Asymmetric dihydroxylation of cinnamonitrile to trans-3-[(5S,6R)-5,6- dihydroxycyclohexa-1,3-dienyl]-acrylonitrile using chlorobenzene dioxygenase in Escherichia coli (pTEZ30)

Article abstract of DOI:10.1002/adsc.200404032

Asymmetric cis-dihydroxylations of aromatic compounds are catalyzed by bacterial dioxygenases. In order to prevent through conversion, either dihydrodiol dehydrogenase blocked mutant strains or recombinant bacterial cells are used as biocatalysts for synthetic purposes. We characterized the cis-dihydroxylation of cinnamonitrile by chlorobenzene dioxygenase (CDO) in recombinant E. coli on different reaction scales. The absolute stereochemistry of the product was determined to be trans-3-[(5S,6R)-5,6-dihydroxycyclohexa-1,3- dienyl]-acrylonitrile. The cells showed a maximum specific activity of 3.76 U/g cdw in shake-flask experiments. Stable expression of the dioxygenase genes in E. coli JM101 (pTEZ30) resulted in increasing volumetric productivities. The maximum volumetric productivities of 80 and 92 mg product/L/h were achieved on 2-L and 30-L scales, respeclively. The specific growth rate correlated with the volumetric productivity during the biotransformations. An average volumetric productivity of 40 mg product/L/h in reactors on 2-L and 30-L scales resulted in 0.96 and 16.4 g of isolated product at the end of the biotransformations. This points out the need for metabolically active cells and controllable expression systems for achieving high volumetric productivities for cofactor dependent biooxidations. We have now applied this concept for the asymmetric dihydroxylation of the non-natural substrate cinnamonitrile using multicomponent CDO in tailored E. coli JM101 in long-term reactions.

Full text of DOI:10.1002/adsc.200404032

FULL PAPERS
Asymmetric Dihydroxylation of Cinnamonitrile to
trans-3-[(5S,6R)-5,6-Dihydroxycyclohexa-1,3-dienyl]-acrylonitrile
using Chlorobenzene Dioxygenase in Escherichia coli (pTEZ30)
Selcuk Yildirim,a Josef Zezula,b Tomas Hudlicky,b Bernard Witholt,a
Andreas Schmida,*
a
Institute of Biotechnology, HPT D 73, ETH-Zurich Hoenggerberg, 8093 Zurich, Switzerland
Phone: (þ41)-1-633-3691, fax: (þ41)-1-633-1051, e-mail: andreas.schmid@biotech.biol.ethz.ch
b
Department of Chemistry, Brock University, St. Catharines, Ontario, L2S3A1 Canada
Received: January 31, 2004: Accepted: May 21, 2004
Abstract: Asymmetric cis-dihydroxylations of aromat- tively. The specific growth rate correlated with the
ic compounds are catalyzed by bacterial dioxygenases. volumetric productivity duringthe biotransforma-
In order to prevent through conversion, either dihy- tions. An average volumetric productivity of 40 mg
drodiol dehydrogenase blocked mutant strains or re- product/L/h in reactors on 2-L and 30-L scales result-
combinant bacterial cells are used as biocatalysts for ed in 0.96 and 16.4 gof isolated product at the end of
synthetic purposes. We characterized the cis-dihy- the biotransformations. This points out the need for
droxylation of cinnamonitrile by chlorobenzene dioxy- metabolically active cells and controllable expression
genase (CDO) in recombinant E. coli on different systems for achievinghigh volumetric productivities
reaction scales. The absolute stereochemistry of the for cofactor dependent biooxidations. We have now
product was determined to be trans-3-[(5S,6R)-5,6-di- applied this concept for the asymmetric dihydroxyla-
hydroxycyclohexa-1,3-dienyl]-acrylonitrile. The cells tion of the non-natural substrate cinnamonitrile using
showed a maximum specific activity of 3.76 U/gcdw
multicomponent CDO in tailored E. coli JM101 in
in shake-flask experiments. Stable expression of the long-term reactions.
dioxygenase genes in E. coli JM101 (pTEZ30) result-
ed in increasingvolumetric productivities. The maxi-
mum volumetric productivities of 80 and 92 mgprod-
Keywords: asymmetric synthesis; biocatalysis; bio-
transformations; chlorobenzene dioxygenase; dihy-
uct/L/h were achieved on 2-L and 30-L scales, respec- droxylation; scale-up
Introduction
Oxidation of aromatic compounds to cis-dihydrodiols is
catalyzed by multicomponent bacterial dioxygenases
(Figure 1).[1,2] These reactions are of special interest,
since cis-dihydrodiols are valuable buildingblocks for
the stereoselective synthesis of molecules containing
Figure 1. Dioxygenase catalyzed biotransformation of aro-
matic compounds to the corresponding cis-dihydrodiols.
5]
multiple chiral centers.[3
Thus far, several ring-hydroxylating dioxygenases naphthalene dioxygenase,[22
biphenyl dioxyge-
24]
have been identified[6] includingnaphthalene,
tolu- nase,[25] chlorobenzene dioxygenase,[16] and phenan-
[7 9]
ate,[10] benzene,[11] toluene,[12] benzoate,[13] biphenyl,[14] threne dioxygenase[26,27]. Some of the biotransformation
2-nitrotoluene,[15] chlorobenzene[16] and tetrachloroben- processes usingmutant strains have been well character-
zene dioxygenases.[17] Since Gibson and co-workers first ized, developed and scaled up.[28 31] Biotransformations
described the isolation of an optically active diol,[18] with recombinant cells expressing dioxygenase genes
more than 300 vicinal arene cis-diols have been identi- were performed on laboratory and industrial production
fied.[4,19] Biocatalysts were cis-dihydrodiol dehydrogen- scale yielding gram to multi-kilogram amounts of cis-cy-
ase blocked mutants of Pseudomonas species, Sphingo- clohexadiene diols.[32]
monas yanoikuyae and Alcaligenes eutrophus.[2,20] Alter-
In general, aromatic nitriles are interesting synthons,
natively, the biotransformations used recombinant cells because the enzymatic dihydroxylation may be coupled
containing dioxygenases like toluene dioxygenase,[12,21] to a selective hydrolysis of the nitrile group or subse-
Adv. Synth. Catal. 2004, 346, 933 942
DOI: 10.1002/adsc.200404032
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
933

Selcuk Yildirim et al.
FULL PAPERS
quent cycloaddition reactions to increase the pool of chi- fore substrate addition. The substrate concentration de-
ral synthons. We now examine the efficiency of a biopro- creased to 1.1 mM (Figure 3). After addition of another
cess for the highly selective asymmetric dihydroxylation 1 mM substrate, the total substrate concentration in-
of non-natural cinnamonitrile to trans-3-[(5S,6R)-5,6-di- creased to 2.1 mM and the cell concentration stayed al-
hydroxycyclohexa-1,3-dienyl]-acrylonitrile (1) by chloro- most constant afterwards. At time 9 h and 23 h addition-
benzene dioxygenase (CDO) in recombinant E. coli al 2 mM and 1 mM substrate was added, respectively.
JM101 (pTEZ30).[33] We examined the efficiency of
Three distinct phases of product formation could be
the biocatalyst with a focus on the volumetric productiv- distinguished: a rapid linear product formation over
ity by comparingbiotransformations on different scales. 3.5 h up to a product concentration of 1.4 mM with an
We discuss advantages and limitations of our recombi- average specific activity of 0.9 U/g cdw, a slower linear
nant expression system based on the alkane oxidation increase in product concentration to 2.16 mM with an
promoter PalkB for arene cis-diol production as well as average specific activity of 0.27 U/g cdw for the next
the determination of the absolute stereochemistry of 5 h, and a period towards the end of the biotransforma-
the title compound.
tion where the product concentration did not increase
any further. The second phase might be longer as there
is product formation between 9 and 21 hours. However,
an average volumetric productivity of 0.14 U/g cdw in
this time range points to a decrease in volumetric pro-
Results
The specific activity of CDO in E. coli JM101 (pTEZ30) ductivity towards the third phase. The product forma-
for cinnamonitrile was determined in 20-mL shake tion rate decreased continuously after the cells stopped
flasks (Figure 2). CDO showed a maximum specific ac- growing (Figure 5). This suggests a relationship between
tivity of 3.8 Units per gram cell dry weight (U/g cdw) and the metabolic state of the cells and specific activity of the
decreased to 0.3 U/gcdw duringthe biotransformation
resultingin an average specific activity of 2.25 U/gcdw.
cells. In total, 5.6 mM of substrate was added and the
product concentration reached 2.93 mM after 21 h,
The maximum specific activity of CDO in E. coli which corresponds to 0.48 g/L. Total mass balance
JM101 (pTEZ30) for cinnamonitrile is in the same range showed that part of the substrate was stripped during
as the specific activities of CDO for benzonitrile and the biotransformation and this resulted in a 52% yield
naphthalene (3.9 and 2.86 U/gcdw, respectively). [33]
after 25 h.
Biotransformation of Cinnamonitrile on a 2-L Scale
Biotransformation of Cinnamonitrile on a 30 L-Scale
Specific activities of whole cell biocatalysts are fre- Scale-up of whole cell biocatalytic reactions requires not
quently lower under process conditions in bioreactors only an increase in volume but also establishment of ap-
than in short term assays, e.g., in shake flasks.[34 38] The propriate conditions for cell growth, efficient synthesis
biotransformation of cinnamonitrile was therefore per- of the relevant enzyme, and a bioreactor environment
formed in a 2-L reactor and process parameters were de- that ensures stable bioconversion of substrate over a
termined. The biotransformation was started by the ad-
dition of 1.6 mM cinnamonitrile, after which the cells
grew to 8.2 g cdw/L at a growth rate similar to that be-
Figure 3. Biotransformation of cinnamonitrile using E. coli
JM101 (pTEZ30) in a bioreactor on a 2-L scale: cell amount
&
&
^
( ); product concentration ( ); substrate concentration ( ).
After a batch fermentation, glycerol was fed to the reactor
and the cells were induced for 3 h with DCPK. Substrate
was added sequentially.
Figure 2. Biotransformation of cinnamonitrile using E. coli
JM101 (pTEZ30) in shake flasks. Reaction was performed
usingrestingcells (1.29 gcdw/L) in 20-mL shake flasks.
934
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 933 942

Asymmetric Dihydroxylation of Cinnamonitrile
FULL PAPERS
longperiod of time. In order to evaluate the scale-up po-
tential of cis-dihydroxylations usingrecombinant cells
The maximum and the average volumetric productiv-
ities, as well as the activity profile of 1 formation are sim-
we carried out the biotransformation of cinnamonitrile ilar on the 2-L and 30-L scales. The decrease of produc-
using E. coli JM101 (pTEZ30) in a reactor with a 30-L tivity towards the end of the biotransformation was also
working volume. Four stages of growth could be distin- observed in the shake flask experiment and may have
guished (Figure 4): a rapid linear growth with a specific been the consequence of toxicity of the substrate and
growth rate of 0.74 hÀ1 for 3 h to 15.2 gcdw/L before the
product. The decrease in specific activity of the biocata-
biotransformationwasstarted,astationaryperiodduring lyst correlated with the decrease in volumetric produc-
the first hour of biotransformation at almost 14 gcdw/L, tivity (Figure 5). After cell growth, the dissolved oxygen
a less rapid but still linear growth with a specific growth tension was constant (20 40% of saturation) through-
rate of 0.47 hÀ1 for the next 5 hours to 25.5 gcdw/L,
and a final stationary phase at almost 23 gcdw/L. During
out the biotransformation. This constant oxygen con-
sumption indicates metabolic activity of the biocatalyst
the first period, the cell growth did not appear to be even in the stationary phase with concomitant product
affected by induction. Cell growth deceased in the sec- formation. This relationship points to a couplingof the
ond phase after the addition of the substrate, resumed metabolic state of the cells with the specific activity of
in the third phase, and leveled off in the last stage, which CDO as depicted in Figure 5.
is probably because of the toxic effect of the product
HPLC analysis of the products did not point to the for-
which accumulated in the reactor. Toxic effects of diene- mation of by-products duringthe reaction. cis-Dihydro-
diols on cell growth have been observed before.[31,39,40] diols are highly soluble in aqueous solutions but can de-
[42]
At toxic substrate and product concentrations, the pores hydrate to the correspondingphenols
under acidic
formed in the cytoplasmic membrane may be large conditions and high temperatures. This happened dur-
enough to cause complete loss of the membrane integri- ingthe evaporation step for product recovery: less
ty, resultingin cell death. [41]
than 2% of 1 were dehydrated to the phenol. Here, con-
Formation of trans-3-[(5S,6R)-5,6-dihydroxycyclo- tinuous phase extraction with ethyl acetate might be
hexa-1,3-dienyl]-acrylonitrile (1) started directly after used as an alternative to the evaporation of the aqueous
the addition of cinnamonitrile with a rate of 0.29 U/g phase.[32] Table 1 lists the process parameters of our cin-
cdw until 6.5 h reaction time. Then net cell growth stop- namonitrile biotransformation using E. coli JM101
ped, but the rate of product formation decreased only (pTEZ30) on different scales. The downstream process-
slowly until the end of the transformation period giving ingand the purification yields with the corresponding
a final product concentration of 3.32 mM (0.55 g/L). The purities are given in Table 2. The cis-dihydroxylation
initial specific activity increased from 0.23 U/gcdw after
induction to a maximum of 0.52 U/gcdw at 2.5 h then
of cinnamonitrile using E. coli JM101 (pTEZ30) is a
scaleable process as cell growth and stability of the re-
decreased again to 0.07 U/g cdw at 12 h. During this combinant expression system are not significantly
time, E. coli JM101 (pTEZ30) produced 14.6 g 1. This affected by the scale-up. The process is limited by the
corresponded to an average volumetric activity of 4.2 volumetric productivity, which can be improved by
U/L or an average volumetric productivity of 40 mg/L/h keepingthe cells in a metabolically active state. High
of 1.
specific activities usinggrowingcells compared to rest-
ingcells were pointed out by Walton and Stewart. [43] It
Figure 4. Biotransformation of cinnamonitrile using E. coli
JM101 (pTEZ30) in a bioreactor with a 30-L workingvol-
&
^
ume: cell amount ( ); product concentration ( ); substrate
Figure 5. Volumetric productivities of biotransformation of
cinnamonitrile using E. coli JM101 (pTEZ30) in shake-flasks,
and in reactors on 2-L and 30-L scales, and the specific
growth rate of the cells in a 30-L reactor.
&
concentration ( ). Cells were induced with DCPK and the
substrate concentration in the reactor was maintained below
1.2 mM through manual additions of it.
Adv. Synth. Catal. 2004, 346, 933 942
asc.wiley-vch.de
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
935

Selcuk Yildirim et al.
FULL PAPERS
was shown that growing E. coli BL21(DE3) pMM4 have
higher activities for the oxidation of cyclohexanone to e-
caprolactone than the restingcells. Higher volumetric
productivities were obtained with restingcells, but this
is mainly due to high cell densities used during the bio-
transformations. Thus, induction and start of biotrans-
formation at high density of growing cells may result
higher volumetric activities.
Proof of Absolute Stereochemistry
The crude diol 1 (sticky, crystalline solid in appearance)
was purified by recrystallization from CH2Cl2-EtOAc/
hexane to yield beige microscopic needles. The pure ma-
terial had a meltingpoint of 119 121 8C and aD þ337.98
(EtOH, c 1.0).
To determine the absolute stereochemistry of 1 we de-
cided on a convergent match with a compound that
could be rapidly synthesized from diol 2 whose absolute
configuration[44] as well as enantiomeric purity[45,46] are
Figure 6. Determination of the absolute stereochemistry of
diol 1 by chemical correlation.
53]
well known and proven by several total syntheses.[47
The proton NMR as well as IR of 4 obtained from diol
To this end a crude sample of 1 (as isolated from the fer- 1 matched exactly those of the synthetically derived 4.
mentation) was subjected to a reduction with potassium Mixed 1H-NMR showed no deviation from the spectra
azodicarboxylate (PAD) in order to increase the stabil- of either sample. The optical rotation of synthetic 4
ity of subsequently synthesized intermediates. The PAD was almost identical atþ207.58 (CH2Cl2, c 1.0).
reduction provided selectively nitrile 3, which was puri-
Thus the absolute stereochemistry of the hydroxyl
fied and converted to acetonide 4 as shown in Figure 6. groups in diol 1 is b, as expected from the toluene dioxy-
This material, purified by flash chromatography, gave genase processing of monosubstituted aromatics. To
an aD of þ208.38 (CH2Cl2, c 1.0). Diol 2 was obtained date there has been no deviation observed in the b-con-
from the whole cell fermentation of bromobenzene figuration of the diols derived from such compounds
with E. coli JM109 (pDTG 601), a strain developed by (exceptingthe diols of various benzoic acids which
David Gibson and expressingthe toluene dioxygenase
genes.[12,32] This material was also reduced with PAD
and the diol protected as acetonide 5.[54] Heck reac-
also possess different 1,2-regiochemistry).[47]
tion[55,56] of 5 with freshly distilled acrylonitrile in the Discussion
presence of Pd(OAc)2, PPh3, Et3N in DMF at 1008C
gave, after 20 h, a 20% yield of 4. The yield of 4 was im- Biocatalytic oxidations use oxygen as a cheap and envi-
proved to 76% [takinginto account recovered starting
material (54%)] by usingan excess of acrylonitrile
(3.0 equivs.) and a longer reaction time (68 h).
ronmentally friendly oxidant, and often have high regio-,
enantio- and stereoselectivities.[57] The oxidation of ar-
omatic compounds to cis-dihydrodiols catalyzed by bac-
Table 1. Bioprocess data of the cis-dihydroxylation of cinnamonitrile on different scales
Parameter
Unit
Biotransformation of cinnamonitrile in
Shake flask
Bioreactor, lab scale
Bioreactor, technical scale
30
Workingvolume
L
0.02
0.96
0.17
49
2
Total amount of product formed
End product concentration
Bioconversion
g0.0034
g/L
%
16.4
0.48
72[a]
0.14
0.97
0.54[a]
0.55
74[b]
Maximum growth rate
1/h
0.24
0.52
Maximum specific activity
Average specific activity
Maximum volumetric productivity
Average volumetric productivity
U/gcdw
U/g cdw
mgprod/L/h
mg prod/L/h
3.76
2.25
0.94
0.49
0.24[b]
80
92
40[a]
40[b]
[a]
Value in 9 h.
Value in 12 h.
[b]
936
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 933 942

Asymmetric Dihydroxylation of Cinnamonitrile
FULL PAPERS
Table 2. Downstream processingdata.
Amount of product [g]
Yield [%]
Purity
[b]
[b]
[b]
Reactor
Volume reduction
16.4
13.7
11.6
9.2
3.7
2.1
74[a]
62
52
Acetonitrile precipitation
Purification on silica gel column[c]
62
72
76
69
2
11.4
Absolute amount[d]
Recrystallization
52
48[e]
100
[a]
Bioconversion yield.
Not isolated.
Four different fractions collected.
Total pure product amount.
[b]
[c]
[d]
[e]
Yield of the recrystallization step.
terial dioxygenases is especially interesting since the over biotransformation times between 7 96 hours (Ta-
chemical equivalent of such a reaction is unknown. ble 3).
There is only a single report of tetrahydroxylation of
Although the mutant cells have high activities towards
benzene under photolytic osmylation conditions.[58] their natural substrates, they often show lower activities
The vast majority of the reported cis-dihydroxylations for non-physiological compounds.[20,59] As presented in
were performed with whole cells as biocatalysts, due to Table 3 average volumetric productivities of wild-type
a low stability of the multicomponent system of dioxyge- strains towards their natural substrates like toluene
nases and requirements of expensive cofactor.[2] As a re- and naphthalene were as high as 400 mg prod/L/h,
sult of such extensive studies the average volumetric whereas the productivities of the P. putida and Rhodo-
productivity of the (mutated) wild-type cells on their coccus sp. B264-1 towards a non-natural substrate in-
natural substrates could be increased to 0.24 2 g/L/h dene were 44 and 32 mgprod/L/h, respectively. We ob-
Table 3. Microbial cis-dihydroxylation processes.
Strain
Enzyme Substrate
Scale Biotrans. Max. sp
cdw
Avg. prod. Product
Ref.
[L]
time [h] activity
[g/L] [g/L/h]
conc. [g/L]
[U/gcdw]
P. putida UV4
P. putida UV4
P. putida UV4
P putida UV4
TDO
TDO
TDO
TDO
TDO
n.m.
TDO
NDO
NDO
n.m.
Toluene
Toluene
Toluene
Toluene
Indene
Toluene
Toluene
Naphthalene 1.0
Dibenzofuran 0.8
Indene
Toluene
Toluene
Indene
Indene
Diverse
96
20
12.5
7
n.m
26 33
158
1.2[a]
4 6[c] 0.41
4 6
1 1.5 1.0
0.24[b]
22.7[b]
8.2
25
6.8
0.35
18
24.0
6.5
1.6
[59]
[31]
[30]
[72]
[28]
[73]
[63]
[20]
[74]
[28]
[59]
[75]
[76]
[28]
1.7
1.7[d]
1.7
2
132
P. putida
70[e]
1.5
3.5
8
n.m.
n.m.
105
n.m.
n.m.
0.044
0.25
Rhodococcus MA7249
P. putida (strain NG1)
P. fluorescens TTC1
P. fluorescens TTC1
Rhodococcus sp. B264-1
E. coli JM109 (pKST11)
E. coli TG2 (p1/1)
E. coli D160-1
70
24
15
10
150
96
7.9
24
24
n.m
1.5]f, g] 0.8
40
2.1
0.4
n.m.
n.m
n.m.
35
n.m.
n.m
n.m
n.m.
n.m.
0.160
0.032
23[e]
1.5
5
4.8
TDO
TDO
TDO
TDO
0.03[a] 0.014[b]
1.32[b]
1.02
1.2
1.2[g]
n.m
n.m
n.m
0.13
0.05
0.042
n.m
15
E. coli D160-1
E. coli JM109 (pDTG601) TDO
23[e]
10 15
1
0.06 15[h] [32]
[a]
Average cell dry weight.
[b]
Average cell dry weight was used for the calculation of the value.
4 6 g/L paste of cells was resuspended.
Two phase system consistingof 510 mL organic (10% toluene) phase and 1190 mL aqueous phase.
Total volume of the reactor.
Assumption: 50% of the cell dry weight is carbon.
Average cell dry weight during the biotransformation time given.
The minimum and the highest values obtained for the given substrates.
[c]
[d]
[e]
[f]
[g]
[h]
Adv. Synth. Catal. 2004, 346, 933 942
asc.wiley-vch.de
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
937

Selcuk Yildirim et al.
FULL PAPERS
served 40 mgprod/L/h with E. coli JM101 and the non-
The activity of E. coli JM101 (pTEZ30) for cinnamo-
natural substrate cinnamonitrile. Overexpression in a nitrile in a reactor on a 2-L scale was very low (0.97 U/g
recombinant host does not necessarily result in higher cdw), compared to the activity of the (mutated) wild-
activities than in the wild-type (Table 3).[59] However, type strains for toluene conversion (26 158 U/gcdw)
recombinant hosts like Escherichia coli are well defined in small bioreactors. However, the average volumetric
and therefore easy to handle. They can be grown to high productivity of 40 mgprod/L/h in both 2-L and 30-L
cell densities on glucose as a cheap and non-toxic carbon scales made it possible to obtain 0.96 and 16.4 gamounts
source.[60,61] Induction of gene expression can be control- of 1 respectively, with a clear perspective of improve-
led by easily controllable promoters usingnon-aromatic
ment in productivity. We expect that tightly controllable
inducers,[62] which are not substrates for the dioxygenas- recombinant whole cell biocatalysts will allow optimiza-
es and can easily be separated from the dihydrodiols. tion of the expression and enzyme activity. This will be
Scale-up of recombinant cell biocatalysis is significant very useful for cis-dihydroxylations of non-natural sub-
for synthetic purposes. The key points in scale-up of strates like aromatic nitriles. This will also have a high
such systems are: efficient synthesis of the enzyme to impact onthe synthetic applicabilityofengineereddioxy-
high specific activities; high biocatalyst stability, namely genases, which are steadily increasing in number.[66
69]
high volumetric productivity over time; substrate and
product toxicity; and downstream processing, product
recovery. The last two points have been improved using
Conclusion
(mutated) wild-type cells based on the use of different
[63]
feedingstrateiges,
regulated substrate addition,[31]
Recombinant cells have generally been used to overex-
press dioxygenases for the production of cis-diols from
various different substrates. Here we characterized the
product and described the biotransformation of cinna-
monitrile to trans-3-[(5S,6R)-5,6-dihydroxycyclohexa-
1,3-dienyl]-acrylonitrile (1) using E. coli JM101
(pTEZ30) in different scales. The results show that re-
combinant hosts with a strongexpression system, such
asE.coliJM101 (pTEZ30), can be used to produce dioxy-
genases efficiently to perform cis-dihydroxylations up to
technical scales. The cells should be maintained in a met-
abolically active state duringthe biotransformation in
order to increase and maintain the volumetric produc-
tivity for long-term reactions. Controlled substrate addi-
tion and in situ product removal techniques can be inte-
grated into the cis-dihydroxylation processes usingre-
combinant cells in order to scale-up and improve the
performance of these processes.
two-liquid phase reaction media,[30] and in situ product
recovery.[64,65] Here we have described, on three differ-
ent scales, the enzyme synthesis, and the biocatalyst sta-
bility for the production of an interestingproduct 1 with
a focus on specific activity, and volumetric productivity
over the reaction time, which determine the efficiency
of the cis-dihydroxylation process (1). We determined
the absolute stereochemistry of the hydroxyl groups in
title compound 1 as b, as expected from the toluene di-
oxygenase processing of monosubstituted aromatics by
matchinga derivative with a material prepared by inde-
pendent synthesis from a compound of known absolute
stereochemistry (i.e., diol from bromobenzene). Effi-
cient expression of the dioxygenase genes under the
control of the alk promoter PalkB in recombinant E.
coli JM101 (pTEZ30) resulted in maximum volumetric
productivities of 80 and 92 mgprod/L/h in a 2-L and a
30-L reactor, respectively. On both scales, the average
volumetric productivity was 40 mgproduct/L/h (over
9 h in 2 L and over 12 h in 30 L). This points to a straight-
forward scalability of the cis-dihydroxylation of cinna-
monitrile usingrecombinant E. coli JM101 (pTEZ30).
Yet, the decrease in volumetric productivity at the end
of the biotransformation indicated that the biocatalyst
stability is limited and correlates with the metabolic ac-
tivity of the cells which therefore should be examined
indetail duringthebiotransformation. Thelinkofgrowth
rateofE. coliexpressingalkgeneswithitsbiocatalytic ac-
tivity has also been described earlier.[60] In our case, the
factor limitingmetabolic activity ofthe cells is most prob-
ably the presence of the product. Therefore, the cis-dihy-
droxylation processes usingrecombinant cells might be
further improved by the regulated addition of the sub-
strate[31] or the integration of in situ product removal[64,65]
to minimize the toxic effect of the substrateand the prod-
uct to the cells. This would allow a further increase of the
volumetric productivity resultingin higher yields.
Experimental Section
Strain, Media, Cultivation and Chemicals
We used E. coli JM101 {supE thi D(lac-proAB) F’[traD36 lacIq-
D(lacZ)M15 proABþ]}[70] as a recombinant host strain and
pTEZ30 as an expression vector[33] containingchlorobenzene
dioxygenase genes tcbAa, tcbAb, tcbAc, and tcbAd under the
control of the alk regulatory system of P. oleovorans GPo1.[71]
Luria-Bertani broth (LB) complex (Difco, Detroit, MI.) or
M9 minimal medium[70] supplemented with 50 mg/L kanamy-
cin were used as a medium. When necessary the medium was
solidified with 1.5% agarose (Difco). Glucose, thiamine, and
US* the composition of which is described elsewhere[35] were
added to M9 medium at a final concentration of 0.5%,
0.001%, and 0.1%, respectively. All incubation and growth
steps were carried out at 308C. Liquid media was incubated
on a shaker at 220 rpm. Chemicals were obtained from Fluka
AG (Buchs, Switzerland).
938
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 933 942

Asymmetric Dihydroxylation of Cinnamonitrile
FULL PAPERS
for 40 min, the reactor contents were supplemented aseptically
with glucose solution, US*, magnesium sulfate, thiamine, ka-
namycin, and antifoamingagent. The volume was adjusted to
27 L with water. The reactor was inoculated with 3 L of M9 pre-
culture (prepared as M9 preculture for biotransformations in
shake flasks) and batch cultivation was performed at a stirring
speed of 500 rpm and an aeration of 15 L/min for about 10 h.
Silicone oil-based antifoam agent (Sigma) was used to control
foaming. Fed-batch was started with pumping of the feed con-
sisting of 46% glucose and 1% MgSO4 with a rate of 10 g/L/h
into the reactor. Stirringspeed was increased to 700 rpm. After
1 h of fed-batch growth, additional kanamycin (1.5 g) US*
(30 mL) and thiamine (0.3 g) were added and the expression
of CDO genes was induced by the addition of 0.05% DCPK.
3 h later the biotransformation was started by the addition of
substrate. Substrate was added repeatedly to keep the concen-
trations below 1.5 mM. Samples were taken at different time
points and diluted 10 times with ice-cold acetonitrile and cen-
trifuged at 5000 rpm at 48C. Supernatants were filtered
through a cellulose membrane filter (Spartan, Schleicher &
Schuell GmbH, Germany) and analyzed. Volumetric produc-
tivities and specific growth rates were calculated for the inter-
vals between two samplingpoints.
Biotransformations in Shake-Flasks
Freshly transformed E. coli JM101 cells were plated on LB agar
plates. After incubation overnight, 5 mL LB medium was ino-
culated with a single colony and incubated for 10 h. 1 mL of this
LB preculture was used to inoculate 100 mL of M9 medium.
After incubation overnight, 100 mL of M9 medium were ino-
culated with the M9 preculture to a final cell concentration
of A450 of 0.1, induced with 0.05% dicyclopropyl ketone
(DCPK) and further incubated for 4 h. Then the cells were cen-
trifuged at 10,000Âgfor 10 minutes and the pellet was resus-
pended in 20 mL 50 mM potassium phosphate buffer
(pH 7.4) supplemented with 1% glucose to a final cell concen-
tration of 1.29 gcdw/L and further incubated. Substrate was
added to a final concentration of 2 mM. Samples were taken,
centrifuged at 20,000Âgfor 5 minutes and supernatant was an-
alyzed. Activity of the cells was calculated based on the prod-
uct formation. One unit (IU) is defined as 1 mmol of product
formed in one minute.
Biotransformation of Benzonitrile in a 2-L Reactor
Biotransformation of cinnamonitrile was carried out in a stir-
red tank reactor with two turbine impellers, four baffles, and
a total volume of 3 liters. The reactor contained 1775 mL of
M9 medium, 2 mL of trace elements solution US*, thiamine
and antibiotics. The pH was adjusted to 7.2 with 25% NH4
OH and 25% phosphoric acid. Temperature was kept at
308C. Silicone oil-based antifoam agent (Sigma) was used to
control foaming. After inoculation with 200 mL of the precul-
ture (prepared as M9 preculture for biotransformations in
shake flasks) the reactor was aerated at a rate of 2 liters per mi-
nute and stirred at 1,500 rpm for ca. 10 h (overnight), which re-
sulted in a culture in the stationary phase which contained
3.48 g(dry weight) of cells per liter. The medium was then sup-
plemented with 8 mL of trace elements solution US* and 8 mg
of thiamine. Subsequently, the culture was fed at a rate of 6 mL/
h with an aqueous solution containing50% glycerol and 1%
MgSO4. Half an hour later, the cells were induced by addition
of 0.05% DCPK to the reactor. Three hours after induction,
substrate was added to a final concentration of 1.57 mM and
then added repeatedly afterwards. Samples were taken at dif-
ferent time points and diluted 10 times with ice-cold acetoni-
trile and centrifuged at 5000 rpm at 48C. Supernatants were fil-
tered through a cellulose membrane filter (Spartan, Schleicher
& Schuell GmbH, Germany) and analyzed. Biotransforma-
tions were carried out twice and nearly identical results were
obtained.
Downstream Processing
Reaction medium (30 L) was centrifuged at 5000 rpm and at
48C for 15 minutes. The supernatant was removed and evapo-
rated in a rotary evaporator at room temperature to a volume
of approximately 1 L. The dissolved salts and proteins were
precipitated with the addition of 4 L of acetonitrile. The mix-
ture was centrifuged again under the same conditions and the
supernatant was evaporated down to 20 mL. Product was run
on a silica gel (60) column with 60% ethyl acetateand 40% hex-
ane as eluent and fractions with different purities were collect-
ed. For the absolute stereochemistry determination, a part of
the product was further purified by recrystallization from
CH2Cl2-EtOAc/hexanes to yield beige microscopic needles.
Part of the product from the biotransformation on a 2-L scale
was purified by recrystallization from EtOAc/hexane and
used as a standard for quantifications. Purities were quantified
based on the comparison of the UV absorption area of the
standard and the samples from the fractions.
Chemical Analysis
To analyze the metabolites, a nucleosil C18 RP column (pore
size, 100 ä; particle size 5 mm; inner diameter, 12.5 cmÂ
2 mm) (Macherey-Nagel AG, Oensingen, Switzerland) was
used with a mobile phase of 69.93% of H2O-30% acetoni-
trile-0.07% H3PO4. The UV detector was set at a wavelength
range of 210 600 nm. NMR spectra were recorded on a Bruk-
er DPX 300 in CDCl3, IR measurements were performed with
a Perkin-Elmer ™Spectrum One∫ FT-IR. The optical rotations
were measured on a Perkin-Elmer 341 polarimeter.
Biotransformation of Benzonitrile in a 30-L Reactor
The 30-L biotransformations were performed in a 42-L stirred
tank reactor with regulated temperature, pH, stirrer speed, and
internal pressure (New MBR, Zurich, Switzerland). The pH
was kept at 7.2 by the addition of 25% (wt/vol) ammonium wa-
ter or 30% phosphoric acid. Data collection occurred every
30 s with the Caroline II software (PCS, Wetzikon, Switzer-
land) on an OS/9 operatingsystem. Mixingwas achieved
with two 6-bladed Rushton turbine impellers. Before steriliza-
tion the reactor contained salts equivalent to 30 L of M9 me-
dium dissolved in 20 L of water. After sterilization at 1218C
Determination of Absolute Stereochemistry
trans-3-[(5S,6R)-5,6-Dihydroxycyclohexa-1,3-dienyl]-acrylo-
nitrile (1): Crude diol (0.250 g), a pasty beige solid, was purified
Adv. Synth. Catal. 2004, 346, 933 942
asc.wiley-vch.de
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
939

Selcuk Yildirim et al.
FULL PAPERS
by recrystallization from CH2Cl2-EtOAc/hexanes to give beige
microscopic needles; yield: 0.120 g(48%); mp 119 121 8C;
[a]2D6:þ337.98 (c 1.0, EtOH); Rf (EtOAc)¼0.57; 1H NMR:
d¼6.99 (d, 1H, J¼16.5 Hz), 6.19 (d, 1H, J¼3.0 Hz), 5.98 (bs,
2H), 5.60 (d, 1H, J¼16.4 Hz), 4.37 (d, 1H, J¼6.2 Hz), 4.17
(d, 1H, J¼6.2 Hz), 3.37 (s, 2H); 13C NMR: d¼149.9, 136.3,
134.5, 132.1, 123.2, 118.6, 95.3, 69.7, 64.8. IR: n¼3313, 3056,
151.0, 141.0, 134.1, 118.9, 109.4, 96.4, 73.1, 70.6, 28.0, 26.6,
24.9, 21.8; IR: n¼3064, 3017, 2986, 2216, 1632, 1600 cmÀ1
anal. calcd.: C 70.22%, H 7.37%; found: C 69.97%, H 7.39%.
trans-3-[(3aR,7aS)-2,2-Dimethyl-3a,6,7,7a tetrahydro-
;
benzo[1,3]dioxol-4-yl]-acrylonitrile (4): To a solution of (3aS,
7aS)-7-bromo-2,2-dimethyl-3a,4,5,7a tetrahydrobenzo[1,3]-
dioxole (5; 0.680 g, 2.92 mmol) in dry dimethylformamide
(5 mL) was added Pd(OAc)2 (0.066 g, 0.292 mmol, 0.1 equiv.),
triphenylphosphine (0.153 g, 0.584 mmol, 0.2 equivs.), dry tri-
ethylamine (3.25 mL, 23.30 mmol, 8 equivs.) and, finally, freshly
distilled acrylonitrile (576 mL, 8.76 mmol, 3.0 equivs.). The re-
action mixture was briefly degassed by a stream of argon and
heated in a sealed tube to 1008C for 68 h. Upon coolingthe re-
action mixture was poured into water (20 mL) and extracted
with ether (10Â20 mL). Organic layers were washed with
brine (10 mL) and dried over anhydrous magnesium sulfate.
After removal of dryingagent, the solvent was evaporated un-
der reduced pressure and the crude mixture was separated by
column chromatography to give unreacted starting material
(0.364 g, 54%) as colorless oil and the desired product; yield:
0.212 g(76% based on the recovered startingmaterial); mp
62 63.58C, [a]2D7:þ207.58 (c 1.0, CH2Cl2).
3036, 2974, 2210, 1590, 1557 cmÀ1
.
trans-3-[(5S,6R)-5,6-Dihydroxycyclohex-1-enyl]-acryloni-
trile (3): Potassium azodicarboxylate (0.968 g, 4.99 mmol, 1.8
equivs.) was added to a cooled (À48C) stirred solution of
trans-3-[(5S,6R)-5,6-dihydroxycyclohexa-1,3-dienyl]-acryloni-
trile (1; 0.452 g, 2.77 mmol) in methanol (10 mL). Then a solu-
tion of glacial acetic acid (630 mL, 0.658 g, 10.97 mmol, 2.2
equivs. to PAD) in methanol (10 mL) was added dropwise
over a 2 hour period while maintainingthe temperature of
the reaction mixture below 08C. Upon completion of the addi-
tion the reaction mixture was allowed to slowly warm up to
room temperature over 5 h at which time the TLC analysis in-
dicated completion. Excess acetic acid was neutralized by the
addition of saturated sodium bicarbonate solution (5 mL)
and most of methanol was removed under reduced pressure.
The residue was diluted with water (5 mL) and aqueous phase
was extracted with ethyl acetate (6Â20 mL). Organic extracts
were washed with brine (2.5 mL) and dried over anhydrous so-
dium sulfate. Dryingagent was then removed by filtration and
the solvent was removed under reduced pressure. The residue
was almost completely dissolved in minimum amount of hot
ethyl acetate (ca. 508C), and a small amount of brown solid,
which remained was decanted. The supernatant was then care-
fully reheated and hexanes were added until the hot solution
appeared cloudy. Upon coolingto room temperature, the flask
was placed in freezer. The product was isolated as yellowish
needles by filtration and dried under vacuum; yield: 0.186 g
(41%); mp 84 858C; [a]2D7: À14.38 (c 1.0, CH2Cl2); Rf
Acknowledgements
We would like to thank Dr. Roland Wohlgemuth and Dr. Hans-
Peter E. Kohler for critical discussions and Dipl.-Ing. Ulrich Ba-
uer for his excellent technical support. Funding by the Swiss Pri-
ority Program Biotechnology (SPP Biotechnology) and Fluka
AG (Buchs, Switzerland) is gratefully acknowledged. We (JZ
and TH) are grateful for financial assistance from NSERC
and Brock University.
1
(EtOAc)¼0.47; H NMR: d¼6.96 (d, 1H, J¼16.5 Hz), 6.22
(t, 1H, J¼4.0 Hz), 5.64 (d, 1H, J¼16.5 Hz), 4.29 (d, 1H, J¼
4.0 Hz), 3.78 (m, 1H), 2.29 (m, 2H), 1.76 (m, 2H); 13C NMR:
d¼151.0, 141.1, 134.7, 118.5, 95.5, 68.9, 65.0, 25.3, 24.6; IR:
References
n¼3400, 3061, 3017, 2935, 2216, 1631, 1598 cmÀ1
.
[1] D. T. Gibson (Ed.), Microbial degradation ofaromatic
hydrocarbons, in: D. T. Gibson, V. Subramanian (Eds.),
Microbial Degradation ofOrganic Compounds , Marcel
Deckker, New York, 1984, pp. 181 252.
[2] D. R. Boyd, G. N. Sheldrake, Nat. Prod. Rep. 1998, 15,
309 324.
[3] S. M. Brown, T. Hudlicky, Organic Synthesis: Theory and
Practice, Vol. 2, Greenwich, JAI Press, 1993.
[4] T. Hudlicky, D. Gonzales, D. T. Gibson, Aldrichimica
Acta 1999, 32, 35 62.
[5] G. N. Sheldrake, Biologically Derived Arene cis-Diols as
Synthetic Building Blocks, in Chirality in Industry, (Eds.:
A. N. Collins, G. N. Sheldrake, J. Crosby), John Wiley &
Sons Ltd., Chichester, 1992, pp. 127 166.
[6] C. S. Butler, J. R. Mason, Adv. Microb. Physiol. 1997, 38,
47 84.
trans-3-[(3aR,7aS)-2,2-Dimethyl-3a,6,7,7a tetrahydro-
benzo[1,3]dioxol-4-yl]-acrylonitrile (4): trans-3-[(5S,6R)-5,6-
Dihydroxycyclohex-1-enyl]-acrylonitrile(3; 0.300 g , 1.82 mmol)
was dissolved in 2,2-dimethoxypropane (DMP) (10 mL,
81.5 mmol, 45 equivs.). To a stirred solution was then added a
catalytic amount of p-toluenesulfonic acid monohydrate, and
the reaction progress was monitored by TLC. Upon comple-
tion the reaction was quenched with saturated sodium bicar-
bonate solution (5 mL) and water (5 mL), excess DMP was re-
moved under reduced pressure. The aqueous phase was ex-
tracted with methylene chloride (5Â20 mL) and the organic
extracts were dried over anhydrous sodium sulfate overnight.
After removingthe dryingagent the filtrate was evaporated
to dryness with small amount of silica. Thus pre-adsorbed
crude product was then purified by column chromatography
(silica, hexane:EtOAc, 90:10). A colorless oil was obtained,
which crystallized upon standing; yield: 0.176 g (47%); mp
61 628C; [a]2D7: þ208.38 (c 1.0, CH2Cl2); Rf (hexane:EtOAc,
[7] B. D. Ensley, D. T. Gibson, A. L. Laborde, J. Bacteriol.
1982, 149, 948 954.
[8] B. D. Ensley, D. T. Gibson, J. Bacteriol. 1983, 155, 505
511.
[9] S. Kurkela, H. Lehv‰slaiho, E. T. Palva, T. H. Teeri, Gene
1988, 73, 355 362.
1
40:60)¼0.72; H NMR: d¼6.91 (d, 1H, J¼16.5 Hz), 6.25 (t,
1H, J¼4.0 Hz), 5.59 (d, 1H, J¼16.5 Hz), 4.56 (d, 1H, J¼
5.5 Hz), 4.37 (m, 1H), 2.36 (m, 1H), 2.13 (m, 1H), 1.94 (m,
1H), 1.75 (m, 1H), 1.37 (s, 3H), 1.31 (s, 3H); 13C NMR: d¼
940
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 933 942

Asymmetric Dihydroxylation of Cinnamonitrile
FULL PAPERS
[10] S. Harayama, M. Rekik, K. Timmis, Mol. Gen. Genet.
1986, 202, 226 234.
[36] B. B¸hler, I. Bollhalder, B. Hauer, B. Witholt, A. Schmid,
Biotechnol. Bioeng. 2003, 82, 833 842.
[11] S. Irie, S. Doi, T. Yorifuji, M. Takagi, K. Yano, J. Bacter-
iol. 1987, 169, 5174 5179.
[12] G. J. Zylstra, D. T. Gibson, J. Biol. Chem. 1989, 264,
14940 14946.
[13] E. Neidle, C. Hartnett, L. Ornston, A. Bairoch, M. Re-
kik, S. Harayama, J. Bacteriol. 1991, 173, 5385 5395.
[14] B. D. Erickson, F. J. Mondello, J. Bacteriol. 1992, 174,
2903 2912.
[37] L. E. H¸sken, H. H. Beeftink, J. A. M. de Bont, J. Wery,
Appl. Microbiol. Biotechnol. 2001, 55, 571 577.
[38] L. E. H¸sken, M. C. F. Dalm, J. Tramper, J. Wery,
J. A. M. de Bont, R. Beeftink, J. Biotechnol. 2001, 88,
11 19.
[39] A. Amanullah, C. J. Hewitt, A. W. Nienow, C. Lee, M.
Chartrain, B. C. Buckland, S. W. Drew, J. M. Woodley,
Biotechnol. Bioeng. 2002, 80, 239 249.
[15] J. V. Parales, A. Kumar, R. E. Parales, D. T. Gibson,
Gene 1996, 181, 57 61.
[16] C. Werlen, H.-P. E. Kohler, J. R. van der Meer, J. Biol.
Chem. 1996, 271, 4009 4016.
[17] S. Beil, B. Happe, K. N. Timmis, D. H. Pieper, Eur. J. Bi-
ochem. 1997, 247, 190 199.
[18] D. T. Gibson, J. R. Koch, C. L. Schuld, R. E. Kallio, Bio-
chemistry 1968, 7, 3795 3802.
[19] D. T. Gibson, E. R. Parales, Curr. Opin. Biotechnol. 2000,
11, 236 243.
[20] G. Bestetti, D. Bianchi, A. Bosetti, P. Di Gennaro, E.
Galli, B. Leoni, F. Pelizzoni, G. Sello, Appl. Microbiol.
Biotechnol. 1995, 44, 306 313.
[40] J. M. Carragher, W. S. McClean, J. M. Woodley, C. J.
Hack, Enzyme Microb. Technol. 2001, 28, 183 188.
[41] A. Amanullah, C. J. Hewitt, A. W. Nienow, C. Lee, M.
Chartrain, B. C. Buckland, S. W. Drew, J. M. Woodley,
Biotechnol. Bioeng. 2003, 81, 405 420.
[42] D. R. Boyd, J. Blacker, B. Byrne, H. Dalton, M. V. Hand,
S. C. Kelly, R. A. M. Oferrall, S. N. Rao, N. D. Sharma,
G. N. Sheldrake, J. Chem. Soc. Chem. Commun. 1994,
3, 313 314.
[43] A. Z. Walton, J. D. Stewart, Biotechnol. Prog. 2002, 18,
262 268.
[44] T. Hudlicky, E. E. Boros, H. F. Olivo, J. S. Merola, J. Org.
Chem. 1992, 57, 1026 1028.
[45] 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.
[46] D. R. Boyd, M. R. J. Dorrity, M. V. Hand, J. F. Malone,
N. D. Sharma, H. Dalton, D. J. Gray, G. N. Sheldrake, J.
Am. Chem. Soc. 1991, 113, 666 667.
[47] T. Hudlicky, H. F. Olivo, J. Am. Chem. Soc. 1992, 114,
9694 9696.
[48] X. Tian, T. Hudlicky, K. Kˆnigsberger, J. Am. Chem. Soc.
1995, 117, 3643 3644.
[21] C. C. R. Allen, D. R. Boyd, H. Dalton, N. D. Sharma,
S. A. Haughey, R. A. S. McMordie, B. T. McMurray,
G. N. Sheldrake, K. Sproule, J. Chem. Soc. Chem. Com-
mun. 1995, 119 120.
[22] M. J. Simon, T. D. Osslund, R. Saunders, B. D. Ensley, S.
Suggs, A. Harcourt, W.-C. Suen, D. L. Cruden, D. T. Gib-
son, G. J. Zylstra, Gene 1993, 127, 31 37.
[23] W.-C. Suen, D. T. Gibson, Gene 1994, 143, 67 71.
[24] P. Di Gennaro, G. Sello, D. Bianchi, P. D×Amico, J. Biol.
Chem. 1997, 272, 30254 30260.
[49] T. Hudlicky, X. Tian, K. Kˆnigsberger, R. Maurya, J.
Rouden, B. Fan, J. Am. Chem. Soc. 1996, 118, 10752
10765.
[25] F. J. Mondello, J. Bacteriol. 1989, 171, 1725 1732.
[26] A. Saito, T. Iwabuchi, S. Harayama, J. Bacteriol. 2000,
182, 2134 2141.
[50] M. Mandel, T. Hudlicky, L. D. Kwart, G. M. Whited, J.
Org. Chem. 1993, 58, 2331 2333.
[51] T. Hudlicky, H. Luna, J. D. Price, F. Rulin, Tetrahedron
Lett. 1989, 30, 4053 4054.
[52] M. Mandel, T. Hudlicky, L. Kwart, G. M. Whited, Col-
lect. Czech. Chem. Commun. 1993, 58, 2517 2522.
[53] M. Mandel, T. Hudlicky, J. Chem. Soc. Perkin Trans. 1
1993, 7, 741 743.
[54] V. Bui, T. V. Hansen, Y. Stenstrom, D. W. Ribbons, T.
Hudlicky, J. Chem. Soc. Perkin Trans. 1 2000, 1669 1672.
[55] I. P. Beletskaya, A. V. Cheprakov, Chem. Rev. 2000, 100,
3009 3066.
[56] X. Xu, G. Fakha, D. Sinou, Tetrahedron 2002, 58, 7539
7544.
[57] Z. Li, J. B. vanBeilen, W. A. Duetz, A. Schmid, A. de-
Raadt, H. Griengl, B. Witholt, Curr. Opin. Chem. Biol.
2002, 6, 136 144.
[27] H.-K. Chun, Y. Ohnishi, N. Misawa, K. Shindo, M. Hay-
ashi, S. Harayama, S. Horinouchi, Biosci. Biotechnol. Bi-
ochem. 2001, 65, 1774 1781.
[28] B. C. Buckland, S. W. Drew, N. C. Connors, M. M. Char-
train, C. Lee, P. M. Salmon, K. Gbewonyo, W. Zhou, P.
Gailliot, R. Singhvi, R. C. Olewinski, Jr. , W.-J. Sun, J.
Reddy, J. Zhang, B. A. Jackey, C. Taylor, K. E. Goklen,
B. Junker, R. L. Greasham, Metab. Eng. 1999, 1, 63 74.
[29] A. Amanullah, C. J. Hewitt, A. W. Nienow, C. Lee, M.
Chartrain, B. C. Buckland, S. W. Drew, J. M. Woodley,
Enzyme Microb. Technol. 2002, 31, 954 967.
[30] A. M. Collins, J. M. Woodley, J. M. Liddell, J. Ind. Micro-
biol. 1995, 14, 382 388.
[31] C. J. Hack, J. M. Woodley, M. D. Lilly, J. M. Liddell, En-
zyme Microb. Technol. 2000, 26, 530 536.
[32] M. A. Endoma, V. P. Bui, J. Hansen, T. Hudlicky, Org.
Process Res. Dev. 2002, 6, 525 532.
[33] S. Yildirim, T. T. Franco, B. Witholt, H.-P. E. Kohler, A.
Schmid, unpublished results.
[58] W. B. Motherwell, A. S. Williams, Angew. Chem. Int. Ed.
1995, 34, 2031 2033.
[34] S. Panke, M. G. Wubbolts, A. Schmid, B. Witholt, Bio-
technol. Bioeng. 2000, 69, 91 100.
[59] M. G. Quintana, H. Dalton, Enzyme Microb. Technol.
1999, 24, 232 236.
[35] B. B¸hler, B. Witholt, B. Hauer, A. Schmid, Appl. Envi-
ron. Microbiol. 2002, 68, 560 568.
[60] O. Favre-Bulle, B. Witholt, Enzyme Microb. Technol.
1992, 14, 931 937.
Adv. Synth. Catal. 2004, 346, 933 942
asc.wiley-vch.de
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
941

Selcuk Yildirim et al.
FULL PAPERS
[61] M. G. Wubbolts, O. Favre-Bulle, B. Witholt, Biotechnol.
Bioeng. 1996, 52, 301 308.
[62] O. Favre-Bulle, E. Weenink, T. Vos, H. Preusting, B.
Witholt, Biotechnol. Bioeng. 1993, 41, 263 272.
[63] R. O. Jenkins, G. M. Stephens, H. Dalton, Biotechnol. Bi-
oeng. 1986, 29, 873 883.
[70] J. Sambrook, E. F. Fritsch, T. Maniatis, Molecular Clon-
ing A laboratory Manual. 2nd edn, (Ed.: C. Nolan),
New York, Cold SpringHarbor Laboratory Press, 1989.
[71] S. Panke, A. Meyer, C. M. Huber, B. Witholt, M. G.
Wubbolts, Appl. Environ. Microbiol. 1999, 65, 2324
2332.
[64] A. A. Garcia, D. H. Kim, G. Whited, L. Kwart, W. An-
thony, C. Downie, Isolation & Purification 1994, 2, 19
25.
[65] R. M. Lynch, J. M. Woodley, M. D. Lilly, J. Biotechnol.
1997, 58, 167 175.
[66] R. E. Parales, K. Lee, S. M. Resnick, H. Jiang, D. J. Less-
ner, D. T. Gibson, J. Bacteriol. 2000, 182, 1641 1649.
[67] K. Pollmann, V. Wray, H. J. Hecht, D. H. Pieper, Micro-
biology-Sgm 2003, 149, 903 913.
[68] H. Suenaga, T. Watanabe, M. Sato, Ngadiman, K. Furu-
kawa, J. Bacteriol. 2002, 184, 3682 3688.
[72] A. J. Brazier, M. D. Lilly, Enzyme Microb. Technol. 1990,
12, 90 94.
[73] M. Chartrain, N. Ikemoto, C. Taylor, S. Stahl, V. Sanford,
K. Gbewonyo, C. Chirdo, C. Maxwell, J. Osoria, B. Buck-
land, G. Randolph, J. Biosci. Bioeng. 2000, 90, 321 327.
[74] D. Bianchi, A. Bosetti, D. Cidaria, A. Bernardi, I. Ga-
gliardi, P. D×Amico, Appl. Microbiol. Biotechnol. 1997,
47, 596 599.
[75] L. P. Wahbi, D. Gokhale, S. Minter, G. M. Stephens, En-
zyme Microb. Technol. 1996, 19, 297 306.
[76] J. Reddy, C. Lee, M. Neeper, R. Greasham, J. Zhang,
Appl. Microbiol. Biotechnol. 1999, 51, 614 620.
[69] N. Zhang, B. G. Stewart, J. C. Moore, R. L. Greasham,
D. K. Robinson, B. C. Buckland, C. Lee, Metab. Eng.
2000, 2, 339 348.
942
¹ 2004 WILEY-VCH VerlagGmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 933 942