Influence of Backbone Fluorine Substitution upon the Folding

Jun 1, 2009 - (1) Mathad, R. I.; Gessier, F.; Seebach, D.; Jaun, B. The effect of backbone-heteroatom substitution on the folding of peptides- a singl...
1 downloads 0 Views 2MB Size
J. Phys. Chem. B 2009, 113, 8695–8703

8695

Influence of Backbone Fluorine Substitution upon the Folding Equilibrium of a β-Heptapeptide Zrinka Gattin† and Wilfred F. van Gunsteren*,‡ Laboratory of Physical Chemistry, Swiss Federal Institute of Technology, ETH, 8093 Zu¨rich, Switzerland, and Laboratory of Organic Chemistry, Swiss Federal Institute of Technology, ETH, CH-8093 Zu¨rich, Switzerland ReceiVed: December 16, 2008; ReVised Manuscript ReceiVed: April 21, 2009

Explicit solvent molecular dynamics (MD) simulations of three β-heptapeptides with a central β- HAla(R-F) amino acid (Figure 1) in methanol are reported. They aim at an analysis of the conformational consequences of CR carbon atom bound fluoro atoms, and the particular configuration of the central fluoro-β-amino acid: peptide 3 with an S configuration of the CR bound fluor atom, peptide 4 with an R configuration of the CR bound fluor atom, and peptide 5 with a difluoro substitution at the CR atom of residue 4. The NMR and CD spectra of these three β-peptides were earlier (Mathad et al. HelV. Chim. Acta 2005, 88, 266-280) interpreted to indicate a decrease in propensity of 314-helical structure from peptide 4 to peptide 5 to peptide 3. This result was at odds with previous experimental data for β-heptapeptides with a central β-HAla(R-Me) amino acid which showed that the β-heptapeptide with the S,S configuration of the central β-HAla(R-Me) was the most 314-helical, whereas the S,R configuration did not lead to any detected helicity. The reported MD simulations resolve this paradox. The MD trajectories of all three peptides do agree with the primary, measured data: NMR nuclear Overhauser effect (NOE) atom-atom distance bounds and 3J-coupling constants. A conformational analysis of the MD trajectory conformations shows, however, a decrease in 314-helical character from peptide 3 to peptide 5 to peptide 4, which is in line with the results for the nonfluorinated peptides. It is shown that interpretation of NMR NOE data using single-structure refinement in vacuo based on local (along the sequence) and limited atom-atom distance data as in ref 1 (Mathad et al. HelV. Chim. Acta 2005, 88, 266-280) may lead to molecular structures that are not representative for the ensemble of molecular conformations. Introduction β-Peptides, the structural analogues of naturally occurring R-peptides, are known to be useful test cases for the investigation of the influence of molecular composition on the folding properties of polypeptides.2,3 Their resistance to proteases (in ViVo and in Vitro)4 and ability to permeate cell membranes5 assign β-peptides a promising role in pharmaceutical applications as antibacterial or antimicrobial agents,6 inhibitors of fat and cholesterol absorption,7 or somatostatin-mimicking agents.8 The specific biological activity of β-peptides depends on the formation of well-defined secondary and tertiary structures, which requires detailed knowledge about the relationship between the β-amino acid sequence and the prevailing threedimensional structure of the peptide. The additional carbon atom in the backbone of β-amino acid residues offers additional substitution patterns regarding the backbone carbon atom sidechain attachment site (CR and/or Cβ) and configurations of the backbone chiral centers.9,10 Experimental and molecular dynamics (MD) studies show that differences in composition, site, and orientation of the side-chain11,12 and backbone atom replacements, as in aminoxyor depsipeptides,13,14 have a direct influence on the dominant secondary structure of the investigated peptide. * Corresponding author. E-mail: [email protected], [email protected]. Phone: 0041-44-632-5502. Fax: 0041-44632-1039. † Laboratory of Physical Chemistry. ‡ Laboratory of Organic Chemistry.

Less is known about the influence of backbone heteroatom (OH, NH2, F, Cl, S, etc.) substitutions on the conformational ensemble of a polypeptide.15,16 This is partly due to the conformational instability of such compounds. However, some β-peptides with backbone bound heteroatoms are natural compounds, such as 2,3-diamino butanoic acid,17 a key structural moiety of some peptidic antibiotics like aspartocin18 and aminodeoxybestatin.19 Previous studies by Seebach et al.20 and Daura et al.21 have shown an increase in the propensity of 314-helical structure for β-peptides containing a central β-alanine-R-methyl amino acid (β-HAla(R-Me)). This 314-helical fold is known as the predominant stable fold of β-peptides. We notice that these studies have shown that only the S,S configuration yields a stable 314-helical fold,20 whereas the S,R configuration does not result in any 314helicity. Three β-heptapeptides containing a fluoro analogue of the central β-HAla(R-Me) residue were synthesized and investigated with CD22 and NMR techniques.1 These experiments indicated different propensities of 314-helical structure for the three peptides 3, 4, and 5 of Figure 1: peptide 4 (S,R (fluoro) configuration) was thought to show the most 314-helical structure, peptide 5 (difluoro substitution on CR) less 314-helical structure, and peptide 3 (S,S (fluoro) configuration) the least 314-helical structure. This experimental suggestion that the S,R configuration of β-HAla(R-F) induces more helicity than the S,S configuration is contrary to the earlier mentioned observations for β-HAla(RMe). Even more puzzling is that the S,R configuration makes a

10.1021/jp811106e CCC: $40.75  2009 American Chemical Society Published on Web 06/01/2009

8696

J. Phys. Chem. B, Vol. 113, No. 25, 2009

Gattin and van Gunsteren

Figure 1. Chemical formulas of the three β-heptapeptides studied (peptide 3, H-β-HVal-β-HAla-β-HLeu-(S,S)-β-HAla(R-F)-β-HVal-β-HAla-βHLeu-OH; peptide 4, H-β-HVal-β-HAla-β-HLeu-(S,R)-β-HAla(R-F)-β-HVal-β-HAla-β-HLeu-OH; peptide 5, H-β-HVal-β-HAla-β-HLeu-(S)-βHAla(di-R-F)-β-HVal-β-HAla-β-HLeu-OH). In the simulations, both end groups were protonated in line with experiment. The force-field parameters for the atoms of the β-HAla(R-F) amino acid residue are shown in Table 1.

TABLE 1: Force-Field Parameters for the r-Fluorinated β-Alanine Residuea atom

description

IAC

partial charge (e)

CA F, F1, F2

carbonyl C fluor F

11 16

0.2 -0.2

bond length

b0 (nm)

Kb (106 kJ mol-1 nm-4)

CA-F

0.136

4184

angle

θ0 (deg)

Kθ (kJ mol-1)

CB-CA-F C-CA-F F1-CA-F2

111.4 111.4 107.6

460.2 460.2 460.2

a The atom names are defined in Figure 1; parameters (bond lengths, angles, and dihedrals) of carbonyl, amide, and methyl groups (not given) were chosen in analogy to GROMOS96 45A3 force-field parameters. The GROMOS96 integer atom code (IAC) which describes the Lennard-Jones parameters of the corresponding atoms for fluor atom were taken from Fioroni et al.29

314-helical fold conformationally unfavorable due to steric clashes. The molecular model structure for peptide 3 ((S,S)βHAla(R-F)) derived from the NMR data1 shows no 314-helical structure, due to the 180° value of the dihedral angle (FsCRsCdO), a value expected from theory for R-fluoroamide compounds.23 The preference for this dihedral angle value is incompatible with 314-helical structure. Nevertheless, last year, Mathad et al.24 observed for a longer β-tridecapeptide which had a central (S,S)β-HAla(R-F) configuration 314-helical structure but concomitant with a dihedral angle (FsCRsCdO) value of -90° (the synclinal orientation). In ref 1, the structural refinements for peptides 3 and 4 were done without long-range nuclear Overhauser effects (NOEs) because these were not detected in the NMR experiment. This makes the derived model structures ambiguous: they are just one out of many possible structures that are compatible with the short-range NOE data. In this paper, we report MD simulation studies on the conformational behavior of the three mentioned centrally β-HAla, R-fluoro-substituted β-heptapeptides; see Figure 1. All three peptides were simulated in methanol solution at an elevated temperature of 340 K to enhance the sampling of the

conformational space and at a pressure of 1 atm using the GROMOS biomolecular simulation software25,26 and the GROMOS 45A3 force field.25,27 The ensembles of trajectory structures were analyzed in terms of conformational space sampled by the peptide, structural properties such as hydrogen bonding and in terms of the level of agreement with the available experimental NMR data, NOE (nuclear Overhauser effect) atom-atom distance bounds and 3J-coupling constants.1 Methods Molecular Model. The simulations of the three β-heptapeptides (Figure 1) were carried out in explicit solvent methanol using the GROMOS96 biomolecular simulation software25,26 and 45A3 force-field parameter set.25,27 The methanol solvent molecules were represented using a rigid three-site model belonging to the standard GROMOS96 set of solvents.25 Aliphatic CHn groups were treated as united atoms, both in the solute and solvent. The β-heptapeptides were protonated at the C- and N-termini, yielding a positive charge of +1 e. No counterions were used. The β-HAla(R-F) molecular-topology building block was constructed in analogy to the β-HAla(R-Me) building block. The parameters are those of the GROMOS 45A3 force field,27,28 with the exception of the repulsive van der Waals parameter of 1/2 the fluor atoms: C12 (F) ) 1.081 × 10-3 (kJ mol-1 nm12)1/2.29 Several simulations were performed for peptide 4 with different charges on the central CR and F atoms (data not shown). We did not observe significant charge induced effects on the conformational ensemble which would justify the use of a charge larger than -0.2 e on the F atoms, which was the value used for 1,1,1,3,3,3-hexafluoro-propan-2-ol (HFIP).29 In addition, we performed an ab initio conformational study with the Gaussian 0330 program using the DFT/6-311+G** method on the NH2-CβH(Me)-CRF2-CONHMe fragment. The torsional angle energy profiles (-CR-C- angle and -Cβ-CR- angle) obtained using the GROMOS force field were in good agreement with the QM profiles; see Figures S1 and S2 of the Supporting Information. Simulation Setup. The NMR model structures were taken as starting configurations for all three peptides. These were placed at the center of a rectangular periodic box enforcing a

Folding Equilibrium of a β-Heptapeptide

J. Phys. Chem. B, Vol. 113, No. 25, 2009 8697

minimum distance of 1.4 nm between any peptide atom and the closest box wall. This box was then filled with methanol solvent molecules such that the distance between (non-hydrogen) solvent and solute atoms was bigger than 0.23 nm. The resulting numbers of solvent molecules were 1353, 1073, and 1046 for peptides 3, 4, and 5, respectively. To relax the systems prior to the MD simulations, a steepest-descent energy minimization was performed. After an initial equilibration period of 140 ps, the simulations were continued for 100 ns. The analysis was based on configurations saved every 0.5 ps. The simulations of the three β-heptapeptides were performed at 340 K. For this purpose, solute and solvent temperatures were maintained independently by weak coupling to two temperature baths with relaxation times of 0.1 ps.31 The pressure was calculated using a molecular virial and maintained by weak coupling to a pressure bath (isotropic coordinate scaling) with a relaxation time of 0.5 ps, using an isothermal compressibility of 4.575 × 10-4 (kJ mol-1 nm-3)-1. Bond lengths and the bond angle in methanol were constrained by application of the SHAKE algorithm32 with a relative geometric tolerance of 10-4. The equations of motion were integrated using the leapfrog algorithm based on a time step of 2 fs. Long-range interactions were handled using a triplerange cutoff scheme25,26 with cutoff radii of 0.8 nm (interactions updated every time step) and 1.4 nm (interactions updated every five time steps). The mean effect of omitted electrostatic interactions beyond the long-range cutoff distance (1.4 nm) was accounted for by the inclusion of a Barker-Watts reactionfield force33,34 based on a dielectric permittivity of 17.7 for methanol.35 Analysis. The atom-positional root-mean-square deviations (rmsd’s) from the energy-minimized NMR model structure for peptides 3, 4, and 5 were evaluated on the basis of all backbone atoms (N, CR, Cβ, and C) of residues 2-6. These atoms were also used in the translational superpositioning and rotational fitting of the trajectory structures. A conformational clustering analysis, performed as described in previous studies,36 was carried out on the trajectory structures of the β-heptapeptide using 2 × 105 trajectory structures. A rmsd similarity criterion of 0.1 nm for atoms N, C, CR, and Cβ of residues 2-6 of the β-heptapeptide was used, as previously described in refs 36 and 37. For all systems, the hydrogen-bond analysis uses as criterion for defining a hydrogen bond a maximum hydrogen acceptor distance of 0.25 nm and a minimum donor atom-hydrogen acceptor angle of 135°. Interproton distance bounds derived from the NOE cross-peak intensities at 298 K for the β-heptapeptides were compared to the corresponding average effective interproton distances in the simulations, 〈r-6〉-1/6, as appropriate for small rapidly tumbling molecules.38,39 The interproton distances involving aliphatic hydrogen atoms were calculated by defining virtual (CH1), prochiral (stereospecific CH2), and pseudo (CH3 and nonstereospecific CH2) atomic positions during the analysis.25,26 3J-coupling constants, 3J(HN-Hβ), were calculated for the simulated and model structures using the Karplus relation40

J(HN, Hβ) ) A cos2 θ + B cos θ + C

3

(1)

where θ is the dihedral angle between the planes defined by the atoms (H, N, Cβ) and the atoms (N, Cβ, Hβ). The parameters A, B, and C for the calculation of 3J(HN-Hβ) were 6.4, -1.4, and 1.9 Hz, respectively.41 Results For each of the three β-heptapeptides (Figure 1), a 100 ns MD simulation at 340 K in methanol was performed with the

Figure 2. Black line: backbone atom-positional root-mean-square difference, rmsd (backbone atoms N, C, CR, and Cβ of residues 2-6 only), of simulated molecules with respect to the energy-minimized NMR model structure with lowest energy as a function of time for peptide 3 (upper panel), peptide 4 (middle panel), and peptide 5 (bottom panel). Black bars (hbond): occurrence of NH(i)-O(i+2) hydrogen bonds (residue sequence number i ) 1-5, from top) characteristic of a 314-helix.

aim of studying the effect of the presence of a central β-HAla(RF) amino acid on the folding behavior of β-peptides. Peptide 3, (S,S)β-HAla(r-F). The upper panel of Figure 2 shows the backbone atom-positional root-mean-square deviation (rmsd) (backbone atoms N, C, CR, and Cβ of residues 2-6 only) of the trajectory structures of the MD simulation of peptide 3 in methanol from the NMR model structure as a function of time along with the occurrence of the most dominant intramolecular hydrogen bonds. The hydrogen bonds observed in this simulation correspond to 14-membered rings characteristic of 314-helical structure. Peptide 3 is most of the simulation time (49%) in a 314-helical conformation. The upper left and right panels of Figure 3 and Tables S1 and S4 in the Supporting Information show the proton-proton NOE distance bound violations and HN-Hβ 3J-coupling constants calculated from the trajectory structures for peptide 3 (black bars) in comparison with NOE distance bound violations and 3J-values calculated from the NMR bundle of 10 model structures (gray bars), respectively. A total of 33 NOE distance bounds and three 3J-coupling values were available.1 The MD simulation does not violate significantly any of the experimentally derived NOE distance bounds. Interestingly, the MD ensemble agrees better with the measured experimental data than the set of 10 NMR model structures derived on the basis of these experimental data. The average 3J-coupling constants calculated for the trajectory structures using the Karplus relation40 agree very well with the experimentally measured

8698

J. Phys. Chem. B, Vol. 113, No. 25, 2009

Gattin and van Gunsteren

Figure 3. Averaged (r-6-averaging) distances minus the NMR NOE upper bound distances calculated over all recorded structures (2 × 105) in the MD simulation at 340 K (black) and for the 10 model structures of the NMR bundle (gray),1 for peptide 3 (upper left panel), peptide 4 (middle left panel), and peptide 5 (bottom left panel). A number of 33 and 28 upper bound distances were derived from experimental NOE intensities observed in the ROESY NMR spectra at 298 K for peptides 3 and 4 and for peptide 5, respectively.1 Three 3J-coupling constants (HN-HCβ) for peptide 3 (upper right panel) and six for peptide 4 (middle right panel) and peptide 5 (bottom right panel), averaged over all structures from the MD simulation at 340 K (black) and averaged over the 10 model structures of each NMR bundle (gray).1 The Karplus equation40 with a ) 6.4 Hz, b ) -1.4 Hz, and c ) 1.9 Hz41 was used. The corresponding values are presented in Tables S1-S6 in the Supporting Information.

Figure 4. Conformational preferences of the β-heptapeptide peptide 3. The five most populated conformers (central structures of the five most populated clusters) with a backbone (residues 2-6) rmsd similarity criterion of 0.1 nm (free clustering analysis) observed in the 100 ns simulation at 340 K are shown. For each cluster, its corresponding population (in %) and description of the type of helicity are indicated. Atom coloring: C, yellow; H, white; N, dark blue; O, red.

values. All deviations are below 1 Hz and generally much lower than those for the set of NMR model structures. During the folding-unfolding process, the peptide adopts different conformations. To identify the most populated ones, a conformational clustering analysis was performed which clusters the trajectory structures according to their atompositional rmsd (backbone atoms of residues 2-6 only) for each pair of structures. The rmsd similarity criterion was set to 0.1 nm. The central member structures of the five most populated clusters (conformers) of peptide 3 are displayed in Figure 4, together with the corresponding population (in %) and a description of the type of helicity. The most populated conformer (49%) of peptide 3 corresponds to a 314-helix. Because the NMR data did not indicate the presence of a 314-helix for peptide 3, we performed an analysis of proton-proton NOE distances and HN-Hβ 3J-coupling constants for the five ensembles of structures of the five most populated clusters. Only two NOE violations

are bigger than 0.1 nm for any of the five investigated clusters. These are, together with the three 3J-values, presented in Figure 5. For comparison, the NOEs and 3J-values obtained by averaging over all trajectory structures are given in black. The complete data of the performed analyses is given in Tables S1 and S4 in the Supporting Information. The NOE analysis presented in Figure 5 shows that the nonlinear r-6 averaging of NOE distances may lead to lowly populated conformers contributing significantly to a particular NOE signal. Or in other words, the most dominant conformer may violate the NOE bounds, while the complete conformational ensemble shows no violations. Mathad et al.1 assumed that the absence of 314-helical structure in their NMR model structures for peptide 3 is due to the conformation of approximately 180° of the FsCRsCdO dihedral angle which prevents 314-helix formation. An analysis of the trajectory structures with respect to the FsCRsCdO dihedral angle conformation confirms this correlation. In Figure 6, the time series (left upper panel) and the normalized distribution (right upper panel) of the FsCRsCdO dihedral angle is presented along with the rmsd with respect to a 314helical conformation (bottom panel) for peptide 3. When the 314-helix is present, the FsCRsCdO dihedral angle has a value around 280° (-80°). However, the MD data for peptide 3 show that an ensemble dominated by a 314-helical conformation is fully compatible with the measured NOE and 3J-coupling constant data. In other words, from the latter, one may not conclude that the 314-helix is not the dominant conformation. Peptide 4, (S,R)β-HAla(r-F). The middle panel of Figure 2 shows the backbone atom-positional rmsd of the trajectory structures of the MD simulation (at 340 K) of peptide 4 in methanol from the 314-helical NMR model structure as a

Folding Equilibrium of a β-Heptapeptide

J. Phys. Chem. B, Vol. 113, No. 25, 2009 8699

Figure 5. Averaged (r-6-averaging) atom-atom distances minus the NMR NOE upper bound distances for NOE 27 (Hβ(3)-HN(4)) and NOE 30 (Hβ(4)-HN(5)) calculated for the MD trajectory structures (black bars) and for the first five most populated clusters (first cluster, red; second cluster, green; third cluster, blue; fourth cluster, cyan; fifth cluster, magenta) for peptide 3. Three 3J-coupling constants (HN-Hβ) averaged over all MD trajectory structures (black circles) and over the structures of each of the first five most populated clusters. The corresponding values are presented in Tables S1 and S4 in the Supporting Information.

Figure 6. Upper left panel: Torsional dihedral angle FsCRsCdO as a function of time for the peptide 3 simulation. Upper right panel: Distribution of the values of the torsional dihedral angle FsCRsCdO. Bottom panel: Backbone atom-positional root-mean-square deviation from the 314helical conformation as described in the caption of Figure 2.

function of time along with the occurrence of the most dominant intramolecular hydrogen bonds. The hydrogen bonds observed in this simulation correspond to 14-membered rings characteristic of a 314-helical structure, and peptide 4 is approximately 12% of the simulation time in a 314-helical conformation. The middle left and right panels of Figure 3 and Tables S2 and S5 in the Supporting Information show the proton-proton NOE distance bound violations and HN-Hβ 3J-coupling constants calculated from the trajectory structures for peptide 4 (black bars) and the NOE distance bound violations and 3Jvalues calculated from the NMR bundle of 10 model structures (gray bars), respectively. A total of 33 NOE distance bounds and six 3J-coupling values were available.1 The MD simulation

does not violate significantly any of the experimentally derived NOE distance bounds. As for peptide 3, the MD ensemble agrees better with the measured experimental data than the set of 10 NMR model structures derived on the basis of these experimental data. The average 3J-coupling constants calculated for the trajectory structures using the Karplus relation40 agree very well with the experimentally measured values. All deviations are below 1 Hz and much lower than those for the set of NMR model structures. The central member structures of the five most populated conformers of peptide 4 are displayed in Figure 7, together with the corresponding population (in %) and a description of the type of helicity. The most populated conformer with only 12%

8700

J. Phys. Chem. B, Vol. 113, No. 25, 2009

Figure 7. Conformational preferences of the β-heptapeptide peptide 4. The five most populated conformers (central structures of the five most populated clusters) with a backbone (residues 2-6) rmsd similarity criterion of 0.1 nm (free clustering analysis) observed in the 100 ns simulation at 340 K are shown. For each cluster, its corresponding population (in %) and description of the type of helicity are indicated. Atom coloring: C, yellow; H, white; N, dark blue; O, red.

corresponds to a 314-helix. On the basis of this NMR data, Mathad et al.1 characterized the dominant conformer of peptide 4 as a 314-helical structure in spite of the absence of inter-residual NOEs between non-neighbor (along the sequence) residues. Although the S,R configuration of β-HAla(R-F) is disfavoring a 314-helical structure because of steric clashes, and is therefore expected to result in an unstructured conformational ensemble, in the MD trajectory, the 314-helical structure was found as the most populated conformation. As for peptide 3, we performed an analysis of proton-proton NOE distances and HN-Hβ 3Jcoupling constants for the five ensembles of structures of the five most populated clusters. Only two NOE violations are bigger than 0.1 nm for any of the five investigated clusters. These are together with the five 3J-values presented in Figure 8. The complete data is given in Tables S2 and S5 in the Supporting Information. The NOE analysis presented in Figure 8 shows, as for peptide 3 in Figure 5, that the nonlinear r-6 averaging of NOE distances may lead to lowly populated conformers contributing significantly to a particular NOE signal. Or in other words, the most populated conformer may violate the NOE bounds, while the complete conformational ensemble shows no violations. As for peptide 3, we performed an analysis of the FsCRsCdO dihedral angle conformation. In Figure 9, the time series (left upper panel) and the normalized distribution (right upper panel) of the FsCRsCdO dihedral angle is presented along with the rmsd with respect to a 314-helical conformation (bottom panel) for peptide 4. The FsCRsCdO dihedral angle distribution for peptide 4 is rather different from and shows more diversity than that for peptide 3 (Figure 6). When the 314helix is present, the FsCRsCdO dihedral angle has a value around 170°. In summary, the MD data for peptide 4 show that an ensemble that is not dominated by any particular (helical or other) conformation is fully compatible with the measured NOE and 3J-coupling constant data. In other words, from the latter, one may not conclude that the 314-helix is the dominant conformer. Peptide 5, (S)β-HAla(r-F2). The bottom panel of Figure 2 shows the backbone atom-positional rmsd of the trajectory structures of the MD simulation (at 340 K) of peptide 5 in methanol from the 314-helical NMR model structure as a function of time along with the occurrence of the most dominant intramolecular hydrogen bonds. The hydrogen bonds observed in this simulation correspond to 14-membered rings characteristic of a 314-helical structure, and peptide 5 is approximately 37% of the simulation time in a 314-helical conformation. The bottom left and right panels of Figure 3 and Tables S3 and S6 in the Supporting Information show the proton-proton NOE distance bound violations and HN-Hβ 3J-coupling con-

Gattin and van Gunsteren stants calculated from the trajectory structures for peptide 5 (black bars) and the NOE distance bound violations and 3Jvalues calculated from the NMR bundle of 10 model structures (gray bars), respectively. A total of 28 NOE distance bounds and six 3J-coupling values were available.1 The MD simulation significantly violates two NOE distances (NOE 3, HN(2)-Hβ(3); NOE 27, HN(5)-Hβ(7)) with 0.103 and 0.121 nm. In contrast to peptides 3 and 4, here the MD ensemble agrees slightly worse with the measured experimental data than the set of 10 NMR structures derived on the basis of these experimental data. The average 3J-coupling constants calculated for the trajectory structures using the Karplus relation40 agree very well with the experimentally measured values. All deviations are below 1 Hz and generally lower than those for the set of NMR model structures. The central member structures of the five most populated conformers of peptide 5 are displayed in Figure 10, together with the corresponding population (in %) and a description of the type of helicity. The most populated conformer with 37% corresponds to a 314-helix. We performed an analysis of the proton-proton NOE distances and HN-Hβ 3J-coupling constants for the five ensembles of structures of the five most populated clusters. Fourteen NOE violations are bigger than 0.1 nm for any of the five investigated clusters and are presented in Figure 11. The complete data is given in Tables S3 and S6 in the Supporting Information. The NOE analysis presented in Figure 11 shows that the most populated cluster, which has 314-helical character, satisfies the NOE bounds better than any other conformational cluster and also better than the ensemble average. This suggests that the measured NOEs are those for a 314-helix. Figure 12 shows the time series (left upper panel) and the normalized distribution (right upper panel) of the FSisCRsCdO dihedral angle along with the rmsd with respect to a 314-helical conformation (bottom panel) for peptide 5. The FSisCRsCdO dihedral angle distribution for peptide 5 is rather different from that for peptide 4 but very similar to that for peptide 3. When the 314-helix is present, the FSisCRsCdO dihedral angle has a value around 280° for peptide 3, whereas the FResCRsCdO dihedral angle is around 160° for peptide 4. In summary, the MD data for peptide 5 suggests that the conformational ensemble has considerable 314-helical character. Discussion We have investigated the effect of one centrally placed β-HAla(R-F) amino acid on the folding-unfolding equilibria of three β-heptapeptides and on their propensity for folding into a 314-helix by performing MD simulations at 340 K starting from an energy-minimized NMR model structure.1 The three investigated β-heptapeptides carry different configurations at CR of the β-HAla(R-F) amino acid residue: peptide 3 (S,S), peptide 4 (S,R), and peptide 5 (S at Cβ, di-F at CR). The MD results are in agreement with primary, i.e., measured, experimental data (NOEs and 3J-values), but for peptides 3 and 4, the conformational interpretation that we can infer from the MD trajectories contradicts the molecular conformations derived1 from the experimental data by single-structure refinement using the program X-PLOR.42 The reason for this contradiction is likely to rest with the absence of long-range NOEs for peptide 3 and for peptide 4. Second, single-structure refinement in Vacuo based on only local and a limited number of atom-atom distance bounds can easily lead to a nonrepresentative structure for the molecule. In contrast, the MD simulations using a thermodynamically equilibrated force field and explicit solvent molecules lead, even without using the experimental data as

Folding Equilibrium of a β-Heptapeptide

J. Phys. Chem. B, Vol. 113, No. 25, 2009 8701

Figure 8. Averaged (r-6-averaging) atom-atom distances minus the NMR NOE upper bound distances for NOE 27 (HN(2)-HN(3)) and NOE 33 (HN(5)-Hβ(4)) calculated for the MD trajectory structures (black bars) and for the first five most populated clusters (first cluster, red; second cluster, green; third cluster, blue; fourth cluster, cyan; fifth cluster, magenta) for peptide 4. Three 3J-coupling constants (HN-Hβ) averaged over all MD trajectory structures (black circles) and over the structures of each of the first five most populated clusters. The corresponding values are presented in Tables S2 and S5 in the Supporting Information.

Figure 9. Upper left panel: Torsional dihedral angle FsCRsCdO as a function of time for the peptide 4 simulation. Upper right panel: Distribution of the values of the torsional dihedral angle FsCRsCdO. Bottom panel: Backbone atom-positional root-mean-square deviation from the 314helical conformation as described in the caption of Figure 2.

restraints, to conformational ensembles that better satisfy these experimentally measured data. Among the structures of the five most populated conformational clusters of the MD trajectories of peptides 3 and 4 (which comprise more than 60% of all trajectory structures), there are structures that satisfy the given NOE upper bounds. For peptide 5, both MD simulation and single-structure refinement in Vacuo using NOE restraints suggest predominantly 314-helical character of the conformational ensemble. In summary, the MD simulation results resolve the paradox between the conformational interpretations of two sets of NMR experimental data1,20 regarding centrally β-H-Ala, R-substituted β-heptapeptides. They show a decrease in 314-helical character from peptide 3 to peptide 5 to peptide 4, in accordance with ref 20 but at variance with ref 1. It was shown that the conformational interpretation presented in ref 1 is not well founded in the measured experimental data, due to the use of singlestructure refinement techniques.

Figure 10. Conformational preferences of the β-heptapeptide peptide 5. The five most populated conformers (central structures of the five most populated clusters) with a backbone (residues 2-6) rmsd similarity criterion of 0.1 nm (free clustering analysis) observed in the 100 ns simulation at 340 K are shown. For each cluster, its corresponding population (in %) and description of the type of helicity are indicated. Atom coloring: C, yellow; H, white; N, dark blue; O, red.

Finally, one might speculate about the reasons behind the observed conformational differences between the differently

8702

J. Phys. Chem. B, Vol. 113, No. 25, 2009

Gattin and van Gunsteren

Figure 11. Averaged (r-6-averaging) atom-atom distances minus the NMR NOE upper bound distances for NOEs 3, 7, 14-17, 19-20, and 23-28 (see Table S3 in the Supporting Information) calculated for the MD trajectory structures (black bars) and for the first five most populated clusters (first cluster, red; second cluster, green; third cluster, blue; fourth cluster, cyan; fifth cluster, magenta) for peptide 5. Three 3J-coupling constants (HN-Hβ) averaged over all MD trajectory structures (black circles) and over the structures of each of the first five most populated clusters. The corresponding values are presented in Tables S3 and S6 in the Supporting Information.

Figure 12. Upper left panel: Torsional dihedral angle FSisCRsCdO as a function of time for the peptide 5 simulation. Upper right panel: Distribution of the values of the torsional dihedral angle FSisCRsCdO. Bottom panel: Backbone atom-positional root-mean-square deviation from the 314helical conformation as described in the caption of Figure 2.

F-substituted peptides. Which are energetic and entropic contributions due to quantum-mechanical stereoelectronic effects, due to steric atom-atom clashes, hydrogen-bonding, or solvent accessibility?43 Since such effects may mutually influence each other, they cannot be inferred from one simulation. One would have to systematically vary each factor and perform simulations. Regarding the time required to establish folding equilibria for short peptides, such an investigation is currently beyond our capacity.

Supporting Information Available: Tables with averaged (r-6-averaging) atom-atom distances minus the NMR NOE upper bound distances and 3J-coupling constants (HN-Hβ, in Hz) for peptides 3, 4, and 5 which are the corresponding data for Figures 3, 5, 8, and 11. Torsional angle energy profiles of a model molecule NH2-CβH(Me)-CRF2-CONHMe in vacuo calculated with DFT/6-311+G** and GROMOS96 using the 45A3 force-field parameter set. This material is available free of charge via the Internet at http://pubs.acs.org.

Acknowledgment. We thank Dr. Bernhard Jaun and Dr. Raveendra Mathad for providing us with the NOE data and for fruitful discussions and Halvor S. Hansen for help with QM calculations. Financial support was obtained from the National Centre of Competence in Research (NCCR) Structural Biology of the Swiss National Science Foundation, which is gratefully acknowledged.

References and Notes (1) Mathad, R. I.; Gessier, F.; Seebach, D.; Jaun, B. The effect of backbone-heteroatom substitution on the folding of peptides- a single fluorine substituent prevents a β-heptapeptide from folding into a 314-helix (NMR analysis. HelV. Chim. Acta 2005, 88, 266–280. (2) Goodman, C. M.; Choi, S.; Shandler, S.; DeGrado, W. F. Foldamers as versatile frameworks for the design and evolution of function. Nat. Chem. Biol. 2007, 3 (5), 252–262.

Folding Equilibrium of a β-Heptapeptide (3) van Gunsteren, W. F.; Gattin, Z. Simulation of folding equilibria. In Foldamers: Structure, Properties and Applications; Hecht, S., Huc, I., Eds.; Wiley: Weinheim, Germany 2007; pp 173-192. (4) Hook, D. F.; Gessier, F.; Noti, C.; Kast, P.; Seebach, D. Probing the proteolytic stability of β-peptides containing R-fluoro and R-hydroxyβ-amino acids. ChemBioChem 2004, 5, 691–706. (5) Rueping, M.; Mahajan, Y. R.; Sauer, M.; Seebach, D. Cellular uptake studies with β-peptides. ChemBioChem 2002, 3, 257–259. (6) Arvidsson, P. I.; Ryder, N. S.; Weiss, H. M.; Gross, G.; Kretz, O.; Woessner, R.; Seebach, D. Antibiotic and hemolytic activity of a β2/β3 peptide capable of folding into a 12/10-helical secondary structure. ChemBioChem 2003, 4, 1345–1347. (7) Werder, M.; Hauser, H.; Abele, S.; Seebach, D. β-peptides as inhibitors of small-intestinal cholesterol and fat absorption. HelV. Chim. Acta 1999, 82, 1774–1783. (8) Gademann, K.; Ernst, M.; Seebach, D. The cyclo- β-tetrapeptide (β-HPhe- β-HThr- β-HLys- β-HTrp): Synthesis, NMR structure in methanol solution and affinity for human somatostatin receptors. HelV. Chim. Acta 2000, 83, 16–33. (9) Gu¨nther, R.; Hofmann, H.-J. Theoretical prediction of substituent effect on the intrinsic folding properties of β-peptides. HelV. Chim. Acta 2002, 85, 2149–2168. (10) Martinek, T. A.; Fu¨lo¨p, F. Side-chain control of β-peptide secondary structure- Design principles. Eur. J. Biochem. 2003, 270, 3657–3666. (11) Raguse, T. L.; Lai, J. R.; LePlae, P. R.; Gellman, S. H. Toward β-peptide tertiary structure: Self-association of an amphiphilic 14-helix in aqueous solution. Org. Lett. 2001, 3 (24), 3963–3966. (12) Gla¨ttli, A.; Seebach, D.; van Gunsteren, W. F. Do valine side chains have an influence on the folding behaviour of β-substituted β-peptides. HelV. Chim. Acta 2004, 87, 2487–2506. (13) Peter, C.; Daura, X.; van Gunsteren, W. F. Peptides of aminoxy acids: A molecular dynamics simulation study of conformational equilibria under various conditions. J. Am. Chem. Soc. 2000, 122, 7461–7466. (14) Gattin, Z.; Gla¨ttli, A.; Jaun, B.; van Gunsteren, W. F. Simulation of β-depsipeptides: The effect of missing hydrogen-bond donors on their folding equilibria. Biopolymers 2007, 85 (4), 318–332. (15) Gla¨ttli, A.; van Gunsteren, W. F. Are NMR-derived model structures for β-peptides representative for the ensemble of structures adopted in solution. Angew. Chem., Int. Ed. Engl. 2004, 43, 6312–6316. (16) Gademann, K.; Ha¨ne, A.; Rueping, M.; Jaun, B.; Seebach, D. The fourth helical secondary structure of β-peptides: The (P)-2 8-helix of a β-hexapeptide consisting of (2r,3s)-3-amino-2-hydroxy acid residues. Angew. Chem., Int. Ed. Engl. 2003, 42, 1534–1537. (17) Capone, S.; Guarangna, A.; Palumbo, G.; Pedatella, S. Efficient synthesis of orthogonally protected anti-2,3-diamino acids. Tetrahedron 2005, 61, 6575–6579. (18) Martin, J. H.; Hausmann, W. K. Isolation and identification of DR-pipecolic acid, R [L], β-methylaspartic acid and R,β-diaminobutyric acid from the polypeptide antibiotic aspartocin. J. Am. Chem. Soc. 1960, 82, 2079. (19) Herranz, R.; Vinuesa, S.; Castro-Pichel, J.; Pe´rez, C.; Garcı´a-Lo´pez, M. T. Aminodeoxybestatin and epi-aminodeoxybestatin: stereospecific synthesis and aminopeptidase inhibition. J. Chem. Soc., Perkin Trans 1 1992, 1825–1830. (20) Seebach, D.; Ciceri, P. E.; Overhand, M.; Jaun, B.; Rigo, D.; Oberer, L.; Hommel, U.; Amstutz, R.; Widmer, H. Probing the Helical Secondary Structure of Short-Chain β-peptides. HelV. Chim. Acta 1996, 79, 2043– 2066. (21) Daura, X.; van Gunsteren, W. F.; Rigo, D.; Jaun, B.; Seebach, D. Studying the stability of a helical β-heptapeptide by molecular dynamics simulations. Chem.sEur. J. 1997, 3, 1410–1417. (22) Gessier, F.; Noti, C.; Rueping, M.; Seebach, D. Synthesis and CD spectra of fluoro- and hydroxy-substituted β-peptides. HelV. Chim. Acta 2003, 86, 1862–1870. (23) Banks, J. W.; Batsanov, A. S.; Howard, J. A. K.; O’Hagan, D.; Rzepa, H. S.; Martin-Santamaria, S. The preferred conformation of R-fluoroamides. J. Chem. Soc., Perkin Trans 2 1999, 2409–2411. (24) Mathad, R. I.; Jaun, B.; Flo¨gel, O.; Gardiner, J.; Lo¨weneck, M.; Code´e, J. D. C.; Seeberger, P. H.; Seebach, D.; Edmonds, M. K.; Graichen, F. H. M.; Abell, A. D. NMR-solution structures of fluoro-substituent β-peptides: A 314-helix and a hairpin turn. The first case of a 90° OdCC-F dihedral angle in an R-fluoro-amide group. HelV. Chim. Acta 2007, 90, 2251–2273.

J. Phys. Chem. B, Vol. 113, No. 25, 2009 8703 (25) van Gunsteren, W. F. Billeter, S. R. Eising, A. A. Hu¨nenberger, P. H. Kru¨ger, P. Mark, A. E. Scott, W. R. P., and Tironi, I. G. Biomolecular Simulation: The GROMOS96 Manual and User Guide; vdf Hochschulverlag, ETH Zu¨rich: Zu¨rich, Switzerland, 1996. (26) Scott, W. R. P.; Hu¨nenberger, P. H.; Tironi, I. G.; Mark, A. E.; Billeter, S. R.; Fennen, J.; Torda, A. E.; Huber, T.; Kru¨ger, P.; van Gunsteren, W. F. The GROMOS biomolecular simulation program package. J. Phys. Chem. A 1999, 103 (19), 3596–3607. (27) Schuler, L. D.; Daura, X.; van Gunsteren, W. F. An improved GROMOS96 force field for aliphatic hydrocarbons in the condense phase. J. Comput. Chem. 2001, 22, 1205–1218. (28) Oostenbrink, C.; Villa, A.; Mark, A. E.; van Gunsteren, W. F. A biomolecular force field based on the free enthalpy of hydration and solvation: The GROMOS force-field parameter sets 53A5 and 53A6. J. Comput. Chem. 2004, 25, 1656–1676. (29) Fioroni, M.; Burger, K.; Mark, A. E.; Roccatano, D. Model of 1,1,1,3,3,3-hexafluoro-propan-2-ol for molecular dynamics simulations. J. Phys. Chem. B 2001, 105, 10967–10975. (30) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A.; Vreven, T., Jr.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, J.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03; Gaussian, Inc.: Wallingford, CT, 2004. (31) Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren, W. F.; DiNola, A.; Haak, J. R. Molecular dynamiccs with coupling to an external bath. J. Chem. Phys. 1984, 81 (8), 3684–3690. (32) Ryckaert, J. P.; Ciccotti, G.; Berendsen, H. J. C. Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes. J. Comput. Phys. 1977, 23, 327–341. (33) Barker, J. A.; Watts, R. O. Monte Carlo studies of the dielectric properties of water-like models. Mol. Phys. 1973, 26, 789–792. (34) Tironi, I. G.; Sperb, R.; Smith, P. E.; van Gunsteren, W. F. A generalized reaction field method for molecular dynamics simulations. J. Chem. Phys. 1995, 102 (13), 5451–5459. (35) Walser, R.; Mark, A. E.; van Gunsteren, W. F.; Lauterbach, M.; Wipff, G. J. The effect of force-field parameters on properties of liquids: Parametrisation of a simple three site model for methanol. J. Chem. Phys. 2000, 112, 10450–10459. (36) Daura, X.; van Gunsteren, W. F.; Mark, A. E. Folding-unfolding thermodynamics of a β-heptapeptide from equuilibrium simulations. Proteins: Struct., Funct., Genet. 1999, 34, 269–280. (37) Daura, X.; Jaun, B.; Seebach, D.; van Gunsteren, W. F.; Mark, A. E. Reversible peptide folding in solution by molecular dynamics simulation. J. Mol. Biol. 1998, 280, 925–932. (38) Tropp, J. Dipolar relaxation and nuclear Overhauser effects in nonrigid molecules: The effect of fluctuating internuclear distances. J. Chem. Phys. 1980, 72 (11), 6035–6043. (39) Daura, X.; Antes, I.; van Gunsteren, W. F.; Thiel, W.; Mark, A. E. The effect of motional averaging on the calculation of NMR-derived structural properties. Proteins 1999, 36, 542–555. (40) Karplus, M. Contact electron-spin coupling of nuclear magnetic moments. J. Chem. Phys. 1959, 30, 11–15. (41) Pardi, A.; Billeter, M.; Wu¨thrich, K. Calibration of the angular dependence of the amide proton - CR proton coupling constants, 3JHNR, in a globular protein. J. Mol. Biol. 1984, 180, 741–751. (42) Schwieters, C. D.; Kuszewski, J. J.; Tjandra, N.; Clore, G. M. The Xplor-NIH NMR molecular structure determination package. J. Magn. Reson. 2003, 160, 65–73. (43) Shoulders, M. D.; Hodges, J. A.; Raines, R. T. Reciprocity of steric and stereoelectronic effects in the collagen triple helix. J. Am. Chem. Soc. 2006, 128, 8112–8113.

JP811106E