Stereochemical and mechanistic aspects of dioxygenase-catalysed benzylic hydroxylation of indene and chromane substrates

Article abstract of DOI:10.1039/b300898c

Toluene dioxygenase (TDO)-catalysed benzylic hydroxylation of indene substrates (8, 16 and 17), using whole cell cultures of Pseudomonas putida UV4, was found to yield inden-1-ol (14 and 223) and indan-1-one bioproducts (15 and 23). The formation of these bioproducts is consistent with the involvement of carbon-centred radical intermediates. TDO-catalysed oxidation of indenes 8 and 16 also gave cis-diols 13 and 18 respectively. TDO and naphthalene dioxygenase (NDO), used as both whole-cell preparation and as purified enzymes, were found to catalyse the benzylic hydroxylation of chromane 30, deuteriated (±)-chromane 30_D and enantiomers (4S)-30_D and (4R)-30_D to yield (4R)- and (4S)-chroman-4-ols 31/31_D respectively. The mechanism of benzylic hydroxylation of chromane 30/30_D involves the stereoselective abstraction of a pro-R (with TDO) or a pro-S (with NDO) hydrogen atom at C-4 and a marked preference for retention of configuration.

Full text of DOI:10.1039/b300898c

View Article Online / Journal Homepage / Table of Contents for this issue
Stereochemical and mechanistic aspects of dioxygenase-catalysed
benzylic hydroxylation of indene and chromane substrates
Derek R. Boyd,*a Narain D. Sharma,a Nigel I. Bowers,a Rosemary Boyle,a John S. Harrison,a
Kyoung Lee,b,d Timothy D. H. Buggc and David T. Gibson*d
a School of Chemistry, The Queen’s University of Belfast, Belfast, UK BT9 5AG.
E-mail: dr.boyd@qub.ac.uk; Fax: 44 (0) 028 90382117
b Department of Microbiology, Changwon National University, Kyongnam, South Korea
c Department of Chemistry, The University of Warwick, Coventry, UK CV4 7AL
d Department of Microbiology and the Center for Biocatalysis and Bioprocessing,
The University of Iowa, Iowa City, IA 52240, USA
Received 27th January 2003, Accepted 20th February 2003
First published as an Advance Article on the web 18th March 2003
Toluene dioxygenase (TDO)-catalysed benzylic hydroxylation of indene substrates (8, 16 and 17), using whole cell
cultures of Pseudomonas putida UV4, was found to yield inden-1-ol (14 and 22) and indan-1-one bioproducts (15 and
23). The formation of these bioproducts is consistent with the involvement of carbon-centred radical intermediates.
TDO-catalysed oxidation of indenes 8 and 16 also gave cis-diols 13 and 18 respectively. TDO and naphthalene
dioxygenase (NDO), used as both whole-cell preparations and as purified enzymes, were found to catalyse the
benzylic hydroxylation of chromane 30, deuteriated ( )-chromane 30D and enantiomers (4S)-30D and (4R)-30D to
yield (4R)- and (4S)-chroman-4-ols 31/31D respectively. The mechanism of benzylic hydroxylation of chromane
30/30D involves the stereoselective abstraction of a pro-R (with TDO) or a pro-S (with NDO) hydrogen atom at
C-4 and a marked preference for retention of configuration.
Introduction
Ring hydroxylating dioxygenase enzymes (Rieske non-heme
iron oxygenases)1 catalyse a wide range of stereoselective
oxygenations.2–5 These include monooxygenation (e.g. benzylic
hydroxylation6–11 or sulfoxidation12–15), dioxygenation (e.g. cis-
dihydroxylation3,4 or bis-benzylic hydroxylation10), trioxygen-
ation (e.g. benzylic hydroxylation/cis-dihydroxylation11,16,17 or
cis-dihydroxylation/sulfoxidation15,18
)
and tetraoxygenation
(e.g. bis-cis-dihydroxylation19,20).
The mechanism of any of these dioxygenase-catalysed
oxygenations remains to be elucidated. However, it is known
that aromatic ring hydroxylating dioxygenase systems generally
contain two or three components. In naphthalene dioxygenase
(NDO, E.C.1.14.12) these are: (i) an iron–sulfur flavoprotein
reductase, (ii) a Rieske [2Fe–2S] ferredoxin involved in electron
transfer from NADP(H) and (iii) a catalytic oxygenase
component with a mononuclear iron site.1 Recent X-ray crystal-
lographic studies of NDO have greatly increased our under-
standing of the structure of its catalytic domain.21,22 The active
site of NDO was found to contain several hydrophobic amino
acids in a relatively flat and elongated cavity capable of accept-
ing aromatic substrates. An X-ray crystal structure of NDO has
also shown indole 1, an aromatic substrate, bound at the active
site as an Fe() peroxo species 2 linked at C-3, that may be an
intermediate in the oxidation of indole 1 to indigo 5 (Scheme 1).
The product, released from the binding site, is assumed to
be the transient cis/trans-1,2-dihydrodiol metabolite of indole,
3cis/3trans, that spontaneously dehydrates yielding indoxyl 4 and
the autoxidation product indigo 5.
Structures similar to the Fe() peroxo species 2 or involving
an Fe()OOH species have been postulated as intermediates in
the mechanism of mono- and di-oxygenation reactions of
comparable substrates catalysed by dioxygenases.21–23 Biotrans-
formation studies of the indole isosteres, benzo[b]thiophene 6,
benzo[b]furan 7 and indene 8, have been carried out in our
laboratories.6,24–27 In contrast with the transient equilibrating
cis/trans diol metabolites of indole, 3cis/3trans shown in Scheme 1,
the corresponding heterocyclic dihydrodiols, 9cis/9trans and 10cis/
10trans, proved to be sufficiently stable to isolate and assign
stereochemically.26,27 The carbocyclic cis-dihydrodiols 11 and
12, derived from benzo[b]thiophene26 and benzo[b]furan27
respectively, and ring opened products from dihydrodiols, 9cis/
9trans and 10cis/10trans, were also isolated. Furthermore, dioxygen-
ase-catalysed sulfoxidation was observed in benzo[b]thiophene
substrates, yielding their sulfoxides and the derived cyclo-
adducts.26,27 Thus for mechanism studies, the indenes 8, 16 and
17 were considered as simpler models for the dioxygenase-
catalysed oxidation of indole 1.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 2 9 8 – 1 3 0 7
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
1298

View Article Online
Scheme 1
Scheme 2
Scheme 3
indene 8 and this strain had shown that (Ϫ)-(1R)-inden-1-ol 14
Results and discussion
was enantiopure (> 98% ee). This change in ee value may have
been due to asymmetric destruction of one enantiomer of
inden-1-ol and/or a primary kinetic isotope effect. Stereo-
selective removal of one enantiomer of indan-1-ol had been
observed earlier10 with P. putida UV4. Spontaneous rearrange-
ment of the 2H-labelled samples of inden-1-ol 14D and 14ЈD gave
both labelled [3-2H] and unlabelled forms of indan-1-one 15
Initial studies with indene 8 were carried out using whole cell
mutant strains of bacteria containing TDO (P. putida 39/D6
and P. putida UV47,25). The cis-dihydrodiol dehydrogenase
enzyme, responsible for the desaturation of the substituted
benzene cis-dihydrodiol metabolites to the corresponding
catechols, was blocked in the UV4 and 39/D strains thus
allowing the cis-dihydrodiol metabolites to accumulate. With
indene substrate 8, both strains were found to yield cis-1,2-di-
hydroxyindane 13 (cis-dihydroxylation product), inden-1-ol
14 (benzylic hydroxylation product) and indan-1-one 15
(rearrangement product of inden-1-ol 14) as major bioproducts
(Scheme 2).
2
(cf. Scheme 2). The equilibration of the H atom between the
C-1 and C-3 positions in indene 8D is evident from inden-1-ol
metabolites 14D/14ЈD found using UV4 or 39/D mutant strains
of P. putida. This is consistent with the intermediacy of a stabil-
ؒ
ised allylic/benzylic carbon-centred radical 8D (Scheme 3).
Further evidence for the involvement of delocalised carbon-
The earlier report6 on the metabolism of [3-2H]- or [3-D]-
indene 8D using TDO (P. putida F 39/D) showed that (Ϫ)-cis-
diol 13D was formed with an enantiomeric excess (ee) favouring
the (1S,2R) configuration (30% ee) (Scheme 3). Diol 13D also
centred radicals similar to 8 during TDO-catalysed benzylic
ؒ
hydroxylation of 5-bromoindene 16 and 6-bromoindene 17
substrates was sought (Scheme 2). The possible indane cis-diol
(13,18,21), inden-1-ol (14,19,22) and indan-1-one (15,20,23)
metabolites, from the corresponding indene substrates
(8,16,17), are shown in Scheme 2. In practice, the (ϩ)-(1R,2S)-
dihydrodiol metabolite 18 (46% ee) was isolated from 5-bromo-
indene 16 as the major (42% yield) metabolite (Scheme 4).
This result is comparable to that found earlier using indene
substrate 8D where cis-1,2-dihydroxyindane 13D of opposite
configuration (1S,2R, 38% ee) was obtained. The (1R,2S)
absolute configuration of cis-diol 18 was determined by circular
dichroism (CD) comparison with (1S,2R)-indene cis-diol
metabolite 13 where the strongest CD absorption peaks (196–
203 nm) were of opposite signs. The ee value of cis-diol 18 was
2
had the H atom located exclusively at the C-1 position while
2
the (ϩ)-(1S)-inden-1-ol (26% ee) showed the H atom to be
distributed between positions C-3 (60% 2H, 14D) and C-1 (40%
2H, 14ЈD). A comparative study, again with [3-2H]-indene 8 and
TDO (whole cells of P. putida UV4), gave comparable results.
2
Thus, the H atom was found only at C-1 in (Ϫ)-(1S,2R)-diol
13D (38% ee) while inden-1-ol showed the 2H atom to be located
at positions C-3 (43% 2H, 14D) and C-1 (57% 2H, 14ЈD) by
1H-NMR analysis. A preference for the opposite (Ϫ)-(1R)
configuration of inden-1-ol 14D/14ЈD (28% ee) was, however,
observed using P. putida UV4. Earlier studies with unlabelled
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 2 9 8 – 1 3 0 7
1299

View Article Online
Scheme 4
1
estimated by H-NMR analysis of the (Ϫ)-(S)-[2-(1Ј-methoxy-
ethyl)benzene]boronic acid (MEBBA) derivative.28
A minor (4% yield) metabolite from 5-bromoindene 16 was
identified as 6-bromoindan-1-one 23; it was assumed to have
been obtained by spontaneous rearrangement of the relatively
unstable 6-bromoinden-1-ol intermediate 22 (Scheme 4). This
isolation of metabolite 23 is significant since it provides further
Scheme 5
destruction.10 The configuration of the minor metabolite
(ϩ)-28 (R = Cl) shown in Scheme 5 was earlier10 incorrectly
assigned as (1R,2R) instead of (1S,2S).
ؒ
evidence of an equilibrating allylic/benzylic radical (24 ). In
this case abstraction of a hydrogen atom from C-1 in
5-bromoindene 16 was followed by double bond migration and
hydroxylation at the original C-3 position; no evidence of either
inden-1-ol 19 or ketone 20 was found.
Biotransformation of 2-iodoindane 25 (R = I) was of particu-
lar interest in the context of radical intermediate formation
during dioxygenase-catalysed benzylic hydroxylation; inden-1-
ol 14 (13% yield) and cis-diol 13 (20% yield) were obtained
along with traces (< 1% yield) of the anticipated benzylic
hydroxylation products 27 and 29 (R = I). The former two
major metabolites were structurally and stereochemically indis-
tinguishable from the metabolites of indene 8 (Scheme 2);
their formation can be explained by the initial generation of a
Biotransformation of 6-bromoindene 17 using P. putida UV4
involved benzylic hydroxylation with an exclusive preference for
oxidation at the C-1 position to yield 6-bromoinden-1-ol 22
(16% yield). The derived 6-bromoindan-1-one 23 (28% yield)
was also isolated; no evidence was found of the dihydroxylation
product, cis-diol 21. The observed benzylic hydroxylation of
both 5-bromoindene 16 and 6-bromoindene 17 to yield the
same indan-1-one product 23 appears to be the result of one of
ؒ
benzylic radical 26 followed by the loss of an iodine atom
(Scheme 5). This rapid homolytic cleavage of an iodine atom
β to a carbon-centred radical intermediate, e.g. 26 , is well
ؒ
the resonance forms of the allylic/benzylic radical 24 being
ؒ
preferred for hydroxyl-group-transfer at the active site. The
absolute configuration of 6-bromoinden-1-ol 22 was deter-
mined as (1S) by comparison of its CD spectrum with
(Ϫ)-inden-1-ol 14 (from indene 8 with P. putida UV4), which
had the opposite (1R) absolute configuration. Benzylic alcohol
22 was found to have an ee value of > 98% by chiral stationary
phase high-pressure liquid chromatography (CSPHPLC). The
examination of relative contributions of enzyme-catalysed
asymmetric synthesis, or stereoselective removal of one
enantiomer, on the absolute configurations and ee values of
cis-diol 18 and inden-1-ol 22 was not carried out.
precedented. Thus, the rate of homolysis of an iodoethyl
radical to yield propene is 700 times faster than the ring open-
ing of the ethylcyclopropyl radical to yield the methylallyl
radical, a favoured process widely used to detect radicals.31
The instability of inden-1-ol bioproducts, 14 and 22,
obtained from indenes 8 and 17 respectively (Schemes 2 and 4),
and the range of oxidation products from indanes 25 (R = H,
Cl, Br and I, Scheme 5), rendered them less desirable substrates
for dioxygenase mechanism studies. 1,2-Dihydrobenzo[b]-
thiophene and 1,2-dihydrobenzo[b]furan have bicyclic struc-
tures similar to indenes, and contain only one benzylic site with
a heteroatom at the other benzylic position. However, with
TDO (P. putida UV4) as biocatalyst, benzylic hydroxylation,
sulfoxidation, desaturation and cis-dihydroxylation reactions
occur and hence a mixture of bioproducts was also found with
these substrates.26,27,32
Chromane 30 was next adopted as a model for further
mechanism studies, initially using whole cells of P. putida UV4.
Chroman-4-ol 31 (13% yield) and chroman-4-one 32 (5%) were
isolated as the sole metabolites (Scheme 6, a). The more polar
alcohol 31, isolated by PLC (Rf 0.28, 50% Et2O in hexane), was
found to have a large excess of the (4R) enantiomer (95% ee) by
CSPHPLC analysis. While this ee value could be due to select-
ive substitution of the pro-R benzylic hydrogen atom during
hydroxylation, the possible role of stereoselective destruction of
one enantiomer (kinetic resolution) occurring during further
oxidation to ketone in P. putida UV4 must also be considered.
This process was indeed observed using racemic chroman-4-ol
31 as substrate. Chroman-4-one 32 was formed (18% yield) and
the residual alcohol 31 (21% recovered yield) was found to have
an excess of the (4R) enantiomer (84% ee). Kinetic resolution
involving the selective oxidation of the (4S) enantiomer of
alcohol 31 to yield ketone 32 is assumed to have arisen
from an alcohol dehydrogenase enzyme present in whole cells
of P. putida UV4. Since the (4R) enantiomer of chroman-4-ol
The results obtained using TDO (P. putida UV4) as bio-
catalyst with [3-2H]-indene 8, 5-bromoindene 16 and 6-bromo-
indene 17 are compatible with the initial formation of
ؒ
ؒ
equilibrating radical intermediates (8 or 24 ) shown in Schemes
3 and 4. The hydrogen-atom-abstraction process is facilitated
by the relatively weak allylic/benzylic C–H bond in substrates 8,
16 and 17 and by the resulting delocalisation of a single electron
ؒ
ؒ
between C-1 and C-3 in intermediate 8 or 24 . It is evident that
the precise regioselectivity of dioxygenase-catalysed oxidation
at the indene ring would be determined by orientation of the
substrate within the active site; this would, in turn, depend
upon the relative positions of the indene substituent and the
radical intermediate.
Earlier studies of the metabolism of indane 25 (R = H),6,7,29
2-substituted indanes 25 (R = Cl, Br, I),10 and indanone sub-
strates30 also lend support to the concept of stereoselective
TDO- and NDO-catalysed monohydroxylation via a benzylic
ؒ
radical e.g. 26 (Scheme 5). The biotransformation pathways
(P. putida UV4) of indanes 25 (R = H, Cl, Br, I) include benzylic
mono-hydroxylation to yield monols 27 (R = H, Cl, Br) and 28
(R = Cl) and benzylic bis-hydroxylation to yield 1,3-diols 29
(R = H, Cl, Br).10 The enantiopure indan-1-ol products 27–29
obtained had identical configurations at the benzylic stereo-
genic centres (Scheme 5), and could have resulted from
either asymmetric oxidation or stereoselective asymmetric
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 2 9 8 – 1 3 0 7
1300

View Article Online
Scheme 6
Scheme 7
Scheme 8
31 was preferentially formed, using either chromane 30 or
over non-deuteriated (4R)-chroman-4-ol 31 may be due to a
racemic chroman-4-ol 31 as substrates, the relative contribu-
tions of asymmetric synthesis and kinetic resolution during the
whole cell biotransformation could not be estimated.
primary kinetic isotope effect operating during the enzyme-
catalysed hydroxylation and/or ketone formation. The presence
of deuterium at a benzylic position had earlier been found to
exert a marked influence on the ratio of bioproducts obtained
from TDO-catalysed (P. putida UV4) monohydroxylation of
benzyl cyanide and was attributed to a primary kinetic isotope
effect.11
Racemic and enantiopure forms of 4-deuteriated chromane
30D were required as substrates to investigate further aspects of
the dioxygenase-catalysed benzylic hydroxylation mechanism
(Scheme 6). These were synthesised by LiAl2H4 reduction of
racemic, (Ϫ)-(3R,4S)- and (ϩ)-(3S,4R)-3,4-epoxychromane 34
to yield chroman-3-ol 35D (Scheme 7). Reaction of alcohol 35D
with p-toluenesulfonyl chloride to yield the tosylate 36D and
reductive cleavage with LiAlH4 gave the mono-deuteriated
racemic, (4S)- and (4R)-forms of chromane 30D. The 3,4-epoxy-
chromane enantiomers (3R,4S)-34 ([α]D Ϫ71) and (3S,4R)-34
([α]D ϩ 69), obtained earlier from their 3-bromo-4-methoxy-
(phenyl)trifluoromethylacetate (bromo-MTPA) esters (3S,4S)-
33 and (3R,4R)-33,33 gave the corresponding [4-2H]-chromane
enantiomers (4S)-30D ([α]D Ϫ1.7 0.5) and (4R)-30D ([α]D ϩ 1.0
0.5) respectively. Enantiocomplementary CD spectra with
maximum absorption at ca. λ 228 nm were also obtained for
each enantiomer.
Chromane enantiomers (4S)-30D and (4R)-30D were used, as
substrates, to address the stereochemical/mechanistic question
of whether dioxygenase-catalysed benzylic hydroxylation
occurs with retention or inversion of configuration (Schemes 6
and 8). Biotransformation (P. putida UV4) of racemic [4-2H]-
chromane (4R/4S)-30D (Scheme 6, b) gave enantiopure deuteri-
ated (4R)-chroman-4-ol 31D (7% yield) and non-deuteriated
(4R)-chroman-4-ol 31 (4% yield) and chroman-4-one 32 (7%
yield). The slight excess of deuteriated (4R) enantiomer 31D
Addition of (Ϫ)-(4S)-chromane 30D to whole cells of
P. putida UV4 gave (ϩ)-(4R)-chroman-4-ol 31D (14% yield,
> 98% ee) with the deuterium atom retained at C-4 (> 98% 2H,
by 1H- and 2H-NMR analysis) and 4-chromanone 32 (9% yield)
(Scheme 6, c). Similarly (ϩ)-(4R)-chromane 30D gave (ϩ)-(4R)-
chroman-4-ol 31 (> 98% ee and 14% yield) with almost total
removal of the benzylic 2H atom (< 2% 2H by 1H- and 2H-NMR
analysis) and chroman-4-one 32 (9% yield) (Scheme 6, d). These
results are consistent with a mechanism involving the benzylic
hydroxylation of chromane 30 through exclusive abstraction of
the pro-R benzylic hydrogen atom followed by complete reten-
tion of configuration during insertion of the hydroxyl group.
However, if the formation and subsequent stereoselective loss
of the (Ϫ)-(4S)-enantiomer of alcohol 31 were to occur, it
would challenge this explanation since chroman-4-one 32 was
obtained as a significant metabolite (9–14% yield). In order to
reduce the role of further oxidation of chroman-4-ol 31 to
chroman-4-one 32, biotransformations with purified TDO
(P. putida 39/D) were carried out on substrates (Ϫ)-(4S)-
chromane 30D and (ϩ)-(4R)-chromane 30D (Scheme 8).
As anticipated, with a purified form of TDO, benzylic
hydroxylation of racemic (4S/4R)-, and enantiopure (4S)- and
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 2 9 8 – 1 3 0 7
1301

View Article Online
Table 1 Relative ratio of (4R)-chroman-4-ol 31/31D and chroman-2-ol
37D hydroxylation products from pure TDO- and NDO-catalysed
hydroxylation of chromane 30D
(4R)-Chroman-4-ol
31/31D
Substrate 30D
Enzyme
(R or S, % ee)
Chroman-2-ol 37D
(R/S)-30D
(S)-30D
TDO
TDO
100 (R, 72)
100 (R, 68)
100 (R, 75)a
100 (R, 47)
100 (R, 65)a
51 (S, 92)b
46 (S, 88)b
48 (S, 80)c
55 (S, 99)b
48 (S, 96)c
(R)-30D
TDO
(R/S)-30D
(S)-30D
NDO
NDO
49b
54b
52a
45b
52
(R)-30D
NDO
a Repeat experiment using a larger quantity of substrate 30D. b Accom-
panied by a minor amount (3–9%) of 2,3-dihydroxychromane 40/40D.
c Repeat experiment using a larger quantity of substrate and with little
evidence of 2,3-dihydroxychromane 40/40D formation.
Scheme 9
(4R)-chromane 30D was observed without evidence of further
oxidation to ketone 32 (Scheme 8, Table 1). In all cases
chroman-4-ol 31 or 31D was enriched (47–75% ee) in the
(ϩ)-(4R) enantiomer based on CSPHPLC analysis. This obser-
vation confirmed that a significant proportion of the ee was due
to a TDO-catalysed asymmetric hydroxylation process. GC-MS
analysis of the individual (4R)- and (4S)-enantiomers of
chroman-4-ol 31D, obtained from substrate (Ϫ)-(4S)-chromane
30D (separated by CSPHPLC), showed that the 2H atom at C-4
was retained (> 90% 2H) in each enantiomer. Conversely when
(ϩ)-(4R)-chromane 30D was added as substrate to purified
TDO, the resulting (4R)- and (4S)-enantiomers of chroman-4-
ol 31 showed loss of the 2H-atom at C-4 (> 90% H). The
experimental results obtained using purified TDO (Scheme 8)
with (Ϫ)-(4S)-chromane 30D and (ϩ)-(4R)-chromane 30D con-
firm the earlier observation using P. putida UV4 whole cells
(Scheme 6), i.e. the pro-R benzylic hydrogen atom is cleaved in a
highly stereoselective manner. The foregoing results support
the premise that the formation of benzylic hydroxylation bio-
product with a strong preference (74–87%) for the (R) enantio-
mer means that the hydroxylation has largely occurred with the
retention of configuration. A summary of results obtained
using pure TDO with racemic and enantiopure forms of
chromane 30D is presented in Table 1.
Finally, a comparative study was conducted using purified
NDO with racemic and enantiopure (4R)- and (4S)-chromane
substrates 30D (Scheme 8). GC-MS analysis indicated that in
each case chroman-4-ol 31 or 31D was formed as a major
metabolite (46–55% relative yield, Table 1). As expected from
other benzylic hydroxylations catalysed by NDO,5,34 a marked
preference (> 80%) was found for the (S)-configuration of
chroman-4-ol 31/31D. The absolute configurations of the major
chroman-4-ol enantiomers 31/31D formed with enzymes TDO
(R) and NDO (S)5–11 were found to be enantiocomplementary.
While
a single bioproduct, chroman-4-ol 31/31D, was
obtained with pure TDO, the use of pure NDO resulted in the
formation of an additional monohydroxylation product; it was
identified as chroman-2-ol 37D (45–54% relative yield) by GC-
MS comparison with an authentic sample (Scheme 9). A prece-
dent for NDO-catalysed hydroxylation on a carbon adjacent to
an oxygen atom was observed during earlier studies with
methoxybenzene and ethoxybenzene substrates. Dealkylation
of the latter substrates was found as a result of the initial
hydroxylation of the activated methyl and methylene groups
adjacent to the oxygen atom.34 A third metabolite, cis-3,4-di-
hydroxychromane 40/40D (< 10% relative yield), was also
detected by GC-MS analysis using pure NDO. This NDO bio-
product had been found earlier in the whole cell study (P. putida
9816/11). A major proportion of the original deuterium at C-4
(73% 2H from (4R)- and 72% 2H from (4S)-chromane 30D) was
retained in cis-3,4-dihydroxychromane 40/40D. The formation
of cis-diol 40/40D from chromane 30D using both whole cells
and purified NDO could be accounted for by NDO-catalysed
desaturation to yield chromene 39/39D followed by cis-di-
hydroxylation (Scheme 9). The retention of a similar major
proportion of deuterium at C-4 from both (4R)- and (4S)-
chromane 30D substrates during this desaturation process was
unexpected. The mechanism may involve a kinetic isotope effect
rather than stereoselective removal of one benzylic hydrogen
atom. The NDO enzyme system has previously been found to
catalyse the desaturation of indane 25 (R = H) to yield indene
8,5 conversion of chromane 30D to chromene 39/39D may have
occurred via a similar mechanism.
A whole cell biotransformation (P. putida 9816/11 containing
NDO) of racemic substrate [4-2H]-chromane 30D was also
carried out (Scheme 9). This NDO-catalysed hydroxylation was
mainly found to occur at C-4 to give (4S)-chroman-4-ol 31/31D
(40% relative yield) and chroman-4-one 32 (53% relative yield)
metabolites based on GC-MS and 1H-NMR analysis. (4S)-
Chroman-4-ol 31/31D was found to have lost ca. 75% of the
original deuterium atom. It could be accounted for by a
metabolic sequence involving: (i) direct enzyme-catalysed
hydroxylation of [4-2H]-chromane 30D to (4S)-chroman-4-ol
31/31D; (ii) oxidation of a portion of chroman-4-ol 31/31D to
chroman-4-one 32; and (iii) reduction of chroman-4-one 32 to
yield unlabelled chroman-4-ol 31. The relatively low ee value
found for chroman-4-ol 31/31D (24% compared with > 90% ee
obtained later using the purified NDO, Table 1) is consistent
with its reformation by reduction of chroman-4-one 32. A
minor metabolite of chromane 30D (7% relative yield) showed
similar GC-MS characteristics to an authentic sample of
cis-3,4-dihydroxychromane 40/40D. A significant proportion of
the original deuterium atom present in [4-2H]-chromane 30D
was found to be retained at C-4 in cis-3,4-dihydroxychromane
When (4S)-chromane 30D was used as substrate with purified
NDO (Scheme 8), the deuterium atom from the upper face was
2
selectively removed (> 90% H loss, equivalent to the benzylic
pro-S hydrogen atom at C-4 in chromane 30, Scheme 6). Addi-
tion of the hydroxyl radical was again found to occur mainly
(> 90%) from the upper face to yield (4S)-chroman-4-ol 31/31D
(80–88% ee, Table 1). Use of (4R)-chromane 30D led to
2
40D (64% H). Direct spectral comparison with an authentic
sample of chroman-2-ol 37D indicated that it was not present
among the monohydroxylated bioproducts.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 2 9 8 – 1 3 0 7
1302

View Article Online
Scheme 10
preferential cleavage of the benzylic hydrogen atom from
iron() hydroperoxide intermediate could function as an
the upper face (> 90% 2H retention) and hydroxyl radical addi-
tion almost exclusively (> 98%) from the same face to yield
(4S)-chroman-4-ol 31/31D (96–99% ee).
electrophilic oxidant for P-450-catalysed epoxidation and
hydroxylation.49,50 Therefore, a second alternative is that the
initial oxidant in the cis-dihydroxylation mechanism might be
an electrophilic iron hydroperoxide species. This hypothesis is
supported by the observation of an indole hydroperoxide in the
active site of naphthalene dioxygenase by X-ray crystallo-
graphy.51 The interpretation of the single turnover experi-
ments,45 and recent results showing that hydrogen peroxide
can replace the requirement for oxygen and a reduced Rieske
[2Fe–2S] cluster,52 also suggest the formation of a hydroperoxo
intermediate.
The observed monohydroxylations can be explained by
reaction of a carbon-centred radical with an activated oxygen
species, which might be either a high-valent iron–oxo inter-
mediate (Scheme 10, A), or an iron hydroperoxide species
(Scheme 10, B), which presumably is an intermediate in the
cis-dihydroxylation reaction. Recent crystallographic studies
with NDO have identified oxygen bound in a side-on position
near iron, as found in a cyclic peroxide of the type shown in
Scheme 10B.53 This species could be readily converted via a one
electron reduction process into its more reactive hydroperoxide.
Either species could be responsible for the monohydroxyl-
ation shown in Scheme 10B and for the diverse range of
mono- and poly-oxygenation reactions catalysed by the Rieske
dioxygenases.
Conclusion
The TDO-catalysed formation of benzylic monohydroxylation
products 14 and 22, using whole cells of P. putida UV4 and
indene substrates 8 and 17 respectively, is consistent with the
intermediacy of delocalised allylic/benzylic radicals. Using
TDO and NDO from both whole cell and purified enzyme
sources, benzylic hydroxylation of chromane 30 to yield
chroman-4-ol 31 was found to occur mainly with retention of
configuration (74–87% R with TDO and 90–100% S with
NDO). This was accompanied by an initial stereospecific
abstraction at a benzylic C–H bond favouring either the pro-R
(TDO) or pro-S (NDO) hydrogen atom. A similar result
(72–78% retention of configuration) was recently obtained dur-
ing the monohydroxylation of specifically labelled ethane and
butane substrates using methane monooxygenase containing a
diiron centre.35 NDO-catalysed oxidation of chromane 30 also
resulted in non-benzylic hydroxylation to yield chroman-2-ol 37
and via desaturation/cis-dihydroxylation to yield cis-3,4-di-
hydroxychromane 40.
The mechanism of dihydroxylation, the nature of the active
oxygen species, and the intermediates in the catalytic cycle are
still unknown for the cis-dihydroxylating dioxygenases. One of
the earliest mechanism studies confirmed that both atoms from
18O2 were incorporated during the cis-dihydroxylation of
arenes.44 The involvement of Rieske [2Fe–2S] clusters in cataly-
sis implies that one-electron transfers are involved, which has
been demonstrated experimentally by single turnover studies on
NDO.45 The isolation of monohydroxylated products, reported
previously6–11 and in this work, implies that the dihydroxylation
reaction occurs by stepwise insertion of two oxygen atoms.
The hydroxylation reactions catalysed by cytochrome P-450
monooxygenases46 and α-ketoglutarate-dependent dioxygen-
ases47 proceed via retention of configuration, as has been
observed in this study. For P-450 monooxygenases46 and
Experimental
1H-NMR spectra were recorded using Bruker Avance DPX-300
and DPX-500 instruments and 2H-NMR spectra using a
400 MHz Bruker WP400 instrument. Optical rotation ([α]D)
measurements were carried out with a Perkin-Elmer 214 polar-
imeter at ambient temperature (ca. 20 ЊC) and are given in units
of 10Ϫ1 deg cm2 gϪ1. Enantiopurity was determined by CSPH-
PLC using specified Chiralcel columns and hexane–2-propanol
(9 : 1) as eluent at a flow rate of 0.5 cm3 minϪ1. CD spectra were
recorded using a Jasco J-720 instrument and acetonitrile as
solvent.
High resolution MS were recorded at 70 eV on a VG
Autospec mass spectrometer using a heated inlet system.
Accurate MW values were determined by the peak matching
method with perfluorokerosene as standard. GC-MS analysis
of the concentrated crude EtOAc extract, obtained from
metabolism of chromane 30D using P. putida 9816/11, was
carried out using a Hewlett Packard 6890 gas chromato-
graph directly attached to a Hewlett Packard model 5973 Mass
α-ketoglutarate-dependent dioxygenases,47,48
a high-valent
iron–oxo species has been proposed as the active hydroxylating
agent. Thus, a possible explanation of the observed data is the
intermediacy of an iron–oxo species during cis-diol formation
(e.g. Scheme 1).
Recent work on site-directed mutants of cytochrome P-450
monooxygenase ∆2B4 have led Vaz et al. to propose that an
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 2 9 8 – 1 3 0 7
1303

View Article Online
Selective Detector. The GC oven was fitted with a wall-coated
open-tubular capillary column (12 m × 0.2 mm) with 100%
dimethyl polysiloxane (0.33 µm) as the bonded phase. The
injector port was held at 250 ЊC and the oven was programmed
3Ј-H), 5.20 (1 H, d, J1,2 5.2, 1-H), 7.11 (1 H, d, J4,5 8.0, 4-H),
7.35 (1 H, dd, J5,4 8.0, J5,7 1.8, 5-H), 7.52 (1 H, d, J7,5 1.8, 7-H).
Synthesis of 5-bromoindene 16 and 6-bromoindene 17
at 30 ЊC for 1 min and then ramped to 300 ЊC at 10 ЊC minϪ1
.
A solution of 5-bromoindan-1-ol or 6-bromoindan-1-ol (4.3 g,
20 mmol) and p-toluenesulfonic acid (ca. 0.075 g) in benzene
(150 cm3) was refluxed (2 h) and the resulting water continu-
ously removed from the reaction mixture with a Dean–Stark
trap. The cooled benzene solution was washed with water, dried
(MgSO4) and concentrated under reduced pressure; purification
of the residue by flash chromatography (hexane) yielded pure
5-bromoindene 16 or 6-bromoindene 17.
The products obtained using chromane 30D as the substrate for
the purified TDO and NDO enzymes were analysed by GC-MS
using a Hewlett Packard model 5890 gas chromatograph
equipped with an HP Ultra-1 capillary column (25 m × 0.2 mm,
0.33 µm film thickness). The column temperature was pro-
grammed from 70–240 ЊC at 10 ЊC minϪ1 with a helium flow of
25 cm3 sϪ1. Temperatures of the injection port and transfer line
were 220 and 280 ЊC, respectively. Samples (1 µl) were injected
at a split ratio of 50 : 1 and MS were obtained using an HP
model 5970 Mass Selective Detector with electron impact
ionization (70 eV).
5-Bromoindene 16. Colourless crystals (3.43 g, 87% yield), mp
39–40 ЊC (hexane) (lit.,38 mp 41 ЊC); δH(500 MHz, CDCl3) 3.35
(2 H, dd, J1,2 2.0, J1,3 1.9, 1-H), 6.59 (1 H, dt, J2,1 2.0, J2,3 5.6,
2-H), 6.81(1 H, dt, J3,2 5.6, J3,1 1.9, 3-H), 7.29–7.33 (2 H, m,
Ar-H), 7.52 (1 H, d, J4,5 1.4, Ar-H).
Substrates were metabolized (18
h unless otherwise
mentioned) using whole cell cultures of the mutant strain
Pseudomonas putida UV415 and Pseudomonas putida 9816/1129
under conditions reported earlier. The bioproducts were gener-
ally harvested by repeated solvent extraction (EtOAc) of
the sodium chloride-saturated aqueous solution containing the
biotransformed material, followed by concentration of the
6-Bromoindene 17. An oil (3.62 g, 92% yield), bp 75 ЊC/1 mm
Hg (lit.,39 bp 50 ЊC/0.1 mm Hg) (Found: Mϩ, 193.9733, C9H7Br
requires Mϩ 193.9731); δH(300 MHz, CDCl3) 3.38 (2 H, d,
J1,2 2.0, 1-H), 6.54 (1 H, dt, J1,2 2.0, J2,3 5.5, 2-H), 6.83 (1 H, m,
3-H), 7.26 (1 H, d, J4,5 8.1, 4-H), 7.41 (1 H, d, J5,4 8.1, 5-H), 7.61
(1 H, s, 7-H).
1
combined organic extracts under reduced pressure. H NMR
spectra of the crude mixture of bioproducts obtained from
each biotransformation were routinely examined prior to any
purification.
Synthesis of chromane 30
Indan-1-one 15, chroman-4-ol 31 and 5-bromoindan-1-one
20 were available commercially. Samples of chroman-2-ol 37
and cis-3,4-dihydroxychromane 40 were available from earlier
studies.33 (Ϫ)-(3R,4S)- and (ϩ)-(3S,4R)-3,4-Epoxychromane
34 were prepared from the corresponding bromo-MTPA
diastereoisomers (3S,4S)- and (3R,4R)-trans-3-bromo-4-
[methoxy(phenyl)trifluoromethylacetoxy]chromane 33 by the
reported procedure.33
p-Toluenesulfonic acid (ca. 0.030 g) and hydroquinone
(ca. 0.05 g) were added to a solution of racemic chroman-4-ol
31 (8.0 g, 53 mmol) in benzene (150 cm3). The reaction mixture
was heated under reflux using a Dean–Stark trap (2 h), washed
with water, dried (Na2SO4) and concentrated under reduced
pressure to give chrom-3-ene (5.1 g, 74%), bp 30 ЊC/0.2 mm Hg
(lit.,40 51–53 ЊC/1.2 mm Hg); δH(300 MHz, CDCl3) 4.81 (2 H, m,
2-H), 5.76 (1 H, m, 3-H), 6.41 (1 H, d, J4,3 9.6, 4-H), 6.76 (1 H,
d, J8,7 8.0, 8-H), 6.85 (1 H, dd, J6,7 8.6, J6,5 7.4, 6-H), 6.95 (1 H,
dd, J5,6 7.4, J5,7 1.5, 5-H), 7.09 (1 H, m, 7-H).
A solution of chrom-3-ene (1.06 g, 8.03 mmol) in CHCl3 (35
cm3) was stirred (12 h) in the presence of Pd/C (10%, 0.10 g)
under an atmosphere of hydrogen at ambient temperature and
pressure. The catalyst was filtered off and the filtrate concen-
trated to give the crude hydrogenated product 30. Distillation
under reduced pressure yielded chromane 30 (0.70 g, 67%), bp
50–51 ЊC/0.1 mm Hg (lit.,41 97–98 ЊC/19 mm Hg); δH (300 MHz,
CDCl3) 1.95 (2 H, m, 3-H), 2.74 (2 H, t, J4,3 6.6, 4-H), 4.13 (2 H,
t, J2,3 5.1, 2-H), 6.76 (2 H, m, 6-H and 8-H), 7.00 (2 H, m, 5-H
and 7-H).
Synthesis of indene substrates 8, 16 and 17
[3-2H]-indene 8. Sodium borodeuteride reduction of indan-1-
one 15 to yield [1-2H]-indan-1-ol followed by acid-catalysed
dehydration yielded [3-2H]-indene 8 (45% overall yield) using
the literature method.6 1H-NMR and MS analysis showed that
[3-2H]-indene 8 had a deuterium atom incorporated (> 98% 2H)
only at the C-3 position.
6-Bromoindan-1-one 23. 6-Bromoindan-1-one 23 (4.7 g, 44%)
was obtained by the literature procedure,36 mp 106–107 ЊC
(hexane) (lit.,36 108–109 ЊC); δH(300 MHz, CDCl3) 2.72 (2 H, t,
J2,3 5.8, 2-H), 3.10 (2 H, t, J3,2 5.8, 3-H), 7.37 (1 H, d, J4,5 8.1,
4-H), 7.68 (1 H, d, J5,4 8.1, J5,7 1.7, 5-H), 7.87 (1 H, d, J7,5 1.7,
7-H).
( )-[4-2H]-Chroman-3-ol 35D. To a solution of racemic
epoxide 34 (0.75 g, 5.1 mmol) in dry Et2O (15 cm3) was added
lithium aluminium deuteride (0.54 g, 13 mmol, 98 atom % 2H)
and the reaction mixture stirred (18 h) at room temperature.
The reaction was quenched by the dropwise addition of water.
Filtration and concentration of the dried (Na2SO4) ether solu-
tion yielded ( )-[4-2H]-chroman-3-ol 35D (0.70 g, 91%), mp
78–79 ЊC (from Et2O–hexane); δH (300 MHz, CDCl3) 2.09 (1 H,
br s, OH), 2.77 (1 H, br s, 4-H), 4.08 (2 H, m, 2-H), 4.23 (1 H, m,
3-H), 6.87 (2 H, m, 6-H and 8-H), 7.09 (2 H, m, 5-H and 7-H).
The deuterium incorporation was estimated to be ≥ 98% at C-4
by 1H-NMR spectral analysis.
Synthesis of 5-bromoindan-1-ol and 6-bromoindan-1-ol
Sodium borohydride (50 mmol) was added portion-wise (0.5 h)
to a solution of indanone (20 or 23, 25 mmol) in MeOH
(50 cm3) at 0 ЊC. The solution was stirred at room temperature
for a further 2 h and the solvent removed under reduced pres-
sure. The residue was extracted into CH2Cl2 (3 × 50 cm3), the
extract dried (MgSO4) and the solvent evaporated off to yield
the crude indanol.
5-Bromoindan-1-ol. White crystals (4.8 g, 90%); mp 77–79 ЊC
(from hexane) (Found: C, 50.7; H, 4.1. C9H9OBr requires C,
50.7; H, 4.3%); δH(500 MHz, CDCl3) 1.94 (1 H, m, 2Ј-H), 2.48
(1 H, m, 2-H), 2.80 (1 H, m, 3-H), 3.03 (1 H, m, 3Ј-H), 5.18
(1 H, m, 1-H), 7.26 (1 H, m, Ar-H), 7.36 (2 H, m, Ar-H).
(؊)-(3S,4R)-[4-2H]-Chroman-3-ol 35D. Reductive ring open-
ing of the (Ϫ)-(3R,4S)-epoxide 34 (1.34 g, 9.1 mmol, [α]D Ϫ71)
with lithium aluminium deuteride (0.95 g, 22.7 mmol) yielded
(Ϫ)-(3S,4R)-[4-2H]-chroman-3-ol 35D (1.26 g, 92%), mp 95–96
ЊC (from Et2O–hexane); [α]D Ϫ20.6 (c 2.0, CHCl3).
6-Bromoindan-1-ol. White crystals (4.7 g, 88%); mp 85–86 ЊC
(hexane) (lit.,37 84–86 ЊC); δH(300 MHz, CDCl3) 1.94 (1 H, m,
2-H), 2.49 (1 H, m, 2Ј-H), 2.75 (1 H, m, 3-H), 2.98 (1 H, m,
(؉)-(3R,4S )-[4-2H]-Chroman-3-ol 35D. Similar treatment of
the (ϩ)-(3S,4R)-enantiomer of epoxide 34 (1.09 g, 7.4 mmol)
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 2 9 8 – 1 3 0 7
1304

View Article Online
gave (ϩ)-(3R,4S)-[4-2H]-chroman-3-ol 35D (1.04 g, 94%), mp
95–96 ЊC; [α]D ϩ21.5 (c 2.1, CHCl3). Enantiomers of alcohol
35D were spectrally indistinguishable from the racemic sample.
The deuterium incorporation was found to be > 98% (by
1H-NMR) in both the enantiomers.
hexane) separation gave (Ϫ)-[1-2H]-(1S,2R)-1,2-dihydroxy-
indane 13D (Rf 0.1), [1-2H]- and [3-2H]-(1R)-inden-1-ol 14D/14ЈD
(Rf 0.5) and 1-indanone 15, identified by comparison with an
authentic sample (0.02 g, 4%, Rf 0.8).
(؊)-[1-2H]-(1S,2R)-1,2-Dihydroxyindane 13D. Colourless
crystals (0.11 g, 18%), mp 95–96 ЊC (from CHCl3–hexane) (lit.,5
mp 95–96 ЊC); [α]D Ϫ21 (c 0.4, CHCl3); δH (500 MHz; CDCl3)
2.95 (1H, dd, J3,2 3.7, J3,3Ј 16.3, 3-H), 3.12 (1H, dd, J3Ј,2 5.7,
J3Ј,3 16.3, 3Ј-H), 4.42 (1H, dd, J2,3 3.7, J2,3Ј 5.7, 2-H), 7.21–7.28
(3H, m, Ar-H) and 7.39–7.41 (1H, m, Ar-H); ee (ca. 38%) by
CSPHPLC analysis (Chiralcel OD column).
( )-[4-2H]-3-(p-Toluenesulfonyloxy)chromane 36D. p-Toluene-
sulfonyloxy chloride (1.17 g, 6.1 mmol) was added to a cooled
solution of [4-2H]-3-chromanol 35D (0.62 g, 4.1 mmol) in dry
pyridine (4 cm3); the reaction mixture was stirred (24 h). Pyrid-
ine was removed (reduced pressure), water (10 cm3) was added
to the residue and the reaction mixture extracted with CH2Cl2
(2 × 30 cm3). The extract was dried (Na2SO4) and concentrated
under reduced pressure to yield ( )-[4-2H]-3-(p-toluene-
sulfonyloxy)chromane 36D (0.94 g, 75%), mp 91–92 ЊC (from
Et2O–hexane) (Found: C, 62.9; H, 5.2. C16H15DO4S requires C,
63.0; H, 5.6%); δH (500 MHz, CDCl3) 2.46 (3 H, s, Me), 2.95
(1 H, br s, 4-H), 4.10 (2 H, m, 2-H), 4.95 (1 H, m, 3-H), 6.81
(1 H, d, J8,7 7.9, 8-H), 6.87 (1 H, m, 6-H), 6.96 (1 H, d, J5,6 7.1,
5-H), 7.11 (1 H, m, 7-H), 7.36 (2 H, d, J 8.0, Ar-H), 7.81 (2 H, d,
J 8.0, Ar-H).
[1-2H]- and [3-2H]-(1R)-Inden-1-ol 14D/14ЈD. Colourless
needles (0.045 g, 8%), unstable sample, mp 83–84 ЊC (from
ether–hexane) (Found: C, 81.5; H, 6.5. C9H7DO requires C
81.2; H, 6.8%); [α]D Ϫ70 (c 0.4, CHCl3) (lit.,24 [α]D Ϫ178 but
found to increase on purification to [α]D Ϫ249); δH(500 MHz;
CDCl3) 5.21 (1H, 43% D, br s, 1-H), 6.42 (1H, dd, J2,1 1.8,
J2,3 5.6, 2-H), 6.75 (1H, 57% D, d, J3,2 5.6, 3-H), 7.24 (3H, m,
Ar-H), 7.52 (1H, d, J 6.9, Ar-H); ee (ca. 28%) by CSPHPLC
analysis (Chiralcel OB column).
(؉)-(3S,4R)-[4-2H]-3-(p-Toluenesulfonyloxy)chromane 36D.
(Ϫ)-(3S,4R)-[4-2H]-3-chromanol 35D (1.30 g, 8.6 mmol) when
reacted with p-toluenesulfonyl chloride gave (ϩ)-(3S,4R)-[4-
2H]-3-(p-toluenesulfonyloxy)chromane 36D (2.00 g, 76%), mp
94–95 ЊC, [α]D ϩ 6.4 0.5 (c 1.65, CHCl3).
Biotransformation of 5-bromoindene 16
Biotransformation (P. putida UV4, 7 h) of 5-bromoindene 16
(0.150 g, 0.77 mmol) and extraction (EtOAc) of the centrifuged
culture medium yielded two products; these on separation by
PLC (60% Et2O in pentane) gave 6-bromoindan-1-one 23 (0.007
g, 4%, Rf 0.5) and (1R,2S)-5-bromoindane-1,2-diol 18 (Rf 0.1).
(؊)-(3R,4S )-[4-2H]-3-(p-Toluenesulfonyloxy)chromane 36D.
(ϩ)-(3R,4S)-[4-2H]-3-chromanol 35D (1.06 g, 7.0 mmol) on
reaction with p-toluenesulfonyl chloride gave (Ϫ)-(3R,4S)-[4-
2H]-3-(p-toluenesulfonyloxy)chromane 36D, mp 94–95 ЊC; [α]D
Ϫ6.0 0.5 (c 1.44, CHCl3).
6-Bromoindan-1-one 23. It was spectrally identical to a
sample obtained from the biotransformation of 6-bromoindene
17.
( )-[4-2H]-Chromane 30D. Lithium aluminium hydride (0.60
g, 15.8 mmol) was added to a solution of racemic [4-2H]-3-
(p-toluenesulfonyloxy)chromane 36D (0.66 g, 2.2 mmol) in dry
THF (10 cm3) and the reaction mixture heated (24 h) under
reflux. The cooled reaction mixture was filtered after decom-
position of the complex by addition of water (1.5 cm3). The
filtrate, combined with the Et2O washings, was dried (Na2SO4),
concentrated under reduced pressure and the residue purified
by flash column chromatography on silica gel (Et2O–hexane)
followed by distillation under reduced pressure to yield ( )-[4-
2H]-chromane 30D (0.21 g, 72%), bp 80 ЊC/0.02 mm Hg (Found:
Mϩ, 135.0794. C9H9DO requires 135.0794); δH (300 MHz,
CDCl3) 2.01 (2 H, m, 3-H), 2.77 (1 H, br s, 4-H), 4.19 (2 H, m,
2-H), 6.81 (2 H, m, 6-H and 8-H), 7.05 (2 H, m, 5-H and 7-H);
m/z (%) 136 (M ϩ 1ϩ, 18), 135 (Mϩ, 100), 134 (23), 120 (11), 107
(18), 79 (28). The deuterium atom incorporation (≥ 98%) at C-4
was determined by 1H- and 2H-NMR spectral analyses.
(1R,2S )-5-Bromo-1,2-dihydroxyindane 18. White crystals
(0.07 g, 42%), mp 160–161 ЊC (from CH2Cl2); [α ]D ϩ 54.1
(c 0.77, MeOH) (Found: C, 46.8; H, 3.7. C9H9BrO2 requires C
46.8; H, 3.9%); δH(500 MHz, CDCl3) 2.30 (1 H, d, JOH,1 5.2,
OH), 2.54 (1 H, d, JOH,2 7.2, OH), 2.91(1 H, dd, J3,2 3.1,
J3,3Ј 16.4, 3-H), 3.06 (1 H, dd, J3Ј,2 5.6, J3Ј,3 16.4, 3Ј-H), 4.54 (1 H,
m, 2-H), 5.02 (1 H, dd, J1,2 6.1, J1,OH 5.2, 1-H), 7.12 (1 H, d,
J4,5 8.0, 4-H), 7.40 (1 H, m, J5,4 8.0, 5-H), 7.57 (1 H, s, 7-H);
electronic CD data: λ 226 nm (∆ε Ϫ2.571), 203 nm (∆ε 13.58);
ee (ca. 46%) by 1H-NMR analysis of the MEBBA derivative.
Biotransformation of 6-bromoindene 17
Biotransformation (P. putida UV4, 7 h) of 6-bromoindene 17
(0.200 g, 1 mmol) and extraction (EtOAc) of the centrifuged
culture medium yielded two products; separation by PLC
(30% Et2O in hexane) gave the less polar (Rf 0.45) bioproduct,
6-bromoindan-1-one 23 (0.061 g, 28%) which was identified by
spectral comparison with an authentic sample. The second,
more polar (Rf 0.35) product was found to be (1S)-6-bromo-
inden-1-ol 22.
(؊)-(4S )-[4-2H]-Chromane
30D.
(ϩ)-(3S,4R)-[4-2H]-3-
(p-Toluenesulfonyloxy)chromane 36D (1.95 g, 6.4 mmol, [α]D ϩ
6.4) was treated with excess lithium aluminium hydride to yield
(4S)-[4-2H]-chromane 30D (0.64 g, 74%), bp 80 ЊC/0.02 mm Hg;
[α]D Ϫ1.7 (c 0.32, CHCl3).
(؉)-(1S )-6-Bromoinden-1-ol 22. Colourless crystals (0.035 g,
16%), mp 56 ЊC (hexane) (Found: C, 51.1; H, 3.4. C9H7BrO
requires C, 51.2; H, 3.3%); [α]D ϩ 185.3 (c 0.8, CHCl3); δH(300
MHz, CDCl3) 1.72 (1 H, d, JOH,1 8.9, OH), 5.16 (1 H, m,
J1,OH 7.4, 1-H), 6.39 (1 H, dd, J2,1 1.6, J2,3 5.6, 2-H), 6.68 (1 H, d,
J3,2 5.6, 3-H), 7.09 (1 H, d, J4,5 7.9, 4-H), 7.40 (1 H, dd, J5,4 7.9,
J5,7 1.4, 5-H), 7.64 (1 H, d, J7,5 1.4, 7-H); electronic CD data:
λ 270 nm (∆ε 0.928), 230 nm (∆ε Ϫ1.796), 204 nm (∆ε 7.779),
198 nm (∆ε 7.298); > 98% ee by CSPHPLC analysis (Chiralcel
OB column).
(؉)-(4R)-[4-2H]-Chromane
30D.
(Ϫ)-(3R,4S)-[4-2H]-3-
(p-Toluenesulfonyloxy)chromane 36D (1.45 g, 4.8 mmol, [α]D
Ϫ6.0) afforded (4R)-[4-2H]-chromane 30D (0.50 g, 78%), [α]D
ϩ1.0 (c 0.54, CHCl3). (Ϫ)-(4S)- and (ϩ)-(4R)-enantiomers of
1
[4-2H]-chromane 30D had identical H-NMR spectra (≥ 98%
2H) and mirror image CD spectra.
Biotransformation of [3-2H]-indene 8D
Biotransformation of chromane 30
Biotransformation (P. putida UV4, 8 h) of [3-2H]-indene 8D
(0.47 g, 4.1 mmol) and extraction (EtOAc) of the centrifuged
culture medium yielded three products. PLC (40% Et2O in
Biotransformation (P. putida UV4) of chromane 30 (0.150 g,
1.12 mmol) and extraction (EtOAc) of the centrifuged culture
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 2 9 8 – 1 3 0 7
1305

View Article Online
medium yielded a mixture of two bioproducts; these on separ-
ation by PLC (50% Et2O in pentane) afforded (ϩ)-(4R)-
chroman-4-ol 31 (Rf 0.28) and chroman-4-one 32 (Rf 0.5).
sulfoxidation of sulfide14 and benzylic hydroxylation of
indanone substrates.28 Biotransformations on
medium
a
scale were conducted in flasks (250 cm3) which were shaken
(150 rpm) in a water bath (28 ЊC). The reaction mixture in each
flask used for TDO or NDO biotransformations contained
30 cm3 of 2-(4-morpholino)ethanesulfonic acid (50 mmol, pH
6.8), NADH (2 mmol), ferrous ammonium sulfate (0.1 mmol),
catalase (0.005 g) and the respective reductase (0.003 g), ferre-
doxin (0.0016 g) and ISP (0.00175 g) components of TDO and
NDO. The quantities of substrates added to the pure enzyme
preparations were (ϩ)-(4R)-[4-2H]-chromane 30D (0.004 g)
and (Ϫ)-(4S)-[4-2H]-chromane 30D (0.0067 g) respectively. The
biotransformations were conducted over a 4 h period. The
metabolites were recovered by extraction (EtOAc) and concen-
tration under reduced pressure. The product mixtures were sub-
jected to GC-MS analysis and the relative ratios of products
obtained are shown in Table 1. Enantiopurity values of
chroman-4-ol 31/31D were determined, after purification (TLC),
by CSPHPLC analysis (Chiralcel OJ column).
(؉)-(4R)-Chroman-4-ol 31. Colourless crystals (0.022 g,
13%), mp 68–70 ЊC (from Et2O–pentane) (lit.,40 73–75 ЊC); [α]D
ϩ 63 (c 0.38, EtOH) (lit.,41 ϩ 67.0); δH (300 MHz, CDCl3) 1.96–
2.08 (2 H, m, 3-H), 2.18 (1 H, br s, OH), 4.22 (2 H, m, 2-H), 4.73
(1 H, m, 4-H), 6.81–6.93 (2 H, m, 6-H and 8-H), 7.16–7.29 (2 H,
m, 5-H and 7-H); m/z (Mϩ 150, 100), 149 (59); ee (ca. 95%) by
CSPHPLC analysis (Chiralcel OB column).
Chroman-4-one 32
Viscous oil (0.008 g, 5%) (lit.,40 bp 78–80 ЊC at 0.3 mm Hg);
δH (300 MHz, CDCl3) 2.81 (2 H, t, J3,2, 3-H), 4.54 (2 H, t,
J2,3 6.6, 2-H), 7.00 (2 H, m, 6-H and 8-H), 7.47 (1 H, m, 7-H),
7.90 (1 H, dd, J5,6 7.9, J5,7 1.7, 5-H); m/z (Mϩ 148, 70), 120 (100),
92 (63), 63 (19).
Biotransformation of ( )-[4-2H]-chromane 30D
Acknowledgements
Biotransformation (P. putida UV4) of racemic deuterium-
labelled chromane 30D (0.050 g, 0.37 mmol, ≥ 98 atom % H)
2
We thank the Biotechnology and Biological Sciences Research
Council (NDS), the Department of Education for Northern
Ireland (NIB, RB), the European Social Fund (JSH) and
the National Institute of General Medical Sciences Grant
GM29909 (DTG) for funding and Dr J. T. G. Hamilton for his
valuable help with GC-MS analysis.
and isolation of metabolites was carried out in a similar manner
as for substrate 30; (ϩ)-(4R)-chroman-4-ol 31/31D (0.006 g,
11%) and chroman-4-one 32 (0.004 g, 7%) were obtained.
1H-NMR spectral analysis indicated that the ratio of non-
deuteriated alcohol 31 to deuteriated alcohol 31D was 37 : 63;
ee ≥ 98% by CSPHPLC analysis.
Biotransformation of (؊)-(4S )-[4-2H]-chromane 30D
References
Biotransformation (P. putida UV4) of (Ϫ)-(4S)-[4-2H]-
1 D. T. Gibson and R. E. Parales, Curr. Opin. Biotechnol., 2000, 11,
236.
2 D. R. Boyd, N. D. Sharma and C. C. R. Allen, Curr. Opin.
Biotechnol., 2001, 12, 564.
3 T. Hudlicky, D. Gonzalez and D. T. Gibson, Aldrichimica Acta,
1999, 32, 33.
4 D. R. Boyd and G. N. Sheldrake, Nat. Prod. Rep., 1998, 15,
309.
5 S. M. Resnick, K. Lee and D. T. Gibson, J. Ind. Microbiol., 1996, 17,
438.
6 L. P. Wackett, L. D. Kwart and D. T. Gibson, Biochemistry, 1988, 27,
1360.
7 D. R. Boyd, N. D. Sharma, P. J. Stevenson, J. Chima, D. J. Gray and
H. Dalton, Tetrahedron Lett., 1991, 32, 3887.
2
chromane 30D (0.050 g, 0.37 mmol, ≥ 98 atom % H) yielded
[4-2H]-chroman-4-ol 31D (0.08 g, 14%) and chroman-4-one 32
(0.005 g, 9%). Metabolite 31D was shown to be deuteriated at
C-4 (ca. 98%) by 1H-NMR spectral analysis.1H- and 2H-NMR
spectral analysis confirmed that deuterium was located at C-4;
ee > 98% by CSPHPLC analysis.
Biotransformation of (؉)-(4R)-[4-2H]-chromane 30D
Biotransformation (P. putida UV4) of (ϩ)-(4R)-[4-2H]-
2
chromane 30D (0.05 g, 0.37 mmol, ≥ 98 atom % H) yielded
chroman-4-ol 31 (0.008 g, 14%) and chroman-4-one 32 (0.005 g,
9%). Chroman-4-ol 31 was found to be essentially non-deuteri-
ated (ca. 1% deuterium at C-4) by 1H-NMR spectral analysis; ee
>98% by CSPHPLC analysis.
8 J. M. Brand, D. L. Cruden, G. J. Zylstra and D. T. Gibson,
Appl. Environ. Microbiol., 1992, 58, 3407.
9 D. R. Boyd, N. D. Sharma, N. I. Bowers, P. A. Goodrich,
M. R. Groocock, A. J. Blacker, D. A. Clarke, T. Howard and
H. Dalton, Tetrahedron: Asymmetry, 1996, 7, 1559.
10 N. I. Bowers, D. R. Boyd, N. D. Sharma, P. A. Goodrich,
M. R. Groocock, A. J. Blacker, P. Goode and H. Dalton, J. Chem.
Soc., Perkin Trans. 1, 1999, 1453.
11 D. R. Boyd, N. D. Sharma, N. I. Bowers, J. Duffy, J. S. Harrison and
H. Dalton, J. Chem. Soc., Perkin Trans. 1, 2000, 1345.
12 C. C. R. Allen, D. R. Boyd, H. Dalton, N. D. Sharma,
S. A. Haughey, R. A. S. McMordie, B. T. McMurray, K. Sproule and
G. N. Sheldrake, J. Chem. Soc., Chem. Commun., 1995, 119.
13 K. Lee, J. M. Brand and D. T. Gibson, Biochem. Biophys. Res.
Commun., 1995, 212, 9.
14 D. R. Boyd, N. D. Sharma, S. A. Haughey, M. A. Kennedy,
B. T. McMurray, G. N. Sheldrake, C. C. R. Allen, H. Dalton and
K. Sproule, J. Chem. Soc., Perkin Trans. 1, 1998, 1929.
15 D. R. Boyd, N. D. Sharma, S. A. Haughey, J. F. Malone, A. King,
B. T. McMurray, R. Holt and H. Dalton, J. Chem. Soc., Perkin
Trans. 1, 2001, 3288.
16 D. T. Gibson, B. Gschwendt, W. K. Yeh and V. M. Kobal,
Biochemistry, 1973, 12, 1520.
17 D. R. Boyd, N. D. Sharma, M. Groocock, J. F. Malone and
H. Dalton, J. Chem. Soc., Perkin Trans. 1, 1997, 1897.
18 D. R. Boyd, N. D. Sharma, V. Ljubez, B. E. Byrne, S. D. Shepherd,
C. C. R. Allen, L. A. Kulakov, M. I. Larkin and H. Dalton,
Chem. Commun., 2002, 1914.
19 D. R. Boyd, N. D. Sharma, M. A. Kennedy, J. F. Malone, C. C. R.
Allen, S. M. Resnick and D. T. Gibson, J. Org. Chem., 1999, 64,
4005.
Biotransformation of ( )-[4-2H]-chromane 30D
Biotransformation (P. putida 9816/11, 18 h) of the racemic
deuterium-labelled chromane 30D (0.20 g, 1.48 mmol, ≥ 98%
2H) was carried out according to the reported method.43 EtOAc
extraction (2 × 30 cm3) of the centrifuged culture medium
yielded a mixture of bioproducts. GC-MS analysis, using the
earlier specified conditions, showed the following bioproducts:
substrate 30D (8.6 min), chroman-4-one 32 (10.5 min, 53%),
(Ϫ)-(4S)-chroman-4-ol 31/31D (10.8 min, 40%) and cis-3,4-di-
hydroxychromane 40/40D (12.6 min, 7%). Bioproducts 31/31D
(0.03 g, 13% isolated yield) and 32D (0.035 g, 16% isolated yield)
were separated and purified by PLC (40% Et2O in hexane).
Chromanol 31/31D, [α]D Ϫ16.0 (c 1.0, EtOH) was found to have
an ee of ca. 24% by CSPHPLC analysis. The cis-3,4-dihydroxy-
chromane 40/40D was not isolated.
Biotransformation of (؊)-(4S )-[4-2H]- and (؉)-(4R)-[4-2H]-
chromane 30D using purified TDO and NDO enzymes
Biotransformations of (Ϫ)-(4S)-[4-2H]- and (ϩ)-(4R)-[4-2H]-
chromane 30D with purified TDO and NDO enzymes were
carried out by the same procedure as was used for the
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 2 9 8 – 1 3 0 7
1306

View Article Online
20 D. R. Boyd, N. D. Sharma, J. G. Carroll, C. C. R. Allen,
D. A. Clarke and D. T. Gibson, Chem. Commun., 1999, 1345.
21 B. Kauppi, K. Lee, E. Carredano, R. E. Parales, D. T. Gibson,
H. Eklund and S. Ramaswamy, Structure, 1998, 6, 571.
22 E. Carredano, E. Karlsson, B. Kauppi, D. Chaudhury, R. E. Parales,
J. V. Parales, K. Lee, D. T. Gibson, H. Eklund and S. Ramaswamy,
J. Mol. Biol., 2000, 182, 5495.
23 K. Chen, M. Costas, J. Kim, A. K. Tipton and L. Que, J. Am. Chem.
Soc., 2002, 124, 3026.
24 D. R. Boyd, N. D. Sharma, R. Boyle, B. T. McMurray, T. A. Evans,
J. F. Malone, J. Chima, H. Dalton and G. N. Sheldrake, J. Chem.
Soc., Chem. Commun., 1993, 49.
35 A. M. Valentine, B. Wilkinson, K. E. Lieu, S. Komar-Panicucci,
N. D. Priestly, P. G. Williams, H. Morimoto, H. G. Floss and
S. J. Lippard, J. Am. Chem. Soc., 1997, 119, 1818.
36 L. A. M. Cornelius and D. W. Combs, Synth. Commun., 1994, 24,
2777.
37 W. J. Houlihan, J. H. Gogerty, V. A. Parrino and E. Ryan, J. Med.
Chem., 1983, 26, 765.
38 M. Adamczyk, D. S. Watt and D. A. Netzel, J. Org. Chem., 1984, 49,
4226.
39 M. Quere, Bull. Soc. Chim. Fr., 1971, 2983.
40 W. E. Parham and L. D. Huestis, J. Am. Chem. Soc., 1962, 84,
813.
25 D. R. Boyd, R. A. S. McMordie, N. D. Sharma, H. Dalton,
P. Williams and R. O. Jenkins, J. Chem. Soc., Chem. Commun., 1989,
339.
26 D. R. Boyd, N. D. Sharma, I. N. Brannigan, D. A. Clarke,
H. Dalton, S. A. Haughey and J. F. Malone, Chem. Commun., 1996,
2361.
27 D. R. Boyd, N. D. Sharma, S. A. Haughey, J. F. Malone,
B. T. McMurray, G. N. Sheldrake, C. C. R. Allen and H. Dalton,
Chem. Commun., 1996, 2363.
41 J. Gripenberg and H. Tapio, Acta Chem. Scand., 1966, 20, 1561.
42 W. D. Still, Jr. and D. J. Goldsmith, J. Org. Chem., 1970, 35, 2282.
43 D. R. Boyd, N. D. Sharma, L. V. Modyanova, J. G. Carroll,
J. F. Malone, C. C. R. Allen, J. T. G. Hamilton, D. T. Gibson,
R. E. Parales and H. Dalton, Can. J. Chem., 2002, 80, 589.
44 D. T. Gibson, G. E. Cardini, F. C. Maseles and R. E. Kallio,
Biochemistry, 1970, 9, 1631.
45 M. D. Wolfe, J. V. Parales, D. T. Gibson and J. D. Lipscomb, J. Biol.
Chem., 2001, 276, 1945.
28 S. M. Resnick, D. S. Torok and D. T. Gibson, J. Org. Chem., 1995,
60, 3546.
29 D. T. Gibson, S. M. Resnick, K. Lee, J. M. Brand, D. S. Torok,
L. P. Wackett, M. J. Schocken and B. E. Haigler, J. Bacteriol., 1995,
177, 2615.
30 S. M. Resnick, D. S. Torok, K. Lee, J. M. Brand and D. T. Gibson,
Appl. Environ. Microbiol., 1994, 60, 3323.
31 D. Burdi and T. P. Begley, J. Am. Chem. Soc., 1991, 113, 776.
32 D. R. Boyd, N. D. Sharma, R. Boyle, J. F. Malone, J. Chima and
H. Dalton, Tetrahedron: Asymmetry, 1993, 4, 1307.
33 D. R. Boyd, N. D. Sharma, R. Boyle, T. A. Evans, J. F. Malone,
K. M. McCombe, H. Dalton and J. Chima, J. Chem. Soc.,
Perkin Trans. 1, 1996, 1757.
46 J. T. Groves, G. A. McClusky, R. E. White and M. J. Coon, Biochem.
Biophys. Res. Commun., 1978, 81, 154.
47 C. J. Schofield and Z. Zhang, Curr. Opin. Chem. Biol., 1999, 9, 722.
48 B. Siegel, Bio-Org. Chem., 1979, 8, 2129.
49 A. D. N. Vaz, S. J. Pernecky, G. M. Raner and M. J. Coon, Proc.
Natl. Acad. Sci. USA, 1996, 93, 4644.
50 A. D. N. Vaz, D. F. McGinnity and M. J. Coon, Proc. Natl. Acad.
Sci. USA, 1998, 95, 3555.
51 E. Carredano, A. Karlsson, B. Kauppi, D. Choudhury, R. E.
Parales, J. V. Parales, K. Lee, D. T. Gibson, H. Eklund and
S. Ramaswamy, J. Mol. Biol., 2000, 296, 701.
52 M. D. Wolfe and J. D. Lipscomb, J. Biol. Chem., 2003, 278, 829.
53 A. Karlsson, J. V. Parales, R. E. Parales, D. T. Gibson, H. Eklund
and S. Ramaswamy, Science, 2003, 299, 1039.
34 S. M. Resnick and D. T. Gibson, Biodegradation, 1993, 4, 195.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 2 9 8 – 1 3 0 7
1307