Selective aliphatic carbon-hydrogen bond activation of protected alcohol substrates by cytochrome P450 enzymes

Article abstract of DOI:10.1039/c3ob42417k

Protected cyclohexanol and cyclohex-2-enol substrates, containing benzyl ether and benzoate ester moieties, were designed to fit into the active site of the Tyr96Ala mutant of cytochrome P450cam. The protected cyclohexanol substrates were efficiently and selectively hydroxylated by the mutant enzyme at the trans C-H bond of C-4 on the cyclohexyl ring. The selectivity of oxidation of the benzoate ester protected cyclohexanol could be altered by making alternative amino acid substitutions in the P450cam active site. The addition of the double bond in the cyclohexyl ring of the benzoate ester protected cyclohex-2-enol has a debilitative effect on the activity of the Tyr96Ala mutant with this substrate. However, the Phe87Ala/Tyr96Phe double mutant, which introduces space at a different location in the active site than the Tyr96Ala mutant, was able to efficiently hydroxylate the C-H bonds of 1-cyclohex-2-enyl benzoate at the allylic C-4 position. Mutations at Phe87 improved the selectivity of the oxidation of 1-phenyl-1-cyclohexylethylene to trans-4-phenyl-ethenylcyclohexanol (92%) when compared to single mutants at Tyr96 of P450cam. the Partner Organisations 2014.

Full text of DOI:10.1039/c3ob42417k

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Selective aliphatic carbon–hydrogen bond
activation of protected alcohol substrates by
cytochrome P450 enzymes†
Cite this: Org. Biomol. Chem., 2014,
12, 2479
Stephen G. Bell,*a,b Justin T. J. Spence,a Shenglan Liu,a Jonathan H. Georgea and
Luet-Lok Wongb
Protected cyclohexanol and cyclohex-2-enol substrates, containing benzyl ether and benzoate ester
moieties, were designed to fit into the active site of the Tyr96Ala mutant of cytochrome P450cam. The
protected cyclohexanol substrates were efficiently and selectively hydroxylated by the mutant enzyme at
the trans C–H bond of C-4 on the cyclohexyl ring. The selectivity of oxidation of the benzoate ester pro-
tected cyclohexanol could be altered by making alternative amino acid substitutions in the P450cam
active site. The addition of the double bond in the cyclohexyl ring of the benzoate ester protected cyclo-
hex-2-enol has a debilitative effect on the activity of the Tyr96Ala mutant with this substrate. However,
the Phe87Ala/Tyr96Phe double mutant, which introduces space at a different location in the active site
than the Tyr96Ala mutant, was able to efficiently hydroxylate the C–H bonds of 1-cyclohex-2-enyl ben-
zoate at the allylic C-4 position. Mutations at Phe87 improved the selectivity of the oxidation of 1-phenyl-
1-cyclohexylethylene to trans-4-phenyl-ethenylcyclohexanol (92%) when compared to single mutants at
Tyr96 of P450cam.
Received 3rd December 2013,
Accepted 25th February 2014
DOI: 10.1039/c3ob42417k
www.rsc.org/obc
form (1R)-5-exo-hydroxycamphor, wild-type (WT) P450cam oxi-
Introduction
dises related molecules including norcamphor,15–17 thio-
Cytochrome P450 (CYP) enzymes are heme-dependent mono- camphor, camphene,18 adamantane, adamantanone,19 5,5-
oxygenases that catalyse the insertion of one oxygen atom difluorocamphor,20 5,6-dehydrocamphor,21 and 5-methylene-
from atmospheric dioxygen into carbon–hydrogen bonds in a camphor.22 All of these analogues show lower spin state shifts,
diverse range of organic compounds.1,2 They have many phys- binding affinities, NADH oxidation rates and coupling efficien-
iological roles including hormone and secondary metabolite cies than (1R)-camphor with the WT enzyme. Helms and co-
biosynthesis, as well as xenobiotic metabolism.3,4 The selective workers designed substrate analogues to fill extra space in the
oxidation of non-functionalised hydrocarbons using oxygen is P450cam substrate binding pocket.23 The camphor carbonyl
of great interest for applications in organic synthesis and the was replaced by an ether oxygen linked to a hydrocarbon chain
utilisation of CYP enzymes as biocatalysts is an active field of of varying length. The best of these analogues was found to
research.5–9 Recent work has reported how P450 enzymes can bind with a higher affinity than camphor and with only a
be optimised in order to catalyse synthetically relevant regio-, slightly lower spin state shift. They also showed increased
stereo- and even enantioselective oxidations.10–12
binding to the Tyr96Phe (Y96F) mutant over the WT enzyme.
One of the best characterised cytochrome P450 monooxy- Though no turnover data has been reported this work shows
genases is P450cam (CYP101A1) from Pseudomonas putida. that substrates can be designed to fit the active site.
P450cam catalyses the hydroxylation of camphor to 5-exo-
The tyrosine 96 residue of P450cam forms a hydrogen bond
hydroxycamphor, with 100% stereoselectivity.13,14 In addition with the carbonyl oxygen of camphor and the substrate range
to hydroxylating its physiological substrate, (1R)-camphor to of the enzyme can be broadened by mutant variants at this
position. The Tyr96Ala (Y96A) mutant can accept diphenyl-
methane, benzylcyclohexane, phenylcyclohexane and 1-phenyl-
1-cyclohexylethylene as substrates and oxidise them.24–30 The
mixed cyclohexyl and phenyl ring substrates above were solely
oxidised on the cyclohexyl ring. The substrate binding and
monooxygenase activity of P450cam for different non-natural
substrates can be optimised by introducing further mutations
aSchool of Chemistry and Physics, University of Adelaide, SA 5005, Australia.
E-mail: stephen.bell@adelaide.edu.au
bDepartment of Chemistry, University of Oxford, Inorganic Chemistry Laboratory,
South Parks Road, Oxford, OX1 3QR, UK
†Electronic supplementary information (ESI) available. See DOI:
10.1039/c3ob42417k
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 2479–2488 | 2479

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 1 Protected alcohol substrates and the methods of protection and
deprotection.
around the active site, examples of substrates which have
been successfully and selectively oxidised include terpenes,
gaseous alkanes, polychlorinated benzenes and polyaromatic
hydrocarbons.31–37
Fig. 2 The active site of camphor-bound WT P450cam.
Alternatively, the substrate can be engineered using a pro-
tecting group. This approach first introduces a protecting
group before the biotransformation is performed. This group
can then be removed to generate the compound of interest.
This strategy has been developed to improve activity and
selectivity, prevent undesired side reactions and therefore
improve the performance of biocatalysts in chemical
synthesis.38–41
Here we report the oxidation of substrates with protected
alcohol functionalities.42,43 The protecting groups used were
benzyl ether and benzoate esters, both of which can be easily
added to and removed from alcohol functionalities (Fig. 1).
These substrates, which contain a cyclohexyl and a phenyl
ring, were designed to fill the active site of the Y96A mutant of
P450cam. It is hoped that by protecting the alcohol functional-
ity, with a group containing a phenyl ring, that the substrates
would bind in the active site mutants of P450cam with the
aliphatic ring positioned close to the ferryl oxygen for oxi-
dation. This could result in selective oxidation on this ring
and offer a route to diols of cyclohexane and cyclohexene
which are versatile synthetic intermediates.44,45
but smaller active site than the Y96A variant, and the
Phe87Leu/Tyr96Phe (F87L/Y96F) and Phe87Ala/Tyr96Phe
(F87A/Y96F) double mutants which would generate additional
space in the P450cam active site but at a different location
(Fig. 2). The wild type P450cam (WT, C334A variant) was also
used as a reference.48 Cyclohexyl benzoate (2) induced larger
spin state shifts with the WT enzyme than cyclohexyloxymethyl
benzene (1) and small shifts were observed with the Y96F
mutant. For all three substrates the Y96A or the F87A/Y96F
mutants underwent the largest shift to the high-spin form on
substrate binding. 1-Cyclohex-2-enyl benzoate (3) induced
smaller shifts than either 1 or 2 with the mutant enzymes. The
binding of the benzyl ether (1) to the F87L/Y96F mutant
resulted in a significantly greater spin state shift to the high-
spin form than the binding of either benzoate ester (2 and 3).
The NADH oxidation rates of the substrate/mutant combi-
nations were determined under standard conditions and
product formation was analysed qualitatively by gas chromato-
graphy (GC, Fig. S1 and S2†). Even though all of the substrates
(1–3) induced a spin state shift with the WT enzyme the NADH
oxidation activities were low and the levels of product for-
mation were not significant, indicating that substrate binding
to the WT enzyme may be weak. All of the mutants showed
Results
Substrate binding and turnover assays
Cyclohexyloxymethyl benzene (1), cyclohexyl benzoate (2) and higher activity for the oxidation of the protected cyclohexanol
1-cyclohex-2-enyl benzoate (3) were synthesised and purified substrates (1 and 2). The Y96F mutant was more active than
using standard methods.46,47 To monitor the substrate the WT enzyme but the Y96A and F87A/Y96F mutants gave the
binding the spin state shifts of the Y96A P450cam mutant with highest NADH oxidation rates and the best substrate conver-
the protected alcohols were determined. In addition we also sions (Fig. S1†). For the best mutants the NADH oxidation
tested the Tyr96Phe (Y96F) mutant, which has a hydrophobic rates were markedly higher for cyclohexyl benzoate (2) than for
2480 | Org. Biomol. Chem., 2014, 12, 2479–2488
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 1 Substrate spin state shifts (% HS, 5%) and NADH oxidation
rates (N, nmol per nmol P450 per min) of the P450cam mutants and the
protected alcohol substrates cyclohexylmethyl benzene (1), cyclohexyl
benzoate (2) and 1-cyclohex-2-enyl benzoate (3). Cyclohexanol and
cyclohex-2-enol showed negligible (<5%) spin state shifts with the WT
and Y96F enzymes and none with the Y96A and the F87A/Y96F mutants.
(—) Turnover rate is too low to be reliably calculated above the leak rate
and (n.d.) not determined due to low rate or product formation levels
Table 2 Product distributions for protected alcohols 1 and 2 with
P450cam mutants. The products 4 and 6 have a shorter retention time
in the normal phase HPLC analysis compared to the products 5 and 7,
respectively (Fig. S3)
1
2
Product 4
Product 5
Product 6
Product 7
P450cam enzyme RT 16 min RT 18 min RT 16 min RT 18 min
1
2
3
WT
Y96A
Y96F
F87L/Y96F
F87A/Y96F
V247A
None
<5%
60%
9%a
None
>95%
40%
70%
<5%
35%
10%
5%
85%
85%
30%
>95%
65%
90%
95%
15%
15%
P450cam
% HS
N
% HS
N
% HS
N
82%a
82%a
>95%
>95%
WT
Y96A
Y96F
F87L/Y96F
F87A/Y96F
V247A
30%
90%
25%
90%
≥95%
15%
30%
—
45
60
35
n.d.
—
n.d.
55%
≥95%
40%
50%
90%
90%
60%
—
280
50
205
380
n.d.
180
30%
70%
5%
20%
70%
45%
25%
—
50
—
20
180
n.d.
60
11%a
<5%
<5%
Y96F/V247A
a For substrate 1 there were small amounts of additional products
detected in the turnover (total 7% for F87A/Y96F and 9% for F87L/
Y96F). The retention times given are those for the normal phase HPLC
analysis (Fig. S3).
Y96F/V247A
cyclohexyloxymethyl benzene (1). The NADH oxidation activity
and levels of product formation of cyclohexanol and cyclohex-
2-en-1-ol were negligible for the WT and P450cam mutant
enzymes. Taken together these results indicate that the pro-
tecting groups are important in generating efficient bio-
catalysts when using this P450 system.
more of one product (4 and 7, Retention time (RT) 16 and
18 min) being formed over the other (5 and 6, RT 18 and
16 min, Table 2). For substrate 1 the F87L/Y96F and F87A/Y96F
mutant changed the selectivity to favour one product (5) over
the other (82%) and resulted in the low level formation of two
additional minor products (total 7–9%). The Y96A mutant
resulted in the almost exclusive formation of product 5 (>95%).
For substrate 2, both the Y96A and F87A/Y96F mutants
enhanced the selectivity of the product (7) somewhat favoured
by the Y96F mutant resulting in one majority product (>95% for
Y96A). Even though the WT enzyme resulted in very low overall
product yields it shifted the selectivity of product formation
with substrate 2 to favour formation of the minor product (6)
generated by the other mutants (Table 2 and Fig. S1†).
To further explore the altered selectivity of the WT enzyme
with 2 and the Y96F mutant with 1 we generated the Val247Ala
(V247A) and Tyr96Phe/Val247Ala (Y96F/V247A) mutants. These
mutations would create additional space closer to the heme
rather than higher up in the active site where tyrosine 96 and
phenylalanine 87 are located. Both of these mutations resulted
in lower spin state shifts with 1 and did not result in a signifi-
cant increase in activity or product formation above the Y96F
mutant (Table 1). However, they did increase the selectivity of
product formation resulting in the oxidation to the product, 5,
observed with the Y96A mutant. The effect of the V247A
and Y96F/V247A mutants, on 2 was more dramatic and both
increased the spin state shift, turnover activity and levels of
product formation over the parent enzymes (WT and the Y96F
mutant, respectively, Table 1). Significantly the selectivity of
product formation was also shifted to favour product, 6, (85%)
over the product, 7, which is the major formed by the other
mutants (Table 2 and Fig. S3†).
The total amount of product formed with the protected
cyclohexanol substrates, 1 and 2, correlated well between the
P450cam mutants, i.e. if one mutant is better than another for
1 it is also better with 2. The benzoate ester protected com-
pounds gave more product (and higher NADH oxidation rates)
than the benzyl ether compounds. For both substrates the
F87A/Y96F and F87L/Y96F mutants had total overall turnovers
which were greater than the Y96F mutant but lower than the
Y96A mutant (Fig. S1†). Despite the greater NADH oxidation
rates observed for the F87A/Y96F mutant with 2 and Y96F with
1 the levels of product formation were lower than the Y96A
mutant indicating that unproductive uncoupling reactions
must be greater for these substrate/mutant combinations.37,49
Overall the trends in the spin state shift, NADH oxidation
activity and product formation levels are similar to those
obtained with comparably sized substrates such as benzyl-
cyclohexane and 1-phenyl-1-cyclohexylethylene.24
1-Cyclohex-2-enyl benzoate (3) yielded little or no product
with the WT enzyme and Y96F mutant. By way of contrast this
was also a poor substrate for the Y96A mutant with very low
levels of product formation being observed. However, 3 was
efficiently oxidised by both the F87L/Y96F and F87A/Y96F
mutants (Fig. S2†). The F87A/Y96F mutant resulted in the
highest NADH oxidation rate and level of product formation
(Table 1 and Fig. S2†).
Characterisation of the products
Products, 5 and 7, of substrates, 1 and 2, respectively, were
generated using a whole-cell oxidation system comprising of
the genes of the Y96A mutant of P450cam, putidaredoxin
(Pdx) and putidaredoxin reductase (PdR).50 The products
were extracted and purified by silica gel chromatography
The products of the monooxygenase reactions were further
analysed by HPLC which better highlighted variations in the
product distribution between different substrate and enzyme
combinations (Fig. S3†). For both 1 and 2 two product peaks
were observed with the Y96F mutant (4–7) with marginally
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 2479–2488 | 2481

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 3 The major products, 5 and 7, isolated from the oxidation of substrates 1 and 2 using the Y96A mutant. Both were characterised by 1H and 13
C
NMR spectroscopy. An overall scheme for oxidation of the cyclohexanol substrates including the addition of the benzyl and benzoate protecting
groups, oxidation by P450cam mutants and the deprotection steps. The carbons on products 5 and 7 have been numbered as per the NMR assign-
ments (see Experimental). DIPEA; diisopropylethylamine.
Fig. 4 The product (8) isolated from turnover of substrate 3 by the F87A/Y96F mutant of P450cam. An overall scheme for oxidation of cyclohexenol
including the addition of the benzoate protecting group, oxidation by the P450cam F87A/Y96F mutant and the deprotection steps. The carbons on
product 8 have been numbered as per the NMR assignments (see Experimental).
(34–106 mg of purified product). The highest yield after purifi- carbon in product 5 in its 13C spectra (Fig. S5†). Both cyclo-
cation (48%) was obtained for the cyclohexylbenzoate oxi- hexyl substrates (1 and 2) showed C-1 hydrogen 3J coupling
dation product 7. The products were characterised by NMR constants of 8.9–9.4 Hz, which is indicative of axial–axial
spectroscopy and both products 5 and 7, arose from hydroxy- 3J coupling which is larger than axial–equatorial and equator-
lation of the trans C–H bond at the C-4 position of the cyclo- ial–equatorial 3J couplings. Therefore the protected alcohol
hexyl ring (Fig. 3).
group substituent is found in the equatorial position and the
Oxidation at C-4 is easily identifiable due to the symmetric germinal hydrogen in the axial position. The alcohol groups
1
nature of the cyclohexyl ring and therefore the equivalent H can also be assigned as equatorial due to the large axial–axial
and 13C resonances of the species at C-2 and C-3 in the 3J couplings of the C-4 hydrogens (3J = 9.2–9.6 Hz, Fig. 3,
1
product (Fig. 3, S4 and S5†). The benzylic carbon and hydro- S4 and S5†). All assignments were confirmed by H–1H COSY
gens in product 5 (C-5) have characteristic signals in the and 1H–13C HMBC and HSQC methods. The deprotection
1H and 13C NMR spectrum (Fig. 4 and S4†) as does the ester of 5 using H2/Pd/C resulted in the isolation of trans-1,4-
2482 | Org. Biomol. Chem., 2014, 12, 2479–2488
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
cyclohexanediol, which was identified by NMR from the three This indicates that hydroxylation of the racemic mixture of 3
resonances. The mass of the compounds was also measured occurs diastereoselectively on each enantiomer. The minor
using GC-MS and was as expected (see Experimental).
products (<10%) observed via HPLC and GC (Fig. S2 and S7†)
The low levels of product formation of the Y96F, Y96F/ could be diastereomers of the racemic oxidation product, 8,
V247A, V247A and WT enzymes hampered attempts to isolate or arise from oxidation at a different carbon atom of the
and identify the oxidation products, 4 and 6. Product, 4, from protected cyclohexenol ring.
the oxidation of 1 was not produced in sufficient quantities for
1-Phenyl-1-cyclohexylethylene turnovers
NMR characterisation. A NMR spectrum was obtained for the
partially purified minor product 6 arising from oxidation of 2,
which showed that it was hydroxylated on the cyclohexyl ring
but the hydrogens could not be assigned because of extensive
overlap of the resonances in the aliphatic region. However, the
number of resonances strongly suggested that the compound
was not symmetrical and must have arisen from oxidation at
C-2 or C-3 of the substrate with C-3 being more likely by com-
parison with the product selectivity of phenylcyclohexane and
1-phenyl-1-cyclohexylethylene.24
Given the significant improvements in the product formation
of the F87L/Y96F and F87A/Y96F variants with 3 we also tested
1-phenyl-1-cyclohexylethylene (9) with these mutants and with
Y96F/V247A variant in order to assess the effects of these
mutations on product selectivity. This substrate is predomi-
nantly oxidised at C-4 by the Y96A mutant but the stereo-
selectivity is low and roughly equal amounts of cis- and trans-
4-phenyl-ethenyl cyclohexanol are produced (10 and 11,
Table 3).24
HPLC analysis of the turnover products of 1-cyclohex-2-enyl
benzoate, 3, revealed one major product (>90%) with the F87A/
Y96F mutant (Fig. S2†). After whole-cell oxidation turnovers,
purification and isolation the product, 8, was characterised by
1H and 13C NMR methods (1H–1H COSY and ROESY and
1H–13C HMBC and HSQC) as having arisen from hydroxylation
at the C-4 position of the cyclohexenyl ring (Fig. 4 and S6†).
The two olefinic hydrogen signals (at C-2 and C-3) at
approximately 5.9 and 6.0 ppm show clearly that the double
The spin state shifts with the F87L/Y96F and F87A/Y96F
mutants were high yet the activities and product formation of
all the mutants tested were lower than those of the Y96A
mutant. However, the selectivity of product formation was sig-
nificantly altered. The F87L/Y96F double mutant was the most
selective yielding 92% of the trans-4-phenyl-ethenylcyclohexa-
nol (11). The F87A/Y96F double mutant also favoured for-
mation of this product (79%) over the cis isomer but generated
additional minor products (Table 3). On the other hand the
Y96F/V247A mutant shifted the selectivity from the C-4 posi-
tion to generate the cis-3-phenyl-ethenylcyclohexanol (12) as
the major product.
3
bond is intact in 8. The J couplings of the hydrogens of C-1
and C-4 are small (4.7 and 2.4 Hz, Fig. S6†) but are larger than
for C1 of the substrate (3, <3.0 and 1.8 Hz). The analysis of
cyclohexene ring substrates is more complex than cyclohexyl
rings. C-1 and C-4 are in one plane forcing the ring into a half
chair configuration. As a result allylic substituents reside in
pseudoequatorial and pseudoaxial positions and the energy
Discussion
difference between having the bulkier substituents in the pseu- Cyclohexanol can be oxidised by small-molecule iron catalysts
doequatorial rather than the pseudoaxial position is lower to give cyclohexanone and a mixture of cyclohexanediols.52
than for the equatorial and axial positions in a cyclohexane Cyclohexenediols are usually synthesised from cyclohexa-
ring. Electronegative substituents such as OH and OR groups dienes.53,54 Cyclohexane turnover has been investigated with
often occupy the pseudoaxial position.51 This has been ration- WT P450cam and the Y96A mutant. Both gave low levels of
alised by an anomeric effect from a stabilising resonance cyclohexanol as the main product with the Y96A mutant also
between the double bond π system and the allylic bond anti- producing some of the further oxidation product, cyclohexa-
3
bonding σ orbital. The J coupling constants of the C-1 H of 3 none. Cyclohexane and cyclohexene oxidation by metallo-
are small, 1.8 and 3.0 Hz, suggesting a pseudoequatorial posi- porphyrins,55,56 biomimetic methods,57,58 and photochemical
tion. In addition the chemical shift of C-5 of the cyclohexane and electrochemical approaches,59,60 have also been studied.
ring in the 13C NMR spectra (δ 18.9) is indicative of a γ effect For cyclohexane these methods usually give cyclohexanol and
caused by a diaxial interaction between the substituent at C-1 cyclohexanone as the products while cyclohexene is usually
and a proton on the C-5 γ carbon. In the product, 8, the line- converted to the epoxide with some allylic alcohol and ketone
3
width at half-height and the J couplings of the hydrogens at formation.
C-1 and C-4, 2.4 and 4.7 Hz, indicate that the hydrogens
occupy pseudoequatorial positions with the substituents benzyl ether moiety linked to a cyclohexyl ring. The protected
pseudoaxial (Fig. 4).51
substrates (1–3) are of a similar size to the mixed cyclohexyl
The compounds investigated all contain a benzoate ester or
1-Cyclohex-2-enyl benzoate, 3, contains a chiral centre and and phenyl ring substrates, benzylcyclohexane and 1-phenyl-
was prepared as a racemic mixture. The hydroxylated product, 1-cyclohexylethylene, studied previously,24 and have been
8, has two chiral centres and was analysed via normal phase designed to fit the active site of the Y96A mutant of P450cam.
chiral HPLC and found to be essentially a racemic mixture This mutant is expected to have a more spacious and hydro-
(49 : 51, Fig. S7†). The NMR data showed that 8 was predomi- phobic active site than the WT enzyme and has been shown
nantly a single compound (>90%) and not diastereomeric. to accommodate large hydrophobic substrates such as
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 2479–2488 | 2483

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 3 Substrate-induced high-spin heme content (% HS, 5%), NADH oxidation rates (N, nmol per nmol P450 per min) and product distributions
of the P450cam mutants and 1-phenyl-1-cyclohexane. (n.d.) not determined
P450cam mutant
% HS
N
10
11
12
7%
13%
0%
7%
66%
Y96A
Y96F
F87L/Y96F
F87A/Y96Fa
Y96F/V247Aa
85%
45%
≥95%
≥95%
n.d.
168
100
40
160
50
47%
39%
8%
3%
6%
46%
48%
92%
79%
19%
a There were small amounts of a second minor product in the turnover (total 11% for F87A/Y96F and 9% for Y96F/V247A).
diphenylmethane and benzylcyclohexane. A preliminary report most of the mutants tested in this work. Moreover, the selecti-
of the crystal structure of the Y96A mutant, bound with vity for 6 could be increased to 85% by creating additional
4-hydroxydiphenylmethane, revealed that the active site resi- space at the bottom of the active site by adding the V247A
dues undergo rearrangement.41 Phenylalanine 87 moves into mutation.
the space vacated by the tyrosine residue in Y96A. Therefore
With all the substrates the WT protein showed a lower
while space is created in the active site it is not of the form of NADH oxidation activity and level of product formation than
an aromatic pocket where the tyrosine was located. The lin- the Y96F mutant indicating that increasing the hydrophobicity
kages between the phenyl and cyclohexyl ring of 1 to 3 are of the active site is a major factor in the substrate binding and
longer compared to those in benzylcyclohexane and 1-phenyl- subsequent turnover of these substrates. The lower activities
1-cyclohexylethylene and require the additional space gener- observed with 1 and benzylcyclohexane compared with 2
ated by the Y96A or the F87A/Y96F mutants for optimal suggest that the ester linkage facilitates C–H bond activation.
binding and activity.24 These protecting groups also test if the This could be due to a modified binding orientation of the
polarity of the linking group (1 and 2) affects substrate binding substrate in the active site leading to more efficient catalysis.
activity and product selectivity. The spin state shifts of 1 with The low levels of product formation of the Y96F, Y96F/V247A,
the WT and mutant enzyme are similar to those obtained with V247A and WT enzymes are likely due to the smaller size of
benzylcyclohexane and 1-phenyl-1-cyclohexylethylene. The car- the active sites of these enzymes. Mutants with larger and
bonyl group in 2 increased binding of the substrate to the WT more hydrophobic active sites (e.g. Y96A and F87A/Y96F) were
enzyme compared to the benzyl ether. This could be due to better biocatalysts. The formation of a more spacious active
polar interactions between the carbonyl oxygen of the sub- site appears to be more critical for efficient turnover of these
strate and the more hydrophilic active-site of the WT enzyme substrates than for phenylcyclohexane and benzylcyclohexane
in which the tyrosine at position 96 is still intact. Despite the oxidation as both of these substrates showed appreciable
substrate induced spin state shift the NADH oxidation rate and activity with the Y96F mutant.24,25
product formation rate for the WT enzyme with 2 were nuga-
Greater NADH oxidation rates were observed for certain
tory. Under the conditions used for the turnovers the first elec- mutant/substrate combinations, e.g. F87A/Y96F and cyclohexyl-
tron transfer is rate limiting for the overall reaction. A linear benzoate, 2, yet resulted in lower levels of product formation
free-energy relationship has been observed between the indicating that unproductive uncoupling reactions must
P450cam spin state equilibrium and heme redox potential and account for a larger proportion of the reducing equivalents in
the change in redox potential has been linked to thermodyn- these turnovers. Uncoupling of the reducing equivalents from
amic control of the first electron-transfer step.61 Kinetic NAD(P)H into non-productive oxygen reduction pathways can
control of this step via the reorganisation energy has also been occur at three points during the P450 catalytic cycle. Oxy-
hypothesised to be important.62 The smaller shift in the spin ferrous P450 is unstable to auto-oxidation to form superoxide.
state for the WT enzyme coupled with the low activities and If excess water molecules are present in the active site during
levels of product formation indicates that water binding to the the substrate turnover then protonation of the iron-bound
ferric iron is still significant in the substrate-bound species. oxygen in the ferric hydroperoxo intermediate may occur
This would slow down the rate of the first electron transfer resulting in dissociation of the hydroperoxo ligand and hydro-
and the NADH oxidation rate by either or both the thermodyn- gen peroxide formation. Lastly if the substrate is bound too far
amic or kinetic control mechanisms.1
away from the compound I oxygen, or if the functional group
The polar interactions between tyrosine 96 and the benzo- closest to this oxygen is inert to oxidation by compound I
ate ester moiety may also be the reason for the modified then electron-transfer reduction can compete with C–H bond
product selectivity for 2, with the WT enzyme which favoured insertion. The overall result is the use of two molecules of
the formation of the product, 6. This is the minor product for NAD(P)H to reduce O2 to two water molecules.37,49
2484 | Org. Biomol. Chem., 2014, 12, 2479–2488
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
The formation of products, 5 and 7, which are hydroxylated substrates and P450 enzymes. The size of the protecting group
on the C-4 position of the cyclohexyl ring suggests that the pro- could be modified conjointly with tailoring the active site of
tected substrates bind to the P450cam mutants with the the enzyme via rational design or directed evolution to create
phenyl group in the upper part of the active site and the cyclo- optimised biocatalysts for a range of organic molecules with
hexyl ring closer to the oxygen of the ferryl, compound I inter- applications in organic synthesis.
mediate. However, the longer linker between the cyclohexyl
and benzene rings in substrates 1 and 2 relative to those in
benzylcyclohexane and 1-phenyl-1-cyclohexylethylene results in
Experimental
changes in the selectivity of the C–H bond oxidations. For
General
both 1 and 2 the Y96A mutant showed superior product for-
General chemicals and HPLC solvents and were purchased
from Sigma-Aldrich and silica gel for chromatography was
from Grace Davison, Australia. The P450cam mutants were
generated on the C334A mutant (WT) as previously
described.25,26,34,37,48 The mutated P450cam variants, PdR and
Pdx proteins were produced and purified using standard
methods and stored in 50% glycerol at −20 °C prior to
use.63–66 Substrates 1–3 were synthesised using standard
methods (Fig. 1) and purified via flash chromatography46,47
The purity and identity of each was confirmed via NMR and
GC-MS (the numbering system is identical to that used in
Fig. 3 and 4).
mation and selectivity (>95%) than all the other mutants
tested including the F87A/Y96F. This contrasts with only 15%
selectivity for attack at the equatorial C–H bond in the oxi-
dation of phenylcyclohexane and 90% with benzylcyclohexane
by the Y96A mutant. 1-Phenyl-1-cyclohexylethylene oxidation
with Y96A results in high regioselectivity at the 4 position but
there is no preference for the trans C–H bond over the cis. The
high selectivity observed with substrates 1 and 2 suggests that
these protected substrates bind in the active site of the Y96A
with a single C–H bond in a desirable position for hydrogen
abstraction by compound I. The products 5 and 7 would both
yield trans-1,4-cyclohexanediol on deprotection.
Data for 1: Rf: 0.4 (petroleum ether–EtOAc, 100 : 1); BP
142 °C (1.3 mm Hg); 1H NMR (600 MHz, CDCl3) δ 1.21–1.29
(3H, m, H-3a and H-4a), 1.33–1.39 (2H, m, H-2a), 1.51–1.54
(1H, m, H-4e), 1.72–1.78 (2H, m, H-3e), 1.93–1.95 (2H, m,
H-2e), 3.35 (1H, tt, J = 9.4 and 3.8 Hz, H-1), 4.54 (2H, s, H-5),
7.23–7.26 (1H, m, H-9), 7.31–7.35 (4H, m, H-8 and H-7); 13C
NMR (150 MHz, CDCl3) δ 24.1 (C-3), 25.8 (C-4), 32.2 (C-2), 69.6
(C-2), 76.9 (C-5), 127.2 (C-7), 127.4 (C-9), 128.3 (C-8), 139.3
(C-6). GC-MS (C13H18O): calculated 190.13576; mass found
190.1.
The introduction of the double bond in 3 increases the
rigidity and changes the conformation of the cyclohexyl ring
and this has a significant effect on substrate binding and turn-
over. This substrate induced lower spin state shifts and
reduced NADH oxidation activities across the mutant enzymes
tested. The effect on product formation was more dramatic
with only the F87L/Y96F and F87A/Y96F mutants showing
appreciable catalytic turnover. The different conformation of
the cyclohexene ring must alter the position of the C–H bonds
relative to the heme iron and have a negative effect on the
amount of product formed with the Y96A mutant. By way of
comparison the oxidation of 1-phenyl-1-cyclohexylethylene,
which introduces an alkene double bond in the linker region,
is less stereoselective at the C-4 position than the oxidation of
benzylcyclohexane. However, the selectivity was improved
using the F87L/Y96F and F87A/Y96F mutants. These mutants
generate space in a different location in the active site and are
able to accommodate these double bond containing substrates
in a more favourable position for selective and efficient C–H
bond activation.
The products generated are chemically useful building
blocks in organic synthesis and could be routinely produced
in tens to a hundred milligram quantities (34–106 mg) for
characterisation, using a whole-cell oxidation system in shake
flasks. The yields are of the whole-cell oxidation step are mod-
erate (17–48%) as the substrates were added in excess to maxi-
mise the total turnover and to minimise any potential further
oxidation of the hydroxylated product, which is more soluble
in the aqueous buffer. Further optimisation of these whole-cell
systems, such as addition of co-solvents, increasing cell
density, increased control of aeration, carbon source and pH
and improved substrate addition and product removal regimes
1
Data for 2: Rf: 0.4 (petroleum ether–EtOAc, 100 : 1) H NMR
(600 MHz, CDCl3) δ 1.32–1.38 (1H, m, H-4a), 1.42–1.48 (2H, m,
H-3a), 1.55–1.64 (3H, m, H-2a and H-4e), 1.77–1.82 (2H, m,
H-3e) 1.93–1.96 (2H, m, H-2e), 5.03 (1H, tt, J = 8.9 and 3.8 Hz,
H-1), 7.43 (2H, t, J = 7.8 Hz, H-8), 7.54 (1H, t, J = 7.8 Hz, H-9),
8.05 (2H, d, J = 7.8 Hz, H-7); 13C NMR (150 MHz, CDCl3) δ 23.6
(C-3), 25.5 (C-4), 31.6 (C-2), 73.0 (C-1), 128.2 (C-8), 129.5 (C-7),
131.0 (C-6), 132.6 (C-9), 166.0 (C-5). GC-MS (C13H16O2): calcu-
lated 204.11503; mass found 204.1.
Data for 3: Rf: 0.8 (petroleum ether–EtOAc, 2 : 1) 1H NMR
(600 MHz, CDCl3) δ 1.67–1.73 (1H, m, H-5a), 1.81–1.91 (2H, m,
H-4a and H-6a), 1.95–2.01 (1H, m, H-5e), 2.02–2.08 (1H, m,
H-6e), 2.11–2.17 (1H, m, H-4e), 5.51 (1H, dd, J = 3.0, 1.8 Hz,
H-1), 5.83 (1H, dq, J = 9.6, 1.8 Hz, H-3), 6.00 (1H, dt, J = 9.6, 3.0
Hz, H-2), 7.43 (2H, t, J = 7.8 Hz, H-10), 7.54 (1H, t, J = 7.8 Hz,
H-11), 8.05 (2H, d, J = 7.8 Hz, H-9); 13C NMR (150 MHz, CDCl3)
δ 18.9 (C-5), 24.9 (C−-6), 28.4 (C-4), 68.6 (C-1), 125.7 (C-10),
128.4 (C-9), 128.8 (C-2), 129.6 (C-8), 132.7 (C-11), 132.8 (C-3),
166.2 (C-7). GC-MS (C13H14O2): calculated 202.09938; mass
found 202.1.
Substrate binding and turnover analysis
all would increase the yield of product obtained. More impor- Glycerol was removed immediately before use by gel filtration
tantly the results from this study could be extended to other on a 5 ml PD-10 column (GE Healthcare, UK) by eluting with
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 2479–2488 | 2485

View Article Online
Paper
Organic & Biomolecular Chemistry
50 mM Tris, pH 7.4. UV/Vis spectra and spectroscopic activity colony was inoculated into 2.5 ml 2YT broth (2YTamp) in a
assays were recorded at 30 0.5 °C on a Varian CARY-50 or 10 ml tube and grown at 37 °C for 6 hours when 2 ml of this
CARY-1E spectrophotometer. The P450cam mutants were sample was added to 500 ml 2YTamp in a two-litre flask and
diluted to ∼2 µM in 50 mM Tris, pH 7.4 (2.5 ml). Aliquots of grown until the O.D. reached 1. Protein expression was
substrate (0.5–2 μL) were added using a Hamilton syringe from induced by the addition of 100 μM IPTG (from a 0.5 M stock
a 100 mM stock solutions in DMSO. The sample was mixed in H2O) and the temperature was reduced to 25 °C and the
and the spectrum was obtained between 700 nm and 250 nm. shaker speed reduced to 140 rpm. The growths were allowed to
Further aliquots of substrate were added until the spin state continue for another 16 hours before the cell pellet was
did not shift further. The high-spin heme content was esti- harvested by centrifugation. The pellet was washed in EMM
mated (to approximately 5%) by comparison with a set of media and resuspended in an equal volume of EMMamp. The
spectra generated from the sum of the appropriate percentages cell suspension was divided into 250 ml aliquots and was
of the spectra of the substrate-free (>95% low-spin, Soret peak added to a 1 L flask and the oxidation started with the
at 418 nm) and camphor-bound (>95% high-spin, Soret peak addition of the substrate. The substrates (4 mM from a
at 392 nm) forms of WT P450cam.
100 mM stock in DMSO) were added to 250 ml of cell suspen-
Steady state NADH oxidation assays were performed with sion in a 1 L flask and the whole-cell reactions were then
mixtures (1.5 ml) containing 50 mM Tris, pH 7.4, 0.05 μM shaken at 220 rpm and 30 °C for a further 20 hours. The
P450cam, 10 μM Pdx, 0.5 μM PdR, 200 mM KCl and 100 μg whole-cell turnovers were analysed on a reverse phase silica
ml−1 bovine liver catalase. The mixtures were equilibrated at column (5 mm i.d. × 250 mm) using a mobile phase of water
30 °C for 2 min. NADH was added to ca. 320 μM (final A340
2.00) and the absorbance at 340 nm was monitored. Substrates 20–100% acetonitrile being developed over 30 min at a flow
were added from a 100 mM stock solution in DMSO to a final rate of 1 ml min−1
=
(0.01% trifluoroacetic acid)–acetonitrile with a gradient of
.
concentration of up to 0.5 mM. The rate of NADH turnover was
For product isolation the cell pellet was removed by cen-
calculated using ε340 = 6.22 mM−1 cm−1. 1-Phenyl-1-cyclohexyl- trifugation and supernatant was extracted using EtOAc. The
ethylene turnovers and product distributions were analysed as organic layer was separated and the extraction of the aqueous
described previously.24
phase was repeated with EtOAc (2 × 100 ml). The combined
Turnovers (1 ml) were extracted with 400 μL of ethyl acetate. organic layers were washed with brine (3 × 100 ml), then dried
Analytical HPLC work was carried out on a normal phase silica over MgSO4, filtered and concentrated in vacuo. The resultant
column (5 mm i.d. × 250 mm) using a mobile phase of mixture of substrate and product oil was purified using silica
hexane–isopropanol with a gradient of 0–8% isopropanol gel flash chromatography (petroleum ether–EtOAc, 2 : 1). The
being developed over 30 min at a flow rate of 1 ml min−1 products were then isolated and characterised by NMR and
(Fig. S3†). Chiral HPLC was performed using a Daicel Chiralcel GC-MS as described below.
OJ-H column (4.6 mm i.d. × 250 mm) using a hexane–isopro-
panol mobile phase (95 : 5) held for 4 minutes with a gradient
of 5–40% isopropanol being developed over 30 minutes at a
Enzymatic oxidation of (cyclohexyloxymethyl)benzene 1 to 5
flow rate of 0.5 ml min−1 (Fig. S7†). For gas chromatography After extraction and silica gel chromatography the majority
analysis a 1 or 2 μl sample of the ethyl acetate extracts of turn- product of cyclohexyloxymethylbenzene oxidation was isolated
over reactions were injected onto a J & W scientific 0.25 mm × as a colourless oil (34 mg).
1
30 m DB-1 fused silica gas chromatography column in a
Data for 5: H NMR (600 MHz, CDCl3) δ 1.25–1.34 (2H, m,
Fisons GC 8000 series gas chromatograph. The samples were H-3a), 1.37–1.43 (2H, m, H-2a), 1.95–1.98 (2H, m, H-3e),
carried through the column using helium carrier gas and the 2.04–2.06 (2H, m, H-2e), 3.38 (1H, tt, J = 9.7, 4.0 Hz, H-1), 3.68
compounds present were detected using a flame ionisation (1H, tt, J = 9.6, 4.1 Hz, H-4), 4.53 (2H, s, H−-5), 7.25–7.29 (1H,
detector. The injector was held at 150 °C, the detector at m, H-9), 7.33 (4H, d, J = 4.2 Hz, H-8 and H-7); 13C NMR
250 °C and the oven at 200 °C isotherm (Fig. S1 and S2†). (150 MHz, CDCl3) δ 29.3 (C-2), 32.6 (C-3), 69.6 (C-1), 70.1 (C-4),
GC-MS was carried out on a Shimadzu GC17A coupled to a 76.1 (C-5), 127.38 (C-7), 127.41 (C-9), 128.3 (C-8), 138.9 (C-6).
QP505A GC-MS detector equipped with a DB-5 MS column
(30 m × 0.25 mm × 0.25 μm). The injector and interface temp-
eratures were 250° and 280 °C, respectively and helium carrier
Synthesis of trans-cyclohexane-1,4-diol
gas was used.
To a solution of 5 (34 mg, 0.17 mmol) in ethanol (1 ml) was
added Pd/C (5 mg), followed by H2 (1 balloon, 1 atm), and the
reaction was stirred at room temperature for 16 hours. The
Whole-cell substrate oxidation
To isolate and identify the products a whole-cell oxidation reaction was filtered through a pad of celite, and then concen-
system utilising the plasmid pCWSGB++[P450cam mutant], trated in vacuo to give crude diol (18 mg, 95%) as a white solid,
which contains the genes for putidaredoxin reductase, putidar- which was pure enough for characterisation.
edoxin and the Y96A or the F87A/Y96F mutant, was trans-
Data for trans-cyclohexane-1,4-diol: 1H NMR (300 MHz,
formed into competent BL21(DE3) cells and grown on LB CD3OD) δ 1.17–1.23 (4H, m, H-2a), 1.78–1.81 (4H, m, H-2e),
plates containing ampicillin, 100 μg ml−1 (LBamp).50 A single 3.54 (2H, tt, J = 6.9, 3.5 Hz, H-1).
2486 | Org. Biomol. Chem., 2014, 12, 2479–2488
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Enzymatic oxidation of cyclohexyl benzoate 2 to 7
amenable to oxidation by this or other monooxygenase
enzymes.
After extraction and silica gel chromatography the majority
product of cyclohexyl benzoate oxidation was isolated as a pale
yellow solid (106 mg).
1
Data for 7: H NMR (600 MHz, CDCl3) δ 1.46–1.56 (2H, m,
Acknowledgements
H-3a), 1.61–1.65 (2H, m, H-2a), 2.03–2.06 (2H, m, H-3e),
2.11–2.14 (2H, m, H-2e), 3.80 (1H, tt, J = 9.2, 3.8 Hz, H-4), 5.02
(1H, tt, J = 9.4, 4.0 Hz, H-1), 7.43 (2H, t, J = 7.8 Hz, H-1), 7.55
(1H, t, J = 7.8 Hz, H-1), 8.03 (2H, d, J = 7.8 Hz, H-1); 13C NMR
(150 MHz, CDCl3) δ 28.4 (C-2), 32.2 (C-3), 68.8 (C-4), 72.2 (C-1),
128.2 (C-8), 129.4 (C-7), 130.7 (C-6), 132.8 (C-9), 166.0 (C-5).
GC-MS (C13H16O3): calculated 220.10994; mass found 220.1.
The authors thank Prof. James de Voss (University of Queens-
land) for chiral HPLC analysis. This work was partially funded
by the University of Adelaide and the BBSRC through a stu-
dentship. LLW and SGB thank Prof. Sabine Flitsch, University
of Edinburgh (now at the University of Manchester) for supply-
ing samples of protected alcohols for preliminary substrate
binding studies.
Enzymatic oxidation of cyclohex-2-en-1-yl benzoate 3 to 8
After extraction and silica gel chromatography the majority
product of cyclohexyl benzoate oxidation was isolated as a red
oil (F87A/Y96F = 52 mg).
References
1
Data for 8: H NMR (600 MHz, CDCl3) δ 1.65–1.70 (1H, m,
1 Cytochrome P450: Structure, Mechanism, and Biochemistry,
ed. P. R. Ortiz de Montellano, Kluwer Academic/Plenum
Publishers, New York, 3rd edn, 2005.
2 A. Sigel, H. Sigel and R. Sigel, The Ubiquitous Roles of
Cytochrome P450 Proteins, John Wiley & Sons, 1st edn, 2007,
vol. 3.
3 F. P. Guengerich, Chem. Res. Toxicol., 2001, 14, 611–650.
4 E. M. Isin and F. P. Guengerich, Biochim. Biophys. Acta,
2007, 1770, 314–329.
H-5), 1.77–1.82 (1H, m, H-6a), 2.19–2.24 (1H, m, H-5e),
2.25–2.29 (1H, m, H-6e), 4.33–4.36 (1H, m, J = 4.7, 2.4 Hz, H-4),
5.56–5.58 (1H, m, J = 4.7, 2.4 Hz, H-1), 5.91 (1H, d, J = 10.8 Hz,
H-2), 5.99 (1H, d, J = 10.8 Hz, H-3), 7.43 (2H, t, J = 7.8 Hz,
H-10), 7.55 (1H, t, J = 7.8 Hz, H-11), 8.03 (2H, d, J = 7.8 Hz,
H-9); 13C NMR (150 MHz, CDCl3) δ 26.2 (C-5), 29.9 (C-6), 65.6
(C-4), 68.9 (C-4), 128.3 (C-10), 128.5 (C-2), 129.6 (C-9), 130.3
(C-8), 132.9 (C-11), 134.5 (C-3), 166.1 (C-7). GC-MS (C13H14O3):
calculated 218.09430; mass found 218.1.
5 S. G. Bell, N. Hoskins, C. J. C. Whitehouse and L. L. Wong,
Design and Engineering of Cytochrome P450 Systems, John
Wiley & Sons, 1st edn, 2007, vol. 3, ch. 14, pp. 437–476.
6 R. Bernhardt, J. Biotechnol., 2006, 124, 128–145.
7 F. Hollmann, I. W. C. E. Arends, K. Buehler, A. Schallmey
and B. Buhler, Green Chem., 2011, 13, 226–265.
8 D. Monti, G. Ottolina, G. Carrea and S. Riva, Chem. Rev.,
2011, 111, 4111–4140.
Conclusion
The protected cyclohexanols and cyclohex-2-enols were poor
substrates for WT P450cam and the Y96F mutant resulted in
oxidations of low yield and selectivity. However, the Y96A or
the F87A/Y96F mutants showed improved activity with all of
the protected substrates. The introduction of a double bond in
the cyclohexyl ring appears to have a marked effect on the sub-
9 C. J. Whitehouse, S. G. Bell and L. L. Wong, Chem. Soc.
Rev., 2012, 41, 1218–1260.
strate binding and turnover efficiency. The Y96A mutant 10 R. Agudo, G. D. Roiban and M. T. Reetz, ChemBioChem,
resulted in the optimal product turnover and selectivity for 2012, 13, 1465–1473.
the protected cyclohexanol substrates, (cyclohexyloxymethyl)- 11 S. Q. Pham, P. Gao and Z. Li, Biotechnol. Bioeng., 2013, 110,
benzene (1) and cyclohexyl benzoate (2) producing >95% of a 363–373.
single majority product for both substrates which were identi- 12 H. Venkataraman, S. B. A. de Beer, D. P. Geerke,
fied as 4-benzyloxy-cyclohexanol (5) and benzoic acid-
4-hydroxy cyclohexyl ester (7), respectively. Both of these would
N. P. E. Vermeulen and J. N. M. Commandeur, Adv. Synth.
Catal., 2012, 354, 2172–2184.
yield trans-1,4-cyclohexanediol on deprotection. The F87A/ 13 M. Katagiri, B. N. Ganguli and I. C. Gunsalus, J. Biol.
Y96F mutant was a superior biocatalyst for the protected cyclo- Chem., 1968, 243, 3543–3546.
hexenol substrate, 1-cyclohex-2-enyl benzoate (3). The F87A/ 14 T. L. Poulos, B. C. Finzel and A. J. Howard, J. Mol. Biol.,
Y96F mutant produced >90% of benzoic acid 4-hydroxy-cyclo- 1987, 195, 687–700.
hex-2-enyl ester (8) which would result in the formation of 15 W. M. Atkins and S. G. Sligar, J. Am. Chem. Soc., 1987, 109,
trans-cyclohex-2-ene-1,4-diol. Overall these results confirm that 3754–3760.
changing the active site volume and hydrophobicity can 16 W. M. Atkins and S. G. Sligar, Biochemistry, 1988, 27, 1610–
increase the binding of large hydrophobic substrates. Design- 1616.
ing the substrate, by changing the protecting group to match 17 P. J. Loida, S. G. Sligar, M. D. Paulsen, G. E. Arnold and
the size and shape of the enzyme active site pocket could R. L. Ornstein, J. Biol. Chem., 1995, 270, 5326–5330.
allow the development of synthetic strategies for the selective 18 W. M. Atkins and S. G. Sligar, J. Biol. Chem., 1988, 263,
hydroxylation of smaller more hydrophilic substrates not
18842–18849.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 2479–2488 | 2487

View Article Online
Paper
Organic & Biomolecular Chemistry
19 R. E. White, M.-B. McCarthy, K. D. Egeberg and S. G. Sligar, 40 S. J. Aitken, G. Grogan, C. S. Y. Chow, N. J. Turner and
Arch. Biochem. Biophys., 1984, 228, 493–502.
20 K. S. Eble and J. H. Dawson, Biochemistry, 1984, 23, 2068–
2073.
21 M. H. Gelb, P. Malkonen and S. G. Sligar, Biochem. Biophys.
Res. Commun., 1982, 104, 853–858.
S. L. Flitsch, J. Chem. Soc., Perkin Trans. 1, 1998, 3365–
3370.
41 S. L. Flitsch, S. J. Aitken, C. S. Y. Chow, G. Grogan and
A. Staines, Bioorg. Chem., 1999, 27, 81–90.
42 H. Kunz and H. Waldmann, in Comprehensive Organic Syn-
thesis, ed. B. M. Trost, Pergamon, Oxford, UK, 1991, vol. 6,
pp. 631–701.
22 D. M. Maryniak, S. Kadkhodayan, G. B. Crull, T. A. Bryson
and J. H. Dawson, The syntheses of 1R- and 1S-5-methyle-
nylcamphor and their epoxidation by cytochrome P-450- 43 C. B. Reese, in Protective groups in organic chemistry, ed.
CAM, Tetrahedron, 1993, 49, 9373–9384.
J. F. W. McOmie, Plenum Press, New York, 1973, pp.
95–143.
44 M. G. Banwell, J. N. Lambert and S. L. Richards,
Aust. J. Chem., 1991, 44, 939–950.
45 G. C. Bloomfield, T. J. Ritchie and R. Wrigglesworth,
J. Chem. Soc., Perkin Trans. 1, 1992, 1229–1236.
23 V. Helms, E. Deprez, E. Gill, C. Barret, G. H. B. Hoa and
R. C. Wade, Biochemistry, 1996, 35, 1485–1499.
24 S. G. Bell, D. A. Rouch and L.-L. Wong, J. Mol. Catal. B:
Enzym., 1997, 3, 293–302.
25 P. A. England, D. A. Rouch, A. C. G. Westlake, S. G. Bell,
D. P. Nickerson, M. Webberley, S. L. Flitsch and 46 J. W. Gathirwa and T. Maki, Tetrahedron, 2012, 68, 370–375.
L.-L. Wong, Chem. Commun., 1996, 357–358.
26 S. M. Fowler, P. A. England, A. C. G. Westlake, D. R. Rouch,
47 K. Yasui, K. Fugami, S. Tanaka and Y. Tamaru, J. Org.
Chem., 1995, 60, 1365–1380.
D. P. Nickerson, C. Blunt, D. Braybrook, S. West, 48 D. P. Nickerson and L. L. Wong, Protein Eng., 1997, 10,
L. L. Wong and S. L. Flitsch, J. Chem. Soc., Chem. Commun.,
1994, 2761–2762.
27 R. E. Speight, F. E. Hancock, C. Winkel, H. S. Bevinakatti,
M. Sarkar, S. L. Flitsch and N. J. Turner, Tetrahedron: Asym-
metry, 2004, 15, 2829–2831.
1357–1361.
49 S. Kadkhodayan, E. D. Coulter, D. M. Maryniak,
T. A. Bryson and J. H. Dawson, J. Biol. Chem., 1995, 270,
28042–28048.
50 S. G. Bell, C. F. Harford-Cross and L.-L. Wong, Protein Eng.,
2001, 14, 797–802.
28 P. A. England, C. F. Harford-Cross, J. A. Stevenson,
D. A. Rouch and L. L. Wong, FEBS Lett., 1998, 424, 271–274. 51 J. Lessard, P. V. M. Tan, R. Martino and J. K. Saunders,
29 D. P. Nickerson, C. F. Harford-Cross, S. R. Fulcher and
L. L. Wong, FEBS Lett., 1997, 405, 153–156.
30 J. A. Stevenson, A. C. G. Westlake, C. Whittock and
L. L. Wong, J. Am. Chem. Soc., 1996, 118, 12846–12847.
31 S. G. Bell, X. Chen, R. J. Sowden, F. Xu, J. N. Williams,
Can. J. Chem., 1977, 55, 1015–1023.
52 J. T. Groves and M. Vanderpu, J. Am. Chem. Soc., 1974, 96,
5274–5275.
53 C. Kaneko, A. Sugimoto and S. Tanaka, Synthesis, 1974,
876–877.
L. L. Wong and Z. Rao, J. Am. Chem. Soc., 2003, 125, 705– 54 T. Tschamber, F. Backenstrass, H. Fritz and J. Streith, Helv.
714. Chim. Acta, 1992, 75, 1052–1060.
32 S. G. Bell, X. Chen, F. Xu, Z. Rao and L. L. Wong, Biochem. 55 E. R. Birnbaum, R. M. Le Lacheur, A. C. Horton and
Soc. Trans., 2003, 31, 558–562. W. Tumas, J. Mol. Catal. A: Chem., 1999, 139, 11–24.
33 S. G. Bell, E. Orton, H. Boyd, J.-A. Stevenson, A. Riddle, 56 B. Meunier, Chem. Rev., 1992, 92, 1411–1456.
S. Campbell and L.-L. Wong, Dalton Trans., 2003, 2133– 57 Y. Moro-oka and M. Akita, Catal. Today, 1998, 41, 327–338.
2140.
58 S. I. Murahashi and N. Komiya, Catal. Today, 1998, 41,
34 S. G. Bell, R. J. Sowden and L.-L. Wong, Chem. Commun.,
2001, 635–636.
35 S. G. Bell, J.-A. Stevenson, H. D. Boyd, S. Campbell,
339–349.
59 F. Blatter, H. Sun, S. Vasenkov and H. Frei, Catal. Today,
1998, 41, 297–309.
A. D. Riddle, E. L. Orton and L.-L. Wong, Chem. Commun., 60 K. Otsuka and I. Yamanaka, Catal. Today, 1998, 41, 311–
2002, 490–491. 325.
36 X. Chen, A. Christopher, J. P. Jones, S. G. Bell, Q. Guo, 61 S. G. Sligar, Biochemistry, 1976, 15, 5399–5406.
F. Xu, Z. Rao and L. L. Wong, J. Biol. Chem., 2002, 277, 62 M. J. Honeychurch, H. A. O. Hill and L. L. Wong, FEBS
37519–37526.
Lett., 1999, 451, 351–353.
37 C. F. Harford-Cross, A. B. Carmichael, F. K. Allan, 63 J. A. Peterson, M. C. Lorence and B. Amarneh, J. Biol.
P. A. England, D. A. Rouch and L. L. Wong, Protein Eng.,
2000, 13, 121–128.
38 D. F. Munzer, P. Meinhold, M. W. Peters, S. Feichtenhofer,
Chem., 1990, 265, 6066–6073.
64 A. C. Westlake, C. F. Harford-Cross, J. Donovan and
L. L. Wong, Eur. J. Biochem., 1999, 265, 929–935.
H. Griengl, F. H. Arnold, A. Glieder and A. de Raadt, Chem. 65 T. Yasukochi, O. Okada, T. Hara, Y. Sagara, K. Sekimizu
Commun., 2005, 2597–2599.
39 A. Robin, V. Kohler, A. Jones, A. Ali, P. P. Kelly, E. O’Reilly,
and T. Horiuchi, Biochim. Biophys. Acta, 1994, 1204,
84–90.
N. J. Turner and S. L. Flitsch, Beilstein J. Org. Chem., 2011, 66 F. Xu, S. G. Bell, Z. Rao and L. L. Wong, Protein Eng., Des.
7, 1494–1498.
Sel., 2007, 20, 473–480.
2488 | Org. Biomol. Chem., 2014, 12, 2479–2488
This journal is © The Royal Society of Chemistry 2014