Intermolecular Electron-Transfer Catalyzed on Nanoparticle Surfaces

Abstract

Surface-functionalized nanoparticles enhance the rate of electron transfer (ET) between Cyt c(Fe²⁺) and Co(phen)₃³⁺ by a factor of 10⁵ through simultaneous electrostatic binding of ET donor and acceptor.

Interprotein electron transfer (ET) is crucial to energy transduction in photosynthesis and respiration, and is a direct example of protein recognition coupled with chemistry. The effects of distance and driving force are well understood, and can be manipulated in synthetic and natural systems to enhance ET kinetics.^1,2 In contrast, the effects of surface binding and dynamics on ET rates are harder to interpret,^3-6 making new model systems crucial for understanding the role of dynamics in ET reactivity.

Nanoparticles are capable of catalyzing selected reactions, by acting as artificial receptors for substrates.^7,8 For example, the rates of peptide ligation⁷ and phosphodiester cleavage were enhanced by up to 10³–fold on functionalized nanoparticle surfaces.⁹ Here, we show a 10⁵-fold enhancement of an intermolecular ET rate through use of surface-functionalized nanoparticles (Au-TX) as artificial receptors. This catalysis arises through bringing ET donor and acceptor into close proximity, demonstrating the ability to modulate protein ET at surfaces, and provides a tool for observing and controlling the interplay between binding, dynamics, and reactivity.

Bimolecular ET kinetics follow Marcus theory,² with driving force, reorganization, and steric factors determining rate. Cytochrome c (Cyt c) undergoes rapid ET with other proteins, as well as with small molecules. The dominant ET pathway is over the exposed heme edge; however heme access is limited by surface Lys residues,¹⁰ making ET selective for partners that have the appropriate binding surface.

Au nanoparticles functionalized with thiol ligands containing the free-carboxylate form of amino acids (Au-TX)⁸ bind to Cyt c (pI = 10) with K_S ∼ 10⁷ M^-1. ^11,12 Moreover Au-TX binds selectively to the surface of Cyt c near the heme edge.¹¹ As the charges of both Co(phen)₃³⁺ and Cyt c are complementary to that of Au-TX, we tested whether concurrent binding could catalyze ET. The kinetic scheme accounting for the pre-equilibrium binding of Cyt c to Au-TX, followed by Co(phen)₃³⁺ binding (K_D) and ET (k_ET):

$$\text{Cyt}{\text{c}}^{2+}+\text{Au}-\text{TX}\stackrel{K_{\text{S}}}{\rightleftharpoons }\text{Cyt}{\text{c}}^{2+}:\text{Au}-\text{TX}+{\text{Co}(\text{phen})}_3^{3+}\stackrel{K_{\text{D}}}{\rightleftharpoons }\stackrel{k_{\text{ET}}}{\to }$$ (1)

Cyt c was reduced with dithionite followed by gel filtration, and Co(phen)₃³⁺ synthesized as per the literature. Cyt c oxidation was monitored as a single exponential decrease in A₅₅₀ following the mixing of Cyt c (5 μM after mixing, 10mM Tris, pH 7.40, I = 20 mM, 12 mM NaCl) with Co(phen)₃³⁺ (0.083 – 1.00 mM after mixing, in same buffer) in a stopped-flow spectrometer. Cyt c oxidation by Co(phen)₃³⁺ was nearly independent of [Co³⁺], with k_ET = 0.66 (±0.02) s^-1 (Figure 2),¹³ in good agreement with prior reports.¹⁴ This has been attributed to a rate-limiting conformational change at the heme edge of Cyt c; however, under moderate ionic strength the bimolecular ET rate is k_ET/K_D = 10³ M^-1 s^-1.¹⁵

Figure 2.

Cyt c oxidation (5 μM) by Co(phen)₃³⁺ in the presence of varied Au-TAsp (0 – 0.52 μM); buffer is 10 mM Tris, pH 7.40, 12 mM NaCl, I = 20 mM, 25.0°C.

The Cyt c:Au-TAsp adduct was formed by pre-incubating Cyt c (10 μM) with Au-TAsp (0.16 – 1.04 μM; higher concentrations were not used due to high optical density), then mixing with Co(phen)₃³⁺ as above. Cyt c:Au-TAsp oxidation was evident as single-exponential decays in A₅₅₀, indicating that Cyt c binding to Au-TAsp was in a rapid pre-equilibrium (Eq. 1). As the position of this equilibrium favored unbound Cyt c, the ET process may be considered as coupled.¹⁶ The observed rates for Cyt c:Au-TAsp exhibited saturation due to Co(phen)³⁺ binding, with k_obs as high as 65 s^-1 (Figure 2). Analogous data sets for Cyt c:Au-TPhe exhibited similar saturation kinetics.¹³ The second order kinetics were fitted to obtain apparent rate constants for Cyt c oxidation: k_obs = k_ET(app)[Co]₀/([Co]₀+K_D).

The apparent maximum ET rate, (k_ET)_app, was a linear function of nanoparticle concentration, indicating that the binding of Cyt c to Au-TX (K_S of Eq. 1) was far from equilibrium (Figure 3). Under rapid pre-equilibrium conditions, (k_ET/K_S)[Au-TX] = (k_ET)_app; the slope of Figure 3 is the absolute bimolecular rate constant for Co(phen)₃³⁺ reacting with Cyt c on the Au-TX surface. Linear least-squares fitting for Cyt c:Au-TX (X = Asp, Phe) yielded slopes that were nearly identical: k_ET/K_S = 1.35 (±0.03) × 10⁸ and 1.20 (±0.03) × 10⁸ M^-1 s^-1, respectively. This is significantly larger than the ET rate in the absence of Au-TX (1 × 10³ M^-1 s^-1),¹⁵ indicating that complexation by Au-TX catalyzed the ET reaction by 10⁵.

Figure 3.

Bimolecular rate plot for Cyt c oxidation (5 μM, 10mM Tris, pH 7.4, I = 20 mM) by Co(phen)₃³⁺ in the presence of Au-TX (X = Asp, Phe). Linear fitting to (k_ET/K_S)[Au-TX] = k_ET(app) yielded k_ET/K_S = 1.35 (±0.03) × 10⁸ (X = Asp) and 1.20 (±0.03) × 10⁸ M^-1 s^-1 (X = Phe). Error bars are approximately the size of each data point.

This catalysis can be understood within the context of Marcus theory,² which states that the rate of bimolecular ET depends on the collision rate (Z ∼ 1 × 10¹¹ M^-1 s^-1) and the activation energy (ΔG*): k = Zexp(-ΔG*/RT). In the limit of low driving force (ΔG⁰ ≪ λ), ΔG* = λ/4 + w, where w is the work to bring the reactants together. The reaction of Cyt c(Fe²⁺) with Co(phen)₃³⁺ (k = 1 × 10³ M^-1 s^-1) can be attributed to λ = 44 kcal/mol, assuming that w = 0. This λ is consistent with the self-exchange reaction rate for each reagent.² The faster rate in the presence of Au-TX may arise from changes in either λ or w; should λ remain unchanged, the data could be accommodated with w = - 7 kcal/mol. This value for work is attributed to the electrostatic attraction of Co(phen)₃³⁺ and Cyt c to Au-TX.

This electrostatic attraction increases the local concentration of each reagent. Calorimetric and kinetic data indicate that the Co(phen)₃³⁺:Au-TX binding equilibrium was saturated at 1 mM [Co(phen)₃³⁺], leading to the plateau in k_obs (Figure 2); in contrast, Cyt c binding to Au-TX was sub-saturating, leading to the linear dependence of k_ET on [Au-TX] (Figure 3). Entropic factors (i.e. the “Circe effect”¹⁷) presumably dominate intermolecular ET catalysis in the Cyt c:Au-TX system, however enthalpic contributions cannot be excluded.

Kinetic complexity masked the rate of electron flow within encounter complexes in previous studies of the relationship between conformational dynamics and interprotein ET. In some cases, the overall rate was limited by a conformational change, leading to gated ET.^16,18,19 In other cases, ET was rate-limiting but occurred following a disfavored conformational change with concomitant coupled ET.^3,4 Although amide H/D exchange showed that Au-TPhe bound to a smaller Cyt c face than Au-TAsp,¹¹ which could increase the rate of conformational dynamics, this did not translate into faster ET kinetics for Cyt c:Au-TPhe. This suggests that conformational dynamics within the {Cyt c:Au-TX:Co³⁺} encounter complex are much faster than ET, and that dynamics may be uncoupled from ET under these conditions.

In summary, we have demonstrated the use of nanoparticles as highly efficient catalysts for intermolecular ET. These catalysts function by reversibly binding the ET donor and acceptor, thus increasing the local concentration of the redox partners. This process may provide a model for understanding intermolecular ET, in which binding is coupled to reactivity.

Figure 1.

a) Encounter complex for election transfer from Cyt c to Co(phen)₃³⁺ on the surface of Au-TX; b) structure of TX ligands.