Biocatalytic oxidation of alcohols using galactose oxidase and a manganese(iii) activator for the synthesis of islatravir

Article abstract of DOI:10.1039/d0ob02395g

Galactose oxidase (GOase) is a Cu-dependent metalloenzyme that catalyzes the oxidation of alcohols to aldehydes. An evolved GOase variant was recently shown to catalyze a desymmetrizing oxidation as the first enzymatic step in the biocatalytic synthesis of islatravir. Horseradish peroxidase (HRP) is required to activate the GOase, introducing cost and protein burden to the process. Herein we describe that complexes of earth-abundant Mn(iii) (e.g.Mn(OAc)₃) can be used at low loadings (2 mol%) as small molecule alternatives to HRP, providing similar yields and purity profiles. While an induction period is observed when using Mn(OAc)₃as the activator, employment of alternative Mn(iii) sources, such as Mn(acac)₃and K₃[Mn(C₂O₄)₃], eliminates the induction period and provides higher conversions to product. We demonstrate that use of the Mn(OAc)₃additive is also compatible with subsequent biocatalytic steps in the islatravir-forming cascade. Finally, to exhibit the wider utility of Mn(OAc)₃, we show that Mn(OAc)₃functions as a suitable activator for several commercially available variants of GOase with a series of alcohol substrates.

Full text of DOI:10.1039/d0ob02395g

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Biocatalytic oxidation of alcohols using galactose
oxidase and a manganese(III) activator for the
synthesis of islatravir†
Cite this: Org. Biomol. Chem., 2021,
19, 1620
Heather C. Johnson, * Shaoguang Zhang,
Anna Fryszkowska,
Sandra A. Robaire, Artis Klapars, Niki R. Patel,
Mark A. Huffman and Neil A. Strotman
Serge Ruccolo,
Aaron M. Whittaker,
Galactose oxidase (GOase) is a Cu-dependent metalloenzyme that catalyzes the oxidation of alcohols to
aldehydes. An evolved GOase variant was recently shown to catalyze a desymmetrizing oxidation as the
first enzymatic step in the biocatalytic synthesis of islatravir. Horseradish peroxidase (HRP) is required to
activate the GOase, introducing cost and protein burden to the process. Herein we describe that com-
plexes of earth-abundant Mn(III) (e.g. Mn(OAc)3) can be used at low loadings (2 mol%) as small molecule
alternatives to HRP, providing similar yields and purity profiles. While an induction period is observed
when using Mn(OAc)3 as the activator, employment of alternative Mn(III) sources, such as Mn(acac)3 and
K3[Mn(C2O4)3], eliminates the induction period and provides higher conversions to product. We demon-
strate that use of the Mn(OAc)3 additive is also compatible with subsequent biocatalytic steps in the isla-
travir-forming cascade. Finally, to exhibit the wider utility of Mn(OAc)3, we show that Mn(OAc)3 functions
as a suitable activator for several commercially available variants of GOase with a series of alcohol
substrates.
Received 1st December 2020,
Accepted 19th January 2021
DOI: 10.1039/d0ob02395g
rsc.li/obc
oxygen as a terminal oxidant. Synthetic applications of GOase
require the use of two additional metalloenzymes: catalase
Introduction
Biocatalysis has been increasingly adopted by the pharma- serves to decompose in situ generated H2O2, and horseradish
ceutical industry in the synthesis of complex molecules.1–3 peroxidase (HRP) serves to activate GOase to reach the desired
The exquisite chemo-, regio- and stereoselectivity conferred oxidation state of the Cu active site (Scheme 2).
by enzymes render them an attractive option as green and
The generally accepted catalytic cycle for GOase7,8 is shown
sustainable catalysts for many synthetic transformations. in Scheme 2. GOase can exist in three unique states, of which
Recently, our laboratories disclosed a fully biocatalytic route to two (II and III) are catalytically active, while species I is a cata-
synthesize islatravir, an HIV reverse transcriptase translocation lytically inactive Cu(II) complex. One-electron oxidation of
inhibitor (Scheme 1).4 Islatravir was formed from 2-ethynylgly- complex I generates the Cu(II) radical species II, which pro-
cerol (1) in a three-step cascade (single aqueous stream) using motes the oxidation of the alcohol substrate to form the
nine enzymes, five of which were engineered via directed desired aldehyde via a sequential two-electron, two-proton
evolution. The first step in the cascade utilized an evolved transfer process. The concomitant Cu(I) species III can then
variant of galactose oxidase (GOase, EC 1.1.3.9) from Fusarium reform complex II via oxidation with molecular oxygen. It has
graminearum, which catalyzed the desymmetrizing oxidation been suggested that the reactive species II can undergo a one
of prochiral 2-ethynylglycerol to form (R)-2-ethynylglycer- electron reduction deactivation pathway and fall off the cata-
aldehyde (2).4
lytic cycle,9 resulting in a slower rate and incomplete substrate
GOase belongs to the copper radical oxidase enzyme conversion. Generally, peroxidase enzymes, such as HRP, have
family5 (AA5 in the CAZy enzyme classification6) that catalyzes been utilized as enzymatic activators to efficiently (re)generate
two-electron oxidations of alcohols to aldehydes and uses oxidized species II from reduced
I via one electron
oxidation,10,11 although this protein–protein interaction is still
not fully understood.12 At large scale, use of HRP can add sig-
nificant cost,11 as well as increase protein burden, adding con-
siderably to the challenges of process development. As our
original disclosure utilized 3 wt% of a purified HRP,4 we recog-
Department of Process Research and Development, Merck & Co., Inc., Rahway,
NJ 07065, USA. E-mail: heather.johnson@merck.com
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0ob02395g
1620 | Org. Biomol. Chem., 2021, 19, 1620–1625
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 1 Overview of the biocatalytic cascade used to synthesize islatravir.
dants such as persulfates, organic peroxides and high-valent
transition-metal complexes. HRP (6 wt%) and activator-free
conditions were tested as positive and negative controls,
respectively. A selection of the screening results is shown in
Chart 1. Only marginal conversion (3%) was observed in the
reaction in the absence of an activator, presumably due to pre-
dominant formation of inactive species I; upon addition of
CuSO4 to copper-free purified GOase-1, reduced species I is
observed by UV-vis spectroscopy (see ESI†) indicating this
variant primarily forms I in the absence of an activator.
Conditions using most activators shown in Chart 1 achieved
some conversion with similar enantioselectivities to the HRP
control, showing the oxidations are primarily GOase-mediated
in these cases. Despite this, inclusion of the majority of
screened oxidants resulted in low conversions to the desired
aldehyde, likely due to one or more of the following reasons:
(1) poor solubility of the activator, (2) undesired reactions with
enzymes, (3) side product formation (e.g. over-oxidation to
2-ethynylglyceric acid, or degradation of 2-ethynylglyceralde-
hyde to formic acid),4 or (4) insufficient redox potential.
Interestingly, employment of the most commonly used activa-
tors in the literature, K3Fe(CN)6 and K3IrCl6, led to poor yields
of 2-ethynylglyceraldehyde (17% and 36%, respectively).
Promisingly, Mn(III) complexes Mn(OAc)3 and MnF3 emerged
as viable leads, helping to provide conversions (56% and 59%,
respectively) comparable to HRP (62% conversion) and with
similar purity. To avoid the possibility of in situ HF formation
on scale, Mn(OAc)3 was selected as our lead for further investi-
gation. In an early literature report, the Mn(III)-EDTA complex
was utilized as a GOase activator in the oxidation of galactose,
but was observed to activate more slowly than chloroiridate.17
Further explorations of Mn(III) as a GOase activator have not
Scheme 2 Simplified catalytic cycle for GOase, as proposed by
Whittaker.7 Adapted from ref. 11.
nized the opportunity for improvement by identifying a small
molecule chemical oxidant as a low-cost, non-enzymatic
GOase activator. In previous studies, K3Fe(CN)6 and K3IrCl6
have been used as activators for GOase to catalyze the oxi-
dation of galactose13 and to generate fully activated GOase for
spectroscopic studies.14 Moreover, K3Fe(CN)6 has been used to
activate oxidases in biocatalytic oxidations of benzyl alcohol,
although high concentrations relative to substrate were
required (40 mol% to equimolar or above, depending on the
oxidase).11,12,15 Despite these advances, there remains a need
for a practical and economical small molecule GOase activator
for synthetic scale reactions. Herein, we report that catalytic
quantities of Mn(III) salts can be used as attractive alternatives
to HRP to activate GOase for the biocatalytic oxidation of alco-
hols, providing significant cost savings while maintaining
high levels of catalytic activity.
been reported.18 Mn(OAc)3 has been commonly used as an
19
Results and discussion
oxidant in organic reactions, often using acetic acid as the
To identify suitable one-electron oxidant activators for GOase, solvent,20 and has been shown to act as both a Lewis acid and
we turned to high-throughput experimentation (HTE). HTE electron transfer reagent for radical initiation.21 In aqueous
screens utilized the best performing GOase variant at the time solution, the Mn(III) oxidation state is susceptible to dispropor-
of initiating screening, GOase-1;16 using HRP as an activator, tionation to Mn(IV) (i.e. MnO2) and Mn(II), although it is stabil-
this variant yielded ca. 60% conversion to the desired aldehyde ized by O-donor ligands,22 and Mn(OAc)3 was recently shown
in 85% ee, with minimal over-oxidation. We sought GOase- to operate as an oxidant in aqueous systems for the oxidation
compatible oxidants that have suitable oxidation potential to of sulfinates.23
activate species I, but are unable to directly oxidize the triol to
Following the results from HTE screening, the Mn(OAc)3
the racemic aldehyde. Over 100 small molecule oxidants were conditions were tested on gram-scale to verify the screening
selected and screened at various loadings (10 to 100 mol% result (Fig. 1). For this experiment, a further-evolved GOase
relative to 2-ethynylglycerol, see ESI† for details).
variant, GOase-2, was used;16 during development, process
In addition to the oxidants shown in the literature to acti- optimization and enzyme evolution were progressing in paral-
vate GOase, we evaluated a wide variety of commonly used oxi- lel, and GOase-2 emerged as a catalytically more active variant
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 1620–1625 | 1621

View Article Online
Paper
Organic & Biomolecular Chemistry
Chart 1 Selected HTE screening results on small molecule activators for GOase-1. Conversion to aldehyde is represented by bars, and the aldehyde
ee by circles.
with improved enantioselectivity after two rounds of directed
evolution from GOase-1. With 14 wt% of GOase-2 powder and
10 mol% Mn(OAc)3, 2-ethynylglyceraldehyde was obtained in
60% yield, compared with 57% yield using HRP. The enantio-
meric excess was comparable between the activators (96% and
95% ee for HRP and Mn(OAc)3, respectively), and purity pro-
files were similar (see ESI†). The Mn(OAc)3 loading was
reduced to only 2 mol% without impacting the 2-ethynylglycer-
aldehyde yield after ca. 20 hours reaction, suggesting that the
rate of catalysis is greater than 50 times the rate of enzyme de-
activation. Without an activator, 2-ethynylglyeraldehyde was
formed in 29% yield after 16 hours (in 82% ee), suggesting
that limited activation of this GOase variant occurs in the
absence of an activator (Fig. 1).19 No oxidation was observed in
the control reaction of 2-ethynylglycerol with Mn(OAc)3 in the
absence of GOase, showing Mn(OAc)3 does not directly
Fig. 1 Desymmetrizing oxidation of 2-ethynylglycerol (0.258 M) using
promote the oxidation of 2-ethynylglycerol.
6 wt% HRP or Mn(OAc)3 (2 or 10 mol%) as activators for GOase-2. Mol%
and wt% are relative to 2-ethynylglycerol.
Since desymmetrizing oxidation is the first step in the bio-
catalytic cascade to synthesize islatravir,4 any chosen alterna-
tive to HRP for this process must be compatible with the bioca-
talytic transformations carried out downstream. Gratifyingly, a
representative stream of 2-ethynylglyceraldehyde formed in the
With Mn(OAc)3 identified as a competent HRP alternative,
presence of Mn(OAc)3 (5 mol% relative to 2-ethynylglycerol) we sought to understand more about the nature of the GOase
performed well in the subsequent reactions. The stream gave activation. Closer examination of the reaction profile using
96% conversion to 2-ethynylglyceraldehyde 3-phosphate (3) in 2 mol% Mn(OAc)3 revealed an induction period in the product
the kinase-catalyzed phosphorylation, and the four-enzyme formation. Profiling of the initial regime was conducted to
aldol/glycosylation step produced islatravir in 86% isolated further probe the activation period using Mn(OAc)3 at 1 to
yield (relative to 2-fluoroadenine, see ESI† for details).
10 mol% loading. Upon increasing the loading, the induction
1622 | Org. Biomol. Chem., 2021, 19, 1620–1625
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
Varying the loading of water-soluble K3[Mn(C2O4)3] from 1
to 10 mol% (relative to 2-ethynylglycerol) resulted in almost
identical initial reaction rates (see ESI†), as was observed post-
induction with Mn(OAc)3, suggesting that the activator loading
in this range does not impact the initial catalytic turnover of
GOase. While use of 2 mol% and 4 mol% of K3[Mn(C2O4)3] led
to essentially the same yields of 2-ethynylglyceraldehyde (74%
and 75%, respectively), decreasing the loading to 1 mol%
resulted in a somewhat lower 68% yield. As the initial rates are
similar across each loading, the difference in yield is likely due
to formation of inactive I via off-cycle reduction of active II
during catalysis (Scheme 2). The lower yield observed with
1 mol% K3[Mn(C2O4)3] may suggest that an off-cycle reduction
occurs at ca. 1/50th to 1/100th the rate of catalytic turnover with
this variant,25 although some decomposition of K3[Mn
Fig. 2 Initial 2-ethynylglyceraldehyde formation from oxidation of
2-ethynylglycerol (0.258 M) using GOase-2 (14 wt%), catalase (14 wt%),
and various Mn(OAc)3 loadings. Mol% and wt% are relative to
2-ethynylglycerol.
period was observed to decrease (Fig. 2). However, following (C2O4)3], which would reduce the re-activation, and therefore
the induction period, the productive regime of the catalysis overall yield, cannot be discounted. At 10 mol% loading, only
was shown to proceed at similar rates independent of Mn 51% 2-ethynylglyceraldehyde was produced, possibly
(OAc)3 loading.
suggesting additional mechanistic complexity in which higher
The presence of an induction period suggests that the concentrations of oxalate may irreversibly deactivate GOase or
active state of GOase is not formed immediately on addition of catalase. Moreover, while K3[Mn(C2O4)3] is a soluble source of
Mn(OAc)3, and also that the activating state of Mn(OAc)3 might Mn(III) and effective activator, its sensitivity to light and heat26
not be formed immediately. To further understand the nature and lack of commercial availability render it a less attractive
of the induction period, several other Mn species were investi- candidate for incorporation into a manufacturing process.
gated as activators, with the insoluble oxides MnO2 and
With Mn(OAc)3 established as a convenient alternative to
Mn2O3 proving ineffective activators of GOase-2 (Fig. 3). This HRP for our substrate and engineered GOase, we next aimed
observation suggests that MnO2, potentially formed in situ by to show that Mn(OAc)3 is an effective HRP replacement for
disproportionation of Mn(III), is unlikely to be the active alternative GOase-mediated oxidation reactions. Commercially
oxidant derived from Mn(OAc)3. Mn(acac)3 or K3[Mn(C2O4)3], available GOase variants M1 and M3–5 were examined to
Mn(III) complexes with distorted octahedral coordination geo- demonstrate that the Mn(OAc)3 activation is not specific to
metry,22 are competent activators at 2 mol% loading, providing enzymes engineered for islatravir synthesis. GOase M1 was
comparable initial catalysis as Mn(OAc)3 (post-induction developed by Arnold and co-workers as a variant with higher
period). After 16 hours, the aldehyde yield with Mn(acac)3 was stability and improved expression in E. coli compared with
62% (compared with 59% with Mn(OAc)3 under parallel con- wild-type GOase,27 while GOase M3–5 was evolved by Turner to
ditions), while K3[Mn(C2O4)3] enables higher aldehyde yield of oxidize secondary alcohols,28 but has also been shown to
73%. Furthermore, the induction period is not observed with operate on a variety of primary alcohol substrates.29–32 Table 1
these activators, suggesting that the Mn(III) oxidation state is shows the comparisons at selected time points between identi-
responsible for the oxidation of I to II, as has been previously cal conditions proceeding with either no activator, HRP
proposed for Mn(III)-EDTA-mediated activation.17 We speculate (3 wt%) or Mn(OAc)3 (5 mol%). These examples demonstrate
that the induction period observed with Mn(OAc)3 may be that aldehyde yields with Mn(OAc)3 as the GOase activator are
attributed to modifications of the trimeric structure19 required comparable with, or superior to, results obtained using HRP,
to undergo electron transfer with GOase, and/or the limited confirming the viability of using Mn(OAc)3 as a cost-effective
solubility of Mn(OAc)3.24
and widely available HRP replacement.33
Fig. 3 Initial 2-ethynylglyceraldehyde formation from oxidation of 2-ethynylglycerol (0.258 M) using GOase-2 (14 wt%), catalase (14 wt%), and
various activators (2 mol%). Mol% and wt% are relative to 2-ethynylglycerol.
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 1620–1625 | 1623

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 Oxidation of various alcohols with GOase comparing no activator, Mn(OAc)3 (5 mol%) or HRP (3 wt%)
Entry Substrate
1
GOase variant/loading Reaction time/h % aldehyde (no activator) % aldehyde (Mn(OAc)3) % aldehyde (HRP)
M3–5 10 wt%
24
8
1
59
88
54
92
2
GOase-2 20 wt%
5.5
3
4
M1 10 wt%
2.7
16
20
8
55
73
35
72
GOase-2 20 wt%
5
6
GOase-2 20 wt%
M3–5 20 wt%
8
1
46
89
54
81
2.7
47
% yields were determined in situ by 1H NMR spectroscopy using an internal standard. 10 vol% DMSO was used in all entries except 2. Mol% and
wt% are relative to the primary alcohol substrate.
Schultz for useful discussions. We thank Codexis, Inc. for
directed evolution and supplying the enzymes.
We have demonstrated that Mn(III) complexes are effective
Conclusions
activators of GOase for biocatalytic alcohol oxidations. In
seeking a convenient and economical replacement for HRP,
over 100 potential activators were rapidly screened via high
References
1 A. Fryszkowska and P. N. Devine, Curr. Opin. Chem. Biol.,
2020, 55, 151–160.
2 R. A. Sheldon and J. M. Woodley, Chem. Rev., 2018, 118,
801–838.
throughput experimentation. While incorporation of many
activators provided the desired aldehyde in poor yields, pre-
sumably due to incompatibility with GOase and/or overoxi-
dation of the product, we showed that loadings as low as
2 mol% Mn(OAc)3 enabled the formation of (R)-2-ethynylgly-
ceraldehyde from 2-ethynylglycerol in comparable yield to
HRP. As a result, this readily available activator can enable
a lower cost process and reduces the protein burden, which
simplifies downstream purifications. An induction period
was observed when using Mn(OAc)3, but could be elimi-
nated using alternative Mn(III) sources such as Mn(acac)3 or
K3[Mn(C2O4)3]. The broad generality of Mn(III) salts as
single-electron activators was demonstrated with both the
M1 and M3–5 variants of GOase and could be applied to the
oxidation of six representative examples of primary alcohol
substrates. We expect that this combination of bio- and
chemo-catalysis will be a useful method for the synthetic
community.
3 K. R. Campos, P. J. Coleman, J. C. Alvarez, S. D. Dreher,
R. M. Garbaccio, N. K. Terrett, R. D. Tillyer, M. D. Truppo
and E. R. Parmee, Science, 2019, 363, eaat0805.
4 M. A. Huffman, A. Fryszkowska, O. Alvizo, M. Borra-Garske,
K. R. Campos, K. A. Canada, P. N. Devine, D. Duan,
J. H. Forstater, S. T. Grosser, H. M. Halsey, G. J. Hughes,
J. Jo, L. A. Joyce, J. N. Kolev, J. Liang, K. M. Maloney,
B. F. Mann, N. M. Marshall, M. McLaughlin, J. C. Moore,
G. S. Murphy, C. C. Nawrat, J. Nazor, S. Novick, N. R. Patel,
A. Rodriguez-Granillo, S. A. Robaire, E. C. Sherer,
M. D. Truppo, A. M. Whittaker, D. Verma, L. Xiao, Y. Xu
and H. Yang, Science, 2019, 366, 1255–1259.
5 P. Kersten and D. Cullen, Fungal Genet. Biol., 2014, 72, 124–130.
6 A. Levasseur, E. Drula, V. Lombard, P. M. Coutinho and
B. Henrissat, Biotechnol. Biofuels, 2013, 6, 41.
7 J. W. Whittaker, Chem. Rev., 2003, 103, 2347–2363.
8 J. W. Whittaker, Arch. Biochem. Biophys., 2005, 433, 227–239.
9 M. M. Whittaker and J. W. Whittaker, J. Biol. Chem., 1988,
263, 6074–6080.
Conflicts of interest
There are no conflicts to declare.
10 L. D. Kwiatkowski and D. J. Kosman, Biochem. Biophys. Res.
Commun., 1973, 53, 715–721.
11 A. T. Pedersen, W. R. Birmingham, G. Rehn, S. J. Charnock,
N. J. Turner and J. M. Woodley, Org. Process Res. Dev., 2015,
19, 1580–1589.
Acknowledgements
We thank Margaret Miller for assistance with setup for scaling
the oxidation reactions, Gregory Hughes and Rose Mathew for 12 S. M. Forget, H. Brumer, F. R. Xia and J. E. Hein, Org.
assistance with 2-ethynylglycerol preparation, and Danielle
Biomol. Chem., 2020, 18, 2076–2084.
1624 | Org. Biomol. Chem., 2021, 19, 1620–1625
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
13 L. D. Kwiatkowski, M. Adelman, R. Pennelly and
D. J. Kosman, J. Inorg. Biochem., 1981, 14, 209–222.
14 C. G. Saysell, C. D. Borman, A. J. Baron, M. J. McPherson
and A. G. Sykes, Inorg. Chem., 1997, 36, 4520–4525.
15 K. Parikka and M. Tenkanen, Carbohydr. Res., 2009, 344,
14–20.
K3[Mn(C2O4)3] is fully soluble at all loadings used in these
studies (>11 mg mL−1). As an induction period was not
observed using Mn(acac)3, it is likely that structural modifi-
cations to Mn(OAc)3 are responsible for the observed induc-
tion period, but differing solubility kinetics in the reaction
mixture also cannot be discounted.
16 In this manuscript, GOase-1 and GOase-2 refer to 25 While this variant appears to form I every 50–100 cycles,
GALO-104 and GALO-105, respectively, which are the com-
mercial names of these enzymes available for purchase
from Codexis. These were described in ref. 4 as GOase
Rd10bb and GOase Rd12bb, respectively.
the frequency of I formation is highly dependent on the
variant of GOase. See: M. S. Rogers, E. M. Tyler,
N. Akyumani, C. R. Kurtis, R. S. Spooner, S. E. Deacon,
S. Tamber, S. J. Firbank, K. Mahmoud, P. F. Knowles,
S. E. V. Phillips, M. J. McPherson and D. M. Dooley,
Biochemistry, 2007, 46(15), 4606–4618; G. R. Dyrkacz,
R. D. Libby and G. A. Hamilton, J. Am. Chem. Soc., 1976, 98,
626–628. In the latter example, I is proposed to form every
few thousand turnovers.
17 G. A. Hamilton, P. K. Adolf, J. de Jersey, G. C. DuBois,
G. R. Dyrkacz and R. D. Libby, J. Am. Chem. Soc., 1978, 100,
1899–1912.
18 Mn(III) has been shown to reactivate the catalytically inac-
tive form of glyoxal oxidase, another copper radical oxidase
from the AA5 enzyme family in the CAZy classification 26 G. H. Cartledge and W. P. Ericks, J. Am. Chem. Soc., 1936,
system (ref. 6); in this case, glyoxal oxidase was proposed to 58, 2061–2065.
undergo a one-electron oxidative (re)activation by water- 27 L. Sun, I. P. Petrounia, M. Yagasaki, G. Bandara and
soluble Mn(III) oxalate, of which the regeneration from Mn F. H. Arnold, Protein Eng., Des. Sel., 2001, 14, 699–704.
(II) was mediated by peroxidase. See: T. Roncal, C. Muñoz, 28 F. Escalettes and N. J. Turner, ChemBioChem, 2008, 9, 857–
L. Lorenzo, B. Maestro and M. del M. Díaz de Guereñu,
Enzyme Microb. Technol., 2012, 50(2), 143–150.
860.
29 S. Herter, S. M. McKenna, A. R. Frazer, S. Leimkühler,
A. J. Carnell and N. J. Turner, ChemCatChem, 2015, 7,
2313–2317.
19 Unlike most Mn(III) complexes, which exist as Jahn–Teller
distorted octahedral complexes, Mn(OAc)3 refers to com-
pounds based upon a trinuclear [Mn3O(OAc)6·2H2O]+ unit 30 B. Bechi, S. Herter, S. McKenna, C. Riley, S. Leimkühler,
(see ref. 22); for simplicity, we will continue to refer to the
compound as Mn(OAc)3.
N. J. Turner and A. J. Carnell, Green Chem., 2014, 16, 4524–
4529.
20 M. Mondal and U. Bora, RSC Adv., 2013, 3, 18716–18754.
21 G. W. Wang, Y. W. Dong, P. Wu, T. T. Yuan and Y. B. Shen,
J. Org. Chem., 2008, 73, 7088–7095.
31 S. Staniland, B. Yuan, N. Giménez-Agulló, T. Marcelli,
S. C. Willies, D. M. Grainger, N. J. Turner and J. Clayden,
Chem. – Eur. J., 2014, 20, 13084–13088.
22 N. N. Greenwood and A. Earnshaw, Chemistry of the 32 B. Yuan, A. Page, C. P. Worrall, F. Escalettes, S. C. Willies,
Elements, Butterworth-Heinemann, Oxford, 2nd edn, 1997.
23 Y. S. Xiong, J. Weng and G. Lu, Adv. Synth. Catal., 2018,
360, 1611–1616.
24 Freshly prepared mixtures of Mn(OAc)3·2H2O or Mn(acac)3
in water showed <0.5 mg mL−1 solubility. By contrast,
J. J. W. McDouall, N. J. Turner and J. Clayden, Angew.
Chem., Int. Ed., 2010, 49, 7010–7013.
33 Reactivity without a co-activator is observed in some cases,
although the reactions are nonetheless accelerated by
addition of the co-activator.
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 1620–1625 | 1625