Báo cáo khoa học: Kinetically controlled refolding of a heat-denatured hyperthermostable protein pot

Kinetically controlled refolding of a heat-denatured
hyperthermostable protein
Sotirios Koutsopoulos
1
, John van der Oost
2
and Willem Norde
1,3
1 Laboratory of Physical Chemistry and Colloid Science, Wageningen University, the Netherlands
2 Laboratory of Microbiology, Wageningen University, the Netherlands
3 Department of Biomedical Engineering, University of Groningen, the Netherlands
Hyperthermophilic microorganisms, which often
belong to the Archaea, are able to grow optimally at
100 °C or higher [1]. After their discovery, it was nec-
essary to revise our ideas about the mechanisms
involved in the maintenance of protein structural integ-
rity and function at elevated temperatures [2,3]. For
the stabilization of proteins at high temperatures,
a concerted optimization of structural features is
employed. These include reduced solvent-exposed sur-
face area [4], increased packing density [5–7], increased
core hydrophobicity [8,9], decreased length of surface
loops [6] and extended ion-pair networks [10–13]. Nat-
ure uses different combinations of the same structural
features to stabilize proteins that are adjusted to other
environmental conditions [14].
In this work, we have investigated the unfolding ⁄
refolding process of the extracellular endo-b-1,3-glu-
canase (LamA) from the hyperthermophilic micro-
organism Pyrococcus furiosus that flourishes in the
surroundings of low-depth undersea volcanic areas at

temperatures ranging from 70 to 103 °C [15]. Proteins
evolve through a balanced compromise between struc-
tural rigidity, allowing for the maintenance of the
native conformation at the physiological temperature
of the organism, and flexibility, which is required for
functionality. LamA is inactive at room temperature
and shows maximum enzymatic activity at 104 °C,
where ‘normal’ proteins from mesophilic organisms are
already denatured [2,16]. In the simplified picture
introduced 70 years ago by Anson and Mirsky [17],
protein heat denaturation was described as a two-state
transition between the native and the denatured state.
Nowadays, the idea of one or more intermediate states
is well established in many cases of heat-induced
Keywords
calorimetry; endo-b-1,3-glucanase;
hyperthermostable enzyme; protein
refolding
Correspondence
S. Koutsopoulos, Center for Biomedical
Engineering, Massachusetts Institute of
Technology, NE47-Room 307,
500 Technology Square, Cambridge,
MA 02139-4307, USA
Fax: +31 617 258 5239
Tel: +31 617 324 7612
E-mail:
(Received 30 July 2007, accepted
21 September 2007)
doi:10.1111/j.1742-4658.2007.06114.x

The thermal denaturation of endo-b-1,3-glucanase from the hyperthermo-
philic microorganism Pyrococcus furiosus was studied by calorimetry. The
calorimetric profile revealed two transitions at 109 and 144 °C, correspond-
ing to protein denaturation and complete unfolding, respectively, as shown
by circular dichroism and fluorescence spectroscopy data. Calorimetric
studies also showed that the denatured state did not refold to the native
state unless the cooling temperature rate was very slow. Furthermore, pre-
viously denatured protein samples gave well-resolved denaturation transi-
tion peaks and showed enzymatic activity after 3 and 9 months of storage,
indicating slow refolding to the native conformation over time.
Abbreviations
DSC, differential scanning calorimetry; DNS, 3,5-dinitrosalicylic acid; DH
cal
, calorimetrically determined enthalpy change; DH
vH
, van’t Hoff
enthalpy change; LamA, endo-b-1,3-glucanase; T
d
, denaturation temperature.
FEBS Journal 274 (2007) 5915–5923 ª 2007 The Authors Journal compilation ª 2007 FEBS 5915
protein denaturation: the native state is tightly folded,
the intermediate state(s) is functionally inactive and
partially folded in a non-native conformation(s), and
the unfolded state is characterized by significant
amounts of loosely structured domains. In terms of
biological functioning, the last two states are dena-
tured, but only the latter resembles the random coiled
conformation. A major question is whether the transi-
tions between these states belong to a sequence of
reversible processes that can be described thermody-

namically.
Apart from the profound applications of thermo-
zymes in biocatalysis and biotechnology, by studying
the thermal resistance properties of proteins we aim to
tackle one of the most challenging problems in modern
biophysics: what is the mechanism used by these pro-
teins to stabilize their three-dimensional structure and
sustain biological function at anthropocentrically
extreme temperatures?
Results
Calorimetry
Calorimetric studies of LamA in 0.01 m phosphate
buffer showed a single denaturation transition peak.
The denaturation temperature T
d
was dependent on
pH and shifted from c. 112 °C at pH 4.0 to 109 °Cat
pH 7.0 to 104 °C at pH 8.5 [18]. A typical thermogram
of LamA is shown in Fig. 1A (line a). Variation of the
scanning rate between 6 and 90 °CÆh
)1
did not affect
the T
d
, the shape of the endothermic peak or the
enthalpy associated with the transition. This suggests
that the thermal denaturation of LamA is not kineti-
cally controlled [19,20].
The calorimetric criterion introduced by Privalov &
Khechinashvili [21] to judge a two-state transition

requires that the calorimetrically determined enthalpy
change DH
cal
is equal to the van’t Hoff enthalpy
change DH
vH
, which may be calculated from the dif-
ferential scanning calorimetry (DSC) thermogram
using the equation
DH
vH
¼ 4RT
2
d
C
p;max
DH
cal
ð1Þ
where R is the ideal gas constant, T
d
is the denatur-
ation temperature and c
p,max
is the maximum heat
capacity, with regard to the peak baseline, which is
observed at the denaturation temperature. A two-state
model implies that transient intermediate states, which
should be distinguished from thermodynamically stable
intermediates such as the molten globule, are not pop-

ulated at the transition temperature [22]. The validity
of this criterion has often been argued and therefore
care should be taken when it is applied [20]. In the
case of LamA, the DH
cal
⁄ DH
vH
ratio deviates from
unity, yielding a value of c. 0.5, suggesting a non-two-
state transition. Furthermore, the standard functions
integrated in the microcal origin dsc software
(MicroCal Inc., Northampton, MA, USA) could not
fit the endotherms as a two-state transition.
The reversibility of the thermal transitions was
tested by cooling the protein sample to room tempera-
ture. Using different cooling rates between 15 and
90 °CÆh
)1
, no exothermic transition suggesting protein
refolding was observed (Fig. 1A, line a¢). Reheating
the protein solution in the calorimeter cell, after
cooling to room temperature, did not show an endo-
thermic peak (Fig. 1A, line b). Heating LamA to
C
B
A
Fig. 1. Heat capacity as a function of temperature for 0.5 mgÆmL
)1
LamA in 0.01 M phosphate buffer at pH 7.0. (A, B) Cooling and
reheating of LamA shows no reversible denaturation peaks. (C)

Transition observed when the first run is stopped just below the
denaturation temperature. The scan rates tested were between 6
and 90 °CÆh
)1
and, after each heating step, the sample was
allowed to cool to room temperature.
Kinetically controlled refolding of a denatured protein S. Koutsopoulos et al.
5916 FEBS Journal 274 (2007) 5915–5923 ª 2007 The Authors Journal compilation ª 2007 FEBS
temperatures just above T
d
did not result in a revers-
ible transition peak (Fig. 1B). In another experiment,
the sample was heated up to exactly T
d
and was then
allowed to cool to 25 °C. Subsequent reheating
revealed a transition at the same T
d
value (Fig. 1C),
which, however, was characterized by a heat exchange
which was c. 50% decreased relative to that associated
with the first peak. This suggests that approximately
half of the LamA molecules are irreversibly denatured
during the first partial transition (curve a) [20]. These
are standard experiments from which we may conclude
irreversible unfolding.
The first experimental evidence of the reversible
thermal denaturation of LamA was observed at very
slow cooling rates. As mentioned previously, fast cool-
ing did not result in an exothermic refolding peak.

However, when very slow cooling of 0.1 °CÆh
)1
was
applied, a refolding transition peak was observed at
the same temperature at which denaturation occurred
during the first heating step (Fig. 2). Furthermore, the
enthalpy released during refolding was similar to that
absorbed on denaturation. Notably, this protein sam-
ple, which was obtained during slow cooling, was
found to be enzymatically active and, on reheating, a
denaturation peak of slightly lower intensity was
observed at the same temperature as before.
Calorimetric tests of long-stored samples of dena-
tured LamA (i.e. after 3 and 9 months of storage at
) 20 ° C) showed well-resolved transition peaks at the
same temperatures as those observed during the first
heating step (Fig. 3). However, the enthalpies associ-
ated with these transitions were considerably lower.
This indicates that the number of refolded LamA mol-
ecules is smaller than that which initially gave the first
strong endothermic peak and, furthermore, that refold-
ing is time dependent. It is intriguing to suggest that, if
the system were given more time, more denatured
enzyme molecules would be natively refolded. How-
ever, it was not possible to test this because, in the
absence of antimicrobial agents, the denatured protein
sample was not stable for longer periods.
Transition phenomena similar to those shown in
Figs 1–3 were also observed for LamA in solutions at
pH 6.5 and pH 8.5.

Enzymatic activity
The specific enzymatic activity of LamA is 1547.6
unitsÆmg
)1
at 90 °C. LamA samples derived from very
slow cooling experiments were tested and showed
recovered activity up to 83% (Fig. 4) compared with
the activity of the untreated enzyme. In the fast-cooled
denatured samples of LamA from a standard DSC
experiment, the enzymatic activity was completely sup-
pressed. However, storage of these samples for 3 and
9 months resulted in a notable increase in the enzymatic
activity by 8% and 19%, respectively. The relatively
long time required for the LamA molecules to show
measurable activity reflects the slow kinetics of the
refolding process to the native conformation.
After heat incubation at 150 °C, samples of LamA
did not show detectable activity, even after 6 months
of storage at 4 °C.
Circular dichroism
The secondary structural characteristics of LamA in
solution were determined using far-UV CD (Fig. 5,
Fig. 2. Effect of cooling rate on the refolding of 0.5 mgÆmL
)1
LamA
in 0.01
M phosphate buffer at pH 7.0. Denaturation on heating (a)
at a heating rate of 0.1 °CÆh
)1
and the exothermic peak (b)

observed on cooling with the same scanning rate.
Fig. 3. LamA refolding as a function of time: (a) heat denaturation
of 0.5 mgÆmL
)1
native LamA in 0.01 M phosphate buffer at pH 7.0;
(b) reheating the same sample 3 months later; (c) reheating the
same sample 9 months later. Scan rate was 30 °CÆh
)1
.
S. Koutsopoulos et al. Kinetically controlled refolding of a denatured protein
FEBS Journal 274 (2007) 5915–5923 ª 2007 The Authors Journal compilation ª 2007 FEBS 5917
top panel). Spectral analysis suggested that the second-
ary structure of native LamA mainly consisted of
b-sheets and turns, up to c. 96%. Heat-denatured
LamA at 110 °C still contained 86% of b-structures,
but the amount of a-helices and random coils
increased to 4% and 10%, respectively. The far-UV
CD spectrum of LamA after heat treatment in the CD
cell unit at 110 °C resembled that obtained from the
same protein sample after cooling to 25 °C. Prolonged
heat incubation for 30 min at 150 °C resulted in the
collapse of the secondary structure, and the polypep-
tide chain of LamA appeared to be unordered (Fig. 5,
top panel, curve c).
Following the ellipticity of LamA in a closed cell at
220 nm as a function of temperature, we observed a
transition between 105 and 110 °C (Fig. 5, top panel,
inset). This transition could not be monitored further
because of the instrument’s temperature limitations. At
pH 7.0, the denaturation of LamA occurs at c. 109 °C,

which is just below the maximum scanning tempera-
ture of the instrument.
Fluorescence spectroscopy
The tryptophan fluorescence emission spectrum of
native LamA shows a maximum at 335 nm (Fig. 5,
bottom panel, curve a). Increasing the temperature of
the solution resulted in a gradual decrease in the fluo-
rescence intensity without a shift in the emission maxi-
mum. Such a decrease in the intensity is attributed to
increased tryptophan quenching as a result of thermal
motion [23]. In comparison with the spectrum of
native LamA at 25 °C, the spectral profile of dena-
tured LamA at 110 °C (Fig. 5, bottom panel, curve b,
corrected for the temperature effect on the intensity)
showed slightly decreased intensity with a red shift in
the emission maximum to 344 nm. This indicates a
structural distortion, accompanied by partial exposure
of previously confined tryptophan(s) to the solvent.
After incubation for 30 min at 150 °C and cooling
to 25 °C, the emission maximum shifted to 357 nm
(curve c) and the intensity decreased three-fold, sug-
gesting a collapsed tertiary structure [24].
Discussion
The unfolding ⁄ refolding process of LamA was moni-
tored using calorimetry. Analysis of the thermograms
Fig. 4. Enzymatic activity of native and heat-treated LamA at differ-
ent storage times. The activity was measured at 90 °C in 0.01
M
phosphate buffer at pH 7.0.
Fig. 5. Far-UV CD (top panel) and fluorescence emission (bottom

panel) spectra of 0.25 mgÆmL
)1
LamA in 0.01 M phosphate buffer
at pH 7.0. Curve a, native state (recorded at 25 °C); curve b, heat-
denatured at 110 °C (recorded at 110 °C; fluorescence spectrum
was corrected for the tryptophan emission yield which decreases
as a function of temperature); curve c, after heat incubation for
30 min at 150 °C (recorded at 25 °C). Top panel inset: thermal tran-
sition of LamA monitored by the molar ellipticity at 220 nm.
Kinetically controlled refolding of a denatured protein S. Koutsopoulos et al.
5918 FEBS Journal 274 (2007) 5915–5923 ª 2007 The Authors Journal compilation ª 2007 FEBS
suggested that the denaturation of LamA could not be
described as a two-state transition, because the calori-
metric criterion, i.e. the DH
cal
⁄ DH
vH
ratio, deviated
from unity. However, the difference between DH
cal
and DH
vH
may be caused by inherent difficulties in the
precise determination of the peak height and the inte-
grated area of the peak, which has an unusual shape.
This point needs further discussion. The heat capacity
of denatured LamA, represented by the post-transition
baseline, was unexpectedly low when compared with
the heat capacity of the native state (pre-transition
baseline). It should be noted that, in Fig. 1C, line b,

the second up-scan of pre-heated LamA and, in Fig. 3
line c, the slowly refolded protein showed the same
negative-like Dc
p
profile. Unless we assume that this is
caused by an instrument artefact operating at the
extreme end of its detection limit (i.e. 127 °C), such a
profile is commonly attributed to aggregation of dena-
tured protein molecules.
Previously reported time-resolved anisotropy data
did not show a significant increase in the hydrody-
namic radius of the heat-denatured LamA molecule
compared with the size of the native protein [25]. In
addition, the fact that slow cooling resulted in an
exothermic refolding transition suggests that the
aggregation of LamA is not very likely (unless we
assume that protein aggregation is a reversible pro-
cess). In the absence of aggregation, similar denatur-
ation profiles have been reported, but not discussed,
for other hyperthermostable proteins [26–30]. The
observed negative-like Dc
p
upon thermal denaturation
of LamA at 109 °C may stem, in part, from the
physicochemical properties of liquid water at temper-
atures approaching 110 °C (H. Klump, University of
Cape Town, South Africa, personal communication)
[31]. Several lines of evidence support this hypothesis:
(a) extrapolation of calorimetric data by Privalov
[32], in his review on calorimetry in 1979, showed

that the specific entropy of unfolding of several pro-
teins intersects at c. 110 °C; (b) Baldwin’s hydrocar-
bon model predicted that the entropy of mixing DS°
of a nonpolar compound with water is negative at
ambient temperatures, but approaches zero at
c. 113 °C [33] (when DS° is zero, the solution shows
ideal entropy of mixing and hydrophobic moieties
may be readily dissolved in hydrophilic medium); (c)
Shinoda [34] suggested that the increased solubility of
hydrocarbons at high temperatures also depends on
the enthalpy, which reflects the changes in the hydro-
gen bonding interactions in water surrounding the
nonpolar compound; (d) according to Ne
´
methy and
Scheraga [35–37], the interaction of hydrocarbons
with water at high temperatures results in changes in
the local structure of the clustered water molecules
adjacent to the hydrophobic surfaces (e.g. these water
molecules show less hydrogen bonding and, therefore,
are less hydrophilic). If the exposure of hydrophobic
groups to water on denaturation does not contribute
much to Dc
p
, perhaps other factors, such as solvation
of the protein’s polar groups, become more impor-
tant.
Moreover, we cannot exclude the possibility that, on
denaturation, the conformational changes in the pro-
tein molecule are such that, from the protein core,

more polar (compared to hydrophobic) amino acids
are exposed to the polar solvent. The resulting struc-
turally distorted, partially unfolded equilibrium inter-
mediates are probably related to the kinetic folding
intermediate reported by Park et al. [38]. Indeed, the
surface of the native LamA molecule contains a large
nonpolar fraction. A similar post-transition decrease in
heat capacity, lower than that expected for a com-
pletely unfolded polypeptide, was also observed in the
denaturation of the recombinant human growth hor-
mone [39]. Therein, it was suggested that the protein
retained residual structure and, hence, was not fully
hydrated after thermal denaturation.
Whatever the case may be in the LamA system,
these conjectures suggest that a negative Dc
p
on pro-
tein denaturation and unfolding may be possible. It is
also possible that the observed transition profile rests
on an eluding component or mechanism that has not
been considered so far.
A more detailed characterization of the state of the
protein at temperatures beyond 110 °C was not possi-
ble because of limitations in the existing instrumenta-
tion, which is not designed to operate at such
biologically extreme temperatures. We were able to
show that the state of denatured LamA was signifi-
cantly different from that of the native protein with
regard to secondary and tertiary structural elements
(Fig. 5). However, as the transition could not be

completely monitored by CD or fluorescence spec-
troscopy, we could not unambiguously determine
whether the post-transitional state of the protein rep-
resents unfolding, or if it is just an intermediate
which unfolds completely only upon heating to even
higher temperatures.
To answer the question about the state of LamA
at temperatures above the denaturation point (i.e. at
109 °C), we used a calorimeter with a scanning tem-
perature efficiency up to 200 °C (MC-DSC 4100, Cal-
orimetry Sciences Corporation, Lindon, UT, USA).
This experiment revealed that, following the main
transition peak at 109 °C, another small exothermic
peak appeared at 144 °C (Fig. 6), which suggests that
S. Koutsopoulos et al. Kinetically controlled refolding of a denatured protein
FEBS Journal 274 (2007) 5915–5923 ª 2007 The Authors Journal compilation ª 2007 FEBS 5919
the first peak does not represent complete unfolding,
and that specific protein domains remain folded up to
the temperature of the second transition. This implies
that refolding proceeds through an intermediate state.
Notably, this partially folded non-native state of
LamA, after thermal denaturation at 109 °C, may
refold to the native conformation either by ultra-slow
cooling immediately after denaturation or by storing
it for prolonged times. Slow cooling resulted in an
exothermic refolding peak, indicating protein refold-
ing. This was confirmed not only by enzymatic activ-
ity tests but also by calorimetric studies: (a) on
reheating the refolded protein, a DSC denaturation
profile similar to that observed in the first heating

step was found; and (b) heating a previously dena-
tured LamA sample gave, after long storage, well-
resolved denaturation peaks with partial recovery of
the enthalpy exchange.
In this work, we have shown that the temperature-
induced transition of LamA from the native to the
denatured state can be reversed if sufficient time is
given for the system to equilibrate. Irreversible dena-
turation is commonly observed on heating of pro-
teins. However, the effect of time is rarely considered,
even though theoretical studies have predicted that, in
the absence of aggregation, reversible transitions are
possible when slow relaxation is involved [40]. When
the system relaxes more slowly than the time window
of the measurement, i.e. the duration of the experi-
ment, we ‘see’ the process as irreversible, but, given
sufficient time, it may well be restored to the original
state. This is the case for the refolding of LamA. In
the article by Kaushik et al. [41], the unfold-
ing ⁄ refolding kinetics were investigated, and it was
shown that a predenatured hyperthermophilic pepti-
dase from P. furiosus could refold completely after
36 h of incubation at 32 °C. Refolding required a few
days on incubation at lower temperatures. These
experiments resemble those presented here, where we
showed reversible transition in long-stored frozen
LamA samples. It is interesting to speculate that this
behaviour may also be found in other heat-denatured
proteins: refolding to the native conformation may be
possible if sufficient time is given to the system.

By contrast with the above-mentioned study, we
were able to observe by calorimetry an exothermic
refolding peak on very slow cooling of denatured
LamA. It is not clear yet whether the denatured, par-
tially folded state of LamA is kinetically trapped as a
result of slow relaxation refolding kinetics, or whether
this state is a thermodynamically stable form trapped
in a local energy minimum of the energy distribution
funnel. We can speculate that one of the reasons for
the slow refolding process may be the relatively slow
cis ⁄ trans isomerization of one or more of the 18
prolines of the protein [42]. This process requires
hundreds of seconds to be completed [43]. LamA con-
tains only one cysteine, and therefore post-transitional
improper intramolecular disulfide bond formation is
not possible, which would result in irreversible pro-
tein denaturation. Chemical modification of amino
acids on protein denaturation is possible in denatur-
ation processes occurring at such high temperatures.
However, temperature-induced deamidation of gluta-
mines and asparagines could not be detected by mass
spectroscopy, because this chemical reaction did not
lead to significant changes in the protein mass. It is
conceivable that this possibility is not very likely
because LamA was able to refold to the active pro-
tein conformation: extensive chemical changes on
deamidation would irreversibly prevent correct protein
folding.
In conclusion, the calorimetric analysis showed that
the transition of LamA from the native state to a par-

tially unfolded intermediate was reversible if conditions
were selected to give the system sufficient time. There
is a strong biological argument supporting the conclu-
sions presented here: extracellular LamA is exposed to
temperature changes occurring in the microorganism’s
environment (e.g. volcanic underwater milieu). During
an environmental temperature change, such a system
relaxes to the initial state very slowly. During this per-
iod, denatured LamA molecules may recover their
native conformation and biological activity.
Fig. 6. Heat capacity as a function of temperature up to 200 °Cof
0.7 mgÆmL
)1
LamA in 0.01 M phosphate buffer at pH 7.0 (scan rate
was 30 °CÆh
)1
).
Kinetically controlled refolding of a denatured protein S. Koutsopoulos et al.
5920 FEBS Journal 274 (2007) 5915–5923 ª 2007 The Authors Journal compilation ª 2007 FEBS
Experimental procedures
Purification of LamA
The gene encoding LamA (GenBank accession
no.AF013169) was expressed in Escherichia coli (strain BL 21
DE3) and cloned into pGEF+ under control of the T7 pro-
moter [16]. Further purification was achieved by gel filtration
(Superdex 200, Amersham Pharmacia, Uppsala, Sweden).
The purity of the enzyme was tested using high-performance
liquid chromatography and matrix-assisted laser desorption
ionization-time of flight mass spectroscopy (PerSeptive Bio-
systems Voyager DE-RP mass spectrometer, Framingham,

MA, using sinapinic acid crystallized on a gold-coated well
plate; spectra were calibrated with protein standards). LamA
is a single domain globular-ellipsoid protein with a molecular
mass of 30 085 Da. Its isoelectric point is at pH 4.4, as deter-
mined by isoelectric focusing. Pure LamA was stored at
) 20 °C in 0.01 m phosphate buffer at pH 7.0, without anti-
microbial agents, which might affect the protein’s physico-
chemical characteristics.
Differential scanning calorimetry
Calorimetric studies were carried out in a VP-DSC calori-
meter (MicroCal Inc.). Very small heat exchanges of LamA
were recorded between 20 and 130 °C using, as reference,
the buffer solution. All samples were degassed under vac-
uum for 15 min prior to loading the cells, which were main-
tained under a pressure of 2.5 bar to avoid boiling of the
sample. The concentration of LamA was 0.5 mgÆmL
)1
;
experiments were also performed at different concentrations
between 0.1 and 2 mgÆmL
)1
with the same results normal-
ized per mass of the enzyme. Unless stated otherwise, the
temperature was increased at a rate of 30 °CÆh
)1
and, after
reaching the maximum desired value, the sample was
allowed to cool to room temperature. Heating rates
between 6 and 90 °CÆh
)1

were also used. A very slow cool-
ing rate of 0.1 °CÆh
)1
(i.e. 0.002 °CÆmin
)1
) was also tested.
The normalized excess heat capacity functions were
obtained after baseline subtraction and data processing
using the formulation of Privalov [32].
Enzymatic activity tests
The enzymatic activity of LamA before and after heat treat-
ment was measured using the colorimetric reagent 3,5-dini-
trosalicylic acid (DNS) [44]. This method is based on the
spectrophotometric determination of the hydrolysed ends of
oligosaccharides resulting from degradation of the substrate
(i.e. laminarin). For the assay, the hyperthermostable
enzyme and the substrate in 0.01 m phosphate buffer at
pH 6.5 were incubated for 10 min at 90 °C. The enzymatic
reaction was stopped by rapidly cooling the sample at room
temperature. After the addition of DNS, the sample was
incubated at 100 °C for 5 min and diluted (1 : 5, v ⁄ v) in
water. The sample was then allowed to cool to room tem-
perature and the absorbance was measured at 595 nm.
Circular dichroism measurements
CD spectroscopy was used to investigate the secondary
structure of LamA before and after heat treatment. Far-
UV (190–260 nm) CD spectra of 0.25 mgÆmL
)1
LamA in
quartz cuvettes (path length, 0.1 cm) were recorded in a

J-715 spectrophotometer (JASCO, Tokyo, Japan). The scan
rate was 100 nmÆmin
)1
, with a resolution of 0.2 nm and
response time of 0.25 s. Spectra were recorded in a closed
metal-caged quartz cuvette under pressure to prevent the
evaporation of water. The CD spectra of LamA after heat
incubation at 150 °C were collected on samples which had
been previously heated and then cooled to room tempera-
ture (heat incubation for 30 min at 150 °C was performed
in a temperature-controlled oil bath using thick-walled glass
tubes with a lid capable of withstanding the vapour pres-
sure of water). After subtraction of blank spectra, data
analysis was performed by fitting the spectra to reference
spectra using contin software [45,46].
Fluorescence spectroscopy
Fluorescence emission spectra of 0.025 mgÆmL
)1
LamA in
quartz cuvettes (path length, 1 cm) were recorded in the range
300–400 nm in a Varian Cary Eclipse spectrophotometer
(Palo Alto, CA). Spectra of denatured LamA at 110 °C were
recorded in a closed cuvette under pressure to prevent solvent
evaporation. Fluorescence spectra of LamA after heat incu-
bation at 150 °C were collected on samples that had been
cooled to room temperature. Excitation was set at 300 nm to
excite only the tryptophans. The excitation and emission slit
widths were 5.0 and 2.5 nm, respectively. All spectra were
corrected for the background emission peak of water.
Acknowledgements

The authors gratefully acknowledge discussions with
Horst Klump (University of Cape Town, South Africa)
on the shape of the denaturation profile of the protein
from calorimetric data. This research was supported
by an Individual Marie Curie Fellowship of the Euro-
pean Community programme ‘Improving Human
Research Potential and the Socio-Economic Knowl-
edge Base’ to S.K.
References
1 Brown JR & Doolittle WF (1997) Archaea and the pro-
karyote-to-eukaryote transition. Microbiol Mol Biol Rev
61, 456–502.
S. Koutsopoulos et al. Kinetically controlled refolding of a denatured protein
FEBS Journal 274 (2007) 5915–5923 ª 2007 The Authors Journal compilation ª 2007 FEBS 5921
2 Koutsopoulos S, van der Oost J & Norde W (2005)
Temperature dependent structural and functional
features of a hyperthermostable enzyme using elastic
neutron scattering. Proteins: Structure, Function,
Bioinformatics 61, 377–384.
3 Koutsopoulos S (2007) Protein hyperthermostability –
current status and beyond. FEBS J 274, 4011–4011.
4 Chan MK, Mukund S, Kletzin A, Adams MW & Rees
DC (1995) Structure of a hyperthermophilic tungstop-
terin enzyme, aldehyde ferredoxin oxidoreductase.
Science 267, 1463–1469.
5 Anderson DE, Hurley JH, Nicholson H, Baase WA &
Matthews BW (1993) Hydrophobic core repacking and
aromatic interaction in the thermostable mutant of T4
lysozyme Ser 117 fi Phe. Protein Sci 2, 1285–1290.
6 Britton KL, Baker PJ, Borges KM, Engel PC, Pasquo

A, Rice DW, Robb FT, Scandurra R, Stillman TJ &
Yip KSP (1995) Insights into thermal-stability from a
comparison of the glutamate-dehydrogenases from
Pyrococcus-furiosus and Thermococcus-litoralis. Eur J
Biochem 229, 688–695.
7 Russell RJM, Hough DW, Danson MJ & Taylor GL
(1994) The crystal structure of citrate synthase from the
thermophilic archaeon Thermoplasma-acidophilum.
Structure 2, 1157–1167.
8 Spassov VZ, Karshikoff AD & Ladenstein R (1995)
The optimization of protein–solvent interactions. Ther-
mostability and the role of hydrophobic and electro-
static interactions. Protein Sci 4, 1516–1527.
9 Schumann J, Bohm G, Schumacher G, Rudolph R &
Jaenicke R (1993) Stabilization of creatinase from Pseu-
domonas-putida by random mutagenesis. Protein Sci 10,
1612–1620.
10 Perutz MF & Raidt H (1975) Stereochemical basis of
heat stability in bacterial ferredoxins and in hemoglobin
A2. Nature 255, 256–259.
11 Korndo
¨
rfer I, Steipe B, Huber R, Tomschy A &
Jaenicke R (1995) The crystal-structure of holo-glyceral-
dehyde-3-phosphate dehydrogenase from the hyper-
thermophilic bacterium Thermotoga-maritima at 2.5-
angstrom resolution. J Mol Biol 246, 511–521.
12 Russell RJM, Ferguson JMC, Hough DW, Danson MJ
& Taylor GL (1997) The crystal structure of citrate syn-
thase from the hyperthermophilic archaeon Pyrococcus

furiosus at 1.9 angstrom resolution. Biochemistry 36,
9983–9994.
13 Matsui I & Harata K (2007) Implication for buried
polar contacts and ion pairs in hyperthermostable
enzymes. FEBS J 274, 4012–4022.
14 Unsworth LD, van der Oost J & Koutsopoulos S (2007)
Hyperthermophilic enzymes – stability, activity and
implementation strategies for high temperature applica-
tions. FEBS J 274, 4044–4056.
15 Fiala G & Stetter KO (1986) Pyrococcus furiosus sp.
nov. represents a novel genus of marine heterotrophic
archaebacteria growing optimally at 100°C. Arch Micro-
biol 145, 56–61.
16 Gueguen YW, Voorhorst GB, van der Oost J & de Vos
WM (1997) Molecular and biochemical characterization
of an endo-b-1,3-glucanase of the hyperthermophilic
archaeon Pyrococcus furiosus. J Biol Chem 272, 31 258–
31 264.
17 Anson ML & Mirsky AE (1934) The equilibrium
between active native trypsin and inactive denatured
trypsin. J Gen Physiol 17, 393–398.
18 Koutsopoulos S, van der Oost J & Norde W (2004)
Adsorption of an endoglucanase from the hyperthermo-
philic Pyrococcus furiosus on hydrophobic (polystyrene)
and hydrophilic (silica) surfaces increases protein heat
stability. Langmuir 20, 6401–6406.
19 Lepock JR, Ritchie KP, Kolios MC, Rodahl AM, Heinz
KA & Kruuv J (1992) Influence of transition rates and
scan rate on kinetic simulations of differential scanning
calorimetry profiles of reversible and irreversible protein

denaturation. Biochemistry 31, 12 706–12 712.
20 Shriver JW, Peters WB, Szary N, Clark AT & Edmond-
son AP (2001) Calorimetric analyses of hyperthermo-
phile proteins. Methods Enzymol 334, 389–422.
21 Privalov PL & Khechinashvili NN (1974) Thermody-
namic approach to the problem of stabilization of glob-
ular protein structure. A calorimetric study. J Mol Biol
86, 65–684.
22 Zhou Y, Hall CK & Karplus M (1999) The calorimetric
criterion for a two-state process revisited. Protein Sci 8,
1064–1074.
23 Gally JA & Edelman GM (1962) Effect of temperature
on fluorescence of some aromatic amino acids and pro-
teins. Biochim Biophys Acta 60, 499–509.
24 Lakowicz JR (1999) Principles of Fluorescence Spectros-
copy, 2nd edn. Kluwer Academic ⁄ Plenum Publishers,
New York.
25 Koutsopoulos S, van der Oost J & Norde W (2005)
Conformational studies of a hyperthermostable enzyme.
FEBS J 272, 5484–5496.
26 Klump HH, DiRuggiero J, Park J-B, Adams MWW &
Robb FT (1992) Glutamate dehydrogenase from the
hyperthermophile Pyrococcus furiosus. Thermal denatur-
ation and activation. J Biol Chem 267, 22 681–22 685.
27 DeDecker BS, O’Brien R, Fleming PJ, Geiger JH, Jack-
son SP & Sigler PB (1996) The crystal structure of a
hyperthermophilic archaeal TATA-box binding protein.
J Mol Biol 264 , 1072–1084.
28 Vetriani C, Maeder DL, Tolliday N, Yip KS-P, Stillman
TJ, Britton KL, Rice DW, Klump HH & Robb FT

(1998) Protein thermostability above 100°C. A key role
for ionic interactions. Proc Natl Acad Sci USA 95,12
300–12 305.
29 Wassenberg D, Welker C & Jaenicke R (1999) Thermo-
dynamics of the unfolding of the cold-shock protein
from Thermotoga maritima. J Mol Biol 289, 187–193.
Kinetically controlled refolding of a denatured protein S. Koutsopoulos et al.
5922 FEBS Journal 274 (2007) 5915–5923 ª 2007 The Authors Journal compilation ª 2007 FEBS
30 Karantzeni I, Ruiz C, Liu C-C & LiCata VJ (2003)
Comparative thermal denaturation of Thermus aquaticus
and Escherichia coli type 1 DNA polymerases. Biochem
J 374, 785–792.
31 Klump HH, Adams MWW & Robb FT (1994) Life in
the pressure cooker. The thermal unfolding of proteins
from hyperthermophiles. Pure Appl Chem 66, 485–489.
32 Privalov PL (1979) Stability of proteins. Small globular
proteins. Adv Protein Chem 33, 167–241.
33 Baldwin RL (1986) Temperature dependence of the
hydrophobic interaction in protein folding. Proc Natl
Acad Sci USA 83, 8069–8072.
34 Shinoda K (1977) ‘Iceberg’ formation and solubility.
J Phys Chem 81, 1300–1302.
35 Ne
´
methy G & Scheraga HA (1962) Structure of water
and hydrophobic bonding of proteins: I. A model for
the thermodynamic properties of liquid water. J Chem
Phys 36, 3382–3400.
36 Ne
´

methy G & Scheraga HA (1962) Structure of water
and hydrophobic bonding of proteins. II. Model for the
thermodynamic properties of aqueous solutions of
hydrocarbons. J Chem Phys 36, 3401–3417.
37 Ne
´
methy G & Scheraga HA (1962) Structure of water
and hydrophobic bonding of proteins. III. The thermo-
dynamic properties of hydrophobic properties in pro-
teins. J Phys Chem 66, 1773–1789.
38 Park S-H, Shastry MCR & Roder H (1999) Folding
dynamics of the B1 domain of protein C explored by
ultrarapid mixing. Nat Struct Biol 6, 943–947.
39 Kasimova MR, Milstein SJ & Freire E (1998) The con-
formational equilibrium of human growth hormone.
J Mol Biol 277 , 4009–4418.
40 Potekhin SA & Kovrigin EL (1998) Folding under in-
equilibrium conditions as a possible reason for partial
irreversibility of heat-denatured proteins: computer sim-
ulation study. Biophys Chem 73, 241–248.
41 Kaushik JK, Ogasahara K & Yutani K (2002) The
unusually slow relaxation kinetics of the folding–
unfolding of pyrrolidone carboxyl peptidase from a
hyperthermophile, Pyrococcus furiosus. J Mol Biol 316,
991–1003.
42 Privalov PL & Pothekin SA (1986) Scanning microcal-
orimetry in studying temperature-induced changes in
proteins. Methods Enzymol 131, 4–51.
43 No
¨

lting B (1999) Protein Folding Kinetics. Springer-
Verlag, Berlin.
44 Miller GL (1959) Dinitrosalicylic acid reagent for
determination of reducing sugar. Anal Chem 31, 426–
428.
45 Venyaminov SY, Baikalov IA, Shen ZM, Wu C-SC &
Yang JT (1993) Circular dichroic analysis of denatured
proteins – inclusion of denatured proteins in the refer-
ence set. Anal Biochem 214, 17–24.
46 Provencher SW & Glo
¨
ckner J (1981) Estimation of
globular protein secondary structure from circular-
dichroism. Biochemistry 20, 33–37.
S. Koutsopoulos et al. Kinetically controlled refolding of a denatured protein
FEBS Journal 274 (2007) 5915–5923 ª 2007 The Authors Journal compilation ª 2007 FEBS 5923