Co-crystallization is the gold standard in structure-based drug discovery: it reveals exactly how a small-molecule inhibitor binds to a protein target at atomic resolution, enabling medicinal chemists to design more potent and selective drugs. From HIV protease inhibitors approved in 1995 to nirmatrelvir (Paxlovid) in 2021, every breakthrough in structure-based drug design traces back to a high-quality protein–inhibitor co-crystal structure that told chemists where and how to optimize their compounds.
What Is Co-Crystallization?
Co-crystallization is the process of crystallizing a protein and its ligand together in the same solution, allowing the small molecule to bind before crystal lattice contacts form. The resulting crystal diffracts X-rays to yield an electron density map that reveals the precise 3D binding mode of the inhibitor in the protein active site.
Co-crystallization differs from soaking (where ligand is added to pre-formed apo-protein crystals) in several ways:
- Higher occupancy: Ligand is present during nucleation and growth, ensuring full binding-site saturation.
- Avoids cracking and strain: Soaking can cause crystal lattice disruption when large conformational changes accompany binding; co-crystallization avoids this artifact.
- Unambiguous electron density: Full occupancy produces clearer ligand density for precise binding-mode interpretation.
The co-crystallization workflow requires four key inputs:
- High-quality recombinant protein – ≥95% purity by SDS-PAGE, PDI <0.2 by DLS, Tm >45°C.
- Inhibitor at molar excess – typically 2–10× over protein to saturate binding sites before nucleation.
- Optimized crystallization conditions – sparse-matrix screens, pH/precipitant fine-tuning, cryoprotectants.
- Synchrotron X-ray access – beamtime for data collection and atomic-resolution structure determination.
Why Protein Quality Matters Most
Protein quality is the single biggest predictor of co-crystallization success. Even the most optimized crystallization conditions cannot compensate for a sample that aggregates, is heterogeneously glycosylated, or carries flexible loops. The essential quality benchmarks are:
- Purity ≥95% by SDS-PAGE; ≥99% by reverse-phase HPLC for challenging targets.
- Monodispersity (PDI <0.2) by dynamic light scattering (DLS); SEC-MALS confirms molecular weight and oligomeric state.
- Thermal stability (Tm >45°C) by differential scanning fluorimetry (DSF); instability leads to denaturation during multi-day crystallization.
- Enzymatic activity confirmed in biochemical or biophysical assays (SPR, ITC, TR-FRET) – crystallography-grade implies active, properly folded protein.
- Surface entropy reduction (SER) engineering: Strategic K→A or E→A mutations at high-entropy surface patches dramatically improve crystallizability without affecting the folded core.
BioHippo distributes crystallography-grade recombinant proteins from Aurora Biolabs, BPS Bioscience, and other verified suppliers, all with confirmed SDS-PAGE purity, specific activity, and DLS profiles suitable for co-crystallization experiments.
Practical Co-Crystallization Strategies
Modern co-crystallization campaigns combine screening scale (rapid condition sampling) with optimization scale (fine-tuning of promising leads):
- Sparse-matrix screening: Commercial kits (Hampton Crystal Screen, Molecular Dimensions JCSG+) test ~100 conditions in parallel via vapor diffusion (hanging-drop or sitting-drop format).
- Hit detection and classification: Initial crystal hits are scored by morphology and reproducibility. Promising leads advance to optimization.
- Fine-screening: PEG concentration, pH, precipitant type, temperature, and additives (cryoprotectants, heavy-atom derivatives) are systematically varied.
- Resolution targets: 1.5–2.5 Å is ideal for medicinal chemistry (atomic binding-mode detail); 2.5–3.5 Å is acceptable for hit triage.
Fragment-based drug discovery (FBLD) applies the same logic to very small molecules (MW <300 Da). Fragments are screened by biophysical assays (SPR, DSF, 1D NMR), then co-crystallized to reveal binding modes. Fragment hits can be grown, merged, or linked into drug-like leads – a strategy that contributed to vemurafenib and venetoclax.
From Co-Crystal to Lead Optimization
Once a co-crystal structure is solved, it becomes the blueprint for computational design:
- Molecular docking: Virtual screens of compound libraries against the crystallographically defined active site identify additional hit series without synthesis.
- Molecular dynamics (MD) simulations: Nanosecond to microsecond MD trajectories test whether the binding pose is stable under physiological conditions and reveal cryptic allosteric pockets.
- Fragment growing: Co-crystal structures of fragments in adjacent sub-pockets guide synthetic elaboration into higher-affinity leads.
- Free energy perturbation (FEP+): Alchemical calculations predict relative binding free energy (ΔΔG) between analogs with ±0.5 kcal/mol accuracy, enabling prioritization before synthesis.
Recent advances in cryo-electron microscopy (cryo-EM) provide structural alternatives for large multimeric complexes and membrane proteins that resist crystallization – enabling co-complex structure determination at 2–3 Å resolution directly from frozen grids.
Co-Crystallization vs. Soaking and Other Methods
How does co-crystallization compare to alternative structural approaches?
| Method | Timeline | Occupancy | Best For |
|---|---|---|---|
| Co-crystallization | 4–8 weeks | Near 100% | High-affinity inhibitors; conformational changes; unambiguous binding modes |
| Soaking | 1–2 weeks | Variable (<100%) | Rapid screening of fragment or compound series; parallel workflows |
| Cryo-EM | 3–6 weeks | High (in vitreous ice) | Large complexes, membrane proteins, flexible targets; 2–4 Å resolution |
| NMR spectroscopy | 4–12 weeks | Dynamic ensemble | Intrinsically disordered proteins; dynamics; <40 kDa optimal |
FAQ – Co-Crystallization Essentials
How do I know if my protein is suitable for co-crystallization?
Run an initial quality assessment: (1) SDS-PAGE to confirm ≥95% purity; (2) DLS to measure PDI (target <0.2); (3) DSF/ThermoFluor to measure Tm (target >45°C); (4) biophysical assay (SPR, ITC, or TR-FRET) to confirm ligand binding and activity. If all four pass, your protein is crystallography-grade and ready to screen.
What's the difference between co-crystallization and soaking?
In co-crystallization, the protein and ligand crystallize together; ligand is present from the start and achieves near-100% occupancy in binding sites. In soaking, apo-protein crystals are transferred into ligand solution afterward – faster, but occupancy is often lower and conformational changes can crack the crystal. Choose co-crystallization for high-affinity inhibitors and unambiguous binding modes; soaking for rapid screening of many compounds.
How long does a co-crystallization campaign take?
Typically 4–8 weeks from initial screening to a diffraction-quality crystal, plus 2–4 weeks for synchrotron data collection and structure determination. High-throughput fragment screening followed by soaking can compress this to 2–3 weeks, but unambiguous binding modes usually require co-crystallization (longer timeline).
Where can I source crystallography-grade proteins?
BioHippo carries crystallography-grade recombinant proteins from Aurora Biolabs, BPS Bioscience, Cytion, and other verified suppliers. All are validated for SDS-PAGE purity, DLS profile, and specific activity. Request a quote for your target or explore custom protein expression services.
The proteins listed above are for research use only (RUO). For product questions or to request a quote, visit our Request a Quote page.
