attempted using as homogeneous a preparation as possible, such preparations having
a greater chance of yielding crystals than material that contains impurities. Because
of our inadequate understanding of the physical processes involved in crystallisation,
methods for growing protein crystals are generally empirical, but basically all
involve varying the physical parameters that affect solubility of the protein–for
example pH, ionic strength, temperature, presence of precipitating agents–to produce
a state of supersaturation. The process involves extensive trial and error to find a
procedure that results in crystals for a particular protein. Initially this involves a
systematic screen of methods to identify those conditions that indicate crystallinity,
followed by subsequent experiments that involve fine-tuning of these conditions.
Basically, nucleation sites of crystal growth are formed by chance collisions of
molecules forming molecular aggregates, and the probability that these aggregates will
occur will be greater in a saturated solution. Clearly, to produce saturated solutions,
tens of milligrams of proteins are required. This used to represent a considerable
challenge for other than the most abundant proteins, but nowadays genetic engineering
methodology allows the overproduction of most proteins from cloned genes almost
on demand. The following are some of the methods that have proved successful.
(a)Dialysis. A state of supersaturation is achieved by dialysis of the protein solution
against a solution containing a precipitant, or by a gradual change in pH or ionic
strength. Because of frequent limitations on the amount of protein available, this
approach often uses small volumes ( >50 mm^3 ) for which a number of
microdialysis techniques exist.
(b)Vapour diffusion. This process relies on controlled equilibration through the
vapour phase to produce supersaturation in the sample. For example, in the hanging-
drop method, a microdroplet (2–20 mm^3 ) of protein is deposited on a glass coverslip;
then the coverslip is inverted and placed over a sealed reservoir containing a
precipitant solution, with the droplet initially having a precipitant concentration
lower than that in the reservoir. Vapour diffusion will then gradually increase the
concentration of the protein solution. Because of the small volumes involved this
method readily lends itself to screening large numbers of different conditions.
When produced, crystals may not be of sufficient size for analysis. In this case
larger crystals can be obtained by using a small crystal to seed a supersaturated
protein solution, which will result in a larger crystal.- Once prepared, the crystal (which is extremely fragile) is mounted inside a quartz
or glass capillary tube, with a drop of either mother liquor (the solution from which it
was crystallised) or a stabilising solution drawn into one end of the capillary tube to
prevent the crystal from drying out. The tube is then sealed and the crystal exposed
to a beam of X-rays. Since the wavelength of X-rays is comparable to the planar
separation of atoms in a crystal lattice, the crystal can be considered to act as a
three-dimensional grating. The X-rays are therefore diffracted, interfering both in
phase and out of phase to produce a diffraction pattern as shown in Fig. 8.4.
Data collection technology necessary for recording the diffraction pattern is now
highly sophisticated. Originally, conventional diffractometers and photographic film
were used to detect diffracted X-rays. This involved wet developing of the film and
subsequent digital scanning of the negative. Data collection by this method took
many weeks. By contrast, modern area-detectors can collect data in under 24 h.
337 8.4 Protein structure determination