Recent Advances in Chemical Proteomics for Protein Profiling and Targeted Degradation

Abstract

Chemical proteomics has emerged as a powerful approach to decipher protein function, interactions, and targeted degradation pathways in complex biological systems. Recent advances in chemical labeling strategies, including activity-based protein profiling (ABPP), proximity labeling (PL), and proteolysis-targeting chimeras (PROTACs), have facilitated a deeper understanding of protein function and interaction networks. First, ABPP employs covalent probes to selectively label active enzymes, uncovering functional proteomics and drug-target interactions. Innovations such as PhosID-ABPP and streamlined cysteine ABPP have improved site-specific quantification and throughput, enabling proteome-wide analysis of enzyme activity and small-molecule interactions. Second, PL enables the characterization of transient protein–protein interactions using enzymatic or chemically triggered approaches. Advances including TurboID and TransitID enhanced the spatiotemporal resolution of PL. Third, PROTACs expand the scope of targeted protein degradation by leveraging the ubiquitin–proteasome system. Collectively, we highlight recent advancements in integrating mass spectrometry (MS) with these methodologies in the field of chemical proteomics.

Introduction

Mass spectrometry (MS) has emerged as an indispensable tool for identifying and quantifying proteins in complex biological systems [1,2]. Chemical approaches enable selective labeling and enrichment of target proteins, improving the detection of low-abundance species and enhancing the specificity of proteomic analyses [3–5]. Chemical proteomics employs diverse reaction mechanisms for selective protein labeling: (1) Covalent activity-based reactions target specific amino acid residues, such as cysteine and serine, providing insights into enzyme activity and substrate specificity [6,7]; (2) Bioorthogonal reactions, including copper-catalyzed azide-alkyne cycloaddition (CuAAC) and strain-promoted azide-alkyne cycloaddition (SPAAC), allow protein labeling without interfering with native biological processes [8]; (3) Photoactivation reactions utilize light-activated probes, such as aryl azides and diazirines, to capture dynamic protein interactions [9,10]; and (4) Proximity-dependent biotinylation, such as BioID and TurboID, employs biotin ligases to label and identify the molecular neighborhood of a protein of interest [11,12]. These strategies have significantly advanced the study of protein function and dynamics in biological systems. This review discusses recent applications in integrating MS with these methodologies, including activity-based protein profiling (ABPP), proximity labeling (PL), and proteolysis-targeting chimeras (PROTACs) (Figure 1).

Figure 1. Overview of labeling strategies in chemical proteomics.

(a) Activity-Based Protein Profiling (ABPP). ABPP utilizes activity-based probes containing a reactive group, a linker, and a tag (e.g., biotin or fluorophore) to selectively label active enzymes. Labeled proteins are subsequently enriched and identified, facilitating studies on drug–target interactions and target visualization. (b) Proximity Labeling (PL). PL employs an engineered enzyme, such as BirA* (a mutant biotin ligase), fused to a bait protein or protein of interest (POI). Upon labeling, nearby proteins become biotinylated and are subsequently enriched for identification, enabling high spatial resolution and intracellular mapping of protein–protein interactions. (c) Proteolysis-Targeting Chimeras (PROTACs). PROTACs are heterobifunctional molecules that induce targeted protein degradation. They contain an E3 ligase ligand, a POI ligand, and a linker, facilitating the formation of a ternary complex with the POI and an E3 ligase. This interaction triggers ubiquitination and subsequent proteasomal degradation of the POI, enabling targeted degradation and overcoming drug resistance.

Affinity-Based Protein Profiling

Activity-based protein profiling (ABPP) has evolved from targeting specific enzyme families into a broader platform for mapping drug and electrophilic compound interactions in biological systems [13–16]. Unlike conventional proteomics, ABPP measures protein functionality using activity-based probes (ABPs), which consist of a reactive warhead, a linker, and a reporter group. Under near-physiological conditions, ABPP selectively identifies active proteins in specific experimental or pathological conditions (Figure 1a) [17]. Recent advances in ABPP have improved enrichment and quantification workflows, enabling site-specific and multiplexed proteome-wide analyses. Kleinpenning et al. developed the PhosID approach, which utilizes phosphonate-based capture with immobilized metal affinity chromatography (IMAC) to selectively enrich probe-modified peptides while preserving site information [18]. This enrichment strategy overcomes problems with harsh elution, high non-specific background, and incomplete localization common to traditional biotin pull-downs. Later, Bergen et al. applied PhosID-ABPP to lysates and intact cells, achieving reliable residue-level probe identification [19]. Recently, this workflow was further optimized to illuminate on-target and off-target interactions in kinases, revealing intricate covalent engagements [20]. Collectively, these studies show that phosphonate tagging enables superior site localization compared with biotin-based approaches. However, due to its reliance on metal coordination, strongly acidic, non-phosphorylated peptides can over-bind [21], leading to false positives. Therefore, IMAC conditions must be carefully optimized.

As enrichment strategies have undergone further development, quantitative ABPP methods have evolved to support higher throughput, greater multiplexing, and reduced sample requirements [22]. For instance, isotopic tandem orthogonal proteolysis (isoTOP) ABPP enables single-residue cysteine quantification using isotopically encoded, cleavable tags. However, it requires mass shifts greater than 4 Da to prevent isotopic overlap, which limits multiplexing and necessitates custom probe synthesis [23–25]. Similarly, stable isotope labeling by amino acids in cell culture (SILAC) ABPP relies on metabolic labeling with heavy amino acids, restricting its application to cultured cells (Figure 2a) [26]. Although these precursor-ion-based approaches have improved throughput, they are generally limited to duplex or triplex quantification. To address these limitations, product-ion-based quantification using tandem mass tags (TMT/TMTpro) enables the simultaneous analysis of up to 35 samples in a single LC–MS/MS run, significantly boosting throughput while reducing the sample requirement (Figure 2b) [27,28]. SLC-ABPP (streamlined cysteine ABPP) further optimized this process by incorporating an iodoacetamide-based probe for labeling and applying post-proteolysis TMT labeling [29]. Recently, a plate-based single-pot solid-phase-enhanced sample preparation (SP3)-TMT workflow coupled with high-field asymmetric ion mobility spectrometry (FAIMS) enabled the quantification of 18,000–24,000 reactive cysteines from just 10–20 μg of proteome (Figure 2c) [30].

Figure 2. Overview of quantitative ABPP strategies.

(a) Precursor ion-based quantitative ABPP. SILAC-ABPP utilizes metabolic labeling with heavy and light amino acids, and isoTOP-ABPP employs isotopically encoded probes (>4 Da shift) detectable in MS1 spectra. Quantification relies on extracted ion chromatograms (EIC). (b) Product ion-based quantitative ABPP. Isobaric tags (e.g., TMTpro) enable up to 35-plex multiplexing by introducing a mass reporter and balancer that fragment to release reporter ions in MS/MS scans. This approach enables higher throughput and reduces isotopic overlap compared to precursor ion-based methods. (c) SLC-TMT workflow (using 10 samples as an example). Streamlined cysteine ABPP (SLC-ABPP) labels cysteines at the peptide level following enrichment and digestion, then applies TMT labeling to the resulting peptides. This allows high-throughput screening with advanced MS instrumentation and reduces the amount of sample needed per experiment. (d) sCIP-TMT workflow. Silane-based Cleavable Isotopically labeled Proteomics with TMT (sCIP-TMT) embeds the isobaric label before proteolysis at the protein level, allowing samples to be pooled earlier. This reduces sample-to-sample variability, simplifies downstream sample handling, and preserves the high multiplexing capability for efficiently identifying and quantifying reactive cysteines.

Despite these advances, peptide-level tagging can still suffer from variability arising from the enrichment and enzymatic digestion process. To address this shortcoming, protein-level labeling workflows have been developed to allow earlier incorporation of mass tags (Figure 2d). For instance, sCIP (silane-based cleavable isotopically labeled proteomics) employs a dialkoxydiphenylsilane (DADPS) acid-cleavable linker, which seamlessly incorporates stable isotopes into capture reagents that react with target cysteines at the protein stage [31]. By allowing sample pooling immediately after labeling, sCIP improved quantitative precision over the standard peptide-level workflow and simplifies sample preparation. More recently, sCIP-TMT merges the DADPS-based platform with readily available TMT reagents, alleviating the need for extensive custom syntheses while retaining high multiplex capability [32]. Also, advanced MS techniques, including FAIMS and real-time search-MS3, were applied to effectively filter out interfering ions and mitigate ratio compression effects in quantification [29,30]. Taken together, these methods have enabled ABPP to profile enzyme activities, drug–target engagements, and off-target liabilities at the proteome scale with unprecedented throughput and accuracy [14].

Proximity Labeling

Proximity labeling (PL) has emerged as a powerful approach for characterizing protein–protein interactions by covalently tagging proteins near a bait molecule in living systems [33–35]. Unlike traditional affinity purification methods that often lose weak transient interactions during wash steps, PL captures these fleeting contacts under near-physiological conditions. Classic PL methods like BioID use a mutant biotin ligase (BirA*) fused to a bait protein, which generates biotinoyl-5’-AMP intermediates that react with lysine residues on nearby protein (Figure 1b) [11]. However, the large size (~35 kDa) of BirA* and its long labeling time (16–18 hours) limit its effectiveness in capturing transient and rapidly fluctuating interactions [12,36]. To address these challenges, engineered variants such as TurboID and miniTurbo offer faster labeling kinetics and reduced steric interference [12,36–39]. TurboID enables rapid tagging within minutes, and miniTurbo, due to its smaller size, minimizes disruption of the bait protein’s function. Beyond biotin ligase-based systems, peroxidase-based methods like APEX and HRP enable rapid covalent tagging of nearby proteins by generating short-lived radicals from biotin-phenol or biotin-tyramide in the presence of hydrogen peroxide. APEX, active in reducing environments, is widely used for intracellular labeling, while HRP, less active in the cytosol, is mainly applied to the cell surface or fixed samples. [12,36]

Building on these advances, Lee et al. developed LOV-Turbo by integrating the light-sensitive LOV domain into the proximity labeling enzyme TurboID, enabling rapid and reversible control of its labeling activity using low-power blue light [40]. In the absence of light, the LOV domain clamps TurboID’s active site, suppressing background labeling. Upon exposure to low-intensity blue light, a rapid conformational change reactivates the enzyme, allowing precise pulse-chase labeling simply by toggling the light. This reversible control makes LOV-Turbo particularly suitable for capturing protein interactions that occur over short timescales or in biotin-rich environments, such as neural systems.

Beyond enzyme engineering, innovative PL strategies have expanded to track protein movement and trafficking. Qin et al. developed TransitID, which employs tandem labeling with two orthogonal enzymes TurboID and engineered ascorbate peroxidase (APEX) to map proteome dynamics in living cells (Figure 3a) [41]. This dual-labeling method distinguishes proteins that traffic between compartments, such as from the cytosol to mitochondria or from the nucleolus to stress granules. For instance, TransitID revealed that stress granules play a protective role by preventing aggregation of the transcription factor JUN during oxidative stress. Furthermore, TransitID was applied to map intercellular protein trafficking, exemplified by its application in profiling communication between tumor cells and macrophages.

Figure 3. Novel proximity labeling (PL) strategies for capturing dynamic proteomes in living systems.

(a) TransitID: dynamic mapping of proteome trafficking. This dual-enzyme method leverages TurboID (for biotin tagging in a “source” location) and APEX2 (for alkyne-phenol labeling in a “destination” location), enabling spatiotemporal mapping of protein movement within and between cells under near-physiological conditions. Adapted from Ref. [41] with permissions. (b) CAT-S: proximity labeling in primary living samples. In this approach, a targeting photocatalyst and caged thio-quinone methide (thioQM) probe undergo blue-light–driven decaging, releasing a reactive thioQM warhead that tags nearby proteins with biotin in non-genetically modified, difficult-to-transfect samples, including primary mouse tissues and human T cells. Adapted from Ref. [42] with permissions. (c) MultiMap: multiscale PL with tunable resolution. A single photocatalyst (Eosin Y) activates three types of photo-probes (phenol, arylazide, and diazirine), each generating a different labeling radius—from high resolution (~100 Å) to low resolution (~1000 Å). This platform reveals local membrane interactomes on non-engineered live cells or across intercellular synapses, providing multiscale insights into protein–protein networks when paired with biochemical validation and AI-assisted structural analysis. Adapted from Ref. [44] with permissions.

Emerging non-genetic chemical approaches now extend proximity labeling to primary samples that are difficult to manipulate genetically. Liu et al. introduced CAT-S, which enables light-triggered activation of a thio-quinone methide (thioQM) warhead for spatially restricted labeling in hard-to-transfect cells, including native mouse tissues and human T cells (Figure 3b) [42]. Similarly, Takato et al. developed PhoxID, which uses singlet oxygen-mediated photooxidation for rapid PL in genetically unmodified mouse brains, offering unprecedented temporal resolution for receptor-proximal protein mapping [43]. Furthermore, the MultiMap platform enhances PL resolution by employing a single photocatalyst (Eosin Y) to activate different photo probes (diazirine, aryl azide, and phenol derivatives) (Figure 3c) [44]. This allows fine-tuning of labeling radii from ~100 Å for local interactomes to >1000 Å for broader interaction networks. The method’s adjustable resolution allows users to interrogate both local receptor microenvironments and broader intercellular contacts. Deployed in situ on living specimens, MultiMap can capture endogenous plasma-membrane interactomes—even across cell–cell contact zones such as immunological synapses—thereby providing a dynamic blueprint of the cell-surface microenvironment. Moreover, coupling MultiMap data with computational tools like AlphaFold-Multimer provides structural insights into newly identified protein–protein interactions, enriching both proteomic and structural analyses of complex networks. Collectively, these chemistries provide spatial control without genetic manipulation, but they require precise light delivery and thorough evaluation of phototoxicity.

In sum, proximity labeling now enables researchers to genetically encode labeling enzymes (e.g., TurboID, LOV-Turbo, TransitID) or introduce photoactivatable warheads (e.g., CAT-S, PhoxID, MultiMap) for real-time spatial proteome profiling. Taken together, these innovations have cemented proximity labeling as a critical technique in modern chemical proteomics, uncovering dynamic cellular behaviors inaccessible to conventional immunoprecipitation assays.

Proteolysis-targeting chimeras

Mass spectrometry-based proteomics plays a critical role in evaluating PROTAC-mediated degradation, assessing degradation efficiency, and identifying off-target effects. In this section, we discuss the concept of PROTACs and highlight recent chemical proteomics strategies. PROTACs are heterobifunctional molecules that harness the cell’s natural degradation machinery to eliminate disease-causing proteins [45,46]. A PROTAC consists of three key components: a ligand that binds to the protein of interest (POI), a ligand that recruits an E3 ubiquitin ligase, and a linker that connects them (Figure 1c). By simultaneously binding to the POI and the E3 ligase, PROTACs facilitate the formation of a ternary complex, inducing proximity-driven ubiquitination of the POI and marking it for proteasomal degradation [47–49]. Unlike traditional inhibitors that merely block catalytic sites, PROTACs selectively degrade target proteins through the ubiquitin–proteasome pathway. As a result, PROTACs can target undruggable proteins, including those that lack well-defined active sites for conventional inhibitors [50]. While PROTACs are highly effective at inducing protein degradation and hold significant therapeutic potential, they often suffer from limited water solubility, poor tissue permeability, and suboptimal pharmacokinetics due to their molecular structure [51,52]. Additionally, challenges such as achieving high specificity, ensuring efficient cellular uptake, and minimizing off-target effects impact delivery strategies and metabolic stability. Future advances in modular and stimuli-responsive PROTAC designs, such as conditionally activatable or tumor-microenvironment-specific PROTACs, could further enhance their clinical applicability [51,52]. Several emerging strategies continue to broaden the landscape of targeted protein manipulation. Autophagy-targeting chimeras (AUTACs) and lysosome-targeting chimeras (LYTACs) redirect proteins to lysosomes via autophagy or receptor-mediated endocytosis, enabling the degradation of cytosolic and membrane proteins that are difficult to target [47]. Additionally, post-translational modification regulators, including PhosTACs and DUBTACs, manipulate protein stability by recruiting phosphatases or deubiquitinases, respectively [53,54]. These approaches expand the chemical biology toolkit beyond the conventional ubiquitin–proteasome pathway, opening new avenues for selective degradation or stabilization of challenging protein targets. In recent years, several innovative platforms—such as antibody-PROTACs, aptamer-PROTACs, and nano-PROTACs—have been developed to address the delivery and pharmacokinetic challenges associated with traditional PROTACs.

Antibody-PROTACs (AbTACs) leverage the high specificity of monoclonal antibodies to direct PROTACs to target cells (Figure 4a) [55,56]. Conjugation to antibodies enables selective recognition of cell surface antigens, followed by receptor-mediated endocytosis and intracellular release of the PROTAC molecule into the cytoplasm. Once inside, the PROTAC binds to the POI and an E3 ubiquitin ligase, facilitating targeted protein degradation via ubiquitin–proteasome system. Recently, Wang et al. developed a ROR1-targeting AbTAC by linking a BRD4-degrading PROTAC to a ROR1 antibody, which demonstrated significant antitumor activity both in vitro and in vivo [57]. Moreover, Cotton et al. developed bispecific AbTAC (Figure 4b) that can simultaneously recruit a membrane-bound E3 ligase and a cell surface POI to induce lysosomal degradation, effectively targeting membrane proteins such as PD-L1 and EGFR [58]. Similarly, aptamer-PROTACs utilize single-stranded DNA or RNA aptamers to improve tumor selectivity and solubility (Figure 4c) [59,60]. Aptamers exhibit low immunogenicity, ease of synthesis, rapid tissue penetration, and high stability, binding targets with high affinity through unique three-dimensional structures. Recently, Liu et al. developed a new archetype of PROTAC (PS-ApTCs), a novel PROTAC platform that utilizes phosphorothioate-modified aptamers for tumor-targeted and spatial-selective degradation of nucleocytoplasmic shuttling proteins, effectively disrupting nucleolin trafficking and enhancing antiproliferative effects [61]. Furthermore, combining PS-ApTCs-mediated nucleolin degradation with aptamer–drug conjugate-based chemotherapy achieves a synergistic tumor inhibition effect.

Figure 4. Overview of engineered PROTAC delivery strategies.

(a) Antibody-PROTACs (AbTACs). Antibody-PROTAC conjugates selectively bind to antigens on target cells, facilitating endocytosis and PROTAC release. The released PROTAC recruits an E3 ligase, leading to POI ubiquitination and degradation via the ubiquitin–proteasome system (b) Bispecific antibody PROTACs. The bispecific antibody recognizes both the POI and a membrane-bound E3 ligase, inducing POI ubiquitination. The complex undergoes endocytosis, resulting in POI degradation through the lysosomal pathway. (c) Aptamer-based PROTACs. Aptamer-PROTAC conjugates use aptamers for POI recognition and E3 ligase recruitment, enabling ubiquitination and subsequent POI degradation via the ubiquitin–proteasome system. (d) Nano-PROTACs. PROTACs are encapsulated in pH- and cathepsin B-responsive nanoparticles (PSRNs). Upon exposure to acidic conditions and cathepsin B in the tumor microenvironment, the PROTACs are released, facilitating POI degradation.

In addition to these targeted approaches, another common design strategy is the incorporation of nanomaterials into delivery systems. Nano-PROTACs have emerged as a promising approach by conjugating PROTACs with nanomaterials to overcome conventional limitations by improving water solubility, tissue permeability, and metabolic stability [62,63]. Conjugating PROTAC molecules with nanomaterials allows for the utilization of nanoscale delivery systems to enhance the therapeutic efficacy of targeted protein degradation [64]. These systems provide precise spatiotemporal control over drug release, enabling the on-demand activation of PROTACs in target tissues while minimizing off-target effects. For example, Yang et al. developed a type of polymer PROTAC conjugated and pH/cathepsin B sequential responsive nanoparticles (PSRNs) for colorectal cancer therapy (Figure 4d) [64]. These nanoparticles utilized a pH-sensitive design and a cathepsin B-cleavable linker to achieve targeted delivery and precise intracellular release of PROTACs, effectively degrading cyclin-dependent kinase 4 and 6 (CDK4/6). This approach enhanced tumor targeting, improved immune checkpoint blockade efficacy, and demonstrated significant anti-tumor outcomes in colorectal cancer models.

Conclusion

Innovative chemical proteomics approaches continue to transform our ability to decode protein function and interactions in complex biological systems. Recent methodological advances, spanning activity-based protein profiling, proximity labeling, and proteolysis-targeting chimeras (PROTACs), have expanded the toolkit for comprehensive protein analysis. High-throughput enrichment strategies, isobaric labeling, and next-generation instrumentation facilitate deeper and proteome-wide investigations. For example, refined protein-level tagging and enrichment techniques enhance enzyme activity profiling, while light-controlled and chemically triggered proximity labeling methods offer improved spatial and temporal resolution for mapping dynamic protein interactions. In parallel, emerging PROTAC-based approaches, including antibody-, aptamer-, and nano-PROTACs, open new avenues for targeted protein degradation with greater therapeutic precision and efficacy. Moreover, the integration of computational tools, such as machine learning-based structural prediction, with advanced analytical techniques like real-time search-MS3, further optimizes data accuracy and throughput.

We believe that the future of chemical proteomics lies in the integrated application of ABPP, PL, and PROTACs, each offering distinct biological insights. ABPP provides direct information on enzyme activity and functional proteoforms, enabling the identification of catalytically active proteins under physiological or pathological conditions. PL enables spatially resolved interactome mapping, capturing transient and low-affinity protein–protein interactions within defined subcellular microenvironments. PROTACs, in turn, enable the selective degradation of target proteins, supporting perturbation-based studies to uncover causal relationships within signaling networks. Together, they constitute a comprehensive toolkit for dissecting proteome dynamics, regulatory mechanisms, and therapeutic vulnerabilities with unprecedented resolution and information. We anticipate that the convergence of these strategies will not only deepen mechanistic insights into proteome dynamics but also accelerate the discovery of novel druggable targets.