Strategies and Tools to Explore Protein S-Nitrosylation

Abstract

Background

A biochemical pathway by which nitric oxide accomplishes functional diversity is the specific modification of protein cysteine residues to form S-nitrosocysteine. This post-translational modification, S-nitrosylation, impacts protein function, interactions and location. However, comprehensive studies exploring protein signaling pathways or interrelated protein clusters that are regulated by S-nitrosylation have not been performed on a global scale.

Scope of Review

To provide insights to these important biological questions, sensitive, validated and quantitative proteomic approaches are required. This review summarizes current approaches for the global identification of S-nitrosylated proteins.

Major Conclusions

The application of novel methods for identifying S-nitrosylated proteins, especially when combined with mass-spectrometry based proteomics to provide site-specific identification of the modified cysteine residues, promises to deliver critical clues for the regulatory role of this dynamic posttranslational modification in cellular processes.

General Significance

Though several studies have established S-nitrosylation as a regulator of protein function in individual proteins, the biological chemistry and the structural elements that govern the specificity of this modification in vivo are vastly unknown. Additionally, a gap in knowledge exists concerning the potential global regulatory role(s) this modification may play in cellular physiology. By further studying S-nitrosylation at a global scale, a greater appreciation of nitric oxide and protein S-nitrosylation in cellular function can be achieved.

Introduction

Since its discovery nitric oxide has become increasingly evident as a major regulator of physiological function. The effects of nitric oxide on physiology are exerted primarily through two molecular mechanisms, comprised of cyclic GMP (cGMP)-dependent signaling cascades and post-translational modification (PTM) of proteins. Initially, nitrosylation of the heme iron in soluble guanylate cyclase was found to activate the enzyme to generate cGMP, thereby regulating the function of a number of cGMP-dependent signaling pathways. However, in recent years the importance of nitric oxide-derived post-translational modifications of proteins has gained recognition as mediators of protein function.

One of these modifications, S-nitrosylation, is defined as the covalent addition of a nitric oxide equivalent to the thiol side chain of cysteine [1]. This modification has been shown to alter protein activity, protein-protein interactions, and sub-cellular localization under physiological and pathological conditions [2, 3, 4, 5]. Additionally, evidence indicates that S-nitrosylation is reversible and is regulated in a temporal and spatial sense, reminiscent of other post-translational modifications such as phosphorylation [6, 7, 8]. Despite these indications of the emerging significance of S-nitrosylation, little is known regarding the proximal mechanisms of in vivo formation as well as how selectivity, in terms of directing modification to specific cysteine residues, is achieved. To improve our understanding of the formation and selectivity of this post-translational modification in vivo, global interrogation of S-nitrosoproteomes can be exceptionally valuable. Below we review the methodologies for the global identification of S-nitrosylated proteins (Table 1), and discuss potential utilities of proteomic-derived data.

Table 1.

Techniques for Identifying S-Nitrosylated Proteins

Method	Principles	Comments
1)Biotin Switch Technique (BST): The original method and several variations thereafter: -Detergent-free BST -SNOSID -d -Switch -His-Tag Switch -Fluorescent Switch/2D-DIGE -SNO-RAC	Blockage of free thiols with thiosulfonate MMTS Ascorbate reduction of Snitrosocysteine. Reaction of ascorbate-generated free cysteine with biotin-HPDP (or other thiol-reactive compounds). Enrichment of labeled proteins or peptides, and/or 2D-DIGE to assess relative changes in S-nitrosylation levels. Western blot or LC-MS/MS to identify protein targets. Peptide enrichment allows site-specific identification of modified cysteine residues.	-Collectively 135 endogenously-modified protein targets have identified -82 endogenous sites on 63 proteins were also identified -Proteomics applicability -Reliance on selective reduction of SNO moiety prior to labeling, which has been discussed previously [20–26]
2)Direct Detection by ESI-QTOF Mass Spectrometry	Direct detection of SNO moiety on modified proteins	-Site-specific identification -Limited to isolated proteins or peptides
3)Phosphine-based derivatization	Specific reaction of SNO moiety on proteins with phosphine-based compounds without blocking free thiols	-Direct reaction of phosphine compounds with S –nitrosocysteine provides selective one-step labeling of modified residues -Potential applicability for mass-spectrometry-based proteomics needs to be explored
4)Gold Nanoparticles (AuNPs)	Blockage of free thiols with IAM. Enrichment of SNO-proteins via reaction of S-nitrosocysteine with AuNPs. Trypsin digestion and elution of bound peptides. LC-MS/MS to identify sites of modification	-Site-specific identification -AuNP reactivity toward S-nitrosocysteine and S-glutathionylated cysteine residues prevents absolute assignment of correct PTM -Applied only to purified protein preparations, not complex mixtures
5)Organomercury-based Affinity Enrichment	Blockage of free thiols with MMTS. Reaction with organomercury reagents. LC-MS/MS to identify independently protein targets and sites of modification.	-Site-specific identification of 328 endogenous S-nitrosylated peptides in 192 proteins in a single experiment -Reliance on organomercury-SNO chemistry; requires negative controls

1.0 The Biotin Switch and Its Variations

1.1 The Biotin Switch

The biotin switch [9] was the first and is the most commonly used technique for identifying S-nitrosylated proteins. Three steps constitute the basic principles of the method: 1) initially, proteins in a purified preparation or complex mixture are denatured by SDS (or urea, which is more amenable to downstream mass spectrometry analysis, 10); to expose protein cysteine residues. Subsequently, the reduced thiols in these proteins are blocked by reaction with reagents such as methyl methanethiosulfonate (MMTS), N-ethylmaleimide (NEM), or iodoacetic acid (IAA). 2) Following this blocking step, S-nitrosylated -cysteine residues are selectively reduced by treatment with ascorbate. 3) The ascorbate-reduced cysteine residues are then reacted with biotin-HPDP (N-[6-(biotinamido)hexyl]-3’-(2’-pyridyldithio)-propionamide). The conjugation of these protein cysteine residues to biotin permits enrichment of the modified targets from a complex mixture using avidin-based affinity capture. The affinity enriched preparations can be probed with antibodies, if the protein targets of interest are known, or digested with trypsin and subjected to liquid chromatography tandem mass spectrometry (LC-MS/MS). One added benefit of this method is that it is amenable to stable isotope labeling with amino acids in cell culture (SILAC) in order to quantitatively determine the levels of S-nitrosylation for specific cysteine residues via mass spectrometry-based approaches [11].

1.2 Biotin Switch-based Peptide Identification

The original biotin-switch method did not provide the means to identify the site(s) of modification in S-nitrosylated proteins. Identification of these sites is the ultimate qualifier for the unambiguous assignment of S-nitrosylated proteins. Site-specific identification enables mutational analysis to explore the functional role of this modification. Two similar approaches to achieve site-specific identification of modified cysteine residues have been developed based on the biotin switch [12, 13]. For both methods, proteins are first digested with trypsin and the resultant peptides are incubated with avidin or its derivatives. These peptides are then eluted and subjected to LC-MS/MS to determine the site of modification.

1.3 The His-Tag Switch

Since the original description of the biotin switch methods, several variations have been developed. One of these variants, the His-Tag switch, [14] begins with the blocking of free thiols by NEM, followed by ascorbate reduction of S-nitrosylated cysteine residues. It diverges from the biotin switch, however, by treating these ascorbate-reduced cysteines with a conjugate of iodoacetate and a His-containing peptide. Proteins containing this “His-tag” are then enriched through affinity chromatography in a nickel column, after which they are eluted and subjected to 1-dimensional (1D) gel electrophoresis. In-gel digestion with trypsin is then performed. Trypsin, in addition to cleaving the peptide backbone also facilitates cleavage of part of the alkylating agent, resulting in a mass increase of the cysteine residue by 271.12 Da, which is used to identify the modified peptide by LC-MS/MS. This method thus provides another approach for identifying the site(s) of modification in S-nitrosylated proteins.

1.4 Fluorescence-based Detection (DyLight, Cyanine, and AMCA-based Methods)

Another method based on the biotin switch relies on 2-dimensional differential gel electrophoresis (2D-DIGE, 15). The initial steps of this approach are identical to the biotin switch. After blocking by MMTS (or NEM) S-nitrosylated-cysteine residues are reduced by ascorbate. However, instead of biotin-HPDP, a set of DyLight maleimide sulfhydryl reactive fluorescent compounds are used to react with the newly reduced cysteine residues. Individual samples labeled with a DyLight fluorescent compound are separated on a 2D gel. By comparing the fluorescent intensity of a single spot in the gel, a relative assessment of protein S-nitrosylation levels can be made. Another variation of this method employs fluorescent cyanine maleimide sulfhydryl reactive compounds instead of the DyLight compounds [16].

An additional iteration of this “fluorescence switch” utilizes 7-amino-4-methyl coumarin-3- acetic-acid (AMCA)-HPDP to label S-nitrosylated cysteine residues after ascorbate reduction [17]. In this method, samples are resolved on either 1D or 2D gels after labeling. Subsequently, the gel is subjected to UV illumination, which activates the AMCA fluorophore and allows it to be directly visualized. All three versions of the “fluorescence switch” method allow for relative quantification of S-nitrosylation levels of proteins between samples through comparison of fluorescent intensities. However, in each case the site(s) of S-nitrosylation in specific proteins remains elusive, thereby rendering further studies of the functional impact of this modification challenging.

1.5 d-Switch

In addition to the capabilities offered by a combination of SILAC labeling and the biotin switch, another quantitative method was developed and tested in a purified protein preparation [18]. In this technique, reduced cysteine residues in recombinant glutathione S-transferase P1 (GST-P1) are labeled with NEM. S-nitrosylated cysteine residues in GST-P1 are then subjected to ascorbate reduction, followed by treatment with deuterated (d₅)-NEM and in-gel trypsin digestion. Mass spectrometry is then used to determine the relative amounts of peptides containing only d₅-NEM versus those containing both d₅-NEM and NEM, providing a quantitative assessment of S-nitrosylation.

1.6 SNO-RAC

Recently, another approach was developed by Stamler and colleagues to enrich for S-nitrosylated proteins from a complex mixture [19]. Referred to as “SNO-RAC,” this method relies on the conjugation of reduced cysteine residues to a solid support, such as thiopropyl sepharose. The first two steps of the protocol are identical to the biotin switch. However, the ascorbate-reduced cysteine residues are incubated with thiol-reactive resins, resulting in a covalent disulfide linkage. At this point, the proteins can be eluted and analyzed by western blotting. Alternatively, the disulfide linkage allows for on-resin trypsin digestion of bound proteins, resulting in site-specific identification of modified cysteine residues by LC-MS/MS. Compared to the traditional biotin switch, SNO-RAC has a better sensitivity for proteins with higher mass (>100 KDa). When combined with iTRAQ labeling, it can also report on the S-nitrosylation/de-nitrosylation of specific cysteine residues on a global scale [20].

1.7 Overview of the Biotin Switch-based Methods

Application of the biotin switch method and its various iterations to the identification of endogenously S-nitrosylated proteins (defined here as proteins in cell lysate or tissue homogenate without nitric oxide or trans-nitrosylating donor treatment) has yielded 135 targets from multiple organs and cell types. Within this subset of proteins, 82 sites have been reported from 63 proteins, illustrating the importance of the biotin switch in exploring the S-nitrosoproteome. However, three potential issues have been discussed with regard to these methods. First, the efficiency/sensitivity of this assay relies on complete blocking of reduced cysteine residues. Incomplete blocking will result in false identification, which can be minimized by the inclusion of negative controls such as pretreatment with ultraviolet (UV) photolysis or dithiothreitol (DTT). Second, the efficiency of ascorbate reduction has been questioned [21, 22, 23], suggesting decreased sensitivity of the method. The concerns over ascorbate reduction are compounded by its potential ability to reduce disulfides [24, 25], leading to false identification. Again, the inclusion of negative controls (listed above) as well as samples not treated with ascorbate can be employed. Alternatively, sinapinic acid has been used to treat cell lysates in place of ascorbate as a more selective method of reducing S-nitrosylated cysteine residues [26]. Third, there is the possibility of disulfide exchange after ascorbate reduction, leading to false identification of modified cysteine residues [20]. The use of methods that rely on a direct reaction with S-nitrosylated residues without requiring ascorbate reduction may overcome this concern. Taken together, these potential methodological challenges stimulated the development of alternative approaches for exploring the S-nitrosoproteome.

2.0 Direct Detection of S-nitrosylation by Mass Spectrometry

Theoretically, many of the concerns noted with the biotin switch and other chemical derivatizations can be avoided by direct detection of S-nitrosocysteine by mass spectrometry. In practice, such an approach remains challenging and has been possible primarily for isolated proteins. Certain MS approaches such as matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) MS, where the energy needed for peptide ionization also causes the loss of nitric oxide from the cysteine residues, can be employed. The reduction in mass by 29 Da in peptides after ionization is then diagnostic of S-nitrosylated sites. Within the last five years, a few groups [27, 28] have been able to directly identify S-nitrosylated cysteine residues by electrospray ionization quadrupole time-of-flight (ESI-QTOF) MS. However, such investigations were limited to synthetic proteins and/or peptides, or were only able to identify a few sites of S-nitrosylation from a complex mixture, rendering them unsuitable in their current form to proteome-wide identification of S-nitrosylated protein targets.

3.0 Gold Nanoparticle-based Enrichment

A method that relies on a direct reaction of S-nitrosocysteine residues with gold nanoparticles (AuNPs) has been developed for identifying sites of S-nitrosylation in purified protein preparations [29]. In this method, reduced cysteine residues are initially alkylated by iodoacetamide (IAM), after which the protein is subjected to proteolysis. Following this digestion, peptides are incubated with AuNPs to selectively react with the S-nitrosylated cysteine residues, generating free nitric oxide and peptides conjugated to the AuNPs. The nanoparticles are then treated with DTT to elute bound peptides, which are subsequently analyzed by mass spectrometry to identify sites of modification. Despite the potential advantages offered by this method, two drawbacks exist. First, the AuNPs react with both S-nitrosylated and S-glutathionylated cysteines, providing a challenge for absolute assignment of specific post-translational modifications to these residues. Second, this approach has not yet been applied to complex mixtures. Nonetheless, it holds excellent promise for site-specific identification of modified cysteine residues in S-nitrosylated proteins.

4.0 Phosphine-based Direct Labeling of S-Nitrosylated Proteins

To overcome the reliance of previous methods on the complete blocking of reduced cysteine residues (either by MMTS, NEM, or other reagents), Zhang and colleagues [30] introduced another approach to directly label S-nitrosylated cysteine residues in cell extracts. For this particular method, S-nitrosylated cysteine residues are reductively ligated (in the presence of water) with a biotin-labeled phosphine substrate, resulting in the generation of a sulfenamide product and thiolate. The sulfenamide and thiolate then spontaneously react to provide a stable disulfide (conjugated to biotin) at the formerly S-nitrosylated cysteine residue. Avidin-based enrichment of these labeled proteins is performed, and captured proteins are then resolved on a 1D gel. This particular approach is advantageous in its reactive specificity toward S-nitrosylated cysteine residues in the absence of blocking. Following the introduction of this method, a number of phosphine-based compounds have been developed that directly react with S-nitrosylated -cysteine residues [31, 32, 33]. Although these approaches have yet to identify endogenously modified sites, these compounds offer intriguing possibilities to explore protein S-nitrosylation in vivo.

5.0 Organomercury- Based Capture

After considering some of the limitations inherent in the biotin switch technique and its successors, we introduced novel complementary methods for capturing S-nitrosylated proteins and identifying their sites of modification [34]. In this protocol, phenylmercury compounds (either conjugated to an agarose solid support or to polyethylene glycol-biotin) react directly with S-nitrosocysteine residues to form a stable thiol-mercury bond [35]. The first step of this procedure is the same as that of the biotin switch. Following the blocking step, proteins are incubated with either the organomercury-conjugated resin (MRC) or a soluble phenylmercury-polyethyleneglycol-biotin (mPEGb) compound. After formation of the thiol-mercury bond, a number of options are available. S-nitrosylated proteins can be enriched by either the MRC approach or by avidin-based affinity capture (for mPEGb), and eluted using beta-mercaptoethanol to reduce the thiol-mercury bond. Eluted proteins are then subjected to 1D gel electrophoresis, in-gel trypsin digestion, and LC-MS/MS analysis. In order to identify the specific cysteine residues modified by S-nitrosylation, a slightly altered protocol is applied. After incubation of blocked proteins with the phenylmercury-based reagents, proteins are then digested with trypsin on-resin or in-solution. The resultant peptides are either eluted with mild performic acid to oxidize the cysteine in the thiol-mercury bond to cysteic acid (for the MRC approach), or subjected to a combination of avidin-based affinity capture and performic acid oxidation. In both cases, the cysteic acid generated in captured peptides is used as the MS signature to identify the cysteine modified by S-nitrosylation. Application of these methods yielded 328 endogenous sites of S-nitrosylation in 192 proteins in the wild-type mouse liver [34]. The method relies on the inclusion of negative controls and the reporting of the percentage of peptides identified as false positives (those shared between negative controls and experimental samples).

Advantages of this approach include the fact that it circumvents the ascorbate reduction step. Additionally, it provides an opportunity to pinpoint the modified cysteine residue using the MS/MS signature of the cysteic acid (C+48). Since both proteins and peptides are identified independently, the peptides can be matched to the proteins and thus not rely on a single peptide for protein identification, as in previous methods [12, 13].

6.0 Overview of mass-spectrometry-based methodologies for detection of S-nitrosylated proteins

As with any other PTM-based proteomic studies, there are some potential caveats associated with global analyses of in vivo S-nitrosylated proteins. (1) Up to this point, only a few methods afford sensitivity for in vivo detection, and often investigators rely on the induction of S-nitrosylation by applying exogenous nitric oxide donors or trans-nitrosylating agents to cells and protein preparations. Unfortunately, such studies provide only putative sites of modification, not necessarily those modified in vivo [36]. (2) Some methods, but not all, do not provide the site of modification. The identification of the site not only provides confidence for the correct identification of the protein but also enables follow-up studies, such as mutational analysis, to explore biological function. (3) Negative controls and the reporting of false identification rate (FIR) must be routinely evaluated. (4) We must also consider the possibility that even a method which identifies the sites of modification, with appropriate negative controls and low FIR, may still only report a subset of the modified proteins that are either most abundant or more stable. Despite these limitations, it is still worthwhile to pursue inquiries into the S-nitrosoproteome, if only to provide more information about the structural and functional importance of this modification in vivo.

Conclusions

Compared to widespread studies of PTMs such as protein phosphorylation, glycosylation, or even acetylation, the field of S-nitrosylation represents relatively uncharted territory. Within the last few years, however, this area of research has become quite prominent due to improvements in the detection of S-nitrosylated cysteine residues. By applying mass spectrometry-based proteomics, a greater appreciation of the biological significance of this modification is emerging. Additional work is needed to improve the sensitivity of these methods. Armed with sensitive and specific techniques, the S-nitrosoproteome of multiple organs and/or cell types can be investigated. The rich data from such studies can be analyzed in a variety of ways. Structurally, it can reveal significant new insights on elements that guide the selectivity of the modification. Additionally, it can also provide clues to potential biochemical reactions that derive the modification as well as explore the relative stability of different sites of modification. Functional analyses can uncover pathways and functional clusters of S-nitrosylated proteins, as well as their cellular location(s) [37], allowing for development of novel hypotheses that can be tested using more targeted approaches.

Highlights.

• S-nitrosylation of cysteine residues an important regulator of protein function

• Description of methods to identify S-nitrosylated proteins

• Mass spectrometry-based proteomics essential for further studies of S-nitrosylation biology