Vasopressin Type 1A Receptor Up-regulation by Cyclosporin A in Vascular Smooth Muscle Cells Is Mediated by Superoxide*

Based on our previous results, we investigated whether cyclosporin A (CsA)-induced vasopressin type 1A receptor up-regulation was mediated by free radicals. We report that CsA analogues with different affinities for cyclophilin and calcineurin were able to up-regulate vasopressin type 1A receptor and to generate free radicals in smooth muscle cells independently of calcineurin. Further, we demonstrate that the antioxidant N -acetyl- L -cysteine blocked the increase in vaso- pressin type 1A receptor mRNA and protein levels induced by CsA and that low concentrations of pro-oxidants were able to directly increase vasopressin type 1A receptor mRNA and protein levels. In addition, short exposure to CsA or pro-oxidants was sufficient to significantly increase vasopressin type 1A receptor mRNA and protein levels. Using cell-permeable forms of superoxide dismutase and catalase, we finally show that superoxide mediates the CsA-induced effects on vasopressin type 1A receptor. These results provide strong evidence that CsA-induced superoxide generation is causally involved in vasopressin type 1A receptor expression and demonstrate for the first

The cyclic undecapeptide cyclosporin A (CsA) 1 is the most widely used immunosuppressive drug to prevent transplant rejection and in the therapy of autoimmune diseases (1). CsA acts by binding to cyclophilin to inhibit calcineurin phosphatase activity, NF-AT dephosphorylation, and interleukin-2 expression, thus preventing T-lymphocyte proliferation (1)(2)(3).
The use of CsA is accompanied by mild to severe side effects, and the clinically most important are nephrotoxicity and hypertension (4 -6). Both are likely caused by CsA-induced local vasoconstriction (7,8). We have shown previously (9,10) that CsA caused an elevation in cytosolic free calcium concentrations in rat and human aortic smooth muscle cells and enhanced vasoconstriction of rat aortic smooth muscle cells (RASMC) when these were stimulated with vasoconstrictor hormones such as endothelin-1, serotonin, angiotensin II, and vasopressin. At least for vasopressin and angiotensin II, the expression of their respective cell surface receptors was shown to be increased by CsA in rat and human aortic smooth muscle cells (10 -12). Recently, we have shown that CsA up-regulated the vasopressin type 1A (V 1A ) receptor via an increase in the corresponding mRNA levels in RASMC (13). Up-regulation of vasoconstrictor hormone receptors may be responsible for the enhanced vasoconstriction under in vivo conditions, thus leading, via an increase in peripheral resistance, to hypertension and to a decrease in glomerular filtration (4 -6). However, the exact mechanisms by which CsA enhances vasoconstriction have not yet been clarified.
Several lines of evidence point to a possible role of reactive oxygen species (ROS) as mediators leading to the side effects of CsA (14 -17). Some studies have shown that CsA is able to produce ROS in vascular endothelial and mesangial cells (16,18,19). In addition, we have shown that CsA generates ROS in RASMC that were inhibited by antioxidants (20). ROS have been assigned a role of biological mediators in cellular signaling (21)(22)(23). For example, it has been shown that plateletderived growth factor-induced cell proliferation was dependent on the cellular production of H 2 O 2 (24). In addition, it has been demonstrated that ROS are able to activate several transcription factors, such as nuclear factor-B and activated protein-1 (25)(26)(27)(28). Recent studies have illustrated that not only H 2 O 2 but also superoxide is involved in cellular signaling (22, 29 -32). Thus, ROS seem to play a role as activators of signaling pathways leading to altered gene expression.
Using CsA analogues with different capacities to bind cyclophilin and calcineurin we demonstrate that CsA generates ROS and up-regulates V 1A receptor in RASMC independently of cyclophilin and calcineurin. Further, we report that CsA generates superoxide and that it is likely the superoxide anion radical that is able to directly increase V 1A receptor expression at the mRNA and protein levels.
* This work was supported by grants from the Swiss National Science Foundation (31-68315.02 and 3100A0-100513), the Roche Research Foundation, and the Fondation Herbette of the University of Lausanne. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
ʈ To whom correspondence should be addressed. Cultures of Smooth Muscle Cells-RASMC were prepared from aortae of male Wistar Kyoto rats (200 -300 g) as described (9). RASMC were cultured in DMEM supplemented with essential and non-essential amino acids, vitamins, 10 g/ml ciproxin, and 10% fetal calf serum and kept at 37°C in a humidified atmosphere of 5% CO 2 in air. For studies with CsA, cells were used at confluence (after 7 to 9 days of culture) between passages 6 and 11. Twenty-four h before experiments, culture media were replaced with fresh DMEM without fetal calf serum. Treatment of RASMC with CsA and other agents was always performed in serum-free DMEM.
[ 3 H]AVP Binding-Binding experiments were carried out as described (13). Results are expressed in percent of control to normalize for inter-experimental variations. Control values ranged from 0.5 to 4.2 fmol/well in different experiments, and the intra-experimental variability was less than 5%.
Measurements of ROS by 2Ј,7Ј-DCFH-ROS measurements were performed as outlined in Ref. 20. Results are expressed in percent of control to normalize for inter-experimental variations. Because of the variability between experiments, the gain applied to the photomultiplier tube was adjusted before each experiment.
Isolation of Total RNA, DNase Treatment, and Reverse Transcription-Total RNA from RASMC cultured in 6-well plates was isolated using a RNeasy Mini Kit according to the manufacturer's instructions. All RNAs were quantified by spectrophotometrical determination of the absorption at 260 nm. The RNA integrity was assessed by electrophoresis in non-denaturing 1.2% agarose gels stained with ethidium bromide. DNase treatment and reverse transcription were executed as described (13).
Real-time Quantitative PCR Analysis-Real-time quantitative PCR analysis was performed with a PerkinElmer Life Sciences 7700 Sequence Detector (PerkinElmer Life Sciences) as illustrated (13). Results were normalized for GAPDH expression for each sample. Results are expressed in percent of control to normalize for inter-experimental variations. V 1A receptor/GAPDH ratios of controls varied from 0.51 to 1.04 in different experiments, and the intra-experimental variability was less than 5%.
Data Analysis-Results are presented as the means of at least three independent experiments with bars indicating S.E. Statistical evaluation was performed by means of one-way analysis of variance followed by Newman Keuls or Dunnett post-tests using the software lnPlotPrism (GraphPad Software, San Diego, CA). Differences with a value of p Ͻ 0.05 were considered significant.  Fig. 1 illustrates that all analogues (tested at 1 M) were able to significantly increase V 1A receptor expression in RASMC after a 20-h pretreatment at 37°C. This response was of similar magnitude as for CsA (about 3-fold), thus suggesting that CsAinduced V 1A receptor up-regulation occurred independently of cyclophilin and/or calcineurin inhibition.

CsA and Analogues
ROS Formation by CsA Analogues in RASMC-As V 1A receptor up-regulation appears to occur independently of the calcineurin/NF-AT pathway, we investigated whether ROS could be involved. To assess whether CsA requires cyclophilin and/or calcineurin for the generation of ROS, we tested whether analogues of CsA would be able to generate ROS. Because the oxidation of 2Ј,7Ј-DCFH has been used in many laboratories to detect cellular radical formation (33-36), we used this approach to measure ROS in RASMC. Cells were treated with CsA or the CsA analogues tested for V 1A receptor up-regulation and 2Ј,7Ј-dichlorofluorescein (DCF) fluorescence was measured after 1 h. Fig. 2 demonstrates that all analogues were able to produce ROS. From these results, we conclude that CsA and analogues generate ROS independently of their ability to bind to cyclophilin and/or calcineurin. In experiments that are not shown, we verified that ROS formation by CsA analogues were blocked by antioxidants.
Effect of Antioxidants on CsA-induced Increase in V 1A Receptor mRNA and Protein Levels in RASMC-As CsA and all analogues tested generated ROS and increased V 1A receptor expression, we tested whether the CsA-induced increase in V 1A receptor expression could be blocked by antioxidants. Because it is known that several transcription factors are regulated by changes in the redox balance, the two antioxidants N-acetyl-Lcysteine (NAC) and ␣-tocopherol were tested on CsA-induced increase in V 1A receptor mRNA and protein levels. Fig. 3  receptor mRNA levels induced by CsA after a 6-h treatment (A), but this antioxidant also decreased CsA-induced V 1A receptor protein up-regulation after a 20-h treatment (B). The apparently higher levels of [ 3 H]AVP binding in cells treated with CsA and NAC were statistically not different from control. Co-treatment of RASMC with CsA and ␣-tocopherol was equally effective in blocking CsA-induced V 1A receptor up-regulation (data not shown).
Effect of Pro-oxidants on V 1A Receptor mRNA and Protein Levels-The fact that antioxidants were able to block CsAmediated increase in V 1A receptor mRNA and protein levels suggests that CsA-induced ROS generation may be responsible for this effect. To address this question, RASMC were treated with the peroxides hydrogen peroxide (H 2 O 2 ) or tert-butylhydroperoxide (BHP) or the superoxide-generating systems dimethylnaphthoquinone (DMNQ) or hypoxanthine (HX)/xanthine oxidase (XO). V 1A receptor mRNA expression was determined 6 h after treatment, and V 1A receptor protein expression was determined 20 h after treatment. As shown in Fig. 4, all pro-oxidants were able to significantly increase V 1A receptor mRNA (A) and V 1A receptor protein (B) levels in RASMC.
Effect of Incubation Time on CsA-or Pro-oxidants-induced V 1A Receptor Expression-The number of V 1A receptor was earlier found to be maximal after 20 h of treatment even though cell surface receptor expression was already significantly elevated after 8 -12 h (13). In addition, we have shown that the effect of CsA on receptor protein expression was preceded by an increase in V 1A receptor mRNA levels starting at 2 h and peaking at 6 -10 h after CsA addition (13). These data suggest that shorter times of CsA exposure may be sufficient to induce V 1A receptor up-regulation. If ROS are the critical mediators of this response, it is questionable whether ROS are produced for the full 20 h. Our data have shown that CsAinduced ROS generation was increased within the first hour, peaking at 45 min, followed by a steady decline up to 2 h (20). Additional investigations have shown that after 20 h CsA does no longer generate any detectable levels of ROS when incubated with 2Ј,7Ј-DCFH for the last hour (data not shown). Also, the pro-oxidants used in Fig. 4 very likely produced ROS only for a limited time before either the substrate was exhausted (HX), or the chemical was inactivated (H 2 O 2 , BHP, DMNQ). Therefore, we investigated whether short exposure of RASMC to CsA or pro-oxidants was sufficient to trigger V 1A receptor up-regulation detected after 20 h. For this, cells were incubated with either CsA or DMNQ or HX/XO for 30 min or 1 h. Then the compounds were removed, and the cells were incubated in medium without compounds for a total time of 20 h before [ 3 H]AVP binding was measured. Fig. 5 shows that exposure of RASMC to CsA or pro-oxidants for as little as 30 min was sufficient to significantly up-regulate V 1A receptors after 20 h.

The Effect of PEG-SOD or PEG-CAT on CsA-induced ROS Formation and V 1A Receptor Up-regulation-
To test what kind of radical was generated by CsA and served as mediator of CsA-induced V 1A receptor expression, we examined the effects of PEG-SOD that dismutates superoxide to hydrogen peroxide and PEG-CAT that reduces hydrogen peroxide to water (37,38). RASMC were pretreated for 6 h with PEG-SOD (100 units/ ml) or PEG-CAT (50 units/ml), and DCF-fluorescence and [ 3 H]AVP binding were determined in the presence or absence of CsA as described under "Experimental Procedures." Fig. 6A shows that PEG-SOD increased the CsA-induced DCF-signal whereas PEG-CAT decreased this signal. On the other hand, PEG-SOD decreased CsA-mediated V 1A receptor up-regulation whereas PEG-CAT had no effect (Fig. 6B)

DISCUSSION
In this study, we provide evidence that superoxide serves as mediator in CsA-induced increase in V 1A receptor mRNA and protein expression whereas the contribution of the calcineurin/ NF-AT pathway in this effect is negligible. In antigen-triggered receptor signaling, calcium release will activate calcineurin to dephosphorylate NF-AT, which will translocate into the nucleus and induce transcription of genes such as interleukin-2. To exert its immunosuppressive activity, CsA has to bind to cyclophilin, and this complex will inhibit the protein phosphatase activity of calcineurin and consequently prevent the dephosphorylation of NF-AT and interleukin-2 expression (1).
We have shown earlier that 1 M CsA was able to increase the number of V 1A cell surface receptors (12) via an increase in the corresponding mRNA levels (13). This effect could be mediated by transcriptional up-regulation of the V 1A receptor promoter involving the calcineurin/NF-AT pathway. Calcineurin has also been shown to play an important role in the regulation of other transcription factors, including NF-B (39), Jun (40,41), cAMP-response element-binding protein (42), Elk-1 (43), and others. Therefore, we studied whether CsAinduced V 1A receptor up-regulation depended on calcineurin using CsA analogues with different affinities for cyclophilin and calcineurin. Our results demonstrate that all CsA analogues tested were able to increase [ 3 H]AVP binding independently of their ability to bind to cyclophilin and/or calcineurin. In addition, as calcineurin is activated by calcium, sequestering of cytosolic calcium with the cell-permeable calcium chelator 1,2bis-(o-aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid tetra-(acetoxymethyl) ester (BAPTA/AM) would inhibit calcineurindependent signaling. In independent experiments, we saw that treatment of RASMC with 10 M BAPTA/AM was not able to block CsA-induced [ 3 H]AVP binding (data not shown). Together, these results suggest that a role of calcineurin in CsAinduced V 1A receptor up-regulation is highly unlikely.
In previous work we have shown that 1 M CsA (representing peak plasma CsA concentrations encountered in CsA-treated patients (44)) produced significant amounts of ROS in RASMC that could be blocked by antioxidants (20). Another study has also found that 1 M CsA generated ROS in cardiomyocytes (45). Small amounts of ROS are indispensable for many biochemical processes (23,46). In recent years, considerable evidence has accumulated supporting a role for ROS as activators of signaling pathways and transcription factors (22,28,46). When the balance between pro-and antioxidants is disturbed in favor of the former, a situation of oxidative stress ensues (23). It is now well established that oxidative stress is involved in the pathogenesis of numerous diseases (22).
Because CsA-induced V 1A receptor up-regulation occurred independently of calcineurin, our next goal was to elucidate whether the generation of ROS by CsA is implicated in the mechanisms leading to CsA-induced vasoconstriction, because oxidative stress has been proposed as a causative factor in the toxic side effects of CsA (14 -16, 47 (48).
Our focus was to first investigate whether CsA-induced ROS formation could lead to V 1A receptor up-regulation. Because all CsA analogues increased [ 3 H]AVP binding, we tested whether they were able to produce ROS. Our results show that all CsA analogues tested were indeed able to generate significant amounts of ROS, independently of cyclophilin and/or calcineurin binding. Even if the relative fluorescence intensity varied between the analogues, ROS formation by CsA appears to be independent from cyclophilin or calcineurin. The small differences in fluorescence intensity could be related to the chemical structure of the CsA analogues. Because ROS can interfere directly (49) or indirectly via calmodulin kinase II (50) with calcineurin signaling it may be possible that ROS generated by CsA analogues could still act via calcineurin signaling. However, these two studies used considerably higher concentrations of pro-oxidants compared with this study. In fact, in additional experiments using DMNQ to inhibit calcineurin in an interleukin-2 reporter gene assay (51) concentrations in excess of 1 M were needed to inhibit calcineurin signaling (IC 50 ϭ 6 M) as compared with 10 nM DMNQ to increase V 1A receptor expression (data not shown). These results suggest that ROS could serve as mediators in CsA-induced V 1A receptor up-regulation but act independently of calcineurin.
To further investigate whether ROS are involved in CsAinduced vasoconstriction, the effect of antioxidants was assessed on CsA-mediated increase in V 1A receptor at the mRNA and protein levels. Our findings demonstrate that NAC blocked the increase in V 1A receptor mRNA and protein levels induced by CsA. Similar data were obtained with ␣-tocopherol (data not shown). Taken together, these data further support a causal role of ROS in CsA-mediated increase in V 1A receptor mRNA and protein levels. It is interesting to note that both hydro-and lipophilic antioxidants were able to block the CsA-induced V 1A receptor up-regulation suggesting a possible membrane component in CsA-induced ROS signaling. It should also be noted that it is unlikely that CsA can react to form radicals directly.
Our  6A). The apparent increase in DCF fluorescence with PEG-SOD treatment suggests that the dismutation reaction was the rate-limiting step. It is interesting that this occurs despite the presence of large amounts of endogenous SOD in the cells and rapid spontaneous superoxide dismutation rates. Alternatively, it may be considered that excess SOD diverts superoxide from other redox reactions (for instance with nitrogen monoxide to form peroxynitrite), but we suggest that nitrogen monoxide and peroxynitrite do not play a role in CsA-induced DCFH oxidation. 2 Second, if superoxide mediated the CsA-induced V 1A receptor up-regulation, it would be expected that PEG-SOD would decrease whereas PEG-CAT would not affect the CsA effect. Our results confirm this notion and suggest that superoxide but not H 2 O 2 was responsible for the CsA effect on V 1A receptor upregulation. In line with these results are additional data using the superoxide scavengers 5-diethoxyphosphoryl-5-methyl-1pyrroline-N-oxide (1 mM) and coelenterazine ( . is most probably directly mediating the increase in V 1A receptor mRNA and protein levels in RASMC.