Moss Tom. DNA-protein interactions: principles and protocols
Подождите немного. Документ загружается.
554 Colin and Brown
3. Customized gradients can be supplied to standard laboratory chromatography col-
umns using a programmed FPLC control unit and pumps. Flow rates of 1–50 mL/h
are possible without significant back pressure.
4. The protein binds to heparin–Sepharose in 0.25 M NaCl and 7 M urea. A reverse
gradient is applied to remove urea. The protein is eluted in a 0.25 M NaCl to 2 M
NaCl gradient and 10 µM zinc acetate.
5. The collodion bag is pre-equilibrated in buffer and contains a small magnetic
stirrer (5 mm × 2 mm) to aid dialysis and recovery of the protein.
6. This assay can be used to measure protein concentrations between 50 µg/mL and
50 mg/mL according to the manufacturer’s directions.
7. Data may be conveniently analyzed by linear regression using commercial soft-
ware like KaleidaGraph.
8. The protein concentration obtained by the Bradford colorimetric assay is rela-
tive. An absolute value requires multiplication by a conversion factor. This fac-
tor can be derived from simultaneous amino acid analysis or microKjeldahl
nitrogen determination (18). In the case of the TFIIIA 22-kDa fragment, the
deduced concentration of BSA is multiplied by 0.39.
9. TFIIIA and its fragments have anomalous electrophoretic mobilities. Reduction
and carboxymethylation with 50 mM iodoacetamide restores the predicted gel
mobility to the intact protein. Fragments containing zinc fingers generally run
slower than expected after this treatment.
10. Chemical detritylation can be efficiently performed by passing 0.5% trifluoro-
acetic acid for 3 min at room temperature through a reverse-phase ProRPC
HR column (Pharmacia). Following deprotection of the bases (30% [w/v]
NH
4
OH, 16 h at 65°C), the oligomer is purified with a Mono Q HR column
(Pharmacia) and eluted in a NaCl gradient with 7 M urea according to the
manufacturer’s instructions.
11. After the addition of water, the Corex tube is sealed with parafilm to avoid spillage.
The water is rolled around the walls of the silanized tube as well as on the pellet.
12. This is conveniently done in a thermal cycler PCR machine at a linear cooling
rate of 1°C/3 min.
13. A stock acrylamide solution (30:0.8 acrylamide:bis) is deionized with Amberlite
MB–150 resin (10 g/100 mL) for 1 h, filtered through a NYL/50 filter unit (0.2 µm),
and stored at 4°C. This treatment minimizes the sequestering of the protein from
protein–DNA complexes during gel electrophoresis.
14. Glass plates, combs, spacers and the gel apparatus must be cleaned and washed
completely free of detergent to avoid disruption of the protein–DNA complex.
15. The polyacrylamide gel is 1.5 mm thick and contains 50 mM NaCl, 40 mM
HEPES–NaOH (pH 7.5), and 5% (v/v) glycerol. The running buffer consists of
50 mM NaCl, 20 mM HEPES–NaOH (pH 7.5).
16. A duplex with an unrelated sequence of the same length may be used to monitor
the level of nonspecific protein binding. The affinity of the protein for each of the
single strands of the duplex can also be tested.
17. Samples can be applied smoothly to the gel by slowly winding the Pipetman
volume control back to zero.
Reconstitution of Complexes 555
18. At high salt concentration, above 0.65 M NaCl, the protein–DNA complex is
dissociated and both components are soluble. The salt concentration is lowered
stepwise to 0.225 M NaCl to reconstitute the complex (19). At the same time, the
protein is diluted and becomes less likely to aggregate or bind nonspecifically to
the duplex.
19. Wash dust off the cover slips with deionized water and dry in a suitable rack at
50°C in an oven. Immerse cover slips for 15 min in dilute AquaSil (1:40) and
leave to dry overnight. Wash in deionized water and dry at 50°C. Keep in a closed
container to avoid contact with dust.
20. It is customary to equilibrate against a well solution that contains double the
concentration of the droplet components. One volume of the well solution con-
taining appropriate salts, buffers, and precipitant is added to the droplet of the
protein–DNA complex.
21. The grease can be applied accurately to the rim with a 10cc plastic syringe filled
with high-vacuum grease.
References
1. Steitz, T. A. (1990) Structural studies of protein-nucleic acid interaction: the
sources of sequence-specific binding. Quart. Rev. Biophys. 23, 205–280.
2. Dubendorff, J. W., and Studier, F. W. (1991) Controlling basal expression in an
inducible T7 expression system by blocking the target T7 promoter with lac
repressor. J. Mol. Biol. 219, 45–59.
3. Studier, F. W., Rosenberg, A. G., Dunn, J. J., and Dubendorff, J. W. (1990) Use
of T7 RNA polymerase to direct the expression of cloned genes. Methods
Enzymol. 85, 60–89.
4. Engelke, D. R., Ng, S.-Y., Shastry, B. S., and Roeder, R. G. (1980) Specific inter-
action of a purified transcription factor with an internal control regionof 5S RNA
genes. Cell 19, 717–728.
5. Gao, Y. G., Su, S. Y., Robinson, H., Padmanabhan, S., Lim, L., MacCrary, B. S.,
et al. (1998) The crystal structure of the hyperthermophile chromosomal protein
Sso7d bound to DNA. Nature Struct. Biol. 5, 782–786.
6. Deutscher, M. P. (ed.) (1990) Guide to Protein Purification, Methods in Enzy-
mology, Academic, New York, p. 182.
7. Oliver, R. W. A. (ed.) (1989) HPLC of macromolecules: a practical approach.
IRL, Oxford.
8. Carter, C. W.,Jr. (1992) Crystallization of Nucleic Acids and Proteins: A Practi-
cal Approach (Ducruix, A. and Giege, R., eds.), IRL, NY, pp. 47–71.
9. Chen, F. E., Huang, D.-B., Chen, Y.-Q., and Ghosh, G. (1998) Crystal structure of
p50/p65 heterodimer. Nature 391, 410–413.
10. Fraenkel, E. and Pabo, C. O. (1998) Comparison of X-ray and NMR structure
for the Antennapedia homodomain–DNA complex. Nature Struct. Biol. 5,
692–696.
11. Becker, S., Groner, B., and Muller, C. W. (1998) Three-dimensional structure of
the Stat3beta homodimer bound to DNA. Nature 394, 145–151.
556 Colin and Brown
12. Chen, L., Glover, J. N. M., Hogan, P. G., Rao, A., and Harrison, S. C. (1998)
Structure of the DNA-binding domains from NFAT, Fos and Jun bound specifi-
cally to DNA. Nature 392, 42–48.
13. Batchelor, A. H., Piper, D. E., de la Brousse, F. C., McKnight, S. L., and
Wolberger, C. (1998) The structure of GABPalpha/beta: an ETS domain-ankyrin
repeat heterodimer bound to DNA. Science 279, 1037–1041.
14. Redinbo, M. R., Stewart, L., Kuhn, P., Champoux, J. J., and Hol, W. G. J. (1998)
Crystal structures of human topoisomerase I in covalent and noncovalent com-
plexes with DNA. Science 279, 1504–1513.
15. Doublie, S., Tabor, S., Long, A. M., Richardson, C. C., and Ellenberger, T. (1998)
Crystal structure of a bacteriophage T7 DNA replication complex at 2. 2Å resolu-
tion. Nature 391, 251–258.
16. Nolte, R. T, Conlin, R. M., Harrison, S. C., and Brown, R. S. (1998) Differing
roles for zinc fingers in DNA recognition: structure of a six-finger transcription
factor IIIA complex. Proc. Natl. Acad. Sci. USA 95, 2938–2943.
17. White, A., Ding, X., vanderSpek, J. C., Murphy, J. R., and Ringe, D. (1998) Struc-
ture of the metal-ion-activated diphtheria toxin repressor/tox operator complex.
Nature 394, 502–506.
18. Jaenicke, L. (1974) A rapid micromethod for the determination of nitrogen and
phosphate in biological material. Anal. Biochem. 61, 623–627.
19. Zwieb, C. and Brown, R. S. (1990) Absence of substantial bending in Xenopus
laevis transcription factor IIIA-DNA complexes. Nucleic Acids Res. 18, 583–587.
2-D Crystallization of Protein Complexes 557
557
From:
Methods in Molecular Biology, vol. 148: DNA–Protein Interactions: Principles and Protocols, 2nd ed.
Edited by: T. Moss © Humana Press Inc., Totowa, NJ
38
Two-Dimensional Crystallization
of Soluble Protein Complexes
Patrick Schultz, Nicolas Bischler, and Luc Lebeau
1. Introduction
Structural data on biological macromolecules provide invaluable insights
into the interactions of proteins with nucleic acids. Data obtained at atomic
resolution by X-ray diffraction or nuclear magnetic resonance (NMR) studies
ultimately describes the exact folding of polypeptide chains and the contacts
between proteins and DNA. However, many complexes are difficult to analyze
at atomic resolution because either they are too large or too difficult to crystal-
lize in three dimensions (3-D). Recent progress in electron microscopy, speci-
men preservation, and image processing provides the possibility to calculate
detailed molecular envelopes that are complementary to X-ray crystallography
with little theoretical limit on specimen size. In the case of membrane proteins
organized into one-molecule-thick two-dimensional (2-D) arrays, atomic mod-
els could be elaborated from electron microscopy data (1,2). It is beyond the
scope of this chapter to describe in detail the electron crystallographic methods
and computer image analysis required to calculate a 3-D model of a crystal-
lized protein complex; these aspects are described elsewhere (3). Here, we will
focus on the formation of 2-D crystals of soluble proteins, an essential
preliminary step in high-resolution structure determination by electron micros-
copy, and provide the necessary information to set up, screen, and evaluate
crystallization experiments.
The remarkable achievements in the field of membrane protein 2-D crystals
prompted the pioneering work of Kornberg and collaborators aimed at
transposing the crystallization mechanisms occurring in a lipid bilayer to
soluble proteins (4). The method consists in targeting the protein of interest to a
lipid surface self-assembled as a monomolecular film at an air–water interface.
558 Schultz, Bischler, and Lebeau
The buffer-exposed hydrophilic part of the lipid molecule may carry charged
groups (5,6) or be chemically modified (7,8) in order to interact with the pro-
tein. Consequently, the protein is concentrated at the lipid plane and adopts
only a few orientations relatively to it. If the lipid molecules are free to diffuse
in the monolayer plane, the system permits the 2-D crystallization of the mac-
romolecule. This achieved, the lipid–protein film is transferred onto a solid
support and processed for electron microscopy observations. As was demon-
strated by the study of streptavidin 2-D crystals (9), which revealed structural
information down to 3 Å in projection, such an approach can yield high-resolu-
tion structural data.
Two categories of lipid derivatives can be used according to the specificity
of the interaction to be established with the protein complex (10). On the one
hand, a charged surface can be created by using lipids with positive or negative
charges that will interact with the surface potential of the protein to be crystal-
lized (5). The characteristic behavior of the protein in ion-exchange chroma-
tography may give a hint as to which type of charged lipids should be used.
Typically used lipids are phosphatidyl serine which carries a negative charge,
and stearyl amine or alkyl trimethyl ammonium, both of which are positively
charged. On the other hand the lipid film may be derivatized by a specific
ligand recognized by the protein of interest. The ligand is chemically grafted to
the lipid moiety through a linker whose length modulates the accessibility for
the protein. The grafted molecule can be a natural ligand of the protein such as
dinitrophenol for specific anti-DNP antibodies (4), novobiocin for gyrase B
subunit (8) (Fig. 1), or biotin for streptavidin (11). More recently lipid mol-
ecules have been designed to interact with specific tags (such as polyhistidines)
introduced genetically into the sequence of the protein of interest (12,13). The
polar head group of the lipid molecule carries a nitrilotriacetate moiety, which
chelates nickel ions and interacts with the histidine tag.
Stability and fluidity of the spread lipid layer are mainly provided by the
hydrophobic part of the lipid (Fig. 1). The lipid layer has to be in the fluid
phase at the incubation temperature because crystallization of the lipid chains
was shown to prevent protein organization probably by lowering its 2-D
mobility (11,13). In most cases, a cis double bond in the alkyl chains provides
sufficient fluidity. In addition, the lateral cohesion of the lipid molecules has to
be strong enough to allow the spreading of a stable monolayer and prevent the
solubilization of the lipid as micelles or liposomes. A double alkyl chain con-
taining 18 carbon atoms (dioleyl) generally fulfills these requirements.
When protein–lipid interactions occur, the proteins are rapidly concentrated
close to the lipid layer and are likely to be partly oriented. In the case of a
specific interaction with a functionalized lipid, the macromolecules are teth-
ered to the lipid film by a unique site and the extent of orientation is deter-
2-D Crystallization of Protein Complexes 559
mined by the length and flexibility of the linker region (8) (Fig. 1). Increased
concentration, possible preferential orientation, and in-plane mobility facili-
tate contacts between macromolecules, which results in their increased organi-
Fig. 1. Schematic representation of a lipid molecule used for 2-D crystallization of
the b-subunit of DNA gyrase. The hydrophobic alkyl chain with a cis double bond
confers fluidity and stability to the spread lipid layer. The linker region provides accessi-
bility of the ligand for the protein of interest. The recognition function, here a novobio-
cin molecule, determines the interaction properties between the lipid and the protein.
560 Schultz, Bischler, and Lebeau
zation when an ordered network of interactions is established. However, a dead
end may be reached if the macromolecules interact too strongly or too rapidly
with each other. As a consequence, 2-D aggregates can form, which appear as
close-packed, noncrystalline assemblies.
The method requires limited amounts of protein because a single experi-
ment may only need 300 ng of protein. However, a quantity of 1 mg is more
realistic because any new project implies systematic trials starting without
a priori knowledge of the factors affecting the lipid–protein interactions. When
a specific interaction of the protein with the lipid is involved, the degree of
purity of the biological sample appears to be of less crucial importance than
when charged lipid are used (13,14).
2. Materials
1. A suitable lipid to interact with the protein of interest.
2. Teflon
®
or Nylon
®
blocks in which cylindrical wells 4 mm in diameter and 1 mm
deep have been milled such that each well can contain about 15 µL of aqueous
solution (Fig. 2).
3. Standard Cu or Cu/Rh 300 mesh electron microscopy grids.
4. Mica sheets 2.5 × 9 cm in size.
5. A carbon evaporator.
6. A 2% uranyl acetate solution.
7. A control electron microscope.
8. An optical diffraction bench.
3. Methods
3.1. Preparation of Electron Microscopy Supports
The lipid–protein assemblies will have to be transferred onto a standard electron
microscopy grid. The grids need to be coated with a thin hydrophobic carbon film to
support the assemblies and to allow the adsorption of the hydrophobic lipid alkyl chains.
1. The mica sheets are freshly cleaved to create a clean and flat surface.
2. The mica sheets are placed into a carbon evaporator and a 10 to 50-nm thick
carbon film is evaporated under vacuum onto the cleaved face of the sheets.
3. Electron microscopy grids are placed on a supporting filter paper below the sur-
face of a water bath. The size of the filter paper matches that of the mica sheet
and will hold a total of 75–100 grids.
4. The carbon foil is floated at the clean air–water interface by dipping the mica
sheet in the water bath with an angle of about 30°.
5. Finally, the carbon foil is gently lowered onto the grids by removing the water
using a vacuum pump.
It was observed with some systems, such as streptavidin, that the contact of
the lipid layer with the carbon foil interferes with the quality of protein arrays
2-D Crystallization of Protein Complexes 561
(15). “Holey” carbon grids can then be used to transfer the crystals without
interactions with a carbon surface, the lipid–protein layer being spread over
the holes. A protocol to prepare “holey” carbon films is as follows (16):
1. The surface of an optical microscope glass slide is extensively cleaned by boiling
in an aqueous detergent solution and extensive rinsing with demineralized
water (H
2
Od).
2. The slide is immersed in a 0.1% Triton X405 (Sigma) solution for 30 min, briefly
rinsed with H
2
Od to remove the excess of detergent and left to dry. This will
result in a clean hydrophobic surface.
3. The slide is placed on a precooled aluminum block to allow minute water drop-
lets to form on the surface by condensation. The size of the droplets depends on
the humidity of the room and on the condensation time.
Fig. 2. Design of a Teflon
®
block for 2-D crystallization experiments. (A) A Teflon
®
cylinder 4 cm in diameter truncated into 1 cm-thick slices and 16 wells, 4 mm in
diameter and 1 mm deep, are milled into the block such that each well can contain
about 15 µL of solution. (B) During the crystallization experiments performed in the
wells (a), the Teflon
®
block (b) is placed into a humid chamber consisting of a reverted
Petri dish containing some buffer (c).
562 Schultz, Bischler, and Lebeau
4. One mL of a cellulose acetate or cellulose butyrate solution (0.4% [w/v] in ethyl
acetate) is poured with a pipet over the surface, excess solution is removed from
one end of the slide by touching a filter paper and left to dry. Upon drying, the
cellulose forms a thin film around the water droplets, thus forming holes.
5. At this point, an optical microscope can be used to check the size of the holes and
their distribution.
6. The slide is then immersed 30 min in a 0.5% (w/v) sodium dodecyl sulfosuccinate
solution to peal off the cellulose film.
7. Electron microscopy grids are deposited on a supporting filter paper just below
the surface of a water bath.
8. The holey cellulose film is floated on the clean air–water interface and deposited
onto the grids.
9. A 50-nm-thick carbon film is evaporated onto the cellulose-coated grids.
10. The cellulose film is finally dissolved by placing the grids on a ethyl acetate-
soaked filter paper.
3.2. Crystallization Experiments
1. The lipids are best stored as a dry powder under an argon atmosphere at –80°C. A
mother solution at a concentration of 10 mg/mL is produced by solubilizing the
lipids in an organic solvent such as a 1:1 mixture of chloroform:hexane. This
solution can be stored under argon up to 1 yr at –20°C. The working lipid solu-
tion is at a concentration of 0.5–1 mg/L in an organic solvent. All solution are
stored in 2 mL glass vials with Teflon
®
caps to prevent solvent evaporation.
2. The Teflon
®
wells have to be cleaned prior to use for crystallization experiments
to remove residues of proteins or lipids. The Teflon
®
support should be dipped
into a sulfochromic acid solution for 1 h, rinsed 10 times with H
2
Od, dipped for
1 h into H
2
Od, and rinsed again three times with H
2
Od. Alternatively, the support
can be rinsed 10 times with methanol to eliminate proteins, 10 times with a
chloroform:methanol 2:1 or hexane:methanol 9:1 solution to remove lipids and
rinsed 10 times with hexane to remove fatty acids. Finally, the support is brought
into contact with a filter paper to remove the excess of H
2
Od or organic solvent,
without wiping to avoid electrostatic charging, and is allowed to dry in a dust-
free chamber.
3. Incubations are performed in a humid chamber to prevent buffer evaporation
(Fig. 2B). The Teflon
®
block is placed in a reverted Petri dish containing some
buffer with an opening in the top to let air in and out during removal of the lid.
4. In each well, 10 µL of buffer are added (Fig. 3B).
5. To spread the lipid at the air-buffer interface, 1 µL of the lipid solution at a con-
centration of 0.5–1 mg/mL is placed on the top of the drop of buffer with a
micropipet. At this moment it can be observed that the surface tension of the drop
is released (Fig. 3C).
6. The organic solvent is allowed to evaporate for 5 min.
7. The protein solution (5 µL) is injected into the aqueous phase (Fig. 3D). The
final protein concentration is generally set between 20 and 200 µg/mL.
2-D Crystallization of Protein Complexes 563
8. The incubation chamber is closed, and if oxidation is a problem air is replaced by argon.
9. The incubation time will vary from one system to another but is generally in the
range of 1–36 h. Most experiments can be performed at room temperature, but
longer incubation times at 4°C may improve crystal quality in some cases.
3.3. Electron Microscopy
1. The 2-D crystal is transferred to the electron microscopy grid through hydropho-
bic contacts between the lipid chains and the carbon foil (Fig. 3E). This is simply
done by placing the grid over the well for 1–2 min. The grid is then withdrawn
and prepared for observation (see Notes 1 and 2).
2. To be visualized by electron microscopy, the specimen has to be contrasted by
creating a mould of heavy atoms around the proteins, a process named negative
Fig. 3. Setup of a crystallization experiment. In each well (A), a 10-µL drop of
buffer is placed (B). Because the Teflon well is hydrophobic, the drop does not wet the
surface. Upon the addition of 1 µL of the lipid solution at a concentration of 0.5–1 mg/mL,
the surface tension of the drop is released (C). After evaporation of the organic sol-
vent, 5 µL of the protein solution is injected into the well (D) and is allowed to interact
with the lipid layer. The resulting lipid–protein assembly is transferred to a carbon-
coated electron microscopy grid placed on the top of the drop (E).