Mendeleev Communications Electronic Version, Issue 2, 2002 1 Superacidic cyclisation.lipase-mediated kinetic resolution as a short route from achiral linear isoprenoid alcohols to scalemic cyclic isomers Edward P. Serebryakov,*a Galina D. Gamalevich,a Veacheslav N. Kulcitki,b Nicon D. Ungurb and Pavel F. Vlad*b a N. D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, 119991 Moscow, Russian Federation.Fax: +7 095 135 5328 b Institute of Chemistry, Academy of Sciences of Republic of Moldova, MD 2028 Kishinev, Republic of Moldova. Fax: + 373 2 73 9775 10.1070/MC2002v012n02ABEH001566 (¡¾)-¥á-Cyclogeraniol and (¡¾)-drim-7-en-11-ol acetates obtained via the FSO3H-induced cyclisation of geraniol and (E)-farnesol and subsequent acetylation were hydrolysed in the presence of hog pancreas lipase (PPL) to afford (R)-(+)-¥á-cyclogeraniol (ee ~30% at the optimal conversion C = 20¡¾2%) and (5R,9R,10R)-(+)-drim-7-en-11-ol (ee 78.5% at C = 30%), respectively; (¡¾)-15-acetoxyisoagath- 12-ene, obtained similarly from all-E-geranylgeraniol, is resistant to PPL-mediated hydrolysis, but is hydrolysed in the presence of lipase from Candida cylindracea to afford (10S,14R)-(.)-isoagath-12-en-15-ol of 69.80% ee in ~3% yield.Cyclic isoprenoids with a common structure and common relative configuration can occur in the nature in both enantiomeric forms. Typically, the species abundant in one or another enantiomer belong to rather remote taxons. To study the biomedical properties of enantiomeric terpenoids and, all the more so, to convert the latter into chiral building blocks, considerable amounts of both enanitomers are needed, but it is not always easy to obtain each of the two from natural sources or by chemical synthesis.However, both enantiomers of certain cyclic isoprenoids could be obtained in only three steps via a stereodivergent protocol that includes (i) the stereospecific superacid-induced cyclisation of geometrically pure achiral alcohols of the type H[CH2C(Me)= CHCH2]nOH into racemic cyclic isomers, (ii) the acetylation of the latter to produce the corresponding (¡¾)-acetates, and (iii) the subsequent lipase-mediated kinetic resolution of the acetates into differently functionalised (+)- and (.)-enantiomers by partial hydrolysis.Cyclisation of the lower members of the above type (n = 2.4) is well known,1 while enzymatic optical resolution of chiral alcohols or esters is such an obvious solution,2 that one can only wonder, (or just not wonder) why this protocol has not been actuated so far.¢Ó This work demonstrates the feasibility of stereodivergent transformation of geraniol 1 and (E,E)-farnesol 2 into scalemic ¥á-cyclogeraniols [(R)-3, (S)-3] and drim-7-en-11-ols [(all-R)-4, (all-S)-4].Similar transformation of all-E geranylgeraniol 5 into (.)-isoagath-12-en-15-ol [(10S,14R)-6] is also reported. By means of superacidic cyclisation alcohols 1, 2 and 5 were converted into corresponding cyclic isomers (¡¾)-3, (¡¾)-4, and (¡¾)-6,¢Ô from which acetates (¡¾)-3a, (¡¾)-4a and (¡¾)-6a were prepared¡× and subjected to partial hydrolysis in the presence of hog pancreas lipase (PPL) or lipase from Candida rugosa (�� C.cylindracea, CCL). In order to attain the highest possible ee of scalemic alcohols 3 and 4 without sacrificing too much the yield, conversions of (¡¾)-3a and (¡¾)-4a were arrested at C ¡Ì 30%. Acetate (¡¾)-6a could not be hydrolysed in the presence of PPL even upon repetitive addition of the enzyme up to a twofold excess (w/w).The scalemic form of alcohol 6 was obtained only using CCL as the catalyst.¢Ò The PPL-mediated hydrolysis of (¡¾)-3a proceeded with low enantioselectivity. In the best variant (C = 20% in 32 h), it afforded in a 20% yield a specimen of (+)-¥á-cyclogeraniol [(R)-3] with [a]D 20 +34.6¡Æ (c 0.96, EtOH), which corresponds to ~30% ee.For the specimens of 3 with ee ~100%, [a]D 25 +122¡Æ (EtOH)4(a) for the R enantiomer and .116¡Æ or .109.2¡Æ (EtOH) for its S antipode were reported.4(b),(c) Mild alkaline hydrolysis of the fraction of unconverted acetate gave a specimen of alcohol (S)-3 with [a]D 20 .10.4% (c 1.05, EtOH), which corresponds to ee ~10%. The hydrolysis of (¡¾)-4a was quicker (C = 30% in 22 h) and more selective.The isolated drim-7-en-11-ol with all-R configuration [(R)-4] had [a]D 21 +13.1¡Æ (c 1.1, PhH), which corresponds to ee 78.5%. Lit., [a]D 25 +15.8¡Æ (PhH)5(a) for a specimen of (R)-4 with ee 88% and [a]D rt .18¡Æ or .18.2¡Æ (PhH)5(a),(b) for the specimens of all-S drim-7-en-11-ol [(S)-4] of ~100% ee. The fraction of unconverted acetate upon alkaline hydrolysis gave alcohol (S)-4 with [a]D 20 .5.81¡Æ (c 0.89, PhH), which corresponds to ee ~32% (Scheme 1).The enhanced hydrolysis enantioselectivity in the case of (¡¾)-4a was considered earlier.6 The transition from (¡¾)-4a to (¡¾)-6a resulted in a dramatical deceleration of the PPL-catalysed hydrolysis, the recovery of acetate (¡¾)-6a after 96 h of incubation being almost quantitative. The hydrolysis of (¡¾)-6a in the presence of CCL was also very slow (C ~10% after 148 h). The reaction mass was workedup as usual and chromatographed twice, first on SiO2 and then on Florisil¢ç, to give an alcohol in 3% yield with [a]D 21 .7.2¡Æ (c 0.23, CHCl3).It was identical to the (.)-enantiomer of 6 with the 10S,14R configuration in 1H and 13C NMR spectra, GC MS data7(a),(b) and the sign of [a]D. Specimens of the latter with ~100% ee, obtained earlier from natural sources7(a),(b) or by partial synthesis from grindellic acid7(a) and sclareol7(c), displayed [a]D rt .9 to .10.5¡Æ (CHCl3).7 Consequently, the ee of this speci- ¢Ó A conceptually similar approach to Ambrox¢ç.from farnesyl acetate via lipase-mediated kinetic resolution of (¡¾)-drimane-8,10-diol3 is unsuitable for obtaining unsaturated cyclic alcohols with unambiguously positioned double bonds. ¢Ô Starting alcohols 1.3 were cyclised on treatment with fluorosulfonic acid (20 equiv.) in 2-nitropropane at .78¡ÆC, the reaction mass was neutralised with NEt3, washed with water, extracted with Et2O, dried (Na2SO4), and concentrated. The remainder was subjected to column chromatography on SiO2 using hexane.Et2O gradient elution (80:20 ¢ç 0:100) to give, in agreement with known procedures, (¡¾)-¥á-cyclogeraniol [(¡¾)-3, yield 73%],1(a) (¡¾)-drim-7-en-15-ol [(¡¾)-4, yield 71%]1(b) and (¡¾)-isoagath-12-en-15-ol [(¡¾)-6, yield 72%].1(c) ¡× The solutions of (¡¾)-4, (¡¾)-5 and (¡¾)-6 (0.25 mmol in hexane) were treated with Ac2O.Py (1:1, v/v; ~20 ¡ÆC, 12.18 h) in the presence of 4-DMAP (4.5 mg). After standard work-up the acetates were purified by column chromatography on SiO2 to give (¡¾)-3a, (¡¾)-4a and (¡¾)-6a in 92, 94 and 79% yields, respectively. ¢Ò Powdered PPL (48 units per mg protein, Olainfarm, Latvia) or CCL (type VII with lactose, 1200 units per mg solid, Sigma) and a substrate (1:2, w/w) were suspended in a 0.1 M aqueous phosphate buffer solution [pH 6.5 for PPL or 7.0 for CCL, 1.3 ml per 0.25 mmol of (¡¾)-acetate] and vigorously stirred at 20.22 ¡ÆC.The duration of stirring for (¡¾)-3a, (¡¾)-4a and (¡¾)-6a was 28.32 and 22 h (using PPL), or 148 h (using CCL), respectively. After standard work-up the only products of PPLmediated hydrolysis, the required alcohol and acetate (TLC monitoring), were separated by column chromatography on SiO2 [in the case of Cdiated hydrolysis of (¡¾)-6a this was followed by additional chromatography on Florisil¢ç] and identified spectroscopically.The enantioselectivity of hydrolysis was estimated by comparing the signs and magnitudes of [a]D of chromatographically pure alcohols obtained at C ¡Ì 30% with those reported in the literature for the same alcohols of ~100% ee.The fractions of unconverted acetates left after the hydrolysis of (¡¾)-3a or (¡¾)-4a were saponified (1 N NaOH/MeOH.H2O, ~20 ¡ÆC, 2.4 h) to give alcohols with the opposite signs of [a]D.Mendeleev Communications Electronic Version, Issue 2, 2002 2 men of (R)-6 is 69�C80%. The unconverted acetate afforded ¡®entisoagath- 7-en-11-ol¡� [(10R,14S)-6] with only ~1.5% ee {[a]D 21 +0.13¡ã (c 0.4, CHCl3)} upon alkaline hydrolysis.Attempts to enhance the enantioselectivity of enzymatic resolution of alcohols (¡À)-3 and (¡À)-4 by acylating them in the Ac2O�CPPL/hexane or vinyl acetate�CCCL/Et2O systems were unsuccessful. The PPL-mediated acylation of alcohol (¡À)-3 to 15% conversion gave acetate (R)-3a with only ~1.3% ee. When CCL was used, the fractions of unconverted alcohols recovered at C 55�C75% [predominantly (S)-3 and (S)-4] were contaminated by ¦Â-cyclogeraniol and drim-8-en-12-ol, respectively, and had ~9�C13% ee (as estimated by correlating the observed [a]D values with the respective 1H and 13C NMR data).8 Our results suggest that the superacidic cyclisation�Cacetylation �Cenzymatic hydrolysis protocol is synthetically useful.Its optimisation by lipase screening and/or by other known biocatalytic methods2,9 seems to be a feasible task. This work was supported by the Russian Foundation for Basic Research and the Federal Programme of Supporting Leading Scientific Schools of Russia (grant nos. 96-03-33396, 99-03-32992 and 96-15-97461) and by INTAS (grant no. 96-1109). References 1 (a) P.F. Vlad, N. D. Ungur, V. H. Nguen and V. B. Perutsky, Izv. Akad. Nauk, Ser. Khim., 1995, 2494 (Russ. Chem. Bull., 1995, 44, 2390); (b) P. F. Vlad, N. D. Ungur and V. B. Perutsky, Khim. Prir. Soedin., 1986, 793 [Chem. Nat. Compd. (Engl. Transl.), 1986, 12, 737]; (c) P. F. Vlad, N. D. Ungur and V. B. Perutsky, Khim. Prir. Soedin., 1986, 514 [Chem. Nat. Compd. (Engl. Transl.), 1986, 12, 485]; (d) P.F. Vlad, Pure Appl. Chem., 1993, 65, 1329. 2 (a) Enzyme Catalysis in Organic Synthesis, eds. K. Drauz and H.Waldmann, VCH, Weinheim, 1995, vol. 1, pp. 178�C261; (b) C. H. Wong and G. M. Whitesides, Enzymes in Synthetic Organic Chemistry, 2nd edn., Pergamon Press, Throwbridge, 1995, pp. 9�C13, 60�C130; (c) K. Faber, Biotransformations in Organic Chemistry.A Textbook, 2nd edn, Springer, Berlin, 1995, pp. 1�C23, 24�C144; (d) E. P. Serebryakov, Izv. Akad. Nauk, Ser. Khim., 2001, 1986 (Russ. Chem. Bull., Int. Ed., 2001, 50, 1984). 3 H. Tanimoto and T. Oritani, Tetrahedron: Asymmetry, 1996, 7, 1695. 4 (a) R. Buchecker, R. Egli, H. Regel-Wild, C. Tscharner, C. H. Eugster, G. Uhde and G. Ohloff, Helv. Chim. Acta, 1973, 56, 2544; (b) K.Mori, M. Amiake and M. Itou, Tetrahedron, 1993, 49, 1871; (c) H. Meyer and A. R��ttimann, Helv. Chim. Acta, 1980, 63, 1451. 5 (a) H. Akita, M. Nozawa, A. Mitsuda and H. Ohsawa, Tetrahedron: Asymmetry, 2000, 11, 1375; (b) H. H. Appel, C. J. Brooks and K. H. Overton, J. Chem. Soc., 1959, 3322. 6 E. P. Serebryakov and G. D. Gamalevich, Mendeleev Commun., 1996, 221. [There, due to erroneous assignment of ligands seniority at the atom C(9) in molecules (R)-4 and (S)-4a, the configurations of the dextrorotatory alcohol and ¡®unconverted acetate¡� were denoted wrongly]. 7 (a) G.Cimino, D. de Rosa, S. de Stefano and L. Minale, Tetrahedron, 1974, 30, 645; (b) K. Gustafson and R. J. Andersen, Tetrahedron, 1985, 41, 1101; (c) D. K. Manh Duc, M. Fetizon and M. Kone, Bull. Soc. Chim. Fr., 1975, 2351. 8 G. D. Gamalevich and E. P. Serebryakov, Izv. Akad, Nauk, Ser. Khim., 1997, 175 (Russ. Chem. Bull., 1997, 46, 171). 9 F. Theil, Tetrahedron, 2000, 56, 2905. Scheme 1 Reagents and conditions: i, FSO3H (20 equiv.)/Me2CHNO2, �C75 ¡ãC; ii, Ac2O�CPy (1:1, v/v)/4-DMAP (cat.), room temperature; iii, H2O (pH 6.5)/PPL (substrate:enzyme = 2:1, w/w), room temperature; iv, KOH�C MeOH (aqueous), room temperature; v, H2O (pH 7.0)/CCL (substrate: enzyme = 2:1, w/w), room temperature, 148 h. OH 1 OR (¡À)-3 H (¡À)-3a ii (92%) i (73%) iii (C ¡Ö 20%) OH H OR H R S (R)-(+)-3 ee ¡Ö 30% (S)-3a (R = Ac) (S)-(�C)-3 (R = H) ee ¡Ö 10% iv OH OAc H i (71%) ii (94%) 2 (¡À)-4a iii (C = 30%) OH H OR H R R R S S S (R)-(+)-4 ee 78.5% (S)-4a (R = Ac) (S)-(�C)-4 (R = H) ee ¡Ö 32% iv H i (72%) ii (79%) 5 (¡À)-6a v (3%) (C ¡Ö 10%) OH H OAc H H OH H R S (10S,14R)-6 ee 69�C80