Catch-and-Release: The Assembly, Immobilization, and Recycling of Redox-Reversible Artificial Metalloenzymes

Technologies to improve the applicability of artificial metalloenzymes (ArMs) are gaining considerable interest; one such approach is the immobilization of these biohybrid catalysts on support materials to enhance stability and enable their retention, recovery, and reuse. Here, we describe the immobilization of polyhistidine-tagged ArMs that allow the redox-controlled replacement of catalytic cofactors that have lost activity, e.g., due to poisoning or decomposition, on immobilized metal affinity chromatography resins. By using periplasmic siderophore-binding protein scaffolds that originate from thermophilic bacteria (GstCeuE and PthCeuE) in combination with a siderophore-linked imine reduction catalyst, reaction rates were achieved that are about 3.5 times faster than those previously obtained with CjCeuE, the analogous protein of Campylobacter jejuni. Upon immobilization, the GstCeuE-derived ArM showed a decrease in turnover frequency in the reduction of dehydrosalsolidine by 3.4-fold, while retaining enantioselectivity (36%) and showing improved stability that allowed repeat recovery and recycling cycles. Catalytic activity was preserved over the initial four cycles. In subsequent cycles, a gradual reduction of activity was evident. Once the initial activity decreased to around 40% of the initial activity (23rd recycling cycle), the redox-triggered artificial cofactor release permitted the subsequent recharging of the immobilized protein scaffold with fresh, active cofactor, thereby restoring the initial catalytic activity of the immobilized ArM and allowing its reuse for several more cycles. Furthermore, the ArM could be assembled directly from protein present in crude cell extracts, avoiding time-consuming and costly protein purification steps. Overall, this study demonstrates that the immobilization of redox-reversible ArMs facilitates their “catch-and-release” assembly and disassembly and the recycling of their components, improving their potential commercial viability and environmental footprint.


Materials
Unless otherwise noted, reagents were used as received from commercial suppliers and used as supplied unless otherwise stated.All expression media and buffers were prepared using ddH2O (purification system, Millipore).Solvents for chromatography were HPLC grade.
Instrumentation 1 H and 13 C{ 1 H} NMR spectra were recorded on Jeol EX and ES 400 MHz instruments ( 1 H NMR 400 MHz, 13 C NMR 101 MHz).Electrospray ionisation mass spectrometry (ESI-MS) was performed on a Bruker compact ® TOF mass spectrometer.Fluorescence spectra were recorded on a Hitachi F-4500 fluorescence spectrophotometer.UV-vis spectra were recorded on a Shimadzu UV-1800 in a quartz cuvette (Starna scientific).HPLC measurements were performed on an Agilent 1200 infinity II quaternary system equipped with a 1260 Quaternary Pump G7111B, G7116A multicolumn thermostat, G7165A multiwavelength detector and G7129A Vialsampler using the specified eluent gradients.Inductively coupled plasma-optical emission spectroscopy (ICP-OES) was carried out on an Agilent 5100 spectrometer and analysis was performed in ICP Expert version 7.6.2.12331.(ICP-MS) was carried out on an Agilent 7700 (Biorenewables Development Centre, York, United Kingdom).

Data accessibility
The data supporting this research is available for download from the research data repository of the University of York at https://doi.org/10.15124/85cd4daf-de96-4192-9712-1180496c51f9.

Cloning, expression and purification of 6His-tagged proteins
The periplasmic binding protein CjCeuE (44-330) and its thermostable homologues GstCeuE protein (39-319) and PthCeuE protein (37-318) (constructs with the signal peptide removed) were chosen, cloned into Lic-adapted pET 28a vector (YSBLic3C) and purified with the presence and absence of a 6-terminal His-tag, as described recently. 3

Binding affinity determination by intrinsic fluorescence quenching
Fluorescence spectra were recorded on a Hitachi F-4500 fluorescence spectrophotometer with an excitation wavelength of 280 nm, emission range of 295-410 nm, 10 nm excitation slit width, 20 nm emission slit width, 60 nm/min scanning speed, automatic response, corrected spectra and 950 V detector voltage.Preparations and measurements were carried out following the procedures previously reported. 3The only change was that the [Fe III (3)Cp*Ir III ] stock solution was prepared in DMF to improve solubility (avoid precipitation) and hence the accuracy of quantities being dispensed by DOSTAL DOSY.

Stock solutions MES/formate buffer ('catalytic' buffer):
The buffer solution was prepared by dissolving 17.06 g of 2-morpholin-4-ylethanesulfonic acid monohydrate (MES monohydrate) and 27.20 g of sodium formate (HCOONa) in 80 ml water and then adjusting the volume to 100 mL (final concentration: 0.8 M MES, 4 M HCOONa).7.5 mL of the stock solution were transferred to a flask and the pH adjusted to either 6, 6.5.7, 7.5 or 8 by addition of 5 M NaOH before the solution volume was brought up to 10 ml (final concentrations 0.6 M MES, 3 M sodium formate).
Isoquinoline stock: 200 or 20 mM stock solutions of isoquinoline (1) were prepared in the catalytic buffer of choice.
Harmaline stock: 12 mM stocks of harmaline (5) were prepared in the catalytic buffer pH 7.

S6
Quenching solution: 12.5 mM solution of L-glutathione in Water:MeOH (1:2) mixture was prepared, volumes as required.L-Glutathione was firstly dissolved in water before addition of MeOH.Quenching solutions were stored at 4 °C and used on the day of preparation.

Homogeneous batch reactionscontrols
Reactions were carried out analogously to those previously reported, 2 however using a thermos-shaker for microtubes (Grant-bio) instead of magnetic stirrer.For isoquinoline the reactions were run in catalytic buffer at pH 6 (0.6 M MES / 3 M HCOONa), a 50 mM substrate or HPLC analysis (methods detailed in the next section).

Heterogeneous batch reactions
Generally, batch reactions with the immobilised ArMs were carried out by using adapted batch reactors, following the 'General Batch Reaction Procedure' provided in the main text.For catalytic tests with isoquinoline as the substrate, a substrate concentration 2 mM was used, unless stated otherwise.Once completed, reaction mixtures were separated from the immobilised ArMs via spin filtration.100 µL aliquots were withdrawn from each collection vials and mixed with 900 µL of quenching solution.For catalytic tests with harmaline, a substrate concentration of 10 mM was used, unless stated otherwise.Upon completion of the reaction, reaction mixtures were separated from the immobilised ArMs via spin filtration.25 µL aliquots were withdrawn from each collection vial and mixed with 1975 µL of quenching solution.Quenched samples were submitted for UV-vis or HPLC analysis (methods detailed in the next section).
For the thermostability assay, immobilised ArMs were incubated in the catalytic buffer at 60 °C, under static conditions, for 4, 8 or 18 h, before the catalytic test was started with the addition Supplementary information

S7
of substrate and start shaking at 800 rpm for the duration of 1 h.A control with no preincubation was carried out and used as reference for relative activity calculations.

Product analysis
The progress of the imine reduction reactions was monitored by UV/vis spectroscopy via the decrease of the absorption bands with maxima at around 350 nm (isoquinoline) and 375 nm (harmaline).UV/vis was found to be sufficient for monitoring reaction evolution, with analogous analytical accuracy when compared to HPLC (Figure S7).
Representative non-chiral and chiral chromatograms are presented in Figures S9 and S10, respectively.

UV-Vis analysis
After quenching, the respective reaction aliquots were transferred to a quartz cuvette and spectra recorded from 500 to 250 nm, 1 nm interval, fast scan.

Chiral HPLC analysis
Chiral  a Final dilution of digested samples before analysis (made up to 100 mL).
b Weighed sample for digestion.Digestion carried out as described in the respective supplementary method (Supplementary information, section 9) c Calculated Fe and Ir content (nmol) normalised by the weighed sample (g).