Contrasting Effects of Guanidinium Chloride and Urea on the Activity and Unfolding of Lysozyme.

Cosolvents play an important role in regulating the stability and function of proteins present in the cell. We studied the role of cosolvents, urea and guanidinium chloride (GdmCl), which act as protein denaturants, in the catalytic activity and structural stability of the protein lysozyme using activity measurements, spectroscopy, and molecular dynamics simulations. We find that the activity of lysozyme increases on the addition of urea, whereas it decreases sharply on the addition of GdmCl. At low GdmCl concentrations ([GdmCl] < 4 M), the activity of lysozyme decreases, even though there is no significant perturbation in the structure of the lysozyme folded state. We find that this is due to the strong interaction of the Gdm+ ion with the residues Asp52 and Glu35, which are present in the lysozyme catalytic site. In contrast, urea interacts with Trp63 present in the loop region present near the active site of lysozyme, inducing minor conformational changes in lysozyme, which can increase the activity of lysozyme. At higher denaturant concentrations, experiments show that GdmCl completely denatures the protein, whereas the folded state is stable in the presence of urea. We further show that GdmCl denatures lysozyme with the disulfide bonds intact in the protein, whereas urea denatures the protein only when the disulfide bonds are broken using reducing agents.

Introduction

To investigate the mechanism of protein folding, the native folded structure of proteins is perturbed using chemical denaturants in addition to temperature and force.[1−5] Urea and guanidinium chloride (GdmCl) are the most commonly used chemical denaturants.[6−15] To contrast the denaturing mechanisms of urea and GdmCl, we studied the unfolding and enzymatic activity of a very important protein, lysozyme, in the presence of urea and GdmCl. Lysozyme is a 129-residue protein[16] possessing antibacterial activity.[17] It catalyzes hydrolysis of peptidoglycan, which acts as the backbone for bacterial cell wall.[18,19] The bacterial cell wall components bind to the active site in lysozyme to form enzyme–substrate complexes. The residues Asp52 and Glu35, present in the active site of lysozyme, are found to play an important role in the catalysis mechanism.[16,18,20] The side chains of the catalytic residues Asp52 and Glu35 preferentially cleave the glycosidic bond connecting the two saccharide rings present in the substrate (Figure S1).[18−20] The hen egg white lysozyme (Protein Data Bank (PDB) ID: 1AKI)[21] in the folded state contains four disulfide bonds, two β sheets, and an α-helix core composed of five helices, which are more than five-residue long, in addition to small helices[21] (Figure ). The disulfide bonds enhance the stability of the folded state of the protein.

Figure 1

Crystal structure of lysozyme in the folded state (PDB ID:1AKI[21]). The β-sheets and helices are shown in blue and red, respectively. The unstructured part of the protein is shown as a gray coil. Four disulfide bonds are shown in yellow.

Even though lysozyme is the first biological enzyme whose structure is known, there are a few experimental[22−25] studies on the denaturation of lysozyme in physiological conditions. The main objective of this study is to analyze the effect of urea and GdmCl on lysozyme at neutral pH and in the absence of reducing agents. Experiments show that at neutral pH urea can denature lysozyme only when the disulfide bonds are broken by adding reducing agents like dithiothreitol.[26,27] In neutral conditions, no significant structural change in the protein is observed when urea is added to the solution, whereas complete protein unfolding is observed when GdmCl is used as a denaturant.[22−25] A comprehensive understanding of chemical denaturation of lysozyme by urea and GdmCl is still elusive, and the role of disulfide bonds in lysozyme unfolding in the presence of denaturants is unclear.

In this article, we have used spectroscopic measurements and molecular dynamics (MD) simulations to understand the effect of denaturants, urea and GdmCl, on the activity and denaturation of lysozyme. Analysis of experiments and simulations shows that urea does not unfold lysozyme at neutral pH and in the absence of reducing agents when the disulfide bonds are intact. The activity of lysozyme does not decrease at all in these conditions. In reducing conditions, when the disulfide bonds are broken, the activity of lysozyme decreases with the urea concentration and also the tertiary structure falls apart, linking the activity of lysozyme with the folded structure. However, in the presence of low GdmCl concentration and in close-to-neutral pH conditions, the activity of lysozyme decreases, but this is primarily due to the strong binding of the Gdm+ ion with the catalytic residues Asp52 and Glu35 in the active region of lysozyme. At high GdmCl concentrations, the activity falls sharply due to both the interaction of Gdm+ ion with the catalytic residues and also due to the breakdown in the folded structure. GdmCl unfolds lysozyme without the addition of reducing agents, and the disulfide bonds are intact in the denatured ensemble of lysozyme obtained by adding GdmCl.

Experimental and Computational Methods

Materials

Lysozyme and Micrococcus lysodeikticus were purchased from Sigma Aldrich. Urea (99% purity) and GdmCl (99.8% purity) were purchased from Spectro-Chem (India) and Merck, respectively. Purified water (H2O, Millipore, resistivity: 18.2 MΩ cm) is used to make different samples.

Activity Measurements

Lysozyme activity is determined according to the following enzymatic assay. Aliquots (∼100 μL) of the samples containing 200–400 units/mL of lysozyme are added to 1.4 mL of a ∼0.015% (w/w) suspension of M. lysodeikticus in a phosphate buffer of pH 6.24. The protein activity is measured by monitoring the decrease in the absorbance at λ = 450 nm (Thermo Fischer Scientific Evolution 201 spectrophotometer) at 25 °C due to the lysis of M. lysodeikticus cells (the measured absorbance is corrected by the absorbance of pure phosphate buffer, pH 6.24). The activity measurements have been performed 10 times to check the reliability and reproducibility of the data.

Spectroscopic Measurements

Steady-state fluorescence study is performed at 25 °C using a HORIBA JOBIN YVON Fluoromax-4 spectrofluorometer. The lysozyme (∼10 μM) samples are prepared in a buffer of pH 7 (25 mM Tris–HCl, 150 mM NaCl) containing 0–8 M urea and GdmCl solutions, respectively. The excitation wavelength at 280 nm with slit width of both the excitation and emission are at 2 nm. Circular dichroism (CD) spectra are recorded on a JASCO J-815 CD spectrometer (model number J-815-1505) using 28 ± 1 μM lysozyme for the far-UV (180–260 nm) and 80 ± 1 μM lysozyme for the near-UV region (260–330 nm). We have also calculated mean residue ellipticity (MRE) for both the bands of the CD spectra in the far-UV region.

Molecular Dynamics Simulations

We have performed molecular dynamics simulations to study the effect of denaturants, urea and GdmCl, on the stability of lysozyme. The initial protein conformations to start the simulations are generated using the crystal[21] structure of lysozyme (PDB ID: 1AKI) (Figure ). The simulations are performed using the GROMACS molecular dynamics (MD) simulation package[28] and optimized potential for liquid simulations (OPLS) all-atom force field.[29,30] The SPC/E water model is used to mimic the interaction of water with the protein.[31] The force-field parameters derived using the Kirkwood–Buff equations by Weerasinghe and Smith are used for urea and GdmCl molecules[32,33] (see the Supporting Information (SI), Figures S2, S3 and Table S1). The simulation cell is a cubic box, and periodic boundary condition is applied in the x, y, and z directions. The desired urea or GdmCl concentration is achieved by randomly inserting the required number of denaturant molecules in the simulation box with the protein, and finally, the box is solvated with water. The initial configuration is minimized using the steepest-descent method to remove unfavorable interactions between atoms. After minimization, the system is equilibrated for 1 ns by applying harmonic constraints to the heavy atoms of the protein. To dissect the role of disulfide bonds in the stability of the protein, we also performed simulations with and without the disulfide linkages starting from the protein crystal structure. Simulations are performed in the NPT ensemble. The temperature, T, and pressure, P, of the simulation box are maintained at 300 K (or 325 K) and 1 bar, respectively, using the modified Berendsen thermostat,[34] which is a stochastic velocity rescaling method[35] with a time constant of 0.1 ps, and Parrinello–Rahman barostat[36] with a time constant of 2 ps. The isothermal compressibility of the system is set to 4.5 × 10–5 bar–1. The nonbonded interactions between pairs of atoms are calculated according to the Verlet cut-off scheme.[37] Short-range van der Waals and electrostatic interactions are truncated at 10 Å. Long-range electrostatic interactions are calculated using the particle mesh Ewald method, with a grid size of 1.6 Å.[38] All of the covalent bonds are kept rigid using the LINCS algorithm,[39] and a time step of 2 fs is used to integrate the equations of motion.

Urea and GdmCl Force Field

The force field derived by Weerasinghe et al. is used in the current study for urea[33] and GdmCl[32] along with the SPC/E water model.[31] The force field is parameterized to reproduce the experimental density and Kirkwood–Buff integrals[40] as a function of composition. The computed surface tension of water/urea solutions using this force field is in good agreement with the experimental values,[41] and these force fields are found to be reasonable to model urea[41−57]and GdmCl[54,56,58−60] denaturant effects in protein folding. We performed a simulation of 8 M urea and 6 M GdmCl solution in a 40 Å water box in the NPT ensemble. All of the other simulation parameters and conditions are described in the section above. The radial distribution functions and first shell coordination numbers computed from the simulation are in agreement with the earlier reported values[32,33] (Figures S2, S3, and Table S1).

Data Analysis

The spatial density maps are computed using the volmap plugin in the Visual Molecular Dynamics package.[61] The space around the protein residues is divided into a grid of size 0.1 Å.[29] The grid points take the values 0 or 1, depending on the occupancy of atoms. The central carbon atom in urea or Gdm+ is treated as a sphere of radius equal to its van der Waals radius. A grid point is considered to be occupied if the sphere lies on the grid. The value at a grid point is averaged over multiple simulation frames. The radius of gyration (Rg) of the protein is calculated to estimate its size using the following equationwhere is the vector connecting the atom i to the geometric center of the protein. The hydrogen bond is defined to exist between two atoms if the distance between the donor atom (O or N) and the acceptor atom (O or N) is less than 3.5 Å and the angle between the donor, hydrogen, and acceptor atoms is greater than 135°. To compute the contact number between the denaturant and protein, a contact is defined to exist when any atom of the denaturant is within 5 Å of any atom of the protein residues.

Results and Discussion

The enzymatic activity of lysozyme as a function of time for different concentrations of urea and GdmCl solutions is shown in Figure . The dotted line in both the figure panels corresponds to the control experiment when lysozyme is not present in the cell suspension, whereas the solid lines show the activity of lysozyme with different concentrations of denaturants in the cell suspension. The enzymatic activity data clearly shows that the activity of lysozyme increases with the addition of urea, whereas it decreases monotonically with the addition of GdmCl compared with the native condition (Figure ). The increase in the activity of lysozyme induced by urea is ∼20%, and it is reproduced in multiple experiments, whereas the activity of lysozyme decreases sharply in the presence of GdmCl and is almost lost completely above 4.0 M GdmCl. The enzymatic activity data clearly suggests that the structural changes induced by urea and GdmCl in lysozyme are drastically different. The activity of lysozyme depends on the binding of the cell (substrate) to the active region of lysozyme, as shown in Figure S1. Both the availability of the active residues and the structural integrity of the protein will affect the binding. The change in activity data suggests that GdmCl interacts with the residues present in the active core region of the lysozyme, whereas urea does not directly interact with the residues in the active region. Urea probably enhances the interaction of Asp52 and Glu35 with the substrate, which can contribute to the slightly enhanced activity of lysozyme.

Figure 2

Time-dependent enzymatic activity of lysozyme in solutions having different concentrations of urea (a) and GdmCl (b). The dotted line in both the panels corresponds to the control experiment, which is performed in the absence of lysozyme in the cell suspension.

To understand the structural changes in lysozyme induced by urea and GdmCl, we have measured the CD spectra of lysozyme in far- and near-UV regions representing the secondary and tertiary structures of proteins, as shown in Figure . The protein shows two bands centered at around 210–219 and 222–232 nm corresponding to the β-sheet and α-helix structures, respectively. On addition of urea and GdmCl, the peak position of the bands centered at ∼210 nm gets shifted. Hence, we have considered the ellipticity of the peak maxima of the two bands to assess the change in the ellipticity of the bands centered at 210–219 and 222–232 nm regions. On addition of urea, ellipticity of the first band decreases slightly, whereas it increases marginally for the second band at ∼222 nm (inset of Figure a,b), indicating the decrease in β-sheet content on addition of urea (the mean residue ellipticity (MRE) at peak maxima of two bands is shown in Figure S4, which follows the same behavior). Unlike the case for urea, the ellipticity of both the bands decreases sharply with the addition of GdmCl compared with the native structure, indicating that the secondary structure of lysozyme changes markedly in the presence of GdmCl. Figure c,d shows the CD spectra of lysozyme in the tertiary region (contribution from the tryptophan as well as the disulfide bonds) in the presence of different concentrations of urea and GdmCl, respectively. The ellipticity of the tryptophan residue at 293 nm increases with the addition of urea (Figure c), whereas it decreases sharply with the addition of GdmCl and is ∼0 in 4 M GdmCl solution (Figure d). The CD data in the tryptophan region indicates that the exposure to and interaction of tryptophan residues with urea and GdmCl are completely different. The CD data of lysozyme in far- and near-UV regions clearly show that the addition of ∼4.0 M GdmCl unfolds lysozyme by drastically decreasing the secondary and tertiary structures, leading to the complete loss of its enzymatic activity. Unlike GdmCl, urea does not significantly perturb the secondary structure and increases the ellipticity in the tertiary region, which probably increases its enzymatic activity.

Figure 3

Far-UV (upper panels) and near-UV (lower panels) CD spectra of lysozyme in urea (a and c) and GdmCl (b and d). Ellipticity changes at the peak maxima of the bands in the ranges of 210–218 and 222–225 nm and at 293 nm with the addition of urea (black) and GdmCl (red) are shown in the insets of (a), (b), and (d), respectively.

We have also measured the intrinsic fluorescence of lysozyme in different concentrations of urea (Figure a) and GdmCl (Figure b) to understand the local solvent environment around the tryptophan residues and its effect on folding. The native form of lysozyme shows emission maxima (λmax) at 335 nm (black line, Figure a,b), and with the addition of urea, the change in λmax is not prominent and is almost constant within the experimental uncertainty (Figure c). In contrast to urea, λmax of lysozyme is constant up to 3 M GdmCl, and after that, it gets red-shifted ∼10 nm sharply up to 6 M and then becomes constant (Figure c). The change in λmax on addition of GdmCl follows a sigmoidal-shaped curve, which is trademark of protein denaturation. It is clear from this plot that lysozyme completely denatures when GdmCl concentration is greater than 4.0 M. The fluorescence intensity of lysozyme decreased on addition of urea, whereas it increased with the addition of GdmCl (Figure d). Lysozyme has six tryptophan residues, of which Trp62, Trp63, and Trp108 residues contribute prominently to the intrinsic fluorescence in the native form of lysozyme. Trp62 and Trp63 are in the irregular loop region and solvent-exposed, whereas Trp108 is buried in the active core region of lysozyme. The trend of λmax as well as the change in the emission intensity data clearly indicates that the addition of urea does not significantly perturb the polarity of the local environment. However, in the case of GdmCl (∼4 M), Trp108 is highly exposed to GdmCl/water and realizes more polar surroundings (Figure S5). The experimental data clearly suggests that the interactions of urea and GdmCl and their mechanisms of inducing structural change in lysozyme are distinctly different.

Figure 4

Fluorescence emission spectra of lysozyme in the presence of urea (a) and GdmCl (b). Change in λmax (c) and intensity (d) as a function of denaturant concentration.

We performed MD simulations to understand the effects of urea and GdmCl on the activity of and structural changes in lysozyme. There is a substantial decrease in the activity of lysozyme when the concentration of GdmCl is less than 4.0 M (Figure ), even though significant structural changes are not observed (Figure c,d). We computed the spatial density distribution of Gdm+ and urea within 5 Å of the protein (Figure a,b) to understand the interaction of the denaturants with different residues in lysozyme. In 2 M GdmCl solution, we find that the maximum occupancy of Gdm+ is near the residues Asp52 and Glu35, which are important for the catalytic activity of lysozyme (Figure a,c). This shows that Gdm+ interacts strongly with the active residues and can inhibit the interaction of the substrate with these residues, diminishing the catalytic activity of lysozyme. This also explains the decrease in lysozyme activity even at low GdmCl concentrations (Figure ). The interaction of Gdm+ with the catalytic residues is further verified by computing the radial distribution function, g(r), between the central C atom of urea or Gdm+ and acidic O atoms of Asp52 or Glu35. The computed g(r) shows that Gdm+ interacts with these residues strongly compared with urea (Figure e). The strong interaction between Gdm+ and the catalytic residues, Asp52 and Glu35 anions, is due to the electrostatic interaction. The average numbers of hydrogen bonds between Gdm+ and residues Asp52 and Glu35 are 3.3 and 2.8, respectively, in 2 M GdmCl solution, whereas urea is found to barely form hydrogen bonds with these residues. Asp and Glu are hydrogen bond acceptors, and Gdm+ is preferred over urea because of the following reasons: (1) Gdm+ is more polar than urea; (2) urea has only four hydrogen atoms that can be part of hydrogen bonding, whereas Gdm+ has six hydrogen atoms; (3) the geometry of Gdm+ is apt to make maximum number of hydrogen bonds with Asp52 and Glu35 (Figure c). The decrease in the activity of lysozyme in a low-concentration GdmCl solution must be due to the less accessibility of catalytic residues Asp52 and Glu35 rather than the structural change in lysozyme due to unfolding. At a higher concentration, the loss in activity is associated with the structural change. A similar observation has been reported in the case of papain.[62,63] At low denaturant concentrations, there was a loss in the activity of papain without any structural change, and this was attributed to the complex formation of the denaturant molecules with the active residues of papain.[62,63]

Figure 5

Spatial density maps of (a) Gdm+ and (b) urea within 5 Å of the protein with fractional occupancy more than 0.75 in 2 M denaturant solution at T = 300 K. The protein is shown in gray, and orange dots show Gdm+ or urea fractional occupancy. (c) Interaction of Gdm+ with Asp52 and Glu35 side chains. (d) Interaction of urea with Trp63 and Thr118 side chains. In (a)–(d), red spheres represent oxygen, blue spheres represent nitrogen, and white spheres represent hydrogen. (e) Radial distribution function, g(r), between the central C atom of urea or Gdm+ and acidic O atoms of Asp52 and Glu35. (f) Radial distribution function, g(r), between the amine nitrogen of Trp62 and Trp63 residues with the central C atom of urea and Gdm+.

From the spatial density distribution of urea, we find that urea strongly interacts with the residues Trp63 and Thr118 with an occupancy ∼0.8 (Figure b) and it does not interact strongly with the catalytic residues, Asp52 and Glu35 (Figure e). In urea solution, most of the catalytic residues are freely accessible due to their weak interaction with urea except Trp63, which is in the vicinity of the substrate-binding pocket. Trp63 is present in the loop region (Trp62 to Cys76) close to the active site, whereas Thr118 is far from the active region. The radial distribution function shows that Gdm+ does not interact very strongly with either Trp63 or Trp62, whereas urea interacts strongly with Trp63 but not with Trp62 (Figure f). The sharp peak in g(r) between Trp63 and Thr118 with urea implies that urea interacts strongly with these residues. The strong interaction with Trp63 could be due to the NH−π interaction as reported recently,[64] and direct hydrogen bonding between urea and Thr118 is also observed in the X-ray diffraction studies.[65] The partial changes in the activity of lysozyme with the increase in urea concentration can be possible due to the small conformational changes near the substrate-binding pocket due to the interaction of urea with Trp63 (Figure S6). The conformational changes are inferred from the root-mean-square deviation (RMSD) of Trp63 and the loop region compared with their position in the crystal structure (Figure S6). Higher RMSD values for Trp63 and the protein loop region are observed only in urea solutions and not in GdmCl solutions.

To understand the effect of denaturant and disulfide bonds on structural change, we have performed molecular dynamics simulations of lysozyme starting with the folded structure as the initial conformation in the presence (native) and absence (reduced) of disulfide bonds at different denaturant concentrations. For each denaturant concentration, the time scale of the simulation is 300 ns and the temperature is maintained at T = 325 K. The simulations show that in the presence of all of the four disulfide bonds the protein is very stable on a time scale of 300 ns and the average radius of gyration (Rg) of the protein is constant within the error bars as the denaturant (urea or GdmCl) concentration is varied from 0 to 8 M (Figure S7). In the absence of all of the four disulfide bonds, the average Rg of the protein is found to increase with an increase in the denaturant concentration (Figure S7). The simulation results indicate that the disulfide bonds enhance the stability of the folded state of lysozyme. We also performed a long simulation of ∼2.5 μs in 8 M GdmCl and urea solutions (Figure S8). The protein did not unfold on this time scale. To check for the force field effects, we also performed a long simulation (∼500 ns) using the CHARMM36 force field,[42] and the protein did not unfold on this time scale (Figure S9). Due to this reason, we cannot conclusively comment on the unfolding mechanism of lysozyme in the presence of urea and GdmCl from simulations.

To understand the role of disulfide bonds in lysozyme unfolding in the presence of denaturants, we have performed an assay experiment with 5,5′-dithio-bis(2-nitrobenzoic acid) (DTNB), which has been successively used for the identification of free −SH groups in proteins (details of the method are given in the SI).[66] The absorption spectra of DTNB with protein in the presence and absence of 6 M GdmCl and urea (Figure S10) are same and do not show any feature at ∼412 nm, suggesting that the addition of GdmCl and urea does not perturb the disulfide bonds present in lysozyme. This indicates that the disulfide bonds are intact in the denatured ensemble of lysozyme obtained by the addition of GdmCl.

The experimental activity data as well as the CD data in the tertiary region clearly indicates that the disulfide bonds present in lysozyme are not broken in the presence of urea. Furthermore, we have also performed the activity experiment of lysozyme in the reduced condition, which shows a gradual decrease in the activity (Figure S11) similar to the case of GdmCl, clearly suggesting that urea denatures the protein only when the disulfide bonds are broken using reducing agents.

In summary, lysozyme loses its activity at low GdmCl concentration due to the strong interaction between Gdm+ and residues present in the active site of lysozyme. At high GdmCl concentration, structural changes induced by GdmCl also contribute to the loss in lysozyme activity. In the urea solution, no loss in activity is observed as there is no strong interaction between urea and active residues. Fluorescence and CD spectra show that the protein unfolds in the presence of GdmCl and not in urea solution, and GdmCl unfolds lysozyme, without breaking the disulfide bonds present in the protein.