Electrochemical studies are reported on a series of peptides constrained into either a 310-helix (1-6) or β-strand (7-9) conformation, with variable numbers of electron rich alkene containing side chains. Peptides (1 and 2) and (7 and 8) are further constrained into these geometries with a suitable side chain tether introduced by ring closing metathesis (RCM). Peptides 1, 4 and 5, each containing a single alkene side chain reveal a direct link between backbone rigidity and electron transfer, in isolation from any effects due to the electronic properties of the electron rich side-chains. Further studies on the linear peptides 3-6 confirm the ability of the alkene to facilitate electron transfer through the peptide. A comparison of the electrochemical data for the unsaturated tethered peptides (1 and 7) and saturated tethered peptides (2 and 8) reveals an interplay between backbone rigidity and effects arising from the electron rich alkene side-chains on electron transfer. Theoretical calculations on β-strand models analogous to 7, 8 and 9 provide further insights into the relative roles of backbone rigidity and electron rich side-chains on intramolecular electron transfer. Furthermore, electron population analysis confirms the role of the alkene as a "stepping stone" for electron transfer. These findings provide a new approach for fine-tuning the electronic properties of peptides by controlling backbone rigidity, and through the inclusion of electron rich side-chains. This allows for manipulation of energy barriers and hence conductance in peptides, a crucial step in the design and fabrication of molecular-based electronic devices.
Electrochemical studies are reported on a series of peptidesconstrained into either a 310-helix (1-6) or β-strand (7-9) conformation, with variable numbers of electron rich alkenecontaining side chains. Peptides (1 and 2) and (7 and 8) are further constrained into these geometries with a suitable side chain tether introduced by ring closing metathesis (RCM). Peptides 1, 4 and 5, each containing a single alkene side chain reveal a direct link between backbone rigidity and electron transfer, in isolation from any effects due to the electronic properties of the electron rich side-chains. Further studies on the linear peptides 3-6 confirm the ability of the alkene to facilitate electron transfer through the peptide. A comparison of the electrochemical data for the unsaturated tethered peptides (1 and 7) and saturated tethered peptides (2 and 8) reveals an interplay between backbone rigidity and effects arising from the electron rich alkene side-chains on electron transfer. Theoretical calculations on β-strand models analogous to 7, 8 and 9 provide further insights into the relative roles of backbone rigidity and electron rich side-chains on intramolecular electron transfer. Furthermore, electron population analysis confirms the role of the alkene as a "stepping stone" for electron transfer. These findings provide a new approach for fine-tuning the electronic properties of peptides by controlling backbone rigidity, and through the inclusion of electron rich side-chains. This allows for manipulation of energy barriers and hence conductance in peptides, a crucial step in the design and fabrication of molecular-based electronic devices.
Electron transfer in
proteins plays an important role in a wide
range of metabolic processes at the cellular level.[1,2] Many
factors have been shown to influence this electron transfer, including
the distance separating the electron donor and acceptor,[3−5] the extent of secondary structure,[6,7] dipole moment,[8−10] and the nature of the constituent amino acid side chains.[11−14] Of particular significance is the suggestion that peptidescan undergo
electron transfer via either a bridge-assisted superexchange or an
electron hopping mechanism.[15] While electron
superexchange is a one step process that is exponentially dependent
on distance, the hopping model involves a multistep process for electron
translocations across long distances, whereby the charge temporarily
resides on the bridge between redox centers.[16,17] Consequently, electron hopping through a peptidecan be facilitated
by redox-active amino acid side chains in the sequence that act as
“stepping stones” for electron transfer.[18,19] Studies on model peptides have confirmed this, where the rate of
electron transfer increases significantly with the introduction of
electron rich side-chains into the peptide.[11] For example, Kimura and co-workers demonstrated that linearly spaced
electron rich naphthyl groups within syntheticpeptides increase the
photocurrent by efficient electron hopping between the moieties, compared
to reference peptidescontaining one or no naphthyl groups.[20] An electron rich tryptophan side chain has also
been shown to act as a “relay station” to facilitate
multistep electron transfer in an azurin metallo-protein isolated
from .[21] Multiple sequence alignment
of genomes from the respiratory oxidoreductase enzyme NDH1, have revealed
the conservation of specificaromatic amino acids from simple prokaryotes
through to man, that may serve as candidates for transient charge
localization between metalclusters.[22] The
majority of research conducted thus far has focused on aromatic amino
acids as the source of electron rich “stepping stones”.Recent work with an Aib (α-aminoisobutyric acid) rich hexapeptide,
constrained into a 310-helix by a triazole-containing covalent
tether linking its i to i + 3 residues,[23] has shown that backbone rigidity also plays
a significant role in defining the rate of electron transfer in peptides.
Increased rigidity restricts backbone torsional motion, resulting
in an additional reorganization energy barrier to electron transfer.
In the study reported here, a series of alkenecontaining peptides,
both linear (see 4, 5, 6, and 9, Figures 1 and 2) and also a series of alkene tethered peptides (see 1 and 7, Figures 1 and 2), is used to begin to unravel the interplay of
peptidebackbone rigidity and the nature of the amino acid side chains
in defining the rate of electron transfer, where until now these effects
have been considered without factoring in the other variable. Electrochemical
and theoretical studies are presented on peptidesconstrained into
both a 310-helix (see 1 and 2, Figure 1) and a β-strand (see 7 and 8, Figure 2). The
alkene group in the peptides is shown to promote electron transfer,
with its influence on backbone rigidity and its role as an electron
rich “stepping stone” discussed to explore the generality
and connectivity of these effects.
Figure 1
Structures of helical peptides 1–6.
Figure 2
Structures of β-strand peptides 7–9.
Structures of helical peptides 1–6.Structures of β-strand peptides 7–9.
Experimental Methods
Chemicals
Fmoc-Aib-OH, Boc-Aib-OH, Boc-Ser-OH, Fmoc-OSu,
2-chlorotrityl chloride polystyrene resin, 1-hydroxy-7-azabenzotrizole
(HOAt) and 2-(1H-7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyl
uronium hexafluorophosphate methanaminium (HATU) were purchased from
GL Biochem (Shanghai) Ltd., China. Dichloromethane (DCM), diethyl
ether (Et2O), ethyl acetate (EtOAc), methanol and ethanol
were purchased from Ajax Finechem Pty Ltd. (Australia). Piperidine,
acetonitrile, propan-2-ol and N,N-dimethylformamide (DMF) were purchased from Merck, Australia. Anhydrous N,N-dimethylformamide (DMF), dimethyl sulfoxide
(DMSO), tetrahydrafuran (THF), Second-generation Grubbs’ catalyst,
Pd/Ccatalyst, ethyl vinyl ether, 2,2,2-trifluoroethanol (TFE), trifluoroacetic
acid (TFA), 4 M HCl/dioxane solution, N,N′-dicyclohexylcarbodiimide (DCC), dimethylaminopyridine (DMAP),
cysteamine and diisopropylethylamine (DIPEA) were purchased from Sigma-Aldrich,
Australia. NaOH was purchased from Chem Supply, Australia. Single-walled
carbon nanotubes (P2-SWCNTs) were purchased from Carbon Solutions,
Inc., USA. Boc-Ser(Al)-OH[24] and ferrocenylmethylamine[25,26] were prepared as published. All solvents and reagents were used
without purification unless noted.
High-Performance Liquid
Chromatography
The syntheticpeptides were analyzed and purified by reverse phase 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 (100/0.1 by v/v)
and ACN/TFA (100/0.08 by v/v) solutions were used as aqueous and organic
buffers.
NMR Spectroscopy
1HNMR spectra were recorded
in DMSO-d6 or CDCl3-d solutions using a Varian Gemini-300 NMR. 13CNMR and two-dimensional NMR experiments utilizing COSY, ROESY,
HSQC and HMBC, were obtained on 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),
t (triplet) or m (multiplet).
Mass Spectroscopy
Low resolution mass spectral data
were analyzed using a Finnigan MAT LCQ spectrometer with MS/MS and
ESI probe, utilizing XCalibur software. High resolution mass spectral
data were analyzed using an Ultimate 3000 RSL HPLC (Thermo Fisher
Scientific Inc., MA) and an LTQ Orbitrap XL ETD using a flow injection
method, with a flow rate of 5 μL/min. The HPLC flow is interfaced
with the mass spectrometer using the Electrospray source (Thermo Fisher
Scientific Inc., MA). Mass spectra were obtained over a range of 100 < m/z < 1000. Data were analyzed using
XCalibur software (Version 2.0.7, Thermo Fisher Scientific).
FTIR Spectroscopy
Infrared spectra were collected on
a PerkinElmer Spectrum 100 FT-IR spectrometer, with attenuated total
reflectance (ATR) imaging capabilities, fitted with a ZnSe crystal,
with an average reading taken from 4 scans at 4 cm–1 resolution.
Peptide Synthesis
The linear hexapeptides
(3,[23]4, 5 and 6) were synthesized using solid phase peptide
synthesis on
2-chlorotrityl chloride resin using Fmoc-Aib-OH, Fmoc-protected allyl
serine, and HATU/DIPEAcoupling conditions as detailed in the Supporting Information. Cleavage from the resin
was followed by C-terminal coupling with ferrocenylmethylamine, and
the N-terminal Boc group was removed to give the
freeamine for coupling to a single-walled carbon nanotube (SWCNTs)/Au
electrode assembly. The 310-helical (1) and
β-strand (7) macrocycles were prepared by ring
closing metathesis of the appropriate dienes as detailed in the Supporting Information. Peptide 1 involves linking the i to i +
3 residues using a strategy previously reported by O’Leary
and co-workers.[24] In this study the geometry
of a related sequence was confirmed by X-ray crystallography.[24] For both of these peptides a single alkene isomer
was obtained which was assigned the (E)-configuration
based on the alkenecoupling constant[27] (15.8 Hz for 7). The C-terminal ferrocenyl group and N-terminal freeamine were introduced as above. The saturated
analogues (2 and 8) were prepared by hydrogenation
of the macrocyclicalkenes and the linear peptide 9 was
prepared by simple peptidecoupling in solution. All peptides (1–9) were purified using reverse phase
HPLC prior to attachment to the SWCNTs/Au electrode by HATU/DIPEA.
General
Procedure for N-Boc Cleavage
The N-Boc protected
peptides 1–9 were
dissolved in trifluoroethanol (TFE), and 4 M HCl in 1,4-dioxane was
added. The reaction solution was stirred at rt for 20–30 min,
and the solvent removed in vacuo.
All electrochemical measurements were
taken with a CHI 650D electrochemical analyzer (CH Instruments, Inc.)
with ohmic-drop correction at room temperature. A peptide modified
gold surface formed the working electrode[23] (geometric area of 0.33 cm2), with a platinum mesh and
Ag/AgCl wire used as the counter and reference electrodes, respectively.
The Ag/AgCl reference electrode was calibrated after each experiment
against the ferrocene/ferriceniumcouple. Ferrocene-derivatized peptide
electrodes were electrochemically characterized in 0.1 mol L–1 tetra-n-butylammonium hexafluorophosphate (TBAPF6)/CH3CN solutions. The digitized, background-subtracted
curves were analyzed using a Data Master 2003 program.
Computational
Methods
The lowest energy conformers
for all of the N-protected peptides were determined
in Gaussian 09, with tight convergence criteria using a hybrid B3LYP
method with 6-31G** basis set for all C, H, N, O atoms, and Lanl2dz
basis set for the Fe atom in order to define the backbone conformations
of all peptides. The geometry of each diabatic state was optimized
using the latest constrained density functional theory (cDFT)[28] as implemented in NWChem 6.1.1[29] using the B3LYP density functional method with 6-31G**
basis set for all C, H, N, O atoms, and Lanl2dz basis set for the
Fe atom. 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.[30] 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. In these calculations, the solvent effects were taken into
account approximately by the COSMO approach. The Löwdin electron
population analysis for uncharged and charged amino acids was conducted
by respectively placing the charge of 0 and +1 on the individual residue
within the linear helical peptide, namely an Aib or a modified serine
with electron rich alkene side-chain, using the B3LYP density functional
method with 6-31G** basis set for all C, H, N, O atoms, and Lanl2dz
basis set for the Fe atom.[7]
Results
and Discussion
Peptide Design
Aib residues were
incorporated into
peptides 1–6 in order to promote
the formation of a unifying 310-helical secondary structure.[23] The peptides 1, 4, 5, and 6 also contain alkenes as potential hopping
sites for electron transfer, where this group is part of the macrocycle
of 1. The peptides 2 and 3 lack
an alkene and hence provide suitable controls. The diene of 6 is positioned in the i and i + 3 residues, to locate the alkenes on the same side of the molecule
in the helix in a proximal arrangement to promote electron transfer.
This diene also allows cyclization by ring closing metathesis (RCM)
to introduce a covalent tether to further constrain the peptide backbone
into a 310-helix and to rigidify the backbone into this
geometry. The β-strand constrained peptides (7 and 8) have the covalent tether linking the i and i + 2 residues. Such a 17-membered ring, with
an aryl group at the N-terminus is known to stabilize
a β-strand geometry with the associated rigidification of the
backbone.[27] The linear diene 9 was also prepared as a control for the electrochemical studies.
Conformational Analysis of Peptides
The geometry of
peptides 1–6 was confirmed as 310-helical by 1HNMR spectroscopy. In particular,
strong NH (i) to NH (i + 1) ROESY correlations were found for peptides 1–6, together with CαH (i) to NH (i +
1) and medium range CαH (i) to NH (i + 2) correlations, as
shown in Figure 3 and the Supporting Information. A CαH (i) to NH (i + 2) cross
peak is only possible for a 310-helix,[31] as the distance between these two hydrogen atoms is in
the order of 3.5 Å, whereas in an α-helix the distance
between CαH (i) to NH (i + 2) atoms is approximately 4.5 Å,
and near the limit of detection.[32] An absence
of CαH (i) to NH (i + 4) correlations was noted for all peptides,
thus excluding the possibility of an α-helical structure, which
is characterized by (i to i + 4)
hydrogen bonds.[33] Strong correlations were
also evident for CβH2 (i) and NH (i) in peptides 1 and 2.[34] Hence the
cumulative 1HNMR data confirms the presence of 310-helical structures for each of peptides 1–6.
Figure 3
1H NMR ROESY spectrum representative of peptide 2, showing CαH (i)
to NH (i + 1), CαH (i) to NH (i +
2) and CβH2 (i)
and NH (i) crosspeaks, indicative
of a 310-helical conformation.
1HNMR ROESY spectrum representative of peptide 2, showing CαH (i)
to NH (i + 1), CαH (i) to NH (i +
2) and CβH2 (i)
and NH (i) crosspeaks, indicative
of a 310-helical conformation.The conformations of peptides 7–9 were confirmed as β-strand by a combination of 1HNMR and IR spectroscopy. CαH (i) to NH (i + 1) and CβH (i) to NH (i + 1) ROESY correlations were found for all three peptides, indicative
of a β-strand geometry[35] (see Supporting Information). Furthermore, 1HNMR JNcoupling constants[35] of 8–10
Hz were observed for these peptides. Amide I and II bands, used extensively
in peptide/protein structural determination, were found to be in the
range assigned to a β-strand conformation[36] for all three tripeptides. Amide A (N–H stretching)
frequencies between 3277 and 3293 cm–1 were also
observed in the IR spectra of peptides 7–9, indicative of the presence of hydrogen bonding within ordered
β-sheets[37] (see Supporting Information).Lowest energy conformers for the N-protected analogues
of 1–6 (peptides 10–15). The N-Boc protection group is circled
in Peptide 10.The lowest energy conformers for the N-protected
analogues of 1–9 (see peptides 10–18, Figures 4 and 6) were determined by molecular modeling
in order to further define the backbone geometries. The N-protected peptides were used in these studies, as freeamines are
known to give rise to unrealistic electrostatic interactions, resulting
in unstable lowest energy conformers.[38] The lowest energy conformers for the N-protected
helical hexapeptides 10–15 (see Figure 4) were calculated. The resulting models indicate
that the backbone lengths (from first to last carbonyl carbons) are
almost identical, differing by no more than 0.04 Å. The mean
hydrogen bond lengths in the constrained helical peptides 10 and 11 is 2.10 Å, which is in accordance with
similar 310-helical structures,[23,24,34] and 2.12 Å in the unconstrained helical
peptides 12–15 (see Supporting Information), also similar to those reported elsewhere.[23,24,34] The most significant difference
in the intramolecular hydrogen bond lengths for each of the helical
peptides is only 0.15 Å, between residues 2 and 5 in peptides 10 and 15, which correspond to the i and i + 3 positions of the constraint. The average
dihedral angles for residues 1–6 in each of the N-Boc protected analogues, deviate from an ideal 310-helix
by no more than 3.6° and 5.9° for Φ and ψ, respectively.
Figure 5 shows the lowest energy conformer
for 15, revealing that the side-chains are positioned
on the same side of the molecule, with the terminal alkenes separated
by 6 Å.
Figure 4
Lowest energy conformers for the N-protected analogues
of 1–6 (peptides 10–15). The N-Boc protection group is circled
in Peptide 10.
Figure 6
Lowest energy conformers for the N-protected
analogues
of 7–9 (peptides 16–18).
Figure 5
(a) The lowest energy conformer for peptide 15 (analogue
of 6) showing the two side-chains facing each other in
a proximal arrangement (circled) and (b) the view looking down the
helix, which indicates that the two side-chains are in the same plane
(circled). (Optimized by the hybrid B3LYP method with 6-31G** basis
set for all C, H, O, N atoms and Lanl2dz for Fe atom.)
(a) The lowest energy conformer for peptide 15 (analogue
of 6) showing the two side-chains facing each other in
a proximal arrangement (circled) and (b) the view looking down the
helix, which indicates that the two side-chains are in the same plane
(circled). (Optimized by the hybrid B3LYP method with 6-31G** basis
set for all C, H, O, N atoms and Lanl2dz for Fe atom.)Lowest energy conformers for the N-protected
analogues
of 7–9 (peptides 16–18).The calculated lowest
energy conformers for the N-protected β-strand
peptides 16, 17 and 18, (see
Figure 6) indicate
that the backbone length (from first
to last carbonyl carbons) are once again almost identical, with the
backbone length of the constrained peptides differing by only 0.05
Å. The largest variation in backbone length is 0.30 Å, between
the linear analogue (18) and the unsaturated peptide
(16). All other dimensions critical to the characterization
of a β-strand conformation, such as NH (i) to NH (i + 1), CαH (i) to NH (i + 1) and CβH2 (i) to NH (i + 1) distances (see Supporting Information) are in accordance with
literature values.[39] Figure 7 shows the lowest energy conformers for peptides 16 and 17, highlighting the structural difference between
the side-chains of the saturated and unsaturated molecules.
Figure 7
Lowest energy
conformers for peptides 16 and 17, analogues
of 7 and 8 (overlapped),
optimized by the hybrid B3LYP method with 6-31G** basis set for all
C, H, O, N atoms and Lanl2dz for Fe atom. The side-chain of the unsaturated 16 is depicted in white, with that of peptide 17 in pink. The saturated/unsaturated models overlap fittingly, with
the exception of the highlighted region about the double bond.
A combination of the molecular modeling studies and the 1HNMR and IR data demonstrates that peptides 10–15 share remarkably similar 310-helical conformations,
while peptides 16–18 exhibit a common
β-strand geometry. Thus, the prominent structural differences
between each of these peptides and hence the analogues (1–6 and 7–9)
are simply the variation in the number of electron rich alkenes, the
presence (or absence) of the side-bridge constraint, and the associated
effect that this has on backbone rigidity as discussed below.Lowest energy
conformers for peptides 16 and 17, analogues
of 7 and 8 (overlapped),
optimized by the hybrid B3LYP method with 6-31G** basis set for all
C, H, O, N atoms and Lanl2dz for Fe atom. The side-chain of the unsaturated 16 is depicted in white, with that of peptide 17 in pink. The saturated/unsaturated models overlap fittingly, with
the exception of the highlighted region about the double bond.
Electrochemical Analysis
of Intramolecular Electron Transfer
Each of the peptides 1–9 was separately
attached to vertically aligned single-walled carbon nanotube array/gold
(SWCNTs/Au) electrodes[40] in order to study
their electron transfer kinetics. SWCNTs/Au electrodes were used in
this study to provide a high surface concentration of redox probes,
with an associated significant increase in sensitivity and reproducibility
of the electrochemical measurement over bare Au electrodes.[40] Analysis of the electrochemical results for
the helical peptides 1–6 reveal a
pair of redox peaks in each cyclic voltammogram, characteristic of
a one-electron oxidation/reduction reaction (Fc+/Fc) (see
Figure 8). The formal potentials (E0) and apparent electron transfer rate constants (kapp) were estimated using Laviron’s formalism,[41] and given in Table 1.
Figure 8
(a) Cyclic voltammograms for peptides 1–6 immobilized on SWCNTs/Au electrodes taken at 5 V s–1. (b) Peak potential versus ln (scan rate) for peptides 1–6 after background current subtraction.
Table 1
Electron Transfer Rate Constants (kapp), Surface Concentrations and Formal Potentials
(Eo) for the Helical Peptides (1–6)
peptide
surface concentration ( × 10–10 mol cm–2)
Eo (V vs AgCl/Ag)
kapp/s–1
1
4.37 ± 0.43
0.844
17.49 ± 1.46
2
4.19 ± 0.35
0.881
31.88 ± 2.82
3
9.79 ± 0.21
0.508
62.90 ± 5.35
4
4.02 ± 0.41
0.380
260.38 ± 25.32
5
4.12 ± 0.48
0.379
307.11 ± 30.61
6
3.58 ± 0.37
0.375
388.44 ± 37.94
(a) Cyclic voltammograms for peptides 1–6 immobilized on SWCNTs/Au electrodes taken at 5 V s–1. (b) Peak potential versus ln (scan rate) for peptides 1–6 after background current subtraction.A comparison of the data for the
peptides 1, 4 and 5 provides
some insight into the influence
of backbone rigidity, where these peptides share a common 310-helical geometry and the presence of a single alkene. Peptide 1 is constrained and hence rigidified by its tether. Peptide 4 contains five Aib residues, while peptide 5 would be the most flexible of the three with an Ala residue in place
of one Aib at the site of cyclization in 1. The data
on these compounds reveals an electron transfer rate constant for
the macrocyclicpeptide 1 of 17 s–1, a clear 15–20 fold lower than that of peptides 4 and 5. Peptide 4 gave the next lowest
electron transfer rate constant (260 s–1), with
the most flexible peptide 5 displaying a value of 307
s–1. Thus, there is a clear correlation between
the electron transfer rate constant and the flexibility of the peptide
backbone. Increased rigidity impedes electron transfer, presumably
by restricting the precise torsional motions required by a hopping
mechanism, that lead to facile intramolecular electron transfer along
the peptide.[23,42]A dramatic shift to the
positive in the formal potential of the
constrained peptide 1, compared to those of the linear
analogues 4 and 5, was also observed. The
difference between the formal potentials of the constrained (1) and unconstrained (4 and 5) peptides
was a significant 465 mV. This is similar to results from our previous
study involving a hexapeptide that was also stapled i to i + 3, but by a triazolecontaining linker introduced
by an alternative Huisgen cycloaddition strategy (480 mV).[23] Such a marked disparity between the formal potentials
of these linear and macrocyclicpeptides is further evidence of the
additional backbone rigidity imparted by the side-bridge constraint.
The observed effect on electron transfer is the result of cyclization
and the associated rigidification, rather than by the makeup of the
component macrocycle.A comparison of the data for the three
linear hexapeptides (3, 4, and 6) provides a measure
of the influence of the electron rich alkene side-chains on the rate
of electron transfer somewhat in isolation from the effects of backbone
rigidity. Peptide 6, with alkenes at both the i and i + 3 positions, exhibited the largest
electron transfer rate constant of 388 s–1. The
peptidecontaining one alkene side chain (4) gave an
electron transfer rate constant of 260 s–1. Peptide 3, which lacks an alkene side chain in its sequence, gave
a much reduced electron transfer rate constant of 62 s–1. The electron transfer rate constant clearly increases with the
increasing number of electron rich alkenes in the peptides, which
presumably facilitate electron transfer, by way of a hopping mechanism
utilizing the alkenes as “stepping stones”.It
is important to note that the relative rigidity of the backbones
of peptides 3, 4, and 6 may
also contribute to the rate of electron transfer, which would be expected
to decrease with increasing numbers of Aib units through the series
as discussed above and elsewhere.[23] A comparison
of the data for peptides 5 and 6 sheds further
light on this suggestion. These two peptidescontain the same number
of Aib units and differ only in the number of alkenyl groups to act
as potential “stepping stones”. The observed electron
transfer rate constant for 6 was 388 s–1, 20% higher compared to that for 5 (307 s–1). This clearly demonstrates the ability of the alkene groups to
facilitate electron transfer through the peptide by acting as a “stepping
stone”. It is thus clear that a combination of both the electronic
properties and the extent of backbone rigidity determines the rate
of electron transfer in peptides.(a) Cyclic voltammograms for β-strand peptides 7–9 immobilized on SWCNTs/Au electrodes taken
at 5 V s–1. (b) Peak potential versus ln (scan rate)
for peptides 7–9 after background
current subtraction.A comparison of the unsaturated and saturated macrocyclicpeptides 1 and 2 provides further insights
into the role
of these two effects on the efficiency of electron transfer. Unlike
the linear peptides 4, 5 and 6, the planar alkene of 1 is able to influence both the
backbone rigidity and potentially the electronic properties with its
inclusion in a ring. One might expect this alkene to enhance backbone
rigidity, while at the same time providing a potential “stepping
stone” for electron transfer. These effects are opposing, with
the first expected to decrease the electron transfer rate and the
second to increase it. Interestingly, the unsaturated macrocycle 1 gave an approximate 2-fold decrease in the electron transfer
rate relative to 2, with values of 17 and 31 s–1, respectively. This observation is reinforced for the two β-strand
constrained peptides, the unsaturated macrocycle 7 and
the saturated analogue 8 which displayed electron transfer
rates of 11 and 23 s–1, respectively (see Tables 1 and 2). Clearly, increasing
backbone rigidity in both secondary structures (310-helix
and β-strand) decreases the efficiency of electron transfer.
Curiously, the saturated helical peptide 2 exhibited
a formal potential shift to the positive of 37 mV compared to the
unsaturated analogue 1. The saturated β-strand
peptide 8 recorded an even greater formal potential shift
to the positive of 151 mV relative to its unsaturated analogue 7 (see Figure 9). Thus, oxidation/reduction
of the ferrocene moiety in both of the saturated peptides is energetically
less favorable than in the corresponding unsaturated peptides. However,
the observed electron transfer rate constants for both saturated peptides
are almost double that of their unsaturated counterparts. As noted
here for 1, 2, 7, 8, and elsewhere,[23] peptidesconstrained
by a side-chain tether give rise to a significant increase in the
formal potentials relative to their linear analogues, reflecting the
associated increase in backbone rigidity. Thus, any formal potential
shift to the positive is usually combined with a reduction in the
electron transfer rate constant. Therefore, while the effect of backbone
rigidity appears to be the dominant factor in this case, it would
also be expected that the electron rich alkene in the tether of peptides 1 and 7 should enhance electroniccoupling. However,
further investigation is required to substantiate this notion, and
this is developed further in the following computational study section.
Table 2
Electron Transfer
Rate Constants (kapp), Surface Concentrations
and Formal Potentials
(Eo) for the β-Strand Peptides (7–9)
peptide
surface concentration ( × 10–10 mol cm–2)
Eo (V vs AgCl/Ag)
kapp/s–1
7
9.21 ± 0.89
0.676
11.72 ± 1.16
8
7.13 ± 0.68
0.827
23.62 ± 2.13
9
5.56 ± 0.31
0.408
421.36 ± 41.51
Figure 9
(a) Cyclic voltammograms for β-strand peptides 7–9 immobilized on SWCNTs/Au electrodes taken
at 5 V s–1. (b) Peak potential versus ln (scan rate)
for peptides 7–9 after background
current subtraction.
Computational Study on Intramolecular Electron Transfer
High level theoretical calculations, using the latest constrained
density functional theory (cDFT), were conducted on β-strand
models 19, 20, and 21 in order
to provide further insights into the relative roles of backbone rigidity
and electron rich side-chains on intramolecular electron transfer
(see Figure 10). These peptides are analogous
to 7, 8 and 9, but with ferrocene
units included at both termini to act as both donor and acceptor.
Diabatic states were constructed by individually localizing an overall
charge of +1 on each of the amino acids and ferrocene units,[28] as shown in Figure 10. Reorganization energies (λ) for electron transfer along the
backbone were calculated, together with electroniccoupling constants
(Hab) in order to provide an insight into
the overall intramolecular electron transfer dynamics.
Figure 10
Constructed
diabatic states in model peptides 19 (top), 20 (middle) and 21 (bottom).
Constructed
diabatic states in model peptides 19 (top), 20 (middle) and 21 (bottom).A trend between the electroniccoupling constants (Hab) and the number of double bonds in each side-chain
is evident. The structure with the greatest number of electron rich
side-chains (unconstrained peptide 21) has the largest
coupling constant (0.106 eV). This is sequentially followed by the
unsaturated peptide 19 (0.087 eV), and the saturated
peptide 20 (0.049 eV) (see Table 3). Significantly, the electroniccoupling constant for unsaturated 19 (0.087 eV) is almost double that of the saturated analogue 20 (0.049 eV). This clearly indicates that the electron rich
alkene in the tether of the unsaturated peptide does indeed enhance
electroniccoupling.
Table 3
Electronic Coupling
Constants (Hab), the Number of Double
Bonds in Each Side-Chain
and Average Reorganization Energies (λ) for Peptides 19–21, and the Formal Potentials (Eo) and Electron Transfer Rate Constants (kapp) for Their Analogues, Peptides 7–9
peptide
formal potential
(Eo) (V vs AgCl/Ag)
Hab (eV)
number of C=C in side chains
average reorganization
energy (λ) (eV)
kapp/s–1
20 (8)
0.827
0.049
0
0.65
23
19 (7)
0.676
0.087
1
0.74
11
21 (9)
0.408
0.106
2
0.35
421
Furthermore, the formal potential observed
experimentally for the
unsaturated peptide 7 is significantly lower than that
observed for the saturated 8 (see Table 3). This lower potential is clearly attributable to the effect
of the electron rich alkene since the only structural difference between
these constrained peptides is the presence or otherwise of the electron
rich π-bond in the side-chain of 7. In contrast,
the higher reorganization energy calculated for the derivative of 7 (unsaturated analogue 19 (0.74 eV)), relative
to the derivative of 8 (saturated analogue 20 (0.65 eV)), is likely the direct consequence of the lack of rotational
freedom available in the side-chain of the unsaturated peptide. This
leads to an increase in the rigidity of the backbone and consequently
to the lower rate of electron transfer observed for unsaturated 7, compared to saturated 8.Additionally,
a large difference of up to 0.49 eV is apparent between
the reorganization energies of the constrained peptides (19 and 20), and those of the unconstrained 21 (see Table S13, Supporting Information). The unconstrained peptide 21 gave rise to the highest
calculated (Hab) and the lowest calculated
(λ) based on Marcus theory,[43] which
suggests that oxidation/reduction of the ferrocene moiety is energetically
more favorable in the linear peptide, than in either of the constrained
compounds. This supports the earlier experimental observation where
the linear peptide 9 exhibited the lowest formal potential
relative to the two constrained peptides 7 and 8 (by between 268 and 419 mV), and the highest electron transfer
rate constant (an 18–38 fold increase) relative to 7 and 8 (see Table 3). This is
indicative of the additional backbone rigidity imparted by the constraint.
Thus, the influence of the electron rich alkenes and any effects arising
from a change in backbone rigiditycan be studied in isolation using
a combination of experimental and theoretical studies. Both factors
clearly contribute to the rate of electron transfer in peptides.Electron population analysis (EPA) was conducted to further elucidate
the role of electron rich side-chains as “stepping stones”
for electron transfer. The amino acid residues used in the synthesis
of the linear helical peptides (4, 5 and 6), namely, an Aib and a modified serine with electron rich
alkene side-chain were considered. A Löwdin analysis of the
charge distribution from the cDFT calculations on the charged and
uncharged amino acid residues is shown in Table 4. Approximately 88% of the extra charge is distributed on the amide
region when the positive charge (+1) was injected into the Aib residue.
This emphasizes the significant contribution made by the amide region
to intramolecular electron transfer through the peptide backbone,
clearly demonstrating the participation of a through-bond hopping
mechanism.[7] However, when the positive
charge (+1) was injected into the modified serine residue, only 68%
of the extra charge was distributed on the amide region, with the
electron rich alkene side-chain holding approximately 20% of the extra
charge localized on the residue (see Table 4). Thus, these results confirm the role of the electron rich alkene
side-chain as a “stepping stone” for electron transfer.
Table 4
Löwdin Analysis of the Charge
Distribution on Uncharged and Charged Amino Acid Residues (1) Aib
and (2) Modified Serine with Electron Rich Alkene Side-Chain
Conclusion
Electrochemical
studies are reported on a series of peptides (1–9) in order to elucidate the effect
of backbone rigidity and the nature of the amino acid side chains
in defining the rate of electron transfer. Aib residues were incorporated
into peptides 1–6 to promote the
formation of a unifying 310-helical secondary structure,
with the number of alkenes in their side-chains varying from 0 to
2. The backbones of peptides 1 and 2 were
further constrained into a 310-helix with a side chain
tether introduced by RCM. The side-chain of 1 contains
a single C=C double bond, while peptide 2 is fully
saturated. Peptides 7–9 share a common
β-strand conformation, with 7 (unsaturated) and 8 (saturated) further rigidified into this geometry via cyclization
by RCM. Electrochemical studies conducted on peptides 1, 4 and 5, each containing a single alkene
in their structure, revealed a direct link between backbone rigidity
and the efficiency of electron transfer. The significant difference
in the formal potentials of the constrained 1 and unconstrained 4 and 5, (465 mV) is in accordance with our previous
study (480 mV).[23] This demonstrates a general
observation, where a tether hinders electron transfer in peptides
by restricting backbone flexibility. Further studies on the linear
peptides 3–6, containing between
0 and 2 electron rich side-chains in their structure, confirmed the
ability of the alkene to facilitate electron transfer through the
peptide, while nullifying the effects of backbone rigidity.The macrocyclic helical peptides reveal a formal potential shift
to the positive and subsequent reduction of the electron transfer
rate constant for unsaturated 1, relative to saturated 2. Comparable results were also evident for the unsaturated
(7) and saturated (8) β-strand peptides.The only structural difference between the unsaturated and saturated
peptides is the presence or otherwise of the electron rich π-bond
in the side-chain, so the lower the electron transfer rate constants
observed for both unsaturated peptides (1 and 7) are likely a direct consequence of the lack of rotational freedom
about this double bond, which results in further rigidification of
the peptide backbone. High level calculations performed on peptides 19 and 20 (analogues of 7 and 8) confirmed that the reorganization energy is greater in
the unsaturated peptide (19), thus supporting the observed
lower electron transfer rate constant of 7, relative
to saturated 8. However, the lower formal potential observed
experimentally for the unsaturated 7, suggests that oxidation/reduction
of the ferrocene moiety is energetically more favorable in 7. Theoretical calculations show that the derivative of the unsaturated
peptide 7 (19), exhibited a higher electroniccoupling constant than its saturated counterpart 20,
which helps to explain this paradox, while further demonstrating the
ability of the alkene to facilitate electron transfer. Thus, the theoretical
electroniccoupling constants and reorganization energies, together
with the formal potentials and electron transfer rate constants observed
experimentally for the macrocyclicpeptides, reveal for the first
time an interplay between electron rich alkene side-chains and backbone
rigidity, with both factors clearly shown to contribute to the efficiency
of electron transfer in peptides. Additional high level calculations
were also performed on the amino acid residues used in the synthesis
of the linear helical peptides (4, 5 and 6), namely, an Aib and a modified serine with electron rich
alkene side-chain. Injection of a positive charge into the modified
serine residue shows that approximately 20% of the extra charge is
localized on the electron rich side-chain, so confirming the role
of the alkene as a “stepping stone” for electron transfer.These findings provide a new approach to fine-tune the electronic
properties of peptides through chemical modification of the backbone
to increase/decrease rigidity, and through the inclusion of electron
rich side-chains. Such structurally diverse peptides with controllable
electronic functions open new avenues in the design and fabrication
of efficient components for molecular-based electronic devices.
Authors: Bushra R Chaudhry; James D E T Wilton-Ely; Alethea B Tabor; Daren J Caruana Journal: Phys Chem Chem Phys Date: 2010-07-14 Impact factor: 3.676
Authors: Amie K Boal; Ivan Guryanov; Alessandro Moretto; Marco Crisma; Erica L Lanni; Claudio Toniolo; Robert H Grubbs; Daniel J O'Leary Journal: J Am Chem Soc Date: 2007-05-11 Impact factor: 15.419
Figure 1
Figure 2
Figure 3
Figure 4
Figure 6
Figure 5
Figure 7
Figure 8
Figure 9
Figure 10