# Protein Crystallography: Improving Crystals for Better Diffraction

## Introduction to Protein Crystallography and Crystal Quality

Protein crystallography is the [dominant](/blog/careers/dominant-definition-biology) method for determining the three-dimensional structures of proteins at atomic resolution. The technique relies on the ability to grow a well-ordered three-dimensional crystal of the protein of interest, irradiate it with X-rays, and measure the diffraction pattern that results from the constructive and destructive interference of X-rays scattered by the electron clouds of atoms within the crystal. The diffraction pattern is then mathematically transformed into an electron density map, into which the [amino acid sequence](/blog/guides/amino-acid-sequence) is fitted to build an atomic model.

The fundamental relationship between crystal quality and structural information is straightforward: the more ordered the crystal lattice, the higher the resolution of the diffraction data, and the more detailed the structural model. A crystal that diffracts to 3.0 Å resolution will reveal the overall fold of the protein and the positions of large side chains, but it will not reliably show water molecules, small ligand interactions, or subtle conformational changes. A crystal that diffracts to 1.5 Å or better allows visualization of individual atoms, ordered solvent, and precise hydrogen-bonding networks. For [structure-based drug design](/knowledge/bioinformatics/structure-based-drug-design-bioinformatics), where the goal is to understand how a small molecule binds to a protein pocket with sub-angstrom precision, high-resolution data are not a luxury—they are a requirement.

### Why Crystal Quality Matters

Crystal quality is not a binary property. A crystal can be large but poorly ordered, diffracting weakly to low resolution. Conversely, a crystal can be small—perhaps 50 micrometers in each dimension—yet diffract to atomic resolution if its internal lattice is highly regular. The key parameters that define crystal quality are:

- **Resolution**: The minimum interplanar spacing (d-spacing) from which measurable diffraction is observed. Higher resolution means more reflections and a more detailed electron density map.
- **Mosaicity**: The degree of angular misalignment between the microscopic blocks (mosaic blocks) that make up the crystal. Low mosaicity (typically 0.1–0.5°) indicates a well-ordered crystal; high mosaicity (above 1°) indicates lattice disorder and broad, overlapping diffraction spots.
- **Wilson B-factor**: A measure of the average atomic displacement (thermal motion and static disorder) within the crystal, estimated from the fall-off of diffraction intensities with resolution. Lower Wilson B-factors (20–40 Ų) indicate a more rigid, well-ordered crystal.
- **Diffraction limit and spot quality**: The signal-to-noise ratio of the diffraction spots, their sharpness, and the absence of diffuse scattering.

A crystal that is visually beautiful—large, with sharp facets and uniform extinction under polarized light—can still be a poor diffractor. Conversely, a small, unremarkable-looking crystal can be an excellent diffractor. The only reliable way to assess crystal quality is to expose it to an X-ray beam.

### Overview of the Crystallization Process

[Protein crystallization](/knowledge/molecular-biology/protein-crystallization) is a phase transition in which the protein leaves a supersaturated solution and forms an ordered solid. The process is governed by the phase diagram, which plots protein concentration against precipitant concentration (or another crystallization parameter). The diagram contains four regions: the undersaturated zone (where crystals dissolve), the metastable zone (where existing crystals grow but no new nuclei form), the labile zone (where nucleation occurs spontaneously), and the precipitation zone (where the protein forms amorphous aggregates).

Crystallization proceeds in two stages: nucleation and growth. Nucleation is the formation of a critical nucleus—a small, ordered cluster of protein molecules that is thermodynamically stable and can grow. Growth is the addition of protein molecules to the surface of the nucleus in an ordered manner. The goal of crystallization is to achieve a balance: enough supersaturation to drive nucleation, but not so much that the protein precipitates or that too many nuclei form, depleting the protein supply and producing many small, poorly ordered crystals.

The standard method for initial crystallization screening is vapor diffusion, in which a drop containing protein and precipitant is equilibrated against a reservoir containing a higher concentration of precipitant. Water vapor diffuses from the drop to the reservoir, concentrating the protein and precipitant in the drop, driving the solution into the supersaturated region. Two common vapor-diffusion geometries are hanging drop and sitting drop. The hanging-drop geometry, in which the drop is suspended from a siliconized coverslip, is favored for its optical clarity and ease of manipulation. The sitting-drop geometry, in which the drop sits on a small pedestal, is more amenable to automation and is the standard format for commercial robotic screening.

## Key Factors Affecting Crystal Growth

The success of crystallization depends on a small number of critical variables. Understanding how each variable influences the phase behavior of the protein is essential for rational optimization.

### Protein Purity and Homogeneity

Protein purity is the single most important factor in crystallization. A protein preparation that is 95% pure by SDS-PAGE may still contain contaminants that inhibit crystallization. Contaminants can act as heterogeneous nucleants, promoting the formation of amorphous precipitate or multiple small crystals. They can also incorporate into the crystal lattice, disrupting order and degrading diffraction quality.

The purity requirement for crystallization is stringent: the protein should be >97% pure, ideally >99%. This means going beyond a single affinity purification step. For example, a His-tagged protein purified by immobilized metal affinity chromatography (IMAC) should be subjected to a second purification step, such as size-exclusion chromatography (SEC) or ion-exchange chromatography. SEC is particularly valuable because it not only removes contaminants but also exchanges the protein into a well-defined buffer and removes aggregates. Aggregates are a common cause of crystallization failure; they act as nucleants for amorphous precipitate and deplete the concentration of soluble, monodisperse protein. See [His Tag Protein Purification](/knowledge/molecular-biology/his-tag-protein-purification) for a detailed protocol on IMAC, and [Custom Protein Purification](/knowledge/molecular-biology/custom-protein-purification) for strategies to design a multi-step purification scheme.

Protein homogeneity extends beyond purity. A protein that is heterogeneous due to post-translational modifications, partial proteolysis, or conformational flexibility will be difficult to crystallize. Mass spectrometry can detect such heterogeneity; [Intact Protein Mass Spectrometry](/knowledge/molecular-biology/intact-protein-mass-spectrometry) is the method of choice for confirming that the molecular weight of the purified protein matches the predicted sequence and that no truncation or modification has occurred. If the protein is a mixture of conformations, adding a ligand that stabilizes one conformation—such as a substrate analog, inhibitor, or cofactor—can improve homogeneity and crystallizability.

### Precipitant and pH

The precipitant is the reagent that drives the protein out of solution. Precipitants are classified by their mechanism of action. Salts such as ammonium sulfate and sodium citrate act by salting-out: they compete with the protein for water of hydration, reducing the protein's solubility. Polymers such as polyethylene glycol (PEG) of various molecular weights (e.g., PEG 3350, PEG 4000, PEG 8000) act by volume exclusion: they occupy space in the solution, effectively increasing the protein concentration and promoting protein-protein contacts. Organic solvents such as 2-methyl-2,4-pentanediol (MPD) and ethanol lower the dielectric constant of the solution, enhancing electrostatic interactions between protein molecules.

The choice of precipitant and its concentration determines the position of the solution in the phase diagram. At low precipitant concentration, the solution is undersaturated and crystals dissolve. As precipitant concentration increases, the solution enters the metastable zone, then the labile zone, and finally the precipitation zone. The goal is to find a precipitant concentration that places the drop in the labile zone for nucleation but that, as the drop equilibrates and protein is consumed, moves it into the metastable zone for growth.

pH is equally critical because it determines the ionization state of amino acid side chains on the protein surface, which in turn determines the electrostatic interactions that drive crystal packing. The pH of the crystallization buffer should be within 0.5–1.0 pH units of the protein's isoelectric point (pI) for optimal crystallization, because at the pI the protein has no net charge and protein-protein contacts are favored. However, the protein must remain stable and soluble at that pH, which is not always the case. A pH screen spanning 4.0 to 9.0 in increments of 0.5 is standard.

### Temperature and Other Physical Parameters

Temperature affects protein solubility, the kinetics of nucleation, and the rate of crystal growth. Most crystallization screens are performed at 20°C or 4°C. Lower temperatures generally reduce protein solubility and slow the kinetics of both nucleation and growth. For proteins that are unstable at room temperature, crystallization at 4°C may be the only option. For proteins that nucleate too rapidly at 20°C, reducing the temperature can slow nucleation and allow fewer, larger crystals to form.

Other physical parameters include the drop volume (typically 0.1–2 µL in modern robotic screening), the ratio of protein to precipitant in the drop (typically 1:1, but varying this ratio can be useful), the rate of equilibration (controlled by the reservoir-to-drop concentration difference and by the use of oils to slow vapor diffusion), and the surface on which the drop sits. The latter is often overlooked: the hydrophobicity and charge of the crystallization surface can influence nucleation. Silanized (hydrophobic) coverslips are standard for hanging drops, while sitting-drop plates have either polystyrene or glass surfaces, each with different nucleation propensities.

## Systematic Optimization of Crystallization Conditions

Initial crystallization screens are designed to sample a broad range of conditions with the goal of producing a "hit"—a condition that yields crystals, microcrystals, or crystalline precipitate. These hits are rarely optimal; they are starting points for systematic optimization.

### Grid Screening and Fine Screening

Once a hit is obtained, the most powerful optimization strategy is grid screening: varying one or two parameters at a time in a systematic matrix. The most common grid screen varies precipitant concentration against pH. For example, if a hit was obtained in 0.1 M sodium citrate pH 5.5, 20% PEG 3350, a grid screen would vary PEG 3350 from 12% to 28% in 2% increments and pH from 4.5 to 6.5 in 0.5 increments. This produces a 9 × 5 matrix of 45 conditions, each tested in a separate drop.

The logic of grid screening is that the initial hit may be in a region of the phase diagram that is suboptimal. The hit condition may be too close to the precipitation zone, producing many small crystals, or too close to the metastable zone, producing no crystals at all. By sampling a range of precipitant concentrations and pH values around the hit, the optimal condition—where a few large, well-ordered crystals grow—can be found.

A second dimension of optimization is the protein concentration. The initial screen typically uses a protein concentration of 10–20 mg/mL. If the hit produced many small crystals, reducing the protein concentration to 5–8 mg/mL may reduce nucleation and allow larger crystals to grow. If the hit produced no crystals but only precipitate, increasing the protein concentration may push the solution into the labile zone.

### Using Sparse Matrix Screens as a Starting Point

Sparse matrix screens, such as the commercially available Index, Crystal Screen, and JCSG+ kits, are designed to sample a diverse range of chemical conditions—different salts, polymers, pH buffers, and additives—in a single 96-condition screen. The conditions are not a systematic grid; they are a "sparse" sampling of the vast chemical space of crystallization conditions, based on empirical knowledge of conditions that have successfully crystallized other proteins.

The value of sparse matrix screens is that they efficiently identify a starting point. A typical protein will produce crystals in 1–5% of the conditions in a sparse matrix screen. The hits are then optimized by grid screening around the hit conditions. It is important to recognize that a sparse matrix screen is not an optimization tool; it is a discovery tool. The conditions in the screen are deliberately broad and are rarely optimal for any given protein.

When a hit is obtained, it is worth repeating the hit condition several times to confirm reproducibility before investing effort in optimization. Some hits are irreproducible due to stochastic nucleation events; others are reproducible but produce crystals of variable quality. The reproducibility of a hit is itself a diagnostic: highly reproducible hits tend to be in conditions that are robust to small variations in protein concentration, temperature, and drop volume, and these conditions are often easier to optimize.

## Additives and Their Role in Crystal Improvement

Additives are small molecules, salts, or other compounds that are added to the crystallization drop at low concentration (typically 1–100 mM) to modulate protein-protein interactions and improve crystal quality. Additives can act by several mechanisms: they can bind to specific sites on the protein surface, stabilizing a particular conformation; they can alter the dielectric constant or ionic strength of the solution; they can compete with the protein for water, effectively increasing the local protein concentration; or they can act as mild denaturants, increasing the conformational flexibility of surface loops and allowing better crystal packing.

### Common Additives and Their Mechanisms

A standard additive screen includes a diverse set of small molecules:

| Additive Class | Examples | Mechanism of Action |
|---|---|---|
| Divalent cations | MgCl₂, CaCl₂, ZnCl₂ | Coordinate to surface residues (e.g., His, Asp, Glu), bridging protein molecules in the lattice |
| Reducing agents | DTT, TCEP, β-mercaptoethanol | Prevent oxidation of surface cysteines, maintain protein in a homogeneous reduced state |
| Non-detergent sulfobetaines | NDSB-195, NDSB-201 | Modulate protein solubility and reduce aggregation without denaturing |
| Polyamines | Spermine, spermidine | Neutralize negative charges on DNA-binding proteins, promote lattice contacts |
| Glycerol and other polyols | Glycerol, ethylene glycol, sucrose | Stabilize the protein, reduce solvent entropy, promote ordered crystal contacts |
| Metal ions | Co²⁺, Ni²⁺, Cd²⁺ | Form specific coordination complexes with surface His residues |
| Coenzymes and substrates | NAD⁺, ATP, S-adenosylmethionine | Stabilize a specific conformational state of the enzyme |

The choice of additive is guided by the properties of the protein. For a protein with surface cysteines, a reducing agent is essential. For a protein that is prone to aggregation, NDSB-201 at 100 mM can be effective. For an enzyme that undergoes a conformational change upon ligand binding, adding the substrate or a substrate analog can lock the protein into a single conformation, improving lattice order.

### Detergents and Lipids for Membrane Proteins

Membrane proteins present a special challenge: they have large hydrophobic surfaces that must be shielded from the aqueous solvent. The standard approach is to solubilize the membrane protein in a detergent micelle, and to crystallize the protein-detergent complex. The choice of detergent is critical. Short-chain detergents such as n-octyl-β-D-glucopyranoside (β-OG) and n-nonyl-β-D-maltoside (NM) form small micelles and are often more amenable to crystallization than long-chain detergents such as dodecyl-β-D-maltoside (DDM), which form large micelles that obscure the hydrophilic surface available for crystal contacts.

For membrane proteins that resist crystallization in detergents, lipidic cubic phase (LCP) crystallization is an alternative. In LCP, the protein is reconstituted into a lipid bilayer formed by monoolein, and crystallization is driven by the addition of precipitant, which causes the protein to leave the lipid bilayer and form crystals. LCP has been essential for the structure determination of G protein-coupled receptors (GPCRs), including the β₂-adrenergic receptor and rhodopsin. The method is technically demanding but can yield crystals of membrane proteins that are refractory to detergent-based crystallization.

## Seeding Techniques to Enhance Crystal Quality

Seeding is the introduction of pre-formed crystal nuclei into a crystallization drop to control nucleation and promote growth. The principle is to separate nucleation from growth: nucleation is difficult to control and often produces too many crystals, while growth from a seed can be carefully controlled to produce fewer, larger, and better-ordered crystals. Seeding is particularly useful when initial screens produce many small crystals or clusters of crystals, or when crystals appear but do not grow to a usable size.

### Microseeding

Microseeding involves the transfer of microscopic crystal fragments into a fresh crystallization drop. The seed stock is prepared by crushing a crystal (or crystalline precipitate) in a small volume of mother liquor, typically 20–50 µL, and then serially diluting the suspension. A 1:1000 or 1:10,000 dilution is often appropriate. A small volume of the diluted seed stock—typically 0.1–0.5 µL—is added to a fresh drop that is in the metastable zone, where nucleation would not occur spontaneously but where existing nuclei can grow.

The advantage of microseeding is that it decouples nucleation from growth. The seed provides the nucleus, and the drop conditions are chosen to favor growth, not nucleation. This often produces a single large crystal per drop, or a few large crystals, rather than a shower of microcrystals. Microseeding is also useful for reproducing crystallization in a new condition: a seed from a crystal grown in condition A can be used to nucleate growth in condition B, which may produce better-ordered crystals.

### Macroseeding

Macroseeding is the transfer of a single, well-formed crystal into a fresh drop. The crystal is washed briefly in a stabilizing solution (mother liquor with a slightly lower precipitant concentration) to remove surface impurities, then transferred to a fresh drop that is in the metastable zone. The crystal continues to grow, and the fresh drop provides a clean environment free of the impurities and precipitate that may have accumulated in the original drop.

Macroseeding is technically more demanding than microseeding because the crystal must be handled without damage. It is used when a crystal grows to a certain size and then stops, or when a crystal is of good quality but too small for diffraction experiments. Macroseeding can be repeated: a crystal can be transferred through several fresh drops, growing larger with each transfer.

### Streak Seeding

Streak seeding is a variant of microseeding in which a cat whisker or a fine glass fiber is drawn across a crystal or crystalline precipitate to pick up microscopic nuclei, and then streaked across a fresh drop. The streak deposits a gradient of nuclei: more nuclei at the start of the streak, fewer at the end. Crystals often grow along the streak line, with the largest crystals at the end of the streak where the nucleus density is lowest. Streak seeding is quick and simple, requiring no dilution series, and is often the first seeding technique to try.

## Counter-Diffusion and Other Advanced Methods

Vapor diffusion is the workhorse of [protein crystallization](/knowledge/molecular-biology/protein-crystallization), but it is not the only method. Alternative methods can produce crystals of superior quality, particularly for proteins that are difficult to crystallize by vapor diffusion.

### Counter-Diffusion in Capillaries

Counter-diffusion is a crystallization method in which the protein and precipitant diffuse toward each other in a capillary, creating a gradient of supersaturation along the length of the capillary. The capillary, typically made of glass or quartz with an inner diameter of 0.2–0.5 mm, is filled with a gel (such as agarose or silica gel) containing the protein. One end of the capillary is sealed, and the other end is placed in contact with a reservoir of precipitant. The precipitant diffuses into the gel, creating a gradient of precipitant concentration along the capillary. At each point along the capillary, the supersaturation is different; somewhere along the gradient, the conditions will be optimal for nucleation and growth.

The advantage of counter-diffusion is that it samples a continuous range of conditions in a single experiment. It also produces crystals in a convection-free environment (the gel suppresses convection), which can improve crystal order. Counter-diffusion is particularly well suited for proteins that are sensitive to handling or that crystallize only in a very narrow range of conditions.

### Gel Crystallization

Gel crystallization is a method in which the protein is crystallized in a gel matrix, such as agarose, silica gel, or polyacrylamide. The gel suppresses convection and sedimentation, creating a quiescent environment in which crystals grow slowly and with high order. The gel also provides a physical support that prevents crystals from settling and sticking to surfaces.

Gel crystallization is often combined with counter-diffusion or with vapor diffusion. In the gel-acupuncture method, a capillary filled with gel and protein is inserted into a reservoir of precipitant, and crystals grow along the concentration gradient. The method is particularly effective for proteins that produce crystals with high mosaicity by conventional vapor diffusion, as the convection-free environment can reduce lattice disorder.

## Evaluating Crystal Quality and Diffraction

The ultimate test of crystal quality is diffraction, but there are useful preliminary assessments that can be made before a crystal is exposed to X-rays.

### Visual Inspection and Birefringence

Visual inspection under a stereomicroscope can reveal a great deal about crystal quality. Well-ordered crystals have sharp edges, flat faces, and uniform extinction when viewed under crossed polarizers. Protein crystals are birefringent: they rotate the plane of polarized light, and when viewed between crossed polarizers, they appear bright against a dark background. The brightness and uniformity of the birefringence are indicators of internal order. A crystal that appears uniformly bright when rotated is likely well ordered; a crystal that shows irregular extinction or appears dark may be disordered or may be a salt crystal.

Salt crystals are a common artifact. They are often birefringent as well, but they can be distinguished from protein crystals by their hardness (they are difficult to crush with a needle), their high density (they sink in the mother liquor), and their appearance (they often have a regular, geometric habit). Protein crystals are soft, have a lower density, and often have a more irregular habit. The definitive test is to crush a crystal and examine it under a microscope: protein crystals crush easily and appear as a fine powder, while salt crystals resist crushing.

### Diffraction Screening and Data Collection

The definitive assessment of crystal quality is X-ray diffraction. Modern synchrotron beamlines allow rapid screening of crystals: a crystal is mounted, exposed to the X-ray beam for a few seconds, and the diffraction pattern is recorded. The key metrics are:

- **Resolution**: The highest angle at which diffraction spots are observed above background. This is the single most important metric.
- **Spot shape**: Diffraction spots should be sharp and round. Elongated or streaky spots indicate mosaicity or lattice disorder.
- **Diffuse scattering**: A high background of diffuse scatter indicates disorder in the crystal, often due to high solvent content or thermal motion.
- **Wilson B-factor**: Estimated from the intensity fall-off, this gives a measure of the average atomic displacement in the crystal.

A crystal that diffracts to 2.5 Å or better is generally worth pursuing for structure determination. A crystal that diffracts to 3.5 Å or worse may still be useful for a low-resolution structure, but it is likely to require optimization. See [X Ray Crystallography](/knowledge/molecular-biology/x-ray-crystallography) for a detailed description of the diffraction experiment and data processing.

## Common Pitfalls and Troubleshooting

Even experienced crystallographers encounter failures. The most common pitfalls are described below, along with strategies for troubleshooting.

### Over-Nucleation and How to Avoid It

The most common problem in crystallization is over-nucleation: the drop produces many small crystals, or a shower of microcrystals, but no large crystals. This occurs when the drop is in the labile zone at a high supersaturation, where nucleation is rapid and many nuclei form simultaneously. Each nucleus competes for the available protein, and none grows to a usable size.

The solution is to reduce supersaturation. This can be done by:

1. Reducing the protein concentration from 15 mg/mL to 8 mg/mL.
2. Reducing the precipitant concentration by 2–5% (e.g., from 20% PEG 3350 to 16%).
3. Changing the pH by 0.5 units away from the pI.
4. Lowering the temperature from 20°C to 4°C, which slows nucleation.
5. Using seeding to introduce a controlled number of nuclei into a drop that is in the metastable zone.

### Crystal Handling and Mounting Issues

Protein crystals are fragile. They are held together by weak non-covalent interactions—hydrogen bonds, salt bridges, and van der Waals contacts—and they are typically 50–90% solvent by volume. They are easily damaged by mechanical stress, changes in temperature, and changes in the composition of the surrounding solution.

Common handling mistakes include:

- **Crystal dehydration**: Exposing a crystal to air causes water to evaporate, which can crack the crystal or change the unit cell dimensions. Crystals must be transferred quickly and kept in mother liquor at all times.
- **Crystal crushing**: Using a metal needle to manipulate crystals can crush them. A nylon loop or a glass fiber is gentler.
- **Incorrect cryoprotection**: For data collection at cryogenic temperatures (100 K), crystals must be transferred to a cryoprotectant solution (e.g., mother liquor plus 20–30% glycerol or ethylene glycol) before flash-cooling. If the cryoprotectant concentration is too low, ice forms and destroys the crystal; if it is too high, the crystal may crack or the lattice may be disrupted.
- **Temperature shock**: Rapid changes in temperature can crack crystals. Crystals should be equilibrated to the temperature of the crystallization room before handling.

## Practical Summary: Steps to Improve Your Crystals

The following workflow summarizes the process of improving crystal quality, from initial screening to final optimization.

### Step-by-Step Optimization Workflow

1. **Confirm protein purity and homogeneity.** Run SDS-PAGE to confirm >97% purity. Run [Automated Protein Quantification](/knowledge/molecular-biology/automated-protein-quantification) to confirm the protein concentration. Use [Protein Quantification Mass Spectrometry](/knowledge/molecular-biology/protein-quantification-mass-spectrometry) to check for post-translational modifications or truncation. If the protein is heterogeneous, consider adding a ligand or cofactor to stabilize a single conformation.
2. **Perform initial sparse matrix screening.** Set up a 96-condition screen at 20°C with a protein concentration of 10–15 mg/mL. Incubate and inspect daily for the first week, then weekly thereafter.
3. **Identify hits and assess their quality.** Crystals, microcrystals, and crystalline precipitate are all hits. Score them by size, number, and morphology.
4. **Optimize the hit by grid screening.** Vary precipitant concentration and pH around the hit condition. Use 2% increments in precipitant and 0.5 pH units. Include a protein concentration series (8, 12, 16 mg/mL).
5. **Test additives.** If grid screening does not improve crystal quality, add a small-molecule additive screen. Test additives at 10 mM and 100 mM.
6. **Try seeding.** If crystals are numerous but small, or if they are clusters, prepare a seed stock and perform streak seeding or microseeding into fresh drops in the metastable zone.
7. **Evaluate crystals by diffraction.** Mount the best crystals, cryoprotect them, and screen them at a synchrotron beamline. Assess resolution, spot shape, and mosaicity.
8. **Iterate.** If diffraction is poor, return to the grid screen and refine further. Consider alternative methods such as counter-diffusion or gel crystallization.

### When to Consider Alternative Approaches

If conventional vapor diffusion and optimization fail to produce diffraction-quality crystals, consider the following:

- **Change the protein construct**: Truncate flexible N- or C-terminal regions, or remove disordered loops. Limited proteolysis can identify stable domains.
- **Change the protein surface**: Site-directed mutagenesis of surface residues (e.g., replacing Lys with Ala or Glu) can create new crystal contacts. This is the basis of surface entropy reduction.
- **Crystallize a complex**: If the protein is a kinase, crystallize it with a nucleotide analog. If it is a receptor, crystallize it with its ligand. The complex is often more rigid and more crystallizable than the apo protein.
- **Try a different crystallization method**: Counter-diffusion, gel crystallization, or lipidic cubic phase may succeed where vapor diffusion fails.

## Frequently Asked Questions

### How can I improve the quality of my protein crystals?

Improve protein purity first: aim for >97% purity by adding a second purification step such as size-exclusion chromatography. Then systematically optimize the crystallization condition by varying precipitant concentration, pH, and protein concentration in a grid screen. Add small-molecule additives to modulate protein-protein interactions. If crystals are numerous but small, use seeding to control nucleation. Finally, screen crystals by X-ray diffraction and iterate based on the results.

### What are the most common reasons for poor crystal diffraction?

The most common reasons are: (1) protein heterogeneity—the sample contains multiple conformations, truncations, or post-translational variants; (2) high solvent content—crystals with >70% solvent often diffract poorly; (3) lattice disorder—high mosaicity due to weak crystal contacts; (4) radiation damage—crystals degrade during data collection; and (5) improper cryoprotection—ice formation or lattice disruption during flash-cooling.

### What is microseeding and how does it help?

Microseeding is the transfer of microscopic crystal fragments into a fresh crystallization drop. The seed stock is prepared by crushing a crystal in mother liquor and serially diluting it. A small volume of the diluted seed stock is added to a drop in the metastable zone, where nucleation would not occur spontaneously. The seed provides a nucleus for growth, allowing the drop conditions to be optimized for growth rather than nucleation. This often produces fewer, larger, and better-ordered crystals.

### Why is protein purity important for crystallization?

Protein purity is important because contaminants interfere with crystallization in several ways. They can act as heterogeneous nucleants, promoting amorphous precipitate or multiple small crystals. They can incorporate into the crystal lattice, disrupting order and degrading diffraction quality. They can also bind to the protein surface, blocking the crystal contacts that are necessary for lattice formation. A protein preparation that is >97% pure is generally required for successful crystallization.

### What additives can improve protein crystals?

Common additives include divalent cations (MgCl₂, CaCl₂) that bridge protein molecules in the lattice; reducing agents (DTT, TCEP) that prevent oxidation of surface cysteines; non-detergent sulfobetaines (NDSB-201) that reduce aggregation; polyamines (spermine) that neutralize surface charges; and glycerol that stabilizes the protein. The choice of additive depends on the properties of the protein. A standard additive screen tests 96 different small molecules at two concentrations.

### How do I choose the right precipitant concentration?

The right precipitant concentration is one that places the drop in the labile zone for nucleation but that, as the drop equilibrates and protein is consumed, moves it into the metastable zone for growth. In practice, this is found by grid screening: vary the precipitant concentration in 2% increments around the initial hit. If the drop produces many small crystals, reduce the precipitant concentration. If the drop produces no crystals, increase it. The goal is to find the concentration that produces a few large crystals.

### What is the difference between macroseeding and microseeding?

Macroseeding is the transfer of a single, well-formed crystal into a fresh drop in the metastable zone. The crystal continues to grow in the fresh drop, which is free of the impurities and precipitate that accumulated in the original drop. Microseeding is the transfer of microscopic crystal fragments into a fresh drop. Microseeding is easier to perform and is more commonly used; macroseeding is used when a crystal is of good quality but too small, and it requires careful handling to avoid damaging the crystal.

### How can I tell if my crystal is good enough for X-ray diffraction?

Visual inspection is a poor predictor of diffraction quality. The only reliable way to assess a crystal is to expose it to an X-ray beam. At a synchrotron beamline, a crystal can be screened in a few seconds. A crystal that diffracts to 2.5 Å or better with sharp spots and low mosaicity is generally good enough for structure determination. A crystal that diffracts to 3.5 Å or worse may still be useful for a low-resolution structure, but it is likely to require optimization.

## Key Takeaways

- Protein purity and homogeneity are the most critical factors for successful crystallization; aim for >97% purity and confirm homogeneity by mass spectrometry.
- Crystal quality is defined by resolution, mosaicity, and Wilson B-factor, not by crystal size or appearance; the only reliable assessment is X-ray diffraction.
- Systematic optimization by grid screening—varying precipitant concentration, pH, and protein concentration—is the most effective way to improve crystal quality.
- Additives such as divalent cations, reducing agents, and non-detergent sulfobetaines can modulate protein-protein interactions and improve crystal packing.
- Seeding techniques, including microseeding, macroseeding, and streak seeding, separate nucleation from growth and are powerful tools for producing fewer, larger, better-ordered crystals.
- Alternative crystallization methods, including counter-diffusion and gel crystallization, can yield higher-quality crystals for proteins that resist conventional vapor diffusion.
- Common pitfalls include over-nucleation, protein heterogeneity, and improper crystal handling; troubleshooting requires a systematic approach that addresses each variable in turn.

## Further Reading

- McPherson A, DeLucas LJ. *Microgravity protein crystallization*. NPJ microgravity. 2015. [PubMed 28725714](https://doi.org/10.1038/npjmgrav.2015.10)
- Smialowski P, Wong P. *Protein Crystallizability*. Methods in [molecular biology](/blog/careers/molecular-biology) (Clifton, N.J.). 2016. [PubMed 27115641](https://doi.org/10.1007/978-1-4939-3572-7_17)
- Schönherr R, Rudolph JM, Redecke L. *Protein crystallization in living cells*. Biological chemistry. 2018. [PubMed 29894295](https://doi.org/10.1515/hsz-2018-0158)
- Matinyan S, Filipcik P, Abrahams JP. *Deep learning applications in protein crystallography*. Acta crystallographica. Section A, Foundations and advances. 2024. [PubMed 38189437](https://doi.org/10.1107/S2053273323009300)
- Stevens RC. *High-throughput protein crystallization*. Current opinion in structural biology. 2000. [PubMed 11042454](https://doi.org/10.1016/s0959-440x(00)00131-7)
- Blakeley MP, Hasnain SS, Antonyuk SV. *Sub-atomic resolution [X-ray crystallography](/knowledge/molecular-biology/x-ray-crystallography) and neutron crystallography: promise, challenges and potential*. IUCrJ. 2015. [PubMed 26175905](https://doi.org/10.1107/S2052252515011239)



<div data-calculator="toxicity"></div>

## Related Clinical & Scientific Guides

* [MAPK Pathway: Mechanism, Function, and Clinical Relevance](/knowledge/molecular-biology/mapk-pathway)
* [Mammalian Cell Culture Bioreactors: A Practical Guide](/knowledge/molecular-biology/mammalian-cell-culture-bioreactor)
* [Nucleotide Formation: Biosynthesis and Assembly of DNA/RNA Building Blocks](/knowledge/molecular-biology/nucleotide-formation)