The power of electrified nanoconfinement for energising, controlling and observing long enzyme cascades

Multistep enzyme-catalyzed cascade reactions are highly efficient in nature due to the confinement and concentration of the enzymes within nanocompartments. In this way, rates are exceptionally high, and loss of intermediates minimised. Similarly, extended enzyme cascades trapped and crowded within the nanoconfined environment of a porous conducting metal oxide electrode material form the basis of a powerful way to study and exploit myriad complex biocatalytic reactions and pathways. One of the confined enzymes, ferredoxin-NADP+ reductase, serves as a transducer, rapidly and reversibly recycling nicotinamide cofactors electrochemically for immediate delivery to the next enzyme along the chain, thereby making it possible to energize, control and observe extended cascade reactions driven in either direction depending on the electrode potential that is applied. Here we show as proof of concept the synthesis of aspartic acid from pyruvic acid or its reverse oxidative decarboxylation/deamination, involving five nanoconfined enzymes.


Supplementary Figure 1. Calculation of FNR coverage from non-turnover peaks.
By integrating the charge passed (coulombs) under each of the non-turnover peaks (and taking an average) it is possible to determine the amount of electroactive FNR absorbed on the electrode using the relationship -Quantity (in moles) = charge/nF, where n is the number of electrons involved (2) and F is the Faraday constant.

Supplementary Figure 3. Cyclic voltammograms of the 4-enzyme cascade in
CO2-saturated buffer. Enzyme ratio in droplet applied to the electrode: 1 FNR / 5 ME / 1 FumC / 1 AspA. Buffer: 0.2 M MOPS, 0.1 M KHCO3, 4 mM MgCl2, 1 mM MnCl2, 20 mM pyruvate in 100% CO2 (initial pH 7.5). Scan rate: 1 mV/s; 25°C. Grey: Blank, no cofactor present. Red: after injection of NADP + to a final concentration of 20 μM. Blue: after injecting L-aspartate to a final concentration of 20 mM. Figure 4. Reductive amination/carboxylation of pyruvate to aspartate. Conditions: stirring, electrode potential = -0.45 V vs SHE, 25°C. Buffer: 0.05 M HEPES, 4 mM MgCl2, 1 mM MnCl2, 20 μM NADP + , pH 7.5. A high surface area electrode (12 cm 2 ) was used to increase yield for detection of products by 1 H-NMR. The nanoporous indium tin oxide was loaded with the following enzyme ratio: 0.1 CA / 1 FNR / 2 ME /1 FumC / 1 AspA. Injection of KHCO3 (to 0.1 M) triggered the start of the cascade reaction. The rate (reduction current) rapidly increased and remained stable for 5 hours after which it decreased over the course of 1 day. After 20 hours (blue arrow) the solution (4.5 mL) was tested by 1 H-NMR and contained 6.80 mM aspartate, 0.06 mM fumarate and 1.4 mM malate. Thus, a conversion of 34% to aspartate was obtained after 24 hours with a total turnover number (TTN, , 20 μM NADP+, pH 7.5. Electrode surface area: 3.5 cm 2 . The nanoporous indium tin oxide (ITO) electrode was loaded with a droplet having the following enzyme ratio: 0.1 CA / 1 FNR / 1 ME / 1 FumC / 1 AspA. Injection of L-aspartate (to 20 mM) triggered the start of the cascade reaction; the rate rapidly increased (oxidation current became more positive). After 16 hours (blue arrow) the solution (4 mL) was tested by 1 H-NMR. Stability of the confined enzyme cascade was confirmed since replacement of the solution with fresh buffer and recharging with substrate resulted in a rate close to the original value, and the same distribution of intermediates and product was maintained (Figure 4 in main paper).

Supplementary Figure 12. Control experiment showing that aspartate is not electroactive at the potential applied and injection of E1 to E4 initiates catalysis.
Conditions: stirring, electrode potential = +0.1 V vs SHE, 25°C. Electrode surface area 2.9 cm 2 . Buffer: 0.05 M MOPS, 50 mM L-aspartate, 4 mM MgCl2, 20 mM MnCl2, 20 μM NADP + , pH 7.5. Arrow indicates injection of E1 to E4 at a ratio of 1 FNR / 5 ME / 1 FumC / 1 AspA at a final concentration of 0.5 µM in the case of FNR, FumC and AspA, while ME was injected to a final concentration of 2.5 µM. Figure 13. Cascade activity in solution. a: Full view. b: Truncated view to show the lag; inset: magnification of the start of experiment 1 (red). Benzyl viologen (50 mM) was used as a reporter and mediator for cofactor recycling by FNR; activity monitored using the absorbance at 600 nm due to reduced benzyl viologen. 20 µM NADP + , 20 mM aspartic acid, activity initiated by the addition of aspartic acid. c: Magnification of the initiation by aspartic acid addition in the electrochemistry experiment shown in Figure 3d, note the immediacy and absence of a lag. For a and b: experiment 1 was a single measurement; experiment 2 was performed in duplicate with the reaction initiated by addition of aspartate; experiment 3 was a single measurement; experiment 4 was performed in duplicate.  Figure S8); overall molecular mass of cascade enzymes = 340,000 Da. In electrochemistry experiment 6 (corresponding to Figure 3) the cascade ratio used was 1FNR:5ME:1FumC:1AspA; total molecular mass of cascade enzymes = 532,000 Da. Molecular masses based on monomers of each enzyme.

Supplementary
* Mediated experiments in solution: 50 mM benzyl viologen, 20 µM NADP + , 20 mM aspartic acid * Rate given as moles product per total number of moles of enzyme.
* Rates in brackets were obtained as a linear fit to the steepest part of the traces in Figure S13 and as such have not yet reached steady state.
* The volume of a porous electrode of 1 cm x 1 cm x 3 µm is 0.3 µL; for the hypothetical [cascade] in an electrode of these dimensions, the calculation is based on the total volume occupied by the ITO layer: this therefore is an overestimation of the true volume of the pores and as such the predicted concentrations would be even higher.