Residue-Level Distance Restraints for Cryo-EM Validation, Integrative Modeling & PPI Interface Mapping — Without Crystallization
XL-MS services provide residue-pair distance restraints for protein structure validation and PPI interface mapping — directly from purified complexes or living cells. The output is a set of experimental distance constraints that validate cryo-EM models, refine AlphaFold predictions, and map interaction interfaces at amino acid resolution — without crystallization.
XL-MS occupies a unique position between HDX-MS and cryo-EM: HDX-MS identifies protected interface regions; XL-MS provides residue-pair distance constraints within those regions. Together, they deliver the most complete experimental picture of a protein complex interface available without crystallization — HDX defining the footprint, XL-MS providing the crosslinks that constrain integrative models at amino acid resolution.
Core Capabilities:
XL-MS uses bifunctional chemical crosslinkers to covalently connect two amino acid residues (typically lysines) that are within the crosslinker's spacer distance in the native 3D structure. The crosslinked protein is digested into peptides; crosslinked peptide pairs — characterized by their combined mass and fragmentation pattern — are identified by LC-MS/MS. Because the crosslinker spacer arm imposes a defined maximum distance (e.g., DSS/BS3: ~11.4 Å extended, corresponding to ~26–30 Å between Cα atoms of the linked lysines, validated by MD simulation benchmarks), each identified crosslink provides a quantitative distance restraint that constrains the protein's fold or complex architecture.
Unlike cryo-EM — which requires purified complexes at high concentration and often struggles with flexible regions — XL-MS works on purified proteins, crude cell lysates, or intact cells using membrane-permeable crosslinkers.
If your structural biology or protein interaction research requires experimental distance restraints — to validate models, resolve ambiguous density, or confirm computational predictions — XL-MS provides the data.
No Crystallization. No Size Limit. No Freezing.
XL-MS works on proteins and complexes in solution — at native pH, native salt, and physiological temperature. Membrane proteins, flexible multi-domain assemblies, and complexes that resist crystallization are all accessible. Crosslinks are applied directly to your sample in amine-free buffer; no grids, no vitrification, no crystal trays.
Captures Transient Interactions and Flexible Regions That Cryo-EM Misses
Cryo-EM density fades where proteins are flexible. XL-MS crosslinks those regions — providing distance restraints for loops, disordered tails, and dynamic interfaces that are invisible in density maps. In-cell crosslinking goes further: interactions are captured in their native environment before lysis, eliminating the post-lysis reassortment that confounds AP-MS.
Quantitative Distance Restraints — Not Subjective Density Interpretation
Each identified crosslink is a number: two residues linked by DSS were ≤30 Å apart in solution. Applied as HADDOCK or Rosetta distance bounds, these restraints constrain your model with experimental data — not visual judgment of cryo-EM density or AlphaFold confidence scores. A typical purified complex yields 50–200 inter-protein restraints, enough to resolve subunit arrangement and interface architecture.
Five XL-MS capabilities spanning standard DSS/BS3 crosslinking through MS-cleavable chemistry, in-cell applications, integrative modeling, and custom crosslinker development.
XL-MS provides distance restraints that complement — rather than replace — cryo-EM, HDX-MS, and computational methods. Choose based on what type of structural information your project needs.
| Feature | XL-MS | HDX-MS | Cryo-EM | AlphaFold |
| What It Provides | Residue-pair distance restraints (26–30 Å) | Protected interface regions (peptide-level) | 3D density map (2–4 Å resolution) | Predicted 3D structure (in silico) |
| Requires Crystallization | No — solution-phase or in-cell | No — solution-phase | No — but requires purified complex at high concentration | No — computational only |
| Distinguishes Direct Contact from Allostery | Yes — crosslink = direct proximity | No — protection can be allosteric | Yes — at sufficient resolution | Depends on prediction confidence |
| Works on Flexible / Disordered Regions | Partially — depends on lysine availability | Yes — flexible regions exchange rapidly | No — flexible regions are invisible | Predicts with low confidence (pLDDT |
| In-Cell / In Vivo Compatible | Yes — membrane-permeable crosslinkers | No — requires purified protein | No — purified complex only | N/A |
| Throughput | 1–3 samples per MS run | 1 sample per experiment | 1 sample per grid (days of data collection) | Minutes to hours (GPU-dependent) |
| Sample Required | 10–50 µg purified protein | 0.5–1 mg purified protein | ~0.1 mg at ~1 mg/mL | Sequence only |
| Best For | Experimental distance restraints for model validation, PPI interface mapping, integrative modeling, in-cell interactomes | Interface footprint identification, conformational dynamics, biosimilar comparability | High-resolution 3D structure when purification and grid preparation succeed | Hypothesis generation, initial model building — requires experimental validation |
Crosslinker selection & optimization
Crosslinking reaction & digestion
LC-MS/MS acquisition
Crosslinked peptide identification
Distance restraint validation & modeling
Report delivery
Thermo Orbitrap Fusion Lumos with HCD/ETD & Crosslinker Portfolio
XL-MS requires high-resolution MS with multiple fragmentation modes for confident crosslinked peptide identification. Our platform combines Orbitrap instrumentation with a curated crosslinker library covering the most commonly used chemistries.
QC & Crosslinker Optimization
| Step | What We Check |
| Crosslinker Titration | SDS-PAGE: optimal crosslinker:protein ratio produces clear MW shift without aggregation or precipitation |
| Digestion Efficiency | Peptide yield and missed cleavage rate; suboptimal digestion re-optimized with alternative enzyme or extended incubation |
| Crosslink Identification FDR | 1% at crosslink level (target-decoy); low identification count triggers crosslinker or digestion re-optimization |
| Distance Violation Rate | >80% of crosslinks should satisfy Cα–Cα ≤30 Å when mapped onto the correct structure — lower rates flag model or crosslinker issues |
Thermo Orbitrap Fusion Lumos
| Parameter | Specification |
| Purified Protein / Complex | 10–50 µg at 0.5–5 mg/mL; ≥90% purity by SDS-PAGE; amine-free buffer (20 mM HEPES, pH 7.8 preferred; avoid Tris, glycine, and primary amines) |
| Crosslinker Preference | Specify during consultation; we maintain a library of DSS, BS3, DSSO, BS2G, EDC, SDA, and can source custom crosslinkers upon request |
| In-Cell XL-MS | ≥2 × 10⁷ cells per condition; cells treated with membrane-permeable crosslinker, washed, and pelleted; pellets shipped on dry ice |
| Structure/Model (Optional) | PDB file or AlphaFold prediction for crosslink violation analysis; if no structure is available, crosslinks are provided as a standalone restraint list for downstream modeling |
| Shipping | Purified protein: ship on dry ice in amine-free buffer; cell pellets: dry ice; lyophilized protein: ambient (resuspend in amine-free buffer upon receipt) |
Buffer compatibility is the single most common source of XL-MS project delays. Tris, glycine, and other amine-containing buffers react with NHS-ester crosslinkers, consuming the crosslinker before it can react with protein lysines. We strongly recommend dialyzing or buffer-exchanging your sample into 20 mM HEPES (pH 7.8) before submission. Our team can perform buffer exchange upon sample receipt if needed — please coordinate during consultation.
Crosslink Lists, Distance Violation Analysis & Restraint Files for Integrative Modeling
Every XL-MS project includes identified crosslink tables, distance violation analysis on user-provided structures, and modeling-ready restraint files.
Crosslinks on 3D Structure
All identified crosslinks rendered as dashed lines on your protein structure — inter-protein crosslinks in red, intra-protein in blue. Each line annotated with linked residue numbers. The primary figure for publication — showing the experimental distance restraints that define your complex architecture.
Distance Violation Analysis
Crosslinks mapped onto user-provided PDB or AlphaFold structure. Cα–Cα distances calculated; violations (>30 Å for DSS/BS3) highlighted in red, satisfied crosslinks in blue. Violation report with residue-level annotation.
Crosslink Coverage Map & Restraint Files
Sequence coverage map showing crosslink density across your protein — every lysine pair identified with its Cα–Cα distance. HADDOCK/Rosetta/IMP-compatible restraint files (upper bound = 30 Å). PyMOL session file with all crosslinks pre-rendered on your structure.
What distance does a DSS/BS3 crosslink constrain — and how was this validated?
DSS and BS3 have an 11.4 Å spacer arm (fully extended). Merkley et al. (Protein Science) mined an 807-protein molecular dynamics simulation database to empirically validate the appropriate Cα–Cα distance constraint — determining that 26–30 Å between Cα atoms of crosslinked lysines is the appropriate upper bound. This accounts for lysine side-chain flexibility (~6 Å per side chain), local backbone dynamics, and the crosslinker's full extension. In practice, >80% of correctly identified crosslinks should satisfy the ≤30 Å threshold when mapped onto the correct structure. We apply a 30 Å upper bound as the default distance restraint for integrative modeling.
How does XL-MS complement HDX-MS — and when should I use both?
HDX-MS identifies regions of a protein protected from solvent exchange upon binding — defining the interaction footprint at peptide resolution. However, HDX-MS cannot distinguish residues directly at the interface from those undergoing allosteric conformational changes elsewhere. XL-MS provides positive evidence of residue proximity — two residues crosslinked by DSS must have been within ~30 Å. The combination is powerful: HDX-MS maps the protected footprint; XL-MS identifies crosslinks within that footprint, confirming which residues are at the interface and providing distance restraints for modeling. For projects where reviewer-ready structural validation is required — particularly for cryo-EM models or AlphaFold-predicted complexes — the HDX + XL-MS combination provides orthogonal experimental evidence that neither technique alone supplies. Our HDX-MS service can be coordinated with XL-MS from the same sample submission.
How many crosslinks can I expect from a typical XL-MS experiment?
For a purified protein complex at 10–20 µM, a standard DSS crosslinking experiment typically yields 50–200 inter-protein crosslinks and 100–500 intra-protein crosslinks, depending on lysine content, complex size, and crosslinker accessibility. Lysine-poor proteins (e.g., some membrane proteins) may yield fewer crosslinks; lysine-rich, well-folded complexes yield more. The crosslinker titration step during optimization identifies the crosslinker:protein ratio that maximizes inter-protein crosslinks while minimizing over-crosslinking and aggregation. For quantitative benchmarks, we report the number of crosslinks at 1% FDR with the identification software used.
Can XL-MS validate my AlphaFold or cryo-EM model?
Yes — this is one of XL-MS's primary applications. Crosslinks identified by XL-MS are mapped onto your model and Cα–Cα distances are calculated. Crosslinks satisfying the ≤30 Å threshold are classified as "satisfied" — they support the model. Crosslinks exceeding 30 Å are classified as "violations" — they indicate regions where the model disagrees with solution-phase experimental data. A high satisfaction rate (>80%) provides strong experimental support for the model. Systematic violations in a specific domain or interface flag regions requiring model refinement. This orthogonal validation is increasingly requested by reviewers for structures solved by cryo-EM at moderate resolution (3–4 Å) or for AlphaFold-predicted complexes deposited without experimental validation. Our deliverables include a PyMOL session with crosslinks rendered on your structure and a violation report with residue-level annotation.
Which crosslinker should I choose — and can you help me decide?
Yes — crosslinker selection is part of every project consultation. DSS (non-cleavable, hydrophobic) is the most widely used for purified complexes in structural biology — it produces the highest crosslink yield and is supported by the most mature data analysis tools (pLink2). BS3 (water-soluble, membrane-impermeable) is preferred when DMSO is incompatible with your protein. DSSO (MS-cleavable) simplifies data analysis via signature fragment ions — recommended for complex samples (cell lysates, multi-subunit assemblies) where non-cleavable crosslinker analysis may produce higher false discovery rates. For specific applications — shorter distances, cysteine-targeted, or photoreactive crosslinking — we consult on the appropriate chemistry from our crosslinker library during experimental design.
Can XL-MS be performed in living cells — not just on purified proteins?
Yes — using membrane-permeable crosslinkers applied directly to living cells. The crosslinker enters cells and covalently links proximal proteins in their native environment before lysis — eliminating the post-lysis reassortment and dissociation artifacts that affect AP-MS. In-cell XL-MS captures conformation-dependent interactions, weak/transient complexes, and compartment-specific interactomes that are inaccessible to purified-protein methods. Welp et al. (2024) demonstrated in-cell XL-MS in E. coli, identifying >1,500 crosslinked proteins with a specialized NuXL search engine. Our in-cell XL-MS protocol includes crosslinker concentration and incubation time optimization for mammalian, yeast, and bacterial systems.
Merkley, E.D. et al. Distance restraints from crosslinking mass spectrometry: Mining a molecular dynamics simulation database to evaluate lysine-lysine distances. Protein Science. 23(6):747-759. DOI: 10.1002/pro.2464
— Definitive validation of the 26–30 Å Cα–Cα distance constraint for DSS/BS3 crosslinkers, derived from 807-protein MD simulation database. The basis for applying XL-MS crosslinks as quantitative distance restraints in integrative modeling.
Botticelli, L. et al. (2024). Chemical cross-linking and mass spectrometry enabled systems-level structural biology. Current Opinion in Structural Biology. 87:102872. DOI: 10.1016/j.sbi.2024.102872
— Reviews XL-MS as a transformative technology for proteome-wide PPI networks and quantitative in vivo interactome dynamics — positioning the technique as a cornerstone of systems structural biology.
Welp, L.M. et al. (2024). Chemical crosslinking extends and complements UV crosslinking in analysis of RNA/DNA nucleic acid–protein interaction sites by mass spectrometry. bioRxiv. DOI: 10.1101/2024.08.29.610268
— Extends XL-MS to nucleic acid–protein interactions. Introduces NuXL search engine; identifies >1,500 crosslinked proteins in E. coli — demonstrating in-cell XL-MS at proteome scale.
Online Inquiry
