Hydration and Packing Effects on Prion Folding and -Sheet Conversion

0 downloads 0 Views 237KB Size Report
Apr 19, 2004 - Nandi, P. K., Leclerc, E., Nicole, J. C., and Takahashi, M. (2002) J. Mol. Biol. 322, 153–161. 48. Cordeiro, Y., Lima, L. M., Gomes, M. P., Foguel, ...
THE JOURNAL OF BIOLOGICAL CHEMISTRY © 2004 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 279, No. 31, Issue of July 30, pp. 32354 –32359, 2004 Printed in U.S.A.

Hydration and Packing Effects on Prion Folding and ␤-Sheet Conversion HIGH PRESSURE SPECTROSCOPY AND PRESSURE PERTURBATION CALORIMETRY STUDIES* Received for publication, April 19, 2004, and in revised form, May 25, 2004 Published, JBC Papers in Press, June 1, 2004, DOI 10.1074/jbc.M404295200

Yraima Cordeiro‡§, Julia Kraineva§¶, Revanur Ravindra¶, Luı´s Maurı´cio T. R. Lima储, Mariana P. B. Gomes‡, Debora Foguel‡, Roland Winter¶, and Jerson L. Silva‡** From the ‡Departamento de Bioquı´mica Me´dica, Centro Nacional de Ressona៮ ncia Magne´tica Nuclear de Macromole´culas, Instituto de Cieˆncias Biome´dicas, Universidade Federal do Rio de Janeiro, Rio de Janeiro 21941-590, Brazil, the ¶Department of Chemistry, Physical Chemistry I, University of Dortmund, Otto-Hahn Strasse 6, D-44227 Dortmund, Germany, and the 储Departamento de Medicamentos, Faculdade de Farma´cia, Universidade Federal do Rio de Janeiro, Rio de Janeiro, 21941-590, Brazil

The main hypothesis for prion diseases proposes that the cellular protein (PrPC) can be altered into a misfolded, ␤-sheet-rich isoform (PrPSc), which undergoes aggregation and triggers the onset of transmissible spongiform encephalopathies. Here, we compare the stability against pressure and the thermomechanical properties of the ␣-helical and ␤-sheet conformations of recombinant murine prion protein, designated as ␣-rPrP and ␤-rPrP, respectively. High temperature induces aggregates and a large gain in intermolecular antiparallel ␤-sheet (␤-rPrP), a conformation that shares structural similarity with PrPSc. ␣-rPrP is highly stable, and only pressures above 5 kilobars (1 kilobar ⴝ 100 MegaPascals) cause reversible denaturation, a process that leads to a random and turnrich conformation with concomitant loss of ␣-helix, as measured by Fourier transform infrared spectroscopy. In contrast, aggregates of ␤-rPrP are very sensitive to pressure, undergoing transition into a dissociated species that differs from the denatured form derived from ␣-rPrP. The higher susceptibility to pressure of ␤-rPrP can be explained by its less hydrated structure. Pressure perturbation calorimetry supports the view that the accessible surface area of ␣-rPrP is much higher than that of ␤-rPrP, which explains the lower degree of hydration of ␤-rPrP. Our findings shed new light on the mechanism of prion conversion and show how water plays a prominent role. Our results allow us to propose a volume and free energy diagram of the different species involved in the conversion and aggregation. The existence of different folded conformations as well as different denatured states of PrP may explain the elusive character of its conversion into a pathogenic form.

* This work was supported by grants from Conselho Nacional de Desenvolvimento Cientifíco e Tecnolo´gico (CNPq); Financiadora de Estudos e Projetos (FINEP); Fundaça˜o de Amparo a` Pesquisa do Estado do Rio de Janeiro (FAPERJ); and Coordenaça˜o de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) of Brazil (to J. L. S. and D. F.), by an international grant from the International Centre for Genetic Engineering and Biotechnology (to J. L. S.), by a grant from Fundaça˜o Universita´ria Jose´ Bonifa´cio (FUJB) (to L. M. T. R. L.), and by a grant from the Deutsche Forschungsgemeinschaft (to R. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. § These authors contributed equally to this work. ** To whom correspondence should be addressed: Dept. de Bioquı´mica Me´dica, Instituto de Cieˆncias Biome´dicas, Universidade Federal do Rio de Janeiro, Av. Bauhı´nia 400, Bloco E S10, Rio de Janeiro, 21941-590, Brazil. Tel.: 55-21-25626756; Fax: 55-21-25626756; E-mail: [email protected].

The prion protein (PrP)1 is known to be the major agent that causes transmissible spongiform encephalopathies (1). The conversion from an ␣-helical (cellular PrP or denoted PrPC) to a ␤-sheet-rich structure (Prion Scrapie, denoted PrPSc) determines the onset of these particular diseases (1– 4). The PrPC is anchored in the cell membrane by a glycosylphosphatidylinositol bridge and is rich in ␣-helical structure and highly soluble (5). In contrast, PrPSc is mostly insoluble and has a greater ␤-sheet content than PrPC (5, 6). However, in most cases, both prion isoforms are derived from PrP molecules displaying the same primary sequence, although the PrPSc may exist in variable truncated forms in vivo, because of its partial resistance to proteolysis (7). The mechanism of PrPC to PrPSc conversion is not yet completely understood. However, recent studies show that other macromolecules, such as nucleic acids, may participate in the conversion (8 –10). Since the discovery that PrP is the main agent responsible for transmissible spongiform encephalopathies (1), several studies have been published on the thermodynamic and structural properties of this unusual protein (11–15). Most of these studies have used denaturing agents or an increase in temperature to assess the dissociation/denaturation of the prion protein (16, 17). Another tool for investigating the unfolding of PrP is high pressure, which has an advantage over other methods because its perturbation of macromolecules in solution depends solely on the volume change of the process under study (18). High pressure favors the formation of structures with lower volume, and application of pressure generally hydrates the hydrophobic interior of proteins (19, 20). Moreover, proteins with a large volume fraction of solvent-excluded cavities are highly sensitive to pressure (18, 21–23). Thus, high pressure is a unique tool for exploring hydration and packing of proteins. More recently, misfolded proteins that form aggregates and amyloids, which are derived from partially folding intermediates at the junction between productive and off-pathway folding, have been studied as well (17, 23–25). Other groups have utilized high pressure to study prion folding and inactivation but always in combination with other denaturing conditions (17, 26 –28). Here, we describe the effects of high pressure on the full-length recombinant murine prion protein (rPrP23–231) and on the 1 The abbreviations used are: PrP, prion protein; rPrP, murine recombinant PrP; FT-IR, Fourier transform infrared; rPrP23–231, full-length recombinant murine PrP; PPC, pressure perturbation calorimetry; kb, kilobar.

32354

This paper is available on line at http://www.jbc.org

Different Hydrated Species of the Prion Protein ␤-sheet-rich rPrP aggregates (␤-rPrP), which are amyloid-like structures obtained by thermal treatment (29). We use Fourier transform infrared (FT-IR) spectroscopy (30 –33) and pressure perturbation calorimetry (32, 34) to probe the secondary structure and hydration of the protein, respectively. We find that, whereas ␣-helical rPrP undergoes aggregation at high temperature into a ␤-sheet-rich structure, it is markedly resistant to pressure, displaying almost no change in secondary structure up to 4 kilobars (kb). Denaturation of ␣-rPrP occurs at higher pressures and is totally reversible. Unlike native ␣-rPrP, the ␤-rPrP aggregates are highly susceptible to pressure and are readily dissociated at pressures below 4 kbar. In contrast to pressure, thermal denaturation of rPrP is an irreversible process. The overall hydration of ␣-rPrP and ␤-rPrP was evaluated by pressure perturbation calorimetry (PPC), which provides information on the hydration of the proteins during thermal denaturation through measurement of the heat induced by small periodic changes of gas pressure (32, 34). Overall, our results show denaturation of recombinant full-length prion protein by high pressure without the use of temperature or denaturants; newly formed aggregates are less hydrated and have more cavities than native prion protein and late aggregates. MATERIALS AND METHODS

Reagents and Protein Samples—All of the reagents used were of analytical grade. D2O for FT-IR spectroscopy was purchased from Aldrich. The rPrP23–231 cloned in pRSET plasmid (kindly given by Prof. R. Brentani and Prof. Vilma R. Martins) was expressed in Escherichia coli and purified by high affinity column refolding as described (35). For the FT-IR spectroscopy experiments, purified rPrP was dissolved in D2O and then lyophilized three times to remove all H2O from the sample. The dried protein was dissolved to a concentration of 4% (w/v) in 10 mM sodium phosphate buffer (pD 6.5) in 99.9% D2O for all thermal experiments. Pressure-insensitive Tris-DCl buffer (10 mM, pD 7.5) was used for high pressure experiments. rPrP aggregates were obtained by incubation of the native protein (4%) at 50 °C for 2 and 48 h and are denoted ␤-rPrP and ␤-rPrPlate, respectively. rPrP at 2 mg/ml in 10 mM sodium phosphate buffer at pH 6.5 was used for PPC experiments. High Pressure Fourier Transform Infrared Spectroscopy—For the temperature studies, protein solutions were loaded into a FT-IR cell with 4-mm-thick CaF2 windows separated by 50-␮m mylar spacers. The temperature in the cell was controlled by an external water circuit and was increased gradually between 20 and 80 °C at 10 °C/h. A diamond anvil cell (High Pressure Diamond Optics Inc.) was used for the measurements under pressure. The samples were mixed with powdered ␣-quartz in a stainless-steel compartment (0.5 mm in diameter, 0.05 mm deep) (31, 33), and changes in pressure were quantified by the shift of the quartz phonon band at ⬃798 cm⫺1 (36). The pressure studies were carried out between 1 bar and ⬃10 kbar. An equilibration time of ⬃15 min was given before taking data at each temperature and pressure. FT-IR spectra were collected on a Nicolet Magna 550 FT-IR spectrometer equipped with a mercury cadmium telluride detector operated at ⫺196 °C. Spectra were generated by co-adding 256 interferograms collected at 2 cm⫺1 resolution and apodized with a Happ-Genzel function. Base-line data were obtained using buffer in D2O. The sample chamber was purged with dry, CO2-free air. Determination of peak position and curve fitting were performed with OMNIC (Nicolet, Madison, WI) and GRAMS (Galactic, Salem, NH) software, respectively. Integral intensities of the secondary structure elements were calculated by analysis of the deconvoluted amide I⬘ (the prime indicates that the solvent is D2O) vibrational mode of the IR spectrum with a band fitting procedure assuming a Gaussian-Lorentzian line shape function (30, 33). Fourier self-deconvolution of the IR spectra was performed with a resolution enhancement factor of 1.8 and a bandwidth of 15 cm⫺1. We note that the results of this method need to be treated with caution for the determination of absolute values of the secondary structure elements, because their transition dipole moments may be different and because theoretical predictions of the absorbance frequencies of model polypeptide secondary structures may be influenced by structural distortions, variable hydrogen/deuterium exchange, etc. No problems arise, however, from the application of the fitting method to the study of relative changes in conformations of the protein backbone, which was the primary goal of this study. The goodness of the fit reached using the

32355

TABLE I Approximate wave numbers of ␣-rPrP and ␤-rPrP secondary structure elements in the amide I⬘ region (1700 –1600 cm⫺1) of the infrared spectra (in parentheses are the wavenumbers for mature ␤-rPrPlate) Secondary structure

Amide I⬘ wavenumber (cm⫺1)

Antiparallel ␤-sheets ␤-Sheets Turns ␣-Helices Random coil

⬃1680 (1689) and ⬃1613 (1616) ⬃1628 (1635) ⬃1668 (1673) ⬃1651 (1659) ⬃1642 (1649)

Levenberg-Marquardt method was very satisfactory (R2 values were 98.7–99.6%; the noise level was ⬃0.1%) when peak fitting of the amide I⬘ band was done with six mixed Gaussian and Lorentzian peak functions. The error in determination of the secondary structure elements from the relative peak areas of the amide I⬘ band (integral intensities) from different runs is approximately ⫾2%. For analysis of pressure-induced transitions of ␣- and ␤-rPrP, two conformational states were assumed, native (N) and unfolded (U). The standard Gibbs free energy change (⌬G0 ⫽ ⫺RTlnK) and the equilibrium constant for the reaction (K) depend on the standard volume change of the reaction according to the relation K(p) ⫽ K0exp(⫺p⌬V/ RT), where K(p) and K0 are the equilibrium constants of unfolding at pressure p and atmospheric pressure, respectively, and ⌬V is the volume change of the reaction (18). Pressure Perturbation Calorimetry—PPC measurements were performed with a Valerian Plotnikov differential scanning calorimeter (MicroCal Inc., Northampton, MA) equipped with PPC accessory from MicroCal. The PPC technique is described in detail elsewhere (32). In short, one can obtain valuable thermodynamic information on protein hydration, expansivity, and relative volume changes upon thermal denaturation (34) by measuring the heat change ⌬Q, which is released upon small pressure changes, ⌬p, at temperature T. Through several reference measurements and knowing the thermal expansion coefficient of the solvent (␣s) as well as the mass (m) and partial specific volume of the solute (V), it is possible to calculate the apparent thermal expansion coefficient (␣) of the protein in solution (34), ␣ ⫽ ␣s ⫺ ⌬Q/(T ⌬p m V). The relative volume change of the protein during unfolding or aggregation can also be calculated by measuring ␣(T) in a system undergoing a heat-induced transition and integrating the ␣ versus. T plot over temperature (32, 34). At every temperature step, an identical small pressure jump of ⫹5 bar was applied to both sample and reference cells using pressurized N2 gas. For the analysis, signals from the bufferbuffer, buffer-water, and water-water runs are subtracted from the sample buffer PPC run. The partial specific volume of mouse rPrP used for the volumetric calculations was 0.73 cm3 g⫺1. RESULTS

The stability of rPrP was investigated by FT-IR spectroscopy using temperature and pressure as physical perturbations. The most useful IR band for the analysis of the secondary structure of proteins in aqueous media is the amide I band (which downshifts by ⬃5 cm⫺1 when in D2O as solvent and is then labeled amide I⬘ band), which occurs between ⬃1700 and ⬃1600 cm⫺1 (31). The amide I band represents 76% of the C⫽O stretching vibration of the amide group, coupled to the C–N stretching (14%) and C–C–N deformation (10%) mode. The exact frequency of this vibration depends on the nature of the hydrogen bonding involving the amide group, and this is determined by the particular secondary structure adopted by the protein. Using the analysis described under “Material and Methods,” the component bands representing ␣-helices, ␤-sheets, turns, and random structures can be determined. The percentages of these structures are estimated by expressing the areas of the component bands as a fraction of the total amide I area. (Table I). Because of the unknown transition dipole moments of the various secondary structure elements, no absolute values for the population of conformational states can be given. Curve fitting of the amide I⬘ spectra of rPrP revealed six bands, appearing at ⬃1668 (turns), 1651 (␣-helices), 1642 (random coil), 1628 (␤sheets), 1680, and 1613 (antiparallel ␤-sheets) cm⫺1 (Table I).

32356

Different Hydrated Species of the Prion Protein

FIG. 1. Temperature-induced denaturation/aggregation of rPrP. A, FT-IR spectra of ␣-rPrP at pD 6.5 upon gradual heating from 20 to 80 °C at 10 °C/h (only selected temperatures are shown). B, relative intensity of ␣-rPrP secondary structure components as a function of temperature. The open symbols refer to the respective secondary structure component after return to 25 °C. The error in determination of the secondary structure elements from the relative peak areas of the amide I⬘ band (integral intensities) from different runs is approximately ⫾2%, smaller than the size of the symbols.

Effect of Temperature on the Structure and Stability of rPrP—To investigate the effect of temperature on the secondary structure of rPrP, we measured the heat-induced changes in the amide I⬘ region of the infrared spectrum in the temperature range from 20 to 80 °C. Selected FT-IR spectra of the prion protein are shown in Fig. 1A. The amide I⬘ band maximum in the native protein occurs at ⬃1648 cm⫺1, which is typical for random coil-rich proteins (30). The temperatureinduced denaturation is indicated by the appearance of new IR bands at 1613 and 1680 cm⫺1 and occurs at 43 ⫾ 3 °C. These new IR bands are characteristic of the intermolecular antiparallel ␤-sheet resulting from aggregation of the denatured protein (37). The evolution of changes in secondary structures with temperature is shown in Fig. 1B. Based on fitting the amide I⬘ band to obtain the subcomponents (Table I), the structure of the native protein at 25 °C contains 11% turns, 28% ␣-helices, 46% random coil, and 14% antiparallel ␤-sheet with ⫾ 2% deviations (Fig. 1B). The NMR data for mouse rPrP23–231 yield were 10.0% turns, 25.8% ␣-helices, 62.2% random structures, and 1.9% ␤-strands, as calculated from mouse rPrP121–231 (1AG2.pdb) after adding the N-terminal segment 23–120 and assuming this domain to be completely random (38, 39). rPrP loses ␣-helical structure and acquires a higher antiparallel ␤-sheet content during the temperature-induced denaturation process (Fig. 1). We use the term denaturation in a broad sense, because the loss of ␣-helices and the gain of ␤-sheet cannot be considered a typical unfolding transition. The relative content of ␣-helices decreases from 28% at 25 °C to 16% at 70 °C. Concomitantly, the protein also loses random coil structure

FIG. 2. Pressure-induced denaturation of ␣-rPrP. A, FT-IR spectra of ␣-rPrP at pD 7.5 as a function of pressure at 25 °C (only selected pressures are shown). B, relative intensity of ␣-rPrP secondary structure components as a function of pressure. All open symbols correspond to the respective secondary structure component after return to atmospheric pressure. The error in determination of the secondary structure elements from the relative peak areas of the amide I⬘ band (integral intensities) from different runs is approximately ⫾2%, smaller than the size of the symbols.

(from 46% at 25 °C to 18% at 70 °C) and gains intermolecular ␤-sheet structure (up to 43%; Fig. 1B). Upon cooling from 80 °C to room temperature, the aggregation bands remain unchanged, indicating that the rPrP23–231 temperature-induced transition cannot be reversed by a simple temperature decrease (Fig. 1B, open symbols). The thermal transition was also observed by CD spectroscopy, which revealed a process highly dependent on protein concentration.2 We also found that aggregation, as measured by light scattering, paralleled the gain in ␤-sheet structure (data not shown). Effect of Pressure on the Structure and Stability of Native ␣-rPrP and Aggregated ␤-rPrP—To reach a better understanding of the PrP folding pathway, we also analyzed the stability of native rPrP against pressure, and we describe the differences between the pressure- and temperature-denatured states. To investigate the effect of pressure on the secondary structure of rPrP, we measured the pressure-induced changes in the amide I⬘ vibration mode in the range from 0.001 to 10 kbar at two selected temperatures (25 and 50 °C). At 25 °C, the prion protein (␣-rPrP) is native, predominantly ␣-helical, and random. Fig. 2A shows spectra of ␣-rPrP as a function of pressure at 25 °C. Pressure-induced unfolding occurs above 4.5 kbar with a p1⁄2 of 5.4 ⫾ 0.2 kbar and is indicated by changes in the amide I⬘ band region; the band becomes broader, and the intensity of the main peak decreases and shifts to lower wavenumbers, indicating a decrease in ␣-helical structures. The behavior of secondary structural changes upon isothermal sample pressurization is presented in Fig. 2B. During unfolding, which is 2

Y. Cordeiro and J. L. Silva, unpublished results.

Different Hydrated Species of the Prion Protein

32357

FIG. 4. ␣-rPrP and ␤-rPrP display different stabilities against pressure. The f-values as a function of pressure were obtained from native ␣-rPrP ␣-helical secondary structure (see Fig. 2) and from ␤-sheet changes for ␤-rPrP (see Fig. 3). The extent of denaturation (ƒ) was calculated as follows: ƒ ⫽ (IRobs ⫺ IRinicial/IRfinal ⫺ IRinicial), where IRobs is the observed IR value at pressure p, IRinicial is the IR value at 1 bar, and IRfinal is the final IR value of the pressure-induced unfolding curve.

FIG. 3. Pressure treatment disrupts rPrP aggregates (␤-rPrP). A, relative intensity of ␤-rPrP secondary structure components as a function of pressure. Inset, FT-IR spectra of ␤-rPrP at 50 °C (pD 7.5) at 1 bar (solid line) and at 3.8 kbar (dashed line). B, relative intensity changes of secondary structure elements of ␤-rPrPlate (at pD 7.5, after incubation for 48 h at 50 °C) as a function of pressure at 50 °C. For A and B, the open symbols refer to the respective secondary structure component after return to atmospheric pressure. The error in determination of the secondary structure elements from the relative peak areas of the amide I⬘ band (integral intensities) from different runs is approximately ⫾2%, smaller than the size of the symbols.

reversible (Fig. 2, open symbols), ␣-helical structure decreases significantly, from 26 to 12%, and the content of turns increases from 12 to 19%. No marked changes are observed in random and ␤-sheet conformations within 12 h. In addition, we compared the stability of native ␣-rPrP with the aggregated ␤-sheet form obtained by thermal treatment. rPrP was incubated above the transition temperature (50 °C) for 2 h (referred to as ␤-rPrP) and 48 h (␤-rPrPlate). The IR bands of the ␤-rPrPlate aggregate (at ⬃1680 and ⬃1613 cm⫺1) exhibit both a blue shift of ⬃7 cm⫺1 compared with the ␤-rPrP aggregate (Table I), which might be due to the formation of an aggregated structure with weaker/fewer hydrogen bonding of the C⫽O groups. We subjected both types of aggregates (␤-rPrP and ␤-rPrPlate) to pressure at 50 and 25 °C, respectively. The effect of high pressure on aggregated ␤-rPrP at a temperature above the thermal unfolding temperature was investigated up to 10 kbar. The inset in Fig. 3A shows ␤-rPrP spectra at 50 °C at 1 and 3.8 kbar; the amide I⬘ band has a shape typical of aggregated rPrP (Fig. 1A) with a broad maximum at 1641 cm⫺1 and two shoulders at 1613 and 1680 cm⫺1, a pattern that is characteristic of antiparallel ␤-sheets. The effect of pressure on the ␤-rPrP aggregates at 50 °C is dramatic. The aggregate dissociates at pressures between 2.5 and 4 kbar, with a p1⁄2 of 2.8 ⫾ 0.1 kbar. The relative peak area of ␤-sheet structures decreases from ⬃43 to ⬃29%. However, the intermolecular

␤-sheet structures are not completely disrupted by pressure (Fig. 3A), resulting in a denatured species that is different from that obtained by pressure denaturation of ␣-rPrP. The content of random structures increases markedly (up to ⬃41%) at the transition, whereas the population of ␣-helices decreases to ⬃10% at 10 kbar. After pressure release and return to 1 bar, the protein acquires the same amount of intermolecular antiparallel ␤-sheet as it had before pressure treatment. Hence, unlike the thermal denaturation of the prion protein, the pressure-induced transition is reversible. Because the final data points (at 10 –12 kbar) are different for the ␣-rPrP and ␤-rPrP structures, they clearly belong to different denatured states. The differences in pressure susceptibility between ␣-rPrP and ␤-rPrP are clearly visualized in Fig. 4. ␣-rPrP undergoes pressure denaturation at much higher pressures and with a smaller volume change than ␤-rPrP. Overall, the data indicate that formation of ␤-rPrP induces the appearance of solventexcluded cavities, highly sensitive to pressure. We also studied the stability of ␤-rPrPlate against high pressure up to 12 kbar at two different temperatures, 25 and 50 °C (see Fig. 3B for the 50 °C data). Within experimental error, we could not detect any significant changes in the FT-IR spectra or in the population of secondary structure elements at either temperature over the whole pressure range covered, which indicates a marked pressure stability of these mature aggregates. Pressure Perturbation Calorimetry—We performed PPC measurements to further explore the changes in thermal expansion and volume of native rPrP and of the ␤-rPrP aggregate. The apparent thermal expansion coefficient ␣ is depicted as a function of temperature between 10 and 95 °C for both samples (Fig. 5). These data clearly show that ␣ of rPrP decreases markedly from 1.75 ⫻ 10⫺3 K⫺1 to a value near 0.3 ⫻ 10⫺3 K⫺1. Above ⬃80 °C, ␣ decreases only slightly. High ␣-values at low temperature and their decrease with increasing temperature are indicative of the presence of a significant number of structure-breaking hydrophilic groups on the surface of rPrP (34, 40). In other words, the solvent-accessible surface area is very high for ␣-rPrP. No identifiable peak occurs in the ␣(T) curve, indicating that the temperature-induced aggregation process does not involve detectable volume changes during the PPC run, very similar to the final insulin aggregation step (32). The changes in ␣(T) for ␤-rPrP (Fig. 5B) are much smaller than for ␣-rPrP (Fig. 5A), which is obviously because of the decrease in hydrated accessible surface area of the aggregated chains.

32358

Different Hydrated Species of the Prion Protein

FIG. 5. Pressure perturbation calorimetry curves of ␣-rPrP and ␤-rPrP. Shown is a PPC scan of 2 mg/ml ␣-rPrP (A, filled squares) and ␤-rPrP (B, open circles) at pH 6.5 as a function of temperature. A stepwise temperature increment was employed to maintain a scan rate of ⬃10 °C/h. DISCUSSION

Here, we present clear-cut evidence that the conversion of ␣-rPrP into ␤-rPrP involves a substantial change in hydration and accessible surface area as determined by high pressure FT-IR spectroscopy and pressure perturbation calorimetry. Fourier transform infrared spectroscopy has been widely used to probe transitions in secondary conformation of proteins (6, 17, 31, 41), and the use of a diamond anvil cell provides access to secondary structure components of proteins also upon pressurization (31, 33). Thermally induced unfolding of ␣-rPrP gave rise to an increased ␤-sheet content at temperatures above ⬃41 °C. The simultaneous increase in two IR bands at low and high wavenumbers (1613 and 1680 cm⫺1) upon temperature-induced denaturation is due to intermolecular antiparallel ␤-sheet formation, as a result of protein aggregation (37). Formation of intermolecular ␤-sheets during PrP aggregation has also been inferred from attenuated total reflection FT-IR spectroscopy analysis of prion rods after limited proteolysis of hamster PrPSc (41) and from electron microscopy of hamster prion peptide fibrils (42). In agreement with experimental results and computational approaches reported by other authors, our data show that the increase in ␤-sheet content occurs at the expense of both ␣-helical structure (5, 6, 43– 45) and random coil content, principally the former. There are some discrepancies between our results and previous data for secondary structure components of PrP aggregates measured by different IR techniques and by deconvolution of CD spectra (6, 41, 46). However, it is important to address that some previous analysis were done with brainderived highly infectious PrPSc, which is found oligomerized/ aggregated in vivo and is never completely pure. These PrPSc preparations are not always homogeneous and may contain different polypeptide composition, N-linked glycans, metal ions, and glycosylphosphatidylinositol anchors among other molecules that, until now, were impossible to dissociate from the PrP molecules without denaturation. Furthermore, PrPSc extracted from infected brain, without further in vitro proteolysis, is also a mixture of fully protease-sensitive and partially protease-resistant molecules that have different conformations. Also, the biochemical conditions used in other works to prepare the aggregates differ from our procedure. All of these authors have investigated hamster PrP aggregation using different sample pH, purification procedures, and lengths of PrP, and also the glycosylation level was variable (6, 41, 46). Furthermore, the aggregates obtained by raising the temperature, as in this study, are not necessarily similar to aggregates

FIG. 6. Volume (A) and Gibbs free energy (B) diagrams of ␣-rPrP and ␤-rPrP. ␣-rPrP (circle) and ␤-rPrP (triangle) pressure-denatured states are denoted as U and U⬘, respectively. The numbers indicate volume (ml/mol) (A) and Gibbs free energy of unfolding (kcal/mol) (B) determined values for ␣-rPrP and ␤-rPrP pressure-induced unfolding.

obtained by procedures such as incubation at lower pH, the use of detergents, or incorporation into liposomes. Nonetheless, we also cannot discharge the possibility that the PrP aggregates obtained by us differ significantly from in vivo infectious PrPSc, especially in the light of the impossibility of producing infectious PrPSc from recombinant PrPC in vitro. Native rPrP at 21 °C was highly resistant to pressure up to 4 kbar (Fig. 2), and at higher pressures, the amount of ␣-helix decreased, whereas the turns increased. rPrP tryptophan fluorescence revealed no significant changes in tertiary structure up to 3 kbar (data not shown). Pressure denaturation monitored by FT-IR spectroscopy was totally reversible, which allowed us to obtain the thermodynamic parameters of the ␣-rPrP and ␤-rPrP denaturation processes, as depicted in the volume and free energy diagrams of Fig. 6. ␣-rPrP is more hydrated and has a larger solvent-accessible surface area than aggregated ␤-rPrP. The smaller accessible surface area of ␤-rPrP is inferred from the pressure perturbation calorimetry experiments, and the presence of a greater volume of waterexcluded cavities is revealed by its much greater pressure sensitivity (⌬V ⫽ ⫺43.6 ⫾ 7.0 ml/mol; ⌬G0 ⫽ 2.81 ⫾ 0.10 kcal/mol). The occlusion of hydrophobic surfaces related to ␤-sheet formation can also be acquired by binding to nucleic acids (8, 47) or to 4,4⬘-dianilino-1,1⬘-binaphthyl-5,5⬘-sulfonate (bis-ANS) (48). Another feature revealed in both diagrams is that the denaturation transitions engender different denatured states (U and U⬘). They appear to arise from different folding routes: ␣-rPrP denatures into U with smaller changes in volume, whereas ␤-rPrP denatures into U⬘ with a larger volume change. There is a clear kinetic barrier, both in the volume (activation volume) and in the Gibbs free energy (activation energy) between U and U⬘. This unusual property is probably related to both the slow in vivo conversion and to the infectious nature of prion diseases. It may also explain the inability to show that any ␤-sheet-rich form obtained from recombinant PrP is an efficient infectious agent. Nevertheless, in vitro ␤-sheet isoforms have physical properties similar to PrPSc, and amyloid-like aggregates exhibit epitopes equivalent to those of Scrapie PrP (15). Indeed, it is quite intriguing why infectious PrPSc cannot be refolded in vitro, without a PrPSc template, as demonstrated by Kocisko et al. (49). The greater pressure stability of ␣-rPrP, as determined by its higher standard Gibbs free energy change of unfolding (⌬G0 ⫽ 5.37 ⫾ 0.15 kcal/mol) in contrast to that of ␤-rPrP (⌬G0 ⫽ 2.81 ⫾ 0.10 kcal/mol) seems paradoxical at first glance, because the chemical potential of these two forms shows the opposite (Fig. 6). For the free energy diagram, we assumed a metastability model for the conversion of ␣-rPrP into ␤-rPrP (14). The

Different Hydrated Species of the Prion Protein apparent contradiction is resolved by the finding that their N 43 U transitions are not connected at equilibrium and by the fact that ␣-rPrP is converted into ␤-rPrP by increasing the temperature, and no further unfolding is caused by temperatures as high as 95 °C. The volume and free energy diagrams revealed by pressure agree with recent thermodynamic and kinetic data showing that partially structured intermediates are essential in the folding pathway (50, 51). To investigate the time dependence of rPrP aggregation, we incubated rPrP at 50 °C for 2 h (forming ␤-rPrP) and 48 h (forming ␤-rPrPlate). The aggregation of mouse rPrP proved to be highly time-dependent, because there are marked differences in secondary structure between ␤-rPrP and ␤-rPrPlate. The population of turns and random structures was essentially the same, but the ␤-sheet content increased, and the ␣-helix content decreased in the ␤-rPrPlate sample. This observation suggests a gain of ␤-sheet structure at the expense of ␣-helix. Interestingly, although ␤-rPrP was highly pressure-sensitive (Fig. 3), ␤-rPrPlate displayed no secondary structural changes over the entire pressure range (1 bar to 12 kbar). According to the principle of Le Cha៮ telier, pressure shifts the equilibrium to conformational states that occupy smaller volumes (21, 22), leading normally to dissociation of oligomeric proteins (18) or to protein unfolding (21). Generally, oligomeric proteins (21, 22) and amyloid aggregates (23–24) dissociate in the pressure range from 1 to 3 kbar. However, because ␤-rPrPlate exhibited no changes in the IR absorption even up to 12 kbar, this mature aggregate no longer contains internal cavities susceptible to pressurization and hence can be considered to be rather densely packed. The presence of a significant number of structure-breaking hydrophilic groups on the surface of rPrP (34, 40) was revealed by positive ␣-values at low temperatures and a marked decrease in ␣ at higher temperatures. Proteins with highly charged polar surfaces exhibit larger ␣-values and steeper slopes of ␣(T) (40). No identifiable peak is observed in the ␣(T) curve of rPrP, indicating that the transition from ␣-rPrP to ␤-rPrP (see volume diagram in Fig. 6) does not lead to significant net changes in volume. By way of comparison, Dzwolak et al. (32) also observed no significant volume change upon aggregation of insulin, which takes place at ⬃85 °C. Compared with ␣-rPrP, there are much smaller changes in ␣(T) for ␤-rPrP (Fig. 5), which is obviously because of the absence of a large hydrated accessible surface area of the aggregated chains. In conclusion, we describe for the first time denaturation of recombinant prion protein by high pressure without the concomitant use of temperature or chemical denaturants, and we show that rPrP aggregates display different susceptibilities to high pressure, depending on the time of exposure to high temperature during aggregation. We show that different folded conformations as well as different denatured states of rPrP can be distinguished on the basis of hydration, surface exposure, and cavities. These dissimilarities result in the paradox that ␤-rPrP is highly resistant to temperature, whereas it is very sensitive to pressure; the opposite occurs with the native ␣-rPrP. Acknowledgments—We thank Emerson R. Gonc¸ alves for excellent technical support and M. Sorenson for careful reading of the manuscript. REFERENCES 1. 2. 3. 4.

Prusiner, S. B. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 13363–13383 Weissmann, C. (1999) J. Biol. Chem. 274, 3– 6 Caughey, B (2001) Trends Biochem. Sci. 26, 235–242 Aguzzi, A., and Polymenidou, M. (2004) Cell 116, 313–327

32359

5. Pan, K.-M., Baldwin, M., Nguyen, J., Gasset, M., Serban, A., Groth, D., Mehlhorn, I., Huang, Z., Fletterick, R. J., Cohen, F. E., and Prusiner, S. B. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 10962–10966 6. Caughey, B. W., Dong, A., Bhat, K. S., Ernst, D., Hayes, S. F., and Caughey, W. S. (1991) Biochemistry 30, 7672–7680 7. Caughey, B., Raymond, G. J., Ernst, D., and Race, R. E. (1991) J. Virol. 65, 6597– 6603 8. Cordeiro, Y., Machado, F., Juliano, L., Juliano, M. A., Brentani, R. R., Foguel, D., and Silva, J. L. (2001) J. Biol. Chem. 276, 49400 – 49409 9. Deleault, N. R., Lucassen, R. W., and Supattapone, S. (2003) Nature 425, 717–720 10. Caughey, B., and Kocisko, D. A. (2003) Nature 425, 673– 674 11. Swietnicki, W., Petersen, R., Gambetti, P., and Surewicz, W. K. (1997) J. Biol. Chem. 272, 27517–27520 12. Riek, R., Wider, G., Billeter, M., Hornemann, S., Glockshuber, R., and Wu¨ thrich, K. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 11667–11672 13. Cohen, F. E. (1999) J. Mol. Biol. 293, 313–320 14. Baskakov, I. V., Legname, G., Prusiner, S. B., and Cohen, F. E. (2001) J. Biol. Chem. 276, 19687–19690 15. Baskakov, I. V., Legname, G., Baldwin, M. A., Prusiner, S. B., and Cohen, F. E. (2002) J. Biol. Chem. 277, 21140 –21148 16. Safar, J., Roller, P. P., Gajdusek, D. C., and Gibbs, C. J., Jr. (1993) Protein Sci. 2, 2206 –2216 17. Torrent, J., Alvarez-Martinez, M. T., Heitz, F., Liautard, J.-P., Balny, C., and Lange, R. (2003) Biochemistry 42, 1318 –1325 18. Silva, J. L., Foguel, D., and Royer, C. (2001) Trends Biochem. Sci. 26, 612– 618 19. Oliveira, A. C., Gaspar, L. P., Da Poian, A. T., and Silva, J. L. (1994) J. Mol. Biol. 240, 184 –187 20. Royer, C. A. (2002) Biochim. Biophys. Acta. 25, 201–209 21. Silva, J. L., and Weber, G. (1993) Annu. Rev. Phys. Chem. 44, 89 –113 22. Mozhaev, V. V., Heremans, K., Frank, J., Masson, P., and Balny, C. (1996) Proteins 24, 81–91 23. Foguel, D., Suarez, M. C., Ferra˜ o-Gonzales, A. D., Porto, T. C., Palmieri, L., Einsiedler, C. M., Andrade, L. R., Lashuel, H. A., Lansbury, P. T., Kelly, J. W., and Silva, J. L. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 9831–9836 24. Ferra˜ o-Gonzales, A. D., Souto, S. O., Silva, J. L., and Foguel, D. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 6445– 6450 25. Ishimaru, D., Andrade, L. R., Teixeira, L. S., Quesado, P. A., Maiolino, L. M., Lopez, P. M., Cordeiro, Y., Costa, L. T., Heckl, W. M., Weissmuller, G., Foguel, D., and Silva, J. L. (2003) Biochemistry 42, 9022–9027 26. Kuwata, K., Li, H., Yamada, H., Legname, G., Prusiner, S. B., Akasaka, K., and James, T. L. (2002) Biochemistry 41, 12277–12283 27. Brown, P., Meyer, R., Cardone, F., and Pocchiari, M. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 6093– 6097 28. Martins, S. M., Chapeaurouge, A., and Ferreira, S. T. (2003) J. Biol. Chem. 278, 50449 –50455 29. Rezaei, H., Choiset, Y., Eghiaian, F., Treguer, E., Mentre, P., Debey, P., Grosclaude, J., and Haertle, T. (2002) J. Mol. Biol. 322, 799 – 814 30. Byler, D. M., and Susi, H. (1986) Biopolymers 25, 469 – 487 31. Panick, G., Malessa, R., and Winter, R. (1999) Biochemistry 38, 6512– 6519 32. Dzwolak, W., Ravindra, R., Lendermann, J., and Winter, R. (2003) Biochemistry 42, 11347–11355 33. Herberhold, H., Marchal, S., Lange, R., Scheyhing, C. H., Vogel, R. F., and Winter, R. (2003) J. Mol. Biol. 330, 1153–1164 34. Lin, L. N., Brandts, J. F., Brandts, J. M., and Plotnikov, V. (2002) Anal. Biochem. 302, 144 –160 35. Zahn, R., von Schro¨ tter, C., and Wu¨ thrich, K. (1997) FEBS Lett. 417, 400 – 404 36. Wong, P. T. T., Moffatt, D. J., and Baudais, F. L. (1985) Appl. Spectrosc. 39, 733–735 37. Ismail, A. A., Mantsch, H. H., and Wong, P. T. T. (1992) Biochim. Biophys. Acta 1121, 183–188 38. Riek, R., Hornemann, S., Wider, G., Billeter, M., Glockshuber, R., and Wu¨ thrich, K. (1996) Nature 382, 180 39. Riek, R., Hornemann, S., Wider, G., Glockshuber, R., and Wu¨ thrich, K. (1997) FEBS Lett. 413, 282–288 40. Ravindra, R., and Winter, R. (2003) Chemphyschem. 4, 359 –365 41. Gasset, M., Baldwin, M. A., Fletterick, R. J., and Prusiner, S. B. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 1–5 42. Zhang, H., Kaneko, K., Nguyen, J. T., Livshits, T. L., Baldwin, M. A., Cohen, F E., James, T. L., and Prusiner, S. B. (1995) J. Mol. Biol. 250, 514 –526 43. Huang, Z., Prusiner, S. B., and Cohen, F. E. (1996) Fold. Des. 1, 13–19 44. Inouye, H., and Kirschner, D. A. (1998) J. Struct. Biol. 122, 247–255 45. Jamin, N., Coı¨c, Y.-M., Landon, C., Ovtracht, L., Baleux, F., Neumann, J.-M., and Sanson, A. (2002) FEBS Lett. 529, 256 –260 46. Safar, J., Roller, P. P., Gajdusek, D. C., and Gibbs, C. J., Jr. (1993) J. Biol. Chem. 268, 20276 –20284 47. Nandi, P. K., Leclerc, E., Nicole, J. C., and Takahashi, M. (2002) J. Mol. Biol. 322, 153–161 48. Cordeiro, Y., Lima, L. M., Gomes, M. P., Foguel, D., and Silva, J. L. (2004) J. Biol. Chem. 279, 5346 –5352 49. Kocisko, D. A., Come, J. H., Priola, S. A., Chesebro, B., Raymond, G. J., Lansbury, P. T., and Caughey, B. (1994) Nature 370, 471– 474 50. Apetri, A. C., and Surewicz, W. K. (2003) J. Biol. Chem. 278, 22187–22192 51. Apetri, A. C., Surewicz, K. A., and Surewicz, W. K. (2004) J. Biol. Chem. 279, 18008 –18014