Chapter 2

Protein Crystallization
Alexander McPherson

Abstract
Protein crystallization was discovered by chance nearly 200 years ago and was developed in the late nine-
teenth century as a powerful purification tool, and a demonstration of chemical purity. The crystallization
of proteins, nucleic acids, and large biological complexes, such as viruses, depends on the creation of a
solution that is supersaturated in the macromolecule, but exhibits conditions that do not significantly
perturb its natural state. Supersaturation is produced through the addition of mild precipitating agents
such as neutral salts or polymers, and by manipulation of various parameters that include temperature,
ionic strength, and pH. Also important in the crystallization process are factors that can affect the struc-
tural state of the macromolecule, such as metal ions, inhibitors, cofactors, or other conventional small
molecules. A variety of approaches have been developed that combine the spectrum of factors that effect
and promote crystallization, and among the most widely used are vapor diffusion, dialysis, batch, and
liquid–liquid diffusion. Successes in macromolecular crystallization have multiplied rapidly in recent years
due to the advent of practical, easy-to-use screening kits, and the application of laboratory robotics.

Key words Crystals, Supersaturation, Growth mechanisms, Homogeneity, X-ray diffraction,

Precipitants, Crystallization methods, Vapor diffusion, Dialysis, Mother liquor

1 Introduction

Although the technologies of nuclear magnetic resonance and,

more recently, cryogenic electron microscopy, have made signifi-
cant inroads, presently the only technique that can yield atomic
level structural images of biological macromolecules is X-ray dif-
fraction analysis as applied to single crystals. While other methods
may produce important structural and dynamic data only X-ray
crystallography is adequate to precisely define atomic coordinates.
The application of X-ray crystallography is absolutely dependent
on crystals of the macromolecule, and not simply crystals, but crys-
tals of sufficient size and quality to permit accurate data collection.
The quality of the final structural image is directly determined by
the perfection and physical properties of the crystalline specimen.
The crystals, therefore, become the keystone element of the entire
process, and the ultimate determinant of its success. The crystals

Alexander Wlodawer et al. (eds.), Protein Crystallography: Methods and Protocols, Methods in Molecular Biology, vol. 1607,
DOI 10.1007/978-1-4939-7000-1_2, © Springer Science+Business Media LLC 2017

17
18 Alexander McPherson

themselves have no medicinal or pharmaceutical value, but provide

the X-ray diffraction patterns that serve as the fundamental data,
which through Fourier synthesis, allow the direct visualization of
the macromolecules or their complexes composing the crystals.
When crystallizing proteins for X-ray diffraction analysis, one
is usually dealing with homogenous, often exceptionally pure mac-
romolecules, and the objective is to grow only a few large, perfect
crystals. The proteins themselves may be purified from natural
sources, microbes or tissues of plants and animals, or it may be
produced by recombinant DNA techniques. The number of crys-
tals needed for recording data may be few, but often the amount of
protein available is severely limited. This in turn places constraints
on the approaches and strategies that can be used to obtain those
crystals. While new methodologies such as synchrotron radiation
[1, 2] and cryocrystallography [3–5] have driven the necessary size
of specimen crystals consistently downward, they have not elimi-
nated the need for crystal perfection.

2 The Nature of Protein Crystals

Protein crystals are composed of approximately 50% solvent on

average. Those seen in Fig. 1 vary from 33% solvent for monoclinic
lysozyme up to 61% for concanavalin B. At the extremes one finds
insulin at about 25% and tropomyosin at 90%. Protein occupies the
remaining volume so that the entire crystal may be thought of as
an ordered gel permeated by extensive networks of channels and
interstitial spaces filled with solvent, through which small mole-
cules can diffuse. There does not appear to be a direct correlation
between the solvent volume of a protein crystal and its diffraction
properties. It has, however, been noted that transitions of a crystal-
lographic unit cell to smaller volume, with concomitant reduction
of included solvent, has frequently produced an improvement in
diffraction resolution [6, 7].
In proportion to molecular mass, the number of contacts (salt
bridges, hydrogen bonds, hydrophobic interactions) that a con-
ventional organic molecule forms with its neighbors in a crystal far
exceeds the very few exhibited by crystalline macromolecules.
Since these contacts provide the lattice interactions essential for
crystal integrity, this largely explains the differences in properties
between crystals of salts or small molecules and macromolecules. It
may also explain why the introduction of a few additional contacts,
or even one uniquely strong interaction, can profoundly affect the
diffraction resolution of a protein crystal.
Living systems are based almost exclusively on aqueous chem-
istry within narrow ranges of temperature and pH. Macromolecules,
thus, have evolved an appropriate compatibility and dependency.
Serious deviations or perturbations are rarely tolerated. As a conse-
quence, all protein crystals are grown from aqueous media, ones to
 Protein Crystallization 19

Fig. 1 An array of protein crystals showing the range of habits they may assume: in (a) thaumatin, (b) bovine
trypsin, (c) tetragonal lysozyme, (d) monoclinic lysozyme, (e) beef liver catalase, (f–h) three different crystal
forms of bovine RNase S, (i) beta-lactoglobulin, (j) concanavalin B, (k) satellite tobacco mosaic virus, (l) glucose
isomerase, (m) concanavalin A, (n) rhombohedral canavalin, and (o) orthorhombic canavalin

which they are tolerant, and these solutions are called mother
liquors. As described below, crystals can be made to grow from
these mother liquors when the mother liquors are made supersatu-
rated in protein.
There are important physical and chemical differences between
ionic crystals, or those of most low-molecular-mass compounds,
and crystals of proteins. For example, protein crystals generally
have fairly simple morphologies, or habits as they are called, while
conventional crystals often display very complex polyhedral or pris-
matic appearances. This is mainly due to the absence of centers of
symmetry, mirror planes, and glide planes in protein crystals.
Proteins exist in only one enantiomeric form and, therefore, ­cannot
have such symmetry elements in their space groups. As a further
consequence, protein crystals can fall into only 65 space groups
rather than the 230 space groups allowed mixtures of enantiomers,
and these 65 tend to have rather simple point group symmetries
that are reflected in the habits.
20 Alexander McPherson

Conventional crystals are characterized by firm lattice interac-

tions, are usually well ordered, physically hard and brittle in general,
relatively easy to manipulate, usually can be exposed to air, have
strong optical properties, and diffract X-rays intensely.
Macromolecular crystals are by comparison usually more limited in
size, are very soft and crush easily, disintegrate if allowed to dehy-
drate, exhibit weak optical properties and diffract X-rays poorly.
Protein crystals are temperature sensitive and undergo extensive
damage after prolonged exposure to radiation. Frequently, several
crystals must be analyzed for a structure determination to be suc-
cessful although the advent of cryocrystallography [3–5, 8] pixel
area detectors of very high photon counting efficiency [9], high
intensity synchrotron X-ray sources [1, 8], and new phasing meth-
ods [10] have greatly lessened this constraint. Those same advance-
ments have also reduced the size (volume) of crystals useful for
X-ray diffraction analysis. Until the 1990s, crystals in the range of
dimensions 0.25–1.0 mm were commonly required. Currently,
structures can be determined from crystals in the range of 20–50 μm.
The extent of the diffraction pattern from a crystal is directly
correlated with its degree of internal order. The more vast the pat-
tern, or the higher the resolution to which it extends, the more
structurally uniform are the molecules in the crystal and the more
precise is their periodic arrangement. The level of detail to which
atomic positions can be determined by crystal structure analysis in
turn corresponds closely with that degree of crystalline order. While
conventional crystals often diffract to their theoretical limit of reso-
lution, protein crystals, by comparison, produce diffraction patterns
of more limited extent. Protein crystals, all crystals in fact, are not
uniform, flawless solids, but exhibit many defects and dislocations
that produce a mosaic pattern of slightly misaligned sectors, or
domains. Domain boundaries, often refered to as stacking faults or
grain boundaries in conventional crystals, are far more numerous in
protein crystals than conventional crystals, probably by several orders
of magnitude [11]. These features contribute further to the limita-
tion of diffraction quality. Some defects seen in protein and virus
crystals by atomic force microscopy (AFM) are presented in Fig. 2.
The liquid channels and solvent filled cavities that permeate
macromolecular crystals and the lack of order they engender are
primarily responsible for the limited resolution of the diffraction
patterns. Because of the relatively large solvent spaces between
adjacent molecules and the consequent weak lattice forces, all mol-
ecules in the crystal may not occupy exactly equivalent orientations
and positions but may vary slightly within or between unit cells.
Furthermore, because of their structural complexity and their
potential for conformational dynamics, protein molecules in the
aqueous environment of a crystal may exhibit slight variations in
the course of their polypeptide chains or the dispositions of side
groups from one to another.
 Protein Crystallization 21

Fig. 2 Planar defects (stacking faults) in crystals of proteins and viruses. (a), (c), (d), and (e) are surfaces of
satellite tobacco mosaic virus crystals, (b) canavalin and (f) a crystal of cucumber mosaic virus. The planar
defects, homologous to grain boundaries in conventional crystals, divide the crystal into domains, which in turn
are responsible for the mosaicity of the crystals

Although the presence of extensive solvent regions is a major

contributor to the generally modest diffraction quality of protein
crystals, it is also largely responsible for their value to biochemists
as platforms for expermentation. Because of the high solvent con-
tent, the individual macromolecules in protein crystals are sur-
rounded by layers of water that maintain their structure virtually
unchanged from that found in solution. As a consequence, ligand
binding, enzymatic activity, spectroscopic characteristics, and most
other biochemical features are essentially the same as for the fully
solvated molecule. Conventional chemical compounds, which may
be ions, ligands, substrates, coenzymes, inhibitors, drugs, or other
effector molecules, may be freely diffused into and out of the crys-
tals. Crystalline enzymes, though immobilized, are frequently
accessible for experimentation simply through alteration of the
surrounding mother liquor.
Figure 3 shows, in a representative manner, how small organic
molecules diffuse into a protein crystal through its network of sol-
vent channels. The blue dye xylene cyanol, a molecule in the
22 Alexander McPherson

Fig. 3 A large crystal (about 1.5 mm in length) of the protein canavalin has had its mother liquor replaced with
an equivalent mother liquor containing the blue dye xylene cyanol. This series of photographs taken over about
8 h shows the diffusion of the dye molecules into the protein crystal

molecular weight range of a biological coenzyme or a possible

drug, was added directly to the mother liquor of a large canavalin
crystal. The dye front, as it diffuses into the crystal is clearly evi-
dent, and its progress could be recorded and measured. From this
it could be estimated that the dye, the small molecule, diffused
through the crystal lattice at a rate of about 60 μm/h.
Figure 4 illustrates another experiment where a large thauma-
tin crystal was saturated with the pH sensative dye m-cresol purple
at high pH (pH 8) giving it a blue color. The mother liquor was
then replaced with with an equivalent one but at low pH (pH 6).
As H3O+ ions diffused into the crystal, the dye internal to the crys-
tal changed to a yellow color. Again, the dye transition front and
its movement through the crystal was photographically recorded
and measured. From this experiment it could be estimated that
when a gradient of H3O+ of 10−8 > 10−6 exists between the interior
of the crystal and its mother liquor, H3O+ ions diffuse to the center
of the crystal with an average rate of about 1000 μm/h.
A diversity of crystallographic unit cells and habits that we refer
to as polymorphism are common phenomena with macromolecular
 Protein Crystallization 23

Fig. 4 A crystal of thaumatin was saturated with the pH sensitive dye

m-­bromocresol purple at pH 8 where the dye is blue. The mother liquor was then
replaced with an equivalent mother liquor at pH 6. As the H3O+ ions diffused into
the crystals, the internal m-bromocresol purple dye molecules changed color to
yellow. The experiment, photographed over a period of about 35 min, shows the
progress of the diffusion of the H3O+ ions into the protein crystal

crystals. In Fig. 1 alone we see crystals with three different unit cells
of RNase S, two of both lysozyme and canavalin. Glucose isomer-
ase, catalase, and trypsin can also be induced to crystallize in addi-
tional unit cells. Presumably this is a consequence of the protein’s
conformational dynamic range and the sensitivity of the lattice con-
tacts involved. Thus, different unit cells and different symmetries
may arise under what, by most standards, would be called identical
conditions. In fact, multiple crystal forms are s­ ometimes seen coex-
isting in the same sample of mother liquor as in Fig. 5.

3 Energetics, Kinetics, and Mechanisms of Protein Crystallization

There are further differences that complicate the crystallization of

proteins as compared with conventional, small molecules [12–18].
First, proteins may coalesce to form several solid or dense liquid
states that include amorphous precipitates, oils, or gels, as well as
crystals, and most of these other forms are kinetically favored as
supersaturation rises. Second, unlike most conventional crystals,
24 Alexander McPherson

Fig. 5 In this vapor diffusion droplet containing sodium nitrate and a trace amount
of sodium chloride, the protein lysozyme has crystallized in two distinctly differ-
ent crystal forms. The large cluster of thin laths is the monoclinic crystal form,
while the many smaller, darker crystals are of tetragonal symmetry

protein crystals nucleate, or initiate development, only at very high

levels of supersaturation, often two to three orders of magnitude
more than required to sustain crystal growth. This often leads to
massive showers of microcrystals, or even more often, precipitate.
Further, the kinetics of macromolecular crystal nucleation and
growth are generally two to three orders of magnitude slower than
for conventional molecules [19–23]. The latter difference arises
from the considerably larger size, lowered diffusivity, and weaker
association tendencies compared with small molecules or ions, as
well as a lower overall probability of incorporation of an incoming
protein molecule into a growth step [24].
Crystals, including protein crystals, grow by successive layer
addition [18, 22, 23, 25]. The rate limiting step in crystal growth is
not, however, the completion of an active layer by recruitment of
molecules from solution into growth steps and kinks at the edges of
expanding layers (refered to as tangential growth), as this is energeti-
cally favorable and rapid [25, 26]. The rate limiting step in crystal
growth is the initiation of new, superior growth layers. This is more
demanding in terms of self organization, less probable, and by far the
slower process. It is refered to as growth in the normal direction.
There are four mechanisms that have been described for pro-
tein crystals to provide growth in the normal direction. These were
deduced by application of AFM to actively growing crystals [27].
 Protein Crystallization 25

The various mechanisms have been treated in detail elsewhere [23,

27–30], but two of the four predominate. At higher levels of super-
saturation the principal mechanism, illustrated by the examples in
Fig. 6, is layer initiation by two dimensional nucleation on existing
crystal surfaces. Like the formation of a three dimensional critical
nucleus (see below), this requires molecules otherwise free in solu-
tion to self organize on a surface to form a small, crystallographi-
cally ordered array that may then expand by tangential growth. The
major difference between the formation of three and two dimen-
sional nuclei is that in the latter case the molecules are confined to
a surface. This restriction of their freedom encourages their associa-
tion. In addition, their self organization is guided in an epitaxial
manner by the molecular lattice of the underlying, existing layer.

Fig. 6 A major source of growth steps and layers on the surfaces of growing macromolecular crystals, particu-
larly at medium to high levels of supersaturation, are two dimensional nuclei that exceed critical nuclear size
and subsequently develop into two dimensional islands. Shown here are two-dimensional islands on a variety
of protein and virus crystals. This is the dominant mechanism for face normal growth for most macromolecular
crystals. In (b) the arrow denotes a triangular nucleus that reflects the symmetry of the crystal face
26 Alexander McPherson

The other important mechanism is nucleation of new layers in

a continuous manner through the occurance and activity of screw,
or spiral dislocations. Examples of such dislocations on a variety of
protein crystals are shown in Fig. 7. They appear as both left and
right handed spirals, as simple and compound screws, and they
exhibit a variety of appearances dependent on the symmetry of the
crystal and various kinetic factors. Because they do not require the
improbable ordering of free molecules from solution, screw dislo-
cations produce new layers even at low supersaturation. Together
the two mechanisms of two dimensional nucleation and screw dis-
location growth account for virtually all protein crystal growth.
Relevant to the practice of crystallization, the specific operable
growth mechanism is principally determined by the crystallization
conditions and the degree of supersaturation they produce. Often
one mechanism may supersede another as supersaturation changes,
and occassionally multiple mechanisms may operate simultaneously

Fig. 7 A major source of growth steps on growing crystals, particularly at lower supersaturation, are screw, or
spiral dislocations. Shown here are a variety of screw dislocations on the surfaces of macromolecular crystals
that illustrates their diverse character
 Protein Crystallization 27

[16, 29]. The mechanisms of growth are further important because

they may determine the amount and distribution of i­mpurities
incorporated in the crystal, the crystal defect structure, the ultimate
size, and even certain diffraction properties such as mosaicity and
resolution limit. It is important in practice to be aware that physical
perturbations, such as vibrations, jarring, or temperature fluctua-
tions can disrupt a growth mechanism or produce a shift from one
mechanism to another.

4 Supersaturation, Nucleation, and Growth

Crystallization of any molecule, or any chemical species including

proteins, proceeds in two distinct but inseparable steps, nucleation
and growth. Nucleation is the most difficult problem to address
both theoretically and experimentally because it represents a mys-
terious first order phase transition by which molecules pass from a
wholly disordered state to an ordered one. We believe that this
likely occurs through the initial formation of partially ordered, or
paracrystalline intermediates, protein aggregates having only short-­
range order, that through internal rearrangement ultimately yield
small, crystallographically ordered assemblies that we refer to as
critical nuclei [16, 29, 31]. A critical nucleus is an ordered cluster
of molecules that is of sufficient size (has a surface to volume ratio)
such that it acquires new molecules at a rate greater than that of
losing molecules.
The size, or number of molecules making up a critical nucleus
is dependent on the molecular dimensions, the extent of supersatu-
ration, and the surface free energy of molecular addition. Currently
the critical nuclear size has only been described for a few proteins,
and for some cases, these were only investigated in terms of two-­
dimensional nuclei developing on the surfaces of already existent
crystals [20, 21]. Recently, a theory has emerged which attempts to
explain the nucleation phenomenon in terms of statistical fluctua-
tions in solution properties [32–34]. This idea holds that a distinc-
tive “liquid protein phase” forms in concentrated protein solutions,
and that this “phase” ultimately gives rise to critical nuclei with
comprehensive order. The hypothesis is supported by observations,
using both atomic force microscopy and quasi-elastic light scatter-
ing, of a third mode of crystal growth in the normal direction
termed growth by three dimensional nucleation [27, 31, 35].
Growth of macromolecular crystals is a better-characterized
process than nucleation, and its mechanisms are reasonably well
understood. As described above, protein crystals grow principally
by the classical mechanisms of dislocation growth, and growth by
two-dimensional nucleation, along with two other less common
mechanisms known as normal growth and three dimensional
nucleation [27, 29, 31]. A common feature of nucleation and
28 Alexander McPherson

growth is that both are critically dependent on the supersaturation

of the mother liquor giving rise to the crystals. Supersaturation is
the variable that drives both processes and determines their occur-
rence, extent, and the kinetics that govern them.
Crystallization of any molecules, including proteins, absolutely
requires the creation of a supersaturated state. This is best illus-
trated by the phase diagram for crystallization shown in Fig. 8.
Supersaturation is a nonequilibrium condition in which some
quantity of the macromolecule in excess of the solubility limit,
under specific chemical and physical conditions, is nonetheless
present in solution. Equilibrium is reestablished by formation and
development of a solid state, such as crystals, as the system returns
to the saturation limit. To produce a supersaturated state, the
properties of an undersaturated, or saturated solution must be

Precipitation

Labile
(supersaturated)
Protein concentration

Metastable
(supersaturated)

Stable
(undersaturated)

Salt concentration

Fig. 8 The phase diagram for the crystallization of macromolecules. The solubility
diagram is divided sharply into a region of undersaturation and a region of super-
saturation by the line denoting maximum solubility at specific concentrations of
a precipitant, which may be salt or a polymer. The line represents the equilibrium
between existence of solid phase and free molecule phase. The region of super-
saturation is further divided in a more uncertain way into the metastable and
labile regions. In the metastable region nuclei will develop into crystals, but no
nucleation will occur. In the labile region, both might be expected to occur. The
final region, at very high supersaturation is denoted the precipitation region
where that result might be most probable
 Protein Crystallization 29

modified to reduce the ability of the medium to sustain the solubil-

ity of the protein (i.e., reduce the chemical activity of the solvent).
Alternatively, some property of the protein molecules must be
altered to reduce protein solubility and/or increase the attraction
of one protein molecule for another, thereby inducing association.
In any case, relationships between solvent and solute, or between
the molecules in solution, are perturbed so as to promote forma-
tion of the solid state.
If no crystals or other condensed phase is present as conditions
are changed, then solute will not immediately produce a new
phase, and the solution will enter and remain in the supersaturated
state. The solid state, hopefully a crystal nucleus, but other incipi-
ent states as well, does not develop spontaneously as the saturation
limit is exceeded. Energy, analogous to the activation energy of a
chemical reaction, is required to initiate the second phase, the sta-
ble critical nucleus of a crystal, or perhaps an unfortunate precipi-
tate. Thus, a kinetic, or energy (or probability) barrier allows
conditions with time to proceed more distant from equilibrium
and further into the zone of supersaturation. Once a critical nucleus
does appear in a supersaturated solution, however, it will proceed
to accumulate molecules from solution and grow until the system
regains equilibrium at saturation. So long as nonequilibrium forces
prevail and some degree of supersaturation exists to drive events, a
crystal will grow or precipitate continue to form.

5 General Approach

Protein crystallization is based on a diverse set of principles, unique

experiences and evolving ideas. There is no comprehensive theory,
or even an organized, extensive base of fundamental data to guide
an investigator, though that is an effort in progress. As a conse-
quence, protein crystal growth is largely empirical in nature, and
demands patience, perseverance and intuition.
What complicates the crystallization process, in addition to our
limited understanding of the phenomena involved, is the intimi-
dating complexity and range of the macromolecules before us.
Even in the case of rather small proteins, such as cytochrome c or
myoglobin for example, there are roughly a thousand atoms with
hundreds of bonds and thousands of degrees of freedom. For
viruses of molecular weights measured in the millions of Daltons,
and for large multi-protein complexes, the possibilities for confor-
mation, interaction, and dynamics are almost unlimited.
We are, however, beginning to develop rational approaches to
protein crystallization based on an understanding of the funda-
mental properties of the systems. We are now increasingly using, in
a systematic manner, the classical methods of physical chemistry to
determine the energetic and kinetic characteristics of the
30 Alexander McPherson

mechanisms responsible for the self-organization of large biologi-

cal molecules into crystal lattices. As an alternative to the precise
and reasoned strategies that we commonly apply to scientific prob-
lems, we, nevertheless, still rely primarily on what is fundamentally
a trial and error approach. Protein crystallization is generally a mat-
ter of searching, as systematically and intelligently as possible, the
ranges of the individual parameters that influence crystal forma-
tion, finding a set, or multiple sets of factors that yield some kind
of crystals, even of poor quality, and then optimizing the individual
variables to obtain the best possible crystals. This is usually achieved
by carrying out an extensive series, or establishing a vast matrix of
crystallization trials, evaluating the results, and using what infor-
mation is obtained to improve conditions in successive rounds of
trials. Because the number of variables is so large, and the ranges so
broad, experience and insight in designing and evaluating the indi-
vidual and collective trials becomes an important consideration.

6 Screening for Initial Crystallization Conditions

As noted above, there are usually two phases in the creation of pro-
tein crystals for an X-ray diffraction investigation, and these are (a)
the identification of chemical, biochemical, and physical conditions
that yield a crystalline material, though it may initially be inade-
quate for X-ray analysis, and (b) the systematic alteration of those
initial conditions by incremental amounts to obtain optimal sam-
ples for diffraction. The first of these is fraught with the greater risk,
as some proteins simply refuse to form crystals, and clues as to why
are elusive or absent. Optimization, however, often proves the more
demanding of effort, more time consuming, and frustrating.
There are two fundamental approaches to searching for crys-
tallization conditions. The first is a systematic variation of what are
believed to be the most important variables, i.e., precipitant type
(salt, polymer, organic liquid) and its concentration, pH, tempera-
ture, protein concentration, and potential ligands. The second is
what we might term a shotgun approach, but a shotgun aimed
with intelligence, experience, and accumulated wisdom. While far
more thorough in scope and more congenial to the scientific mind,
the first method usually requires more effort and a greater amount
of protein. In those cases where the quantity of material is limiting,
it may simply be impractical. The second technique, however, pro-
vides more opportunity for useful conditions to escape discovery.
In general, though, it requires less precious material.
The second approach also has, presently at least, one other
major advantage, and that is availability and convenience. There is
currently on the commercial market, from numerous companies, a
wide variety of crystallization screening kits. The availability and
ease of use of these relatively inexpensive kits, which may be used in
 Protein Crystallization 31

conjunction with a variety of crystallization methods (hanging and

sitting drop vapor diffusion, dialysis, etc., see below) make them the
most popular approach for attacking, at least initially, a new crystal-
lization problem. With these kits, nothing more is required than
combining a series of potential crystallization solutions with one’s
protein of interest using a micropipette, sealing the samples, and
waiting for good fortune to smile. Occasionally it does, but some-
times not, and that is when the crystal grower must begin using his
own intelligence to diagnose problems and devise remedies.
Once some crystals, even if only microcrystals, are observed
and shown to be of protein origin, then optimization begins. Every
component in the solution yielding crystals must be noted and
considered (buffer, salt, ions, etc.), along with pH, temperature,
and whatever other factors might have an impact on the quality of
the results (see below). Each of these parameters or factors is then
carefully incremented in additional trial matrices that encompass
ranges spanning the conditions which gave the “hit.” Because the
problem is nonlinear, that is, one variable may be coupled to
another, this process is often more complex and difficult than one
might anticipate [15, 16, 36–39]. It is here that the amount of
protein and the limits of the investigator’s patience may prove a
formidable constraint.

7 Creating Supersaturation in Practice

In practice, one begins with a solution, a potential mother liquor,

which contains some concentration of the protein below its solu-
bility limit, or alternatively at its solubility maximum (an exception
being the batch method, see below). The objective is then to grad-
ually alter conditions so that the solubility of the protein in the
sample is significantly reduced, thereby rendering the solution
supersaturated. This may be done through a variety of approaches.
Principally, these depend upon (a) altering the protein itself (e.g.,
by change of pH, which alters the ionization state of surface amino
acid residues, by binding a ligand, or by introducing mutations),
(b) altering the chemical activity of the water (e.g., by addition of
salt or organic solvent), (c) altering the degree of attraction of one
protein molecule for another (e.g., addition of bridging ions or
molecules), or (d) altering the nature of the interactions between
the protein molecules and the solvent (e.g., addition of polymers
such as PEG), which also reduces the chemical activity of water.
Table 1 is a compilation of approaches upon which one might
develop strategies for crystallizing a protein for the first time. Indeed,
there are doubtless others that hopefully emerge from the imagina-
tion and cunningness of the investigator. The details of the various
approaches have been described elsewhere [15, 36, 38, 39, 41] and
need receive no extensive treatment here. It is probably sufficient to
32 Alexander McPherson

Table 1
Approaches to creating supersaturation

1. Direct Mixing of Protein and Precipitant: A protein and precipitant solution are thoroughly
mixed so that the final solution is immediately supersaturated in protein. This relies on the
energy, or probability barrier to critical nucleus formation to restrain the system and limit the
number of nuclei and the time of their appearance. The most common application is in
microdrops under oil

2. Temperature Alteration: Refers to a raising or lowering of the temperature of a protein–

precipitant solution that is very near supersaturation. The temperature change is made in a
direction that reduces the solubility of the protein. Most proteins in high salt solutions are more
soluble at cold temperature, while protein–PEG combinations and low ionic strength solutions of
protein are generally more soluble at warm temperatures

3. Alteration of Ionic Strength: Salt is added to a protein solution to high concentration so that
competition for water lowers the solubility of the protein, referred to as “salting out.” Salt ions
can also be removed by dialysis to create a low ionic strength state where, because of deprivation
of cations, the protein is less soluble, referred to as “salting in.” See the phase diagram in Fig. 8
and the illustration in Fig. 10

4. pH Alteration: As the pH of a protein solution is changed, certain amino acid side chains on the
protein molecules’ surfaces alter their ionization, and therefore their charge state. As a
consequence the electrostatic surface of the protein molecules change. If this produces charge
complementary surfaces and additional favorable interactions, or removes unfavorable interactions,
then the molecules will be encouraged to associate and the solubility of the protein will be
reduced. See Fig. 11

5. Ligand Binding: The solubility of most proteins, due to both long range and local
conformational changes, may be altered as a consequence of ligand binding. The ligands may be
coenzymes or other prosthetic groups, inhibitors, or ions. The last of these, particularly divalent
cations, can also form bridges between otherwise repulsive groups on protein molecules and
transform unfavorable interactions into geometry specific, favorable interactions

6. Alteration of the Dielectric Constant of the Medium: This is usually effected by the direct addition,
or addition by dialysis, of an organic liquid of low dielectric constant into the protein solution.
This encourages electrostatic and hydrogen bonding interactions between macromolecules

7. Direct Removal of Water: This can be brought about by simple evaporation or by concentration
of the protein that reduces the water molecules available for solvation of the protein. Any method
that produces dehydration of the protein falls in this category

8. Addition of a Polymer: PEG is most commonly used as a polymeric precipitant and it is

hypothesized to act principally through the mechanism of “volume exclusion.” Because of its very
large hydrodynamic radius, the disordered polymer restricts the volume of solvent that the
protein can access, essentially depriving it of solvating water molecules. It effectively concentrates
and dehydrates the protein and thereby reduces its solubility. There is a possibility that PEG may
also act as an adhesive intermediary between protein molecules to enhance their association

9. Removal of a Solubilizing Agent: Some proteins can be concentrated to an enhanced degree by

inclusion in the solution of some agent that increases its solubility such as a chaotrope or
osmolyte [40]. Removal of the agent after concentrating then leaves the protein at reduced
solubility and perhaps at supersaturation. The agent may be removed by dialysis

10. A
 ddition of Non-volatile Alcohols and Low Molecular Weight Polymers: Liquid compounds such as
MPD, hexanediol, PEGs 200 Da, 400 Da, and low molecular weight Jeffamines reduce the
solubility of proteins, probably by competing for water and altering dielectric constants, but their
true mechanism remains obscure. They may also incorporate into crystals and favorably alter the
solvent structure inside the crystals
 Protein Crystallization 33

say that if a protein has a propensity to crystallize, it can probably be

accomplished by variation of precipitant type, precipitant concentra-
tion, pH, to a lesser extent temperature, but with all due consider-
ation to the biochemical properties and eccentricities of the protein
under investigation. Finally, we are all advised that with real estate
there are three important factors, and they are location, location,
and location. With protein crystallization there are similarly three,
and they are purity, purity, and homogeneity.

8 Methodology

The growth of protein crystals must be carried out in some physi-

cal apparatus that allows the investigator to reduce the solubility of
the protein by altering the properties of the mother liquor, using,
perhaps, one of the strategies in Table 1. Currently, these involve,
almost exclusively, microtechniques. Crystallization “trials” with a
matrix of 48 or 96 conditions may be carried out with volumes of
only a fraction of a milliliter if done manually, a few microliters or
less with some robotic or microfluidic systems. These employ plas-
tic, multichambered trays for hanging and sitting drops, plexiglass
buttons for dialysis, or microdrops under oil. Other methods are
found in Table 2.
Crystallization devices and the associated methodologies have
also been described in detail elsewhere [14, 36, 39, 45]. Detailed
instructions and web sites are frequently provided by the manufac-
turers of the crystallization kits, supplies, and plasticware, along
with many helpful illustrations. The hanging drop and sitting drop
procedures for vapor diffusion, and the batch method using micro-
drops under oil are now most in favor, and are recommended for
most investigations. In those cases where mother liquor compo-
nents cannot be transported through the vapor phase (e.g., metal
ions, detergents) then microdialysis may be the only recourse. An
important point, however, is that the best method for screening
conditions and obtaining an initial set of crystallization parameters
may not be the best means for optimization. Thus one may start
with one technique but ultimately find that another gives larger
crystals of higher quality.
Vapor diffusion, in either the “sitting drop” or “hanging drop”
arrangements is the most popular approach. This is illustrated for
both arrangements in Fig. 9a, b. Vapor diffusion relies on the
equilibration of a small droplet, 1–10 μl in volume usually, against
a larger liquid reservoir of 0.5–1.0 ml. The droplet is initially a
mixture, most commonly 1:1, of a stock protein solution at
10–30 mg/ml, with the reservoir that may contain, for example, a
buffered, concentrated salt or PEG solution. Loss of water from
the droplet to the reservoir and equilibration of the two over time,
hours to days, restores the droplet (almost) to the concentration of
34 Alexander McPherson

Table 2
Physical and chemical procedures for producing solubility minima

1. Bulk (visual) Crystallization: This is a term used by old biochemists to describe the direct
addition of salt to a protein solution while watching intently for the appearance of an opalescent
sheen (Tyndall effect) that indicates incipient crystallization. This was a skill acquired by
researchers for about a century when crystallization was principally used as a powerful purification
tool, and ultimately as a demonstration of protein purity

2. Batch Method in Vials: Once a precipitant concentration that produces crystals is identified for a
particular protein, then a protein solution is simply mixed with the precipitant at 1–3% less. The
protein–salt solution is then dispersed into small screw cap vials and allowed to stand. Very slow
evaporation around the caps causes the precipitant concentration to gradually increase in the vials
until supersaturation is achieved. This was the most widely used technique until about 1970

3. Evaporation: Probably the oldest method in existence, it was used to produce the first reported
protein crystals, those of earthworm hemoglobin [42]. Very slow and controlled evaporation may
still be used and effected through the use of narrow capillaries or wicks

4. Dialysis and Microdialysis: This procedure can be performed in bulk using dialysis tubing, or on a
microscale using capillaries or dialysis buttons enclosed by a dialysis membrane. Dialysis has the
great advantage that it allows investigation of a continuum of precipitant concentrations or a
range of pH. It permits the introduction or removal of ligands, coenzymes, inhibitors or ions. It
can be used with the same protein sample to carry out multiple, independent experiments simply
by changing the exterior solution

5. Concentration Dialysis: Dialysis as described above in (4) but dissolving in the outside solution a
high concentration of high molecular weight PEG (PEG 20,000 for example). Water is
withdrawn as dialysis proceeds and the protein becomes increasingly concentrated as conditions
are altered. Apparatus was once available for performing the concentration function by drawing a
vacuum on the system simultaneous with dialysis, in which case no PEG was required

6. Liquid Bridge: A kind of direct, but slow mixing of a protein solution with a precipitant solution.
Drops of each are placed in a sealed container maintained at 100% humidity, and a needle is used
to draw out a thin connecting liquid bridge between the two drops. Protein diffuses very slowly,
thus it is precipitant that gradually diffuses into the protein drop and promotes supersaturation

7. Free Interface Diffusion: In a tube or capillary, a lighter protein solution is gently layered upon a
heavier precipitant solution (or vice versa depending on relative densities). Slow diffusion along
with some convection across the interface produces a diversity of local concentration gradients
that may promote crystal nucleation and support crystal growth. This technique has been
particularly useful in microgravity where pure diffusive transport prevails and convection is absent

8. Vapor Diffusion: This refers to any arrangement, microdroplets “sitting” on a plastic support, for
example, or “hanging” from a glass cover slip, equilibrating through the vapor phase with a
larger volume reservoir solution containing a higher concentration of precipitant. Over time the
osmolarity of the drop asymptotically approaches that of the reservoir because of water loss from
the protein–precipitant drop. This approach in one manifestation or another is currently in widest
and most popular use.

9. Sequential Extraction: This method [43] is primarily used to produce microcrystals to be later used
for seeding. It depends on the sequential extraction of a salt induced, protein precipitate (centrifuge
pellet) by solutions of decreasing precipitant concentration at 4 °C. The drops of extract are
subsequently placed at 25 °C where the protein (in salt solution) is less soluble and supersaturation is
achieved
(continued)
 Protein Crystallization 35

Table 2
(continued)

10. pH Induced Crystallization: This is a powerful approach and can be accomplished through direct
addition of acid or base, by dialysis, or by vapor diffusion. This takes advantage of the frequent
strong dependence of protein solubility as a function of H3O+ concentration.

11. Temperature Induced Crystallization: This method takes advantage of the difference in solubility
of some proteins as a function of temperature within the range of 0 –37 °C. For most, but not all
proteins, this dependence is rather weak so that the technique is used infrequently with proteins.
It is extremely important in the crystallization of conventional molecules

12. Effector Addition: This depends on the difference in solubility of a protein when it has a
coenzyme, inhibitor, or other ligand bound to it. Dialysis or direct addition can be used to
introduce or remove a ligand and thereby affect protein solubility
13. Crystallization in Gels: Diffusion through gels, such as silica or agarose gels, of a mobile
precipitant into a protein containing gel, essentially free interface diffusion in a gel, can be used
to produce the supersaturated state. Currently in popular use for the crystallization of membrane
and lipophilic proteins, the “lipidic cubic phase” [44] for crystallization takes advantage of the
complex structure of the gel (mesophase) itself to induce nucleation and allow controlled growth

Cover slide (or sealing tape)

(a) (b)
Crystallization droplet

Well of VDX
or Linbro plate
H2O

[ppt]reservoir
[ppt]drop =
2

Micro-bridge
Reservoir Solution

Reservoir solution

Fig. 9 The sitting drop vapor diffusion method is illustrated in this schematic diagram (a). The drop on the
elevated platform, which is commonly 2–10 μl, consists of half stock protein solution and half the reservoir
solution, which contains some concentration of a salt or polymer precipitant. About 0.5 ml of the reservoir
solution is added to the bottom of the cell before sealing. By water equilibration through the vapor phase the
drop ultimately approaches the reservoir in osmolarity both raising the concentration of the precipitant in the
drop and increasing the protein concentration. The hanging drop vapor diffusion method is illustrated sche-
matically in (b). The components of the drop and reservoir, and the physical equilibration process are the same
here as for the sitting drop. The exception is that the protein drop is suspended from a cover slip over the
reservoir rather than resting on a surface. Plasticware for carrying out both sitting and hanging drop vapor
diffusion are widely, and commercially available in numerous formats
36 Alexander McPherson

the protein in the stock solution, and increases the precipitant con-
centration to that of the reservoir. Hopefully this process produces
a droplet supersaturated in protein.
Though proportions and volumes are different, dialysis has the
same objective, to gradually raise the salt or PEG concentration of
a protein solution to the point where it becomes saturated in pro-
tein. Drops under oil [46], or the batch method as it is called, uses
no equilibration. A protein solution and a potential crystallization
promoting solution (salt, PEG, buffer) are simply mixed in some
reasonable proportion and droplets dispersed under oil. This
method relies on the nucleation energy barrier to allow immediate
establishment of supersaturated drops. Other techniques, such as
free interface diffusion [47], slow diffusion and mixing of protein
and precipitant solutions across a liquid–liquid interface, are also in
use, but see less application than batch or vapor diffusion.
In high throughput laboratories, screening for crystallization
conditions, and even optimization in some cases, has generally been
consigned to robotic devices [48–50]. This is particularly true in
those of large pharmaceutical companies where many proteins may
be under simultaneous structural investigation. Automated systems
have the advantages of exceptional record maintenance, most can
deploy sub microliter amounts of mother liquor, and they can be
used to screen vast matrices of conditions that might otherwise be
impossible in a practical sense for a lone investigator using manual
techniques. Robotic systems are, in addition, now being used to
examine and evaluate the results of crystallization trials using opti-
cal subsystems and image processing techniques [51–53]. Evaluation
of trial arrays of conditions, however, continues to be problematic
because of the continuing difficulty in devising meaningful criteria
for progress in the absence of actual crystals. That is, the sole pres-
ence of various kinds of precipitates or other phases in an individual
crystallization trial gives only very murky indications of how near
the conditions were to a successful mother liquor.

9 Precipitants

One of the most important components in a mother liquor

intended to crystallize a protein is sometimes called the precipi-
tant, or other times the crystallization agent. It is generally, but not
always, the chemical component that reduces the solubility of the
protein or reduces the chemical activity of water. Salts, such as
ammonium sulfate or potassium phosphate, or polymers such as
PEG are classic examples. The mechanisms by which they act (see
below) may be different, but their essential role is to deprive the
protein of solvating water and to promote protein association.
If one were to examine the reagents utilized in any of the com-
mercial crystallization screens that are based on shotgun approaches,
 Protein Crystallization 37

or examined the crystallization conditions that have been compiled

into data bases [54, 55], then it would become apparent that a
wide variety of precipitating (crystallizing) agents have been used.
Indeed many agents have been employed, and some, such as
ammonium sulfate and polyethylene glycol have produced a great
number of successes. It is often necessary, however, to explore
many precipitants, and it is difficult to know initially which might
offer the greatest likelihood of obtaining crystals.
Individual precipitants and their properties have also been
reviewed in some detail [16] and are not extensively discussed here.
To summarize, however, it is possible to group the precipitants into
categories based on their mechanisms for promoting crystallization.
The majority of precipitants of proteins fall into four broad catego-
ries (1) salts, (2) organic solvents, (3) long chain polymers, and (4)
low molecular weight polymers and non-volatile alcohols. The first
two classes are typified by ammonium sulfate and ethyl alcohol
respectively, and higher polymers such as polyethylene glycol 4000
are characteristic of the third. In the fourth category we might place
compounds such as methylpentanediol (MPD) and polyethylene
glycols of molecular weight less than about 1000.
The solubility of proteins in concentrated salt solutions is com-
plicated, but it can be viewed naively as a competition between salt
ions, principally the anions, and the protein for the binding of
water molecules, which are essential for the maintenance of solu-
bility [56–60]. At sufficiently high salt concentrations the protein
molecules become so uncomfortably deprived of solvent that they
seek association with one another in order to satisfy their electro-
static and amphipathic requirements. In this environment large,
semi ordered aggregates that could lead to critical crystal nuclei, as
well as disordered amorphous precipitate may form. Other salt
ions, chiefly cations, also may be necessary to insure protein solu-
bility. At low ionic strengths, cation availability may be insufficient
to maintain protein solubility, and under those conditions too,
crystals may form. The behavior of typical proteins over the entire
range of salt concentrations, including both the “salting in” and
“salting out” regions is illustrated by Fig. 10.
Salts exert their effect principally by dehydrating proteins
through competition for water molecules, and a measure of their
efficiency in this is the ionic strength, whose value is the product of
the molarity of each ion in solution with the square of their valences.
Thus, multivalent ions are the most efficient precipitants. Sulfates,
phosphates, citrates, and more recently malonates [61] and mixtures
of the salts of dicarboxylic acids have traditionally been employed.
One might anticipate little variation among different salts so
long as the valences of their ions were the same. Thus there should
be little expected variation between two different sulfates such as
Li2SO4 and (NH4)2SO4 if only ionic strength were involved. This is
often observed not to be the case. In addition to salting out, which
38 Alexander McPherson

Fig. 10 The curve shown here represents a typical solubility curve for a protein
and divides the region of undersaturation from that of supersaturation. It also
illustrates the existence of the classical “salting in” and a “salting out” region for
the protein. By taking advantage of the latter effects, supersaturation may be
achieved by equilibrating a system from a point of maximum solubility (P0) to one
of reduced solubility (P1 or P2) by adjusting the precipitant concentration

is a general dehydration effect not really much different than evap-

oration or concentration (except that water is not physically
removed) there are also specific protein–ion interactions that may
have further consequences [58, 60]. This is perhaps not unex-
pected given the varied hydration properties of different ions and
the unique polyvalent character of individual proteins, protein
structural and dynamic complexity, and the intimate dependence
of their physical properties on their surroundings. It is inadequate,
therefore, when attempting to crystallize a protein to examine only
one or two salts and ignore the broader range. Alternative salts can
sometimes produce crystals of varied quality, morphology, and in
some cases diffraction properties.
It is usually not possible to predict the degree of saturation or
molarity of a precipitating agent required for the crystallization of
a particular protein without some prior knowledge of its solubility
behavior. In general, however, it is a concentration of the precipi-
tant just a few percent less than that which yields an amorphous
precipitate [62], and this can be determined for a macromolecule
under a given set of conditions using only minute amounts of
material [15]. To determine the approximate insolubility points
with a particular precipitant a 10 μl droplet of a 5–15 mg/ml pro-
tein solution can be placed in the well of a depression slide and
observed under a low-power light microscope as increasing
 Protein Crystallization 39

amounts of saturated salt solution or organic solvent (in 1- or 2-μl

increments) are added. If the well is sealed between additions with
a coverslip, the increases can be made over a period of many hours.
Along with precipitant type and concentration, pH is usually the
most important variable influencing the solubility of proteins. As
such, it provides a powerful approach to creating supersaturated solu-
tions, and hence effecting crystallization. Its manipulation at various
ionic strengths and in the presence of diverse precipitants is a founda-
tional concept in formulating screening matrices and discovering suc-
cessful crystallization conditions. An example of how pH might be
used to effect crystallization of a protein is illustrated in Fig. 11.
Organic solvents reduce the dielectric constsnt of the medium,
hence the screening of the electric fields that mediate macromo-
lecular interactions in solution. A danger, however, is that they also
tend to destabilize protein structure. As the concentration of
organic solvent is increased, interaction between protein molecules
increases, solvent becomes less effective (the activity coefficient of
water is reduced) and the solid state becomes more favored [63, 64].
Organic solvents should be used at low temperature, at or below
0 °C, and they should be added very slowly with good mixing
[16]. Since they are usually volatile, vapor diffusion techniques are
equally applicable. Ionic strength should, in general, be maintained

Fig. 11 As shown here, most proteins have specific solubility minima as a func-
tion of pH. One can take advantage of this property to produce supersaturation
by altering a system from a pH permitting high solubility (P1 or P2) to a point of
low solubility (P0). This is a powerful approach to promoting crystallization of
macromolecules
40 Alexander McPherson

low and whatever means are otherwise available should be pursued

to protect against denaturation.
Some polymers, among which polyethylene glycols (PEG) are
most popular [65, 66], produce volume exclusion effects that
induce separation of proteins from solution [65, 67]. Polymeric
precipitants, unlike proteins, have no consistent, fixed conforma-
tion. They writhe and twist randomly in solution and, as a conse-
quence, occupy far more space than their molecular weights would
suggest. This effect, refered to as volume exclusion, results in less
solvent available space for the protein molecules that then segre-
gate, aggregate, and ultimately form a solid state, in favorable cases
crystals. PEG is also extremely hydrophilic and binds water mole-
cules avidly (about 2.3 water molecules per monomeric unit [68, 69]).
It, like salts, competes for solvent, thereby dehydrating the protein
molecules.
Evidence has emerged recently that suggests PEG, and some
related polymeric precipitants, may not exert their effects on pro-
tein crystallization exclusively by the mechanism of volume exclu-
sion and dehydration. Some observations indicate that PEG, at
least fragments and lower molecular weight components (all PEG
preparations exhibit a distribution of lengths about a mean) may
actually co-crystallize with the protein due to positive, associative
interactions [68, 69], and thus occupy interstitial spaces and chan-
nels otherwise filled by solvent alone. Inside the crystal, PEG likely
remains disordered, or at best partially ordered, and thereby
escapes detection by X-ray analysis. If this PEG incorporation is
valid, then it has important ramifications, as PEG could well influ-
ence protein association and crystallization by both altering inter-
stitial water structure, and possibly by providing a soft superstructure
that helps guide crystal growth. More remains to be done to test
this intriguing idea.
A large number of protein structures have now been solved
using crystals grown from polyethylene glycol. These confirm that
the protein molecules are in as native condition in this medium as
in any other. This is reasonable because the larger molecular weight
polyethylene glycols probably do not even enter the crystals and
therefore do not directly contact the interior molecules. In addi-
tion, it appears that crystals of many proteins when grown from
polyethylene glycol are essentially isomorphous with, and exhibit
the same unit cell symmetry and dimensions as those grown by
other means.
PEG sizes from Mr = 200 to 20,000 Da have successfully pro-
vided protein crystals, but the most useful seem to be those in the
range 2000–8000 Da. A large number of reports have appeared,
however, in which a protein could not easily be crystallized using
this range but yielded in the presence of PEG 200 Da or 20,000 Da.
The molecular weight sizes may not be completely interchangeable
for a given protein even within the mid range. Some produce the
best-formed and largest crystals only at, say, Mr = 4000 Da, and
 Protein Crystallization 41

less perfect examples at other weights. This is a parameter that is

best optimized by empirical means along with concentration. The
very low molecular weight PEGs such as 200 and 400 Da are
somewhat similar in character to MPD and hexanediol. There does
not appear to be any correlation between the molecular weight of
a protein and that of the PEG best used for its crystallization. The
higher molecular weight PEGs do, however, have a proportionally
greater capacity to force proteins from solution.
An advantage of PEG over most other precipitating agents is
that proteins crystallize within a fairly narrow range of PEG con-
centration; this being from about 4% to 20% (although there are
numerous examples where either higher or lower concentrations
were necessary). In addition, the exact PEG concentration at which
crystals form is rather insensitive. If one is within a few percent of
the optimal value (in some cases even more), some success is likely
to be achieved. With most crystallizations from high ionic strength
solutions or from organic solvents, one must be within 1% or 2% of
an optimum lying anywhere between 15% and 85% saturation. The
advantage of PEG, then, is that when conducting a series of initial
trials to determine what conditions will give crystals, one can use a
fairly coarse selection of concentrations and over a rather narrow
total range.
Since PEG solutions are not volatile, PEG must be used like
salt or MPD and equilibrated with the protein by dialysis, slow
mixing, free interface diffusion, or vapor equilibration. When the
reservoir concentration of PEG is in the range of 5–12%, the pro-
tein solution to be equilibrated should be at an initial concentra-
tion of about half, conveniently obtained by mixing equal volumes
of the reservoir and protein solution. When the final PEG concen-
tration to be attained is much higher than 12%, it is probably advis-
able to initiate the mother liquor at no more than 5–10% below the
desired final value.

10 Factors Affecting Crystallization

There are many factors that can affect the crystallization of proteins
and these, too, have been reviewed elsewhere [16, 36, 38, 39]. They
fall into categories of physical factors such as temperature, chemical
influences such as pH or ionic strength, and biochemical factors that
include, among many others, purity and monodispersity. Any one,
or any combination may affect the liklihood of crystallization occur-
ing at all, the nucleation probability and rate, crystal growth rate and
mechanism, and the ultimate sizes and quality of the products. As
noted above, pH and the concentrations of salt and other precipi-
tants are virtually always of importance. The concentration of the
protein, which may vary from as low as 2 mg/ml for viruses and
large complexes [70], to as much as a hundred mg/ml for some
highly soluble proteins, is an additional, significant variable.
42 Alexander McPherson

Other parameters may be unimportant in some cases but play a

crucial role in others. In particular the presence or absence of
ligands, coenzymes, or inhibitors, the variety of salt or buffer, the
equilibration technique used, temperature fluctuations, and the
presence of detergents and chaotropes [40] are all pertinent factors.
Parameters of somewhat lesser and largely obscure significance are
things like gravity, electric and magnetic fields, or viscosity. It can
not be predicted which of these many variables may be of impor-
tance for a particular macromolecule, and the influence of any one
must, in general, be investigated through empirical trials.
An intriguing problem, or opportunity depending on one’s
perspective, is what additional components or compounds should
be included in the mother liquor in addition to solvent, buffer,
protein, and precipitating agent [16, 36, 40, 71, 72]. The most
desirable effectors, it would seem, are those which maintain the
protein in a single, homogeneous, and invariant state. Reducing
agents such as glutathione, β-mercaptoethanol, and dithithreitol
are useful to preserve sulfhydryl groups and prevent oxidation.
EDTA and EGTA are effective if one wishes to protect the protein
from heavy or transition metal ions. Inclusion of these components
may be particularly important when crystallization requires a long
period of time to reach completion. When crystallization is carried
out at room temperature in polyethylene glycol or low ionic
strength solutions, then attention must be given to preventing the
growth of microbes. These generally secrete proteolytic enzymes
that may have serious effects on the integrity of the protein under
study. Inclusion of sodium azide, thymol or chlorobutanol at low
levels may be necessary to suppress invasive bacteria and fungi.
Substrates, coenzymes and inhibitors often serve to maintain an
enzyme in a more compact and stable form. Thus a greater degree of
structural homogeneity may be imposed on a population of protein
molecules and a reduced level of statistical variation achieved by com-
plexing the protein with a natural ligand before attempting its crystal-
lization. In some cases an apoprotein and its ligand complexes may
be significantly different in their physical behavior and can, in terms
of crystallization, be treated as almost entirely separate problems.
Complexes may provide additional opportunities for growing crystals
if the native apoprotein is refractory. It is worthwhile, therefore,
when searching for crystallization conditions, to explore complexes
of the macromolecule with substrates, coenzymes, and inhibitors at
an early stage. Such complexes are, in addition, often inherently more
interesting in a biochemical sense than the apoprotein.
Various metal ions have occasionally been observed to pro-
mote the crystallization of proteins. Bacterial glucose isomerase,
for example, crystallizes readily from PEG solutions in the pres-
ence of Mg++, but only with difficulty in its absence. Cadmium and
some other divalent cations induce immediate crystallization of the
iron storage protein ferritin [73]. In some instances ions are essen-
tial for activity. It is, therefore, reasonable to expect that they might
 Protein Crystallization 43

aid in maintaining certain structural features of the molecule.

There are other examples, however, where metal ions, particularly
divalent metal ions of the transition series such as Ca++, were found
to encourage crystal growth but played no recognized role in the
protein’s activity or structure. They likely serve as bridging agents
between molecules in the crystal lattice.

11 The Protein as a Variable

A factor of particular importance to crystallization is the homoge-

neity and monodispersity of the protein [74, 75] and this deserves
special emphasis. Some proteins may crystallize even from very
heterogeneous mixtures (egg albumin, lysozyme, canavalin,
α-amylase, for example), and indeed, crystallization has long been
used as a powerful purification tool. It is the reason, in fact, why it
originated as a technique and has been held in such high regard. In
general, however, the likelihood of success in crystal growth is
greatly advanced by increased homogeneity of the protein sample.
Investment in further purification is always warranted, and usually
profitable. When every effort to crystallize a protein fails, the best
recourse is to further purify.
Recombinant DNA technology provided an enormous impetus
to crystal growth research and X-ray crystallography 35 years ago, as
it provided crystallographers access to proteins found in very low
abundance that nevertheless played important roles in cells. Indeed
it may be on the verge of providing another advance at this very
time. Arguably, the most important parameter in protein crystalliza-
tion is the protein itself. Until recently we have had little or no direct
control over most of the important features of that parameter.
Modification at the genetic level, however, provides us that oppor-
tunity, and its possibilities are only now being realized [76–78].
Through truncations, mutations, chimeric conjugates, and
many other protein engineering contrivances, the probability of
crystallization has been significantly enhanced. If we can learn how
to go about this in a rational and systematic manner then advances
may occur in future years that match the progress of the past.
Approaches to application of mutation will be addressed and elab-
orated by others elsewhere in this volume.

12 Optimization of Crystallization Conditions

Optimization means adjusting the parameters of crystallization con-

ditions, initially estimated from screening matrices [16, 37, 39], with
the objective of discovering improved conditions that ultimately yield
the best crystals for diffraction data collection. Optimization is in a
sense refinement, but it is complicated somewhat because the param-
eters almost certainly are not independent of one another. They may
44 Alexander McPherson

be linked or correlated. Furthermore, solubility diagrams, which

would have many dimensions, do not exist for specific proteins. Every
protein has a unique length and amino acid sequence, and a unique
three-dimensional c­ onformation. Every protein is an individual with
its own eccentricities and peculiarities. A further complication is that
there can be an “embarrassment of riches” where many “hits” are
obtained initially and the question arises as to which deserve the
effort required for further improvement.
Optimization, as it is often practiced is in principle relatively
straightforward. The parameters that define the initial conditions
are first identified (pH, precipitant type, precipitant concentration,
temperature, ion concentration, etc.). Following this, solutions are
made that incrementally and systematically vary the parameters
about the initial values. That is, if the pH of the initial hit was 7.0,
then the same mother liquor might be composed but at pH 6, 6.2,
6.4, 6.6, etc. up to pH 8.0. This does not guarantee that one will
arrive at optimal conditions, parameters may be correlated, but it is
the best approach that we have.
While simple in principle, optimization becomes demanding in
the laboratory. First of all, the number of parameters or effecting
conditions may be large [15, 16, 36, 37]. It may not be clear which
parameters are actually important, or what the range for explora-
tion should be. Thus we have as an initial goal of optimization to
deduce what variables are relevant and how to prioritize each rela-
tive to another so that adjustments can be made, all the while mini-
mizing or neglecting the least or irrelevant factors.
Optimization may require a substantial amount of protein sam-
ple, and this may be severely limited. Thus, efficiency and economy
become essential, and the use of very small volume trials [48, 50, 52]
will be tempting. Small volumes, however, should be treated with
caution. One seldom obtains large crystals from nanoliter volumes of
mother liquor, and when promising results from very small drops are
scaled up to larger volumes to grow larger crystals (which larger vol-
umes tend to yield) the increase in scale fails to materialize.
The greatest obstacle to success in optimization is most fre-
quently an absence of sufficient commitment, or a lack of effort on
the part of the investigator. Screening for new crystallization condi-
tions can be made almost, but not quite, painless. Commercial kits
can be purchased that contain precisely prepared solutions. Robotics
are now employed to dispense samples into plates, further robotic
devices categorize and store the plates, and automated photographic
systems present images of the many drops for viewing [49, 50, 52].
Automated systems, however, cannot make optimization
effortless, and that is because optimization requires composition of
a vast number of solutions that must be formulated or purchased,
and the use of robotics in optimization presents as many problems
as it solves, at least at this point in time. Making up a myriad of
solutions, adjusting their pHs to exact values, and so on is tedious.
 Protein Crystallization 45

In other words, doing a lot of basic laboratory chemistry demands

a lot of hard labor. Many investigators would rather struggle with
marginal, or even miserable crystals obtained from the first hit than
undertake the optimization effort.

13 Membrane Proteins

Proteins that are naturally membrane associated, or that are other-

wise unusually hydrophobic or lipophilic in nature present unique
problems. Such proteins are, in general, only sparingly soluble in
normal aqueous media, some virtually insoluble, others lose their
active conformations, and this in turn makes the application of con-
ventional protein crystallization techniques problematic. Problems
are difficult but not intractable. To address these difficulties the use
of detergents, particularly non-ionic detergents, has been devel-
oped [79–83]. No attempt will be made here to describe the vari-
ous techniques or the combinations of detergents and accessory
molecules that have been used, as that involves a number of com-
plexities and considerations that are covered in another chapter.
An essential difficulty associated with inclusion of a solubilization
agent, such as a detergent, is that it adds an additional dimension to
the matrix of conditions that must otherwise be evaluated. For exam-
ple, if one is content to use a standard 48 drop screen of conditions, at
least initially, then the additional search for a useful detergent means
that the 48 trial screen must be multiplied by the number of detergent
candidates. A further problem is that there are a great number of
potentially useful detergents. Hampton Research (Aliso Viejo, CA), a
major source of screening reagents, offers three different detergent
kits of 24 compounds each. Were one to simply apply a basic 48-well
screen with each detergent, then that would require a total of 3456
individual trials. While this may actually be possible with highly auto-
mated, nanoscale systems, and where a substantial amount of material
is available, it is impractical for most laboratories.
Basic crystal screens, whether they are systematic screens or
shotgun screens, should not, however, be abandoned. It becomes
essential though to reduce, at least initially, the number of deter-
gents to be considered. If, for example, a set of six highly promis-
ing detergents could be identified, then less than 300 trials would
be called for initially, an undertaking well within the capabilities
of most labs. No one, however, has yet reduced the set to a
favored few. Everyone has an opinion as to which detergents
should be favored, and no consensus has yet emerged from data
bases and analyses of experiments. To make matters even more
challenging, it appears that some, perhaps many detergents func-
tion best when accompanied by small amphiphilic molecules such
as LDAO. This would of course add yet another dimension to the
screening problem.
46 Alexander McPherson

While not as valuable as naming actual candidate detergents,

the author can point to a number of useful reviews and discussions
that illustrate the properties and virtues of various detergents for
membrane crystallization. Reference 83 is a good review of workup
until that time, and more recently, there are fine discourses by Loll
[80], Caffrey [44], Garavito and Ferguson-Miller [84], Hunte,
et al. [85], and Wiener [79, 86], as well as a chapter in this book.

14 Some Important Concepts

Although approaches to protein crystallization remain largely

empirical, substantial progress has been made. We have now iden-
tified useful reagents, devised a host of physical–chemical tech-
niques for studying the crystallization process, and gained a better
understanding of the unique features of proteins and their complex
assemblies that affect their capacity to crystallize. Some principles
now stand out regarding the crystallization problem, and these are
summarized in Table 3.

Table 3
Some important concepts in protein crystallization

1. Protein Purity—Crystallization occurs because a population of structurally and chemically

homogeneous molecules are made amenable to the formation of periodic bonding arrangements.
Molecular misfits create disruptions of order and inhibit critical nucleus formation and crystal
growth. Efforts to make the most pure and uniform protein sample as possible are never wasted

2. Solubility and Monodispersity—High protein concentration generally means more reliable

crystallization and a greater overall chance of success in initial screens, and this depends on
solubility of the protein. Solubility, however, also implies protein monodispersity and the absence
of arbitrary oligomers and aggregates in the sample that are little more than contaminants
3. Stability—A foundational concept in crystallization is the unchanging nature of the molecules with
regard to conformation and physical–chemical properties. It is now a given that the more stable a
protein, the more likely it is to crystallize. The investigator must do whatever possible to insure
that the protein molecules remain in their native state
4. Supersaturation—This is the crucial, controlling factor in determining nucleation probability, and
both the mechanisms and kinetics of crystal growth. It can be achieved in many ways, and the path
by which it is reached is as important as the ultimate value. A solution supersaturated in protein is a
physical necessity for crystallization
5. Association—Supersaturation can be reached in many cases by enhancing attractive, specific
interactions between protein molecules and thereby reducing their solubility. Additives, ions,
protein modifications are traditional approaches. Reducing the chemical activity of the solvent
abets this process and is the mechanism by which most precipitating agents operate
6. Nucleation—This is essential to start the crystallization process, and it is largely dependent on
probability. That in turn depends on the degree of supersaturation and the path (through the
phase diagram) by which supersaturation is reached. Competition from other condensed phases,
such as precipitate, is the primary adversary. Enough supersaturation is necessary; too much
supersaturation is a damper.
(continued)
 Protein Crystallization 47

Table 3
(continued)
7. Variety—Because of the stochastic elements involved in crystallization, chance is an important
factor. The more chances one has, the more likely is success. Explore as many possibilities and
opportunities as possible in terms of sample source, sample conformation, physical, chemical, and
biochemical parameters
8. Constancy—Physical and/or chemical perturbations can inject energy into a dynamic, crystallizing
system and cause deviations of otherwise ordered growth mechanisms. Disruptions from
mechanical jarring, evaporation, or from temperature fluctuations can be devastating. Maintain
the crystallizing samples at an optimal state during the full course
9. Impurities—The incorporation of impurities, not only molecules present in the protein sample,
but in the reagents, apparatus, or from the environment can seriously contribute to unwanted
nucleation and growth termination
10. Preservation—Crystals vary in their long term stability once they have reached terminal size. It
may be necessary to take “post crystallization” measures to insure that the crystals maintain their
quality until X-ray data collection can begin. These may include lowering temperature, increasing
the precipitant concentration, prevention of evaporation through plastics, addition of stabilizers,
cryo-vitrification, or mounting in sealed capillaries. Shock and handling must be avoided

References

1. Helliwell JR (1992) Macromolecular crystal- 9. Gruner SM, Eikenberry EF, Tate MW (2001)
lography with synchrotron radiation. Comparison of X-ray detectors. In: Rossmann
Cambridge University Press, Cambridge, MG, Arnold E (eds) International tables for
UK crystallography, vol F. Kluwer Academic
2. Bingel-Erlenmeyer R, Olieric V, Grimshaw Publishers, Dordrecht, pp 143–153
J et al (2011) SLS crystallization platform at 10. Rossmann MG, Arnold E (2001)
Beamline X06DA-A fully automated pipeline Crystallography of biological macromolecules,
enabling in situ X-ray diffraction screening. vol F. International tables for crystallography.
Cryst Growth Des 11:916–923 Dordrecht, Kluwer Academic Publishers
3. Garman EF, Schneider TR (1997) 11. Malkin AJ, Kuznetsov YG, McPherson A
Macromolecular cryocrystallography. J Appl (1996) Defect structure of macromolecular
Crystallogr 30:211–237 crystals. J Struct Biol 117:124–137
4. Pflugrath JW (2004) Macromolecular cryo- 12. Feigelson RS (1988) The relevance of small
crystallography—methods for cooling and molecule crystal growth theories and tech-
mounting protein crystals at cryogenic tem- niques to the growth of biological macromol-
peratures. Methods 34:415–423 ecules. J Cryst Growth 90:1–13
5. Pflugrath JW (2015) Practical macromolecular 13. Feher G (1986) Mechanisms of nucleation and
cryocrystallography. Acta Crystallogr F Struct growth of protein crystals. J Cryst Growth
Biol Commun 71:622–642 76:545–546
6. Heras B, Edeling MA, Byriel KA et al (2003) 14. Durbin SD, Feher G (1996) Protein crystalli-
Dehydration converts DsbG crystal diffraction from zation. Annu Rev Phys Chem 47:171–204
low to high resolution. Structure 11:139–145 15. McPherson A (1982) The preparation and
7. Kiefersauer R, Than ME (2000) A novel free-­ analysis of protein crystals. Wiley, New York
mounting system for protein crystals: transfor- 16. McPherson A (1999) Crystallization of bio-
mation and improvement of diffraction power logical macromolecules. Cold Spring Harbor
by accurately controlled humidity changes. Laboratory Press, Cold Spring Harbor, NY
J Appl Crystallogr 33:1223–1230 17. McPherson A, Malkin AJ, Kuznetsov YG
8. Pflugrath JW (1992) Developments in X-ray (1995) The science of macromolecular crystal-
detectors. Curr Opin Struct Biol 2:811–815 lization. Structure 3(8):759–768
48 Alexander McPherson

18. Rosenberger F (1986) Inorganic and protein 34. Piazza R (1999) Interactions in protein solu-
crystal growth—similarities and differences. tions near crystallization: a colloid physics
J Cryst Growth 76:618 approach. J Cryst Growth 196:415–423
19. Kuznetsov YG, Malkin AJ, Greenwood A et al 35. Malkin A, McPherson A (1994) Light scatter-
(1995) Interferometric studies of growth ing investigations of nucleation processes and
kinetics and surface morphology in macromo- kinetics of crystallization in macromolecular
lecular crystal growth. Canavalin, thaumatin systems. Acta Crystallogr D Biol Crystallogr
and turnip yellow mosaic virus. J Struct Biol 50:385–395
114:184–196 36. McPherson A, Gavira JA (2014) Introduction
20. Malkin AJ, Kuznetsov YG, Glantz W et al to protein crystallization. Acta Crystallogr F
(1996) Atomic force microscopy studies of sur- Struct Biol Commun 70:2–20
face morphology and growth kinetics in thau- 37. McPherson A, Cudney B (2014) Optimization
matin crystallization. J Phys Chem of crystallization conditions for biological mac-
100:11736–11743 romolecules. Acta Crystallogr F Struct Biol
21. Malkin AJ, Kuznetsov YG, McPherson A Commun 70:1445–1467
(1997) An in situ investigation of catalase crys- 38. Ducruix A, Giége R (1992) Crystallization of
tallization. Surf Sci 393:95–107 nucleic acids and proteins, a practical approach.
22. Chernov AA, Komatsu H (1995) Principles of IRL Press, Oxford
crystal growth in protein crystallization. In: 39. Bergfors TM (1999) Protein crystallization:
Bruinsma JPEOSL (ed) Science and technol- techniques, strategies and tips. International
ogy of crystal growth. Kluwer, Dordrecht, The University Line, La Jolla, CA
Netherlands, p 67 40. Bolen DW (2004) Effects of naturally occur-
23. Vekilov PG, Chernov AA (2002) The physics ring osmolytes on protein stability and solubil-
of protein crystallization. Solid State Phys ity: issues important in protein crystallization.
57:2–147 Methods 34:312–322
24. Chernov AA (2003) Protein crystals and their 41. McPherson A, Cudney R, Patel S (2003) The
growth. J Struct Biol 142:3–21 crystallization of proteins, nucleic acids, and
25. Rosenberger A (1979) Fundamentals of crystal viruses for X-ray diffraction analysis. In:
growth. Springer-Verlag, Berlin Fahnestock SR, Steinbuchel A (eds)
26. Chernov AA (1984) Modern crystallography Biopolymers, vol 18(16), pp 427–468
III. Crystal growth. Springer-Verlag, Berlin. 42. Hünefeld FL (1840) Der Chemismus in der
27. Malkin AJ, Kuznetsov YG, Land TA et al thierischen organisation. FA Brockhaus, Leipzig
(1995) Mechanisms of growth for protein and 43. Jacoby WB (1968) A technique for the crystal-
virus crystals. Nat Struct Biol 2:956–959 lization of proteins. Anal Biochem 26:295
28. McPherson A, Malkin AJ, Kuznetsov YG et al 44. Caffrey M (2003) Membrane protein crystalli-
(2001) Atomic force microscopy applications zation. J Struct Biol 142:108–132
in macromolecular crystallography. Acta 45. McPherson A (1976) The growth and prelimi-
Crystallogr D Biol Crystallogr 57:1053–1060 nary investigation of protein and nucleic acid
29. McPherson A, Malkin AJ, Kuznetsov YG crystals for X-ray diffraction analysis. Methods
(2000) Atomic force microscopy in the study Biochem Anal 23:249–345
of macromolecular crystal growth. Annu Rev 46. Chayen NE, Shaw-Stuart PD, Blow DM
Biophys Biomol Struct 29:361–410 (1992) Microbatch crystallization under oil: a
30. McPherson A, Kuznetsov YG (2014) new technique allowing many small volume
Mechanisms, kinetics, impurities and defects: crystallization trials. J Cryst Growth
consequences in macromolecular crystalliza- 122:176–180
tion. Acta Crystallogr F Struct Biol Commun 47. Salemme FR (1972) A free interface diffusion
70:384–403 technique for the crystallization of proteins for
31. Kuznetsov Yu G, Malkin A, McPherson A X-ray crystallography. Arch Biochem Biophys
(1998) Atomic force microscopy studies of 151:533–539
phase separations in macromolecular systems. 48. Bard J, Ercolani K, Svenson K, Olland A,
Phys Rev B 58:6097–6103 Somers W (2004) Automated systems for pro-
32. Ten Wolde R, Frenkel D (1997) Enhancement tein crystallization. Methods 34:329–347
of protein crystal nucleation by critical density 49. Hui R, Edwards A (2003) High-throughput pro-
fluctuations. Science 277:1975–1978 tein crystallization. J Struct Biol 142:154–161
33. Haas C, Drenth J (1999) Understanding pro- 50. Santarsiero BD, Yegian DT, Lee CC et al
tein crystallization on the basis of the phase (2002) An approach to rapid protein crystalli-
diagram. J Cryst Growth 196:388–394
 Protein Crystallization 49

zation using nanodroplets. J Appl Crystallogr macromolecules. Biochem Biophys Res

35:278–281 Commun 207:819–828
51. Hosfield D, Palan J, Hilger M et al (2003) A 67. Ingham KC (1990) Precipitation of proteins
fully integrated protein crystallization platform with polyethylene glycol. Methods Enzymol
for small-molecule drug discovery. J Struct Biol 182:301–306
142:207–217 68. Israelachvili J (1997) The different faces of
52. DeLucas LJ, Bray TL, Nagy L, McCombs D, poly(ehylene glycol). Proc Natl Acad Sci U S A
Chernov N, Hamrick D, Cosenza L, Belgovskiy 94:8378–8379
A, Stoops B, Chait A (2003) Efficient protein 69. Sheth SR, Leckband D (1997) Measurements
crystallization. J Struct Biol 142:188–206 of attractive forces between proteins and end-­
53. Luft JR, Collins RJ, Fehrman NA et al (2003) grafted poly(ethylene glycol) chains. Proc Natl
A deliberate approach to screening for initial Acad Sci U S A 94:8399–8404
crystallization conditions of biological macro- 70. McPherson A, Larson SB (2015) A guide to
molecules. J Struct Biol 142:170–179 the crystallographic analysis of icosahedral
54. Gilliland GL, Tung M, Blakeselee DM et al viruses. Crystallogr Rev 21:4–55
(1994) Biological macromolecules crystalliza- 71. Timasheff SN, Arakawa T (1988) Mechanism
tion database, version 3.0: new features, data of protein precipitation and stabilization by co-­
and the NASA archive for protein crystal solvents. J Cryst Growth 90:39–46
growth data. Acta Crystallogr D Biol 72. McPherson A, Cudney B (2006) Searching for
Crystallogr 50:408–413 silver bullets: an alternative strategy for crystal-
55. Gilliland GL (1988) A biological macromole- lizing macromolecules. J Struct Biol
cule crystallization database: a basis for a crys- 156:387–406
tallization strategy. J Cryst Growth 90:51–60 73. Granick S (1942) Ferritin: I. Physical and
56. Cohn EJ, Ferry JD (1943) The interactions of chemical properties of horse spleen ferritin.
proteins with ions and dipolar ions. In: Cohn J Biol Chem 146:451–461
EJ, Edsall JT (eds) Proteins, amino acids and 74. Giege R, Lorber B, Theobald-Dietrich A
peptides as ions and dipolar ions. Reinhold, (1994) Crystallogenesis of biological macro-
New York, pp 586–622 molecules: facts and perspectives. Acta
57. Cohn EJ, Edsall JT (eds) (1943) Proteins, Crystallogr D Biol Crystallogr 50:339–350
amino acids and peptides as ions and dipolar 75. McPherson A, Malkin A, Kuznetsov YG et al
ions. Van Nostrand-Reinhold, Princeton, NJ (1996) Incorporation of impurities into mac-
58. Collins KD (2004) Ions from the Hofmeister romolecular crystals. J Cryst Growth
series and osmolytes: effects on proteins in 168:74–92
solution and in the crystallization process. 76. Dale GE, Oefner C, D'Arcy A (2003) The pro-
Methods 34:300–311 tein as a variable in protein crystallization.
59. Herriott RM (1942) Solubility methods in the J Struct Biol 142:88–97
study of proteins. Chem Rev 30:413 77. Derewenda ZS (2004) The use of recombinant
60. Hofmeister F (1888) Zur Lehre von der methods and molecular engineering in protein
Wirkung der Saltz. Nauyn—Schniedebergs crystallization. Methods 34:354–363
Arch Exp Pathol Pharmakol 24:247 78. Derewenda ZS, Vekilov PG (2006) Entropy
61. McPherson A (2001) A comparison of salts for and surface engineering in protein crystalliza-
the crystallization of macromolecules. Protein tion. Acta Crystallogr D Biol Crystallogr
Sci 10:418–422 62:116–124
62. Sumner JB, Somers GF (1943) The enzymes. 79. DeLucas L (ed) (2009) Membrane protein
Academic Press, New York crystallization: current topics in membranes,
63. Englard S, Seifter S (1990) Precipitation tech- vol 63. Elsevier, Amsterdam
niques. Methods Enzymol 182:301–306 80. Loll PJ (2003) Membrane protein structural
64. Cohn EJ, Hughes WL, Weare JH (1974) biology: the high throughput challenge.
Crystallization of serum albumin from ethanol J Struct Biol 142:144–153
water mixtures. J Am Chem Soc 81. Michel H (ed) (1990) General and practical
69:1753–1761 aspects of membrane protein crystallization.
65. McPherson A (1976) Crystallization of pro- Crystallization of membrane proteins. CRC
teins from polyethylene glycol. J Biol Chem Press, Boca Raton, FL
251:3600–6303 82. Wiener MC (2004) A pedestrian guide to
66. Patel S, Cudney R, McPherson A (1995) membrane protein crystallization. Methods
Polymeric precipitants for the crystallization of 34:364–372
50 Alexander McPherson

83. Zulauf M (1990) Detergent phenomena in 85. Hunte C, Jagow G, Schagger H (2003)
membrane protein crystallization. In: Michel Membrane protein purification and crystalliza-
H (ed) Crystallization of membrane proteins. tion: a practical guide. Academic Press, San Diego
CRC Press, Boca Raton, FL 86. Weiner MC (2001) Existing and emergent
84. Garavito RM, Ferguson-Miller S (2001) roles for surfactants in the three-dimensional
Detergents as tools in membrane biochemistry. crystallization of integral membrane proteins.
J Biol Chem 276:32403–32406 Curr Opin Struct Biol 6:412–419