Biocatalytic asymmetric dihydroxylation of conjugated mono-and poly-alkenes to yield enantiopure cyclic cis-diols

Article abstract of DOI:10.1002/adsc.200505033

Dioxygenase-catalysed asymmetric dihydroxylation, of a series of conjugated monoalkenes and polyenes, was found to yield the corresponding monols and 1,2-dihydrodiols. The diol metabolites were obtained from monosubstituted, gem-disubstituted, cis-disubstituted, and trisubstituted alkene substrates, using whole cells of Pseudomonas putida strains containing toluene and naphthalene dioxygenases. Dioxygenase selection and alkene type were established as important factors, in the preference for dioxygenase-catalysed 1,2-dihydroxylation of conjugated alkene or arene groups, and monohydroxylation at benzylic or allylic centres. Competition from allylic hydroxylation of methyl groups was observed only when naphthalene dioxygenase was used as biocatalyst. The structures, enantiomeric excess values and absolute configurations of the bioproducts, were determined by a combination of stereochemical correlation, spectroscopy (NMR and CD) and X-ray diffraction methods. cis-1,2-Diol metabolites from arenes, cyclic alkenes and dienes were generally observed to be enantiopure (> 98% ee), while 1,2-diols from acyclic alkenes had lower enantiomeric excess values (< 88% ee). The enantiopure cis-diol metabolite of a gem-disubstituted fulvene was used as precursor in a new chemoenzymatic route to a novel C₂-symmetrical ketone.

Full text of DOI:10.1002/adsc.200505033

FULL PAPERS
Biocatalytic Asymmetric Dihydroxylation of Conjugated Mono-
and Poly-alkenes to Yield Enantiopure Cyclic cis-Diols
a, b,
a, b
b
b
Derek R. Boyd, * Narain D. Sharma, Nigel I. Bowers, Ian N. Brannigan,
b
b
b
Melanie R. Groocock, John F. Malone, Gareth McConville,
Christopher C. R. Allen
c
a
Centre for the Theory and Application of Catalysis, The Queenꢀs University of Belfast, Belfast, BT95AG, UK
Fax: (þ44)-232-9097-4687, e-mail: dr.boyd@qub.ac.uk
b
c
School of Chemistry, The Queenꢀs University of Belfast, Belfast, BT95AG, UK
QUESTOR Centre, The Queenꢀs University of Belfast, Belfast, BT95AG, UK
Received: January 21, 2005: Accepted: April 1, 2005
Supporting Information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: Dioxygenase-catalysed asymmetric dihy- absolute configurations of the bioproducts, were de-
droxylation, of a series of conjugated monoalkenes termined by a combination of stereochemical correla-
and polyenes, was found to yield the corresponding tion, spectroscopy (NMR and CD) and X-ray diffrac-
monols and 1,2-dihydrodiols. The diol metabolites tion methods. cis-1,2-Diol metabolites from arenes,
were obtained from monosubstituted, gem-disubsti- cyclic alkenes and dienes were generally observed to
tuted, cis-disubstituted, and trisubstituted alkene sub- be enantiopure (>98% ee), while 1,2-diols from acy-
strates, using whole cells of Pseudomonas putida clic alkenes had lower enantiomeric excess values
strains containing toluene and naphthalene dioxyge- (<88% ee). The enantiopure cis-diol metabolite of
nases. Dioxygenase selection and alkene type were es- a gem-disubstituted fulvene was used as precursor in
tablished as important factors, in the preference for a new chemoenzymatic route to a novel C -symmetri-
2
dioxygenase-catalysed 1,2-dihydroxylation of conju- cal ketone.
gated alkene or arene groups, and monohydroxylation
at benzylic or allylic centres. Competition from allylic
hydroxylation of methyl groups was observed only Keywords: allylic/benzylic hydroxylation; asymmetric
when naphthalene dioxygenase was used as biocata- alkene/arene dihydroxylation; chemoenzymatic syn-
lyst. The structures, enantiomeric excess values and thesis; dioxygenases
[
5]
Introduction
arenes. Similar guidelines have not yet been establish-
ed for dioxygenase-catalysed 1,2-dihydroxylations of all
Ring hydroxylating dioxygenase enzymes, present in possible substituted alkene types. Thus, it is difficult to
mutant strains of the soil bacterium Pseudomonas puti- predict the regioselectivity and stereoselectivity of the
da, have been found to catalyse the formation of a wide biocatalytic asymmetric dihydroxylation reactions of al-
range of enantiopure cis-dihydrodiol metabolites, from kene substrates.
the corresponding arene substrates. To date, more than
It should be emphasised that non-conjugated acyclic
three hundred arene cis-dihydrodiols have been ob- alkenes are generally very poor substrates for dioxyge-
tained using toluene dioxygenase as biocatalyst, and nase-catalysed 1,2-dihydroxylation, and few examples
[
24,25]
these have been extensively used as chiral precursors of this type have been reported.
in synthesis.
However, the asym-
Benzene dioxygenase (BDO), toluene metric 1,2-dihydroxylation of several acyclic substrates
dioxygenase (TDO) and naphthalene dioxygenase bearing a monosubstituted alkene group has been ob-
[1–8]
[
13–15]
(
NDO) enzymes have also been reported to catalyse served.
It is noteworthy that in these examples
the alkene was conjugated to an aryl group, e.g., oxida-
[
9–23]
the dihydroxylation of a small number of alkenes.
The considerable potential of enantiopure alkene bio- tion of chlorostyrene 1c to yield arene-cis-dihydrodiol
[
14]
products as synthetic precursors has yet to be exploited. 2c and alkene-1,2-diol 3c (Scheme 1).
Simple models and guidelines are now available to Similar studies have been reported on the dioxyge-
predict the regio- and stereochemical outcome of dioxy- nase-catalysed 1,2-dihydroxylation of the benzo-fused
genase-catalysed cis-dihydroxylation of substituted cyclic alkenes 6a–c to give the corresponding 1,2-diols
Adv. Synth. Catal. 2005, 347, 1081–1089
DOI: 10.1002/adsc.200505033
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1081

FULL PAPERS
Derek R. Boyd et al.
8a–c, but without any evidence of the corresponding
arene-cis-dihydrodiols.
The dioxygenase-catalysed 1,2-dihydroxylation of
conjugated dienes, in common with acyclic alkene sub-
strates, has to date also received little attention. Iso-
prene, a typical example of a conjugated diene, is freely
liberated into the environment from plants and, in view
of the large quantities produced, exerts an important in-
fluence on atmospheric chemistry. This observation
prompted our recent study of the BDO-catalysed 1,2-di-
hydroxylation of acyclic C and C dienes including bu-
4
5
tadiene and the methyl-substituted derivatives isoprene,
cis-piperylene and trans-piperylene using P. putida
[
22]
ML2. Our preliminary results from the dioxygenase-
catalysed 1,2-dihydroxylation of several conjugated cy-
Scheme 1. TDO- and NDO-catalysed dihydroxylation of
monosubstituted alkenes 1a–g.
[21]
clic C -C dienes and C trienes, have been reported.
5
8
7
Chemical methods are currently available for catalytic
asymmetric dihydroxylation of all six alkene types
7
a–c of opposite absolute configurations (enantiocom- (mono-, gem-di-, cis-di-, trans-di-, tri- and tetrasubstitut-
[10–12,18,19] [26]
plementarity) usingTDOandNDO (Scheme 2).
ed alkenes). Enantomeric excess (ee) values of diols,
Alkenediols 7a–c were also accompanied by enantio- synthesised by using the most commonly employed
complementary benzylic monohydroxylation products chiral dihydroxylating agents, e.g., K OsO (OH) ,
2
4
4
K Fe(CN) , (DHQ) PHAL (AD-mix-b), are generally
3
6
2
Scheme 2. TDO- and NDO-catalysed mono- and di-hydroxylation of alkenes 6a–c to yield cis-diols 7a–c and monoalcohols
a–c.
8
1082
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1081–1089

Biocatalytic Asymmetric Dihydroxylation of Conjugated Mono- and Polyalkenes
FULL PAPERS
high (>95% ee) for five of the alkene types; the degree tida UV4 (a mutant strain containing TDO without diol
of stereoselectivity for the sixth, cis-disubstituted al- dehydrogenase activity), and P. putida NCIMB 8859 (a
[
26]
kenes, is lower (20–80% ee). During this study, par- wild-type strain containing NDO and a diol dehydro-
ticular emphasis has been placed on: (i) a systematic genase) (Scheme 1). Substrates 1b–d were selected on
evaluation of the regioselective and stereoselective po- the basis of our earlier observations that meta-disubsti-
tential of dioxygenase-catalysed 1,2-dioxygenation tuted benzenes were, generally, poorer substrates for
when challenged with substrates containing all six al- TDO-catalysed arene cis-dihydroxylation, in compari-
kene types (ii) establishing general trends that would fa- son to ortho- and para-disubstituted benzenes. It was an-
cilitate prediction of the facial and regiochemical as- ticipated that this factor might favour the formation of
pects of dioxygenase-catalysed 1,2-dihydroxylation of monosubstituted alkene-1,2-diols 3b–d over arene-cis-
aryl-substituted alkenes 1a–g, 9 (Schemes 1 and 3) dihydrodiols 2b–d. The other styrene substrates 1e–g
and conjugated cyclic dienes 14a–i (Scheme 4) and were chosen as typical examples of gem-disubstituted
(
iii) demonstrating the value of dioxygenase-catalysed 1e, cis-disubstituted 1f and trans-disubstituted 1g al-
asymmetric dihydroxylation of cis-disubstituted al- kenes.
kenes, in comparison with chemical asymmetric dihy-
Each of the styrene substrates 1a–g was biotrans-
droxylation, by using the enantiopure alkene 1,2-dihy- formed under the catalytic influence of TDO into the
drodiol bioproduct 15f as precursor in a chemoenzymat- corresponding (1S,2R)-arene-cis-dihydrodiols 2a–g
ic synthesis of the useful C -symmetrical ketone 21 (Table 1). Arene-cis-diols 2a–d were accompanied by
2
(
Scheme 5).
the corresponding alkene-1,2-diols 3a–d. After PLC
separation and purification, the ee values and absolute
configurations of the cis-dihydrodiols 2a–g (2–42%
yield) were determined using both (R) and (S) enan-
tiomers of 2-(1-methoxyethyl)-phenylboronic acids
Results and Discussion
Non-conjugated monoalkenes appear to be very poor (MEPBAs) to form the corresponding diastereoisomer-
[
27–30]
substrates for bacterial dioxygenase-catalysed 1,2-dihy- ic boronates according to the reported method
droxylation, based on the few literature reports availa- by comparison of CD spectra. The cis-dihydrodiol bio-
and
[24,25]
ble.
This premise is supported by our unpublished products 2a–g were all found to be enantiopure
observations of very low isolated yields of the corre- (>98% ee) and compounds 2a, 2e–g showed the typical
sponding alkene-1,2-diols from as series of non-conju- CD spectra associated with a (1S,2R) absolute con-
gated acyclic alkene (e.g., mono-, di- and trisubstituted figuration (l¼218–235 nm [Deꢀve], 280–282 nm
allylbenzenes) and cyclic alkene substrates (e.g., cyclo- [Deþve]); in accord with expectations from earlier bio-
hexene and 1,4-dihydronaphthalene) in the presence transformations of monosubstituted and meta-disubsti-
[14]
of TDO biocatalyst. The remit of the present study, to tuted benzene substrates. The earlier report
of
elucidate factors that influence dioxygenase-catalysed TDO-catalysed dihydroxylation (P. putida 39/D) of
alkene mono- and dihydroxylation versus other oxida- meta-chlorostyrene 1c to yield cis-dihydrodiol 2c with
tion types, was therefore, confined to conjugated al- a lower enantiopurity value (1% yield, 54% ee), was
kenes.
Styrene 1a and a series of meta-substituted styrene
not reproduced during this study using P. putida UV4.
While the monosubstituted (1R)-alkenediols 3a–d
substrates 1b–d were added to whole cells of both P. pu- were consistently formed (3–22% yield) using TDO as
Table 1. Arene- (2a–g) and alkene (3a–f)-diol metabolites of styrenes (1a–g).
[a, b]
[b]
[c]
Alkene
Arene-cis-1,2 dihydrodiol
Alkene-1,2-diol ( )
[
b]
[d]
[d]
Yield
Config.
Yield
ee
Config.
1
1
1
1
1
1
1
a
b
c
d
e
f
2a
2b
2c
2d
2e
2f
32
27
2
22
24
42
37
1S,2R
1S,2R
1S,2R
1S,2R
1S,2R
1S,2R
1S,2R
3a
3b
3c
3d
3e, 4
3f
3 (60)
22 (14)
14 (18)
8 (12)
88 (80)
62 (62)
42 (56)
48 (56)
(46)
1R
1R
1R
1R
1R
1R,2S
–
[e]
(20), (12)
(15)[
(82)
f]
g
2g
5
(52)
–
[
[
[
[
[
[
a]
>
98% ee.
b]
Using TDO.
c]
d]
e]
f]
(
Using NDO).
Absolute configuration.
Yield of allylic alcohol 4.
Yield of a,b-unsaturated carboxylic acid 5.
Adv. Synth. Catal. 2005, 347, 1081–1089
asc.wiley-vch.de
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1083

FULL PAPERS
Derek R. Boyd et al.
biocatalyst, no evidence was found for the alkene-1,2- sions could be drawn. The present study of TDO and
diol metabolites 3e–g from the corresponding disubsti- NDO-catalysed 1,2-dihydroxylation of styrene sub-
tuted alkenes 1e–g under these conditions. The ee val- strates 1a–g, supported the earlier observed trend that
ues and absolute configurations of the alkene-1,2-diol dioxygenase-catalysed 1,2-dihydroxylation occurred
bioproducts 3a–d were determined by: (i) comparison more readily in the sequence: monosubstituted al-
of optical rotation values with those reported in the lit- kene>cis-disubstituted alkene>gem-disubstituted al-
erature and (ii) formation of the corresponding diaster- kene>trans-disubstituted alkene.
eoisomeric di-MTPA esters using (R) and (S) forms of 2-
In view of the total preference shown for TDO-cata-
methoxy-2-phenyltrifluoromethylacetyl chloride, fol- lysed cis-dihydroxylation of the phenyl ring over the dis-
1
lowed by their H NMR spectral analysis. Based on ubstituted alkene group in styrene substrates, 1e–g
this method, the alkene-1,2-diols 3a–d were all found (Scheme 1), neither trisubstituted nor tetrasubstituted
to have an excess of the (1R) enantiomer (42–88% alkene analogues of these compounds were considered
[
14]
ee). The earlier observation that 1,2-diol 3c (2% yield, as appropriate substrates. A precedent for 1,2-dihydrox-
5% ee), a metabolite of meta-chlorostyrene 1c using P. ylation of an acyclic trisubstituted alkene bond was
9
putida 39D (TDO), had an excess of the (1S) configura- found when isoprene was oxidised using BDO as bioca-
tion, is at variance with the consistent (1R) trend found talyst, although this was only a minor pathway relative to
using P. putida UV4.
NDO had previously been found to catalyse the pro- bond.
1,2-dihydroxylation of the monosubstituted alkene
[
22]
duction of cis-dihydrodiols of naphthalene, and biphen-
Benzo-fused cyclic alkenes, e.g., 6a–c, in the presence
[
9]
yl (a monosubstituted benzene). One disadvantage of of TDO and NDO, had been known to give the corre-
NDO was its inability to catalyse the formation of cis-di- sponding 1,2-diols without arene cis-dihydroxylation
hydrodiols from any monosubstituted benzene sub- (Scheme 2).Thus, as a logical extension of our earlier
[
9]
strates except for biphenyl. However, this limitation studies of the biotransformation of substituted in-
[
23]
[31]
was turned to advantage, when styrene substrates denes and indanes, the methyl-substituted deriva-
a–g were used in the present study with NDO. Thus, tives of bicyclic alkenes 6a (2-methyl-, 3-methyl- and
1
in the absence of arene-cis-dihydroxylation, the alkene- 2,3-dimethylindenes) and 6b (3-methyl- and 4-methyl-
diol bioproducts 3a–f were formed exclusively (12– 1,2-dihydronaphthalenes) were used as substrates with
60% yield, Table 1), thereby avoiding any separation TDO and NDO. Unfortunately only one of the sub-
problem of alkene- and arene-1,2-diols. Furthermore, strates, 2-methylindene 9, which contains a cyclic trisub-
both TDO and NDO enzyme systems yielded alkene- stituted alkene bond, was found to give the correspond-
1
,2-diols 3a–f with an excess (42–88% ee) of the (1R) ing alkene 1,2-diol metabolite (using P. putida UV4,
absolute configuration, but failed to catalyse the forma- TDO, Scheme 3).
tion of the 1,2-diol from trans-alkene 1g.
Asymmetric dihydroxylation of 2-methylindene 9
When a-methylstyrene 1e, was added as substrate to thus occurred to give the cis-diol 10 (26% yield,
whole cells of P. putida NCIMB 8859, the expected al- Scheme 3). The diol 10 was found to be enantiopure
1
kene-1,2-diol 3e (20% yield), and an unusual allylic and of the (1S,2R) configuration, based on the H
monohydroxylation product 4 (12% yield) were isolat- NMR spectral analysis of the corresponding di-MTPA
ed. It was interesting to note that trans-b-methylstyrene esters. The major bioproduct (57% yield) was identified
1g showed no evidence of 1,2-dihydroxylation and in- as the racemic 2-methyl-1-indanone 12, which was as-
stead gave the a,b-unsaturated carboxylic acid 5 (52% sumed to have been formed during the biotransforma-
yield) as the sole metabolite, presumably as a result of tion or isolation/purification phases, via a facile isomer-
initial allylic hydroxylation of the methyl group fol- isation of the unstable benzylic alcohol metabolite 11
lowed by further oxidation. NDO-catalysed allylic hy- followed by racemisation of the indanone product 12.
[
8,9]
droxylation appears to be without precedent. Howev- This was confirmed by the isolation of 2-methylindenol
er, allylic hydroxylation of substrates 1e and 1g in the 11 as an unstable minor metabolite (1% yield) which
presence of NDO was observed to be the preferred met- readily isomerised to ketone 12. A repeat experiment
abolic pathway, when the rate of 1,2-dihydroxylation again gave ketone 12 with a slight enrichment (7% ee)
was either slowed (e.g., on the gem-disubstituted alkene of the (2R) enantiomer consistent with a partial racemi-
1
1
e) or halted (e.g., on the arene rings of styrenes 1e and sation process. It was possible to form an MTPA ester of
g and the trans-alkene group of styrene 1g). indenol 11 for determination of its enantiopurity
The yields of acyclic alkene-1,2-diols, obtained from a (>98% ee). A (2R) configuration was assigned to indenol
earlier biotransformation study of methyl-substituted 11 by analysis of its CD spectrum. It is noteworthy that
butadienes with a wild-type strain (ML2) of P. putida, both the mono- and dihydroxylation products 11 and
were used as a crude indicator of the relative rates of di- 10, obtained using TDO, had identical absolute configu-
oxygenase-catalysed dihydroxylation of four alkene rations to those obtained earlier (8a–c and 7a–c) using
[22]
types. Due to further biotransformations of the 1,2- the same biocatalyst and the corresponding unsubstitut-
[
10,12,13,18,19]
diol bioproducts also occurring, only tentative conclu- ed benzocycloalkenes 6a–c (Scheme 2).
1084
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1081–1089

Biocatalytic Asymmetric Dihydroxylation of Conjugated Mono- and Polyalkenes
FULL PAPERS
Scheme 3. TDO-and NDO-catalysed mono- and dihydroxylation of 2-methylindene 9.
When the biotransformation of 2-methylindene 9 was of NDO-catalysed 1,2-dihydroxylation of either the al-
repeated using P putida NCIMB 8859 (NDO) , benzylic kene or arene groups in substrate 9, it is evident that
hydroxylation was found to be dominant and (1S)-2- monohydroxylation is preferred at the cyclic benzylic/
methylinden-1-ol 11 was isolated in good yield (67%,
allylic (C-1) centre rather than at the acyclic allylic
>
98% ee) without isomerisation to ketone 12. Isomeri- (Me) centre.
sation of the enantiopure compound, under mild basic
The last part of the study was focused on the biotrans-
conditions, yielded ketone 12 which was again found formation of four cyclic dienes 14a–d and five trienes
to have racemised spontaneously. Allylic hydroxylation 14e–i, using TDO and NDO as biocatalysts (Scheme 4,
of the methyl group was also observed to give 2-hydroxy- Table 2). Enantiomerically enriched samples (5–38%
methylindene 13 as a minor metabolite (5% yield). ee) of the alkene-1,2-diols 15a–f had been obtained ear-
When considered with the earlier NDO metabolites 4 lier from the corresponding dienes 14a–f, using AD-
[
26,32]
and 5, bioproduct 13 provided a third example of mix-[DHQD) -PHAL].
The absolute configura-
2
NDO-catalysed allylic hydroxylation. In the absence tions of cis-diols 15a–f were tentatively assigned on
the basis of the mnemonic device proposed by Sharp-
[
26,32]
less.
In a preliminary report of the present work, us-
[21]
ing TDO as biocatalyst, it was shown that several of
these cyclicpolyenesyielded the corresponding enantio-
pure cis-diols; the absolute configurations of these were
[32]
based solely on the earlier tentative assignments. As
the preliminary results indicated a marked increase in
the ee values of alkene 1,2-diols 14b, 14c,14f and 14i, ob-
tained by the biocatalytic route over the chemical meth-
[
32]
od, both these and other polyene substrates were in-
vestigated further using TDO and NDO.
Cyclopentadiene 14a, was biotransformed, by P. puti-
Scheme 4. TDO-catalysed 1,2-dihydroxylation of dienes da strains containing either TDO or NDO, to give the
14a–i.
corresponding cis-diol metabolite 15a with a relatively
Table 2. cis-Diol metabolites (15a–i) of dienes (14a–d) and trienes (14e–i).
[a]
[b]
[b]
Polyene
cis-Diol
Yield [%]
Config.
Polyene
cis-Diol
Yield [%]
Config.
[
c]
[d]
14a
14b
14c
14d
14e
15a
15b
15c
15d
15e
32[ (14)
1R,2S
1R,2S
1R,2S
1R,2S
1R,2S
14e
14f
14g
14h
14i
15e’
15f
15g
15h
15i
15 (0)
19 (7)
7
5
5
e]
12 (8)
1R,2S
1R,2S
1R,2S
1R,2S
20 (24)
4 (4)
29 (24)
[
[
[
[
[
a]
Using TDO; (using NDO).
b]
c]
d]
e]
>
98% ee.
2
4
0% ee.
0% ee.
cis-1,2-Dihydroxycyclohexa-3,5-diene also formed (10% yield).
Adv. Synth. Catal. 2005, 347, 1081–1089
asc.wiley-vch.de
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1085

FULL PAPERS
Derek R. Boyd et al.
low excess of the (1R,2S) enantiomer (20% and 40% ee,
The fulvenes 14f–i, containing both cis-disubstituted
respectively). The other monocyclic diene substrates and tetrasubstituted alkene groups, presented the
1
4b–d similarly yielded the corresponding cis-diols TDO enzyme with an opportunity to differentiate be-
1
5b–d, but these were found to be enantiopure, based tween alkene bond types and prochiral faces. The corre-
1
on H NMR spectral analyses of their diastereoisomeric sponding single enantiomer cis-dihydrodiols 15f–i were
MEPBA boronates and di-MTPA esters. The latter de- formed exclusively (5–19% yield). Thus, the TDO en-
rivatives provided two independent methods for con- zyme showed a marked preference for dihydroxylation
firming the (1R,2S) configuration tentatively assigned of a cis-disubstituted alkene over a tetrasubstituted al-
[32]
earlier.
kene, in accord with expectations based onthe earlier re-
In addition to the expected cis-1,2-dihydrodiol 15b sults obtained using indene and 2,3-dimethylindene sub-
12% yield), TDO-catalysed oxidation of 1,3-cyclohex- strates.
(
adiene 14b also gave cis-1,2-dihydroxycyclohexa-3,5-di-
As the assignment of absolute configuration of the ful-
ene 16 (10% yield). Diol 16 resulted from a TDO-cata- vene-cis-diol metabolite 15f, derived from the Sharpless
[
26]
lysed desaturation of 1,3-cyclohexadiene 14b to yield mnemonic device method,
was described as tenta-
[
32]
benzene followed by its cis-dihydroxylation. An earlier tive, an X-ray crystal structure analysis of the crystal-
precedent for this TDO-catalysed desaturation/cis-di- line dicamphanate derivative 17 was used to provide an
hydroxylation sequence was the formation of cis-1,2-di- unequivocal determination of absolute configuration,
hydroxy-1,2-dihydronaphthalene from 1,2-dihydro- relative to the known (1S) configuration of the campha-
nate groups (Figure 1). This analysis confirmed the
The ability of the TDO system to stereodifferentiate (1R,2S) absolute configuration originally proposed
[18]
naphthalene.
[26]
between different alkene faces and groups was evaluat- and thus provided an anchor structure for the other ful-
ed with the help of the triene substrates 14e–i. Biotrans-
formation of cycloheptatriene 14e with P. putida UV4
resulted in TDO-catalysed dihydroxylation at two dif-
ferent cis-disubstituted alkene bonds to yield an insepa-
rable mixture of chiral diol 15e (29% yield) and achiral
diol 15e’ (15% yield). The former chiral diol contained a
conjugated diene group andreactedwith 4-phenyl-1,2,4-
triazoline-3,5-dione to yield a cycloadduct, while diol
1
5e’ remained unaffected. Formation of di-MTPA ester
diastereoisomers from the cycloadduct confirmed that
diol 15e was enantiopure (>98% ee). NDO-catalysed
cis-dihydroxylation of cycloheptatriene 14e using P. pu-
tida NCIMB 8859 gave the enantiopure chiral diene 15e
as the sole metabolite of identical (1R,2S) configuration
to that isolated using TDO. The lack of NDO-catalysed
allylic hydroxylation products from any of the cyclic di-
ene substrates 14a–i despite the availability of doubly
activated allylic carbon atoms in two substrates (14a
and 14e), indicated that this type of oxidation was par-
ticularly slow at cyclic allylic centres compared with
1,2-dihydroxylation of an alkene bond.
Figure 1. X-ray crystal structure of compound 17.
1086
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1081–1089

Biocatalytic Asymmetric Dihydroxylation of Conjugated Mono- and Polyalkenes
FULL PAPERS
venediols 15g–i. A similar (1R, 2S) configuration was, Conclusion
therefore, assigned to the other fulvene-cis-diol metab-
olites 15g–i, by comparison of their CD spectra (l¼ Examples of dioxygenase-catalysed asymmetric 1,2-di-
2
05–210 nm [Deꢀve], 238–242 nm [Deþve]). Further hydroxylation of monosubstituted (1a–d; with TDO
evidence of identical absolute configurations and enan- and NDO), gem-disubstituted (1e; with NDO), cis-dis-
tiopurity values (>98% ee) for each of the fulvenediols ubstituted (1f, 14a–i; with NDO) and trisubstituted al-
1
1
5f–i was obtained from H NMR analysis of the corre- kenes (9; with TDO) have been found during the current
sponding boronates using (R) and (S)-MEPBA.
study. Neither the trans-disubstituted alkene 1g nor the
The final phase of the study was linked to our current tetrasubstituted alkene bonds in fulvenes 14f–i were di-
interest in using dioxygenase bioproducts for the syn- hydroxylated using TDO or NDO. From an estimate of
thesis of chiral ligands and auxiliaries. The (1R,2S)- the product yields obtained, the relative ease of 1,2-di-
diol metabolite 15f of dimethylfulvene 14f was em- hydroxylation of the six alkene types, using BDO,
ployed as a precursor of the potentially useful chiral ke- TDO and NDO, appears to be in the order: mono-substi-
tone 21 that was required in these and other laboratories tuted alkene>cis-disubstituted alkene>gem-disubsti-
[
33]
(
Scheme 5). (1R,2S)-Diol metabolite 15f was protect- tuted
ed as the acetonide derivative 18 prior to subjecting it to ubstituted>tetrasubstituted alkene.
Competing dioxygenase-catalysed arene dihydroxy-
alkene>trisubstituted
alkene>trans-dis-
the reaction sequence: osmylation (18 ! 19), bis-aceto-
nide formation (19 ! 20) and ozonolysis (20 ! 21). lations often occurred simultaneously with monosubsti-
The chemical dihydroxylation of the cis-disubstituted al- tuted alkene 1,2-dihydroxylation, e.g., TDO-catalysed
kene bond of acetonide 18 to yield diol 19, in common arene-cis-dihydroxylation was also observed with acy-
with the TDO-catalysed dihydroxylation of the cyclic al- clic alkenes bearing an aryl substituent 1a–g. With a cy-
kene bond in fulvene 14f to yield diol 15f, occurred with- clic trisubstituted alkene (e.g., 9) competition from mon-
out evidence of dihydroxylation of the corresponding ohydroxylation (benzylic and allylic) was observed
tetrasubstituted alkene bond.
without any evidence of arene dihydroxylation. Allylic
This chemoenzymatic approach to ketone 21 hydroxylation of both a gem-disubstituted alkene (e.g.,
(
Scheme 5) has advantages over the earlier chemical 1e) and a trans-disubstituted alkene (e.g., 1g) was again
[33]
method
from n-heptylfulvene in terms of the higher observed using NDO. The TDO enzyme was assumed to
stereoselectivity during the biocatalytic asymmetric di- be responsible for the competing tandem desaturation
hydroxylation (>98% ee compared with 78% ee using of 1,2-cyclohexadiene 14b and cis-dihydroxylation of
AD-mix) and regioselectivity for the cis-disubstituted the resulting benzene intermediate to yield cis-dihydro-
alkene bond during the chemical dihydroxylation step diol 16.
(
>98% compared with 48%, at the cyclic disubstituted
As shown herein, dioxygenase selection has a marked
effect on the proportion of alkene 1,2-dihydroxylation
relative to alternative oxidation pathways. The locations
of substituents on the alkene substrate also exert a
alkene bond).
Scheme 5. Chemoenzymatic synthesis of chiral ketone 21 from fulvene 14f.
Adv. Synth. Catal. 2005, 347, 1081–1089
asc.wiley-vch.de
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1087

FULL PAPERS
Derek R. Boyd et al.
strong influence on the regioselectivity. Thus, the pro- Typical Small-Scale Biotransformation Procedures
portion of TDO-catalysed alkene dihydroxylation, rela- using P. putida NCIMB 8859 and P. putida UV4
tive to arene dihydroxylation of styrene substrates, was
The P. putida NCIMB 8859 wild-type inducible strain was
grown on naphthalene (10 g L ) as sole carbon source and in-
ducer. Small-scale biotransformations were carried out using
whole cells in the late exponential phase of growth. The excess
generally found to be higher when a meta-substituent
was present on the arene ring (e.g., 1b–d) and lower
when the alkene had a methyl substitutent (e.g., 1e–g).
ꢀ1
The synthetic potential of the dioxygenase-catalysed naphthalene was filtered off, the filtrate was centrifuged and
asymmetric dihydroxylation of alkenes depends upon the collected cells were resuspended in potassium phosphate
the availability of 1,2-diol bioproducts in enantiopure buffer (50 mM) at pH 7.2. Sodium pyruvate (0.5%) was added
ꢀ
1
as co-substrate with the substrates (0.2–1.0 g L ) which were
added in a minimum volume of ethanol to 500-mL Erlenmeyer
flasks. These were in turn incubated at 308C on an orbital shak-
er (200 rpm) for 24 h. The contents of the flasks were centri-
fuged (9000 rpm, 10 min). The aqueous supernatant was de-
canted off and the bioproducts harvested by concentration un-
der reduced pressure and solvent extraction (EtOAc) of the so-
dium chloride-saturated solution.
form and of known absolute configuration. While the
acyclic alkene substrates 1a–g yielded enantiopure
arene-cis-dihydrodiols 2a–g, of identical configuration
(
1
>98% ee, 1S,2R), the corresponding acyclic alkene
,2-diols 3a–f had lower ee values and similar configu-
rations (42–88%, 1R) using both TDO and NDO. Based
on the results obtained with acyclic alkenes, the dioxyge-
nase-catalysed asymmetric dihydroxylation route does
The P. putida UV 4 constitutive mutant strain required that
not appear to have any significant advantages over the the most TDO-active colonies (determined by the indole
[
26]
[43]
chemical method. However, with the notable excep- test) be selected before use since reversion to the wild-
type state occurred readily. The whole cells were again grown
in Erlenmeyer flasks containing minimal salts growth medium
100 mL) and sodium pyruvate (0.5%) and substrates added to
cell suspensions in phosphate buffer. The biotransformation
24 h) and isolation of bioproducts (concentration and EtOAc
tion of cyclopentadiene 14a, the other cyclic cis-disubsti-
tuted alkenes (9 and 14b–i), were all found to undergo
TDO-catalysed asymmetric 1,2-dihydroxylation to
give cis-diols 10 and 15b–i as single enantiomers having
similar (S)-configurations at the benzylic and allylic cen-
tres.
(
(
extraction) was as described for the P. putida NCIMB 8859
strain.
The successful biocatalytic formation of enantiopure
cis-diols (7b, 7c, 10 and 15b–i), from the corresponding
cyclic cis-disubstituted alkenes, provides an attractive
alternative route to the reported chemical asymmetric
dihydroxylation method where ee values obtained
Acknowledgements
Financial support from the BBSRC (NDS), DENI (INB,
were much lower. Initial studies to realise the synthetic GMcC), and TBNI (MRG), is gratefully acknowledged.
potential of these enantiopure 1,2-diols include the use
of cis-diol 15f as a precursor of enantiopure ketone 21.
References
[
1] D. A. Widdowson, D. W. Ribbons, Janssen Chim. Acta
990, 8, 3–9.
1
[
[
2] H. J. Carless, Tetrahedron Asymmetry 1992, 3, 795–823.
3] G. N. Sheldrake, in: Chirality in Industry, (Eds.: A. N.
Collins, G. N. Sheldrake, J. Crosby), John Wiley, Chiches-
ter, 1992, Chap. 6, pp. 128–166.
Experimental Section
Characterisation data for all bioproducts is available in the
Supporting Information.
[
4] S. M. Brown, T. Hudlicky, in: Organic Synthesis: Theory
and Applications, JAI Press Inc., Greenwich, 1993, pp.
1
13–176.
5] D. R. Boyd, G. N. Sheldrake, Nat. Prod. Rep. 1998, 15,
09–324.
6] T. Hudlicky, D. Gonzalez, D. T. Gibson, Aldrichim. Acta
999, 32, 35–62.
Wild-type strains of P. putida (NCIB 11767 and NCIMB
[
[
[
8859) used in this work were obtained from the National Col-
3
lections of Industrial and Marine Bacteria, 23 St. Machar
Drive, Aberdeen, Scotland, UK. The UV4 mutant strain of P.
putida was obtained from Dr. S. C. Taylor, Avecia Pharmaceut-
icals, Billingham, Cleveland, UK. The UV4 mutant strain, con-
taining TDO, was derived from the NCIB 11767 wild-type
strain using the methods described in the following patents:
1
7] D. R. Boyd, N. D. Sharma, C. C. R. Allen, Curr. Opin.
Biotechnol. 2001, 12, 564–573.
8] R. A. Johnson, Org. React. 2004, 63, 117–264.
9] S. M. Resnick, K. Lee, D. T. Gibson, J. Indust. Microbiol.
[
[
[42]
EP 76606, EP 253,485 and US 5,073,640 and in the literature.
The mutant strain P. putida 39/D (ATCC 700008) also contains
1
996, 17, 438–457.
TDO and was generally found to give comparable results for a
wide range of substrates in these and other laboratories.
[10] L. P. Wackett, L. D. Kwart, D. T. Gibson, Biochemistry
1988, 27, 1360–1367.
[8]
Comparative experimental details for the biotransformations
of arenes using TDO, NDO and other dioxygenases with a
range of P. putida strains have recently been reported.
[11] R. A. S. McMordie, D. R. Boyd, N. D. Sharma, H. Dal-
ton, P. Williams, R. O. Jenkins, J. Chem. Soc. Chem.
Commun. 1989, 339–340.
[8]
1088
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1081–1089

Biocatalytic Asymmetric Dihydroxylation of Conjugated Mono- and Polyalkenes
FULL PAPERS
[
12] D. R. Boyd, M. R. J. Dorrity, J. F. Malone, R. A. S.
McMordie, N. D. Sharma, H. Dalton, P. J. Williams, J.
Chem. Soc. Perkin Trans. 1 1990, 489–494.
[28] D. R. Boyd, N. D. Sharma, F. Hempenstatl, M. A. Ken-
nedy, J. F. Malone, C. C. R. Allen, S. M. Resnick, D. T.
Gibson, J. Org. Chem. 1999, 64, 4005–4011.
[29] D. R. Boyd, N. D. Sharma, S. A. Haughey, J. F. Malone,
A. King, B. T. McMurray, R. Holt, H. Dalton, J. Chem.
Soc. Perkin Trans. 1 2001, 3288–3296.
[30] D. R. Boyd, N. D. Sharma, A. King, B. Byrne, S. A.
Haughey, M. A. Kennedy, C. C. R. Allen, Org. Biomol.
Chem. 2004, 2, 2530–2537.
[31] N. I. Bowers D. R. Boyd, N. D. Sharma, P. A. Goodrich,
M. R. Groocock, A. J. Blacker, P. Goode, H. Dalton, J.
Chem. Soc. Perkin Trans. 1 1999, 1453–1461.
[32] Z. M. Wang, K. Kakiuchi, K. B. Sharpless, J. Org. Chem.
1994, 59, 6895–6897.
[
[
[
[
13] T. Hudlicky, H. Luna, G. Barbieri, L. D. Kwart, J. Am.
Chem. Soc. 1988, 110, 4735–4741.
14] T. Hudlicky, E. E. Boros, C. H. Boros, Tetrahedron:
Asymmetry 1993, 4, 1365–1386.
15] K. Konigsberger, T. Hudlicky, Tetrahedron: Asymmetry
1
993, 4, 2469–2474.
16] D. T. Gibson, S. M. Resnick, K. Lee, J. M. Brand, D. S.
Torok, L. P. Wackett, M. J. Schocken, B. E. Haigler, J.
Bacteriol. 1995, 177, 2615–2621.
17] D. S. Torok, S. M. Resnick, J. M. Brand, D. L.Cruden,
D. T. Gibson, J. Bacteriol. 1995, 177, 5799–5805.
18] D. R. Boyd, N. D. Sharma, N. A. Kerley, R. A. S. McMor-
die, G. N. Sheldrake, P. Williams, H. Dalton, J. Chem.
Soc. Perkin Trans. 1 1996, 67–74.
[
[
[33] A. Armstrong, B. R. Hayter, Tetrahedron: Asymmetry
1997, 8, 1677–1684.
[34] M. Adamczyk, D. S. Watt, D. A. Netzel, J. Org. Chem.
1984, 49, 4226–4237.
[
19] S. M. Resnick; D. T. Gibson, Appl. Environ. Microbiol.
1
996, 62, 1364–1368.
[35] K. J. Stone, R. D. Little, J. Org. Chem. 1984, 49, 1849–
1853.
[
20] D. R. Boyd, N. D. Sharma, B. Byrne, M. V. Hand, J. F.
Malone, G. N. Sheldrake, J. Blacker, H. Dalton, J.
Chem. Soc. Perkin Trans. 1 1998, 1935–1943.
[36] H. Banks, H. Ziffer, J. Org. Chem. 1982, 47, 3743–3747.
[37] D. R. Boyd, N. D. Sharma, S. Haughey, M. A. Kennedy,
B. T. McMurray, G. N. Sheldrake, C. C. R. Allen, H. Dal-
ton, K. Sproule J. Chem. Soc. Perkin Trans. 1 1998,1929–
1933.
[
21] N. I. Bowers, D. R. Boyd, N. D. Sharma, M. A. Kennedy,
G. N. Sheldrake. H. Dalton, Tetrahedron: Asymmetry
1
998, 9, 1831–1834.
[
22] D. R. Boyd, D. Clarke, M. C. Cleij, J. T. G. Hamilton,
G. N. Sheldrake, Monatsh. Chem. 2000, 131, 673–685.
23] D. R. Boyd, N. D. Sharma, N. I. Bowers, R. Boyle, J. S.
Harrison, T. Bugg, K. Lee, D. T. Gibson, Org. Biomol.
Chem. 2003, 1, 1298–1307.
24] H. Ziffer, K. Kabuto, D. T. Gibson, V. M. Kobal, D. M.
Jerina, Tetrahedron 1977, 33, 2491–2496.
25] P. J. Geary, R. J. Pryce, S. M. Roberts, G. Ryback, J. A.
Winders, J. Chem. Soc. Chem. Commun. 1990, 204–205.
26] H. C. Kolb, M. S. van Nieuwenhze, K. B. Sharpless,
Chem. Rev. 1994, 94, 2483–2547.
[38] M. Goodall, P. M. Kelly, D. Parker, K. Gloe, H. Stephan,
J. Chem. Soc. Perkin Trans. 2 1997, 59–70.
[39] J. E. Lynch, E. L. Eliel, J. Am. Chem. Soc. 1984, 106,
2943–2948.
[40] O. C. Kreutz, P. J. S. Moran, J. Rodigues, R. Augusto, Tet-
rahedron: Asymmetry 1997, 8, 2649–2654.
[41] G. Jaouen, A. Meyer, J. Am. Chem. Soc. 1975, 97, 4667–
4672.
[42] D. G. Ballard, A. J. Blacker, J. M. Woodley, S. C. Taylor,
in: Plastics from Microbes, (Ed.: D. P. Mobley), Hanser,
Munich, 1994, 139.
[43] R. O. Jenkins, H. Dalton, FEMS Microbiology Letters
1985, 30, 227–231.
[
[
[
[
[
27] S. M Resnick, D. S. Torok, D. T. Gibson, J. Org. Chem.
1
995, 60, 3546–3549.
Adv. Synth. Catal. 2005, 347, 1081–1089
asc.wiley-vch.de
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1089