E. coli-Based Semi-Artificial Photosynthesis: Biocompatibility of Redox Mediators and Electron Donors in [FeFe] Hydrogenase Driven Hydrogen Evolution

Powering E. coli with sunlight: the optimization of a synthetic photocatalytic system allows E. coli to function as a robust and efficient platform for semi-artificial photosynthesis.


I. Experimental section I.1. General
All chemicals were purchased from Sigma-Aldrich, Merck, and VWR, and used as received unless otherwise stated.All anaerobic work was performed in an MBRAUN glovebox under argon atmosphere.

I.2. Strains and plasmids
The expression vector and the gene coding for CrHydA1 were kindly provided by Prof.
Marc Fontecave (College de France, Paris/CEA, Grenoble).The gene encoding for Tam HydS was synthesized and cloned in pET-11a(+) by Genscript® using restriction sites NdeI and BamHI following codon optimization for expression in E. coli.The expression vector and the gene coding for CpI were kindly provided by Dr. Jifu Duan and Prof. Thomas Happe (Ruhr-Universität Bochum).The plasmids were used to transform the BL21(DE3) E. coli strain, which was used in the presented study.

I.4. Whole-cell sample preparation
Escheria coli whole-cell holo-CrHydA1 samples were prepared as previously reported. 7 short, cell cultures were grown in supplemented M9 media (70 mM phosphate buffer, 0.5 g/L NaCl, 1 g/L NH4Cl, 1 mM MgSO4, 100 µM CaCl2, 4 g/L glucose, 100 µg/mL ampicilline) at 37 °C until the optical density at 600 nm was 0.6-0.8.The expression of the enzyme was induced by addition of IPTG (1 mM) and FeSO4 (100 µM) at 20 °C for 14-16 hours.After washing, the cells are transferred into a glovebox and all subsequent steps are carried out under anaerobic conditions.The cells are artificially matured by addition of [2Fe] adt (80 µM) for 1 h at 37 °C.The artificially matured cell suspensions were washed three times in Tris-HCl buffer supplemented with 150 mM NaCl (pH 8).The washed cells were then suspended in phosphate buffered saline (PBS, 100 mM NaH2PO4/Na2HPO4 with 150 mM NaCl, pH 6.5 or 7.5), supplemented with 100 mM of a sacrificial electron donor (SED), to yield an OD600 of 50.
Photocatalytic assays were prepared by taking 100 µL of the cell suspension, adding 16.7 µL of photosensitizer (PS, 6 mM) in PBS, redox mediator (RM, 10-20 mM) in PBS to yield a final concentration of 1 mM, and diluting to 1 mL final volume with PBS, supplemented with SED (100 mM, 200 mM or 500 mM as stated).This procedure yields a specific cell density OD600 of 5, a photosensitizer concentration of 100 µM, a RM concentration of 1 mM, and a SED concentration in the final sample of 90-100 mM.
Samples using TamHydS or CpI were prepared analogously, using the respective transformed cells.Hydrogen production was quantified by gas chromatography (PerkinElmer LLC, MA, USA) equipped with a thermal conductivity detector (TCD) and a stainless-steel column packed with molecular sieve (60/80 mesh).The operational temperatures of the injection port, the oven, and the detector were 100 °C, 80 °C and 100 °C, respectively.Argon was used as carrier gas at a flow rate of 35 mL min -1 and nitrogen as a reference gas.

I.5. Photocatalytic assays
The headspace of each sample was manually sampled by taking out 50 µL or 100 µL of the headspace volume with a gas-tight syringe.The amount of H2 present in the headspace was estimated using a calibration curve (see below).Each reported sample was carried out in at least two biological replicates and at least two technical replicates each.

I.5.1. Addition experiments
For addition experiments, solutions containing 1 mM PS, 10 mM RM or 1 M SED or combinations thereof were prepared in PBS buffer (pH 6.5 for cysteine samples, pH 7.5 for TEOA samples).The solutions were degassed and ~100 µL were added to the photocatalytic samples at indicated timepoints.The samples were then incubated in the dark at 30 °C for 30 minutes before starting light irradiation.

I.5.2. Buffer separation assays
Photocatalytic samples were taken at different timepoints as indicated and transferred into a glovebox.Under inert gas atmosphere, the vials were opened and the cell-suspension was centrifuged at 14 500 rpm for 5 minutes.The supernatant was removed and the cell pellet was subsequently resuspended in PBS buffer with 100 mM SED (L-cysteine at pH 6.5, or TEOA at pH 7.5), adding 16.7 µL of PS (6 mM) in PBS, MV (10-20 mM) in PBS to yield a final concentration of 1 mM, yielding a final total volume of 1 mL.This procedure yields the same PS, SED and RM concentrations as in the photocatalytic sample prepared from fresh cell suspension.The samples were then incubated in the dark at 30 °C for 30 minutes before starting light irradiation.

I.6. H2 calibration curve
To convert the observed H2 peaks into the actual amount of H2 produced by the sample, a calibration curve was conducted.The vials used for the photocatalytic assays were filled with 1 mL PBS in the glovebox, sealed, and placed on shakers at 30 °C.Using a syringe, discrete volumes (10-1000 µL) were taken out of each vial and replaced with the same volume of pure hydrogen gas.The samples were then placed on the shaker for 10-20 minutes and 50 or 100 µL of the headspace volume were sampled via GC.Each hydrogen concentration was carried out in triplicates.The resulting data points as well as a linear fit are shown in Figure S.I. 6.
The observed hydrogen peak in the GC measurements was correlated to the amount of hydrogen present in the injected sample n(H2) using the Ideal Gas Equation: with Vinj(H2) being the percentual injected volume of H2, (atmospheric) pressure p, temperature T and the Ideal Gas Constant R. The calculated n(H2) then refers to a discrete value of the integral of the experimentally determined H2-peak (Ac) in the gas chromatogram.Linear regression of a plot of Ac with the y-intercept being fixed to 0 assigned to n(H2) in nmol yields a calibration equation of the form whereby the herein obtained slope b 2 668.216 66 functions as conversion factor for attributing a raw integral value (Ac) to a discrete amount of H2 in the injected headspace sample.
The total amount of H2 produced by a sample can then be calculated by multiplying with a factor of 70, accounting for the total headspace volume: All hydrogen production values are calculated this way, hence representing the total amount of hydrogen produced by a 1 mL sample with a cell density of OD600 = 5.Generally, the H2-production of these systems is low and the observed amounts of hydrogen production are not significantly above those observed for samples that lack an active hydrogenase, i.e. without the co-factor.This complicates a detailed analysis of the influence of the various photocatalytic components under these specific conditions.

III. Cell viability studies
Cell viability studies were carried out by spreading cell suspension on agar plates.These   3 of the main manuscript: -no cell growth (note that little dots stem from bubbles in the agar plate); + some well isolated cell colonies are observed; ++ usually well isolated cell colonies covering most of the plate; +++ most of the plate is covered in cell colonies; ++++ the plate is completely overgrown with no distinguishable cell colonies.

IV. Cyclic voltammetry of diquat derivatives
Electrochemical measurements were carried out under inert glovebox atmosphere using   Table S.I. 2. Summary of absorption parameters of diquat derivatives in their oxidized state (DQ-X 2+ ) and their singly-reduced state (DQ-X + ). a absorption wavelength λ / nm (extinction coefficient ε / 10 a maximum absorption wavelengths λ are reported in nm and extinction coefficients ε in 10 3 M -1 cm -1 ; measurements carried out under inert-gas atmosphere in aqueous Tris-HCl buffer (1 M, pH 8).

VI.1. Nanosecond-laser pump probe transient absorption spectroscopy
The set-up has previously been described described. 9The sample was excited using a Nd/YAG laser (Quantel, Brilliant) passed through an OPO tuned to 460 nm for experiments using [Ru(bpy)3] 2+ and 516 nm for experiments using EY.The excitation energies varied from 14 to 17 mJ/pulse.The sample was probed using an unpulsed (for time-scales > 100 µs) or pulsed (for time-scales < 100 µs) Xe arc lamp perpendicular to the excitation light.The probe light was passed through one or two monochromators (Applied Photophysics, pbp Spectra Kinetic Monochromator 05-109) with one after the sample (2-mm slit opening) and an optional one before the sample (6-mm slit opening).The signal was detected using a photomultiplier tube (PMT, Hamamatsu R928), digitized using an oscilloscope (Agilent Technologies Infiniium 600 MHz) and processed using the Applied Photophysics LKS software.The measurements were carried out in 2 mm × 10 mm or 10 mm × 10 mm cuvettes, with the probe lamp passing through the 10 mm path in both cases.
All spectroscopic samples were prepared in a glovebox under anaerobic atmosphere in gas-tight cuvettes.Experiments are carried out in PBS solution (pH 6.5 or 7.5).1][12][13][14][15][16] These values are typically determined from quenching from the [Ru(bpy)3] 2+ * emission by MV.In the presence of different electron donors, a rate for photoreduction of MV by [Ru(bpy)3] 2+ * of 2.5•10 8 M -1 s -1 (with mercaptoethanol) to 2.9 •10 8 M -1 s -1 (with TEOA) has been reported. 17These values were gained by studying the formation of MVred absorption bands.slightly ranging from 6.5•10 6 M -1 s -1 to have been reported. 10,17 ble S.I. 3. Summary of the rate constants reported in this study.Values are calculated from averages of the mono-exponential fits of at least three individual traces.

VI.2. Reduced redox mediator accumulation
Experiments were carried out either on a Varian Cary 50 UV-Vis Spectrophotometer, using a 530 nm pulsed LED lamp for excitation, or an Agilent 8453 diode array UV-vis spectrometer, using a 447.5 nm LED (Luxeon Star, Rebel premounted LED fitted with carlco 29.8/10 mm lens) that was controlled by an HP 8116A 50 MHz pulse/function generator.In both set-ups, the LED was set to continuous illumination (447.5 nm LED)/ continuous pulsing (530 nm LED) and the accumulation of RMred was followed by either the absorption at 400 nm (MV and DQ derivatives) or 600 nm (MV).
Samples contained 4 µM PS, 20-25 µM RM, 100 mM SED in PBS (pH 6.5 for samples with L-cysteine, pH 7.5 samples with TEOA).Samples were prepared under inert-gas atmosphere and experiments were carried out in 10x10 mm quartz glass cuvettes.For in vivo experiments, cell suspensions were prepared as described above (section I.4).The final OD600 was 0.17.
The solution potentials were calculated using the Nernst equation.Similar to the [Ru(bpy)3] 2+ +L-cysteine+MV system discussed in the main manuscript, the maximum TON reached for EY+TEOA+MV systems is close to the expected maximum amount of hydrogen.However, TEOA is considered a two-electron donor 18 , which should allow for much higher TON.
Addition of TEOA after the H2 plateau is reached does not lead to a reactivation of H2 production, similarly to the [Ru(bpy)3] 2+ +L-cysteine+MV system.However, even TEOA addition before the plateau is reached as well as starting with a higher TEOA concentration (see Figure 2c, main manuscript) does not lead to substantially higher TON.We have stated previously that decomposition of EYred competes with the electron transfer to an acceptor. 5nce, early addition or higher starting concentrations of TEOA could lead to more efficient EY reduction, which then decomposes rather than being re-oxidized efficiently.Indeed, addition of both EY and TEOA yields an increased TON, albeit not as high as expected, based on overall TEOA available.Only when all three components of the photosystem are added, i.e.
EY, TEOA and MV, the hydrogen evolution yields TONs that match the doubling of TEOA in the system.This excludes the build-up of toxic or inhibiting by-products in the assay solution as a limiting factor.Yet, it is unclear how exactly the decompositions of the single components affect each other and what role the potential second electron coming from TEOA plays.

Figure
Figure S.I. 1. 1 H-NMR-spectrum of DQ-COOH recorded in D2O.Chemical shifts are referenced to the signal of the solvent (δD2O = 4.79 ppm)
Whole-cell samples (1 mL sample volume) were prepared in SureSTART™ glass vials (total volume 8 mL) and sealed with crimp cap and a rubber septum.The vials were placed on a shaker at 30 °C in the dark for 30 min.They were subsequently placed under a triphosphor fluorescent white lamp (Aura long life universal, 58W/865, 4000 lx, main emission 400-650 nm, see Figure S.I. 5) and kept under constant irradiation at 30 °C.

Figure S.I. 5 .
Figure S.I. 5. Lamp profile.Normalized emission spectrum of the triphosphor fluorescent white lamp used in photocatalytic experiments.

Figure S.I. 7 .Figure S.I. 8 .
Figure S.I. 6: H2-Production Calibration Curve.Calibration curve to examine the discrete amount of H2 produced by photocatalytic samples via GC.The experimentally reported integral of the H2-peak in gas chromatograms, Ac, is attributed to an amount of H2 in nmol.
plates were prepared with 20 g/L LB medium, and 100 mg/L ampicillin was added to each plate to select for E. coli cells harbouring the CrHydA1 containing plasmid.The samples were prepared analogously to the photocatalytic samples and were kept under continuous irradiation at 30 °C.They were taken at different time points and 100 µL of the cell suspension was spread on the agar plates.The plates were incubated over night at 37 °C or over several days at room temperature.The cell growth was evaluated qualitatively by comparing the coverage of grown cells and putting each sample in one of five categories, which are shown in Figure S.I. 13 on representative examples.The results from these plating studies are summarized in Table3in the main manuscript.

Figure S.I. 13 .
Figure S.I. 13.Photographs of representative examples of plating experiments corresponding to the different ratings in Table3of the main manuscript: -no cell growth (note that little dots stem from bubbles in the agar plate); + some well isolated cell colonies are observed; ++ usually well isolated cell colonies covering most of the plate; +++ most of the plate is covered in cell colonies; ++++ the plate is completely overgrown with no distinguishable cell colonies.
Figure S.I. 16.UV-vis absorption spectra of diquat derivatives in their oxidized and singly-reduced form; measurements carried out under inert-gas atmosphere in aqueous Tris-HCl buffer (1 M, pH 8).*artefact from lamp.

[
Scheme S.I. 2. Schematic representation of the reductive quenching photocatalytic cycle involving EY as photosensitizer (4 µM), TEOA as sacrificial electron donor (100 mM), and MV (20 µM), as electron acceptor; pseudo-first order rate constants as obtained by transient absorption spectroscopy.Reported values are averages of three or more individual traces.

FigureFigure
Figure S.I. 26.Ratio of RMred over total amount of RM upon continuous illumination.Sample solutions contain 4 µM PS, 20 µM MV or 25 µM DQ-OH, and 100 mM SED in PBS solution (pH 7.5for TEOA samples, or 6.5 for cysteine samples), in the absence (orange) or presence (green) of fresh CrHydA1 containing E. coli cells (OD600 = 0.2); measurements are carried out under an inert gas atmosphere.
(bromomethyl)propanoic acid ( 4.22 g, 17.16 mmol) were dissolved in 8 ml acetonitrile in a flame-dried round-bottom flask, equipped with a stirrer.The mixture was refluxed at 90°C for 48h.The product was precipitated from diethyl ether and washed several times with ethyl acetate to obtain the product with a 87,7% yield.

Table S .
I. 1. Summary of the half-wave reduction potentials E1/2 in V vs SHE of the first reduction of the redox mediators in PBS (pH = 6.82); peak separation ΔEP in mV in brackets.
E0 being the standard redox potential of the RM, R being the universal gas constant, T the temperature in Kelvin, F the Faraday constant, n the number of electrons transferred, [red] being the concentration of RMred when a (pseudo) steady state is reached and [ox] being the concentration of remaining RM in solution.It should be noted that the solution potential cannot be directly correlated to the driving force of the electron transfer step from the RM to the cellular acceptor.Yet, it can act as a measure which combines thermodynamic driving force (i.e. the standard redox potential of the RM) and kinetic effects stemming from the different concentrations of RMred present in the solution.The results are summarized in Figure S.I. 26 and Table2in the main manuscript.