The intracellular formation of ROS was detected with the fluorescent probe 5- (and 6-) chloromethyl-2’,7’-dichlorodihydrofluorescein diacetate (CM-H2DCFDA; Molecular Probes). NRK-52E cells (106 cells/ml) were cultured in the presence of 100 μM carboplatin for 24 h or pretreated for 30 min with 50 μM pravastatin and then loaded with 0.1 μg/ml
CM-H2DCFDA at 37°C for 30 min in the dark. Cells were then washed twice with Hank balanced salt solution (HBSS) containing calcium and magnesium. Dichlorodihydrofluorescein (DCF) fluorescence was measured immediately with a charge-couple device (CCD) camera (DP72, Olympus, Melville, NY) coupled to a microscope system (BX51, Olympus) at 100×
magnification.
Statistical analysis
Continuous variables were compared by one-way analysis of variance (ANOVA). When there was a significant difference between groups, multiple comparisons of means were performed with the Bonferroni procedure and type-I error adjustment. All statistical assessments were two-sided and evaluated at the 0.05 level of significance. Statistical analyses were performed with SPSS 15.0 statistics software (SPSS Inc., Chicago, IL). Survival analysis was performed according to the
Kaplan-Meier method, and between-group differences in survival were tested by log-rank test.
31
RESULTS (2)
Carboplatin stimulated apoptosis and caspase activation in vitro and in vivo.
To determine if carboplatin-induced nephrotoxicity occurred via a caspase-dependent pathway in vitro, NRK-52E cells were treated with increasing carboplatin concentrations (50 ~ 800 μM), which resulted in a concentration-dependent increase in the level of caspase-3 (Fig. 8A); this increase continued for at least 48 h (Fig. 8B). We explored the effects of carboplatin in vivo by determining the expression of caspase-3 in kidney cells of mice injected with 100 mg/kg carboplatin. Immunoblotting results showed significantly increased staining for caspase-3 in the kidneys of carboplatin-treated mice (2 days after single treatment; Fig. 8C). Apoptosis in kidney cells of mice treated with carboplatin was also significantly increased to 16% compared to that of control mice (2%) (Fig. 8D).
Pravastatin attenuated carboplatin-induced ROS production, renal cell apoptosis, and caspase-3 expression.
Exposure of NRK-52E cells to 100 μM carboplatin for 24 h induced production of ROS approximately 25-fold that in saline-treated cells, as shown by DCF staining (Fig. 9A). This increase in the level of ROS was significantly attenuated by pretreatment with 50 μM pravastatin for 30 min.
Pravastatin also inhibited carboplatin-induced caspase-3 expression in NRK-52E cells (Fig. 9B).
In vivo experiments showed that injection of mice with pravastatin
reduced not only the rate of apoptosis in kidney cells, from 14% to 3% (Fig.
9C), but also the expression of cleaved, active caspase-3 compared to mice treated with carboplatin alone (Fig. 9D).
Pravastatin improved renal function and survival in mice treated with carboplatin.
Carboplatin treatment resulted in abnormal renal function, as revealed by the levels of serum urea nitrogen and creatinine measured 5 days after injection. This effect was improved by pretreatment with pravastatin (Fig.
10B). In accordance with these markers of renal function, control mice and mice treated with pravastatin showed healthy, histologically comparable tubular systems (Fig. 10Ca, b). However, kidneys of mice treated with carboplatin showed substantial histopathologic changes such as tubular necrosis and dilation, protein casts, and loss of brush borders (Fig. 10Cc).
Injection with pravastatin in addition to carboplatin produced a marked decrease in these features (Fig. 10Cd).
Survival of mice injected with carboplatin was markedly decreased compared to that of control mice. Treatment with pravastatin in addition to carboplatin significantly increased 7-day survival compared to mice treated with carboplatin alone, as determined by Kaplan-Meier analysis (Fig. 10A).
Pravastatin induction of HO-1 in NRK-52E cells involved cyclooxygenase-2 and a PPAR-α pathway.
To test whether the renal protective effect of pravastatin was mediated
33
by HO-1, NRK-52E cells were treated with 20 μM pravastatin for 24 or 48 h, and the expression of HO-1 was determined by immunoblotting. As shown in Fig. 11A, the expression of HO-1 was increased after treatment with pravastatin for 24 h. However, HO-1 expression decreased when pravastatin treatment was continued for 48 h.
A similar temporal response was observed when we examined the effect of pravastatin on the expression of COX-2. To determine whether
overexpression of COX-2 might increase the intracellular level of 15d- PGI2, a ligand for PPARα, we examined the effect of 20 μM pravastatin on the level of 6-keto PGF1α (the stable hydrolysis product of PGI2) with an enzyme immunoassay detection system. As observed for the expression of COX-2, pravastatin also induced an increase in 6-keto PGF1α at 24 h (Fig.
11B). Similarly, pravastatin increased the level of PPAR-α and HO-1 in NRK-52E cells (Fig. 11C).
We next examined whether inhibition of PPAR-α and COX-2 affected the expression of HO-1. When PPAR-α was inhibited by transfection of pravastatin-treated cells with plasmid containing PPAR-α siRNA, the expressions of PPAR-α and HO-1 were reduced (Fig. 11C). When the COX-2 inhibitor NS-398 (Cayman Michigan, USA) was added to
pravastatin-treated cells, the increase in HO-1 level was again significantly decreased, and administration of NS-398 and PPAR-α siRNA in
combination showed a synergistic reduction in the level of HO-1 (Fig.
11D).
Pravastatin enhanced PPAR-α nuclear translocation in NRK-52E cells.
We next examined whether pravastatin enhanced nuclear translocation of PPAR-α. As shown in Fig. 12A, the level of cytosolic PPAR-α was not significantly affected by pravastatin in NRK-52E cells transfected with plasmid containing FLAG-PPAR-α. However, the level of nuclear PPAR-α was significantly increased in pravastatin-treated cells (Fig. 12A lane 3 vs lane 4). Immunostaining also showed that FLAG-PPAR-α translocated from the cytosol to the nucleus in NRK-52E cells treated with pravastatin (Fig. 12Bc). This effect was significantly decreased in cells treated with the COX-2 inhibitor NS-398 (Fig. 12Bd).
Pravastatin-activated HO-1 gene expression is involved in a PPAR-α-dependent pathway in NRK-52E cells.
To clarify whether the increase in expression of HO-1 induced by
pravastatin was dependent on translocation of PPAR-α, a PPRE primer for electrophoretic mobility shift assay was designed using a PPAR-binding sequence located in the HO-1 promoter (position –621 to –500) from rat DNA. PPAR-α DNA-binding activity was observed in nuclear extracts of pravastatin-treated NRK-52E cells, and this was markedly suppressed in cells pretreated with unlabeled primer (Fig. 13A). To further examine whether the PPAR-α protein was located in the nucleus and associated with PPRE, a ChIP assay was performed with control and pravastatin-treated cells. As shown in Fig. 13B, the association of PPAR-α with the PPRE region of the HO-1 promoter was increased in pravastatin-treated cells.
35
Finally, to determine the role of pravastatin-induced PPAR-α in activating PPRE and HO-1 promoter activity, NRK-52E cells were cotransfected with increasing amounts of a luciferase expression vector with the PPRE reporter or a 4.5-kb human HO-1 promoter-reporter
construct, along with pcDNA3-RXR and pcDNA3-FLAG- PPAR-α or the PPAR-α agonist WY14643(Cayman Michigan, USA). As shown in Fig.
13C, PPAR-α or RXR-α alone slightly increased luciferase activity, but combined transfection with PPAR-α and RXR-α increased luciferase activity by approximately 50-fold in cells with the PPRE reporter and
approximately 6-fold in cells with the HO-1 promoter construct. In addition, treatment with pravastatin augmented luciferase activity in both the PPRE and HO-1 promoters compared to cells untreated with pravastatin or transfected with PPAR-α or RXR-α alone. Treatment with WY14643 also increased luciferase activity by approximately 50-fold in cells with the PPRE reporter and approximately 8-fold in cells with the HO-1 promoter.
Pravastatin cotreatment further increased luciferase activity in both the PPRE and HO-1 promoters compared to untreated cells.
Pravastatin induced HO-1 expression via a PPAR-α-dependent pathway in mice.
We next subjected C57/B6, H129 wild type (PPAR-α+/+), and PPAR-α–/–
mice to pravastatin injection. After injection of pravastatin for 2 days, expression of PPAR-α and HO-1 was increased in renal extracts of wild-type C57/B6 (Fig. 14A; mRNA and protein) and wild-type W129
mice (Fig. 14B; protein). However, the increase in HO-1 expression was markedly attenuated in PPAR-α–/– mice (Fig. 14B).
37
DISCUSSION (2)
It has been report that prostacyclin can be induced by statins and mediate its cardioprotective effects(Yochai et al. 2005). The cardioprotective effects of atorvastatin are mediated by increased production of PGs achieved by upregulation of cPLA2, COX-2, PGI2 synthase and 6-keto-PGF1a
(Birnbaum et al. 2005). This attract our interest to use the statin as a PGI2 activator to evaluate the implication of PGI2/PPAR-α pathway to renal injury other than I/R.
Statins are competitive inhibitors of 3-hydroxy-3-methylglutaryl- coenzyme A (HMG-CoA) reductase, which regulates the synthesis of cholesterol from mevalonic acid (Goldstein et al. 1990). Because mevalonate is the precursor not only of cholesterol but also of many
nonsteroidal compounds; inhibition of HMG-CoA reductase by statins may result in pleiotropic effects such as anti-inflammatory and
antiarteriosclerotic actions beyond lipid reduction (Faggiotto et al. 1999, Rosenson et al. 1998, Ni et al. 2001). The renoprotective effects of statins have been reported in ischemia-reperfusion injury (Joyce et al. 2001), subtotal renal ablation (Lee et al. 1993), diabetic nephropathy (Kim et al.
2000), and unilateral ureteral obstruction (Moriyama et al. 2001).
In a previous study, we showed that carboplatin-induced cardiotoxicity was characterized by marked increases in apoptosis (3- to 5-fold) and in the levels of caspase-3 and ROS in cardiomyocytes and that these effects could be ameliorated by treatment with pravastatin (Cheng et al. 2008).
Nevertheless, the benefits of statins in a mouse model of cisplatin-induced nephropathy have yet to be reported.
In the present study, we extended our work to include the effects of carboplatin in the kidney. Our findings of increased apoptosis and cleaved caspase-3 in cultured kidney cells and in vivo demonstrated that
carboplatin had similar effects to those observed in the cardiac system. In addition, we showed that the levels of ROS also increased in kidney cells exposed to carboplatin, suggesting the involvement of ROS in
carboplatin-induced apoptosis. Results of the present study indicate that pretreatment with pravastatin markedly decreased carboplatin-induced renal tissue damage and ameliorated renal dysfunction, consistent with the results of Li et al. (Li et al. 2004), who showed that pravastatin normalized serum creatinine levels in a rat model of chronic cyclosporine-induced nephropathy.
The cardioprotective effects of atorvastatin have been shown to involve the upregulation of COX-2 and increased production n of PGI2 (Birnbaum
et al. 2005). When we examined the effects of pravastatin on the
expression of COX-2 and PGI2,we found that both were increased in response to pravastatin, suggesting that in kidney cells also, pravastatin stimulates overexpression of COX-2, resulting in increased PGI2.
PGI2 is known to mediate the translation of PPAR-α (Chen et al. 2009).
In the previous work, we showed that nuclear PPAR-α was increased in 39
NRK-52E renal epithelial cells in response to treatment with pravastatin and that this increase resulted from translocation of PPAR-α from the cytoplasm to the nucleus. The observation that translocation was blocked by the addition of the COX-2 inhibitor NS-398 supports the hypothesis that the translocation was mediated by PGI2.
Many lines of inquiry have pointed to a prominent cytoprotective role for HO-1 in modulating tissue responses to injury (Sikorski et al. 2004). In a rat model of renal ischemia-reperfusion injury, Gueler et al. (Gueler et al.
2007) showed that pretreatment with cervistatin decreased renal damage and dysfunction after ischemia-reperfusion. These authors found that HO-1 expression was increased after ischemia-reperfusion, that the increased expression was significantly greater in rats pretreated with cervistatin, and that the protective effect of statin was completely abolished by cotreatment with a competitive inhibitor of HO-1. Grosser et al. reported that treatment of endothelial cells with statins resulted in activation of the HO-1 promoter, along with accumulation of HO-1 transcript and protein (Grosser et al.
2004). In a study of vascular smooth muscles cells, Lee et al. reported that simvastatin increased the level of HO-1 and suggested that p38 and the phosphatidylinositol-3-kinase and protein kinase B (PI3K-Akt) pathway might be involved (Lee et al. 2004). Our present findings of significantly increased levels of HO-1 transcript and protein in the kidneys of
pravastatin-pretreated mice after carboplatin-induced renal injury are consistent with these previous findings in rats. In addition, we
demonstrated that this effect was mediated by PGI2 and PPAR-α. This was
further corroborated by our in vitro data, in which we showed that
pravastatin treatment increased expression of PGI2, PPAR-α, and HO-1 in cultured renal tubule cells.
The work of Krönke et al. in human vascular cells showed that the HO-1 promoter contains a PPRE and that HO-1 is transcriptionally regulated by PPAR-α (Krönke et al. 2007). We showed PPRE-binding activity in nuclear extracts of pravastatin-treated NRK-52E cells, that PPAR-α was associated with the HO-1 promoter, and that the association was stimulated by pravastatin. Finally we showed that PPAR-α and RXR-α increased expression of a plasmid construct containing the HO-1 promoter and that this expression was stimulated by pravastatin.
Our in vitro results showing inhibition of pravastatin-induced HO-1 activation by NS-398 and PPAR-α siRNA further support the hypothesis that, in kidney cells, activation of HO-1 by pravastatin involves in a PPAR-α-dependent pathway. Our experiments in PPAR-α-knockout mice confirmed that in vivo, as in vitro, the increase in HO-1 expression
stimulated by treatment with pravastatin was, to some extent, dependent on a PPAR-α pathway.
Taken together, our data support a relation between pravastatin, PGI2, PPAR-α, and HO-1. Pravastatin induces overexpression of COX-2, which upregulates PGI2. PGI2 then promotes the translocation of PPAR-α to the nucleus, where it binds to the PPRE of HO-1 promoter and induces the expression of HO-1.
41
Because HO-1 plays a cytoprotective role in modulating the responses of many tissues to different types of injury and pathologic states, it is not surprising that there would be numerous complex tissue-specific inducers that lead to its increased expression. The cardioprotective effects of
atorvastatin have been shown to be mediated by increased production of multiple prostaglandins, including cytosolic phospholipase A2 (cPLA2), PGI2, and PGE2, via upregulation of COX-2 (Birnbaum et al., 2005).
Whether prostaglandins other than PGI2, such as PGE2 and PGJ2, augment PPAR-α translocation and activate HO-1 expression in our
carboplatin-induced renal injury model will require further study.
Extension our study by useing PGIS knock-out animal model or PGIS inhibitor is carefully considered.
In addition, PPARs other than PPAR-α can be activated by statins. For example, Yano et al. reported that statins activate PPAR-γ in macrophages (Yano et al. 2007). We have not ruled out a role for other PPARs such as PPAR-β/δ or PPAR-γ in our model.
HO-1 degrades heme into CO and biliverdin, which have powerful anti-inflammatory, antiapoptotic, and antioxidant effects. This likely accounts for the beneficial effect of pravastatin in inhibiting
carboplatin-induced nephrotoxicity. However, these heme degradation products are also potentially injurious. Indeed, simvastatin induction of HO-1 has been shown to be mediated by nuclear factor erythroid 2-related
factor 2 (Nrf2) in Neuro 2A cells, and its upregulation was significantly associated with increased apoptosis in that system (Hsieh et al. 2008).
To assay for possible iron toxicity induced by overexpression of HO-1, which can result in increased generation of ROS and inflammation, we overexpressed HO-1 by adenovirus infection in neuronal PC-12 cells.
Results showed that overexpression of HO-1 induced ferric iron deposition, which was decreased by treatment with the iron chelator deferoxamine. We also assayed PC-12 cell cytotoxicity and showed that overexpression of HO-1 induced more cytotoxicity in PC-12 cells than in NRK-52 cells, indicating that the effect of HO-1 may be tissue dependent. This implies the limited use of HO-1 activator, or it may benefit to kidney but toxic to other tissues or organs. Thus, the multiple pathways for induction of HO-1 may work synergistically to optimize its expression under many different circumstances.
Besides, the pathway we have proposed is likely to be one of several, if not many, that regulates the expression of HO-1 and one of several that mediates the pleiotropic effects of statins. More evidences and data are required for conclusion.
Our present data showed that carboplatin at a single high dose (100 mg/kg) induced renal dysfunction in mice, as evidenced by elevated levels of blood urea nitrogen and creatinine. It is important to note that these parameters of nephrotoxicity were observed at least 5 days after
43
administration of carboplatin; whereas the survival of mice continued to decline until day 7. The decrease in survival produced by carboplatin in mice may be the result of both nephrotoxicity and cardiomyopathy (Cheng et al., 2008). In addition, impaired activity and suppressed expression of antioxidant enzymes in the kidneys of the mice may also be involved (Husain et al. 2004). Our present results showed that lower doses of carboplatin (50 mg/kg) did not alter the level of blood urea nitrogen or creatinine (data not shown) or 7-day survival in mice. Therefore, both dose and time of carboplatin exposure are probably important in causing
nephrotoxicity.
An interesting observation from the present study is that pravastatin did not completely protect mice against carboplatin-induced nephrotoxicity (Figs. 2 and 3). Similar partial effects of other statins, including cervistatin and atorvastatin, on renal ischemia-reperfusion injury have been reported (Gueler et al. 2007; Gottman et al. 2007). Although the exact reason for the insufficient effect of pravastatin is not clear at present, one possible
explanation involves an inhibitory activity of pravastatin on the mammalian target of rapamycin (mTOR) pathway (Roudier et al. 2006). Therefore, increases in the mTOR response as a consequence of pravastatin treatment may partially mask pravastatin’s beneficial effect on carboplatin-induced kidney injury. Additional studies of mTOR activity are needed to further explore whether these mechanisms are important for the antiapoptotic effect mediated by pravastatin in vitro and in vivo.
In the present study, we administered pravastatin to mice at a dose of 1 mg/kg. This dose was lower than the daily therapeutic dose in humans (20–80 mg/kg/day) (McLean et al. 2008). However, when we increased the dose to 10 mg/kg in mice subjected to carboplatin treatment, the survival rate was significantly decreased compared to that in mice receiving doses of 5 mg/kg or 1 mg/kg (data not shown). This result is in conflict with that of a previous study that used pravastatin (100–150 mg/kg) to inhibit
ischemia-reperfusion–induced nephrotoxicity (Sharyo et al. 2008). It is possible that the high dose of carboplatin used in our present study may have augmented the expression of inducible nitric oxide synthase (Ikeda et
al. 2001) and contributed to more production of oxidative stress than that in
the ischemia-reperfusion model. On the basis of our results, we conclude that low doses of pravastatin have the potential to provide effectiveprotection against nephrotoxicity induced by carboplatin. However, higher doses must be administrated carefully. Further clinical evaluations in
humans are needed to determine whether pravastatin would be an attractive drug for the treatment of acute renal injury.
Besides, the types and dosage of statins are major limiting factors in the use of these drugs, because of adverse effects such as myositis and
rhabdomyolysis, especially when renal function was impaired. Unlike lipophilic HMG-CoA reductase inhibitors, pravastatin is hydrophilic, and its metabolism is independent of that of cytochrome P-450 3A4 in the liver (Williams et al. 2002). As a result, pravastatin may have fewer toxic effects than other statins. It has been reported to be efficacious and tolerated well
45
by transplant recipients undergoing cyclosporine therapy (Yoshimura et al.
1992).
In another study, two doses of pravastatin (5 or 20 mg/kg) were used in cyclosporine-induced nephropathy. Interestingly, the lower dose only improved renal histopathology, whereas the higher dose improved both histopathology and renal function. This observation suggests that, different dose of pravastatin may activate different signaling pathway. (Can et al.
2004.)
The functions of pravastatin also extend to immunomodulation, as demonstrated by the inhibition of the expression of class II major histocompatibility antigens and of natural killer cell activity
(Weitz-Schmidt et al. 2001). Indeed, experimental and clinical studies in cardiac and renal transplant recipients show that the administration of pravastatin successfully decreases acute or chronic rejection episodes and improves graft survival (Ji et al. 2002, Katznelson et al. 1996,
Kobashigawa et al. 1995). Furthermore, combined treatment with
Kobashigawa et al. 1995). Furthermore, combined treatment with