Switching from (R)- to (S)-selective chemoenzymatic DKR of amines involving sulfanyl radical-mediated racemization

Article abstract of DOI:10.1039/c0ob00054j

Chemoenzymatic dynamic kinetic resolution (DKR) of amines involving sulfanyl radical-induced racemization happened to be the very first switchable DKR process allowing the synthesis of either (R)- or (S)-amides, in good yield and high enantiomeric excess,

Full text of DOI:10.1039/c0ob00054j

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Switching from (R)- to (S)-selective chemoenzymatic DKR of amines
involving sulfanyl radical-mediated racemization†
Lahssen El Blidi,a Nicolas Vanthuyne,b Didier Siri,c Ste´phane Gastaldi,*a Miche`le P. Bertrand*a and
Ge´rard Gil*b
Received 28th April 2010, Accepted 29th June 2010
First published as an Advance Article on the web 27th July 2010
DOI: 10.1039/c0ob00054j
Chemoenzymatic dynamic kinetic resolution (DKR) of amines involving sulfanyl radical-induced
racemization happened to be the very first switchable DKR process allowing the synthesis of either (R)-
or (S)-amides, in good yield and high enantiomeric excess, depending on the nature of the enzyme; the
different steps of the development of (S)-selective DKR are discussed.
Introduction
Owing to the prime importance of optically active amines in
the synthesis of natural products and bioactive compounds for
both academics and industrial chemists,1,2 amines dynamic kinetic
resolution (DKR) that allows the synthesis of optically pure
amides from racemic amines has stimulated numerous strategies.3,4
The chemoenzymatic process involving sulfanyl radical induced
racemization of aliphatic amines is currently investigated in our
group.5,6 Until now, the DKR of amines has specifically been
achieved with the thermostable polymer supported lipase B from
Candida antartica (Novozyme 435).4,6b These processes lead to
Scheme 1 Preliminary results.
(R)-amides. Proteases have a catalytic triad that is nearly the
mirror image of that of lipases7 which formally enables to devise
switchable chemoenzymatic DKR processes by changing the
nature of the enzyme.
pentanol. A compromise was found by changing the nature of
thiol and by adding a co-solvent.13
The compatibility between the amine resolution optimized
with CAL-B and the sulfanyl radical promoted racemization
enabled (R)-selective DKR to be achieved in good yield and high
A serious drawback remained, that is, the low stability of
alkaline protease under irradiation at 300 nm. An acceptable one-
pot three-step procedure was devised that consisted in performing
first a KR period in 3-methyl-3-pentanol, followed by addition of
trifluoroethane thiol and THF while irradiating for 2 h, and after
having stopped irradiation, a new cycle of enzymatic resolution
was performed by adding fresh portions of enzyme and acyl
donor (Scheme 1). These conditions allowed the preparation of
(S)-amides in yields ranging from 58% to 80% with enantiomeric
excesses ranging between 78% and 94%.6c However, the goal of a
“real” DKR process was still not reached.
◦
enantiomeric at 80 C (Scheme 1).6b However, the route to (S)-
selective DKR was not straightforward.
Most difficulties were inherent in protease characteristics.
Proteases are far less selective and less stable than lipases in organic
solvents.7,8 The screening of various serine proteases led us to select
the commercially available alkaline protease.9,10 This subtilisin like
protease, isolated from Bacillus licheniformis, was stabilized by
coating.8,11 The simultaneous screening of acyl donors led to the
selection of an N-octanoylglycine derivative.9
Efficient racemization could be achieved at low temperature
upon irradiation at 300 nm.12 However, racemization was slowed
down when it was carried out in KR solvent, i.e. 3-methyl-3-
Results and discussion
Fortunately, we found out that alkaline protease efficacy under
irradiation was greatly improved when initiating the formation
of sulfanyl radical at 350 nm in the presence of AIBN. Under
these conditions, AIBN was photochemically decomposed and
2-cyanopropyl radicals initiated the S–H bond homolysis.
The most significant results of racemization tests performed
under these new conditions on amine 2a are reported in Table 1.
The reproducibility of these racemizations was greatly improved
in the presence of an aprotic co-solvent. Meanwhile, the use of
t-butanol was reinvestigated, which was justified by the significant
improvement of the enzyme enantioselectivity in the KR experi-
ments that were performed simultaneously (Table 2).
aEquipe Chimie Mole´culaire Organique, LCP UMR 6264, Boite 562,
Universite´ Paul Ce´zanne, Faculte´ des Sciences St Je´roˆme, Avenue
Escadrille Normandie-Niemen, 13397, Marseille Cedex 20, (France).
E-mail: michele.bertrand@univ-cezanne.fr
bLaboratoire de Ste´re´ochimie Dynamique et Chiralite´, ISM2, UMR 6263,
Universite´ Paul Ce´zanne, Faculte´ des Sciences St Je´roˆme, Avenue Escadrille
Normandie-Niemen, 13397, Marseille Cedex 20, (France)
cEquipe Chimie The´orique et Mode´lisation Mole´culaire, LCP UMR 6264,
Universite´ de Provence, Faculte´ des Sciences St Je´roˆme, Avenue Escadrille
Normandie-Niemen, 13397, Marseille Cedex 20, (France)
† Electronic supplementary information (ESI) available: Experimental
procedures, and characterization data for new compounds. See DOI:
10.1039/c0ob00054j
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 4165–4168 | 4165
©

View Article Online
Table 1 Racemization of amine 2a upon irradiation for 2.5 h at 350 nm
in the presence of AIBN (24–30 ◦C)
Solventa
2a ee (%)
3-Me-3-pentanol
t-BuOH
t-BuOH/Toluene (2/1)
t-BuOH/THF (5/1)
38–76
53–92
22–28
50–57
a AIBN was added in 3 portions (20 or 30 mol% each) every 45 min to a
0.1 M solution of 2a.
Fig. 1 Structures of amines 2a–g and acyl donor 4.
The substantial increase in enantioselectivity gained by selecting
N-acylalanine trifluoroethyl esters as acyl donors10 led us to check
if these donors were racemized by the same process as the amines,
using trifluoroethane thiol. Indeed no racemization of N-octanoyl
alanine trifluoroethyl ester (4) was detected.14 According to G3B3-
MP2 calculations, going from ethyl to trifluoroethyl esters lowers
the C–H BDE at the captodative position by 5.7 kJ mol-1 for the N-
acetylglycine esters (354.5/348.8 kJ mol-1), and by 8.9 kJ mol-1 for
the N-acetylalanine esters (355.2/346.3 kJ mol-1). Going from N-
acetylalanine trifluoroethyl ester to the corresponding N-i-propyl
amide increases the C–H BDE by 14.6 kJ mol-1. The variations in
the series are significant. They suggest that the alanine derivatives
are not poorer hydrogen atom donors than the glycine derivatives.
In addition, in the case of alanine derived acyl donor (4) the N-i-
propyl amide is even more difficult to racemize than the ester. This
might explain why no trace of (S,R)-diastereomeric amide 5a was
detected in the reaction product.15
the rate of racemization when irradiating at 300 nm6c), happened
to be acceptable for racemizations performed under irradiation at
350 nm. In addition, the results show that methyl-b-cyclodextrin
R
could be as efficient or even superior to Brijꢀ 56 when associated
to a,b-D-glucopyranoside as coating agent. Good results were
obtained for KR performed in the presence of the enzyme
coated with a,b-D-glucopyranoside and methyl-b-cyclodextrin. A
significant increase in E factor was registered when replacing 3-
methyl-3-pentanol by t-butanol under these conditions (entries 1
to 3).
The use of a co-solvent appeared as a good compromise to
run concomitantly, an efficient KR and an efficient racemization
procedure.
Therefore, the chiral acyl donor 4 was tested in the DKR17
process. These tests were first achieved on amine 2a. The most
significant results are given in Table 3. High ees were observed in
entries 2–6. Although differences are tiny (all the combinations
mentioned below are nearly equivalent) the highest yield in amide
5a was recorded in entry 6, that is in 5/1 mixture of t-BuOH and
THF.
All conditions were gathered to generalize the (S)-selective real
DKR process to the series of aliphatic amines in Fig. 1. The results
are reported in Table 4. The main limitation of the methodology
KR resolution with this acyl donor was reinvestigated in
different solvent mixtures as reported in Table 2. This allowed
us to define which mixtures would lead to both acceptable
enantioselectivities and rates of reaction with amines 2a–g
(Fig. 1).
It must be underlined that t-butanol, that was once discarded
as compared to 3-methyl-3-pentanol (for the sake of improving
Table 2 Alkaline protease-catalyzed KR of amines 2a–g using 4 as the acyl donor
2
5
Amine
Solvent
t/h
ee (yield)c
de (yield)
C%d
Ee
1
2
3
4
5
6
7
8
2a
2a
2a
2b
2c
2d
2e
2f
2g
2a
2a
2a
3-Me-3-pentanola
t-BuOHa
1.5
2
1
1
1.5
1
1.5
1
3
0.75
2
1
86.5 (-)
>99.5 (-)
99 (48)
97 (46)
98 (44)
98 (47)
98 (40)
92 (35)
94 (28)
98 (-)
94.3 (-)
87 (-)
47.8
53.3
51.5
51.3
52.6
52.1
55.1
50.0
53.4
50.5
52.5
50.2
99
86
150
101
73
88
40
79
35
t-BuOHb
93 (50)
92 (45)
98 (47)
90 (52)
80 (52)
92 (50)
82 (48)
96 (-)
t-BuOHb
t-BuOHb
t-BuOHb
t-BuOHb
t-BuOHb
9
t-BuOHb
10
11
12
t-BuOH/toluene (2/1)a
t-BuOH/toluene (5/1)a
t-BuOH/THF (5/1)b
229
112
101
>99.5 (-)
94 (48)
90 (-)
93 (51)
a Enzyme coated with BrijꢀR 56 and a,b-D-glucopyranoside (8/1/1, w/w/w). b Enzyme coated with a,b-D-glucopyranoside and Me-b-cyclodextrin
(8/1/1, w/w/w). c Isolated after derivatization as a Boc-carbamate. d Calculated according to C = eeamine/(eeamine+ eeamide). e Enantioselectivity factors were
calculated according to E = ln[(1 - C)(1 - eeamine)]/ln[(1 - C)(1+eeamine)].16
4166 | Org. Biomol. Chem., 2010, 8, 4165–4168
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Table 3 Improvement of (S)-selective DKR of amine 2a
Experimental
Immobilization of alkaline protease
In a 100 mL flask, 60 mg of octyl a,b-D-glycopyranoside and 60 mg
of methyl-b-cyclodextrin were diluted at 60 ◦C in 50 mL of buffer
(Phosphate 0.1M, pH 7), this solution was then cooled to RT. The
alkaline protease liquid extract (500 mg) from Valley Research
was added to 25 mL of the previous solution. The mixture was
rapidly frozen in liquid N2 and lyophilized for 20 h to give 822 mg
of enzymatic powder.
Solvent
5a de (%)
Yield (%)
1
2
3
4
5
6
3-Me-3-pentanola
t-BuOHa
70
94
94
94
94
93
81
68
70
72
76
78
t-BuOHb
General procedure for the kinetic resolution of amines
t-BuOH/Tolueneb (2/1)
t-BuOH/Tolueneb (5/1)
t-BuOH/THF b (5/1)
To a solution of acyl donor (4) (104 mg, 0.35 mmol) in t-BuOH
(5 mL) was added coated alkaline protease (25 mg) and amine (2)
(0.5 mmol). The resulting mixture was stirred at room temperature
for 60 min. The amine ee was determined by GC after derivatiza-
tion of an aliquot of the crude mixture in trifluoroacetamide with
1.5 equiv of N-methyl-bis-trifluoroacetamide. The enzyme was
filtered off from the solution and washed with dichloromethane
(10 mL). Then Boc2O (66 mg, 0.3 mmol) was added to the filtrate,
and the mixture was stirred until complete consumption of the
amine (TLC monitoring). The solvent was then evaporated and
the crude mixture was purified on gel-silica (pentane–diethyl ether
0 to 10% to afford the Boc-amine, then dichloromethane–MeOH
0 to 1% to afford the amide).
a Enzyme coated with BrijꢀR 56 and a,b-D-glucopyranoside (8/1/1,
w/w/w). b Enzyme coated with a,b-D-glucopyranoside and Me-b-
cyclodextrin (8/1/1, w/w/w).
Table 4 (S)-Selective DKR of amines 2b–ga
Amine
Amide 5 de (%)
Yield (%)
2b
2c
2d
2e
2f
92
88
90
78
88
73
68
73
66
72
68
65
General Procedure for the dynamic kinetic resolution of amines
2g
To a solution of N-octanoyl-L-alanine trifluoroethyl ester (4)
(125 mg, 0.42 mmol) in t-BuOH/THF:5/1 (3 mL, 0.1 M) was
added coated alkaline protease (15 mg) and the amine (0.3 mmol).
The mixture was irradiated at 24–28 ◦C in a quartz tube (diameter
1.5 cm) in a Rayonet apparatus (RPR-200, 16 UV lamps Sylvania
Blacklight 351 F15W/T5/BL350) for 2 h 30 min and AIBN
(45 mg, 0.25 mmol) was added by portion every 45 min (3 ¥ 15 mg).
The enzyme was filtered off from the solution and washed with
dichloromethane (5 mL). The solvent was then evaporated and the
crude material was purified by flash column chromatography on
silica gel (dichloromethane–methanol 0 to 5%) to give pure amide.
a Experimental conditions identical to Table 3 entry 6
is inherent in the rate of the KR that must not be too slow as
compared to the rate of the radical racemization in order to limit
competitive radical degradation processes. The DKR of amines for
which the resolution process is too slow could not be achieved.18
Conclusions
Acknowledgements
Reversible hydrogen atom abstraction from the stereogenic center
adjacent to nitrogen atom by sulfanyl radical happened to be the
very first racemization procedure allowing to achieve either (R)-
or (S)-selective chemoenzymatic DKR of chiral aliphatic amines.
Moreover, it must be emphasized that no metal was involved in
the overall process. If (R)-selective DKR mediated with CAL-B
lipase was rather easily devised, the achievement of (S)-selective
DKR using alkaline protease was a real challenge. The stability of
alkaline protease required racemization to be induced at room
temperature photochemically(350 nm). This led us to change
parameters such as the solvent (t-butanol/THF mixture) and the
acyl donor (peptide mimetic 4). Under these new experimental
conditions, the synthesis of (S)-amides was achieved in good yield
and high enantiomeric excess, through the chemoenzymatic DKR
of a series of structurally related amines bearing various functional
groups. Further progress in view of making the process viable for
industrial applications will be reported in due course.
Gift of Alkaline protease from Valley Research is gratefully
acknowledged. This work was supported by the Agence Nationale
de la Recherche (ANR 06-Blan-0137).
References
1 (a) S. A. Lawrence, In Amines: Synthesis, Properties and Applications,
Cambridge University Press, Cambridge, U.K, 2004; (b) M. Breuer, K.
Ditrich, T. Habicher, B. Hauer, M. Keeler, R. Stu¨rmer and T. Zelinski,
Angew. Chem., Int. Ed., 2004, 43, 788; (c) G. Hieber and K. Ditrich,
Chim. Oggi, 2001, 19, 16.
2 For selected reviews dealing with the synthesis of chiral amines,
see:(a) T. C. Nugent and M. El-Shazly, Adv. Synth. Catal., 2010, 352,
753; (b) D. Ferraris, Tetrahedron, 2007, 63, 9581; (c) F. van Rantwijk
and R. A. Sheldon, Tetrahedron, 2004, 60, 501.
3 For general reviews, see:(a) K. Faber, Chem.–Eur. J., 2001, 7, 5004;
(b) H. Pellissier, Tetrahedron, 2003, 59, 8291; (c) H. Pellissier, Tetrahe-
dron, 2008, 64, 1563; (d) R. S. Ward, Tetrahedron: Asymmetry, 1995, 6,
1475.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 4165–4168 | 4167
©

View Article Online
4 For chemoenzymatic approach, see:(a) A. N. Parvulescu, P. A. Jacobs
and D. E. De Vos, Adv. Synth. Catal., 2008, 350, 113; (b) C.
Roengpithya, D. A. Patterson, A. G. Livingston, P. C. Taylor, J. L. Irwin
and M. R. Parrett, Chem. Commun., 2007, 3462; (c) A. N. Parvulescu,
P. A. Jacobs and D. E. De Vos, Chem.–Eur. J., 2007, 13, 2034; (d) J.
B. Crawford, R. T. Skerlj and G. J. Bridger, J. Org. Chem., 2007, 72,
669; (e) A. J. Blacker, M. J. Stirling and M. I. Page, Org. Process Res.
Dev., 2007, 11, 642; (f) M. A. J. Veld, K. Hult, A. R. A. Palmans and
E. W. Meijer, Eur. J. Org. Chem., 2007, 5416; (g) M.-J. Kim, W.-H.
Kim, K. Han, Y. K. Choi and J. Park, Org. Lett., 2007, 9, 1157; (h) M.
Stirling, J. Blackerb and M. I. Page, Tetrahedron Lett., 2007, 48, 1247;
(i) C. E. Hoben, L. Kanupp and J.-E. Ba¨ckvall, Tetrahedron Lett.,
2008, 49, 977; (j) J. Paetzold and J.-E. Ba¨ckvall, J. Am. Chem. Soc.,
2005, 127, 17620; (k) L. K. Tha`len, D. Zhao, J.-B. Sortais, J. Paetzold,
C. Hoben and J.-E. Ba¨ckvall, Chem.–Eur. J., 2009, 15, 3403; (l) A. N.
Parvulescu, P. A. Jacobs and D. E. De Vos, Appl. Catal., A, 2009, 368,
9; (m) L. H. Andrade, A. V. Silva and E. C. Pedrozo, Tetrahedron Lett.,
2009, 50, 4331; (n) M. T. Reetz and K. Schimossek, Chimia, 1996, 50,
668.
5 For preliminary reports on amines racemization, see:(a) S. Escoubet,
S. Gastaldi, N. Vanthuyne, G. Gil, D. Siri and M. P. Bertrand, J. Org.
Chem., 2006, 71, 7288; (b) S. Escoubet, S. Gastaldi, N. Vanthuyne,
G. Gil, D. Siri and M. P. Bertrand, Eur. J. Org. Chem., 2006,
3242.
6 (a) M. Nechab, N. Azzi, N. Vanthuyne, M. P. Bertrand, S. Gastaldi
and G. Gil, J. Org. Chem., 2007, 72, 6918; (b) S. Gastaldi, S. Escoubet,
N. Vanthuyne, G. Gil and M. P. Bertrand, Org. Lett., 2007, 9, 837;
(c) L. El Blidi, M. Nechab, N. Vanthuyne, S. Gastaldi, G. Gil and M.
P. Bertrand, J. Org. Chem., 2009, 74, 2901.
8 For a review, see: F. Bordera, Chem. Rev., 2002, 102, 4817 and refs cited
therein.
9 M. Nechab, L. El Blidi, N. Vanthuyne, S. Gastaldi, M. P. Bertrand and
G. Gil, Org. Biomol. Chem., 2008, 6, 3917.
10 A.-L. Bottalla, S. Queyroy, N. Azzi-Schue, N. Vanthuyne, S. Gastaldi,
M. P. Bertrand and G. Gil, Tetrahedron: Asymmetry, 2009, 20, 2823.
11 For examples of use of coated subtilisin see: (a) L. Bore´n, B. Mart´ın-
Matute, Y. Xu, A. Co´rdova and J.-E. Ba¨ckvall, Chem.–Eur. J., 2006,
12, 225; (b) M.-J. Kim, Y. I. Chung, Y. K. Choi, H. K. Li, D. Kim and
J. Park, J. Am. Chem. Soc., 2003, 125, 11494; (c) S. G. Martinez, E.
Alvira, L. Vergara Cordero, A. Ferrer, I. Montane´s-Clemente and G.
Barletta, Biotechnol. Prog., 2002, 18, 1462.
12 L. Routaboul, N. Vanthuyne, S. Gastaldi, G. Gil and M. P. Bertrand,
J. Org. Chem., 2008, 73, 364.
13 The rate slow-down was assigned to the solvation of the amine
by hydrogen bonding that would increase the a-CH BDE. For the
incidence of intramolecular hydrogen bonding on the a-CH BDE in
amines, see: (a) J. Laleve´e, X. Allonas and J.-P. Fouassier, J. Am. Chem.
Soc., 2002, 124, 9613; (b) See also: ref.5b.
14 Furthermore, 1H-NMR analysis in the presence of an internal standard
showed that no degradation occurred. The results of the DKR
experiments achieved with this acyl donor further confirmed that 5a was
not epimerized, since no trace of any diastereomer of 5a was detected.
15 It must be underlined however, that the S–H BDE in trifluoroethane
thiol is stronger than all C–H BDEs mentioned in the text. The
direct comparison of the calculated values for BDEs of different types
might be hazardous. C–H BDE in 4 : 346.3 kJ mol-1; C–H BDE in N-
acetylalanine N-i-propyl amide : 360.9 kJ mol-1; trifluoroethane thiol
S–H BDE: 365.6 kJ mol-1 (at 298 K according to G3B3MP2 method).
16 C. S. Chen, Y. Fujimoto, C. J. Girdaukas and C. J. Sih, J. Am. Chem.
Soc., 1982, 104, 7294.
7 (a) H. Kitaguchi, P. A. Fitzpatrick, J. E. Huber and A. M. Klibanov,
J. Am. Chem. Soc., 1989, 111, 3094; (b) R. J. Kazlauskas and A. N. E.
Weissfloch, J. Mol. Catal. B: Enzym., 1997, 3, 65; (c) C. K. Savile and
R. J. Kazlauskas, J. Am. Chem. Soc., 2005, 127, 2104; (d) P. F. Mugford,
U. G. Wagner, Y. Jiang, K. Faber and R. J. Kazlauskas, Angew. Chem.,
Int. Ed., 2008, 47, 8782.
17 Only the natural S-isomer is recognized by the enzyme, therefore the
racemic acyl donor can be used without changing the result but the
quantity has to be doubled.
18 This happened to be the case for 1-phenyl-2-amino-propane.
4168 | Org. Biomol. Chem., 2010, 8, 4165–4168
This journal is
The Royal Society of Chemistry 2010
©