Biotransformation with whole microbial systems in a continuous flow reactor: Resolution of (RS)-flurbiprofen using Aspergillus oryzae by direct esterification with ethanol in organic solvent

Article abstract of DOI:10.1016/j.tetlet.2013.08.119

Cell-bound lipases of dry mycelium of Aspergillus oryzae were used in organic solvent for the resolution of racemic flurbiprofen by direct esterification with ethanol in a flow-chemistry reactor. Under flow conditions a significant reduction of the reaction time and an increase of the enantioselectivity were achieved compared to the batch mode. Moreover, the process was implemented by adding an in-line purification step integrated with the racemization of the unreacted flurbiprofen directly into a polymer-supported resin.

Full text of DOI:10.1016/j.tetlet.2013.08.119

Tetrahedron Letters 54 (2013) 6090–6093
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Biotransformation with whole microbial systems in a continuous
flow reactor: resolution of (RS)-flurbiprofen using Aspergillus oryzae
by direct esterification with ethanol in organic solvent
b
a
a
a
a
Lucia Tamborini a, , Diego Romano , Andrea Pinto , Martina Contente , Maria C. Iannuzzi , Paola Conti ,
⇑
Francesco Molinari b
a Dipartimento di Scienze Farmaceutiche, Universitá degli Studi di Milano, via Mangiagalli 25, 20133 Milano, Italy
b Dipartimento di Scienze per gli Alimenti, la Nutrizione e l’Ambiente (DEFENS), Universitá 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:
Cell-bound lipases of dry mycelium of Aspergillus oryzae were used in organic solvent for the resolution of
racemic flurbiprofen by direct esterification with ethanol in a flow-chemistry reactor. Under flow condi-
tions a significant reduction of the reaction time and an increase of the enantioselectivity were achieved
compared to the batch mode. Moreover, the process was implemented by adding an in-line purification
step integrated with the racemization of the unreacted flurbiprofen directly into a polymer-supported
resin.
Received 22 July 2013
Revised 2 August 2013
Accepted 28 August 2013
Available online 6 September 2013
Keywords:
Enzymatic resolution
Dry mycelia
Ó 2013 Elsevier Ltd. All rights reserved.
Flow reactor
Flurbiprofen
Aspergillus oryzae
Introduction
morphological structure allows for easy filtration and re-utiliza-
tion; this last feature favors simple set-up of continuous
Carboxylesterases (mostly lipases) have been used in monopha-
sic organic solution under controlled values of water activity (aw)
for catalyzing ester formation; the strategies more frequently
employed for shifting the equilibrium toward ester formation
imply the use of interesterification or transesterification, because
direct esterification is often hampered by water formation, which
negatively influences the equilibrium.1 The use of cell-bound
enzymes of different microorganisms has proved an effective
method for direct esterification of different alcohols and carboxylic
acids in organic solvent.2 Dry whole mycelia of filamentous eumy-
cetes can be directly used as biocatalyst, showing few advantages,
such as: high stability in organic solvents, high resistance to the
inactivation due to free carboxylic acids (including acetic acid),
and high molar conversions enabled by favorable partition of
water.3 Previous observations suggested that the mycelium con-
tributes marginally to the water take-up: rather than sequestrating
water inside the cell wall, it seems that the mycelium provides a
micro-environment where the water produced during the esterifi-
cation is promptly removed.4 Furthermore, mycelial microorgan-
isms can be employed without immobilization, because their
bioreactors.
Direct esterification of racemic alcohols or carboxylic acids with
dry mycelia of strains of Aspergillus oryzae and Rhizopus oryzae
often resulted in an efficient kinetic resolution.5 For example,
A. oryzae has been used in pure organic solvent for the resolution
of (RS)-flurbiprofen,6 displaying good enantioselectivity toward
(R)-flurbiprofen and furnishing results competitive with the data
obtained using commercial enzymes.7 However, the biotransfor-
mations showed limited productivity and space time yield due to
substrate inhibition effects and low solubility of flurbiprofen in
the solvents where the best results, in terms of activity and enanti-
oselectivity, were observed (e.g. aliphatic hydrocarbons).6 Kinetic
resolution of racemic flurbiprofen is attractive because (S)-flurbi-
profen is a nonsteroidal anti-inflammatory drug, whereas its
R-enantiomer shows anticancer effects in vivo and in vitro.8
Moreover, as recently reviewed by Kourist and co-workers,9 de-
spite the considerable number of biocatalytic routes developed
so far for the production of (S)-profens, there is still room for
improvement with the perspective of more efficient and sustain-
able processes.
Flow reactors can dramatically improve the performances of li-
pase-catalyzed reactions.10 Recently, the kinetic resolution of flur-
biprofen has been performed in a continuous flow reactor using an
⇑
Corresponding author. Tel.: +39 0250319322; fax: +39 0250313926.
E-mail address: lucia.tamborini@unimi.it (L. Tamborini).
0040-4039/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.08.119

L. Tamborini et al. / Tetrahedron Letters 54 (2013) 6090–6093
6091
CH3
O
CH3
CH3
O
OH
OEt
OH
+
EtOH
+
O
1 equiv
F
F
F
Dry micelia
1
2
(R)-
1
(S)-
(RS)-flurbiprofen
1 equiv
Scheme 1. Esterification of (RS)-flurbiprofen 1 with EtOH using mycelium of Aspergillus oryzae MIM in a continuous flow reactor.
Table 1
Screening of different flow rates in the enantioselective esterification of (RS)-flurbiprofen 1 with EtOH using Aspergillus oryzae MIM in a continuous flow reactor
Molar conversiona (%)
ees (%)
eep (%)
Ea
rc
(lmol/min g)
b
b
Entry
Time (min)
Flow rate (ll/min)
Batch
1440
15
8
6.5
5
—
46
85
105
137
20
58
45
28
19
22
90
66
34
20
86
66
82
88
88
16
18
19
21
19
0.14
1.48
2.12
1.63
1.44
1
2
3
4
Reaction conditions: 10 mM solution of (RS)-flurbiprofen in n-heptane, 1 equiv of EtOH, 180 mg of lyophilized mycelium of Aspergillus oryzae MIM, T = 50 °C.
a
Calculated according to Ref. 17.
b
Determined by chiral HPLC.14
c
Calculated according to Ref. 18.
immobilized lipase B from Candida antarctica (Novozym 435Ò)
allowing a significant reduction of the reaction time compared to
the classical batch method.11
Temperature was kept constant at 50 °C, while flow rate, which
sets the residence time, was varied and the best results are re-
ported in (Table 1).16
In the present work, we have studied the use of dry mycelium of
A. oryzae12 in an organic solvent for the resolution of racemic flur-
biprofen in a continuous flow reactor, combining the advantages of
an easy to produce (and use) biocatalyst with a process-intensifica-
tion technology. The process was implemented by adding an in-
line purification step integrated with the racemization of the unre-
acted flurbiprofen directly into a polymer-supported resin.
The data reported in Table 1 indicate that the use of a flow reac-
tor dramatically reduced the reaction time and slightly increased
the enantioselectivity. In fact, a 19% conversion was reached in
only 5 min of residence time with 88% enantiomeric excess (ee)
of the product (Table 1, entry 4). On the other hand, a residence
time of 15 min (entry 1) resulted in a marked increase of the con-
version (58%) allowing, in this case, the obtainment of (S)-flurbi-
profen with a good enantiomeric excess (90% ee). Therefore, by
simply modulating the flow rate, optically enriched substrate
[(S)-flurbiprofen, entry 1] or product [(R)-flurbiprofen ethyl ester,
entries 3 and 4] could be obtained.
Results and discussion
The esterification reaction catalyzed by dry mycelia of A. oryzae
was firstly performed in a batch mode using stoichiometric
amounts of racemic flurbiprofen and ethanol in n-heptane at
50 °C.13 The progress and stereobias of the reaction was monitored
by chiral HPLC,14 which showed that after 8 h the enantiomeric ex-
cess of (S)-flurbiprofen 1 was 22%, the enantiomeric excess of the
product (R)-flurbiprofen ethyl ester 2 was 86%, and the conversion
was 20%.
Flow experiments were initially focused on reaching a degree of
conversion similar to the one obtained in the batch process.15 To
this aim, a 10 mM solution of (RS)-flurbiprofen 1 in n-heptane con-
taining 1 equiv of EtOH was made to flow through a glass column
loaded with dry mycelium of A. oryzae MIM (180 mg, Scheme 1).
The specific reaction rate (r) of the batch reaction was
0.14 lmol/min g, whereas in the flow process, using the conditions
reported in Table 1, entry 4, which provided a similar degree of
conversion of the batch reaction, the specific reaction rate was
1.44 lmol/min g, that means an increase of the productivity of
about 10 times. It must be noted that the resolution is obtained
through direct esterification with formation of water, which can af-
fect the equilibrium of the reaction and, consequently, the overall
stereoselectivity. Therefore, experiments in the presence of molec-
ular sieves were carried out. For this purpose, we prepared a mixed
bed column filled with an equal weight of lyophilized mycelium of
A. oryzae MIM and molecular sieves. Keeping constant the temper-
Table 2
Screening of different temperatures in the enantioselective esterification of (RS)-flurbiprofen with EtOH using Aspergillus oryzae MIM in the presence of molecular sieves in a
continuous flow reactor
b
b
Entry
T (°C)
Molar conversiona (%)
ees (%)
eep (%)
Ea
rc
(lmol/min g)
1
2
3
4
5
50
30
40
60
70
38
30
32
41
42
54
38
42
62
60
88
88
88
90
84
26
22
23
35
21
2.45
1.93
2.06
2.64
2.70
Reaction conditions: 10 mM solution of (RS)-flurbiprofen in n-heptane, 1 equiv of EtOH, 180 mg of lyophilized mycelium of Aspergillus oryzae MIM, and 180 mg of molecular
sieves (powder, 4 Å). Residence time: 6.5 min; flow rate 116
lL/min.
a
Calculated according to Ref. 17.
b
Determined by chiral HPLC.14
c
Calculated according to Ref. 18.

6092
L. Tamborini et al. / Tetrahedron Letters 54 (2013) 6090–6093
STEP B:
1) 10% DBU in n-heptane, Δ
2) n-heptane
STEP A
3) 5%AcOH in n-heptane
CH3
O
OEt
116 μl min−1
100 psi
CH3
F
Dry micelia
A21
Molecular sieves
(R)-2
1
(RS)-flurbiprofen + EtOH
(10 mM in n-heptane)
OH
O
F
(RS)-flurbiprofen 1
Scheme 2. Schematic representation of the overall process.
ature and the residence time at 6.5 min, it was possible to observe
an increase of the molar conversion from 28% (Table 1, entry 3) to
38% (Table 2, entry 1). Different temperatures were then tested
(Table 2, entries 2–5). A slight increase of both the molar conver-
sion and ee of the product was achieved working at 60 °C (Table 2,
entry 4). No significant differences in the conversion or in the ee of
the product were observed by increasing the equivalents of EtOH.
The addition of molecular sieves generally increased the specific
reaction rates and enantioselectivity; under the best conditions
(Table 2, entry 4), the specific rate was 19 times higher than the
one registered in the batch mode and the enantiomeric ratio (E)
was increased from 16 to 35.
Moreover, the stability of the lyophilized mycelia was evalu-
ated. The thermal stability was already demonstrated by us in a
previous work,3a whereas the mycelia stability over continuous
work was checked by performing the resolution under the condi-
tions reported in Table 2 entry 4, and running the reactor for 3 h.
Samples were collected and analyzed every hour and the matching
of the results obtained over the time was verified both in terms of
conversion and enantiomeric purity of substrate and product.
As previously described,11 the overall process can be imple-
mented by adding an in-line purification step of the exiting solu-
tion, consisting in a catch and release protocol, which allows the
easy separation and recovery of both (S)-flurbiprofen and (R)-flur-
biprofen ethyl ester. Following the reported procedure, the unre-
acted carboxylic acid was trapped by flowing the exiting solution
into a column containing the polymer-supported base Amberlyst
A21 (Scheme 2). To further improve the process, the racemization
of the trapped (S)-flurbiprofen 1 was included. The complete race-
mization was achieved by filling the column with a 10% solution of
DBU in n-heptane (prepared through dilution of the DBU charged
in one injection loop with the n-heptane flow stream) and, after
stopping the flow stream, heating the column at 115 °C for 1 h
(lower concentrations of DBU did not allow the complete racemi-
zation of the trapped acid). The column was then washed with n-
B from Candida antarctica,11 the use of a continuous flow reactor al-
lows a significant reduction of the reaction time compared to the
classical batch method and dramatically improved the productivity
of the batch biotransformation, with beneficial effects also on the
enantioselectivity. The procedure here reported appears to be
notably simple because dry mycelium can be directly used in the
continuous reactor without immobilization. To the best of our
knowledge, this is the first application of microbial whole cells as
an enantioselective biocatalyst in a continuous flow reactor.
References and notes
1. (a) Faber, K. In Biotransformations in Organic Chemistry; Springer Verlag: Berlin,
2004. 5th edition; (b) Bornscheuer, U. T.; Kazlauskas, R. J. In Hydrolases in
Organic Synthesis; Wiley-VCH: Weinheim, 1999; (c) Conti, P.; Dallanoce, C.; De
Amici, M.; De Micheli, C.; Carrea, G.; Zambianchi, F. Tetrahedron: Asymmetry
1998, 9, 657–665.
2. (a) Bell, G.; Blain, J. A.; Paterson, J. D. E.; Shaw, C. E. L.; Todd, R. J. FEMS Microbiol.
Lett. 1978, 3, 223–225; (b) Molinari, F.; Marianelli, G.; Aragozzini, F. Appl.
Microbiol. Biotechnol. 1995, 43, 967–973; (c) Molinari, F.; Gandolfi, R.;
Aragozzini, F. Biotechnol. Tech. 1996, 10, 103–108; (d) Molinari, F.; Villa, R.;
Aragozzini, F. Biotechnol. Lett. 1998, 20, 41–44; (e) Gandolfi, R.; Marinelli, F.;
Lazzarini, A.; Molinari, F. J. Appl. Microbiol. 2000, 89, 870–875; (f) Gandolfi, R.;
Gaspari, F.; Franzetti, L.; Molinari, F. Ann. Microbiol. 2000, 50, 183–189.
3. (a) Molinari, F.; Gandolfi, R.; Converti, A.; Zilli, M. Enzyme Microb. Technol. 2000,
27, 626–630; (b) Converti, A.; Del Borghi, A.; Gandolfi, R.; Lodi, A.; Molinari, F.;
Palazzi, E. Biotechnol. Bioeng. 2002, 77, 232–237; (c) Converti, A.; Gandolfi, R.;
Zilli, M.; Molinari, F.; Binaghi, L.; Perego, P.; Borghi, M. Appl. Microbiol.
Biotechnol. 2005, 67, 637–640.
4. Molinari, F.; Romano, D.; Gandolfi, R.; Gardossi, L.; Hanefeld, U.; Converti, A.;
Spizzo, P. In Modern Biocatalysis; Fessner, W.-D., Anthonsen, T., Eds.; Wiley-
VCH: Weinheim, 2009; pp 79–91.
5. (a) Gandolfi, R.; Converti, A.; Pirozzi, D.; Molinari, F. J. Biotechnol. 2001, 92, 21–
36; (b) Gandolfi, R.; Gualandris, R.; Zanchi, C.; Molinari, F. Tetrahedron:
Asymmetry 2001, 12, 501–504; (c) Romano, D.; Ferrario, V.; Molinari, F.;
Gardossi, L.; Sanchez-Montero, J. M.; Converti, A. J. Mol. Catal. B: Enzym. 2006,
41, 71–74.
6. Spizzo, P.; Basso, A.; Ebert, C.; Gardossi, L.; Ferrario, V.; Romano, D.; Molinari, F.
Tetrahedron 2007, 63, 11005–11010.
7. (a) Morrone, G.; Nicolosi, G.; Patti, A.; Piattelli, M. Tetrahedron: Asymmetry
1995, 6, 1773–1778; (b) Duan, G.; Ching, C. B. Biochem. Eng. J. 1998, 2, 237–245;
(c) Ghanem, A. Chirality 2010, 22, 597–603.
8. (a) Flower, R. J. Nat. Rev. Drug Disc. 2003, 2, 179–191; (b) Wynne, S.; Djakiew, D.
Mol. Cancer Res. 2010, 8, 1656–1664.
9. Kourist, R.; Dominguez de Maria, P.; Miyamoto, K. Green Chemistry 2011, 13,
2607–2618.
heptane for 10 min at a total flow rate of 300 lL/min, and subse-
quently, flurbiprofen was released with a 5% solution of AcOH in
n-heptane and analyzed by chiral HPLC.
10. Itabaiana, I.; de Mariz e Miranda, L. S.; de Souza, R. O. M. A. J. Mol. Catal. B:
Enzym. 2013, 85–86, 1–9.
11. Tamborini, L.; Romano, D.; Pinto, A.; Bertolani, A.; Molinari, F.; Conti, P. J. Mol.
Catal. B: Enzym. 2012, 84, 78–82.
Conclusion
12. Aspergillus oryzae MIM was maintained on malt extract (8 g/L, agar 15 g/L, pH
5.5), cultivated in 500 mL-Erlenmeyer flasks containing 100 mL medium and
incubated for 48 h at 28 °C on a reciprocal shaker (100 rpm). Liquid media
contained a basic medium (yeast extract 1 g/L, (NH4)2SO4 5 g/L, K2HPO4 1 g/L,
MgSO4Á7H2O 0.2 g/L, pH 5.8) supplemented with Tween 80 (0.5%). Suspensions
of spores (1.6 Â 104) were used as inocula. Mycelia grown for 48 h in
Dry mycelia of A. oryzae can be effectively used for enantiose-
lective esterification of racemic flurbiprofen in organic solvent in
a flow reactor. As previously observed in the kinetic resolution of
flurbiprofen using the commercially available immobilized lipase

L. Tamborini et al. / Tetrahedron Letters 54 (2013) 6090–6093
6093
submerged cultures were harvested by filtration at 4 °C and resuspended in
deionized water, homogenized for 5 min (Silverson L2R, Silverson Machines,
Waterside, UK), frozen at À20 °C, and finally lyophilized (Alfa Criosec, Milan,
Italy) at plate temperature of 25 °C.
(S)-flurbiprofen 1: retention time 5.7 min; (R)-flurbiprofen ethyl ester 2:
retention time 14.4 min; (S)-flurbiprofen ethyl ester 2: retention time
15.9 min.
15. The continuous flow biotransformations were performed using a R2+/R4 flow
reactor commercially available from Vapourtec equipped with Omnifit glass
column (6.6 mm i.d. Â 100 mm length).
13. Resolution of (RS)-flurbiprofen in batch: The dry mycelium of Aspergillus oryzae
MIM (30 mg/mL) was suspended in 6.0 mL of n-heptane. Equimolar
concentrations (10 mM) of racemic flurbiprofen (14.6 mg, 0.06 mmol) and
ethanol (2.88 mg, 0.06 mmol) were added to the suspension and the reaction
mixture was maintained in a thermostated bath at 50 °C under constant
magnetic stirring for 8 h. The biotransformation medium was then centrifuged
and mycelium removed; the organic phase was treated with a NaOH aqueous
solution (4 mL, 0.1 N) and the organic extract washed with water (4 mL), dried
over Na2SO4, and evaporated to give flurbiprofen ethyl ester (3.3 mg, 20%
conversion, eep = 86%); the alkaline solution was brought to pH 1 with HCl
(0.5 N) and extracted twice with EtOAc and the resulting organic phase was
dried over Na2SO4 and evaporated to give unreacted flurbiprofen (11.7 mg,
16. Resolution of (RS)-flurbiprofen in flow: Racemic flurbiprofen
1 (14.6 mg,
0.06 mmol) and EtOH (1 equiv) were dissolved in n-heptane (6 mL) and the
solution was flowed through a glass column (6.6 mm i.d. Â 100 mm length)
filled with lyophilized cells of Aspergillus oryzae MIM (180 mg). When
molecular sieves were used (Table 2), a mixed bed column was prepared
using 180 mg of lyophilized cells of Aspergillus oryzae MIM and 180 mg of
molecular sieves (powder, 4 Å). A 100 psi back-pressure regulator was applied
to the system. Temperature and flow rate were varied and their values are
reported in Table 1 and Table 2 After collecting a total volume of 8 mL, the
reaction outcome was analyzed by chiral HPLC. To this aim, a sample of 200
was withdrawn, evaporated, and re-dissolved in acetonitrile (150 L).
lL
ees = 22%). The reaction was followed by HPLC. Withdrawals of 100
taken from the reaction mixture, filtered, evaporated, re-dissolved in
acetonitrile (100 L) and analyzed by chiral HPLC.
lL were
l
17. Chen, C.-S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J. Am. Chem. Soc. 1982, 104,
l
7294–7299.
}
14. HPLC analyses were performed with a Jasco PU-980 pump equipped with a
UV–vis detector Jasco UV-975 (wavelength: 254 nm). Column: Lux Amylose-2,
4.60 mm i.d. Â 150 mm, Phenomenex. Eluent: acetonitrile/water/formic acid
(1:1:0.2, v/v/v). Flow: 1.0 mL/min. (R)-flurbiprofen 1: retention time 4.3 min;
18. Csajái, C.; Szatzker, G.; Toke, E. R.; Ürge, L.; Darvas, F.; Poppe, L. Tetrahedron:
Asymmetry 2008, 19, 237–246.