A. Tomin et al. / Process Biochemistry 45 (2010) 859–865
861
Table 1
2.5. Kinetic resolution of rac-1b, rac-1d and rac-3 in continuous-flow bioreactors
Screening the enzymes for kinetic resolution of racemic cycloalkanols (rac-1a–e
and rac-3) by acylation with vinyl acetate in a shake flask.
Screening the operation conditions for the continuous-flow mode reactions: The
solution of racemic alcohol (10 mg mlÀ1; 0.090 M for rac-1b, 0.077 M for rac-1d,
0.057 M for rac-3) in a hexane-THF-vinyl acetate 2:1:1 mixture was pumped through
the lipase-filled CatCartTM column (lipases: CaLB for rac-1b and rac-3; sol–gel LAK for
rac-1b). Runs were performed under various conditions [temperature (20–60 8C) and
flow rate (0.1–0.3 ml minÀ1) were varied]. Samples were collected and analyzed
during stationary operation (GC every 10 min between 30 min and 90 min after
changing the conditions). The results of these experiments are listed in Table 3.
Preparative kinetic resolutions in continuous-flow mode: Solutions of rac-1b, rac-1d
and rac-3 were treated similarly as described for the previous screen, but the lipase-
filled columns (lipases: CaLB for rac-1b and rac-3; sol–gel LAK for rac-1d) were
operated at50 8Cand theproductswerecollected ataflowrateof0.2 ml minÀ1 for3 h.
The solvent was removed from the resulting reaction mixture (36 ml) under vacuum.
Theresidue containingthe mixtureoftheformedacetate[(+)-(R)-2b, (+)-(S)-2dor(+)-
(R,R)-4]and unreactedalcohol[(À)-(S)-1b, (À)-(R)-1dor(À)-(S,S)-3]wasseparatedby
column chromatography (silica gel, 5–40% gradient ofEtOAcin hexane). The results of
these experiments are shown in Panel B of Table 4.
No.
Substratea
Enzymea
c (%)
eeb (%)
Ec
1
2
3
4
5
6
rac-1a
rac-1a
rac-1a
rac-1a
rac-1a
rac-1a
BUTE 3b
Lipase PS
Lipase AK
CaLB
50
47
53
54
<5
40
96.0
87.2
99.7
>100
ꢀ200
>100
99.7
d
75
d
Lipozyme TL IM
PLE
54.5
18
7
8
rac-1b
rac-1b
rac-1b
rac-1b
rac-1b
rac-1b
BUTE 3b
Lipase PS
Lipase AK
CaLB
51
52
53
51
42
<5
99.7
99.8
97.1
94.2
>200
>200
61
9
10
11
12
78
Lipozyme TL IM
PLE
64.8
d
30
d
13
14
15
16
17
18
rac-1c
rac-1c
rac-1c
rac-1c
rac-1c
rac-1c
BUTE 3b
Lipase PS
Lipase AK
CaLB
30
39
56
25
16
10
30.6
55.5
99.8
23.8
13.1
4.0
7.6
22
55
3. Results and discussion
7.9
Lipozyme TL IM
PLE
6.4
2.1
In this study our major aim was to investigate the lipase-
catalyzed kinetic resolutions of novel cyclic secondary alcohols in
batch and continuous-flow reactions. Because the enantiomer
selectivity of the lipase-catalyzed kinetic resolutions proved to not
be sensitive to the mode of reaction [24], batch mode reactions
were used to select the most selective processes as the best
candidates for comparison of the two reaction modes.
19
20
21
22
23
24
25
rac-1d
rac-1d
rac-1d
rac-1d
rac-1d
rac-1d
rac-1d
BUTE 3b
Lipase PS
Lipase AK
Sol–gel LAK
CaLB
39
52
58.8
97.2
99.0
50
>200
55
48
48
89.0
d
>200
d
>98
50
Lipozyme TL IM
PPL
98.7
40.8
>200
>200
31
26
27
28
29
30
31
32
rac-1e
rac-1e
rac-1e
rac-1e
rac-1e
rac-1e
rac-1e
BUTE 3b
Lipase PS
Lipase AK
Sol–gel LAK
CaLB
35
<5
5
12.0
1.8
3.1. Selecting the substrates for comparative studies
d
d
42.0
d
1.7
d
First, racemic cycloalkanols rac-1a–e and rac-3 were synthesized
and preliminary enzyme screening for the kinetic resolutions of
these racemic substrates in batch mode was performed (Fig. 1). The
effect of the size and nature of the ring on the enantiomer selectivity
was investigated using three different racemic C2-methylene
substituted cycloalkan-1-ols (2-methylenecyclopentan-1-ol, 2-
methylenecyclohexan-1-ol and 2-methylenecycloheptan-1-ol; rac-
1a–c, respectively) and two methylene-dioxepanols (6-methylene-
[1,3]dioxepan-5-ol and 2,2-dimethyl-6-methylene-[1,3]dioxepan-
5-ol; rac-1d,e, respectively). For comparison, a conformationally
more flexible racemic cycloalkan-1-ol, the trans-2-bromocyclo-
hexan-1-ol rac-3 with a bromine substituent of comparable size to
the methylene group was also investigated (Fig. 1).
<5
64
<5
5
20.5
d
1.6
d
Lipozyme TL IM
PPL
1.3
1.9
33
34
35
36
37
38
rac-3
rac-3
rac-3
rac-3
rac-3
rac-3
BUTE 3b
Lipase PS
Lipase AK
CaLB
51
12
49
55
9
99.2
12.8
86.9
98.8
9.3
>100
87
64
51
20
Lipozyme TL IM
PLE
29
12.7
2.2
a
The reactions (for experimental details see Section 2.3.) were analyzed at various
times (rac-1a: 4 h, rac-1b: 8h, rac-1c: 8h, rac-1d: 24h, rac-1e: 24h, rac-3: 24h).
b
The ee(S)-1a–c, ee(R)-1d,e and ee(S,S)-3 values were determined by GC (see
Supplementary Information).
c
The enantiomer selectivity (E) was calculated from c and ee(S)-1a–c/ee(S,S)-3 [31]
The enzyme-catalyzed kinetic resolutions of the racemic
secondary alcohols rac-1a–e and rac-3 were screened with
different lipases (Table 1) using vinyl acetate as the acyl donor
in hexane-THF solvent [28,30] in batch mode (Fig. 1). This screen
identified a number of highly enantiomer-selective enzymes
(Table 1 [31,32]). Similar to earlier studies [13,30], the best
selectivities were observed with lipases from the Pseudomonas
strains (Lipase AK and Lipase PS), lipase B from C. antarctica and
or from ee(R)-1d,e and ee(S)-2d,e [32]. Due to sensitivity to experimental errors, E
values calculated in the 100–200 range are reported as >100, values in the 200–500
range are reported as >200 and values calculated above 500 are given as ꢀ200.
d
E and ee values were not determined due to the inappropriate conversion (<5%
or >98%).
2.3. Screening the enzymes for kinetic resolution of racemic cycloalkanols (rac-1a–e
and rac-3) in batch mode
lipase BUTE 3b.
a-Chymotrypsin and papain were inactive in the
Enzyme (10 mg) was added to a solution of racemic alcohols (rac-1a–e and rac-3,
20 mg each) in hexane – THF 2:1 (1 ml) and vinyl acetate (100
ml), and the mixture
acylation of rac-1a–e and rac-3. Because the benzyl-protected rac-
1e was not accepted as a substrate by the enzymes tested in this
study, this racemic alcohol was not investigated further.
was shaken in a sealed glass vial at 1000 rpm at room temperature. Data on the
conversion and enantiomeric composition of the products [(R)-2b and (R,R)-4 with
CaLB; (S)-2d with sol–gel LAK] from various enzymes for different reaction times
are presented in Table 1.
The absolute configuration of the slower reacting six-membered
ring alcohol (+)-1b was determined to be (S)- by comparing the
optical rotation of (+)-1b, isolated from the kinetic resolution with
CaLB, to the reported (+)-optical rotation of (S)-1b [10,33]. The
absolute configuration of the enantiopure esters of rac-1a and rac-1c
were assigned as (R)- on the basis of Kaslauskas rule [34]. Due to the
reversaloftheligandhierarchy,thesamegeometryshouldbelabeled
as (S)- for the two different methylene-dioxepanols 1d and 1e.
In accordance with the previous results from the other 2-
substituted-cycloalkan-1-ols [13,16], most of the hydrolases
exhibited the highest enantiomer selectivity in kinetic resolution
of the five- and six-membered ring substrates, rac-1a and rac-1b
2.4. Lipase-catalyzed kinetic resolution of rac-1a–d and rac-3 in batch mode
Enzyme (CaLB, 300 mg, for rac-1a–c and rac-3; sol–gel LAK, 400 mg, for rac-1d)
was added to a solution of racemic alcohol (rac-1a–d or rac-3, 300 mg each) in
hexane-THF 2:1 (30 ml) and vinyl acetate (3 ml), and the mixture was stirred at
room temperature (25 8C). After the proper reaction time (rac-1a: 2 h; rac-1b: 6 h;
rac-1c: 16 h; rac-1d: 24 h; rac-3: 8 h) the enzyme was filtered off and the filtrate
was concentrated under vacuum. The residue containing a mixture of the formed
acetate [(+)-(R)-2a–c, (+)-(S)-2d or (+)-(R,R)-4] and unreacted alcohol [(À)-(S)-1a–c,
(À)-(R)-1d or (À)-(S,S)-3] was separated by column chromatography (silica gel, 5–
40% gradient of EtOAc in hexane). The results of the batch mode kinetic resolutions
are shown in Panel A of Table 4.