Electrochemical and Computational Studies on Intramolecular

School of Chemistry and Physics, The University of Adelaide, Adelaide, SA 5005, ... E-mail: [email protected]; [email protected]...
0 downloads 0 Views 513KB Size
Article pubs.acs.org/JPCC

Electrochemical and Computational Studies on Intramolecular Dissociative Electron Transfer in β‑Peptides Jingxian Yu,*,† David M. Huang,† Joe G. Shapter,‡ and Andrew D. Abell*,† †

School of Chemistry and Physics, The University of Adelaide, Adelaide, SA 5005, Australia School of Chemical & Physical Science, Flinders University, Bedford Park, SA 5042, Australia



S Supporting Information *

ABSTRACT: The preparation of the β-peptides PCB-β3Val-β3Ala-β3LeuNHC(CH3)2OOtBu and PCB-(β3Val-β3Ala-β3Leu)2-NHC(CH3)2OOtBu, with a specific donor and acceptor at each terminus, is described. Circular dichroism, 2D NMR, and density functional theory calculations confirmed that PCB-(β3Val-β3Ala-β3Leu)2-NHC(CH3)2OOtBu adopts a 14-helix conformation, whereas PCB-β3Val-β3Ala-β3Leu-NHC(CH3)2OOtBu has an ill-defined secondary structure. The electron-transfer rate constants in the two peptides were found to be 2580 and 9.8 s−1 respectively. Computational simulations based on Marcus theory coupled to constrained density functional theory provide clear theoretical evidence that different electron-transport pathways occur in the two peptides due to their different conformations: sequential hopping within PCB(β3Val-β3Ala-β3Leu)2-NHC(CH3)2OOtBu and superexchange within PCB-β3Val-β3Ala-β3Leu-NHC(CH3)2OOtBu. Electron population analysis provides the first clear theoretical evidence that amide groups can act as hopping sites in long-range electron transfer.



INTRODUCTION The study of electron transfer in proteins is of considerable interest because of its central role in key biological processes1,2 such as photosynthesis and respiration3 and also as a basis of potential applications in molecular electronics.4,5 Despite significant endeavors, predominantly using model α-peptides,6−12 the mechanisms of electron transfer are still not well understood.13 However, two distinct mechanisms have been proposed to explain electron transfer in peptides, namely, electron superexchange and electron hopping.8,14,15 For the hopping mechanism, an electron temporarily resides on the bridge between two redox centers during its passing from one redox center to the next. By contrast, the bridge serves as a medium to pass the electron between the donor and acceptor in the superexchange mechanism. Secondary structure is thought to play an important role in the process of electron transfer; however, there are few systematic studies that address this issue.16 β-Peptides offer some advantages over natural αpeptides in this context because they form helical and sheet-like secondary structures with high predictability, structural diversity, and stability.17−21 As such, these structures are ideal for studying the fundamental mechanisms of electron transport, but they have as yet received scant attention.16,22 Most studies on electron transfer report electrochemical measurements on peptides immobilized on surfaces. The problem here is that the conformational freedom of the peptide is confined in a 2D quasicrystalline arrangement.14,23 Maran and coworkers24 reported a pioneering solution-based study using peptides containing an N-terminal p-cyanobenzamide or a phthalimide donor and a C-terminal tert© 2012 American Chemical Society

butylperoxide acceptor to begin to overcome these shortcomings. Here the peroxide group acts as a concerted dissociative-type acceptor that is strongly influenced by the large inner reorganization energy associated with elongation of the cleaving O−O bond.25−29 Thus the reduction of the peroxide is irreversible and kinetically very slow.9,27,30 Because of these unfavorable kinetics, the electron is first injected by the electrode onto the donor, leading to the corresponding radical anion. The rate-determining step then involves intramolecular electron transfer across the peptide bridge to the acceptor.9,31 A schematic of this process is shown in Figure 1. The pcyanobenzamide donor has a formal potential of −1.75 V, which is 0.35 V more negative than the phthalimide donor. Hence the p-cyanobenzamide donor provides a more powerful driving force to induce an electron across the bridge from donor (D) to acceptor (A).9 In this study, we carry out electrochemical and computational studies on a β3-hexapeptide, (β3Val-β3Ala-β3Leu)2 that adopts a 14-helix and also a shorter nonhelical tripeptide, (β3Val-β3Ala-β3Leu) to understand the dependence of electrontransfer mechanisms on secondary structure in solution. Both structures were capped with an N-terminal p-cyanobenzamide donor, with a C-terminal α-aminoisobutyric acid (Aib) unit to allow the synthetic attachment of a tert-butylperoxide acceptor.32 The electron-transfer rates within both β3-peptides were determined electrochemically and also using Marcus electronReceived: August 20, 2012 Revised: November 3, 2012 Published: November 28, 2012 26608

dx.doi.org/10.1021/jp3082563 | J. Phys. Chem. C 2012, 116, 26608−26617

The Journal of Physical Chemistry C

Article

taken with the following control parameters: scan speed, 10 nm min−1; slit width, 0.2 nm; bandwidth, 1 nm; response time, 0.5 s; and N2 purging flow rate, 10 L min−1. These scans were averaged, and a blank solvent spectrum was subtracted to generate the corrected spectrum. Peptide Synthesis on 2-Chlorotrityl Chloride Resin. 2Chlorotrityl chloride resin (200−400 mesh, 5.00 g) was dried under vacuum overnight and suspended in freshly distilled DCM (25 mL). Fmoc-Aib-OH (2.50 g) was dried in vacuo for 24 h and then dissolved in anhydrous DCM (15 mL) to which DMF (0.5 mL) was added. This solution was poured in the resin suspension; then, DIPEA (5 mL) was added. The mixture was stirred gently at rt overnight and then transferred to a sintered funnel fitted with a Teflon stopcock. The resin was drained and rinsed successively with DCM (3 × 50 mL), DMF (3 × 50 mL), and finally more DCM (3 × 50 mL). A solution of DCM, methanol and DIPEA (17:2:1 respectively, 2 × 30 min) was added to the resin to cap any unreacted 2-chlorotrityl chloride linker. The resin was rinsed successively with DCM (3 × 50 mL), DMF (3 × 50 mL), and further DCM (3 × 50 mL) and then dried in vacuo overnight to give 6.52 g of Aib-loaded resin (Fmoc-Aib-OH loading = 0.72 mmol g−1). PCB-β3Val-β3Ala-β3Leu-NHC(CH3)2OOtBu.32 The FmocAib−OH-loaded 2-chlorotrityl chloride resin (2.00 g) was

Figure 1. Schematic electron-transfer process in the D-Sp-A structure.

transfer theory33 coupled to constrained density functional theory (cDFT).34 Marcus theory provides an accurate description of the kinetics of electron-transfer reactions in terms of physically appealing parameters − the extent of molecular reorganization, the strength of the electronic coupling, and the free energy gap associated with the reaction.35 cDFT provides a straightforward and efficient method for extracting the relevant parameters by constructing chargelocalized diabatic states and then optimizing the geometries of constrained systems.36,37



EXPERIMENTAL METHODS Chemicals. Fmoc-Aib-OH, Fmoc-Val-OH, Fmoc-Leu-OH, 2-chlorotrityl chloride polystyrene resin, and 2-(1H-7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyl uronium hexafluorophosphate methanaminium (HATU) were purchased from GL Biochem (Shanghai), China. Acetic anhydride, 2-methylpropan-2-OL, dichloromethane (DCM), methanol, and ethanol were purchased from Ajax Finechem (Australia). Piperidine and N,N-dimethylformamide (DMF) were purchased from Merck, Australia. Acetonitrile was purchased from Optigen Scientific, Australia. Anhydrous DMF, trifluoroacetic acid (TFA), 4cyanobenzoic acid (99%), and diisopropylethyl amine (DIPEA) were purchased from Sigma-Aldrich, Australia. Fmoc-β3-AlaOH was purchased from Hangzhou Sage Chemical, China. Fmoc-β3-Val-OH and Fmoc-β3-Leu-OH were prepared as published.38 All solvents and reagents were used without purification unless noted. High-Performance Liquid Chromatography. The synthetic peptides were analyzed and purified by reverse-phase high-performance liquid chromatography (HPLC), using an HP 1100 LC system equipped with a Phenomenex C18 column (250 × 4.6 mm) for analytical traces and a Phenomenex C18 column (250 × 21.2 mm) for purification, a photodiode array detector, and a Sedex evaporative light scattering detector. Water/TFA (100/0.001 by v/v) and water/TFA (100/0.001 by v/v) solutions were used as organic and aqueous buffers. Spectroscopic Measurements. 1H NMR spectra were recorded in DMSO-d6 solutions using a Varian Inova 600 MHz spectrometer. Chemical shifts are reported in ppm (δ) using TMS (0.00 ppm) as the internal standard. Signals are reported as s (singlet), d (doublet), or m (multiplet). Mass spectral data were collected on a Finnigan’s LCQ mass spectrometer. Circular Dichroism. Circular dichroism (CD) spectra were acquired with a JASCO J-815 CD spectrometer (JASCO, U.K.) using an optical cell of 0.1 cm optical path length at the residue concentration of 2.7 mM in methanol at 22 °C. Ten scans were

transferred into a sintered funnel fitted with a Teflon stopcock and then rinsed with DCM (2 × 20 mL). After air drying, the Fmoc group was removed by reaction with a solution of 25% piperidine in DMF (20 mL) for 30 min, followed by washing successively with DCM (3 × 20 mL), DMF (3 × 20 mL), and DCM (3 × 20 mL). To a solution of Fmoc-β3-Leu-OH (1.00 g, 2 equiv) in DMF (4 mL) was added a 0.5 M solution of HATU in DMF (2 mL), followed by DIPEA (1.2 mL, four-fold excess), and the resulting solution was added to the deprotected resin. The mixture was left for 2 h, with occasional stirring. The resin was isolated by filtration and rinsed successively with DCM (3 × 50 mL), DMF (3 × 50 mL), and DCM (3 × 50 mL). The sequence was repeated two more times to ensure complete coupling. Successive couplings of Fmoc-β3-Ala-OH and Fmocβ3-Val-OH were carried out in order, using this protocol, to give the appropriate peptide. The peptide was capped with the addition of 4-cyanobenzoic acid (1.00 g) in the last cycle (using the same protocol) and then cleaved from the resin on treatment with 2% TFA/DCM (v/v). Purification of the resulting oil by HPLC gave PCB-β3Val-β3Ala-β3Leu-Aib-OH. A solution of this peptide (0.5 mmol) and HATU (0.65 mmol), in anhydrous DCM containing several drops of anhydrous DMF, was stirred overnight at rt. The solvent was removed under reduced vacuum, the residue was dissolved in ethyl acetate, and the solution was washed with 10% aqueous KHSO4 followed by water. The organic phase was evaporated to dryness under reduced pressure, and the oxazolone intermediate was dissolved in anhydrous DCM (5 mL). To this was added 4-dimethylaminopyridine (DMAP) (2 mmol) and a 5.5 M solution of tert-butyl hydroperoxide in decane (0.65 mL, 6 mmol). The mixture was refluxed for 2 days. The solvent was removed under reduced pressure and the peptide peroxide was 26609

dx.doi.org/10.1021/jp3082563 | J. Phys. Chem. C 2012, 116, 26608−26617

The Journal of Physical Chemistry C

Article

09 with tight convergence criteria.40 (See the Supporting Information for the optimized geometries and corresponding atomic coordinates.) Figure 2 illustrates the constructed diabatic states in both peptide systems, as suggested by May15,41 and Schlag,42,43 obtained by localizing an overall charge of −1 on the donor, each β3-amino acid, and the acceptor individually. (See the Supporting Information for details of the atom groups to which the excess electron was localized in the diabatic states.) The geometry of each diabatic state was optimized with the excess electron constrained to the indicated part of the molecule using cDFT as implemented in NWChem 6.044 with the B3LYP density functional and basis sets of increasing size from 3−21G up to 6-31G*, starting from the optimized geometry of the uncharged molecular system. The group of atoms over which the excess electron was localized in each diabatic state is detailed in the Supporting Information. Electronic coupling matrix elements Hab and diabatic potential energy surfaces were computed for these diabatic states using cDFT at the B3LYP/6-31G* level with NWChem 6.0.36,45 Diabatic potential profiles were determined by assuming that during an electron-transfer step the nuclear configuration changes smoothly between the optimized geometries of the diabatic states in which the excess electron is localized before and after electron transfer.46 Thus, the energy of each of the two diabatic states along the electron-transfer reaction coordinate was taken as the energy for geometries linearly interpolated between the optimized geometries of the two diabatic states, with the excess electron localized to the part of the molecule corresponding to the diabatic state in question.

purified by reverse-phase HPLC. 1H NMR (600 MHz, DMSOd6) δ 8.35 (d, J = 8.9 Hz, 1H, NH), 8.02−7.92 (m, 4H, C6H4), 7.97 (s, 1H, NHC(CH3)2), 7.74 (d, J = 7.9 Hz, 1H, NH), 7.60 (d, J = 8.8 Hz, 1H, NH), 4.19 (m, 1H, NHCHCH2), 4.09 (m, 1H, NHCHCH3), 3.95 (m, 1H, NHCHCH), 2.38−2.22 (m, 2H, CH(CH2CH(CH3)2)CHH, CH(CH2CH(CH3)2)CHH), 2.22−1.90 (m, 4H, CH(CH(CH3)2)CHH, CH(CH(CH3)2)CHH, CH(CH3)CHH, CH(CH3)CHH), 1.80 (m, 1H, CHCH(CH3)2), 1.5 (m, 1H, NHCHCH2CH(CH3)2), 1.31 (d, 6H, C(CH3)2O), 1.30−1.18 (m, 2H, NHCHCH2CH), 1.18 (s, 9H, tBu), 0.94 (d, 3H, NHCHCH3), 0.90−0.84 (dd, 6H, CHCH(CH3)2), 0.84−0.76 (dd, 6H, CH2CH(CH3)2). MS: [M +Na]+calcd = 624.8, [M+Na]+found = 624.7. PCB-(β3Val-β3Ala-β3Leu)2-NHC(CH3)2OOtBu. PCB-(β3Val3 β Ala-β3Leu)2-NHC(CH3)2OOtBu was similarly synthesized

and purified. 1H NMR (600 MHz, DMSO-d6) δ 8.37 (d, J = 8.7 Hz, 1H, NH), 7.98 (s, 1H, NHC(CH3)2), 7.97−7.92 (d, J = 8.4 Hz, 1H, NH), 7.93−7.91 (s, 4H, C6H4), 7.75 (d, J = 8.1 Hz, 1H, NH), 7.61 (m, 2H, 2 × NH), 7.47 (d, J = 5.6 Hz, 1H, NH), 4.24−4.14 (m, 1H, NHCHCH2), 4.14−3.98 (m, 4H, 4 × NHCHCH2), 3.98−3.90 (m, 1H, NHCHCH2), 2.38−1.90 (m, 12H, 6 × CH2CO), 1.85−1.62 (m, 2H, 2 × CHCH(CH3)2), 1.52 (m, 2H, 2 × NHCHCH2CH), 1.31 (m, 6H, C(CH3)2CO), 1.30−1.18 (m, 4H, 2 × NHCHCH2CH), 1.20 (s, 9H, tBu), 1.00−0.70 (m, 30H, 2 × NHCHCH3, 2 x CHCH(CH3)2, 2 × CH2CH(CH3)2). MS: [M+Na]+calcd = 950.2, [M+Na]+found = 950.2. Electrochemistry. The cyclic voltammetry (CV) measurements on the synthetic peptides were carried out in 0.1 mol L−1 tetra-n-butylammonium hexafluorophosphate (TBAPF 6 , Sigma)/DMF (Isocratic HPLC grade, Scharlau Chemie) solutions. Glassy carbon was the working electrode (Φ = 1.0 mm, purchased from BAS, Tokyo, Japan) with a 1 cm2 Pt mesh as the counter electrode. The reference electrode (Ag/AgCl,) was calibrated after each measurement against the ferrocene/ ferricenium couple. All potentials are reported against the KCl saturated calomel electrode (SCE), using E°Fc/Fc+ = 0.464 V versus SCE. All electrochemical experiments were performed inside a drybox filled with high purity N2 using a computercontrolled PGSTAT100 electrochemical workstation (Autolab, The Netherlands) with ohmic-drop correction at rt. The digitized, background-subtracted curves were analyzed using the UTILS package programmed by Dirk Heering and compared with the corresponding digital simulations. The Electrochemical Simulation Package (ESP 2.4)39 was used to estimate intramolecular electron-transfer rate constants using a step increment of 1 mV and the ‘Exact’ mode for the chemical reaction approximation. The intermolecular electron transfer rate constant, kinter, was set to the value of 103.6 s−1 obtained in a previous publication.9 Computational Methods. The initial geometry of PCB(β3Val-β3Ala-β3Leu)2-NHC(CH3)2OOtBu β3-hexapeptide was constructed based on the crystal structure of a 14-helix βpeptide bundle.18 The geometries of uncharged PCB-β3Valβ 3 Ala-β 3 Leu-NHC(CH 3 ) 2 OOtBu and PCB-(β 3 Val-β 3 Alaβ 3 Leu) 2 -NHC(CH 3 ) 2 OOtBu were optimized using the B3LYP density functional and STO-3G basis set in Gaussian



RESULTS AND DISCUSSION Structure of β-Peptides in Solution. (β3Val-β3Alaβ3Leu)2 has been shown by Seebach47 and Abell16,22 to adopt a 14-helix with approximately three residues per helix turn that positions the side chains in a stacked arrangement along the axis of the helix. This results in a hydrophilic inner core and a hydrophobic outer surface. A ROESY spectrum of PCB-(β3Valβ3Ala-β3Leu)2−OH (as shown in Figure S1 in the Supporting Information) indicates three unambiguous correlations that are consistent with this geometry: one between the β3-methine and β2-methylene protons on different β-Ala residues and a second between the β3-methine and β2-methylene protons of the different β-Val residues, and a third between the β3-methine and β2-methylene protons of the different β-Leu residues. Furthermore, the intensity of the three correlations indicates that the alanine protons are closer in space than the valine and leucine protons.16,22 The CD spectrum of PCB-(β3Val-β3Alaβ3Leu)2-NHC(CH3)2OOtBu revealed a strong Cotton effect, as shown in Figure 3, with a minimum and a shoulder maximum at the expected wavelengths (203 and 193 nm). This is also consistent with a helical conformation in solution.22 Gas-phase geometry optimization using DFT on PCB-(β3Val-β3Alaβ3Leu)2-NHC(CH3)2OOtBu shows that it readily folds into a low-energy hydrogen-bonded structure (as shown in Figure S2a in the Supporting Information), which is consistent with the CD spectrum and the backbone correlations determined by NMR. β3Val-β3Ala-β3Leu is too short to give a helical structure.22 Furthermore, the gas-phase geometry optimization for PCB-β3Val-β3Ala-β3Leu-NHC(CH3)2OOtBu suggests that the peptide strand is kinked, as shown in Figure S2b in the Supporting Information. Electrochemical Analysis of Intramolecular Electron Transfer. Figure 4 shows a series of cyclic voltammograms 26610

dx.doi.org/10.1021/jp3082563 | J. Phys. Chem. C 2012, 116, 26608−26617

The Journal of Physical Chemistry C

Article

Figure 2. Schematic showing the parts of the molecule where the electron charge is localized in the diabatic states, the sequential electron-transfer pathway, and the electron-transfer superexchange pathway for both (a) PCB-β3Val-β3Ala-β3Leu-NHC(CH3)2OOtBu and (b) PCB-(β3Val-β3Alaβ3Leu)2-NHC(CH3)2OOtBu.

obtained for 2.5 mM PCB-β 3 Val-β 3 Ala-β 3 Leu-NHC(CH3)2OOtBu and 1.4 mM PCB-(β3Val-β3Ala-β3Leu)2-NHC(CH3)2OOtBu on a glassy carbon electrode in 0.1 mol L−1 TBAPF6/DMF solutions. The potential sweep was carried out from −1.5 V, going in the cathodic direction. Reduction of both PCB-β3Val-β3Ala-β3Leu-NHC(CH3)2OOtBu and PCB-(β3Valβ3Ala-β3Leu)2-NHC(CH3)2OOtBu showed an irreversible signal, with reduction current peaks at −1.87 V for PCBβ3Val-β3Ala-β3Leu-NHC(CH3)2OOtBu and at −1.84 V for PCB-(β3Val-β3Ala-β3Leu)2-NHC(CH3)2OOtBu at a scan rate of 500 mV s−1. As expected, reduction current peaks grew with increasing scan rate. The peroxide-type dissociative electron transfer occurs via the following mechanism:9 D‐Sp ‐A + e− ⇌ •−D‐Sp ‐A •− 3

3

(fast)

D‐Sp ‐A → D‐Sp‐ +A•

3

Figure 3. CD spectrum of PCB-(β Val-β Ala-β Leu) 2 -NHC(CH3)2OOtBu at 2.7 mM in methanol.

(intramolecular electron transfer) 26611

dx.doi.org/10.1021/jp3082563 | J. Phys. Chem. C 2012, 116, 26608−26617

The Journal of Physical Chemistry C

Article

Figure 4. Cyclic voltammograms of (a) PCB-β3Val-β3Ala-β3Leu-NHC(CH3)2OOtBu (2.5 mM) and (b) PCB-(β3Val-β3Ala-β3Leu)2-NHC(CH3)2OOtBu (1.4 mM) on a glassy carbon electrode in 0.1 mol L−1 TBAPF6/DMF solutions, with a scan rate of 100, 200, and 500 mV s−1, respectively, from top to bottom, as indicated by the arrow. Insets: Experimental and simulated cyclic voltammetric curves at a scan rate of 500 mV s−1 for (a) PCB-β3Val-β3Ala-β3Leu-NHC(CH3)2OOtBu and (b) PCB-(β3Val-β3Ala-β3Leu)2-NHC(CH3)2OOtBu. •−

the diabatic potential profiles for sequential and superexchange electron-transfer pathways for both PCB-β3Val-β3Ala-β3LeuNHC(CH3)2OOtBu and PCB-(β3Val-β3Ala-β3Leu)2-NHC(CH3)2OOtBu. The energy of diabatic state A (as shown in Figure 2) is lower than that of diabatic state D in all of the diabatic profiles. This theoretically confirms that the cyanobenzamide (electron donor)−tert-butylperoxide (electron acceptor) pair has good energy compatibility, which provides a suitable driving force to move an electron from the donor to the acceptor within the investigated β-peptide anions. The molecular reorganization energy (λ) and the energy gap (ΔE) between two neighboring diabatic states was estimated from these diabatic potential profiles, as indicated schematically in Figure 5a,b. The calculated λ and ΔE values are given in Table 2 for PCB-β3Val-β3Ala-β3Leu-NHC(CH3)2OOtBu and in Table 3 for PCB-(β3Val-β3Ala-β3Leu)2-NHC(CH3)2OOtBu along with the strength of the electronic coupling (Hab)ij calculated from the electronic structures of neighboring diabatic states.45 It can be seen from Figure 5 that there is a significantly larger energy gap ΔE for the electron-transfer step from diabatic state D to B1 for PCB-β3Val-β3Ala-β3Leu-NHC(CH3)2OOtBu (∼0.07 au) than for PCB-(β 3 Val-β 3 Ala-β 3 Leu) 2 -NHC(CH3)2OOtBu (−0.0007 au). It is difficult to rigorously quantify the contributions to ΔE and the reorganization energy λ from different sources such as changes in bond strain, van der Waals interactions, or electrostatic interactions, but the computational data do provide potential explanations for the observed differences between PCB-β3Val-β3Ala-β3Leu-NHC(CH 3 ) 2 OOtBu and PCB-(β 3 Val-β 3 Ala-β 3 Leu) 2 -NHC(CH3)2OOtBu. A significant contribution to the difference in ΔE for the electron-transfer step from the diabatic state D to B1 appears to be electrostatic in nature. The intramolecular electrostatic interaction energy Eelec between atom pairs was estimated as Eelec = ∑i