Neira1,2and Javier Go´mez1 1 Instituto de Biologı´a Molecular y Celular, Universidad Miguel Herna´ndez, Elche Alicante; 2 Instituto de Biocomputacio´n y Fı´sica de los Sistemas complejos
Trang 1The conformational stability of the Streptomyces coelicolor
histidine-phosphocarrier protein
Characterization of cold denaturation and urea–protein interactions
Jose´ L Neira1,2and Javier Go´mez1
1 Instituto de Biologı´a Molecular y Celular, Universidad Miguel Herna´ndez, Elche (Alicante); 2 Instituto de Biocomputacio´n y Fı´sica
de los Sistemas complejos, Zaragoza, Spain
Thermodynamic parameters describing the conformational
stability of the histidine-containing phosphocarrier protein
from Streptomyces coelicolor, scHPr, have been determined
by steady-state fluorescence measurements of isothermal
urea-denaturations, differential scanning calorimetry at
different guanidinium hydrochloride concentrations and,
independently, by far-UV circular dichroism measurements
of isothermal urea-denaturations, and thermal
denatura-tions at fixed urea concentradenatura-tions The equilibrium
unfold-ing transitions are described adequately by the two-state
model and they validate the linear free-energy extrapolation
model, over the large temperature range explored, and the
urea concentrations used At moderate urea concentrations
(from 2 to 3M), scHPr undergoes both high- and
low-temperature unfolding The free-energy stability curves have
been obtained for the whole temperature range and values of
the thermodynamic parameters governing the heat- and
cold-denaturation processes have been obtained
Cold-denaturation of the protein is the result of the combination
of an unusually high heat capacity change (1.4 ± 0.3
kcalÆmol)1ÆK)1, at 0Murea, being the average of the fluor-escence, circular dichroism and differential scanning calori-metry measurements) and a fairly low enthalpy change upon unfolding at the midpoint temperature of heat-denaturation (59 ± 4 kcalÆmol)1, the average of the fluorescence, circular dichroism and differential scanning calorimetry measure-ments) The changes in enthalpy (mDHi), entropy (mDSi) and heat capacity (mDC pi), which occur upon preferential urea binding to the unfolded state vs the folded state of the protein, have also been determined The mDHi and the mDSi are negative at low temperatures, but as the temperature is increased, mDHi makes a less favourable contribution than
mDS i to the change in free energy upon urea binding The
mDCpiis larger than those observed for other proteins; how-ever, its contribution to the global heat capacity change upon unfolding is small
Keywords: calorimetry; denaturant–binding interactions; histidine-phosphocarrier; protein stability
A full understanding of the physical interactions underlying the structure, folding and the function of a protein requires
a detailed description of its conformational stability in terms of the free energy of unfolding Such a thermo-dynamic description relies on the quantitative analysis of denaturant-induced or thermally induced folding-unfolding transitions, measured either spectroscopically or calorimet-rically In both cases, data analyses involves the extra-polation of the thermodynamic parameters to standard conditions, usually 298 K in the absence of denaturant To extrapolate thermal denaturation data, the change in DCp, and its temperature dependence must be known [1,2] The extrapolation of data from chemical-denaturation [with either urea or guanidinium hydrochloride (Gdm Cl) as denaturants] is carried out using either the linear free energy model, LEM [3–5], or the binding model [6] The LEM is by far th e most commonly used model, and it has been found to be valid for several proteins [7–9] Combined analysis of the LEM with thermal denaturation data, assuming a temperature-independent DCpand the thermo-dynamic equivalence between the thermally and chemically denatured states, have been reported for several proteins [10, 7 and references therein] These analyses yield the thermodynamic parameters governing the conformational
Correspondence to J L Neira and J Go´mez, Instituto de Biologı´a
Molecular y Celular, Edificio Torregaita´n, Universidad Miguel
Herna´ndez, Avda del Ferrocarril s/n, 03202, Elche (Alicante), Spain.
Fax: + 34 966658459, + 34 966658459, Tel.: + 34 966658467,
E-mail: jlneira@umh.es and jgomez@umh.es
Abbreviations: CD, circular dichroism; DSC, differential scanning
calorimetry; Gdm Cl, guanidinium hydrochloride; DC p , the heat
capacity change; mDCpi, the heat capacity change upon preferential
urea-binding to the unfolded protein vs the protein folded state; DH m ,
the calorimetric enthalpy change at T m ; m DH i , the enthalpy change
upon preferential urea-binding to the unfolded protein vs the protein
folded state; HPr, histidine phosphocarrier protein of the PTS;
scHPr, HPr from S coelicolor; bsHPr, HPr from B subtilis;
ecHPr, HPr from E coli; LEM, linear extrapolation method;
PTS, the phosphoenolpyruvate-dependent sugar
phosphotransferase system; DS m , the calorimetric entropy
change at T m ; m DS i , the entropy change upon preferential
urea-binding to the unfolded protein vs the protein folded
state; T m , thermal denaturation midpoint.
Dedication: This paper is dedicated to the memory of Jose´ Laynez.
(Received 27 January 2004, revised 24 March 2004,
accepted 2 April 2004)
Trang 2stability of the corresponding proteins, namely, the
enthalpy, entropy and the heat capacity changes Recently,
other chemical-denaturation models have been proposed
based on: (a) a local-bulk partitioning, where the
distribu-tion of denaturant between the surface of the protein and
the bulk solution is described by a partition coefficient
[11,12] or (b) a model-independent nonlinear extrapolation
procedure which considers a truncated Taylor expansion of
the Gibbs energy function [13] Both approaches have been
tested with model proteins and found to yield identical
conformational free-energies to those obtained by using the
LEM [11,13]
An essential step in the transport of carbohydrates across
the cell membrane of bacteria via the
phosphoenolpyruvate-dependent sugar phosphotransferase system (PTS) [14,15]
involves the transfer of a phosphoryl group from EI (enzyme
I of the PTS), the first protein in the cascade of proteins
forming the PTS, to HPr, the histidine-phosphocarrier
protein HPr is the smallest protein in the protein cascade
of the PTS and it is thought to be a key element in the
regulation of PTS as it is always present in the
phosphory-lation of any sugar [16] and is involved in gene reguphosphory-lation [14]
The structures of HPr proteins from several species have been
described by NMR spectroscopy [17–19 and references
therein] and X-ray diffraction [20,21] Those structures show
a classical open-face b-sandwich fold consisting of three
a-helices packed against a four-stranded antiparallel b-sheet
This fold is also shown by proteins with no apparent
involvement in any phosphorylation reaction [22,23]
Streptomyces species are soil-dwelling actinomycetes
which grow on a variety of carbon sources, such as cellulose
and several types of mono- and di-saccharides They are the
source of approximately two thirds of all natural antibiotics
currently produced by the pharmaceutical industry The
complete genome of Streptomyces coelicolor has been
sequenced, showing the largest number of genes found in
any bacteria [24] The presence of the different components
of the PTS in S coelicolor has been reported, and the
corresponding proteins cloned and expressed [25–27] HPr
of S coelicolor, scHPr, contains 93 amino acid residues; it
lacks cysteine and tyrosine residues, and it only contains one
tryptophan and one phenylalanine residue Assignment and
preliminary NMR studies of the HPr of S coelicolor
indicate that its structure is similar to that observed in
other members of the HPr family (J L Neira, unpublished
results) As scHPr has a similar structure, but a different
amino acid sequence to HPrs of Escherichia coli, ecHPr,
or Bacillus subtilis, bsHPr, whose structures and folding
properties have been described previously [9,22,28–30], it is
important to understand whether the structure, sequence or
both determine the conformational stability in the HPr
family There is a growing interest in determining to what
extent related proteins share the same conformational
stability features [31] For instance, bsHPr seems to fold
via a two-state process [28], as was thought to occur also in
ecHPr [9,22] Recently, however, the presence of non-native
contacts during ecHPr folding has been detected [30], and
structural rearrangements occurring upon folding around
an engineered tryptophan mutant have been observed [29]
Interestingly, scHPr seems to unfold at low pH via a
partially folded state [32] The similarity between both
partially folded states in both HPr species remains to be
elucidated, and it is also not yet known whether there is any relationship between the presence of those states and the different conformational stability between the HPr species [32] Thus, the study of the stability among the different HPr members will allow one to establish whether there is a common mechanism for the conformational stabilization in this important family In addition, the determination of these conformational stabilities could provide the evaluation
of the thermodynamic parameters governing protein– denaturant interactions, which, in turn, would shed light
on the still poorly understood mechanism of protein denaturation Attempting insight into those questions, we use a two-part strategy in this work First, we aimed to determine the thermodynamic parameters governing the conformational stability of scHPr (namely, DS, DH and
DCp), using different biophysical techniques [fluorescence, circular dichroism (CD) and differential scanning calorime-try (DSC)] and to compare these with the thermodynamic parameters obtained for other members of the HPr family, that is, bsHPr (where only CD measurements were performed) and ecHPr, where several biophysical tech-niques were also used by two independent groups [9,22] The use of different biophysical techniques allows comparison between the different thermodynamic data obtained and, thus, an assessment of the quality of the measurements Second, we aimed to determine the thermodynamic param-eters governing the urea–scHPr interactions and their temperature-dependence, and to compare them with those obtained in other proteins
Herein, it is shown that scHPr is only moderately stable
in aqueous solution Its DG upon unfolding is only 4.0 kcalÆmol)1at pH 7.5 at the temperature of maximum stability The analysis of the data performed at different temperatures validate the LEM The presence of moderate concentrations of urea as a denaturant agent (2–3M) strongly destabilizes the native state of the protein with cold-denaturation detectable at temperatures above 273 K The possibility to study both cold- and heat-denaturation over a range of urea concentrations has made possible the determination of the thermodynamic parameters governing first, the HPr unfolding and, second, the urea–protein interactions The combination of denaturant and heat-induced denaturation experiments gave proof that cold denaturation was a consequence of the combination of a large heat capacity change (1.4 ± 0.3 kcalÆmol)1ÆK)1, at 0
Murea, being the average of the fluorescence, CD and DSC measurements) and a low enthalpy change upon unfolding
at the midpoint temperature of heat-denaturation (59 ± 4 kcalÆmol)1, the average of the fluorescence, CD and DSC measurements) On the other hand, the enthalpy and entropy changes upon preferential urea-binding to the unfolded state vs the folded state are negative at low temperatures, but as the temperature is increased the enthalpy makes a less favourable contribution than the entropy to the free energy change upon urea–protein interaction Finally, the change in heat capacity and enthalpy upon urea-binding is larger (116 ± 4 calÆmol)1Æ
K)1ÆM )1), than those observed in ecHPr [9] and bsHPr [28], suggesting differential residual structure in the presence of urea among the three proteins However, the contribution
of mDCpi to the global heat capacity change upon unfolding
is small in the three proteins
Trang 3Experimental procedures
Materials
Ultra-pure urea used in fluorescence and CD, and the
Gdm Cl used in the DSC experiments were from ICN
Biochemicals Imidazole, trizma acid, its base, and NaCl
were from Sigma 2-Mercaptoethanol was from Bio-Rad,
and the Ni2+-resin was from Invitrogen Dialysis tubing
was from Spectrapore (Los Angeles, CA, USA) with a
molecular mass cut-off of 3500 Da Standard suppliers were
used for all other chemicals Water was deionized and
purified using a Millipore system Urea and Gdm Cl stock
solutions were prepared gravimetrically and filtered using
0.22 lm syringe-driven filters from Millipore Exact
con-centrations of urea and Gdm Cl stock solutions were
calculated from the refractive index of the solution using an
Abbe 325 refractometer [33]
Protein expression and purification
The HPr clone comprises residues 1–93, with an extra
methionine and a His6-tag at the N terminus We have
performed all the studies with this construction as its
structure, as observed by NMR (J L Neira, unpublished
results), is similar to that found in other members of the
HPr family and the His6-tag is disordered in solution,
making no contacts with the rest of the protein
Further-more, stability measurements and biophysical
characteri-zation have shown no differences between the His-tagged
protein and that where the tag had been removed [32]
Recombinant protein was expressed and purified as
described elsewhere [32] Protein was more than 99%
pure as judged by SDS protein-denaturing gels Also,
mass spectrometry analysis was performed in a
MALDI-TOF instrument, and only one peak was observed (data
not shown) The samples were dialysed extensively against
water and stored at )80 C Protein concentration was
calculated from the absorbance of stock solutions
meas-ured at 280 nm, using the extinction coefficients of model
compounds [34]
Fluorescence measurements
All fluorescence spectra were collected on a SLM 8000
spectrofluorometer (Spectronics Instruments, Urbana, IL,
USA), interfaced with a Haake water bath A 0.5-cm
path-length quartz cell (Hellma) was used
Urea-unfolding experiments were acquired by excitation
at 280 nm The slit width was typically equal to 4 nm for
the excitation light, and 8 nm for the emission light The
fluorescence experiments were recorded between 300 and
400 nm The signal was acquired for 1 s and the wavelength
increment was set to 1 nm Blank corrections were made in
all spectra The unfolding curves were obtained in 10 mM
phosphate buffer, pH 7.5, either by direct titration of the
protein solution with urea or by preparation of different
solutions containing a constant concentration of protein
and different urea concentrations (between 0 and 6M) Both
methods yielded superimposable sigmoidal plots for the
fraction of folded protein vs urea concentration and
identical m- and transition midpoint-values
Fluorescence spectra at different urea concentrations were processed using the wavelength averaged emission intensity, <k> [32]
Circular dichroism measurements
CD spectra were collected on a Jasco J810 spectropolari-meter fitted with a thermostated cell holder and interfaced with a Neslab RTE-111 water bath The instrument was calibrated periodically with (+)10-camphorsulphonic acid Isothermal wavelength spectra at different urea concentra-tions (between 0 and 6M) were acquired at a scan speed of
50 nmÆmin)1with a response time of 2 s and averaged over four scans at the desired temperature Far-UV measure-ments were performed using 20–40 lMof protein in 10 mM phosphate buffer (pH 7.5), using 0.1- or 0.2-cm path-length cells All spectra were corrected by subtracting the proper baseline The mean CD signal, [Q], was obtained from the raw ellipticity data,Q [32]
Thermal-denaturation experiments were performed at constant heating rates of 60CÆh)1and a response time of
8 s Measurements were acquired every 0.2C Thermal scans were collected in the far-UV region at 222 nm from
278 to 363 K in 0.1-cm path-length cells with a total protein concentration of 20 lM The reversibility of thermal tran-sitions was tested by recording a new scan after cooling down to 278 or 283 K the thermally denatured sample, and comparing the thermal denaturation curve with that obtained in the first scan In all studies carried out here, the experiments were fully reversible either for the heat- or cold-denaturation processes Thermal denaturation meas-urements were performed in the presence of different amounts of urea ranging from 0 to 3M (with maximum temperatures of 363 K) and from 5 to 6M(with maximum temperatures of 323 K) Sample exposure to high temper-atures was kept short to minimize any protein modification
by urea decomposition products and consequent irrevers-ibility The possibility of drifting of the CD spectropola-rimeter was tested by running two samples containing buffer, before and after the thermal experiments No difference was observed between both scans In all cases, after the reheating experiment, the samples were transparent and no precipitation was observed Care was taken to avoid loss of volume due to evaporation by using a cap that sealed the cuvette
To asses the reproducibility of trends in the data and fitted parameters, each of the CD measurements (either thermal or chemical denaturation experiments) was repea-ted twice in two independent sets in the temperature range explored In all the experiments both set of data yielded identical results
Differential scanning calorimetry DSC experiments were performed with a MicroCal MC-2 differential scanning calorimeter interfaced to a computer equipped with a Data Translation DT-2801 A/D converter board for instrument control and automatic data collection Lyophilized protein was dissolved in buffer (10 mM phos-phate buffer, pH 7.5) and dialysed extensively against 2 L
of the same buffer at 277 K All calorimetric experiments were performed at concentrations of 1 mgÆmL)1 Samples
Trang 4were degassed under vacuum for 10 min with gentle stirring
prior to being loaded onto the calorimetric cell Samples
were heated at a constant scan rate of 60CÆh)1and held
under a constant external pressure of 1 bar in order to avoid
both bubble formation and evaporation at high
tempera-tures Before rescanning, the samples were cooled in situ to
293 K for 40 min Experimental data were corrected from
small mismatches between the two cells by subtracting a
buffer vs buffer baseline, prior to data analysis After
normalizing to concentration, a chemical baseline calculated
from the progress of the unfolding transition was
subtrac-ted The excess heat capacity functions were then analysed
using the software packageORIGIN(Microcal Software, Inc.,
Northampton, MA, USA), supplied with the instrument
For the experiments in the presence of Gdm Cl, a stock
solution of 6M Gdm Cl concentration was used and
the corresponding amount of Gdm Cl was added Gdm Cl
was used in the DSC measurements, instead of the urea
employed in the CD measurements (see above), to avoid
deamidation processes Small concentrations of Gdm Cl
were used, because, as it has been shown [32], scHPr is
highly destabilized by the presence of Gdm Cl
No differences were observed in the thermodynamic
parameters obtained when different scan rates were used
(data not shown and [32])
Data analyses
Fitting of any equation described in this paragraph was
performed by using KALEIDAGRAPH (Synergy Software,
Reading, PA, USA) working on a PC
Data analysis of the isothermal urea denaturation
curves Far-UV CD and fluorescence chemical
denatura-tion data were analysed using the two-state model for the
native (N) to denatured (U) protein equilibrium According
to that model, the free energy governing the folding reaction
(DG) at a temperature T (in Kelvin), and the monitored
observable, X (either [Q] or <k>), are related [7–13, and
references therein] by:
X¼ðXNþ XUe
ðDG=RTÞÞ ð1 þ eðDG=RTÞÞ ð1Þ where XNand XUare the signals for the fully native and
fully unfolded states, respectively (the so-called baselines),
and correspond to the pre- and post-transition plateau
regions The complete analysis of the thermal- (see below)
and urea-denaturation data requires an accurate
determin-ation of both baselines, which can be described as linear
functions of temperature (in K) and urea concentration
[7,28]:
XNðT; ½ureaÞ ¼ X0
Nþ aN0Tþ bN0½urea ð2Þ
XUðT; ½ureaÞ ¼ X0
Uþ aU 0Tþ bU 0½urea ð3Þ where the first term in each equation is the corresponding
observable value at 273 K in the absence of urea, for the
native and the unfolded states, respectively; the second term
is the linear slope of the observable change with the
temperature; and, the last term is the effect of urea on the
baselines To allow for comparisons among the
thermal-(see below) and chemical-denaturations, the average emis-sion intensity or the mean CD signal were converted to plots of fU, the fraction of unfolded protein, which was then given by
fU¼ X XNðT; ½ureaÞ
XUðT; ½ureaÞ XNðT; ½ureaÞ ð4Þ Thermodynamic equations either in the presence or in the absence of chemical denaturant For a two-state unfolding reaction characterized by a temperature-independent heat capacity change, DCp, within the temperature interval under study, the equations for the dependence of the change in enthalpy (DH), entropy (DS) and free energy (DG) are given
by [1,2,35]:
DHðTÞ ¼ DHmþ DCpðT TmÞ ð5Þ
DSðTÞ ¼ DSmþ DCpln T
Tm
ð6Þ
DGðTÞ ¼ DHm 1 T
Tm
þ DCp T Tm T ln T
Tm
ð7Þ
In the above equations, Tmis the midpoint of the thermal transition [i.e the temperature at which DG(T)¼ 0], which
is taken as the standard reference temperature DHmand
DSmare the values of DH and DS at Tm, respectively Following the linear free-energy extrapolation model [3–5,36] the changes in DH, DS, DG and DCphave a linear dependence with denaturant concentration (the primes denote the corresponding values of the thermodynamic magnitude in the presence of urea):
DH0¼ DH mDHi½urea ð8Þ
DS0¼ DS mDSi½urea ð9Þ
DG0¼ DG m½urea ð10Þ
DC0p¼ DCp mDCpi½urea ð11Þ where m, mDHi, mDSi and mDCpi are the changes in free energy, enthalpy, entropy and heat capacity, respectively, associated with the preferential interaction of urea with the unfolded form of the protein relative to the folded form Assuming that mDCpi is temperature-independent, the tem-perature dependencies of m, mDHi, mDSi are then given by [8]:
mDHiðTÞ ¼ mmiDHiþ mDCpiðT Tm0Þ ð12Þ
mDSiðTÞ ¼ mmiDSiþ mDCpiln T
T0 m
ð13Þ
mðTÞ ¼ mDH i TmDS i¼ mmi
DH i
1 T
T0 m
þ mDC pi
T Tm0 T ln
T
T0 m
ð14Þ which are similar to Eqns 5, 6 and 7, respectively Here,
mmi
DH and mmi
DS are the values of mDHi and mDSi at the
Trang 5reference temperature Tm¢, which has been chosen as
the midpoint of the thermal denaturation (i.e where
m(T)¼ 0) Equation 14 indicates that the protein–urea
interactions are temperature-dependent
The temperature dependencies of the free energy, DG¢,
the enthalpy, DH¢, and the entropy, DS¢, at any urea
concentrations are given by equations identical to Eqns
(7, 5 and 6), respectively The temperature dependence of
DG¢ can also be described by the characteristic
temper-atures: Tm¢, Ts¢ and Th¢ [1,3], which are the temperatures
where DG¢, DS¢ and DH¢, respectively, are equal to zero
The equations describing the relationships between those
characteristic temperatures, and DH¢ and DCp¢ are
described elsewhere [8,37]
Thermally induced denaturation curves monitored by
far-UV CD The thermal denaturation curves obtained in the
presence of urea can be obtained by using Eqn (1) and
the expression of the DG¢ (which is analogous to Eqn 7)
The thermally induced denaturation data were converted to
plots of the fraction of protein in the unfolded state
according to Eqn (4) From this equation, the equilibrium
constant can be obtained in the folding transition region
and then DG¢ (i.e the free energy at a given urea
concentration) is determined as a function of T (in K) A
plot of DG¢ vs T at the melting temperature, Tm¢, yields a
slope equals to DHm¢/Tm¢ ¼ DSm¢, that is, the change in
entropy accompanying the unfolding reaction
The temperature at which scHPr was denatured by
cooling is described in the literature [1,37]
Results
In scHPr, the spectroscopic and chromatographic studies,
and the coincident equilibrium unfolding curves obtained
with different spectroscopic probes [32] are consistent with a
two-state folding behaviour The isothermal fluorescence
and far-UV CD urea-denaturation curves were, in all cases,
reversible Isothermal urea denaturation curves were
obtained at 10 different constant temperatures from 278
to 323 K, when followed by fluorescence, and at eight different temperatures from 278 to 318 K, when followed by far-UV CD However, because of the absence of a transition when thermal-denaturation was followed by fluorescence [32], the thermal-denaturation experiments were carried out
by observing the changes in ellipticity at 222 nm, using
far-UV CD
Fig 1 Urea-induced unfolding of scHPr monitored by the change in
intrinsic fluorescence spectra at 10 m M phosphate buffer, pH 7.5 In (A)
and (B), f U is plotted vs the concentration of denaturant (urea) at
selected temperatures ranging from 278 to 323 K The lines through
the data points represent the nonlinear least square fits to Eqn (1)
yielding the m- and [urea] half-values at each temperature (C)
Tem-perature dependence of the m-value from fluorescence measurements.
The errors bars are fitting errors to Eqn (1) The dotted line is the
linear temperature-dependence of the m-value, with a slope of
)8.6 ± 0.9 10 )3 kcalÆmol)1Æ M )1 ÆK)1 (D) Temperature dependence
of the Gibbs free-energy of unfolding The solid line represents
the nonlinear least square fit of the data to:
DGðT Þ ¼ DH ðT 0 Þ T DSðT 0 Þ þ DC p T T 0 T ln T
0
, which
is similar to Eqn (7) except that here T 0 , the reference temperature, was
taken as 298 K At 298 K, the enthalpy, DH, entropy, DS, and free
energy changes, DG, upon unfolding of scHPr obtained from the
fit-ting were 6.7 ± 0.5 kcalÆmol)1, 9.9 ± 1.5 cal mol)1ÆK)1 and
3.9 ± 0.2 kcalÆmol)1, respectively The temperature dependence of
DG was consistent with a temperature-independent heat capacity
change, DC p , of 1.57 ± 0.29 kcalÆmol)1ÆK)1 The inset represents the
average energy obtained at 298 K.
Trang 6Isothermal urea-denaturation monitored by changes
in the intrinsic fluorescence of the protein
Figure 1A,B shows fUas a function of urea concentration
for selected temperatures ranging from 278 to 323 K The
unfolding mechanism was consistent with the two-state
model at all temperatures Baselines were calculated from
Eqns (2 and 3) considering only the corresponding
observ-able value in the absence of urea (either X0
Nor X0
U), and the urea coefficient (either bN0 or bU0) for the folded and
unfolded species, respectively Each set of data was analysed
according to the LEM The calculated free energy changes
upon protein unfolding are plotted vs temperature in
Fig 1D
The m-values exhibited a slight tendency to decrease as
the temperature was raised from 1.43 ± 0.20 at 278 K to
0.98 ± 0.15 kcalÆmol)1ÆM )1 at 323 K This decrease was
linear within the temperature range explored, yielding
a slope of)8.6 ± 0.9 · 10)3kcalÆmol)1ÆM )1ÆK)1(Fig 1C)
Conversely, the [urea]½([urea]½is the urea concentration at
the transition midpoint) revealed a similar trend as
that observed for the temperature dependence of DG
(Fig 1D)
As shown in Fig 1D, the conformational stability of
scHPr was only moderate at neutral pH, reaching a
maximum of 4.0 ± 0.1 kcalÆmol)1 The free energy change
upon unfolding, DG, decreased both at higher and lower
temperatures The temperature dependence of DG was
consistent with an enthalpy change, DH(298 K) of
6.7 ± 0.5 kcalÆmol)1, an entropy change, DS(298 K) of
9.9 ± 0.9 calÆmol)1ÆK)1 and a temperature-independent
heat capacity change equal to 1.57 ± 0.29 kcalÆmol)1ÆK)1,
which is in good agreement with that determined by the
analysis of the thermal- and urea-denaturation data
obtained by far UV-CD (1.05 ± 0.08 kcalÆmol)1ÆK)1 at
0M urea, see below) The conformational stability vs
temperature curve predicted a temperature of 259 K for the
cold-denaturation and 335 K for heat-denaturation The
latter value is in good agreement with the results obtained
for the heat-induced denaturation experiments monitored either by DSC (Tm¼ 333.3 ± 3 K), by far-UV CD (Tm¼
340 ± 2 K) (see below), and even the predicted cold-denaturation temperature agrees with that determined by far-UV CD (see below)
Finally, the temperature dependence of the conforma-tional stability of scHPr by fluorescence reveals that the Th was 294 ± 2 K, while the Tswas 296 ± 2 K (using the equations described in the literature [8,37]) Both values are
in good agreement with those calculated from the heat-denaturation experiments in the presence of different urea concentrations followed by far-UV CD (see below) Heat-denaturation monitored by DSC
Figure 2 shows the excess heat capacity functions for the heat-induced denaturation of scHPr in 10 mMphosphate buffer, pH 7.5, in the presence of small quantities of Gdm
Cl, ranging from 0 to 0.2M The protein unfolds reversibly via a two-state mechanism The midpoint temperature of the transition, Tm, as well as its enthalpy change upon unfolding, DH(Tm), decreased as the concentration of Gdm
Cl increased from 0 [Tm¼ 338 ± 3 K and DH(Tm)¼
60 ± 2 kcalÆmol)1] to 0.2M [Tm¼ 328 ± 3 K and DH(Tm)¼ 47 ± 3 kcalÆmol)1) The unfolding of the protein is consistent with a DCp value of 1.4 ± 0.2 kcalÆ mol)1ÆK)1, which was determined from the slope of the linear plot of DH(Tm) vs Tm This value is in good agreement with that determined by far-UV CD at 0Murea (1.05 ± 0.08 kcalÆmol)1ÆK)1, at 0M urea, see below) fol-lowing the thermal- and chemical-denaturation curves and with that determined from the fluorescence experiments, following the urea-denaturation curves (see above) The thermodynamic parameters for the unfolding of the protein extrapolated at 298 K indicate that the conforma-tional stability of the protein is only moderate (DG¼ 3.8 ± 0.3 kcalÆmol)1), the enthalpic contribution is still favourable for the native state (DH¼ 5.4 ± 0.5 kcalÆ mol)1) and the entropic contribution unfavourable for the
Fig 2 The excess heat capacity function of scHPr at pH 7.5 in 10 m M phosphate buffer containing small quantities of Gdm Cl as destabilizing agent (0–0.2 M ) In all the condi-tions tested, the protein was shown to unfold reversibly by reheating the sample once it was cooled down The constant scanning rate was
60 CÆh)1and samples were heated up to
368 K Both excess heat capacity functions, from heating and re-heating scans, yielded virtually identical T m values, while the calori-metric enthalpy for the second scan was over 85% the value obtained for the first one The continuous lines represent the fitting of the experimental data to a two-state reversible model Inset: temperature dependence of the enthalpy change upon unfolding In this tem-perature range, the unfolding of scHPr was characterized by a temperature-independent heat capacity change upon unfolding of 1.4 ± 0.2 kcalÆmol)1ÆK)1.
Trang 7folded state (–TDS¼)1.6 ± 0.3 kcalÆmol)1) These values
are in close agreement with those obtained from the
isothermal urea-denaturations followed either by
fluores-cence (see above) or far-UV CD (see below)
Isothermal urea-denaturations followed by far-UV CD
Experimental data at selected temperatures, plotted as the
fraction of unfolded protein, fU, are shown in Fig 3A Also
at 298 K, the raw data at selected urea concentrations are
shown (Fig 3B) The m-values did show, over the examined
range, a slight linear temperature dependence, with a slope
of)5 ± 3 · 10)3kcalÆmol)1ÆM )1ÆK)1(Fig 3C) The value
of this slope agrees, within the error, with that observed
in the urea-denaturations followed by fluorescence (see
above) The larger error in the CD measurements could be
due to the inherent larger errors (when compared to
fluorescence) obtained in the determination of the m-values
by using CD, as it has been shown in other proteins [38]
The [urea]½, which are more accurately determined and
therefore less susceptible to experimental errors than the
m-values, did show a temperature dependence (Fig 3D)
similar to that observed for the free energy change of
unfolding in water, DG (Fig 3D)
Thermal denaturation at fixed urea concentrations
as monitored by far-UV CD
Figure 4 illustrates the effects of urea on the thermostability
of the protein At urea concentrations lower than 2M,
scHPr showed a single conformational transition within the
temperature interval studied Conversely, at urea
concen-trations larger or equal than 2M, scHPr showed two
conformational transitions (both following a two-state
mechanism): one at low temperatures and other at high
temperatures, corresponding to cold- and heat-denaturation
of the protein, respectively As urea concentration was
increased, the temperature for the heat-denaturation was
shifted at lower temperatures, while the midpoint
tempera-ture for the cold-denaturation increased (Fig 4B) Above
3.5M of urea, no significant thermal-transition was
observed (data not shown), which agrees with the results
obtained for isothermal chemical denaturation experiments
monitored by both steady-state fluorescence and far-UV
CD These thermal denaturation data were analysed to determine the DCponce the folded and unfolded baselines were determined, as discussed below The heat capacity change is, in principle, assumed to be temperature-inde-pendent, although it changes to a small extent with temperature [39,40] (see Discussion)
Determination of the pre- and post-transition regions The baseline for the fully folded protein in the CD experiments was generated as follows In 0M urea, the
CD data in the pretransition region (278–310 K) were a
Fig 3 Urea-induced unfolding of scHPr monitored by the change in
CD (A) Urea-denaturation curves at selected temperatures at pH 7.5
(10 m M phosphate) as monitored by the change in ellipticity at 222 nm
in the far-UV CD spectra The fraction of protein in the unfolded
form, f U , calculated using Eqn (4) is plotted as a function of urea
concentration at 278 K (s), 303 K (d) and 313 K (h) The inset
represents the changes in the raw ellipticity at 222 nm, 298 K The
solid lines through the data are the nonlinear least squares fits to
Eqn (1) (B) Raw CD data at 298 K at different urea concentrations.
(C) Temperature dependence of the m-value from CD measurements.
The error bars are the fitting errors to Eqn (1) The dotted line is the
linear temperature-dependence of the m-value, with a slope of
)5 ± 3 · 10 )3 kcalÆmol)1Æ M )1 ÆK)1 (D) The temperature dependence
of the [urea] ½ (right side, s) and DG (left side, h) The error bars are
fitting errors to Eqn (1) The line through the DG data is the fitting to
Eqn (7) The errors are larger at the high temperatures, because the
pretransition regions were shorter.
Trang 8function of temperature exclusively, and a linear fit provided
the intercept, X0
N, and temperature coefficient, aN0, as
defined in Eqn (2) These two parameters were combined
with the pretransition CD data (278–310 K) obtained in
the presence of urea concentrations lower than 2.0M to
evaluate the coefficient for the urea-dependent term The
native baseline was then (Eqn 2):
HN
where HNðT; ½ureaÞ is in units of degreeÆcm)2Ædmol)1at
222 nm, T is in K and [urea] is inM The indicated errors in
the above expression are the fitting errors to Eqn (1) The
urea-dependent term, bN0essentially shifted the baselines by
a constant amount in the thermal denaturation curves, and
it was very small for all the urea concentrations explored
The above expression was used for all the thermal
denatur-ation curves obtained, including those at 2, 2.5 and 3M
urea, where the protein was either not completely folded at
any temperature (3M) or was only folded for a small range
of temperatures (2 and 2.5M) (Fig 4B)
The baseline for the unfolded protein in the CD
experiments was obtained using the same approach
des-cribed by other authors [7,8,36] CD data for thermal
transitions in the presence of 2.5 or 3Murea in the
post-transition region (where the baseline was large enough,
Fig 4B) and those obtained for the fully unfolded protein at
concentrations larger than 5Murea, heated up to maximum
temperatures of 323 K (data not shown), were fitted
individually as linear functions of temperature, yielding
basically the same slope (the aU 0 parameter in Eqn 3) with
a value of)0.025 ± 0.005 degÆcm)2Ædmol)1ÆK)1 The fits among the data for the different urea concentrations are then parallel to each other, with an offset corres-ponding to the urea contribution (the bU 0 parameter in Eqn 3) This value was added individually for each thermal denaturation curve, including those carried out
at concentrations lower than 2M urea where the post-transition baseline is not defined over a wide enough temperature range (Fig 4B) The coincidence of the slope of the post-transition regions for the thermally and chemically unfolded forms of the scHPr indicate that both unfolded forms are thermodynamically equivalent Similar findings have been observed for both unfolded forms in other proteins [7–9,28]
Determination of DCp Once the native and unfolded transition regions were determined for all the thermal denaturation experiments, three different approaches were used to determine DCp
(a) Fitting of the CD thermal denaturation data at each urea concentration to Eqn (1) yielded, for the heat-denaturation, the DHm¢ and Tm¢ These values were estimated from a van’t Hoff analysis over a narrow temperature range (usually lower than 5C), where the unfolding transition occurs (i.e for fUbetween 0.4 and 0.6) All the thermal denaturation experiments were used in the plot, except that of 3M, where it was not possible to determine the pretransition region as the protein was not completely folded at any explored temperature (Fig 4B) The slope of a linear plot of DHm¢ vs Tm¢ was the DCp (a similar procedure has been used before in the determin-ation of DCpfrom the DSC measurements) The linear fit yielded a value of 1.3 ± 0.2 kcalÆmol)1ÆK)1(Fig 5A) (b) For the 2, 2.5 and 3Murea concentrations, the CD thermal transitions revealed both heat- and cold-denatura-tions In these cases, it is possible to obtain the complete free energy stability curve as the curve of DG¢ changes its sign twice (i.e DG¢ equals zero twice) It can be shown that Eqn (7) can also be written as [36]:
lnðK0apÞ ¼ ðDC
0
pþ DS0ðT0ÞÞ R
þ
ðDC0
pþ DS0ðT0ÞÞ
0ðT0Þ
RT0
T 0 T
DC
0 p
R ln
T0 T
ð15Þ where T0 is any chosen temperature reference If this temperature reference corresponded to either of the cold-,
Tc 0
m, or the heat-denaturation, Tm¢, temperatures (i.e the temperatures where DG¢ ¼ 0 kcalÆmol)1), then the first two terms in Eqn (15) are equal but of opposite sign Thus, if the chosen temperature is Tm¢ Eqn 15 is:
lnðK0apÞ ¼ðDC
0
pþ DS0
mÞ
ðDC0
pþ DS0
mÞ R
Tm T
DC
0 p
R ln
Tm T
ð16Þ The fitting parameters for the cold and heat-denaturation data of 2.5 and 3.0 urea are listed in Table 1, and Fig 5B
Fig 4 Temperature- and urea-concentration-dependence of the mean
residue ellipticity at 222 nm (A) s, 0 M ); h, 1 M and d, 1.5 M (B) s,
2 M ; h, 2.5 M and d (3 M ) Continuous lines through the data are the
fittings to Eqn (1), and the thermodynamic parameters governing such
transitions are given in Table 2.
Trang 9shows the free-energy stability curves for those urea
concentrations and 2M The fittings for the unfolding at
both temperatures (cold and heat) at 2.5 and 3.0M, yielded
the same DCp¢ (Table 1) Fitting of data at 2Murea did not
yield good results for the cold-denaturation, probably
because this process was only observed at its early stages
(Fig 5B) It is interesting to note that the DCP¢ obtained for
the heat-denaturation was the same, within the error,
among the three urea concentrations (Table 1)
Further-more, they were similar to that determined using the other two approaches (see before and the following paragraph) (c) Pace and Laurents have described a method where it is possible to obtain the DCp¢ by using the isothermal ation curves at any temperature, and the thermal denatur-ation data for any of the urea concentrdenatur-ations explored [39] In addition, the method also provides a validation of LEM Following that method, the entire DG¢ curve was determined
at any of the explored urea concentrations (from 0 to 2.5M) over a wide temperature range The results from fitting the experimental CD data with Eqn (7), at any urea concentra-tion, with DHm¢, Tm¢ and DCp¢ as variable parameters, are shown in Table 2 and Fig 5C Data at 3.0Murea were not taken into account because at this concentration, the folded protein is not present at any explored temperature (Fig 4B) The DCp¢ at 0M urea had the value of 1.05 ± 0.08 kcalÆ mol)1ÆK)1, at 0M urea, and the corresponding DCp (the y-axis intercept in Eqn 11), is 990 ± 60 calÆmol)1ÆK)1 From the data in Table 2, it seems that there was a slight trend in
DCp to increase with urea concentration, although this tendency was small and fitting the data (DCp vs urea concentration) to Eqn (11) yielded a slope of 123 ± 40 calÆ mol)1ÆK)1ÆM )1 (see Discussion) The value at 0 M urea (1.05 ± 0.08 kcalÆmol)1ÆK)1) is in good agreement with that determined from the van’t Hoff analysis and those deter-mined by fluorescence and DSC (see before)
Determination of the Th¢ and Ts¢ The values of the temperatures where the enthalpy and entropy are zero, Th¢ and Ts¢, respectively, were calculated by using equations described in the literature [8,37] and are listed in Table 2 Th¢ was observed to increase with the urea concentrations (Table 2) The vatiation of Th¢ is given by:
dT0 h d½urea¼
oðDHm=DCpÞ o½urea
T and then if DCp shows a small variation with urea concentration, and thus, it can be assumed to be nearly constant over the concentration range explored, the above equation yields:
dT0 h d½urea
1
DCp
oDHm o½urea
T
In scHPr, the value of DCpis [denaturant]-dependent (see Discussion), and the last approximation can not be strictly applied; this shows why Th¢ increased, but also why it did not change in a linear manner, even though there was a linear relationship between DHmand [urea] (Fig 5A) Conversely, the Ts¢ remained constant, within the error,
at any urea concentration This is due to the fact that Ts¢ is predicted to increase nonlinearly with urea concentration, according to:
dT0 s d½urea¼ T
0 s
oðDSm=DCpÞ o½urea
T
In the region of 2.5M urea, both temperatures become equal; at this temperature ( 296 K) the fully folded and unfolded states do not differ in enthalpy, entropy or in free energy Similar findings have been observed in ecHPr [9], barstar [8], and a lac repressor DNA-binding domain [11]
Fig 5 Analysis of thermal unfolding curves of scHPr monitored by
far-UV CD signal at 222 nm at various urea concentrations, pH 7.5 (10 m M
phosphate) (A) DH m ¢ vs T m ¢ obtained by the vant Hoff analysis of
thermal denaturation data The error bars are fitting errors to Eqn (1).
(B) Analysis of thermal denaturation data at 2 M (s), 2.5 M (h) and
3 M (d) using the method of Chen and Schellman [36] as described in
the text For the sake of clarity, only the data fit at 2.5 M to Eqn (16) is
shown (continuous line) (C) Fitting (solid lines) of DG¢ to Eqn (7)
according to the method of Pace and Laurents [39] for 0.5 M of urea
(j) and 1.5 M (d) Data at low temperatures were obtained by using
the LEM data (unfilled symbols), and those data at higher
tempera-tures were obtained from the thermal denaturation experiments at the
specified urea concentration (filled symbols) Error bars are from the
fitting to Eqn (1).
Trang 10Evaluation of the thermodynamical parameters
governing urea–protein interactions: m,mDHiandmDSi
By using Eqns (5–7), at different urea concentrations, and
the values of the DHm¢, DSm¢, Tm¢ and DCp¢ obtained by the
approach of Pace and Laurents [39] (Table 2), a detailed
analysis of the dependencies of DHm¢, DSm¢ and DCp¢ upon
urea concentration could be obtained Only the
thermo-dynamic parameters obtained from the analysis at 0, 0.5, 1
and 1.5Murea concentrations were taken into account, due
to the errors associated in the determination of the free
energy curve at the highest urea concentrations (2.0, 2.5 and
3M urea), where cold-denaturation was clearly observed
(Figs 4B and 5B)
In Fig 6, the dependencies of DH¢ and DS¢ are shown
at two selected temperatures, 293 and 318 K In both
cases, the errors in the determination of the
thermody-namic parameters are large, due to the large scattering of
the measured data At 293 K, DH¢ and DS¢ were
positive, and its absolute value increased linearly as urea
concentration changed This suggests a favourable
inter-action of urea with scHPr at this temperature (Fig 6A,B),
as it has been shown in other proteins [40] The compensation between both magnitudes leaded to a resultant value of DG¢ that decreased linearly as the urea concentration increased (Fig 6C) Conversely, at 318 K, DH¢ and DS¢ also had a positive value, but the absolute magnitude decreased slightly as the concentration of urea increased Here, DG¢ also decreased linearly as urea concentration increased, but it showed a better fit (Fig 6C) than those observed for DH¢ and DS¢ This observation does not imply any thermodynamic feature
of the so-called enthalpy–entropy compensation as besides the mainly artefactual nature of this correlation [41], large errors in the determination of both DG and
DH have been invoked as the main reason why this phenomenon is usually observed [42] The behaviour of DG¢ at the rest of the temperatures analysed was similar
to those described here for 293 and 318 K (data not shown) It is worth mentioning here that: (a) the DG¢ values obtained from the linear fits agreed well with the DG¢ values obtained directly from the LEM at the chosen temperatures (data not shown) and (b) the slopes
of DG¢ agree, within the experimental uncertainty, with
Table 1 Thermodynamic parameters of the cold- and heat-denaturation of scHPr at different urea concentrations Parameters were obtained by using the method of Chen and Schellman [36] at pH 7.5 (10 m M phosphate) Errors are fitting errors to Eqn (1).
Urea ( M )
Heat-denaturation Cold-denaturation
T m (K) DC p ¢ (kcalÆmol)1ÆK)1) DS m ¢ (calÆmol)1ÆK)1) T c
m (K) DC c 0
p (kcalÆmol)1ÆK)1) DS m ¢ (calÆmol)1ÆK)1)
2a,b 322.9 ± 0.2 1.7 ± 0.3 142 ± 37
2.5 b 320.1 ± 0.2 1.43 ± 0.04 103 ± 46 276.5 ± 0.5 1.43 ± 0.04 )108 ± 45
3 306.0 ± 0.4 1.8 ± 0.1 33 ± 43 295.4 ± 0.4 1.8 ± 0.1 )30 ± 44
a
Attempts to determine the cold-denaturation at this urea concentration were unsuccessful, probably because of the absence of enough data
in the cold-denaturation region of the curve (Fig 5B) b The values of the thermodynamic magnitudes for these two urea concentrations agree, within the error, with those determined in Table 2.
Table 2 Thermodynamic parameters of the thermally induced denaturation of scHPr at pH 7.5 (10 mM phosphate) at fixed urea concentrations T m ¢ was obtained from the fitting of the CD thermal denaturation data to Eqn (1) Fitting of the thermal and urea-denaturation data (using the approach of Pace and Laurents [39]) yielded similar values, within the error T m
c
¢ was determined using equations described in the literature [1,37] The errors are calculated from the propagation of fitting errors DH m ¢ , the enthalpy of the cold denaturation, was obtained from fitting to Eqn (1)
of the thermal denaturation CD data Fitting of the thermal and urea-denaturation data (using the approach of Pace and Laurents [39]) yielded similar values, within the error H m ¢ was determined from Eqn (5) with the value of DC p ¢ and T m c¢ listed in the table DS m ¢ and DS m c¢ were calculated from the ratesDHm0
T m orDHmc0
T mc0 , respectively (see text) The errors are calculated from the propagation of fitting errors T h ¢ was determined by using equations described in the literature [37] The errors are calculated from the propagation of fitting errors T s ¢ was determined by using equations described in the literature [8,37] The errors are calculated from the propagation of fitting errors DC p ¢ was obtained from fitting of the thermal- and urea-denaturation data (using the approach of Pace and Laurents [39]) Indicated errors are fitting errors to Eqn (7) at different urea concentrations.
Urea
( M ) T m ¢ (K) T m
c
¢ (K)
DH m ¢ (kcalÆmol)1)
DH mc¢ (kcalÆmol)1)
DS m ¢ (calÆmol-1ÆK-1)
DS mc¢ (calÆmol)1ÆK)1) T h ¢ (K) T s ¢ (K)
DC p ¢ (kcalÆmol)1ÆK)1)
0 340 ± 2 256 ± 2 58 ± 3 )58 ± 5 170 ± 9 )228 ± 10 285 ± 3 289 ± 2 1.05 ± 0.08 0.5 333.3 ± 0.3 262 ± 2 48 ± 2 )45 ± 3 144 ± 6 )171 ± 7 288 ± 3 291 ± 3 1.06 ± 0.09 1.0 332.6 ± 0.1 266 ± 2 43± 2 )40 ± 3 129 ± 6 )153 ± 8 291 ± 3 293 ± 3 1.03 ± 0.09 1.5 327.6 ± 0.3 267 ± 3 43 ± 2 )40 ± 4 131 ± 6 )151 ± 9 290 ± 1 292 ± 2 1.16 ± 0.09 2.0 a 322.2 ± 0.5 275 ± 2 31 ± 2 )29 ± 3 96 ± 10 )106 ± 8 294 ± 3 296 ± 3 1.15 ± 0.08 2.5a 313 ± 7 282 ± 3 24 ± 3 )23 ± 5 77 ± 10 )81 ± 9 296 ± 4 297 ± 4 1.4 ± 0.3
a The values of the thermodynamic magnitudes for these two urea concentrations agree, within the error, with those determined by using the approach of Chen and Schellman [36] ( Table 1).