1
Growing Protein CrystalsUsing Calcium-Integrin Binding
Protein as a ModelPresented byChad Blamey
FBPwww.scripps.edu/~arvai/ xtals/xtals.html
Goals
• What are good crystals– Why getting good crystals is important
• Understand how crystals grow• Discus techniques for crystallizing proteins
– Application type of discussion• Strategies for optimizing crystal growth
– Understanding your favorite protein• CIB as a model
– My favorite protein!!!• Lysozyme demonstration
2
Lysozyme Demonstration
2µl 2µl
4µl
Buffer30% w/v Polyethelene glycol 50001 M NaCl50 mM NaAcetate pH 4.5
Lysozyme Protein100mg/ml50 mM Na Acetate pH 4.5
glass slide
Should make large crystals in less than 15 minutes.We will watch it for the hour of the lecture.
Everyone Should Know
• Protein crystals are precipitated protein in solutions
• You can think of them highly concentrated aqueous solutions (usually about 500 mg/ml)
• Amorphous precipitation is random• Crystals are ordered
– This is the property we are interested in– Figure 3.1 CMCC– Gray areas!
Crystals PrecipitationCrystalline
3
Crystallization: Needs
• Obtaining quality crystals is by far the limiting step to solving a structure
• Crystals need to be of sufficient size and quality to diffract x-rays
• Size: Normally should be 100 µm in smallest dimension• Quality: Reflections collected from diffraction data are the
primary source of data to build an electron density map, therefore quality of protein model depends greatly on crystal quality– Growing good crystals is key to a good structure
Crystallization
• With enzymes is is often important to maintain enzymatic activity in crystal– Some enzymes can function in crystal
• Best way to test crystal quality is by mounting a crystal and attempting to diffract x-rays– Visual inspection helpful too– May not be meaningful
4
Low vs High Data
• Difference between 9.0 Å and 4.5 Å
• The higher the resolution the better!
• CIB crystal spots
9 Å
4.5 Å
Good vs Poor Data
Ca+007 M035
4.5 ÅPoor, smeary spots
Notice ‘twined’ spots
4.1 ÅGood! Round spots
Higher order visible (circle)
5
Spot PredictionCrystal M035
Do spots match mathematical predictions?
How Do Proteins Crystallize?
• For crystallization to occur it has to be thermodynamically favorable
• Precipitants remove available water forcing proteins to associate with each other
• Hopefully in a organized fashion
Water
Protein
polyethelyene glycolsaltssugarsorganic solvents
precipitantprecipitant- -
--
++
++
6
Growing Crystals:Hanging Drop Method
• Widely used• Vapor diffusion• Drop equalizes with
reservoir• Volume of drop slowly
decreases• Protein concentration
slowly increases• CMCC Figure 3.2 A1 A2 A3 A4 A5 A6
B1 B2 B3 B4 B5 B6C1 C2 C3 C4 C5 C6D1 D2 D3 D4 D5 D6
[X]
[Y]Sitting drop
reservoir
drop
crystals
Phases of Proteins In SolutionNot to be confused with phases of light
[Protein]
Solu
bilit
y
SupersaturationMetastabile
Undersaturated
PrecipitationCrystals
Growth & Nucleation
Growth only
Barrier of Nucleation
Soluble protein
Figure 3.3
CMCC figure 3.3
7
[Protein]
Solu
bilit
y
SupersaturationMetastabile
Undersaturated
Figure 3.3 Phase diagram 1. Concentrate solution enough so nucleation occurs in only a few cases
2. Initial growth pulls some protein out of solution
3. Reducing [protein] back into metastable range
4. Grow only a few large crystals
Basic concept:
Nucleation & Growth
Optimize Crystal Growth
• The number of factors can be overwhelming
• Focus on those factors which most effect growth
• Set up arrays to vary two different conditions at once
• Cross your fingers
A1 A2 A3 A4 A5 A6B1 B2 B3 B4 B5 B6C1 C2 C3 C4 C5 C6D1 D2 D3 D4 D5 D6
[X]
[Y]
8
Crystal Quality
The Tricky Part
• Conditions for crystallization are dependent on each-other
• Crystal quality will change as you vary growth conditions
• Figure 3.4[A]
[B]
For solution made up of three parts A, B and C.Changing [C] will effect the quality of the crystal interms of [A].
Growing Crystals:Other Techniques
• Spin at extremely high speeds, hundreds of thousands of g’s
• Slowly increases the relative protein concentration
Spin
0 hours 6 hours 12 hours
Ulatacentrifugation
Dialysis•Uses liquid-liquid diffusion•Diffusion is slow•Rate controlled by membrane
9
Crystal Screens
• Hampton Research screen tests a wide assortment of conditions of salts, buffers, pH’s and additives– Best conditions from
literature• Often first hits with
screens are small poor quality crystals
• Do not use the absence of crystals as a gauge of conditions rather use solubility
Factors Effecting Crystal Growth
• Ionic Strength*• Specific Ions (Ca2+)• Protein Concentration*• Detergents• Inorganic Precipitant• pH*• Temperature*• Time• Monodispersion*
• Vibrations• Pressure• Gravity• Relative Proportion of
Conditions• Purity Of Protein*• Access to water*• Ligands• Binding partners
*Most important
10
Characterization of Protein
• Of course the more you know about your protein the easier it is to manipulate
• Cystine is often the most critical a.a.– CIB has three and no disulfide bonds, but cause
multimers• Key ligands and metals, like Ca (for CIB)• Stability in certain solutions
– Hydrodynamic radius (NMR)– Stability (CD is great)– Dynamic light scattering– Mass spec
CIB Protein CharacterizationIEF
Western-Blot
IEF CIB purified w/o reducing agentsSDS-PAGE
No DTT DTT
Mar
ker
22CIB
pH 5.6
CIB purified w/ reducing agents
Further characterization of a protein can improve purity and therefore crystal quality
11
Ligands and Co-crystallization
• Try to obtain crystals with different ligands and/or co-crystallize with another protein– Metals, peptides & binding proteins– Proteins with a known structure can simplify the
process• Enzymes and Substrate complexes
– Non-competitive inhibitors– Substrate analogs
• Often changes protein conformation– Two structures!– This gives information about how the protein function
Additives
• Often designed to reduce strength of protein-protein interaction
• Detergents important category• Reducing agents• Organic solvents
12
Small Crystals
• Often small crystals can be made larger by microseeding new drops with previously grown crystals or adding more protein solution
• Multinucleation can be avoided by reducing the temperature or adding glycerol
• Crystals only need to be large enough to diffract x-rays well
Radical Approaches
• Remove either N- or C-terminus by weak proteolysis or by molecular cloning– Often termini can be disordered which interferes with
lattice formation• Crystallize with a fusion protein
– Fusion proteins are well documented with a solved structure that easily from a lattice, example: GST
– “Pull” the fusion protein into an ordered crystal– Can use the protein for molecular replacement to solve
phase– Many recombinant proteins are purified using fusions
anyways, i.e. not hard to try
13
Radical Approaches, Cont
• Mutants:– Specific residues problem residues can be mutated
using recombinant DNA technology– Domains can be crystallized separately
• Issues:– Different conformations from the native state likely– Domains can only be part of the story– Changing the means starting over in terms of
crystallization solution
Discussion
• Given a hypothetical protein that doesn’t give you any positive hits in your first screen what could you do to obtain quality crystals?– Meaning: almost no precipitation in each drop!
• Likewise what if you get lots of precipitation?• Say on the other hand at room temperature you
have a condition with lots of tiny crystals, what can be done to reduce the amount of nucleation?
14
Lysozyme
• Vapor diffusion takes about 12 hours to complete. Where on phase diagram did the [lysozyme] start given almost no vapor diffusion occurred?
• Lysozyme in other solutions crystallizes much more slowly (period of days). Which solutions would yield higher quality crystals? Why?
• What are some of the properties of lysozyme that allow it to crystallize so quickly?
CIB
• Similar to a ubiquitous protein Calmodulin
• Binds calcium• Regulates other
proteins (13 so far)• Found in most tissues
types: brain, muscle etc.
• No enzymatic activity
15
CIB Discussion
• What are some possible techniques that could be used to obtain CIB crystals?
• What about the cystines of CIB?• Why is it important that the radius of CIB is
smaller when it is bound to calcium?• CIB contains a surface exposed hydrophobic
patch, how could this information change your crystallization conditions?