Macromolecular Crystallization N' Sukumar NECAT, Building 436 Argonne National Lab Argonne, Illinois

1
Macromolecular CrystallizationN.
SukumarNE-CAT, Building 436Argonne National Lab
Argonne, Illinois
2
Crystallization

• Crystals Purified protein undergoes slow
  precipitation from an aqueous solution, resulted
  in alignment of protein molecules in a repeating
  series of unit cell by adopting consistent
  orientation. The molecules are held together by
  non-covalent interactions.
• May Call this Black Art then a Science.
  Because, some people will get the crystals, while
  others can not reproduce the crystals exactly at
  the same condition.
• Now, the process is more understood than 1980s
  but still far away from understanding the
  mechanism behind the crystallization.
• Major Bottleneck in the structure determination
  of macromolecules.

3
Pre-crystallization necessities

• Pure protein gt 95 (at least gt 90) (In
  SDS-PAGE)
• Use Dynamic Light Scattering (DLS) and Mass
  Spectroscopy to study the homogeneity and
  stability problems of the sample
• Concentration of protein gt5mg/ml. At least, a
  concentration , where precipitation occurs 35
  of times during initial screening.
• Proteins can be stored at -70C. Split the
  samples in small aliquots to avoid repeated
  freezing and thawing.

4
Theory of Crystallization
5
Types of crystallization setup

• Frequently used techniques
• - Sitting drop and hanging drop methods
• Sparingly used techniques
• - Free Interface Diffusion (FID)
• - Sandwich Drop,
• - Batch, Microbatch (with and w/o oil)
• - Microdialysis

6
Hanging drop and Sitting drop method

• The protein is mixed with a crystallization
  solution containing various precipitants.
  (usually 2µl and 2µl) (called the drop)
• The diluted protein and crystallization mixture
  is incubated over the concentrated
  crystallizationsolution (called the well).
• Over time, the drop will equilibrate with the
  well concentrating both the protein and the
  precipitants.
• As the protein concentrates it can come out of
  solution as either a precipitate, a phase
  transition (oil), or as a crystal.

7
Free Interface Diffusion (FID)
8
Sandwich Drop
9
Batch method
10
Microbatch Under Oil
11
Microdialysis
12
Theory of Crystallization(Chayen, Cur. Op. Str.
Bio. 14, 577-583 (2004)
13
Well Solution

• Buffers (either non-zwitterionic or zwitterionic)
  of concentration 100mM in the pH range 4-9.
• 2. Salts which includes monovalent, divalent
  salts - in wide range of concentration
• 3, Precipitants
• Poly Ethylene Glycols (PEG) of different
  molecular weight
• Salts (Ammonium Sulfate , LiSo4, LiCl, Na
  Formate, Na Citrate)
• Ethanol, MPD, iso-propanol

14
Commercial Screens

• Hampton Research (HR) www.hamptonresearch.com
• Emerald BioSystems www.emeraldbiosystems.com
• Sigma-Aldrich Screens www.sigma-aldrich.com/x
  raycrystallography
• Each screen usually contains 24-50 solutions
• All the companies produce variety of specialty
  screens such as Membrane protein kit, CRYO
  (conditions ready for freezing), MPD etc.
• - For every new protein, start with at least
  HRs Crystal Screen I II and
• Index Screen.

15
How to analyze the drops?
16
How to analyze the drops?

• Drop - Clear, Light, Medium or Heavy
  precipitate
• Note whether precipitate is always in a
  certain type of condition like gtpH 7, PEGS,
  specific salts etc.
• - microcrystals, Phase separation, salt
  crystals
• Microcrystalline or phase separation
• Good starting point - large crystals can
  be obtained by varying protein concentration and
  various components present in the well .
• Sometimes it might be difficult to distinguish
  micro-crystals from precipitate.

17
Good Crystals
18
Quality of good crystals

• Single (not attached to others, grown in
  cluster etc. this might be fine, or might be
  evidence of problems)
• - When you pick a crystal from a cluster,
  but should be very careful during the process.
• Straight edges.Curves are generally bad
• Reasonable size, 100-400 microns
• But, not all the good looking crystals will
  diffract well. Reason nobody can tell

19
What to do when there is a hit

• For example, there is a very small or micro-
  crystals or phase separation at
• 20PEG 4000, 200mM NaCl, 100mM Tris-HCl
  Buffer pH 7.5
• 1. While scanning around the condition,
  vary only one parameter at a time
• 2. Scan the concentration of PEG4K in the
  range 0 - 40
• 3. Scan PEG 3000, PEG 3500, PEG 5000 PEG
  6000 in the same concentration range. Usually,
  higher mol. wt. PEG less concentration
• and vice versa.
• 4. Scan NaCl concentration in the range
  from 0- 400mM
• 5. Scan with different monovalent salts
  like ammonium chloride, lithium chloride,
  potaasium chloride etc.

20
What to do when there is a hit

• 6. Scan with salt like Na Formate, Na
  sulfate, Na acetate etc.
• 7 . Before choosing the salts, check again
  your screen result whether any of
• the salts presence always leads to
  precipitation or not.
• 8. check the whole buffer active range with
  D0.2.
• 9. Change the buffer suitable in that range
  and scan it.
• 10. Check everything you can think of.
• 11. Once stable conditions found and
  reproducible, then try Additives.
• 12. Try Seedings

21
Additives

• Additives additional chemical added in small
  amount to modify the crystallization conditions
• Sugars (like Xylitol, Sorbital etc), Salts, MPD,
  Ethyene Glycol can be used as an additive.
• Hampton Research sells additive screen (96
  tubes).
• Added only to the drop.
• Adding 10mM of any of the additives to the drop
  is a good start.

22
Procedure for Macroseeding
23
Procedure for microseeding
24
Procedure for Streak Seeding
25
What to do when there is no hit

• Evaluate the Screen carefully.
• What pHs are soluble, preferred salts
• Does your protein need more salt/glycerol to
  remainstable?
• Crystallize in the presence of substrates,
  inhibitors.
• - try multiple substrates at a time
  adding substrates
• or inhibitors can stabilize the
  conformation of protein
• Change the stabilization buffer

26
What to do when there is no hit

• Did you really check the purity thoroughly
  before?
• - Try site directed mutagenesis - Mutating
  Asn to Asp will increase solubility
• - Change hydrophobic residues to
  hydrophilic residues especially surface
• residues of the protein
• - remove if any biologically insignificant
  portion present at the N-terminal or C-terminal
  of the protein
• - Try different constructs
• - With or without Tags
• - Change the source

27
Examples
28
TdT (Terminal deoxynucleotidyl transferase)-
belongs to family of DNA polymerase ß (pol ß)
family (Sukumar et al. Acta Cryst., D56,
1662-1664 (2000)

• Contains two functionally independent region.
  N-terminal domain (130 aa) and C-terminal domain
  (399 aa). Crystallizing full length protein was
  difficult.
• Chopping off N-terminal domain gave the crystals
  in 24 hours (i.e. C-terminal domain which
  contains catalytic core of the protein).

29
Amicyanin crystal grown by macroseeding in 6
months time- size 2x1.3x1mm (Ref. Sukumar et al,
Acta Cryst. D61, 640-642 (2005)
30
Macroseeding Unknown spoiling factor(s)
involved
31
Membrane protein crystallization

• Membrane proteins represent close to 40 of
  currently sequenced genomes.
• Currently, PDB contain 31721 entries.
• Only 170 PDB entries are accounted for membrane
  proteins (just 0.5) Out of 170 entries, only
  93 are unique proteins.

32
Membrane protein Crystallization

• Difficult to handle because of amphipathic
  nature. Possess hydrophobic surface where they
  in contact with lipid bilayers and polar surface
  where they in contact with aqueous solution as
  well as polar head group of lipids.
• To solubilize and purify membrane proteins,
  detergents required amphiphilic molecules that
  form micelles above their critical miceller
  concentration (CMC).
• Various detergents (like Triton, C8E4, Octyl
  Glucoside etc) need to be scanned for
  purification.
• Figure shows monotopic membrane protein binds to
  the membrane

33
Membrane Protein Crystallization

• Getting large quantity of pure protein necessary
  for crystallization is a problem.
• Needs detergent to get crystallized in addition
  to all the components discussed before for
  globular proteins.
• Finding right detergent is a major problem
• Even the good crystals may not give good
  diffraction. Such situation arises more
  frequently in membrane protein than soluble
  proteins.

34
Example Monotopic membrane protein - Mandelate
Dehydrogenase (MDH)

• MDH - Got crystals with at least three different
  detergents.
• Improved diffraction from 12? to 3.7?.
• As the diffraction could not be improved beyond
  3.7?, alternate approach was tried.

35
MDH Vs GOX (Glycolate Oxidase) sequence

• WT MDH MSQNLFNVEDYRKLAQKRLPKMVYDYLEGGAEDEYGV
  KHNRDVFQQWRFKPKRLVDVSRRSLQAEVLGKRQSMPLLIGPT
  80
• WT GOX MEIT--NVNEYEAIAKQKLPKMVYDYYASGAEDQWTL
  AENRNAFSRILFRPRILIDVTNIDMTTTILGFKISMPIMIAPT
  78
• WT MDH GLNGALWPKGDLALARAATKAGIPFVLSTASNMSIED
  LARQCDGDLWFQLYV-IHREIAQGMVLKALHTGYTTLVLTTDV
  159
• WT GOX AMQKMAHPEGEYATARAASAAGTIMTLSSWATSSVEE
  VASTGPGIRFFQLYVYKDRNVVAQLVRRAERAGFKAIALTVDT
  158
• WT MDH AVNGYRERDLHNRFKIPMSYSAKVVLDGCLHPRWSLD
  FVRHGMPQLANFVSSQTSSLEMQAALMSRQMDASFNWEALRWL
  239
• WT GOX PRLGRREADIKNRFVLPPFLTLKNFEGI---------
  ----------DLGKMDKANDSGLSSYVAGQIDRSLSWKDVAWL
  219
• WT MDH RDLWPHKLLVKGLLSAEDADRCIAEGADGVILSNHGG
  RQLDCAISPMEVLAQSV-AKTGK-PVLIDSGFRRGSDIVKALA
  317
• WT GOX QTITSLPILVKGVITAEDARLAVQHGAAGIIVSNHGA
  RQLDYVPATIMALEEVVKAAQGRIPVFLDGGVRRGTDVFKALA
  299
• WT MDH LGAEAVLLGRATLYGLAARGETGVDEVLTLLKADIDR
  TLAQIGCPDITSLSPDYLQNE-GVTNTAPVDHLIGKGTHA 393
• WT GOX LGAAGVFIGRPVVFSLAAEGEAGVKKVLQMMRDEFEL
  TMALSGCRSLKEISRSHIAADWDGPSSRAVARL 369

36
MDH-GOX2 crystal

• Crystallized in sitting drop method Drop
  Protein MESAmm. Sulfate Et.Glycol and FMN
• Reservoir solution 4M NaCl
• Diffracted to 1.35? resolution
• For the first time, an intrachain
  membrane-binding segment is replaced with its
  soluble homologue to facilitate crystallization,
  but still retains catalytic activity.

37
Chimera of Mandelate Dehydrogenase (MDH-GOX2)
GOX (Ref. Sukumar et al, Biochemistry, 40,
9870-9878 (2001)

• GOX insert in MDH closely similar to
  corresponding segment in GOX.
• In GOX, 189-195 residues is disordered.
• 198-206 in MDH (its own residues) which is
  a-helical as predicted but different from
  corresponding segment of GOX(198-205) by 100.
• MDH-GOX2 (Blue) and GOX (Red). Poorly conserved
  segments of MDH-GOX2 and GOX are in Yellow and
  Green respectively.

38
MDH-GOX2 crystal

• Once you get crystals and collect good data, rest
  of your project will be around the computers.
• Sometime, you may need to come back to wet lab
  for crystallization.. When??
• MDH-GOX2 is one such example. The crystals are
  twinned. Though there are many powerful programs
  are available at present, it is always better to
  remove twinning if possible.
• In this case of MDH-GOX2, twinning removed by
  cutting the crystals along the crack, into two
  equal halves.
• In one of its mutants, twinning removed by adding
  substrate to it.

39
References

• Crystallization of Biological Macromolecules A.
  Mcpherson, Cold Spring Harbor Laboratory press
  (ISBN 0-87969-527-7)
• Methods and Results in Crystallization of
  Membrane proteins Ed. So Iwata, IUL
  Biotechmology Series (ISBN No. 0-9636817-9-6)
• Methods in Enzymology, Vol 276, Part A (ISBN
  0-12-182177-3) and Vol 368 Part C (ISBN
  0-12-182177-3) Eds. C.W. Carter, Jr R.M.
  Sweet, Academic Press
• International Tables of Crystallography Vol. F
  Eds. M.G. Rossmann and E.Arnold (ISBN
  0-7923-6857-6)
• www.hamptonresearch.com
• http//wwwstructure.llnl.gov/crystal_lab/crystall.
  htm
• http//www.biochem.utah.edu/heidi/crystallization
  .htm
• http//photoscience.la.asu.edu/photosyn/faculty/Al
  len/protein-structures/membrane-proteins/crystal.h
  tml