Effective Potentials for Folding Proteins

Abstract

A coarse-grained off-lattice model that is not biased in any way to the native state is proposed to fold proteins. To predict the native structure in a reasonable time, the model has included the essential effects of water in an effective potential. Two new ingredients, the dipole-dipole interaction and the local hydrophobic interaction, are introduced and are shown to be as crucial as the hydrogen bonding. The model allows successful folding of the wild-type sequence of protein G and may have provided important hints to the study of protein folding.PACS numbers: 87.15. Aa, 87.14.Ee The problem of predicting the native structure of a protein for a given sequence has been of great interest due to its relevancy to many fields in biology. In the crudest level, lattice models are proposed and have provided important insights [1, 2]; however, due to the oversimplification, they are far from real applications. On the other hand, all-atom simulations deliver more details for the folding process, but the requirement of computational resources tends to be realistically unaffordable [3]. Developing models of coarse graining thus becomes the next step. For this purpose, off-lattice models [4] using Gō-type [5] potentials have been used to explore the folding dynamics. Since the relevant interactions are based on native structures, the Gō-type potentials can not be used to predict structures. There are also models that succeeded in separately folding helix bundles or folding beta hairpins [6]. Nevertheless, the interacting potentials employed are also biased towards the native states. So far, there is no model that can fold proteins using realistic potentials, it is therefore desirable to construct a coarse-grained model that can fold proteins without being biased in any way to the native state.In this paper, based on microscopic considerations, we propose a coarse-grained model with realistic potentials. The model has been tested successfully on more than 16 small proteins, of sizes from 12 to 56 amino acids [7]. For most examples, even without particularly optimizing our code, the computing time is reasonably short and is within the order of hours on ordinary desktop computers. Here, instead of exploring its predicting ability, we shall be focusing on only one protein (one of the protein G families with PDB ID : 1GB4) to illustrate the folding mechanism embedded in the proposed model. A brief summary of other important proteins is given in [7].In our model, side-chains are coarse-grained as spheres but explicit structures are kept in backbones [8]. On the other hand, water molecules are not included explicitly but their effects are incorporated in effective potentials among side-chains and backbones. The hydrophobic (HP) interaction has been known as the most important effect due to water. Recently, it is realized that the length-scale of water molecules has to be kept at short distances to prevent proteins collapsing prematurely [10]. Therefore, the desolvation model [10] combined with the Miyazawa-Jernigan (MJ) matri...

Effects of surface hydrophobicity on the conformational changes of polypeptides of different length

“…The backbone hydrogen bonding interaction plays a crucial role in the formation and stabilization of ordered secondary structures of polypeptide. To accurately describe the geometry of the hydrogen bond, the hydrogen bonding potential developed by Chen et al [32] was employed here:…”

Role of Backbone Dipole Interactions in the Formation of Secondary and Supersecondary Structures of Proteins

“…13,14 The inclusion of explicit dipole–dipole interactions in these types of models has been shown to stabilize β-sheets. 1,15,16 However, all the above-mentioned models renormalize the role of the solvent through effective short-range, inter-residue interactions. Thus, they are not appropriate for studying the effect of the environment in protein folding because explicit solvent is needed.…”

Recent Advances in Transferable Coarse-Grained Modeling of Proteins

“…Furthermore, a backbone hydrogen bonding potential as a function of the relative distance and orientation of an amide and a carbonyl group was added following Irbäck et al 39 , where a radial 12-10 Lennard-Jones potential is combined with an angular term. It has been shown previously 98 that carbonyl and amide groups at the peptide bond form dipoles that interact with each other and that the nearest-neighbor interaction is enough to favor β over α content 99 . Therefore, an extra term describing dipole-dipole interactions of neighboring residues is also included in this model.…”