Chemoenzymatic synthesis of both enantiomers of 2-tert-butyl-2-methyl-1,3- benzodioxole-4-carboxylic (TBMB) acid

Article abstract of DOI:10.1002/adsc.201000446

Both enantiomers of 2-tert-butyl-2-methyl-1,3-benzodioxole-4-carboxylic (TBMB) acid, which has a unique quaternary chiral center located on the acetal carbon, were prepared from a racemic ketone in which the carboxy group was replaced by a trifluoroacetyl

Full text of DOI:10.1002/adsc.201000446

FULL PAPERS
DOI: 10.1002/adsc.201000446
Chemoenzymatic Synthesis of Both Enantiomers of 2-tert-Butyl-
2-methyl-1,3-benzodioxole-4-carboxylic (TBMB) Acid
Toshinori Higashi,a,* Chika Abe,a Keiko Ninomiya,a Takuya Machida,a
Nobuyuki Chishima,a Shohei Taketomi,a Miyu Furuta,a Yoko Komaki,a
Yoshihiko Senba,a Taeko Tokuda,a Mitsuru Shoji,a and Takeshi Sugaia,*
a
Faculty of Pharmacy, Keio University, 1-5-30, Shibakoen, Minato-ku, Tokyo 105-8512, Japan
Fax : (+81)-3-5400-2665; phone: (+81)-3-5400-2665; e-mail: higashi-ts@pha.keio.ac.jp or sugai-tk@pha.keio.ac.jp
Received: June 8, 2010; Revised: August 4, 2010; Published online: September 24, 2010
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.201000446.
Abstract: Both enantiomers of 2-tert-butyl-2-methyl- even though it was six atoms removed from the ester
1,3-benzodioxole-4-carboxylic (TBMB) acid, which carbonyl carbon. Although the two substrates had
has a unique quaternary chiral center located on the the same absolute configuration at the secondary al-
acetal carbon, were prepared from a racemic ketone cohol, the reaction rate of one stereoisomer was 72
in which the carboxy group was replaced by a tri- times greater than that of the other isomer. The
fluoroacetyl group. First, reduction with fungi, Geo- reason for this differential reactivity was attributed
trichum candidum provided (2S,1’S)- and (2R,1’S)- mainly to a large difference in Vmax(app) between the
1’-(2-tert-butyl-2-methyl-1,3-benzodioxol-4-yl)-2’,2’,2’- stereoisomers. The products, acetate and alcohol,
trifluoroethanol in a highly enantiofacially selective were easily separated by chromatography, and each
manner. After acetylation, the resulting diastereo- was then derivatized to (R)- and (S)-TBMB acid,
meric mixture was submitted to Candida antarctica with >99.2% ee, respectively.
lipase B-catalyzed transesterification. The reaction
proceeded in a stereoselective manner under the in- Keywords: asymmetric synthesis; enzyme catalysis;
fluence of the chiral center at the acetal carbon, kinetic resolution; transesterification
Introduction
the derivation and analytical processes. Application of
this reagent in HPLC analysis is also promising by
virtue of its strong fluorescence.
2-tert-Butyl-2-methyl-1,3-benzodioxole-4-carboxylic
(TBMB) acid (1a, Figure 1) has been developed by
So far, however, a shortage of supply of enantio-
Ohrui and Nishida, as a unique chiral derivatizing merically pure 1a has been a drawback. Due to the lo-
agent for analyzing biomolecules.[1] Simple singlet cation of the chirality at the quaternary acetal posi-
1H NMR signals of the acetal substituents are very tion, direct construction by means of asymmetric syn-
useful for analyzing the stereoisomeric purities and thesis is difficult. To date, only preferential crystalliza-
absolute configurations of amino acids and mono- or tion of the naturally occurring cinchonidine salt of
diglycerides in lipid metabolites with chiral centers. In racemic 1a has been used to produce one of the two
TBMB, the unique chirality is located at a quaternary enantiomers, (S)-1a. No practical method for the
acetal carbon, and there is no chance for racemization preparation of (R)-1a has yet been developed because
under physiological to alkaline conditions throughout no chiral amine has been found that will produce a
salt suitable for preferential crystallization. In this
paper, we describe chemoenzymatic approaches
which enable the preparation of both enantiomers of
1a.
Figure 1. Both enantiomers of TBMB acid 1a.
Adv. Synth. Catal. 2010, 352, 2549 – 2558
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2549

Toshinori Higashi et al.
FULL PAPERS
Results and Discussion
Approaches to the Direct Kinetic Resolution
Our first attempt to produce both isomers of TBMB
acid was based on kinetic resolution via enzyme-cata-
lyzed hydrolysis of the racemic mixture of methyl
ester 1b as shown in Scheme 1.
Among several hydrolytic enzymes, Candida rugosa
lipase (Meito OF) preferentially hydrolyzed (R)-1b
with an enantiomeric ratio (E value)[2] of 20, and the
selectivity was higher than that of Candida antarctica
lipase (Novozym 435, E=11). The hydrolysis, howev-
er, was very slow in both cases, probably due to steric
hindrance from the adjacent quaternary acetal. The
minimum extent of conversion required for practical
separation of the enantiomers was never achieved,
even at a raised reaction temperature (508C). To en-
hance the enzyme activity of the ester, particularly of
the “fast” (R)-isomer, the methyl substituent was re-
placed with the more electron-withdrawing chloroeth-
yl ester group.[3]
Scheme 2. Synthetic plan – indirect resolution.
The reaction with substrate 1c proceeded at lower The trifluoromethyl ketone 2 was selected as a precur-
À
temperature (308C), indicating that the activity was sor, because the CF3 CO bond is readily cleavable
indeed enhanced; however, the selectivities were dis- under basic conditions. This compound has the added
appointing. All of the hydrolytic enzymes (Table 1, advantage that an easily accessible chiral center
entries 3–5) showed low enantioselectivities. This would be introduced, for example, by a simple deriva-
result suggests that the reaction rate for both enantio- tion to the secondary alcohol 3a. The control of the
mers of 1c was enhanced for C. rugosa and C. antarc- stereochemistry of this newly created chiral center
tica lipases, leading to a decrease in the difference in would provide two indirect ways to resolve the stereo-
reaction rate between the two enantiomers.
isomers. If the two isomers turned out to be enantio-
mers, then an enantioselective treatment would be
successful (Scheme 2, path A). In contrast, if the two
isomers existed as diastereomers, chromatographic
separation techniques would be available (path B).
Toward that end, acetal 5 was ortho-metallated
Approaches for Indirect Resolution
We then turned our attention to “indirect resolution”
of the quaternary chiral center as shown in Scheme 2. using n-BuLi/TMEDA.[4] The resulting anion was very
stable, but the subsequent trifluoroacetylation step[5]
was highly exothermic and thus required careful con-
trol of the reaction temperature below À408C via
slow addition of ethyl trifluoroacetate as the electro-
phile. The desired ketone 2 was obtained in 88%
yield and the reaction was easily scaled-up to a 30-
gram level in a highly reproducible manner. Next, the
reduction of 2 with NaBH4 in EtOH provided the al-
cohol 3a as a set of diastereomers as shown in
Scheme 1. Attempts for enzyme-catalyzed resolution of rac-
emic esters.
Table 1. Hydrolase-catalyzed hydrolysis of (Æ)-1b, c.[a]
Entry
Substrate
Enzyme
Conversion [%]
ee (S) [%]
E value[2]
1
2
3
4
5
1b
1b
1c
1c
1c
C. antarctica lipase B (Novozym 435)
C. rugosa lipase (Meito, OF)
C. antarctica lipase
C. rugosa lipase
pig liver esterase (Sigma, E2884)
21
8
25
11
38
21
8
20
0
11
20
5
1
1
0
[a]
For reaction conditions, see Experimental Section.
2550
asc.wiley-vch.de
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 2549 – 2558

Chemoenzymatic Synthesis of Both Enantiomers
Scheme 3. Synthesis of four diastereomers of 3a.
Scheme 3. TLC analysis clearly indicated two spots
(Rf =0.50 and 0.47). The relative stereochemistry was
temporarily assigned to be (2R*,1’R*)-for the less
polar isomer and (2R*,1’S*)-for the more polar
isomer. Confirmation of this assignment is described
below.
First, less polar (2R*,1’R*)-3a was acetylated to
provide a racemic acetate 3b. As this product is an
equimolar mixture of (2S,1’S)-3b and (2R,1’R)-3b in
which the relative stereochemistry is perfectly con-
trolled, an enzyme-catalyzed enantioselective transes-
terification would enable an “indirect resolution” of
the inaccessible remote quaternary center, as suggest-
ed in Scheme 2, path A. Literature reports indicated
Scheme 4. Kinetic resolution of trifluoromethyl acetates.
that trifluoroethoxy esters with a substituent aromatic for acetate 3b were more advantageous compared to
ring on the secondary position are very good candi- hydrolysis with the same lipase, in which the pro-
dates for lipase-catalyzed enantioselective hydrolysis longed reaction lowered the ee of (2S,1’S)-3a. In an-
and transesterification.[6]
other attempt, the acylation of the racemic alcohol 3a
The acetal 3b was thus treated with C. antarctica with vinyl acetate, the reaction was very slow due to
lipase B and cyclopentanol at 308C under transesteri- the electron-withdrawing effect of the trifluoromethyl
fication conditions, to give an alcohol 3a (>99.9% ee) group.
and an unreacted recovery 3b (>99.9% ee) as shown
The absolute stereochemistry at the secondary alco-
in Scheme 4. Compared with primary alcohols, the hol of the “fast” isomer was determined to be (S)
use of a secondary alcohol such as cyclopentanol is by 1H NMR analysis after the derivatization to
reasonable considering its higher selectivity.[7] The by- a-methoxy-a-trifluoromethylphenylacetate (MTPA)
product accompanied with the progress of the transes- esters[8] (see Supporting Information). The stereo-
terification is the secondary alkoxy acetate (cyclopen- chemistry of the preferred enantiomer was in good
tyl acetate in Scheme 4), which acts as a poor acyl accordance with previously reported, similar trifluoro-
donor. Primary alkoxy acetates suffer from reverse methyl-substituted substrates. The stereochemistry of
transesterification with “fast” alcohols and interfere the acetal chiral center was determined as follows:
with the complete consumption of the fast isomer, (1’S)-alcohol 3a was oxidized with 2-iodoxybenzoic
À
which is indispensable for effective kinetic resolution. acid (IBX), and C C bond subsequently cleaved
In our case, the transesterification proceeded in a under aqueous basic conditions to yield the acid 1a
highly enantioselective manner (E>200 at 50% con- with a positive rotation. In this way, the absolute con-
version). Moreover, the transesterification conditions figuration of the acetal chiral center in the “fast”
Adv. Synth. Catal. 2010, 352, 2549 – 2558
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2551

Toshinori Higashi et al.
FULL PAPERS
enantiomer was determined to be (S), and thus the lective reduction of (Æ)-2 is used, all of the alcohols
stereochemistry for the less polar diastereomer in are resolvable. Our first attempt involved such a dia-
Scheme 3 and Scheme 4 was unambiguously fixed.
stereoselective reduction of 2 with a variety of hy-
We were surprised that in the attempted transes- dride reducing agents (Scheme 6, case A). The results
terification of (2R,1’S)- and (2S,1’R)-3b, which were were not promising, as in most cases such as DIBAL-
derived from the more polar diastereomer of 3a as H formation of the undesired diastereomer [(2R,1’S)
shown in Scheme 3, the reactivity was quite low (9% and (2S,1’R)] was preferred based on the TLC analy-
conversion) although the enantioselectivity was high sis of the crude mixture.
(E=100). Even at a reaction temperature as high as
608C, both transesterification under the above-men-
tioned conditions and hydrolysis in a buffer solution Enantiofacially Selective Reduction by
were very slow (Scheme 5). Several commercially Microorganisms
available lipases such as Burkholderia cepacia lipase
(Amano PS-IM) also exhibited very poor activity An alternative solution involved the kinetic resolution
against this pair of enantiomers. It was concluded that of another pair of fast (2S,1’S)- and slow (2R,1’S)-iso-
there was only one “fast” isomer [(2S,1’S)-], and the mers (Scheme 6, case B). Thus we tried an enantiofa-
other three, (2R,1’R)-, (2R,1’S)- and (2S,1’R)-, were cially selective reduction of (Æ)-2, expecting the for-
slow isomers.
mation of (2S,1’S) and (2R,1’S)-3a by applying such
Observation of the lipase-catalyzed reactions sug- modified borohydride, using cobaltACHTNUGTRNE(NUG III) ketoiminato
gested two solutions for stereochemical separation complex as asymmetric catalysts.[9] In this particular
with regard to the chirality located at the acetal case, however, probably due to the very high steric
carbon in Scheme 6, cases A and B. As already shown hindrance and extremely electron-deficient nature of
in Scheme 4, if only a mixture of (2S,1’S) and the trifluoromethyl ketone, the reduction was very
(2R,1’R)-isomers (case A) produced via a diastereose- slow and all four stereoisomers were obtained.
At this stage, we turned our attention to a whole-
cell microorganism-catalyzed reduction of trifluoro-
methyl ketone 2. From the screening of yeasts and
fungi, Geotrichum candidum NBRC 5767[10] provided
us with delightful results. The reduction proceeded in
a highly stereoselective manner to give less polar
(2S,1’S)-3a (45%, 98.8% ee) and more polar (2R,1’S)-
3a (45%, 96.1% ee) (Scheme 7). The diastereomeric,
non-antipodal relationship between the two products
was proven by NMR and HPLC analyses, and the ste-
reochemistry of the secondary alcohol was deter-
mined to be (S) in both products. The reduction of
the carbonyl group proceeded under the “anti-Prelog”
rule, a result that is somewhat controversial in previ-
ous reports.[10] Recent studies have revealed that mul-
tiple carbonyl reductases with different selectivities
Scheme 5. Low activity of (2R*,1’S*)-3b against lipase-cata-
lyzed transesterification.
Scheme 6. Diastereoselective and enantiofacially selective
reduction of (Æ)-2.
Scheme 7. Enantiofacially selective reduction of 2 with the
whole-cell microorganism.
2552
asc.wiley-vch.de
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 2549 – 2558

Chemoenzymatic Synthesis of Both Enantiomers
Table 2. Kinetic properties for (2S,1’S)- and (2R,1’S)-3b.
are expressed in cultured cells of G. candidum NBRC
5767, and one of these enzymes works predominantly
depending upon the reaction conditions. The recov-
ered, unreacted ketone 2 (1%) from the above reac-
tion was racemic, thus indicating that the reduction
proceeded in a non-enantioselective manner with
regard to the pre-existing chirality in the substrate.
The present reaction conditions were optimized,
and the use of glucose (5%) as a carbon source at the
Substrate
G
ACHTUNGTRENUN(NG 2R,1’S)-3b
Vmax(app) [mmol/h]
Km [M]
9.87
0.97
0.39
2.39
pre-incubation stage and for cofactor-recycling in the the “fast” (2S,1’S)-isomer, with a ratio of 0.41:1 com-
reduction step increased the yield to 91%, even at pared to the “slow” (2R,1’S)-isomer. The ratio of
1% (w/v) of the substrate 2 on a gram scale. Glucose Vmax(app)/Km between the diastereomers was estimated
was found to be more effective than glycerol or sec- to be 64:1, and was in good agreement with the iso-
ondary alcohols such as isopropyl alcohol, which have meric ratio (72:1) mentioned before.
also been reportedly used for this purpose.[11] The re-
Given that previous reports indicated that the (S)-
duction was also successfully scaled up in an open air configured trifluoromethyl-substituted secondary al-
tank.
cohols and their corresponding acyl derivatives are
always good substrates, the poor affinity of the pres-
ent (2R,1’S)-isomer was quite strange. The three-di-
mensional structure of C. antarctica lipase B has been
extensively studied, especially for the catalytic site
Kinetic Resolution on Remote Chirality
With the mixture of (2S,1’S)- and (2R,1’S)-3b in hand, (Ser105, His224, Asp187) and the cavity where a
the transesterifcation reaction was then attempted. large substituent in the substrate such as an aromatic
When the mixture was incubated with C. antarctica ring can be accepted.[13] The “fast” (2S,1’S)-isomer
lipase B under transesterification conditions, the reac- really fits very well with the enzyme. In contrast, in
tion readily proceeded in a diastereoselective manner. all attempts to apply the “slow” (2R,1’S)-isomer to
The alcohol (2S,1’S)-3a (41%, 93.2% de) and un- either the ground state or the tetrahedral transition
changed (2R,1’S)-3b (55%, 81.6% de), were obtained state,[14] the bulky substituent on the benzodioxole
as shown in Scheme 8. The “stereoisomeric ratio”, ring suffered from steric hindrance due to the a-helix,
which was calculated in a manner similar to that used which is located adjacent to the cavity, and particular-
for determining the E value in enantiomeric kinetic ly by Ala282 (see Supporting Information). This
resolutions, reached as high as 72, at 47% conversion. mode of recognition suggests the possibility for the
The recognition of such a remote chiral center was substitution of the bulky tert-butyl group with a b-
of significant interest. Only a few examples have been naphthyl group for recognition of remote chirality,
reported to date where the reaction rates were influ- which is related to the analogous chiral derivatizing
enced by the stereochemistry of a remote chiral agent, 2-methyl-2-b-naphthyl-1,3-benzodioxole-4-car-
center beyond the aromatic ring.[12] Therefore, boxylic (MNB) acid.[15]
enzyme kinetic studies on the transesterification of
pure (2S,1’S)-3b and (2R,1’S)-3b were performed in-
dependently. The results indicated a large difference Synthesis of (S)-TBMB Acid
in Vmax(app) between the two diastereomers in a ratio
of ca. 25:1 (Table 2). Km was also advantageous for The removal of trace amounts of contaminating enan-
tiomers, which is indispensable for chiral derivatizing
reagents, was the final task, and lipase-catalyzed reac-
tions were again the key for solving this problem. A
mixture of (2S,1’S)- and (2R,1’S)-3a (96.6:3.4) was
acetylated and the resulting mixture was re-submitted
to C. antarctica lipase B-catalyzed transesterification
(Scheme 9). This second resolution successfully pro-
vided enantiomerically and diastereomerically pure
(2S,1’S)-3a (60%), leaving the (2R,1’S)-isomer as the
acetate.
Pure (2S,1’S)-3a was oxidized with IBX and the
À
CO CF3 bond subsequently cleaved by alkaline hy-
drolysis to give (S)-TBMB acid 1a (87%) with an ee
of >99.9% as determined by HPLC analysis after
conversion to 1b.
Scheme 8. Recognition of remote stereochemistry in the
lipase-catalyzed transesterification.
Adv. Synth. Catal. 2010, 352, 2549 – 2558
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2553

Toshinori Higashi et al.
FULL PAPERS
Scheme 9. Repetition of resolution for fast isomer (2S,1’S)-
3b, and derivatization to (S)-1a.
Synthesis of (R)-TBMB Acid
In turn, the mixture containing mainly the (2R,1’S)-
isomer was treated again with C. antarctica lipase B
to remove the contaminating “fast” (2S,1’S)-isomer.
The “slow” (2R,1’S)-isomer, however, still contained
two intractable (1’R)- stereoisomers which were inert
to C. antarctica lipase-B catalyzed reactions.
Scheme 10. Resolution of chloroacetate (2R,1’S)-3c and the
derivatization to (R)-1a.
The solution to this problem involved introduction
of an electron-withdrawing chloroacetate group and
catalysis with a different lipase. Although B. cepacia
lipase was practically inert to any other stereoisomers
of the acetate 3b as previously mentioned, the com-
bined use of B. cepacia lipase and the chloroacetate
group provided a highly pure sample. The reaction
rate of (2R,1’S)-3c was dramatically enhanced among
those of three slow isomers. After chromatographic
purification, (2R,1’S)-3a was obtained in diastereo-
merically pure form. Subsequently, in a manner simi-
lar to that used for the (S)-isomer, (R)-TBMB acid 1a
À
(73%, 99.2% ee) was obtained by oxidation and C C
bond cleavage (Scheme 10).
Conclusions
An indirect resolution of the quaternary chiral center
in TBMB acid was achieved by the introduction of a
trifluoromethyl-substituted secondary alcohol moiety.
The tandem use of microbial enantiofacially selective
reduction in an anti-Prelog fashion followed by lipase-
catalyzed resolution of the remote chiral center, sum-
marized in Scheme 11, was effectively demonstrated
to provide pure enantiomers of TBMB acid (1a) on a
preparative scale. With the development of this prac-
tical synthesis of both enantiomers, TBMB acid may
now be used not only for analysis, but also as a chiral
auxiliary in asymmetric syntheses.
Scheme 11. Summary – tandem use of enantiofacially selec-
tive reduction and kinetic resolution of remote chiral center
toward both enantiomers of TBMB acid.
Experimental Section
Characterization data of all of the compounds including
spectroscopic and analytical data, general remarks, some ex-
2554
asc.wiley-vch.de
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 2549 – 2558

Chemoenzymatic Synthesis of Both Enantiomers
perimental details and characterization data and docking
studies are available as Supporting Information.
column chromatography (10 g). Elution with hexane-EtOAc
(5:1) afforded 1c as a pale yellow oil; yield: 450 mg (87%).
Methyl 2-tert-Butyl-2-methyl-1,3-benzodioxole-4-
carboxylate (1b)
Enzyme-Catalyzed Hydrolysis of (Æ)-1c
To a mixture of 1c (10.2 mg, 0.034 mmol) and phosphate
buffer solution (0.1M, 0.5 mL, pH 7.0) was added C. antarc-
tica lipase B (20.2 mg) and the resultant mixture was stirred
for 24 h at 308C. After acidification, the mixture was filtered
and extracted with EtOAc. The extract was washed with
water and brine, dried over Na2SO4, and concentrated under
To a stirred solution of 2 (19.5 mg, 0.068 mmol) in DMF
(140 mL) was added aqueous NaOH solution (8M, 26 mL,
0.21 mmol) and the mixture was stirred for 3 h at room tem-
perature. Then, MeI (18 mL, 0.28 mmol) was added and the
mixture was stirred at room temperature. After 3 h, the re-
action was quenched by the addition of water. The mixture
was extracted with EtOAc. The extract was washed with
brine, dried over Na2SO4, and concentrated under vacuum.
The residue was purified by preparative TLC [developed
with hexane-EtOAc (4:1)] to afford 1b as colorless solid;
yield: 16.8 mg (99%); mp 77–788C.
1
vacuum. The conversion was determined by H NMR analy-
sis of crude mixture. The residue was purified by preparative
TLC [developed with hexane-EtOAc (4:1)] to afford 1c
(yield: 7.0 mg, 69%), which was analyzed by HPLC [Chiral-
cel OJ, 0.46 cmꢁ25 cm; hexane-isopropyl alcohol (20:1);
flow rate 0.5 mLminÀ1; detected at 320 nm]: tR (min)=13.3
[(S)-, 59.8%], 16.8 [(R)-, 40.2%]. Results with other en-
zymes are listed in Table 1.
Enzyme-Catalyzed Hydrolysis of (Æ)-1b
As the representative procedure, the hydrolysis with C. ant-
arctica lipase B (Novozym 435) is described. To a mixture of
1b (20.0 mg, 0.080 mmol), phosphate buffer solution (0.1M,
0.8 mL, pH 7.0), and n-decane (80 mL) was added C. antarc-
tica lipase B (40.0 mg), and the mixture was stirred at 308C
for 25 h and then heated to 508C for 48 h. After acidifica-
tion to pH 3.5 with citric acid aqueous solution (1M), the
mixture was filtered with a pad of Celite and the filtrate was
extracted with EtOAc. The extract was washed with water
and brine, dried over Na2SO4, and concentrated under
vacuum. The residue was purified by silica gel column chro-
matography (1 g). Elution with hexane-Et2O (4:1) firstly af-
forded 1b (yield: 11.6 mg, 58%), which was analyzed by
HPLC [column, Daicel Chiralcel OJ, 0.46 cmꢁ25 cm;
hexane-isopropyl alcohol (24:1); flow rate 0.5 mLminÀ1; de-
tected at 320 nm]: tR (min)=10.4 [(S)-, 60.4%], 13.1 [(R)-,
39.6%].
2-tert-Butyl-2-methyl-1,3-benzodioxole (5)
To a solution of catechol (28.5 g, 259 mmol) in toluene
(270 mL) was added tert-butyl methyl ketone (54.2 g,
528 mmol) and a catalytic amount of p-TsOH monohydrate
(660 mg, 3.47 mmol). The mixture was refluxed for 24 h with
azeotropic removal of water. After being cooled to room
temperature, the reaction was neutralized by the addition of
aqueous NaOH solution (2M). The mixture was extracted
with Et2O. The extract was washed with water and brine,
dried over Na2SO4, and concentrated under vacuum. The
residue was distilled under reduced pressure to give 5 which,
upon standing at room temperature, solidified as colorless
needles; yield: 35.6 g (71%); mp 29–308C; bp 85–908C/
11 Torr. Its IR and NMR spectra were identical with those
reported previously.[16]
Further elution gave the hydrolyzate 1a (yield: 2.3 mg,
12%). To a solution of this sample in MeOH (100 mL) was
added TMSCHN2 (2.0M in hexane, 25 mL, 0.050 mmol). The
mixture was stirred at room temperature for 10 min. After
confirming the disappearance of the starting material by
TLC, the mixture was concentrated under vacuum to give
1b. This was analyzed by HPLC: tR (min)=10.4 [(S)-,
10.2%], 13.1 [(R)-, 89.8%]. The result obtained by applying
C. rugosa lipase (Meito, OF) is shown in Table 1. No hydrol-
ysis was observed by applying either K. oxytoca esterase
(Nagase, SNSM-87) or B. cepacia lipases (Amano PS or Su-
mitomo SC) under the same conditions.
1’-(2-tert-Butyl-2-methyl-1,3-benzodioxol-4-yl)-2’,2’,2’-
trifluoroethanone (2)
To a stirred solution of n-butyllithium (2.6M in hexane,
100 mL, 260 mmol) was added dropwise tetramethylethy-
AHCTUNGTREGlNNUN enediamine (38.6 mL, 266 mmol) at room temperature.
Then, acetal 4 (31.6 g, 165 mmol) was added dropwise at
room temperature to give a homogeneous solution. The
mixture was then cooled to À508C, and to that CF3CO2Et
(52.0 mL, 431 mmol) was carefully added to keep that tem-
perature, as the reaction is exothermic. The addition com-
pleted within 1.5 h and the mixture was stirred at room tem-
perature for 24 h. Then, the reaction was quenched by the
addition of saturated aqueous NH4Cl solution, and the mix-
ture was extracted with EtOAc. The extract was washed
with brine, dried over Na2SO4, and concentrated under
vacuum. The residue was distilled under reduced pressure to
give 2 as a yellow oil; yield: 42.0 g (88%); bp 95–988C/
2 Torr.
2’-Chloroethyl 2-tert-Butyl-2-methyl-1,3-
benzodioxole-4-carboxylate (1c)
To a stirred solution of 2 (500 mg, 1.74 mmol) in DMF
(8.7 mL) was added KOH (244 mg, 4.38 mmol), and the mix-
ture was stirred for 30 min at room temperature then heated
to 408C for 1.5 h. To this, 1,2-dichloroethane (540 mL,
6.94 mmol) was added and the reaction temperature was
raised to 508C. After 22 h, the reaction was quenched by
the addition of phosphate buffer solution (0.1M, 10 mL,
pH 7.0) and the mixture was extracted with EtOAc. The ex-
tract was washed with brine, dried over Na2SO4, and concen-
trated under vacuum. The residue was purified by silica gel
AHCTUNGTREG(NNNU 2R*,1’R*)- and (2R*,1’S*)-1’-(2-tert-Butyl-2-methyl-
1,3-benzodioxol-4-yl)-2’,2,’2’-trifluoroethanol (3a)
To a solution of 2 (30.0 mg, 0.104 mmol) in EtOH (1.0 mL)
was added NaBH4 (23.5 mg, 0.618 mmol) and the mixture
was stirred at room temperature for 1 h. Then, the reaction
Adv. Synth. Catal. 2010, 352, 2549 – 2558
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2555

Toshinori Higashi et al.
FULL PAPERS
was quenched by the addition of water and AcOH, and the
mixture was extracted with EtOAc. The extract was washed
with water and brine, dried over Na2SO4, and concentrated
under vacuum. The residue was purified by preparative
TLC [developed with hexane-EtOAc (4:1)] to afford
(2R*,1’R*)-3a (yield: 13.7 mg, 45%), Rf 0.50 [hexane-EtOAc
(4:1)] and (2R*,1S*)-3a (yield: 14.0 mg, 46%), Rf 0.47
[hexane-EtOAc (4:1)].
filter and washed with water. The weight of combined wet
cells was ca. 41.6 g from 800 mL of the broth.
Harvested cells of G. candidum (30.5 g) grown in a glu-
cose medium were re-suspended in phosphate buffer solu-
tion (0.2M, 100 mL, pH 6.0), together with a substrate 2
(1.0 g, 3.47 mmol) and glucose (15.1 g), in a 300-mL round-
bottomed flask which was loosely capped with sterilized
filter paper for a plentiful supply of oxygen. The mixture
was stirred at 308C for 24 h. The mixture was centrifuged
(3000 rpm) to remove cell mass and the supernatant was fur-
ther filtered through a pad of Celite. The filtrate was satu-
rated with NaCl and extracted with EtOAc. The cell debris
and the insoluble materials on the filter was mixed with ace-
tone (30 mL). After being sonicated for 15 min, the mixture
was filtered and the filtrate was concentrated under vacuum.
The combined cell debris extract was mixed with EtOAc ex-
tract from the filtrate of broth, and washed with brine, dried
over Na2SO4, and concentrated under vacuum. The residue
was purified by silica gel column chromatography (30 g).
Elution with hexane-EtOAc (20:1 to 5:1) afforded 3a as a
pale yellow oil; yield: 918.8 mg (91%). HPLC: tR (min)=
13.2 [(2R,1’R)-, 0.3%], 13.9 [(2R,1’S)-, 50.0%], 14.9
[(2S,1’S)-, 48.6%], 20.8 [(2S,1’R)-, 1.0%].
ACHTUNGTRENNUNG(2R*,1’R*)- and (2R*,1’S*)-1’-(2-tert-Butyl-2-methyl-
1,3-benzodioxol-4-yl)-2’,2’,2’-trifluoroethyl Acetate
(3b)
A mixture of (2R*,1’R*)-3a (13.7 mg, 0.047 mmol), acetic
anhydride (120 mL, 1.27 mmol), pyridine (120 mL) and a cat-
alytic amount of DMAP was stirred at room temperature
for 1 h. Then the reaction was quenched by the addition of
water, and the mixture was extracted with EtOAc. The ex-
tract was washed with water and brine, dried over Na2SO4,
and concentrated under vacuum. The residue was purified
by silica gel column chromatography (1 g). Elution with
hexane-EtOAc (50:1 to 10:1) provided (2R*,1’R*)-3b as a
yellow oil; yield: 14.7 mg (94%); Rf 0.53 [hexane-EtOAc
(9:1)].
In a similar manner, (2R*,1’S*)-3b was obtained from
(2R*,1’S*)-3a, as a yellow oil; yield: 14.7 mg (92%); Rf 0.49
[hexane-EtOAc (9:1)].
Scaled-Up Reduction of (Æ)-2 with Whole Cells of
G. candidum NBRC 5767
A broth [peptone (200 g), yeast extract (50 g), KH2PO4
(30 g), K2HPO4 (20 g), at pH 6.5, total volume of 5 L] was
sterilized in a stainless steel cylinder (30 cm i.d.ꢁ40 cm tall).
This was mixed with separately sterilized glucose solution
(500 g in 5 L deionized water), and immediately a seed cul-
ture (100 mLꢁ4, pre-incubated for 2 days) was added. The
mixture was stirred at 180 rpm with a mechanical stirrer
with an aseptic aeration (4 LminÀ1) at 26.58C for 4 days.
The mycelia was harvested by filtration and the weight
reached 1.2 kg.
C. antarctica Lipase B-Catalyzed Transesterification
of (2R*,1’R*)- and (2R*,1’S*)-3b
To a solution of (2R*,1’R*)-3b (20.6 mg, 0.062 mmol) in cy-
clopentanol (300 mL) was added C. antarctica lipase B
(81.6 mg). The mixture was stirred for 3 days at 308C. The
mixture was filtered with a pad of Celite and the filtrate was
concentrated under vacuum. The residue was purified by
preparative TLC [developed with hexane-EtOAc (4:1)] to
afford 3a (yield: 7.8 mg, 44%) and 3b (yield: 7.6 mg, 37%).
The alcohol was analyzed by HPLC [Chiralcel OJ-H,
0.46 cmꢁ25 cm; hexane-isopropyl alcohol (40:1); flow rate
0.5 mLminÀ1; detected at 288 nm]: tR (min)=14.9 [(2S,1’S)-,
single peak]. The recovered acetate as above was hydro-
lyzed to give 3a, which was analyzed by HPLC: tR (min)=
13.2 [(2R,1’R)-, single peak].
In a similar manner as for the transesterification of
(2R*,1’R*)-3b, alcohol 3a (yield: 1.4 mg, 9%) and acetate 3b
(yield: 13.6 mg, 66%) were obtained from (2R*,1’S*)-3b.
HPLC analysis of 3a: tR (min)=13.9 [(2R,1’S)-, 99.1%], 20.8
[(2S,1’R)-, 0.9%]. HPLC analysis of 3a originating from the
recovered acetate: tR (min)=13.9 [(2R,1’S)-, 45.4%], 20.8
[(2S,1’R)-, 54.6%].
The reduction was performed in a similar manner. To a
suspension of compressed cells of G. candidum (1.2 kg) in
phosphate buffer solution (0.2M, 8 L, pH 6.5) containing
glucose (1.2 kg) was added dropwise a solution of 2 (39.4 g,
136.8 mmol) in EtOH (35 mL) over 1 h with stirring
(180 rpm). The mixture was stirred for 5 days at 26.58C.
Work-up, extraction and the partial purification of the crude
residue (34.2 g) by a silica gel column chromatography
(200 g) provided a mixture containing 3a (1:1 diastereomeric
mixture, total yield: 17.7 g, 61.0 mmol, 45%) and 2 (yield:
9.80 g, 34.0 mmol, 25%). The yields of 3a and 2 were esti-
1
mated by measuring the H NMR spectrum with an internal
standard [(Æ)-(1R*,2R*,3R*,6R*,7S*)-2-iodo-4,8-dioxatricy-
clo[4.2.1.03,7]nonan-5-one, whose purity is guaranteed by its
crystalline nature, by comparing with its standard signal at
d=4.80 (d, J=5.2 Hz, 1H)].
Whole Cell-Catalyzed Reduction of (Æ)-2 with
G. candidum NBRC 5767
C. antarctica Lipase B-Catalyzed Transesterification
of (2S,1’S)- and (2R,1’S)-3b
A loopful of the strain grown on the agar-plate culture was
aseptically inoculated to a glucose medium [glucose (40.0 g),
peptone (16.0 g), yeast extract (4.0 g), KH2PO4 (2.4 g),
K2HPO4 (1.6 g), at pH 6.2, total volume of 800 mL] in eight
500-mL baffled Erlenmeyer cultivating flasks. The flask was
shaken on a gyratory shaker (160 rpm) for 24 h at 308C. The
wet cells were harvested by filtration with a rough paper
To a solution of (2S,1’S)- and (2R,1’S)-3b (1:1, 9.70 g, total
29.2 mmol) in cyclopentanol (150 mL) was added C. antarc-
tica lipase B (20.0 g). The mixture was stirred for 3 days at
308C, then filtered with a pad of Celite, and the filtrate was
concentrated under vacuum. The residue was purified by
2556
asc.wiley-vch.de
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 2549 – 2558

Chemoenzymatic Synthesis of Both Enantiomers
silica gel column chromatography (200 g). Elution with
hexane-EtOAc (10:1) afforded 3a (yield: 3.40 g, 41%) and
3b (yield: 5.30 g, 55%). HPLC analysis of 3a: tR (min)=13.9
[(2R,1’S)-, 3.4%], 14.9 [(2S,1’S)-, 96.6%]. HPLC analysis of
3a originating from the recovered acetate: tR (min)=13.2
[(2R,1’R)-, 1.2%], 13.9 [(2R,1’S)-, 86.7%], 14.9 [(2S,1’S)-,
8.8%], 20.8 [(2S,1’R)-, 3.3%].
Repetition of the above-mentioned lipase-catalyzed trans-
esterification was carried out for further kinetic resolution,
as follows. The recovered acetate 3b (5.30 g) in the first res-
olution was treated with C. antarctica lipase B (11 g) in cy-
clopentanol (80 mL) in the same manner. The work-up and
purification gave 3b as recovery (yield: 4.60 g, 88%), whose
stereochemical profile was analyzed after derivation to 3a:
tR (min)=13.2 [(2R,1’R)-, 1.5%], 13.9 [(2R,1’S)-, 93.7%],
14.9 [(2S,1’S)-, 2.1%], 20.8 [(2S,1’R)-, 2.6%].
(2R,1’S)-3a; yield: 587 mg (59% over 2 steps). HPLC: tR
(min)=13.9 [(2R,1’S)-, 99.6%], 20.8 [(2S,1’R)-, 0.4%].
(R)-TBMB Acid (1a)
In a similar manner as for (S)-1a, (R)-TBMB acid (1a) was
obtained as a colorless solid; yield: 122 mg, (73% over two
steps, 99.2% ee). HPLC analysis after derivatizion to (R)-
1b: tR (min)=10.4 [(S)-, 0.4%], 13.1 [(R)-, 99.6%]. Analyti-
cal sample for (R)-1a: mp 171–1728C; [a]2D5: À35.0 (c 1.00,
MeOH). IR, NMR and mass spectra were in good accord-
ance with those of (S)-1a.
Acknowledgements
On the other hand, the alcohol 3a obtained in the first
resolution was acetylated to give 3b (yield: 3.70 g, 95%).
This was treated with C. antarctica lipase B (7.5 g) in cyclo-
pentanol (50 mL) in the same manner. The work-up and pu-
rification gave 3a (yield: 1.90 g, 60%). HPLC: tR (min)=
14.9 [(2S,1’S)-3a, single peak].
The authors thank Professor Kaoru Nakamura of the Insti-
tute for Chemical Research University of Kyoto, Dr. Mikio
Fujii of Toho University and Professor Tohru Yamada of
Keio University, for their valuable discussion and encourage-
ment of this study, and Mr. Yusaku Iwanagaꢀs contribution in
the early phase of this work. Amano Enzyme Inc. for gener-
ous gift of PS-IM, Novozymes Japan for Novozym 435, Su-
mitomo Chemical Co., for B. cepacia lipase and Nagase &
Co., for several proteases and esterase are acknowledged with
thanks. This work was supported both by a Grant-in-Aid for
Scientific Research (No. 18580106) and formation of a re-
search centre in cell signaling drug discovery for molecular
targeting therapies: matching fund subsidy 2009–2013 from
the Ministry of Education, Culture, Sports, Science and Tech-
nology, Japan, which are acknowledged with thanks.
(S)-TBMB Acid (1a)
To a solution of (2S,1’S)-3a (602 mg, 2.08 mmol) in EtOAc
(2.0 mL) was added IBX (1.2 g, 4.92 mmol), and the mixture
was stirred at 808C for 24 h. The mixture was filtered with a
pad of Celite and the filtrate was concentrated under
vacuum. The residue was dissolved in DMF (4.0 mL) and
aqueous NaOH solution (8M, 2.0 mL, 16.0 mmol) added.
The mixture was stirred at room temperature for 1 h and
was then carefully acidified with hydrochloric acid (2M) to
pH 3.5. The precipitated solids were collected by filtration
to give (S)-1a (yield: 424 mg, 87%) as a colorless solid,
whose ee was >99.9% by the HPLC analysis of 1b obtained
by treatment with TMSCHN2: tR (min)=10.4 [(S)-, single
peak]. Recrystallization of (S)-1a from acetone-water af-
forded an analytical sample; mp 171–1728C; [a]2D5: +35.6 (c
1.00, MeOH) {lit.[1] [a]2D2: +30.5 (c 0.1, MeOH)}. Its
1H NMR spectrum was identical with that reported previ-
ously.[1]
References
[1] a) Y. Nishida, H. Ohrui, H. Meguro, Tetrahedron Lett.
1989, 30, 5277–5280; b) Y. Nishida, H. Ohrui, H.
Meguro, C. Kabuto, Anal. Sci. 1991, 7, 349–350; c) Y.
Nishida, M. Abe, H. Ohrui, H. Meguro, Tetrahedron:
Asymmetry 1993, 4, 1431–1434; d) Y. Nishida, E. Itoh,
M. Abe, H. Ohrui, H. Meguro, Anal. Sci. 1995, 11,
213–220.
ACHTUNGTRENNUNG(2R,1’S)-1’-(2-tert-Butyl-2-methyl-1,3-benzodioxol-4-
yl)-2’2’2’-trifluoroethyl Chloroacetate (3c)
[2] C.-S. Chen, Y. Fujimoto, G. Girdaukas, C. J. Sih, J. Am.
Chem. Soc. 1982, 104, 7294–7299.
[3] I. Fatakopoulou, E. Barbayianni, V. Constantinou-Ko-
kodou, U. T. Bornscheuer, G. Kokotos, J. Org. Chem.
2007, 72, 782–786.
The mixture of (2R,1’S)-3a (1.0 g, 3.45 mmol, 95.6% de,
94.6% ee), chloroacetic anhydride (590 mg, 3.45 mmol), pyr-
idine (10 mL), and a catalytic amount of DMAP was stirred
at room temperature for 20 min. The same work-up for 3b
provided chloroacetate 3c as a colorless oil; yield: 1.1 g. This
was employed for the next step without further purification.
[4] R. J. Crawford, W. F. Erman, C. D. Broaddus, J. Am.
Chem. Soc. 1972, 94, 4298–4396.
[5] L. S. Chen, G. J. Chen, C. Tamborski, J. Fluorine Chem.
1981, 18, 117–129.
[6] K. Kato, Y. Gong, R. Irimescu, T. Saito, Y. Yokogawa,
B. cepacia Lipase-Catalyzed Transesterification of
(2R,1’S)-3c
Biotechnol. Lett. 2002, 24, 1623–1630.
[7] K. Kitsuda, J. Calveras, Y. Nagai, T. Higashi, T. Sugai,
J. Mol. Catal. B: Enz. 2009, 59, 197–200.
To a solution of (2R,1’S)-3c (1.10 g, 3.29 mmol, 95.6% de,
94.6% ee) in cyclopentanol (16.0 mL) was added B. cepacia
lipase (Amano PS-IM, 2.2 g). The mixture was stirred for 3
days at 508C. The mixture was filtered with a pad of Celite
and the filtrate was concentrated under vacuum. The residue
was purified by silica gel column chromatography (30 g).
Elution with hexane-EtOAc (50:1 to 10:1) afforded
[8] I. Ohtani, T. Kusumi, Y. Kashman, H. Kakisawa, J.
Am. Chem. Soc. 1991, 113, 4092–4096.
[9] T. Yamada, T. Nagata, Y. Ohtsuka, T. Ikeno, T. Mu-
kaiyama, J. Synth. Org. Chem. Jpn. 2003, 61, 843–856.
[10] a) K. Nakamura, K. Takenaka, M. Fujii, Y. Ida, Tetra-
hedron Lett. 2002, 43, 3629–3631; b) T. Harada, Y.
Adv. Synth. Catal. 2010, 352, 2549 – 2558
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2557

Toshinori Higashi et al.
FULL PAPERS
Kubota, T. Kamitanaka, K. Nakamura, T. Matsuda, Tet-
rahedron Lett. 2009, 50, 4934–4936.
[11] K. Nakamura, Y. Inoue, A. Ohno, Tetrahedron Lett.
e) S.-Z. Wang, J.-P. Qu, G. Xu, R.-L. Yang, Biochem.
Eng. J. 2009, 45, 113–119.
[13] J. Uppenberg, N. ꢃhrner, M. Norin, K. Hult, G. J. Kley-
wegt, S. Patkar, V. Waagen, T. Anthonsen, T. A. Jones,
Biochemistry 1995, 34, 16838–16851.
1995, 36, 265–266.
[12] a) E. Mizuguchi, M. Takemoto, K. Achiwa, Tetrahe-
dron: Asymmetry 1993, 4, 1961–1964; b) J. Y. Goujon,
F. Zammatio, B. Kirschleger, Tetrahedron: Asymmetry
2000, 11, 2409–2420; c) D. G. Gascoyne, H. L. Fink-
beiner, K. P. Chan, J. L. Gordon, K. R. Stewart, R. J.
Kazlauskas, Org. Chem. 2001, 66, 3041–3048; d) L. F.
Solares, R. Brieva, M. Quirꢂs, I. Llorente, M. Bayod,
V. Gotor, Tetrahedron: Asymmetry 2004, 15, 341–345;
[14] T. Ema, J. Kobayashi, S. Maeno, T. Sakai, M. Utaka,
Bull. Chem. Soc. Jpn. 1998, 71, 443–453.
[15] E. Ito, Y. Nishida, H. Horie, H. Ohrui, H. Meguro,
Bunseki Kagaku (Analytical Science) 1995, 44, 739–
746.
[16] E. Napolitano, R. Fiaschi, E. Mastrorilli, Synthesis
1986, 122–125.
2558
asc.wiley-vch.de
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2010, 352, 2549 – 2558