ARTICLE pubs.acs.org/JPCB
Denaturation Behaviors of Two-State and Non-Two-State Proteins Examined by an InterruptionIncubation Protocol Jun-Jie Luo, Fu-Gen Wu, Ji-Sheng Yu, Rui Wang, and Zhi-Wu Yu* Key Laboratory of Bioorganic Phosphorous Chemistry and Chemical Biology (Ministry of Education), Department of Chemistry, Tsinghua University, Beijing 100084, P. R. China ABSTRACT: We established an interruptionincubation protocol to study the denaturation and aggregation behaviors of two-state and non-two-state proteins, represented by hen egg-white lysozyme (lysozyme) and bovine serum albumin (BSA), by differential scanning calorimetry (DSC) and Fourier-transform infrared (FTIR) spectroscopy. After incubation at selected interrupting temperatures (Tint), the onset temperature (Tonset) and denaturation temperature (Td) of the reheating scans were found to pose distinct contrasts between the two proteins. BSA shows increasing Tonset and Td as Tint increases, whereas lysozyme exhibits invariable Tonset and Td. After long-time incubation, the reheating scans of BSA still show residual peaks in the DSC curves. Moderate aggregation upon incubation restricts the refolding of the thermodynamically stable unfolding intermediates in the denaturation of BSA. On the contrary, prolonged incubation at selected Tint causes complete denaturation of lysozyme. The results were also supported by FTIR experiments. It is concluded that the variable Tonset and Td indicate the unfolding of the “trapped” intermediates in the second heating scans, and the invariable Tonset and Td values indicate the nonexistence of such intermediates. The examination of two additional proteins, ribonuclease A and γ-globulin, shows a similar phenomenon. The results may represent the general features of two-state and non-two-state denaturations and the protocol would serve to study protein denaturation cooperativity.
1. INTRODUCTION Proteins, serving as the principal structural and functional units in living organisms, have highly ordered structures and conformations folded from peptide chains. When subjected to stress factors such as heat or chemical treatment, proteins may denature. The unfolded peptide chains have the potential to refold when these factors are removed. The folding/unfolding behaviors of proteins have attracted significant interest, with efforts aimed at elucidating the folding/unfolding mechanisms,1 kinetics,2,3 and the folding energy barrier.46 For decades, differential scanning calorimetry (DSC), Fourier-transform infrared (FTIR) spectroscopy, Raman spectroscopy, circular dichroism (CD) spectroscopy, and nuclear magnetic resonance (NMR) spectroscopy have been widely employed to study protein thermal denaturation,712 thermal denaturation in the presence of sugars13,14 and surfactants,1517 and denaturant-induced protein denaturation.9,18,19 On the other hand, proteins tend to aggregate at high temperatures, characterized by the formation of intermolecular β-sheets. In most cases, aggregation competes with the folding/unfolding processes and disturbs or even disrupts the whole pathway.3,20,21 Therefore, aggregation is often considered as an obstacle to be avoided when protein denaturation is studied. Cooperativity, as an important phenomenon observed in biological systems, such as molecular recognition events22 and phase transitions of biomembranes,23,24 also exists in the protein denaturation realm. The cooperativity issue in protein denaturation is represented as two-state and non-two-state transitions. In general, “non-two-state” transition includes downhill and multistate r 2011 American Chemical Society
transitions. The former has only one thermodynamic state in denaturation process with properties that change gradually from folded to unfolded-like as protein stability decreases.25 It has been intensively studied and a thermodynamic approach has been established to discriminate it from the two-state transition.2527 The latter involves thermodynamically stable intermediates in the unfolding process,2830 which is often referred to as nontwo-state denaturation in a narrow sense. The “van’t Hoff analysis” was established to distinguish between two-state and multistate reversible transition.31,32 By using this approach, it has been found that some small proteins such as lysozyme and ribonuclease A (RNase A) adopt two-state denaturation processes.3236 Many proteins, however, undergo non-two-state processes. One typical example is human plasminogen.37 The denaturation of the intact plasminogen is a multistate reversible transition composed of seven two-state subtransitions. Another important issue is that if aggregation is involved in a protein unfolding process, the reversible unfolding transitions including two-state and multistate transitions will be converted to irreversible two-state and multistate transitions.3,38 Bovine serum albumin (BSA), composed of 582 amino acids, with molecular weight of 66.5 kDa,39,40 is considered to take an irreversible nontwo-state denaturation.41,42 Previous IR studies have also shown that BSA tends to form intermolecular β-sheet.43,44 Received: January 11, 2011 Revised: June 5, 2011 Published: June 14, 2011 8901
dx.doi.org/10.1021/jp200296v | J. Phys. Chem. B 2011, 115, 8901–8909
The Journal of Physical Chemistry B The denaturation behaviors of two-state and non-two-state proteins are quite different, and the absence/presence of thermodynamically stable intermediates is the key point to differentiate them. Therefore, capturing the unfolding intermediates of nontwo-state proteins is an important but challenging work in studies of protein denaturation. An extreme example is immunoglobulin G (IgG) which has two domains (Fab and Fc) which denature independently and exhibit two single denaturation peaks in DSC thermal curves. Two intermediates that correspond to the independent denaturation of either of the two domains have been obtained by different heat treatment and pH adjustment.45 However, to identify the unfolding intermediates of BSA by DSC is much more difficult because the denaturation peak can be deconvolved into two or three overlapped peaks mathematically.41,42 A new approach, the interruptionincubation experimental protocol, has been developed in this work to capture stable intermediates. Using the new protocol, the denaturation/aggregation behaviors of two model proteins, BSA and lysozyme, that take irreversible non-two-state and two-state denaturations, respectively, have been studied by DSC and FTIR. The result poses distinct contrast between each other. In the case of BSA, the refolding of the thermodynamically stable intermediates was blocked by partial aggregation, thereby allowing the detection of these “trapped” intermediates in the DSC second heating scan. The changing onset (Tonset) and denaturation (Td) temperatures of the DSC peaks in the second heating scans suggest the existence of some stable intermediates in the thermal denaturation of BSA. In the case of lysozyme, the invariable Tonset and Td suggest the absence of unfolding intermediates. Similar results have also been obtained by comparative studies on another pair of non-two-state and two-state model proteins, γ-globulin and RNase A. The thermodynamic approach proposed in this work reveals the different behaviors of irreversible two-state and non-two-state denaturation coupled with aggregation, and thus provides a new way to study/predict the protein denaturation cooperativity.
2. EXPERIMENTAL SECTION 2.1. Sample Preparation. BSA, RNase A, and γ-globulin were purchased from Sigma Chemical Co. (St. Louis, MO). Lysozyme was purchased from Amresco (Solon, OH). All of the proteins were supplied as lyophilized powders and were used without further purification. Double deionized water with a resistivity of 18.2 MΩ cm or D2O (99.9% of deuterium) purchased from Cambridge Isotope Laboratories, Inc. (Andover, MA) were used to prepare protein solutions. For DSC experiments, BSA, lysozyme, and γ-globulin were dissolved in 0.02 M sodium phosphate buffer (pH 7.4), and RNase A was dissolved in 0.02 M sodium acetic buffer (pH 4.8) to obtain the respective protein solutions with a concentration of 5 mg/mL. The ionic strength of all of the solutions was adjusted to 0.15 M using NaCl. In FTIR experiments, the solutions of lysozyme and BSA were prepared in the same manner except that the buffer was in D2O (pD 7.4), and the pD value was corrected according to pD = pH* + 0.4 for the deuterium isotope effect.46 The concentration of protein solutions used in FTIR experiments was 25 mg/mL. 2.2. DSC. All the calorimetric data were obtained by a CSC model 6300 Nano III differential scanning calorimeter (Calorimetry Sciences Corp., Lindon, UT) with a capillary cell volume of 0.3 mL. The heating and cooling scans were performed at a scan rate of 1 °C/min. The concentration of protein samples is
ARTICLE
5 mg/mL (except otherwise specified), which is the optimal concentration for nanoDSC measurements, and a constant pressure of 3 atm was applied to avoid bubble formation following the literature works.47,48 All data analyses were carried out on the NanoAnalyze software. The temperature intervals recorded by DSC are 0.02 °C in our measurement and the error of the analyzed Tonset and Td values is (0.2 °C. 2.3. FTIR. FTIR spectra in the range of 4000900 cm1 were recorded using a Nicolet 5700 Fourier-transform infrared spectrometer equipped with a DTGS detector. A spectral resolution of 2 cm1 and a zero filling factor of 2 were used in the experiments, and the precision of the wavenumber was better than 0.1 cm1. The protein solution was placed between a pair of CaF2 windows separated by a 25 μm Teflon spacer, which were mounted on a Linkam heatingcooling stage (Linkam Scientific Instruments, U.K.) for temperature control ((0.1 °C). The FTIR heating or cooling scans were performed at a scan rate of 1 °C/min, which is the same as in the DSC scans. For each spectrum, 32 scans were performed. All the spectra were analyzed after deducting the contributions from residual water vapor.49 The absorbance of the buffer was then eliminated by subtracting the spectrum of the buffer from the spectra of the samples under the same experimental conditions. Band-narrowing by Fourier self-deconvolution (FSD) using a half-bandwidth of 20 cm1 and a band-narrowing factor k = 2.0 was performed to resolve overlapping IR bands.50 The number and positions of IR bands were taken from the second derivatives of the FSD spectra. All of the subtraction, FSD, and second derivative calculation were carried out using the OMNIC 7.2 software (Thermo Electron Co.). The secondary structures were determined by fitting the amide I0 band of the FSD IR spectra with Gaussian+Lorentz functions using the PeakFit 4.05 software (AISN Software Inc.). 2.4. InterruptionIncubation Experimental Protocol. A special experimental protocol was designed for both DSC and FTIR measurements: Heating a protein sample from an initial temperature, T0 (e20 °C), to a predesignated interrupting temperature (Tint), incubating for a period of time τ, then cooling to T0, and then reheating to 90 °C or above. The temperature scan rate in both DSC and FTIR experiments was maintained at 1 °C/min. The protocol is shown schematically as follows: ðiÞ first heating
ðiiÞ incubation time τ
ðiiiÞ cooling
T0 s f Tint s f Tint s f T0 ðivÞ second heating
s f 90 °C The Tint and τ are variables and can be predesignated in different experiments. Additionally, the actual experiments can include all of the procedures or only (i) and (ii). We refer to these experiments as “interruptionincubation experiments”.
3. RESULTS AND DISCUSSION 3.1. Protein Thermal Denaturation Detected by DSC and FTIR. DSC results of the first and second heating scans from 30 to
90 °C are shown in Figure 1. Both BSA and lysozyme are featured by a single endothermic peak in the first heating process. The Tonset, Td, and integrated ΔHcal of the endothermic peaks are listed in Table 1. After immediately cooling from 90 °C, BSA and lysozyme give only baselines in the second heating scans, indicating both proteins have undergone irreversible denaturation. 8902
dx.doi.org/10.1021/jp200296v |J. Phys. Chem. B 2011, 115, 8901–8909
The Journal of Physical Chemistry B
ARTICLE
Table 1. Thermodynamic Parameters of the Thermal Denaturation Processes of BSA and Lysozymea Tonset (°C)
Td (°C)
ΔHcal (kJ/mol)
BSA
61.1
66.5
923.2
lysozyme
64.6
71.1
424.5
a
Sodium phosphate buffer in H2O, pH 7.4; the protein concentrations are 5 mg/mL.
Figure 1. DSC results of 5 mg/mL (A) BSA and (B) lysozyme at pH 7.4 (phosphate buffer in H2O) and a scan rate 1 °C/min. The black solid lines are the first heating scans and the red dash lines are the second heating scans.
These results are similar to previous studies of BSA41,42 and lysozyme,32,51,52 with respect to the thermodynamic parameters and the shape of the peak. The structural changes during the thermal denaturation processes of BSA and lysozyme over a temperature range of 2090 °C were examined by FTIR. The IR bands from 1590 to 1710 cm1 (the amide I0 bands) upon heating, cooling, and reheating are presented in Figure 2. The amide I0 bands upon heating were deconvolved to reveal the overlapping subcomponent bands, and each subcomponent IR band was assigned (see Table 2) according to previous IR studies of BSA,8,43,44 lysozyme,8,53,54 and other proteins or polypeptides.55,56 The secondary structure of BSA is mainly composed of R-helices (1653 cm1), β-turns (1678 cm1), and short segments connecting R-helices (1632 cm1). Native lysozyme consists of R-helices (1653 cm 1 ), β-turns (1674 cm 1 ), intramolecular β-sheets (1638 cm1), and antiparallel or high-energy β-sheets (1686 cm1). As shown in Figure 2, from 70 to 90 °C for BSA or 60 to 80 °C for lysozyme in the first heating scans, the peaks representing R-helices and other secondary structures show remarkable red shifts. This is due to the loosening of the compact hydrophobic cores and the exposure of peptide chains to solvent, thereby resulting in the formation of more hydrogen bonds between CdO groups and water molecules. The secondary structural changes in BSA and lysozyme upon heating, expressed as the changes of peak area percentages, can be obtained and the results are shown in Figure 3. For the thermal denaturation of BSA (Figure 3A), accompanying the loosening and solvation of hydrophobic chains, a peak for more hydrogen bonded β-turns (1666 cm1) is observed at ∼60 °C, whereas the peak for the original β-turns (1678 cm1) remains essentially unperturbed over the whole temperature range examined (data not shown). Subsequently, BSA takes an abrupt structural change from 70 to 80 °C: R-helices and short segments connecting R-helices collapse, while inter- and intramolecular β-sheets (1616 and 1638 cm1, respectively), as well as random coils (1643 cm1), emerge. Lysozyme (Figure 3D) takes an apparently “two-step” denaturation process. The collapse of R-helices and the formation of β-turns take place first at 5866 °C and then at 7886 °C. In the higher temperature range, the formation of antiparallel β-sheets (1686 cm1 ) and random
coils (1646 cm1) accompany the transition from R-helices to β-turns. Protein aggregation is characterized by the formation of intermolecular β-sheets (1616 cm1). For BSA (Figures 2A and 3A), both inter- and intramolecular β-sheets form at 7090 °C. This indicates that for BSA, aggregation accompanies the unfolding process within the transition temperature range. Moreover, increasing and then decreasing trend of random coils can be observed around 80 °C (Figure 3A). This implies that at least part of the random coils serve as an “intermediate structure” in the transformation from helices to strands. At >85 °C the protein completely transforms to the denatured state with aggregated structures. However, in the case of lysozyme, by comparison, intermolecular β-sheets form in abundance upon cooling (Figure 2B), which is consistent with a previous study.57 This implies that aggregation of lysozyme takes place after the unfolding process. A peak centered at ∼1516 cm1 is due to the aromatic ring stretching vibration of tyrosine. According to previous studies,49,58,59 the shift of this peak can serve as an indicator of protein tertiary structural changes. Figure 4, panels A and C, shows the Tyr peak shifts of BSA and lysozyme, respectively. The Tyr bands in the two proteins undergo two-step downward shifts in wavenumber during the first heating, which is consistent with a previous study of BSA.43 The red shifts are attributed to the exposure of Tyr residues to solvent and therefore the strengthened hydrogen bonding between Tyr OH moieties and D2O. It is worth noting that, for BSA, the tertiary structural changes at 4050 °C (Figure 4A) are much earlier than the secondary structural changes starting at 70 °C (Figure 3A). Similar phenomenon has also been observed in lysozyme. It is in line with a previous study, in which the tertiary structural changes were found to be a precondition of the secondary structural changes.60 Then upon cooling and the second heating, the partially reversible shifts of the band show that proteins regain some extent of tertiary structures, which may not be the original tertiary structures. 3.2. Protein Denaturation/Aggregation Behaviors Revealed by FTIR InterruptionIncubation Experiments. The structural changes of BSA were examined by incubating the sample at a predesignated Tint for a sufficient period of time to allow the process to reach the maximum extent, or the equilibrium state, where protein unfolds to a thermodynamically stable state and cannot proceed any further. This experiment involved steps (i) and (ii) of the specially designed interruptionincubation protocol (section 2.4) with Tint = 71 and 75 °C. The obtained secondary structural changes are shown in Figure 3, panels B and C. Compared to Figure 3A, it is clear that, after it reaches Tint, BSA continues to unfold upon incubation as evidenced by the continued changes in secondary structures. The structural changes proceed to the maximum extent after incubation for ∼2 h. Assuming the completely denatured state has been reached 8903
dx.doi.org/10.1021/jp200296v |J. Phys. Chem. B 2011, 115, 8901–8909
The Journal of Physical Chemistry B
ARTICLE
Figure 2. Representative Fourier self-deconvolved FTIR absorbance spectra of 25 mg/mL (A) BSA and (B) lysozyme at pD 7.4 (phosphate buffer in D2O). The spectra from bottom to top were measured during the first heating, cooling, and immediate reheating scans, respectively.
Table 2. Infrared Absorption Band Positions, Assignments, and Corresponding Structures of BSA and Lysozymea wavenumber (cm1) BSA
lysozyme
assignment
1516
1516
tyrosine residues
tertiary structure
1616
1616
intermolecular
secondary structure
β-sheet 1632
structure
(amide I0 )
short segments connecting R-helices
1638
1638
intramolecular β-sheet
1643
1646
random coil
1653
1653
(disordered structure) R-helix β-turn
1666 1678
1674
β-turn
1684
1686
antiparallel β-sheet
a Sodium phosphate buffer in D2O, pD 7.4; the protein concentrations are 25 mg/mL.
at 90 °C in Figure 3A, BSA cannot reach that state when incubated at 71 and 75 °C for 6 h. The equilibrium states appear to represent particular intermediate states, and the state BSA reaches upon incubation at 75 °C approaches the completely denatured state more closely than that upon incubation at 71 °C. Moreover, the time scales that BSA requires to reach equilibrium state at the two Tint are shorter than 2 h. Similar investigations on lysozyme were also carried out. In contrast to BSA, the equilibrium states that lysozyme reaches upon incubation at 72 and 75 °C (Figure 3, panels E and F) are just the completely denatured state as shown by the peak area percentages at 90 °C (Figure 3D). Different time scales are required for lysozyme to reach the completely denatured state: 12 h upon incubation at 72 °C and 2 h at 75 °C. This sharp
contrast in time scale is very interesting, as the discrepancy of the secondary structures at 72 and 75 °C in the consecutive scan is quite small (Figure 3D). Aggregation, signaled by the 1616 cm1 band, was seen to be enhanced/induced in the incubation process for BSA/lysozyme. For BSA, although aggregation started in the unfolding process upon heating (Figure 3A), it needed longer time to reach the maximum aggregated state as evidenced by the increasing peak area percentages of the 1616 cm1 band upon incubation. For lysozyme, aggregation was not observed in the heating scan. However, it was found to take place after long-time incubation (6 h in Figure 3E and 0.5 h in Figure 3F). Another feature can be seen in Figure 3 is that aggregation of the two proteins is moderate. For BSA, the peak area percentages of the 1616 cm1 band are
