Two previously described plasma sources, the PlasmaDerm DBD (Cinogy, Germany) (Baldus et al. 2015) and the atmospheric pressure capillary plasma jet (Winzer et al. 2022; Schüttler et al. 2024b), were used. The DBD had a copper electrode with a diameter of 20 mm. It was driven at VRMS = 13.5 kV and a trigger frequency of 300 Hz (Baldus et al. 2015). For plasma treatments, 40 µl samples were placed on grounded stainless-steel supports at a distance of 1 mm from the DBD.

The RF-driven atmospheric pressure capillary plasma jet has a plasma volume of 4 mm × 0.88 mm × 40 mm (outer capillary dimension 5 mm × 1.32 m × 40 mm), and H2O2 production by the jet was previously characterized [22]. Here, an input power of 6.6 ± 0.6 W was used to ignite the plasma, and to increase H2O2 production, the feed gas (2 slm He flow) was (partially) routed through a bubbler with cooled deionized water. The distance between the nozzle of the capillary jet and the sample was approx. 16 mm.

The cypC overexpression plasmid was designed based on pET28b using NdeI and HindIII restriction sites at the beginning and the end of the gene, respectively. The B. subtilis 168 gene was codon optimized for Escherichia coli (E. coli) (the nucleotide sequence is provided in Supplementary Information). The vector-encoded C-terminal His-tag was silenced by preserving the natural stop codon of the cypC gene. Gene synthesis, cloning, and sequencing were performed by Genscript (USA). The plasmid enabling isopropyl-β-D-thiogalactopyranoside (IPTG)-dependent expression of the His6-cypC construct was transformed into chemically competent E. coli DH5α for amplification and then re-isolated. LB agar containing 50 µg ml−1 kanamycin was used for selection.

The plasmid pET28b::cypC was freshly transformed into chemically competent BL21 (DE3) E. coli cells. An overnight culture was prepared using LB medium supplemented with 50 µg ml−1 kanamycin for selection. CypC was initially overproduced in 1 L cultures in LB medium containing 50 µg ml−1 kanamycin, 200 µmol l−1 hemin chloride (stock solution dissolved in 100 mmol l−1 NaOH), and 500 µmol l−1 δ-aminolevulinic acid. After inoculating to an OD600 of 0.05, the culture was incubated at 37 °C until an OD600 of 0.5–0.6 was reached. IPTG was then added to 100 µmol l−1, and cells were harvested for protein purification after a further 4-h incubation at 30 °C.

To improve heme loading, ZYM5052 auto-induction medium was used (Studier 2005; Linde et al. 2020). For the main culture, 1 L ZYM5052 medium was supplemented with 50 µg ml−1 kanamycin, 200 µmol l−1 hemin chloride (see above), and 500 µmol l−1 δ-aminolevulinic acid and inoculated with 10 ml of a preculture. Overexpression was allowed to proceed for 5 days at 16 °C and 120 rpm. Every 24 h, samples were withdrawn to monitor expression by SDS-PAGE and western blot analysis according to standard protocols and cell densities of the samples were normalized. His6-tagged proteins were detected using a fluorescence-based 6 × His-tag monoclonal antibody (ThermoFisher Scientific, USA) and a ChemiDoc MP Imaging System (Bio-Rad, USA). On day 6 (120 h incubation), cells were harvested by centrifugation, washed with 100 mmol l−1 potassium phosphate buffer (pH 7), and stored at − 80 °C until further use.

To lyse the cells, samples were resuspended in lysis buffer (0.2 mg ml−1 DNase, 0.2 mg ml−1 RNase, 0.35 mg ml−1 lysozyme, 2 mmol l−1 DTT, and cOmplete protease inhibitor (Roche, Switzerland) in 100 mmol l−1 potassium phosphate buffer containing 300 mmol l−1 potassium chloride and 20% glycerol, pH 7) (Dirks et al. 2023). Cells were disrupted using a pressure cell homogenizer (FPG12800, Homogenising Systems, UK) for six cycles. After 30 min of centrifugation at 21.000 g, the supernatant was collected and loaded onto a HisTrap FF crude 5 ml column (GE Healthcare, USA) for purification of the His6-tagged CypC with an ÄKTA pure25 system (GE Healthcare, USA). Proteins were eluted using a three-step gradient with increasing imidazole concentrations (three column volumes each of 50 mmol l−1, 75 mmol l−1, 200 mmol l−1). The HisTrap FF crude 5 ml column was finally washed with 10 column volumes of 500 mmol l−1 imidazole to remove any remaining bound protein. Elution fractions of interest were pooled and concentrated using centrifugal filter units (10 kDa molecular weight cut-off) prior to reconstitution.

The concentration of the purified CypC was determined using the Bradford method. To increase heme loading, the protein was then incubated with a two-fold molar excess of hemin chloride (in a total volume of approx. 50 ml) at 25 °C for 1 h. Afterwards, unbound hemin chloride was removed by overnight dialysis against 100 mmol l−1 potassium phosphate buffer containing 300 mmol l−1 potassium chloride and 20% glycerol, pH 7. The dialyzed and reconstituted CypC was aliquoted and stored at − 80 °C.

Absorption spectra of CypC were recorded using an enzyme concentration of 12 µmol l−1 in a total volume of 100 µl (100 mmol l−1 potassium phosphate buffer containing 300 mmol l−1 potassium chloride and 20% glycerol, pH 7) in a UV/VIS spectrophotometer (V-750 spectrophotometer, Jasco, Germany). Buffer was used as a blank. R/z values were calculated by relating the absorption of the Soret peak at its maximum intensity (~ 420 nm) to absorption at 280 nm (Chance and Maehly 1955; Shannon et al. 1966).

Enzyme activity of CypC was determined based on the conversion of the natural substrate myristic acid to α- and β-hydroxylated myristic acid (Girhard et al. 2013). To this end, 120 µmol l−1 myristic acid was mixed with 0.1 µmol l−1 CypC in potassium phosphate buffer (100 mmol l−1, pH 7) in a total volume of 200 µl. To start the reaction, 0.5 mmol l−1 H2O2 was added, and the reaction was allowed to proceed at 37 °C for 30 min under constant agitation (500 rpm). Afterwards, the analysis was performed following a protocol described by Girhard et al. (Girhard et al. 2013). Briefly, samples were mixed twice with 500 µl diethyl ether and centrifuged. The organic phase was transferred into a new vial and dried using anhydrous MgSO4. The supernatant was again transferred into a new vial to then evaporate the organic phase at 34.6 °C for 20 min. Residual was resuspended in 50 µl N,O-bis(trimethylsilyl)trifluoroacetamide with 1% trimethylchlorosilane (BSTFA-TMCS, TCI Chemicals, Germany) and incubated at 80 °C for 30 min. Derivatized samples were immediately measured by gas chromatography (Shimadzu GC-2030 Nexis, Japan) along with electron impact mass spectrometry (EI-MS) using a GC/MS-QP2020 NX (Shimadzu, Japan) equipped with an FS-Supreme-5 ms column (30 m × 0.25 mm × 0.25 µm, Chromatographie Service, Germany). The temperature of the column was held at 160 °C for 1 min prior to ramping to 260 °C at 10 °C min−1. Finally, the temperature was increased to 300 °C (40 °C min−1) and held for 3 min. Helium was used as the carrier gas (flow rate of 1 ml min−1), and the mass detector was operated in electron impact (EI) mode at 70 eV with an electron multiplier voltage of 1.25 kV. Total mass and fragment masses were compared to previous reports (Girhard et al. 2013).

Conversion of non-natural substrates was analyzed using 2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) or guaiacol together with heptanoic acid as a decoy molecule according to Shoji and Watanabe (2017). Activity assays were performed following a basic protocol with slight modifications for decoy molecule addition. 5 mmol l−1 2,2′-azino-bis(3-ethylthiazoline 6-sulfonate) (ABTS; ε = 36.8 mmol l−1 cm−1) or 50 mmol l−1 guaiacol (ε = 26.6 mmol l−1 cm−1) were used. The substrates were dissolved in 100 mmol l−1 sodium citrate buffer (pH 5) and 100 mmol l−1 potassium phosphate buffer (containing 300 mmol l−1 potassium chloride and 20% glycerol, pH 7), respectively, while 20 mmol l−1 heptanoic acid (dissolved in EtOH) was added. Reactions were initiated by always adding the same volume of H2O2 solution prepared by dilution in distilled water (A. dest.) to yield concentrations ranging from 0 to 10 mmol l−1. For 2 min, product formation was monitored at 405 nm or 470 nm using a microplate reader (Biotek, Epoch, Germany). The enzyme concentration was 1 μmol l−1. Activity was calculated from the initial reaction velocity within the first seconds of the kinetic (formula: ∆absorption/∆time).

The temperature optimum was tested in a range of 10 to 40 °C in 10 °C increments using a UV/VIS spectrophotometer with built-in Peltier element (V-750 spectrophotometer, Jasco) in a total volume of 100 µl. All reaction components were incubated at the respective temperature for 10 min before starting the reaction.

Immobilization with ReliZyme HA403 M beads was carried out as described previously for rAaeUPO (Yayci et al. 2020b; Dirks et al. 2023) by applying 10 µmol l−1 CypC to 200 mg HA403 M beads in a volume of 5 ml 100 mmol l−1 potassium phosphate buffer containing 300 mmol l−1 potassium chloride and 20% glycerol (pH 7). Binding efficiencies were determined by comparing enzyme activity of the supernatant and of the solution initially used for immobilization. Plasma-driven biocatalysis was performed as described earlier for rAaeUPO (Yayci et al. 2020b). In short, protein-loaded beads (100 mg) were transferred to a rotating bed reactor (built in-house, dimensions: ∅2 cm × 0.7 cm). The reactor was placed in a vessel filled with 5 ml potassium phosphate buffer (100 mmol l−1, pH 7) containing 50 mmol l−1 ethylbenzene and 20 mmol l−1 heptanoic acid (dissolved in EtOH). Plasma treatment was performed for up to 60 min, and 150 µl samples were withdrawn every 5 min for analysis of product formation (Yayci et al. 2020a, 2020b). To compensate for the sample withdrawal and evaporation, 300 µl of potassium phosphate buffer (100 mmol l−1, pH 7) containing 50 mmol l−1 ethylbenzene and 20 mmol l−1 heptanoic acid were resupplied.

To avoid H2O2 accumulation, the entire reaction solution was exchanged every 10 min. Product accumulation in the reaction volume was measured by GC (Yayci et al. 2020a, 2020b). Plasma treatment was continued for a total of 120 min.