NADP(H)-dependent biocatalysis without adding NADP(H)

Abstract

Isocitrate dehydrogenase 1 (IDH1) naturally copurifies and crystallizes in a resting state with a molecule of its exchangeable cofactor, NADP⁺/NADPH, bound in each monomer of the homodimer. We report electrochemical studies with IDH1 that exploit this property to reveal the massive advantage of nanoconfinement to increase the efficiency of multistep enzyme-catalyzed cascade reactions. When coloaded with ferredoxin NADP⁺ reductase in a nanoporous conducting indium tin oxide film, IDH1 carries out the complete electrochemical oxidation of 6 mM isocitrate (in 4mL) to 2-oxoglutarate (2OG), using only the NADP(H) that copurified with IDH1 and was carried into the electrode pores as cargo-the system remains active for days. The entrapped cofactor, now quantifiable by cyclic voltammetry, undergoes ~160,000 turnovers during the process. The results from a variety of electrocatalysis experiments imply that the local concentrations of the two nanoconfined enzymes lie around the millimolar range. The combination of crowding and entrapment results in a 10² to 10³-fold increase in the efficiency of NADP(H) redox cycling. The ability of the method to drive cascade catalysis in either direction (oxidation or reduction) and remove and replace substrates was exploited to study redox-state dependent differences in cofactor binding between wild-type IDH1 and the cancer-linked R132H variant that catalyzes the "gain of function" reduction of 2OG to 2-hydroxyglutarate instead of isocitrate oxidation. The combined results demonstrate the power of nanoconfinement for facilitating multistep enzyme catalysis (in this case energized and verified electrochemically) and reveal insights into the dynamic role of nicotinamide cofactors as redox (hydride) carriers.

Keywords: NADPH; biocatalysis; electrocatalysis; isocitrate dehydrogenase; nanoconfinement.

Scheme 1.

Electrochemical Leaf and wild-type and variant IDH1 reactions. (A) Principle of the Electrochemical Leaf: NADP(H) is recycled between FNR and an NADP(H)-dependent dehydrogenase; “D” represents the “dashboard,” the equipment used to control and monitor the reaction, which is composed of a potentiostat, a computer, and an electrode rotator. (B) The reversible wild-type IDH1 reaction. (C) The neomorphic reaction catalyzed by cancer-associated IDH1 variants, including IDH1 R132H. Molecules are shown as the forms predominating at pH = 7 to 8 (8).

Fig. 1.

¹H NMR data showing the release of copurified enzyme-bound NADP(H) from wild-type IDH1 (A) and IDH1 R132H (B) upon thermal denaturation (see SI Appendix, Fig. S3 for the equivalent experiment with FNR). Reading vertically from the lower spectra i) folded IDH1 protein—NADP(H) is initially enzyme-bound, hence not observed; ii) denaturation releases NADP(H) into solution. (A) Wild-type IDH1 copurifies with an approximately 2:1 mixture of NADP⁺ (red asterisk) and NADPH (blue asterisk). (B) IDH1 R132H apparently copurifies exclusively with NADPH.

Fig. 2.

Electrochemical nanoconfinement experiments demonstrating that wild-type IDH1 activity is clearly observed using only IDH1-copurified NADP(H) (no NADP(H) was added). (A) Chronoamperogram showing IDH1 activity as increasing concentrations of DL-isocitrate are titrated into the solution. The injection concentrations shown are the final concentrations of each addition of isocitrate (not cumulative). (B) Cyclic voltammetry showing wild-type IDH1 activity at different concentrations of isocitrate. The gray trace shows IDH1 activity when 10 µM NADP⁺ was added to the solution. Conditions (A and B): (FNR+IDH1)@ITO/PGE electrode, temperature 25 °C, volume 4 mL, pH = 8 (100 mM HEPES, 10 mM MgCl₂), and enzyme loading ratios (molar): FNR/IDH1; 2/1. (A): electrode area 0.06 cm², 1,000 rpm, potential E (vs. standard hydrogen electrode, SHE) = +0.2 V. (B): 0.03 cm² electrode area, scan rate 1 mV/s, stationary electrode. Racemic DL-isocitrate was used for both A and B.

Fig. 3.

Scaled-up and time-extended wild-type IDH1 experiment (4 cm² electrode, 4 mL stirred solution) showing quantitative conversion of 6 mM D-isocitrate to 2OG using only IDH1-copurified NADP(H). (A) Chronoamperogram showing IDH1 activity over 5 d. (B) Bar chart comparing predicted yields (based on charge passed) with ¹H NMR-quantified substrate (D-isocitrate) and product (2OG) at each time point (TP). The NMR measurements correspond to samples taken from the working electrode solution, and the concentrations shown are corrected for the volumes of samples taken/injections made. Conditions for (A): (FNR + IDH1)@ITO/Ti foil electrode, area 4 cm², solution agitated by stirring, temperature 25 °C, volume 4 mL, potential E (vs. SHE) = +0.2 V, pH = 8 (100 mM HEPES, 10 mM MgCl₂). Enzyme loading ratios (molar): FNR/IDH1; 2/1. *Note: After completion of the electrochemical experiment, NMR analysis of the counter electrode solution showed that 0.78 µmol of 2OG and 2.6 µmol of isocitrate were present, having crossed through the glass frit from the working electrode chamber during the experiment. The “crossed-over” 2OG was added to the time point 4 NMR value to show the total product made. When the isocitrate in the counter electrode solution is included, the residual isocitrate (0.661 mM) + 2OG produced (5.88 mM) = 6.541 mM at time point 4, a value close to the total 6.5 mM of isocitrate initially added.

Fig. 4.

Scan rate-dependent Faradaic capacity (A and B) and trumpet plots (C and D) from thin-film voltammetry experiments using electrodes coloaded with FNR and wild-type IDH1 compared to an FNR-only electrode. At low scan rates, the NADP(H) carried in with IDH1 can be detected; peaks collapse to the FNR-only signal at high scan rates. (A and B) Scan rate-dependent coverage plots fitted with an asymptotic exponential equation to allow extrapolation to 0 mV/s. (C and D) Trumpet plots showing the changes in oxidation and reduction peak potentials as a function of scan rate (see trend lines). Conditions: stationary (FNR + IDH1)@ITO/PGE electrode (except for FNR-only data, which did not contain IDH1), electrode area 0.06 cm², temperature 25 °C, volume 4 mL, pH = 8 (100 mM HEPES), and enzyme loading ratios (molar): (A and C): FNR/IDH1; 2/1; (B and D): FNR/IDH1; 2/5.

Fig. 5.

Chronoamperometry experiments showing steady-state catalysis by wild-type IDH1 (isocitrate oxidation to 2OG) and IDH1 R132H (2OG reduction to 2HG) using only copurified enzyme-bound NADP(H), interrupted by live buffer exchanges and potential switches. In panels A–D, the first buffer exchange (>1,000-fold dilution) used the starting buffer solution with 5 mM substrate (DL-isocitrate or 2OG) added to maintain IDH1 catalysis, while the second buffer exchange (~55,000-fold dilution) was performed using the starting buffer solution without any added substrate. (A) and (C) are equivalent experiments except that the potential in (C) was switched from oxidizing (+0.2 V) to reducing (−0.5 V) (timespans indicated by gray boxes) to convert all of the nicotinamide in the pores to NADPH. (B) and (D) are equivalent experiments with the exception of the potential switch to an oxidizing value (+0.2 V). (E and F) Oscillating potential switch experiments showing that wild-type IDH1 has a high affinity for both NADP⁺ and NADPH (panel E), whereas IDH1 R132H has a much lower affinity for NADP⁺ compared to NADPH (panel F). Conditions: (FNR + E2)@ITO/PGE electrode (where E2 represents wild-type IDH1 or IDH1 R132H), area 0.06 cm², rotated at 1,000 rpm, temperature 25 °C, volume 2.6 mL, potential E (vs. SHE) = +0.2 V for wild-type IDH1 or –0.5 V for IDH1 R132H (except where potential switches are indicated by gray boxes), and pH = 8 (100 mM HEPES, 10 mM MgCl₂). Enzyme loading ratios (molar): FNR/IDH1 WT; 2/1 or FNR/IDH1 R132H; 1:2.

Fig. 6.

Nanoconfined enzyme kinetics measuring the activity of the IDH1-FNR cascade (A) and FNR alone (B) fitted using derived electrochemical-kinetic equations [Eq. 2 (A) and Eq. 1 (B)]. The same electrode (with FNR and IDH1 loaded in a 2:1 ratio) was used in (A) and (B): (A) Rate of isocitrate oxidation at increasing concentrations of NADP⁺, where NADP(H) is recycled between IDH1 and FNR. (B) Rate of NADPH oxidation by nanoconfined FNR at increasing concentrations of NADPH. (A and B) The enzyme coverages (shown in black) were used as inputs for the fitted equations: results (shown in blue) were determined by fitting the equations to the data. The data in (A) were fitted using Eq. 2 to account for the slight substrate inhibition observed (K_i(nano) = 11 mM). Both equations fitted to the data sets had an R² > 0.99. (C) Data from (A) and (B) compared on a semilog plot. (D) The first four data points in panel A were extrapolated to the x-intercept to estimate the effective NADP(H) solution concentration of the nanoconfined IDH1-copurified NADP(H) (see text). Conditions (A and B): (FNR + IDH1)@ITO/PGE electrode, area 0.06 cm², temperature 25 °C, volume 4 mL, potential E (vs. SHE) = +0.2 V, pH = 8 (100 mM HEPES), and enzyme loading ratios (molar): FNR/IDH1; 2/1. (A): buffer also contained: 10 mM MgCl₂, 10 mM DL-isocitrate; rotated at 1,000 rpm. (B): rotated at 4,000 rpm.