Functionalizing tandem mass tags for streamlining click-based quantitative chemoproteomics

The study addresses the need for scalable, high-throughput, and high-coverage methods to map ligandable or druggable residues across the proteome using covalent chemoproteomics. Traditional workflows rely on late-stage isobaric labeling after proteolysis, which increases sample-to-sample variance and prolongs processing. Advances in covalent labeling chemistries and sample preparation have expanded targetable residue types, but cysteine remains a favored target due to its functional roles, available chemistries, and clinical success of cysteine-targeting drugs. The purpose of this work is to introduce sCIP-TMT, an isobaric labeling platform that enables early sample pooling via a fully functionalized clickable enrichment handle, thereby streamlining workflow, increasing throughput, and maintaining accurate quantification for chemoproteomic screening.

The paper reviews developments in covalent labeling chemistries that map diverse nucleophilic residues (serine, lysine, tyrosine, methionine, aspartate/glutamate, arginine, cysteine) and platforms analyzing reversible binders. It highlights improvements in sample prep (e.g., SP3) and software (pLink, MSFragger, SAGE) that reduce processing time, as well as high-throughput plate-based workflows. Isobaric labeling (iTRAQ, TMT, DiLeu) boosts acquisition speed and multiplexing, commonly applied after proteolysis in workflows like SLC-ABPP. Hyperplexing strategies and thermal proteome profiling are cited. However, late labeling increases variance and time. Prior early-labeling approaches include sCIP (6-plex) with built-in isobaric label and azidoTMT (11-plex) using anti-TMT antibody enrichment; limitations include smaller multiplexing for sCIP and variable performance or lack of TMTpro-compatible antibodies for azidoTMT, motivating robust enrichment-compatible reagents that allow early pooling.

Reagent design and synthesis: A solid-phase peptide synthesis route produced a minimalist sCIP capture reagent with a free N-terminus (sCIP-Gly-NH2) containing biotin, a DADPS cleavable linker, and an azide. Glycine was included as a spacer to facilitate high-yield conjugation with isobaric reagents. The capture reagent was obtained in 53% yield and high purity. Late-stage functionalization with commercially available TMT activated ester reagents was performed in situ by mixing sCIP-Gly-NH2 with TMT (e.g., TMTzero or TMT10plex) at 1:1 stoichiometry for 1 h at ambient temperature, followed by quenching with 0.5 equivalents hydroxylamine for 15 min. LC-MS verified >99% conversion to sCIP-TMT.

Chemoproteomic workflow (sCIP-TMT): Cell lysates were labeled with iodoacetamide-alkyne (IAA, 500 µM, 1 h), then subjected to CuAAC with pre-formed sCIP-TMT conjugates, enabling early pooling immediately after click. Samples underwent SP3 cleanup, reduction/alkylation, trypsin digestion, streptavidin enrichment, and mild acid cleavage at DADPS to elute peptides for LC-MS/MS.

Acquisition and analysis: Data were acquired on Orbitrap Eclipse with FAIMS-MS2 or SPS-MS3 where specified, using 70–180 min gradients on in-house packed C18 columns. NCE optimization was performed (HCD 25–45%). Data were processed with MSFragger/FragPipe, including labile search and diagnostic ion mining. TMT integrator settings tagged intact sCIP-TMT cysteine modification mass. Performance was benchmarked against an IA-DTB cysteine capture workflow followed by peptide-level TMT labeling. Spike-in ratio mixtures (1:1 and 1:5:10:15) assessed quantitative accuracy under FAIMS-MS2 and SPS-MS3.

Screening application: A focused panel of four electrophilic fragments (KB02, KB10, methylphenyl propiolate, methyl cinnamate) was screened in HEK293T lysates (500 µM, duplicates) using the sCIP-TMT10 workflow. Liganded cysteines were defined by log2 ratios >1 relative to vehicle.

Extension to N-terminomics: Predigested lysates were labeled with ethynyl-2PCA to modify peptide N-termini, then clicked to six different sCIP-TMT channels, enriched, and mixed in defined ratios for LC-MS/MS to assess compatibility with N-terminal peptide profiling.

Detailed step-by-step methods for reagent preparation, CuAAC, SP3, enrichment, LC-MS gradients, instrument settings, database searching, and validation are provided in the Methods and Supplementary sections.

• Efficient in situ conjugation: sCIP-Gly-NH2 reacts with TMT reagents to give >99% conversion to sCIP-TMT at 1:1 stoichiometry.
• Coverage with sCIP-TMTzero: 3,856 proteins, 11,219 peptides, and 8,543 unique cysteines identified. Modified peptides showed higher charge states due to the +633.3957 Da modification and protonatable TMT piperidine.
• Diagnostic ions: TMTzero reporter ion (m/z 126.1277) was detected in nearly 100% of modified spectra with >80% average intensity; a cysteine desulfurization diagnostic ion (m/z ~668.3896) was found in >97% of modified PSMs with ~70% mean intensity.
• Fragmentation optimization: HCD NCE ~35% maximized reporter intensity and identification coverage; the desulfurization ion dominated at lower NCEs, while TMT reporter dominated above 30% NCE.
• TMT10plex performance: Early pooling with sCIP-TMT10 yielded high coverage on a 3 h gradient, comparable to bulk TMT cysteine-peptide labeling studies without extensive fractionation.
• Throughput gains: Estimated ~6 h reduction in active sample preparation time and 86 fewer containers versus established 18-plex TMTpro workflows, while using comparable amounts of TMT reagent.
• Quantitative accuracy: Spike-in tests showed measured reporter ratios closely matched expected values for equimolar and 1:5:10:15 mixes under FAIMS-MS2; SPS-MS3 produced tighter ratio distributions at the cost of lower coverage.
• Reproducibility versus IA-DTB/TMT: Median CVs for sCIP-TMT were 5.2% (1:1) and 7.8% (1:4) versus 9.6% and 9.7% for peptide-level TMT labeling with IA-DTB. Capture efficiency was similar (>80% biotin/desthiobiotin-modified peptides captured); IA-DTB samples showed >99% TMT modification including non-desthiobiotin peptides likely due to non-specific enrichment.
• Fragment screen outcomes: Across the four-compound panel, 10,733 cysteine peptides corresponding to 8,515 unique cysteines and 3,787 proteins were identified; 789 high-confidence liganded cysteines (log2 > 1) were detected. Reactivity fractions: KB02 liganded 11.4% of total cysteines; KB10 3.6%; MPP 3.8%; MC showed very attenuated reactivity in prior analyses (~0.2%). Over 85% of MPP-engaged cysteines overlapped with KB02/KB10 targets; 493 cysteines were uniquely liganded by a single compound. Examples: GSTO1 Cys32 strongly labeled by KB02/KB10; PFAS Cys270 and LIG3 Cys929 unique to KB10; CAPN2 Cys301 and TLE1 Cys526 unique to MPP.
• Concordance with prior datasets: sCIP-TMT FAIMS-MS2 ratios for KB02 correlated with prior MS1-based data (r^2 = 0.63) with expected ratio compression; results were consistent with CysDB aggregated ligandability.
• Novel sites: 760/789 liganded cysteines overlapped with CysDB; 29 sites were uniquely identified as liganded by sCIP-TMT, including MEIOC Cys342 and Akirin-2 Cys3. Diagnostic desulfurization ion supported confident cysteine PSM assignment.
• N-terminomics compatibility: >700 unique peptide N-termini were identified from ethynyl-2PCA-labeled samples, and TMT reporter intensities reflected expected mixing ratios, demonstrating applicability beyond cysteine chemoproteomics.

The sCIP-TMT platform addresses key bottlenecks in isobaric chemoproteomics by moving labeling upstream of proteolysis and enabling early pooling at the protein level after click conjugation. This streamlines workflows, reduces handling and active preparation time, and decreases inter-sample variance while maintaining high coverage and quantitative fidelity. Unlike azidoTMT/iadoTMT approaches that require anti-TMT antibody enrichment, sCIP-TMT uses established avidin-based enrichment via biotin and leverages the DADPS linker for mild acid elution and broad proteome coverage. The platform is compatible with standard TMT data acquisition and analysis pipelines. Benchmarking showed accurate ratio measurements in multiplexed spike-ins, improved CVs versus peptide-level TMT workflows, and strong overlap with prior MS1-based ligandability datasets despite some MS2 ratio compression. The frequently observed desulfurization diagnostic ion in sCIP-TMT spectra adds confidence for identifying cysteine-modified peptides and may be useful for filtering or validating novel identifications. Screening a small panel of electrophiles demonstrated that sCIP-TMT effectively captures proteome-wide reactivity differences and SAR, validating known targets (e.g., GSTO1 Cys32) and revealing unique liganded sites and potential druggable residues (e.g., CRKL Cys249), with orthogonal gel-based ABPP corroboration. The method’s extension to N-terminomics further indicates its versatility for probes that block amine-based labeling. Overall, sCIP-TMT provides a practical, scalable, and accurate platform for multiplexed chemoproteomics, readily integrable with existing enrichment and data analysis practices and adaptable to broader residue targets and higher multiplexing levels.

The study introduces sCIP-TMT, a click-compatible, biotin- and DADPS-containing isobaric platform that enables early sample pooling for quantitative chemoproteomics. A 10-plex sCIP-TMT set was synthesized in situ with high conversion and applied to cysteine profiling, achieving high proteome coverage, accurate quantification, reduced sample preparation time, and lower variance compared to conventional peptide-level TMT workflows. Application to electrophile screening identified 789 liganded cysteines, recapitulated known targets, provided SAR across distinct chemotypes, and uncovered previously unreported ligandable sites. The consistent observation of a cysteine desulfurization diagnostic ion supports confident identification of modified peptides. The approach extends to N-terminal proteomics with ethynyl-2PCA labeling. Future directions include scaling to TMTpro 18-plex and hyperplexing via isotopic variants, deeper coverage via offline fractionation, increased compatibility with additional residue-selective probes, and potential integration of diagnostic ions for improved identification stringency.

Reported limitations include the modest synthetic yield (53%) of the sCIP-Gly-NH2 reagent, which may be improved by optimizing resin loading. Early sample pooling may be less amenable to some established automation workflows, though plate-based adaptations could mitigate this. MS2-based TMT quantification exhibits some ratio compression compared to SPS-MS3, while SPS-MS3 reduces coverage. Resampling similar cell line models may yield modest gains in novel cysteine coverage. The current work focused on TMT10; broader validation with TMTpro and large-scale screens remains to be demonstrated.