<<

. 25
( 61 .)



>>

Bacterlostatlc agents (sodmm azlde)
Detergents octyl+glucoslde, CHAPS (3-[(3-cholamldopropyl)
dlmethylammomo]- 1-propane-sulfonate)
Polyvalent cations and amons, such as spermine or spermidine, Mg*+, Ca*+,
[CoO™W,13+, Eu3+

Table 2
Frequently Used Buffers for Crystallization Trials
Buffering
Buffer range, pH
Na acetate 3.6-5 6
Sodium/potassium phosphate0 5.0-8.2
5.2-7 2
Sodium cacodylate
Bzs-tns: bu[2-hydroxyethl]imlnotrrs[hydroxymethyl]-methane 5 8-7.2
PIPES plperazine-N,N™-bu[2-ethanesulfonic acid] 6 l-7.5
MOPS 3-[N-morpholmo]propanesulfomc acid 6 5-7 9
HEPES. N-[2-hydroxyethyl]piperazine-N™-[2-ethanesulfonlc acid] 6 8-8.2
Tris: tns[hydroxymethyl]ammomethane 70-90
CAPS* 3-[cyclohexylaminol-l-propanesulfomc acid 97-11 1
“Forms morgamc crystals with traces of barium or calcium, so care must be taken when
choosing additives.
159
Luisi, Anderson, and Hope
160


droplet suspended from silinated
silicone grease seal




reservoir, containing buffered
precipitating agent




Hanging droplet




droplet of protein resting on
tbe top of a column




reservoir, containing buffered
precipitating agent




Sitting droplet

Fig. 1. Schematic diagram of vapor diffusion crystallization setup. A droplet contain-
ing the protein solution and %-% vol of the external reservoir is suspended from a
silinated cover slip over a reservoir containing buffered precipitating agent. Water mol-
ecules and other volatile species will exchange between the droplet and reservoir and, at
equilibrium, their chemical activities in the two compartments will become equal.

tallization kits are commercially available (Hampton Research) and useful for
quickly surveying many conditions. Crystallization trials using vapor diffu-
sion crystallization can be undertaken with disposable plastic tissue-culture
plates (Limbo model FB16-24-TC) that have 24 wells (2 cm diameter x 2 cm
deep) with flat rims. Circular glass cover slips with 2 cm diameter are required
for suspending a droplet of protein solution over the reservoir. Application of
silicone vacuum grease around the rim of each well enables airtight sealing
(see Fig. 1). Pretreat the cover slips with silination solution (3% by volume of
dimethylsilane in dichloromethane) for one-half hour and then air-dry in a fume
hood after washing briefly with dichloromethane. Water will not adhere to
properly silinated slides. The cover slips can be easily cleaned before use by
immersing in water.
161
Crystallization of Macromolecules
2.4. Quantitation of Protein Concentration for Crystallization
Commercial kits are available that use the Bradford, bicmchoninic acid
(BCA), or Lowry assay,and these have different degrees of sensttivity, depend-
ing on the sample. As described m Section 3., UV spectrophometry is also a
convenient means of qwantitation.
2.5. Two-Dimensional Crystallizations
for Electron Microscopy
Prepare fresh solutions of a ltpid mixture contammg 0.05 mg/mL of
octadecylamine and 0.45 mg/mL of egg L-a-phosphatidylcholine in 1: 1 chloro-
form/hexane (5). Store anaerobically at -20°C. Two-dimensional crystals can
also be prepared on mica, and good-quality mica sheets are required. Use 1%
weight uranyl acetate solution for negative staining to evaluate the trials by
electron microscopy.
3. Met hod+s
3. I. Growing Crystals: Introduction
Protein crystals are most likely to grow under conditions where the effective
protein concentration exceeds its solubility limit. This condition is known as
supersaturation, and it is a nonequilibrium state.Crystallization conditions lurk
where the system relaxes to equilibrium, and the protein molecules are excluded
into the solid state.The most common strategy in growing crystals is to achieve
supersaturation by modifying the solvent with precipitating agents, adjusting
the pH, or by altering a physical property, such as the temperature.
Using light-scattering measurements to infer particle sizes,it has been exper-
imentally observed that crystals are most likely to grow when the precipitating
agent does not cause the protein molecules to aggregate randomly (6) For
instance, lysozyme remains monodispersed over a high concentration range of
sodium chloride, but in the presence of ammonium sulfate, it aggregates. This
observation appears to explain why lysozyme crystals can be readily prepared
from solutions of sodium chloride, but not from ammonium sulfate (see Sec-
tion 3.5.). Therefore, in preparing crystals, one is ideally searching for condi-
tions of supersaturation without aggregation, where nucleation and then crystal
growth will occur. If, however, the system is pushed to supersaturation under
conditions where aggregation results, only noncrystalline precipitate may develop.
To achieve supersaturation, it is necessary to screen a range of conditions,
including variations in concentration of precipitating agents, pH, protein con-
centration, and temperature (given in rough order of importance for the typical
case). Summaries of suitable precipitating agents and biological buffers for
crystallization trials are presented m Tables 1 and 2. The number of possible
Luisi, Anderson, and Hope
762
combmations of crystalhzatton condttions is immense; however, the supply of
protein is usually limited. One would ideally like to cover as wide a range of
conditions as possible in the Initial searches,but still be able to correlate trends
between trials for subsequent crystallization screens. The best strategy ts to
strike a balance between these requirements, and thts approach is known as
factorial analysis (7) A number of general crystalhzattonsetshave been designed
in part on this prmciple and also on the use of aposterlorr mformation from the
generally most successful crystallizatron conditions (8-11). In these general
screens, a carefully selected, diverse set of the most successful conditions is
chosen as the preliminary crystallrzation screen
3.2. Effectors of Crystal Growth
3 2. I. Preclpltatmng Agent
The most common procedure to achieve supersaturation is to change the
concentratton of salts gradually. At very low tonic strength, protein solubtlity
1slow. Indeed, some proteins crystallize m distilled water as they are being
concentrated. For most proteins, the solubthty becomes greater with increasing
ionic strength, a phenomenon known as “salting-in.” However, the solubihza-
non occurs up to a point when a reverse solubtlity effect known as “saltmg-
out” is manifested, and both salt ions and protein molecules compete for
hydration structures, which are required to maintain then solubility Further-
more, the high iomc strength of the solution may shield unfavorable electro-
static interactions between protein molecules, thereby permtttmg assoclatton.
It IS important that a large range of salts be used m attempting to crystallize
a macromolecule, since, m addttton to saltmg-out effects, there may also be
specific protein-ion mteracttons that may have secondary consequences. Poly-
meric alcohols, such as the range of polyethylene glycols, are very popular as
effective crystallizattons agents. Organic solvents can also be used to induce
crystalltzatton, although they are not as popular as salts and polyethylene gly-
col. Proton concentratton can have a tremendous effect on crystal growth. Solu-
b&y tends to be minimal at the tsoelectric point, but not all proteins crystallize
at the correspondmg ˜1. Lattice contacts might be made by salt bridges, which
can be affected by pH
3.2.2. Temperature
Temperature can have a strong effect on crystal growth owing to its mflu-
ence on protein solubihty. Proteins can have either negative or positive solu-
bihty coefficients with temperature. For instance, msulm crystals can be grown
by an inverse temperature gradient (for example, see ref. 12). If there is suffi-
cient material available, crystallization trays should be set out at least in duph-
163
Crystallization of Macromolecules
cate, with one tray being kept at room temperature and the second maintamed
at 4OCin a cold room Optimally, a bank of controlled temperature mcubators
should be used at as many different temperatures as possible. Precisely because
of the effect of temperature on crystallization, it is a good pohcy to mamtain
the crystalbzation trays at a constant temperature, preferably by usmg a well-
insulated container or a controlled temperature incubator.
3.2.3. Additives
Divalent cations and spermme are proven crystallization agents for nucleic
acids (4,6), Detergents are required for some proteins, such as membrane pro-
ten-is.In some cases,proteins or protein-DNA complexes can be cocrystalhzed
with a heavy atom, such as lanthamde salts (13)
3.3. Protein Quantification
Protein concentration can be conveniently quantified using the Bradford,
BCA, or Lowry assay,and these have different degrees of accuracy, depending
on the sample. Amino acid composition analysis is a reliable method for quan-
titating protein concentration, providing that accurate amino actd standards are
available. UV spectroscopy IS yet another convenient method of estrmatmg
protein concentratton, and the relationship between concentratton vs absorp-
tion is roughly linear up to about two optical units. The protem concentration is
estimated from the absorbance at 280 nm usmg the calculated extmction coef-
ficient based on the composition of the aromatic ammo acids. This estimation
ISusually made more accurate by using a buffer containing a denaturmg agent,
such as 6M guamdmmm hydrochloride, Spectroscopic measurement is also a
convement means of quantifying nucleic acid concentration, again using cal-
culated extinction coefficients based on the base composition. Empirical cor-
rections for hypochromic effect must be made, usually by heating the sample
or digesting with nuclease.
3.4. Crystal Preparation by Vapor Diffusion
The most popular method of growing crystals is by vapor diffusion, where
the protein solution is mixed with buffered solution of the precipitatmg agent,
and the mixture is permitted to equilibrate against a large reservoir containing
buffered precipitatmg agent at a higher concentration. There are two methods
of achieving vapor diffusion: hanging droplets and sitting droplets (Fig. 1). In
the hanging droplet method, a microdroplet of protein (as little as 2 pL) is
placed on a stlinated microscope cover slip, which is then inverted and placed
over a well containing 1 mL of buffered precipitating solution. It is important
that the cover slip is silinated, since this ensures proper drop formation and
prevents droplet spreading. A large number of conditions can be screened by
Luisi, Anderson, and Hope
164
this method using only a small amount of protein. Disposable plastic tissue-cul-
ture plates that have 24 wells (2 cm diameter x 2 cm deep) with flat rims enable
airtight sealmg by application of silicone vacuum grease around the ctrcumfer-
ence of each rim. These plates provide a further advantage in that they can be
easily examined under a dissectmg microscope and allow compact storage.
The design of the sittmg droplet is similar to the hanging droplet, except that
the drop sits on an elevated table above the precipitatmg agent (Fig. 1). The
sitting droplet approach is advantageous in cases where the protem droplet
spreads extensively on coverslips or when detergents are bemg used in crystal-
lizations, since these cause droplet spreading.
3.5. Crystal Preparation by Dialysis
A macromolecule may be guided gradually toward crystalhzation by dtaly-
sis against a solutton of buffered precipitating agent. This method has the
advantage that the rate of equilibration can be controlled by adjusting the con-
ditions of the external side of the membrane. If the concentration of precipitant
is too high and an amorphous preciptate results rather than crystals, the sample
may be redissolved and new conditions established simply by adjustmg the
external solution.
Various methods have been designed for microdialysis cells or vessels. In
most cases, only l&20 pL of protein solution are injected mto a short glass
capillary or tube. The vessel is then sealed at one end, and the other 1scovered
with a dialysis membrane. The entire assembly is submerged m a container hold-
mg the solution of buffered precipitating agent. Gradual equihbration then occurs
through the membrane. One convenient aspect of this procedure is that the pro-
tein can be maintained under anaerobtc conditions. The droplets can be set out
under nitrogen with reducing agent in the degassedbuffer.
3.6. Crystal Nucleation by Microseeding
Microseeding is used to induce directed crystal growth to increase crystal
size or improve crystal quality. A supersaturated protem solution that contains
small crystals from earlier trials may be dtluted into a fresh solution that is only
shghtly supersaturated, so that slow growth of crystals will occur. One prob-
lem with seedmg with microcrystals is that, if too many crystals are induced
into the fresh supersaturated solution, twined or poorly formed crystals may
form that are unsuitable for data collection.
3.7. Practical Example: Growing Crystals of Lysozyme
We recommend that the hopeful crystallographer practice the hanging drop-
let method by crystallizing hen egg lysozyme, which will yield beautiful crys-
tals m l-2 d. These crystals can be used to practice some of the more difficult
165
Crystallization of Macromolecules

manipulations, such as mounting crystals in caplllarles for data collection or
flash-freezing in liquid nitrogen.
1. Preparea stock buffer solution of 0.M Na acetate, pH 4.7.
2. Preparea stocksolution of 50 mg/mL hen eggwhite lysozyme(use good-quality
protein). Spin in an Eppendorfcentrifuge at 4OCto remove any insolublematerial.
3. Preparea tray in which the precipitatingagent,NaCl, variesm the wells from 24%
wt/vol m the Na actetate buffer. Apply siliconegrease the rim of the reservoir
to
4 On a silinated cover slip, mix 5 pL of the lysozymesolutton with 5 pL of the
preclpltating solution from the reservoir. Carefully invert the cover slip and seal
againstthe rim of eachreservoir
5. Leave undisturbed m an Insulated storagebox at room temperatureor m a tem-
perature-controlled incubator at 25™C for l-2 d.
3.8. Evaluation of Crystals
The best means of examining crystallization trials IS with a good-quality
dissecting microscope. A convenient method of vlsuallzmg the smallest micro-
crystals (I.e., crystals with longest dimensions of 5 pm) is with a set of polariz-
mg filters. Because crystals consist of ordered chn-al molecules, hght that
passesthrough them will become plane-polarized. (This will occur for all crys-
tals, except those with cubic symmetry.) Consequently, light that 1s already
plane-polarized will change its polarization state on passmg through a protein
crystal. The field of view will be perceived to change from brightness to dark-
ness as one of the filters IS rotated through 90”. Microcrystal that may be too
small to vrsuallze directly under the dissecting mtcroscope may be seen as a
“glowing” effect under rotating polarizers. Inexpensive polarizers may be con-
structed from a sheet of Polaroid filter, placing one between the lllummatmg
light source and specimen, and the second between any of the objective lenses.
One of the most dlsheartenmg aspects of crystallizatlons trials is that mor-
ganic salt crystals can often be mistaken for protein crystals. There are a num-
ber of simple tests to check if crystals are composed of protein or salt. Protein
crystals tend not to have the same mechanical rigidity as salt crystals and ˜111

<<

. 25
( 61 .)



>>