J Mol Model (2015) 21:88 DOI 10.1007/s00894-015-2621-5

ORIGINAL PAPER

Environmental polarity induces conformational transitions in a helical peptide sequence from bacteriophage T4 lysozyme and its tandem duplicate: a molecular dynamics simulation study Harpreet Kaur & Yellamraju U. Sasidhar

Received: 20 June 2014 / Accepted: 15 February 2015 # Springer-Verlag Berlin Heidelberg 2015

Abstract Our recent molecular dynamics (MD) simulation of an insertion/duplication mutant ‘L20’ of bacteriophage T4 lysozyme demonstrated a solvent induced α→β transition in a loosely held duplicate helical region, while α-helical conformation in the parent region was relatively stabilized by its tertiary interactions with the neighboring residues. The solution NMR of the parent helical sequence, sans its protein context, showed no inherent tendency to adopt a particular secondary structure in pure water but showed α-helical propensity in TFE/water and SDS micelles. In this study we investigate the conformational preference of the ‘parent’ and ‘duplicate’ sequences, sans the protein context, in pure water and an apolar TFE/water solution. Apolar TFE/water solution is a model for non-polar protein context. We performed MD simulations of the two peptides, in explicit water and 80 % (v/v) TFE/water, using GROMOS 53a6 force field, at 300 K and 1 bar (under NPT conditions). We show that in TFE/water mixture, salt bridges are stabilized by apolar TFE molecules and main chain-main chain hydrogen bonds promote the αhelical conformation, particularly in the duplicate peptide. Solvent exposure, in pure water, resulted in an α→β transition to form a triple stranded β-sheet structure in the ‘duplicate’ sequence, with a rare psi-loop topology, while a mixture of turn/bend conformations were adopted by the ‘parent’

Electronic supplementary material The online version of this article (doi:10.1007/s00894-015-2621-5) contains supplementary material, which is available to authorized users. H. Kaur : Y. U. Sasidhar (*) Department of Chemistry, Indian Institute of Technology Bombay, Powai, Mumbai 400076, India e-mail: [email protected] Present Address: H. Kaur Fachgebiet Protein Modelling, Technische Universität München, Emil-Erlenmeyer-Forum 8, D-85354 Freising, Germany

sequence. Thus the differences in conformational preference of the parent and duplicate sequence sans protein context, in pure water and TFE/water, implicate the importance of the environment polarity in dictating the peptide conformation. Mechanism of folding of the observed psi-loop in the duplicate sequence gives insights into folding of this rare β-sheet topology. Keywords α→β transition . Bacteriophage T4 lysozyme . Salt bridges . Solvent shielding . TFE . Triple stranded β-hairpin

Introduction Anfinsen postulated that the fold that a protein adopts is governed only by its primary amino acid sequence [1]. Since then sequence-structure relationship has been considered to play a substantial role in governing the conformational preference of a protein/peptide. For instance, peptide fragments involved in nucleation of their parent protein folding have been shown to adopt their inherent native fold even in the absence of the protein context. A series of molecular dynamics studies on the folding a β-hairpin sequence (21DTVKLMYKGQPMTFR35) [2], from staphylococcal nuclease, and different fragments corresponding to its turn region (YKGQP [3] and YKGQ [4]) showed that their intrinsic hairpin and turn propensities, respectively, were maintained even without the complete protein context indicating the importance of the primary sequence in determining the conformation adopted by the peptide fragment. It is also known that apart from the primary sequence, environmental factors like tertiary contacts within the folded protein, solvent polarity, ions, pH, temperature etc. have also been considered to be of

88

Page 2 of 14

great significance in determining the conformational preference of a sequence [5, 6]. Studies have shown that upon decreasing the solvent polarity an α- to 310-helix conformational switching occurs, as in the case of an acylated heptapep tide amide comprising six Cα-tetrasubstituted amino acids and one histidine [7] and an amphiphilic peptide of the triacylglycerol lipase derived from Pseudomonas aeruginosa [8]. NMR and molecular dynamics (MD) studies showed that a βoctapeptide adopts a 314-helical structure in methanol while it folds to a hairpin in water [9]. An MD study of α- and βpeptides shows that they show a greater stability for an αhelical conformation (stabilized by solute–solvent interaction and intra-solute entropy) in water and 314-helical conformation in methanol (stabilized by intra-solute interaction) [10]. In order to probe the effect of solvent polarity, organic cosolvents like TFE have also been used. For example, TFE stabilizes an amphiphilic α-helical intermediate, which was hypothesized to enhance the folding of a β-sheet protein in the apolar solvent [11]. This intermediate was found to be involved in a rate-limiting step of the folding reaction. CD and NMR studies of two antimalarial peptides have also been shown to promote α-helical conformations in TFE [12]. Comparison of folding free energy of a of poly-alanine peptide sequence in different alcohols showed that TFE stabilizes αhelical conformation better than ethanol, methanol, and pure water, as shown by CD and UV resonance Raman spectroscopy. Apart from stabilizing α-helical conformations, TFE has also been found to stabilize β structures in peptides [13]. An MD study of a designed triple stranded β-sheet peptide betanova and the β-hairpin sequence 41–56 from the B1 domain of protein G displayed an enhanced β structure propensity in TFE/water in comparison to pure water [14]. β-hairpin structure of five designed peptides was seen to be better stabilized in TFE/water than in aqueous solution as determined by NMR experiments [15]. In a study by Mandal et al., it has been shown that TFE denatures bovine spleen galectin-1 protein, leading to formation of a non-native α-helical structure [16]. Thus TFE promotes secondary structure formation in peptides, by mimicking a protein-like apolar environment, while it induces denaturation in proteins. Solution NMR of the peptide sequence corresponding to a helix in wild type (WT) T4 lysozyme (T4L), showed that the peptide possessed no inherent tendency to adopt any particular secondary structure in pure water but showed α-helical propensity (a maximum of 30 % helicity, in 50 % (v/v) TFE/ water) in TFE/water and SDS micelles [17]. In a later study by Sagermann et al., an insertion/duplication mutant of T4L was crystallized in which a duplicate of the same α-helical peptide sequence from the wild type T4L [18] was inserted in tandem at the N-terminal of the ‘parent’ α-helix (40 to 50) (Fig. 1). As a result, in the mutant T4L (called as ‘L20’) crystal structure, the ‘duplicate’ sequence extended the parent helix by two helical turns. However, our recent molecular dynamics

J Mol Model (2015) 21:88

Fig. 1 Schematic representation of conformations observed by the parent helical sequence (residues 40 to 50) and inserted duplicate sequence (residues 40i to 50i) in wild type bacteriophage T4 lysozyme (WT T4L) and its mutant L20, respectively, along with flanking loop residues. The inserted duplicate sequence extends the parent helix N-terminally by two helical turns (comprising residues K43i to I50i). Starting conformation of duplicate and parent peptide for molecular dynamics simulation (in this study) is also shown. Helical and loop conformations are represented by cylinders and tubes, respectively

(MD) simulation of the L20 crystal structure (in explicit water) displayed an α→β transition in the duplicate region of the mutant protein [19]. The parent helical region was seen to maintain its α-helical conformation relatively well throughout the L20 simulation. Apart from this, a control simulation of the WT crystal structure also displayed intact α-helical conformation in the parent region. According to Sagermann and Matthews, in the case of WT protein, the parent helix interacts with a β-sheet region through hydrophobic interactions [20]. In our simulation study we observed that a C-terminal β-sheet anchors the parent helix in L20 while the loosely held Nterminal loop in duplicate region is vulnerable to solvent attack and thus undergoes an α→β transition, thus indicating the importance of non-local tertiary interactions in stabilizing the parent α-helical conformation. To summarize, these studies show that the ‘parent’ 11reside peptide sequence 40NAAKSELDKAIN50 adopts a native α-helical conformation in its protein context [18, 19] and the protein like environment provided by apolar cosolvents [17]. However, the same sequence unfolds to form a random structure in water without protein context [17]. Also the same sequence when duplicated forms a β-structure that interacts with polar solvent water, in the presence of protein context [19]. Thus environmental polarity plays an important role in the conformation adopted by the parent peptide. However, the

J Mol Model (2015) 21:88

conformational features of its 22-residue ‘duplicate’ alone are not known. In order to examine the conformational preference of this sequence (from L20), we performed MD simulation of the ‘duplicate’ peptide in pure water. This simulation is also expected to provide insights into the results of our previous MD simulation [19] of full protein L20. Also, to test its conformational propensity in an apolar environment simulation of the peptide in a TFE/water mixed solution was also performed. This may be compared to the case when the helix packs with proximal residues in the full protein context through non-polar interactions. For this, an 80 % (v/v) TFE/water mixed solution was employed, which mimics a protein-like apolar environment. As a control, simulation of the sequence corresponding to the parent peptide was also performed in pure water as well as in 80 % TFE/water mixed solution. It has been suggested that the concentration of TFE required to maximally stabilize helical conformation in peptides depends on their intrinsic helical propensity [21]. Thus for peptides with low helical propensity 30–50 % TFE concentration is required to maximally stabilize their helical conformation while that for peptides with smaller helical propensity is 80 %. Also, it has been shown that the effect of solvent polarity influences the peptide conformations based on its chain length, as suggested by NMR, CD and molecular modeling studies of dehydropeptides containing (Z)-dehydrophenylalanine [22]. Apart from this it has been suggested that the helix formation propensity in TFE depends on its amino acid sequence [23]. Thus a comparison of conformational features of parent and duplicate peptides in water and TFE/water mixture, sheds light on the relevance of factors like amino acid sequence and length of the peptide in determining their conformational propensity in different solvents. Our MD results, presented in this paper, indicate that starting from the respective native α-helical conformations (observed in crystal structure of L20) for the parent and duplicate peptides, TFE seems to maintain the helical conformations in both the peptides to varying degrees as a function of time. An α→β transition was observed for the duplicate peptide in pure water. However, a mixture of bends and turn conformations were adopted by the parent peptide in pure water. These results indicate that a balance in surrounding apolar or polar environment, provided by TFE and water respectively, dictate the conformational preference of the peptides. Thus this study examines the conformational propensity of an α-helical peptide sequence and its tandem duplicate, sans the protein context, in pure water and an apolar environment modeled by TFE, to further understand the conformational preference of these peptides as a part of their parent protein. Further, the mechanism of α→β transition as observed in duplicate peptide in pure water may have implications for folding of a rarely observed psi-loop topology and is discussed in detail.

Page 3 of 14 88

Methods Materials The coordinates of the parent and duplicate peptide sequence were taken from the crystal structure of T4L (PDB ID 262 L, chain B), [18] with the ends protected by acetyl (−CH3CO) group at the N-terminus and N-methyl (−NHCH3) group at the C-terminus. Thus the sequence of parent peptide simulated was Ac1NAAKSELDKAINMe13 and that of the duplicate peptide was Ac1NAAKSELDKAINAAKSELDKAINMe24 (Fig. 1). For both the peptides helical conformation adopted in the crystal structure of T4L was used as the initial conformation for all the simulations. Pair of simulations — in pure water and 80 % (v/v) TFE/water solution — were performed for each of the parent and the duplicate peptides. Both the peptides were placed in cubic boxes of edge length 4.97 and 5.48 nm for the parent and the duplicate peptide, respectively. Periodic boundary conditions were used. A SPC model was used for the water molecules [24] and for the TFE simulations a model optimized by Fioroni et al. [25] was used. Molecular dynamics parameters MD simulations were performed using GROMACS package of programs [26] (version 4.0.4) with united atom GROMOS 53a6 force field [27] on Intel Xeon quad core processor based machines running CentOS 4.3 (www.centos.org). The electrostatic interactions were treated by the Particle mesh Ewald (PME) method [28, 29], with a Coulomb cut off of 1.1 nm, fourier spacing of 0.12 nm and an interpolation order of 4. The van der Waals interactions were treated using the Lennard-Jones potential and a switching function with a cutoff distance of 1 nm and a switching distance of 0.9 nm. The system was energy minimized in two steps — first using the steepest descent algorithm and then using conjugate gradient algorithm, until the maximum force on the atoms is smaller than 100 kJ mol−1 nm−1 and convergence was obtained. After energy minimization position restrained MD (keeping the peptide atoms restrained to a fixed position) were carried out for all four systems at 300 K each. The position restrained MD was performed in two steps, first the system was equilibrated under NVT conditions for 200 ps, followed by another 200 ps equilibration under NPT conditions. This distributes the solvent molecules homogeneously around the protein removing any voids present. Initial velocities required to start the simulations were drawn from Maxwell velocity distribution at 300 K. Final productive MD was then performed with an integration time step of 2 fs. To constrain the bonds LINCS algorithm [30] was used. Coordinates were saved after every 0.5 ps and velocities were saved after every 20 ps. Temperature and pressure coupling were done considering protein, water and TFE as separate groups, respectively,

88

J Mol Model (2015) 21:88

Page 4 of 14

using Berendsen method [31] for each, with a reference temperature of 300 K and time constant (tau_t) of 0.1 ps and reference pressure of 1 bar and time constant (tau_p) of 1 ps respectively.

peak at 0.4 nm in the distance (between side chain amine nitrogen of lysine and side chain carbonyl carbon of aspartic and glutamic acid residues) distribution plots was considered [35]. All figures were generated using Matlab, VMD [36] and Xmgrace (http://plasma-gate.weizmann.ac.il/Grace).

Analysis of molecular dynamics trajectories Analyses of the obtained trajectories were carried out using analysis tools provided by GROMACS package. For clustering the conformations sampled in all the simulations, an algorithm described by Daura et al. [32] was used, as implemented in GROMACS. To ensure adequate sampling of the conformational space we have plotted the fractional cluster frequency for the top two clusters, for each simulation, with respect to the simulation time (Fig. S1). We found that both the parent and duplicate peptide reached conformational equilibrium in all the cases after ~600 ns of the simulation time. To ensure equilibration of the 80 % (v/v) TFE/water mixture, we analyzed the density of the solution and also the partial density of TFE and water in the mixed solvent (see Fig. S2for detailed calculations and density plots). It was observed that the TFE/ water mixture displayed a density of ~1277 kg/m3, which is close to the calculated value of 1260 kg/m3. A similar corroboration was obtained for observed (during MD) and calculated partial density values of TFE and water, respectively. Thus the TFE/water mixture used in our simulations showed equilibration. For secondary structure assignment, DSSP program [33] was used as implemented in GROMACS. For formation of a hydrogen bond, a cut-off distance of 0.26 nm, which is the sum of the van der Waals radii of hydrogen (0.12 nm) and acceptor (oxygen=0.14 nm), and an acceptor-donor-hydrogen angle of 30° was considered. For the hydrogen bond formation analysis in the observed triple stranded β-sheet conformation in the duplicate peptide in pure water, the hydrogen bonds persisting for more than 10 % of the complete simulation time (600 ns) have been considered as significant. Calculation of radius of gyration (Rg) of a peptide was done for the backbone atoms. While hydrophobic radius of gyration (Rgh) was calculated by considering the side chain atoms of the hydrophobic residues present in the peptide. For formation of a four residue β-turn, the distance between Cα atoms of its first and fourth residues is considered to be less than 0.7 nm [34]. Further, turns can be both closed (with a 4→1 hydrogen bond) and open (without a 4→1 hydrogen bond). For the peptide-water interaction analysis, number of contacts between water (oxygen) and peptide backbone were calculated in the first hydration shell (0.4 nm, see Fig. 7 and Fig. S3) around main chain of each residue. Side chain-water interactions were also investigated by calculating radial distribution functions (RDF) of water oxygen and TFE fluorine atoms around each peptide side chain, by using the g_rdf program incorporated in GROMACS. For formation of salt bridges, a

Results α-helical conformation persists in duplicate peptide after partial unfolding, in TFE/water, while it is lost in the case of parent peptide The variation of secondary structure of the two peptides in pure water and TFE/water solution, as a function of simulation time, was monitored by the DSSP program. The duplicate peptide maintained significant α-helical propensity in TFE/ water solution for the entire 600 ns of simulation, after partial unfolding at its C-terminal (Fig. 2a). Whereas the α-helical conformation in the parent peptide was lost completely after ~150 ns (Fig. 2c). This indicates that the α-helical conformation of the duplicate peptide is much better stabilized, in TFE/ water solution, in comparison to that for the parent peptide. However, in comparison to the DSSP plots of the two peptides in pure water (Fig. 2b and d), relatively greater sampling of α-helical conformation was observed for both the peptides in TFE/water. In pure water, the duplicate peptide showed an α→β transition at ~70 ns and adopted a triple stranded βhairpin conformation after ~140 ns. In the case of parent sequence (in pure water) the peptide lost its initial α-helical conformation in the early stages (~25 ns) of simulation and adopted a loose π-helix conformation initially which then uncoiled to form bends and turns in the later stages of the simulation. Therefore the α-helix stabilizing effect of TFE seems to be more pronounced in case of the duplicate peptide. Pure water solvates duplicate peptide backbone more than TFE/water We calculated the total number of peptide-solvent hydrogen bonds formed in pure water and TFE/water solution. Note that by peptide-solvent hydrogen bonds, here, we mean total number of hydrogen bonds formed by TFE + water with the peptide, in the case of TFE/water. It was observed that the average number of main chain-solvent hydrogen bonds formed by the parent peptide was almost the same in pure water (16±4) and TFE/water (15±3, Fig. 3a). Whereas, for the duplicate peptide average number of main chain-solvent hydrogen bonds formed is greater in pure water (25±4) than that in TFE/ water (14±4, Fig. 3b), indicating greater solvation of the peptide backbone in pure water. The number of side chain-solvent hydrogen bonds formed by parent and duplicate peptide, in the two solvent systems,

J Mol Model (2015) 21:88

Page 5 of 14 88

Fig. 2 Variation of secondary structure as a function of time shown by DSSP plots for duplicate (a and b) and parent peptide (c and d) in TFE/ water solution and pure water, respectively. Better retention of α-helical conformation was observed in TFE solution for both the peptides. In pure water, duplicate peptide undergoes an α→β transition at ~70 ns and

obtains a triple stranded β-hairpin conformation while parent peptide samples π-helix/turn/bend conformations. Conformations corresponding to different snapshots are also shown for each simulation. N-terminal of peptides is marked by letter ‘N’

i.e., pure water and TFE/water solution, were found to be almost the same in the respective cases (Fig. 3c and d). It was seen that the number of side chain-solvent hydrogen bonds formed by the duplicate peptide, in both pure water and TFE/water, is approximately double of that formed by the parent peptide. This may be attributed to the duplication of the number of donors and acceptors in the side chain of the duplicate peptide. Similarly the number of main chain-solvent hydrogen bonds formed by duplicate peptide in pure water is also almost double that formed by the parent peptide. However, the number of main chain-solvent hydrogen bonds formed by the duplicate peptide in TFE/water was seen to be either lesser or equal to that formed by the parent peptide.

dynamics trajectories) in pure water and TFE/water, for all residues individually (Fig. 4). It was observed that, for both the peptides, number of water contacts decreased in TFE/ water in comparison to pure water suggesting exclusion of water molecules from the vicinity of peptide backbone. Comparison of number of water contacts in parent and duplicate peptide reveals that the duplicate peptide displayed relatively fewer peptide-water contacts in TFE/water. It may be seen that for the duplicate peptide, region of residues involved in αhelical conformation (Ala4 to Ala15) in TFE/water showed minimum number of water contacts. Thus it may be said that exposure of peptide backbone to water seems to destabilize its α-helical conformation, particularly in duplicate peptide.

Water was excluded from the neighborhood of the peptide backbone in TFE/water

TFE shields peptide chain from solvent water in TFE/water

In order to further investigate the possible role of water in destabilizing α-helical conformations in parent and duplicate peptide we calculated the average number of water (oxygen) contacts with peptide backbone (see Analysis of molecular

In order to probe the preferential solvation of peptide chain by TFE in TFE/water solution, we calculated the number of hydrogen bonds formed by water and TFE individually with the peptide main chain and side chain, in TFE/water simulations. It was observed that TFE formed a greater number of

88

Page 6 of 14

J Mol Model (2015) 21:88

Fig. 3 Time variation of total number of hydrogen bonds formed by main chain (a and b) and side chain (c and d) of parent and duplicate peptide with solvent, respectively, in pure water and TFE/water (TFE + water) indicates greater solvation of duplicate peptide backbone in pure water

Fig. 5 Time variation of number of hydrogen bonds formed by main chain (a and b) and side chain (c and d) of parent and duplicate peptide with water and TFE individually in TFE/water indicated a greater number of peptide-TFE hydrogen bond formation suggesting exclusion of solvent water from the vicinity of peptide chain

hydrogen bonds than water with both parent and duplicate peptide (Fig. 5). It may be noted that the parent peptide forms relatively fewer main chain-TFE hydrogen bonds (~5±2) in the initial stages (up to ~60 ns, corresponding to α-helical conformation adopted by the parent peptide, see Fig. 2c), while the number increases in later stages (~10±2). Similarly an increase in number of main chain-TFE hydrogen bonds was observed in the case of the duplicate peptide, after ~450 ns (corresponding to a more random conformation adopted by the C-terminal region of the duplicate peptide, see Fig. 2a). Thus it seems that TFE shields the peptide chain

from solvent water in TFE/water solution and hence promotes α-helical conformation. Figure 6 shows representative structures of parent and duplicate peptide depicting relative shielding of peptide chain from water by TFE in TFE/water solution. Central structure of the two most populated clusters, obtained by cluster analysis (Analysis of molecular dynamics trajectories) of TFE/water trajectory (600 ns), for each peptide are shown.

Fig. 4 Average number of contacts between water (oxygen) and peptide backbone for (a) parent and (b) duplicate peptide, in pure water and TFE/ water, shows relatively greater exclusion of water from the vicinity of peptide backbone in TFE/water

TFE preferentially solvates hydrophobic side chains over polar side chains To further investigate the nature of interactions between solvent molecules and the peptide side chains we calculated the radial distribution function (RDF) of TFE (fluorine) and water (oxygen) around the side chains of two peptides in both pure water and TFE/water solution. Similar trends were observed for the distribution of solvent molecules around the two peptides as shown in Fig. 7. The interaction of different types of residue side chains (i.e., polar/hydrophobic) was seen to vary for water, in pure water and TFE/water, and TFE. This has been explained as follows. For the hydrophobic side chains (comprising of Ala, Leu, and Ile), TFE displayed a sharp and strong (high) peak at ~0.4 nm, particularly for Ala side chains in the two peptides (Fig. 7a and Fig. S3). Sharp peaks were also observed for water molecules around Ala side chains, in both pure water and TFE/water, though the height of the peak

J Mol Model (2015) 21:88

Fig. 6 Schematic illustrating the shielding of peptide chain from solvent water by TFE in (a) parent and (b) duplicate peptide in TFE/water, by representative structures (central structure of two most populated clusters) obtained from clustering complete peptide trajectories. Left and right panels display structures with and without TFE, respectively, for the same cluster. TFE is shown in wireframe surf representation, water in wireframe representation and the polar and hydrophobic side chains of the two peptides are shown in licorice and surf representations, respectively. Dotted circles represent TFE shielded salt bridges, discussed later in Non-native salt bridges were observed in duplicate helical sequence in TFE/water

was relatively smaller in comparison to TFE. It is suspected that the small sized Ala side chain possesses marginal hydrophobicity and hence fails to bury itself away from solvent water. However, in the duplicate peptide the peak for water molecules in TFE/water solution was completely lost for the Ala residues present at the middle of the peptide chain (Ala11, Ala14, and Ala15), unlike the terminal Ala residues (Ala3, Ala4, and Ala15, see Fig. S3). Since the middle region is involved in formation of α-helix in the duplicate peptide in TFE/water it is suspected that TFE shields the side chains of this region from water. Relatively broader peaks were observed for TFE around Ile and Leu side chains owing to their bigger size in comparison to Ala (Fig. 7b and c). Water molecules also displayed broad peaks around Ile and Leu side chains with a diminished height in pure water with respect to the TFE peak in TFE/water. The height of the water peak

Page 7 of 14 88

Fig. 7 Radial distribution function (RDF) plots of TFE (fluorine) and water (oxygen) molecules around hydrophobic (a to c), positively charged (d), negatively charged (e to f), and polar uncharged (g to h) side chains of duplicate peptide indicates that TFE preferentially interacts with hydrophobic side chains. TFE also seems to interact with hydrophobic tail of Lys residues. Only one example of each type of side chain has been shown

around these residues was seen to be even smaller in TFE/ water solution. For the polar side chains, the RDF plots may be described for three groups of residues: negatively charged (Glu and Asp), positively charged (Lys), and uncharged polar residues (Asn and Ser). In the case of positively charged Lys side chain broader peak was observed for both TFE (in TFE/water) and pure water with almost comparable heights (Fig. 7d and Fig. S3). This may be observed due to its amphipathic nature, due to a polar head group and long hydrophobic tail, of the Lys side chain that may interact with water and TFE, respectively. For the negatively charged side chains, TFE and water in TFE/water displayed an insignificant population around them, as indicated by the small peak at 0.4 nm in the RDF

88

Page 8 of 14

plots (Fig. 7e and f). In contrast to this, radial distribution of pure water around negatively charged side chains showed a sharp peak. Sharp and strong (narrow) peaks were observed for water molecules, in both pure water and TFE/water, around polar uncharged side chains of Asn and Ser in contrast to the broader peaks observed for TFE (Fig. 7g and h). Thus, TFE preferentially solvates hydrophobic side chains.

Non-native salt bridges were observed in duplicate helical sequence in TFE/water We also analyzed the salt bridge formation between side chains of positively charged lysine residues and negatively charged residues, glutamate and aspartate, in the parent and duplicate peptide sequences. We found that the duplicate peptide showed relatively greater sampling of four salt bridges — K5:D9, K10:E7, K16:D20 and K21:E18 — in TFE/water in comparison to that in pure water. This has been indicated by the peak observed at 0.4 nm (see Analysis of molecular dynamics trajectories) in the corresponding salt bridge distance distribution plots (Fig. 8c to f and Fig. S4). Except for K48:E45 salt bridge, corresponding to K21:E18 in the duplicate peptide and K10:E7 in the parent peptide, all other salt bridges were seen to be absent in the native duplicate helical sequence in the L20 crystal structure. In the case of parent peptide, two salt bridges — K5:D9 and K10:E7 — displayed significant sampling in TFE/water (Fig. 8a to b). Thus TFE seems to stabilize salt bridge formation in the two peptide sequences. However, relatively smaller sampling was Fig. 8 Frequency distribution of distance between salt bridge forming residues in parent (a to b) and duplicate peptide (c to f) in pure water and TFE/water indicates greater sampling of salt bridges in TFE/water in comparison to pure water. Right panel displays representative structures indicating salt bridges in the parent and the duplicate peptide, marked as a to b for the parent and c to f for the duplicate, corresponding to the salt bridge distribution as in left panels. Central structure of the top most populated cluster, obtained from clustering complete peptide trajectories, for each peptide is shown. N-terminal of peptides is marked by letter ‘N’

J Mol Model (2015) 21:88

observed for the salt bridge interactions in the parent peptide than the duplicate peptide, as indicated by a smaller frequency of the peak at 0.4 nm in the salt bridge distance distribution plots for parent peptide.

Mechanism of α→β transition in duplicate peptide in pure water To investigate the role of hydrophobic association in β-sheet formation in the duplicate peptide in pure water, comparison of its backbone radius of gyration (Rg), side chain hydrophobic radius gyration (Rgh), and hydrophobic solvent accessible surface area (SASA) were monitored. The three quantities showed concerted variation in their values (Fig. 9a to c). Starting from the initial α-helical structure, with average values of 1 nm for both Rg and Rgh and 15 nm2 for SASA (structure (i) in Fig. 9), a rise in the Rg, Rgh, and SASA values (1.7, 1.8, and 17 nm2, respectively, structure (iii) in Fig. 9) was observed upon complete uncoiling of the peptide. The hydrophobic side chains of the duplicate peptide then buckled together to give rise to a number of collapsed conformations (structure (iv) to (vi) in Fig. 9). A collapsed conformation, displaying R g , R gh , and SASA values of 0.8,0.8, and 13 nm2, respectively (structure (vi) in Fig. 9) served as a nucleus for formation of the β-sheet structure. Apart from the hydrophobic association, formation of few (2±1) hydrogen bonds was also observed in the β-sheet nucleating collapsed structure (Fig. 9d and structure (iv) to (vi) in Fig. 9). However, the number of hydrogen bonds increased (to 8±2) in the fully

J Mol Model (2015) 21:88

Page 9 of 14 88

Lys21, respectively, folded by formation of K5NH:E18O, E18NH:K3O and D20NH:K5O hydrogen bonds (Fig. 10a and d). Then the hydrogen bonds — A14NH:S17O and S17NH:A14O — formed at ~75 ns with a simultaneous decrease in A14C α -S17C α distance to a value < 0. 7 nm, resulting in formation of turn 1 (Fig. 10b, d, and e, also see Analysis of molecular dynamics trajectories for the criterion for turn formation). Subsequently formation of third antiparallel β-strand, i.e., strand B, involving residues Ala11 to Ala14, was seen (at ~170 ns) by formation of E7NH:D20O, I12NH:L19O, L19NH:I12O, and A22NH:E7O (Fig. 10c and f). Along with the formation of strand B, an elongation of strand C by one more residue (Ser17) was also observed. Simultaneously E7Cα-K10Cα distance also decreased (to a value

Environmental polarity induces conformational transitions in a helical peptide sequence from bacteriophage T4 lysozyme and its tandem duplicate: a molecular dynamics simulation study.

Our recent molecular dynamics (MD) simulation of an insertion/duplication mutant 'L20' of bacteriophage T4 lysozyme demonstrated a solvent induced α→β...
4MB Sizes 2 Downloads 4 Views