Profiling the landscape of cysteine posttranslational modifications in brain aging and neurodegeneration

Abstract

Cysteine residues occupy a unique position in the proteome: their thiolate side chain combines high nucleophilicity with redox sensitivity, making them prime targets for a diverse and ever-expanding array of post-translational modifications (PTMs). This review provides an overview of recent methodological developments for chemoselective site-specific detection and quantitation of the major cysteine PTMs—sulfenylation (RSOH), sulfinylation (RSO₂H), sulfonylation (RSO₃H), persulfidation (RSSH), S-nitrosylation (RSNO), and S-palmitoylation—emphasizing applications in brain aging and neurodegeneration. In neural tissues, these approaches have begun to map age-dependent increases in sulfenylation and sulfonylation, declines in persulfidation, and aberrant S-nitrosylation and palmitoylation linked to Alzheimer’s, Parkinson’s, and Huntington’s disease. However, significant challenges remain. Further improvements in sensitivity, specificity, and quantitative accuracy are essential to capture low-abundance and labile modifications in complex neural tissues. These attempts should be coupled to more detailed anatomical dissection of these modifications in different parts of the brain, enabling region- and cell-type–specific insights. Advancing analytical workflows, integrating multi-dimensional data, and linking chemical modifications to biological outcomes will pave the way for innovative therapeutic strategies targeting cysteine chemistry in neurological disease.

Introduction

Owing to rich and versatile sulfur chemistry, cysteine residues occupy a unique position in the proteome: their thiolate side chain combines nucleophilicity with redox sensitivity, making them prime targets for a diverse and ever-expanding array of post-translational modifications (PTMs). One of the first widely reported regulatory cysteine PTMs, aside from disulfide formation, was S-nitrosylation—discovered in the early 1990s—which revealed nitric oxide’s role in reversible thiol modulation (Fig. 1) [1,2]. The importance of reactive oxygen species (ROS) and their cellular impact spurred research into sulfenylation (as well as sulfinylation and sulfonylation), illuminating hydrogen peroxide (H₂O₂)-driven signaling [3]. With the discovery of hydrogen sulfide (H₂S) as a new gasotransmitter, focus shifted to persulfidation, which proved to be the primary mechanism through which H₂S signals [4]. More recently, novel modifications such as cyanylation [5] or lysine–cysteine redox switches [6] have emerged, underscoring that cysteine PTMs continue to grow in number and complexity beyond classical lipidation and oxidation events.

Fig. 1.

Overview of Major Regulatory Cysteine PTMs and Their Inducing Biological Signals. H₂O₂: hydrogen peroxide; NO: nitric oxide; H₂S: hydrogen sulfide; RSH: any thiol, including glutathione; HCN: hydrogen cyanide; palmitoyl-CoA: palmitoyl coenzyme A.

In the brain—where intense metabolic activity generates reactive oxygen, nitrogen, and sulfur species, and where antioxidant defenses decline with age—cysteine PTMs serve as both dynamic signaling switches and potential drivers of dysfunction. Reversible modifications (sulfenylation, S-nitrosylation, persulfidation, palmitoylation) can fine-tune enzyme activity, metal binding, and protein–protein interactions, while irreversible modifications (sulfinylation, sulfonylation) often mark proteins for degradation or contribute to aggregation. Dysregulation of this redox code is implicated in Alzheimer’s, Parkinson’s, and Huntington’s diseases, where aberrant cysteine chemistry intersects with protein misfolding, synaptic failure, and altered condensate dynamics.

Initially viewed as part of a simple reduced/oxidized cysteine cycle, thiol-redoxome analyses were constrained to measurements of “reduced/oxidized thiol ratios” (± dithiothreitol (DTT) or ​± ​tris(2-carboxyethyl)phosphine (TCEP)). However, these treatments cannot distinguish between disulfides, sulfenic acids, persulfides, and S-nitrosothiols—all of which fall under the DTT/TCEP-reducible category [7]. Thus, one might observe “no change” in overall thiol oxidation [8] while significant shifts occur among specific cysteine PTMs [[9], [10], [11]]. In fact, large changes in “reduced/oxidized” thiol ratios likely occur only under extreme pathological conditions, and such assays fail to detect the subtle but significant changes that constitute cysteine PTM signaling pathways.

Mapping these modifications site-specifically—especially in neural tissue—is challenging. Many cysteine PTMs are labile: sulfenic acids and persulfides can interconvert or over-oxidize during lysis, and S-nitrosothiols are thermolabile and light-sensitive. A better understanding of the chemical nature of these modifications [[12], [13], [14], [15]] has driven the development of tailored chemoselective probes and workflows for proteome-wide PTM detection. These technological advances have been propelled by the expansion of click chemistry in biology and improvements in mass spectrometry (MS). Quantitation has likewise evolved: tandem mass tags enable multiplexed comparisons across disease models or treatment conditions, while data-independent acquisition MS delivers reproducible, high-throughput profiling with deeper proteome coverage [16,17]. Sophisticated bioinformatic pipelines now integrate site-localization scoring, relative-occupancy estimation, and network analysis, enabling researchers to discern co-regulation patterns and infer functional modules [8]. Despite these gains, significant gaps remain. New tools that provide better spatial resolution at the level of brain subregions or individual cell types are still needed, though emerging approaches—such as laser-capture microdissection combined with micro-scale chemoproteomics—show promise [[18], [19], [20]].

This review will focus on these methodological advances—chemistries, enrichment strategies and quantitative workflows—as they have been applied to brain tissue, summarizing current capabilities and highlighting persistent hurdles in decoding the expanding cysteine PTM landscape in neurobiology.

Protein Sulfenylation, Sulfinylation and Sulfonylation in Brain Aging and Neurodegeneration

Protein sulfenylation

Protein sulfenylation, the one-electron oxidation of a cysteine thiolate (RS⁻) by H₂O₂ or related ROS to form sulfenic acid (RSOH) (Fig. 1), is a reversible and highly labile post-translational modification that underlies redox signaling in diverse biological contexts [12,21]. As a chemical “switch,” sulfenylation transiently alters protein structure, activity, and interactions, yet it also serves as a gateway to irreversible over-oxidation (sulfinic, RSO₂H, and sulfonic acids, R–SO₃H) under conditions of excessive oxidative stress (Fig. 1).

Regulation of the sulfenylation cycle relies on the interplay between thiol-oxidizing and -reducing systems. Once formed, protein RSOH can undergo intramolecular disulfide exchange (RSSR), react with glutathione to yield mixed disulfides (RSSG) or react with hydrogen sulfide (H₂S) to yield protein persulfides (RSSH) (Fig. 2a) modulating enzyme activities and protein–protein interactions [14,22]. Recent studies suggest that RSOH could also react with endogenously produced hydrogen cyanide (HCN), resulting in protein cyanylation (RSCN) (Fig. 2a) [5]. Peroxiredoxins (Prx) exemplify a key regulatory node: their catalytic cysteine is transiently sulfenylated during peroxide detoxification, transformed to disulfide and subsequently recycled by thioredoxin [[23], [24], [25]], or if overoxidized to sulfinic acid, repaired by sulfiredoxin (Srx), thereby closing the redox cycle [26]. Failure of these repair pathways skews the balance toward irreversible oxidation and proteome dysfunction [10].

Fig. 2.

Principal methods for RSOH detection. A) Sulfenylated residues have multiple fates, each leading to a different cysteine PTM. Further reaction with H₂O₂ results in higher oxidation states (RSOxH, where x ​= ​2, 3). Reaction with thiols (RSH), such as glutathione (GSH), leads to disulfide formation. Reaction with H₂S results in protein persulfidation, while reaction with HCN yields S-cyanylation. B) Common methods for RSOH detection rely on the electrophilic nature of sulfur. Dimedone and its derivatives have seen the widest application. Anti-dimedone antibodies enable selective enrichment of proteins or peptides, followed by proteomics analysis and/or Western blotting. A biotinylated dimedone, DCP-Bio1, permits streptavidin-based enrichment of labeled proteins or peptides and subsequent MS analysis. “Clickable” probes—such as DYn-2 or DAz-2 (including deuterated forms), and newer reagents like BTD—are widely used for site-centric, proteome-wide analysis of sulfenylation. The most common click-chemistry tag is UV-cleavable biotin. C) Water-compatible Wittig reagents (WYneC, WYneO, WYneN) undergo rapid, chemoselective ligation to sulfenic acids. After initial labeling, proteins are trypsinized and the labeled peptides are coupled—via click chemistry—to either light or ¹³C-labeled heavy UV-cleavable biotin reagents. By mixing light and heavy samples in a 1:1 ratio, quantitative comparisons can be made. Labeled peptides are then enriched on streptavidin beads and selectively released by UV light for MS analysis.

The chemistry and signaling roles of protein sulfenylation have been subject of research for more than two decades and have been extensively covered by excellent review articles [12,21,27,28]. Notable examples of RSOH signaling roles are growth-factor receptor signaling [29], glycolysis [30] and circadian rhythmicity [31]. Carroll’s lab showed that EGFR-activated production of H₂O₂ converts EGFR catalytic Cys797 to sulfenic acid, further enhancing its kinase activity, while parallel sulfenylation of the antagonist phosphatases PTP1B and SHP2 delays their re-activation, together prolonging Ras–ERK output [29]. Quantitative “sulfenome” mapping with high-rate benzothiazine probes revealed that oxidation of the glycolytic gatekeeper GAPDH at Cys152 rises sharply under oxidative stress [32], which could result in transient halting of glycolysis and diverting of carbon into the NADPH-producing pentose-phosphate pathway to bolster antioxidant defense [33]. Endogenous H₂O₂ oscillations, produced by p66 Shc rhythmically sulfenylate CLOCK at Cys195; loss of this modification in p66 Shc-null mice lengthens circadian period and rewires hepatic transcriptomes, establishing sulfenylation as an integral cog in mammalian timekeeping [31].

Methods and challenges in detecting cysteine sulfenylation

The methodological foundation of sulfenylation analysis was laid in 1974, when dimedone (5,5-dimethyl-1,3-cyclohexanedione) was shown to react chemoselectively with sulfenic acids to form a stable thioether adduct that survives proteolysis and MS read-out (Fig. 2b) [34]. Owing to its unrivalled selectivity, dimedone—and later diketone analogs—became the benchmark probe for protein-wide sulfenylation surveys [12,35,36]. A key limitation, however, is its sluggish reaction rate [37], which restricts capture of transient oxidative events in living cells.

A major leap in analytical sensitivity came from “DCP family” (Fig. 2b). By tethering fluorescein, rhodamine or biotin to a 1,3-cyclohexanedione core, seven reagents (DCP-FL1/2, DCP-Rho1/2, DCP-Bio1/2/3) were generated that enable in-gel fluorescence, affinity enrichment and site-specific MS mapping [38]. The hydrolysable amide bond in DCP-Bio1 allows for selective protein of peptide release after streptavidin enrichment, but the adduct suffers from poor ionizability which hinders site-specific identification of modified cysteines [28,39].

By immunising rabbits with the unique thioether hapten, antibody against dimedone was generated, allowing easy detection of ROSH by Western blotting and efficient enrichment for subsequent proteomic analysis (Fig. 2b) [40].

Integrating the high selectivity of dimedone probes with bioorthogonal chemistry Carrol’s group produced a series of probes that link dimedone to azide or alkyne handles—DAz-1/2, DYn-1/2—for bio-orthogonal conjugation (Fig. 2b) [29,41,42]. Alkyne-tagged DYn-2, in particular, shows minimal cytotoxicity and delivers ∼40 ​% higher labeling signal than its azido counterpart. By mapping hundreds of modification sites in live cells, the study transformed the field from protein-level detection to site-resolved “sulfenomics,” providing the template still used for global analyses [32]. With a systematic kinetic screen of ∼100 cyclic carbon-nucleophiles, Carroll’s team showed that strategic benzofusion and electron-withdrawing substitutions could boost reaction rates more than 200-fold over classical dimedone while preserving exquisite –SOH selectivity, which resulted in use of BTD probes for site-centric chemoproteomic analysis to generate the first proteome-wide, site-resolved “sulfenome” maps in living cells (Fig. 2b) [42,43]. BTD’s intrinsic alkyne handle enabled low-dose labeling, and isotopically coded biotin tags permitted simultaneous identification and occupancy quantification for thousands of cysteines, linking oxidative cues to specific functional hotspots in metabolism, signaling and cytoskeletal regulation [43]. Most recently, a report harnessed the polarity reversal of sulfur upon oxidation to invent an “umpolung” covalent-ligand strategy. Competitive BTD chemoproteomics mapped more than 500 ligandable sulfenic acids across the human proteome, >80 ​% of which are invisible to electrophilic fragments [44].

A major chemical leap followed with the discovery that water-compatible Wittig reagents undergo rapid, chemoselective ligation to sulfenic acids (Fig. 2c) [45]. The reaction produced stable adducts, enabling global stoichiometry analysis and even redox-controlled mitochondrial targeting of folded proteins, thereby extending bioconjugation beyond detection to functional manipulation of oxidized cysteines. For the first time, the sulfenylation stoichiometry of 6623 cysteines was estimated in the cell. Surprisingly, ∼1/2 of those cysteine sites have steady-state protein sulfenylation at >10 ​%, with some sites existing >70 ​% in sulfenylated state [45].

Complementing proteomics, fluorogenic phenaline-dione probes that become fluorescent only after reacting with RSOH were also reported, allowing real-time imaging of sulfenylation waves inside living cells. Application to kinase-inhibitor screens revealed unexpected cross-talk between growth-factor signaling and cysteine oxidation, illustrating how spatially resolved redox data can refine drug mechanism-of-action studies [46].

Most recently, Furdui’s group reported [¹⁸F]Fluoro-DCP ([¹⁸F]F-DCP), the first positron-emission-tomography radiotracer for sulfenylation [47]. In mouse xenografts, [¹⁸F]F-DCP discriminates radio-resistant from radio-sensitive head-and-neck tumors, highlighting sulfenylation as a non-invasive biomarker of oxidative stress and therapeutic response. Kidney and liver accumulation remains a hurdle, but targeted formulations or alternative isotopes may overcome this limitation [47].

RSOH in brain aging and neurodegeneration

During brain aging, elevated mitochondrial ROS production and declining antioxidant capacity are expected to drive a global increase in protein sulfenylation. However, despite the most extensive methodological advancements, the role of RSOH in brain aging and in neurodegenerative diseases remains the most understudied of all “major” cysteine PTMs. Recent reports by Vignane et al., provides the first comprehensive analysis of protein sulfenylation in aging mouse brain [10]. The data suggest that ∼1000 proteins show age-dependent increase in protein sulfenylation, with pathways such as Alzheimer’s disease, Amyotrophic lateral sclerosis, Huntington’s disease and Parkinson’s disease being particularly enriched. More importantly, the study suggests that biophysical properties, such as liquid-liquid phase separation of the protein might be controlled by this modification defining the duration that protein can spend in these biomolecular condensates and potentially undergo aggregation [10]. Collectively, these insights underscore protein sulfenylation as a dynamic modulator of neuronal health and a compelling target for interventions to slow brain aging and halt neurodegenerative progression.

Cysteine sulfinylation

Cysteine sulfinylation (Cys-SO₂H) arises when sulfenic acids undergo a second two-electron oxidation (Fig. 1). Long regarded as a biochemical “dead end,” sulfinylation is now recognised as a partially regulated cysteine PTM that modulates peroxidase catalysis, metabolic flux and redox signaling [48]. Its cellular abundance is expected to be low relative to sulfenic acid, but accumulation could be prominent in tissues subjected to chronic oxidative stress, positioning Cys-SO₂H as a potential trigger and marker of neurodegenerative pathology [49]. Contrary to the dogma that sulfinylation is irreversible, typical 2-Cys peroxiredoxins can be returned to their active thiol state by Srx, an ATP-dependent cysteine-sulfinic-acid reductase. Structural and kinetic analyses show that Srx phosphorylates the Prx-SO₂H, forms a thiosulfinate intermediate and completes reduction via intramolecular disulfide transfer [50]. Recent work by Akter et al., identified >55 previously-unknown protein substrates of the Srx, extending its function well beyond those of 2-cysteine peroxiredoxins [48].

Due to their relatively inert chemical nature, sulfinic acids represent a challenge for selective labeling and detection. The major analytical breakthrough was the Carroll laboratory’s C-nitroso warhead probe NO-Bio [51], which undergoes rapid, chemoselective nitrosothiol transfer with sulfinic acids while sparing thiols and sulfenic acids (Fig. 3); the built-in biotin handle permits streptavidin enrichment and site-resolved LC-MS/MS mapping. Proteomic implementation in mammalian cells revealed ∼250 Cys-SO₂H sites, including peroxiredoxin (Prx) catalytic cysteines, glycolytic enzymes and several neuronal RNA-binding proteins. Subsequent optimization produced the DiaAlk diazene probe with a higher second-order rate constant and an alkyne tag compatible with click chemistry (Fig. 3) [48]. Coupled to isotopic “light/heavy” cleavable biotin linkers, DiaAlk enabled quantitative sulfinome profiling and uncovered redox-regulated hotspots in mitochondrial complex I, pyruvate dehydrogenase and the Parkinson’s-linked chaperone DJ-1 [48].

Fig. 3.

Recent Developments in RSO₂H Detection. Following initial thiol blocking, sulfinic acids are selectively labeled with probes carrying specific payloads—such as biotin or a “clickable” handle. The nitrogen atom subjected to the nucleophilic attack by the sulfinic group is highlighted in blue. The first-generation reagents (NO-Bio and NO-Ph) have been superseded by the newer Dia-Alk probes, which enable site-centric proteomic analysis of protein sulfinylation. As in the RSOH workflow, UV-cleavable biotin remains the most common click-chemistry tag.

The Parkinson’s-disease protein DJ-1 (PARK7) contains an unusually nucleophilic Cys106 that is selectively oxidized to the sulfinic state under physiological peroxide loads [52,53]. Crystal structures reveal that Cys106-SO₂H stabilises the DJ-1 dimer and exposes a mitochondrial-targeting surface, thereby promoting organelle translocation and peroxide detoxification [52,53]. Carroll’s DiaAlk profiling confirmed DJ-1 Cys106 as a bona-fide sulfinylation site in situ and identified neighboring redox-coupled lysines that modulate its chaperone function [48]. Over-oxidation to sulfonic acid abrogates activity and is enriched in post-mortem substantia nigra, suggesting that failure to maintain the sulfinic state marks transition from protective to pathogenic DJ-1 signaling (Fig. 4) [54,55].

Fig. 4.

Cysteine oxidation states control the fate and function of DJ-1. DJ-1, or PARK7, is postulated to serve as a buffer for ROS, as well as a signaling molecule [52,55]. Sulfinylation of its C106 results in dimer stabilization and mitochondrial translocation, where it supports the whole life of mitochondria, from bioenergetics to dynamics and motility. Hyperoxidation of C106 leads to a dimer destabilization and aggregation and this form is reported to accumulate in Parkinson’s disease [56].

Cysteine sulfonylation: chemistry, detection and emerging roles in neuro-ageing

Cysteine sulfonic acid results from the four-electron oxidation of a thiolate and represents the thermodynamic terminus of the thiol redox ladder (Fig. 1). Because the reaction reduces nucleophilicity at sulfur, RSO₃H is generally regarded as irreversible under biological conditions and therefore marks proteins for functional inactivation or proteolytic turnover [12]. Nevertheless, the modification is far from random damage: recent large-scale datasets show that specific, structurally privileged cysteines are reproducibly trioxidised in aging tissues, suggesting regulated accumulation that may contribute to chronic pathophysiology [57].

The sulfonylated residues are relatively inert and therefore represent a significant challenge for chemoselective labeling, one that has not been overcome yet. So far, the only reliable approach has been a direct shotgun identification by a +48 ​Da mass shift (Fig. 5). Two recent proteomic efforts have begun to quantify RSO₃H at the organismal level. Sánchez-Milán et al. profiled mouse dermal proteomes across the lifespan and found a progressive, site-specific rise in sulfonylation; the targets were strongly enriched for serine/threonine-rich motifs, and modelling showed that the trioxidised cysteine introduced negative charge and hydrogen-bonding capacity analogous to phosphoserine, thereby rewiring protein–protein interactions in aged skin [57]. In parallel, Vignane et al. used deep redox-proteomics in aging mouse cortex to identify a few hundred sites that show clear increase with aging [10]. The novelty of that approach was that the thiols were initially blocked by 4-choloro-7-nitro-beznofurazan probes, which not only block free cysteines but also sulfenic acids, minimizing artifactual thiol oxidation that could occur during sample preparation. Nonetheless, these approaches rely on detection of modified peptides without any specific enrichment which underrepresent the real number of modified sites. Attempts to enrich sulfonic acids with ionic affinity capture using polyarginine-coated nanodiamonds have been made, but this approach is limited by the nonspecific enrichment of phospho-peptides, which are more abundant (Fig. 5) [58].

Fig. 5.

Methods for detection of cysteine sulfonylation. There are currently only two approaches for RSO₃H detection. The first one relies on detection of mass shift of +48 ​Da. Fast blocking of all cysteines with commonly available thiol blocking reagents minimizes the artifactual cysteine oxidation during sample preparation. Enrichment of RSO₃H peptides with ionic affinity capture using polyarginine-coated nanodiamonds has been proposed.

Unlike sulfenic or sulfinic acids, sulfonic acid cannot be reduced by Srx; Prx-SO₃H therefore accumulates under sustained oxidative load and acts as a molecular timer that shuts down peroxidase buffering once repair capacity is exceeded [23,24,59,60]. Loss of Prx activity has been documented in aging brain and renders neurons hypersensitive to glutamate excitotoxicity and mitochondrial ROS bursts [61,62]. More broadly, trioxidation of active-site cysteines is expected to abolish catalytic function in phosphatases and metabolic enzymes, irreversibly diverting signaling and carbon fluxes during chronic stress [63]. Structurally, sulfonylation is expected to have similar effects on proteins like phosphorylation [57]. Indeed, recent work by Vignane et al., suggest that cysteine sites that undergo sulfonylation in aging brain are located within betta sheet domains, leading to increased protein’s ability to aggregate. The study identifies GAPDH sulfonylation, as a main driver of its aggregation [10].

As mentioned before, DJ-1 contains a hyper-reactive Cys106 that cycles between thiol, sulfenic and protective sulfinic states [52]. Under sustained oxidative insults, Cys106 progresses to sulfonic acid, destabilising the dimer, promoting aggregation and abolishing mitochondrial translocation (Fig. 4) [55]. Post-mortem substantia nigra from PD patients is enriched in the DJ-1-SO₃H isoform, and mouse models lacking functional DJ-1 display dopaminergic loss and motor deficits [64]. Although DJ-1 sulfinylation is partially reversible, the sulfonic state is terminal, effectively converting a neuroprotective chaperone into an inert aggregate nucleus.

Sulfonylation is therefore not just an “oxidative graveyard” but a bona-fide regulatory endpoint that encodes the history of redox stress in proteins critical for metabolism, proteostasis and neuronal resilience. Further method development which would allow for selective chemoproteomic site-centric peptide enrichment and analysis would provide deeper insights into this modification and its role in regulating protein structure and function.

Protein persulfidation

Protein persulfides (RSSH/RSS⁻) are obligatory intermediates in sulfur-transfer pathways that assemble iron–sulfur clusters and synthesize biotin, thiamine, lipoic acid, molybdopterin, and thiolated RNA bases [22,65]. Seminal work by Snyder’s team showed that persulfidation is not confined to these specialist enzymes but is a widespread post-translational modification, thereby laying the chemical foundation for many physiological and pharmacological actions of H₂S [4,66]. “Persulfidation” (preferred to the earlier, etymologically incorrect “sulfhydration”) will be used here for the RSSH/RSS⁻ couple; alternative IUPAC descriptors include hydridodisulfide and disulfane.

The S–H bond in RSSH is weaker than in RSH, lowering pK_a by ∼3–4 units [14,67]. Cysteine persulfide is thus ∼99 ​% deprotonated at physiological pH which coupled to the “α-effect” makes RSS⁻ more nucleophilic than RS⁻ [14,68,69]. Because persulfides bear “sulfane sulfur”,a protonated RSSH can itself serve as a soft electrophile. Nucleophilic attack by other thiols for examples would therefore occur at the inner sulfur, releasing H₂S (preferred over thiol expulsion owing to the lower pK_a of H₂S) [68,69]. This dual nature makes persulfides intrinsically unstable in solution as they can react with each other (half-lives of a few minutes for model compounds) [14,22,70].

Persulfides also outpace thiols in two-electron redox chemistry, reacting 20-fold faster with H₂O₂ and an order of magnitude faster with peroxynitrite [69]. Oxidation yields perthiosulfenic, -sulfinic and -sulfonic acids (RSSOH/RSSO₂H/RSSO₃H). Crucially, the internal S–S bond allows thioredoxin or reactive thiols to reduce even the highest oxide back to RSH [9,22]. This “rescue loop” (see section 6) distinguishes persulfidation from classical cysteine over-oxidation and underpins the hypothesis that persulfides serve as evolutionarily conserved sacrificial, yet ultimately repairable, redox buffers that limit irreversible protein damage during aging and oxidative stress [9].

Like phosphorylation/dephosphorylation, protein persulfidation is a regulated process. While the search for global persulfidases continues, persulfides are readily formed in the reaction of H₂S with oxidized thiols, most notably RSOH (Fig. 2) [9,69,71]. Protein-SOH, generated during H₂O₂ signaling, reacts several hundred-fold faster with HS⁻ than with GSH (Fig. 2), converting transient sulfenylation into persulfidation [69]. In cells lacking one of the H₂S-producing enzymes, cystathionine γ-lyase (CSE), H₂O₂ triggers exaggerated sulfenylation; H₂S donors restore normal signaling kinetics, showing that the SOH→SSH switch resolves redox signals [9]. A paradigmatic case is EGFR Cys797: initial sulfenylation increases kinase activity, whereas subsequent persulfidation, driven by inducible H₂S synthesis, dampens the signal [9]. Other mechanisms, such as disulfide reduction [72], metal-centered radical chemistry [73,74], and enzymatic generation of low-molecular weight persulfides [75] have been demonstrated to lead to localized changes in protein persulfidation.

Regulation requires efficient removal of RSSH (Fig. 6A). The thioredoxin (Trx)/thioredoxin-reductase (TrxR) system is the dominant depersulfidase: Trx reduces protein-SSH ∼10-fold faster than disulfides, releasing H₂S and forming the canonical Trx disulfide that is recycled by TrxR ​+ ​NADPH [71,76]. Global persulfide levels fall rapidly during cell lysis unless TrxR is inhibited by auranofin, confirming in-cell turnover. The Trx-related protein TRP14 provides backup under oxidative stress when Trx is diverted to peroxiredoxin repair [76]. A parallel glutathione cycle operates through glutaredoxin and glutathione reductase [76]. Because both enzyme families are NADPH-dependent, metabolic control of the NADPH/NADP ​⁺ ​ratio likely dictates the lifetime of persulfidated signaling states in vivo.

Fig. 6.

Methods for Persulfide Detection. A) Protein persulfides are formed from oxidized cysteines, e.g. by reacting with PSOH generated during H₂O₂-induced thiol oxidation. Persulfides are removed by the action of Trx/TrxR couple, which requires NADPH. B) Biotin Thiol Assay (BTA) and Variants. Free cysteines and persulfides are initially blocked using either iodoacetamide or N-ethylmaleimide derivatives bearing a biotin tag. After streptavidin enrichment, peptides containing persulfidated cysteines are released with DTT or TCEP and then analyzed by MS. A major caveat is that sulfenic acids also react with these thiol-blocking reagents, producing DTT/TCEP-cleavable adducts that compromise selectivity. C) Low-pH Quantitative Thiol Reactivity Profiling (QTRP). Alkylation is performed with iodoacetamide alkyne (IPM) at low pH, which keeps persulfides fully deprotonated (and thus highly reactive) while most free thiols remain protonated (and less reactive). Following click chemistry with UV-cleavable, biotinylated probes, labeled peptides are recovered without a reduction step. Quantitation is achieved by comparing the m/z of peptides, accounting for the extra sulfur atom in persulfide-containing species. D) Dimedone-Switch Method. In the first step, RSH, RSSH, and RSOH are all blocked with 4-chloro-7-nitrobenzofurazan (NBF–Cl). In the second step, persulfides are selectively labeled with a dimedone-based probe (e.g., DCP-Bio1), enabling streptavidin enrichment and downstream MS analysis.

Methodological advancements and limitations in persulfide detection

Detecting and quantifying persulfides: Reliable measurement of protein persulfides (RSSH) is technically demanding. The earliest estimate—that roughly one-quarter of protein cysteines exist in the persulfidated state [4]—now appears inflated. Historically, investigators quantified RSSH as part of the broader sulfane-sulfur pool, which includes protein persulfides, low-molecular weight persulfides and inorganic polysulfides. Two studies performed a quarter-century apart, but using the same chemical principle, report nearly identical values in the low-micromolar range [22].

Numerous labeling strategies have been described, but most fall short in either specificity or sensitivity. Contemporary proteome-scale workflows exploit the high nucleophilicity of persulfides after an initial electrophilic “block,” and three families of methods now dominate.

Biotin-thiol assay (BTA) and its modifications (ProPerDP, qPerS-SID): [[76], [77], [78]] In these methods, a biotinylated electrophile is used to alkylate both thiols (RSH → thioether) and persulfides (RSSH → mixed disulfide) (Fig. 6B). Streptavidin enrichment followed by DTT or TCEP reduction is expected to selectively release RSSH-derived species. Specificity, however, is questionable: sulfenic acids (RSOH) also react with N-ethyl-maleimide or iodoacetamide to form DTT-cleavable adducts [7], and thioethers generated with N-ethyl-maleimide undergo retro-Michael cleavage [79]. To address this, Bibli et al. pre-trapped RSOH with dimedone before labeling and identified 1536 persulfidated cysteines in endothelial cells [80].

QTRP (low-pH quantitative thiol reactivity profiling): [81] Relying on the same chemistry, but not using reduction step Yang’s group developed a MS method for site-centric RSSH detection (Fig. 6C). Alkylation at pH 5.0 capitalises on the lower pK_a and higher nucleophilicity of persulfides, while most thiols remain protonated and inert. Clickable, UV-cleavable tags enable direct MS comparison of “+S” versus “no-S” peptides; 1547 persulfidated sites on 994 proteins were reported using this approach. Limitation of this approach is that it simultaneously detects both, free thiols and persulfides, with the former often saturating MS as more abundant and widespread. Still, this method remains the only selective site-centric approach for RSSH detection. Recent study has suggested that mixed disulfides formed during labeling can undergo thiosulfoxide tautomerisation and upon reaction with a nucleophiles to thioether transformation, limiting quantification [82]. However, conditions used in that study are seldom encountered in routine cell-lysis protocols and are not used in any of the MS protocols.

Tag-switch methods: Of all the methods, tag-switch approach has found the widest application, owing to it versatile use and high selectivity. The method relies on thiol blocking with aromatic thiol blocking reagents, such as methyl-sulfonyl-benzothiazole (MSBT), under strong denaturing condition. This step transforms persulfides into reactive mixed disulfides, with inner sulfur being electrophilic, unlike in naturally occurring inter- and intra-molecular disulfides. This feature can be exploited for selective labeling with nucleophiles such as cyanoacetic acid-derivatives [83]. More than 2000 persulfidated proteins in Arabidopsis thaliana have been identified using this approach [84]. However, poor solubility of MSBT and MS-labile cyanoacetate tags render analysis challenging. A “dimedone-switch” variant (Fig. 6D) was introduced recently, where more soluble blocker 4-chloro-7-nitro-benzofuran (NBF–Cl) replaces MSBT. NBF–Cl blocks not only the thiols and persulfides but sulfenic acids and amino groups as well, giving a green fluorescence with the latter. This feature has proven useful as a total protein load for in-gel detection of RSSH, as it provides greater sensitivity over traditional Coomassie blue-based staining [9]. As a nucleophile the method uses dimedone-based reporters, which are commercially abundant. These improvements support in-gel detection, fluorescence microscopy and large-scale MS of persulfidated targets. ∼2000 proteins have been found to be endogenously persulfidated in mouse frontotemporal regions, with almost one half showing age-related changes [10].

Continued methodological refinement—ideally integrating orthogonal blocking, enrichment and cleavage chemistries—will be required before absolute quantification of the persulfidome becomes routine.

Protein persulfidation: emerging regulator of aging and neurodegeneration

Brain expresses three key H₂S producing enzymes, cystathionine γ-lyase (CSE), cystathionine β-synthase (CBS) and mercaptopyruvate sulfur-transferase (MPST). While CSE is predominantly found in neurons, CBS in astrocytes and glia, MPST is ubiquitously present in all cells [9]. Aging brains display a progressive fall in CSE, CBS and MPST, leading to a global decline in protein RSSH [9]. The drop in RSSH and increase of RSO₃H seems to be predominantly regulated by CSE.

CSE knockout have been shown to display Huntington-like motor phenotypes, and human HD striatum shows an 85–90 ​% reduction in CSE with marked depletion of P-SSH [85]. Mutant huntingtin suppresses CSE transcription via SP1 and ATF4 [85]; activating the PERK–ATF4 pathway with monensin restores H₂S, elevates persulfidation and increases oxidative-stress resistance in HD cells [86]. In spinocerebellar ataxia-3, brain CSE and P-SSH are similarly reduced; Drosophila over-expressing CSE show rescue of eye degeneration, underscoring a causal link between the CSE→H₂S→P-SSH axis and polyQ toxicity [87,88].

CSE-catalyzed H₂S formation and subsequent persulfidation play important roles in Parkinson’s disease (PD) as well. Basal persulfidation of the E3-ligase parkin at C59/C95/C182 activates ubiquitination and proteasomal clearance of damaged proteins. PD patient striatum shows >70 ​% loss of parkin-SSH with reciprocal S-nitrosylation increase, correlating with toxic α-synuclein accumulation [89]. Persulfidation of DJ-1 at C106 limits formation of irreparable sulfonic acid, supporting its ROS-scavenging role; declining P-SSH on DJ-1 accompanies disease progression [9,87].

Links between CSE, persulfidation and Alzheimer’s disease have been made as well. [90,91] In healthy neurons, tau binds CSE, boosting H₂S output and persulfidating glycogen-synthase-kinase-3β (GSK3β) at C218, which suppresses its kinase activity and limits tau phosphorylation [92]. Aging or AD brains show decrease of CSE expression and subsequently GSK3β persulfidation, resulting in hyper-phosphorylated tau and tangle formation (Fig. 7). Slow-release H₂S donor GYY4137 restores global P-SSH and improves cognition in 3 ​× ​Tg-AD mice [92].

Fig. 7.

Persulfidation controls Tau phosphorylation in Alzheimer’s disease. Tau interacts with cystathionine gamma lyase, CSE, stimulating its H₂S producing activity which results in persulfidation of glycogen synthase kinase 3β (GSK3β) and inhibition of its activity. With aging and in Alzheimer’s disease (AD), there is a decline in CSE levels. GSK3β becomes more active leading to hyperphosphorylation of tau and its aggregation.

Across disorders, diminished CSE expression and persulfidation converge with elevated cysteine oxidation, impairing redox-sensitive proteins that govern proteostasis (parkin), mitochondrial quality control (DJ-1), kinase signaling (GSK3β) and autophagy. Recent studies suggest that this may be because H₂S-mediated protein persulfidation dissolves biomolecular condensates formed in the cells upon the exposure to stress, preventing these hyperconcentrated micro compartments to enforce protein aggregation [10]. Restoring persulfidation—genetically or with brain-penetrant H₂S donors—therefore emerges as a unified therapeutic strategy for neurodegeneration. Indeed, ergothioneine, an alternative substrate for CSE and booster of RSSH levels [93], improves cognitive performance of patients with early signs of dementia [94].

Protein S-Nitrosylation

Protein S-nitros(yl)ation is the covalent, reversible attachment of a nitroso group to a cysteine thiol, forming S-nitrosothiols (RSNO) (Fig. 1). This PTM translates nitric oxide (NO) signals into discrete functional changes on target proteins. The most studied cysteine PTM, protein S-nitrosylation has been subject of numerous review articles [13,[95], [96], [97], [98], [99]]. The seminal works by Stamler’s [1,2,100,101], Lipton’s [[102], [103], [104]] and Snyder’s group [66,[105], [106], [107]] laid foundation for the current understanding of the signaling roles of S-nitrosylation.

While several chemical pathways for this reaction exist, the most biologically relevant mechanism, particularly within the central nervous system, involves an intermediate with nitrosonium cation (NO⁺) character [99]. This species is typically generated through the interaction of nitric oxide (⋅NO) with transition metals like iron (Fe³⁺) or copper (Cu²⁺). The resulting NO⁺-like electrophile then reacts with a nucleophilic thiolate anion (RS⁻) on the target protein to form the RSNO adduct [108]. Other potential mechanisms, such as the direct radical recombination of ⋅NO with a thiyl radical (RS⋅) or nitrosation via N₂O₃, are considered less common under typical physiological conditions [108].

The generation of NO is orchestrated by three distinct nitric oxide synthase (NOS) isoforms. Neuronal NOS (nNOS/NOS1) and endothelial NOS (eNOS/NOS3) are constitutively expressed and activated by intracellular calcium (Ca²⁺) to produce physiological levels of NO that regulate normal signaling. The third isoform, inducible NOS (iNOS), is typically expressed in glial cells following inflammatory stimuli, often leading to excessive NO production. While early research linked NO signaling primarily to the activation of soluble guanylate cyclase (sGC) [109], subsequent studies have established protein S-nitrosylation as the major sGC-independent pathway for NO’s biological functions [110]. Pathological conditions arise when NO production becomes excessive, for instance, through hyperactivation of nNOS by extrasynaptic NMDA receptors or the induction of iNOS, leading to aberrant S-nitrosylation and neurotoxicity [97,99,111].

The functional consequences of S-nitrosylation are profound, as it often targets cysteine residues that are vital for a protein’s structure and activity. Similar to phosphorylation or acetylation, this modification can modulate a wide range of cellular processes, including enzymatic activity, ion channel conductance, protein trafficking, and intermolecular interactions [96,99,112].

A key mechanism for propagating the signal is trans-nitrosylation, the transfer of an NO group from one SNO-protein to a cysteine on another protein. This process occurs via a reversible nucleophilic attack by a thiolate anion on the nitroso nitrogen of the donor SNO-protein [113]. The specificity and efficiency of both initial S-nitrosylation and subsequent trans-nitrosylation are tightly regulated. This regulation occurs through both enzymatic control and non-enzymatic factors, including the target cysteine’s pKa, the local redox environment, protein conformation, hydrophobic microenvironments, and the presence of specific S-nitrosylation motifs within the amino acid sequence [99,114]. The dynamic nature of this modification is maintained by a corresponding process of denitrosylation, catalyzed by enzymes like the thioredoxin system [95,100]. An imbalance where S-nitrosylation outpaces denitrosylation, often due to the pathological overproduction of NO, can disrupt protein function, contributing to cellular stress and the progression of neurodegenerative disorders.

Methods and challenges in RSNO detection

Over the past two decades, several chemoselective enrichment techniques have emerged, each leveraging distinct reaction chemistries to stabilize, capture, and quantify S–NO sites.

The Biotin-Switch Assay: Originally conceived by Jaffrey and Snyder in 2001 [115], the biotin-switch assay revolutionized S–NO detection by converting transient S-nitrosothiols into stable biotin adducts. In its canonical three-step workflow, (i) all free thiols are alkylated with methyl methanethiosulfonate (MMTS) under denaturing conditions; (ii) S-nitrosothiols are then selectively reduced by ascorbate to liberated thiols; and (iii) these newly exposed thiols are tagged with biotin–HPDP for avidin-based enrichment and downstream immunoblot or MS analysis (Fig. 8A).

Fig. 8.

Methods for RSNO detection. A) Biotin-Switch Assay. Thiol groups are first blocked with methyl methanethiosulfonate (MMTS). S-nitrosothiols (RSNOs) are then selectively reduced by ascorbate, liberating free thiols that are trapped by HPDP-biotin (a biotinylated pyridyldithiol reagent). Labeled proteins are enriched on streptavidin beads and analyzed by mass spectrometry. B) Resin-Assisted Capture (SNO-RAC). This variant replaces HPDP-biotin with thiopropyl Sepharose resin. After ascorbate-mediated reduction of RSNOs, the newly exposed thiols bind directly to the resin, allowing for on-bead enrichment and subsequent MS analysis. C) Cys-BOOST Method. Free thiols are initially alkylated with iodoacetamide (IAM). RSNOs are then reduced by ascorbate, and the resulting thiols are tagged with iodoacetamide-alkyne. Remaining disulfides are reduced with TCEP and blocked with IAM. A biotin tag is introduced via click chemistry, and labeled peptides are enriched on streptavidin beads for MS analysis. D) Organo-Mercury Capture. RSNOs react directly with an organomercury resin, enabling selective capture of S-nitrosylated proteins or peptides for downstream analysis. E) SNOTRAP. Thiol groups are first blocked with IAM. RSNOs are then directly labeled using a triaryl-phosphine probe bearing a biotin moiety. The biotinylated proteins or peptides are enriched on streptavidin beads and identified by MS.

While broadly adopted, the assay suffers from key shortcomings. MMTS may incompletely block sterically buried thiols, leading to artefactual labeling and false positives upon biotinylation. Ascorbate, though relatively selective for S–NO, can also reduce sulfenic acids and disulfides, confounding true S–NO quantitation.

Resin-Assisted Capture (SNO-RAC): To streamline the biotin-switch protocol, Forrester et al. introduced SNO-RAC [116]. After identical MMTS blocking and ascorbate reduction, the newly formed thiols are captured directly on a thiol-reactive resin (e.g. thiopropyl sepharose), obviating both biotinylation and streptavidin steps (Fig. 8B). Bound proteins are eluted under reducing conditions and identified by MS.

SNO-RAC typically yields higher recovery and lower background than the biotin-switch, with fewer handling steps reducing sample loss. However, it remains a binary capture: a protein bearing a single S–NO site binds once, precluding direct assessment of multi-site occupancy. Moreover, resin capacity must be carefully titrated to avoid under- or over-loading, and harsh elution can unintentionally release non-SNO proteins via disulfide cleavage.

Bioorthogonal Tagging with Cys-BOOST: Cys-BOOST refines ascorbate-driven workflows through click-chemistry tagging [117]. Following free thiol blocking with iodoacetamide and RSNO reduction by ascorbate, free thiols are alkylated with an iodoacetamide-alkyne probe; a copper-catalyzed azide–alkyne cycloaddition then appends a cleavable biotin handle. Peptides can be released from streptavidin in a traceless manner, enhancing site localization and quantitative precision (Fig. 8C). The high specificity and efficiency of click chemistry translate into improved enrichment yields. This method provides the in-depth site-centric quantitation of S-nitrosylation.

Organo-Mercury Capture: Bypassing reductive steps altogether, organo-mercury resins exploit the high affinity of mercury for S-nitrosothiols [118]. Proteins bearing S–NO react directly with the resin, forming stable mercury–thiolate bonds; subsequent on-column digestion or elution facilitates MS identification (Fig. 8D). This approach affords high chemical selectivity for SNO versus other cysteine oxo-forms and avoids ascorbate artifacts. However, the toxicity and environmental hazards of mercury reagents pose practical drawbacks. Moreover, the robust Hg–S bond can hinder proteolysis, compromising peptide recovery and sequence coverage.

Triaryl-Phosphine Based Capture (SNOTRAP): Triaryl-phosphine reagents (e.g. TPP) react chemoselectively with S-nitrosocysteine to form biotin-tagged phosphine oxide adducts, enabling direct pulldown without reduction (Fig. 8E) [119]. Labeled proteins are enriched on streptavidin and quantified by MS, providing semi-quantitative readouts of S–NO levels. SNOTRAP’s direct labeling minimizes indirect reduction artifacts and false positives. Yet phosphine reagents are costly, sensitive to oxidation, and their reaction kinetics under physiological pH can be slow—prolonged incubations risk spontaneous S–NO decay. Off-target reaction with sulfenic acids, though slower, also necessitates parallel controls.

Recent MS developments in quantitation (SILAC: Stable Isotope Labeling by Amino acids in Cell culture, or isobaric probes: iodoTMT, iTRAQ) have been coupled to above-mentioned approaches to allow better measurement of dynamic changes in a multiplex manner and with drastically reduced sample amounts.

No single S–NO detection platform is universally optimal. In practice, rigorous validation often combines orthogonal chemistries (e.g. biotin-switch plus SNOTRAP) and orthogonal readouts (Western blotting, MS-based site mapping, and activity assays). A tiered approach—screening with a high-throughput assay (e.g. SNO-RAC), followed by site-specific validation (e.g. Cys-BOOST/MS or SNOTRAP), and functional assessment—remains the gold standard for comprehensive S-nitrosoproteomic profiling.

S-nitrosylation: the key cysteine PTM driving neurodegeneration

In the healthy central nervous system, nNOS-derived NO diffuses locally to modulate synaptic efficacy: S-nitrosylation of the NMDA receptor NR2A subunit reduces Ca²⁺ influx to prevent excitotoxicity [120], while transient RSNO on dynamin-related protein 1 (Drp1) fine-tunes mitochondrial fission to match energetic demands of synaptic activity [121]. Beyond ion channels and mitochondrial regulators, RSNO modulates transcription factors (e.g., S-nitrosylation of CREB enhances its DNA binding) [122], ion pumps (e.g., SERCA), and cytoskeletal proteins, orchestrating neurovascular coupling and plasticity [99].

Age-related elevation of RSNO levels has been documented in rodent hippocampus, where cumulative nitrosative stress impairs synaptic vesicle recycling and long-term potentiation, correlating with memory decline [11]. Specifically, increased S-nitrosylation of synaptic scaffolding proteins such as PSD-95 disrupts postsynaptic signaling complexes [107]. S-nitrosylation is a critical post-translational modification driving mammalian aging, primarily by disrupting mitochondrial homeostasis [123]. An age-related increase in the S-nitrosylation of key proteins, such as dynamin-related protein 1 (Drp1), triggers excessive mitochondrial fission and impairs mitophagy [123,124]. This results in the accumulation of dysfunctional mitochondria, a core mechanism that directly promotes cellular senescence. Complementing this, systems biology analyses of the aging brain reveal a more complex, systemic role [11]. Rather than a simple global increase, aging involves a significant and tissue-specific reprogramming of the S-nitroso-proteome, as observed in the cortical and striatal regions of mice. This targeted shift indicates that aging is characterized by distinct changes in which protein networks are regulated by S-nitrosylation [11]. Collectively, these findings establish that S-nitrosylation contributes to aging both by directly impairing critical organellar quality control and through a broader, systemic reprogramming of cellular protein function.

However, chronic or dysregulated RSNO underlies neurodegeneration. In Alzheimer’s disease, iNOS induction in glia yields high NO flux, leading to S-nitrosylation of protein disulfide isomerase (PDI), compromising its chaperone activity and exacerbating amyloid-β misfolding [102]. In Parkinson’s disease, parkin S-nitrosylation at Cys323 inactivates its E3 ubiquitin ligase function, causing build-up of misfolded proteins and dopaminergic neuron death [106]. Persistent S-nitrosylation of Drp1 also drives pathological mitochondrial fragmentation in multiple disorders [125]. Thus, RSNO operates as a molecular toggle: essential for synaptic homeostasis when tightly regulated, yet neurotoxic when sustained.

Emerging therapeutic strategies aim to rebalance S-nitrosylation: small-molecule denitrosylases (e.g., thioredoxin mimetics) and selective RSNO scavengers have shown neuroprotective effects in preclinical models, restoring mitochondrial integrity and reducing protein aggregates [126].

Role of Protein S-Palmitoylation in Brain Signaling and Neurodegenerative Diseases

Protein S-palmitoylation is a reversible lipid modification (generally called S-acylation) in which a 16-carbon palmitate moiety is covalently attached to cysteine residues via a thioester bond (Fig. 9A) [127]. Although not studied proteome wide as much as other cysteine PTMs, this dynamic post-translational modification orchestrates membrane association, subcellular trafficking, and protein–protein interactions, thereby finely tuning synaptic signaling [[128], [129], [130], [131]].

Fig. 9.

Methods for Detecting Cysteine Palmitoylation. A) Palmitoylation is a reversible modification in which a palmitic acid moiety is transferred from palmitoyl-CoA to a protein cysteine thiol, forming a thioester. This reaction is catalyzed by palmitoyl acyltransferases (PATs), while depalmitoylation is mediated by palmitoyl thioesterases (APTs). B) A site-specific detection method involves incubating cells with an alkyne-modified palmitate analog (17-ODYA), which is incorporated into proteins. A click-chemistry handle then allows attachment of biotin for selective enrichment via streptavidin, followed by mass-spectrometry analysis. C) A more general approach uses hydroxylamine (NH₂OH) to cleave thioester bonds, regenerating free thiols. These thiols are subsequently trapped either by HPDP-biotin with streptavidin enrichment or directly enriched on thiopropyl Sepharose resin, similar to RSNO workflows.

At excitatory synapses, S-palmitoylation of PSD-95 promotes its postsynaptic membrane targeting and clustering of AMPA and NMDA receptors, enhancing synaptic strength and plasticity [107,132]. Conversely, depalmitoylation by thioesterases, such as APT1, facilitates PSD-95 dispersal during long-term depression, highlighting a bidirectional switch that underlies learning and memory [133]. Similarly, palmitoylation of G protein-coupled receptors and voltage-gated ion channels modulates receptor desensitization and current density, respectively, thus broadening the functional repertoire of neuronal excitability [129,130]. Palmitoylation of gephyrin on Cys212 and Cys284 by the palmitoyl acyltransferase DHHC-12, promotes its postsynaptic membrane association and clustering to stabilize GABAA receptors at inhibitory synapses [134]. Inhibition or mutation of these palmitoylation sites reduces gephyrin cluster size and weakens inhibitory synaptic transmission, demonstrating that reversible palmitoylation of gephyrin dynamically regulates synaptic strength [134].

In neurodegenerative contexts, aberrant palmitoylation has emerged as a contributing factor. In Alzheimer’s disease, altered palmitoylation of amyloid precursor protein and its secretases shifts processing toward amyloidogenic pathways, exacerbating Aβ accumulation and synaptotoxicity [135]. Huntington’s disease models reveal that mutant huntingtin exhibits reduced palmitoylation by HIP14, impairing its targeting to synaptic membranes and triggering downstream transcriptional dysregulation [136]. In Parkinson’s disease, palmitoylation of the Parkinson’s disease–associated protein synaptotagmin-11 causes greater stability of synaptotagmin-11 and its interaction with specific portions of membranes. Palmitoylation of Syt11 disrupts α-synuclein homeostasis in neuron and may therefore promote Lewy body formation and dopaminergic neuron loss [137].

Collectively, these findings position S-palmitoylation as a critical regulator of neuronal homeostasis and as a nexus between synaptic physiology and pathology. Strategies aimed at restoring balanced palmitoylation cycles may thus hold therapeutic promise across a spectrum of neurodegenerative disorders.

Methodological approaches to Detect S-palmitoylation

Detecting S-palmitoylation presents challenges due to the thioester linkage and the hydrophobic nature of palmitate, but also because it is often impossible to distinguish it from other acylations such as mirisotylation or acetylation. Several biochemical and proteomic strategies have been developed (Fig. 9).

Metabolic Labeling with Clickable Fatty Acids: Incorporation of alkyne- or azide-modified palmitate analogs (e.g., 17-ODYA) into live cells enables subsequent copper-catalyzed azide–alkyne cycloaddition (CuAAC) to biotin or fluorophores (Fig. 9B) [128]. Labeled proteins can be enriched via streptavidin affinity and visualized by in-gel fluorescence or identified by MS. This approach offers high specificity and temporal resolution, but requires optimization of analogue concentration to minimize perturbation of endogenous lipid metabolism, but it is limited only to cell culture experiments.

Acyl-Biotin Exchange (ABE): ABE leverages the selective cleavage of thioester bonds by hydroxylamine (NH₂OH) to expose free thiols, which are then labeled with biotin-HPDP (Fig. 9C) [129]. After removal of hydroxylamine, biotinylated proteins are affinity-purified and analyzed by Western blot or MS. ABE is robust and amenable to complex lysates but may suffer from incomplete blocking of non-palmitoylated cysteines, leading to background biotinylation.

Acyl-Resin Assisted Capture (Acyl-RAC): Similar to ABE, Acyl-RAC uses hydroxylamine to cleave thioesters, followed by capture of the liberated thiols on thiol-reactive resin. Bound proteins are eluted and identified by MS (Fig. 9C). Acyl-RAC simplifies workflow by eliminating biotinylation steps and reducing nonspecific background, making it advantageous for high-throughput studies [138]. However, neither of the approaches can distinguish palmitoylation from other acylations.

Coupling any enrichment strategy with quantitative MS methods (SILAC, TMT, label-free) enables site-specific mapping and comparative analysis across conditions [139]. By integrating orthogonal techniques—metabolic labeling for live-cell dynamics, ABE/RAC for bulk proteomics, and MS for quantitation—researchers can achieve a comprehensive and accurate picture of the palmitoylome. Rigorous controls, such as samples where hydroxylamine is omitted and mutant cysteine constructs, are essential to confirm specificity.

PTM symphony or cacophony: how different modifications affect the same proteins

The cysteine thiols are the focal points of a complex regulatory mechanisms, where different PTMs compete for, or cooperate in, the control of protein function. This PTM crosstalk allows neurons to integrate diverse signaling inputs onto a single protein substrate. For example, the chemical state of the cysteine residue—whether it is a reduced thiol, S-nitrosylated, or persulfidated—dictates its availability for subsequent modifications like S-palmitoylation or sulfenylation. This interplay is fundamental for sculpting synaptic strength and governing protein quality control, and its dysregulation is a recurring theme in neurological disease.

A prime example of this biochemical antagonism is found at the synapse, in the dynamic regulation of scaffolding proteins by S-nitrosylation and S-palmitoylation. At excitatory synapses, the master scaffold PSD-95 requires palmitoylation of N-terminal cysteines for its trafficking and anchoring to the postsynaptic density, a process essential for glutamatergic synapse maturation (Fig. 10). However, nitric oxide (NO) signaling can lead to the S-nitrosylation of these same cysteines. Since both modifications target the same thiol, they are mutually exclusive. S-nitrosylation acts as a dominant-negative signal, physically blocking the protein acyltransferases (PATs) from palmitoylating PSD-95, thereby preventing its synaptic localization (Fig. 10) [107]. This mechanism can serve as a neuroprotective brake during excitotoxicity, uncoupling NMDA receptors from downstream effectors. A parallel story unfolds at inhibitory synapses with the scaffold protein gephyrin. Gephyrin’s clustering of GABA and glycine receptors is also dependent on its palmitoylation status. S-nitrosylation of gephyrin displaces it from the synapse, leading to a rapid reduction in inhibitory neurotransmission, providing a powerful mechanism for the activity-dependent remodeling of inhibitory circuits [140].

Fig. 10.

Differential Regulation of PSD-95 by S-Nitrosylation and Palmitoylation. Palmitoylated PSD-95 (yellow wavy lines) is stably associated with the postsynaptic dendritic membrane. NMDA receptor stimulation promotes depalmitoylation via a putative palmitoyl-protein thioesterase (PPT). Simultaneously, calcium influx through the NMDA receptor activates neuronal nitric oxide synthase (nNOS) by binding to calmodulin, generating NO. NO produced in close proximity to PSD-95 ​S-nitrosylates the protein, blocking its free cysteines and preventing re-palmitoylation by PSD-95 palmitoyl acyltransferases (PATs), thereby inhibiting its membrane association.

Another striking case study in redox cross-regulation is parkin. Under conditions of nitrosative stress, a hallmark of PD, catalytic cysteines in Parkin become S-nitrosylated (Fig. 11). This modification inhibits its E3 ligase activity, preventing it to tag proteins and damaged mitochondria for clearance via mitophagy and leading to the accumulation of damaged proteins [106]. Contrary, persulfidation of the same critical cysteines has been shown to enhance Parkin’s catalytic activity and protect it from pathogenic S-nitrosylation (Fig. 11) [89]. This places parkin at the center of a battle between two endogenous gases, NO and H₂S, where the balance determines its neuroprotective capacity. These examples reveal that cysteine PTM crosstalk is not a random occurrence but a precise and evolutionarily honed mechanism for integrating competing cellular signals, the disruption of which lies at the heart of neurological dysfunction.

Fig. 11.

Differential Regulation of Parkin Activity by RSSH and RSNO. Persulfidation increases Parkin’s E3 ligase activity, resulting in more efficient ubiquitination and degradation of damaged proteins. In contrast, in Parkinson’s disease—where elevated S-nitrosylation is observed—S-nitrosylation of Parkin inhibits its activity, leading to the accumulation and aggregation of damaged proteins.

In addition to activation/inactivation of enzyme’s activity, cysteine PTMs could have a general protective effect. Protein persulfidation is a good example. During oxidative stress, cysteine thiols are converted to sulfenic acids. If a sulfenic acid forms within a buried pocket, it can become stabilized and less accessible to reducing agents. H₂S, by contrast, is small enough to penetrate deep into protein cores and convert RSOH to a persulfide. These persulfides can then be reduced back to free thiols by the thioredoxin system [71,76]. When oxidative stress persists (such as in aging and many ROS-related diseases), sulfenylated cysteines oxidize further to RSO₂H and RSO₃H [141], oxidations that are generally considered irreversible (although some RSO₂H groups can be reduced back to thiols) [48]. Persulfidated residues are expected to act as better scavengers of ROS than regular thiols, resulting in the formation of RSSO₃H. The existence of the S–S bond makes this species a potential target for Trx and the restoration of native thiolate via this rescue loop (Fig. 12) [9,142]. This process might appear stochastic and therefore can be disregarded as a nonsignaling process; however, that presumption is inaccurate. The overall structure and half-life of thiol-containing proteins are preserved, which is one way of regulating protein function.

Fig. 12.

Persulfidation as a rescue loop that protects thiols form hyperoxidation. Trx: thioredoxin. ROS: reactive oxygen species. Sulfonylated cysteine residues are expected to lead to the loss of activity and increased aggregability.

Lessons learned and open questions from chemoproteomic cysteine PTM analysis

In depth, site-centric proteome analysis of individual cysteine PTMs has revolutionized our understanding of the role that these modifications have in health and diseases, but they have also been in discord with individual functional characterization that these chemical changes impose on identified targets. While they provide overview of how specific cysteine PTM changes under certain conditions, these studies do not always provide answers as to how those modifications affect protein structure and function. Frequent question in this field is: what percentage of a given cysteine site is occupied by a specific modification? One might argue that a relative five-fold change—from 1 ​% to 5 ​% occupancy—would have negligible biological impact. However, this assumption warrants re-examination.

Can we ever estimate site occupancy for any cysteine PTM? To calculate occupancy, one would need to generate a fully reduced “cysteinylome.” However, cysteine is a site of plethora of PTMs. For example, sulfinylation and sulfonlyation, mentioned in this review, are irreversible and cannot be reduced back. In addition to cysteine palmitoylation, there are reports of cysteine prenylation [143], and acetylation [144], all of which require harsh chemical cleavage with hydroxylamine. Furthermore, cysteines are targets of variety of naturally occurring electrophiles such as HNE [145], itaconate [146], fumarate [147] etc, forming thioether bond which is not easily cleavable with reducing agents. Recent report has suggested cyanylation as a new and widespread cysteine PTM [5], while Pappenheim et al., 2022 reported widespread occurrence of covalent lysine-cysteine redox switches [148]. None of those can be easily reduced to cysteine. Therefore, the relative change between the conditions remains the only reliable way of reporting cysteine PTMs.

In fact, even for the phosphoproteome, where almost full de-phosphorylation could be achieved, it is very difficult to calculate and estimate the occupancy of each S, T or Y [149,150]. Recent studies have challenged the importance of determining site occupancy in assigning the biological role of specific PTMs, revealing that the average site occupancy for ubiquitination is only 0.005 ​% [151], while acetylation occupies 0.1 ​% of lysine residues [152], despite both modifications being essential for cellular function.

Moreover, even if occupancy could be accurately measured, it would represent an average across the total protein pool. Proteins are often unevenly distributed among subcellular compartments, and a 1 ​% global occupancy could correspond to 50 ​% occupancy within a specific locale—dramatically altering local signaling dynamics.

A more complex question arising from proteome-wide cysteine PTM analysis is whether—and how—these modifications cross-talk. Although a few well-characterized examples exist, comprehensive, global comparisons are lacking. In a recent systematic meta-analysis, Li and colleagues collated all experimentally identified human RSNO and RSSH sites and assessed their sequence, structural, and functional contexts [153]. They report that, while both modifications converge on a broad spectrum of proteins to maintain redox homeostasis, they display distinct preferences and impacts. S-nitrosylation sites tend to reside in solvent-exposed regions and cluster within pathways involved in cellular damage repair and regulatory PTM crosstalk (e.g., phosphorylation, ubiquitination), whereas persulfidation more often modifies cysteines engaged in disulfide bonds and influences tissue-development and immune pathways. Importantly, when RSNO and RSSH occur at the same cysteine, they most commonly act synergistically to suppress protein function; at distinct sites, however, RSNO generally inhibits while RSSH activates target proteins. This global reciprocity underscores a finely balanced interplay between NO and H₂S signaling and provides a valuable framework for dissecting gasotransmitter-driven biological effects in health and disease [153]. Further comprehensive analyses of different cysteine PTMs performed at the same time could change our one-sided views to how these modifications control cellular function.

Conclusion

In conclusion, recent methodological advances have dramatically expanded our ability to profile cysteine post-translational modifications in the brain, yet significant challenges remain. Further improvements in sensitivity, specificity, and quantitative accuracy are essential to capture low-abundance and labile modifications in complex neural tissues. These attempts should be coupled to more detailed anatomical dissection of these modifications in different parts of the brain, enabling region- and cell-type–specific insights. Holistic integration of all cysteine PTMs—such as S-nitrosylation, persulfidation, sulfe/i/o/nylation, palmitoylation, and emerging modifications like cyanylation—will provide a unified map of cysteine reactivity and regulation. Furthermore, coupling high-throughput chemoproteomic workflows with computational modeling and machine-learning approaches can predict modification sites, dynamic cross-talk, and structural impacts, guiding targeted functional experiments. Systematic incorporation of functional analyses—ranging from enzymatic assays to live-cell imaging—will validate the biological significance of identified sites and elucidate mechanistic underpinnings. Finally, exploring how cysteine PTMs influence biomolecular phase separation and condensate formation may uncover novel mechanisms of synaptic organization, stress granule dynamics, and neurodegenerative pathology. Advancing analytical workflows, integrating multi-dimensional data, and linking chemical modifications to biological outcomes will pave the way for innovative therapeutic strategies targeting cysteine chemistry in neurological disease.

Author contribution

Prof Milos Filipovic wrote this manuscript.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work the authors used ChatGPT in order to improve language and correct grammar. After using this tool/service, the author reviewed and edited the content as needed and takes full responsibility for the content of the publication.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Milos Filipovic reports financial support was provided by University of Glasgow. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.