SERS, Molecular Dynamics and Molecular Orbital Studies of the

ARTICLE pubs.acs.org/JPCC

SERS, Molecular Dynamics and Molecular Orbital Studies of the MRKDV Peptide on Silver and Membrane Surfaces Alvaro E. Aliaga,† Hern
an Ahumada,§ Karen Sep
ulveda,§ Juan S. Gomez-Jeria,† Carlos Garrido,† Boris E. Weiss-L
opez,*,§ and Marcelo M. Campos-Vallette*,† †

Laboratory of Vibrational Spectroscopy and §Laboratory of Molecular Physical Chemistry, Faculty of Science, University of Chile, PO Box 653 Santiago, Chile ABSTRACT: The MRKDV peptide, structurally associated with an immunomodulatory protein, was studied using surface enhanced Raman scattering (SERS), molecular dynamics (MD) simulations, and quantum chemical calculations. The SERS spectrum of the MRKDV peptide adsorbed on the silver surface is dominated by signals coming from the guanidinium moiety of the arginine amino acid (R). Guanidinium is the intrinsic probe that drives the orientation of the peptide onto the silver surface. Molecular mechanics and extended H€uckel calculations of a model of MRKDV interacting with a silver surface support the experimental results. MD calculations representing the evolution of the peptide toward a model membrane were also performed. The guanidinium moiety interacts with the phospholipidic membrane surface. A hydrophobic C-terminal modification favors the peptide membrane affinity.

1. INTRODUCTION Surface enhanced Raman scattering SERS has been an active research area with important applications ranging from surface chemistry to biological chemistry and biomedical analysis. Biosensors are intrinsically associated with peptide-metal nanoparticle systems.1-3 The peptide-metal nanoparticle interaction also has important consequences on the optical properties of the nanoparticle itself. SERS of a number of amino acids, peptides, and proteins, acquired with various SERS active substrates, has been reported. Recent SERS results by Di Foggia et al.4 indicate that in the case of four alternating polar/nonpolar peptides derived from the self-assembling peptide EAK 16 (Ac-AEAKAEAKAEAKAEAK-NH2), the interaction with a TiO surface takes place mainly through carboxylate groups, and the selfassembled structure of the peptide on the metal surface is predominantly β-sheet. Mitchell et al.5 have recently reported that the SERS methodology is a key factor related to the success of the application of metal nanoparticles in the appearance of strong SERS features from oligopeptides; they also used statistical analysis methods for the peptide detection by SERS. Wei et al.6 obtained the SERS spectra of three cysteine (Cys)-containing aromatic peptides and penetratin bound to nanoshell substrates. They concluded that the aromatic amino acid residues provide the dominant features in the SERS and Raman spectra. From SERS spectra of several peptides composed of different combinations of proline, tryptophan, and tyrosine, Seballos et al.7 concluded that the binding with a silver surface occurs through both the carboxylic end and the aromatic amino acid moieties. r 2011 American Chemical Society

Recent results on the applications of SERS to peptides have been published by Podstawka et al.,8 who compared the adsorption behavior of bombesin (BN) (pGlu-Gln-Arg-Leu-Gly-AsnGln-Trp-Ala-Val-Gly-His-Leu-Met-NH2) and five BN-related peptides in silver colloidal solution. The peptide metal interaction is mainly verified through the pyrrole ring of tryptophan and the aromatic ring of the phenylalanine components. They also infer a weak interaction through particular skeletal fragments of the peptidic chain. These results are slightly modified in the series of peptides. Differences in organization and orientation of the peptides on the metal surface by changing the silver colloidal surface to a silver electrode surface were inferred. Infrared techniques such as PM, RAIRS, and ATR allowed one to propose that the peptide L-glutathione (γ-Gln-Cys-Gly) interacts with Au through the cysteine thiol group in a goldwater interface and through the glycine carboxylate in a goldethanol interface.9 A reflection absorption infrared spectroscopic (RAIRS) study on Ag was performed by Itoh et al.10 for Langmuir-Blodgett films of the palmitoyl ornithine (PO) and palmitoyl lysine (PL) peptides. The authors concluded a particular organization of the peptides at the surface. The interaction of peptides and oligopeptides with metal surfaces can occur through at least two different mechanisms: (a) electrostatic interaction between charges in the oligopeptide and residual charges existing on the metal surface and (b) coordination complexes formed between very active groups, Received: November 9, 2010 Revised: December 29, 2010 Published: February 25, 2011 3982

dx.doi.org/10.1021/jp1107153 | J. Phys. Chem. C 2011, 115, 3982–3989

The Journal of Physical Chemistry C

ARTICLE

Figure 1. MRKDV-Ag final optimized geometry. Ball-and-stick molecular model displays the chemical element symbols. The silver surface appears in gray, and the peptide atoms are carbons, light blue; nitrogen, dark blue; oxygen, red; sulfur, yellow; and hydrogen, white.

such as -SH, -COO-, imidazol in hystidine (His),11 and indole in tryptophan (Trp),12 with surface metal atoms. In addition, intermolecular interactions between oligopeptides should be considered. In this sense, the hydrophobic interactions between nonpolar amino acid residues in the oligopeptides also play an important role in both the three-dimensional conformation and the mechanism of adsorption on the metal surface. The metal ion binding properties of two fluorescent analogues of trichogin GA IV, which is a natural undecapeptide showing significant antimicrobial activity, were studied by circular dichroism, time-resolved optical spectroscopy, and molecular mechanics calculations.13 Binding of Ca(II) and Gd(III) to the studied peptide was shown to promote a structural transition from highly helical conformations to folded structures characterized by the formation of a loop that embedded the metal ion. Time-resolved fluorescence spectroscopy revealed that the peptide dynamics is also remarkably affected by ion binding: peptidebackbone motions slowed down to the microsecond time scale. Finally, molecular mechanics calculations emphasized the role of the central Gly5-Gly6 motif, which allowed for the twisting of the peptide segment that gave rise to the formation of the binding cavity. Several bioactive peptides exert their biological function by interacting with cellular membranes. Structural data and information about the distribution inside lipid bilayers become essential for a detailed understanding of their mechanism of action. A detailed biophysical characterization of the interaction of the cell-penetrating S413-PV peptide with model membranes was performed.14 It has been demonstrated that the interactions are essentially of electrostatic type. As a consequence of this interaction with negatively charged model membranes, the S413-PV peptide becomes buried into the lipid bilayer, which occurs concomitantly with significant peptide conformational changes consistent with the formation of a helical structure. Using surface plasmon resonance, the binding properties of a new class of immunomodulating peptides derived from the transmembrane region of the T cell antigen receptor on model membranes were determined.15 The dibasic core peptide was found to bind both zwitterionic and anionic model membranes as well as to T cell membrane preparations. By contrast, switching

one or both of the basic residues to acidic residues led to a complete loss of binding to model membranes. The position of the charged amino acids in the sequence, the number of hydrophobic amino acids between the charged residues, and substitution of one or both basic to neutral amino acids were found to affect binding. Molecular dynamics simulations were performed on the sequence-symmetric cyclic decapeptide antibiotic gramicidin S (GS), interacting with a hydrated dimyristoylphosphatidylcholine (DMPC) bilayer; Results were compared with a control simulation of the system in the absence of GS.16 The area per lipid, lipid tail order parameters, and quadrupole spin-lattice relaxation times data of the control simulation are in agreement with experiments. The GS has little effect on the membrane dipole potential or water permeability. Atomic force microscopy has been used to probe the organization of lipid monolayers and bilayers.17 The authors explored peptide-membrane interactions, with emphasis on microbial lipopeptides and tilted peptides. By means of a combined approach of fluorescence spectroscopy and molecular dynamics, B. Orioni et al.18 studied the mechanism of membrane perturbation by the antimicrobial peptide PMAP-23. Depth-dependent fluorescence quenching experiments and peptide-translocation assays were employed to determine the location of the peptide inside the membrane. Molecular dynamics (MD) simulations were performed starting from a random mixture of water, lipids, and peptide following the spontaneous self-assembly of the bilayer. Both experimental and theoretical data indicated a peptide location just below the polar head groups of the membrane, with an orientation essentially parallel to the bilayer plane. The atomic detail provided by the simulations suggested the occurrence of an additional, more specific, novel mechanism of bilayer destabilization by PMAP-23 involving the unusual insertion of charged side chains into the hydrophobic core of the membrane. The aim of the present contribution deals with the study by SERS spectroscopy of the N-terminal motif peptide, methionylarginyllysyl-R-aspartylvaline MRKDV (see Figure 1), from the subunit A of the immunogenic Concholepas concholepas hemocyanin.19 To complete the analysis of the SERS experiments, a 3983

dx.doi.org/10.1021/jp1107153 |J. Phys. Chem. C 2011, 115, 3982–3989

The Journal of Physical Chemistry C theoretical study based on the extended H€uckel theory and 6-31G* methods for a simplified molecular model of the analyte-metal surface interaction is proposed. Different experimental techniques provide evidence that supports the assumption that the role of subunit A, containing the motif MRKDV, seems to be crucial for the proper immunomodulatory functioning of the protein in vertebrates.20 With this perspective in mind, we extend the SERS study with a MD study of the behavior of the isolated motif peptide in a phospholipid membrane environment. With this contribution, we expect to provide information such as peptide location, interactions with the membrane components, and modifications of the MRKDV conformational structure during the possible process of incorporation that might be essential for understanding the participation of the motif peptide in the subunit-membrane interaction. The MD simulation study is performed for MRKDV in a zwitterionic bilayer made of a mixture of dimyristoylphosphatidylcholine (DMPC) and dihexanoylphosphatidylcholine (DHPC) in water. To account for the possible hydrophobic effects induced by the rest of the protein, a more hydrophobic derivative, the pentapeptide methionylarginylysyl-R-aspartylN1-(1,1 diethylpropyl)valineamide (MRKDVhd), was also studied. Targeting and understanding the biophysical interactions between peptides and membranes at their site of action is paramount for the description of cell functioning and drug design.

2. EXPERIMENTAL SECTION 2.1. Sample. Synthetic water-soluble peptide MRKDV highly purified was purchased from Operon and used as supplied (g98.5% purity). Stock solutions of the peptide in water were prepared to a final concentration of 10-4 M in nanopure water. 2.2. Preparation of Silver Nanoparticles. Silver nanoparticles (AgNPs) were prepared by chemical reduction of silver nitrate with hydroxylamine hydrochloride.21 The resulting colloids display a final pH close to 7. The aqueous solutions used for the AgNPs formation were prepared by using nanopure water. Colloids showed a milky gray color. 2.3. Preparation of SERS Samples. Aliquots of the aqueous peptide solution 10-4 M were added onto a quartz slide (0.51 μL) and dried at room temperature; a film of the dried peptide was obtained. Then the Ag colloidal suspension was added (∼50 μL) onto the dried peptide sample to form peptide-NP aggregates. SERS measurements were performed at room temperature. 2.4. Instrumentation. The Raman and SERS spectrum of the peptide was measured with a Renishaw micro-Raman system (RM1000). The sample was photostable when probed with laser lines at 514, 633, and 785 nm. The best spectral data were obtained by using the 633 nm laser line. This instrument was equipped with a Leica microscope and an electrically refrigerated CCD camera. The signal was calibrated by using the 520 cm-1 line of a Si wafer and a 50 objective. The laser power on the sample was 2 mW. The resolution was set to 4 cm-1, and 5-20 scans of 10 s each were averaged. The spectral scanning conditions were chosen to avoid sample degradation. 2.5. Spectral Reproducibility. No reproducible SERS spectra were obtained by using the traditional way to prepare the sample; that is, by addition of the sample solution to the colloidal suspension or the inverse. The SERS reproducible spectra from batch to batch were obtained by adding the colloidal suspension onto the dried analyte sample.

ARTICLE

3. THEORETICAL CALCULATION 3.1. Molecular Models, Methods and Calculations. To complete the analysis of the SERS experiments, a theoretical study was performed. A simplified molecular model for the analyte surface interaction is proposed. The silver atoms surface was the same employed in our previous studies.22-25 Briefly, a face centered cubic structure with a = 0.408 nm and 9  9  2 cells was built and trimmed to get a planar double layer composed of 324 silver atoms. Since the experimental conditions of the SERS study, the MRKDV peptide was studied as a guanidinium form. Molecular mechanics was employed to optimize the MRKDVAg geometry. The bilayer geometry was kept constant. The peptide was placed at different distances and orientations from the center of the Ag bilayer. Figure 1 shows the final geometry. Extended H€uckel theory (EHT) was used to calculate the wave function of MRKDV as an isolated system and interacting with the metal surface. The Hyperchem program was used.26 EHT calculations produce qualitative or semiquantitative descriptions of molecular orbital and electronic properties.27 The combination of EHT with molecular mechanics was able to give, for example, a qualitative explanation of our previous SERS works in nanotubes,22 humic acids,28 tryptophan,23 lysine,24 and arginine25 interacting with Ag surfaces. 3.2. Molecular Dynamics Simulation. A bilayer made of 72 DMPC and 24 DHPC molecules surrounded by 7049 SPC29 water molecules was assembled in a box of dimensions 5.7  5.3  10.3 nm3, with periodic boundary conditions in all directions of space. This system was equilibrated for more than 10 ns. After that, the simulation box was replicated three times having a total of four simulations. One molecule of MRKDV was incorporated in the hydrophobic core of the bilayer in simulation 1, and one molecule of MRKDV, in the aqueous region of simulation 2. To study the effect of hydrophobic groups, presumably present in the original protein but not in MRKDV, the 3-ethylpentane-3-amine group was bonded directly to the carbonyl atom of valine, forming a peptide bond, to obtain the more hydrophobic derivative MRKDVhd. Following the same procedure described before, one molecule of MRKDVhd was introduced in the hydrophobic region of simulation 3, and one MRKDVhd, in the aqueous region of simulation 4. Figure 2 shows a snapshot of simulation 2 after 5 ns of trajectory. Calculation of all trajectories and their further analyses were made using the software package Gromacs v. 4.030 running in a Super-Micro 4-blades cluster with 2 quad-core processors each. An attached atom representation was used for the protons in the aliphatic chains of the lipids. For trajectory visualization and graphics, the VMD program31 was employed. Bonded and nonbonded interactions were calculated using the Gromos force field.32 LINKS was used to restrict the bond lengths, and SETTLE, to constraint the geometry of water molecules.33,34 A 1 nm cutoff was used for the Lennard-Jones and real space electrostatic potentials. Long-range electrostatic interactions were calculated using PME.35-37 The neighbor list was updated every 10 time steps. Temperature (T) and pressure (P) were kept constant at 300 K (all species independently coupled) and 1 bar using a weak coupling algorithm,38 with time constants of 0.1 and 1.0 ps for T and P, respectively. Trajectories of 6, 6, 18, and 50 ns were calculated for simulations 1, 2, 3, and 4 respectively, all with a time step size of 2 fs. 3984

dx.doi.org/10.1021/jp1107153 |J. Phys. Chem. C 2011, 115, 3982–3989

The Journal of Physical Chemistry C

ARTICLE

Figure 2. Snapshot of simulation 2 after 5 ns of trajectory. DMPC molecules appear in gray; DHPC, in black; and water oxygen and hydrogen are red and white, respectively. DMPC, DHPC, and water are displayed by the bond dynamic drawing method. MRKDV peptide is displayed in the van der Waals and bonds drawing method with carbons in light blue, oxygen in red, sulfur in yellow, and nitrogen in dark blue.

Table 1. SERS Frequencies (cm-1) of the MRKDV Peptide and the Most Probable Band Assignment MRKDV

a b

Figure 3. (A) Raman and (B) SERS spectra of the oligopeptide MRKDV. SERS spectrum was obtained by adding the colloidal suspension onto the dried analyte sample.

4. RESULTS AND DISCUSSION 4.1. SERS Spectra. A unique and reproducible SERS spectrum in colloidal solution for the peptide was not achieved. The

assignmenta

1615

δNH

1559

δNH

1447

D, K, Vb

1353 1307

D, K D, K

1275

D, K, V

1175

K

1060

R

990

R

927

M

845

R

726 685

M, K, V, D M

636

D, K

410

δCCN

M, methionine; R, arginine; K, lysine; D, aspartic acid; and V, valine. For the specific normal modes involved in the vibration, please see ref 39.

SERS spectrum of the peptide in solid, that is, the analyte coated by the metal nanostructure, was successful. The spectral analysis of the SERS spectrum was performed on the basis of the existing Raman data and our previous Raman and SERS band assignments for most of the amino acid residues of the present peptide.23-25,38-41 Raman and SERS spectra of the peptide are displayed in Figure 3. Table 1 contains the most probable assignment expressed in terms of the amino acid symbol. Details of the corresponding normal modes are given by Stewart et al.39 The strongest band of MRKDV at 1060 cm-1 and the weak ones at 990 and 845 cm-1 are ascribed to the guanidinium fragment of Arg (R).25 Other very weak bands are assigned to the amino acids 3985

dx.doi.org/10.1021/jp1107153 |J. Phys. Chem. C 2011, 115, 3982–3989

The Journal of Physical Chemistry C

ARTICLE

Figure 4. Time evolution of the center of mass (c.o.m.) Z-coordinate of all amino acid components of MRKDV (A and B) and MRKDVhd (C and D). Panels A, B, C, and D correspond to simulations 1, 2, 3, and 4, respectively. The limits of the top interface are denoted by horizontal lines.

Met (M), Lys (K), Asp (D), and Val (V) (see Figure 3 and Table 1). The presence of these very weak bands, mostly ascribed to different vibrations of the carboxylic and amino groups, CH2 deformations, and particularly the νCS of Met at 685 cm-1,40 indicates that the corresponding amino acid residues also interact with the metal surface. This interaction should be very weak, even though with a SERS favorable orientation of the chemical groups involved.42 To precise the knowledge of the physical and chemical characteristics of the interaction, it should be necessary to study the nature of the peptide metal interface. The Asp bands of MKRDV at 1447, 1307, and 726 cm-1 suggest that the Arg residue anchors the MRKDV peptide to the surface through its guanidine group. This proposition agrees with the results of Di Costanzo et al.,43 who described the stereochemistry of the guanidine-metal interactions in small molecules and proteins. 4.2. MRKDV Metal Interaction. In the final MRKDV-Ag geometry (see Figure 1), the sulfur atom is located at a distance of 3.0-3.4 Å
from some Ag atoms. One of the N atoms of the guanidinium group is at 2.8-3.2 Å from the metal surface, and their attached H atoms is at distances of ∼2.40 Å from the Ag surface. The N amino and its H-bonded atoms are located at
from the silver surface. Two CO groups are in the range ∼2.4 Å 2.4-2.8 Å
away from the surface. The CO groups are placed at the center of the Ag surface area in which the electronic density of the HOMO is zero. The guanidinium moiety and the sulfur atom are also placed on parts of the metallic surface in which the electron density of the

HOMO is zero. This geometrical situation is the optimal one due to the fact that the interaction of the oxygen lone pairs with the HOMO of the Ag surface must be minimal. As expected from the analysis of the isolated systems, the results for the MRKDV-Ag system indicate that, as in the case of zwitterion arginine, a charge transfer occurs from the valence band of the Ag surface to the pentapeptide. The charge transfer is verified from the valence band of the Ag surface to the empty LUMOs of the guanidinium moiety, the CO oxygen atoms and the sulfur atom closest to the Ag surface. 4.3. Molecular Dynamics. Figure 4 shows the center of mass (c.o.m.) Z-coordinate of the complete peptides as well as the c.o. m. Z-coordinate for each amino acid component as a function of time. Figures 4A-D corresponds to simulations 1-4, respectively. All these figures represent the upper interface of the simulation box (see Figure 2). Three different regions, characteristic of all amphiphilic bilayers, can be distinguished: region I corresponds to the aqueous phase, region II is the interface, and region III is the hydrophobic core; that is, the center of the bilayer. The horizontal lines represent the limits of the interface, estimated according to a previous definition.44 The width of the interface in the absence of guest peptides is 1.1 nm, and the inclusion of MRKDV decreases this value to 0.8-0.9 nm. The inclusion of MRKDVhd has no significant effect on the width of the interface. As mentioned before, in simulation 1 (Figure 4A), MRKDV was originally positioned in region III and remained there for the rest of the simulation. It can be noticed that Val and Lys are positioned closer to the interface, near the aqueous region. 3986

dx.doi.org/10.1021/jp1107153 |J. Phys. Chem. C 2011, 115, 3982–3989

The Journal of Physical Chemistry C

ARTICLE

Figure 5. Mass density profiles of all amino acid components of MRKDV (A and B) and MRKDVhd (C and D) along the Z-axis of the box. Parts A, B, C, and D correspond to simulations 1, 2, 3 and 4, respectively. The limits between regions I, II, and III are denoted by vertical lines.

However, when MRKDV is initially positioned in the aqueous region (Figure 4B), a significant displacement toward the interface and modifications in the distribution of amino acids along the Z-coordinate are observed. The final distribution shows that Val and Arg remained near the aqueous phase, whereas Met, Lys, and Asp become more incorporated into the interface. In these first two simulations, Val contains the complete carboxylate moiety and is always located near the aqueous region of the box. The amine terminal of Met forms a strong hydrogen bond with water at the upper limit of the interface in simulation 2 or with water at the lower limit interface in simulation 1 (the lower interface limits are not indicated in Figure 4). When MRKDVhd is initially positioned in region III, it shows a moderate displacement toward the interface (Figure 4C), without significant amino acids reordering; however, if MRKDVhd is originally placed in region I near the interface (Figure 4D), it experiences a large displacement toward the interior of the interface and a significant modification in the distribution of amino acids along the Z-axis of the box. Met crosses most amino acids to finally reach the lower interface to become hydrogen bonded to water. Arg is located at the interior of the interface, possibly with the guanidine moiety interacting with the phosphate groups. In the last two simulations, Val is directly attached to the hydrophobic substituent, and as expected, it locates closer to region III. The existence of crossings among the amino acids strongly suggests the occurrence of conformational changes in the peptide. This is discussed in more detail later.

To examine the distribution of the amino acids along the Zaxis of the box, mass density profiles of the equilibrated systems were calculated for all simulations. The results are displayed in Figure 5A-D for simulations 1-4, respectively. It can be observed that when either MRKDV or MRKDVhd is initially located in region III (Figure 5A and C), all amino acids appear distributed over this entire region, without significant crossing to region II; however, the distributions of MRKDVhd are narrower than the distributions of MRKDV. This is clearly observed for Lys and Met. If MRKDV or MRKDVhd is initially located in region I (Figure 5B and D), the distributions of MRKDV are also wider than the distributions of MRKDVhd. Furthermore, the residues of MRKDV appear distributed around the interface between regions I and II, whereas the residues of MRKDVhd are distributed around the interface between regions II and III, deeper into the hydrophobic core. This is expected for the more hydrophobic derivative. The results from the simulations show that neither MRKDV nor MRKDVhd is able to cross the interface and penetrate the hydrophobic core of the membrane. When they are initially positioned in the aqueous phase, region I, they spontaneously displace toward the interior of the bilayer, with the final equilibrium location being a function of the hydrophobic character of the peptide. However, when originally located in the hydrophobic core, region III, they remain there. To test for a possible interaction between the guanidine moiety of Arg and the phosphate groups from the phospholipids, the radial distribution function of phosphate groups from DMPC 3987

dx.doi.org/10.1021/jp1107153 |J. Phys. Chem. C 2011, 115, 3982–3989

The Journal of Physical Chemistry C

ARTICLE

Table 2. Average Principal Axis Moments of Inertia of MRKDV and MRKDVhd, (a.m.u.  nm2) When Located in Different Environments: The Interface (region II) and Closer to the Hydrophobic Core (region III)a Ixx region II

MRKDVhd

region III region II

MRKDV

region III a

Figure 6. Radial distribution function of the phosphate groups surrounding the guanidine moiety from arginine. The first maximum at 0.6 nm corresponds to two phosphate groups. The origin is located at the center of mass of the guanidine moiety.

around the guanidine moiety of Arg was calculated from simulation 3, and the result is shown in Figure 6. The first maximum can be estimated at a distance of 0.6 nm and corresponds to two phosphate groups surrounding the guanidine fragment. The existence of a specific electrostatic interaction between the guanidine moiety and the phosphate groups from DMPC cannot be discarded. Even more, a strong and persistent hydrogen bond between the guanidine group and the phosphate head groups from DMPC is detected. The observed crossings among the amino acids along the Zcoordinate, suggest the existence of conformational changes when the environment of the peptide is modified. To detect possible conformational changes of the peptides when located in different regions of the simulation box, we have calculated the average values of the three principal axis inertia moments of MRKDV and MRKDVhd as a function of time in all simulations. An oblate symmetric body possesses two identical moments of inertia, which are smaller than the unique inertia moment, colinear with the symmetry axis of the oblate. A prolate possesses two equal inertia moments that are greater than the unique inertia moment, which is along the symmetry axis of the prolate. A nonsymmetric body has all inertia moments different. Calculated average values of the principal axis inertia moments are displayed in Table 2. When MRKDV approached the interface from either region III or I (simulations 1 and 2), the starting structure was nonsymmetric in both cases. Along the simulations, the peptide becomes a prolate. On the other hand, when MRKDVhd approaches region II from regions III and I (simulations 3 and 4), in both cases, it starts as a prolate symmetric body. Along simulation 3, no significant modifications to the structure were observed and it remains a prolate; however, along simulation 4, the original prolate structure becomes a nonsymmetric body. One of the energy components that affects the three-dimensional structure of proteins and peptides is hydrogen bonds. To explore the origin of the forces responsible for the different conformation, the existence of intramolecular hydrogen bonds between carboxylates and ammonium groups was tested. When MRKDVhd is located near region III, two strong hydrogen bonds were detected between Asp and Arg and between Val and Lys. When it is closer to region II, it becomes a nonsymmetric body with no significant

Iyy

Izz

110 ( 5

149 ( 11

184 ( 10

217 ( 19

255 ( 19

74 ( 6

161 ( 24

192 ( 19

72 ( 10

130 ( 10

153 ( 11

44 ( 4

Errors correspond to standard deviations.

intramolecular hydrogen bond. All the above results are consistent with an important conformational change of MRKDVhd when going from region I to II, with a final average structure that has no symmetry. However, when MRKDVhd goes from region III to II, it remains as a prolate structure, stabilized by intramolecular hydrogen bonds.

5. CONCLUSIONS The peptide MRKDV, structurally associated with an immunomodulatory protein, was vibrationally studied by using surface enhanced Raman scattering SERS and both quantum mechanics and molecular dynamics approaches. No sample degradation (photo or thermal) was observed using 514, 633, and 785 nm laser lines. The sample for SERS was prepared by adding the Ag colloid suspension onto the dried peptide sample, the whole deposited on quartz slide. The MRKDV SERS spectrum is dominated by signals coming from the guanidinium moiety of the arginine amino acid (R); guanidinium is the intrinsic probe that drives the orientation of the peptide on the Ag surface. Theoretical calculations performed by using molecular mechanics and extended H€uckel theory methods for a model of the MRKDV peptide interacting with a Ag cluster support the observed experimental result. We also conclude that when MRKDV, or the N1-(1,1-diethylpropyl)valineamide-substituted peptide (MRKDVhd) is initially located in the aqueous region, it spontaneously displaces toward the interior of the bilayer. The final equilibrium position in the membrane is a function of the hydrophobic character of the peptide. When MRKDV becomes incorporated into the interface, it carries a significant amount of water from the interior to that region, decreasing the width of region II. When the peptides are initially positioned in region III, the center of the bilayer, they remain there. Depending on the environment, the peptide can display different conformations. A persistent H-bond between the guanidine moiety of arginine and the phosphate groups from DMPC was detected. Finally, the amine terminal from methionine forms a strong hydrogen bond with water in either of the interfaces. ’ AUTHOR INFORMATION Corresponding Author

*E-mail: (M.M.C.-V.) [email protected], (B.E.W.-L.) bweiss@ uchile.cl.

’ ACKNOWLEDGMENT This contribution is financially supported by Fondecyt, Grants 1090074 and 1095175. A.E.A. acknowledges AT 24090050 and Doctoral Fellowship CONICYT. 3988

dx.doi.org/10.1021/jp1107153 |J. Phys. Chem. C 2011, 115, 3982–3989

The Journal of Physical Chemistry C

’ REFERENCES (1) Ramanathan, S.; Ensor, M.; Daunert, S. Trends Biotechnol. 1997, 15, 500. (2) Marvin, J. S.; Hellinga, H. W. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 4955. (3) Bontidean, I.; Kumar, A.; Csoregi, E.; Galaev, I. Y.; Mattiasson, B. Angew. Chem., Int. Ed. 2001, 40, 2676. (4) Di Foggia, M.; Taddei, P.; Fagnano, C.; Torreggiani, A.; Dettin, M.; Sanchez-Cortes, S.; Tinti, A. J. Mol. Struct. 2009, 120, 924. (5) Mitchell, B. L.; Patwardhan, A. J.; Ngola, S. M.; Chan, S.; Sundararajan, N. J. Raman Spectrosc. 2008, 39, 380. (6) Wei, F.; Zhang, D.; Halas, N. J.; Hartgerink, J. D. J. Phys. Chem. B. 2008, 112, 9158. (7) Seballos, L.; Richards, N.; Stevens, D. J.; Patel, M.; Kapitzky, L.; Lokey, S.; Millhauser, G.; Zhang, J. Z. Chem. Phys. Lett. 2007, 447, 335. (8) Podstawka, E.; Proniewicz, L. M. J. Phys. Chem. B. 2009, 113, 4978. (9) Bieri, M.; Burgi, T. Langmuir. 2005, 21, 1354. (10) Itoh, K.; Ohe, C.; Tsurumaru, T.; Yasukawa, S.; Yamaguchi, T.; Kasuya, G. Vib. Spectrosc. 2002, 29, 197. (11) Lim, J. K.; Kim, Y.; Lee, S. Y.; Joo, S. W. Spectrochim. Acta A 2008, 69, 286. (12) Sarkar, J.; Chowdhury, J.; Ghosh, M.; De, R.; Talapatra, G. B. J. Phys. Chem. B 2005, 109, 12861. (13) Venanzi, M.; Bocchinfuso, G.; Gatto, E.; Palleschi, A.; Stella, L.; Formaggio, F.; Toniolo, C. ChemBioChem 2009, 10, 91. (14) Mano, M.; Henriques, A.; Paiva, A.; Prieto, M.; Gavilanes, F.; Sim~oes, S.; Pedroso de Lima, M. C. J. Pept. Sci. 2007, 13, 301. (15) Bender, V.; Ali, M.; Amon, M.; Diefenbach, E.; Manolios, N. J. Biol. Chem. 2004, 279, 54002. (16) Mihailescu, D.; Smith, J. C. Biophys. J. 2000, 79, 1718. (17) Brasseur, R.; Deleu, M.; Mingeot-Leclercq, M. P.; Francius, G.; Dufrened, Y. F. Surf. Interface Anal. 2008, 40, 151. (18) Orioni, B.; Bocchinfuso, G.; Kim, J. Y.; Palleschi, A.; Grande, G.; Bobone, S.; Park, Y.; Kim, J. I.; Kahm, K. S.; Stella, L. Biochim. Biophys. Acta 2009, 1788, 1523. (19) De Ioannes, P.; Moltedo, B.; Silva, H.; Pacheco, R.; Faunes, F.; De Ioannes, A. E.; Becker, M. I. J. Biol. Chem. 2004, 279, 26134. (20) Becker, M. I.; Fuentes, A.; Del Campo, M.; Manubens, A.; Nova, E.; Oliva, H.; Faunes, F.; De Ioannes, A. E.; Valenzuela, M. A.; Campos-Vallette, M.; Aliaga, A.; Ferreira, J.; De Ioannes, P.; Moltedo, B. Int. Immunopharmacol. 2009, 9, 330. (21) Leopold, N.; Lendl, B. J. Phys. Chem. B 2003, 107, 5723. (22) Leyton, P.; G
omez-Jeria, J. S.; Sanchez-Cortes, S.; Domingo, C.; Campos-Vallette, M. M. J. Phys. Chem. B 2006, 110, 6470. (23) Aliaga, A. E.; Osorio-Roman, I.; Leyton, P.; Caniulef, C.; C
elis, F.; Diaz Fleming, G.; Clavijo, E.; G
omez-Jeria, J. S.; Campos-Vallette, M. M. J. Raman Spectrosc. 2009, 40, 164. (24) Aliaga, A. E.; Osorio-Roman, I.; Garrido, C.; Leyton, P.; C
arcamo, J.; Clavijo, E.; G
omez-Jeria, J. S.; Diaz Fleming, G.; Campos-Vallette, M. M. Vib. Spectrosc. 2009, 50, 131. (25) Aliaga, A. E.; Garrido, C.; Leyton, P.; Diaz Fleming, G.; Clavijo, E.; Campos-Vallette, M. M.; Sanchez-Cortes, S. Spectrochim. Acta A 2010, 76, 458. (26) Hypercube Inc.: 1115 NW 4th Street, Gainesville, Fl. 32601, USA. (27) G
omez-Jeria, J. S. J. Comput. Theor. Nanosci. 2009, 6, 1361. (28) Leyton, P.; C
ordova, I.; Lizama-Vergara, P. A.; G
omez-Jeria, J. S.; Aliaga, A. E.; Campos-Vallette, M. M.; Clavijo, E.; García-Ramos, J. V.; S
anchez-Cort
es, S. Vib. Spectrosc. 2008, 46, 77. (29) Berendsen, H. J. C.; Postma, J. P. M.; Van Gunsteren, W. F.; Hermans, J. In Intermolecular Forces; Pullman, B., Ed.; The Netherlands, 1981, p 331. (30) Hess, B.; Van der Spoel, D.; Lindahl, E. J. Chem. Theory Comput. 2008, 4, 435. (31) Humphrey, W.; Dalke, A.; Schulten, K. J. Mol. Graphics Modell. 1996, 14, 33.

ARTICLE

(32) Van Gunsteren, W. F.; Berendsen, H. J. C. Angew. Chem., Int. Ed. Engl. 1990, 29, 992.Van Gunsteren, W. F.; Berendsen, H. J. C. Biomos Nijenborgh 4, 9747AG, Groningen, The Netherlands, 1995. (33) Hess, B.; Bekker, H.; Berendsen, H. J. C.; Fraaije, J. G. E. M. J. Comput. Chem. 1997, 18, 1463. (34) Miyamoto, S.; Kollman, P. A. J. Comput. Chem. 1992, 13, 952. (35) Darden, T.; York, D.; Pedersen, L. J. Chem. Phys. 1993, 98, 10089. (36) Essmann, U.; Perera, L.; Berkowitz, M. L. Langmuir. 1995, 11, 4519. (37) Berendsen, H. J. C.; Postma, J. P. M.; Van Gunsteren, W. F.; Dinola, A.; Haak, J. R. J. Chem. Phys. 1984, 81, 3684. (38) Ngola, S. M.; Zhang, J.; Mitchell, B. L.; Sundararajan, N. J. Raman Spectrosc. 2008, 39, 611. (39) Stewart, S.; Fredericks, P. M. Spectrochim. Acta A 1999, 55, 1641. (40) Diaz Fleming, G.; Finnerty, J. J.; Campos-Vallette, M. M.; C
elis, F.; Aliaga, A. E.; Fredes, C.; Koch, R. J. Raman Spectrosc. 2009, 40, 632. (41) De Gelder, J.; De Gussem, K.; Vandenabeele, P.; Moens, L. J. Raman Spectrosc. 2007, 38, 1133. (42) Moskovits, M. Rev. Mod. Phys. 1985, 57, 783. (43) Di Costanzo, L.; Flores, L. V., Jr.; Christianson, D. W. Proteins: Struct. Funct. Bioinf. 2007, 65, 637. (44) Tieleman, D. P.; Van der Spoel, D.; Berendsen, H. J. C. J. Phys. Chem. B 2000, 104, 6380.

3989

dx.doi.org/10.1021/jp1107153 |J. Phys. Chem. C 2011, 115, 3982–3989