Nitric Oxide Down-regulates MKP-3 mRNA Levels

MAP kinase-dependent phosphorylation processes have been shown to interfere with the degradation of the antiapoptotic protein Bcl-2. The cytosolic MAP kinase phosphatase MAP kinase phosphatase-3 (MKP-3) induces apoptosis of endothelial cells in response to tumor necrosis factor α (TNFα) via dephosphorylation of the MAP kinase ERK1/2, leading to Bcl-2 proteolysis. Here we report that the endothelial cell survival factor nitric oxide (NO) down-regulated MKP-3 by destabilization of MKP-3 mRNA. This effect of NO was paralleled by a decrease in MKP-3 protein levels. Moreover, ERK1/2 was found to be protected against TNFα-induced dephosphorylation by coincubation of endothelial cells with the NO donor. Subsequently, both the decrease in Bcl-2 protein levels and the mitochondrial release of cytochrome c in response to TNFα were largely prevented by exogenous NO. In cells overexpressing MKP-3, no differences in phosphatase activity in the presence or absence of NO were found, excluding potential posttranslational modifications of MKP-3 protein by NO. These data demonstrate that upstream of theS-nitrosylation of caspase-3, NO exerts additional antiapoptotic effects in endothelial cells, which rely on the down-regulation of MKP-3 mRNA.

Here we report that the endothelial cell survival factor nitric oxide (NO) down-regulated MKP-3 by destabilization of MKP-3 mRNA. This effect of NO was paralleled by a decrease in MKP-3 protein levels. Moreover, ERK1/2 was found to be protected against TNF␣-induced dephosphorylation by coincubation of endothelial cells with the NO donor. Subsequently, both the decrease in Bcl-2 protein levels and the mitochondrial release of cytochrome c in response to TNF␣ were largely prevented by exogenous NO. In cells overexpressing MKP-3, no differences in phosphatase activity in the presence or absence of NO were found, excluding potential posttranslational modifications of MKP-3 protein by NO. These data demonstrate that upstream of the S-nitrosylation of caspase-3, NO exerts additional antiapoptotic effects in endothelial cells, which rely on the down-regulation of MKP-3 mRNA.
Apoptosis is the enzymatically controlled form of cell death induced by stimulation of distinct cellular signal transduction pathways, as opposed to the lethal cell damage that is known as necrosis (1). In the past few years, several signaling systems have been identified that control apoptotic cell death (2)(3)(4). Whereas the caspase cascade executes the apoptotic pathway, MAP kinases 1 are involved in modulating various regulatory pathways of the cell death machinery (5). Whereas the c-Jun N-terminal MAP kinase (stress-activated protein kinase) promotes apoptosis in various cell types (6), the MAP kinase ERK1/2 exerts prosurvival functions (7). Recently, ERK1/2 was shown to regulate protein levels of the antiapoptotic Bcl-2, thus linking ERK1/2 with the apoptotic signaling complex (8). In detail, by maintaining Bcl-2 in its phosphorylated status, ERK1/2 prevents Bcl-2 from ubiquitination, thereby inhibiting its degradation via the proteasome complex (8,9). Bcl-2 in turn prevents the mitochondrial release of cytochrome c (10), an event that leads to formation of the apoptosome complex ultimately culminating in the activation of the executioner caspase-3 (11).
Besides the well established pro-apoptotic effect elicited by high concentrations of nitric oxide (NO), NO also exerts potent antiapoptotic effects in a variety of cells (12,13). Several interactions of NO with the apoptotic signaling machinery have been postulated to explain the apoptosis inhibitory effects of NO. NO was shown to nitrosate not only the apoptosis executing enzyme caspase-3, where different apoptotic pathways converge (13,14), but also caspase-6, -7, and -8 (15,16). Furthermore, NO has been implicated to inhibit caspase-dependent Bcl-2 cleavage and, consequently, the release of mitochondrial cytochrome c in MCF-7 hepatocytes and endothelial cells (17)(18)(19). Importantly, NO was also reported to interact with p21 ras and MAP kinase signaling (20). In Jurkat cells, NO was shown to activate the MAP kinases c-Jun N-terminal kinase and, though to a lesser extent, p38 and ERK1/2 by S-nitrosation of p21 ras (21).
Here we address the effects of NO to interfere with the dephosphorylation of ERK1/2 as a potential target of the antiapoptotic capacity of NO in endothelial cells. We demonstrate that the down-regulation of the cytosolic MAP kinase phosphatase-3 (MKP-3) (22), which is known to dephosphorylate ERK1/2 (23), maintains ERK1/2 active and, thus, inhibits the execution of apoptosis by preventing Bcl-2 degradation and mitochondrial release of cytochrome c.
Plasmids and Transfection-MKP-3 was amplified by polymerase chain reaction with oligonucleotides that were synthesized to contain BamHI and EcoRV restriction sites and subsequently cloned into the respective sites of the pcDNA3.1-MycHis vector (InVitrogen, the Netherlands). Transient transfection of HUVEC was performed by incubation of 3.0 ϫ 10 5 cells/6-cm well with 3 g of plasmid as described previously (25). To transiently transfect COS-7 cells, 7 g of pcDNA3.1 plasmid containing the respective insert were employed using Superfect (Qiagen, Hilden, Germany).
MKP-3-induced Cell Death-Following the cotransfection of HUVEC with pcDNA3.1-lacZ (1 g) and either pcDNA3.1-MKP-3 (2 g) or the pcDNA3.1 control vector (2 g), the transfected cells were identified by ␤-galactosidase staining. Viable versus dead stained cells were counted by two blinded investigators, and the results were expressed as dead/ viable cells ϫ100. In addition, potential differences in cell death rate because of necrosis were excluded by measuring lactate dehydrogenase release.
Determination of MKP-3 Messenger RNA-RNA was prepared according to Batt et al. (27), and 10 g was loaded on 0.8% formamideagarose gels. RNA was blotted on nylon membranes, and the blots were hybridized with a radioactively labeled full-length human MKP-3 probe and incubated for 24 h. Then the blots were washed (0.1% SDS, 0.2ϫ SSC) and exposed to x-ray films.
Measurement of MKP-3 RNA Transcription Rate-For preparation of nuclei, cells were detached with trypsin and lysed with Nonidet P-40. Nuclei (2 ϫ 10 6 ) were separated by a 20.5% sucrose gradient and incubated in the presence of ATP, GTP, CTP, and [ 32 P]UTP for 30 min at 30°C to allow for the transcription of 32 P-labeled mRNA. Then, RNA was extracted essentially as stated above (27). To prepare hybridization membranes, human full-length MKP-3 cDNA (100 g) or glyceraldehyde-3-phosphate dehydrogenase cDNA (50 g) were blotted onto nylon membranes using a dot blot device (Scotlab, Coatbridge, UK). Blots were cross-linked, hybridized with the radioactively labeled transcripts for 24 h at 65°C, washed (0.1% SDS, 2ϫ SSC), and exposed to x-ray films.
Statistical Analysis-Data are expressed as mean Ϯ S.D. or mean Ϯ S.E. as indicated from at least three independent experiments. Statistical analysis was performed by one-way ANOVA (variance: least significant difference test).

NO Prevents TNF␣-induced ERK1/2 Dephosphorylation, Bcl-2
Degradation, and Mitochondrial Cytochrome c Release-To characterize a potential interference of NO with apoptotic signal transduction involving the MAP kinase p42/p44 (ERK1/2), HUVEC were stimulated with TNF␣ in the presence or absence of the exogenous NO donor SNP. Then, phosphorylation of ERK1/2 was determined by Western blot analysis using a phosphospecific antibody. As shown in Fig. 1A, stimulation of endothelial cells with the proapoptotic cytokine TNF␣ resulted in a time-dependent dephosphorylation of ERK1/2, as described previously (8). In contrast, the exogenous NO donor SNP abrogated ERK1/2 dephosphorylation by TNF␣ at all time points examined (Fig. 1A). Thus, exogenous NO interferes with ERK1/2 dephosphorylation in response to TNF␣.
As ERK1/2 dephosphorylation is known to be a prerequisite for degradation of the antiapoptotic protein Bcl-2, the influence of exogenous NO on Bcl-2 protein degradation was investigated. For this purpose, Bcl-2 levels following exposure to TNF␣ in the presence or absence of the NO donors SNP or SNAP were determined by Western blotting. Fig. 1B illustrates that the degradation of Bcl-2 protein following stimulation with TNF␣ is largely prevented by coincubation with SNP or SNAP. To further confirm the functional significance of the observed protective effect of NO on ERK1/2 phosphorylation status and Bcl-2 protein levels, the release of cytochrome c from mitochondria in TNF␣-stimulated cells was measured under the influence of NO. Subcellular protein fractions were isolated from endothelial cells stimulated with TNF␣ in the presence or absence of SNP to separate the cytosolic fraction from the mitochondrial. As depicted in Fig. 1C, the release of cytochrome c from mitochondria in response to TNF␣ is suppressed by coincubation with SNP. These data suggest an inhibitory role of NO in TNF␣-induced apoptosis signaling upstream of the mitochondria by maintaining ERK1/2 phosphorylation. The inhibition of ERK1/2 dephosphorylation renders Bcl-2 resistant against degradation and, subsequently, inhibits the mitochondrial release of cytochrome c.
NO is known to functionally regulate proteins by S-nitrosation of essential cysteine residues (13, 28). The MAP kinase phosphatase MKP-3, which dephosphorylates ERK1/2, contains an essential cysteine residue located at position 293 within its catalytic domain, the mutation of which was shown to inactivate MKP-3 phosphatase activity (23). Therefore, we investigated whether NO exerts an inhibitory effect on the phosphatase activity of MKP-3 on a posttranslational level, possibly by S-nitrosation of the enzyme. Phosphatase activity of HUVEC and COS-7 cell lysates overexpressing MKP-3 was analyzed in the presence and absence of NO donors by use of a pNPP hydrolysis assay. The phosphatase activity in HUVEC overexpressing MKP-3 was not influenced by an 18-h coincubation with SNP in vivo (93% of the enzymatic activity determined in MKP-3-transfected control cells with no SNP added, Fig. 2A) nor by a 30-min in vitro incubation with the NO donor NOC-15 of COS-7 cell extracts following MKP-3 protein expression (91% of control extracts, Fig. 2B). Thus, NO does not appear to modify MKP-3 phosphatase activity on a posttranslational level. Furthermore, we determined the effect of NO on the major downstream signal event ignited by MKP-3, the release of cytochrome c from mitochondria. In MKP-3-transfected cells, the release of cytochrome c into the cytosol induced by MKP-3 overexpression was not influenced by exogenous NO (Fig. 2C).

Down-regulation of MKP-3 mRNA and Protein Levels by Nitric
Oxide-Having demonstrated that NO does not interfere with MKP-3 activity, we investigated a possible regulatory effect of NO on MKP-3 expression. Therefore, MKP-3 mRNA levels were determined following incubation of HUVEC with the NO donors SNP or SNAP for 2, 4, and 6 h. As shown in Fig. 3, A and B, MKP-3 mRNA is markedly down-regulated in the presence of the NO donor SNP (p Ͻ 0.05). Similar results were obtained using the NO donor SNAP (data not shown). The down-regulation of MKP-3 was confirmed on the protein level, as demonstrated by Western blot analysis (Fig. 3C). To assess the effect of endogenously derived NO, endothelial cells were exposed to shear stress, which activates the endothelial NO synthase (29 -31). Exposure of human endothelial cells to constant laminar flow induced a time-dependent decrease in MKP-3 mRNA levels, as shown in Fig. 3D. Again, this effect was paralleled by a reduction in MKP-3 protein levels following the exposure to constant laminar shear stress, which was entirely prevented by the NO synthase inhibitor NG-monomethyl-L-arginine monoacetate (data not shown). To identify the mechanism by which NO decreases MKP-3 expression, MKP-3 mRNA transcription was analyzed by nuclear run-on experiments. As shown in Fig. 4A, the nuclear transcription rate of MKP-3 mRNA was not altered by the NO donor SNP. Therefore, MKP-3 mRNA stability was additionally analyzed by incubation of HUVEC with actinomycin D. SNP significantly reduced the stability of MKP-3 mRNA (Fig. 4B, p Ͻ 0.01), indicating that NO destabilizes rather than transcriptionally down-regulates MKP-3 mRNA. To assess a functional role of MKP-3 for TNF␣-induced ERK1/2 dephosphorylation, MKP-3 mRNA levels were determined following stimulation of endothelial cell with TNF␣. As shown in Fig. 5, TNF␣ induced a prolonged increase in MKP-3 mRNA levels, which was largely suppressed by NO. Taken together, NO derived from exogenous as well as from endogenous sources down-regulates MKP-3 mRNA levels in endothelial cells.
Effect of NO on MKP-3-induced Cell Death-MKP-3 was demonstrated to induce cell death in HUVEC by inactivating ERK1/2 (8,9). Therefore, we characterized the influence of NO on cell death induction by MKP-3. Death rate of HUVEC was significantly increased by overexpression of MKP-3 compared with control cells that were transfected with the pcDNA3.1 vector with no insert (p Ͻ 0.001, Fig. 6, A and C). When cells were exposed to low levels of NO provided by the NO donor SNP, cell death induction by MKP-3 overexpression was almost completely abolished (p Ͻ 0.001, Fig. 6, A and C). Furthermore, shear stress exposure to augment endogenous NO release also reduced MKP-3-induced cell death even below control levels (p Ͻ 0.005, Fig. 6B). This effect was abolished by the addition of the competitive antagonist of NO synthase, NG-monomethyl-L-arginine monoacetate (Fig. 6B). Thus, death signaling in HUVEC induced by MKP-3, under conditions that do not allow for the modulation of MKP-3 expression, is sensitive to exogenous as well as to endogenous NO. Finally, to exclude a potential influence of NO on the ERK1/2-phosphorylating kinase MEK (MAP kinase kinase), we investigated the effects of NO on cell death induction by MKP-3 during pharmacological inhibition of MEK with 2Ј-amino-3Ј-methoxyflavone. As shown in Fig. 6D, NO was still capable of suppressing cell death induction by MKP-3 overexpression even in the presence of the MEK inhibitor 2Ј-amino-3Ј-methoxyflavone. DISCUSSION NO represents a key regulator of endothelial cell survival (16,32). Various interactions of NO with intracellular signal transduction have been described to explain the prosurvival effects of low levels of NO as produced by the endothelial NO synthase. In this study, we addressed the role of the MAP kinase phosphatase MKP-3 as a potential target of the protective effect of NO in endothelial cells. We demonstrate that NO destabilized MKP-3 mRNA and, thereby, interferes with the TNF␣-induced dephosphorylation of the MAP kinase p44/42 (ERK1/2). Subsequently, NO prevents Bcl-2 degradation and the release of cytochrome c from mitochondria, which results in the protection of endothelial cells from apoptosis. ERK1/2 is an established player in the antiapoptotic defense network (7). The Raf/MEK/ERK pathway has previously been reported to confer protection against apoptosis induced by growth factor withdrawal (33,34). Activation of the Ras/MEK/ ERK pathway was shown to phosphorylate the Bcl-2 family member Bad, resulting in the dissociation of Bad from Bcl-x L , which allows the protection of cells from apoptosis (35). In endothelial cells, ERK1/2-dependent phosphorylation was demonstrated to stabilize Bcl-2 and prevent its proteasome-dependent degradation (8,9). In accordance with the prosurvival influence of phosphorylated ERK1/2, its inactivation by MAP kinase phosphatase signaling in response to cellular stress was shown to induce cytotoxicity (8,9,36). Thus, the maintenance of ERK1/2 activation via the regulation of MAP kinase phosphatase gene expression might represent an additional checkpoint in cellular signaling.
NO is capable of interacting with apoptosis signaling in multiple ways. The variety of interactions of NO with proapoptotic signaling events include the maintenance of sustained Bcl-2 expression in B lymphocytes (37) and the inhibition of caspase activity by S-nitrosation of the catalytic cysteine residue (13,15,38) as well as by a cGMP-dependent anti-apoptotic pathway (39). Here, we report the down-regulation of MKP-3 mRNA levels and, thus, the maintenance of ERK1/2 phosphorylation by NO as a novel mechanism that contributes to the protective effects of NO in endothelial cells. Our data demonstrate that NO exerts an inhibitory effect upstream of the ignition of the proapoptotic signaling cascade by inhibiting up-regulation of MKP-3. These findings extend the redundancy of NO to modulate apoptosis signaling to a rather proximal checkpoint of interference as compared with the integrative antiapoptotic blockade achieved by the inhibition of the downstream effector caspase-3 (Fig. 7). Moreover, by stabilizing ERK1/2 phosphorylation via down-regulation of MKP-3, NO may modulate proliferative signals that involve activation of the transcription factors Elk-1 and ATF-2 by ERK1/2 in addition to the prosurvival maintenance of Bcl-2 levels.
Remarkably, the modulation of MKP-3 protein levels by NO is accomplished by regulation of MKP-3 gene expression as opposed to the posttranslational inhibition of caspase protease activity. The down-regulation of MKP-3 mRNA by NO was independent of the transcription rate as shown by nuclear run-on experiments but is caused by a destabilization of MKP-3 mRNA. NO-induced modulation of mRNA stability has previously been reported. The inhibition of NO production in endothelial cells was shown to mimic the induction of thrombospondin-1 mRNA by hypoxia, which is known to rely on the posttranscriptional stabilization of thrombospondin-1 mRNA (40). Moreover, soluble guanylate cyclase mRNA was demonstrated to be destabilized by NO-dependent up-regulation of a destabilizing protein (41). Although the molecular mechanism underlying the regulation of MKP-3 mRNA stability by NO remains elusive, the induction of a destabilizing protein is rather unlikely given the rapid onset of the NO effect.
In conclusion, we demonstrate that following stimulation with TNF␣, low levels of NO maintain ERK1/2 phosphorylation via down-regulation of MKP-3 mRNA levels, thereby providing constant phosphorylation of the ERK1/2 target Bcl-2, which prevents the degradation of Bcl-2 and, subsequently, the release of cytochrome c from mitochondria. Thus, the antiapoptotic multiplicity of the effects of NO extends from checkpoints up-and downstream of cytochrome c release and includes posttranslational modifications of protein function as well as the regulation of gene expression to control apoptotic signaling events. FIG. 7. Possible interference of NO with apoptotic signaling cascades in endothelial cells at various checkpoints. Cell death stimuli lead to the up-regulation of MKP-3, an effect which is inhibited in the presence of low levels of NO by a destabilizing influence on MKP-3 mRNA. Further downstream the apoptotic signal pathway, distal from Bcl-2 degradation and cytochrome c release, NO is established to interfere with the execution of apoptosis by the inhibition of caspase-3 via S-nitrosation (see "Discussion" for details). APAF-1, apoptotic protease activating factor-1.