The Role of Electrostatics and Folding Kinetics on the Thermostability of Homologous Cold Shock Proteins

Abstract

Understanding which aspects contribute to the thermostability of proteins is a challenge that has persisted for decades, and it is of great relevance for protein engineering. Several types of interactions can influence the thermostability of a protein. Among them, the electrostatic interactions have been a target of particular attention. Aiming to explore how this type of interaction can affect protein thermostability, this paper investigated four homologous cold shock proteins from psychrophilic, mesophilic, thermophilic, and hyperthermophilic organisms using a set of theoretical methodologies. It is well-known that electrostatics as well as hydrophobicity are key-elements for the stabilization of these proteins. Therefore, both interactions were initially analyzed in the native structure of each protein. Electrostatic interactions present in the native structures were calculated with the Tanford-Kirkwood model with solvent accessibility, and the amount of hydrophobic surface area buried upon folding was estimated by measuring both folded and extended structures. On the basis of Energy Landscape Theory, the local frustration and the simplified alpha-carbon structure-based model were modeled with a Debye-Hückel potential to take into account the electrostatics and the effects of an implicit solvent. Thermodynamic data for the structure-based model simulations were collected and analyzed using the Weighted Histogram Analysis and Stochastic Diffusion methods. Kinetic quantities including folding times, transition path times, folding routes, and Φ values were also obtained. As a result, we found that the methods are able to qualitatively infer that electrostatic interactions play an important role on the stabilization of the most stable thermophilic cold shock proteins, showing agreement with the experimental data.

Figure 1:

A) Cold shock proteins from Listeria monocytogenes (Lm-Csp), Bacillus subtilis (Bs-CspB), Bacillus caldolyticus (Bc-Csp) and Thermotoga maritima (Tm-Csp). B) Electrostatic surface of each Csp. C) Sequence alignment of the four cold shock proteins. Conserved residues are presented on a gray background. Charged residues are shown in red and blue, for negative and positive charges, respectively.

Figure 2:

A) Gibbs free energy from charge-charge interactions (ΔG_qq) for each Csp in neutral pH and at 300 K. Red and blue bars represent the negatively and positively charged residues, respectively. The destabilizing residues with surface accessible to solvent higher than 0.5 kj/mol are denoted with an asterisk. B) The same site for every Csp with charged residues displayed in sticks. C) Another site in Lm-Csp with unfavorable interactions. The same site for Bs-CspB, Bc-Csp, and Tm-Csp noted in other works.^,

Figure 3:

Stabilization due to burying hydrophobic surface area. A) Hydrophobic SASA buried upon folding and B) normalized by total SASA in the folded structure. C) ΔG_hϕ−SASA calculated based on buried hydrophobic SASA.

Figure 4:

Mutational and configurational frustration indexes (above and below of the matrix antidiagonal) change induced by adding electrostatic interactions for each protein.

Figure 5:

Single-residue frustration index change induced by adding electrostatic interactions for the chosen Csps.

Figure 6:

A) Heat capacity (C_v(T)) and B) fraction folded (f_N(T)) as a function of temperature for the chosen cold shock protein in simulations with the standard and charged SBM-C_α. For the charged C_α-model, the implicit salt concentration was increased from 0.0003 M to 0.3 M. The horizontal dashed line in B) defines the transition folding temperature (T_f) with f_N(T_f) = 0.5 that coincides with the temperature related with the heat capacity peak in A). The reduced temperature is dimensionless due to the construction of the model in Gromacs reduced units in order to have T_f and the simulated temperature around 1.0.

Figure 7:

A) Diffusion coefficient (D(Q)), B) drift-velocity (v(Q)) and C) free energy profile (F(Q)) as a function of the reaction coordinate fraction of native contacts (Q) for the chosen Csps at their respective T_fs. C_α-model molecular dynamics (MD) simulations were executed with and without charges and by varying the implicit salt concentration. Transition states are represented by the shaded gray area for each Csp. D and v has unit of [Q]²/[MD steps] and [Q]/[MD steps], respectively, which are dimensionless quantities.

Figure 8:

A) Mean first passage time τ_f) and B) mean transition path time (τ_TP) as a function of temperature for each cold shock protein run with SBM-C_α model simulations with and without charges and by varying the implicit salt concentration. Time is in unit of molecular dynamics (MD) steps. The reduced temperature is dimensionless due to the construction of the model that resulted in simulated temperatures around 1.0.

Figure 9:

Distribution of the electrostatic energy along the reaction coordinate Q for each Csp obtained in simulations of the charged SBM-C_α with 0.15 M of salt and at their respective T_fs. The concentrated regions indicates the folded and unfolded states of each protein.

Figure 10:

Routes of folding (R(Q)) as a function of the reaction coordinate Q for the four cold shock proteins simulated with and without charges, and by varying the implicit salt concentration at their respective T_fs. Transition states are represented by the shaded gray area for each Csp.

Figure 11:

Φ-values analysis for the chosen Csps simulated with the standard C_α-model (with no charged residues) and simulated with the charged model with 0.15 M of salt concentration and at their respective T_fs. A) is the difference in the Φ-values calculated with and without charge (ΔΦ) for each residue i in contact with residue j obtained with equation 17. B) shows the Φ− values averaged over native contacts involving residue i from A).