The antioxidant tempol ameliorates arterial medial calcification in uremic rats: Important role of oxidative stress in the pathogenesis of vascular calcification in chronic kidney disease

Authors


Abstract

Vascular calcification is closely related to cardiovascular morbidity and mortality. Accumulating data indicate that oxidative stress is associated with dysfunction of various organs, including cardiovascular diseases in chronic kidney disease (CKD). However, it remains undetermined if oxidative stress induced by uremia promotes arterial medial calcification. The present study investigated the role of oxidative stress in the pathogenesis of arterial medial calcification in uremic rats. Rats with uremia induced by adenine-rich diet progressively developed arterial medial calcification, which was accompanied by time-dependent increases in both aortic and systemic oxidative stress. Immunohistochemical and biochemical analyses showed that the arterial medial calcification progressed in a time-dependent manner that is parallel to the osteogenic transdifferentiation of vascular smooth muscle cells. Accumulation of oxidative stress was also identified in the calcified regions. Time-course studies indicated that both oxidative stress and hyperphosphatemia correlated with arterial medial calcification. Tempol, an antioxidant, ameliorated osteogenic transdifferentiation of vascular smooth muscle cells and arterial medial calcification in uremic rats, together with reduction in aortic and systemic oxidative stress levels, without affecting serum biochemical parameters. Our data suggest that oxidative stress induced by uremia can play a role in the pathogenesis of vascular calcification in CKD, and that antioxidants such as tempol are potentially useful in preventing the progression of vascular calcification in CKD. © 2012 American Society for Bone and Mineral Research

Introduction

Vascular calcification (VC) is highly prevalent in patients with chronic kidney diseases (CKD) and is closely related to cardiovascular disease morbidity and mortality.1, 2 VC is classified into arterial intimal calcification and arterial medial calcification (AMC), and the latter is a characteristic feature of CKD. Recent studies indicate that the pathology of VC is multifactorial and is induced by various complex mechanisms.3 Among them, a phenotypic change in vascular smooth muscle cells (VSMCs), namely the transdifferentiation of VSMCs into osteoblast-like cells, is one of the suggested mechanisms that can promote the development of VC.4, 5 The osteogenic transdifferentiation of VSMC is promoted by a variety of stimuli, such as uremic serum, hyperglycemia, inflammatory cytokines, and hyperphosphatemia.6–11 Hyperphosphatemia, in particular, which has recently been integrated into CKD-mineral and bone disorder (CKD-MBD),12 is reported to be closely related to VC and cardiovascular disease events.13–15 Treatment of CKD-MBD, including treatment of hyperphosphatemia, has therefore been the cornerstone in halting AMC. However, CKD-MBD characterized by hyperphosphatemia and hypercalcemia is less prominent in the early stages of CKD,16 whereas the prevalence of VC is relatively high at this stage.17–19 These results suggest the presence of some undetermined factors that elevate during early CKD and promote VC.

Mounting evidence points to the important pathological role of oxidative stress in various organ dysfunctions.20, 21 Oxidative stress occurs when oxidant production exceeds local antioxidant capacity, resulting in increased oxidation of important macromolecules, including proteins, lipids, carbohydrates, and nucleic acids. Oxidative stress is also elevated in CKD,22, 23 and is involved in the pathogenesis of cardiovascular disease.24–26 However, it remains unknown if oxidative stress also plays a role in the pathogenesis of VC in CKD.

The present study was designed to assess the role of oxidative stress in the pathogenesis of VC in uremia, and the effect of antioxidants on the progression of VC. For this purpose, we undertook protocols to (1) determine the chronological association of AMC with oxidative stress in uremic rats (Protocol 1), and (2) examine the effect of tempol, an antioxidant, on AMC in uremic rats (Protocol 2).

Methods

Animals

All protocols were reviewed and approved by the Ethics Committee on Animal Experimentation, Kyushu University Faculty of Medicine. Male Sprague-Dawley (SD) rats were purchased from Kyudo (Fukuoka, Japan). Animals were housed in a climate-controlled space with 12-hour day/night cycles and allowed free access to food and water. CE-2 (calcium [Ca] 1.0%, phosphorus [P] 1.2%; CLEA Japan Inc., Tokyo, Japan) was used as a standard chow, and adenine-rich chow (0.75% adenine, Ca 1.0%, and P 1.2%; KBT Oriental, Kyoto, Japan) was used to induce CKD-MBD.27 One day before being killed, all rats were housed in metabolic cages, and urine was collected for 24 hours. At the death, blood, abdominal aorta, kidney, urine, and heart were collected. The abdominal aorta were divided into three portions (1 cm for each), and prepared for subsequent analysis. The right kidney; lower portion of the heart, and middle and lower portions of the aorta; and urine were stored at −80°C until analysis. The left kidney, upper portion of the heart, and upper portion of the aorta were fixed with 10% formalin and embedded in paraffin until further use.

Experimental protocols

The time course study, Protocol 1

Male SD rats (8 weeks old, n = 64) were fed standard chow during the first 14 days of adaptation. On day 0 of the postadaptation period, 16 rats were killed, and 24 rats were started on the adenine chow (the adenine-fed renal failure group; ADN), while the remaining 24 rats (control group; CNT) were fed standard chow. All rats were provided with normal tap water ad libitum. Eight rats of each of the ADN and CNT groups were killed on day 14 (week 2), day 28 (week 4), and day 42 (week 6).

Treatment study, Protocol 2

Male SD rats (10 weeks old, n = 36) were fed standard chow for 14 days of adaptation. On day 0 of the postadaptation period, they were randomly subdivided into the following three groups (n = 12 for each group): CNT; standard chow and normal tap water, ADN; adenine chow and normal tap water, TMP; adenine chow and tap water containing 3 mM of 4-hydroxy-2,2,6,6,-tetramethyl piperidinoxyl (tempol) (Sigma, St Louis, MO, USA). The concentration of tempol in the drinking water was determined in a series of preliminary experiments and matched that selected by other investigators.28 Tempol aqueous solution given in an opaque container was changed every 2 days. In this arm of the study, all rats of all three groups were killed at day 42 (week 6).

Biochemical parameters

Serum and urinary levels of Ca, P, albumin, urea nitrogen (UN), and creatinine (Cr) were analyzed using a Hitachi 7170 Autoanalyzer (Hitachi, Tokyo, Japan). Serum parathyroid hormone (PTH) was measured using a two-site rat intact PTH (1-84) ELISA kit (Immutopics, San Clemente, CA, USA). Serum calcitriol was determined using a 1,25-dihydroxyvitamin D RIA kit (TBF; Immunodiagnostic Systems, Boldon, UK). Plasma and urinary levels of 8-hydroxy-2'-deoxyguanosine (8-OHdG) were determined using an 8-OHdG Check ELISA kit (JaICA Nikken Seil Co., Ltd., Shizuoka, Japan).

Examination of tissue calcification

Five-micrometer sections from paraffin-embedded aorta, kidney, and heart were deparaffinized and processed for von Kossa staining. Ca content in the lower part of the abdominal aorta and the right kidney were measured after acid digestion, as described previously.29 Briefly, the tissue sample was weighed and then hydrolyzed in 0.5 mL of 6 M HCl for 24 hours. Ca content was then determined using a commercially available kit (Calcium E-test Wako, Wako, Osaka, Japan). The results were expressed relative to the wet tissue weight. Aortic valve calcification was analyzed semi-quantitatively as follows. Briefly, the extent of aortic valve calcification in each group was scored using a 4-scale system based on the percentage of the calcified area, as assessed by von Kossa staining, relative to the whole valve area; score 0, 0%; score 1+; 1% to 10%, score 2+; 11% to 25%, and score 3+; representing percent calcification of 26% to 50%.

Immunohistochemistry

Immunohistochemistry was performed as described previously.30 Briefly, 2-µM sections were deparaffinized, rehydrated, and microwaved in 0.01 M citrate buffer for antigen retrieval. After inactivation of intrinsic peroxidase by incubation in 0.3% hydrogen peroxidase, the sections were treated with 5% skim milk for 30 minutes at room temperature and then incubated in a humidified chamber at 37°C for an hour with anti-runt-related gene 2 (Runx2) (1:50; Santa Cruz Biotechnology, CA, USA), osteocalcin (1:50; Santa Cruz Biotechnology), nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 4 (Nox4) (1:25; Abcum, London, UK), 8-OHdG (1:200; JaICA), or 4-hydroxy-2-nonenal (4-HNE) (1:100; JaICA) antibodies. After washing, the sections were incubated with biotinylated secondary antibody for 30 minutes at room temperature, followed by incubation with horseradish peroxidase-conjugated streptavidin. The horseradish peroxidase was visualized by reaction with 3,3'-diaminobenzidine tetrahydrochloride and hydrogen peroxide. Digital micrographs of the immunohistochemistry were captured on an Eclipse E800 microscope (Nikon, Tokyo, Japan). The stained areas were estimated and expressed in arbitrary units (AU) using image analysis software.31

RNA isolation and real-time polymerase chain reaction (PCR)

RNA isolation and real-time PCR were performed as described previously.30 Briefly, total RNA was extracted from the aorta using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) and used to prepare complementary DNA by reverse transcription using a PrimeScript RT reagent Kit (Perfect Real Time; Takara Bio Inc., Otsu, Japan). Real-time quantitative PCR was performed using the SYBR Premix Ex Taq (Takara Bio), the Light Cycler System 3302 (Roche Diagnostics, Mannheim, Germany), and the following primers purchased from Takara Bio: glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (RA015380), Runx2 (RA045439), osteocalcin (RA009113), and Nox4 (RA016732). The mRNA levels were expressed relative to those of GAPDH mRNA.

Statistical analysis

All data were analyzed using the StatView v7 program (Abacus Concepts, Berkeley, CA, USA). All values represent mean ± SEM. Differences between two groups were analyzed by the unpaired t-test for each time point or by Dunnett's test. Differences among the three groups were examined by one-way analysis of variance followed by Bonferroni post-hoc test. The Pearson single correlation coefficient was used to determine associations between two parameters. P values < 0.05 were considered statistically significant.

Results

Protocol 1

Serial changes in biochemical parameters in ADN rats

None of the experimental rats died during the study period. Table 1 lists the results of weekly serum and urinary biochemical analyses. CKD-MBD was induced by feeding the rat adenine-rich chow. Compared with CNT, ADN rats showed significant and time-dependent increases in serum levels of Cr, UN, P, and Ca-P product throughout the study period (at weeks 2, 4, and 6); serum levels of Cr, UN, and P were significantly higher at week 2 before VC became evident at week 4 (Fig. 1A).

Table 1. Protocol 1: Serum Values of Various Biochemical Parameters
Serum biochemistryCNTADN
Week 0 (n = 8)Week 2 (n = 8)Week 4 (n = 8)Week 6 (n = 8)Week 0 (n = 8)Week 2 (n = 8)Week 4 (n = 8)Week 6 (n = 8)
  • CNT, control rats; ADN, adenine-fed renal failure rats; Ca, calcium, P, phosphorus.

  • Data are mean ± SEM.

  • Serum samples were collected on day 0 (week 0), day 14 (week 2), day 28 (week 4), and day 42 (week 6).

  • *

    p < 0.05 versus CNT at each corresponding week.

Urea nitrogen (mmol/L)5.7 ± 0.26.0 ± 0.28.3 ± 0.410.1 ± 0.55.7 ± 0.29.8 ± 1.5*14.1 ± 2.2*38.0 ± 12.1*
Creatinine (µmol/L)20.8 ± 1.022.8 ± 1.323.8 ± 0.931.0 ± 0.820.6 ± 1.1116.6 ± 7.9*229.3 ± 13.2*304.1 ± 0.1*
Ca (mmol/L)2.46 ± 0.022.47 ± 0.022.51 ± 0.012.57 ± 0.062.47 ± 0.012.38 ± 0.042.32 ± 0.032.72 ± 0.19
P (mmol/L)2.77 ± 0.062.75 ± 0.062.57 ± 0.042.67 ± 0.042.78 ± 0.053.02 ± 0.11*4.69 ± 0.25*7.11 ± 0.28*
Ca-P product (mmol2/L2)6.80 ± 0.156.79 ± 0.186.47 ± 0.116.87 ± 0.196.78 ± 0.167.29 ± 0.31*10.87 ± 0.60*19.23 ± 1.36*
Figure 1.

Serial changes in vascular calcification in the abdominal aorta (Protocol 1).(A) Representative photomicrographs of the abdominal aorta stained by von Kossa staining at each stage in each group (original magnification, × 40). Calcification is located in the media of the aorta. (B) Ca content in the abdominal aorta was expressed as mg/g wet tissue weight at each stage in each group. Data are mean ± SEM. *p < 0.05 versus CNT.

Serial changes in aortic calcification in ADN rats

Figure 1A shows representative photomicrographs of the abdominal aorta stained with von Kossa stain. Calcification of the media layer of the aorta in ADN increased significantly at weeks 4 and 6, while CNT did not develop VC throughout the study period. The VC was quantified by Ca content in the abdominal aorta at weeks 0, 2, 4, and 6. The aortic Ca contents in ADN at weeks 4 and 6 were significantly higher than those in CNT at the respective time periods (CNT vs AND; 2.80 ± 0.11 vs 4.37 ± 0.44, 2.84 ± 0.13 vs 9.28 ± 1.53 mg/g wet tissue weight, at weeks 4 and 6, respectively) (Fig. 1B).

Serial phenotypic changes in aortic VSMCs in ADN rats

Fig. 2A shows representative images of immunohistochemical staining for Runx2 and osteocalcin in the aorta at week 6. The expression levels of both Runx2 and osteocalcin protein were significantly higher at weeks 4 and 6 (Fig. 2B), as were the mRNA expression levels of both markers assessed by real-time PCR (Fig. 2C).

Figure 2.

Serial osteoblastic-phenotypic changes in the aorta (Protocol 1). (A) Representative photomicrographs of immunohistochemistry for Runx2 and osteocalcin at week 6 in each group. (B) Semi-quantification of the Runx2 and osteocalcin expression assessed by immunohistochemistry at each week in both groups. (C) Relative mRNA expression of Runx2 and osteocalcin for GAPDH in the aorta of ADN at each week compared with CNT at week 0. In each image, the top side is the intima and the bottom side is the adventitia (original magnification, × 200). Data are mean ± SEM. *p < 0.05 versus CNT. AU; arbitrary unit.

Serial changes in aortic and systemic oxidative stress levels

Fig. 3A shows representative microphotographs of the immunohistochemical staining for 8-OHdG and Nox4 at week 6. The accumulation of 8-OHdG in the aorta appeared at week 2, and increased significantly at weeks 4 and 6 (Fig. 3B). The urinary 8-OHdG content, a marker of systemic oxidative stress, also started to increase at week 2, and significant elevations were recognized at week 4 in ADN compared with CNT (Fig. 3C). The semi-quantitative analysis indicated a significant increase in the protein expression of Nox4 at weeks 4 and 6 (Fig. 3B), as did the mRNA expression of Nox4 assessed by real-time PCR (Fig. 3D).

Figure 3.

Serial changes in aortic and systemic oxidative stress levels (Protocol 1). (A) Representative photomicrographs of immunohistochemistry for 8-OHdG and Nox4 in aorta sections at week 6 in each group. (B) Semi-quantification of the 8-OHdG and Nox4 immunohistochemistry conducted at each week in both groups. (C) Urinary 8-OHdG levels at each stage in each group. (D) Relative mRNA expression of Nox4 in the aorta of ADN rats at each week compared with CNT rats at week 0. In each image, the top side is the intima and the bottom side is the adventitia (original magnification, × 200). Data are mean ± SEM. *p < 0.05 versus CNT. AU; arbitrary unit.

Localization of VC, oxidative stress, and osteogenic transdifferentiation in the aorta of ADN rats

Fig. 4 shows microphotographs of immunohistochemistry of serial sections of the aorta from CNT and ADN rats at week 6. Immunoreactivity for Runx2 was positive only in the calcified lesions of the aorta in ADN. Immunoreactivities for 8-OHdG and 4-HNE, which are markers of oxidative stress for nucleic acid and lipid, respectively, were similarly positive in both calcified and non-calcified areas of the arterial media, whereas immunohistochemical staining of Nox4 was stronger in the calcified areas of ADN arterial media at week 6 than in non-calcified areas.

Figure 4.

Comparison of localization of vascular calcification, oxidative stress, and phenotypic changes in the aorta (Protocol 1). Representative photomicrographs of von Kossa staining and immunohistochemistry for Runx2, 8-OHdG, 4-HNE, and Nox4 using serial sections of abdominal aorta from each group. The upper panels represent CNT, and the lower panels represent ADN samples. In each image, the top side is the intima and the bottom side is the adventitia (original magnification, × 200).

Correlation of VC with hyperphosphatemia and oxidative stress

We analyzed the relationship among hyperphosphatemia, systemic oxidative stress, and VC using the data of ADN rats at weeks 4 and 6 (n = 16). Aortic Ca content correlated significantly with both serum P level (r = 0.689) and urinary 8-OHdG level (r = 0.699) (Fig. 5A, B).

Figure 5.

Relationship among urinary 8-OHdG, serum phosphorus, and Ca content in the aorta (Protocol 1). Correlation between urinary 8-OHdG level and Ca content in the aorta (A), or Ca content and serum phosphorus (B). These regression analyses were performed using data from the uremic rats at weeks 4 and 6 (n = 16).

Protocol 2

Effects of tempol on biochemical parameters

Treatment of rats with tempol did not affect food and fluid intake (data not shown). As shown in Table 2, serum levels of UN, Cr, Ca, P, Ca-P product, and PTH in ADN and TMP rats were significantly higher than those in CNT, but there were no significant differences in serum levels of UN, Cr, Ca, P, and Ca-P product, and PTH between TMP and ADN groups. Serum calcitriol levels in ADN and TMP were significantly lower than those in CNT, whereas there were no significant differences between ADN and TMP groups.

Table 2. Protocol 2: Serum Values of Various Biochemical Parameters
Serum biochemistryCNT (n = 12)ADN (n = 12)TMP (n = 12)
  • CNT, control rats; ADN, adenine-fed renal failure rats; TMP, renal failure rats treated with tempol; Ca, calcium; P, phosphorus; PTH, parathyroid hormone; 1,25(OH)2D3, 1,25-dihydroxyvitamin D3.

  • Data are mean ± SEM.

  • Serum samples were collected at day 42 (week 6).

  • *

    p < 0.05 versus CNT.

Urea nitrogen (mmol/L)9.5 ± 0.636.2 ± 20.8*42.1 ± 5.0*
Creatinine (µmol/L)29.2 ± 0.9341.6 ± 24.8*308.8 ± 22.1*
Ca (mmol/L)2.5 ± 0.12.4 ± 0.22.3 ± 0.1
P (mmol/L)2.7 ± 0.16.9 ± 0.5*6.4 ± 0.4*
Ca-P product (mmol2/L2)6.8 ± 0.216.2 ± 1.4*14.8 ± 1.2*
PTH (pmol/L)9.4 ± 2.286.4 ± 7.0*81.9 ± 6.3*
1,25(OH)2D3 (pmol/L)392.1 ± 42.648.1 ± 10.1*52.8 ± 8.4*

Effects of tempol on VC and phenotypic changes

Fig. 6A shows representative microphotographs of von Kossa–stained abdominal aortic calcification in each experimental group. Aortic Ca contents in both ADN (9.8 ± 1.03 mg/g wet tissue weight) and TMP (6.55 ± 0.61 mg/g wet tissue weight) groups were significantly higher than in CNT (2.96 ± 0.14 mg/g wet tissue weight), whereas that in TMP rats was significantly lower compared with ADN rats (Fig. 6B). Immunostaining for Runx2 and osteocalcin showed lower expression in TMP than in ADN (Fig. 6C). Administration of tempol significantly reduced the mRNA expression of Runx2 and osteocalcin in the aorta of TMP rats compared with that in ADN (Fig. 6D).

Figure 6.

Effects of tempol on arterial medial calcification and phenotypic change (Protocol 2). (A) Representative photomicrographs of von Kossa staining in the aorta from each group. (B) Quantification of Ca content in the abdominal aorta. Representative photomicrographs of the immunohistochemistry for Runx2 and osteocalcin in each group (C). In each photograph, the top side is the intima and the bottom side is the adventitia (original magnification, × 200). Relative mRNA expression of Runx2 and osteocalcin in the aorta from each group (D). Data are mean ± SEM. *p < 0.05 versus CNT; **p < 0.05 versus ADN.

Effects of tempol on kidney and aortic valve calcification

Supplemental Fig. S1A and S1C show representative microphotographs of the kidney and aortic valve stained by the von Kossa method in each group. In the ADN and TMP groups, Ca-P crystals deposited mainly in the tubular lumen, and a small amount of Ca-P crystals also deposited in the interstitium. Renal Ca contents of the ADN (1.85 ± 1.23 mg/g wet tissue weight) and TMP rats (1.78 ± 0.89 mg/g wet tissue weight) were significantly higher than that in CNT (0.33 ± 0.07 mg/g wet tissue weight), whereas there were no significant differences between ADN and TMP (Supplemental Fig. S1B). As for aortic valve calcification, the calcified site of the aortic valve was mainly on the base of aortic valve. Semi-quantitative analysis indicated higher calcification scores in the ADN and TMP groups than in the CNT, whereas that of the TMP was lower than that of the ADN group (Supplemental Fig. S1D).

Effects of tempol on aortic and systemic oxidative stress

Tempol reduced both urinary and plasma levels of 8-OHdG induced by uremia (Supplemental Fig. S2A and B). Immunohistochemical staining for both 8-OHdG and 4-HNE in TMP was attenuated compared with that in ADN (Fig. 7A, B, and C), and the mRNA expression of Nox4 was lower in the aorta of TMP compared with ADN (Fig. 7D).

Figure 7.

Effects of tempol on markers of oxidative stress and NADPH oxidase expression in the aorta (Protocol 2). (A) Immunohistochemical staining of the abdominal aorta for 8-OHdG and 4-HNE. Semi-quantification of vessel wall staining for 8-OHdG (B) and 4-HNE (C). In each image, the top side is the intima and the bottom side is the adventitia (original magnification, × 200). (D) Relative mRNA expression of Nox4 by real-time PCR. Data are mean ± SEM. *p < 0.05 versus CNT; **p < 0.05 versus ADN. AU; arbitrary unit.

Discussion

The present in vivo study highlights the role of oxidative stress in promoting uremia-related VC. The results showed that adenine-fed rats developed AMC accompanied by a time-dependent increase in aortic and systemic oxidative stress levels, in which are parallel to the development of progressive renal failure. The extent of AMC correlated with the severity of oxidative stress and serum P levels. The study also showed that treatment with tempol protected against AMC and lowered aortic and systemic oxidative stress levels, probably through inhibition of VSMCs transdifferentiation into osteoblast-like cells.

VC is highly prevalent in CKD,1, 2 and is promoted by various factors such as hyperphosphatemia, hypercalcemia, and high serum PTH. Hyperphosphatemia in particular is one of the most critical accelerators of VC in CKD.4 Consistent with previous studies,32 hyperphosphatemia correlated positively with VC in the present study. Growing evidence indicates that there is increased oxidative stress in CKD patients,22, 23 and recent experimental studies have also confirmed the correlation between oxidative stress and the development of VC.33–37 In this regard, Koleganova et al.38 showed that there were increased levels of 4-HNE and other oxidative stress-related markers in the calcified regions of the aorta in CKD patients. Based on these results, we hypothesized that oxidative stress is also involved in the progression of VC in uremia.

In the present study, we used adenine-rich diet to induce CKD. Adenine is absorbed from the gastrointestinal tract and is excreted by the kidney. During the renal excretion process, adenine forms crystals, which could obstruct the renal tubules, potentially leading to tubulointerstitial fibrosis and chronic renal failure.39 The adenine-fed rats are known to develop hypocalcemia, hyperphosphatemia, calcitriol deficiency, hyperparathyroidism, bone disorder, and extensive AMC.27 Furthermore, because adenine-fed rats show increased oxidative stress in various organs, this model is useful for examining the relationship between oxidative stress and VC.

The most important finding of the present study was that inhibition of oxidative stress protected against VC in chronic renal failure model rat. Tempol, a membrane-permeable antioxidant that exerts multiple anti-oxidative effects, including superoxide dismutase mimetic action and inhibition of phenton reaction,40 improves various organ dysfunctions by reducing tissue oxidative stress levels.41–43 In the present rat model of CKD-MBD, tempol ameliorated AMC, suggesting that the regulation of oxidative stress seems to be important in the prevention of VC. For instance, the phenotypic changes in VSMCs induced by hyperphosphatemia are also promoted by oxidative stress, and inhibition of oxidative stress by antioxidants almost inhibited the calcification.34, 36 In the present study, tempol ameliorated AMC without affecting serum levels of phosphorus, PTH, or vitamin D. In addition, AMC correlated with both oxidative stress and hyperphosphatemia in a similar fashion, indicating that the reduction of oxidative stress is as important as hyperphosphatemia in the prevention of VC in uremia. Further, recent in vitro studies have shown that indoxyl sulfate promoted transdifferentiation of VSMCs into osteoblast-like cells by increasing oxidative stress.43 These results indicate that VC is mediated by oxidative stress induced by uremic milieu.

What is the underlying mechanism whereby tempol protects against VC when it inhibits oxidative stress? Based on the results of the present study and previous reports, there are at least two plausible mechanisms. The pathogenic role of the phenotypic change of VSMCs into osteoblast-like cells has been confirmed in various animal models.44, 45 In the present study, tempol reduced the expression levels of Runx2 and osteocalcin in the rat aorta. A recent in vitro study also showed that tempol blocked the transdifferentiation of aortic valve cells into osteoblast-like cells, leading to decreased Ca deposits.34 Thus, tempol might protect against AMC by reducing oxidative stress and as a consequence inhibiting the VSMC phenotypic change. Another possibility is that tempol attenuates AMC by inhibiting apoptosis of VSMC. Although our study did not address this scenario, Jagadeesha et al.42 reported that tempol attenuated apoptosis of VSMC in the carotid artery with balloon catheter injury in rats, whereas Son et al.46 reported in their in vitro study of VSMC that apoptosis plays a crucial role in the pathogenesis of Ca deposition via Akt intracellular signaling. Thus, evidence so far suggests that tempol seems to reduce VC by inhibiting oxidative stress in the aorta and consequent prevention of both phenotypic changes and apoptosis of VSMCs in the aortic media.

We have also shown in the present study that NADPH oxidase induced by uremia could contribute to the increased ROS production in the aorta. The Nox family is a multisubunit enzyme that catalyzes Omath image by the 1-electron reduction of O2 using NADPH.47 Recently, Nox family expression was detected in vascular cells including VSMC.48, 49 Muteliefu et al.43 showed in vitro that indoxyl sulfate promotes Ca deposition via the production of ROS induced by Nox4 upregulation. A more recent study has reported that even hyperphosphatemia itself could induce osteoblast-like phenotypic changes via ROS production, and that this phenotypic change is almost abolished by an NADPH oxidase inhibitor.34 Our study showed that Nox4, was time-dependently upregulated in the aortic media of uremic rats. These results indicate that VSMCs produce ROS in response to various uremia-related stimuli by upregulating NADPH oxidase. However, the increase in 8-OHdG in the present study preceded upregulation of NADPH oxidase, indicating that the initial increase in oxidative stress in the aorta might be induced by a non-NADPH oxidase-dependent pathway. Further studies are required to determine the underlying mechanism of action of tempol on VC in uremia.

The effects of tempol on other organs should be also elucidated. In the present study, tempol ameliorated aortic valve calcification, whereas it had no effect on renal calcification. These results were plausible. Apart from VC, renal calcification is a simple deposition of Ca-P crystals, and is mainly determined by the extent of urinary Ca and P excretion.50 In fact, urinary Ca excretion levels in ADN and TMP rats were comparable (data not shown). In contrast to renal calcification, aortic valve calcification is a cell-mediated process including phenotypic change into osteoblast-like cells.51 Thus, like VC, tempol might attenuate aortic valve calcification in our study.

Our studies have some limitations that need to be addressed. First, we did not determine the effect of tempol on VC in other models such as 5/6 nephrectomized rat of different oxidative stress level. Liberman et al.35 reported that tempol augmented aortic valve calcification via overproduction of hydrogen peroxide in a rabbit model of relatively low oxidative stress condition, induced by high fat and vitamin D diet. Conversely, tempol inhibited VC in our model of relatively high oxidative stress level. Because tempol exerts its antioxidant capacity by both SOD mimetic and non-SOD mimetic actions,40 we speculate that the net ROS reduction by both actions might determine the effect of tempol on VC, and might depend on the oxidative stress condition when tempol is used. If tempol is used in a low oxidative stress model, increased hydrogen peroxide can aggravate VC. In contrast, if tempol is used in a relatively high oxidative stress model, the net tempol effect is a decrease in the net oxidative stress level, leading to the attenuation of VC. Actually, Nox4 expression was decreased by tempol in the present study, whereas it increased in the Liberman et al. study. Because Nox4 is known to be upregulated under high oxidative stress conditions,52 and known to make vicious circle of producing ROS, these results indicate that tempol in our study lowered the net oxidative stress level, leading to inhibition of VC. Second, we only showed the association of oxidative stress with VC but did not provide direct evidence that oxidative stress promoted VC, although the effect of tempol is mainly mediated through its anti-oxidative properties. Hence, further studies are required to determine the precise mechanism of tempol in the prevention of VC in uremia.

In conclusion, we showed in a CKD-MBD model rat that AMC and oxidative stress progressed parallel to the adenine-induced renal dysfunction, and that AMC showed equivalent and positive correlations with both oxidative stress and hyperphosphatemia. Treatment with anti-oxidant tempol inhibited AMC in the CKD-MBD model. These data provide evidence for the involvement of oxidative stress as a potential mechanism of VC in CKD, and for the clinical application of antioxidants to prevent VC in CKD patients.

Disclosures

All the authors state that they have no conflicts of interest.

Acknowledgements

This study was supported in part by a grant from the Kidney Foundation, Japan (JKFB09-42) to SY. We thank Dr FG Issa (www.word-medex.com.au) for the careful reading and editing of the manuscript.

Authors' roles: Study design: SY. Study conduct: SY and HN. Data collection: SY. Data analysis: SY, M Taniguchi, and M Tokumoto. Data interpretation: M Tokumoto and HN. Drafting manuscript: SY, M Tokumoto, and KF. Revising manuscript content: KT and TK. Approving final version of manuscript: SY, M Taniguchi, M Tokumoto, JT, KF, HN, MI, KT, and TK. KT takes responsibility for the integrity of the data analysis.

Ancillary