Thermal unfolding of Proteins is frequently
accompanied by the formation of
aggregates and therefore behaves as an
irreversible process. It occurs at temperatures
that vary widely according to the
protein, since the temperature of optimum
stability depends on the balance between
hydrogen bonds and hydrophobic interactions.
Generally, the products of thermal
denaturation are not completely unfolded
and retain some structured regions. At the
end of the thermal transition, the addition
of a denaturant such as urea or GdnHCl
frequently induces further unfolding.
An apparent irreversibility at a critical
concentration of denaturant has been observed
during the refolding of monomeric
as well as oligomeric proteins. It was reported
for the first time by M.Goldberg
and coworkers for the refolding of β-
galactosidase, and for tryptophanase. It
was also observed for a two-domain protein,
horse muscle phosphoglycerate kinase
by Yon and coworkers. In the latter
study, when the enzyme activity was used
as a conformational probe of the native
structure, an irreversibility was observed
for a critical concentration of denaturant
equal to 0.7 M± 0.1 MGdnHCl, a concentration
very close to the end of the transition
curve. Such irreversibility was found
to be concentration dependent. For protein
concentrations higher than 30 μM, restoration
of enzyme activity was practically null.
The formation of irreversible nonnative
species was found to be temperature dependent;
it was practically abolished at
4 ◦C, suggesting that aggregation occurs
through hydrophobic interactions. The aggregation
also depends on the time of
exposure of the protein to the denaturant.
When the unfolding–refolding process
was observed using structural signals such
as fluorescence or circular dichroism, it appeared
completely reversible whatever the
final denaturant concentration.
Another example is provided by rhodanese,
a two-domain monomeric protein.
During refolding at low denaturant concentration,
an intermediate accumulates
with partially structured domains and apolar
surfaces exposed to the solvent, leading
to the formation of aggregates. The aggregation
can be prevented by refolding the
protein in the presence of lauryl maltoside.
Most of the examples discussed above
are related to multidomain proteins. Another
degree of complexity appears in the
folding of oligomeric proteins. It is generally
accepted that the early steps of the
process are practically identical to the folding
ofmonomeric proteins. In the last step,
subunit association and subsequent conformational
readjustments yield the native
and functional oligomeric protein. The correct
recognition of subunit interfaces is
required to achieve the process. The overall
process of the folding of oligomeric proteins
was extensively studied by Jaenicke
and his coworkers for several enzymes and
described in reviews. As with monomeric
proteins, the formation of aggregates is
concentration dependent. The kinetics of
aggregation are complex and multiphasic,
indicating that several rate-limiting reactions
are involved in the process. In an
attempt to characterize these aggregates, it
was shown that noncovalent interactions
occur between monomeric species with
30 Aggregation, Protein
partially restored secondary structures.
The aggregates formed by either heat or
pH denaturation can be disrupted in 6 M
GdnHCl into monomeric unfolded species
and then renatured under optimal conditions
to yield an active enzyme.Only strong
denaturants such as high concentrations
of guanidine hydrochloride are efficient in
this disruption process.
The presence of covalent cross-links
such as disulfide bridges in a protein
molecule can complicate the refolding of
the denatured and reduced protein resulting
in the formation of incorrect and
intramolecular disulfide bridges leading
to further aggregation. The first welldocumented
studies were performed by
Anfinsen and his group on the refolding
of reduced ribonuclease. The authors
showed that the reoxidation of the enzyme
produces a great number of species with
incorrectly paired disulfide bonds. This
scrambled ribonuclease is capable of regaining
its native structure in a slow step,
a process that is accelerated by the addition
of a small quantity of reducing reagent
such as β-mercaptoethanol yielding about
100% of active enzyme. The reshuffling
of a protein’s disulfide bonds takes place
through a series of redox equilibria according
to either an intramolecular or
an intermolecular exchange. To prevent
a wrong pairing of half-cystine and further
aggregation, the addition of small
amounts of reducing reagents or redox
mixture is frequently used as investigated
by Wetlaufer.
The detection and characterization of
aggregates represent an important aspect
of folding studies. The aggregation phenomenon
can occur without precipitation.
Indeed, the degree of association of protein
intermediates during folding might
be small, depending on the intermolecular
interactions, and does not necessarily lead
to a visible insolubility. The association
state may be determined in several ways.
The most common methods, available
in any biochemistry laboratory, are gel
permeation and sodium dodecyl sulfate
polyacrylamide gel electrophoresis (SDSPAGE),
used both with and without crosslinking.
The detection of aggregates can
also be monitored by other hydrodynamic
methods suchas analytical ultracentrifugation
or classical light scattering. The latter
method also gives information on the size
of the aggregates. Quasi-elastic light scattering
is a dynamic technique that can be
used to determine macromolecule diffusion
coefficients as a function of time, that
is, to follow the kinetics of aggregation.
Neutron scattering can also be used to
detect protein aggregates, and mass spectrometry
has become a useful tool as well.