Biocatalytic and Chemo-Enzymatic Synthesis of Quinolines and 2-Quinolones by Monoamine Oxidase (MAO-N) and Horseradish Peroxidase (HRP) Biocatalysts

Abstract

The oxidative aromatization of aliphatic N-heterocycles is a fundamental organic transformation for the preparation of a diverse array of heteroaromatic compounds. Despite many attempts to improve the efficiency and practicality of this transformation, most synthetic methodologies still require toxic and expensive reagents as well as harsh conditions. Herein, we describe two enzymatic strategies for the oxidation of 1,2,3,4-tetrahydroquinolines (THQs) and N-cyclopropyl-N-alkylanilines into quinolines and 2-quinolones, respectively. Whole cells and purified monoamine oxidase (MAO-N) enzymes were used to effectively catalyze the biotransformation of THQs into the corresponding aromatic quinoline derivatives, while N-cyclopropyl-N-alkylanilines were converted into 2-quinolone compounds through a horseradish peroxidase (HRP)-catalyzed annulation/aromatization reaction followed by Fe-mediated oxidation.

Introduction

Oxidation is unarguably a fundamental transformation in synthetic chemistry and chemical industry, especially in accessing ubiquitous (hetero)aromatic compounds.¹ Despite endless attempts to improve the efficiency and practicality of this transformation, most of the traditional oxidation reactions heavily rely on transition-metal-catalyzed or promoted strategies or on the use of toxic reagents and harsh reaction conditions.² From a sustainability point of view, the development of chemical processes that extend beyond the traditional oxidations is still highly desirable and remains a long-standing challenge in synthetic chemistry. The use of enzymes as catalysts in oxidations is of great appeal because of their mild, efficient, benign, and highly selective nature.³ In fact, oxidases can catalyze a multitude of oxidative transformations at ambient temperature and pressure and, in some cases, without the need for any external cofactor additive or recycling system. For such reasons, oxidizing biocatalysts have been widely employed in the last decade in asymmetric reactions,⁴ like the deracemization of chiral amines by monoamine oxidases (MAO-N)⁵ or the enantioselective synthesis of alcohols by glucose oxidase (GOase M3–5)⁶ or alcohol dehydrogenases (ADH).⁷ Over the past few years, our group successfully demonstrated the possibility of employing oxidizing biocatalysts such as MAO-N and laccase also in the aromatization of aliphatic or partially saturated cyclic substrates into aromatic pyrroles, pyridines, indoles, and furans under mild reaction conditions.⁸ These transformations well demonstrated the aromatizing properties of MAO-N and laccase enzymes and encouraged us to further expand the scope of aromatic heterocycles accessible via biocatalysis.

Quinolines and 2-quinolones are privileged scaffolds present in many drugs such as lenvatinib, brexpiprazole, bosutinib, and indacaterol (Figure 1a), and consequently, several methodologies have been reported to date for their synthesis. However, the functionalization of these N-heterocycles is a challenge due to the low reactivity of their π-electron-deficient skeleton.⁹ Thus, the development of practical and mild transformations that allow the synthesis of these aromatic rings from readily available starting materials is highly desirable. From a retrosynthetic standpoint, the oxidation of 1,2,3,4-tetrahydroquinolines (THQs) is the straightforward way to obtain aromatic quinolines (Figure 1b).¹⁰ Such an approach enables the synthesis of these heteroaromatic compounds with substitution patterns and/or functional groups that are otherwise difficult to insert via traditional aromatic functionalization reactions. As a result, several examples of oxidation of THQs into quinoline derivatives have been recently described through the use of heterogeneous polymaleimide (PMI),¹¹ cobalt oxide¹² or o-quinone-based¹³ catalysts, or via photocatalysis using Ru-, Ir-, or TiO₂ photosensitizers.¹⁴

Figure 1.

Representative bioactive molecules with the quinoline/2-quinolone motif and our design.

Nevertheless, such methods show drawbacks in terms of reaction yields and, not less important, the costs of the catalysts. Herein, we describe a new, milder, and more sustainable route to quinolines via the oxidative aromatization of THQs by MAO-N biocatalysts (Figure 1b). Although this transformation operates well for N-unsubstituted THQs, it was found that the presence of alkyl substituents on the THQs nitrogen posed hurdles toward the formation of quinolinium derivatives and in turn to 2-quinolone frameworks through further oxidation (Figure 1b).¹⁵ A different biocatalytic strategy was therefore investigated to access N-alkylquinolinium compounds, namely, the oxidative cyclization/aromatization of N-cyclopropyl-N-alkylanilines (NCAs) using horseradish peroxidase (HRP). Even if HRP has been the major focus of numerous structural and mechanistic investigation, due to its unusual stability in aqueous solutions,¹⁶ it has been rarely employed in the area of organic synthesis.¹⁷ Moreover, a chemo-enzymatic approach to access 2-quinolone derivatives from NCAs in a one-pot two-steps cascade, combining HRP with K₃Fe(CN)₆, was developed. To the best of our knowledge, this is the first example describing the direct conversion of N-cyclopropyl-N-alkylanilines into 2-quinolone scaffolds.

Results and Discussion

The MAO-N biocatalyzed oxidation of THQs was first investigated. THQ 1a was initially dissolved in NaPBS buffer (pH = 7.8, 1.0 M) at 37 °C in the presence of DMSO as the cosolvent and treated with three whole cell MAO-N biocatalysts (variants D5, D9, and D11)^5,8 and a whole cell hydroxy-D-nicotine oxidase (HDNO)⁸(Table 1).

Table 1. Optimization of the Reaction Conditions of the Biocatalytic Aromatization of THQs^a.

entry	MAO-N	conv.b
1	D5	22
2	D9	43
3	D11	56
4	HDNO	17
5c	D11	50
6d	D11	53
7e	D11	39
8f	D11	52
9	without MAO-N	0
10g	E. coli BL21(DE3) cells	0

^aReaction conditions: 1a (0.2 mmol), freeze-dried MAO-N whole cells (190 mg), buffer (Na₂HPO₄/NaH₂PO₄, pH = 7.8, 1.0 M) (3.0 mL), DMSO (50 μL), air, 37 °C, 7 days.

^bConversion was determined by ¹H NMR integration of the crude mixture.

^c95 mg of freeze-dried MAO-N D11 whole cells was added at the beginning of the reaction, and then, another 95 mg of freeze-dried MAO-N D11 whole cells was added at the third day of the reaction.

^d190 mg of freeze-dried MAO-N D11 whole cells was added at the beginning of the reaction, and then, another 190 mg of freeze-dried MAO-N D11 whole cells was added after 72 h.

^eThe reaction was stopped after 96 h.

^f2 mg of catalase was added.

^gE. coli BL21(DE3) cells harboring no MAO-N enzymes were used.

The variant MAO-N D11 proved to be the most efficient biocatalyst affording the quinoline 2a with 56% conversion (Table 1, entries 1–4). Increasing the amount of MAO-N in the reaction mixture or adding MAO-N biocatalyst in portions did not affect the reaction efficiency in a significant manner (Table 1, entries 5–6). Shorter reaction times provided the quinoline 2a in lower conversion (Table 1, entry 7), while no improvement in the formation of the quinoline product was observed when the reaction was carried out for more than 7 days. To exclude the possibility that H₂O₂ produced during the biocatalytic oxidation could induce the oxidation of the THQ 1a, catalase was added to the biotransformation mixture (Table 1, entry 8). Quinoline 2a was obtained with 52% conversion in the presence of catalase, showing that H₂O₂ did not affect the aromatization of 1a. Finally, two blank experiments without MAO-N or in the presence of E. coli BL21(DE3) cells harboring no MAO-N enzymes were carried out, confirming that the oxidation of 1a is catalyzed by the MAO-N enzyme (Table 1, entries 9–10).

With the optimal conditions in hand, the reaction scope of the biocatalyzed aromatization of THQs was investigated (Table 2). A series of commercially available THQs bearing different substituents were treated with MAO-N D11 whole cells. Both the electronic nature and the substituents on the THQ backbone dramatically affected the oxidation reaction. Electron-donating substituents favored the aromatization of THQs affording the quinolines 2b–2f with good conversions (up to 84%). On the other hand, the THQs bearing a halogen substituent (F, Cl, Br) on the aromatic ring were poorly converted into the corresponding derivatives 2g, 2h, and 2j. Interestingly, the position of the substituents on the THQs also affects the reaction as shown by the 6-MeO-quinoline 2c which was obtained with 69% conversion, while the analogue 7-MeO-quinoline 2i resulted in only 11% conversion.

Table 2. Substrate Scope of the MAO-N Biocatalytic Aromatization of THQs^a.

^aReaction conditions: 1a (0.2 mmol), freeze-dried MAO-N D11 whole cells (190 mg), buffer (Na₂HPO₄/NaH₂PO₄, pH = 7.8, 1.0 M) (3.0 mL), DMSO (50 μL), air, 37 °C, 7 days; conversion was determined by the ¹H NMR integration of the crude mixture.

^bPurified MAO-N D11 was used.

^cIsolated yield based on the recovery starting material.

^dThe reaction was performed in 15 mL of buffer with 250 μL DMSO.

The effect of the substituent position is clear in the series of the methyl-substituted THQs 1b and 1k–m. The quinolines bearing the methyl group at positions C6 (2b) and C2 (2l) were obtained with good conversions (60 and 52%, respectively) while the derivatives 2k and 2m bearing the methyl substituent at C8 and C4, respectively, were formed in low yields (13% for 2m) or not formed at all (2k). In order to evaluate if the conversion of THQs into quinolines could be affected by the use of the whole cell biocatalyst, the substrates 1b, 1c, and 1f were treated with the purified MAO-N D11 enzyme. In all cases, the quinoline derivatives 2b, 2c, and 2f were formed with conversions similar to those obtained in the whole cell-catalyzed biotransformations. The scalability of this protocol was successfully demonstrated on substrates 1b and 1e by performing the reaction on 1 mmol scale, leading to the desired quinolines 2b and 2e with high yields. Finally, the MAO-N aromatization strategy was extended to the N-methyl derivative of 1a. However, despite several attempts, the desired quinolinium ion 4a was not obtained via MAO-N biocatalyzed aromatization.

In order to rationalize the results of the biocatalytic transformations and to determine if the different conversions observed were due to the diverse binding interactions of the THQs with the MAO-N active site, or to an electronic factor, or a combination of the two, a series of in silico studies was carried out.

According to the generally accepted mechanism of the MAO-N catalyzed oxidation,^18,8c the abstraction of a hydride from the methylene group in the α-position to the nitrogen of the THQ by the FAD cofactor represents the initial step of the catalytic cycle. All the substrates 1b–1h bearing a substituent at C6 showed a similar and consistent binding mode to the MAO-N D11 catalytic site, with the methylene group at the α position of the THQ nitrogen correctly oriented toward the FAD cofactor (Figure 2A,B, compounds 1c and 1h).

Figure 2.

Proposed binding modes for compounds (A) 1c, (B) 1h,(C) 1i, (D) (R)-1l, and (E) (S)-1l in the MAO-N D11 catalytic site (PDB ID: 3ZDN). Carbon atoms of compound 1c are shown in purple, compound 1h in dark green, compound 1i in gold, compound (R)-1l in teal, and compound (S)-1l in green. The binding area of the catalytic site is represented as a transparent surface. FAD is represented as a ball-and-stick model. Nitrogen atoms of 1c, 1h, 1i, (R)-1l, (S)-1l and FAD are shown in blue. The α-methylene group is shown as black ball. The nitrogen atom of 1c and 1h is oriented toward the FAD nitrogen and carbon atom (represented as a black atom) involved in the mechanism of the reaction, whereas the presence of 7-MeO group forces compound 1i to adopt several possible binding modes, but none of them in line with the plausible mechanism of MAO-N biocatalytic aromatization. The methyl substituent on C2 is not impeding a correct binding to the active site for both 1l enantiomers, but only the (S)-enantiomer has the α-hydride facing the FAD cofactor, while the (R)-enantiomer present the α-hydride pointing away.

A closer evaluation of the electrostatic potential surface (EPS) of substrate 1c shows a high electron density localized on its nitrogen-containing ring and on its α-methylene group which may favor the abstraction of the hydride unit by the FAD (Figure 3A).

Figure 3.

Electrostatic potential surface (EPS) for (A) 1c, (B) 1h, and (C) 1i.

In contrast, the presence of a Br substituent in 1h withdraws electrons from the α-methylene group reducing its electron density (Figure 3B). As a consequence, the hydride abstraction step by the FAD cofactor and the following aromatization is facilitated for the THQ 1c as compared to 1h, accounting for the different conversions observed (65 versus 9%). This finding suggests the importance of the electronic factor over the binding site occupation for the substrates bearing a substituent at position C6. Differently, the lower conversions observed for the substrates 1i–1m bearing substituents in different positions than C6 seem to be due mainly to their diverse binding interactions with the MAO-N binding site. In fact, the difference in the abundance of electrons around the nitrogen atom and the α-methylene group between 1c and its C7 analogue 1i is not substantial and both substrates have comparable EPS (Figure 3A,C). However, moving the methoxy group from C6 to C7 forces 1i to modify its occupation of the MAO-N site and to adopt several possible binding modes, none of which is however in line with the plausible mechanism of the biocatalytic aromatization (Figure 2C). To accommodate the 7-MeO group in the MAO-N binding pocket, 1i moves the α-methylene group and the nitrogen atom away from the FAD, thus potentially affecting the ability of the cofactor to abstract the α-hydride and leading in turn to quinoline 2i in poor yields. Similarly, the presence of a substituent on the C8 (1k), C4 (1m), or directly on the THQ nitrogen (N-Met-1a) does not allow the correct orientation of these molecules (Figure S4A–D) in the MAO-N binding site. Also, in these cases, the electronic factor seems to be less relevant for the biotransformation outcomes (Figure S5).

Figure 4 shows how the addition of electron-withdrawing and electron-donating groups to the THQ core influences the nucleophilicity of the α-methylene group and their consequent conversion to quinolines. Overall, there is a linear correspondence between the nucleophilicity index of the α-methylene group and the observed conversions, confirming that the electronic factor plays an important role, at different degrees, in the biotransformation outcomes (e.g., it is the main factor for compound 1c, but it is less relevant for 1i). All these data suggest that the MAO-N aromatization of THQs may be affected both by steric and electronic factors as well as by the binding mode of the substrates within the enzyme catalytic pocket.

Figure 4.

Fukui functions showing the relation between the nucleophilicity N index of THQs and the conversions of the MAO-N biocatalyzed aromatization. The nucleophilicity index N prediction was performed using Multiwfn based on the HOMO energies obtained within the Kohn–Sham scheme and defined as N = E_HOMO(Nu) – E_HOMO(TCE).

Finally, the aromatization of the racemic THQ 1l was analyzed. The docking results obtained show that the methyl substituent on C2 of 1l does not obstruct a correct binding to the active site of MAO-N D11 and suggest that the (S)-enantiomer of 1l could potentially bind the MAO-N binding pocket better than the (R)-enantiomer. In fact, while both enantiomers occupy the MAO-N binding site in a similar manner, only the (S)-enantiomer has the α-hydride facing the FAD cofactor, while the (R)-enantiomer presents the α-hydride pointing away (Figure 2ED). However, the unreacted THQ 1l was recovered at the end of the biocatalytic reaction as a racemic mixture, thus suggesting that the (R)-enantiomer modifies its occupation of the binding site during the biotransformation.

From previous experiments, the biocatalyst MAO-N D11 was unable to convert the substrate N-Me-1a into the corresponding quinolinium product 4a (Table 2). Inspired by former studies on the cytochrome P450 and HRP-mediated metabolism of N-cyclopropyl-anilines¹⁸ as well as on the Pd-mediated synthesis of dihydroquinolines from phenyl cyclopropyl carbamates,¹⁹ we decided to explore an alternative biocatalytic pathway to synthesize N-alkyl-quinolinium compounds 4 through an HRP-catalyzed cyclization/aromatization of N-cyclopropyl-N-alkylanilines substrates (Table 3). N-cyclopropyl-N-methylaniline 3a(20) was initially investigated and treated with HRP and H₂O₂ in NaPBS buffer (pH = 5.5) in the presence of different cosolvents. Although the reaction proceeds in the absence of any organic cosolvent (Table 3, entry 1) affording the quinolinium 4a in 54% NMR yield, the addition of acetone as the cosolvent proved to be beneficial, increasing the yield of 4a to 67% (Table 3, entry 2). Lower yields were observed instead when other cosolvents were used (Table 3, entries 3–-6). Different oxidants than H₂O₂, such as the urea-hydrogen peroxide adduct (UHP), oxone, and meta-chloroperoxybenzoic acid (mCPBA) were then tested. When UHP was used, quinolinium 4a was formed with 56% yield, comparable to the data obtained with H₂O₂ (Table 3, entry 7). On the other hand, no formation of 4a was achieved with oxone and mCPBA (Table 3, entries 8–9). Importantly, no desired product 4a was detected in the absence of H₂O₂, thus suggesting that the reaction is catalyzed by the HRP activated by the peroxide (Table 3, entry 10). Furthermore, when the biocatalytic cyclization/aromatization reaction was carried out without HRP and in the presence of sole H₂O₂, only a low amount of the desired product 4a was observed (Table 3, entry 11). Lowering or increasing the amount of hydrogen peroxide led to inferior results (Table 3, entries 12–13).²¹

Table 3. Optimization of the Reaction Conditions of the Biocatalytic Cyclization/Aromatization of N-Cyclopropyl-N-methylaniline^a.

entry	cosolvent	peroxidec	yield (%)b
1		H2O2	54
2	acetone	H2O2	67
3	THF	H2O2	50
4	DCE	H2O2	28
5	MeCN	H2O2	37
6	DMSO	H2O2	36
7	acetone	UHP	56
8	acetone	oxone	0
9	acetone	mCPBA	0
10	acetone		0
11d	acetone	H2O2	<10
12e	acetone	H2O2	33
13f	acetone	H2O2	52
14g	acetone	H2O2	58
15h	acetone	H2O2	61
16i	acetone	H2O2	60
17i,j	acetone	H2O2	67

^aReaction conditions: 3a (0.08 mmol, 1 equiv), NaPBS (pH = 5.5, 0.4 M, 1 mL), cosolvent (1% v/v), peroxide (2.5 equiv), 100 μL HRP (4 mg per 1 mL NaPBS, 240 U/mg), rt, 2 h.

^bNMR yield determined by ¹H-NMR with sodium 4-methylbenzenesulfonate as the internal standard.

^cUnless otherwise specified, 30% H₂O₂ was used.

^dNo HRP was added.

^e10 μL 30% H₂O₂ was used.

^f30 μL 30% H₂O₂ was used.

^gThe reaction was carried out under N₂.

^h50 μL HRP was added after 0.5 h.

ⁱ100 μL HRP was added after 0.5 h.

^j2 mL NaPBS and 20 μL acetone were used.

The nitrogen atmosphere was also unhelpful to improve the yield of 4a (Table 3, entry 14), as well as increasing the loading of HRP or the volume of the buffer had no obvious effect on the yield of this biotransformation (Table 3, entries 15–17).

The scope and versatility of the HRP-catalyzed cyclization/aromatization strategy were then explored. A range of N-cyclopropyl-N-alkylanilines 3a–n bearing different substituents on the phenyl ring were prepared and treated with HRP/H₂O₂ under the previously identified optimal reaction conditions (Table 4).

Table 4. Substrate Scope of the Cyclization/Aromatization of N-Cyclopropyl-N-alkylanilines 3^a,b.

^aReaction conditions: 3 (0.08 mmol, 1 equiv), 2 mL NaPBS (pH = 5.5, 0.4 M), 20 μL acetone, 20 μL H₂O₂ (30%), 100 μL HRP (4 mg per 1 mL NaPBS, 240 U/mg) and then 50 μL after 0.5 h, rt.

^bNMR yields determined by ¹H NMR with sodium 4-methylbenzenesulfonate as the internal standard.

As a general trend, rather than the nature of the substituent (electron-donating groups, halogens, phenyl), the cyclization/aromatization was more affected by the position of the substituents on the phenyl ring. The para-Me-substituted derivative 3b led to quinolinium 4b with good 50% yield, while the aniline 3c bearing a MeO-group at the para-position on the phenyl ring was converted to 4c with low yield. It was found that 4c was quite unstable even under air conditions, due to its electron-rich properties, and this may have affected the outcome of the reaction. On the other hand, aniline 3g bearing a MeO-group at the meta-position was converted into an inseparable mixture of the isomers 4g/4 g′ (1:1 ratio) with good 58% yield, thus highlighting the effect of the substituents on the outcome of the reaction. Similarly, the di-substituted anilines 3j (ortho-para-dimethyl) and 3k (meta-dimethyl) showed opposite reactivity. No reaction occurred for substrate 3j, showing that a second substituent at the ortho-position was detrimental for the biotransformation. Such an effect, especially if compared with substrate 3b, may be attributed to steric hindrance. In contrast, derivative 3k was converted into quinolinium 4k with excellent 83% yield.

The anilines 3d, 3e, and 3f bearing, respectively, a phenyl, a fluorine, and a chlorine substituent at the para-position of the phenyl ring were all converted into the corresponding quinolinium ions 4d–f with good yields, while both the F- and Cl-meta-substituted anilines 3h and 3i were converted with good yields into the corresponding 4h/4h′ and 3i/3i′ as isomeric mixtures (1:1 ratio of 5- and 7-substituted quinolinium ions). Remarkably, the tricyclic product 4L was smoothly synthesized in 77% yield. Finally, the two N-substituted anilines 3m–n, bearing an Et or Bn group on the N-atom, were also investigated. Both anilines were converted into the N-alkyl quinolinium ions 4m and 4n, respectively, in good-moderate yields.²²

Based on previous reports²³ and our experiments, a proposed mechanism for the HRP biocatalyzed cyclization/aromatization of N-cyclopropyl-N-alkylanilines 3 was proposed (Scheme 1). Initially, the rapid transfer of an oxygen atom from H₂O₂ to the ferric heme cofactor of HRP produced a porphyrin cation-radical/oxoiron complex HRP-Fe=O and a molecule of water. The amine substrates were oxidized by HRP-Fe=O through a single electron transfer (SET) process, giving rise to an N-center radical A and HRP-Fe–O^.–. In situ EPR experiments provided further evidence for the possible involvement of the HRP-Fe–O^. – radical in this transformation. The ring opening of A formed the radical B which collapsed on the aromatic ring leading, after deprotonation, to the intermediate E. A second one-electron oxidation of E forms the enamine F, with the regeneration of HRP. The enamine F is ultimately oxidized by HRP-Fe-O^.- into the final quinolinium ion 4a in a similar way.

Scheme 1. Proposed Mechanism for the HRP-Catalyzed Oxidation Process.

Finally, the synthetic potential of this HRP biocatalyzed transformation was explored through the development of a chemo-enzymatic sequence to convert anilines 3 into 2-quinolone compound 5 in a two-step one-pot manner (Table 5).

Table 5. Chemo-Enzymatic Cascade for the Synthesis of 2-Quinolones 5^a.

^aTwo steps one-pot cascade. K₃Fe(CN)₆ was added after the completion of the first biocatalytic step.

Quinolinium ions are reactive substrates which can be further carbonylated into quinolone derivates through the addition of K₃Fe(CN)₆.²⁴ The four N-cyclopropyl-N-alkylanilines which gave the best results in the substrate scope screening were treated with HRP and H₂O₂ and then added with K₃Fe(CN)₆ and NaOH. The 2-quinolone derivates 5a, 5e, 5k, and 5m (Table 5) were all obtained with good isolated yields from the corresponding anilines in a one-pot two-step chemo-enzymatic cascade. The tricyclic product 5l was also obtained and isolated from the chemo-enzymatic cascade, albeit with a lower yield.

Conclusions

In conclusion, we have developed two new biocatalytic methodologies for the synthesis and construction of quinoline and 2-quinolone heterocycles using two different oxidase enzymes. A series of quinoline derivatives were obtained from 1,2,3,4-tetrahydroquinoline substrates using monoamine oxidase (MAO-N) biocatalysts in good yields. Computational studies highlighted that the MAO-N biotransformation may be affected both by steric and binding effects as well as by the electronic properties of the THQ substrates. In parallel, HPR was successfully employed in the construction of a range of quinolinium derivatives from N-cyclopropyl-N-alkylaniline substrates through a cyclization/aromatization radical cascade, highlighting the possibility to construct heteroaromatic rings from aliphatic substrates. Furthermore, a chemo-enzymatic sequence was designed and successfully developed to convert N-cyclopropyl-N-alkylanilines into 2-quinolones in a one-pot procedure. These results further confirm the key role of biocatalysis in the synthesis of a variety of organic molecules, including nonchiral heteroaromatic compounds.

Glossary

Abbreviations

MAO-N

monoamine oxidase

NCA

N-cyclopropyl-N-alkylaniline

HRP

horseradish peroxidase

UHP

urea-hydrogen peroxide

mCPBA

meta-chloroperoxybenzoic acid

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acscatal.2c05902.

• Experimental details; procedures; and copies of spectra for new compounds (PDF)

Author Present Address

^# College of Chemistry and Chemical Engineering, Central South University, Changsha 410083, P. R. China

Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

K. C. Wong Foundation for funding and financial support to HX. The EPR measurements were performed at the Centre for Pulse EPR at Imperial College London (PEPR), supported by the EPSRC grant EP/T031425/1.

The authors declare no competing financial interest.