The Influence of Secondary Structure on Electron Transfer in Peptides

Electron transfer occurs in proteins over surprisingly long molecular distances of up to 100 Å to facilitate, for example, the function of mitochondria and the process of photosynthesis in green plants and algae.^[1,2] Several factors are thought to influence the kinetics of these processes, including peptide chain length, dipole orientation, backbone conformation, and hydrogen bonding and hence secondary structure.^[3] However, the exact role that these and other factors play in defining the associated mechanisms remains somewhat contentious.^[4] A fundamental understanding of how electron transfer occurs in proteins is needed to progress the field and also to provide an opportunity for the development of applications in the area of molecular electronics.^[5,6] With this in mind, we recently initiated experimental and theoretical studies on electron transfer in a series of model peptides that are known to adopt secondary structure.

Two mechanisms, namely bridge-assisted superexchange and electron-hopping, were originally proposed to rationalise long-range electron transfer in DNA, particularly the effect of the separation of the donor and acceptor (distance dependence) groups.^[7] However, these mechanisms remain somewhat controversial, with conflicting reports on their application to peptides.^[8,9] In an attempt to begin to clarify this we recently reported^[10] electron transfer studies on a series of oligomers [H-(NH-C(CH₃)₂-CO)_n-NH-CH₂-Fc, n = 0–5] of α-aminoisobutyric acid (Aib) that contain a C-terminal ferrocene (Fc) for electrochemical studies and a free N-terminus for attachment to a single-walled carbon nanotube array/p-silicon (100) electrode. Oligomers of geminally disubstituted Aib units were used in this study since relatively short sequences (3 or more residues) are known to induce predictable and stable 3₁₀-helical structures.^[11] The conformations of these peptides were confirmed by 2D NMR to be helical for the longer synthetic peptides (n = 3–5) and ill-defined for the shorter peptides (n = 0–2). The shorter peptides cannot adopt a 3₁₀-helix since the requisite i to i+3 intramolecular hydrogen bonds cannot form.

The electron transfer rate constants (k_app) were determined for each of the peptides and these values were plotted against the Fc iron-to-nitrogen terminus distance as shown in Fig. 1a. A slope transition is apparent for the peptide containing three Aib units (see the 4th data point from the left in this Figure).^[10] It should be noted that the efficiency of electron transfer in peptides is measured by the attenuation constant (β), which provides information on the electron transfer mechanism.^[12] Shorter sequences (n = 0–2) gave rise to a steep decrease on increasing the chain length. The apparent rate attenuation constant (β) for these peptides was calculated to be 0.84 Å^–1.^[10] This is consistent with an electron superexchange mechanism.^[7] The longer helical peptides (n = 3–5) display a weak dependence of the electron transfer rate constant on distance. The rate attenuation constant (β) for these peptides was estimated to be 0.10 Å^–1, as evidenced by the shallow slope of the last 3 points,^[10] indicative of an electron hopping mechanism.^[7]

Fig. 1.  (a) Mechanistic transition in a series of oligomers of α-aminoisobutyric acids. (b) Dependence of k_app on the number of intramolecular hydrogen bonds. Data points from left to right represent an ascending order of n (n = 0–5).

The corresponding electron transfer rate constants (k_app) were plotted against the number of intramolecular hydrogen bonds as shown in Fig. 1b.^[10] The lack of intramolecular hydrogen bonding and hence secondary structure associated with the shorter peptides (n = 0–2) (see the first three data points from Fig. 1a) gives rise to electron transfer via a superexchange mechanism. By comparison, the longer peptides (n = 3–5) have a helical structure as defined by a series of well defined intramolecular hydrogen bonds. Here a strong dependence between k_app and the number of intramolecular hydrogen bonds is observed. It is likely that intramolecular hydrogen bonding facilitates electron transfer in structurally well defined peptides by way of a hopping mechanism. This study suggests that the mechanism of electron transfer is dictated by the extent of secondary structure and hence the degree of intramolecular hydrogen bonding, rather than merely by chain length.

Two β-peptides [PCB-(β³Val-β³Ala-β³Leu)_n–NHC(CH₃)₂OOtBu, n = 1 and 2, designated β3 and β6 respectively; PCB = p-cyanobenzamide] were also synthesised and their electron transfer studied in order to gain further insight and support for our earlier observations.^[13] β-Peptides were used in this study as they are known to adopt helical secondary structures with high predictability, structural diversity, and stability.^[14–18] These structures are ideal for studying the fundamental mechanisms of electron transport, but they have as yet received scant attention. In this case, an N-terminal tert-butyl hydroperoxide was used as the electron acceptor and a C-terminal p-cyanobenzamide as the donor. By using hydroperoxide and cyanobenzamide, the electrochemistry can be determined in solution, whereby the conformational freedom of the peptide is increased.^[4] Circular dichroism (CD) and 2D NMR analysis on β6 revealed that its backbone readily folds into the expected low-energy hydrogen-bonded 3₁₄-helical structure,^[3,19] which is consistent with gas-phase geometry optimisation using the density functional theory (DFT) method. Gas-phase geometry optimisation on β3 suggested that it adopts an ill-defined conformation. Electron transfer rate constants of 2580 s^–1 and 9.8 s^–1 were determined in solution for β3 and β6, respectively.^[13] Based on these values it is clear that the two peptides have different efficiency of electron transfer, suggesting a likely difference in mechanism. Based on this a detailed computational study was undertaken as discussed below.

Latest constrained density functional theory (cDFT) calculations relating to the energy of each diabatic state for both β-peptides gave rise to accurate potential surfaces that were used to estimate the driving force (ΔG), reorganisation energy (λ), and the electronic coupling matrix element (H_ab). These parameters allowed the electron transfer rate constant to be calculated using Marcus theory. The diabatic states in both β-peptides were obtained by individually localising an overall charge of –1 on each of the donor (D), amino acids, and the acceptor (A). Fig. 2 shows constructed diabatic states for both peptides and the possible electron transfer pathways.

Fig. 2.  Constructed diabatic states and the possible electron transfer pathways in (a) β3 and (b) β6. Computed electron transfer rate constants for each electron transfer step are labelled along each arrow, with units of s^–1.

For β3, the electron transfer rate constants for the forward steps D→B1, B3→A, and backward steps A→B3 along the peptide backbone are extremely small, indicating a highly improbable electron transfer route. This is consistent with a superexchange electron transfer pathway for this peptide as depicted in Fig. 3 (top). In comparison, the electron transfer rate constants for the forward (D→A) and backward (A→D) steps for β6 are exceptionally small (see Fig. 2). This implies that a superexchange mechanism does not occur in this case, with the alternative electron hopping pathway operating. These results provide theoretical evidence for the two model peptides following different electron-transfer pathways, presumably again due to different conformations and hence secondary structure.

Fig. 3.  Favoured electron transfer pathways in both β-peptides, superexchange for the ill-defined β3 (top) and hopping for the helical β6 (below) using computed electron transfer rate constants.

Electron transfer studies on a series of oligomers of α-aminoisobutyric acids revealed a change from an electron superexchange to a hopping mechanism on transitioning from an ill-defined to a helical conformation. Further studies involving β–peptides provided theoretical evidence to corroborate these findings. Together these results clearly show that the mechanism of electron transfer in peptides is defined by the extent of secondary structure, rather than simply chain length. Other factors may influence the electron transfer mechanism in peptides, such as dipole orientation and the molecular dynamics of the backbone. Further research is pivotal to advancing the field and to provide opportunities for the use of peptides in molecular electronics applications. All this work is based on an ability to define and control the shape or conformation of peptides.^[20–22]