Spiers Memorial Lecture: Shielding the active site: a streptavidin superoxide-dismutase chimera as a host protein for asymmetric transfer hydrogenation

By anchoring a metal cofactor within a host protein, so-called artificial metalloenzymes can be generated. Such hybrid catalysts combine the versatility of transition metals in catalyzing new-to-nature reactions with the power of genetic-engineering to evolve proteins. With the aim of gaining better control over second coordination-sphere interactions between a streptavidin host-protein (Sav) and a biotinylated cofactor, we engineered a hydrophobic dimerization domain, borrowed from superoxide dismutase C (SOD), on Sav’s biotin-binding vestibule. The influence of the SOD dimerization domain (DD) on the performance of an asymmetric transfer hydrogenase (ATHase) resulting from anchoring a biotinylated Cp*Ir-cofactor – [Cp*Ir(biot-p-L)Cl] (1-Cl) – within Sav-SOD is reported herein. We show that, depending on the nature of the residue at position Sav S112, the introduction of the SOD DD on the biotin-binding vestibule leads to an inversion of configuration of the reduction product, as well as a fivefold increase in catalytic efficiency. The findings are rationalized by QM/MM calculations, combined with X-ray crystallography.


General information. Chemicals were purchased from Sigma Aldrich, Acros
Organics, Alfa Aesar or Fluorochem and used without further purification. Water used for molecular biology and in the catalytic reactions was purified by a Milli-Q Advantage system. All catalytic reactions were carried out with non-degassed solvents under air.
Temperature was maintained using Thermowatch-controlled heating blocks. Gas chromatography-mass spectrometry (GC-MS) analysis was run on a Shimadzu GCMS-QP2020. A normal phase HPLC instrument from Agilent with a Chiralpak IB column (5 μm, 4.6 mm × 250 mm, Daicel) and a UPC2 system (Waters) was used to analyze the samples. High-resolution mass spectrometry (HR-MS) was performed on a Bruker maXis II QTOF ESI mass spectrometer coupled to a Shimadzu LC. The transfer hydrogenation co-factor [Cp*Ir(biot-p-L)Cl] was synthesized as previously reported. [1] Imine reduction yields were determined using an internal standard or substrate to production response factor. Chiral amine retention times and enantiomeric excess (ee) yields were determined using authentic and racemic standards. Molecular biology reagents were purchased from New England Biolabs (NEB), Integrated DNA Technologies (IDT), and Macherey-Nagel using accompanying protocols. Protein concentrations were determined by the calculated A 280 molar absorptivity coefficients using the ProtParam tool (ExPASy) of 41940 M -1 cm -1 and 43430 M -1 cm -1 for wild type Sav and Sav-SOD respectively.
Cloning and expression of Sav and Sav-SOD mutants. The Sav-SOD chimera gene was synthesized and cloned into the NcoI and EagII restriction sites of the pET28 vector by Gene Universal Inc. (Newark, DE). Site directed mutagenesis was achieved using the primers listed in Table S1, followed by BsaI/DpnI digestion and ligated using T4 ligase. Mutations were verified by Sanger sequencing performed by Microsynth (Balgach, Switzerland). The expression and purification of Sav and Sav-SOD proteins were achieved using auto induction media (ZYP-5052) followed by cell lysis and purification using iminobiotin agarose affinity chromatography as previously described. [2] Table S1 Primers used for mutagenesis. ITC analysis. Biotin binding affinities were measured using a Microcal VP-ITC as previously described by Stayton and Coworkers. [3] A 10 M solution of [Cp*Ir(biot-p-L)Cl] in 50 mM sodium phosphate buffer containing 100 mM sodium chloride at pH = 7.75 with 2.5 % DMSO was titrated with 5 L injections of Sav-S112A or Sav-SOD-S112A solutions (100 M, 50 mM sodium phosphate, 100 mM sodium chloride, 2.5% DMSO, pH = 7.75). The reference cell contained the same buffer as the protein and cofactor solution. Measurements were performed at 25 °C. The [Cp*Ir(biot-p-L)Cl] binding constant K a , enthalpy h, and number of binding sites at each temperature were calculated using ITC data analysis origin software (MicroCal).

No signal detected
Panels c-d represents MS spectra collected at a 0.5 eq loading stoichiometry.
QM/MM calculations. In order to calculate the structure of Sav-SOD ATHases, we first estimated by MD simulation the structure of the loops -i.e. the dimerization domain-whose structure could not be resolved from our X-ray crystallography data sets. The sampled loop segment was combined with the crystal structure of streptavidin (pdb: 7ALX, tetramer) [4] to form the initial Sav-SOD structure used for structural optimization. Only two Sav-SOD monomers, which make up the biotinbinding vestibule, were used throughout the computation. A proton at pH 7.4 was assigned to each amino acid residue using the PROPKA server. 6 Here, the [Cp*Ir(biotp-L)H] catalyst structure was appended so that the coordinates of its biotin moiety were the identical to the X-ray structure. Then the structure of the substrate was also appended. Considering the C2-relationship between the two adjacent biotin-binding

Single-crystal X-ray diffraction analysis of [Cp*Ir(biot-p-L)Cl] · Sav-SOD-S112A.
Protein crystals were obtained using a sitting drop vapor diffusion method. The       The program was set up as follows: 80% CO 2 and 20% IPA (with 0.3% diethylamine, DEA) at a constant flow rate (2.5 mL/min). Retention times: 1.77 min for 2methylindoline, 7.9 min for (1)-12 and 8.9 for (2)-12. The compounds were quantified based on the relative peak areas and the internal standard area (absorance at 286 nm) using a response factor of 0.7281. The program was set up as follows: 85% CO 2 and 15% IPA (with 0.3% diethylamine, DEA) from 0-1.5 min; then linear increase to 63% CO 2 and 37% IPA (0.3% DEA) from 1.5-2.3 min; this ratio was kept to 6 min; then the ratio was linear decrease to 85% CO 2 and 15% IPA (0.3% DEA) from 6-6.2 min; keep at 85% CO 2 and 15% IPA (0.3% DEA) from 6.2-6.5 min. Retention times: 2.0 min for 2-methylindoline, 4.6 min for 7, 5.4 min for (R)-13 and 5.8 for (S)-13. The compounds were quantified based on the relative substrate and product peak peak areas and the internal standard area (absorbance at 293 nm) using a response factor of 6.1247.