Structure and Binding of Specifically Mutated Neurotensin Fragments

May 6, 2011 - Structure and Binding of Specifically Mutated Neurotensin Fragments on a Silver Substrate: Vibrational Studies. Edyta Podstawka-Proniewi...
0 downloads 0 Views 4MB Size
ARTICLE pubs.acs.org/JPCB

Structure and Binding of Specifically Mutated Neurotensin Fragments on a Silver Substrate: Vibrational Studies Edyta Podstawka-Proniewicz,*,† Andrzej Kudelski,‡ Younkyoo Kim,§ and Leonard M. Proniewicz† †

Faculty of Chemistry, Jagiellonian University, ul. Ingardena 3, 30-060 Krakow, Poland Faculty of Chemistry, University of Warsaw, ul. L. Pasteura 1, 02-093 Warsaw, Poland § Department of Chemistry, Hankuk University of Foreign Studies, Yongin, Kyunggi-Do 449-791, Korea ‡

ABSTRACT: Here, we report a systematic study showing an analogy between the activities of peptide structural component interactions with both a metal substrate and a G-proteincoupled seven-transmembrane receptor. In the present work, N-terminal fragments of human neurotensin (NT), NT16, NT18, and NT111, and C-terminal fragments of human neurotensin, NT813 and NT913, as well as six specifically mutated analogues with the following modifications, AcetylNT 813 , [Dab 9 ]NT 813 , [Lys 8 ,Lys 9 ]NT 813 , [Lys 8 -(Ò)Lys 9 ]NT 813 , [Lys 9 ,Trp 11 ,Glu 12 ]NT 813 , and Boc[Lys 9 , Leu13OMe]NT913, were immobilized onto an electrochemically roughened silver electrode surface in an aqueous solution. The orientation of the adsorbed molecules and the adsorption mechanism were determined from surface-enhanced Raman scattering (SERS) spectra. A comparison was made between the structures of the mutated fragments to determine how changes in the mutation of the structure influenced the adsorption properties. The contribution of the structural components to the peptides’ ability to interact with the NTR1 receptor was correlated with the SERS patterns. The SERS spectra revealed that the substitution of native amino acids in the investigated peptides slightly influenced their adsorption state on an electrochemically roughened silver surface. Thus, all of the investigated peptides, excluding [Lys9,Trp11,Glu12]NT813, tended to adsorb to the surface mainly via the oxygen atom of the deprotonated phenol group, and the phenyl ring became rearranged in a slightly different edge-on manner (NT18, NT111, NT813, Acetyl-NT813, [Dab9]NT813, [Lys8,Lys9]NT813, [Lys8-(Ò)-Lys9]NT813, NT913, and Boc[Lys9, Leu13OMe]NT913) or in an almost horizontal manner (N16) of the tyrosine residue. Meanwhile, [Lys9,Trp11,Glu12]NT813 bound to this substrate through the tilted phenyl coring of the tryptophan residue. Small changes in the enhancement of the CCNH2, COO, and CONH group modes upon adsorption, which were consistent with the adsorption of these peptides, also occurred (with slightly different strengths) through the nitrogen and oxygen lone pair of electrons in these groups. However, for NT18, a greater preferential interaction between the guanidine group of Arg8 and the roughened silver substrate was observed in comparison to that between the guanidine moiety of the other investigated peptides and the substrate. Vibrational spectroscopy was also used to produce an extensive table of Raman and absorption infrared spectra to allow for a rapid and accurate structural determination of these biomolecules and to allow the reader to easily follow the proposed SERS assignments.

’ INTRODUCTION Pancreatic, colon, and pituitary adenocarcinomas, breast, prostate, and small cell lung cancers, Ewing’s sarcoma, and meningiomas are common human cancers that grow rapidly, disseminate early, and frequently occult metastases.14 Signatures of these cancers include both overexpression and oversecretion of NTR1 (neurotensin subtype-1 G-protein-coupled single seven-transmembrane domain) regulatory receptors.5,6 These signatures of overexpression and oversecretion allow tumor cells to stimulate their own growth and the growth of other neoplastic cells.7 Therefore, these receptors have been proposed as new tumor markers because they are poorly expressed or absent in normal cells.8,9 The biological activity of NTR1 is mediated by its naturally occurring ligand, neurotensin, a 13-aa (NT, pGlu1-Leu2-Tyr3-Glu4-Asn5-Lys6-Pro7-Arg8-Arg9Pro10-Tyr11-Ile12-Leu13 OH, where pGlu denotes 5-oxo-proline; r 2011 American Chemical Society

all amino acids exist in the L conformation) neurotransmitter originally isolated from bovine hypothalami and heterogeneously distributed in both the central nervous system and the gastrointestinal tract.10,11 Understanding the mechanisms that underlie the processes of NT would open up considerable opportunities for the design and development of drugs to control NTR1 receptor expression or disease progression and for the use of prognostic and predictive biomarkers/fluorescent probes, on a case-by-case basis, for tumor imaging. The combined efforts of chemists and biologists have renewed the search for carriers or prodrugs for the delivery of cytotoxic drugs in molecular systems, radiopharmaceuticals for imaging, Received: February 9, 2011 Revised: April 11, 2011 Published: May 06, 2011 7097

dx.doi.org/10.1021/jp201316n | J. Phys. Chem. B 2011, 115, 7097–7108

The Journal of Physical Chemistry B leads, and noninvasive methods to improve preoperative staging.1215 The search for new diagnostic and treatment tools has led to the idea of using peptides that possess selective agonist properties and the ability to interact strongly with NTR1 and overcome the delivery problems associated with poor oral bioavailability and central nervous system (CNS) penetration. Molecules based on N- and C-terminal fragments were among both the nonpeptide and peptide forms of the first NT analogues that were widely investigated.1618 The N-terminal sequence was found to be a very weak contributor to NT biological activity. However, modification of the N-terminus by each of the functional groups completely prevents cleavage at the 89 amide bond.19 It has been widely proven that the active C-terminal fragment in the NT sequence, which mimics the effects of the full-length NT and interacts with NTR1 with high affinity,20 lies between Arg at position 8 and Leu at position 13 and is commonly referred to as NT813.2123 However, NT813 has a longer half-life (10 min) in plasma than NT, but the peptide degrades rapidly in vivo by peptidases.24 This instability has been attributed to rapid enzymatic cleavage of three major sites, the Arg8-Arg9, Pro10-Tyr11, and Tyr11-Ile12 bonds, which leads to neurotensin inactivation.25,26 Therefore, various NT modifications have been designed and introduced to protect these three bonds, including reduced peptide bond analogues of NT813, especially those with the Arg8-Arg9 bond removed,27 and to enhance the physiochemical parameters of the peptide to maintain receptor binding.28 It was originally hypothesized that substitution of the Arg8 side chain with L-Lys in NT813 would favor the ion pairing versus the peptide solvated state, thus resulting in peptides with increased overall lipophilicity that ultimately could translate into higher in vivo activity.16,29,30 NT913 fragment studies have also suggested the requirement of at least one basic residue. The Pro10 and Tyr11 residues are essential for the function of NT in stimulating intracellular cyclic GMP formation in murine neuroblastoma cells (N1E115).31 Additionally, Pro10 has received a special status as a key spacer unit. Arg8 and Arg9 play an important role in receptor recognition because the replacement of these residues by D-Arg and D-Lys, respectively, leads to a reduction in agonistic properties.32 Important to the success of this work is the systematic retroinverso (D-isomer) substitution made in the NT sequence, providing more potent analogues, perhaps through resistance to enzymatic degradation and/or better binding and stabilization of the tertiary structure.7 Thus, neurotensin analogues, in which the tyrosine residue in position 11 is replaced by a D-amino acid, have been synthesized.33,34 Additionally, substitution at positions 8, 9, and 11 by D-amino acids strongly decreases the inhibitory potency of neurotensin. The lipophilic N-terminal Boc-protecting group has been reported to increase the molecule’s hydrolytic stability and improve penetration of the blood brain barrier.35 Finally, a free C-terminal carboxylate also appears to be essential for high receptor affinity.25 Previously, by using surface-enhanced Raman scattering (SERS), we determined the vibrational structure and adsorption mechanism of bombesin and its analogues on silver, gold, and copper surfaces, which underlined an analogous contribution of the bombesin structural components to its ability to interact with both a bombesin-preferred metabotropic seven-transmembrane G-protein-coupled receptor superfamily, rGRP-R, and a metallic surface. We have also examined changes in this adsorption process with the applied electrode potential and the type of mutations in the bombesin neuropeptide, its C-terminal length

ARTICLE

(BNX14) and 614 amino acid length (BN614) fragments, and related peptides.3642 The observed SERS signals correlate well with the contribution of the structural components of these peptides to their ability to interact with rGRP-R. Recently, for human, frog, and pig neurotensin and human single-site mutants of NT, we also correlated the contribution of the structural components to their ability to interact with NTR1 using SERS patterns.43 This correlation has been done on the strength of the comparison of the information gathered from the biological studies, which have shown which amino acids from the peptide backbone were involved in the peptidereceptor interaction, with the information received from SERS studies that have shown which amino acids from the peptide backbone interacted with a metal surface. Thus, we have shown for the system investigated in this work that the same set of the amino acid residues (peptide fragments) that are suggested to interact with the receptor also binds to the electrochemically roughened silver substrate. On the other hand, those that are suggested not to have direct influence on the substratereceptor binding are also not active in SERS, that is, modes of these residues are not seen in the SERS spectra. The present spectroscopic study encompasses the previously described molecules based on N- and C-terminal neurotensin fragments, such as the 16 (NT16), 18 (NT18), and 111 (NT16) N-terminal fragments of neurotensin sequence and eight C-terminal fragments of neurotensin with the following mutations, NT813, Acetyl-NT813, [Dab9]NT813, [Lys8, Lys9]NT813, [Lys8-()-Lys9]NT813, [Lys9,Trp11,Glu12]NT813, NT913, and Boc[Lys9,Leu13OMe]NT913. Table 1 presents the amino acid sequences of these neuropeptides. The implications of this work for understanding the mechanisms of both adsorption at solid/liquid interfaces and substrate/receptor binding3642 could be invaluable. It is clear that, with a sufficiently sensitive and selective SERS method, the peptide behavior at the solid/ liquid interface and changes in this behavior due to specific mutations can be detected in response to specific peptide regions near or on a metal surface. The peptides in these specific regions directly interact with the metal surface and respond to perturbations in these regions caused by chain modifications. This interaction is believed to be of great significance in the context of nanobiomedicine, protein screening, and other therapeutic applications and, hence, should be fully understood. Such an approach has already been adopted by numerous groups investigating physiologically active molecules.44,45 This occurred because in the presence of a solid surface, the process of protein adsorption is often energetically favorable and the adsorption of proteins does not affect their binding capabilities, which implies that their structures are not strongly perturbed on the surface. The amino acid composition and sequence of these regions usually determine the adsorption mechanism of peptides on given metal surfaces. Therefore, analysis of the SERS signal (enhancement, broadness, and wavenumber) coming from constituents’ amino acids of the peptide sequence is useful for understanding possible ways in which a peptide can interact with the surrounding medium, such as how a peptide binds at a solid/liquid interface.3645 One disadvantage of SERS is the difficulty encountered in band assignment because changes in signal enhancement can be very dramatic. Due to adsorbate interactions with metal surfaces, certain bands that are strong in conventional Raman may not be present in SERS and vice versa, that is, weak bands in ordinary Raman spectra may clearly appear in SERS. Although the interpretation, as discussed above, can be 7098

dx.doi.org/10.1021/jp201316n |J. Phys. Chem. B 2011, 115, 7097–7108

The Journal of Physical Chemistry B

ARTICLE

Table 1. Amino Acid Sequences of N-Terminal Fragments of Neurotensin, NT16, NT18, and NT111, and C-Terminal Fragments of Neurotensin, NT813, Acetyl-NT813, [Dab9]NT813, [Lys8,Lys9] NT813, [Lys8-(Ò)-Lys9]NT813, [Lys9,Trp11, Glu12]NT813, NT913, and Boc[Lys9,Leu13OMe]NT913 a

a

The * in the table represent common amino acids in the sequence of all of the investigated peptides.

difficult, SERS is governed by a surface selection rule that states that the vibrations with large tensor components oriented along the vertical axis to the metal surface will be the most enhanced.4650 Therefore, specific information regarding the packing and orientation (adsorbed structure) of biophysical molecules on the metal surface can be gained by utilizing SERS. This study also attempted to correlate the observed changes with changes in biological activity due to the unique ability of these peptides to interact with NTR1. The SERS technique was used to address all of these points. In addition, to support the SERS spectral interpretation, vibrational assignments of infrared and Raman spectra are presented as well.

’ EXPERIMENTAL SECTION Neurotransmitters. N-terminal fragments of neurotensin, NT16, NT18, and NT111, and C-terminal fragments of neurotensin, NT813, Acetyl-NT813, [Dab9]NT813, [Lys8, Lys9]NT813, [Lys8-(Ò)-Lys9]NT813, [Lys9,Trp11,Glu12]NT813, NT913, and Boc[Lys9,Leu13OMe]NT913, were purchased at Bachem Co., Switzerland. Their purity and chemical structures were proven by means of the 1H and 13C NMR spectra (Bruker Avance DRX 300 MHz spectrometer) and electrospray mass spectrometry (Finnigan Mat TSQ 700). FT-IR Measurements. Thin palettes containing ∼0.5 mg of each peptide dispersed in 200 mg of KBr were used for the infrared measurement. The spectra were recorded at room temperature as an average of 30 scans using a Bruker infrared spectrometer (model EQUINOX 55) equipped with a Nernst rod as the excitation source and a DTGS detector in the 4004000 cm1 range with the spectral resolution of 4 cm1.

FT-Raman Measurements. FT-Raman measurements were performed for samples placed on a glass plate. FT-Raman spectra were recorded on a Nicolet spectrometer (model NXR 9650) combined with a liquid-nitrogen-cooled germanium detector. Typically, 1000 scans were collected with a resolution of 4 cm1. Excitation at 1064 nm was used from a continuum-wave Nd3þ: YAG laser. SERS Measurements. To obtain a sufficiently enhanced intensity of the SERS bands, the silver substrates were electrochemically roughened before peptide adsorption. Roughening was carried out in a conventional three-electrode cell with a large platinum sheet as the counter-electrode and an Ag/AgCl (1 M KCl) electrode as a reference (all potentials are quoted versus this electrode). The silver was roughened by three successive negativepositivenegative cycles in a 0.1 M KCl aqueous solution from 0.3 to 0.3 V at a sweep rate of 5 mV s1. The cycling was finished at 0.3 V, and then, the silver electrode was maintained at 0.4 V for 5 min. Afterward, the working electrode was removed at an open-circuit potential and carefully rinsed with water. Figure 1 provides an overview of the morphology of the roughened silver substrates used in this work. Raman spectra were recorded with an ISA T64000 (Jobin Yvon) Raman spectrometer equipped with Kaiser SuperNotchPlus holographic filters, a 600-groove/mm holographic grating, an Olympus BX40 microscope with a 50 long-distance objective, and a 1024  256 pixel liquid-nitrogen-cooled CCD detector. A Laser-Tech model LJ-800 mixed argon/krypton laser provided the excitation radiation at 514.5 nm. The laser power at the sample was set to 1 mW (∼104 W/cm2). The final sample concentration was ∼104 M. 7099

dx.doi.org/10.1021/jp201316n |J. Phys. Chem. B 2011, 115, 7097–7108

The Journal of Physical Chemistry B

ARTICLE

Figure 3. FT-IR spectra of N-terminal fragments of neurotensin in the spectral range of 1750550 cm1. Figure 1. Scanning electron microscopy (SEM) images of the electrochemically roughened silver substrate used in this work. Measurement conditions: A - 15.0kV  100k SE, scale 500 nm; B - 15.0kV  20k SE, scale 20 μm.

Figure 2. FT-Raman spectra of N-terminal fragments of neurotensin in the spectral range of 1750550 cm1.

’ RESULTS AND DISCUSSION FT-Raman and FT-IR Studies. The FT-Raman and FT-IR spectra of solid N-terminal fragments of neurotensin, NT16, NT18, and NT111, are shown in Figures 2 and 3, respectively, for the wavenumber range of 1700550 cm1. Meanwhile, Figures 4 and 5 present FT-Raman and FT-IR spectra, respectively, of solid C-terminal neurotensin fragments, NT813, Acetyl-NT813, [Dab9]NT813, [Lys8,Lys9]NT813, [Lys8-(Ò)Lys9]NT813, [Lys9,Trp11, Glu12]NT813, NT913, and Boc[Lys9,Leu13OMe]NT913 collected in the same spectral range. Because it has been well established and accepted that in an aqueous solution many proteins do not change their secondary structures,51 except in extreme pH conditions, we present solid-state measurements.

Figure 4. FT-Raman spectra of C-terminal fragments of neurotensin in the spectral range of 1750550 cm1.

The values of these band positions are given in inverse centimeters together with the proposed assignments and are listed in Tables 2 and 3. For most of the examined peptides, we observed similar characteristic absorptions. These are illustrated by the example 7100

dx.doi.org/10.1021/jp201316n |J. Phys. Chem. B 2011, 115, 7097–7108

The Journal of Physical Chemistry B

Figure 5. FT-IR spectra of C-terminal fragments of neurotensin in the spectral range of 1750550 cm1.

of NT813 in Figure 5, where the FT-IR spectrum (upper trace) is compared to those of the C-terminus specifically modified fragments. It is apparent that the N-terminal 16 fragment’s FTIR spectrum presented in Figure 3 also exhibited similar spectral features. These included peaks at 16961642, 15571532, ∼1518, ∼839, ∼802, and ∼723 cm1 (see Tables 2 and 3 for accurate wavenumbers). Notably, the ∼1203 , ∼1185, ∼1183, ∼839, ∼802, and ∼723 cm1 bands markedly decreased in intensity when going from NT16 (Figure 2), NT813, [Dab9]NT813, [Lys8,Lys9]NT813, [Lys8-(Ò)-Lys9]NT813, [Lys8,Trp11,Glu12]NT813, and NT913 to Boc[Lys9, Leu13OMe]NT913 and did not appear in the Acetyl-NT813 (Figure 5), NT18, or NT111 FT-IR spectra (Figure 3). However, several additional absorption peaks at 1083, 1023, 931, 889, and 874 cm1 were found in the range of 1100850 cm1 in the FT-IR spectrum of [Lys8-Lys9]NT813 (Figure 5). Tables 2 and 3 give the proposed detailed allocation of the aforementioned bands to the normal coordinates. An exception was an amide I mode (16961642 cm1) commonly used to estimate the secondary structure of investigated peptides and proteins. This normal mode is not an entirely localized vibration, which is clearly shown by the asymmetric shape of the amide I envelope. Hence, the FT-IR spectra in the amide I region clearly indicate the existence of two well-defined sub-bands at 16851668 (typeII turn) and 16521642 cm1 (random coil), whereas only the component at 16691665 cm1 was seen in the FT-Raman spectra. The above-mentioned findings are very similar to those

ARTICLE

found in the NMR solid-state and solution structures of the 813 neurotensin fragment bound to a membrane-mimetic environment and the NTR1 receptor.5254 The structural similarity of the FT-Raman spectra of the N-terminal NT16, NT18, and NT111 fragments (Figure 2), which mainly differed in the presence of a strong, narrow, and symmetric amide I band (at 1669 cm1) for NT16, should be noted. Likewise, the FT-Raman spectra of the C-terminal fragments of neurotensin, NT813, Acetyl-NT813, [Dab9]NT813, [Lys8,Lys9]NT813, [Lys8-(Ò)-Lys9]NT813, NT913, and Boc[Lys9,Leu13OMe]NT913, exhibited a similar spectral fingerprint, with the exception of two bands at 886 and 874 cm1 (ν(CC), δoop(CNH2), and/or Fr(CH2)) for [Lys8, Lys9]NT813 (Figure 4). This was not surprising in light of previous studies that have shown that modes due to aromatic amino acid vibrations dominate the Raman spectra.3643,55,56 Therefore, the characteristic modes of the tyrosine residue (Tyr3 and/or Tyr11) (for example, at 1615, 1598, 13191242, 1207, 1174, 857/828, and 642 cm1 for NT16) or the tryptophan residue (Trp11) (at 1617, 1598, 1549, 1433, 1356, 1162, 994, and 757 cm1 ) (see Table 3 for band assignments) in the case of [Lys9,Trp11,Glu12]NT813 were predominantly enhanced in the FT-Raman spectra of the peptides investigated in this work (Figures 2 and 4). Because the bands of these characteristic modes have been defined, they are not described in detail here; they are listed together with their observed wavenumbers in Tables 2 and 3. Other bands that matched correctly characterized vibrations belonged to the aliphatic Asn, Glu, Gln, Lys, and Arg side-chain vibrations and showed significant absorption/enhancement. These vibrations produced a strong asymmetric broad Raman band at 14491435 cm1 (Figures 2 and 4) and a medium-weak absorption band near 14661436 cm1 (Figures 3 and 5). They were mainly due to deformations of the CH2 groups (δ(CH2)). The methylene group also gave rise to the ∼1341, ∼1318, and 884 cm1 Raman spectral features that were assigned to wagging, twisting, and rocking vibrations, respectively (see Tables 2 and 3 for detailed band assignments). An alternative contribution to the vibrational spectra from the aforementioned nonaromatic amino acids came from the amino (16421628, 11391122, and 1095 1031 cm1) and carbonyl (14061382, ∼722, ∼668, and ∼615 cm1) groups of the polypeptide side chains (see Tables 2 and 3). SERS Studies. Figures 6 and 7 display, for the first time, the SERS spectra of the N-terminal NT16, NT18, and NT111 fragments and C-terminal NT813, Acetyl-NT813, [Dab9]NT813, [Lys8,Lys9]NT813, [Lys8-(Ò)-Lys9]NT813, [Lys9, Trp11,Glu12]NT813, NT913, and Boc[Lys9,Leu13OMe]NT913 fragments, respectively, deposited onto an electrochemically roughened silver substrate in the spectral range of 1700550 cm1. Some general observations can be made concerning these spectra. The [Lys9,Trp11,Glu12]NT813 SERS spectrum (Figure 7) exclusively exhibited the characteristic Raman bands due to the modes of the Trp11 residue vibrations, whereas the modes of the Tyr residue were predominantly enhanced in the SERS spectra of the remaining peptides investigated in this work, excluding the NT18 fragment, for which the most intense SERS signals were due to the Arg8 residue. The bands found at ∼1500 (Y19/W19), ∼1394 (νs(COO)), in the range of 13001210 (Y/W), ∼1145 (νas(CCN)), and 1003 cm1 (Y/W16) strengthened noticeably relative to both the corresponding Raman bands and other SERS signals. These changes indicate that the strongest interaction occurred between the Tyr/Trp, 7101

dx.doi.org/10.1021/jp201316n |J. Phys. Chem. B 2011, 115, 7097–7108

The Journal of Physical Chemistry B

ARTICLE

Table 2. Wavenumbers and Band Assignments for FT-IR, FT-Raman, and SERS Spectra of N-Terminal Fragments of Neurotensin, NT16, NT18, and NT111 a wavenumbers (cm1) NT16 assignment AI

NT18

NT111

FT-IR

FT-Raman

SERS

FT-IR

FT-Raman

SERS

FT-IR

FT-Raman

SERS

1668

1669

1637

1652

1667

1638

1652

1665

1637

1615

1621 1598

1615

1621 1607

1598

1578

1596

1590

1642 Fs(NH2) in N, R, and K Y8a Y8b νaa(COO) in E and/or C-terminal L, δ(CβH2)Y,

1544

1550

1626 1614 1596

1559

1556

1542

1543

1590 1559

ν(CC)Y, and/or AII Y19

1517

Fs(CβH2)Y and/or δ(CH2)

1439

1438

Y19b and/or δ(CH2) ν(CC) and νs(COO) in E and/or C-terminal L

1502

1518

1446

1448

1428 1392

1401

1498

1517

1447

1448

1447

1447

1447

1400

1433 1392

1401

1402

1431 1394

1339

1350

1315

1316

1289

1264

1269

guanido group of R

1364

Fw(CH2)

1347

ν(CC)Y, δ(CβH2)Y, and/or Ft(CH2) ν(CO)Y and δ(CH)Y ν(CO) Y, Fb(COH)Y, Ft(CH2)Y, and/or AIII

1498

1242

1319

1292

1254

1268

1229

1222

1316

1316

1249

1244

1287 1270 1236

1246

1246

1221 Y7a Y9a (Fb(CH)), ν(CC) and/or Fr(CβγδH2)

1202 1185

1207 1174

1197

νas(CCN) and/or Ft(NH2) in N, R, and/or K

1139

1143

1145

ν(CO)Y and/or ν(CC)trans alkyl chain

1173

1207 1174

1206 1173 1140

1122

1128

guanido group of R

1239 1221

1130

1143 1124

1113

1106

Ft(NH2)

1073

ν(CRN)

1054

ν(CRN)

1031

Y13 ν(CN) and/or Fb(NH2)

995

Y (ring bend)

948

1089

1032

1043

1004

1073

1095

1034 1002

985

1074

1049 1008 978

1004

ν(CC=O), ν(CC), and/or Y (ring bend)

930

922

923

921

ν(CC), δoop(CNH2), and/or Fr(CH2)

883

886

885

886

885

886

857

852

854

852

852

852

828

822

830

832

828

832

710

713

714

Y (Fermi resonance)

832

Fb(CH)Y, ν(CC), Fr(CH2), νs(CCN),

800

Fb(COOH), and/or Fb(OH) δ(COO) in E and/or C-terminal L

722

722

Fr(COOH) Y6b

668 642

Fw(COO) Y (ring bend) and/or AVI

646

642

615 597

642 615

614

596

Abbreviations: Y, R, N, E, L, and K are tyrosine, arginine, asparagine, glutamic acid, leucine, and lysine residues; ν is stretching; δ is deformation; and for vibrations, Fw is wagging, Fb is bending, Ft is twisting, Fr is rocking, Fs is scissoring, oop is out-of-plane, s is symmetric, and as is antisymmetric. a

CNH2, and COO moieties and the silver substrate. A more detailed analysis and interpretation of these spectral changes observed for particular moieties will be discussed below. Further information regarding the normal mode assignments, which were mainly based on very recent investigations of human, frog, and pig neurotensins and single-site mutants of human NT,43 is listed in Tables 2 and 3 together with the wavenumbers of the enhanced bands.

The most noteworthy contribution at ∼1500 cm1 (very strong, fwhm = 2022 cm1; fwhm denotes full width at half-maximum) was due to the aromatic ring CC stretching motions of the phenyl ring, namely, W19 for Trp or Y19 for Tyr. Its enhancement remained fairly constant among the spectra, except for that of the NT18 fragment, which had a relative intensity that was reduced by one-third. These point to the comparable strength of interaction 7102

dx.doi.org/10.1021/jp201316n |J. Phys. Chem. B 2011, 115, 7097–7108

FT-

FT-

FT-

FT-

FT-

FT-

Raman SERS IR

[Lys8-(Ò)-

[Lys9,Trp11,

FT-

1558 1554

1556

1614 1594 1598 1557

1650 1614 1594 1596

1617 1600 1614 1598 1547 1549 1547 1532 1551

1648

7103

Y or W16 ν(CC) Y/W (ring bend) and/or ν(CC) ν(CCCOO), ν(CC), and/or Y (ring bend) ν(CC), δoop(CNH2), and/or Fr(CH2)

Y19 or W19 Fs(CβH2)Y, W and/or δ(CH2) Y19b and/or δ(CH2) νa(COO) in E and/or C-terminal L

993 925

932

1207

987

1044

1185 1175 1138 1149 1125 1097

1203 1205

1001

921

921

1204 1207

1003 961 937 886 874

967 931 889 874

1017

1203 1207

675

1003

960 922

998

1203 1207

1003

970 937 896

994

1013

954 915 876

1004

1231 1203 1203 1206 1183 1162 1154 1173 1180 1174 1175 1174 1184 1174 1142 1136 1145 1139 1146 1145 1137 1147 1137 1149 1140 1138 1124 1129 1124 1133 1127 1133 1127 1132 1122 1125 1098 1097 1087 1094 1099 1098 1070 1042 1038 1023 1052 1042 1033 1042

1539

1615 1597 1586

1644

Y, W Fb(CH)W and/or ν(CC)W Y9a (Fb(CH)), ν(CC) and/or Fr(CβγδH2) νas(CCN) and/or Ft(NH2) in N, R, and/or K Fb(CH)W-phenyl ν(CO)Y, ν(CC)trans alkyl chain and/or Fb(CH)W Ft(NH2) of guanido group of R guanido group of R

1615 1596 1586

1681

W7 (Fermi resonance) and/or Fw(CH2) Fw(CH2) ν(CC)Y,W, δ(CβH2)Y,W, and/or Ft(CH2) ν(CO)Y, δ(CH)Y, or ν(CC)W-phenyl ν(CO)Y, Fb(COH)Y, Ft(CH2)Y,W, AIII, and/or W

1541 1518 1501 1517 1454 1461 1451 1450 1439 1435 1433 1387 1388 1403 1384

1556

1614 1597

1652

1540 1541 1497 1518 1500 1518 1501 1519 1502 1510 1516 1462 1446 1454 1463 1449 1457 1461 1449 1457 1461 1450 1456 1459 1466 1460 1447 1428 1440 1436 1431 1448 1430 1436 1437 1433 1437 1433 1439 1438 1402 1391 1391 1396 1376 1378 1396 1394 1396 1387 1402 1369 1382 1356 1360 1341 1352 1342 1340 1349 1339 1351 1336 1359 1348 1334 1352 1345 1326 1318 1317 1289 1318 1290 1322 1290 1292 1293 1299 1328 1267 1272 1267 1266 1270 1271 1266 1271 1251 1278 1265 1249 1229 1246 1251 1219 1245 1221 1247 1249 1222 1224 1216 1246 1247 1222

ν(CC)Y, and/or AII

Y8a or W8a Y8b or W8b W3 (pyrrole ν(C2dC3)) νaa(COO) in E and/or C-terminal L, δ(CβH2)Y,

1651

FT-

FT-

1533

1643 1614 1595

1004

1146

920

1049 1045

1126

1163 1173

1204 1207

1338 1312 1270 1269 1239 1249 1245

1349

998

1141

1345 1287 1266 1218

1497 1447 1428 1393

1583

1643

Raman SERS

1505 1516 1450 1449 1449 1430 1391 1391 1368

1610

Boc[Lys9, Leu13OMe]NT913

Raman SERS IR

FT-

NT913

1668 1668 1636 1679 1648

Raman SERS IR

FT-

Glu12]NT813 FT-

Raman SERS IR

FT-

Lys9]NT813

1686 1667

Raman SERS IR

[Lys8, Lys9]NT813

1642 1659 1636 1680 1667 1647 1680 1667 1643 1696 1669

Raman SERS IR

FT-

[Dab9]NT813

1746 1673 1668

Raman SERS IR

FT-

Acetyl-NT813

AI and Fs(NH2) in N, R, and K

IR

FT-

NT813

wavenumbers (cm1)

ν(CdO)

assignment

Table 3. Wavenumbers and Band Assignments for FT-IR, FT-Raman, and SERS Spectra of C-terminal Fragments of Neurotensin, NT813, Acetyl-NT813, [Dab9]NT813, [Lys8,Lys9]NT813, [Lys8-(Ò)-Lys9]NT813, [Lys9,Trp11,Glu12]NT813, NT913, and Boc[Lys9,Leu13OMe]NT913 a

The Journal of Physical Chemistry B ARTICLE

dx.doi.org/10.1021/jp201316n |J. Phys. Chem. B 2011, 115, 7097–7108

597

642 596

642

723 723 721

598 596

642

598 598

615 625

642

598

642

597

605

642

642

607

708 723 723

Y (ring bend) and/or AVI

Fw(COO)

Y6b/W

723

Fb(CH)Y, ν(CC), Fr(CH2), νs(CCN), Fb(COOH), 802

828 839

800

723

839

722

800

723

798

837

614

722

799

723

838

619

722

757 746

δ(COO) in E and/or C-terminal L

642

801 800 805 801

and/or Fb(OH) W18

850

828 836

839 849 838 838 848 838 849 839 848 839 850 849 839 ν(CC) and/or νs(CNC) secondary amide Y (Fermi resonance)

a Abbreviations: Y, W, R, N, E, L, and K are tyrosine, tryptophan, arginine, asparagines, glutamic acid, leucine, and lysine residues; ν is stretching; δ is deformation; and for vibrations, Fw is wagging, Fb is bending, Ft is twisting, Fr is rocking, Fs is scissoring, oop is out-of-plane, s is symmetric, and as is antisymetric.

ARTICLE

612

Raman SERS

FTFT-

Raman SERS IR

FTFT-

Raman SERS IR

FTFT-

Raman SERS IR

FTFT-

Raman SERS IR

FTFT-

Raman SERS IR

FTFTFTFTFT-

FT-

Acetyl-NT

NT

Raman SERS IR

Lys ]NT Lys ]NT [Dab ]NT

813 9 813 813

Raman SERS IR

NT Glu ]NT

813

[Lys9,Trp11,

12 813

[Lys8-(Ò)-

9 813

[Lys8,

9

wavenumbers (cm1) assignment

Table 3. Continued

IR

Leu OMe]NT913 13 913

Boc[Lys9,

The Journal of Physical Chemistry B

between the phenyl ring of Try or Trp of the NT16, NT111, and C-terminal fragments and the roughened silver substrate and to the weakening of the NT18 Tyr ring coordination to this substrate. The Y19 mode for the protonated Tyr residue was proposed to produce an uncommonly small signal at 15251518 cm1, whereas the deprotonated form at 14961486 cm1 was accompanied by the medium enhanced Y19b mode that, in the SERS spectra presented in Figures 6 and 7, was detected at ∼1430 cm1 (fwhm = 1518 cm1) (Tables 2 and 3).57 Thus, we believe that these two bands are associated with the deprotonated form of the Tyr residue. However, given the respective pKa value for the phenol hydroxy group of Tyr (pKa = 10.10) and the pH conditions of our experiments (pH ≈ 8.5), Tyr should possess a hydroxyl group. However, it would not be surprising for the positively charged silver surface at the open circuit to act as a Lewis acid and facilitate deprotonation of the tyrosine residue. Strikingly, there was a similar relative intensity ratio between these two bands and a SERS signal at ∼1450 cm1 in the spectra of all of the peptides deposited onto the roughened silver substrate, except for the NT18 fragment, where the 1448 and 1433 cm1 bands gained approximately 90% of the 1498 cm1 enhancement. The ∼1450 cm1 band (fwhm = 1620 cm1) was due to vibrations of the tyrosine/tryptophan β-methylene group.58 Either an overlap by the CH2 deformations of the other side chains or a change in the strength of the CH2/silver substrate interactions led to the detected shift in this band wavenumber (1451 T 1446 cm1) between the SERS spectra of the Tyr-containing peptides. This may be supported by a ∼1350 cm1 band (fwhm = 1517 cm1) caused by CH2 wagging motions that was slightly more intense in the SERS spectra than that in the corresponding Raman spectra. It should be ruled out here that the ∼1350 cm1 SERS signal in the NT18 spectrum in Figure 6 was masked by a strong envelope of bands showing two maxima at 1395 and 1362 cm1 that dominated the spectrum. The latter band was rendered an unequivocal sign of direct coordination of the guanidine group of Arg8 with the roughened silver substrate. In the spectral range of 13001210 cm1, the SERS spectra of the investigated peptides have four common spectral features. They show an atypically intense band at ∼1270 cm1, attributed to the ν(CC) mode of the Trp phenyl coring in the case of [Lys9, Trp11,Glu12]NT813 or to the ν(phenol-O) (Y7a) coupled with δ(CH) vibrations of deprotonated Tyr in the case of Tyrcontaining peptides (Tables 2 and 3). The peak had a comparatively narrow width (fwhm = 1416 cm1), which excludes the possibility of contributions from amide III motions to this band, as suggested in the case of Tyr homodipeptides adsorbed on a colloidal silver surface.56 In the case of the Tyr-containing peptides, this SERS signal is indicative of the strength of the interaction between the free-electron pair at the oxygen atom of the Tyr deprotonated hydroxyl group and the roughened silver substrate. The relative intensity was equal to that of the ∼1500 cm1 SERS signal for the N16 (Figure 6) and N813 (Figure 7) fragments, was decreased by one-tenth for N18, N111 (Figure 6), N913, and [Dab9]NT813, was two- to threefold lower for [Lys8-(Ò)-Lys9]NT813, [Lys9,Trp11,Glu12]NT813, Boc[Lys9,Leu13OMe]NT913, and Acetyl-NT813, and was slightly increased for [Lys8,Lys9]NT813 (Figure 7). The corresponding Raman enhancement (Figures 2 and 4) was not as obvious. It was assumed that these scattering variations were due to the distance effect of the phenol-O moiety with respect to the roughened silver substrate. Shoulders at the highwavenumber side of the ∼1270 cm1 band appeared at ∼1290 cm1 (fwhm = 2934 cm1). These shoulders involved 7104

dx.doi.org/10.1021/jp201316n |J. Phys. Chem. B 2011, 115, 7097–7108

The Journal of Physical Chemistry B

ARTICLE

Figure 6. SERS spectra of N-terminal fragments of neurotensin deposited onto an electrochemically roughened silver substrate in the spectral range of 1750550 cm1. Measurement conditions: ∼104 M; excitation wavelength: 514.5 nm; power at sample: ∼1 mW. Insets A and B show an expanded intensity scale.

out-of-plane stretching vibrations of the CC phenyl ring with a contribution from the δ(CβH2) and/or Ft(CH2) modes of Tyr/ Trp. The relative intensity changes of this band and a ∼1220 cm1 spectral feature (fwhm = 1315 cm1) (see Table 3 for assignment) could be used to monitor slight deviations in the parallel orientation of the Tyr/Trp phenyl ring with respect to the normal silver surface because it is well-known that out-of-plane modes of aromatic rings gain enhancement when they adopt an end-on orientation at the silver surface when compared to those for a vertical orientation.4649,59 Figures 6 and 7 indicate the following drop order in the relative intensity of the ∼1290 and 1220 cm1 bands: [Lys8,Lys9]NT813 > [Lys8-(Ò)-Lys9]NT813 = [Dab9]NT813 = N111 g [Lys9,Trp11,Glu12]NT813 g Boc[Lys9,Leu13OMe]NT91 = Acetyl-NT813 ≈ NT813 > NT16 > NT913. This may suggest the order of the tilted orientation of the Tyr aromatic ring. In other words, the tilt angle of the Tyr ring with respect to the surface normal is the highest for [Lys8,Lys9]NT813 and the lowest for NT913. Careful examination of the 1300 to 1210 cm1 strongly overlapping spectral range allowed for the extraction of one more band at ∼1240 cm1 (Figures 6 and 7) that was also observed in the corresponding Raman spectra (Figures 2 and 4). The Raman spectra exhibited remarkably similar scattering around this wavenumber for the investigated peptides deposited onto the

roughened silver substrate. We believe that this SERS spectral feature, barely detectable in some cases due to its overlapping with the ∼1220 cm1 band, arises from the amide III mode. This idea is supported by a scattering at 16501640 cm1, where the amide I mode (Figures 6 and 7, insets A) is observed. The presence of amide modes may indicate that the CONH unit(s) interact with the roughened silver substrate. Moreover, the relative intensity variations of this band observed between the SERS spectra of the investigated peptides either indicate that the amide group for these peptides was close or less parallel to the mean surface plane or that it was too close or too far from the surface to be well-enhanced. Other spectral features of the Tyr and Trp oscillations may serve as excellent markers for the presence of these residues in complex biopolymers at a metal surface and invite further discussion. These spectral features include the bands at 1013998 and ∼850/830 cm1 (Tables 2 and 3). Although both Tyr and Trp contain a phenyl ring, the medium-weak ∼1003 cm1 SERS signal (fwhm = 1115 cm1) was suppressed for the Tyr-containing peptides and shifted to 1013 cm1 (fwhm = 18 cm1) in [Lys9,Trp11,Glu12]NT813 due to the different substitution symmetry of the phenyl ring. The relative intensity of this band in the SERS spectra of the investigated peptides was high when compared to that in corresponding Raman spectra (negligible enhancement) and 7105

dx.doi.org/10.1021/jp201316n |J. Phys. Chem. B 2011, 115, 7097–7108

The Journal of Physical Chemistry B

ARTICLE

Figure 7. SERS spectra of C-terminal fragments of neurotensin deposited onto an electrochemically roughened silver substrate in the spectral range of 1750550 cm1. Measurement conditions: ∼104 M; excitation wavelength: 514.5 nm; power at sample: ∼1 mW. Insets A and B show an expanded intensity scale.

decreased going from [Lys9,Trp11,Glu12]NT813 and [Lys8, Lys9]NT813 to NT913, NT111, and NT18 to [Dab9]NT813, Acetyl-NT813, [Lys8-(Ò)-Lys9]NT813, and Boc[Lys9, Leu13OMe]NT913 to NT813 and NT16. A unique binding mode to the silver surface was observed for NT16. A weak intensity at 1004 cm1 and the appearance of a fairly strong 1032 cm1 feature accompanied by a signal at 646 cm1 and a doublet at 1598 and 1621 cm1 (see Table 2 for band assignments) suggest that this molecule bound to the electrode surface with a strongly tilted geometry. The reasons for such unexpected binding are not currently known. The ∼850/830 cm1 peaks were present only in the spectrum of the para-substituted phenyl ring. Hence, they should be enhanced only in the spectra of Tyr-containing peptides. Indeed, these bands were detected for Tyr3 in the NT16 and N18 SERS

spectra (Figure 6). However, in the case of the NT111 and all of the C-terminal fragments, the doublet appeared with negligible enhancement and was found exclusively in the expanded SERS spectra in Figures 6 and 7 (inset B). It seems that the basic requirement for the Tyr Fermi doublet (Fermi resonance between ring breath at 840 cm1 and out-of-plane ring bend overtone near 420 cm1), the accidental degeneracy or near degeneracy of a fundamental and an overtone of the same symmetry species, was not fulfilled for the Tyr residue at position 11. This anomaly could be attributed to unusual local environments of these residues in the NT111 and all of the C-terminal fragments adsorbed on the roughened silver substrate. Thus far, we have shown evidence for interactions of the Tyr and Trp ring, the oxygen atom of the deprotonated hydroxyl phenol unit, and the CH2 group of N-terminal NT16, 7106

dx.doi.org/10.1021/jp201316n |J. Phys. Chem. B 2011, 115, 7097–7108

The Journal of Physical Chemistry B NT18, and NT111 and C-terminal NT813, Acetyl-NT813, [Dab9]NT813, [Lys8,Lys9]NT813, [Lys8-(Ò)-Lys9]NT813, [Lys9,Trp11,Glu12]NT813, NT913, and Boc[Lys9,Leu13OMe] NT913 fragments with an electrochemically roughened silver substrate. Additionally, the findings reflect similar types of interactions for these peptides, with only slightly different orientations of the Tyr ring in the Tyr-containing fragments, on this substrate. This consistency indicates that the specific mutations introduced on the peptides did not significantly affect the orientation of these fragments on the electrochemically roughened silver substrate, as has been proposed for NTR1 (see Introduction). Several of the bands of the amine and carboxyl group vibrations were absent or that showed weak scattering in the Raman spectra of the N- and C-terminal neurotensin fragments was also present in the spectra of these molecules when chemisorbed onto the roughened silver electrode. However, their band wavenumbers and enhancement underwent small changes between the SERS spectra. For example, a weak band at 1073 cml in the SERS spectrum of the NT16 fragment was due to NH2 group twisting vibrations of Asn5 and/or Lys6, whereas for NT18 and NT111, this weak band may also be allocated to the Arg residue. In the case of C-terminal fragments immobilized on the roughened silver substrate, enhancement was negligible or absent. The asymmetric CCN stretching vibration was enhanced as a medium-intensity asymmetric SERS signal at ∼1145 cm1 (fwhm = 1216 cm1) in all of the SERS spectra presented here (see Tables 2 and 3 for detailed band wavenumbers). Neither of these bands was seen in the corresponding Raman spectra (Figures 2 and 4), and their SERS relative intensities changed only slightly in the order NT111 = NT813 g NT18 = [Dab9]NT813 = Boc[Lys9,Leu13OMe]NT913 = Acetyl-NT813 = [Lys8, Lys9]NT813 > NT16 = [Lys8-(Ò)-Lys9]NT813 = NT913 > [Lys9,Trp11,Glu12]NT813. These enhancement variations are indicative of a direct interaction (bonding) between the CCN molecular fragment of each investigated compound with the silver substrate, in which the CN bond axis is slightly tilted or perpendicular relative to the mean silver surface plane. Such an orientation of this fragment enforced large polarizability changes in this mode, explaining the prominent νas(CCN) mode SERS enhancement when compared to the Raman scattering. The N-terminal NT16, NT18, and NT111 fragments consisted of two carboxyl groups, of glutamic acid at position 4 of the sequence and of the terminal Lys6, Arg8, and Tyr11 residues under the pH conditions of the SERS experiment. All of the C-terminal fragments possessed at least one COO moiety that attached to the positively charged silver surface of the substrate, giving rise to the νs(COO) mode at ∼1394 cm1. Because there was no or negligibly detectable Raman signal for these peptides in the solid state, the SERS spectra of adsorbed peptides on the roughened silver substrate illustrate that the immobilization of the peptides resulted in a significant level of enhancement, the magnitude of which was lost in the following order: NT18 and [Lys9,Trp11,Glu12]NT813 > NT913 = [Lys8,Lys9]NT813 g NT111 = Acetyl-NT813 = [Lys8-(Ò)-Lys9]NT813 > NT16 = [Dab9]NT813 = Boc[Lys9,Leu13OMe]NT913 > NT813. These changes may reflect a decrease in the interaction strength between the carboxyl group and the silver substrate in the same direction as the scattering decreased.

’ CONCLUSIONS The structures of N-terminal fragments of human neurotensin, NT16, NT18, and NT8111, and C-terminal fragments

ARTICLE

of human neurotensin, NT813 and NT913, as well as six specifically mutated analogues with the following modifications, Acetyl-NT813, [Dab9]NT813, [Lys8,Lys9]NT813, [Lys8-(Ò)Lys9]NT813, [Lys9,Trp11,Glu12]NT813, and Boc[Lys9,Leu13OMe]NT913 immobilized onto an electrochemically nanometer-sized roughened silver substrate in an aqueous solution, were investigated. On the basis of the obtained data, adsorption mechanisms were proposed for each case, and suggested adsorbate structures were compared with each other. According to the differences determined in the enhancement of the exhibited features that could be used to identify the specific residues in the adsorbed peptide, we made the following conclusions: 1. The NT16, NT18, NT111, NT813, Acetyl-NT813, [Dab9]NT813 , [Lys8 ,Lys9]NT813, [Lys8-(Ò)-Lys9 ]NT813, NT913, and Boc[Lys9, Leu13OMe]NT913 fragments tended to adsorb to the surface substrate mainly via the oxygen atom of the deprotonated phenol group, and the phenyl ring was arranged in a slightly different edge-on manner for NT16 and NT18, Tyr11 for NT111, NT813, Acetyl-NT813, [Dab9]NT813, [Lys8,Lys9]NT813, [Lys8(Ò)-Lys9]NT813, NT913, and Boc[Lys9, Leu13OMe]NT913 or in an almost horizontal manner of the tyrosine residue for N16. 2. [Lys9,Trp11,Glu12]NT813 also bound to the roughened silver substrate through a tilted phenyl coring; however, it was bound through the tryptophan residue. 3. Small changes in the enhancement of the CCNH2, COO, and CONH group modes upon adsorption, which were consistent with the adsorption of these peptides, also occurred (with slightly different strengths) through the nitrogen and oxygen lone pair of electrons in these groups. However, for NT18, a greater preferential interaction between the guanidine group of Arg8 and the roughened silver substrate was observed in comparison to that between the guanidine moiety of the other investigated peptides and the substrate. Because no significant changes were observed in the SERS fingerprint regions of the investigated fragments, we suggest that the introduced mutations did not change the overall adsorption mechanism but only slightly modified the strength of the interaction, especially via changes in the binding geometry of the phenyl ring(s) with respect to the surface. The results of such a systematic study show a compelling analogy in the activities of a peptide’s structural components when interacting with a metal substrate and a G-protein-coupled seven-transmembrane receptor. This suggests, despite the simplicity of the model, the feasibility of using SERS spectroscopy to probe peptidemetal interactions that, to some extent, may mimic the mechanism of a substrate binding to its receptor. To make it clear, we show that for the investigated compounds, the same set of the amino acid residues that are suggested to interact with the receptor also binds to the silver surface. On the other hand, these parts of the peptides which are suggested not to have direct influence on the substratereceptor binding are also not active in SERS, that is, these modes are not seen in the SERS spectra.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Phone: þ48-12-663-2077. Fax: þ48-12-634-0515. 7107

dx.doi.org/10.1021/jp201316n |J. Phys. Chem. B 2011, 115, 7097–7108

The Journal of Physical Chemistry B

’ ACKNOWLEDGMENT This work was supported by the State Department for Scientific Research of the Ministry of Science and Higher Education (Grant No. N N204 159136 to E.P.). Y.K. gratefully acknowledges HUFS for financial support. The authors are grateful to Prof. Maria Janik-Czachor and Dr. Marcin Pisarek from the Institute of Physical Chemistry, Polish Academy of Sciences, for the SEM images of the roughened silver surface. ’ REFERENCES (1) Souaze’, F.; Dupouy, S.; Viardot-Foucault, V.; Bruyneel, E.; Attoub, S.; Gespach, C.; Gompel, A.; Forgez, P. Cancer Res. 2006, 66, 6243. (2) Pelosi, G.; Volante, M.; Papotti, M.; Sonzogni, A.; Masullo, M.; Viale, G. Q. J. Nucl. Med. Mol. Imaging 2006, 50, 272. (3) Carraway, R. E.; Plona, A. M. Peptides 2006, 27, 2445. (4) Evers, B. M. Peptides 2006, 27, 2424. (5) Thomas, R. P.; Hellmich, M. R.; Townsend, C. M., Jr.; Evers, B. M. Endocr. Rev. 2003, 24, 571. (6) Alshoukr, F.; Rosant, C.; Maes, V.; Abdelhak, J.; Raguin, O.; Burg, S.; Sarda, L.; Barbet, J.; Tourwe, D.; Pelaprat, D.; Gruaz-Guyon, A. Bioconjugate Chem. 2009, 20, 1602. (7) Myers, R. M.; Shearman, J. W.; Kitching, M. O.; RamosMontoya, A.; Neal, D. E.; Ley, S. V. Chem. Biol. 2009, 4, 503. (8) Reubi, J. C.; Waser, B.; Friess, H.; Buchler, M.; Laissue, J. Gut 1998, 42, 546. (9) Souaze, F.; Dupouy, S.; Viardot-Foucault, V.; Bruyneel, E.; Attoub, S.; Gespach, C.; Gompel, A.; Forgez, P. Cancer Res. 2006, 66, 6243. (10) Carraway, R.; Leeman, S. E. J. Biol. Chem. 1973, 248, 6854. (11) Martin, S.; Botto, J.-M.; Vincent, J.-P.; Mazella, J. Mol. Pharmacol. 1999, 55, 210. (12) Deshayes, S.; Morris, M; Heitz, F.; Divita, G. Adv. Drug Delivery Rev. 2008, 60, 537. (13) Demeule, M.; Re’gina, A.; Che’, C.; Poirier, J.; Nguyen, T.; Gabathuler, R.; Castaigne, J.-P.; Be’liveau, R. J. Pharmacol. Exp. Ther. 2008, 324, 1064. (14) Okarvi, S. M. Cancer Treat. Rev. 2008, 34, 13. (15) Leader, B.; Baca, Q. J.; Golan, D. E. Nat. Rev. Drug Discovery 2008, 7, 21. (16) Cusack, B.; McCormick, D. J.; Pang, Y.-P.; Souder, T.; Garcia, R.; Fauq, A.; Richelson, E. J. Biol. Chem. 1995, 270, 18359. (17) St-Pierre, S.; Lalonde, J. M.; Gendreau, M.; Quirion, R.; Regoli, D.; Rioux, F. J. Med. Chem. 1981, 24, 370. (18) Chakravarty, P. K.; Ransom, R. W. Curr. Pharm. Des. 1995, 1, 317. (19) Orwig, K. S.; Lassetter, M. R.; Hadden, M. K.; Dix, T. A. J. Med. Chem. 2009, 52, 1803. (20) Carraway, R.; Leeman, S. E. Structural Requirements for the Biological Activity of Neurotensin, a New Vasoactive Peptide. Peptides: Chemistry, Structure and Biology; Ann Arbor Science: Ann Arbor, MI, 1975; pp 679685. (21) Hong, F.; Cusack, B.; Fauq, A.; Richelson, E. Curr. Med. Chem. 1997, 4, 421. (22) Granier, C.; van Rietschoten, J.; Kitabgi, P.; Poustis, C.; Freychet, P. Eur. J. Biochem. 1982, 124, 117. (23) Tanaka, K.; Masu, M.; Nakanishi, S. Neuron 1990, 4, 847. (24) Bergmann, R.; Scheunemann, M.; Heichert, C.; Mading, P.; Wittrisch, H.; Kretzschmar, M.; Rodig, H.; Tourwe, D.; Iterbeke, K.; Chavatte, K.; Zips, D.; Reubi, J. C.; Johannsen, B. Nucl. Med. Biol. 2002, 29, 61. (25) Garcia-Garayoa, E.; Blauenstein, P.; Bruehlmeier, M.; Blanc, A.; Iterbeke, K.; Conrath, P.; Tourwe, D.; Schubiger, P. A. J. Nucl. Med. 2002, 43, 374. (26) Garcia-Garayoa, E.; Allemann-Tannahill, L.; Blauenstein, P.; Willmann, M.; Carrel-Remy, N.; Tourwe, D.; Iterbeke, K.; Conrath, P.; Schubiger, P. A. Nucl. Med. Biol. 2001, 28, 75.

ARTICLE

(27) Sefler, A. M.; He, J. X.; Sawyer, T. K.; Holub, K. E.; Omecinsky, D. O.; Reily, M. D.; Thanabal, V.; Akunne, H. C.; Cody, W. L. J. Med. Chem. 1995, 38, 249. (28) Lundquist, J. T., IV.; Bullesbach, E. E.; Golden, P. L.; Dix, T. A. J. Pept. Res. 2001, 59, 55. (29) Lundquist, J. T., IV; Orwig, K. S.; Dix, T. A. J. Org. Chem. 1999, 64, 9265. (30) Beeson, C. C.; Dix, T. A. J. Am. Chem. Soc. 1993, 115, 10275. (31) Gilbert, J. A.; Moses, J.; Pfenning, M. A.; Richelson, E. Biochem. Pharmacol. 1986, 35, 391. (32) Gilbert, J. A.; McCormick, D. J.; Pfenning, M. A.; Kanba, K. S.; Enloe, L. J.; Moore, A.; Richelson, E. Biochem. Pharmacol. 1989, 38, 3377. (33) Loosen, P. T.; Nemeroff, C. B.; Bissette, G.; Burnett, G. B.; Prange, A. J.; Lipton, M. A. Neuropharmacology 1978, 17, 109. (34) Jolicoeur, F. B.; Barbeau, A.; Rioux, F.; Quirion, R.; St. Pierre, S. Peptides 1981, 2, 171. (35) Cain, G. A.; Christos, T. E.; Johnson, A. L.; Pottorf, R. S.; Confalone, P. N.; Tam, S. W.; Schmidt, W. K. Bioorg. Med. Chem. Lett. 1993, 3, 2055. (36) Podstawka, E. Biopolymers 2008, 89, 506. (37) Podstawka, E. J. Raman Spectrosc. 2008, 39, 1290. (38) Podstawka, E.; Niaura, G. J. Phys. Chem. B 2009, 113, 10974. (39) Podstawka, E.; Niaura, G.; Proniewicz, L. M. J. Phys. Chem. B. 2010, 114, 1010. (40) Podstawka, E.; Niaura, G.; Proniewicz, L. M. J. Phys. Chem. B 2010, 114, 5117. (41) Podstawka, E. Biopolymers 2008, 89, 980. (42) Podstawka, E.; Ozaki, Y.; Proniewicz, L. M. Langmuir 2008, 24, 10807. (43) Podstawka-Proniewicz, E.; Kudelski, A.; Kim, Y.; Proniewicz, L. M. Structure of Monolayers Formed from Neurotensin and its SingleSite Mutants: Vibrational Spectroscopic Studies. J. Phys. Chem. B 2011submitted. (44) Culha, M.; Kahraman, M.; Karatas, O. F.; Aydin, O. Encyclopedia of Nanoscience and Nanotechnology. In Chapter: Nanoplasmonics in biosensing and nanobiomedicine, Nalwa, H. S., Ed.; American Scientific Publishers, Valencia, CA, 2009. (45) Kim, J. H.; Kim, M. S.; Jeong, D. H.; Lee, Y. S.; Kim, Y. K.; Choi, H. J.; Jun, B. H. Label-Free High Throughput Screening Method by Using Sers Spectroscopic Encoded Bead and Dielectrophoresis. Patent No. 20100072067; Origin: Washington, DC; Agents: VENABLE LLP. (46) Creighton, J. A. In Spectroscopy of Surfaces; Clark, R. J. H., Hester, R. E., Eds.; Wiley and Sons: New York, 1988; Vol. 16, p 37. (47) Moskovits, M. Rev. Mod. Phys. 1985, 57, 783. (48) Gao, X.; Davies, J. P.; Weaver, M. J. J. Phys. Chem. 1990, 94, 6858. (49) Carron, K. T.; Hurley, L. G. J. Phys. Chem. 1991, 95, 9979. (50) Moskovits, M.; Suh, J. S. J. Phys. Chem. 1984, 88, 5526. (51) Eker, F.; Cao, X.; Nafie, L.; Schweitzer-Stenner, R. J. Am. Chem. Soc. 2002, 124, 14330. (52) Coutant, J.; Curmi, P. A.; Toma, F.; Monti, J.-P. Biochemistry 2007, 46, 5656. (53) Barroso, S.; Richard, F.; Nicolas-Etheve, D.; Reversat, J. L.; Bernassau, J. M.; Kitabgi, P.; Labbe-Jullie, C. J. Biol. Chem. 2000, 275, 328. (54) Luca, S.; White, J. F.; Sohal, A. K.; Filippov, D. V.; van Boom, J. H.; Grisshammer, R.; Baldus, M. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 10706. (55) Tu, A. T. Raman Spectroscopy in Biology: Principes and Applications; John Wiley and SonsInterscience: New York, 1982. (56) Herne, T. M.; Ahern, A. M.; Garrell, R. L. J. Am. Chem. Soc. 1991, 113, 846. (57) Hellwig, P.; Pfitzner, U.; Behr, J.; Rost, B.; Pesavento, R. P.; Donk, W. v.; Gennis, R. B.; Michel, H.; Ludwig, B.; Mantele, W. Biochemistry 2002, 41, 9116. (58) Venyaminov, S. Y.; Kalnin, N. N. Biopolymers 1990, 30, 1259. (59) Walls, D.; Bohn, P. W. J. Phys. Chem. 1990, 94, 2039. 7108

dx.doi.org/10.1021/jp201316n |J. Phys. Chem. B 2011, 115, 7097–7108