|Home | About | Journals | Submit | Contact Us | Français|
S-nitrosylation (SNO) is a reversible, thiol-based protein modification that plays an important role in the myocardium by protecting critical cysteine residues from oxidation. However, little is known with regard to the percentage of a given protein that is modified by SNO (ie, SNO occupancy). Current methods allow for the relative quantification of SNO levels, but not for the determination of SNO occupancy.
To develop a method for the measurement of SNO occupancy, and apply this methodology to determine SNO occupancy in the myocardium.
We developed a differential cysteine-reactive tandem mass tag (cysTMT) labeling procedure for the measurement of SNO occupancy. To validate this cysTMT labeling method, we treated whole-heart homogenates with the S-nitrosylating agent S-nitrosoglutathione and determined maximal SNO occupancy. We also examined SNO occupancy under more physiological conditions and observed that SNO occupancy is low for most protein targets at baseline. Following ischemic preconditioning, SNO occupancy increased to an intermediate level compared to baseline and S-nitrosoglutathione treatment, and this is consistent with the ability of SNO to protect against cysteine oxidation.
This novel cysTMT labeling approach provides a method for examining SNO occupancy in the myocardium. Using this approach, we demonstrated that IPC-induced SNO occupancy levels are sufficient to protect against excessive oxidation.
S-nitrosylation (SNO) is a reversible, thiol-based modification that can modulate the activity of myocardial proteins, including those involved with Ca2+-handling and mitochondrial energetics.1–4 Additionally, we and others demonstrated that SNO can protect against cysteine oxidation.4,5 This is important in the setting of ischemia–reperfusion (IR) injury, in which the burst of reactive oxygen species generated at the onset of reperfusion can lead to protein oxidation and degradation. Myocardial ischemic preconditioning (IPC) has been shown to increase protein SNO,2,4 and thus protection from oxidation is expected to be proportionate to the percentage of protein SNO (ie, SNO occupancy). Therefore, it is important to develop a method to determine SNO occupancy.
Cysteine-reactive tandem mass tags (cysTMT) confer the advantage of multiple isobaric tags with reporter ions between 126 and 131 kDa. These labels have been used to measure SNO levels in human pulmonary arterial endothelial cells6 and to profile thiol redox sensitivity.7 There are also a number of additional methods that have been used to quantify relative amounts of SNO,2–4,8,9 but these methods did not measure total free thiols, and therefore did not measure occupancy. Herein, we utilized a novel differential cysTMT labeling strategy to provide a measure of SNO occupancy. In the same sample, free thiols were labeled with one isobaric tag (ie, cysTMTx), while SNO thiols were subsequently labeled with a second isobaric tag (ie, cysTMTy). Free and SNO thiols were then quantified via mass spectrometry. By quantifying both free and SNO thiols, the percentage of a given cysteine residue that was modified via SNO could then be calculated.
Male C57BL/6 mouse hearts were Langendorff-perfused (Figure 1A) and homogenized as previously described.1–4 Homogenates were then subjected to differential cysTMT labeling (Figure 1B). Please see the Online Supplement available at http://circres.ahajournals.com for additional materials and methods related to this study.
GSNO treatment was used to validate our cysTMT labeling procedure and determine maximal SNO occupancy. Wholeheart homogenates were incubated with a supraphysiological concentration of GSNO (1 mmol/L) in the presence of 2.5% SDS, and subjected to differential cysTMT labeling (Figure 1B). As expected, SNO occupancy greatly increased following GSNO treatment (Figure 2, Online Table I). Phosphoglycerate kinase 1 (Cys316) increased from 1.8%±0.1% at baseline to 64.6%±6.3% following GSNO treatment. Similarly, creatine kinase (Cys90) increased from 1.9%±0.2% to 61.5%±9.1%, while mitochondrial malate dehydrogenase (Cys89) increased from 4.5%±0.7% to 56.4%±7.0%. Proteins with high baseline SNO occupancy also showed increased SNO with GSNO treatment. Cytochrome c oxidase 6b1 (Cys65) increased to 63.3%±8.8% from 14.9%±2.6% at baseline, and dihydropyrimidinase-related protein 2 (Cys248) increased to 57.9%±11.3% from 13.6%±2.7%. Although there were several examples of proteins that showed high levels of SNO occupancy at baseline, the majority of proteins showed low levels (1%–10%), and this is consistent with physiological levels of nitric oxide. These results demonstrate that this differential cysTMT labeling procedure can be used for determining SNO occupancy. However, the maximal SNO occupancy observed with GSNO was only 60% to 70%, suggesting that a substantial percentage (≈30%–40%) of this labile modification was lost during labeling. This appears to be the case, as 40% of peptides from GSNO-treated samples incubated for an additional period of 2 hours (at 25°C) prior to cysTMT labeling showed a decrease in SNO occupancy of more than 10% compared to samples that were labeled immediately following GSNO treatment (Online Figure I).
We were also interested in determining SNO occupancy with IPC. Hearts were subjected to IPC (Figure 1A), homogenized, and labeled (Figure 1B). We identified 275 SNO peptides in at least 2 of 7 samples for control and 2 of 5 samples for IPC (Online Table II). Of these peptides, 44 showed a 2-fold or greater increase in SNO with IPC compared to control, with 5 peptides exhibiting an increase in SNO with P<0.05 and an additional 8 peptides with P<0.1. As shown in Figure 3, these peptides included hexokinase-1 (Cys662), short-chain acyl-CoA dehydrogenase (Cys109), mitochondrial malate dehydrogenase (Cys89), and mitochondrial aspartate aminotransferase (Cys106). Hexokinase-1 showed a 4-fold increase in SNO occupancy with IPC, increasing from 3.8%±0.4% to 14.9%±2.4%, while short-chain acyl-CoA dehydrogenase showed a 5-fold increase in SNO occupancy and increased from 4.5%±1.4% to 23.1%±8.4%. These data are consistent with our previous studies showing a ≈2- to 3-fold increase in SNO of selected proteins with IPC.2,4 Previously, we showed a 2.7-fold increase in the SNO level of mitochondrial malate dehydrogenase with IPC,2 and in the current study, mitochondrial malate dehydrogenase showed a 2-fold increase in SNO occupancy. Aspartate aminotransferase and citrate synthase were also observed in previous studies. Interestingly, we detected 12 SNO peptides with occupancies greater than 20% that were only observed with IPC. These identifications included several peptides of titin (Cys14224, Cys21689), dihydrolipoyl dehydrogenase (Cys477), and cysteine- and histidine-rich domain-containing protein 1 (Cys211). Thus, the SNO occupancy levels observed for many proteins with IPC are consistent with the ability of SNO to protect against oxidation.
An IPC-induced increase in oxidative modifications (ie, disulfides, sulfenic acids, etc.) could lead to the overestimation of SNO occupancy with the above approach due to the exclusion of oxidized cysteines from the total cysteine value used as the denominator. Since the cysTMT reagents label cysteine residues through the formation of a disulfide bond, oxidative modifications cannot simply be reduced and labeled with a third cysTMT reagent. Therefore, we measured oxidation in parallel samples using a modified approach (Online Figure II). We identified 187 oxidized peptides in at least 2 of 3 samples for both control and IPC (Online Table III). Under control conditions, oxidation occupancy was between 1% and 15% for more than 65% of the identified cysteine residues. Following IPC, 113 peptides showed increased oxidation, but the majority of these cysteine residues still showed oxidation occupancies of 15% or less. These low levels of oxidation have only a modest effect on the calculation for SNO occupancy. For example, myosin binding protein C (Cys439) showed no change in oxidation occupancy (control: 13.2%±6.5%, versus IPC: 14.1%±10%). Factoring oxidation into the equation for SNO occupancy effectively decreased the calculation of SNO occupancy in IPC from 4.1%±2.0% without cysTMTOx to 3.4%±1.4% with cysT-MTOx, but this change is minimal. There was a small population of peptides that showed larger increases in oxidation with IPC, and these changes have the potential to alter the calculation of SNO occupancy. For example, cytochrome C oxidase 6B1 (Cys65) showed a 9% increase in oxidation with IPC (control: 52.1%±13.9%, versus IPC: 57.0%± 30.6%). Factoring oxidation into the equation for SNO occupancy effectively decreased the calculation of SNO occupancy in IPC from 16.8%±8.5% without cysTMTOx to 7.3%±5.2% with cysTMTOx. These results suggest that oxidative modifications have the potential to alter the calculation of SNO occupancy for certain protein targets (Figure 3), and should be included in order to accurately determine SNO occupancy.
LC-MS/MS summary data for all peptide identifications can be found in Online Table IV and Online Table V. Raw mass spectrometry data can be accessed at Peptide Atlas (http://www.peptideatlas.org/PASS/PASS00085).
The cysTMT labeling protocol described herein provides a novel approach for determining SNO occupancy. Previous studies have developed methods for the relative quantification of SNO, but these methods do not quantify SNO occupancy.2–4,6,8,9 This method was validated using GSNO to induce maximal SNO occupancy (Figure 2) and was then applied to a physiological model of cardioprotection. Consistent with previous studies, we found a ≈2- to 3-fold increase in SNO with IPC.2,4 We identified a number of proteins that showed a significant increase in SNO occupancy with IPC, including mitochondrial malate dehydrogenase, aspartate aminotransferase, and citrate synthase (Figure 3). We also found numerous SNO peptides with occupancies ranging from 20% to 55% that were only present with IPC. These SNO occupancy levels are consistent with the ability of SNO to shield proteins from oxidation. Additionally, the protection afforded by SNO may prevent the degradation of proteins that are excessively oxidized, and therefore, protein function is preserved as we showed in our previous study.4 Indeed, IPC has been shown to prevent the degradation of many mitochondrial proteins following IR injury.10 Interestingly, several of the protein targets that showed decreased degradation following IPC also showed high levels of SNO in our model of cardioprotection, including malate dehydrogenase, α-ketoglutarate dehydrogenase, and voltage-dependent anion channel protein 2.2,4 Although SNO occupancy levels did not exceed 50% for most proteins, the preservation of 25% to 50% of protein function is likely adequate for cardioprotection. For example, consider that heterozygosity is often insufficient to cause disease. The potential also exists for underestimation of SNO occupancy using this approach; we find that even with GSNO treatment, occupancy levels do not exceed 60% to 70%, suggesting that a minimum of 30% to 40% of SNO is lost during labeling. Furthermore, nitric oxide signaling is known to be compartmentalized. This study was performed using whole-heart homogenates, but certain cellular compartments may have enhanced nitric oxide or SNO signaling. Thus, low occupancy does not preclude an important role for SNO in cardioprotection.
The loss of SNO during labeling is a potential confounding factor with this approach, thus leading to the possible underestimation of SNO occupancy. Concurrent cysteine modifications, such as oxidation, are an additional confounding factor. Structural disulfides and other oxidative modifications at baseline are unlikely to affect SNO occupancy, but the stimulation of oxidative modifications could lead to the oxidation of free thiols, and this may alter the calculation for SNO occupancy (Figure 3). Therefore, the inclusion of oxidized thiols is necessary for an accurate determination of SNO occupancy. Protein degradation, as typically occurs with IR,10 may be another confounding factor that could lead to the overestimation of SNO occupancy.
In conclusion, our cysTMT labeling procedure provides a novel approach for the measurement of SNO occupancy in the myocardium. This method was validated using GSNO-treated whole-heart homogenates, and was also used to demonstrate that IPC-induced SNO is sufficient to shield cysteine residues against excessive oxidation.
Herein we examined whether IPC-induced SNO was sufficient to shield cysteine residues from oxidation. To test this hypothesis, we developed a novel differential cysTMT-labeling procedure for determining the percentage of a given cysteine residue that is modified by SNO (ie, SNO occupancy). Using this methodology, we determined that with IPC, SNO occupancy increased to levels that are consistent with the ability of SNO to protect against oxidation. Previous studies have developed methods for the relative quantification of SNO, but this is the first study to examine SNO occupancy. This method can also be adapted to examine other redox-based modifications.
Sources of Funding
This work was supported by National Institutes of Health (NIH) grants 1F32HL096142 (to M.K.) and 5R01HL039752 (to C.S.), and the NHLBI/NIH Intramural Program (to A.A., J.S., M.G., & E.M.).