Application of a promiscuous Arthrobacter sp. from Antarctic in aerobic (R)-selective deracemization and anaerobic (S)-selective reduction

Article abstract of DOI:10.1016/j.molcatb.2014.09.016

Inspired by enzyme-catalyzed reactions with microorganisms found in harsh marine environments, in which the amount of oxygen is restrict, we have shown that Arthrobacter sp. can perform different chemical transformations by switching from anaerobic to aerobic reaction conditions. Depending on the presence or absence of oxygen, either alcohol deracemization or ketone reduction with enantiocomplementary selectivities can be performed by the same microorganism. For example, reactions performed in the presence of oxygen favored the deracemization process, in which a racemic mixture of 1-(4-methylphenyl)ethanol was enriched to the (R)-alcohol in high conversion (94%) and high enantiomeric excess (94%). On the other hand, reaction in the absence of oxygen favored the reduction process, in which 4-methyl-acetophenone was converted to the (S)-alcohol in good conversion (58%) and excellent enantiomeric excess (>99%). These concepts were applied for both deracemization and enantioselective reduction of heteroatom-containing (silicon, phosphorus, tin and boron) molecules. Moreover, preparative scale reactions were also performed for both chemical processes.

Full text of DOI:10.1016/j.molcatb.2014.09.016

Journal of Molecular Catalysis B: Enzymatic 110 (2014) 117–125
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Application of a promiscuous Arthrobacter sp. from Antarctic in
aerobic (R)-selective deracemization and anaerobic (S)-selective
reduction
Dayvson J. Palmeira1, Lidiane S. Araújo1, Juliana C. Abreu, Leandro H. Andrade∗
Institute of Chemistry, University of São Paulo, Av. Prof. Lineu Prestes, 748, 05508-900 São Paulo, SP, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history:
Inspired by enzyme-catalyzed reactions with microorganisms found in harsh marine environments,
in which the amount of oxygen is restrict, we have shown that Arthrobacter sp. can perform differ-
ent chemical transformations by switching from anaerobic to aerobic reaction conditions. Depending
on the presence or absence of oxygen, either alcohol deracemization or ketone reduction with enan-
tiocomplementary selectivities can be performed by the same microorganism. For example, reactions
performed in the presence of oxygen favored the deracemization process, in which a racemic mix-
ture of 1-(4-methylphenyl)ethanol was enriched to the (R)-alcohol in high conversion (94%) and high
enantiomeric excess (94%). On the other hand, reaction in the absence of oxygen favored the reduction
process, in which 4-methyl-acetophenone was converted to the (S)-alcohol in good conversion (58%) and
excellent enantiomeric excess (>99%). These concepts were applied for both deracemization and enantio-
selective reduction of heteroatom-containing (silicon, phosphorus, tin and boron) molecules. Moreover,
preparative scale reactions were also performed for both chemical processes.
Received 11 July 2014
Received in revised form
25 September 2014
Accepted 25 September 2014
Available online 5 October 2014
Keywords:
Deracemization
Enantioselective reduction
Whole cell system
Heteroatom-containing alcohols and
ketones
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
accommodation of the substrate at low temperature, as well as by
facilitating water molecules movement and products release [4].
In the search for novel enzymes to perform selective reactions,
the use of whole cell systems are certainly one of the most attrac-
tive options mainly for oxidoreduction reactions, since no external
cofactor addition/recycling is required [1]. Fungal or bacterial
strains are the most studied so far with this purpose, mainly
isolated from mesophilic environments. However, psychrophilic
and hyper/thermophilic microorganisms present interesting
features, such as specific amino acid composition of their pro-
duced enzymes which implies in different structural properties
(flexibility or rigidity) and substrate specificity. These features,
designed by natural evolution, are directly related to enzyme
activity under extreme environments [2]. For example, enzymes
produced by cold-adapted microorganisms have a significant
low activation energy values and a significant enzyme flexibility
improvement [3], accompanied by increased thermolability, when
compared to its mesophilic homologues. In cold-adapted enzymes,
flexibility appears to play a crucial role in catalysis by helping the
Several cold-adapted enzymes from Antarctic bacterial strains
have been studied [5–7]. Motivated by the interesting features of
cold-adapted microorganisms, we have decided to join in expe-
ditions to Antarctic Peninsula, in which over several hundred
microorganisms were isolated. In this study, it was possible to iso-
late a new Arthrobacter sp. which has demonstrated an efficient
enzymatic source for enantioselective oxidation of several (RS)-1-
(phenyl)ethanols [8].
In a comparison between enzyme-catalyzed systems using
wild-type whole cells and engineering enzymes, the last one have
some advantages since either by directed evolution or rational
design, both substrate chemoselectivity and enantioselectivity
can be modified [9]. Usually, in whole cell systems, once it has
chosen a specific transformation, the screening for enzymes can
be performed. However, if a bifunctional compound is chosen as
substrate for whole cells mediated reactions, additional transfor-
mations catalyzed by different enzymes can be observed. The use of
the same microorganism to perform different reactions in organic
synthesis is more explored with recombinant microorganisms
than wild-type cells [10]. In several cases, the enzyme amount for
the desired reaction in wild-type cells is lower than a recombinant
cell. However, the search for new enzymes from wild-type strains
∗
Corresponding author. Tel.: +55 1130912287.
E-mail address: leandroh@iq.usp.br (L.H. Andrade).
1
These two authors contributed equally to this work.
http://dx.doi.org/10.1016/j.molcatb.2014.09.016
1381-1177/© 2014 Elsevier B.V. All rights reserved.

118
D.J. Palmeira et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 117–125
is also motivated by the well-known great chemical diversity in
metabolites biosynthesis, which certainly is related to interesting
and unknown enzymes [11]. The use of wild-type cells in bio-
transformation of non-natural compounds is an important way to
detect the enzyme responsible for the desired chemical reaction.
Inspired by harsh marine or lake environments in which, for
example, the amount of oxygen can be restrict, and then cellular
chemical process can exhibit interesting enzymes, we decided to
explore a novel, aerobic, Arthrobacter sp. isolated from Antarctic
lake sediments, aiming a whole cell transformation promiscuity
concept under different gaseous compositions (in the presence or
absence of oxygen). In addition, we also decided to select non-
natural compounds as substrates for these enzymes in which
different heteroatoms would be present in their chemical struc-
tures. Besides enzyme broad spectrum, we also would be able to
evaluate the heteroatom stability under the microbial conditions.
In the last decades, heteroatom-containing organic compounds
have attracted wide attention due to its increasing applica-
tion in organic synthesis, in part due to their applications in
metal-catalyzed cross-coupling to form new C C bonds, such as
Hiyama–Denmark cross-coupling of silicon compounds [12–14],
Suzuki–Miyaura cross-coupling of boron compounds [15–17] and
Stille cross-coupling of stannanes [18,19]. Moreover, organo-
phosphorus compounds can be used as ligands for coordination
chemistry [20]. In addition, phosphorus and boron are present in
bioactive molecules [21,22], and silicon organic compounds have
been pointed out as promising isosteres for carbon analogs bioac-
tive molecules [23].
2.2. Synthetic procedures
2.2.1. Synthesis of the silicon-containing racemic alcohols 1b–g
All silicon-containing racemic alcohols 1b–g were prepared
according to the methodology described by us [24].
2.2.2. General procedure for the preparation of silicon-,
phosphorus- and tin-containing ketones (2b, 3a and 3b,
respectively)
To
a 100 mL two-necked round bottom flask the 2-(4-
bromophenyl)-2-methyl-1,3-dioxolane (5 mmol, 1.21 g) was sol-
ubilized in dry THF (17 mL). Then n-butyl lithium (6 mmol,
1.20 mol/L, 5 mL) was added dropwise at −70 ◦C and under N2
atmosphere. The mixture was stirred for 2 h at −70 ◦C and then
the appropriate halide was added dropwise [Me3SiCl (6 mmol,
0.65 g, 0.76 mL), P(O)(OEt)2Cl (6 mmol, 1.03 g, 0.87 mL) or Bu3SnCl
(6 mmol, 1.95 g, 1.63 mL)]. The system was stirred for 30 min at
−70 ◦C, then the temperature was raised to room temperature and
the reaction was maintained overnight. The reaction was inter-
rupted by the addition of saturated solution of NH4Cl (20 mL), the
organic layer was separated and washed with brine (2 × 15 mL).
The aqueous layer was extracted with ethyl acetate (3 × 15 mL),
combined organic layers were dried over MgSO4 and the organic
solvents were evaporated under reduced pressure.
The crude residue was treated with a mixture of HCl (1 mol/L,
13 mL) and acetone (27 mL). The reaction was refluxed during 1 h.
Then the organic solvent was evaporated and CH2Cl2 (10 mL) was
added. The organic layer was separated and the aqueous layer was
extracted with CH2Cl2 (2 × 15 mL). Combined organic layers were
washed with saturated solution of NaHCO3 (2 × 15 mL) and brine
(2 × 15 mL), and dried over MgSO4. The solvent was evaporated
under reduced pressure and the purification was performed by col-
umn chromatography using a mixture of n-hexane/ethyl acetate
(50:50, for phosphorus-containing ketone and 95:5, for silicon- and
tin-containing ketone) as eluent.
Thus, herein we described a new promiscuous behavior of the
Arthrobacter sp. isolated from Antarctic lake sediment, in which
it was able to catalyze both (R)-selective deracemization and (S)-
selective reduction of heteroatom-containing molecules, by only
switching the gaseous composition of the reaction system (from
aerobic to anaerobic conditions).
2. Experimental
2.2.2.1. 4-Acetylphenyltrimethylsilane (2b). Yield: 605 mg (63%).
Colorless oil. 1H NMR (300 MHz, CDCl3) ı = 0.29 (s, 9H), 2.60 (s, 3H),
7.61–7.64 (d, J = 9.0 Hz, 2H), 7.90–7.93 (d, J = 9.0 Hz, 2H). 13C NMR
(75 MHz, CDCl3) ı = −1.20, 26.77, 127.36, 133.66, 137.32, 147.41,
198.57. FT-IR (liquid film) ꢀmax = 2956, 1687, 1428, 1388, 1250,
840 cm−1. LRMS (EI) m/z (%) = 192(9) [M]+, 177(100), 162(1), 149(5),
147(2) 134(5), 119(8), 105(3), 91(3), 73(8), 53(3), 43(13).
2.1. General methods
Unless otherwise noted, commercially available materials were
used without further purification. All solvents were of HPLC or
ACS grade. THF was distilled from Na/benzophenone. Nuclear mag-
netic resonance (NMR) spectra were recorded on a Varian Gemini
200 or a Varian Inova 300 spectrometers operating at frequen-
cies of 200 and 300 MHz (1H NMR) or 50 and 75 MHz (13C NMR),
respectively. The 1H NMR chemical shifts are reported in ppm
relative to the TMS peak. Data are reported as follows: chemical
shift (ı), multiplicity (s = singlet, d = doublet, dd = double doublet,
t = triplet, q = quartet, sex = sextet, m = multiplet), coupling con-
stant (J) in Hertz and integrated intensity. The 13C NMR chemical
shifts are reported in ppm relative to CDCl3 signal. Low-resolution
mass spectra were obtained on a GC/MS Shimadzu spectrom-
eter, operating at 70 eV. Gas chromatography (GC) analysis for
enantiomeric excesses were obtained using a Shimadzu 17-A
Gas Chromatograph, equipped with autosampler, Flame Ioniza-
tion Detector (FID) and a chiral column Varian CP-Chirasil-DEX
CB ␤-cyclodextrin (25 m × 0.25 mm × 0.25 ␮m). The temperature
of the detector and injector was 220 ◦C; pressure = 167 kPa, H2. The
Arthrobacter sp. CCT 7749 (COLEC¸ ÃO DE CULTURAS TROPICAL–CCT;
http://fat.org.br/) was isolated from Antarctic lake marine sedi-
ment [Antarctic Peninsula, Admiralty Bay, King George Island (21 E
423394, UTM 3114505)].
2.2.2.2. Diethyl (4-acetylphenyl)phosphonate (3a). Yield: 538 mg
(42%). Colorless oil. 1H NMR (300 MHz, CDCl3) ı = 1.34 (t, J = 9.0 Hz,
6H), 2.64 (s, 3H), 4.07–4.22 (m, 4H), 7.89–7.96 (m, 2H), 8.01–8.05
(m, 2H). 13C NMR (75 MHz, CDCl3) ı = 16.44 (d, J = 6 Hz), 26.93,
62.56 (d, J = 6 Hz), 128.17 (d, J = 15 Hz), 132.19 (d, J = 9.7), 134.71,
139.95 (d, J = 3 Hz), 197.64. FT-IR (KBr) ꢀmax = 2988, 1695, 1395,
1248, 1018, 961 cm−1. LRMS (EI) m/z (%) = 256(21) [M]+, 241(59),
228(6), 213(100), 185(52) 167(28), 157(8), 147(32), 139(17),
120(5), 104(14), 91(11), 77(19), 65(20), 43(46).
2.2.2.3. 4-Acetylphenyltributylstannane (3b). Yield: 944 mg (46%).
Colorless oil. 1H NMR (300 MHz, CDCl3) ı = 0.88 (t, J = 6 Hz, 9H),
1.09 (t, J = 9 Hz, 6H), 1.27–1.39 (sex, J = 6.0, 6.0 Hz, 6H), 1.48–1.65
(m, 6H), 2.59 (s, 3H), 7.57–7.59 (d, J = 7.5 Hz, 2H), 7.86–7.89 (d,
J = 7.5 Hz, 2H). 13C NMR (75 MHz, CDCl3) ı = 9.81, 13.80, 26.68,
27.47, 29.18, 127.24, 136.72, 136.78, 150.44, 198.81. FT-IR (liquid
film) ꢀmax = 2927, 1687, 1385, 1269, 812, 663 cm−1. LRMS (EI) m/z
(%) = 395(1), 353(82), 297(71), 241(100), 239(99) 196(6), 177(1),
161(1), 137(7), 120(11), 105(12), 91(7), 77(4), 57(2), 43(10).
Sterile materials were used to perform the experiments involv-
ing microorganisms. A laminar flow cabinet Fisher Hamilton (Class
II Biological Safety Cabinet) was used to handle microorganisms.

D.J. Palmeira et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 117–125
119
2.2.3. Synthesis of boronic ester-containing ketone and racemic
alcohol (3c and 4c)
Bacterial cells were harvested by centrifugation at 15 ◦C
(9000 rpm, 40 min) and used in the appropriate reactions.
Boron-containing ketone and racemic alcohol were prepared
according to the methodology described by us [25].
2.4.1. Reactions under aerobic conditions
The appropriate substrate (40 ␮L from a 0.25 mol/L solution
in DMF) was added under sterile conditions into polypropylene
microtiter plates containing 164 mg wet cells resuspended in 2 mL
phosphate buffer solution (pH 7.0, 100 mmol/L). The reaction was
maintained at 20 ◦C and 400 rpm, under air or O2 (for reaction times,
see Tables 1–3 in Section 3). The reaction mixture was extracted by
stirring with tert-butyl methyl ether (2 mL) followed by centrifu-
gation (6000 rpm, 1 min). The organic phase was analyzed by GC
using a chiral capillary column (see Section 2.6).
2.2.4. General procedure for the chemical reduction of
heteroatom-containing ketones 3a and 3b to their corresponding
racemic alcohols (4a and 4b)
To a 100 mL round bottom flask were added the appropriated
heteroatom-containing ketone (2 mmol) and methanol (4 mL). The
mixture was cooled to 0 ◦C. Then the NaBH4 (2.2 mmol, 84 mg)
was added in small portions. The system was stirred for 30 min
at 0 ◦C and then 2.5 h at room temperature. The solvent was evap-
orated under reduced pressure and H2O (4 mL) was added. The pH
was adjusted to 6.0 by the addition of HCl (1 mol/L) and the aque-
ous solution was extracted with CH2Cl2 (3 × 4 mL). The combined
organic layers were dried over MgSO4 and the solvent evaporated
under reduced pressure. Crude product was purified by column
chromatography using a mixture of n-hexane/ethyl acetate (1:1, for
phosphorus-containing alcohol and 8:2, for tin-containing alcohol)
as eluent.
2.4.2. Reactions under anaerobic conditions
A sealed sterile glass tube containing 328 mg of wet cells resus-
pended in 2 mL phosphate buffer solution (pH 7.0, 100 mmol/L)
at room temperature was purged both aqueous solution and gas
atmosphere of the reaction vessel with N2 for 1 h prior to the
substrate addition. Then, the appropriate substrate (40 ␮L from a
0.25 mol/L solution in DMF) was added and the reaction was per-
formed at room temperature, 700 rpm and under N2 (for reaction
times, see Tables 1, 2 and 5 in Section 3). The reaction mix-
ture was extracted by stirring with tert-butyl methyl ether (2 mL)
followed by centrifugation (6000 rpm, 1 min). The organic phase
was analyzed by GC using a chiral capillary column (see Section
2.6).
2.2.4.1. Diethyl
4-(1-hydroxyethyl)-phenyl-phosphonate
(4a).
Yield: 475 mg (92%). Colorless oil. 1H NMR (200 MHz, CDCl3)
ı = 1.32 (t, J = 7.0 Hz, 6H), 1.50 (d, J = 6.6 Hz, 3H), 4.10 (q, J = 7.0 Hz,
4H), 4.94 (q, J = 6.6 Hz, 1H), 7.44 (dd, J = 8.0, 4.0 Hz, 2H), 7.72 (dd,
J = 13.1, 8.3 Hz, 2H). 13C NMR (50 MHz, CDCl3) ı = 16.23, 16.36,
25.29, 62.04, 62.15, 125.29, 125.59, 128.87, 131.87, 132.08, 150.61.
FT-IR (KBr) ꢀmax = 3391, 2981, 1605, 1395, 1028, 970 cm−1. LRMS
(EI) m/z (%) = 258(4) [M]+, 243(67), 229(2), 215(100), 187(63)
169(12), 159(62), 149(8), 141(30), 121(12), 109(43), 91(19),
77(58), 65(13), 43(42).
Table 1
Biotransformation of (RS)-1a mediated by whole cells of Arthrobacter sp. under
different reaction conditions.a
2.2.4.2. 1-[4-(Tributylstannyl)phenyl]ethanol (4b). Yield: 773 mg
(94%). Colorless oil. 1H NMR (200 MHz, CDCl3) ı = 0.88 (t, J = 7.2 Hz,
9H), 1.00–1.61 (m, 21H), 4.86 (q, J = 6.4 Hz, 1H), 7.31 (d, J = 8.0 Hz,
2H), 7.43 (d, J = 8.0 Hz, 2H). 13C NMR (50 MHz, CDCl3) ı = 9.56, 13.66,
24.97, 27.37, 29.08, 70.48, 124.99, 136.64, 141.12, 145.48. FT-IR
(liquid film) ꢀmax = 3345, 2925, 1462, 1067, 816, 663 cm−1. LRMS
(EI) m/z (%) = 397(1), 355(94), 299(71), 243(100), 241(90), 223(7),
195(3), 177(2), 161(1), 137(10), 121(23), 105(41), 91(7), 79(11),
57(5), 41(12).
Entry Time (h) Gaseous system Ketone/alcohol conv. (%)b e.e.alcohol (%)c
1
2
3
4
5
6
3
12
48
120
3
Air
Air
Air
Air
N2
44/56
20/80
14/86
6/94
4/96
46/54
94 (R)
94 (R)
94 (R)
94 (R)
2 (R)
3
O2
96 (R)
a
General conditions: substrate (0.01 mmol, 40 ␮L of 0.25 mol/L DMF solution),
Arthrobacter sp. cells (164 mg), phosphate buffer pH 7.5 (2 mL), 20 ◦C, 400 rpm.
2.3. Culture media
b
Ratio was determined by gas chromatography (GC).
Enantiomeric excess was detemined by chiral GC.
c
The composition of the culture media, MCBE7 and NB, were
(amounts of the ingredients are indicated per liter of distilled
water):
Table 2
Enantioselective reduction of 2a by cells of Arthrobacter sp. under different reaction
conditions.a
MCBE7 (pH = 7.0) = (RS)-1-(phenyl)ethanol (1 mmol, 0.122 g,
122 ␮L), K2HPO4 (0.2% (w/w), 2.0 g), KH2PO4 (0.1%, 1.0 g),
MgSO4·7H2O (0.04%, 0.4 g), NH4Cl (0.04%, 0.4 g), Meat Extract
(Oxoid) (0.025%, 0.25 g), Peptone (Oxoid) (0.025%, 0.25 g), agar
(Oxoid) (2.0%, 20.0 g, when needed for solid culture medium).
Nutrient broth (NB) = 3.0 g Meat Extract (Oxoid) and 5.0 g
Bacteriological Peptone (Oxoid) supplemented with (RS)-1-(4-
methylphenyl)ethanol (0.1 mmol/L).
Entry Time (h) Gaseous system Ketone/alcohol conv. (%)b e.e.alcohol (%)c
1
2
3
4
5
6
3
12
48
120
3
Air
Air
Air
Air
O2
97/3
98/2
93/7
90/10
97/3
42/58
>99 (S)
63 (S)
70 (S)
22 (S)
>99 (S)
>99 (S)
2.4. Resting cells assays
Arthrobacter sp. was inoculated under sterile conditions into
100 mL of appropriate culture media (MCBE7) in Erlenmeyer flasks
(250 mL) and incubated in an orbital shaker (160 rpm, 20 ◦C, 48 h).
After that, the culture was transferred to 2 L Erlenmeyer flasks con-
taining 1 L of sterilized culture media (NB) and incubated again
(160 rpm, 20 ◦C, 48 h).
3
N2
a
General conditions: substrate (0.01 mmol, 40 ␮L of 0.25 mol/L DMF solution),
Arthrobacter sp. cells (164 mg), phosphate buffer pH 7.5 (2 mL), 20 ◦C, 400 rpm.
b
Ratio was determined by gas chromatography (GC).
Enantiomeric excess was detemined by chiral GC.
c

120
D.J. Palmeira et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 117–125
Fig. 1. Selected heteroatom-containing (a) alcohols for deracemization and (b) ketones for enantioselective reduction using Arthrobacter sp. cells.
2.5. Preparative scale reactions
extractor with dichloromethane (300 mL). The organic layer was
dried over MgSO4 and solvent evaporated under reduced pressure.
Crude products were purified by column chromatography using
n-hexane/ethyl acetate as eluent.
Bacterial cells were cultivated as described in Section 2.4.
2.5.1. Reaction under aerobic conditions (deracemization)
The appropriate racemic alcohol (0.5 mmol) diluted in dimethyl-
formamide (2 mL) was added under sterile conditions into 500 mL
Erlenmeyer flask containing 8.2 g of wet cells resuspended in
100 mL phosphate buffer solution (pH 7.0, 100 mmol/L). The reac-
tion was maintained at 20 ◦C and 400 rpm, for 120 h. The reaction
mixture was extracted for 24 h using a continuous liquid–liquid
2.5.2. Reaction under anaerobic conditions (enantioselective
reduction)
A sealed sterile 500 mL round-bottom flask containing 16.4 g of
wet cells resuspended in 100 mL phosphate buffer solution (pH
7.0, 100 mmol/L) at room temperature was purged both aqueous
solution and gas atmosphere of the reaction vessel with N2 for
Fig. 2. Chemical synthesis of racemic alcohols 1b–g, 4a–c and prochiral ketones 2b, 3a–c.

D.J. Palmeira et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 117–125
121
•
1 h prior to the substrate addition. Then, the appropriate ketone
(0.5 mmol) diluted in dimethylformamide (2 mL) was added under
sterile conditions and the reaction was performed at room tem-
perature, 700 rpm and under N2 for 120 h. The reaction mixture
was extracted for 24 h using a continuous liquid-liquid extraction
apparatus with dichloromethane as solvent (300 mL). The organic
layer was dried over MgSO4 and solvent evaporated under reduced
pressure. Crude products were purified by column chromatography
using n-hexane/ethyl acetate as eluent.
(RS)-diethyl 4-(1-hydroxyethyl)-phenyl-phosphonate (4a):
160 ◦C, tR = 22.24 min for (R) and 23.12 min for (S)
•
•
(RS)-1-[4-(tributylstannyl)phenyl]ethanol
(4b)
(acetylated
form): 180 ◦C, tR = 18.36 min for (-) and 19.25 min for (+)
(RS)-1-[4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-
yl)phenyl]ethanol (4c) (acetylated form): 160 ◦C, tR = 9.12 min
for (R) and 8.69 min for (S)
The enantiomeric excess was defined as the ratio of
{(|[R] − [S]|)/([R] + [S])} × 100%, where [R] and [S] are the concen-
trations of (R) and (S) enantiomers, respectively.
2.6. GC analysis
2.7. Absolute configuration
The enantiomeric excesses (e.e.) of the chiral alcohols and their
derivatives (acetylated form) were measured by chiral GC anal-
ysis. The analysis was carried out on a column ␤-cyclodextrin
(25 m × 0.32 mm × 0.25 ␮m, Varian CP-Chirasil-DEX CB). GC con-
ditions (carrier gas – H2, 167 kPa): injector 220 ◦C; detector 220 ◦C.
Isotherm and retention time (tR) for the compounds:
Absolute configurations of silicon- and boron-containing (R)- or
(S)-alcohols were determined by comparison between the signs of
the measured specific rotations with those in the literature [24–26].
Absolute configuration of phosphorus-containing (R)- or (S)-
alcohols were determined by comparison with authentic (R)- and
(S)-enantiopure samples prepared from appropriate enantiopure
(1-(4-bromophenyl)ethoxy)trimethylsilane using methodology as
described in Sections 2.2.2 and 2.6.
4-methyl-acetophenone (2a): 115 ◦C, tR = 1.65 min
4-acetylphenyltrimethylsilane (2b): 128 ◦C, tR = 4.50 min
diethyl (4-acetylphenyl)phosphonate (3a): 160 ◦C, tR = 10.34 min
4-acetylphenyltributylstannane (3b): 180 ◦C, tR = 16.51 min
1-(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-
yl)phenyl)ethanone (3c): 160 ◦C, tR = 5.92 min
•
•
•
•
•
3. Results and discussion
3.1. Biotransformations mediated by Arthrobacter sp. under
aerobic and anaerobic conditions
(RS)-1-(4-methylphenyl)ethanol (1a): 115 ◦C, tR = 2.44 min for (R)
and 2.69 min for (S)
•
•
(RS)-1-(4-trimethylsilyl-phenyl)ethanol
tR = 7.01 min for (R) and 7.47 min for (S)
(1b):
128 ◦C,
Initially, aiming kinetic resolution or deracemization pro-
cess, oxidation of (RS)-1a was investigated by performing this
Table 3
Deracemization of heteroatom-containing alcohols by cells of Arthrobacter sp. under aerobic conditions.a
Entry
Compound
Y group
R2 group
Reaction time (h)
Ketone/alcohol conv. (%)b
e.e.alcohol (%)c
1
2
3
4
5
6
1b
1c
1d
1e
1f
p-SiMe3
m-SiMe3
o-SiMe3
p-SiMe3
m-SiMe3
o-SiMe3
Me
Me
Me
Et
Et
Et
48
48
48
48
48
48
8/92
8/92
0/100
4/96
3/97
0/100
>99 (R)
96 (R)
0
41 (R)
26 (R)
0
1g
7
8
9
10
11
12
1b
1c
1d
1e
1f
p-SiMe3
m-SiMe3
o-SiMe3
p-SiMe3
m-SiMe3
o-SiMe3
Me
Me
Me
Et
Et
Et
120
120
120
120
120
120
3/97
6/94
0/100
4/96
2/98
0/100
98 (R)
96 (R)
0
38 (R)
14 (R)
0
1g
13
14
15
4a
4b
4c
p-P(O)(OEt)2
p-SnBu3
p-Bpin
Me
Me
Me
48
48
48
36/64
13/87
56/44
34 (R)
22 (+)d
78 (R)
16
17
18
4a
4b
4c
p-P(O)(OEt)2
p-SnBu3
p-Bpin
Me
Me
Me
120
120
120
32/68
6/94
44/56
28 (R)
6 (+)d
83 (R)
a
General conditions: substrate (0.01 mmol, 40 ␮L of 0.25 mmol/L DMF solution), Arthrobacter sp. whole cells (164 mg), phosphate buffer pH 7.5 (2 mL), 20 ◦C, 400 rpm.
Ratio was determined by chiral GC (Chiral-Dex-CB).
b
c
Enantiomeric excess was detemined by chiral GC.
Signal of optical rotation ([˛]2D0) of favored enantiomer.
d

122
D.J. Palmeira et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 117–125
transformation at different reaction time under anaerobic and aer-
obic conditions (Table 1).
system), the ketone 2a was produced in 44% and the remaining
alcohol (R)-1a with 96% e.e., showing a usual kinetic resolution
process (entry 6).
As it can be seen in Table 1, by increasing reaction time
under aerobic conditions (air), the alcohol concentration increase.
12 and 48 h reactions showed similar results (entries 2 and
3), with the formation of (R)-1a in 80% and 86% of conver-
sion, respectively, and 94% of enantiomeric excess for both cases.
After 120 h (entry 4), an almost perfect deracemization pro-
cess was achieved in which, the (R)-alcohol 1a concentration
increased to 94% but its enantiomeric excess remained the same
(94%).
Surprisingly, when the same reaction was performed under
anaerobic condition (N2 saturated system), almost no oxidation
was evidenced, since very low ketone concentration (4%) and the
enantiomeric excess for the alcohol 1a was insignificant (entry 5).
It is clear that the presence of an oxygenated atmosphere exerts
a significant and positive influence on the enantioselective oxida-
tion. After 3 h reaction under the aerobic condition (O2 saturated
On the other hand, there are no significant differences between
the two aerobic conditions (O2 saturated or under air systems)
(entries 1 and 6). Both reactions showed similar ketone 2a concen-
tration (46 and 44%, respectively), and very similar enantiomeric
excesses for the remaining alcohol (R)-1a (96 and 94%, respec-
tively). These results indicated that only the natural presence of
O2 in buffer solution was enough to ensure the alcohol oxidation
in the desirable kinetic resolution process. At this point, we can
claim that the oxidoreductase produced by Arthrobacter sp. which
is responsible for the enantioselective oxidation of (RS)-1a, is O2
dependent.
Deracemization of secondary alcohols is a topic of growing inter-
est in biocatalysis, since by mild reaction conditions and cheap
reagents, only one enantiomer can be obtained from racemic mix-
tures [27,28]. This process can be achieved in whole cells systems by
Table 4
Deracemization of silicon-, phosphorus- tin- and boron-containing alcohols in a preparative scale under aerobic conditions.a
Entry
Substrate
Isolated yield % (e.e.%)b
1
(RS)-1b
2
(RS)-4a
3
(RS)-4b
4
(RS)-4c
a
General conditions: substrate (0.5 mmol solubilized in 2 mL of DMF), Arthrobacter sp. cells (8.2 g), phosphate buffer pH 7.5 (100 mL), 20 ◦C, 400 rpm, 120 h.
Enantiomeric excess was detemined by chiral GC.
b
c
Enantiomeric excess was detemined by chiral GC of corresponding acetate derivatives.

D.J. Palmeira et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 117–125
123
tandem reactions, in which enantiocomplementary enzymes pro-
duced either from a single [29] or two different microorganisms
[30] are involved.
Simultaneous reactions involving different and unknown
enzymes in whole cells systems might be inconvenient to fully elu-
cidate mechanism pathway. However, it is possible to speculate
that in deracemization process of (RS)-1a, while an oxidoreduc-
tase is responsible to selectively oxidize alcohol (S)-1a to ketone
2a, another oxidoreductase is then responsible to reduce the pro-
duced ketone 2a to the (R)-1a, increasing both concentration
and enantiomeric excess of alcohol (R)-1a. These oxidoreductases
should be oxygen-dependent since, under anaerobic conditions the
oxidation of (RS)-1a afforded only traces of ketone 2a (Table 1,
entry 5) while reduction of 2a gave alcohol (S)-1a exclusively
(Table 2).
In order to explore this promiscuous feature of Arthrobacter sp.,
we decided to evaluate additional substrates, silicon-, phosphorus-
, tin- and boron-containing alcohols for deracemization process in
the presence of oxygen (Fig. 1a), and the enantioselective reduction
of silicon-, phosphorus-, tin- and boron-containing ketones in the
absence of oxygen (Fig. 1b).
Results described in Table 1 indicates that deracemization pro-
cess take place using a single microorganism (Arthrobacter sp.) in
longer reaction times, but the inability to reach >99% e.e. in all
cases suggest that other oxidoreductases can be involved in this
process. Furthermore, we were intrigued if deracemization pro-
cess was carried out by one or more enzymes, which could lead
us to find other transformations mediated by this microorgan-
ism.
In order to understand this chemical process mediated by
Arthrobacter sp., we performed the reduction of 4-methyl-
acetophenone 2a under anaerobic and aerobic conditions (Table 2).
As a good surprise, it was observed a highly enantioselective
enzyme present in the Arthrobacter sp. which is active only in the
absence of O2 (Table 2, entry 6).
In a similar way as observed in oxidation reactions (Table 1),
bioreduction under N2 and O2 conditions led to different results
(Table 2, entries 5 and 6). After 3 h reaction, enantioselective
reduction of 2a under N2 saturated system produced 58% of cor-
responding alcohol (S)-1a in >99% e.e. On the other hand, O2
saturated reactions despite high enantiomeric excess (>99%) of (S)-
1a product, showed a low conversion. However, none significant
differences were evidenced between O2 saturated and under air
reactions.
In a remarkable feature, while biotransformation of alcohol (RS)-
1a favors formation of (R)-1a (Table 1), reduction of ketone 2a by
the same enzymatic system favors formation of the opposite enan-
tiomer, alcohol (S)-1a (Table 2). Although low conversions achieved
in bioreductions under aerobic conditions (≤10% in all cases), for-
mation of the opposite enantiomer could explain the enantiomeric
excess of 94% for (R)-alcohol in the biotransformation of the racemic
alcohol under the same reaction conditions.
3.2. Deracemization of heteroatom-containing racemic alcohols
by Arthrobacter sp.
The substrates were synthesized as described in Fig. 2 [24].
Different structural modifications for the benzylic alcohols were
considered and silicon-, phosphorus-, tin- and boron-containing
derivatives were synthesized, as well as their corresponding
ketones.
Having all heteroatom-containing alcohols in hand, cells of
Arthrobacter sp. under aerobic conditions (air) were used to achieve
the deracemization process (Table 3).
According to the results showed in Table 3, reactions with m-
and p-silicon substituted alcohols showed similar ketone/alcohol
ratio for all cases, but the enantiomeric excesses for p-silicon sub-
stituted (R)-alcohols (entries 1, 4, 7 and 10) were always higher
than the m-substituted analogs (entries 2, 5, 8 and 11). None of
the o-silicon substituted alcohols (entries 3, 6, 9 and 12) were oxi-
dized by the oxidoreductases present in Arthrobacter sp. under the
reaction conditions evaluated.
Presence or absence of O2 plays an important role on which
type of reaction the microorganism can perform. The oxidoreduc-
tase responsible for enantioselective oxidation is molecular oxygen
dependent, while the oxidoreductase responsible for enantioselec-
tive reduction is not. It is worth mentioning that by switching the
gaseous composition of reaction system, enantiocomplementary
transformations can be achieved.
The increase in size of the side chain attached to the chiral car-
bon also played an important role for the enantiomeric excesses of
the remaining silicon-substituted (R)-alcohols. For example, at the
Table 5
Enantioselective reduction of heteroatom-containing ketones 2b and 3a–c mediated by Arthrobacter sp. cells under anaerobic conditions.a
Entry
Compound
Y group
Time (h)
Ketone/alcohol conv. (%)b
e.e.alcohol (%)c
1
2
3
4
2b
3a
3b
3c
SiMe3
P(O)(OEt)2
SnBu3
3
3
3
3
19/81
11/89
100/0
77/23
>99 (S)
>99 (S)
–
Bpin
>99 (S)
5
6
7
8
2a
3a
3b
3c
SiMe3
P(O)(OEt)2
SnBu3
24
24
24
24
4/96
2/98
100/0
47/53
>99 (S)
>99 (S)
–
Bpin
>99 (S)
a
General conditions: substrate (0.01 mmol, 40 ␮L of 0.25 mmol/L DMF solution), Arthrobacter sp. cells (328 mg), phosphate buffer pH 7.5 (2 mL), 20 ◦C, 400 rpm.
Proportion determined by chiral GC (Chiral-Dex-CB).
Enantiomeric excess detemined by chiral GC. Absolute configurations are presented in parenthesis.
b
c

124
D.J. Palmeira et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 117–125
Table 6
same reaction time, the racemic m-silicon substituted compound
bearing a methyl group attached to the chiral carbon (1c) was der-
acemized to produce the (R)-alcohol in 96% e.e. (entry 2), while
the analog bearing an ethyl group at the chiral carbon (1f) was
deracemized to the (R)-alcohol in only 26% e.e. (entry 5).
When considering compounds bearing an ethyl group attached
to the chiral carbon, by changing the reaction time from 48 h
(entries 4 and 5) to 120 h (entries 10 and 11) a decrease of enan-
tiomeric excess was observed. This result can be related to a
competitive reaction favoring the (S)-enantiomer.
Enzymatic reduction of silicon-, phosphorus- and boron-containing ketones in a
preparative scale under anaerobic conditions.a
Entry
Substrate
Isolated yield % (e.e.%)b
1
2b
In addition to silicon compounds, phosphorus- and tin-
containing alcohols (entries 13 and 14) were also obtained in good
conversions (64 and 87%, respectively) but moderate enantiomeric
excesses (34 and 22%). Increasing the reaction time to 120 h,
similar results to those achieved with silicon compounds 1e and
1f were observed. A small decrease in the enantiomeric excess
of phosphorus compound (R)-4a (28%, entry 16) and a significant
decrease in the enantiomeric excess of tin compound (R)-4b (6%,
entry 17) were obtained.
2
3a
On the other hand, boron-containing alcohol 4c showed more
interesting results when compared to phosphorus and tin analogs.
After 48 h, a conversion of 44% and enantiomeric excess of 78%
(entry 15) for (R)-4c were achieved. Increasing the reaction time to
120 h led to an increase in both conversion (56%) and enantiomeric
excess (83%) for (R)-4c (entry 18).
3
3c
Silicon- and boron-derivatives were more suitable to undergo
the deracemization process catalyzed by enzymes from Arthrobac-
ter sp. than tin- and phosphorus derivatives. Of course, different
factors can be the reason for those results, but in a biocatalytical
point of view, we can point out that the heteroatom nature certainly
have a relevant effect in enzyme activity.
a
General conditions: substrate (0.5 mmol solubilized in 2 mL of DMF),
Arthrobacter sp. cells (16.4 g), phosphate buffer pH 7.5 (100 mL), 20 ◦C, 400 rpm,
120 h.
b
Enantiomeric excess was detemined by chiral GC.
c
Enantiomeric excess was detemined by chiral GC of corresponding acetate
derivatives.
Not detected neither in GC nor after purification.
d
Enzymatic deracemization of silicon-, phosphorus-, tin- and
boron-containing alcohols in a preparative scale were also per-
formed. All reactions were slower than those performed in
analytical scale. These results can be attribute to the oxygen trans-
fer rates which can decrease with reaction scale-up. However, we
were able to achieve good yields and high enantiomeric excesses
as shown in Table 4.
After the screening protocol, enzymatic reduction of silicon-,
phosphorus- and boron-containing ketones in a preparative scale
were also performed (Table 6).
As one can see in Table 6, silicon-, tin- and boron-containing (S)-
alcohols were obtained in good to excellent yields (up to 97%) and
excellent enantiomeric excesses (>99% in all cases).
4. Conclusions
3.3. Enantioselective reduction of heteroatom-containing ketones
by Arthrobacter sp.
An Arthrobacter sp. from Antarctic was successfully used
as promiscuous biocatalyst for either deracemization or
enantioselective reduction of heteroatom-containing compounds.
A new feature of Arthrobacter sp. from Antarctic was discovered.
Mimicking harsh marine environments, in which oxygen is restrict,
either (R)-selective deracemization of alcohols or (S)-selective
reduction of ketones were able to be performed, just by switching
gaseous composition of reaction system from aerobic to anaerobic
condition, respectively.
Deracemization process was studied using a series of silicon-
containing racemic alcohols. The best results were achieved for
silicon-containing alcohol (1b), in which (R)-alcohol was obtained
in 92% of conversion and >99% e.e. Deracemization process was also
applied for phosphorus-, tin- and boron-containing racemic alco-
hols and the corresponding (R)-alcohols were achieved in excellent
yields (up to 94%) and enantiomeric excesses (up to 83%).
As seen in promising results with Arthrobacter sp. cells under
anaerobic conditions, Section 3.1, enantioselective reduction of
silicon-, phosphorus-, tin- and boron-containing ketones (com-
pounds 2b and 3a–c) were explored (Table 5).
In all cases, reduction by Arthrobacter sp. under anaerobic condi-
tions proved to be highly enantioselective to the hetero-containing
(S)-alcohols except for 4-(tributyltin)acetophenone (3b). After 3 h
reactions, silicon- and phosphorus-containing ketones were effi-
ciently converted into its corresponding (S)-alcohols (entries 1 and
2), reaching a conversion of 81% for (S)-1b and 89% for (S)-4a, and
>99% e.e. in both cases. Increasing the reaction time to 24 h, the
(S)-alcohols concentrations were also increased (entries 5 and 6),
in which (S)-1b was obtained in 96% and (S)-4a in 98%, both in
excellent enantiomeric excesses.
Tin-containing ketone 3b did not undergo enzymatic reduction
by Arthrobacter sp. neither in 3 nor 24 h (entries 3 and 7). This can
be related to the large size of this molecule, which may not fit in an
appropriate manner at the enzyme active site.
Enantioselective reduction of silicon-, phosphorus- and boron-
containing ketones to the corresponding (S)-alcohols was per-
formed in good to excellent conversions (up to 98%) and excellent
enantiomeric excesses (>99% in all cases) were observed.
On the other hand, after 3 h (entry 4) boron-containing ketone
3c was moderately converted to its corresponding alcohol (S)-4c in
23% of conversion with >99% of enantiomeric excess. Increasing the
reaction time to 24 h (entry 8), a higher conversion to (S)-4c was
achieved (53%) maintaining the enantiomeric excess, >99%.
Preparative scale reactions were performed for deracemization
process showing to be an efficient protocol to achieve silicon-,
phosphorus-, tin- and boron-containing (R)-alcohols, with good
yields and enantiomeric excesses (up to 97%). Enantioselective
reduction of silicon-, phosphorus and boron-containing ketones

D.J. Palmeira et al. / Journal of Molecular Catalysis B: Enzymatic 110 (2014) 117–125
125
was even better than deracemization process, in which good to
excellent yields (81–97%) and excellent enantiomeric excesses
(>99% in all cases) for (S)-alcohols were achieved.
[9] N.U. Nair, C.A. Denard, H. Zhao, Curr. Org. Chem. 14 (2010) 1870–1882.
[10] T. Ishige, K. Honda, S. Shimizu, Curr. Opin. Chem. Biol. 9 (2005) 174–180.
[11] S. Bertrand, N. Bohni, S. Schnee, O. Schumpp, K. Gindro, J.-L. Wolfender, Bio-
technol. Adv. (2014), http://dx.doi.org/10.1016/j.biotechadv.2014.03.001.
[12] Y. Hatanaka, T. Hiyama, J. Org. Chem. 53 (1988) 918–920.
[13] S.E. Denmark, R.F. Sweis, Acc. Chem. Res. 35 (2002) 835–846.
[14] A. Gordillo, M.A. Ortun˜o, C. López-Mardomingo, A. Lledós, G. Ujaque, E. de Jesús,
J. Am. Chem. Soc. 135 (2013) 13749–13763.
Acknowledgements
Authors thank FAPESP (São Paulo Research Foundation; Grant
Numbers: 2012/14196-1; 2012/11482-3; 2011/01376-9) and CNPq
(National Council for Scientific and Technological Development;
Grant Number: 552834/2010-6) for their fellowship and financial
support.
[15] A. Suzuki, Pure Appl. Chem. 57 (1985) 1749–1758.
[16] N. Miyaura, A. Suzuki, Chem. Rev. 95 (1995) 2457–2483.
[17] K.M. Bjerglund, T. Skrydstrup, G.A. Molander, Org. Lett. 16 (2014) 1888–1891.
[18] W.J. Scott, M.R. Pen˜a, K. Sward, K. Stoessel, J.K. Stille, J. Org. Chem. 50 (1985)
2302–2308.
[19] D. Qiu, S. Wang, S. Tang, H. Meng, L. Jin, F. Mo, Y. Zhang, J. Wang, J. Org. Chem.
79 (2014) 1979–1988.
[20] P.W.N.M. van Leeuwen, P.C.J. Kamer, C. Claver, O. Pàmies, M. Diéguez, Chem.
Rev. 111 (2011) 2077–2118.
References
[21] P. Kafarski, B. Lejczak, Phosphorus Sulfur Silicon Relat. Elem. 63 (1991)
193–215.
[22] B.D. Dorsey, J. Med. Chem. 51 (2008) 1068–1072.
[1] K. Faber, Biotransformations in Organic Chemistry – A Textbook, second ed.,
Springer-Verlag, 1995.
[2] K.S. Siddiqui, T.J. Williams, D. Wilkins, S. Yau, M.A. Allen, M.V. Brown, F.M. Lauro,
R. Cavicchioli, Annu. Rev. Earth Planet. Sci. 41 (2013) 87–115.
[3] S. D’Amico, J.C. Marx, C. Gerday, G. Feller, J. Biol. Chem. 278 (2003) 7891–7896.
[4] J.C. Marx, T. Collins, S. D’Amico, G. Feller, C. Gerday, Mar. Biotechnol. 9 (2007)
293–304.
[5] N.J. Russell, Extremophiles 4 (2000) 83–90.
[6] R. Cavicchioli, K.S. Siddiqui, D. Andrews, K.R. Sowers, Curr. Opin. Biotechnol. 13
(2002) 253–261.
[23] A.K. Franz, S.O. Wilson, J. Med. Chem. 56 (2013) 388–405.
[24] D.J. Palmeira, J.C. Abreu, L.H. Andrade, Molecules 16 (2011) 9697–9713.
[25] L.H. Andrade, T. Barcellos, Org. Lett. 11 (2009) 3052–3055.
[26] T. Barcellos, K. Tauber, W. Kroutil, L.H. Andrade, Tetrahedron: Asymmetry 22
(2011) 1772–1777.
[27] A.J. Carnell, Adv. Biochem. Eng. Biotechnol. 63 (1999) 57–72.
[28] J.H. Schrittwieser, J. Sattler, V. Resch, F.G. Mutti, W. Kroutil, Curr. Opin. Chem.
Biol. 15 (2011) 249–256.
[7] T. Kazuoka, T. Oikawa, I.M.S. Kuroda, K. Soda, Extremophiles 11 (2007) 257–267.
[8] L.S. Araújo, E. Kagohara, T.P. Garcia, V.H. Pellizari, L.H. Andrade, Mar. Drugs 9
(2011) 889–905.
[29] J.V. Comasseto, L.F. Assis, L.H. Andrade, I.H. Schoenlein-Crusius, A.L.M. Porto, J.
Mol. Catal. B: Enzym. 39 (2006) 29–30.
[30] Y.-L. Li, J.-H. Xu, Y. Xu, J. Mol. Catal. B: Enzym. 64 (2010) 48–52.