S-Nitrosylation inhibits the kinase activity of tomato phosphoinositide-dependent kinase 1 (PDK1)

It is well known that the reactive oxygen species NO can trigger cell death in plants and other organisms, but the underlying molecular mechanisms are not well understood. Here we provide evidence that NO may trigger cell death in tomato (Solanum lycopersicum) by inhibiting the activity of phosphoinositide-dependent kinase 1 (SlPDK1), a conserved negative regulator of cell death in yeasts, mammals, and plants, via S-nitrosylation. Biotin-switch assays indicated that SlPDK1 is a target of S-nitrosylation. Moreover, the kinase activity of SlPDK1 was inhibited by S-nitrosoglutathione in a concentration-dependent manner, indicating that SlPDK1 activity is abrogated by S-nitrosylation. The S-nitrosoglutathione–induced inhibition was reversible in the presence of a reducing agent but additively enhanced by hydrogen peroxide (H2O2). Our LC-MS/MS analyses further indicated that SlPDK1 is primarily S-nitrosylated on a cysteine residue at position 128 (Cys128), and substitution of Cys128 with serine completely abolished SlPDK1 kinase activity, suggesting that S-nitrosylation of Cys128 is responsible for SlPDK1 inhibition. In summary, our results establish a potential link between NO-triggered cell death and inhibition of the kinase activity of tomato PDK1.

NO potentiates the induction of hypersensitive cell death in soybean cells by reactive oxygen species (ROS) 2 (1). However, the molecular mechanism of the NO-induced cell death remains an enigma. One potential mechanism is that the activity of proteins that control cell death may be altered by a posttranslational modification, S-nitrosylation. Functional groups containing a nitroso group attached to the sulfur atom of a thiol are called S-nitrosothiols (SNOs) (2). The addition of a nitroso group to a sulfur atom of a Cys residue of a protein is known as S-nitrosylation (3). S-nitrosylation is an enzyme-independent, posttranslational, and labile modification that can function as an on/off switch of protein activity (3)(4)(5). In cells, the extent of S-nitrosylation of proteins is largely determined by the level of S-nitrosoglutathione (GSNO), which serves as a stable and mobile NO pool and effectively transduces NO signaling (3,4). S-nitrosoglutathione reductase (GSNOR) catalyzes GSNO reduction and thus regulates the cellular levels of both GSNO and protein S-nitrosylation (6). Loss of GSNOR leads to increased cellular levels of S-nitrosylated proteins (6,7). Thousands of diverse classes of proteins, both in plants and in mammals, have been identified as targets of S-nitrosylation (8 -12). In plants, proteins with diverse functions are S-nitrosylated at specific Cys residue(s), and their functions are either inhibited or enhanced by this modification (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28).
3-Phosphoinositide-dependent protein kinase 1 (PDK1) and its downstream target, protein kinase B (PKB, also known as Akt), are central regulators of mammalian apoptosis (29 -31). PKB is a member of the AGC family of protein kinases, which is activated by second messengers such as phospholipids and Ca 2ϩ through PDK1. Mammalian PDK1 phosphorylates PKB to promote its function in suppressing programmed cell death (30 -33). PKB negatively regulates apoptosis by phosphorylation and inactivation of pro-apoptotic factors such as BAD (Bcl-2-associated death) and activation of anti-apoptotic factors such as cAMP-response element-binding protein and IB kinase (IKK) (30 -32, 34). Deficiency of the PDK1 gene(s) in Drosophila (35), mice (36), yeast (37,38), and tomato (39), respectively, results in lethality or severe apoptosis. PKB knock-out mice display spontaneous apoptosis in several different tissues (40). In tomato, the PKB/Akt homolog Adi3 (AvrPto-dependent Pto-interacting protein 3) physically interacts with and is phosphorylated by SlPDK1 (39). Silencing both SlPDK1 and Adi3 or treatment with a PDK1 inhibitor results in MAPKK kinase ␣dependent cell death, indicating that Adi3 functions analogously to the mammalian PKB/Akt by negatively regulating cell death via PDK1 phosphorylation (39).
Yasukawa et al. (41) showed that NO donors induced S-nitrosylation and inactivation of PKB/Akt kinase activity in vitro and in vivo and that the mutant Akt1/PKB (224 Cys3 Ser ) was resistant to S-nitrosylation by NO and its kinase inactivation (41). Although the NO and PDK1-PKB/Akt pathways are both key regulators of cell death, the link between these two path-ways has not been firmly established in plants. Here we show that the kinase activity of tomato SlPDK1 was inhibited by GSNO in a concentration-dependent manner and that this inhibition was additively enhanced by H 2 O 2 . Interestingly, we determined that SlPDK1 was S-nitrosylated at Cys 128 . This Cys residue is critical for the function of SlPDK1 because a mutation of Cys 128 to Ser completely abolished its kinase activity. Our results suggest that inhibition of SlPDK1 activity by S-nitrosylation at Cys 128 might be a key molecular event underpinning NO-triggered cell death.

Spontaneous cell death occurs in GSNOR1-silenced tomato plants
NO triggers cell death in soybean suspension cells (1). While investigating the role of GSNOR1 in disease resistance, we observed that spontaneous cell death frequently occurred in the leaves of GSNOR1-silenced tomato plants (Fig. 1), suggesting that GSNOR1 negatively regulates cell death. Increased cell death was also observed in GSNOR1 knock-out mice and Arabidopsis (19,42) that had significantly higher levels of protein S-nitrosylation (7,42). Our experiments, coupled with the previous studies, prompted us to postulate that the cell death observed in GSNOR1 knock-out (19) or knock-down plants ( Fig. 1) was caused by S-nitrosylation of key negative regulator(s) of cell death.

Tomato SlPDK1 but not Adi3 is the target of S-nitrosylation
We took a candidate protein approach to search for negative regulators of cell death with functions potentially regulated by S-nitrosylation. We focused on proteins that were shown previously to be key negative regulators of cell death and S-nitrosylated. PKB rose to the top of our list because it is a conserved cell death regulator in yeasts (43), mammals (34), and plants (39), and it is also a target of S-nitrosylation in mice (41,44). Most importantly, S-nitrosylation of PKB inhibits its kinase activity and, thus, its function (41). To test whether a plant PKB is a target of S-nitrosylation, a biotin-switch assay (45) was performed on tomato Adi3 fused to maltose-binding protein (MBP-Adi3). Adi3 is a homolog of PKB/Akt, which is a known negative regulator of tomato cell death (39). Adi3 was not S-nitrosylated because no band was detectable for MBP-Adi3 in the biotin-switch assay ( Fig. 2A). Unexpectedly, we found that tomato SlPDK1, a protein kinase that acts upstream of Adi3, was S-nitrosylated (Fig. 2B). To confirm this result, the biotinylated MBP-SlPDK1 fusion protein was purified using neutravidin-agarose, and then a Western blot analysis was performed using an antibody against MBP. A band the size of MBP-SlPDK1 was detected (Fig. 2C), confirming that SlPDK1 is a target of S-nitrosylation. To test whether the SlPDK1 is S-nitrosylated in planta, the FLAG-SlPDK1 fusion protein was transiently expressed in NbGSNOR1-silenced Nicotiana benthamiana leaves via Agrobacterium infiltration. The purpose of silencing NbGSNOR1 was to create a high cellular SNO level to facilitate the detection of S-nitrosylation. NbGSNOR1 was silenced using TRV-induced gene silencing (VIGS) (46). 2 days after infiltration, total protein was extracted from the infiltrated leaf area and biotinylated using the biotin-switch-based approach. The biotinylated proteins were then purified using neutravidin-agarose beads followed by Western blotting using an antibody against FLAG. The transiently expressed FLAG-SlPDK1 was indeed S-nitrosylated in planta, further confirming that SlPDK1 is a target of S-nitrosylation (Fig. 2D). It has been reported previously that VIGS of SlPDK1 in tomato or application of a PDK1 inhibitor results in the early death of tomato seedlings and suspension cells, respectively (39). Therefore, SlPDK1 satisfies the two criteria we set for the candidate protein: that it negatively regulates cell death and is S-nitrosylated.  MBP-SlPDK1 (B) were purified from bacterial extracts and treated with 1 mM GSNO. S-nitrosylation was detected by biotin-switch assay using a biotin antibody. Asc, ascorbate. C, biotinylated MBP-SlPDK1 protein was purified with neutravidin-agarose, separated by SDS-PAGE, and detected by immunoblot assay with an anti-MBP antibody. Coomassie Blue was used to stain MBP-Adi3 protein for loading control in A (input), and the nonspecific bands detected in immunoblots were used as controls in B and C. D, the FLAG-SlPDK1 fusion protein transiently expressed in the leaves of NbGSNOR1-silenced N. benthamiana plant is S-nitrosylated in vivo. Equal loading was verified with FLAG antibody against the input protein extracts. These experiments were repeated three times with similar results.

S-Nitrosylation of PDK1 in tomato GSNO inhibits tomato SlPDK1 kinase activity in vitro
Because SlPDK1 is S-nitrosylated, we reasoned that its kinase activity could be altered in the presence of GSNO. To test this possibility, we performed kinase assays in the presence and absence of GSNO using MBP-SlPDK1 purified from Escherichia coli and a kinase-deficient version of Adi3 (Adi3 K337Q ) as substrate. Adi3 K337Q loses most autophosphorylation activity, but it is trans-phosphorylated by PDK1 (39). Trans-phosphorylation of Adi3 K337Q by SlPDK1 was inhibited by GSNO in a concentration-dependent manner (Fig. 3, left), suggesting that S-nitrosylation of SlPDK1 negatively regulates its kinase activity.

Inhibition of SlPDK1 activity is reversible by reducing agent but enhanced additively by H 2 O 2
Because of the labile nature of S-nitrosylation, a reducing agent, such as DTT, can reverse this modification. When DTT was added to the kinase assay buffer, it clearly reversed the inhibition of SlPDK1 kinase activity by GSNO (Fig. 4A). This result suggests that GSNO affects the kinase activity of SlPDK1 through a reversible thiol modification (47).
NO and H 2 O 2 have been shown to synergistically induce cell death in soybean suspension cells (1), and they also act synergistically in inhibiting PKB activity in mammalian cells (41 and GSNO had a stronger inhibitory effect on SlPDK1 kinase activity than either treatment alone (Fig. 3). These observations correlate with the facts that H 2 O 2 enhances the level of NO donor-induced S-nitrosylation of PKB (41) and that NO promotes ROS-induced cell death (1,48).

SlPDK1 kinase activity is additively inhibited by GSNO and H 2 O 2 when transiently expressed in the leaves of N. benthamiana
To explore whether GSNO and H 2 O 2 can additively inhibit SlPDK1 activity in plant, we tested the effect of GSNO and H 2 O 2 on the FLAG-SlPDK1 transiently expressed in N. benthamiana leaves via Agrobacterium infiltration. We purified MBP-Adi3 K337Q from E. coli and used it as a phosphorylation substrate for transiently expressed FLAG-SlPDK1 present in extracts from N. benthamiana leaves. The phosphoryla-tion reaction (purified MBP-Adi3 K337Q ϩ leaf protein extracts) was performed either in the presence of 2 mM GSNO or 1 mM H 2 O 2 individually or both. Consistent with the in vitro assays (Fig. 3), GSNO or H 2 O 2 individually inhibited the activity of MBP-SlPDK1, and when they were added in combination, MBP-SlPDK1 activity was further inhibited (Fig. 4B). These results support the conclusion that SlPDK1 activity is additively inhibited by both GSNO and H 2 O 2 when it was expressed in planta.

Tomato SlPDK1 is S-nitrosylated at Cys 128 and Cys 466
LC-MS/MS approaches based on the biotin-switch assay have been widely used for the identification of S-nitrosylated Cys sites within proteins (49 -51). In our experiment, a differential alkylation strategy modified from a method published previously (52) was applied to the purified wild-type MBP-SlPDK1 fusion protein. S-nitrosylation of SlPDK1 was chemically induced by reaction with GSNO. SlPDK1 was subsequently alkylated with N-ethylmaleimide (NEM) to block free thiols, and after alkylation, the SNO groups were reduced and labeled with iodoacetamide (IAM). In this approach, the Cys residues labeled by IAM are identified as target sites of S-nitrosylation. Following alkylation, the protein sample was enzymatically digested and analyzed by LC-MS/MS. When trypsin was used for protein digestion, we identified peptides containing two of the four Cys residues in SlPDK1, Cys 214 and Cys 466 . Of the two detected Cys residues, only Cys 466 (ICTPKKVMS-FEDAK), but not Cys 214 , was determined to be S-nitrosylated (Fig. 5A).  The kinase activity of the cell extracts from the infiltrated area was determined using purified kinase-dead MBP-Adi3 K337Q as a substrate. Equal loading for MBP-Adi3 K337Q was shown by Coomassie Blue staining, and equal input for FLAG-SlPDK1 was shown by Western blot analysis using FLAG antibody. The experiments were repeated three with similar results.

S-Nitrosylation of PDK1 in tomato
To identify peptides containing Cys 128 or Cys 244 , we used a different protease, Asp-N, for protein digestion based on an in silico digestion prediction (53)(54). With this approach, a peptide that contained Cys 128 with an IAM modification was identified, indicating that Cys 128 (ALESCEGGELF) is a redox-sensitive site susceptible to SNO modification (Fig. 5B). In sum, these results indicated that the Cys 128 and Cys 466 residues of SlPDK1 are S-nitrosylated.

Cys 128 is critical for the kinase activity of SlPDK1
To identify the Cys residues that are important for the kinase activity of SlPDK1, phosphorylation assays were performed for the WT as well as the four Cys3 Ser mutant MBP fusion pro-teins. Mutations at Cys 214 and Cys 466 had no effect on the kinase activity of MBP-SlPDK1 in the absence of GSNO (Fig.  6). However, MBP-SlPDK1 kinase activity was completely abolished by the Cys 128 mutation, whereas some residual activity was observed for the Cys 244 mutant (Fig. 6). These results indicate that Cys 128 and Cys 244 , but not Cys 214 and Cys 466 , are critical for the kinase activity of SlPDK1. As expected, the kinase activities of the WT MBP-SlPDK1 and the 214 Cys3 Ser and 466 Cys3 Ser mutants were severely inhibited by GSNO (Fig.  6), indicating that these two residues are not responsible for the inhibition. The residual kinase activity of the 244 Cys3 Ser mutant was sensitive to GSNO (Fig. 6), suggesting that an additional Cys residue could be responsible for the kinase inhibi-

S-Nitrosylation of PDK1 in tomato
tion, and this is most likely Cys 128 . Therefore, it is highly possible that S-nitrosylation at Cys 128 is responsible for the inhibition of SlPDK1 kinase activity. Thus, our results established a potential link between NO-triggered cell death and the inhibition of SlPDK1 kinase activity by S-nitrosylation at Cys 128 .

Loss of GSNOR1 results in spontaneous cell death in different plant species
It has been reported that loss of GSNOR1 results in a substantial increase in the level of protein S-nitrosylation and increased cell death in response to pathogen infections both in mammals and plants (7,19,42). Unexpectedly, we observed a spontaneous cell death phenotype in SlGSNOR1-silenced plants even in the absence of pathogen infection (Fig. 1), raising the possibility that S-nitrosylation acts as a regulator of cell death. Identification of such a negative regulator(s) with a function that is inhibited by S-nitrosylation will enable the molecular mechanism by which NO induces cell death to be unraveled.

SlPDK1 activity is inhibited by both NO and H 2 O 2 in an S-nitrosylation-dependent manner
Because H 2 O 2 and NO operate together in the cell death signaling cascades (for review, see Ref. 55), identification of factors that interact with both H 2 O 2 and NO could allow the discovery of the point of overlap between the two pathways (4, and 55). In an attempt to identify such factors, we focused on Adi3, a homolog of Akt/PBK in mammals, because it is not only a negative regulator both in tomato and mouse (39,41), but it is also a target of S-nitrosylation in the mouse (39). In addition, S-nitrosylation of mouse PKB/Akt at Cys 224 inactivates its function in vitro and in vivo, and H 2 O 2 has a synergistic effect on the inhibition of Adi3 kinase activity (41). In contrast, we found that, unlike the mouse PKB/Akt, tomato Adi3 was not modified by S-nitrosylation ( Fig. 2A). This result was not totally unexpected because Cys 224 is not conserved in tomato Adi3. Surprisingly, we found that SlPDK1, which functions upstream of Adi3, was S-nitrosylated at Cys 128 and Cys 466 (Figs. 2, B-D,  and 5). These two Cys residues are not conserved in the mammalian PDK1 homolog, which explains why mammalian PDK1 has not been identified as a target of S-nitrosylation. Similar to what has been observed for PKB/Akt in mice (41), GSNO inhibited the SlPDK1 kinase activity in a concentration-dependent manner (Figs. 3 and 4B), and this inhibition was enhanced in the presence of H 2 O 2 (Figs. 3 and 4B) and reversed by addition of the reducing agent DTT (Fig. 4A). The strong additive effect of NO and H 2 O 2 on plant cell death (1) as well as on the inhibition of the SlPDK1 kinase activity (Figs. 3 and 4B) implied that the cell death observed in the SlPDK1-silenced plants might be a consequence of the inhibition of the SlPDK1 kinase activity and that the PDK1 in plants might be one of the long-sought factors targeted by both NO and H 2 O 2 to trigger cell death (1,48). This statement is supported by the fact that silencing SlPDK1 leads to a lethal phenotype (39).
As Cys 128 but not Cys 466 was critical for SlPDK1 kinase activity (Fig. 6), and Cys 128 was the major site of S-nitrosylation (Figs. 5B and 6), it is most likely that inhibition of the kinase activity of SlPDK1 by S-nitrosylation at Cys 128 is responsible for the cell death observed in the SlGSNOR1-silenced tomato plants (Fig. 1). One might argue that abolishing the SlPDK1 activity by the 128 Cys3 Ser substitution could be the result of disrupting the overall structure of SlPDK1. However, it is also possible that S-nitrosylation at Cys 128 could inhibit the kinase activity of SlPDK1 through altering its structure. Together, these results suggest that NO targets different components in the conserved PI3K-PDK1-PKB/Akt signaling pathway to regulate cell death in mice and tomato, demonstrating that distinct differences exist between plants and mammals in the control of NO-induced cell death.
Hu et al. (12) recently analyzed the flanking sequences of 1195 S-nitrosylated Cys residues identified in Arabidopsis proteins and found that EXC was a predominant putative consensus sequence. Interestingly, the two amino acid residues preceding Cys 128 are Glu and Ser, which perfectly matches the consensus EXC sequence and provides additional support for the theory that Cys 128 is the major site of S-nitrosylation in SlPDK1. The other three Cys residues within SlPDK1 do not match this consensus site. Additionally, our result validates the consensus sequence identified in Arabidopsis by Hu et al. (12) and suggests that this consensus sequence is conserved in different plant species.

NO and H 2 O 2 target multiple proteins in the cell death pathway
Given the nonspecific nature of S-nitrosylation, it is possible that multiple proteins that play both positive and negative roles in cell death are simultaneously modified by S-nitrosylation. The final fate of the cells is determined by many factors, including local cellular NO concentration, the subcellular localization of these cell death regulators, and the sensitivity of different cell death regulators to S-nitrosylation (35,56,57). The co-existence of multiple positive and negative regulators of cell death that are regulated by S-nitrosylation enables cells to fine-tune cell death signaling more precisely and with more flexibility.
In mammals, S-nitrosylation of GAPDH triggers binding to Siah1 and mediates the nuclear translocation of GAPDH (58). In the nucleus, GAPDH stabilizes Siah1, enhancing its E3 ubiquitin ligase activity, which ultimately leads to cell death (58). In plants, GAPDH is a target of S-nitrosylation (8,22). These

S-Nitrosylation of PDK1 in tomato
observations raise the possibility that a similar pathway is also operating in plants and that GAPDH could be another key factor in NO-induced cell death. Interestingly, Han et al. (59) recently showed that the cytosolic GAPDH in N. benthamiana plants (GAPCs) negatively regulates autophagy through interacting with autophagy-related protein 3 (ATG3). Silencing of GAPCs enhances N gene-mediated cell death and plant resistance against both incompatible and compatible pathogens. It appears that the GAPDH in plants functions differently from its counterpart in mammals.
The RBOHD subunit of the Arabidopsis NAPDH oxidase complex is S-nitrosylated at Cys 890 , and this modification abolishes its ability to synthesize ROS (19). As a result, when the Cys 890 mutant version of AtRBOHD was transformed into the Arabidopsis rbohd mutant, the transgenic plants underwent increased cell death compared with those transformed with a WT AtRBOHD after being challenged with an avirulent bacterial strain (19). On the other hand, instead of inhibiting the function of a protein, S-nitrosylation of the Arabidopsis cytosolic APX1 at Cys 32 enhances its H 2 O 2 scavenging activity, resulting in increased resistance to oxidative stress (25). The 32 Cys3 Ser mutation reduces the activity of APX1 and abolishes its GSNO-enhanced activity (25). As a result, induction of flg22-induced expression of resistance-related marker genes in the apx1 mutant was rescued by the WT APX1-FLAG but not APX1 C32S -FLAG transgene (25). In mice, the 224 Cys3 Ser mutation of Akt1/PKB does not inhibit its kinase activity, and thus, the mutant Akt1/PKB was resistant to S-nitrosylation and kinase inactivation (41). However, in our case, the 128 Cys3 Ser mutation completely abolished the kinase activity of SlPDK1 (Fig. 6), and, therefore, it is not possible to create a mutant version of SlPDK1 in which the kinase activity is still maintained but no longer inhibited by S-nitrosylation. As a result, this prevents us from generating a transgenic line that could be resistant to NO-induced cell death.

Plant materials
Tomato (S. lycopersicum) and N. benthamiana were used in our experiments. The plants were grown in a growth room at 20°C during the dark and 22°C during the light, with a photoperiod of 16 h.
pTRV1 and pTRV2-SlGSNOR1 or pTRV1 and pTRV2-NbGSNOR1 were co-inoculated into 3-week-old tomato and N. benthamiana seedlings, respectively, by Agrobacterium-mediated infiltration as described previously (46,60), and the phenotypes were assayed 2 to 3 weeks later. Each silencing experiment was carried out a minimum of three times with at least six plants.

In vitro site-directed mutagenesis and protein expression
Site-directed mutagenesis was carried out using the Gen-eTailor site-directed mutagenesis system (Invitrogen). The primers used in the mutagenesis were as follows: Constructs for protein expression were transformed into E. coli strain BL21 Star (DE3) (Invitrogen) and grown overnight in 2 ml of LB at 37°C. A 200-l aliquot of the culture was added to 10 ml of LB, grown at 37°C to A 600 ϭ 0.5, and protein expression was induced with 100 M isopropyl 1-thio-␤-D-galactopyranoside for 3 h at 28°C. Cells were harvested by centrifugation and lysed by sonication in 1 ml of extraction buffer (50 mM Tris (pH 8.0), 50 mM NaCl, 5 mM EDTA, 0.1% Triton X-100, and general protease inhibitors (Sigma)), and debris was pelleted by centrifugation. Proteins were purified from the supernatants by adding amylose resin (New England Biolabs and FLAG M2-agarose (Sigma), incubating at 4°C for 30 min with rotation, pelleting the resin by centrifugation, and washing three times with extraction buffer.

Kinase assays
Phosphorylation assays were performed as described previously (39). Briefly, the MBP translational fusion proteins were immobilized on 50 l of amylose beads, washed three times with kinase buffer without DTT and ATP, and then incubated with or without GSNO of different concentrations at 4°C for 1 h before the in vitro kinase assay was performed. After the incubation, the beads were washed once with kinase buffer without ATP and DTT and then incubated with 5 Ci of [␥-32 P]ATP in 50 l of kinase buffer. The Adi3/Pdk1 kinase buffer contained 10 mM Tris (pH 7.5), 150 mM NaCl, 10 mM MgCl 2 , and 20 M ATP. Reactions were carried out at room temperature for 30 min and analyzed by 10% SDS-PAGE with phosphorimaging (Molecular Dynamics) visualization. For cross-phosphorylation assays, MBP translational fusion proteins were purified separately on 50 l of amylose resin, and proteins left on the resins were mixed and used in a kinase assay,

S-Nitrosylation of PDK1 in tomato
with the buffer of the enzyme carrying out the phosphorylation. Kinase assays were repeated a minimum of three times.
For the kinase assay in planta, SlPDK1 was cloned into pEarleyGate 202 to generate 35S Pro ::FLAG-SlPDK1 and was subsequently transformed into Agrobacterium strain GV2260. The Agrobacterium solution carrying 35S Pro ::FLAG-SlPDK1 was infiltrated into N. benthamiana leaves, and the protein extracts in the infiltrated areas were used for in planta kinase assay with purified MBP-Adi3 K337Q from E. coli as substrate.

Biotin labeling of S-nitrosylated proteins and purification of biotinylated proteins
The Cys residue of a protein can be biotinylated in the presence of biotin (covalently attaching biotin to a protein). The biotin-switch assay has been developed to identify S-nitrosylated proteins (45) and consists of three principle steps: blocking of free cysteine thiols (non-S-nitrosylated thiols) by Smethylthiolation with methyl methanethiosulfonate (MMTS), removing the NO group from the Cys residues with ascorbate, and labeling of thiols by S-biotinylation with biotin-HPDP, a reactive mixed disulfide of biotin. The S-nitrosylated protein can be purified using streptavidin or avidin beads, which bind biotin with an extremely high affinity and specificity. The degree of S-nitrosylation reflected by biotinylation can be determined by Western blotting with a biotin antibody, a protein-specific antibody, or an antibody for a tag.
Arabidopsis cells were frozen in liquid nitrogen and ground to a fine powder. Proteins were extracted by adding 12 ml of HEN buffer (25 mM HEPES (pH 7.7), 1 mM EDTA, and 0.1 mM neocuproine) to 10 g of cell powder. Cell debris was removed by centrifugation (20,000 ϫ g, 10 min at 4°C), and protein concentration was determined according to Bradford with bovine serum albumin as standard.
The in vitro S-nitrosylation and subsequent biotinylation of S-nitrosylated proteins were performed as described in Ref. 45 with minor modifications (8). After treating the supernatant with GSNO for 20 min at room temperature, the proteins were incubated with 20 mM MMTS and 2.5% SDS at 50°C for 20 min with frequent vortexing for blocking non-nitrosylated free Cys residues. Residual MMTS was removed by precipitation with 2 volumes of Ϫ20°C acetone, and the proteins were resuspended in 0.1 ml of HENS buffer (HEN buffer containing 1% SDS)/1 mg of protein. Biotinylation was achieved by adding 2 mM biotin-HPDP and 1 mM ascorbate and incubation at room temperature for 1 h.
After removing biotin-HPDP, the precipitated proteins were resuspended in 0.1 ml of HENS buffer/1 mg of protein and 2 volumes of neutralization buffer (20 mM HEPES (pH 7.7), 100 mM NaCl, 1 mM EDTA, and 0.5% Triton X-100). A total of 15 l of neutravidin-agarose/1 mg of protein was added and incubated for 1 h at room temperature. The matrix was washed extensively with 20 volumes of washing buffer (600 mM NaCl in neutralization buffer), and bound proteins were eluted with 100 mM ␤-mercaptoethanol in neutralization buffer (8).

In vitro S-nitrosylation of PDK1 and differential alkylation of Cys residues
The purified MBP-PDK1 was reconstituted in 250 mM HEPES (pH 7.7), 1 mM EDTA, 0.1 mM neocuproine, and 0.1% SDS and treated with 10 mM DTT for 30 min at 37°C to release all free thiols on the protein. DTT was subsequently removed by using a 0.5-m Amicon filter with 10-kDa molecular weight cutoff (EMD Millipore, Billerica, MA). To induce S-nitrosylation, the protein solution was divided into two aliquots and treated with and without GSNO, respectively, in 1% DMSO at 37°C for 30 min in the dark. After this step, unmodified cysteine residues were blocked with 50 mM NEM at 55°C for 30 min in the dark in 2% SDS, and the excess NEM was removed by cold acetone precipitation. Protein pellets were subsequently dissolved in 50 mM HEPES (pH 7.7) with 0.5% SDS. S-nitrosylated cysteine residues were selectively reduced by adding 5 mM sodium ascorbate and 5 M CuCl. 40 mM IAM was added to label the nascent free thiols in the dark at 37°C for 1 h. The protein samples were then purified and concentrated in 50 mM Tris-HCl (pH 7.5) by using an Amicon filter. Protein digestion was performed at 37°C overnight by adding trypsin or Asp-N with an enzyme-to-protein ratio 1:200. After digestion, 0.1% TFA was added to stop the reaction. Peptide samples were cleaned up with Omix C18 Ziptip (Agilent) and put in 0.1% formic acid prior to LC-MS/MS.

LC-MS/MS and data analysis
LC-MS/MS analysis was performed using the nano-Aquity UPLC system (Waters Corp., Milford, MA) with a homemade 75-m inner diameter ϫ 70 cm reverse-phase capillary column packed with 3 m C18 resin (Phenomenex, Torrance, CA). The LC system was operated at a constant flow rate of 300 nl/min over a 100-min gradient. Mobile phases A and B were 0.1% (v/v) formic acid in water and 0.1% (v/v) formic acid in ACN, respectively. The gradient started with 100% of mobile phase A to 60% of mobile phase B. MS analysis was performed on an LTQ-Orbitrap Velos mass spectrometer (Thermo Scientific, San Jose, CA) coupled with an electrospray ionization interface using a homemade 150-m outer diameter ϫ 20-m inner diameter chemically etched electrospray emitter. The heated capillary temperature and spray voltage were 350°C and 2.2 kV, respectively. Full MS spectra were recorded at a resolution of 60 K over a range of m/z 300 -2000 with an automated gain control value of 1 ϫ 10 6 . The most abundant 10 parent ions were selected for MS/MS using collision-induced dissociation with a normalized collision energy setting of 35. Precursor ion activation was performed with an isolation width of 2.5 Da, a minimal intensity of 1000 counts, and an activation time of 0.1 s. A dynamic exclusion time of 60 s was used.
LC-MS/MS raw data were converted into data files using Bioworks Cluster 3.2 (Thermo Fisher Scientific, Cambridge, MA), and an MSGF plus algorithm (v9979, released in March 2014) was used to search MS/MS spectra against a sequence database containing MBP and PDK1. The key search parameters used were 20-ppm tolerance for precursor ion masses, 0.5-Da tolerance for fragment ions, partial tryptic search with up to two missed cleavages, dynamic oxidation of methionine (15.9949 Da), dynamic NEM modification of Cys (125.0477 Da), and dynamic IAM modification of Cys (ϩ57.0215 Da). Peptides were identified from database search results applying the following criteria: MSGF SpecE value Ͻ10 Ϫ9 , Q value Ͻ0.01 (false discovery rate Ͻ1%), and mass measurement error