Transmembrane Shuttling of Photosynthetically Produced Electrons to Propel Extracellular Biocatalytic Redox Reactions in a Modular Fashion

Abstract Many biocatalytic redox reactions depend on the cofactor NAD(P)H, which may be provided by dedicated recycling systems. Exploiting light and water for NADPH‐regeneration as it is performed, e.g. by cyanobacteria, is conceptually very appealing due to its high atom economy. However, the current use of cyanobacteria is limited, e.g. by challenging and time‐consuming heterologous enzyme expression in cyanobacteria as well as limitations of substrate or product transport through the cell wall. Here we establish a transmembrane electron shuttling system propelled by the cyanobacterial photosynthesis to drive extracellular NAD(P)H‐dependent redox reactions. The modular photo‐electron shuttling (MPS) overcomes the need for cloning and problems associated with enzyme‐ or substrate‐toxicity and substrate uptake. The MPS was demonstrated on four classes of enzymes with 19 enzymes and various types of substrates, reaching conversions of up to 99 % and giving products with >99 % optical purity.


Test of different alcohol/ketone shuttle pairs in the combined Modules B+C with LkADH
Table S3.Test of different alcohol/ketone shuttle pairs in the combined Module B with LkADH, and Module C with the enereductase OPR3 and 1a as substrate.

Expression of enzymes in E. coli
The enzymes used in this study were expressed in E. coli BL21(DE3) or its derivatives (Table S13).

Transformation
Plasmids (100 ng) were mixed with chemically competent E. coli BL21 (DE3) cells (100 µL), rested on ice for 30 minutes and heat-shocked for 30 seconds at 42 °C.SOC medium (200 µL) was added, and the transformed cells were incubated for 1 h at 37 °C and 300 rpm.The cells were then plated on a LBagar plate supplemented with corresponding antibiotic (ampicillin, 100 µg mL -1 or kanamycin, 50 µg mL -1 ) and incubated overnight at 37 °C.

Cultivation
Overnight cultures (ONC) were prepared in LB-medium (10 mL) supplemented with the corresponding antibiotic at 30 °C and 120 rpm.The ONCs were then used for the inoculation of sterile medium (1% v/v) supplemented with the corresponding antibiotic.Precultures were incubated until OD600 of 0.6 was reached, then protein expression was induced, and cells incubated further.Detailed conditions for every enzyme can be found in Table S13.

Harvesting
To harvest the cells, cultures were centrifuged at 3184 g, 20 min, 4 °C, the cell pellet suspended in wash buffer (1 -2.5 g cells per 10 mL phosphate buffer, 10 mM, pH 7), and then centrifuged again under the same conditions.

Preparation of cell-free extracts
Cell pellets were suspended in lysis buffer and sonicated on ice (for conditions see Table S12).The sonicated cells were centrifuged for 25 minutes at 17 000 g and 4 °C.The supernatant (cell-free extract) was shock-frozen (liquid nitrogen) inside a round bottom flask, lyophilized, and stored at -20 °C.The pH of the buffers was adjusted by using hydrochloric acid (HCl) and sodium hydroxide (NaOH).

SDS-PAGE
Protein concentrations of cell pellets and supernatants were determined via a Bradford Assay.
Seed cultures of recombinant Synechococcus elongatus were grown in the presence of spectinomycin (100 µg mL -1 ).

Determination of the cell dry weight and chlorophyll a content
The dry cell weight and the amount of chlorophyl a were determined from samples originating from at least three independent cultivations under growth conditions for working cultures, each measured in triplicates.

Cell dry weight
Working cultures grown and harvested as above were shock-frozen in liquid nitrogen, lyophilized overnight and weighed in three independent experiments.

Chlorophyll a
The chlorophyll a was determined as described. [32]A sample of the cell culture (100 μL) was mixed with cold methanol (900 μL) and incubated in darkness at 4 °C for 2-3 hours or overnight.Then, the samples were centrifuged for 3 min at 14000 g and the absorption of the supernatant was measured at 665 nm.The amount of chlorophyll was determined using the extinction coefficient ε = 78.74L g -1 cm -1 according to Eq. S 1, where the dilution factor corresponds to 10.
Stock solutions of LkADH CFE, NADP + , OPR3, MgCl2 and acetone were prepared in BG11 according to Table S15 (or as indicated at the specific experiments).Reaction mixtures with substrate 8a were supplemented with HEPES (final c. 100 mM, pH 8).Working cultures of Synechococcus elongatus at OD750 between 1 and 2 were harvested by centrifugation (30 °C, 45 min, 2786 g) and the pellet was resuspended in BG11 to reach a stock OD750 (20 to 40).The reactions were performed in 1.5 mL screwtop glass GC-vials.First the substrate stock was added, then BG11 and other components, with the cyanobacteria being added last.Table S15 enlists the sample preparation for reduction of 2a under the optimized conditions.
Table S15.Concentrations of the stock solutions for the fully assembled cofactor recycling system using the optimized conditions for the reduction of 2a.The vials were incubated in a custom photoreactor [33] equipped with cool white LEDs (LED stripes, 5200 K) for the indicated reaction time (16 h) at 600 rpm shaking and at room temperature at an average light intensity of 215 µE m -2 s -1 (photoreactor settings: frequency 100 Hz, duty range 100, duty cycle 5)

Component
or covered with aluminum foil for dark reactions.The workup and analytics are described in section 4.
3.10 50 mL scale reduction of 2a, utilizing the MPS.
2a (188.7 mg, 1 mmol) was added to a 250 mL Erlenmeyer flask with a screw neck.BG11 (31.4 mL), LkADH CFE (5 mg mL -1 , 2.5 mL, final 0.25 mg mL -1 ), NADP + (2 mM, 2.5 mL, final 0.1 mM), OPR3 (8.7 mg mL -1 , 574.7 µL, final 100 µg mL -1 ) and MgCl2 (100 mM, 500 µL, final 1 mM), all dissolved in BG11 were added to the flask.Then, acetone (58.1 µL, final 20 mM) and recombinant Synechococcus elongatus (harvested at OD750 1.9, concentrated to OD750 = 40, 12.5 mL, final OD750 = 9) were added, the flask was closed and placed in an incubator at 24 °C (AQUALYTIC Thermostatically controlled incubators) equipped with a custom photoreactor [33] controlling four cool white LED stripes (5200 K).Two LED stripes were mounted on the left and right wall of the incubator and two LED stripes were mounted above the reaction, on the rack placed in the 7 th slot from the bottom, giving an average light intensity of 300 µE m -2 s -1 at the spot of the reaction (photoreactor settings: frequency 100 Hz, duty range 100, duty cycle 90).The reaction was shaken at 140 rpm.The setup is displayed in the main paper.After 16 h, the mixture was extracted with ethyl acetate (2 x 150 mL), dried over anhydrous Na2SO4, and concentrated on the rotary evaporator.The crude was purified with Biotage® Selekt Flash Purification System using Biotage® Sfär Silica HC Duo column (20 μm, 5 g) and gradient elution starting from cyclohexane to 30% ethyl acetate in twenty column volumes, then to 50% ethyl acetate in five column volumes, yielding 133.4 mg of (R)-2b (70% isolated yield).S9).

2. 6
Tolerance of Synechococcus elongatus to DMSO and DMF At up to 10% DMSO and 5% DMF no color change was observed.At 20% DMSO and 10% DMF, slight color change to turquoise could be observed.DMSO was chosen as the preferred co-solvent as it is generally considered safer and could be added in higher volume if necessary.

Figure S3 .
Figure S3.Detailed reaction scheme of the modular photosynthetic monooxygenation of 12a, using the substrate as alcohol/ketone shuttle pair.

Figure S7 .
Figure S7.Representative GC-FID chromatograms for determining the enantiomeric excess of 1b.(A) Racemic reference compound; (B) illuminated reduction of 1a to 1b using the MPS.

Figure S9 .
Figure S9.Representative HPLC chromatograms for determining the enantiomeric excess of 2b.(A) Racemic reference compound; (B) illuminated reduction of 2a to 2b using the MPS at analytical scale and (C) at 50 mL scale.

Figure S10 .
Figure S10.Representative GC-FID chromatograms for quantification of 3a and 3b.(A) Mixture of reference compounds 3a and 3b; (B) illuminated and (C) dark reduction of 3a using the MPS.

Figure S11 .
Figure S11.Representative GC-FID chromatograms for determining the enantiomeric excess of 3b.(A) Racemic reference compound; (B) illuminated reduction of 3a to 3b using the MPS.

Figure S16 .
Figure S16.Representative GC-FID chromatograms for quantification of 7a and 7b, derivatized as trimethylsilyl esters.(A) Mixture of reference compounds 7a and 7b; (B) illuminated reduction of 7a using the MPS.

Figure S18 .
Figure S18.Representative GC-FID chromatograms for quantification of 8a and 8b, derivatized as trimethylsilyl esters.(A) Mixture of reference compounds 8a and 8b; (B) illuminated reduction of 8a using the MPS.

Figure S20 .
Figure S20.Representative GC-FID chromatograms for quantification of 9a and 9b.(A) Mixture of reference compounds 9a and 9b; (B) illuminated reduction of 9a using the MPS with IRED A.

Figure S22 .
Figure S22.Representative HPLC chromatograms for determining the enantiomeric excess of 10b.(A) illuminated (unselective) reduction of 10a with IRED A, and (B) illuminated (S)-selective reduction of 10a with IRED J, both using the MPS.

Figure S23 .
Figure S23.Representative GC-FID chromatograms for quantification of 11a and 11b.(A) Mixture of reference compounds 11a and 11b; (B) illuminated reduction of 11a using the MPS with IRED A.

Figure S24 .
Figure S24.Representative HPLC chromatograms for determining the enantiomeric excess of 11b.(A) Mixture of reference compounds rac-11b and 11a; (B) illuminated reduction of 11a with IRED A using the MPS.

Figure S27 .
Figure S27.Calibration curves for the quantification of (A) 5a/5b, (B) 6a/6b, (C) 7a/7b and (D) 8a/8b.The corresponding line equations and R 2 values were generated by a simple linear regression and were forced through zero.

Table S1 .
Initial test of alcohol/ketone shuttle pairs for Module A and the combined Modules B+C Initial test of alcohol/ketone shuttle pairs for Module A and the combined Modules B+C with ADH-A and the enereductase OPR3 for different alcohol/ketone shuttle pairs.

Table S4 .
°C and 600 rpm.n.d.= not determined.S6 2.4 Optimization of Module A with the recombinant Synechococcus elongatus strain together and Module B using either ADH-A or LkADH Optimization of Module A with the recombinant Synechococcus elongatus strain together with Module B, using ADH-A or LkADH and Module C with the ene-reductase OPR3 and 1a as substrate.

Table S5 .
Running the fully assembled system with OPR3 at increased concentrations of 2a.

Table S6 .
Biocatalytic reduction of 2a with different ene-reductases, comparing the fully assembled regeneration system with the ADH-based coupled enzyme regeneration system (Modules B+C).

Table S7 .
Substrate scope of ene-reductases driven by the MPS (Modules A+B+C), control reactions and comparison to an ADH-based coupled enzyme recycling system (Modules B+C).

Table S9 .
Performance of the MPS (Modules A+B+C) for the reduction of 9a-11a using IREDs, control reactions and comparison to an ADH-based coupled enzyme recycling system (Modules B+C).

Table S10 .
Performance of the MPS (Modules A+B+C) for the monooxygenation of 12a using a BVMO (CHMO), control reactions and comparison to an ADH-based coupled enzyme recycling system (Modules B+C).

Table S11 .
Enzymes used in this study.

Table S12 .
Buffers and sonication parameters used for preparations of CFEs.

Table S13 .
Cultivation conditions for the expression in E. coli.

Table S14 .
Buffers and sonication parameters used for the purification of enzymes.

Table S16 .
Methods for quantification of compounds by GC-FID.

Table S17 .
Methods for determining the enantiomeric excess of compounds by GC-FID.
Table S 18. Methods for determining the enantiomeric excess of compounds by HPLC.
a: Compounds detected as methyl esters.