Understanding Crowding and Hydration Effects
on Protein Folding
Rationale
In recent years, the
importance of understanding the dominant forces in protein folding has gained
wider appreciation as more scientists have recognized that misfolding of proteins may contribute to the pathology of certain
human diseases. Our primary research goal is to understand the forces that govern
the folding and stability of macromolecules under conditions that mimic their
natural environment in a living cell. We are especially interested in the roles of
macromolecular crowding, confinement, hydration effects due to perturbed water
structure, cotranslational folding, and chaperone interactions. All of these factors may contribute to
the high fidelity of folding in vivo, but these phenomena are difficult
to investigate by the traditional approach of unfolding and refolding
biomolecules in dilute solution. The Eggers laboratory develops alternative experimental methods for testing the effects of crowding and confinement on the structure and stability of
macromolecules, including the use of sol-gel glass encapsulation.
A related goal is to study the effects of crowding agents and biological
surfaces on the properties of water and to evaluate the consequences of
perturbed water structure on protein-protein interactions.
Molecular
Crowding and Molecular Confinement
The cytoplasm of a living cell
is a concentrated mixture of macromolecules. In E.
coli, for example, the combined concentration of protein and RNA is in the
range of 300-400 mg/ml, occupying over 30% of the total volume (see Fig. 1). In addition,
membranes and elements of the cytoskeleton may define local regions of
confinement. The thermodynamic activity of a given macromolecule in this crowded
and confined environment may differ significantly from its activity in a dilute
(ideal) solution. Statistical-thermodynamic models for crowding and confinement
of macromolecules have been thoroughly investigated by
Allen P. Minton
(NIH) using
scaled particle theory. One striking prediction of these models is that the
association constant for a specific protein-protein interaction in vivo may
exceed that measured in dilute solution by 2-3 orders of magnitude! Another
pertinent prediction from the theoretical models is that proteins will tend to
favor compact or globular conformations as the extent of crowding (or uniform
confinement) is increased. This latter prediction has two important implications
for protein biochemists: (i) the initial folding of a nascent polypeptide
in vivo may be promoted by the crowded environment in which it is synthesized,
and (ii) the stability of a mature protein,
i.e. the ability to maintain a native globular state, may be much greater in
vivo than the values measured in vitro. The biological ramifications of
macromolecular crowding have been considered in several reviews by
R. John Ellis
(Warwick).
Fig.
1
The cytoplasm of a living cell is “crowded”. This illustration is
based on known concentrations and dimensions of macromolecules in a yeast cell.
Smaller proteins and metabolites are omitted for clarity. The large object in
the upper lefthand corner is a microtubule. Just below the tubule is an active
ribosome in complex with mRNA and two sock-shaped tRNA molecules. An
intermediate filament is seen on the right, and several smaller actin
filaments cross the figure horizontally. (Figure adapted from The Machinery
of Life, D.S. Goodsell, Springer-Verlag, New York, 1998).
Relatively
few experimental studies have appeared in the literature which address the
theoretical predictions of crowding and confinement. This is not too surprising
when one considers the inherent difficulties in mimicking the crowded
environment of the cytoplasm. Most commonly, investigators have employed a
concentrated solution of a (presumably inert) crowding agent to compare the
properties of a test molecule in the concentrated solution to its properties in
dilute solution. Favored crowding agents include proteins like albumin,
hemoglobin, and ribonuclease or other polymers such as polyethylene glycol
(PEG), dextran, and Ficoll.
One drawback of using concentrated solutions to study protein stability is that
any method you employ to perturb the structure of the "test" protein
will also perturb the structure and properties of the crowding agent, often
leading to irreversible aggregation events. Thus, most investigators employing this system start
with the test protein in a dilute unfolded state and perform a mixing experiment
with the crowded solution to follow the kinetics
and/or efficiency of refolding. By design, this type of experiment does not
assess the stability of the folded state in the crowded environment relative to
dilute solution.
The Sol-Gel Glass System
One
of our approaches is to encapsulate
macromolecules in a porous glass by the sol-gel technique. Sol-gel glasses are
formed in a two-step process from metal alkoxide precursors like
tetramethoxysilane. In the first step, methanol is released by acid hydrolysis
of the precursor forming the “sol”. Secondly, the mixture is neutralized
with aqueous buffer to initiate a condensation reaction that “gels” into a
network of [Si-O-Si] bonds. The sol-gel process is a fast-growing field of research
in material science because it allows one to immobilize labile molecules into
inorganic porous glasses under relatively mild conditions of temperature and pH.
Sol-gels can be formed into optically transparent shapes,
facilitating analysis of the entrapped molecules by many of the same
spectroscopic methods used for solutions.
Fig.
2 Cross-section of hypothetical glass
The
pore housing a macromolecule may be of the same order of magnitude in size as
the molecule itself, forming a confined
microenvironment not unlike that expected to exist in vivo. Two
protein molecules confined within the pores of a hypothetical silica glass
sample are shown in Fig. 2. Water and smaller
solutes may diffuse into the pores of the glass (white oval regions), but the proteins are unable to
escape under most conditions. In a typical wet-aged glass, the pores account for
85% of the total glass volume. Within this environment, the conformation of an encapsulated
protein molecule may be perturbed with heat or chemical denaturant without altering the
physical properties of the crowding agent (silica). Furthermore, intermolecular
aggregation of proteins is prohibited in the glass environment since the
molecules are isolated from each other. Certain experiments/solvents may be tested in this
system that are impossible to test in solution due to irreversible aggregation.
Summary of Results to Date
Two key papers employing the
sol-gel technique were published during postdoctoral studies in the laboratory
of Joan S. Valentine at UCLA (see publications). In this
work, we demonstrated that most proteins retain their native structures
following glass encapsulation, as monitored by circular dichroism spectra. In
addition, thermal denaturation experiments revealed that encapsulated proteins
unfold to a lesser extent upon heating as compared to the same protein in dilute
solution. In fact, for all proteins characterized to date, the fully unfolded
conformation (as defined in dilute solution) is never attained in the glass at
temperatures approaching 100 degrees C! The observed enhancement in thermal stability of glass-encapsulated
proteins may be viewed as strong experimental evidence in support of the theoretical studies
by Allen Minton.
There has been one important
exception to the results summarized above, that being the case of apomyoglobin.
Apomyoglobin is obtained by extraction of the noncovalent heme group and has
served as a model protein for folding studies for decades. Although this protein
maintains a compact globular conformation with much of its helical structure
intact, there is evidence that apomyoglobin is significantly less rigid and less
stable than its heme-bound counterpart. When apomyoglobin is encapsulated in the
silica glass, its structure appears grossly denatured by far-UV CD analysis.
Changing the pH or increasing the ionic strength of KCl has no influence on the
structure of glass-entrapped apomyoglobin, suggesting that electrostatic
interactions between the protein and the negatively-charged glass surface are
not responsible for the unfolded state. Further experiments have led to the
hypothesis that apomyoglobin is unfolded due to the altered properties of water
in the glass. Water structure (i.e., the hydrogen-bonded network of water
molecules) is known to be perturbed at a solid surface relative to its structure
in the bulk solution. We suspect that the glass boundary induces a layer of
water of high free energy (thermodynamically unfavorable). Since the
hydrophobic effect on folding equilibria is directly related to the formation of
unfavorable water structure at the surface of the unfolded protein, and since
this unfavorable water must partition to the protein surface from the
bulk phase, it seems logical
that the driving force for the hydrophobic effect is reduced by increasing the
average free energy of the bulk water. Presumably, apomyoglobin is very
susceptible to this change in the strength of the hydrophobic effect, leading to
its highly unfolded state within the glass pores (see Fig. 3). Deciphering the
subtle relationships between water structure, folding equilibria, solute
effects, and surface effects is the primary goal of our current and future
research.
Fig. 3
Role
of bulk water structure in protein folding equilibria and the hydrophobic
effect. A hypothetical two-state equilibrium is depicted for a folded globular
protein (top panels) and the unfolded random-coil conformation (lower panels) in
three different aqueous environments. (a) In an ideal dilute solution, the
folded conformation is favored due to a strong hydrophobic effect. (b) The
glass-entrapped protein is influenced by the unfavorable water structure at the
silica interface of the surrounding pores. The unfolded state predominates due
to a diminished hydrophobic effect. (c) Addition of compatible solutes reduces
the average free energy of the bulk water to a value that more closely resembles
neat water. The native state is favored due to the restored driving force of the
hydrophobic effect.
return
to top
