho, and downregulated Runx2, Osterix, and OPN expression in rat aortas with VC. T3 may have a powerful protective effect on VC and have potential in preventing and treating VC-accompanied osteoporosis.