⇦Chinese Physics Letters, 2016, Vol. 33, No. 2, Article code 028701⇨ The Structural Stability of Alpha-Helix Determined by the Preference of Amino Acids * Xiao-Xiao Xie(谢潇潇)1‡, Jun-Wei Li(李军委)1‡, Shao-Ying Xiao(肖少英)2, Yu-Zhi Liu(刘玉芝)3, Hui Liu(柳辉)1, Jin-Peng Geng(耿金鹏)1, Su-Hua Zhang(张素花)1, Hui Yu(于慧)1, Yong Zhan(展永)1**, Hai-Long An(安海龙)1** Affiliations 1Key Laboratory of Molecular Biophysics, Institute of Biophysics, School of Sciences, Hebei University of Technology, Tianjin 300401 2School of Architecture & Art Design, Hebei University of Technology, Tianjin 300401 3School of Electrical and Electronics Engineering, Shijiazhuang Tiedao University, Shijiazhuang 050043 Received 27 July 2015 ^*Supported by the National Natural Science Foundation of China under Grant Nos 11247010, 11175055, 11475053 and 11347017, the Natural Science Foundation for Distinguished Young Scholars of Hebei Province under Grant No C2015202340, the Natural Science Foundation of Hebei Province under Grant Nos C2012202079 and C201400305, and the Scientific Innovation Fund for Excellent Young Scientists of Hebei University of Technology under Grant No 2015010.
^**Corresponding author. Email: hailong_an@hebut.edu.cn; zhany@hebut.edu.cn
^‡These authors contributed equally to this work. Citation Text: Xie X X, Li J W, Xiao S Y, Liu Y Z and Liu H et al 2016 Chin. Phys. Lett. 33 028701 Abstract To accomplish their functions, proteins have to achieve different conformations accompanied by conformational transitions. However, the relationship between the preference of amino acids and the stability of the secondary structure is still unclear. Here we perform molecular simulations on a series of helical structures. Our data show that the dissociation energy of the helical structure is related to the preference of amino acids, and the electrostatic repulsion of the residue $i$ and $i+3/4$ with the same sign of charge destabilizes the alpha helix. DOI:10.1088/0256-307X/33/2/028701 PACS:87.15.hp, 87.15.ap, 87.16.Vy © 2016 Chinese Physics Society Article Text It is well known that there are numerous proteins in each cell, which play key roles in a variety of physiological processes. These proteins consist of several kinds of secondary structures, such as helix, loop, $\beta$-sheet. To accomplish their physiological roles, these proteins are often involved with various conformational changes. During the conformational changes, some helices become a loop or semi-helix, while the others are stable and stay as helix. Among these proteins, the linker regions of the calmodulin (CaM), NADP-IDH and Kir channels all have two different conformations (loop or helix).[1-4] The C-linker (K185-T192) of the Kir channels has different conformations corresponding to open and closed states.[5,6] The Kir channel is one of transmembrane proteins, and is activated by phosphatidylinositol 4,5-bisphosphate (PIP$_{2}$). The electrophysiological data and crystal structure show that PIP$_{2}$ binding sites are at the C-linker.[6,7] The crystal structures show that the C-linker could transform between compact helix and flexible loop with binding PIP$_{2}$ or not.[6,7] Most transmembrane helices show high stability when performing their physiological functions. What makes the different behaviors of various helices? The relationship between the preference of amino acids and the stability of the secondary structure is still obscure. In this Letter, we investigate how the amino acids affect the structural stability of the proteins. To achieve the conformational transition from helix to loop, we perform the steered molecular dynamics (SMD)[8] simulations on four helical structures. Our data show that the helices, which are rich in charged residues, are unstable and could be disrupted with small forces, that is, the electrostatic repulsion from charged amino acids with the same sign contributes to destabilizing the helical structures. Homology models of the full-length mouse Kir2.1 channel were performed by using the SWISS-MODEL server based on the crystal structure of chicken Kir2.2 (PDB code: 3SPI).[9-11] For structural optimization, the MD simulations were performed with the NAMD2 program (http://www.ks.uiuc.edu/Research/namd/)[12] and the CHARMM 27 forces filed.[13] The C-linker, the upper part of inner transmembrane helix-M2 of Kir2.1 channel (M2-UP) and the slide helix of Kir2.1 channel (SLIDE HELIX) come from the equilibrated mKir2.1 structure.[14,15] The other helical structure is a fragment of a crystal structure PDB code 1W5L and has been optimized before the SMD simulation. We applied SMD on the helical structures described above to accomplish the helix-coil transition in vacuum, and examined the resulting free energy. Focusing on how the preference of amino acids affects the stability of alpha-helix, we performed the SMD on the helices in vacuum. The pH is set as 7 through all simulations. In our simulations, constant velocity (cv-SMD) stretching has been employed to model mechanical unfolding of the helix. In cv-SMD, the external force originated from a harmonic spring that works on the specific atom and the spring is moved with constant velocity in the direction of the extension vector connecting the two termini. A free energy profile as a function of a coordinate is called a potential of mean force (PMF). Resulting PMF-extension profiles include the information of the stability with measuring the stretching work. We measured the distance which determines the initial value of the $x$-axis between the two N atoms which are from the N-terminus and the C-terminus residue, respectively. We fixed the N atom of the N-terminal residue. The other end (the capping N atom at the C-terminus, SMD atom) is constrained to a dummy atom that moves 16 Å along the direction between the fixed and the SMD atom (Fig. 1). We calculated the PMF involved in the helix-coil transition by employing Jarzynski's identity. With a slow-enough pulling simulation, the reversible work curve can be considered as an exact PMF.[16,17] We performed all simulations with a time step of 2 fs. The pulling velocity was set to 0.0001 Å/ps with the force constant of the harmonic potential as 7.2 kcal/mol/Å$^{2}$. The total number of steps is $8.0\times10^{7}$, and the total time is 160 ns. The simulations were carried out at a constant temperature of 300 K. As a control of the temperature, the Langevin dynamics scheme is used.

Fig. 1. Unfolding of helical C-linker. (a) A folded configuration ($\alpha $-helix). The four hydrogen bonds that stabilize the helix are shown as dotted lines. (b) An extended configuration (coil). The backbone of the peptide is represented as a ribbon. The fixed N atom of the first residue is pointed out in the right panel. The moving guiding potential used in the pulling simulations is represented by a spring which is connected to the C-terminus and pulled with a constant velocity $v$. This figure is made with VMD.[18]

In the cv-SMD, the end-to-end distance was the distance between the fixed and the SMD atom (Fig. 1). As shown in Fig. 1(a), there are four backbone hydrogen bonds which are thought to stabilize the helical structure. However, these hydrogen bonds are broken during the conformation transition achieved by the cv-SMD simulation (Fig. 1(b)). Firstly, we carried out a cv-SMD simulation on stretching of the C-linker (185 KPKKRNET192). The PMF is calculated from the stretching of the C-linker (Fig. 2(a)). The PMF equals to the work which is needed to break the intrahelical hydrogen bonds. Figure 2(b) shows the time course of the variation of the intrahelical hydrogen bonds (H bond). We identified four pairs of backbone H bonds in the static structure which are K185-K188, P186-R189, K187-N190, K188-E191 and thought to stabilize the helical structure. The four backbone H bonds are sensitive to the conformation and broken during the distortion of the helix in the earlier cv-SMD simulation (0–40 ns) (Figs. 2(c)–2(f)). This means that the free-energy difference for the helix-coiled transition is very small. In other words, the C-linker is so unstable that it can be stretched easily to coiled loop with few free energy consumption. We hypothesized that the amino acids with the positive charge have electrostatic repulsion which results in the unstable helix.

Fig. 2. (a) PMF calculated from reversible pulling of C-linker (185KPKKRNET192). The position of the constraint center is varied from 13 to 29 Å. (b) The total number of backbone hydrogen bonds plotted against the time. A minimum heteroatomic distance of 3.5 Å (between N and O) and a minimum bond angle of 135$^{\circ}$ (N–H${\ldots}$O) were used for defining a hydrogen bond. (c)–(f) The numbers of the four backbone hydrogen bonds which stabilize the structure. The four pairs of residues are K185-K188, P186-R189, K187-N190, K188-E191, respectively.

For validation, we performed the cv-SMD simulation which was on the other three short helical structures, the M2-UP, SLIDE HELIX and the fragment of a crystal structure (PDB code: 1W5L). In contrast, the PMF pattern of the M2-UP is significantly different from that of C-linker (Fig. 3(a)). Figure 3(b) shows the time course of the variation of the intrahelical hydrogen bonds. When the four backbone H bonds are all broken at about 50 ns (Figs. 3(c)–3(f)), the PMF achieved the first platform (5 kcal/mol). However, unlike the H bonds of C-linker, new backbone H bonds are emerging during the distortion of the helix as shown in Fig. 3(b). The helix-coil transition of M2-UP suggests a larger free energy difference than the C-linker does, which indicates that the M2-UP helix is difficult to unfold. The interactions between the amino acids may affect the stretching processes, and the corresponding effects could be analyzed based on their sequences.

Fig. 3. (a) PMF calculated from reversible pulling of M2-UP (155 PIAVFMVV 162). The position of the constraint center is varied from 10 to 26 Å. (b) The total number of backbone hydrogen bonds plotted against the time. A minimum heteroatomic distance of 3.5 Å (between N and O) and a minimum bond angle of 135$^{\circ}$ (N–H${\ldots}$O) were used for defining a hydrogen bond. (c)–(f) The numbers of the four backbone hydrogen bonds which stabilize the structure. The four pairs of residues are P155-F159, I156-M160, A157-V161, V158-V162, respectively.

Fig. 4. (a) PMF calculated from reversible of 1W5L-6EDKLEEIL13. The position of the constraint center is varied from 11 to 27 Å. (b) The total number of backbone hydrogen bonds plotted against the time. A minimum heteroatomic distance of 3.5 Å (between N and O) and a minimum bond angle of 135$^{\circ}$ (N–H${\ldots}$O) were used for defining a hydrogen bond. (c)–(f) The numbers of the four backbone hydrogen bonds stabilizing the structure. The four pairs of residues are E6-E10, D7-E11, K8-I12, L9-L13, respectively.

Next, we moved to another two helical structures, the SLIDE-HELIX and 1W5L. The 1W5L (PDB: 1W5L-6EDKLEEIL13) contains five amino acids with charge among 8 amino acids. We performed SMD simulations on both the above helices. As shown in Fig. 4, the PMF and H bonds patterns of 1W5L are similar to the one shown in Fig. 2. The data of the SLIDE-HELIX (Fig. 5) are similar to the data of M2-UP helix, as shown in Fig. 3. These results are consistent with the hypothesis, and the presence of charged residues will destabilize the helices.

Fig. 5. (a) PMF calculated from reversibility of SLIDE-HELIX-72IFTTCVDI79. The position of the constraint center is varied from 10 to 26 Å. (b) The total number of backbone hydrogen bonds plotted against the time. A minimum heteroatomic distance of 3.5 Å (between N and O) and a minimum bond angle of 135$^{\circ}$ (N–H${\ldots}$O) were used for defining a hydrogen bond. (c)–(f) The numbers of the four backbone hydrogen bonds stabilizing the structure. The four pairs of residues are I72-C76, T74-V77, T74-D78, T75-I79, respectively.

Based on the Chou–Fasman method which is a classical prediction method for secondary structures in bioinformatics, the polar residues are generally not likely to appear in helical structures. However, to accomplish their physiological roles, in some times, proteins undergo conformational transitions, some helices transit to loop or semi-helix, while the others stay at helix. Here we perform the SMD simulations on four helical structures aimed at understanding the relationship between the preference of amino acids and the stability of alpha-helix. The results show that the charged residue rich helix is more unstable than the helix with few charged residues which indicates that the electrostatic repulsion of the residue $i$ and $i+3/4$ with the same sign of charge destabilizes the alpha helix. We focus on a class of special helix which could transfer between helix and loops functionally. The present results may guide future research on investigating the preference of amino acids to determine the stability of the alpha helix. References Crystal Structure of the CaV2 IQ Domain in Complex with Ca2+/Calmodulin: High-Resolution Mechanistic Implications for Channel Regulation by Ca2+Insights into voltage-gated calcium channel regulation from the structure of the CaV1.2 IQ domain–Ca2+/calmodulin complexA Closed Compact Structure of Native Ca2+-CalmodulinStructures of Human Cytosolic NADP-dependent Isocitrate Dehydrogenase Reveal a Novel Self-regulatory Mechanism of ActivityCrystal Structure of the Eukaryotic Strong Inward-Rectifier K+ Channel Kir2.2 at 3.1 A ResolutionStructural basis of PIP2 activation of the classical inward rectifier K+ channel Kir2.2Crystal Structure of the Mammalian GIRK2 K+ Channel and Gating Regulation by G Proteins, PIP2, and SodiumSteered molecular dynamics simulations of force-induced protein domain unfoldingSWISS-MODEL and the Swiss-Pdb Viewer: An environment for comparative protein modelingSWISS-MODEL: an automated protein homology-modeling serverThe SWISS-MODEL workspace: a web-based environment for protein structure homology modellingScalable molecular dynamics with NAMD All-Atom Empirical Potential for Molecular Modeling and Dynamics Studies of Proteins ^† Identification of the Conformational transition pathway in PIP2 Opening Kir ChannelsIdentification of Three Interactions to Determine the Conformation Change and to Maintain the Function of Kir2.1 Channel ProteinNonequilibrium Equality for Free Energy DifferencesEquilibrium free-energy differences from nonequilibrium measurements: A master-equation approachVMD: Visual molecular dynamics

[1]	Mori M X, Vander Kooi C W, Leahy D J et al 2008 Structure 16 607
[2]	Van Petegem F, Chatelain F C and Minor D L Jr 2005 Nat. Struct. Mol. Biol. 12 1108
[3]	Fallon J L and Quiocho F A 2003 Structure 11 1303
[4]	Xu X, Zhao J, Xu Z et al 2004 J. Biol. Chem. 279 33946
[5]	Tao X, Avalos J L, Chen J et al 2009 Science 326 1668
[6]	Hansen S B, Tao X and MacKinnon R 2011 Nature 477 495
[7]	Whorton M R and MacKinnon R 2011 Cell 147 199
[8]	Lu H and Schulten K 1999 Proteins 35 453
[9]	Guex N and Peitsch M C 1997 Electrophoresis 18 2714
[10]	Schwede T, Kopp J, Guex N et al 2003 Nucl. Acids Res. 31 3381
[11]	Arnold K, Bordoli L, Kopp J et al 2006 Bioinformatics 22 195
[12]	Phillips J C, Braun R, Wang W et al 2005 J. Comput. Chem. 26 1781
[13]	MacKerell A D, Bashford D, Bellott M et al 1998 J. Phys. Chem. B 102 3586
[14]	Li J, Lü S, Liu Y et al 2015 Sci. Rep. 5 11289
[15]	Li J W, Xiao S Y, Xie X X et al 2015 Chin. Phys. Lett. 32 028702
[16]	Jarzynski C 1997 Phys. Rev. Lett. 78 2690
[17]	Jarzynski C 1997 Phys. Rev. E 56 5018
[18]	Humphrey W, Dalke A and Schulten K 1996 J. Mol. Graphics 14 33