Expanding the scope of alcohol dehydrogenases towards bulkier substrates: Stereo- and enantiopreference for α,α-dihalogenated ketones

Article abstract of DOI:10.1002/cctc.201300834

Alcohol dehydrogenases (ADHs) were identified as suitable enzymes for the reduction of the corresponding α,α-dihalogenated ketones, obtaining optically pure β,β-dichloro- or β,β-dibromohydrins with excellent conversions and enantiomeric excess. Among the different biocatalysts tested, ADHs from Rhodococcus ruber (ADH-A), Ralstonia sp. (RasADH), Lactobacillus brevis (LBADH), and PR2ADH proved to be the most efficient ones in terms of activity and stereoselectivity. In a further study, two racemic α-substituted ketones, namely α-bromo- α-chloro- and α-chloro-α-fluoroacetophenone were investigated to obtain one of the four possible diastereoisomers through a dynamic kinetic process. In the case of the brominated derivative, only the (1R)-enantiomer was obtained by using ADH-A, although with moderate diastereomeric excess (>99 % ee, 63 % de), whereas the fluorinated ketone exhibited a lower stereoselectivity (up to 45 % de). Bulking up: A series of β,β-dihalohydrins are obtained through alcohol dehydrogenase (ADH) catalyzed bioreduction of the synthesized α,α-dihalogenated ketones. Two racemic acetophenone derivatives are also subjected to this protocol to obtain stereoenriched alcohols through dynamic kinetic resolution (DKR).

Full text of DOI:10.1002/cctc.201300834

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201300834
Expanding the Scope of Alcohol Dehydrogenases towards
Bulkier Substrates: Stereo- and Enantiopreference for
a,a-Dihalogenated Ketones
Kinga Ke˛dziora,[a] Fabricio R. Bisogno,[a, b] Ivꢀn Lavandera,[a] Vicente Gotor-Fernꢀndez,[a]
Jose Montejo-Bernardo,[c] Santiago Garcꢁa-Granda,[c] Wolfgang Kroutil,[d] and
Vicente Gotor*[a]
Alcohol dehydrogenases (ADHs) were identified as suitable en-
zymes for the reduction of the corresponding a,a-dihalogenat-
ed ketones, obtaining optically pure b,b-dichloro- or b,b-dibro-
mohydrins with excellent conversions and enantiomeric
excess. Among the different biocatalysts tested, ADHs from
Rhodococcus ruber (ADH-A), Ralstonia sp. (RasADH), Lactobacil-
lus brevis (LBADH), and PR2ADH proved to be the most effi-
cient ones in terms of activity and stereoselectivity. In a further
study, two racemic a-substituted ketones, namely a-bromo-
a-chloro- and a-chloro-a-fluoroacetophenone were investigat-
ed to obtain one of the four possible diastereoisomers through
a dynamic kinetic process. In the case of the brominated deriv-
ative, only the (1R)-enantiomer was obtained by using ADH-A,
although with moderate diastereomeric excess (>99% ee,
63% de), whereas the fluorinated ketone exhibited a lower ste-
reoselectivity (up to 45% de).
Introduction
b,b-Dihalogenated alcohols, also called gem-dihalo alcohols or
b,b-dihalohydrins,[1] are a family of interesting compounds be-
cause of their versatility in organic synthesis,[2] and because of
their role as precursors of biologically active derivatives such
as antineoplastic drugs like mitotane.[3] Thus, owing to the
highly activated nature of these compounds, they can be used
as synthetic intermediates of interesting molecules such as al-
kenes,[4] terminal alkynes,[5] epoxides,[6] and a-methoxy alkyl
acetic acid derivatives.[7] Additionally, owing to their reactivity
in aqueous medium, they have been described as chemical an-
alogues of a-hydroxy aldehydes, opening the scope towards
[a] K. Ke˛dziora, Dr. F. R. Bisogno, Dr. I. Lavandera, Dr. V. Gotor-Fernꢀndez,
Prof. V. Gotor
Scheme 1. Synthetic applicability of the gem-dihalo alcohol core.
Department of Organic and Inorganic Chemistry
Instituto Universitario de Biotecnologꢁa de Asturias, University of Oviedo
Avenida Juliꢀn Claverꢁa s/n, 33006 Oviedo (Spain)
E-mail: vgs@uniovi.es
other types of substrates, for example, a-hydroxy acids
(Scheme 1).[8]
The preparation of the racemic derivatives can be achieved
by means of different synthetic approaches such as the Huns-
diecker reaction,[9] the decarboxylative heterodifunctionalisa-
tion of a,b-unsaturated carboxylic acids[10] or the reduction of
the corresponding ketone precursors.[11] Unfortunately, these
methods usually afford a mixture of products because of the
formation, among others, of dehydrated, hydrolysed or over-
reduced compounds. It is even more difficult to find in the lit-
erature an appropriate methodology to stereoselectively ach-
ieve these chiral precursors. Unfortunately, the selectivities ob-
tained in these processes by reduction of the ketones using
chiral oxaborolidines[12] or borane complexes were moderate
(<83%),[13] and, alternatively, dichlorocarbene CÀH insertion re-
[b] Dr. F. R. Bisogno
INFIQC-CONICET, Dpto. de Quꢁmica Orgꢀnica
Facultad de Ciencias Quꢁmicas
Universidad Nacional de Cꢂrdoba
Ciudad Universitaria, CP 5000, Cꢂrdoba (Argentina)
[c] Dr. J. Montejo-Bernardo, Prof. S. Garcꢁa-Granda
Departamento de Quꢁmica Fꢁsica y Analꢁtica
Universidad de Oviedo
Avenida Juliꢀn Claverꢁa s/n, Oviedo 33006 (Spain)
[d] Prof. W. Kroutil
Department of Chemistry, Organic and Bioorganic Chemistry
University of Graz
Heinrichstrasse 28, 8010 Graz (Austria)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201300834.
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1066 – 1072 1066

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
actions[14] led to incomplete conversions (<90%) starting from
an expensive enantiopure alcohol precursor.
Table 1. Preparation of a,a-dichloroacetophenones 2a–h.
Interestingly, the use of biocatalytic methods under mild re-
action conditions has allowed the selective synthesis of difluor-
ohydrins,[15] However, for the chlorinated or brominated coun-
terparts, and although the formation of by-products was mini-
mised, the enantioselectivities or yields obtained in these pro-
cesses were still not high enough. For instance, (R)-2,2-dibro-
Entry
X
t [h]
2a–h [%][a]
1
2
3
4
5
6
7
8
H
2-Me
2-Cl
3-OMe
3-NO2
3-Cl
4-NO2
4-Cl
16
48
48
16
16
16
16
16
77 (a)
82 (b)
77 (c)
75 (d)
80 (e)
78 (f)
78 (g)
90 (h)
mo-1-(4’-benzyloxy-3’-hydroxymethylphenyl)ethanol
was
obtained in 82% yield and 92% ee by using Rhodotorula rubra
whole cells in the presence of a surfactant.[16] The lipase-cata-
lysed resolution of 2,2-dichloro-1-phenylethanol (3a) was ach-
ieved with Amano Pseudomonas cepacia lipase (PSL), but 44%
conversion of the final product was reached after 142 h.[17] On
the other hand, the bioreduction of the a,a-dihalo ketone pre-
cursor 2a has been tested with whole cells from Geotrichum
candidum APG4[18] and baker’s yeast,[19] but stereoselectivities
remained modest (<55% ee). Based on the high selectivities
displayed by alcohol dehydrogenases (ADHs),[20] and as a-mon-
ohalogenated ketones are excellent substrates for these en-
zymes,[19,21] the bioreduction of a series of bulkier a,a-dihaloge-
nated acetophenones is presented herein. Several partially pu-
rified/overexpressed ADHs were tested to gain access to the
enantiopure b,b-dihalohydrins. Moreover, the reduction of two
racemic derivatives was also tried to study the formation of
two contiguous stereocentres in a dynamic kinetic resolution
(DKR) process catalysed by an ADH through racemisation in
basic conditions.
[a] Isolated yields of prochiral ketones 2a–h after flash chromatography.
In brackets appears the identification of the corresponding a,a-dihalo-
genated acetophenone, 2a–h obtained from 1a–h. For more details see
the Experimental Section.
Bioreduction of prochiral a,a-dihaloketones 2a–j
Once synthesised, the asymmetric bioreduction of a,a-dihalo-
ketones 2a–j was studied by using commercially available and
overexpressed ADHs. Owing to our previous experience with
similar a-halogenated substrates,[15a,21c,f] six enzymes were used
in this study, including also those accepting bulky–bulky ke-
tones as substrates: ADH-A from Rhodococcus ruber,[22] RasADH
from Ralstonia sp.[23] and SyADH from Sphingobium yanoi-
kuyae,[24] which are Prelog enzymes;[25] and on the other hand
LBADH from Lactobacillus brevis,[26] LKADH from Lactobacillus
kefir,[27] and PR2ADH that are anti-Prelog ADHs (see the Experi-
mental Section for more details). All these biocatalysts accept
aromatic ketones bearing a small substituent at alpha position
such as methyl or chloromethyl. Besides, RasADH and SyADH
can also reduce bulkier substrates.[23,24] Except for RasADH and
LKADH, for which glucose and glucose dehydrogenase (GDH)
were used to recycle a catalytic amount of the nicotinamide
cofactor, 2-propanol was employed as the hydrogen donor
(5% v/v). This is owing to the fact that RasADH and LKADH
work better under these conditions, as previously descri-
bed.[23c,28] For clarity, the best results for the synthesis of both
3a–j enantiomers are collected in Table 2, and in the Support-
ing Information more detailed information about the screening
process can be found.
Results and Discussion
Preparation of a,a-dihaloacetophenones and the corre-
sponding alcohols
The synthesis of a,a-dichloroacetophenones 2a–h was per-
formed in good to very high yields starting from commercially
available acetophenones 1a–h, bearing different substitution
pattern in the aromatic ring, by reaction with a 2-fold molar
excess of N-chlorosuccinimide (NCS) in the presence of para-
toluenesulfonic acid (p-TsOH) using acetonitrile as a solvent at
508C (Table 1).
A lower reactivity for the ortho-substituted derivatives 2b–
c was observed, thus longer reaction times were required in
these cases (entries 2 and 3), probably owing to steric hin-
drance. In addition, to achieve the synthesis of 2,2-dichloro-1-
(3,4-dichlorophenyl)ethanone (2i), the corresponding a-chloro-
acetophenone derivative 1i was used as starting material utilis-
ing 1.1 equivalents of NCS, obtaining 2i in 83% yield. Starting
from 1a with a 3-fold molar excess of N-bromosuccinimide
(NBS), 2,2-dibromoacetophenone 2j was achieved in 81% iso-
lated yield. Racemic dihalohydrins 3a–j were obtained in good
to very high yields (70–96%) by reduction of the correspond-
ing ketones with NaBH4 in MeOH at room temperature (see
the Supporting Information).
Satisfyingly, from the twenty possible enantiopure alcohols,
fifteen were obtained in enantiomerically pure form, finding
ADH-A, LBADH and PR2ADH as the most versatile biocatalysts
for the stereoselective reduction of a,a-dihaloacetophenones,
which was achieved in eighteen cases with at least 90% con-
version. Initially, a,a-dichloroacetophenone (2a) was studied
yielding selectively either enantiomer of alcohol 3a in conver-
sions over 95% by using a Prelog enzyme (ADH-A, entry 1) or
anti-Prelog reductases (PR2ADH and LBADH, entries 2 and 3).
Then, the influence of electron-donating (Me or OMe) or elec-
tron-withdrawing substituents (Cl or NO2), at different positions
in the aromatic ring was studied for a,a-dichloroacetophenone
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1066 – 1072 1067

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
reoselectivities for anti-Prelog enzymes PR2ADH and LBADH
(entry 22 and see also the Supporting Information), whereas
RasADH was found as the best Prelog enzyme (entry 21), ADH-
A leading to low conversions and high enantiomeric excess,
and SyADH produced (R)-3j with almost complete conversion
but moderate ee (see the Supporting Information).
Table 2. Asymmetric bioreduction of ketones 2a–j.
Conv. [%][a]
ee [%][b]
Preparation of racemic a,a-dihaloacetophenones
Entry
X
Y
Z
ADH
1 (a)
2 (a)
3 (a)
4 (b)
5 (b)
6 (c)
H
H
H
2-Me
2-Me
2-Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Br
Br
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Br
Br
A
PR2
LB
E. coli/Ras
PR2
E. coli/Ras
LB
E. coli/A
PR2
LB
A
LB
A
LB[c]
A
PR2
A
LB
E. coli/A[c]
LB[c,d]
E. coli/Ras
LB[c,e]
96
99
99
99
43
99
12
99
99
99
99
96
99
99
97
99
98
91
99
90
99
99
>99 (R)
>99 (S)
>99 (S)
>99 (R)
>99 (S)
>99 (R)
91 (S)
>99 (R)
>99 (S)
>99 (S)
>99 (R)
>99 (S)
96 (R)
>99 (S)
>99 (R)
>99 (S)
>99 (R)
>99 (S)
99 (R)
Owing to the high stereoselectivities obtained for ketones
2a–j, and to broaden the applicability of the already tested al-
cohol dehydrogenases, the bioreduction of two halogenated
racemic ketones was envisaged, thus 2-bromo-2-chloro-1-phe-
nylethanone (2k) and 2-chloro-2-fluoro-1-phenylethanone (2l)
were prepared by following standard procedures. For 2k,
a-chloroacetophenone was reacted with NBS in the presence
of p-TsOH in MeCN at 508C for 16 h, yielding the ketone in
85% isolated yield. On the other hand, fluoro ketone 2l was
obtained by following the procedure described by Yamazaki
et al., starting from ethyl chlorofluoroacetate (56% yield).[29]
7 (c)
2-Cl
8 (d)
9 (d)
10 (d)
11 (e)
12 (e)
13 (f)
14 (f)
15 (g)
16 (g)
17 (h)
18 (h)
19 (i)
20 (i)
21 (j)
22 (j)
3-OMe
3-OMe
3-OMe
3-NO2
3-NO2
3-Cl
3-Cl
4-NO2
4-NO2
4-Cl
Bioreduction of racemic a,a-dihaloketones 2k–l
4-Cl
Owing to the acidity of the a-proton, it was expected that ra-
cemisation of the substrates would occur in situ (Scheme 2),
making a DKR process feasible obtaining, in the ideal case, one
3,4-Cl2
3,4-Cl2
H
98 (S)
95 (R)
99 (S)
H
[a] Conversion values calculated by GC. [b] Enantiomeric excess of alco-
hols calculated by using chiral GC or HPLC indicating their absolute con-
figuration in brackets [note the switch in the Cahn–Ingold–Prelog (CIP)
priority]. [c] DMSO (2%v/v) was added. [d] 48 h and 4.5 U of enzyme em-
ployed. [e] 4.5 U of enzyme employed.
Scheme 2. Interconversion of both 2k or 2l enantiomers through an enolate
intermediate.
derivatives 2b–i. Clear trends were observed as follows:
1) ADHs led to good levels of activity and stereoselectivity to
those substrates with the presence of substituents in the meta
or para position (entries 8–20), whereas for ortho-substituted
acetophenones 2b,c (entries 4–7), only the Prelog enzyme
RasADH (entries 4 and 6) allowed the isolation of the corre-
sponding (R)-alcohols in quantitative yield and enantiopure
form. This is especially relevant because the bioreduction of
ortho-substituted acetophenones remains usually hampered. In
a recent contribution, RasADH exhibited good activity for simi-
lar ketones;[23b] 2) bulky–bulky ADH from Ralstonia sp. was
identified as the best enzyme for highly hindered substrates;[23]
3) ADH-A demonstrated also high versatility acting as a very
selective enzyme,[22] only leading to low conversion in the case
of the 2-methyl derivative 2b (see also the Supporting Infor-
mation). Thus, a correct choice between RasADH or ADH-A al-
lowed the synthesis of the (R)-alcohols with excellent stereose-
lectivities and conversions; (iv) the poorest results were gener-
ally attained with LKADH, which seemed to be not suitable for
dihalohydrins synthesis.
of the four possible diastereoisomer products.[28,30] These enan-
tioenriched alcohols would be of high interest because in a fur-
ther step they could be selectively modified to obtain more
complex and valuable structures as a result of the different re-
activity of both halide atoms.
Therefore, we firstly performed a screening with substrate
2k (Table 3). In a first set of experiments, it was observed that
ADH-A, RasADH[23a] and LBADH[26] were the best biocatalysts in
terms of activity and selectivity (entries 1–3). Although still far
away from a perfect diastereoselectivity, it is remarkable that
these enzymes could distinguish between these two halide
atoms, because previous results with structurally similar ke-
tones did not show high induction levels.[23b] As ADH-A was
the enzyme displaying better diastereoselectivity favouring the
formation of the syn diastereomer (59%, entry 1), we tried to
optimise the process by changing several reaction parameters
such as pH or temperature, thus differentially modifying the
rate of the enzymatic and the racemisation reactions leading
to improved diastereomeric excess (de) values. In this regard,
low temperatures had a negative influence in the activity of
Finally, the bioreduction of a bulkier ketone possessing two
bromine atoms at a-position instead of chlorines, a,a-dibro-
moacetophenone (2j), was also analysed, finding complete ste-
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1066 – 1072 1068

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Table 3. Bioreduction of racemic ketone 2k (t=24 h).
Entry
ADH
T [8C] pH 3k [%][a] ee [%][b] de [%][b]
1
2
3
4
5
6
7
E. coli/ADH-A
E. coli/RasADH 30
LBADH
E. coli/ADH-A
E. coli/ADH-A
E. coli/ADH-A
LBADH
30
7.5 99
7.5 95
7.5 65
7.5 18
7.5 86
8.5 98
8.5 70
>99
>99
>99
>99
>99
>99
>99
59 (1R,2R)
1 (1R,2RS)
26 (1S,2R)
58 (1R,2R)
62 (1R,2R)
63 (1R,2R)
33 (1S,2R)
30
4
40
30
30
[a] Conversion values measured by GC. [b] Enantiomeric and diastereo-
meric excess measured by chiral HPLC (note the switch in the CIP priori-
ty).
the biocatalyst (entry 4), but a higher temperature or pH did
not influence the de observed (entries 5–7). Higher pH values
afforded the decomposition of both substrate and product
forming, among other, benzoic acid.
With the aim of gaining a deeper insight in this DKR with
ADH-A, we followed the reaction time course to analyse the
rate of the racemisation step together with the de. In the case
of an inefficient racemisation rate, we would detect a decrease
of the diastereomeric excess of the alcohol product within the
reaction progress (Figure 1). As can be seen, even at low con-
Figure 2. Conversion (&) and de values (&) after 24 h in the E. coli/ADH-A-
catalysed bioreduction of ketone 2k into 3k at 308C and pH 7.5 with differ-
ent proportions of: a) n-hexane; and b) Ammoeng 102, as cosolvents. In all
cases, the ee values were higher than 97%.
best scenario, the DKR outcome.[30a] From the results attained,
it can be summarised that both biphasic and monophasic
media did not have an influence in the enantio- and diastereo-
selectivity of the process, suggesting that a better biocatalyst
to achieve this goal might be constructed by active-site archi-
tecture modification rather than medium engineering.
Crystallisation of acylated alcohol 4
&
^
Figure 1. Conversion ( ) and de values ( ) in the E. coli/ADH-A-catalysed bio-
reduction of ketone 2k into 3k at 308C and pH 7.5. In all cases, ee values
were higher than 99%.
Different attempts were made to obtain suitable crystals for
X-ray diffraction analysis to confirm the relative and absolute
configuration of the stereogenic centres of 3k. The most suc-
cessful approach was achieved, converting the optically active
alcohol obtained from the ADH-A-catalysed bioreduction of 2k
into ester 4, after reaction with 4-nitrobenzoyl chloride in the
presence of triethylamine in dry dichloromethane, and subse-
quent crystallisation using a mixture of diethyl ether and
n-hexane. As can be seen in Figure 3, both stereogenic centres
presented the (R)-configuration, and data for the CÀOH bond
are in agreement with the known stereopreference exhibited
by ADH-A as a Prelog enzyme, which determined the absolute
configuration at position 2 bearing both halogens.[32]
versions, the de values remained almost unaltered during the
whole process, showing that the racemisation rate was fast
enough for the DKR process.
Owing to the fact that strong basic conditions decomposed
both 2k and 3k, several bases in equimolar amounts were
added into the reaction medium to study their effect in the
DKR process as previously described by us.[30a] Thus, DBU (final
pHꢀ9.0), piperidine (final pHꢀ9.0), pyridine (final pHꢀ7.5)
and triethylamine (final pHꢀ8.8) were employed, but no re-
markable improvement in the de was detected (see the Sup-
porting Information for more details).
From the data shown in Table 3, it is remarkable that ADH-A
produced the (1R,2R)-3k diastereoisomer with moderate dia-
stereomeric excess (63% de), which stands for a syn configura-
tion, whereas LBADH slightly preferred the formation of the
(1S,2R)-3k diastereoisomer (33% de), which accounts for an
anti configuration. On the other hand, it is noteworthy that
the reduction of racemic ketone 2k with sodium borohydride
Finally, the effect of a non-miscible organic solvent such as
n-hexane, or a miscible ionic liquid, Ammoeng 102, which al-
ready proved to be compatible with ADH-A,[31] was also mea-
sured (Figure 2). The use of an external additive could influ-
ence both ADH selectivity and reaction rates favouring, in the
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1066 – 1072 1069

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
choice of the enzyme. A series of dihalohydrins bearing differ-
ent substitutions in the phenyl ring were obtained with enan-
tiomeric excess values over 95% for both (R)- and (S)-enantio-
mers by the correct selection of the ADH for the bioreduction
process. In addition, eighteen out of twenty of these alcohol
enantiomers were achieved with over 90% conversion, finding
lower conversions for the ortho-substituted substrates.
Moreover, the asymmetric bioreduction of two racemic ke-
tones, namely a-bromo-a-chloroacetophenone and a-chloro-
a-fluoroacetophenone, was also studied. The use of an organic
base or cosolvent, and the modification of the temperature or
the pH did not have a significant effect on the stereoselectivity
of the DKR process. For the first substrate, the corresponding
enantiopure alcohol was obtained in excellent conversions,
albeit with moderate diastereomeric excess. By using X-ray dif-
fraction, the absolute configuration of the major diastereoiso-
mer obtained could be assigned. For the second racemic
ketone, the ADHs could not differentiate between both halo-
gen atoms. Overall, the bioreduction of a,a-dihaloacetophe-
nones was studied by using different ADHs, giving access to
valuable enantiopure b,b-dihalohydrins selecting the proper
biocatalyst. This study enables the future application of these
enzyme-catalysed processes in the syntheses of more complex
chiral compounds.
Figure 3. X-Ray structure of ester (R,R)-4 synthesised through ADH-A-cata-
lysed bioreduction of racemic ketone 2k.
led to the formation of the racemic mixtures at a proportion
3:1 (50% de) favouring the anti diastereoisomer. The opposite
diastereopreference displayed by ADH-A compared to NaBH4
or LBADH-catalysed reductions is a very interesting feature,
and will be the object of further studies.
Next, the DKR of 2-chloro-2-fluoro-1-phenylethanone (2l)
was also tried under the best conditions found for ketone 2k,
but as shown in Table 4, although excellent conversions and
Table 4. Bioreduction of racemic ketone 2l at pH 7.5 and 308C (t=24 h).
Experimental Section
Entry
ADH
3l [%][a]
ee [%][b]
de [%][b]
Overexpressed ADHs from Rhodococcus ruber (E. coli/ADH-A), from
Ralstonia species (E. coli/RasADH) and Sphingobium yanoikuyae
(E. coli/SyADH) were used as lyophilised cells.[23c,24,33] Glucose dehy-
drogenase (GDH002, 30 UmgÀ1), ADH-A (20 UmgÀ1), PR2ADH
1
2
3
4
E. coli/ADH-A
E. coli/RasADH
LBADH
99
99
99
98
98
97
>99
59
5 (1R,2RS)
<1 (1R,2RS)
1 (1S,2RS)
E. coli/SyADH
45 (1R,2RS)
(0.13 UmgÀ1), and LBADH from Lactobacillus brevis (3.7 UmLÀ1
)
[a] Conversion values measured by GC. [b] Enantiomeric and diastereo-
meric excess measured by chiral HPLC (note the switch in the CIP priori-
ty).
were purchased from Codexis. LKADH from Lactobacillus kefir
(0.42 UmgÀ1) was obtained from Fluka. For the bioreduction pro-
cesses, Tris–H2SO4 buffer was employed in all cases with a-bromi-
nated ketones to avoid undesired SN2 reactions.
enantioselectivities were observed, low de values were ach-
ieved. Sphingobium yanoikuyae ADH overexpressed in E. coli af-
forded the highest value of de (45%), although with low ee
(59%, entry 4), whereas the other biocatalysts studied gave
access to the enantiopure alcohol 3l but with vanished diaste-
reoselectivity (entries 1–3). The effect exerted by the fluorine in
the diastereomeric excess values was remarkable, probably as
a result of a better recognition of the bulky Br atom in sub-
strate 2k as that of fluorine in ketone 2l.
Syntheses
Prochiral ketones 2a–h (general procedure): To a solution of the
corresponding acetophenone 1a–h (3.7 mmol) and p-TsOH
(708.8 mg, 3.7 mmol) in acetonitrile (10 mL), NCS was added (1 g,
7.6 mmol). The reaction mixture was heated at 508C and stirred till
disappearance of the starting material (16–48 h). After completion,
the solvent was evaporated under reduced pressure and the crude
was purified using flash chromatography (50–80% CH2Cl2/hexane)
yielding the corresponding prochiral ketones (see Table 1).[34]
2,2-Dichloro-1-(3,4-dichlorophenyl)ethanone (2i): To a solution of
a-chloroacetophenone 1i (280 mg, 1.25 mmol) and p-TsOH
(238 mg, 1.25 mmol) in acetonitrile (4 mL), NCS was added
(184 mg, 1.37 mmol). The reaction mixture was heated at 508C and
stirred till disappearance of the starting material (16 h). After com-
pletion of the reaction, the solvent was evaporated under reduced
pressure and the crude was purified using flash chromatography
(20% CH2Cl2/hexane) yielding product 2i as a white solid (0.27 g,
83%).
Conclusions
The successful preparation of a,a-dihaloacetophenones as well
as an ADH selection guideline for their stereoselective reduc-
tion was provided. Enantioenriched dihalohydrins are precur-
sors in the chemical synthesis of a wide number of valuable
compounds, but their selective synthesis by traditional chemi-
cal methods is hampered by low asymmetric induction or for-
mation of by-products. Thus, different ADHs under mild reac-
tion conditions in aqueous medium were considered as suit-
able catalysts yielding both enantiomers depending on the
2,2-Dibromo-1-phenylethanone (2j): To a solution of acetophenone
(1a, 250 mg, 2.08 mmol) and p-TsOH (400 mg, 2.08 mmol) in aceto-
nitrile (10 mL), NBS was added (1.11 g, 6.24 mmol). The reaction
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1066 – 1072 1070

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
mixture was heated at 508C and stirred till disappearance of the
starting material (16 h). After completion of the reaction, the sol-
vent was evaporated under reduced pressure and the crude was
purified using flash chromatography (33% CH2Cl2/hexane) yielding
product 2j as a white solid (467 mg, 81%).
Ketones 2a–l with LKADH: In an Eppendorf vial (1.5 mL), LKADH
(7 mg, 3 U) were added to a 510 mL volume of Tris–HCl or Tris–
H2SO4 buffer (50 mm, pH 7.5), NADPH (60 mL of a 10 mm solution,
final concentration: 1 mm), glucose (30 mmol of a 1m solution,
50 mm), glucose dehydrogenase (10 mL, 3 U) and the correspond-
ing ketone (2a–l, 25 mm). The reaction tubes were shaken horizon-
tally at 308C for 24 h and 150 rpm. After that time, the reaction
products were extracted with EtOAc (2ꢃ0.5 mL). The organic layers
were separated by centrifugation (1.5 min, 13000 rpm) and dried
over anhydrous Na2SO4. Conversion and ee values were determined
by GC or HPLC analysis.
Racemic 2-bromo-2-chloro-1-phenylethanone (2k): To a solution of
a-chloroacetophenone 1k (500 mg, 3.2 mmol) and p-TsOH
(615 mg, 3.2 mmol) in acetonitrile (14 mL), NBS was added
(863 mg, 4.85 mmol). The reaction mixture was heated at 508C and
stirred till disappearance of the starting material (16 h). After com-
pletion of the reaction, the solvent was evaporated under reduced
pressure and the crude was purified using flash chromatography
(20% CH2Cl2/hexane) yielding 2k as a white solid (0.64 g, 85%).
Ketones 2a–l with PR2ADH: In an Eppendorf vial (1.5 mL), PR2ADH
(23 mg, 3 U), a 510 mL volume of Tris–HCl or Tris–H2SO4 buffer
(50 mm, pH 7.5) were added, NADH (60 mL of a 10 mm solution,
final concentration: 1 mm), 2-propanol (30 mL, 5% v/v) and the cor-
responding ketone (2a–l, 25 mm). The reaction tubes were shaken
horizontally at 308C for 24 h and 150 rpm. After that time, the re-
action products were extracted with EtOAc (2ꢃ0.5 mL). The organ-
ic layers were separated by centrifugation (1.5 min, 13000 rpm)
and dried over anhydrous Na2SO4. Conversion and ee values were
determined by GC or HPLC analysis.
Racemic 2-chloro-2-fluoro-1-phenylethanone (2l): To a solution of
ethyl chlorofluoroacetate (4.35 mmol, 0.5 mL) in dry toluene (5 mL)
at À788C under nitrogen atmosphere, 1.1 equiv. of phenyl magne-
sium bromide (1.6 mL of a 3m solution in Et2O) was added drop-
wise and the reaction was stirred for one hour. Following that
time, the reaction mixture was warmed up to 08C and then left for
10 min prior to the quenching with ammonium chloride (saturated
solution). The crude was extracted with Et2O (3ꢃ10 mL), dried over
anhydrous Na2SO4 and was slowly evaporated under reduced pres-
sure in an ice bath to prevent the loss of the volatile product. The
crude mixture was purified using flash chromatography (100%
pentane to 70% pentane/CH2Cl2) yielding 2l as a white crystal
solid (0.42 g, 56% yield).[29]
Ketones 2a–l with E. coli/RasADH: To a 15 mg portion of overex-
pressed E. coli/RasADH (lyophilised cells) in an Eppendorf vial
(1.5 mL), Tris–HCl or Tris–H2SO4 buffer (510 mL, 50 mm, pH 7.5) were
added, NADPH (60 mL of a 10 mm solution, final concentration:
1 mm), glucose (30 mmol of a 1m solution, 50 mm), glucose dehy-
drogenase (10 mL, 3 U) and the corresponding ketone (2a–l,
25 mm). The reaction tubes were shaken horizontally at 308C for
24 h and 150 rpm. After that time, the reaction products were ex-
tracted with EtOAc (2ꢃ0.5 mL). The organic layers were separated
by centrifugation (1.5 min, 13000 rpm) and dried over anhydrous
Na2SO4. Conversion and ee values were determined by GC or HPLC
analysis.
Bioreductions
Ketones 2a–l with E. coli/ADH-A: To a 15 mg portion of overex-
pressed E. coli/ADH-A (lyophilised cells) in an Eppendorf vial
(1.5 mL), Tris–HCl or Tris–H2SO4 buffer (510 mL, 50 mm, pH 7.5),
NADH (60 mL of a 10 mm solution, final concentration: 1 mm), 2-
propanol (30 mL, 5% v/v), and the corresponding ketone (2a–l,
25 mm) were added. The reaction tubes were shaken horizontally
at 308C for 24 h and 150 rpm. After that time, the reaction prod-
ucts were extracted with EtOAc (2ꢃ0.5 mL). The organic layers
were separated by centrifugation (1.5 min, 13000 rpm) and dried
over anhydrous Na2SO4. Conversion and ee values were determined
by GC or HPLC analysis.
Ketones 2a–l with E. coli/SyADH: To a 15 mg portion of overex-
pressed E. coli/SyADH (lyophilised cells) in an Eppendorf vial
(1.5 mL), Tris-HCl or Tris-H2SO4 buffer (510 mL, 50 mm, pH 7.5) were
added, NADPH (60 mL of a 10 mm solution, final concentration:
1 mm), 2-propanol (30 mL, 5% v/v) and the corresponding ketone
(2a–l, 25 mm). The reaction tubes were shaken horizontally at
308C for 24 h and 150 rpm. After that time, the reaction products
were extracted with EtOAc (2ꢃ0.5 mL). The organic layers were
separated by centrifugation (1.5 min, 13000 rpm) and dried over
anhydrous Na2SO4. Conversion and ee values were determined by
GC or HPLC analysis.
Scale-up (2a with E. coli/ADH-A): In an Erlenmeyer flask (10 mL),
E. coli/ADH-A (100 mg) was suspended in Tris–HCl buffer (3.6 mL,
50 mm, pH 7.5, 1 mm NADH) and preincubated for 30 min at 308C.
Then, ketone 2a (50 mg, 0.26 mmol) and 2-propanol (0.4 mL, 10%
v/v) were added to the mixture. The reaction was shaken at 308C
and 250 rpm for 48 h. After incubation, the enzymatic reaction was
stopped by extraction with EtOAc (3ꢃ5 mL). The organic layers
were combined and dried over Na2SO4. The solvent was concen-
trated under vacuum, furnishing the enantiopure alcohol (R)-3a
(isolated yield: 65%).
Acknowledgements
This project is supported by the BIOTRAINS Marie Curie Initial
Training Network, financed by the European Union through the
7th Framework People Programme (grant agreement number
238531). Financial support from the Spanish Ministerio de Ciencia
e Innovaciꢂn (MICINN-12-CTQ2011-24237), Ministerio de Econo-
mꢁa y Competitividad (MAT2010-15094, Factorꢁa de Cristaliza-
ciꢂn—Consolider Ingenio 2010), ERDF and the Principado de As-
turias (SV-PA-13-ECOEMP-42 and SV-PA-13-ECOEMP-43) is also
gratefully acknowledged. F.R.B. acknowledges INFIQC-CONICET
and Universidad Nacional de Cꢂrdoba. I.L. acknowledges MICINN
for his research contract under the Ramꢂn y Cajal Program.
Ketones 2a–l with LBADH: In an Eppendorf vial (1.5 mL), LBADH
(10 mL, 3 U) was added to a 450 mL volume of Tris–HCl or Tris–
H2SO4 buffer (50 mm, pH 7.5), followed by NADPH (60 mL of
a 10 mm solution, final concentration: 1 mm), MgCl2 (60 mL of
a 10 mm solution, final concentration: 1 mm), 2-propanol (30 mL,
5% v/v) and the corresponding ketone (2a–l, 25 mm). The reaction
tubes were shaken horizontally at 308C for 24 h and 150 rpm. After
that time, the reaction products were extracted with EtOAc (2ꢃ
0.5 mL). The organic layers were separated by centrifugation
(1.5 min, 13000 rpm) and dried over anhydrous Na2SO4. Conversion
and ee values were determined by GC or HPLC analysis.
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1066 – 1072 1071

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Gotor, Adv. Synth. Catal. 2010, 352, 1657–1661; d) J. H. Schrittwieser, I.
Lavandera, B. Seisser, B. Mautner, W. Kroutil, Eur. J. Org. Chem. 2009,
2293–2298; e) E. Alanvert, C. Doherty, T. S. Moody, N. Nesbit, A. S.
Rowan, S. J. C. Taylor, F. Vaughan, T. Vaughan, J. Wiffen, I. Wilson, Tetra-
hedron: Asymmetry 2009, 20, 2462–2466; f) F. R. Bisogno, I. Lavandera,
W. Kroutil, V. Gotor, J. Org. Chem. 2009, 74, 1730–1732.
Keywords: enantioselectivity · enzyme catalysis · halogens ·
ketones · reduction
[1] These compounds are also described elsewhere as a,a-dihalogenated
alcohols or a,a-dihalohydrins but we consider this naming inappropri-
ate for this family of compounds.
[2] a) K. Shibatomi, A. Narayama, Y. Soga, T. Muto, S. Iwasa, Org. Lett. 2011,
13, 2944–2947; b) J. M. Concellꢄn, P. L. Bernad, J. A. Pꢅrez-Andrꢅs,
Angew. Chem. Int. Ed. 1999, 38, 2384–2386; Angew. Chem. 1999, 111,
2528–2530.
[3] S. Hahner, M. Fassnacht, Curr. Opin. Invest. Drugs 2005, 6, 386–394.
[4] H. Kakiya, H. Shinokubo, K. Oshima, Tetrahedron 2001, 57, 10063–
10069.
[5] Z. Wang, J. Yin, S. Campagna, J. A. Pesti, J. M. Fortunak, J. Org. Chem.
1999, 64, 6918–6920.
[6] R. N. McDonald, R. C. Cousins, J. Org. Chem. 1980, 45, 2976–2984.
[7] a) C. Lee, J. Kim, H. Lee, S. Lee, Y. Kho, J. Nat. Prod. 2001, 64, 659–660;
b) R. Ulrich, D. Martin, J. Med. Chem. 1992, 35, 2238–2243.
[8] B. Singh, P. Gupta, A. Chaubey, R. Parshad, S. Sharma, S. C. Taneja, Tetra-
hedron: Asymmetry 2008, 19, 2579–2588.
[9] a) D. Naskar, S. Roy, Tetrahedron 2000, 56, 1369–1377; b) S. Chowdhury,
S. Roy, Tetrahedron Lett. 1996, 37, 2623–2624.
[10] S. Khamarui, D. Sarkar, P. Pandit, D. K. Maiti, Chem. Commun. 2011, 47,
12667–12669.
[22] a) E. Hamnevik, C. Blikstad, S. Norrehed, M. Widersten, J. Mol. Catal. B,
DOI: 10.1016/j.molcatb.2013.10.023; b) M. Karabec, A. Łyskowski, K. C.
Tauber, G. Steinkellner, W. Kroutil, G. Grogan, K. Gruber, Chem. Commun.
2010, 46, 6314–6316.
[23] a) J. Kulig, A. Frese, W. Kroutil, M. Pohl, D. Rother, Biotechnol. Bioeng.
2013, 110, 1838–1848; b) J. Kulig, R. C. Simon, C. A. Rose, S. M. Husain,
M. Hꢇckh, S. Lꢆdeke, K. Zeitler, W. Kroutil, M. Pohl, D. Rçther, Catal. Sci.
Technol. 2012, 2, 1580–1589; c) I. Lavandera, A. Kern, B. Ferreira-Silva, A.
Glieder, S. de Wildeman, W. Kroutil, J. Org. Chem. 2008, 73, 6003–6005.
[24] I. Lavandera, A. Kern, V. Resch, B. Ferreira-Silva, A. Glieder, W. M. F.
Fabian, S. de Wildeman, W. Kroutil, Org. Lett. 2008, 10, 2155–2158.
[25] This article specifies the stereopreference of some ADHs, see: V. Prelog,
Pure Appl. Chem. 1964, 9, 119–130.
[26] a) S. Leuchs, L. Greiner, Chem. Biochem. Eng. Q. 2011, 25, 267–281; b) M.
Wolberg, W. Hummel, C. Wandrey, M. Mꢆller, Angew. Chem. Int. Ed.
2000, 39, 4306–4308; Angew. Chem. 2000, 112, 4476–4478.
[27] W. Hummel, Appl. Microbiol. Biotechnol. 1990, 34, 15–19.
[28] A. Cuetos, A. Rioz-Martꢁnez, F. R. Bisogno, B. Grischek, I. Lavandera, G. de
Gonzalo, W. Kroutil, V. Gotor, Adv. Synth. Catal. 2012, 354, 1743–1749.
[29] T. Yamazaki, T. Terajima, T. Kawasaki-Takasuka, Tetrahedron 2008, 64,
2419–2424.
[11] B. L. Jensen, J. Jewett-Bronson, S. B. Hadley, L. G. French, Synthesis 1982,
732–735.
[30] For other recent dynamic protocols involving ADHs, see: a) J. Mangas-
Sꢀnchez, E. Busto, V. Gotor, V. Gotor-Fernꢀndez, Org. Lett. 2013, 15,
3872–3875; b) F. Xu, J. Y. L. Chung, J. C. Moore, Z. Liu, N. Yoshikawa,
R. S. Hoerrner, J. Lee, M. Royzen, E. Cleator, A. G. Gibson, R. Dunn, K. M.
Maloney, M. Alam, A. Goodyear, J. Lynch, N. Yasuda, P. N. Devine, Org.
Lett. 2013, 15, 1342–1345; c) D. Kalaitzakis, I. Smonou, Eur. J. Org.
Chem. 2012, 43–46; d) G. A. Applegate, R. W. Cheloha, D. L. Nelson,
D. B. Berkowitz, Chem. Commun. 2011, 47, 2420–2422.
[31] a) G. de Gonzalo, I. Lavandera, K. Faber, W. Kroutil, Org. Lett. 2007, 9,
2163–2166; b) G. de Gonzalo, I. Lavandera, K. Durchschein, D. Wurm, K.
Faber, W. Kroutil, Tetrahedron: Asymmetry 2007, 18, 2541–2546.
[32] Crystallographic data: C15H11BrClNO4, M=384.61, monoclinic, a=
12.2449(7) ꢈ, b=10.0624(8) ꢈ, c=13.1537(7) ꢈ, b=90.428(5)8, V=
1620.7(2) ꢈ3, T=293(2) K, space group P21, Z=4. CCDC 963560 contains
the crystallographic data for the structure of ester (R,R)-4. These data
can be obtained free of charge from the Cambridge Crystallographic
Data Centre through www.ccdc.cam.ac.uk/data_request/cif.
[12] M. Nakagawa, T. Kawate, T. Kakikawa, H. Yamada, T. Matsui, T. Hino, Tet-
rahedron 1993, 49, 1739–1748.
´
[13] P. V. Ramachandran, B. Gong, A. V. Teodorovic, J. Fluorine Chem. 2007,
128, 844–850.
[14] H. Arasaki, M. Iwata, D. Nishimura, A. Itoh, Y. Masaki, Synlett 2004, 546–
548.
[15] See, for instance: a) W. Borze˛cka, I. Lavandera, V. Gotor, J. Org. Chem.
2013, 78, 7312–7317; b) K. Nakamura, R. Yamanaka, Tetrahedron: Asym-
metry 2002, 13, 2529–2533.
[16] A. Goswami, R. L. Bezbaruah, J. Goswami, N. Borthakur, D. Dey, A. K. Ha-
zarika, Tetrahedron: Asymmetry 2000, 11, 3701–3709.
[17] K. Lundell, T. Raijola, L. T. Kanerva, Enzyme Microb. Technol. 1998, 22,
86–93.
[18] T. Matsuda, T. Harada, N. Nakajima, T. Itoh, K. Nakamura, J. Org. Chem.
2000, 65, 157–163.
[19] B. Barkakaty, Y. Takaguchi, S. Tsuboi, Tetrahedron 2007, 63, 970–976.
[20] Recent bibliography: a) F. Hollmann, K. Bꢆhler, B. Bꢆhler, in Enzyme Ca-
talysis in Organic Synthesis (Eds.: K. Drauz, H. Grçger, O. May), Wiley-
VCH, Weinheim, 2012, pp. 1325–1438; b) F. Hollmann, I. W. C. E. Arends,
D. Holtmann, Green Chem. 2011, 13, 2285–2313; c) E. Garcꢁa-Urdiales, I.
Alfonso, V. Gotor, Chem. Rev. 2011, 111, PR110–PR180; d) M. M. Musa,
R. S. Phillips, Catal. Sci. Technol. 2011, 1, 1311–1323.
[33] K. Edegger, C. C. Gruber, T. M. Poessl, S. R. Wallner, I. Lavandera, K. Faber,
F. Niehaus, J. Eck, R. Oehrlein, A. Hafner, W. Kroutil, Chem. Commun.
2006, 2402–2404.
[34] Adapted from: J. C. Lee, Y. H. Bae, S. K. Chang, Bull. Korean Chem. Soc.
2003, 24, 407–408.
[21] Selected recent examples: a) G.-C. Xu, H.-L. Yu, X.-Y. Zhang, J.-H. Xu, ACS
Catal. 2012, 2, 2566–2571; b) J. Mangas-Sꢀnchez, E. Busto, V. Gotor-
Fernꢀndez, F. Malpartida, V. Gotor, J. Org. Chem. 2011, 76, 2115–2121;
c) F. R. Bisogno, A. Cuetos, A. A. Orden, M. Kurina-Sanz, I. Lavandera, V.
Received: September 30, 2013
Published online on February 12, 2014
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 1066 – 1072 1072