Eur J Biochem 271, 2165–2181 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.4142.x The conformational stability of the Streptomyces coelicolor histidine-phosphocarrier protein Characterization of cold denaturation and urea–protein interactions ´ ´ Jose L Neira1,2 and Javier Gomez1 Instituto de Biologı´a Molecular y Celular, Universidad Miguel Herna´ndez, Elche (Alicante); 2Instituto 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 denaturations at fixed urea concentrations The equilibrium unfolding 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 to M), scHPr undergoes both high- and lowtemperature 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 Colddenaturation of the protein is the result of the combination of an unusually high heat capacity change (1.4 ± 0.3 ´ Correspondence to J L Neira and J Gomez, Instituto de Biologı´ a ´ Molecular y Celular, Edificio Torregaitan, Universidad Miguel ´ Hernandez, 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; DCp, the heat capacity change; mDCpi , the heat capacity change upon preferential urea-binding to the unfolded protein vs the protein folded state; DHm, the calorimetric enthalpy change at Tm; mDHi , 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; DSm, the calorimetric entropy change at Tm; mDSi , the entropy change upon preferential urea-binding to the unfolded protein vs the protein folded state; Tm, thermal denaturation midpoint ´ Dedication: This paper is dedicated to the memory of Jose Laynez (Received 27 January 2004, revised 24 March 2004, accepted April 2004) kcalỈmol)1ỈK)1, at M urea, being the average of the fluorescence, circular dichroism and differential scanning calorimetry measurements) and a fairly low enthalpy change upon unfolding at the midpoint temperature of heat-denaturation (59 ± kcalỈmol)1, the average of the fluorescence, circular dichroism and differential scanning calorimetry measurements) The changes in enthalpy (mDHi ), entropy (mDSi ) and heat capacity (mDCpi ), 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 mDSi to the change in free energy upon urea binding The mDCpi is larger than those observed for other proteins; however, 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 thermodynamic description relies on the quantitative analysis of denaturant-induced or thermally induced folding-unfolding transitions, measured either spectroscopically or calorimetrically In both cases, data analyses involves the extrapolation 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 DCp and the thermodynamic equivalence between the thermally and chemically denatured states, have been reported for several proteins [10, and references therein] These analyses yield the thermodynamic parameters governing the conformational ´ 2166 J L Neira and J Gomez (Eur J Biochem 271) stability 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 distribution 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 phosphoenolpyruvatedependent 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 phosphorylation of any sugar [16] and is involved in gene regulation [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 Ó FEBS 2004 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 calorimetry (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 techniques 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 parameters 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 kcalỈmol)1 at 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 (2–3 M) 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 heatinduced denaturation experiments gave proof that cold denaturation was a consequence of the combination of a large heat capacity change (1.4 ± 0.3 kcalỈmol)1ỈK)1, at M urea, 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 ± kcalỈmol)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 ± calỈmol)1Ỉ K)1ỈM)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 Ó FEBS 2004 Conformational stability of HPr of S coelicolor (Eur J Biochem 271) 2167 Experimental 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 concentrations 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 Furthermore, stability measurements and biophysical characterization 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 MALDITOF 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 measured 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 pathlength quartz cell (Hellma) was used Urea-unfolding experiments were acquired by excitation at 280 nm The slit width was typically equal to nm for the excitation light, and nm for the emission light The fluorescence experiments were recorded between 300 and 400 nm The signal was acquired for s and the wavelength increment was set to 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 and M) 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, [32] Circular dichroism measurements CD spectra were collected on a Jasco J810 spectropolarimeter 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 concentrations (between and M) were acquired at a scan speed of 50 nmỈmin)1 with a response time of s and averaged over four scans at the desired temperature Far-UV measurements were performed using 20–40 lM of 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 60 °CỈh)1 and a response time of s Measurements were acquired every 0.2 °C 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 transitions 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 measurements were performed in the presence of different amounts of urea ranging from to M (with maximum temperatures of 363 K) and from to M (with maximum temperatures of 323 K) Sample exposure to high temperatures was kept short to minimize any protein modification by urea decomposition products and consequent irreversibility The possibility of drifting of the CD spectropolarimeter 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 repeated 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 phosphate buffer, pH 7.5) and dialysed extensively against L of the same buffer at 277 K All calorimetric experiments were performed at concentrations of mgỈmL)1 Samples Ĩ FEBS 2004 ´ 2168 J L Neira and J Gomez (Eur J Biochem 271) were degassed under vacuum for 10 with gentle stirring prior to being loaded onto the calorimetric cell Samples were heated at a constant scan rate of 60 °CỈh)1 and held under a constant external pressure of bar in order to avoid both bubble formation and evaporation at high temperatures Before rescanning, the samples were cooled in situ to 293 K for 40 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 subtracted The excess heat capacity functions were then analysed using the software package ORIGIN (Microcal Software, Inc., Northampton, MA, USA), supplied with the instrument For the experiments in the presence of Gdm Cl, a stock solution of M 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 denaturation 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 ), are related [7–13, and references therein] by: (see below) and chemical-denaturations, the average emission intensity or the mean CD signal were converted to plots of fU, the fraction of unfolded protein, which was then given by X XN T; ẵureaị 4ị fU ẳ XU T; ẵureaị XN T; ẵureaị 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]: DHTị ẳ DHm ỵ DCp T Tm ị 5ị   T 6ị DSTị ẳ DSm ỵ DCp ln Tm      T T DGTị ẳ DHm ỵ DCp T Tm À T ln Tm Tm ð7Þ In the above equations, Tm is the midpoint of the thermal transition [i.e the temperature at which DG(T) ¼ 0], which is taken as the standard reference temperature DHm and DSm are 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 DCp have a linear dependence with denaturant concentration (the primes denote the corresponding values of the thermodynamic magnitude in the presence of urea): DH0 ẳ DH mDHi ẵurea 8ị DS0 ẳ DS mDSi ẵurea 9ị DG0 ẳ DG mẵurea 10ị DC0p ẳ DCp mDCpi ẵurea 11ị DG=RTị XN ỵ XU e ị Xẳ ỵ eðÀDG=RTÞ Þ ð1Þ where XN and XU are 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 determination of both baselines, which can be described as linear functions of temperature (in K) and urea concentration [7,28]: XN T; ẵureaị ẳ X0N ỵ aN0 T ỵ bN0 ẵurea 2ị XU T; ẵureaị ẳ X0U ỵ aU0 T ỵ bU0 ẵurea 3ị 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- 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 temperature dependencies of m, mDHi , mDSi are then given by [8]: mDHi Tị ẳ mmi DHi ỵ mDCpi T Tm ị   T mDSi Tị ẳ mmi DSi ỵ mDCpi ln Tm   T mi mTị ẳ mDHi TmDSi ẳ m DHi À Tm    T ỵ mDCpi T Tm T ln Tm ð12Þ ð13Þ ð14Þ which are similar to Eqns 5, and 7, respectively Here, mmiDHi and mmiDSi are the values of mDHi and mDSi at the Ó FEBS 2004 Conformational stability of HPr of S coelicolor (Eur J Biochem 271) 2169 reference 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, and 6), respectively The temperature dependence of DG¢ can also be described by the characteristic temperatures: 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 farUV 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 DHmÂ/Tm ẳ DSmÂ, 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 Fig Urea-induced unfolding of scHPr monitored by the change in intrinsic fluorescence spectra at 10 mM phosphate buffer, pH 7.5 In (A) and (B), fU 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) Temperature 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 kcalỈmol)1ỈM)1ỈK)1 (D) Temperature dependence of the Gibbs free-energy of unfolding The solid line represents the nonlinear least square fit of the  data to: h i T DGT ị ẳ DH T0 ị T DST0 ị ỵ DCp T T0 À T Á ln T0 , which is similar to Eqn (7) except that here T0, 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 fitting were 6.7 ± 0.5 kcalỈmol)1, 9.9 ± 1.5 cal mol)1ỈK)1 and 3.9 ± 0.2 kcalỈmol)1, respectively The temperature dependence of DG was consistent with a temperature-independent heat capacity change, DCp, of 1.57 ± 0.29 kcalỈmol)1ỈK)1 The inset represents the average energy obtained at 298 K 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 farUV CD Ó FEBS 2004 ´ 2170 J L Neira and J Gomez (Eur J Biochem 271) Isothermal urea-denaturation monitored by changes in the intrinsic fluorescence of the protein Figure 1A,B shows fU as 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 observ0 able value in the absence of urea (either XN or XU ), 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 kcalỈmol)1ỈM)1 at 323 K This decrease was linear within the temperature range explored, yielding a slope of )8.6 0.9 à 10)3 kcalặmol)1ặM)1ặK)1 (Fig 1C) Conversely, the [urea]ẵ ([urea]½ is 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 kcalỈmol)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 kcalỈmol)1, an entropy change, DS(298 K) of 9.9 ± 0.9 calỈmol)1ỈK)1 and a temperature-independent heat capacity change equal to 1.57 ± 0.29 kcalỈmol)1ỈK)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 kcalỈmol)1ỈK)1 at M 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 (Tm ¼ 333.3 ± K), by far-UV CD (Tm ¼ 340 ± K) (see below), and even the predicted colddenaturation temperature agrees with that determined by far-UV CD (see below) Finally, the temperature dependence of the conformational stability of scHPr by fluorescence reveals that the Th was 294 ± K, while the Ts was 296 ± K (using the equations described in the literature [8,37]) Both values are in good agreement with those calculated from the heatdenaturation experiments in the presence of different urea concentrations followed by far-UV CD (see below) Heat-denaturation monitored by DSC Figure shows the excess heat capacity functions for the heat-induced denaturation of scHPr in 10 mM phosphate buffer, pH 7.5, in the presence of small quantities of Gdm Cl, ranging from to 0.2 M 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 [Tm ¼ 338 ± K and DH(Tm) ¼ 60 ± kcalỈmol)1] to 0.2 M [Tm ¼ 328 ± K and DH(Tm) ¼ 47 ± kcalỈmol)1) The unfolding of the protein is consistent with a DCp value of 1.4 ± 0.2 kcalỈ mol)1ỈK)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 M urea (1.05 ± 0.08 kcalỈmol)1ỈK)1, at M urea, see below) following 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 conformational stability of the protein is only moderate (DG ẳ 3.8 0.3 kcalặmol)1), the enthalpic contribution is still favourable for the native state (DH ¼ 5.4 ± 0.5 kcalỈ mol)1) and the entropic contribution unfavourable for the Fig The excess heat capacity function of scHPr at pH 7.5 in 10 mM phosphate buffer containing small quantities of Gdm Cl as destabilizing agent (0–0.2 M) In all the conditions tested, the protein was shown to unfold reversibly by reheating the sample once it was cooled down The constant scanning rate was 60 °CỈh)1 and samples were heated up to 368 K Both excess heat capacity functions, from heating and re-heating scans, yielded virtually identical Tm values, while the calorimetric 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 temperature range, the unfolding of scHPr was characterized by a temperature-independent heat capacity change upon unfolding of 1.4 ± 0.2 kcalỈmol)1ỈK)1 Ĩ FEBS 2004 Conformational stability of HPr of S coelicolor (Eur J Biochem 271) 2171 Fig Urea-induced unfolding of scHPr monitored by the change in CD (A) Urea-denaturation curves at selected temperatures at pH 7.5 (10 mM 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, fU, 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 ± · 10)3 kcalặmol)1ặM)1ặK)1 (D) The temperature dependence of the [urea]ẵ (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 folded state (TDS ẳ )1.6 0.3 kcalặmol)1) These values are in close agreement with those obtained from the isothermal urea-denaturations followed either by fluorescence (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 ± · 10)3 kcalỈmol)1ỈM)1ỈK)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 illustrates the effects of urea on the thermostability of the protein At urea concentrations lower than M, scHPr showed a single conformational transition within the temperature interval studied Conversely, at urea concentrations larger or equal than M, 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 temperature for the cold-denaturation increased (Fig 4B) Above 3.5 M 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 DCp once the folded and unfolded baselines were determined, as discussed below The heat capacity change is, in principle, assumed to be temperature-independent, 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 M urea, the CD data in the pretransition region (278–310 K) were a Ó FEBS 2004 ´ 2172 J L Neira and J Gomez (Eur J Biochem 271) basically the same slope (the aU0 parameter in Eqn 3) with a value of )0.025 ± 0.005 degỈcm)2Ỉdmol)1ỈK)1 The fits among the data for the different urea concentrations are then parallel to each other, with an offset corresponding to the urea contribution (the bU0 parameter in Eqn 3) This value was added individually for each thermal denaturation curve, including those carried out at concentrations lower than M urea where the posttransition 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] Fig Temperature- and urea-concentration-dependence of the mean residue ellipticity at 222 nm (A) s, M); h, M and d, 1.5 M (B) s, 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 function of temperature exclusively, and a linear fit provided the intercept, XN , 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 2.0 M to evaluate the coefficient for the urea-dependent term The native baseline was then (Eqn 2): HN T; ẵureaị ẳ 27:7 ặ 0:7ị ỵ 0:025 ặ 0:002ịT ỵ 0:6 ặ 0:2ịẵurea where HN T; ẵureaị is in units of degreeặcm)2ặdmol)1 at 222 nm, T is in K and [urea] is in M The indicated errors in the above expression are the fitting errors to Eqn (1) The urea-dependent term, bN0 essentially 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 denaturation curves obtained, including those at 2, 2.5 and M urea, where the protein was either not completely folded at any temperature (3 M) or was only folded for a small range of temperatures (2 and 2.5 M) (Fig 4B) The baseline for the unfolded protein in the CD experiments was obtained using the same approach described by other authors [7,8,36] CD data for thermal transitions in the presence of 2.5 or M urea in the posttransition region (where the baseline was large enough, Fig 4B) and those obtained for the fully unfolded protein at concentrations larger than M urea, heated up to maximum temperatures of 323 K (data not shown), were fitted individually as linear functions of temperature, yielding 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 heatdenaturation, the DHm¢ and Tm¢ These values were estimated from a van’t Hoff analysis over a narrow temperature range (usually lower than °C), where the unfolding transition occurs (i.e for fU between 0.4 and 0.6) All the thermal denaturation experiments were used in the plot, except that of M, 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 determination of DCp from the DSC measurements) The linear fit yielded a value of 1.3 ± 0.2 kcalỈmol)1ỈK)1 (Fig 5A) (b) For the 2, 2.5 and M urea concentrations, the CD thermal transitions revealed both heat- and cold-denaturations 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]: DC0p ỵ DS0 T0 ịị lnK0ap ị ẳ R      DC0p ỵ DS0 T0 ịị DG0 T0 ị DC0p T0 T0 ỵ ln RT0 R T R T ð15Þ where T0 is any chosen temperature reference If this temperature reference corresponded to either of the cold-, T c m , or the heat-denaturation, TmÂ, temperatures (i.e the temperatures where DG ẳ kcalỈmol)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: lnK0ap ị ẳ    DC0p ỵ DS0m ị DC0p ỵ DS0m ị Tm ỵ R R T   DC0p Tm ln ð16Þ À R T The fitting parameters for the cold and heat-denaturation data of 2.5 and 3.0 M urea are listed in Table 1, and Fig 5B Ó FEBS 2004 Conformational stability of HPr of S coelicolor (Eur J Biochem 271) 2173 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 denaturation curves at any temperature, and the thermal denaturation data for any of the urea concentrations 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 to 2.5 M) over a wide temperature range The results from fitting the experimental CD data with Eqn (7), at any urea concentration, with DHm¢, Tm¢ and DCp¢ as variable parameters, are shown in Table and Fig 5C Data at 3.0 M urea were not taken into account because at this concentration, the folded protein is not present at any explored temperature (Fig 4B) The DCp¢ at M urea had the value of 1.05 ± 0.08 kcalỈ mol)1ỈK)1, at M urea, and the corresponding DCp (the y-axis intercept in Eqn 11), is 990 ± 60 calỈmol)1ỈK)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 calỈ mol)1ỈK)1ỈM)1 (see Discussion) The value at M urea (1.05 ± 0.08 kcalỈmol)1ỈK)1) is in good agreement with that determined from the van’t Hoff analysis and those determined 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 Th¢ was observed to increase with the urea concentrations (Table 2) The vatiation of Th¢ is given by:   dT h oDHm =DCp ị ẳ oẵurea d½ureaŠ T Fig Analysis of thermal unfolding curves of scHPr monitored by farUV CD signal at 222 nm at various urea concentrations, pH 7.5 (10 mM phosphate) (A) DHm¢ vs Tm¢ obtained by the vanÕt Hoff analysis of thermal denaturation data The error bars are fitting errors to Eqn (1) (B) Analysis of thermal denaturation data at M (s), 2.5 M (h) and 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 temperatures were obtained from the thermal denaturation experiments at the specified urea concentration (filled symbols) Error bars are from the fitting to Eqn (1) shows the free-energy stability curves for those urea concentrations and M The fittings for the unfolding at both temperatures (cold and heat) at 2.5 and 3.0 M, yielded the same DCp¢ (Table 1) Fitting of data at M urea 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- 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:   dT0h oDHm %À DCp o½ureaŠ T d½ureaŠ In scHPr, the value of DCp is [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 DHm and [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:   dTs0 oDSm =DCp ị ẳ Ts0 oẵurea dẵurea T In the region of 2.5 M urea, both temperatures become equal; at this temperature (% 296 K) the fully folded and unfolded states 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] Ó FEBS 2004 ´ 2174 J L Neira and J Gomez (Eur J Biochem 271) Table 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 mM phosphate) Errors are fitting errors to Eqn (1) Heat-denaturation Cold-denaturation Urea (M) Tm (K) DCp¢ (kcalặmol)1ặK)1) DSm (calặmol)1ặK)1) T c (K) m DC c (kcalặmol)1ặK)1) p DSm (calặmol)1ặK)1) 2a,b 2.5 b 322.9 0.2 320.1 ± 0.2 306.0 ± 0.4 1.7 ± 0.3 1.43 ± 0.04 1.8 ± 0.1 142 ± 37 103 ± 46 33 ± 43 276.5 ± 0.5 295.4 ± 0.4 1.43 ± 0.04 1.8 ± 0.1 )108 ± 45 )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 Table Thermodynamic parameters of the thermally induced denaturation of scHPr at pH 7.5 (10 mM phosphate) at fixed urea concentrations Tm¢ 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 Tmc¢ was determined using equations described in the literature [1,37] The errors are calculated from the propagation of fitting errors DHm¢ , 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 Hm¢0 was determined from Eqn (5) with the value of DCp¢ and Tmc¢ listed in the table DSm¢ and DSmc¢ were c m calculated from the rates DHm or DHmc0 , respectively (see text) The errors are calculated from the propagation of fitting errors Th¢ was determined by T Tm using equations described in the literature [37] The errors are calculated from the propagation of fitting errors Ts¢ was determined by using equations described in the literature [8,37] The errors are calculated from the propagation of fitting errors DCp¢ 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) 0.5 1.0 1.5 2.0a 2.5a 340 333.3 332.6 327.6 322.2 313 Tm ¢ (K) Tm (K) DHm (kcalặmol)1) DHmc (kcalặmol)1) DSm (calặmol-1ặK-1) DSmc (calặmol)1ặK)1) Th (K) 256 262 266 267 275 282 58 ± 48 ± 43± 43 ± 31 ± 24 ± )58 )45 )40 )40 )29 )23 170 144 129 131 96 77 )228 )171 )153 )151 )106 )81 285 288 291 290 294 296 c 0.3 0.1 0.3 0.5 ± ± ± ± ± ± 2 3 2 2 ± ± ± ± ± ± 3 ± ± ± ± ± ± 6 10 10 ± ± ± ± ± ± 10 9 DCp (kcalặmol)1ặK)1) Ts (K) 3 3 289 291 293 292 296 297 ± ± ± ± ± ± 3 1.05 1.06 1.03 1.16 1.15 1.4 ± ± ± ± ± ± 0.08 0.09 0.09 0.09 0.08 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) Evaluation of the thermodynamical parameters governing urea–protein interactions: m, mDHi and mDSi 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 thermodynamic parameters obtained from the analysis at 0, 0.5, and 1.5 M urea 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 M 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 thermodynamic 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 interaction 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 Ó FEBS 2004 Conformational stability of HPr of S coelicolor (Eur J Biochem 271) 2175 Fig Urea concentration dependencies of DH¢ (A), DS ¢ (B) and DG¢ (C) at pH 7.5 (10 mM phosphate) The values for those three thermodynamic parameters are represented at 293 (h) and 318 K (s) The straight lines through the data are linear least square fittings to Eqs (8– 10), respectively To better appreciate the errors associated with the experimental measurements, the plots of DH¢ (A) and DS¢ (B) show the data on two different axis: 293 K on the left y-axis and 318 K on the right y-axis The y-axis intercepts and slopes of the lines are: (A) at 293 K, 5.1 ± 1.0 kcalỈmol)1 and 1.7 ± 1.0 kcalỈmol)1ỈM)1, respectively; and at 318 K, 33 ± kcalỈmol)1 and 4.4 ± 3.9 kcalỈmol)1ỈM)1, respectively; (B) at 293 K, 3.7 ± 3.3 calỈmol)1ỈK)1 and 0.7 ± calỈ mol)1ỈK)1ỈM)1, respectively; and, at 318 K, 98 ± 11 calỈmol)1ỈK)1 and 10 ± 12 calỈmol)1ỈK)1ỈM)1, respectively; (C), at 293 K, 3.5 ± 1.0 kcalỈmol)1, and 1.5 ± 0.8 kcalỈmol)1ỈM)1; and, at 318 K, 2.2 ± 1.8 kcalỈmol)1, and 1.2 ± 0.9 kcalặmol)1ặM)1 The values of m (the slope of DGÂ) agree, within the experimental uncertainty, with the m-values determined from the isothermal denaturation experiments (see text and Fig 3C) the slopes determined, from the fitting of fU to the urea concentration (Eqn 1, Figs and 3), that is the m-value at the corresponding temperatures An analysis similar to that described in the previous paragraph was performed at the eight chosen temperatures in the range 278–318 K, and the values of mDHi mDSi and m were determined (Fig 7) The slope of the straight line of mDHi vs T yielded mDCpi which had a value of 115 ± calỈ mol)1ỈK)1ỈM)1 mDHi was small at very low temperatures Fig Temperature-dependence of the thermodynamic parameters which govern the interaction of scHPr with urea at pH 7.5 (10 mM phosphate) (A) Temperature dependence of mDHi The straight line through the data is a linear least square fit of the data to Eqn (12), whose slope is mDCpi and it has a value of 115.6 ± 0.5 calỈmol)1Ỉ K)1ỈM)1 (B) Temperature dependence of mDSi The straight line through the data is a linear least square fit of the data to Eqn (13), whose slope is mDCpi and it has a value of 115 ± calỈmol)1ỈK)1ỈM)1 (C) Temperature dependence of m The solid line through the data is a least square fit of the data to Eqn (14) The fitting yield a mDCpi equals to 117 ± calỈmol)1ỈK)1ỈM)1 The errors shown in the values of mDHi , mDSi , and m are the error fits for the slopes in Eqns (8, and 10), respectively (278 K), reaching the zero at 279 K, and becoming increasingly positive as the temperature was raised (Fig 7B) mDSi was also small at low temperatures, and it made favourable contributions to m above 289 K (Fig 7C) Both sets of thermodynamic magnitudes resulted in a free energy curve which had a maximum, as it was expected from Eqn (14) The fitting to this curve yielded a value of mDCpi of 117 ± calỈmol)1ỈK)1ỈM)1, which agreed, within the error, with that determined from the variation of mDHi with the temperature Also, both values agree, although the experimental uncertainty is rather large, with the value ´ 2176 J L Neira and J Gomez (Eur J Biochem 271) determined from the variation of DCp¢ vs urea concentration (Eqn 11), which was 123 ± 40 calỈmol)1ỈK)1ỈM)1 (see above) Discussion The conformational stability and the cold-denaturation of scHPr The conformational stability of a protein, DG, is fully specified when the enthalpy and the entropy changes at a chosen reference temperature (DHm and DSm, respectively), and the heat capacity changes (DCp) are known By using a combination of isothermal urea- and thermal-denaturation experiments followed by CD, fluorescence and DSC, it has been possible to obtain the stability curve of scHPr at different urea concentrations and over a wide temperature range It is important to note here that three different biophysical techniques were used in this work to assess the reproducibility of trends in the experimental data and the fitted parameters Several proteins have been shown to undergo cold denaturation, such as the monomeric k repressor [43], CheY [44], the human fibroblast growth factor [45], other HPr family members [9,22,28], 434 Cro protein [7] and barstar ([8], and references therein) The main difficulty to detect the cold-denaturation has been the temperature at which it happens According to the thermodynamic equations [1,37], cold denaturation should occur at a temperature midpoint, T m c0 , below the freezing point of water for most proteins Then, to observe cold denaturation above the freezing point of water, proteins need their native state to be previously destabilized This means that the study of cold-denaturation is greatly facilitated by the presence of denaturants (chemicals or pH), as denaturants raise the cold-transition temperature and depress the freezing point of water [1,8,9,22,28] In the study of the stability of scHPr shown here, urea has been used to destabilize the folded state The temperature of maximal stability in scHPr, Ts, was 292 ± K (Table 2) (the average of the values obtained at the different urea concentrations) The stability of scHPr decreased as the temperature was lowered (or increased) below (above) 292 K, at any urea concentration (Fig 5C) The decrease in stability at low temperatures observed in the presence of urea can be rationalized by the changes in the thermodynamic parameters governing the urea–protein interactions Briefly, the free energy at any urea concentration, DG¢, depends on the conformational free energy of the isolated protein, DG, and the free energy corresponding to the urea–protein interactions, m, according to Eqn (10) As, at low temperatures, mDHi is small (Fig 7A), making favourable contributions to m (that is, the urea–protein interactions are favoured, Eqn 14), it can be concluded that at low temperatures, DG¢ is high at any urea concentration Comparison among the stability curves at different urea concentrations indicates that the free energy was also decreased when urea concentration was raised (Fig 5C), due to the increase in the number of urea–protein interactions For concentrations < 2.0 M destabilization of the native state of the protein was not enough to detect a significant Ó FEBS 2004 structural unfolding transition at low temperatures (Fig 4A) and only a transition at high temperatures (heat-induced) was observed However, at urea concentrations ‡ 2.0 M two transitions were observed: one at high temperatures (heat-induced) and other at lower temperatures (cold-induced) (Fig 4B) Some conclusions can be drawn from the calculation of the thermodynamic parameters governing both transitions (Table 2) First, it can be observed that as the urea concentration was increased the Tm¢ was reduced and, concomitantly, the T m c0 was increased Also, it is clear from Table that cold-denaturation is accompanied by a substantial decrease in entropy As it is also clear that at M urea the folded scHPr cannot have a higher conformational entropy than the cold-denatured protein, the large decrease in the entropy upon colddenaturation must be accounted for by a change in the entropy of the water This suggests that the hydrophobic effect, as it happens in barstar [8], ecHPr [9,22] and bsHPr [28] is the dominant force stabilizing scHPr On the other hand, it can also be seen from Table that DH m c0 had a negative value, which was similar, but of opposite sign to that observed in the heat-denaturation As it has been suggested [8], this decrease in the enthalpy upon colddenaturation must be due to the change in the enthalpy of interaction of water Why is cold-denaturation observed in scHPr? It has been concluded [1,2,7–9,22,28] that the observation of colddenaturation relies on the magnitude of the DCp and DH(298 K), which are the thermodynamic parameters involved in the determination of the cold-denaturation temperature [1,37] Proteins undergoing cold-denaturation have smaller DH(298 K) and larger DCp values than a protein of similar size The DH(298 K) of scHPr is 10 kcalỈmol)1; this value is smaller than those found in other proteins, which range from 20 to 80 kcalỈmol)1 (see [7–9,28] and references therein) Furthermore, the DCp of scHPr is 1.4 ± 0.3 kcalỈmol)1ỈK)1 (the average of the DSC, CD at M urea and fluorescence measurements), which yields a value per residue (taken into account that the Histag is disordered in solution, as shown by NMR (J L Neira, unpublished results), and that it does not contribute to the stability of the protein [32]), of 14 calỈmol)1ỈK)1, which is as high as that measured in the 434 Cro protein [7], where colddenaturation has also been observed However, that value is not as high as those observed in barstar [8] or other HPrs Cold denaturation was also observed in bsHPr [28] and ecHPr [9,22] and these processes were explained in similar terms to those described here (a small DH and a large DCp) In those HPrs, the complete set of thermodynamic parameters were determined using only the CD measurements, for bsHPr [29], and by using CD [9,22], fluorescence [22] and DSC measurements [22] for ecHPr Among the three HPr members, the DHm are similar for scHPr and bsHPr (59 ± in scHPr, the average of the DSC, the CD and fluorescence measurements, vs 59 ± kcalỈmol)1 in bsHPr [28]), but very different to that observed in ecHPr (76 ± kcalỈmol)1 [9], and 69 ± kcalỈmol)1 [22]) Conversely, the DCp values are the same, within the error, between ecHPr and scHPr (1.4 ± 0.3 in scHPr vs 1.45 ± 0.08 in ecHPr [9]); but they are different to that observed in bsHPr (1.16 ± 0.05 kcalỈmol)1ỈK)1 [28]), although it is within the error of the value measured in Ó FEBS 2004 Conformational stability of HPr of S coelicolor (Eur J Biochem 271) 2177 scHPr The larger error in the DCp of scHPr is due to the fact that it is the average of the three different measurements (DSC, CD and fluorescence measurements) As a consequence of the large DCp and small DH the stability curve of scHPr at M urea also shows an unusual feature The curve has a relatively high Tm (Table 2) and a small conformational stability at 298 K (% kcalỈmol)1, Figs 1C and 3C), at Ts Thus, a large DCp should sharpen the stability curve (and then a low Tm should be expected) [46–48] The small conformational stability at Ts can be explained by the large measured value of DCp, as: DG(Ts) ¼ DHm – (Tm ) Ts) DCp [37,49] To explain the larger Tm measured, the equation which allows the determination of Ts must be used [8,37]:   DHm Ts ¼ Tm exp À Tm DCp If Ts must be around room temperature, as it has been concluded from a statistical survey of the thermodynamic parameters of several proteins [47] and as it has been found here for scHPr (Table 2), the small DHm, the large DCp and the exponential function make Tm shift towards higher values (but not very high as Tm is also in the denominator of the exponential) This displacement of the Tm to higher values has also been explained by considering the changes in water structure upon heating and [50,51] The determination of DCp and its variation with urea and temperature Evaluation of the DCp from stability data requires the determination of the second derivative of the stability from the experimental measurements The possible conclusions should be interpreted with these experimental limitations in mind We describe, first, the temperature and urea dependence of DCp, and, second, this parameter is compared with those found in other proteins We cannot rule out the temperature-dependence of the heat capacity based on our data The use of the approach developed by Pace and Laurents [39] yields a precise estimation of DCp¢ (Table and Fig 5C) as the slope of the free energy curve changes from a large negative value to a large positive value upon decreasing temperature However, this does not prove the temperature-independence of the heat capacity as that dependence is provided by the free energy equation (Eqn 7) The use of the Chen and Schellman approach [36] to determine the DCp¢ at the cold- and heat-denaturation temperatures (Table and Fig 5B) and the fact that the same DCp¢ value was observed for both denaturations (Table 1) does not prove either the temperature-independence of DCp¢, as Eqns (15 and 16) are obtained on such an assumption [36] The consistency of the whole thermodynamic data seem to support, within the error of the experimental measurements, such assumption, but we cannot rule out a small temperature-dependence in the heat capacity, as it has been observed in other proteins [37,39,52] In fact, it has been suggested that the temperature-dependence of DCp follows a broad bell-shaped function with a maximal constant value around 313 K [53], based on the curvature in the heat capacity curve found at lower temperatures for the unfolded state [54] On the other hand, it is possible to conclude the ureadependence of DCp within the experimental uncertainties, based first, on the experimental variation of DCp with urea (Table 2); and, second, on the agreement between the slope obtained from that variation and the mDCpi determined from the variations of mDHi , mDSi and m with temperature (Fig 7) However, those slopes are small, 116 ± calỈ mol)1ỈM)1ỈK)1 (the average obtained from the mDHi and m curves) when compared to the DCp If DCp increases with urea concentration (consistent with a positive mDCpi ) a heat capacity increase should be expected to follow the transfer of model compounds from water to urea solutions Several studies have tried to address this question [55–59] These studies suggest that: (a) the transferral from water to urea of nonpolar groups results in a decrease in the heat capacity, being the process enthalpically disfavoured; (b) transferral of polar groups (from water to urea) results in an increase in the heat capacity, and the process is enthalpically favoured; and (c) the transferral (from water to urea) of the polypeptide backbone has either an increase or nearly zero value in the heat capacity T hese results can be rationalized considering that, upon transferral to urea, there is a disruption of solvent structure around nonpolar groups and formation of solvent structure around the polar groups and the polypeptide backbone It is possible that in scHPr, upon addition of urea, most of nonpolar groups increase their solvent accessibility, and thus the contribution of these groups to the DCp of the protein should decrease as urea concentration is raised; but, on the other hand, the contribution of the backbone should increase as urea concentration is raised as the protein is more disordered This latter increase should compensate for the decrease attained by the exposition of nonpolar groups It is also reasonable to assume that the contribution of ionic groups should be small, as the urea concentration is increased, because the majority of these groups are exposed in the folded (similar to other HPrs structures, J L Neira, unpublished results) and denatured states of the protein In ecHPr [9] and Drosophila Notch ankyrin repeats [60] the DCp decreased upon urea concentration On the other hand, in barnase (in urea) [52] and barstar (in Gdm Cl) [8], DCp increased as the denaturant of concentration was raised It seems that in all the proteins studied so far, except for barnase and scHPr, DCp always decreased with the urea concentrations and increased with the Gdm Cl ([60] and references therein) The increase in the DCp upon urea concentration can be rationalized by considering that at higher concentrations of urea, there would be more denaturant to interact with and then the exothermic enthalpic term superimposed to the intrinsic endothermic term of protein unfolding would be higher On the other hand, in the 434 cro protein [7] and the lac-repressor DNAbinding domain [11], no denaturant-dependence of DCp was observed Although these differences in the behaviour of mDCpi among the different protein models could indicate different compensating effects between the protein–urea interaction effects and the structural changes occurring upon unfolding, it must be borne in mind that the most probable origin of the discrepancies is the experimental uncertainty in determining the third derivative of the stability curve (mDCpi ), obtained, in most of the data, with only one biophysical technique ´ 2178 J L Neira and J Gomez (Eur J Biochem 271) Validation of the LEM in analysis of scHPr data The use of LEM in several proteins has resulted in the complete determination of the thermodynamic parameters governing the folding–unfolding transition; among those proteins are thioredoxin [4], barnase [52], barstar [8], ecHPr [9], bsHPr [28], and the 434 Cro protein ([7] and references therein) In the latter protein, the use of the denaturation binding model [6,61,62] has lead to small differences in the determination of the free energy of 0.4–0.8 kcalỈmol)1, when compared to that obtained by the LEM These small differences have also been observed in other proteins when both models have been used to determine the free energy [39,63] In both models, the basic statement is that denaturant interacts preferentially with the unfolded form of the protein According to the binding model [6,61,62], denaturants bind stoichometrically to proteins, and the preferential interaction is given by the large number of sites upon unfolding; however, there is not very much experimental evidence for such a stoichometric binding Conversely, according to the LEM, denaturants display many weak selective interactions with proteins, and the preferential interaction is given by the larger solvent-accessible surface upon unfolding Recently, two new models have been proposed The first approach is based on a modelindependent approach, yielding, except at low concentrations of denaturant, similar results to those obtained by LEM for the same proteins [13] The second approach uses a local-bulk partitioning model in which the urea distribution between the surface of the protein and the bulk solution is described by a concentration-independent partition coefficient [11,12]; the predicted free energy for the lac repressor DNA-binding domain agrees with that obtained by the LEM Thus, it seems that the LEM is the most widely used method and the other developed approaches agree well with the results provided by the linear free-energy extrapolation Application of the LEM to the thermodynamic data of scHPr accounts well for the measurement of scHPr stability, as shown by the agreement between the chemical- (where LEM was used to analyse the data) and thermal-unfolding data (Fig 5C) However, the determination of the m-value from an independent set of measurements (i.e the thermal denaturation experiments at different urea concentrations) other than those data obtained from the isothermal chemical-denaturations (where the LEM was applied) could provide additional validation of the use of LEM in scHPr It can be easily shown that a Taylor series expansion of the free-energy up to the second term yields [13]:   DHm dTm m1=2 ¼ 17ị Tm dẵurea where midpoints were obtained yields a slope of )9.9 ± KỈM)1 (data not shown) Using this slope, the enthalpies and thermal midpoints obtained by fitting the thermal CD denaturation data (Fig 5A) yield m½ which are between 0.9 ± 0.3 (at M) and 1.7 ± 0.4 (at M) kcalỈmol)1ỈM)1, which agree, within the error, with those values obtained by CD (Fig 3C) in the explored temperature range In scHPr, the m-value is temperature-dependent [either from CD (Fig 3C) or the fluorescence measurements (Fig 1C)], showing a reduction of its value as the temperature is increased (CD measurements) Similar negative slopes in the temperature-dependence of the m-value have been reported in the ovomucoid third domain [10] and ribonuclease A [64] Conversely, only in E coli ribonuclease H a positive slope was observed [65], and in the rest of the proteins studied (included ecHPr [9]) the m-value was not temperature-dependent ([60] and references therein) The decrease in the m-value as the temperature was raised must reflect the unfavourable entropy change and favourable enthalpy change, associated with the accumulation of urea in the vicinity of the protein surface which is exposed upon unfolding (see below) Although it has been suggested that factors other than changes in accessible surface areas play a major role in the determination of the m-values [64], recently, an alternative explanation to the temperaturedependence of m-values has been suggested based on the larger exposure of accessible surface area in the folded state, compared to the unfolded state, due to the increase in the local structural fluctuations upon temperature [11] These suggestions seem to support the findings observed in thermophylic ribonuclease H, where an absence of temperature-dependence in the m-value was observed, when compared to the E coli ribonuclease H (where a temperature-dependence was detected) [65], because local fluctuations in the thermophylic enzyme occurs with a lower frequency Based on those studies, it has been argued that the temperature-independence of the m-value among different proteins could be due to: (a) the small mDCpi [8]; (b) to temperature-compensating effects on the urea activity and the binding constants to the sites exposed upon folding [28]; or (c) the experimental uncertainties From the experiments described in this work, it could be concluded, based exclusively in the CD data, that the m-value in scHPr was nearly temperature-independent (Fig 3B) However, the use of several techniques and several approaches to determine the free stability curve and the repetition of the measurements to asses the reproducibility of trends in data and fitted parameters have allowed to conclude, within the experimental uncertainty, the temperature-dependence of the m-value in scHPr Within the experimental limitations to determine the thermodynamic parameters, we favour the last explanation as the most probable source of the observed temperature-independence behaviour of the m-value in some proteins  m1=2  @DG ẳ @ẵurea ẵureaẳẵurea1=2 ể FEBS 2004 and thus it can be assumed that m1=2 ffi m The plot of the different thermal midpoints (obtained from the thermal denaturation experiments at different urea concentrations, Fig 5A) vs the urea concentrations at which those Thermodynamics of the interaction between scHPr and urea The value of mDCpi in scHPr (116 ± cal mol)1ỈK)1ỈM)1, Fig 7) is larger than that observed in barstar [8] (53 ± 36 calỈmol)1ỈK)1ỈM)1), where a temperature-independence behaviour of the m-value was also observed, Ó FEBS 2004 Conformational stability of HPr of S coelicolor (Eur J Biochem 271) 2179 and also larger than those of ecHPr [9] (50 ± 20 calỈ mol)1ỈK)1ỈM)1), Drosophila Notch ankyrin domain [60] (70 ± 40 calỈmol)1ỈK)1ỈM)1), and a lac-repressor DNAbinding domain (< 20 calỈmol)1ỈK)1ỈM)1) (where temperature-dependence behaviours of the corresponding m-values were observed), but similar to that observed in barnase [52] (160 calỈmol)1ỈK)1ỈM)1, where a negative temperaturedependence of the m-value was also observed) Although it can be argued that these discrepancies are only due to the experimental uncertainty of the determinations, we can also speculate that the differences might reflect either variations in the residual structure present in each protein or larger structural changes upon denaturant-binding Comparison of the other thermodynamic parameters characterizing the binding to urea (mmi i and mmi i ) (Fig 7) DH DS with those from the unfolding of the protein (Table 2) shows that the former are much smaller than the latter Similar results have also been described in other proteins where thermodynamic magnitudes of protein-denaturant binding have also been measured [8,9,28] At 298 K, the enthalpy of urea binding at M of urea to scHPr is )8.2 ± 0.4 kcalỈmol)1 The measurements obtained for ribonuclease A and lysozyme are )8 to )10 kcalỈmol)1 [40] In barstar [8] and ecHPr [9] the enthalpy of interaction was less favourably ()3 kcalỈmol)1) than that of lysozyme, but not as large as that of scHPr It has been argued [9] that those differences among the binding enthalpies could be explained by the different size of the proteins, thus the smaller proteins would have less sites to interact with urea and the enthalpy would be less favourable As the number of amino acids in scHPr and ecHPr is nearly the same, there must be other additional factors affecting the binding to urea, which probably rely on the exact mechanism of the urea–protein interactions and/or the structural changes occurring upon unfolding To date, the exact mechanism of action of chemical denaturants is not known, although it is clear from structural [66] and calorimetric studies [40,52,67] that there is an interaction (a binding) to the folded and unfolded states of proteins, especially to certain side chains [68] Although it must be borne in mind that the urea-binding thermodynamic magnitudes have a large experimental error, it is tempting to suggest that the explanation for the behaviour of the thermodynamic parameters among the three HPrs studied could rely on the residual structure present in the proteins as urea concentration was increased Also, as a consequence of that residual structure, it can be suggested that those differences in the urea-binding thermodynamic parameters are associated with the presence of partially folding species For instance, partially folded species have been observed during the kinetic folding pathway of ecHPr [29,30], where a small favourable-enthalpy interaction with urea was observed [9]; conversely, an equilibrium partially folded species has been observed at low pH in scHPr [32], where a larger enthalpy interaction with urea has been observed (this work); unfortunately, ecHPr precipitated at low pHs, which hampered the detection of any partially folded conformation [22] No partially folded species have been detected in bsHPr either kinetically or at equilibrium, but no thermodynamical denaturant-binding parameters have been reported to date Conclusions To sum up, we have shown here that the LEM can be used to obtain the thermodynamic parameters governing the unfolding of scHPr and its urea-binding properties The scHPr does not have a very high conformational stability, although its heat-temperature of unfolding is high Because of the large change in heat capacity and its small enthalpy of unfolding, a cold-denaturation has been observed in the presence of moderate urea concentrations Conversely to that observed in other proteins of the same family, the values of: (a) the change in the heat capacity upon ureabinding; and (b) the change in the enthalpy upon ureabinding are high, probably due to larger structural changes occurring upon binding to denaturant Acknowledgements We thank Fritz Titgemeyer for the generous gift of the recombinant plasmid encoding the scHPr We thank F N Barrera for critical reading of the manuscript and for stimulating ideas This work was supported by Projects BIO2000-1081 (to J G.), from the Spanish Ministerio de Ciencia y Tecnologı´ a, CTIDIB-2002/6 (to J L N.) from Generalitat Valenciana, and FIS01/0004–02 (to J L N.) from the Spanish Ministerio de Sanidad y Consumo We gratefully thank Marı´ a ´ T Garzon, May Garcı´ a, Marı´ a C Fuster and Javier Casanova for excellent technical assistance We thank the two anonymous reviewers for their critical insights and suggestions References Privalov, P.L (1992) The physical basis of stability of folded conformations of proteins Protein Folding (Creighton, T.E., ed.), pp 83–126 W.H Freeman, New York Freire, E (1995) Differential scanning calorimetry in methods in molecular biology In Protein Stability and Folding: Theory and Practice (Shirley, B.T., ed.), pp 191–218 Humana Press, Towota, New Jersey Schellman, J.A (1987) The thermodynamic stability of proteins Annu Rev Biophys Biophys Chem 16, 115–137 Santoro, M.M & Bolen, D.W (1992) A test of the linear extrapolation of unfolding free energy changes over an extended denaturant concentration range Biochemistry 31, 4901–4907 Greene, R.F & Pace, C.N (1974) Urea and guanidine hydrochloride denaturation of ribonuclease, lysozyme, alpha-chymotrypsin, and beta-lactoglobulin J Biol Chem 249, 5388–5393 Tanford, C (1970) Protein denaturation C Theoretical models for the mechanism of denaturation Adv Prot Chem 21, 1–95 ´ Padmanabhan, S., Laurents, D.V., Fernandez, A.M., Elı´ as-Arnaz, M., Ruiz-Sanz, J., Mateo, P.L., Rico, M & Filimonov, V.V (1999) Thermodynamic analysis of the structural stability of phage 434 Cro protein Biochemistry 38, 15536–15547 Agashe, V.R & Udgaonkar, J.B (1995) Thermodynamics of denaturation of barstar: evidence for cold denaturation and evaluation of the interaction with guanidine hydrochloride Biochemistry 34, 3286–3299 Nicholson, E.M & Scholtz, J.M (1996) Conformational stability of the Escherichia coli HPr protein: test of the linear extrapolation method and a thermodynamic characterization of cold denaturation Biochemistry 35, 11369–11378 10 Swint, L & Robertson, A.D (1993) Thermodynamics of unfolding for turkey ovomucoid third domain: thermal and chemical denaturation Protein Sci 2, 2037–2049 11 Felitsky, D.J & Record, M.T Jr (2003) Thermal and ureainduced unfolding of the marginally stable lac repressor DNA- ´ 2180 J L Neira and J Gomez (Eur J Biochem 271) 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 binding domain: a model system for analysis of solute effects on protein processes Biochemistry 42, 2202–2217 Courtenay, E.S., Capp, M.W., Saceker, R.M & Record, M.T Jr (2000) Thermodynamic analysis of interactions between denaturants and protein surface exposed on unfolding: interpretation of urea and guanidinium chloride m-values and their correlation with changes in accessible surface area (ASA) using preferential interaction coefficients and the local-bulk domain model Proteins 41, 72–85 ´ Ibarra-Molero, B & Sanchez-Ruiz, J.M (1996) A model-independent, nonlinear extrapolation procedure for the characterization of protein folding energetics from solvent-denaturation data Biochemistry 35, 14689–14702 Postma, P.W., Lengeler, J.W & Jacobson, G.R (1993) Phosphoeneolpyruvate carbohydrate phosphotransferase system of bacteria Microbiol Rev 57, 543–594 Bruckner, R & Titgemeyer, F (2002) Carbon catabolite represă sion in bacteria: choice of the carbon source and autoregulatory limitation of sugar utilization FEMS Microbiol Lett 209, 141– 148 Meadow, N.D., Fox, D.K & Roseman, S (1990) The bacterial phosphoenolpyruvate: glycose phosphotransferase system Annu Rev Biochem 59, 497–542 Wittekind, M., Rajagopal, P., Baranchini, B.R., Reizer, J., Saier, M.H Jr & Klevitt, R.E (1992) Solution structure of the phosphocarrier protein HPr from Bacillus subtilis by two-dimensional NMR spectroscopy Protein Sci 1, 1363–1376 Van Nuland, N.A.J., Hangyi, I.W., Van Schaik, R.C., Berendsen, H.J.C., Van Gusteren, W.F., Scheek, R.M & Robillard, G.T (1994) The high-resolution structure of the histidine-containing phosphocarrier protein HPr from Escherichia coli determined by restrained molecular dynamics from nuclear magnetic resonance nuclear Overhauser effect data J Mol Biol 237, 544–559 Maurer, T., Doker, R., Gorler, A., Hengstenberg, W & Kalbitzer, H.R (2001) Three-dimensional structure of the histidine-containing phosphocarrier protein (HPr) from Enterococcus faecalis in solution Eur J Biochem 268, 635–644 Herzberg, O., Reddy, P., Sutrin, S., Saier, M.H Jr, Reizer, J & Kapafia, G (1992) Structure of the histidine-containing phos˚ phocarrier protein HPr from Bacillus subtilis at 2.0- A resolution Proc Natl Acad Sci USA 89, 2499–2503 Jia, Z., Quail, J.W., Waygood, E.B & Delbaere, L.T.J (1993) The ˚ 2.0-A resolution structure of Escherichia coli histidine-containing phosphocarrier protein HPr: a redetermination J Biol Chem 268, 22490–22501 van Nuland, N.A.J., Meijberg, W., Warner, J., Forge, V., Scheek, R.M., Robillard, G.T & Dobson, C.M (1998) Slow cooperative folding of a small globular protein HPr Biochemistry 37, 622–637 Pastore, A., Saudek, V., Ramponi, G & Williams, R.J.P (1992) Three-dimensional structure of acylphosphatase Refinement and structure analysis J Mol Biol 224, 427–440 Bentley, S.D., Chater, K.F et al & Hoopwood, D.A (2002) Complete genome sequence of the model actinomycete Streptomyces coelicolor A3 (2) Nature 417, 141–147 Titgemeyer, F., Walkenhorst, J., Cui, X., Reizer, J & Saier, M.H Jr (1994) Proteins of the phosphoenolpyruvate: sugar phosphotransferase system in Streptomyces: possible involvement in the regulation of antibiotic production Res Microbiol 145, 89–92 Titgemeyer, F., Walkenhorst, J., Reizer, J., Stuiver, M.H., Cui, X & Saier, M.H Jr (1995) Identification and characterization of phosphoenolpyruvate: fructose phosphotransferase systems in three Streptomyces species Micorobiology 141, 51–58 Parche, S., Schmid, R & Titgemeyer, F (1999) The PTS system of Streptomyces coelicolor: identification and biochemical analysis of a histidine phosphocarrier protein HPr encoded by the gene ptsH Eur J Biochem 265, 308–317 Ó FEBS 2004 28 Scholtz, J.M (1995) Conformational stability of HPr: the histidine-containing phosphocarrier protein from Bacillus subtilis Protein Sci 4, 35–43 29 Azuaga, A.I., Canet, D., Smeenk, G., Berends, R., Titgemeyer, F., Duurkens, R., Mateo, P.L., Scheek, R.M., Robillard, G.T., Dobson, C.M & van Nuland, N.A.J (2003) Characterization of single-tryptophan mutants of histidine-containing phosphocarrier protein: evidence for local rearrangements during folding from high concentrations of denaturant Biochemistry 42, 4883–4895 30 Canet, D., Lyon, C.E., Scheek, R.M., Robillard, G.T., Dobson, C.M., Hore, P.J & van Nuland, N.A.J (2003) Rapid formation of non-native contacts during the folding of HPr revealed by realtime photo-CIDNP NMR and stopped-flow fluorescence experiments J Mol Biol 330, 397–407 31 Gunasekaran, K., Eyles, S.J., Haggler, A.T & Gierasch, L.M (2001) Keeping it in the family: folding studies of related proteins Cur Opin Struct Biol 11, 83–93 ´ ´ 32 Fernandez-Ballester, G., Maya, J., Martı´ n, A., Parche, S., Gomez, J., Titgemeyer, F & Neira, J.L (2003) The histidine-phosphocarrier protein of Streptomyces coelicolor folds by a partially folded species at low pH Eur J Biochem 270, 2254–2267 33 Pace, C.N (1986) Determination and analysis of urea and guanidine hydrochloride denaturation curves Methods Enzymol 131, 266–280 34 Pace, C.N & Scholtz, J.M (1997) Measuring the conformational stability of a protein In Protein Structure (Creighton, T.E., ed.), 2nd edn, pp 253–259 Oxford University Press, Oxford 35 Privalov, P.L (1979) Stability of proteins: small globular proteins Adv Protein Chem 33, 167–241 36 Chen, B.L & Schellman, J.A (1989) Low-temperature unfolding of a mutant of phage T4 lysozyme Equilibrium studies Biochemistry 28, 685–691 37 Becktel, W.J & Schellman, J.A (1987) Protein stability curves Biopolymers 26, 1859–1877 38 Vuilleumier, S., Sancho, J., Lowenthal, R & Fersht A.R (1993) Circular dichroism studies of barnase and its mutants: characterization of the contribution of aromatic side chains Biochemistry 32, 10303–10313 39 Pace, C.N & Laurents, D.V (1989) A new method for determining the heat capacity change for protein folding Biochemistry 28, 2520–2525 40 Makhatadze, G.I & Privalov, P.L (1992) Protein interactions with urea and Gdm chloride A calorimetric study J Mol Biol 226, 491–505 41 Cornish-Bowden, A (2002) Enthalpy-entropy compensation: a phantom phenomenon J Biosci 27, 121–126 42 Sharp, K (2001) Enthalpy-entropy compensation: Fact or artefact Prot Sci 10, 661–667 43 Huang, G.S & Oas, T.G (1996) Heat and cold denatured states of monomeric lambda repressor are thermodynamically and conformationally equivalent Biochemistry 35, 6173–6180 44 DeKoster, G.T & Robertson, A.D (1995) Cold denaturation of CheY J Mol Biol 249, 529–534 45 Chi, Y.-H., Kumar, T.K., Wang, H.-M., Ho, M.-C., Chiu, I.-M & C (2001) Thermodynamic characterization of the human acidic fibroblast growth factor: evidence for cold denaturation Biochemistry 40, 7746–7753 46 Rees, D.C & Robertson, A.D (2001) Some thermodynamic implications for the thermostability of proteins Protein Sci 10, 1187–1194 47 Kumar, S., Tsai, C.J & Nussinov, R (2002) Maximal stabilities of reversible two-state proteins Biochemistry 41, 5359–5374 48 Robertson, A.D & Murphy, K.P (1997) Protein Structure and the Energetics of Protein Stability Chem Rev 97, 1251–1267 49 Privalov, P.L (1990) Cold denaturation of proteins Crit Rev Biochem Mol Biol 25, 281–305 Ó FEBS 2004 Conformational stability of HPr of S coelicolor (Eur J Biochem 271) 2181 50 Tsai, C.J., Maizel, J.V & Nussinov, R (2002) The hydrophobic effect A new insight from cold-denaturation and two-state water structure Crit Rev Biochem Mol Biol 37, 55–69 51 Ruelle, P & Kesselring, U.W (1998) The hydrophobic effect A consequence of the mobile order in H-bonded liquids J Pharm Sci 87, 987–997 52 Johnson, C.M & Fersht, A.R (1995) Protein stability as a function of denaturant concentration: the thermal stability of barnase in the presence of urea Biochemistry 34, 6795–6804 53 Privalov, P.L & Makhatadze, G.I (1990) Heat capacity of proteins II Partial molar heat capacity of the unfolded polypeptide chain of proteins: protein unfolding effects J Mol Biol 213, 385–391 54 Privalov, P.L., Tiktopulo, E.I., Venyaminov, S.Y., Griko, Y.V., Makhatadze, G.I & Khechninashvili, N.N (1989) Heat capacity and conformation of proteins in the denatured state J Mol Biol 205, 737–750 55 Zou, Q., Habermann-Rottinghaus, S.M & Murphy, K.P (1998) Urea effects on protein stability: hydrogen bonding and the hydrophobic effect Proteins 31, 107–115 56 Kresheck, G.C & Benjamin, L (1964) Calorimetric studies of the hydrophobic nature of several protein constituents and ovoalbumin in water an aqueous urea J Phys Chem 68, 2476– 2486 57 de Visser, C., Perron, G & Desnoyers, J.E (1977) Volumes and heat capacities of ternary aqueous systems at 25 °C Mixtures of urea, tert-butyl alcohol, dimethylformamide and water J Am Chem Soc 98, 5894–5900 58 Enea, O & Jollcoeur, C (1982) Heat capacity and Volumes of several oligopeptides in urea-water mixtures at 25 °C Some implications for protein unfolding J Phys Chem 86, 3870– 3881 59 Hakin, A.W & Hedgiw, G.R (2001) Group additivity calculations of the thermodynamic properties of unfolded proteins in aqueous solution: a critical comparison of peptide-based and HKF models Biophys Chem 89, 253–264 60 Zweifel, M.E & Barrick, D (2002) Relationships between the temperature dependence of solvent denaturation and the denaturant dependence of protein stability curves Biophys Chem 101–102, 221–237 61 Schellman, J.A (1987) Selective binding and solvent denaturation Biopolymers 26, 549–559 62 Aune, K.C & Tandford, C (1969) Thermodynamics of the denaturation of lysozyme by guanidine hydrochloride I Depdendence on pH at 25 degrees Biochemistry 8, 4586–4590 63 Filimonov, V.V., Azuaga, A.I., Viguera, A.R., Serrano, L & Mateo, P.L (1999) A thermodynamic analysis of a family of small globular proteins: SH3 domains Biophys Chem 77, 195–208 64 De Koster, G.T & Robertson, A.D (1997) Calorimetricallyderived parameters for protein interactions with urea and guanidine-HCl are not consistent with denaturant m-values Biophys Chem 64, 59–68 65 Hollien, S & Marqusee, S (1999) A thermodynamic comparison of mesophilic and thermophilic ribonucleases H Biochemistry 38, 3831–3836 66 Pike, A.C.W & Acharya, K.R (1994) A structural basis for the interaction of urea with lysozyme Protein Sci 3, 706–710 67 Zolkiewski, M., Nosworthy, N.J & Ginsburg, A (1995) Ureainduced dissociation and unfolding of dodecameric glutamine synthetase from Escherichia coli: calorimetric and spectral studies Protein Sci 3, 706–710 68 Tirado-Rives, J., Orozco, M & Jorgensen, W.L (1997) Molecular dynamics simulations of the unfolding of barnase in water and M aqueous urea Biochemistry 36, 7313–7329  ... (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... (see above) Discussion The conformational stability and the cold- denaturation of scHPr The conformational stability of a protein, DG, is fully specified when the enthalpy and the entropy changes... 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