A chemoenzymatic approach to the synthesis of the stereoisomers of a β-adrenergic receptor antagonist

Article abstract of DOI:10.1016/S0957-4166(00)00241-X

The four stereoisomers of Δ²-isoxazoline 2, a β-adrenergic receptor antagonist structurally related to Falintolol 1, were prepared by an enzyme-catalyzed kinetic resolution of the unsaturated secondary alcohol (±)-7 followed by its cycloadditio

Full text of DOI:10.1016/S0957-4166(00)00241-X

Tetrahedron: Asymmetry 11 (2000) 2741±2751
A chemoenzymatic approach to the synthesis of the
stereoisomers of a b-adrenergic receptor antagonist
Clelia Dallanoce,a Marco De Amici,a,* Giacomo Carrea, b Francesco Secundo,b
Sabrina Castellanoc and Carlo De Michelia
a
Á
Istituto di Chimica Farmaceutica e Tossicologica, Universita di Milano, viale Abruzzi 42, 20131 Milano, Italy
bIstituto di Biocatalisi e Riconoscimento Molecolare, CNR, via Mario Bianco 9, 20131 Milano, Italy
c
Á
Dipartimento di Scienze Farmaceutiche, Universita di Trieste, piazzale Europa 1, 34127 Trieste, Italy
Received 11 May 2000; accepted 16 June 2000
Abstract
The four stereoisomers of Á2-isoxazoline 2, a b-adrenergic receptor antagonist structurally related to
Falintolol 1, were prepared by an enzyme-catalyzed kinetic resolution of the unsaturated secondary alcohol
(þ)-7 followed by its cycloaddition to pyruvonitrile oxide. Through this strategy, diastereomeric amino-
alcohols (+)-2a/(^)-2b and (^)-2a/(+)-2b were obtained in 99 and 92% enantiomeric excess, respectively.
The absolute con®guration to the target compounds was assigned via chemical correlation to the enantiomers
of epoxides 4a and 4b, whose stereochemistry had been previously established. # 2000 Elsevier Science
Ltd. All rights reserved.
1. Introduction
The use of the cycloaddition approach in the synthesis of biologically active heterocycles has
characterized our research in the last decade.1 5 Recently, we applied the cycloaddition of nitrile
oxides to the preparation of a set of Á2-isoxazoline derivatives,6 designed as semirigid analogues
of the b-blocking agent Falintolol 1 (Fig. 1).7,8 The 3-isopropenyl stereoisomers anti-(þ)-2a and
syn-(þ)-2b (Fig. 1) were revealed as the most interesting among the investigated compounds. In
particular, (þ)-2a behaved as a non-selective b-adrenergic receptor antagonist, and displayed a
binding anity to b1- and b2-adrenergic receptors very close to that reported for the model
compound 1.6 Since our goal was a deeper investigation of the pharmacological pro®le of deriv-
atives 2, we planned the synthesis of the two enantiomeric pairs (+)-2a/(^)-2a and (+)-2b/(^)-2b.
Any speculation on the relationship between the structure and biological properties of chiral
isomers requires their availability in very high enantiomeric purity. To attain such a goal, we
tackled the synthesis of target compounds by means of biocatalytic methods, which represent an
* Corresponding author. Tel: +39-02-29502263; fax: +39-02-29514197; e-mail: marco.deamici@unimi.it
0957-4166/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(00)00241-X

2742
C. Dallanoce et al. / Tetrahedron: Asymmetry 11 (2000) 2741±2751
Figure 1.
increasingly valuable alternative to common synthetic approaches, i.e. asymmetric synthesis and
resolution procedures.9 11 In the past, we used enzyme-catalyzed reactions as a key step in the
preparation of enantiomerically pure compounds provided with a variety of biological
activities.12 16 As an extension of our e€orts along this research line, we now report the applica-
tion of the chemoenzymatic strategy to the synthesis of the four stereoisomers of 2. In addition,
our approach allowed the assignment of the absolute con®guration to the enantiomers (+)-2a/(^)-
2a and (+)-2b/(^)-2b.
2. Results and discussion
The sequence used to prepare stereoisomers (þ)-2a and (þ)-2b is reported in Scheme 1.6 The
cycloaddition of pyruvonitrile oxide to butadiene yielded 3-acetyl-5-vinyl-Á2-isoxazoline 3,
which, after epoxidation and Wittig methylenation, a€orded a 1:1 mixture of stereoisomeric
epoxides anti-(þ)-4a and syn-(þ)-4b, whose relative con®guration was assigned.6 The desired
amino alcohols (þ)-2a and (þ)-2b were then obtained by reacting (þ)-4a [(þ)-4b] with excess tert-
butylamine.
Scheme 1. (a) AcOEt/NaHCO3; (b) MCPBA/CH2Cl2; (c) Ph3PˆCH2; (d) tBuNH2/MeOH

C. Dallanoce et al. / Tetrahedron: Asymmetry 11 (2000) 2741±2751
2743
Based on these results, we extended the above strategy to the preparation of the four stereo-
isomers of 2 in enantiomerically pure form. For such a purpose, we studied the kinetic resolution
of derivatives related to (RS)-3-butene-1,2-diol through the enzyme-catalyzed transesteri®cation
of their secondary alcohol. The transacetylation of 2-hydroxy-but-3-enyl butyrate (þ)-5 (Scheme
2) under the catalysis of di€erent lipases17 gave a moderate enantioselection, as re¯ected by the
low values of the enantiomeric ratio (E) reported in Table 1. Accordingly, residual monoester was
obtained as a single enantiomer only by increasing the degree of conversion (ꢀ70%). An inver-
sion of the enantiopreference was detected on passing from Candida antarctica lipase B [CALB,
E=18 for (R)-(+)-5] to Pseudomonas ¯uorescens lipase [lipase AK, E=14 for (S)-(^)-5]. The
degree of conversion and the enantiomeric excess (e.e.) of reagent and product were both
evaluated by chiral GLC analysis.17 The absolute con®guration of residual monoester (S)-(^)-5
[(R)-(+)-5] and produced diester (R)-(+)-8 [(S)-(^)-8] was assigned through their conversion
into the enantiomers of 3-butene-1,2-diol, whose absolute con®gurations are known from the
literature.17,18
Scheme 2. (a) Enzyme/vinyl acetate; (b) NaOH/H2O±MeOH
Table 1
Lipase-catalyzed transacetylation of substrates (þ)-5, (þ)-6 and (þ)-7
Since our goal was the synthesis of both enantiomers in very high enantiomeric excess, we
investigated the biotransformation of derivatives (þ)-6 and (þ)-7, which are characterized by the
presence of a sterically hindered ester or silyl ether, respectively. As shown in Table 1, the results
on the transacetylation of substrates (þ)-6 and (þ)-7 revealed that the structural modi®cations
taken into account signi®cantly improved the enantioselectivity. Among the tested enzymes,
Lipase PS was the most selective catalyst of the transacetylation of derivative (þ)-6 (E>100),

2744
C. Dallanoce et al. / Tetrahedron: Asymmetry 11 (2000) 2741±2751
whereas CALB eciently catalyzed the acetyl transfer on substrate (þ)-7 (E>100). Therefore,
both the enzyme/substrate couples were suited for their utilization on a preparative scale.
However, we chose monoester (þ)-7, owing to the higher rate of transacetylation in the presence
of CALB and a better eciency of the chiral HPLC evaluation of the enantiomeric purity.
As shown in Table 1, the CALB-catalyzed transesteri®cation of substrate (þ)-7 with vinyl
acetate allowed the preparation of either residual substrate (S)-(^)-7 and produced ester (R)-(+)-
10 in gram amounts with a very high enantiomeric purity (e.e. higher than 99% and equal to
92%, respectively). The product of the enzyme-catalyzed transesteri®cation (R)-(+)-10 was
submitted to an alkaline hydrolysis to yield secondary alcohol (R)-(+)-7. The two enantiomers
(S)-(^)-7 and (R)-(+)-7 were separately submitted to 1,3-dipolar cycloaddition with pyruvonitrile
oxide. As reported in Scheme 3, the pericyclic reaction gave rise to equimolar amounts of
stereoisomeric 3-acetyl-Á2-isoxazolines (+)-11a/(^)-11b and (^)-11a/(+)-11b, which were separated
Scheme 3. (a) CH3COC(Cl)ˆNOH/AcOEt/NaHCO3; (b) Ph3PˆCH2; (c) TBAF/THF; (d) MeC(OMe)3/pTosOH; (e)
Me3SiCl/CH2Cl2; (f) K2CO3/MeOH; (g) tBuNH2/MeOH

C. Dallanoce et al. / Tetrahedron: Asymmetry 11 (2000) 2741±2751
2745
by column chromatography. Treatment of isomers (+)-11a, (^)-11a, (+)-11b and (^)-11b with
triphenylmethylene phosphorane produced the corresponding 3-isopropenyl derivatives (+)-12a,
(^)-12a, (+)-12b and (^)-12b. Cleavage of the silyl ether was accomplished by reacting (+)-12a [(^)-
12a] and (+)-12b [(^)-12b] with a THF solution of tetrabutyl ammonium ¯uoride, and the
resulting diols (+)-13a [(^)-13a] and (+)-13b [(^)-13b] were submitted to Sharpless' `one pot' pro-
cedure,19 which smoothly a€orded epoxides (+)-4a [(^)-4a] and (+)-4b [(^)-4b] with retention of
con®guration at Ca. Since the con®guration to diastereomeric anti±syn epoxides 4a and 4b was
previously attributed,6 the sequence of steps depicted in Scheme 3 allowed the assignment of
absolute con®gurations to cycloadducts (5S,aS)-(+)-11a and (5R,aS)-(^)-11b.
Treatment of (+)-4a [(^)-4a] and (+)-4b [(^)-4b] with an excess of tert-butylamine gave the
desired aminoalcohols (5S,aR)-(+)-2a [(^)-2a] and (5S,aS)-(+)-2b [(^)-2b], which were transformed
into the corresponding 1:1 oxalates. An HPLC analysis of both diastereomeric pairs (+)-2a/(^)-2b
and (^)-2a/(+)-2b on a chiral stationary phase (Chirobiotec T) demonstrated that the sequence of
reactions applied did not alter the value of enantiomeric purity gained in the enzymatic kinetic
resolution.
In summary, the use of enzymes allowed a convenient synthesis of new chiral heterocyclic
derivatives of interest to medicinal chemistry. The preparation of the four stereoisomers of a
b-adrenoceptor antagonist was achieved in high enantiomeric purity through an ecient CALB-
catalyzed transesteri®cation coupled to a pericyclic reaction involving a nitrile oxide. The
evaluation of the pharmacological pro®le of target compounds will be reported in due course.
3. Experimental
Lipases from Pseudomonas cepacia (lipases PS) and Pseudomonas ¯uorescens (lipase AK) were
purchased from Amano Pharmaceutical Co. Lipase B from Candida antarctica (CALB) was
bought from Novo Nordisk in immobilized form with the commercial name of NOVOZYM 435.
Organic solvents were reagent grade. Racemic 3-butene-1,2-diol was kindly donated by Bayer
AG, Leverkusen. 1H NMR spectra were recorded at 200 MHz in CDCl3 solutions; chemical shifts
(ꢀ) are expressed in ppm and coupling constants (J) in hertz. Chiral GLC analyses were conducted
on a gas chromatograph equipped with a Chrompack CP-Cyclodextrin-2,3,6-M-19 column (50
m, 0.25 mm ID). Chiral HPLC analyses were performed on a Chiralcel OD column (4.6Â250
mm) and on a Chirobiotic T (4.6Â250 mm). The experimental conditions of chromatographic
analyses are speci®ed in the appropriate paragraph. Melting points were determined on a Mod. B
540 Buchi apparatus and are uncorrected. Liquid compounds were characterized by the oven
temperature for bulb to bulb distillations. Rotary power determinations were carried out with a
Perkin±Elmer 241 polarimeter coupled with a Haake N3-B thermostat. TLC analyses were
performed on commercial silica gel 60 F254 aluminum sheets: spots were further evidenced by
spraying with a dilute alkaline potassium permanganate solution. Microanalyses (C, H, N) of
new compounds agreed with the theoretical value þ0.4%.
3.1. Synthesis of (þ)-6 and (þ)-7
A. To a stirred solution of (þ)-3-butene-1,2-diol (2 g, 22.72 mmol) and triethylamine (3.80 mL,
27.26 mmol) in dichloromethane (100 mL) was added dropwise a solution of pivaloyl chloride
(3.08 mL, 25 mmol) in dichloromethane (30 mL) at 0ꢁC. The reaction mixture was stirred at room

2746
C. Dallanoce et al. / Tetrahedron: Asymmetry 11 (2000) 2741±2751
temperature for 3 h, then acidi®ed to pH 3 with 3N HCl. The organic layer was separated and the
aqueous phase was further extracted with dichloromethane (2Â70 mL). The pooled organic
extracts were dried over anhydrous sodium sulfate and the solvent was removed at reduced
pressure. A silica gel column chromatography of the residue (eluant: 10% ethyl acetate/petroleum
ether) gave 3.05 g (78% yield) of the desired monoester.
(þ)-2-Hydroxy-but-3-enyl pivalate 6: colorless oil, bp 155±160ꢁC/20 mmHg, Rf 0.23 (eluant: 15%
1
ethyl acetate/cyclohexane); H NMR: 1.20 (s, 9), 2.43 (bs, 1, OH), 4.07 (dd, 1, CHOCO, J=6.8
and 11.2), 4.10 (dd, 1, CHOCO, J=4.0 and 11.2), 4.36 (m, 1, CHOH), 5.22 (d, 1, CHˆCHCHOH,
J=10.4), 5.36 (d, 1, CHˆCHCHOH, J=17.5), 5.85 (ddd, 1, CH2ˆCH, J=5.6, 10.4 and 17.5).
B. To a stirred solution of (þ)-3-butene-1,2-diol (4.0 g, 45.45 mmol), triethylamine (12.68 mL,
91 mmol) and 4-dimethylaminopyridine (0.556 g, 4.55 mmol) in anhydrous THF (300 mL) was
added dropwise under nitrogen a solution of tert-butyldiphenylchlorosilane (12.85 mL, 50 mmol)
in anhydrous THF (100 mL). The reaction mixture was stirred overnight at room temperature,
the precipitate was ®ltered o€ and the volatiles were evaporated at reduced pressure. The residue
was submitted to a silica gel column chromatography (eluant: 5% ethyl acetate/petroleum ether)
to a€ord 13.04 g (88% yield) of the expected silyl ether.
(þ)-1-(tert-Butyldiphenylsilyloxy)-2-hydroxy-3-butene 7: viscous colorless oil, bp 190±195ꢁC/
0.3 mmHg, Rf 0.44 (eluant: 5% ethyl acetate/cyclohexane); 1H NMR: 1.08 (s, 9), 2.73 (bs, 1, OH),
3.58 (dd, 1, CHOSi, J=7.4 and 9.9), 3.68 (dd, 1, CHOSi, J=3.8 and 9.9), 4.27 (m, 1, CHOH),
5.16 (d, 1, CHˆCHCHOH, J=10.5), 5.32 (d, 1, CHˆCHCHOH, J=17.3), 5.79 (ddd, 1,
CH2ˆCH, J=5.6, 10.5 and 17.3), 7.43 (m, 6, arom.), 7.67 (m, 4, arom.).
3.2. Lipase-catalyzed preparative transacetylation of substrates (þ)-6 and (þ)-7
A. A solution of 1.72 g (10 mmol) of (þ)-6 in petroleum ether (150 mL) was treated with vinyl
acetate (7.5 mL) in the presence of lipase PS (1.5 g). The reaction mixture was stirred (200 rpm) at
room temperature and stopped at 52% conversion (about 60 h). The degree of conversion was
determined by GLC analysis, carried out on a 25 m HP1 capillary silica gel column coated with
methylsilicone gum under the following conditions: oven temperature from 85ꢁC (initial time 5
min) to 130ꢁC (®nal time 10 min). H2 was used as the carrier gas at a heating rate of 1.2ꢁC/min.
The enzyme was ®ltered o€ and the solvent evaporated at reduced pressure. The residue was
submitted to silica gel column chromatography (eluant: 10% ethyl acetate/petroleum ether) to
a€ord 0.757 g (4.4 mmol) of (S)-(^)-6 and 1.017 g (4.75 mmol) of (R)-(^)-9.
(S)-2-Hydroxy-but-3-enyl pivalate (^)-6: ‰ꢁŠ2D0=^5.3 (c 1.040, CHCl3), e.e.>99%. The e.e. was
determined by chiral GLC analysis on a Chrompack CP-Cyclodextrin column, after conversion
into the corresponding (S)-(+)-9, following the temperature program utilized with the achiral
stationary phase. Relative retention times (min): (S)-(+)-9, 30.15; (R)-(^)-9, 30.55.
(R)-2-Acetyloxy-but-3-enyl pivalate (^)-9: colorless oil, bp 165±170ꢁC/20 mmHg, Rf 0.56
20
D
20
436
(eluant: 15% ethyl acetate/cylohexane). ‰ꢁŠ =^0.5 (c 1.0, CHCl3); [ꢁ] =^2.7 (c 1.0, CHCl3);
20
365
[ꢁ] =^5.5 (c 1.0, CHCl3), e.e.=92%, determined by chiral GLC analysis (see above). 1H NMR:
1.18 (s, 9), 2.08 (s, 3, CH3CO), 4.13 (dd, 1, CHOCOtBu, J=6.7 and 11.6), 4.20 (dd, 1, CHOCOtBu,
J=4.3 and 11.6), 5.28 (d, 1, CHˆCHCHOCO, J=10.8), 5.35 (d, 1, CHˆCHCHOCO, J=17.3),
5.51 (m, 1, CHOCOCH3), 5.80 (ddd, 1, CH2ˆCH, J=5.8, 10.8 and 17.3).
The assignment of absolute con®gurations to (S)-(^)-6 and (R)-(^)-9 was con®rmed by their
conversion, through an alkaline hydrolysis, into (S)-(^)- and (R)-(+)-3-butene-1,2-diol, ‰ꢁŠ2D0=^9.5
(c 1.0, CHCl3) and ‰ꢁŠ2D0=+8.8 (c 1.0, CHCl3), respectively.

C. Dallanoce et al. / Tetrahedron: Asymmetry 11 (2000) 2741±2751
2747
B. A solution of 12.08 g (37 mmol) of (þ)-7 in petroleum ether (500 mL) was reacted with vinyl
acetate (27.75 mL) in the presence of CALB (7 g). The reaction mixture was stirred (200 rpm) at room
temperature for about 18 h (52% conversion). The degree of conversion was determined by achiral
GLC analysis (see above): oven temperature from 200ꢁC (initial time 5 min) to 230ꢁC (®nal time 10
min); heating rate, 2ꢁC/min. The enzyme was ®ltered o€ and the solvent evaporated at reduced
pressure. The residue was submitted to silica gel column chromatography (eluant: 3% ethyl acetate/
petroleum ether) to a€ord 5.24 g (16.07 mmol) of (S)-(^)-7 and 6.73 g (18.28 mmol) of (R)-(+)-10.
(S)-1-(tert-Butyldiphenylsilyloxy)-2-hydroxy-3-butene (^)-7: ‰ꢁŠ2D0=^5.0 (c 1.040, CHCl3),
e.e.>99%. The HPLC analysis was carried out on a Chiralcel OD column: eluant 0.2% 2-propanol/
petroleum ether, ¯ow rate 1 mL/min, l 254 nm. Relative retention times (min): (R)-(+)-7, 29.25;
(S)-(^)-7, 40.02. To corroborate the assignment of absolute con®guration, a sample (200 mg) of
(S)-(^)-7 was treated with a 1 M solution of tetrabutylammonium ¯uoride in THF and converted
into known (S)-(^)-3-butene-1,2-diol.
(R)-2-Acetyloxy-1-(tert-butyldiphenylsilyloxy)-3-butene (+)-10: viscous colorless oil, bp 205±
210ꢁC/0.3 mmHg, Rf 0.64 (eluant: 5% ethyl acetate/cyclohexane); ‰ꢁŠD20=+10.8 (c 1.010, CHCl3);
1
e.e.=92% [from HPLC analysis, after conversion into the corresponding (R)-(+)-7]. H NMR:
1.05 (s, 9), 2.06 (s, 3, CH3CO), 3.73 (m, 2, CH2OSi), 5.22 (d, 1, CHˆCHCHOCO, J=10.6), 5.29
(d, 1, CHˆCHCHOCO, J=17.3), 5.41 (m, 1, CHOCO), 5.81 (ddd, 1, CH2ˆCH, J=6.3, 10.6 and
17.3), 7.43 (m, 6, arom.), 7.67 (m, 4, arom.).
3.3. Synthesis of aminoalcohols (+)-2a/(^)-2a and (+)-2b/(^)-2b
A. To an ethyl acetate solution (200 mL) of pyruvohydroximoyl chloride20 (2.67 g, 22 mmol)
and (S)-(^)-7 (4.78 g, 14.67 mmol) was added solid sodium bicarbonate (9.25 g, 0.11 mol). The
mixture was stirred at room temperature until disappearance of the starting dipolarophile (about
18 h). The slurry was then poured into water, the organic layer separated and the aqueous phase
was further extracted with ethyl acetate (2Â50 mL). The pooled organic extracts were dried over
anhydrous sodium sulfate and the solvent was removed under reduced pressure. A silica gel
column chromatography of the residue (eluant: 10% ethyl acetate/petroleum ether) gave 2.38 g of
the anti-isomer (+)-11a and 2.03 g of the syn-isomer (^)-11b (73% overall yield).
(+)-11a (5S,aS): thick colorless oil, Rf 0.27 (eluant: 20% ethyl acetate/n-hexane); ‰ꢁŠ2D0=+66.4 (c
0.990, CHCl3); H NMR: 1.07 (s, 9), 2.48 (s, 3, CH3CO), 2.50 (bs, 1, OH), 3.11 (dd, 1, H-40,
1
J=11.3 and 17.6), 3.14 (dd, 1, H-4, J=7.9 and 17.6), 3.73 (m, 2, CH2OSi), 3.82 (m, 1, CHOH),
4.85 (ddd, 1, H-5, J=5.7, 7.9 and 11.3), 7.41 (d, 6, arom.), 7.66 (d, 4, arom.). Anal. calcd for
C23H29NO4Si: C, 67.12; H, 7.10; N, 3.40. Found: C, 67.55; H, 6.90; N, 3.27.
(^)-11b (5R,aS): thick colorless oil; Rf 0.33 (eluant: 20% ethyl acetate/n-hexane); ‰ꢁŠ2D0=^108.1
1
(c 0.990, CHCl3); H NMR: 1.07 (s, 9), 2.49 (s, 3, CH3CO), 2.65 (bs, 1, OH), 3.13 (m, 2, H-4),
3.68±3.80 (m, 3), 4.89 (ddd, 1, H-5, J=2.4, 9.8 and 9.8), 7.39 (d, 6, arom.), 7.65 (d, 4, arom.).
Anal. calcd for C23H29NO4Si: C, 67.12; H, 7.10; N, 3.40. Found: C, 66.70; H, 6.75; N, 3.71.
B. To a stirred solution of 6.40 g (17.39 mmol) of (R)-(+)-10 in MeOH (120 mL) 40 mL of 2N
NaOH was added. The reaction mixture was stirred for 2 h at room temperature, then methanol was
evaporated under vacuum and the residual aqueous phase was extracted with ethyl acetate (3Â30
mL). After the usual work-up, crude (R)-(+)-7 was puri®ed by distillation (5.025 g, 89% yield).
(R)-1-(tert-Butyldiphenylsilyloxy)-2-hydroxy-3-butene (+)-7: ‰ꢁŠ2D0=+4.4 (c 1.020, CHCl3),
e.e.=92% [HPLC analysis on a Chiralcel OD column, following the conditions reported above
for (^)-7].

2748
C. Dallanoce et al. / Tetrahedron: Asymmetry 11 (2000) 2741±2751
The cycloaddition reaction to (R)-(+)-7, performed following the above described procedure,
allowed the isolation of stereoisomers (^)-11a and (+)-11b.
(^)-11a (5R,aR): thick colorless oil; ‰ꢁŠ2D0=^61.2 (c 0.995, CHCl3). Anal. calcd for
C23H29NO4Si: C, 67.12; H, 7.10; N, 3.40. Found: C, 66.82; H, 7.51; N, 3.19.
(+)-11b (5S,aR): thick colorless oil; ‰ꢁŠ2D0=+98.9 (c 1.015, CHCl3). Anal. calcd for
C23H29NO4Si: C, 67.12; H, 7.10; N, 3.40. Found: C, 66.91; H, 6.87; N, 3.65.
C. To an ice-cooled stirred suspension of potassium tert-butoxide (1.40 g, 12.48 mmol) in
anhydrous toluene (100 mL) was added portionwise methyltriphenylphosphonium bromide (4.78
g, 13.37 mmol). After heating at re¯ux for 1 h, the suspension was cooled at room temperature. A
solution of (+)-11a (2.20 g, 5.35 mmol) in toluene (5 mL) was then added dropwise. The mixture
was stirred at room temperature for about 3 h, until disappearance of the starting material; the
progress of the reaction was monitored by TLC (eluant: 20% ethyl acetate/petroleum ether).
Acetone (10 mL) and water (50 mL) were then added, the organic phase was separated and the
aqueous phase was extracted with ether (3Â30 mL). After the usual work-up, the residue was
submitted to column chromatography (eluant: 5% ethyl acetate/petroleum ether) a€ording 1.16 g
(53% yield) of the desired anti 3-isopropenyl-Á2-isoxazoline (+)-12a.
(+)-12a (5S,aS): thick colorless oil; Rf 0.30 (eluant: 15% ethyl acetate/n-hexane); ‰ꢁŠ2D0=+73.3 (c
0.992, CHCl3); 1H NMR: 1.08 (s, 9), 2.03 (s, 3, CH3-CˆCH2), 2.48 (d, 1, OH, J=3.5), 3.10 (dd, 1,
H-40, J=10.8 and 16.5), 3.18 (dd, 1, H-4, J=7.6 and 16.5), 3.70±3.90 (m, 3), 4.72 (ddd, 1, H-5,
J=5.0, 7.6 and 10.8), 5.21 (bs, 1, CH3-CˆCH), 5.33 (bs, 1, CH3-CˆCH), 7.39 (d, 6, arom.), 7.67
(d, 4, arom.). Anal. calcd for C24H31NO3Si: C, 70.38; H, 7.63; N, 3.42. Found: C, 70.06; H, 7.98;
N, 3.12.
The same procedure carried out on 3-acetyl-Á2-isoxazolines (^)-11a, (^)-11b and (+)-11b gave
the corresponding 3-isopropenyl derivatives (^)-12a, (^)-12b and (+)-12b in comparable yields.
(^)-12a (5R,aR): thick colorless oil; ‰aŠ2D0=^67.1 (c 1.01, CHCl3). Anal. calcd for C24H31NO3Si:
C, 70.38; H, 7.63; N, 3.42. Found: C, 69.95; H, 8.04; N, 3.17.
(^)-12b (5R,aS): thick colorless oil; Rf 0.26 (eluant: 15% ethyl acetate/n-hexane); ‰ꢁŠ2D0=^91.1 (c
1
1.0, CHCl3); H NMR: 1.07 (s, 9), 2.04 (s, 3, CH3-CˆCH2), 2.32 (d, 1, OH, J=5.7), 3.12 (m, 2,
H-4), 3.63±3.82 (m, 3), 4.78 (ddd, 1, H-5, J=3.4, 9.6 and 9.6), 5.18 (bs, 1, CH3-CˆCH), 5.33 (bs,
1, CH3-CˆCH), 7.39 (d, 6, arom.), 7.65 (d, 4, arom.). Anal. calcd for C24H31NO3Si: C, 70.38; H,
7.63; N, 3.42. Found: C, 70.45; H, 8.01; N, 3.15.
(+)-12b (5S,aR): thick colorless oil; ‰ꢁŠ2D0=+82.2 (c 0.990, CHCl3). Anal. calcd for
C24H31NO3Si: C, 70.38; H, 7.63; N, 3.42. Found: C, 70.70; H, 7.91; N, 3.32.
D. To a stirred solution of (+)-12a (1.08 g, 2.64 mmol) in THF (50 mL) 3.5 mL of a 1 M
solution of tetrabutylammonium ¯uoride in THF was added. After stirring for 1 h at room
temperature, the reaction mixture was concentrated and directly submitted to a silica gel column
chromatography (eluant: 30% petroleum ether/ethyl acetate) to give 0.384 g (85% yield) of the
desired diol (+)-13a.
(+)-13a (5S,aR): mp 88.5±90.5ꢁC (colorless lea¯ets from diisopropyl ether); Rf 0.22 (eluant:
20
1
50% ethyl acetate/cyclohexane); ‰ꢁŠD =+172.1 (c 1.050, CHCl3); H NMR: 2.04 (s, 3, CH3), 2.11
(bs, 2, OH), 3.16 (dd, 1, H-40, J=11.0 and 16.4), 3.19 (dd, 1, H-4, J=8.5 and 16.4), 3.71 (dd, 1,
J=6.1 and 11.5), 3.73±3.92 (m, 2), 4.63 (ddd, 1, H-5, J=4.7, 8.5 and 11.0), 5.26 (bs, 1, CH3-
CˆCH), 5.36 (bs, 1, CH3-CˆCH). Anal. calcd for C8H13NO3: C, 56.13; H, 7.65; N, 8.18. Found:
C, 56.27; H, 7.90; N, 8.02.
The same procedure carried out on 3-isopropenyl-Á2-isoxazolines (^)-12a, (^)-12b and (+)-12b
gave the corresponding diols (^)-13a, (^)-13b and (+)-13b in comparable yields.

C. Dallanoce et al. / Tetrahedron: Asymmetry 11 (2000) 2741±2751
2749
(^)-13a (5R,aS): mp 88±90.5ꢁC (colorless lea¯ets from diisopropyl ether); ‰ꢁŠ2D0=^157.1 (c
1.025, CHCl3). Anal. calcd for C8H13NO3: C, 56.13; H, 7.65; N, 8.18. Found: C, 56.38; H, 7.73;
N, 8.41.
(^)-13b (5R,aR): mp 86±88ꢁC (colorless lea¯ets from diisopropyl ether); Rf 0.14 (eluant: 50%
20
1
ethyl acetate/n-hexane); ‰ꢁŠD =^232.4 (c 0.998, CHCl3); H NMR: 2.04 (s, 3, CH3), 2.12 (bs, 1,
OH), 2.48 (d, 1, OH, J=6.2), 3.16 (m, 2, H-4), 3.60±3.85 (m, 3), 4.73 (ddd, 1, H-5, J=4.5, 9.6 and
9.6), 5.24 (bs, 1, CH3-CˆCH), 5.36 (bs, 1, CH3-CˆCH). Anal. calcd for C8H13NO3: C, 56.13; H,
7.65; N, 8.18. Found: C, 55.91; H, 7.97; N, 7.88.
(+)-13b (5S,aS): mp 85±88ꢁC (colorless lea¯ets from diisopropyl ether); ‰ꢁŠ2D0=+211.0 (c 1.0,
CHCl3). Anal. calcd for C8H13NO3: C, 56.13; H, 7.65; N, 8.18. Found: C, 56.37; H, 7.79; N, 8.39.
E. To a solution of 0.350 g (2.05 mmol) of (+)-13a in dichloromethane (5 mL) were added
toluene-4-sulfonic acid monohydrate (5 mg) and trimethyl orthoacetate (310 mL, 2.46 mmol).
After stirring at room temperature for 0.5 h, the volatiles were evaporated at reduced pressure
and the residue was dissolved in dichloromethane (5 mL). Trimethylchlorosilane (365 mL, 2.88
mmol) was then added and the reaction mixture was stirred for 4 h at room temperature. After
removal of the solvent, the residue was taken up with MeOH (10 mL) and treated with potassium
carbonate (0.570 g). The suspension was vigorously stirred for 1 h, then 20 mL of a saturated
aqueous solution of ammonium chloride was added. The reaction mixture was extracted with
dichloromethane (3Â10 mL) and, after the usual work-up, the residue was puri®ed by column
chromatography (eluant: 20% ethyl acetate/petroleum ether), a€ording 0.244 g (78% yield) of the
desired epoxide (+)-4a.
(+)-4a (5S,aR): colorless liquid, bp 80±85ꢁC/1.5 mmHg, Rf 0.51 (eluant: 30% ethyl acetate/
20
1
cyclohexane); ‰ꢁŠD =+167.5 (c 1.040, CHCl3). The H NMR spectrum was identical to that
reported for (þ)-4a.6 Anal. calcd for C8H11NO2: C, 62.73; H, 7.24; N, 9.14. Found: C, 62.44; H,
7.03; N, 9.37.
Epoxides (^)-4a, (^)-4b and (+)-4b were prepared from diols (^)-13a, (^)-13b and (+)-13b in
comparable yields through the same sequence of steps.
(^)-4a (5R,aS): colorless liquid, bp 80±85ꢁC/1.5 mmHg; ‰ꢁŠD20=^155.3 (c 1.150, CHCl3). Anal.
calcd for C8H11NO2: C, 62.73; H, 7.24; N, 9.14. Found: C, 62.59; H, 7.50; N, 8.92.
(^)-4b (5R,aR): colorless liquid, bp 80±85ꢁC/1.5 mmHg (crystalline on standing, mp 51.5±
54.5ꢁC), Rf 0.46 (eluant: 30% ethyl acetate/cyclohexane); ‰ꢁŠ2D0=^211.7 (c 0.990, CHCl3). The 1H
NMR spectrum was identical to that reported for (þ)-4b.6 Anal. calcd for C8H11NO2: C, 62.73;
H, 7.24; N, 9.14. Found: C, 63.01; H, 7.05; N, 9.16.
(+)-4b (5S,aS): colorless liquid, bp 80±85ꢁC/1.5 mmHg; ‰ꢁŠD20=+194.9 (c 0.995, CHCl3). Anal.
calcd for C8H11NO2: C, 62.73; H, 7.24; N, 9.14. Found: C, 62.60; H, 7.41; N, 9.35.
F. A stirred solution of (+)-4a (230 mg, 1.50 mmol) and tert-butylamine (950 mL, 9 mmol) in
methanol (10 mL) was re¯uxed until TLC (eluant: 20% ethyl acetate/petroleum ether) evidenced
the disappearance of the starting material. The solvent and excess reagent were removed under
vacuum, the oily residue was dissolved in 3N HCl (15 mL) and washed with ethyl ether (2Â10
mL). The aqueous layer was alkalinized with solid sodium carbonate and extracted with ethyl
acetate (3Â10 mL). After the usual work-up, the residue of the pooled organic extracts (278 mg,
82% yield) was dissolved in anhydrous ethyl ether and treated with a threefold excess of
anhydrous oxalic acid. The corresponding oxalate, which precipitated immediately, was recov-
ered by suction ®ltration.
(+)-2a (5S,aR)ÂC2H2O4: colorless prisms (from absolute ethanol), mp 191±194ꢁC, dec.;
‰ꢁŠ2D0=+159.6 (c 1.0, MeOH), e.e.>99%, determined by chiral HPLC analysis on a Chirobiotic T

2750
C. Dallanoce et al. / Tetrahedron: Asymmetry 11 (2000) 2741±2751
column [eluant: MeOH/AcOH (0.05%)/TEA (0.05%), injection vol.: 2 mL, ¯ow rate: 0.6 mL/min,
1
l 254 nm]. Relative retention times (min): (^)-2a, 24.52; (+)-2a, 26.22. The H NMR spectrum
(recorded on the corresponding free base) was identical to that reported for (þ)-2a.6 Anal. calcd
for C14H24N2O6: C, 53.15; H, 7.65; N, 8.86. Found: C, 52.92; H, 7.59; N, 8.89.
Isomeric aminoalcohols (^)-2a, (^)-2b and (+)-2b were similarly obtained in comparable yields
from epoxides (^)-4a, (^)-4b and (+)-4b, respectively.
(^)-2a (5R,aS)ÂC2H2O4: colorless prisms (from absolute ethanol), mp 191±193ꢁC, dec.;
‰ꢁŠ2D0=^147.9 (c 1.0, MeOH), e.e.=92% [HPLC analysis on a Chirobiotic T column, following
the conditions reported above for (+)-2a]. Anal. calcd for C14H24N2O6: C, 53.15; H, 7.65; N, 8.86.
Found: C, 53.37; H, 7.45; N, 8.74.
(^)-2b (5R,aR)ÂC2H2O4: colorless prisms (from absolute ethanol), mp 164.5±168ꢁC, dec.;
‰ꢁŠ2D0=^167.9 (c 0.990, MeOH); e.e.>99%, determined by chiral HPLC analysis on a Chirobiotic
T column [eluant: MeOH/AcOH (0.25%)/TEA (0.25%), injection vol.: 2 mL, ¯ow rate: 0.6 mL/
1
min, l 254 nm]. Relative retention times (min): (+)-2b, 13.30; (^)-2b, 14.31. The H NMR spec-
trum (recorded on the corresponding free base) was identical to that reported for (þ)-2b.6 Anal.
calcd for C14H24N2O6: C, 53.15; H, 7.65; N, 8.86. Found: C, 53.10; H, 7.76; N, 8.84.
20
D
(+)-2b (5S,aS)ÂC2H2O4: colorless prisms (from absolute ethanol), mp 164±167.5ꢁC, dec.; ‰ꢁŠ
=+154.3 (c 1.010, MeOH); e.e.=92% [HPLC analysis on a Chirobiotic T column, following the
conditions reported above for (^)-2b]. Anal. calcd for C14H24N2O6: C, 53.15; H, 7.65; N, 8.86.
Found: C, 52.97; H, 7.60; N, 8.71.
Acknowledgements
This research was ®nancially supported by grants from the National Council of Research
(CNR, Target Project on Biotechnology) and from the Ministero dell'Universita Scienti®ca e
Tecnologica (MURST, Rome).
References
1. De Amici, M.; De Micheli, C.; Grana, E.; Rodi, R.; Zonta, F.; Santagostino Barbone, M. G. Eur. J. Med. Chem.
1989, 24, 171±177.
2. De Amici, M.; De Micheli, C.; Grana, E.; Lucchelli, A.; Zonta, F. Il Farmaco 1990, 45, 859±866.
3. De Amici, M.; Magrõ, P.; De Micheli, C.; Cateni, F.; Bovara, R.; Carrea, G.; Riva, S.; Casalone, G. J. Org. Chem.
1992, 57, 2825±2829.
4. Conti, P.; De Amici, M.; De Sarro, G.; Stensbùl, T. B.; Brauner-Osborne, H.; Madsen, U.; De Micheli, C. J. Med.
Chem. 1998, 41, 3759±3762.
5. Conti, P.; De Amici, M.; De Sarro, G.; Rizzo, M.; Stensbùl, T. B.; Brauner-Osborne, H.; Madsen, U.; Toma, L.;
De Micheli, C. J. Med. Chem. 1999, 42, 4099±4107.
6. Conti, P.; Dallanoce, C.; De Amici, M.; De Micheli, C.; Klotz, K. N. Biorg. Med. Chem. 1998, 6, 401±408.
7. Bouzoubaa, M.; Leclerc, G.; Rakhit, S.; Andermann, G. J. Med. Chem. 1985, 28, 896±900, and references cited
therein.
8. Himber, J.; Sallee, V.; Andermann, G.; Bouzoubaa, M.; Leclerc, G.; De Santis, L. J. Ocul. Pharmacol. 1987, 3,
111±120.
9. Sih, C. J.; Wu, S.-H. In Topics in Stereochemistry; Eliel, E. L.; Wilen, S. H., Eds.; John Wiley & Sons: New York,
1989; Vol. 19, pp. 63±125.
10. Wong, C.-H.; Whitesides, G. M. Enzymes in Synthetic Organic Chemistry; Pergamon: Oxford, 1994.
11. Chirality in Industry II; Collins, A. N.; Sheldrake, G. N.; Crosby, J., Eds.; John Wiley & Sons: New York, 1997.

C. Dallanoce et al. / Tetrahedron: Asymmetry 11 (2000) 2741±2751
2751
12. De Amici, M.; De Micheli, C.; Carrea, G.; Spezia, S. J. Org. Chem. 1989, 54, 2646±2650.
13. De Amici, M.; De Micheli, C.; Molteni, G.; Pitre, D.; Carrea, G.; Riva, S.; Spezia, S.; Zetta, L. J. Org. Chem.
1991, 56, 67±72.
14. De Amici, M.; De Micheli, C.; Cateni, F.; Carrea, G.; Ottolina, G. Tetrahedron: Asymmetry 1993, 4, 1073±1080.
15. De Amici, M.; De Micheli, C.; Gianferrara, T.; Stefancich, G. Il Farmaco 1997, 52, 307±311.
16. Conti, P.; Dallanoce, C.; De Amici, M.; De Micheli, C.; Carrea, G.; Zambianchi, F. Tetrahedron: Asymmetry
1998, 9, 657±665.
17. Secundo, F.; Oppizzi, M. L.; Carrea, G. F.; Dallanoce, C.; De Amici, M. Biocat. Biotransf. 1999, 17, 241±250.
18. Suzuki, T.; Kasai, N.; Minamiura, N. Tetrahedron: Asymmetry 1994, 5, 239±246.
19. Kolb, H. C.; Sharpless, K. B. Tetrahedron 1992, 48, 10515±10520, and references cited therein.
20. Armand, J.; Souchay, P.; Valentini, F. Bull. Soc. Chem. France 1968, 4585±4594.