Efficient methodology to produce a duloxetine precursor using whole cells of Rhodotorula rubra

Article abstract of DOI:10.1016/j.tetasy.2016.04.002

Different types of yeasts were employed as biocatalysts in the reduction of β-ketonitriles. The red microorganism, Rhodotorula rubra, was selected as the best performing catalyst in the reduction of different substituted ketonitriles giving total stereoselectivity in most cases (90-99% ee). In particular, its use as fresh and lyophilised cells was expanded to a semi-preparative scale for the production of the duloxetine precursor 1a. R. rubra was screened in the reduction of alkylation products in comparison with Pichia henricii for assignment of configuration of products 2a and 11a after derivatisation with S-MPA.

Full text of DOI:10.1016/j.tetasy.2016.04.002

Tetrahedron: Asymmetry 27 (2016) 389–396
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Efficient methodology to produce a duloxetine precursor using
whole cells of Rhodotorula rubra
Isabella Rimoldi a, Giorgio Facchetti a, Donatella Nava a, Martina Letizia Contente b, Raffella Gandolfi a,
⇑
a Dipartimento di Scienze Farmaceutiche, Sez. Chimica Generale e Organica ‘A. Marchesini’, Università degli Studi di Milano, Via Venezian 21, 20133 Milano, Italy
b Dipartimento di Scienze per gli Alimenti, la Nutrizione, l’Ambiente, Università degli Studi di Milano, Via Mangiagalli 25, 20133 Milano, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Different types of yeasts were employed as biocatalysts in the reduction of b-ketonitriles. The red
microorganism, Rhodotorula rubra, was selected as the best performing catalyst in the reduction of differ-
ent substituted ketonitriles giving total stereoselectivity in most cases (90–99% ee). In particular, its use
as fresh and lyophilised cells was expanded to a semi-preparative scale for the production of the dulox-
etine precursor 1a. R. rubra was screened in the reduction of alkylation products in comparison with
Pichia henricii for assignment of configuration of products 2a and 11a after derivatisation with S-MPA.
Ó 2016 Published by Elsevier Ltd.
Received 11 March 2016
Accepted 11 April 2016
Available online 20 April 2016
1. Introduction
decade different isolated carbonyl reductases have been employed
in the reduction of these valuable precursors although their appli-
b-Hydroxy ketones are important building blocks for the syn-
thesis of chiral 1,3-amino alcohols, 1,3 diamines, 1,3-hydroxy acids
used for the preparation of naturally occurring molecules and
those of pharmaceutical interest as well.1–3 Indeed, they provide
the scaffold for the synthesis of different classes of drugs such as
fluoxetine, duloxetine, atomoxetine, atorvastatin, ezetimibe and
selective inhibitors of the serotonin and norepinephrine reuptake.4
These drugs, in the absence of a valid alternative, are bound to
retain a vast sales volume on pharmaceutical market with about
11.9 $ billions sold in 2011 and with a projected sale of about
13.4 $ billions for 2018.5
Different synthetic approaches have been optimized to produce
b-hydroxynitriles relying on the asymmetric transfer hydrogena-
tion of b-ketonitriles with the employment of chiral organic and
inorganic catalysts yielding the corresponding aminoalcohols after
a simple Pd/C reduction of the nitrile moiety.6–9
Unfortunately, high production costs along with an environ-
mentally hazardous metal toxicity have hampered the use of this
type of catalyst especially for the synthesis of products useful to
the pharmaceutical industry.
A valid alternative emerging for the production of these useful
intermediates is the use of biocatalysts, i.e. the kinetic resolution
starting from racemic mixtures of b-hydroxynitriles catalyzed by
lipases and nitrilases10–12 or the enantioselective reduction of
b-ketonitriles catalyzed by ketoreductases.13–15 Over the last
cation was correlated to the presence of redox cofactors thus inevi-
tably introducing the need for fine tuning the appropriate cofactor
recycle system.16–19 In contrast, the use of whole cells could over-
come this unprofitable drawback.
In this regard the use of bakers’ yeasts in the reduction of
b-ketonitriles allows products with high enantiomeric excesses in
an (S)-configuration to be obtained, but in low yields due to the
competitive non-stereoselective reductive
a-alkylation secondary
reaction leading to the insertion of an ethyl group at the b-posi-
tion; the formation of this by-product is due to the presence of
acetaldehyde deriving from microorganism metabolism.20–23 The
reaction equilibrium could be shifted in favour of the reduction
products over the alkylation ones by using a biphasic system or
by a precise monitoring of the reaction temperature, albeit with
a decrease of reaction rates and yields.23
Dehli et al. had previously studied the reduction of b-ketoni-
triles employing fungus Curvularia lunata whole cells achieving
the S configuration of the product in 77% yield and also in this case
the formation of the alkylation reaction products.24 Moreover, this
biocatalyst was able to reduce in a diastereo- and enantioselective
manner the corresponding alkylation products, in exactly the
opposite way from Saccharomyces cerevisiae, and thus allowing
chiral substituted b-hydroxynitriles to be obtained simply by
modulating growth and biotransformation reaction conditions.
Conversely, alkylation products were totally lacking when recom-
binant Escherichia coli whole cells were employed. This type of cell,
in fact, were proved to over express ketoreductases allowing the
isolation of both enantiomers of the product.16
⇑
Corresponding author. Tel.: +39 02 503 14609; fax: +39 02 503 14615.
E-mail address: raffaella.gandolfi@unimi.it (R. Gandolfi).
http://dx.doi.org/10.1016/j.tetasy.2016.04.002
0957-4166/Ó 2016 Published by Elsevier Ltd.

390
I. Rimoldi et al. / Tetrahedron: Asymmetry 27 (2016) 389–396
Indeed, in the literature the most relevant attempts relative to
(50 g/L) as co-substrate. The results obtained after 48 h of biotrans-
formation are reported in Table 1.
the obtainment of these valuable scaffolds in an enantiopure form
appeared to be focused on the stereoselective reduction of benzoy-
lacetonitrile precursors. For these reasons the selection of new bio-
catalysts able to selectively reduce b-keto nitriles bearing different
aromatic heterocyclic moieties appears extremely appealing, espe-
cially those ones that avoid the use of cofactor recycle systems for
accomplishing the desired products in higher yields.
Similarly to what has been reported in the literature,20 many of
the yeasts screened in this work were able to transform substrate 1
in the reductive alkylation product 2 (entries 1–13), although dif-
ferent behaviours were highlighted. In the case of S. cerevisiae
and of other anaerobic yeasts the substrate 1 is totally converted
into the product 2 as evidenced by entries 1–7. In the other cases
the formation of reduction products 1a and 2a was observed as
in biotransformation by Curvularia lunata CECT 2130, studied by
Dehli and Gotor,13 obtaining an enantiomerically enriched product
nevertheless with poor diastereomeric excesses of product 2a
(entries 8–13). Despite the low molar conversion, P. henricii was
the only yeast able to give a good enantiomeric excess in favour
of diastereomers in with (R,R)- and (S,R)-configurations of sub-
strate 2 (entry 10). Only the red yeasts, known to be aerobic ones,
allowed the reduction product 1a to be obtained with an (S)-con-
figuration (entries 14–17) with moderate-to-high enantiomeric
excess (62–98%) and without the formation of by-products. Alco-
hol 1a was instead obtained with the (R)-configuration using as
biocatalyst Tourolopsis magnoliae, still in good yield and ee (entry
12). On substituting the glucose as co-substrate with ethanol,
which is easily oxidised into acetaldehyde, a general increase in
the reductive alkylation product 2 was observed. This concurrent
reaction was particularly evident when red yeasts were employed,
2. Results and discussion
The use of whole cells in the reduction reactions is attractive as
it gives the possibility of exploiting the microbial metabolism for
the redox cofactor recirculation. On the other hand, the production
of acetaldehyde as a by-product generally results in low yields of
the desired product due to the competitive reductive alkylation
reaction occurring, especially with substrates that are very acti-
vated as in the case of b-ketonitriles (Scheme 1).
Herein selected aerobic and facultative anaerobic yeasts were
evaluated for their ability to reduce a series of substituted
b-ketonitriles. The preliminary screening was realized only with
the aim to address the reaction in favour of the reduction product
1a, using as standard substrate the 3-oxo-3-(thiophen-2-yl)propa-
nenitrile 1, whose reduction product is the valuable precursor of
duloxetine. All the reactions were conducted in presence of glucose
NAD(P)H
NAD(P)+
NAD(P)+
NAD(P)H
O
glucose
pyruvic acid
NAD(P)+
OH
ethanol
H
acetaldehyde
glycolytic pathway
OH
O
NAD(P)H
reduction
CN
S
CN
S
alkylation
S
1a
1
acetaldehyde
OH
O
O
CN
S
CN
NAD(P)H
S
CN
reduction
NAD(P)H
reduction
NAD(P)+
NAD(P)+
2a
2
Scheme 1. Biotransformation of b-ketonitriles with whole-cells.
Table 1
Biotransformation of 3-oxo-3-(thiophen-2-yl)propanenitrile 1 using glucose as co-substrate
Entry
Microorganisms
1a
2
2a
(%)a
ee (%)b
(%)a
(%)a
de (%)a
ee anti (%)c
ee syn (%)c
1
2
3
4
5
6
7
8
Saccharomyces marzianus CBS397
Pichia pini CBS 744
Pichia pastoris CBS 2612
Saccharomyces cerevisiae Zeus
Sporobolomyces holsaticus NCYC 420
Sporobolomyces tsugae CBS 503
Kluyveromyces marxianus CBS1553
Lindnera fabianii CBS 5640
Pichia glucozyma CBS 5766
Pichia henricii CBS 5765
—
—
—
—
—
—
—
18
—
12
19
87
34
53
100
100
100
100
100
100
100
97
100
91
79
89
80
70
0
—
—
—
—
—
—
—
—
12
8
11
13
66
—
—
—
58 (S)
9
10
11
12
13
14
15
16
17
61 (R)
58 (R)
67 (R)
29 (S)
62 (S)
73 (S)
97 (S)
98 (S)
39 syn
1 anti
2 syn
97 RR
14 SS
38 RR
65 SS
95 SR
37 SR
46 SR
46 RS
Pichia etchelsii
Torulopsis magnoliae IMAP 4425
Torulopsis molischiana CBS 837
Sporidiobolus pararoseus SD2
Sporobolomyces salmonicolor MIM
Rhodotorula rubra MIM146
Rhodotorula rubra MIM147
0
0
0
0
18 anti
0
—
a
b
c
Molar conversion and de% were calculated by 1H NMR analysis.
Ee% was determined by HPLC analysis using OJ-H Chiralcel as chiral column (hexane:2-propanol = 90:10, flow = 1.0 mL/min, k = 230 nm).
Ee% was determined by HPLC analysis in accordance to literature conditions.14

I. Rimoldi et al. / Tetrahedron: Asymmetry 27 (2016) 389–396
391
thus affording just a small quantity of product 2a, with good enan-
tiomeric excess in favour of SS diastereomer (Table 2).
substituent (entries 6 and 8) whereas the presence of a deactivat-
ing group in the meta (substrates 6 and 10, entries 5 and 9) or in
the para position (substrates 5 and 8, entries 4 and 7), lowered
the molar conversion Anyway, the ee% of the desired alcohols is
very high (94–99%) in all cases, confirming that R. rubra is a useful
catalyst in the reduction of activated ketonitriles. With the aim of
evaluating the possibility to produce semi-preparative amounts of
Taking into consideration the high enantioselectivity shown by
R. rubra MIM 147 in the reduction of 3-oxo-3-(thiophen-2-yl)
propanenitrile 1 (Table 1, entry 3), the scope of its particularly
interesting biocatalytic performance was expanded to the reduc-
tion of different substituted aryl b-ketonitriles. In particular, the
use of resuspended cells that were grown for 48 h was compared
to the use of lyophilised cells at a concentration of 20 g LÀ1 (dry
weight) successively resuspended in 0.1 M phosphate buffer at
pH = 7 or resuspended in tap water (Table 3).
The reduction of the carbonyl group, independent of the reac-
tion media, proceeded with a high reaction rate in 24 h by employ-
ing resuspended cells and afforded the reduction product only in
the (S)-configuration in all cases. The only exception was observed
in the case of a deactivated substrate as 6 (49%, entry 5). When the
biocatalyst was employed in a lyophilised form a general decrease
in the reaction rate was seen, probably as a consequence of a pos-
sible partial inactivation by the dehydrogenases. As an exception, a
total molar conversion after 72 h of the substrate can be observed
in the case of the unsubstituted aryl 4 (entry 3) or with the hetero-
cyclic aromatic substrates 1 and 3 (entries 2 and 3) where the het-
(S)-3-hydroxy-3-(thiophen-2-yl)propanenitrile,
a
precursor of
duloxetine, biotransformations were carried out using different
substrate concentrations: 2, 4, 6 or 10 g LÀ1 in tap water. Substrate
inhibition was not observed either at lower or higher concentra-
tions up to 10 g LÀ1, even if, as expected, a decrease in molar con-
version was highlighted starting from 6 g LÀ1 (complete conversion
in 48 h). In fact after 12 h the biotransformation was complete for
2 g LÀ1 concentration and in 24 h for 4 g LÀ1 concentration. After
chromatographic purification starting from 1 g of 1, 786 mg of 1a
were obtained. When 10 g LÀ1 substrate concentration was added,
80% yield was achieved in 96 h and the reaction didn’t proceed fur-
ther. In all cases stereoselectivity in favour of the (S)-enantiomer
remained constant (98–99%).
As the results obtained in this preliminary study highlighted
that using ethanol as co-substrate and R. rubra MIM 147 as the bio-
eroatom is in the
a-position to the carbonyl group. When the
catalyst the reductive alkylation product 2 is preferentially
aromatic ring is differently substituted, the best results in terms
of conversion were seen in the presence of a para activating
obtained (Table 2, entry 3), product 2a is not formed. Substrates
2, 11 and 12 were synthesised choosing Saccharomyces cerevisiae,
Table 2
Biotransformations of 3-oxo-3-(thiophen-2-yl)propanenitrile 1 using ethanol as co-substrate
Entry
Microorganisms
1a
2
2a
ee anti (%)c
(%)a
ee (%)b
(%)a
(%)a
de (%)a
ee syn (%)c
1
2
3
4
Sporobolomyces salmonicolor MIM
Rhodotorula rubra MIM146
Rhodotorula rubra MIM147
Sporidiobolus pararoseus SD2
38
17
12
21
61 S
97 S
98 S
82 S
42
78
88
74
20
5
0
71 anti
50 anti
77 SS
68 SS
8 RS
12 RS
5
41 anti
73 SS
52 RS
a
b
c
Molar conversion and de% were calculated by 1H NMR analysis.
Ee% was determined by HPLC analysis using OJ-H Chiralcel as chiral column (hexane:2-propanol = 90:10, flow = 1.0 mL/min, k = 230 nm).
Ee% was determined by HPLC analysis in accordance to literature conditions.14
Table 3
Reductive biotransformation of different aryl substituted b-ketonitriles using R. rubra as biocatalyst
O
OH
Rhodotorula rubra MIM147
N
N
R
R
1-10
1a-10a
(S)-
S
4
R'=H
R=
R'
1
3
R=
R=
5
R'= p-CH3
R'= m-CH3
R'= p-OCH3
R'= p-F
6
O
7
8
9
R'= p-CF3
R'= m-Cl
10
Entry
Substrate
Resuspended cells
Lyophilised cells
Molar conversion (%)b
Molar conversion (%)a
ee (% of S)
ee (% of S)
1
2
3
4
5
6
7
8
9
1
3
4
5
6
7
8
9
10
100
100
100
100
49
100
97
95
97
97
97
97
94
99
96
96
96
100
100
100
31
51
100
15
>99
98
98
96
90
99
99
98
94
90
3
100
a
Molar conversion (%) after 24 h was calculated by 1H NMR analysis.
Molar conversion (%) after 72 h. Ee% was determined by HPLC analysis using OJ-H Chiralcel as chiral column (hexane:2-propanol = 90:10, flow 1 mL/min, k = 216 nm).
b

392
I. Rimoldi et al. / Tetrahedron: Asymmetry 27 (2016) 389–396
O
OH
O
N
N
N
R
S. cerevisiae
Rhodotorula rubra MIM147
R
R
1-3-4
2-11-12
2a-11a-12a
O
S
2
11 R=
12
R=
R=
Scheme 2. Biotransformation by S. cerevisiae affording substrates 2, 11 and 12.
Table 4
enantiomeric excess for the anti-diastereomers was very high
(95–98%) although the diastereomeric excess was influenced by
Reductions of different rac-2-arylbutanenitrile using R. rubra MIM147 as biocatalyst
Entry
Substrate
dea anti (%)
eeb anti (% of S,S)
eeb syn (% of R,S)
the different aryl group in
a position to the carbonyl.
1
2
3
2
11
12
66
76
90
97
95
98
50
0
28
2.1. Assignment of configuration for 2a and 11a
The configuration for diastereomeric mixture of 12a was
reported in literature.9,14 The assignment of the configuration
for diastereomers of 2a and 11a was realised using R. rubra
MIM147 as biocatalyst, able to preferentially form the anti-(S,
S)-diastereomer (Table 4) and Pichia henricii CBS 5765, forming
the anti-(R,R)-diastereomer and one syn-(S,R)-diastereomer with
high ee% (Table 1, entry 10). Employing directly compounds 2
and 11 as substrate for biotransformations the reduction
reaction conducted with P. henricii compared to R. rubra
proceeded slowly but when substrate concentrations was equal
to 1 g LÀ1 and concentration resuspended was 40 g LÀ1 (dry
weight), an increase of diastereomeric excess for 2a was
observed (63% de syn). In the case of substrate 11 a diastere-
omeric excess of 85% syn was achieved. The reactions were
completed after 72 h.
a
de% were calculated from 1H NMR analysis.
Ee% was determined by HPLC analysis using Phenomenex Cellulose 4 as chiral
b
column (hexane:2-propanol = 90:10, flow 1 mL/min, k = 230 nm).
a catalyst reported in literature for its ability to produce this type
of substrates under suitable reaction conditions (Scheme 2).
These substrates are particularly interesting to study due to the
presence of additional stereogenic centre14 so the second part of
this research work focused on the capability to reduce and resolve
the diastereomers of the corresponding reduction products. The
results are summarized in the following table (Table 4).
As reported in our previous work25 for substrate 12, after 48 h
the biotransformations were realized using a substrate concentra-
tion of 1 g LÀ1 for all employed substrates. In all cases the
Figure 1. Comparison among 1H NMR spectra of racemic mixture (a), S-MPA derivatives (b) and HPLC chromatogram (c) for 11a.

I. Rimoldi et al. / Tetrahedron: Asymmetry 27 (2016) 389–396
393
The configurations for 2a and 11a were determined by sequen-
analysis was reported using chiral column Phenomenex Lux
Cellulose 4 (hexane:2-propanol = 90:10, flow 1 mL/min, k = 230 nm).
From 1H NMR spectra, the biotransformation with R. rubra, an
enantiomeric excess of 95% in favour of diastereomers with an
anti-configuration was obtained based on the literature configura-
tion assignment.14 On the contrary on reduction with P. henricii the
syn diastereomers were predominant (63% de). By comparison of
HPLC chromatograms and 1H NMR spectra of MPA-derivatives it
was possible to assign the configurations to all the four diastere-
omers (Fig. 2a and b).
tial comparison of: 1H NMR spectra with the corresponding ones
obtained after derivatisation with (S)-MPA (a-methoxy-phenyl
acetic acid),26 HPLC profiles of racemic mixture respectively of 2a
and 11a (Figs. 1 and 3) and those ones achieved by reduction with
R. rubra and P. henricii (Figs. 2 and 4).
As reported in Figure 1 a racemic mixture of 11a obtained by
reduction with NaBH4 was analyzed by 1H NMR spectroscopy
before (Fig. 1a) and after reaction with S-MPA in presence of
4-DMAP and DCC (Fig. 1b). In Figure 1c the corresponding HPLC
Figure 2. HPLC chromatograms for racemic mixture of 11a, reduction products from R. rubra and from P. henricii (a) compared with the corresponding 1H NMR spectra of
MPA-derivatives (b).
Figure 3. Comparison among 1H NMR spectra of racemic mixture (a), derivatives with MPA (b) and HPLC chromatogram (c) for 2a.

394
I. Rimoldi et al. / Tetrahedron: Asymmetry 27 (2016) 389–396
Figure 4. HPLC chromatograms for racemic mixture of 2a, reduction products from R. rubra and from P. henricii (a) compared with corresponding 1H NMR spectra of MPA-
derivatives (b).
The same approach was used to assign the configurations for
the four diastereomers obtained by reduction of 2a as reported
in the following Figures 3 and 4.
pH = 5.5). In order to obtain cells for the biocatalytic activity tests,
the microorganisms were grown on solid medium for 72 h at 28 °C
and then they were cultured in 1000 mL Erlenmeyer flasks con-
taining 100 mL of the medium (OD530nm mLÀ1 = 0.1 at t0).The
microorganisms were incubated for 48 h at 28 °C on a reciprocal
shaker (100 rpm). The yeasts were grown on malt extract with
5 g LÀ1 Difco yeast extract pH = 5.6, excepting for Sporidiobolus
pararoseus SD2 that grown on YPD medium (10 g LÀ1 Difco yeast
extract, 10 g LÀ1 peptone, 20 g LÀ1 glucose). Fresh cells from sub-
merged cultures were centrifuged (4000Âg for 15 min at 4 °C)
and washed with tap water before using. In the case of Rhodotorula
rubra MIM147 the cells were also lyophilised. The cells used in the
screening biotransformations were concentrated in ratio 1:2.
3. Conclusion
In conclusion the strategy to use the aerobic yeasts allows the
amount of alkylation product during biotransformation of ketoni-
triles to be reduced. Thus, in the presence of ethanol, that was
directly converted into acetaldehyde,
a major production of
alkylated product was seen. Rhodotorula rubra was revealed as a
valid biocatalyst for the reduction of different substituted
ketonitriles using either fresh whole cells or lyophilised ones.
The semipreparative production of dulexetine’s (S)-precursor in
very good enantiomeric was realized using tap water as reaction
solvent in 4 g/L substrate concentration. The assignment of
configuration of the four diastereomeric reduction products of 2a
and 11a was easily realised thanks to the possibility to obtain
the enriched products by Rhodotorula rubra versus Pichia henricii.
4.3. General procedure of biotransformation
The biotransformation screening was carried out in 10 ml-
screw capped test tubes re-suspending the yeast cells in 5 mL of
0.1 M phosphate buffer pH = 7 containing 50 g LÀ1 of glucose (or
ethanol 10 g LÀ1) and adding 2 g LÀ1of substrate 1 dissolved in
DMSO (1%). The reaction mixtures were magnetically stirred at
28 °C for 48 h. The product was extracted with diethyl ether
(2 Â 5 mL), dried with Na2SO4 and the solvent was removed in
vacuo. The samples were analysed by 1H NMR to determinate
the molar conversion and the diastereomeric excess while the
enantiomeric excess was recorded by HPLC equipped with OD-H
Chiralcel, eluent: hexane:2-propanol = 95:5, flow = 0.8 mL/min,
k = 230 nm; rt: 1a: (S) = 49.9 min, (R) = 52.2 min; 2 = 23.9 min;
2a: (S,S) = 26.5 min, (R,S) = 28.8 min, (S,R) = 34.3 min, (R,R) =
36.4 min.14
4. Experimental
4.1. General methods
1H and 13C NMR spectra were recorded in CDCl3 on Bruker
Avance 300 MHz equipped with a non-reverse probe. Chemical
shifts (in ppm) were referenced to residual solvent proton/carbon
peak. Polarimetry analyses were carried out on Perkin Elmer 343
Plus equipped with Na/Hal lamp. MS analyses were performed by
using a Thermo Finnigan (MA, USA) LCQ Advantage system MS
spectrometer with an electronspray ionisation source and an ‘Ion
Trap’ mass analyser. The MS spectra were obtained by direct
infusion of a sample solution in MeOH under ionisation, ESI
positive. Catalytic reactions were monitored by HPLC analysis with
Merck-Hitachi L-7100 equipped with Detector UV6000LP and
chiral column (OJ-H Chiralcel, OD-H Chiralcel or Phenomenex
Lux Cellulose 4). FTIR spectra were collected by using a Perkin
Elmer (MA, USA) FTIR Spectrometer ‘Spectrum One’ in a spectral
region between 4000 and 450 cmÀ1 and analysed by transmittance
technique with 32 scansions and 4 cmÀ1 resolution. Commercially
reagent grade solvents were dried according to standard proce-
dures and freshly distilled under nitrogen before use.
1a: 1H NMR (CDCl3, 300 MHz, 25 °C): d = 1.13 (t, J = 7.3 Hz, 3H),
2.88 (d, J = 5.9 Hz, 2H), 5.27–5.34 (m, 1H), 7.00 (t, J = 4.8 Hz, 1H),
7.10 (d, J = 3.3 Hz, 1H), 7.32 (d, J = 4.7 Hz, 1H) ppm; 13C NMR
(CDCl3, 75 MHz, 25 °C): d = 28.43, 66.38, 117.33, 124.93, 125.95,
127.33, 144.75 ppm. IR
m = 3429, 2922, 2254, 1413, 1040, 850,
709 cmÀ1. MS (ESI) of C7H7NOS (m/z): calcd 153.2, found 175.8
[M+Na+].
2: 1H NMR (CDCl3, 300 MHz, 25 °C): d = 1.14 (t, J = 7.3 Hz, 3H),
1.99–2.15 (m, 2H), 4.11–4.18 (d, J = 6.2, J = 1.4 Hz, 1H), 7.15–7.20
(m, 1H), 7.75 (d, J = 7.6 Hz, 1H) 7.84 (d, J = 6.7 Hz, 1H) ppm; 13C
NMR (CDCl3, 75 MHz, 25 °C): d = 11.82, 24.36, 42.62, 117.37,
128.92, 133.86, 136.82, 141.16, 183.62 ppm. IR
m = 3094, 2974,
2934, 2248, 1671, 1413, 1247, 1206, 1064, 731 cmÀ1. MS (ESI) of
C9H9NOS (m/z): calcd 179.2, found 201.8 [M+Na+].
4.2. Microorganisms: culture conditions
2a: 1H NMR (CDCl3, 300 MHz, 25 °C): d = 1.09 (t, J = 7.4 Hz, 3H,
anti), 1.12 (t, J = 7.4 Hz, 3H, syn), 1.51–1.68 (m, 2H, anti), 1.69–1.8
(m, 2H, syn), 1.58 (br, -OH), 2.82–2.92 (m, 2H, anti), 2.94–2.99
Strains from official collections or from our collection were
routinely maintained on a malt extract (8 g LÀ1 agar 15 g LÀ1
,

I. Rimoldi et al. / Tetrahedron: Asymmetry 27 (2016) 389–396
395
(m, 2H, syn), 5.05 (d, J = 6.3 Hz, 1H, anti), 5.08 (d, J = 6.5 Hz, 1H,
syn), 6.99–7.10 (m, 1H, syn + anti), 7.11 (d, J = 3.6 Hz, 1H, anti),
7.13 (d, J = 3.1 Hz, 1H, syn), 7.32 (d, J = 5.2 Hz, 1H, anti + syn)
ppm; 13C NMR (CDCl3, 75 MHz, 25 °C): d = 11.77 (syn), 11.83
(anti), 21.69 (syn), 22.69 (anti), 42.81 (syn), 43.23 (anti), 69.87
(syn), 70.16 (anti), 120.01 (syn), 120.14 (anti), 125.55 (anti),
125.7 (syn), 125.97 (syn), 126.02 (anti), 127.2 (anti + syn), 143.71
4.6. Synthesis of rac-2-(hydroxy(thiophen-2-yl)methyl)buta-
nenitrile 2a, rac-2(furan-2-yl(hydroxyl)methyl)butanenitrile
11a and rac-2-(hydroxyl(phenyl)methyl)butanenitrile 12a
The rac-2a, rac-11a and rac-12a were obtained by reduction of
the substrates 2, 11 and 12 (1 equiv) in ethanol with NaBH4
(1.5 equiv). The reactions were magnetically stirred at room tem-
perature for 4 h and extracted with water and ethyl acetate
(3 Â 50 mL), dried with Na2SO4 and the solvent was removed in
vacuo.
(syn), 144.05 (anti) ppm. IR
m
= 3435, 2970, 2933, 2245, 1711,
1461, 1032, 706 cmÀ1
. MS (ESI) of C9H11NOS (m/z): calcd
181.27, found 204.8 [M+Na+].
2a: HPLC conditions: column: Phenomenex Lux Cellulose 4,
eluent: hexane:2-propanol = 90:10, flow = 1 mL/min, k = 230 nm;
rt: (R,S) = 6.43 min, (S,S) = 8.73 min, (S,R) = 7.94 min, (R,R) =
10.27 min.
4.4. Screening of different aryl ketonitriles with Rhodotorula
rubra MIM147
11a: 1H NMR (CDCl3, 300 MHz, 25 °C): d = 1.10 (t, J = 7.4 Hz, 3H,
anti), 1.11 (t, J = 7.4 Hz, 3H, syn), 1.54–1.76 (m, 2H, anti + syn),
2.97–3.08 (m, 2H, anti + syn), 4.80 (d, J = 6.3 Hz, 1H, anti), 4.85 (d,
J = 6.8 Hz, 1H, syn), 6.36–6.39 (m, 2H, anti), 6.41 (d, J = 3.3 Hz, 1H,
syn), 6.45 (d, J = 3.3 Hz, 1H, syn), 7.41–7.43 (m, 1H, anti + syn)
ppm;13C NMR (CDCl3, 75 MHz, 25 °C): d = 11.71 (syn), 11.82 (anti),
21.67 (syn), 22.65 (anti), 40.09 (syn), 40.49 (anti), 67.64 (syn), 67.92
(anti), 108.35 (anti) 108.50 (syn), 110.79 (syn + anti), 119.81 (syn),
119.99 (anti), 143.06 (syn + anti), 152.78 (syn + anti) ppm. IR
Fresh cells and lyophilised cells of Rhodotorula rubra MIM147
were used 20 g LÀ1 dry weight re-suspended in phosphate buffer
or in tap water with 50 g LÀ1 of glucose; then the substrates 1,
3–10 were dissolved in DMSO and added at 2 g LÀ1 concentration.
For substrate 1, biotransformations were also carried out at 4, 6 or
10 g LÀ1 concentration.
HPLC conditions:6,8 OJ-H Chiralcel, eluent: hexane:2-propa-
nol = 90:10, flow = 1.0 mL/min, k = 216 nm.
Compound 1a: (S) = 32.1 min, (R) = 38.3 min; compound 3a rt
(S) = 22.2 min, (R) = 25.8 min; compound 4a rt (S) = 23.4 min,
(R) = 29.7 min; compound 5a rt (S) = 20.6 min, (R) = 24.4 min;
compound 6a rt (S) = 18.5 min, (R) = 22.0; compound 10a rt
m
= 3434, 2971, 2936, 2244, 1730, 1462, 1149, 745 cmÀ1. MS (ESI)
of C9H11NO2 (m/z): calcd 165.3, found 187.9 [M+Na+]. HPLC
conditions: column: Phenomenex Lux Cellulose 4, eluent:
hexane:2-propanol = 90:10, flow = 1 mL/min, k = 216 nm; rt: (R,S)
= 6.43 min, (S,S) = 9.95 min, (S,R) = 7.38 min, (R,R) = 11.49 min.
12a: 1H NMR (CDCl3, 300 MHz, 25 °C): d = 1.09 (t, J = 7.7 Hz, 3H,
anti), 1.17 (t, J = 7.6 Hz, 3H, syn), 1.51–1.69 (m, 2H), 2.76–2.83 (m,
2H, anti), 2.87–2.95 (m, 2H, syn), 4.79 (d, J = 6.2 Hz, 1H, anti), 4.83
(d, J = 6.6 Hz, 1H, syn), 7.33–7.56 (m, 5H) ppm; 13C NMR (CDCl3,
75 MHz, 25 °C): d = 9.84 (syn), 10.38 (anti), 24.86 (anti), 25.87
(syn), 41.43 (anti), 42.51 (syn), 76.86 (syn), 77.46 (anti), 127.04
(syn + anti), 128.05 (anti), 128.22 (syn), 128.55 (anti), 128.69
(S) = 23.7 min, (R) = 28.6 min; compound
7 rt (S) = 47.7 min,
(R) = 49.5 min; compound 8a rt (S) = 21 min, (R) = 27.7 min;
compound 9 rt (S) = 15.5 min, (R) = 17.9 min.
4.5. Enzymatic reduction of rac-2-(thiophene-2-carbonyl)
butanenitrile 2, rac-2-(furan-2-carbonyl)butanenitrile 11 and
rac-2-benzoylbutanenitrile 12
Commercial Baker’s yeast (50 g LÀ1) was suspended in a phos-
phate buffer (200 mL, 0.1 M, pH = 7) containing 50 g LÀ1 of glucose
and 5 g LÀ1 of the substrate (1, 3 or 4) was added. The biotransfor-
mation system was shaken with mechanic stirrer at 28 °C. When
the total conversion was achieved, the cells were separated by cen-
trifugation. Both the aqueous phases and the cells mixture were
extracted with diethyl ether (3x50 mL), dried with Na2SO4 and
the solvent was removed in vacuo.
2: The crude product was purified by flash chromatography
(cyclohexane/diisopropyl ether = 7:3) to give 1.04 g of 2 (88%
yield).
11: The crude product was purified by flash chromatography
(CH2Cl2/cyclohexane/ethyl acetate = 4:1:1) to give 929 mg
of 11 (77% yield). 1H NMR (CDCl3, 300 MHz, 25 °C): d = 1.14
(t, J = 7.4 Hz, 3H), 2.02–2.09 (m, 2H), 4.11–4.16 (q, J = 6.2, 4.3 Hz,
1H), 6.61–6.64 (dd, J = 3.6 Hz, J = 1.9 Hz, 1H), 7.40 (d, J = 4.4 Hz,
1H) 7.68 (s, 1H) ppm; 13C NMR (CDCl3, 75 MHz, 25 °C): d = 11.73,
23.66, 41.91, 113.45, 117.06, 119.93, 147.88, 150.57, 179.74 ppm.
(syn), 140.61 (anti), 141.42 (syn) ppm. FTIR
m = 3390, 2964, 1494,
1453, 160, 1103, 1038, 702 cmÀ1. MS (ESI) of C11H11NO (m/z): calcd
175.1, found 198.3 [M+Na+]. HPLC conditions: column: OD-H
Chiralcel, eluent: hexane:2-propanol = 95:5, flow = 0.8 mL/min,
k = 216 nm. rt: (R,S) = 25.6 min, (S,S) = 26.5 min, (S,R) = 34.6 min,
(R,R) = 36.0 min.
4.7. 2-(Hydroxy(thiophen-2-yl)methyl)butanenitrile 2a and 2
(furan-2-yl(hydroxyl)methyl)butanenitrile 11a derivatisation
The enantiomerically enriched mixtures of 2a and 11a were
obtained by biotransformation using fresh cells of Rhodotorula
rubra MIM 147 (20 g LÀ1 dry weight) and Pichia henricii CBS 5765
(40 g LÀ1 dry weight) suspended in phosphate buffer 0.1 M
pH = 7 containing 50 g LÀ1 of glucose. Compounds 2 or 11 were
dissolved in DMSO and added to the biotransformation system
to a final substrate concentration of 1 g LÀ1 in presence of 1%
solvent. After 72 h the reactions were extracted with diethyl
ether (3 Â 50 mL), dried with Na2SO4 and the concentrated at
reduce pressure. The crude products 2a and 11a were purified by
column chromatography on silica gel (CH2Cl2/cyclohexane/ethyl
acetate = 4:1:1).
IR
m = 3135, 2964, 2933, 2249, 1681, 1567, 1464, 1394, 1261,
1021, 771 cmÀ1. MS (ESI) of C9H9NO2 (m/z): calcd 163.2, found
185.8 [M+Na+].
12: The crude product was purified by flash chromatography
(CH2Cl2/hexane/ethyl acetate = 4:1:1) to give 1.02 g of 12 (86%
yield). 1H NMR (CDCl3, 300 MHz, 25 °C): d = 1.16 (t, J = 7.7 Hz,
3H), 2.02–2.15 (m, 2H), 4.30 (dd, J = 6.2, 4.3 Hz, 1H), 7.49–7.56
(m, 2H), 7.65 (d, J = 7.6 Hz, 1H) 7.95 (d, J = 6.7 Hz, 2H) ppm;
13C NMR (CDCl3, 75 MHz, 25 °C): d = 11.71, 23.77, 41.69,
1H NMR of the corresponding derivatives with MPA (
a-meth-
oxy-phenyl acetic acid): To a solution of rac-2a and rac-11a and
the enriched products obtained by biotransformation (1 equiv) in
CDCl3 (0.75 mL), S-(+)-MPA (1 equiv), 4-DMAP (0.5 equiv) and
DCC (1.5 equiv) were added.26
1H NMR for 2a (CDCl3, 300 MHz, 25 °C) d = 6.08 (d, J = 7.1 Hz,
1H, (S,R)-syn), 6.10 (d, J = 7.1 Hz, 1H, (R,S)-syn), 6.14 (d, J = 6.8 Hz,
1H, (R,R)-anti), 6.15 (d, J = 6.8 Hz, 1H, (S,S)-anti) ppm. 1H NMR for
11a (CDCl3, 300 MHz, 25 °C) d = 6.10 (d, J = 3.3 Hz, 1H, (S,S)-anti),
117.41, 128.92–134.63, 170.91, 190.97 ppm. IR
2936, 2249, 1694, 1597, 1449, 1265, 1233, 1208, 1000,
696 cmÀ1
MS (ESI) of C11H11NO (m/z): calcd 173.2, found
196.1 [M+Na+].
m = 3467, 2975,
.

396
I. Rimoldi et al. / Tetrahedron: Asymmetry 27 (2016) 389–396
4. Soni, P.; Banerjee, U. C. Appl. Microbiol. Biotechnol. 2005, 67, 771–777.
5. Every-Palmer, S.; Howick, J. J. Eval. Clin. Pract. 2014, 20, 908–914.
6. Soltani, O.; Ariger, M. A.; Vázquez-Villa, H.; Carreira, E. M. Org. Lett. 2010, 12,
2893–2895.
6.17 (d, J = 3.3 Hz, 1H, (R,S)-syn), 6.20–6.23 (m, 1H, (S,S)-anti),
6.24–6.26 (m, 1H, (R,S)-syn), 6.20–6.23 (m, 1H, (R,R)-anti),
6.34–6.36 (m, 1H, (S,R)-syn), 6.44 (d, J = 3.3 Hz, 1H, (R,R)-anti) ppm.
7. Vázquez-Villa, H.; Reber, S.; Ariger, M. A.; Carreira, E. M. Angew. Chem., Int. Ed.
2011, 50, 8979–8981.
4.8. Synthesis of (S)-3-hydroxy-3-(thiophen-2-yl)propanenitrile
1a with Rhodotorula rubra MIM 147
8. Chen, C.; Kong, L.; Cheng, T.; Jin, R.; Liu, G. Chem. Commun. 2014, 10891–10893.
9. Zerla, D.; Facchetti, G.; Fuse, M.; Pellizzoni, M.; Castellano, C.; Cesarotti, E.;
Gandolfi, R.; Rimoldi, I. Tetrahedron: Asymmetry 2014, 25, 1031–1037.
10. Träff, A.; Lihammar, R.; Bäckvall, J.-E. J. Org. Chem. 2011, 76, 3917–3921.
11. Kamal, A.; Khanna, G. B. R.; Ramu, R. Tetrahedron: Asymmetry 2002, 13, 2039–
2051.
12. Kamila, S.; Zhu, D.; Biehl, E. R.; Hua, L. Org. Lett. 2006, 8, 4429–4431.
13. Dehli, J. R.; Gotor, V. Tetrahedron: Asymmetry 2000, 11, 3693–3700.
14. Dehli, J. R.; Gotor, V. Tetrahedron: Asymmetry 2001, 12, 1485–1492.
15. Zhu, D.; Yang, Y.; Hua, L. J. Org. Chem. 2006, 71, 4202–4205.
16. Xu, G.-C.; Yu, H.-L.; Zhang, Z.-J.; Xu, J.-H. Org. Lett. 2013, 15, 5408–5411.
17. Nowill, R. W.; Patel, T. J.; Beasley, D. L.; Alvarez, J. A.; Jackson Iii, E.; Hizer, T. J.;
Ghiviriga, I.; Mateer, S. C.; Feske, B. D. Tetrahedron Lett. 2011, 52, 2440–2442.
18. Zhu, D.; Ankati, H.; Mukherjee, C.; Yang, Y.; Biehl, E. R.; Hua, L. Org. Lett. 2007, 9,
2561–2563.
19. Hammond, R. J.; Poston, B. W.; Ghiviriga, I.; Feske, B. D. Tetrahedron Lett. 2007,
48, 1217–1219.
20. Smallridge, A. J.; Ten, A.; Trewhella, M. A. Tetrahedron Lett. 1998, 39, 5121–
5124.
21. Itoh, T.; Takagi, Y.; Fujisawa, T. Tetrahedron Lett. 1989, 30, 3811–3812.
22. Fuganti, C.; Pedrocchi-Fantoni, G.; Servi, S. Tetrahedron Lett. 1990, 31, 4195–
4198.
23. Florey, P.; Smallridge, A. J.; Ten, A.; Trewhella, M. A. Org. Lett. 1999, 1, 1879–
1880.
Cells obtained by centrifugation (4000Âg for 15 min at 4 °C) of
the culture broth (500 mL) were washed with tap water
(3 Â 200 mL) and re-suspended in 250 mL of tap water containing
50 g LÀ1 of glucose. 1 g of 1 was dissolved in DMSO and added to
the biotransformation system to give 4 g LÀ1 of substrate concen-
tration and 1% of solvent. The biotransformation system was
shaken with mechanic stirrer at 28 °C for 24 h. The cells were
separated by centrifugation and the two fractions were extracted
with diethyl ether (3 Â 100 mL), the combined organic layer was
dried with Na2SO4 and the solvent was removed in vacuo. The
crude product was purified by flash chromatography (ethyl
acetate/cyclohexane = 7:3) to give 786 mg of (S)-2a (78% yield).
(S)-1a: All characterization data are in agreement with previously
reported. [
a]
20 = À42.1 (c 0.8, CHCl3).
D
References
24. Gotor, V.; Dehli, J. R.; Rebolledo, F. J. Chem. Soc., Perkin Trans. 1 2000, 307–309.
25. Facchetti, G.; Gandolfi, R.; Fuse, M.; Zerla, D.; Cesarotti, E.; Pellizzoni, M.;
Rimoldi, I. New J. Chem. 2015.
1. Simon, R. C.; Mutti, F. G.; Kroutil, W. Drug Discovery Today: Technol. 2013, 10,
e37–e44.
26. Feske, B. D.; Kaluzna, I. A.; Stewart, J. D. J. Org. Chem. 2005, 70, 9654–9657.
2. Müller, M. Angew. Chem., Int. Ed. 2005, 44, 362–365.
3. Kamal, A.; Khanna, G. B. R.; Krishnaji, T.; Ramu, R. Bioorg. Med. Chem. Lett. 2005,
15, 613–615.