Studies towards the synthetic applicability of biocatalytic allylic oxidations with the lyophilisate of Pleurotus sapidus

Article abstract of DOI:10.1016/j.molcatb.2015.07.008

The edible fungus Pleurotus sapidus (PSA) is a particularly interesting biocatalytic system for allylic oxidation and has a remarkably broad substrate range from terpenoids to fatty acids. The oxidations are most likely catalyzed by a lipoxygenase and involve the formation of peroxides via radical intermediates in the first rate-limiting step. We provide herein a rationalization of the observed regioselectivity of these conversions by means of computational determination of bond dissociation enthalpies of a set of tailor-made spirocyclic terpenoids. It was found that only strongly activated allylic positions (BDH₂₉₈ of <80 kcal/mol) with neighboring heteroatoms or with activating alkyl groups are oxidized to the corresponding unsaturated lactones or enones, respectively. With the synthesis and purification of allylic hydroperoxide intermediates, we have been able to characterize the putative direct precursors of enones in PSA oxidations. Our results suggest a two-step oxidation mechanism involving hydroperoxide intermediates which are rapidly converted to the observed enones by an enzymatic reaction.

Full text of DOI:10.1016/j.molcatb.2015.07.008

Journal of Molecular Catalysis B: Enzymatic 121 (2015) 15–21
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Studies towards the synthetic applicability of biocatalytic allylic
oxidations with the lyophilisate of Pleurotus sapidus
Verena Weidmanna, Serge Kliewera, Marko Sicka, Sergej Bycinskija, Margarethe Kleczkab
, Julia Rehbeina, Axel G. Griesbeckb, Holger Zornc, Wolfgang Maisona,∗
a University of Hamburg, Department of Chemistry, Pharmaceutical and Medicinal Chemistry, Bundesstraße 45, 20146 Hamburg, Germany
b Department of Chemistry, University of Cologne, 50939 Köln, Germany
c Institute of Food Chemistry and Food Biotechnology, Justus-Liebig-University Giessen, Heinrich-Buff-Ring 58, 35392 Giessen, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
The edible fungus Pleurotus sapidus (PSA) is a particularly interesting biocatalytic system for allylic oxida-
tion and has a remarkably broad substrate range from terpenoids to fatty acids. The oxidations are most
likely catalyzed by a lipoxygenase and involve the formation of peroxides via radical intermediates in the
first rate-limiting step. We provide herein a rationalization of the observed regioselectivity of these con-
versions by means of computational determination of bond dissociation enthalpies of a set of tailor-made
spirocyclic terpenoids. It was found that only strongly activated allylic positions (BDH298 of <80 kcal/mol)
with neighboring heteroatoms or with activating alkyl groups are oxidized to the corresponding unsat-
urated lactones or enones, respectively. With the synthesis and purification of allylic hydroperoxide
intermediates, we have been able to characterize the putative direct precursors of enones in PSA oxida-
tions. Our results suggest a two-step oxidation mechanism involving hydroperoxide intermediates which
are rapidly converted to the observed enones by an enzymatic reaction.
Received 17 February 2015
Received in revised form 6 July 2015
Accepted 7 July 2015
Available online 23 July 2015
Keywords:
Allylic oxidation
Flavors
Natural products
Terpenoids
Biocatalysis
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
oxidation is the edible fungus Pleurotus sapidus (PSA), which has
a remarkably broad substrate range from terpenoids to fatty acids
Regioselective oxidations of CH-bonds are attractive synthetic
transformations with a broad spectrum of applications in academia
and a high impact on the industrial chemical value chain as they
convert relatively cheap precursors into value-added products
[1,2]. Among these transformations, allylic oxidations are of high
interest, because the olefinic starting materials are readily avail-
able as cheap bulk chemicals and many interesting derivatives,
such as terpenes are available from renewable sources [3,4]. In
addition, the resulting allyl alcohols [5–10] or ␣,␤-unsaturated
carbonyl compounds are attractive synthetic targets of high eco-
nomic and scientific interest [11–17]. Allylic oxidations of olefins
to enones have classically been performed with strong oxidants,
such as chromium or other metal-based reagents [18,19]. In addi-
tion, metal-free and biocatalytic methods have been reported [3].
Several of these biocatalytic protocols have been applied to the syn-
thesis of fine chemicals [20–22], drugs [23] and food ingredients
[24–26]. A particularly interesting biocatalytic system for allylic
[27–31]. We have recently shown that the lyophilisate of PSA is
able to catalyze allylic and benzylic oxidations in a broad range of
olefinic substrates including simple cyclohexene derivatives and
several functionalized terpenoids with preparatively useful yields
[32]. Biocatalytic allylic oxidations with PSA may be performed
with the lyophilisate of commercially available fungal fruiting bod-
ies or with mycelium from submerged cultures. A PSA-derived
dioxygenase has been shown to be responsible for the allylic oxi-
dation of valencene to nootkatone and the same enzyme oxidizes
unsaturated fatty acids [29,31,33–35]. It is thus likely that this
dioxygenase is the major oxidant in other allylic oxidations with
PSA-lyophilisate, too. However, since the lyophilisate is a mixture
of enzymes, alternative oxidation pathways cannot be ruled out for
other substrates. The reaction mechanism was proposed to involve
the initial formation of allyl radicals, which would subsequently
be converted to allylic hydroperoxides 2 (Scheme 1) [24,29]. These
peroxides were found to be the major products of PSA-catalyzed
oxidations of fatty acids and were proposed as intermediates in the
allylic oxidation of valencene to nootkatone [29,31]. The abstrac-
tion of an allylic hydrogen and the resulting formation of radical
intermediates is plausible and is reflected by our recent finding that
∗
Corresponding author. Fax: +49 40 42838 3477.
E-mail address: maison@chemie.uni-hamburg.de (W. Maison).
http://dx.doi.org/10.1016/j.molcatb.2015.07.008
1381-1177/© 2015 Elsevier B.V. All rights reserved.

16
V. Weidmann et al. / Journal of Molecular Catalysis B: Enzymatic 121 (2015) 15–21
(DSMZ 8266), Braunschweig, Germany. Production of biomass and
lyophilisate were described previously by Fraatz et al. [24] For
all new oxidations with PSA, a parallel positive control experi-
ment with the known substrate theaspirane and the same batch of
PSA-lyophilisate was performed, confirming enzymatic activity of
the lyophilisate used. Protein concentrations were determined by
Lowry assay using bovine serum albumin as standard, and peroxi-
dase activity was quantified as previously described in the presence
of 0.2 mM H2O2 [45]:
Scheme 1. Proposed pathways for allylic oxidations of alkenes
lyophilisate of Pleurotus sapidus (PSA).
1 with the
Peroxidase activity: 16.7 U/g lyophilisate (253 mU/mg protein)
Solubilized protein: 66 mg/g lyophilisate
the reactivity of substrates towards PSA oxidation is largely deter-
mined by bond-dissociation energies for the participating allylic
CH-bonds with a threshold of about 80 kcal/mol [36–40]. Interme-
diate allylic peroxides were assumed to be unstable and have not
been thoroughly characterized so far. It is furthermore unclear if
they are converted by “chemical” redox disproportionation or an
enzymatic transformation to the typical reaction products (enones
and/or allylic alcohols). In addition, olefinic terpenoids and unsat-
urated fatty acids are substrates for autoxidations which may be
alternative non-enzymatic reaction pathways.
In this paper, we report our attempts towards a mechanis-
tic understanding of allylic oxidations with the lyophilisate of
PSA. Rationalization of the observed regioselectivity is provided
by computational determination of bond dissociation enthalpies
and correlation with structural and electronic features of selected
tailor-made spirocyclic terpenoids. In addition, we investigate
the role of alternative autoxidations and peroxide intermediates
through synthesis of representative examples and their use as sub-
strates for PSA.
2.2. Experimental procedures and analytical data
Valencene peroxides 29 and 30. A sample of valencene (9)
(204 mg, 1.0 mmol) and 1 mL of a CDCl3 stock solution, of the pho-
tosensitizer meso-tetraphenyl-porphyrine (TPP, 5 × 10−4 m) was
filled into a NMR tube. During the irradiation with a 50 W white LED
lamp at room temperature, oxygen was bubbled through the solu-
tion. The reaction was followed by TLC and NMR spectroscopy. After
complete conversion, the solvent was evaporated under reduced
pressure at 5 ◦C which resulted in a mixture of hydroperoxides
29/30 4:1 (230 mg, 0.97 mmol, 97%) as a green, highly viscous
oil. Column chromatography (pentane/Et2O 100:1.5 → 100:2.5)
gave hydroperoxide 29 (141 mg, 0.6 mmol, 60%) and rearranged
hydroperoxide 30 (36 mg, 0.15 mmol, 15%) as orange oils.
(4R,4aS,6R)-2-Hydroperoxy-4,4a-dimethyl-6-(prop-1-
en-2-yl)-2,3,4,4a,5,6,7,8-octahydronaphthalene29
Rf = 0.35
(pentane/Et2O 8.5:1.5, phosphomolybdic acid). HRMS (ESI): cal-
culated for C15H24O2 + Na+ = 259.1669, found = 259.1676. 1HNMR
(400 MHz, CDCl3): ı (ppm) = 7.86 (s, 1H, OOH), 5.42–5.39 (m,
1H, 1-H), 4.69–4.67 (m, 1H, 13-H), 4.37–4.34 (m, 1H, 2-H), 2.34
2. Experimental
3
(ttt, JH,H = 14.1 Hz, JH,H = 2.5 Hz, JH,H = 2.5 Hz, 1H, 8-H), 2.24 (tt,
3JH,H = 12.5 Hz, JH,H = 3.0 Hz, 1H, 6-H), 2.17 (ddd, JH,H = 14.1 Hz,
3
2
2.1. General methods
3
3JH,H = 4.2 Hz, JH,H = 2.6 Hz, 1H, 8-H), 1.93 (ttt, JH,H = 14.7 Hz,
2
TLC was performed on silica gel aluminum sheets (Macherey
and Nagel). The reagent used for developing TLC plates was phos-
phomolybdic acid (5 g phosphomolybdic acid, 100 mL EtOH). Flash
column chromatography was performed on silica gel (Macherey
and Nagel, 40–60 ␮m).
JH,H = 1.5 Hz, JH,H = 1.5 Hz, 1H, 3-H), 1.87 (td, JH,H = 12.9 Hz,
3JH,H = 2.8 Hz, 1H, 5-H), 1.82–1.76 (m, 1H, 7-H), 1.76–1.72 (m,
1H, 4-H), 1.70 (s, 3H, 12-H), 1.56–1.47 (m, 1H, 3-H), 1.31–1.19
(m, 1H, 7-H), 1.05–0.98 (m, 1H, 5-H), 0.92 (s, 3H, 10-H), 0.90 (d,
3JH,H = 7.0 Hz, 3H, 9-H). 13C NMR (100 MHz, CDCl3): ı (ppm) = 153.5
(C8a), 150.2 (C11), 115.9 (C1), 108.8 (C13), 78.5 (C2), 44.5 (C5),
40.8 (C6), 38.6 (C4a), 35.2 (C4), 32.8 (C8), 32.6 (C7), 30.9 (C3), 20.9
(C12), 17.1 (C10), 15.2 (C9).
1H NMR chemical shifts are calibrated to residual non-
deuterated solvent (CDCl3, ıH = 7.26 ppm). 13C NMR chemical shifts
are referenced to the solvent signal (CDCl3, ı = 77.16 ppm). NMR
spectra were recorded at 300 (75), 400 (100), 500 (125) or 600
(150) MHz on Bruker Avance instruments. The coupling constant
(J) is given in Hz. The chemical shifts ı are reported in ppm and the
signal patterns are indicated as s (singlet), d (doublet), t (triplet), q
(quartet), sext (sextet), m (multiplet), and br. (broad). NMR-signals
have been assigned on the basis of 2D-NMR (HH-COSY, HMBC, and
HSQC) experiments. Relative stereoinformation has been assigned
on the basis of 1D or 2D NOE-experiments.
The atom numbers used for NMR peak assignment do not refer
to IUPAC nomenclature and are available from the structures pro-
vided on the spectra in the supporting information. ESI and APCI
mass spectra were recorded with a Bruker MicroTOF-Q instrument
operating in positive mode. Samples were dissolved in CH3CN H2O
mixtures or pure MeOH and directly injected using a syringe. All of
the reagents were reagent grade and used without further purifi-
cation unless otherwise specified. Solvents for the reactions were
distilled prior to use. All air- or moisture-sensitive reactions were
conducted under nitrogen or argon in flame- or oven-dried glass-
ware and were magnetically stirred.
(4R,4aS,6R)-1-Hydroperoxy-4,4a-dimethyl-6-(prop-1-
en-2-yl)-1,2,3,4,4a,5,6,7-octahydronaphthalene30
Rf = 0.40
(pentane/Et2O 8.5:1.5, phosphomolybdic acid). HRMS (ESI):
calculated for C15H24O2 + Na+ = 259.1669, found = 259.1659. 1H
NMR (400 MHz, CDCl3): ı (ppm) = 7.41 (s, 1H, OOH), 5.76 (dd,
3
3JH,H = 5.2 Hz, JH,H = 2.1 Hz, 1H, 8-H), 4.75–4.72 (m, 13H, 13-
H), 4.35–4.33 (m, 1H, 1-H), 2.39–2.31 (m, 1H, 6-H), 2.20 (dtd,
2JH,H = 17.9 Hz, 3JH,H = 5.2 Hz, 3JH,H = 1.8 Hz, 1H, 7-H), 2.11–2.06 (m,
2
3
3
1H, 2-H), 1.92 (ddd, JH,H = 17.9 Hz, JH,H = 11.5 Hz, JH,H = 2.1 Hz,
2
3
1H, 7-H), 1.79 (td, JH,H = 12.6 Hz, JH,H = 2.1 Hz, 1H, 5-H), 1.75 (s,
3H, 12-H), 1.62-1.57 (m, 2H, 2-H/3-H), 1.34–1.27 (m, 2H, 3-H/4-
3
H), 1.18 (t, JH,H = 12.6 Hz, 1H, 5-H), 1.09 (s, 3H, 10-H), 0.88 (d,
3JH,H = 6.2 Hz, 3H, 9-H). 13C NMR (100 MHz, CDCl3): ı (ppm) = 150.0
(C11), 140.1 (C8a), 130.1 (C8), 108.9 (C13), 87.8 (C1), 43.7/43.6
(C4/C5), 38.1 (C4a), 37.4 (C6), 31.6 (C7), 30.1 (C2), 26.1 (C3), 20.9
(C12), 19.9 (C10), 15.7 (C9).
(2R*,5R*)-7-Hydroperoxy-2,10,10-trimethyl-6-methylene-
1-oxaspiro[4.5]decane 31. A sample of theaspirane (5) (194 mg,
1.0 mmol) and 1 mL of a CDCl3 stock solution, of the photo-
sensitizer meso-tetraphenyl-porphyrine (TPP, 5 × 10−4 m) was
filled into a NMR tube. During the irradiation with a 50 W white
LED lamp at room temperature, oxygen was bubbled through
the solution. The reaction was followed by TLC and NMR spec-
Compounds 20 [41], 24 [42], 26 [42,43], and 28 [44], 32a [40],
were synthesized according to the literature.
The filamentous fungus P. sapidus (PSA) was obtained from
the German Collection of Microorganisms and Cell Cultures

V. Weidmann et al. / Journal of Molecular Catalysis B: Enzymatic 121 (2015) 15–21
17
troscopy. After complete conversion, the solvent was evaporated
under reduced pressure at 5 ◦C which gave hydroperoxide 31
(215 mg, 0.95 mmol, 95%) as a green, highly viscous oil. Rf = 0.35
(pentane/Et2O 8:2, phosphomolybdic acid). HRMS (ESI): cal-
culated for C13H22O3 + Na+ = 249.1461, found = 249.1461. 1H
NMR (400 MHz, CDCl3): ı (ppm) = 7.87 (br. s, 1H, OOH), 5.21
1H, 9-H), 1.68–1.54 (m, 2H, 4-H/9-H), 1.50–1.41 (m, 1H, 3-H), 1.20
(s, 3H, 14-H), 1.03/1.02 (2*s, 2*3H, 12-H/13-H). 13C NMR (100 MHz,
CDCl3): ı (ppm) = 203.4 (C7), 152.3 (C6), 116.9 (C11), 90.4 (C5),
75.3 (C2), 37.0 (C8), 36.8 (C10), 33.8 (C9), 32.6 (C3), 30.6 (C4),
23.7/22.5 (C12/C13), 21.6 (C14).
Biotransformation of valencene hydroperoxide 29. According
to the general procedure hydroperoxide 29 (324 mg, 1.4 mmol) was
treated with 800 mg lyophilisate for 48 h. The crude product was
purified by column chromatography (pentane/Et2O 9:1 → 8:2) to
give the enone 10 (47.5 mg, 0.22 mmol, 16%) and the allylic alcohol
11 (20.4 mg, 0.1 mmol, 7%) as colorless oils. The NMR spectra of 10
and 11 matched those reported in the literature [32].
2
(d, JH,H = 1.7 Hz, 1H, 11-H), 5.13–5.12 (m, 1H, 11-H), 4.62 (t,
3JH,H = 4.8 Hz, 1H, 7-H), 4.10–4.03 (m, 1H, 2-H), 2.02–1.93 (m,
3H, 3-H/4-H/4-H), 1.89–1.81 (m, 1H, 8-H), 1.76–1.69 (m, 1H,
8-H), 1.61–1.54 (m, 1H, 9-H), 1.46–1.37 (m, 2H, 3-H/9-H), 1.22
3
(d, JH,H = 6.1 Hz, 3H, 14-H), 0.90/0.87 (2*s, 2*3H, 12-H/13-H). 13C
NMR (100 MHz, CDCl3): ı (ppm) = 148.1 (C6), 112.0 (C11), 90.0
(C5), 86.4 (C7), 74.4 (C2), 38.0 (C10), 34.3 (C9), 33.8 (C3), 31.2 (C4),
26.3 (C8), 24.1/22.7 (C12/C13), 21.1 (C14).
Biotransformation of spiroether 33. According to the gen-
eral procedure spiroether 33 (50 mg, 0.26 mmol) was treated with
(2R*,5R*)-2,10,10-Trimethyl-6-methylene-1-
600 mg lyophilisate for 96 h. The crude product was analyzed by 1
H
oxaspiro[4.5]decane 33. Allylalcohol 32a (1.00 g, 4.75 mmol)
was converted to the corresponding acetate by treatment with
pyridine (50 mL) and Ac2O (250 mL) at rt for 24 h. The reaction mix-
ture containing 32b was concentrated to dryness under reduced
pressure and the residue was dissolved in dry THF (15 mL). The
solution was treated with formic acid (570 ␮L, 14.25 mmol), Et3N
(2.04 mL, 14.25 mmol), (PPh3)2PdCl2 (70 mg, 0.1 mmol) and PPh3
(162 mg, 0.61 mmol). The reaction was heated to reflux for 24 h
under a nitrogen atmosphere. The mixture was cooled to rt, diluted
with Et2O (20 mL) and washed with water (20 mL), 5% HCl solution
(10 mL) saturated aqueous NaHCO3 solution (15 mL) and brine
(20 mL), dried with Na2SO4 and evaporated. The residue was puri-
fied by column chromatography (pentane/Et2O 100:1.5 → 100:2.5)
to give spiroether 33 (497 mg, 2.56 mmol, 54%) as a colorless oil.
Rf = 0.77 (pentane/Et2O 100:1, phosphomolybdic acid). HRMS
(APCI): calculated for C13H22O + H+ = 195.1743, found = 195.1750.
and 13C NMR and the pure substrate was identified. No conversion
was observed.
Biotransformation of spiroether 20. According to the gen-
eral procedure spiroether 20 (382 mg, 2.5 mmol) was treated with
800 mg lyophilisate for 72 h. The crude product was purified by
column chromatography (pentane/Et2O 10:1 → 7:1) to give lac-
tone 21 (46 mg, 0.28 mmol, 11%), enone 22 (39 mg, 0.24 mmol,
10%) and aldehyde 23 (53 mg, 0.32 mmol, 13%) as colorless oils.
1-Oxaspiro[5.5]undec-3-en-2-one21. Rf = 0.52 (pentane/Et2O 1:1,
phosphomolybdic acid). 1HNMR (400 MHz, CDCl3): ı (ppm) = 6.72
3
3
3
(td, JH,H = 9.8 Hz, JH,H = 4.2 Hz, 1H, 4-H), 5.98 (td, JH,H = 9.8 Hz,
3
4
4JH,H = 2.0 Hz, 1H, 3-H), 2.39 (dd, JH,H = 4.2 Hz, JH,H = 2.0 Hz, 2H,
5-H), 1.98–1.92 (m, 2H, 7-H), 1.78–1.69 (m, 2H, 8-H), 1.58–1.41
(m, 5H, 7-H/8-H/9-H), 1.37-1.29 (m, 1H, 9-H). 13C NMR (100 MHz,
CDCl3): ı (ppm) = 163.9 (C2), 143.2 (C4), 121.1 (C3), 81.3 (C6),
36.4 (C7), 34.6 (C5), 25.4 (C9), 21.6 (C8). 1-Oxaspiro[5.5]undec-2-
en-4-one 22. Rf = 0.65 (pentane/Et2O 1:1, phosphomolybdic acid).
4
1H NMR (400 MHz, CDCl3): ı (ppm) = 4.86 (dd, JH,H = 2.5 Hz,
4
4
3
4JH,H = 1.7 Hz, 1H, 11-H), 4.69 (dd, JH,H = 2.5 Hz, JH,H = 1.7 Hz, 1H,
11-H), 4.08–4.00 (m, 1H, 2-H), 2.31 (td, 2JH,H = 13.4 Hz, 3JH,H = 4.1 Hz,
1H, 7-H), 2.10–1.99 (m, 2H, 4-H/7-H), 1.95–1.87 (m, 1H, 3-H), 1.71
1HNMR (400 MHz, CDCl3): ı (ppm) = 7.23 (d, JH,H = 6.1 Hz, 1H, 2-
H), 5.33 (d, 3JH,H = 6.1 Hz, 1H, 3-H), 2.48 (s, 2H, 5-H), 2.03–1.98 (m,
2H, 7-H), 1.64–1.40 (m, 10H, 7-H/8-H/9-H). 13C NMR (100 MHz,
CDCl3): ı (ppm) = 192.5 (C4), 161.3 (C2), 105.8 (C3), 82.4 (C6), 47.4
(C5), 34.4 (C7), 25.2 (C9), 21.5 (C8). (E)-4-(1-Hydroxycyclohexyl)-
but-2-enal 23. Rf = 0.25 (pentane/Et2O 1:1, phosphomolybdic acid).
2
3
(td, JH,H = 12.7 Hz, JH,H = 8.2 Hz, 1H, 4-H), 1.54–1.47 (m, 2H,
8-H), 1.45–1.40 (m, 2H, 9-H), 1.39–1.31 (m, 1H, 3-H), 1.23 (d,
3JH,H = 6.1 Hz, 3H, 14-H), 0.90/0.88 (2*s, 2*3H, 12-H/13-H). 13C
NMR (100 MHz, CDCl3): ı (ppm) = 151.7 (C6), 105.5 (C11), 90.4
(C5), 74.3 (C2), 38.9 (C9), 38.4 (C10), 34.2 (C3), 33.8 (C7), 31.9 (C4),
24.3 (C12), 23.0 (C8), 22.5 (C13), 20.9 (C14).
3
1H NMR (400 MHz, CDCl3): ı (ppm) = 9.52 (d, JH,H = 7.9 Hz, 1H,
3
3
1-H), 6.98 (td, JH,H = 15.6 Hz, JH,H = 7.5 Hz, 1H, 3-H), 6.15 (tdd,
3
4
3JH,H = 15.6 Hz, JH,H = 7.9 Hz, JH,H = 1.2 Hz, 1H, 2-H), 2.48 (dd,
4
General procedure for the biotransformation: 600 mg PSA
lyophilisate were dissolved in 30 mL Tris–HCl buffer (20 mm, pH
7.5). The dried biomass was rehydrated by stirring at 900 rpm for
10 min. The substrate was added and the solution was stirred at
900 rpm at rt. The reaction progress was controlled by GC-FID.
After 24 h another 400 mg of PSA lyophilisate and 20 mL buffer were
added. After 48 h the reaction was stopped by adding 50 mL of Et2O
and further stirring of the reaction mixture for 30 min. The solution
was filtered and the aqueous phase was extracted with Et2O three
times. The combined organic phases were dried with MgSO4, fil-
tered and concentrated in vacuo. The crude product was analyzed
by NMR and if required purified by column chromatography.
Biotransformation of theaspiran hydroperoxide 31. Accord-
ing to the general procedure hydroperoxide 31 (180 mg, 0.8 mmol)
was treated with 1.3 g lyophilisate for 72 h. The crude product was
purified by column chromatography (pentane/Et2O 9:1 → 8:2) to
give the enone 34 (15.1 mg, 0.07 mmol, 9%) and the allylic alcohol
32a (21.6 mg, 0.1 mmol, 13%) as a colorless oils. The NMR spectra
of 32a matched those reported in the literature [40]. (2R*,5R*)-
2,10,10-Trimethyl-6-methylene-1-oxaspiro[4.5]decan-4-one
34. Rf = 0.35 (pentane/Et2O 8:2, phosphomolybdic acid). HRMS
(ESI): calculated for C13H20O2 + H+ = 209.1536, found = 209.1540.
3JH,H = 7.5 Hz, JH,H = 1.2 Hz, 2H, 4-H), 1.61–1.43 (m, 10H, 2ꢀ-H/3ꢀ-
H/4ꢀ-H). 13C NMR (100 MHz, CDCl3): ı (ppm) = 194.0 (C1), 154.4
(C3), 135.7 (C2), 71.7 (C1ꢀ), 45.6 (C4), 37.9 (C2ꢀ), 25.6 (C4ꢀ), 22.2
(C3ꢀ).
Biotransformation of spiroether 24. According to the gen-
eral procedure spiroether 24 (400 mg, 2.2 mmol) was treated
with 800 mg lyophilisate for 48 h. The crude product was
purified by column chromatography (pentane/Et2O 1:1) to
give lactone 25 (135 mg, 0.7 mmol, 32%) as
a colorless oil.
4,7-Dimethyl-1-oxaspiro[5.5]undec-3,7-dien-2-one25. Rf = 0.20
(pentane/Et2O 2:1, phosphomolybdic acid). HRMS (ESI): calcu-
lated for C12H16O2 + Na+ = 215.1043, found = 215.1048. 1HNMR
(600 MHz, CDCl3): ı (ppm) = 5.82 (s, 1H, 3-H), 5.64–5.62 (m, 1H,
2
2
8-H), 2.71 (d, JH,H = 18.3 Hz, 1H, 5-H), 2.17 (d, JH,H = 18.3 Hz, 1H,
5-H), 2.10–2.04 (m, 1H, 11-H), 1.98–1.92 (m, 3H, 10-H/10-H/11-H),
1.96 (s, 3H, 12-H), 1.81–1.76 (m, 1H, 9-H), 1.75–1.74 (m, 3H, 13-H),
1.55–1.48 (m, 1H, 9-H). 13C NMR (150 MHz, CDCl3): ı (ppm) = 164.8
(C2), 155.4 (C4), 134.1 (C7), 128.4 (C8), 116.0 (C3), 81.6 (C6), 36.6
(C5), 33.9 (10), 25.3 (C11), 23.4 (C12), 19.6 (C9), 18.5 (C13).
Biotransformation of spiroether 26. According to the gen-
eral procedure spiroether 26 (80 mg, 0.45 mmol) was treated
with 1.2 g lyophilisate for 96 h. The crude product was puri-
fied by column chromatography (pentane/Et2O 2:1 → 1:1) to
2
1HNMR (400 MHz, CDCl3): ı (ppm) = 5.71 (d, JH,H = 1.8 Hz, 1H,
2
11-H), 5.29 (d, JH,H = 1.8 Hz, 1H, 11-H), 4.13–4.20 (m, 1H, 2-H),
give enone 27 (10 mg, 0.06 mmol, 14%) as a colorless oil.
2.52–2.38 (m, 2H, 8-H), 2.06–1.95 (m, 2H, 3-H/4-H), 1.78–1.71 (m,
4,7-Dimethyl-1-oxaspiro[5.5]undec-4,7-dien-9-one27. Rf = 0.10

18
V. Weidmann et al. / Journal of Molecular Catalysis B: Enzymatic 121 (2015) 15–21
Scheme 2. Biocatalytic oxidation of theaspirane (5), valencene (9) and linoleic acid
(12). 12, [31,32,40].
(pentane/Et2O 5:1, phosphomolybdic acid). HRMS (APCI): cal-
culated for C12H16O2 + H+ = 193.1216, found = 193.1197. 1HNMR
4
Scheme3. Preparationhydroperoxides29–31 fromvalenceneand theaspiranewith
meso-tetraphenyl-porphyrine (TPP) as photosensitizer and synthesis of an alkene
precursor 33 for biotransformation with PSA.
(400 MHz, CDCl3): ı (ppm) = 5.84 (q, JH,H = 1.3 Hz, 1H, 8-H), 5.34
2
3
3
(s, 1H, 5-H), 3.92 (ddd, JH,H = 11.4 Hz, JH,H = 5.9 Hz, JH,H = 1.7 Hz,
1H, 2-H), 3.77 (ddd, 2JH,H = 11.4 Hz, 3JH,H = 11.0 Hz, 3JH,H = 3.4 Hz, 1H,
2-H), 2.58 (ddd, 2JH,H = 17.2 Hz, 3JH,H = 7.0 Hz, 3JH,H = 5.2 Hz, 1H, 10-
H), 2.38 (ddd, 2JH,H = 17.2 Hz, 3JH,H = 8.8 Hz, 3JH,H = 5.9 Hz, 1H, 10-H),
2.32–2.26 (m, 1H, 3-H), 2.16–2.13 (m, 2H, 11-H), 1.86 (s, 3H, 13-
H), 1.82–1.78 (m, 1H, 3-H), 1.77 (s, 3H, 12-H). 13C NMR (100 MHz,
CDCl3): ı (ppm) = 199.0 (C9), 164.3 (C7), 135.7 (C4), 127.6 (C8),
123.0 (C5), 74.1 (C6), 60.1 (C2), 34.2 (C10), 33.1 (C11), 29.8 (C3),
23.6 (C12), 19.3 (C13).
The results of all biotransformations are summarized in Table 1
together with selected bond dissociation energies which have been
calculated by DFT under implicit consideration of water as solvent
(see Supplementary information for detailed information).
The experimental results from Table 1 confirm our earlier find-
ing of a limiting bond dissociation enthalpy of about 80 kcal/mol
for successful conversion of alkenes with PSA [40]. In consequence,
only strongly activated allylic positions with neighboring het-
eroatoms (entries 1, 3, 5, and 6) or with activating alkyl groups
(entry 7) are oxidized to the corresponding unsaturated lactones
or enones, respectively. Allyl ethers like 14, 17, 20, 24, and 26 are
all good substrates for PSA oxidations and in case of the methyl
substituted derivatives 14, 17, and 24 (entries 1, 3, and 6) a single
lactone isomer is the major product of oxidation, which might be a
consequence of steric hindrance upon the attack of oxygen towards
the intermediate allyl radical or a rearrangement [47,48] of the ter-
tiary alkyl peroxides formed after the attack of oxygen. In contrast,
the unsubstituted allyl ether 20 gave a mixture of three oxidation
products 21–23. Position 9 of spiroethers 24 and 26 seems to be
right at the borderline for reactivity with PSA. In spiroether 24 only
the more reactive position 2 is attacked and lactone 25 is formed
(entry 6). In spiroether 26 position 9 with a slightly lower BDH is
oxidized to the enone 27 (entry 7) although conversion is relatively
low in this case.
In previous studies it was suggested that a lipoxygenase
is responsible for allylic oxidations by PSA [29]. The reaction
mechanism should therefore include allylic hydroperoxides as
intermediates. In fact, hydroperoxides have been found to be the
major products of allylic oxidation of linoleic acid using a recom-
binant lipoxygenase from PSA [29,31]. For other substrates such
as valencene, hydroperoxide intermediates have been found and
analyzed by chromatography and MS [29]. However, a plausible
explanation has not been provided for their spontaneous conver-
sion to enones, which are the major products obtained in this
case. Therefore we synthesized allylic hydroperoxides derived from
valencene and theaspirane via photooxygenation using singlet oxy-
gen (1O2) as reactive species (Scheme 3) [49]. For valencene (9),
this led to a mixture of two isomeric hydroperoxides 29 and 30.
Biotransformation of spiroether 28. According to the general
procedure spiroether 28 (200 mg, 1.33 mmol) was treated with
800 mg lyophilisate for 96 h. The crude product was analyzed by
1H and 13C NMR and the pure substrate was identified.
3. Results and discussion
We have previously shown that allylic spiroethers, such as thea-
spirane (5) (Scheme 2) are suitable substrates for allylic oxidations
with PSA. The conversion of theaspirane (5) according to Scheme 2
was used as a positive control experiment for each new batch of PSA
lyophilisate used in the following conversions to confirm activity of
the biocatalyst. In this context, it is of note that attempts to convert
5 with thermally inactivated lyophilisate failed and no conversion
was observed for 72 h. The same is true for attemptedconversionsof
5 with active lyophilisate under exclusion of oxygen or the addition
of excess H2O2. A table with a summary of these experiments may
be found in the supporting information. The results suggest that (1)
a non-enzymatic autocatalytic formation of oxidation products like
6–8 is slow under the conditions employed, (2) oxygen is a neces-
sary cooxidant as expected for a lipoxygenase catalyzed conversion
and (3) addition of hydrogen peroxide is not favoring the oxidation
of 5.
To evaluate the first step of PSA oxidations, we have synthesized
a number of unsaturated spiroethers (Table 1) by the intramolecu-
lar silyl-modified Sakurai reaction [46]. These spiroethers contain
allylic positions with only slightly differing stereoelectronic prop-
erties and are thus useful model compounds to investigate the first
step of biocatalytic oxidations with PSA, which is the abstraction of
allylic hydrogen and the formation of allyl radicals. All spiroethers
14, 16, 17, 19, 20, 24, 26, and 28 were treated with PSA lyophylisate
according to the protocol for theaspirane (5) depicted in Scheme 2.

V. Weidmann et al. / Journal of Molecular Catalysis B: Enzymatic 121 (2015) 15–21
19
Table 1
Yields and conversions for biotransformations of spirocyclic model compounds 14, 16, 17, 19, 20, 24, 26 and 28 with PSA lyophylisate.
Entry
Starting material
BDH298 (kcal/mol)a
Convb
Productsc
1 [40]
2 [40]
75.7 (2-H)
83.4 (3-H)
83.1 (5-H)
77%
<5%
–
O
O
, 32% yield
18
3 [40]
4 [40]
73.5 (2-H)82.9 (5-H)80.7 (9-H)
>95%
<5%
83.5 (3-H)
80.3 (9-H)
–
2
O
O
O
, 11%
21
O
4
H
OH
, 13%
O
22, 10%
23
5
6
74.9 (2-H)
80.7 (5-H)
>95%
>95%
74.5 (2-H)80.2 (5-H)80.7 (9-H)
7
83.8 (3-H)
80.2 (2-H)
79.1 (9-H)
80.2 (5-H)
50%
<5%
8
–
a
B3LYP/6-31 + G** SCRF = (PCM, solvent = H2O).
Reactions were performed with 0.6–1.2 g PSA lyophilisate in TRIS-buffer at room temperature for 48–96 h. Conversion was determined after work up by GC-FID.
Isolated yields are given after column chromatography.
b
c
Hydroperoxide 29 is the rearrangement product of one of the initial
formed singlet oxygen ene products whereas the second hydroper-
oxide 30 did not rearrange [48]. These compounds are remarkably
stable and were separated by standard column chromatography.
Peroxide 29 is the putative direct precursor of nootkatone (10) upon
biocatalytic oxidation of valencene (9) and thus a particularly inter-
esting compound for the study of PSA oxidations. On the other hand
the photochemical oxidation of theaspirane (5) gave hydroperoxide
31, which is not a direct precursor of theaspiranone (6) and cannot
be formed easily by biocatalytic oxidation of theaspirane (5). It was
thus interesting to check whether 31 would be converted to the cor-
responding enone by PSA. This prompted us to synthesize alkene
33 as a reference substrate for oxidation with PSA according to a
procedure reported by Serra et al. starting from allylic alcohol 32a
via elimination of the intermediate allylic acetate (Scheme 3) [50].
With analytically pure peroxides 29 and 30 in hand we were
able to check the presence of peroxide intermediates in PSA oxida-
tions of valencene. For this purpose, an oxidation of valencene with
PSA lyophilisate was extracted with Et2O after 50% conversion. The
resulting extract was evaporated in vacuo and analyzed by NMR. A
part of the corresponding crude 1H NMR is shown in Fig. 1 along
with the spectra of pure valencene (9), nootkatone (10) and both
peroxides 29 and 30. The methine protons of peroxides 29 and 30
at 4.3–4.4 ppm are well separated from other signals and may thus
be used as markers. In the crude spectrum of a valencene oxidation
(trace A), neither of the peroxides 29 and 30 was observed.
The absence of peroxides in the biocatalytic oxidation of
valencene (9) suggested two possible routes for the generation of
nootkatone (10). The first route involves a rapid enzymatic transfor-
mation of intermediate peroxides, most likely by the same enzyme
responsible for the formation of the peroxides. This combined
lipoxygenase/peroxidase functionality has been described for other
lipoxygenases before [34,51]. The second route involves a rapid
chemical transformation of intermediate peroxides most likely by
redox disproportionation. Given the reasonable stability of perox-
ides 29 and 30, a rapid chemical conversion to nootkatone (10)
seemed unlikely to us. However, it cannot be ruled out due to pos-
sible (non-enzymatic) catalytic effects of metal ions present in the

20
V. Weidmann et al. / Journal of Molecular Catalysis B: Enzymatic 121 (2015) 15–21
Biotransformation of valencene 9 after 50% conversion
(A)
Valencene 9
(B)
Nootkatone 10
(C)
Valenceneperoxide 29
(D)
Valenceneperoxide 30
(E)
5.9
5.8
5.7
5.6
5.5
5.4
5.3
5.2
5.1
5.0
4.9
4.8
4.7
4.6
4.5
4.4
4.3
4.2
4.1
ppm
Fig. 1. Stretches of the 1H NMR spectra of the biotransformation of valencene stopped after 50% conversion (A) with reference NMR spectra of valencene (9) (B), nootkatone
(10) (C) and valencene peroxides 29 (D) and 30 (E).
selectively converted to the corresponding enone. This finding
raised again the question for a possible non-enzymatic conversion
of peroxides like 29 and 31. However, neither reaction of 31 in TRIS
buffer without PSA lyophilisate nor with thermally deactivated PSA
lyophilisate furnished any of the target enone 34 and peroxide 32a
was left unchanged (conversion < 5%). It is worth noting the effect
of oxygen on this reaction. If the conversion of 31 with PSA was
performed under an argon atmosphere, low conversion of 31 (35%)
and the formation of alcohol 32a as single product was observed.
4. Conclusions
Biocatalytic oxidations with PSA show a remarkable broad sub-
strate scope and a number of isoprenoids and fatty acids have been
successfully converted to enones and alcohols as major products.
The oxidations are most likely catalyzed by a lipoxygenase and
involve the formation of peroxides via radical intermediates in the
first rate-limiting step. We provide herein a rationalization of the
observed regioselectivity of these conversions by means of compu-
tational determination of bond dissociation enthalpies of a set of
tailor-made spirocyclic terpenoids 14, 16, 17, 19, 20, 24, 26, and 28.
It was found that only strongly activated allylic positions (BDH298
of <80 kcal/mol) with neighboring heteroatoms or with activating
alkyl groups are oxidized to the corresponding unsaturated lac-
tones or enones respectively.
Hydroperoxide intermediates have been proposed for PSA oxi-
dations in previous studies, but they have never been characterized
nor has a plausible explanation for their spontaneous conversion
to enones been provided. With the synthesis and purification of
Scheme 4. Evaluation of peroxides 29, 31, and alkene 33 as substrates for PSA
allylic hydroperoxide 29 derived from valencene we have been
able to characterize the putative direct precursor of nootkatone
in PSA oxidations of valencene. We have not been able to observe
allylic peroxide 29 during the biocatalytic oxidation of valencene.
However, using pure peroxide 29 as a substrate, we found that
it is converted to nootkatone by an enzymatic reaction. The non-
enzymatic redox disproportionation of 29 was found to be slow
under the conditions employed.
In summary our results suggests a two-step oxidation mech-
anism involving hydroperoxide intermediates which are rapidly
converted to the observed enones. The initial radical reaction deter-
mines the regioselectivity of the conversions. Both steps of the
oxidation.
reaction mixture. Peroxides 29 and 31 were therefore treated with
PSA lyophilisate according to our standard protocol (Scheme 4).
As expected, valencene-derived peroxide 29 was rapidly converted
to nootkatone (10) as a major product along with stereoisomeric
nootkatols. Interestingly, theaspirane-derived peroxide 31 was also
converted to the corresponding enone 34 and alcohol 32a. This con-
version is remarkable, because peroxide 31 cannot be formed from
alkene 33 with PSA (BDH7-H = 81.7 kcal/mol at B3LYP/6-31 + G**
SCRF = PCM). Once a peroxide is formed it is thus obviously non-

V. Weidmann et al. / Journal of Molecular Catalysis B: Enzymatic 121 (2015) 15–21
21
oxidation may be catalyzed by the same enzyme with lipoxygenase
and peroxidase activity.
[24] M.A. Fraatz, S.J.L. Riemer, R. Stöber, R. Kaspera, M. Nimtz, R.G. Berger, H. Zorn,
J. Mol. Catal. B: Enzyme 61 (2009) 202–207.
[25] L. Janssens, H.L. Depooter, N.M. Schamp, E.J. Vandamme, Process Biochem. 27
(1992) 195–215.
Appendix A. Supplementary data
[26] O. Kirk, T.V. Borchert, C.C. Fuglsang, Curr. Opin. Biotechnol. 13 (2002)
345–351.
[27] J. Onken, R.G. Berger, J. Biotechnol. 69 (1999) 163–168.
[28] U. Krings, N. Lehnert, M.A. Fraatz, B. Hardebusch, H. Zorn, R.G. Berger, J. Agric.
Food Chem. 57 (2009) 9944–9950.
[29] S. Krügener, U. Krings, H. Zorn, R.G. Berger, Bioresour. Technol. 101 (2010)
457–462.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.molcatb.2015.07.
008
[30] K. Zelena, U. Krings, R.G. Berger, Bioresour. Technol. 108 (2012) 231–239.
[31] I. Plagemann, K. Zelena, P. Arendt, P.D. Ringel, U. Krings, R.G. Berger, J. Mol.
Catal. B: Enzyme 87 (2013) 99–104.
References
[32] A. Rickert, V. Krombach, O. Hamers, H. Zorn, W. Maison, Green Chem. 14
(2012) 639–644.
[33] C.H. Clapp, S.E. Senchak, T.J. Stover, T.C. Potter, P.M. Findeis, M.J. Novak, J. Am.
Chem. Soc. 123 (2001) 747–748.
[34] C.H. Clapp, M. Strulson, P.C. Rodriguez, R. Lo, M.J. Novak, Biochemistry 45
(2006) 15884–15892.
[1] T. Newhouse, P.S. Baran, Angew. Chem. Int. Ed. 50 (2011) 3362–3374.
[2] E. Roduner, W. Kaim, B. Sarkar, V.B. Urlacher, J. Pleiss, R. Gläser, W.-D. Einicke,
G.A. Sprenger, U. Beifuß, E. Klemm, C. Liebner, H. Hieronymus, S.-F. Hsu, B.
Plietker, S. Laschat, ChemCatChem 5 (2013) 82–112.
[3] V. Weidmann, W. Maison, Synthesis 45 (2013) 2201–2221.
[4] P.C.B. Page, T.J. McCarthy, 2.1–Oxidation adjacent to CC bonds, in: B.M. Trost,
I. Fleming (Eds.), Comprehensive Organic Synthesis, Pergamon, Oxford, 1991,
pp. 83–117.
[5] A. Nakamura, M. Nakada, Synthesis 45 (2013) 1421–1451.
[6] M.S. Kharasch, G. Sosnovsky, J. Am. Chem. Soc. 80 (1958) 756.
[7] M.B. Andrus, J.C. Lashley, Tetrahedron 58 (2002) 845–866.
[8] M.T. Rispens, C. Zondervan, B.L. Feringa, Tetrahedron: Asymetry 6 (1995)
661–664.
[9] I.I. Moiseev, M.N. Vargaftik, Coord. Chem. Rev. 248 (2004) 2381–2391.
[10] S.D.R. Christie, A.D. Warrington, Synthesis (2008) 1325–1341.
[11] T. Joseph, D.P. Sawant, C.S. Gopinath, S.B. Halligudi, J. Mol. Catal. A: Chem. 184
(2002) 289–299.
[12] A.B. Smith, J.P. Konopelski, J. Org. Chem. 49 (1984) 4094–4095.
[13] R. Berger, Biotechnol. Lett. 31 (2009) 1651–1659.
[14] J.R. Hanson, Nat. Prod. Rep. 24 (2007) 1342–1349.
[15] M. Fraatz, R. Berger, H. Zorn, Appl. Microbiol. Biotechnol. 83 (2009) 35–41.
[16] J.A.R. Salvador, S.M. Silvestre, V.M. Moreira, Curr. Org. Chem. 16 (2012)
1243–1276.
[17] J.A.R. Salvador, S.M. Silvestre, V.M. Moreira, Curr. Org. Chem. 10 (2006)
2227–2257.
[18] J. Muzart, Mini-Rev. Org. Chem. 6 (2009) 9–20.
[35] E.H. Oliw, Prostaglandins Other Lipid Mediators 68–69 (2002) 313–323.
[36] K.B. Wiberg, S.D. Nielsen, J. Org. Chem. 29 (1964) 3353–3361.
[37] W.G. Dauben, M.E. Lorber, D.S. Fullerton, J. Org. Chem. 34 (1969) 3587–3592.
[38] G. Rothenberg, Y. Sasson, Tetrahedron 54 (1998) 5417–5422.
[39] J.A. Howard, K.U. Ingold, Can. J. Chem. 45 (1967) 793–802.
[40] V. Weidmann, M. Schaffrath, H. Zorn, J. Rehbein, W. Maison, Beilstein J. Org.
Chem. 9 (2013) 2233–2241.
[41] M.E. Maier, M. Bugl, Synlett (1998) 1390–1392.
[42] V. Weidmann, J. Ploog, S. Kliewer, M. Schaffrath, W. Maison, J. Org. Chem. 79
(2014) 10123–10131.
[43] M. Schaffrath, V. Weidmann, W. Maison, J. Chromatogr. A 1363 (2014)
270–277.
[44] N.R. Natale, R.O. Hutchins, Org. Prep. Proced. Int. 9 (1977) 103–108.
[45] D. Linke, H. Zorn, B. Gerken, H. Parlar, R.G. Berger, Enzyme Microb. Technol. 40
(2007) 273–277.
[46] I.E. Marko, A. Mekhalfia, D.J. Bayston, H. Adams, J. Org. Chem. 57 (1992)
2211–2213.
[47] H.S. Dang, A.G. Davies, I.G.E. Davison, C.H. Schiesser, J. Org. Chem. 55 (1990)
1432–1438.
[48] A.G. Davies, I.G.E. Davison, J. Chem. Soc. Perkin Trans. 2: Phys. Org. Chem.
(1989) 825–830.
[19] J. Muzart, Chem. Rev. 92 (1992) 113–140.
[49] A. Eske, B. Goldfuss, A.G. Griesbeck, A. de Kiff, M. Kleczka, M. Leven, J.M.
Neudorfl, M. Vollmer, J. Org. Chem. 79 (2014) 1818–1829.
[50] S. Serra, C. Fuganti, E. Brenna, Helv. Chim. Acta 89 (2006) 1110–1122.
[51] E.H. Oliw, F. Jerneren, I. Hoffmann, M. Sahlin, U. Garscha, Biochimica Et
Biophysica Acta, Mol. Cell Biol. Lipids 1811 (2011) 138–147.
[20] A. Celik, S.L. Flitsch, N.J. Turner, Org. Biomol. Chem. 3 (2005) 2930–2934.
[21] G. Bellucci, C. Chiappe, L. Pucci, P.G. Gervasi, Chem. Res. Toxicol. 9 (1996)
871–874.
[22] A.J.J. Straathof, S. Panke, A. Schmid, Curr. Opin. Biotech. 13 (2002) 548–556.
[23] J.H. Tao, J.H. Xu, Curr. Opin. Chem. Biol. 13 (2009) 43–50.