Differential activation of mitogen-activated protein kinases by nitric oxide-related species.

Many studies have identified nitric oxide (NO) and related chemical species (NOx) as having critical roles in neurotransmission, vasoregulation, and cellular signaling. Previous work in this laboratory has focused on elucidating the mechanism of NOx signaling in cells. We have demonstrated that NOx-induced activation of the guanine nucleotide-binding protein p21ras leads to nuclear translocation of the transcription factor NFκB. Here, we investigated whether intermediary signaling elements, namely the mitogen-activated protein (MAP) kinases, are involved in mediating NOx signaling. We found that NOx activates the extracellular signal-regulated kinase (ERK), p38, and c-Jun NH2-terminal kinase (JNK) subgroups of MAP kinases in human Jurkat T cells. JNK was found to be 100-fold more sensitive to NOx stimulation than p38 and ERK. In addition, the activation of JNK and p38 by NOx was more rapid than ERK activation. Depletion of intracellular glutathione augmented the NOx-induced increase in kinase activity. Furthermore, endogenous NO, generated from NO synthase, activated ERK, and NOx-induced MAP kinase activation was effectively blocked by the farnesyl transferase inhibitor α-hydroxyfarnesylphosphonic acid. These data support the hypothesis that critical signaling kinases, such as ERK, p38, and JNK, are activated by NO-related species and thus participate in NO signal transduction. These findings establish a role for multiple MAP kinase signaling pathways in the cellular response to NOx.

Many studies have identified nitric oxide (NO) and related chemical species (NO x ) as having critical roles in neurotransmission, vasoregulation, and cellular signaling. Previous work in this laboratory has focused on elucidating the mechanism of NO x signaling in cells. We have demonstrated that NO x -induced activation of the guanine nucleotide-binding protein p21 ras leads to nuclear translocation of the transcription factor NFB. Here, we investigated whether intermediary signaling elements, namely the mitogen-activated protein (MAP) kinases, are involved in mediating NO x signaling. We found that NO x activates the extracellular signal-regulated kinase (ERK), p38, and c-Jun NH 2 -terminal kinase (JNK) subgroups of MAP kinases in human Jurkat T cells. JNK was found to be 100-fold more sensitive to NO x stimulation than p38 and ERK. In addition, the activation of JNK and p38 by NO x was more rapid than ERK activation. Depletion of intracellular glutathione augmented the NO x -induced increase in kinase activity. Furthermore, endogenous NO, generated from NO synthase, activated ERK, and NO x -induced MAP kinase activation was effectively blocked by the farnesyl transferase inhibitor ␣-hydroxyfarnesylphosphonic acid. These data support the hypothesis that critical signaling kinases, such as ERK, p38, and JNK, are activated by NO-related species and thus participate in NO signal transduction. These findings establish a role for multiple MAP kinase signaling pathways in the cellular response to NO x .
Our previous work focused on identifying the signaling cascade responsible for the positive regulatory effects of nitric oxide (NO) 1 and related chemical species (NO x ) on human peripheral blood mononuclear cells (1,2). NO x was found to activate human lymphocytes, as evidenced by increased glucose uptake, tumor necrosis factor ␣ secretion, and nuclear translocation of the transcription factor NFB. Further studies demonstrated that NO x directly activates p21 ras , leading to downstream events such as the activation of NFB (3). We have also identified p21 ras as a general signaling target of reactive free radicals and suggested that it may be a cellular sensor for redox stress (4). In those studies, reactive oxygen species were shown to signal downstream of p21 ras by activation of the extracellular signal-regulated kinases (ERKs). Here, we examined whether activation of low molecular weight G proteins such as p21 ras , Rac1, and Cdc42 by NO x also leads to activation of one or more of the mitogen-activated protein (MAP) kinase family members.
The three identified mammalian MAP kinase subgroups include ERK (5), c-Jun NH 2 -terminal kinase (JNK; Refs. 6 -8), and the recently identified p38 MAP kinase (9 -13). The well studied ERK group is typically activated by growth factors via a p21 ras -dependent signal transduction pathway (5,9,14). JNK and p38 MAP kinases respond to proinflammatory cytokines and environmental stress, although it is not clear whether this occurs via common or parallel pathways (5). The signal transduction pathways leading to ERK, JNK, and p38 kinase activation are biochemically and functionally distinct (7,9). Nevertheless, some extracellular stimuli can activate all of the MAP kinases simultaneously (5). In the present study, we attempt to further clarify the NO x -signaling pathway downstream of low molecular weight G proteins such as p21 ras by exploring the effects of NO-related chemical species on the ERK, p38, and JNK subgroups of MAP kinase.

EXPERIMENTAL PROCEDURES
Materials-A saturated phosphate-buffered saline solution containing 1.25 mM NO (Matheson gas) was prepared as described (3). This solution also contains higher oxides of NO (NO x ). S-Nitroso-Nacetylpenicillamine (SNAP) was synthesized as described previously (15). Anti-ERK1 and anti-ERK2, preconjugated to Sepharose, was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit polyclonal antibodies against the human JNK and p38 MAP kinases and the Jun (residues 1-79) and ATF-2 (residues 1-109) GST fusion proteins were prepared as described previously (6,9). Cell culture media, serum, and supplements were purchased from Life Technologies, Inc. The farnesyl transferase inhibitor ␣-hydroxyfarnesylphosphonic acid was purchased from Biomol (Plymouth Meeting, PA), and [␥-32 P]ATP was obtained from Amersham Corp. All other reagents were purchased from Sigma.
Cell Culture-The Jurkat human T-cell leukemia line was propagated in RPMI 1640 medium supplemented with 10% fetal calf serum and 2% L-glutamine. Human umbilical vein endothelial cells were cultured in medium 199 with 20% fetal calf serum and 90 g/ml endothelial cell growth supplement (PerImmune, Inc., Rockville, MD) and heparin (20 g/ml). The culture and treatment of cells was performed in a 37°C 5% CO 2 incubator.
* This work was supported by National Institutes of Health Grants HL46403 and AI37637 (to H. M. L.) and CA58396 (to R. J. D.). 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.

Effect of NO x Gas, NO-generating Compounds, and
Endogenous NO on ERK1 and ERK2 Activity-We examined whether NO x delivered as a gas-saturated solution or from decomposition of a NO-donor compound could stimulate ERK activity in human Jurkat T cells. To test this hypothesis, we treated Jurkat T cells with various concentrations of NO x -saturated phosphate-buffered saline or SNAP in the presence and absence of L-buthionine-(SR)-sulfoximine (BSO). BSO, a selective inhibitor of ␥-glutamylcysteine synthetase, blocks de novo synthesis of glutathione (17), a vital intracellular antioxidant. We have previously used BSO to demonstrate a redox-sensitive signaling pathway in cells (4). In the presence of BSO, NO x and SNAP increased ERK1 and ERK2 activity 2-3-fold in a concentrationdependent manner (Fig. 1). We found that untreated Jurkat cells possessed a measurable amount of basal ERK activity. It is possible that stress-activated mechanisms arising from cell culture may be responsible for the basal activity of ERK. We have previously found a similar amplitude of ERK activation by reactive oxygen species in these cells (4). It is also possible that nonphysiological depletion of cellular glutathione (as accomplished with BSO treatment) results in cell damage, leading to enzymes that are susceptible to activation by NOrelated species. NO x gas elicited a peak response at a concentration range of 0.3-1 M. The NO-generating compound SNAP required 300-1000 M to produce similar levels of ERK activation. The disparity in the activating concentrations of NO x gas versus SNAP is likely due to their different chemical natures. In the cell, NO can exist as several different redox species (18), each with distinctive chemical properties. The relative abundance of these redox forms may explain the concentration-dependent variations in ERK activation seen between NO x and the various NO-generating compounds. Other factors that influence the relative potencies of exogenous NO x donors include decomposition rates, donor transport, and donor stability (18).
Non-BSO-treated cells showed no increase in ERK activity when stimulated with SNAP and required a higher activating concentration when stimulated with NO-related species (Fig.  1). Our previous work has demonstrated that depletion of the intracellular antioxidant glutathione resulted in enhanced oxygen free radical signaling through p21 ras , MAP kinase, and NFB (4).
To explore the physiological relevance of our studies using exogenous NO donors, we induced NO synthesis in human umbilical vein endothelial cells by treating with lipopolysaccharide and interferon-␥ (19). We found that overnight induction of nitric oxide synthase did not yield detectable ERK1 and ERK2 activation (Fig. 2, open bars), although NO was produced (19). To determine whether a pulse of endogenous NO could activate ERK1 and ERK2, we induced nitric oxide synthase overnight in the presence of a competitive inhibitor, N -methyl-L-arginine. This was followed by a 15-min pulse of substrate, L-arginine. Under these conditions, ERK1 and ERK2 activity was dramatically increased (Fig. 2, hatched bars). It is likely that a burst of NO was effective in stimulating ERK activity, whereas slow overnight production was not due to a desensiti-FIG. 1. Effect of NO-generating compounds and NO x gas on ERK1 and ERK2 activity. Jurkat T cells were pretreated for 18 h in serum-free RPMI 1640 medium with or without 100 g/ml BSO. Cells were treated for 10 min in a 37°C and 5% CO 2 incubator with the indicated concentrations of SNAP or NO x -saturated phosphate-buffered saline. MAP kinase assays were performed as described under "Experimental Procedures" using 1 g of myelin basic protein as substrate. Data points represent the means from at least three separate experiments, each performed in duplicate. Standard deviations were less than 5% in all cases.

FIG. 2. Effect of endogenous NO on ERK1 and ERK2 activity.
Human umbilical vein endothelial cells were either untreated or treated for 16 h with lipopolysaccharide (LPS, 30 g/ml) and interferon-␥ (IFN-␥, 50 ng/ml) or N -methyl-L-arginine (NMA; 3 mM). Then, indicated samples were treated with L-arginine (ARG, 10 mM) for 15 min and assayed for ERK activity. Data represent the means and standard deviations (bars) from three separate experiments. zation effect we have previously observed (3). Together, the data suggest that NO x -induced redox stress may activate the ERK kinases and that this signal may be modulated by cellular antioxidants, similar to reactive oxygen. Furthermore, these data indicate that endogenous NO can trigger ERK activation and thus a downstream cellular response.
Effect of the NO x -generating Compound Sodium Nitroprusside on p38 and JNK Activities-Two other members of the MAP kinase family are JNK and p38. We examined whether these kinases, which are known to be activated by environmental stress (6,9), could also be activated by NO-related species. We treated serum-starved Jurkat cells with various concentrations of sodium nitroprusside (SNP) and observed 2-fold activation of JNK and p38 at concentrations of 3 and 300 M, respectively (Fig. 3). The concentration-response curves for both MAP kinases were biphasic. At high concentrations of SNP, we observed an inhibitory effect, resulting in activities approaching basal levels. We have previously observed an inhibitory component of NO x action in many of the systems we examined (2,3). It is likely that at high concentrations, the toxic and nonspecific effects of NO x are observed.
JNK and p38 have been found to respond to environmental stress and pronflammatory cytokines with strikingly similar dose-and time-dependent kinetics (9). The organization of the signal transduction pathways linking these two kinases is not clear. JNK and p38 may be regulated by parallel or shared pathways. However, our data indicate that either NO and related species activate these kinases through different mechanisms, or they are differentially susceptible to redox activation.
Activation Kinetics of MAP Kinases in Response to H 2 O 2 , Nitroprusside, and NO x Gas-We next studied the kinetics of ERK, p38, and JNK activation in response to SNP, NO x , and H 2 O 2 , which generates a hydroxyl radical in the presence of cellular iron (20) (Fig. 4). Serum-starved cells were treated with a single concentration for 2, 5, 10, 30, or 60 min, and the relative MAP kinase activities were analyzed. ERK1 and ERK2 showed a 3.5-fold increase in activity after a 10-min treatment with H 2 O 2 , whereas a 2-min SNP treatment produced a peak increase of 2.5-fold. We did not observe a defined peak of activation by NO-related species. Instead, ERK activity seemed to rise slowly as the time of exposure to NO x was increased. It is possible that the maximal peak occurred before our 2-min time point. Alternatively, this difference in activation kinetics by SNP and NO x may reflect differential sensitivities to NO species. SNP is known to release the nitrosonium ion (NO ϩ ), an efficient nitrosating species (18). In contrast, our NO gas-saturated solution contains authentic NO as well as other higher oxides. Thus, several upstream activators that are differentially susceptible to activation by reactive nitrogen intermediates are likely to be involved in NO x -mediated MAP kinase activation. Peak increases in p38 occurred at 10 min with H 2 O 2 and 5 min in cells treated with NO x or SNP. JNK showed peak activity after 5 min for all three compounds assayed. From these experiments, we conclude that SNP and NO-related species rapidly induce the activation of the three MAP kinases studied. The effects of the NO x donors took place more rapidly than those attributed to the oxidant H 2 O 2 .
Effect of Farnesyl Transferase Inhibition on the Activation of MAP Kinases in Response to H 2 O 2 , Nitroprusside, and NO x Gas-Previous work done in this laboratory (3, 4) demonstrated the activation of a low molecular weight G protein, most   FIG. 3. Effect of SNP on JNK and p38 activities. Jurkat T cells were pretreated for 18 h in serum-free RPMI 1640 medium. Samples were treated for 10 min in a 37°C and 5% CO 2 incubator with the indicated concentrations of SNP. JNK or p38 assays were performed as described under "Experimental Procedures" using 1 g of the GST-Jun-17 or GST-ATF-2 fusion proteins as substrates for JNK and p38, respectively. Data points represent the means from three separate experiments, each performed in duplicate. Standard deviations were less than 10% in all cases.

FIG. 4. Kinetics of activation of MAP kinases in response to H 2 O 2 , SNP, and NO x gas.
Jurkat T cells were pretreated for 18 h in serum-free RPMI 1640 medium. Samples were treated for 2, 5, 10, 30, and 60 min in a 37°C and 5% CO 2 incubator with 100 M H 2 O 2 , 300 M SNP, or 100 nM NO x . MAP kinase assays were performed as described under "Experimental Procedures" using 1 g of myelin basic protein and GST-ATF-2 or GST-Jun fusion proteins as substrates for ERK, p38, and JNK, respectively. Data points represent the means from three separate experiments, each performed in duplicate. Standard deviations were less than 5% in all cases. likely p21 ras , in the NO x -induced nuclear translocation of NFB. In this study, we have found that ERK, p38, and JNK are activated by NO x and NO-generating compounds. We next studied the role of farnesylated G proteins in the NO x -induced activation of ERK, p38, and JNK. p21 ras has been described as a molecular switch that directs incoming signals from the cell surface toward their appropriate transduction pathways (21). In addition to p21 ras , the related low molecular weight G proteins Rac1 and Cdc42 have been shown to be critical in regulating the activation of JNK and p38 signaling (22)(23)(24)(25)(26). These proteins must undergo a series of posttranslational modifications before they can localize at the cytoplasmic face of the plasma membrane. An important posttranslational modification is the addition of a farnesyl group to the carboxyl-terminal end of the newly synthesized G protein. This step is catalyzed by the enzyme farnesyl transferase, which can be blocked by ␣-hydroxyfarnesylphosphonic acid (27,28). Thus, if not localized to the plasma membrane, these G proteins cannot activate their effector.
We pretreated Jurkat cells with the farnesyl transferase inhibitor for 24 h and then treated them for 10 min with SNP, NO x , or H 2 O 2 (Fig. 5). In the absence of inhibitor, SNP, NO x , and H 2 O 2 increased MAP kinase activity by 1.75-, 1.6-, and 2-fold, respectively. The addition of the inhibitor markedly reduced the ability of these stimuli to activate the three MAP kinases (Fig. 5). These experiments clearly suggest that the MAP kinase response to NO x requires a farnesylated upstream signaling component such as p21 ras , Rac1, and/or Cdc42. DISCUSSION Previous work in this laboratory demonstrated a requirement for functional p21 ras in activation of NFB by NO x (3). We also found that reactive oxygen intermediates signaled through a similar mechanism and thus generalized our findings on signal transduction by NO x to include reactive oxygen species (4). In those studies, we found that reactive oxygen signals through p21 ras , MAP kinase, and NFB. We concluded from these studies that p21 ras is directly targeted by reactive free radicals and may be a cellular sensor for redox status (4).
ERK, JNK, and p38 play critical intermediary roles in mediating signal transduction from the membrane to the nucleus (5,7). Therefore, we investigated whether these signaling enzymes were activated by NO x and the NO x -generating compounds SNP and SNAP in human T cells. The molecular distinction between the different MAP kinase pathways lies in the substrate specificities of the MAP kinase kinases. There are six MAP kinase kinases in mammalian cells. MEK1, MEK2, and MEK5 activate ERK-related MAP kinases (29,30); MKK3 and MKK6 activate p38 (31,32); and MKK4 activates JNK and p38 (31,33,34). These MAP kinase kinases are themselves activated by receptor tyrosine kinases and other mechanisms. Our findings suggest that NO-related species and reactive oxygen intermediates activate the MAP kinases through upstream activation, most likely by activation of a low molecular weight G protein such as p21 ras , Rac1, or Cdc42 (22)(23)(24)(25)(26). The time course of activation of ERK by NO x was slow compared with the rapid effects of NO x on JNK and p38. This differential activation suggests that not all signals go through p21 ras and that other related G proteins may be involved.
JNK activation by UV light has been shown to be potentiated by the activation of p21 Ha-ras (6). This finding demonstrates that JNK signals via at least one p21 ras -dependent pathway. Because the effects of UV irradiation often mimic those of oxidative stress, it was postulated that the UV-induced activation of JNK may be initiated by oxidative stress signaling through an unknown redox sensor. We have previously suggested p21 ras as a redox sensor in the free radical-induced activation of NFB (4). The present study describes the activation of JNK by some of the same reactive species. It is possible that p21 ras or other low molecular weight G proteins (such as Rac1 and CDC42) serve similar roles in JNK activation.
NO is produced in many cell types (35). The effect of NO on cells depends on its local concentration, the redox status of its immediate environment, and the susceptibility of target sites for modification (18,36). Reactive nitrogen and oxygen overproduction is often associated with the synthesis of destructive species, such as superoxide, peroxynitrite, and nitrosonium ion. When these reactive species overwhelm the cell's antioxidant defenses, the resulting redox stress can cause lipid peroxidation, membrane damage, DNA breakage, and enzyme inactivation (20). Our study suggests that cellular glutathione levels regulate a low molecular weight G protein-dependent nuclear signaling pathway in which the MAP kinases play an important role. Thus, the ability of NO x to trigger some signaling events within the cell, such as ERK, JNK, and p38 kinase activation, as shown herein, will likely be revealed under conditions in which cellular glutathione levels are depleted. Jurkat cells were pretreated for 24 h in serum free RPMI 1640 medium with or without 10 M ␣-hydroxyfarnesylphosphonic acid. Samples were treated for 30 min in a 37°C and 5% CO 2 incubator with 300 M SNP, 100 nM NO x , or 100 M H 2 O 2 . MAP kinase assays were performed as described under "Experimental Procedures" using 1 g of myelin basic protein and GST-ATF-2 and GST-Jun fusion proteins as substrates for ERK, p38, and JNK, respectively. Data represent the means and standard deviations (bars) from three separate experiments each performed in duplicate.